JL

£0

sj

ZOOLOGICA

fe

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME XX

DECEMBER 24, 1935 — NOVEMBER 30, 1936
Numbers 1-3 Inclusive

PUBLISHED BY THE SOCIETY
THE ZOOLOGICAL PARK, NEW YORK

Jleto JJotfe Zoological #>octetj>

General Office: 101 Park Avenue, New York City

(Officers

President, W. Redmond Cross

Vice-Presidents, Kermxt Roosevelt and Alfred Ely
Chairman, Executive Committee, W. Redmond Cross

Treasurer, Cornelius R. Agnew
Secretary, Fairfield Osborn

Scientific Staff

Zoological Park

W. Reid Blair, Director

Raymond L. Ditmars, Curator of Mammals and Reptiles
Lee S. Crandall, Curator of Birds
Charles R. Schroeder, Veterinarian
Claude W. Leister, Ass’t to the Director and Curator, Educational Activities

H. C. Raven, Prosector
Edward R. Osterndorff, Photographer
William Bridges, Editor and Curator of Publications

Aquarium

Charles H. Townsend, Director
C. M. Breder, Jr., Assistant Director

William Beebe, Director and Honorary Curator of Birds

department of Croptcal &es;earci)

John Tee-Van, General Associate
Gloria Hollister, Research Associate
Jocelyn Crane, Technical Associate

Cbttorial Committee

* Madison Grant, Chairman

W. Reid Blair
William Beebe

Charles H. Townsend
George Bird Grinnell

William Bridges

* Deceased

TITLES OF PAPERS

Page

1 — Deep-sea fishes of the Bermuda Oceanographic Expeditions. Family

Derichthyidae . William Beebe 1

2 — Deep-sea fishes of the Bermuda Oceanographic Expeditions. Family

Nessorhamphidae William Beebe 25

3 — Deep-sea fishes of the Bermuda Oceanographic Expeditions. Family

Serrivomeridae. Part I: Genus Serrivomer.

William Beebe & Jocelyn Crane 53

-\UG 1 4 19

111

J

LIST OF ILLUSTRATIONS

DEEP-SEA FISHES OF THE BERMUDA OCEANOGRAPHIC EXPE-
DITIONS. FAMILY DERICHTHYIDAE.

Fig.

1.

Fig.

2.

Fig.

3.

Fig.

4.

Fig.

5.

Fig.

6.

Fig.

7.

Fig.

8.

Fig.

9.

Page

Derichthys serpentinus. A photograph of the head of an

adult female, showing pores and striations 2

The geographical and vertical distribution of Derichthys

serpentinus 4

Derichthys serpentinus. Adolescent, transitional adolescent,

adult 6

Heads of Derichthys serpentinus 7

Derichthys serpentinus. Bones of the head of adult female . . 10

Derichthys serpentinus. Hyoid and branchial arches of adult

female 11

Derichthys serpentinus. Vertebrae of adult female 14

Derichthys serpentinus. End of vertebral column and base

of caudal fin in adult female 17

Derichthys serpentinus. Digestive and reproductive systems

in adult female 18

DEEP-SEA FISHES OF THE BERMUDA OCEANOGRAPHIC EXPE-
DITIONS. FAMILY NESSORHAMPHIDAE.

Fig. 10. Stages in the development of Nessorhamphus ingolfianus. .26-27

Fig. 11. Heads of Nessorhamphus ingolfianus 30

Fig. 12. Nessorhamphus ingolfianus. Dentition of upper jaw in transi-
tional adolescent 32

Fig. 13. Nessorhamphus ingolfianus. Skull of larva, dorsal view.... 34

Fig. 14. Nessorhamphus ingolfianus. Skull of larva, lateral view 35

Fig. 15. Nessorhamphus ingolfianus. Hyoid and branchial apparatus

of larva 36

Fig. 16. Nessorhamphus ingolfianus. Skull of transitional adolescent,

dorsal view 40

Fig. 17. Nessorhamphus ingolfianus. Bones of head, pectoral girdle
and anterior part of vertebral column, in transitional

adolescent ; lateral view 40

Fig. 18. Nessorhamphus ingolfianus. Cephalic system in transitional

adolescent, lateral view 41

Fig. 19. Nessorhamphus ingolfianus. Hyoid and branchial apparatus

in transitional adolescent 41

v

Fig. 20. Nessorhamphus ingolfianus. Vertebrae of transitional adoles-
cent 44

Fig. 21. Nessorhamphus ingolfianus. End of vertebral column and

base of caudal fin of transitional adolescent 46

Fig. 22. Nessorhamphus ingolfianus. Alimentary canal of transitional

adolescent 47

DEEP-SEA FISHES OF THE BERMUDA OCEANOGRAPHIC EXPE-

DITIONS. FAMILY SERRIVOMERIDAE.

PART I: GENUS SERRIVOMER .

Fig. 23. The geographical and vertical distribution of the genus

Serrivomer 59

Fig. 24. Serrivomer beanii. Larvae, post-larva, adolescents, transi-
tional adolescent, adult 66-67

Fig. 25. Serrivomer beanii. Larvae, post-larva, adolescents, transi-
tional adolescent, adult 68

Fig. 26. Serrivomer beanii. Cartilaginous elements of larval skull,

dorsal view 70

Fig. 27. Serrivomer beanii. Cartilaginous elements of larval head,

lateral view 71

Fig. 28. Serrivomer beanii. Skull of adult, dorsal view 74

Fig. 29. Same; teeth of upper jaw and vomer, ventral view 74

Fig. 30. Same; bones of head, pectoral girdle and anterior part of

vertebral column, lateral view 74

Fig. 31. Serrivomer beanii. Hyoid and branchial apparatus of adult.. 75

Fig. 32. Serrivomer beanii. Vertebrae.... 76

Fig. 33. Serrivomer beanii. Posterior part of vertebral column and

base of caudal fin in adult female 79

Fig. 34. Serrivomer brevidentatus 94

Fig. 35. Serrivomer brevidentatus. Adolescent, transitional adolescent,

adult 95

Fig. 36. Serrivomer brevidentatus. Adolescent, transitional adolescent,

adult 96

Fig. 37. Serrivomer brevidentatus. Skull of adult, dorsal view 98

Fig. 38. Same; teeth of upper jaw and vomer, ventral view 98

Fig. 39. Same; bones of head, pectoral girdle and anterior part of

vertebral column, lateral view 98

Fig. 40. Serrivomer brevidentatus. Hyoid and branchial apparatus of

adult 99

Fig. 41. Serrivomer brevidentatus. Viscera of adolescent 100

Fig. 42. Serrivomer brevidentatus. Viscera of adult 100

vi

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME XX. NUMBERS 1 AND 2

DEEP-SEA FISHES OF THE BERMUDA
OCEANOGRAPHIC EXPEDITIONS

No. 1— Family DERICHTHYIDAE
No. 2— Family NESSORHAMPHIDAE

William Beebe

PUBLISHED BY THE SOCIETY
THE ZOOLOGICAL PARK, NEW YORK

December 24, 1935

J?ete gods Zoological &orietp

General Office: 101 Park Avenue, New York City

Officers;

President , Madison Grant

Vice-Presidents , W. Redmond Cross and Kermit Roosevelt
Chairman, Executive Committee, Madison Grant
Treasurer, Cornelius R. Agnew
Secretary, Henry Fairfield Osborn, Jr.

Poarb of OTrugt ees

Class of 1936

Madison Grant, Lewis R. Morris, Archer M. Huntington, Cornelius R.
Agnew, Harrison Williams, Marshall Field, Ogden L. Mills,
Vincent Astor, C. Suydam Cutting, Childs Frick,

Alfred Ely, Herbert L. Pratt

GllaSS of 1937

George Bird Grinnell, Frederic C. Walcott, George C. Clark,

W. Redmond Cross, Henry Fairfield Osborn, Jr., George
Gordon Battle, Bayard Dominick, Robert Gordon
McKay, Kermit Roosevelt, John M. Schiff,

Robert L. Gerry, Warren Kinney

Class of 1938

Robert S. Brewster, Edward S. Harkness, Irving K. Taylor, Harry
Payne Bingham, Landon K. Thorne, J. Watson Webb, Oliver
D. Filley, De Forest Grant, George F. Baker

Scientific Staff

W. Reid Blair, Director of the Zoological Park
William T. Hornaday, Director Emeritus
Charles H. Townsend, Director of the Aquarium
C. M. Breder, Jr., Assistant Director , Aquarium
Raymond L. Ditmars, Curator of Mammals and Reptiles
William Beebe, Honorary Curator of Birds and Director of Department of

Tropical Research
Lee S. Crandall, Curator of Birds
H. C. Raven, Prosector
Charles V. Noback, Veterinarian

Claude W. Leister, Assft to the Director and Curator, Educational Activities
Edward R. Osterndorff, Photographer
William Bridges, Editor and Curator of Publications

Cbitorial Committee

Madison Grant, Chairman

W. Reid Blair
William Beebe

William Bridges

Charles H. Townsend
George Bird Grinnell

Zoologica, Volume XX, Number 1

DEEP-SEA FISHES OF THE BERMUDA
OCEANOGRAPHIC EXPEDITIONS

Family DERICHTHYIDAE1
William Beebe

(Figs. 1-9 incl.)

Contents

Introduction p. l

Taxonomy p. 2

Detailed Discussion : Derichthys serpentinus

Bermuda specimens: General trawling data p. 5

Specimens previously recorded p. 5

Description of adult p. 5

Development p. 19

Ecology p. 21

Study material .p. 22

Synonymy and references p. 22

INTRODUCTION

For detailed data of nets, locality, dates, etc., concerning
the capture of the deep-sea eels treated in this monograph, refer
to Zoologica, Vol. XIII, Nos. 1, 2 and 3, and for physical data,
methods of measurement and definitions of growth stages see
Zoologica, Vol. XVI, No. 1. The accounts of deep-sea fishes di-
rectly preceding the present paper comprise ZOOLOGICA, Vol.
XVI, Nos. 2, 3 and 4. Reports on the other families of eels are
approaching completion and will appear shortly, together with
a survey of the Order and a complete resume of present knowl-
edge of the evolution of deep-sea eels.

Ail the material under consideration was taken in the course
of 1,350 nets drawn in one locality, an eight-mile circle, with its

1 Contribution No. 478, Department of Tropical Research, New York Zoological Society.

1

2

Zoologica: N. Y. Zoological Society

[XX; 1

Fig. 1. Derichthys serpentinus. A photograph of the head of an adult female, showing
pores and striations. (x 2.6).

center at 32° 12' North Latitude and 64° 36' West Longitude,
nine and a quarter miles south-southeast of Nonsuch Island,
Bermuda. Vertically this is an imaginary cylinder, considered
as extending from the surface to the bottom of the sea, an ex-
treme range of 1,500 fathoms. Six silk metre-nets were used,
strung at exact intervals along two miles of wire, drawn at an
angle of 30 degrees, at the rate of two knots an hour.

In the present work I have had the cooperation of my entire
staff. Mr. John Tee-Van supervised the capture of the deep-sea
fish. Miss Gloria Hollister cleared and stained specimens for
osteological study. Miss Jocelyn Crane’s part in these papers is
rather that of co-author than of an able assistant; I owe to her
the elaboration of the great mass of details. The drawings are
the work of Mr. George Swanson.

Family DERICHTHYIDAE Gill 1884

Body anguilliform, slender; anus before or behind mid-
body; scales absent, skin smooth; lateral line distinct; head

19351 Beebe: Deep-Sea Fishes of the Bermuda Expeditions 3

oblong, oval; eyes in anterior part of head, well developed; nos-
trils dorso-lateral or lateral, neither pair tubular; mouth with
cleft little oblique, extending at least to posterior part of eye;
jaws strong; maxillaries flattened, firmly articulated with the
expanded pre-vomer ; teeth conical, on jaws and vomer; branchial
apertures small, lateral or oblique slits in front of or below
pectorals, well separated; dorsal commencing behind head; anal
origin before or behind middle of body; caudal, when present,
confluent with dorsal and anal ; membranes of vertical fins thick.
Trewavas (1932, p. 641) has shown that the structure of the
upper jaw in Derichthys is essentially similar to that of other
eels, definitely abolishing the order Carencheli. Three genera.

Key to the Genera

A. Anal origin at or behind middle of body, far behind dorsal

origin (deep-sea, Atlantic) Derichthys Gill 1884

AA. Anal origin well in advance of middle of body, slightly be-
hind dorsal origin.

B. Caudal fin present (deep-sea, Philippines)

Benthenchelys Fowler 1934
BB. Caudal fin absent (shallow water, Panama)

Gorgasia Meek & Hildebrand 1923

Genus Derichthys Gill 1884

With the characteristics of the family. A neck-like con-
striction between head and pectoral fins ; anus sub-median,
slightly behind middle of body ; dorsal, anal and caudal confluent.

It seems almost certain that only one of the three described
species is valid. I agree with Parr (1934 p. 33 if.) that there is
no reason for maintaining D. iselini Borodin 1929 as distinct
from D. serpentinus. As Parr points out, there is not one of the
so-called differences between the species which cannot properly
be laid to the incomplete descriptions and bad preservation of
the type of D. serpentinus, while the terminal, tubular “nos-
trils” of D. iselini are unquestionably merely sensory pores
in advance of the true anterior nostrils. They are very prom-
inent in all the Bermuda specimens.

4 Zoologica: N. Y. Zoological Society [XX; 1

Fig. 2. The geographic and vertical distribution of Derichthys serpentinus. The relative number of specimens taken
at different depths by the Bermuda Oceanographic Expeditions is shown diagrammatically at the left of the column which
gives the vertical range of the genus.

1935] Beebe : Deep-Sea Fishes of the Bermuda Expeditions 5

In addition I am synonymizing D. kempi (Norman 1930)
with D. serpentinus, as our Bermuda examples agree perfectly,
except for the unpaired frontals (see p. 9), with the descrip-
tions and figures of the specimen from the South Atlantic.

Derichthys serpentinus Gill 1884

Specimens Taken by the Bermuda
Oceanographic Expeditions

Eighteen specimens; June to September, 1929 to 1931; 500
to 1,000 fathoms ; from a cylinder of water eight miles in diam-
eter (five to thirteen miles south of Nonsuch Island, Bermuda),
the center of which is at 32° 12' N. Lat., 64° 36' W. Long.;
standard lengths from 55 to 268 mm.

Specimens Previously Recorded

Four specimens; 1,000-0 fathoms; West Indies, North At-
lantic and South Atlantic off Cape Town; recorded standard
lengths from 160 to 200 mm. (Fig. 2).

Description of Adult

Color (fresh specimens) : (Figs. 1, 3C, 4C). Tawny olive to
mouse gray, with glints of bluish sheen on the neck ; fins lightly
pigmented but almost transparent; midway of each web in the
largest specimens is an oval, opaque, whitish patch, possibly lu-
minous; these patches are invisible in preserved fish.

Proportions: Depth in length 16.5 to 22; head in length
6.5 to 8.5; head minus neck2 in length 11.3 to 15; eye in head 7.3
to 8.8 (covered by skin) ; eye in head without neck 4 to 5.2 ; snout
in head 5.5 to 6.5 ; snout in head minus neck 3.2 to 3.6 ; snout to
dorsal origin in length 3.8 to 4; snout to anal origin in length
1.8 to 1.9.

Teeth : Small, conical, in three to five irregular rows on
both jaws, dying out posteriorly. The pre-vomerine band is con-
tinuous with that of the maxillary, and also consists of three to

2 Neck measured from first lateral line pore to posterior end of gill-slit.

6 Zoologica : N. Y. Zoological Society [XX; 1

Fig. 3. Derichthys serpentinus. A, adolescent, 58 mm. ; B, transitional adolescent, 98 mm. ; C, adult, 268 mm. The relative
size of the specimens is indicated by the straight lines.

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 7

Y

-I

C

Fig. 4. Heads of Derichthys serpentinus. A, adolescent, standard length 58 mm. ;
B, transitional adolescent, standard length 98 mm. ; C, adult, standard length 268 mm.
The relative size of the specimens is indicated by the straight lines.

8

Zoologica: N. Y. Zoological Society

[XX; 1

five rows. There is a horseshoe-shaped group on the vomer
proper, separated from the pre-vomerine group; the horseshoe
is composed of an anterior mass of about a dozen teeth, the in-
ner ones sometimes rudimentary, and two posteriorly directed
rows of five to ten teeth each.

Fins : Pectoral 13, slightly longer than the combined lengths
of snout and eye. Dorsal 226 to 256, originating at a point about
one-fifth of the length of the fish from snout; the posterior dor-
sal rays, occupying about the last third of the caudal peduncle,
are abruptly shorter, less than half as long as the others. Anal
155 to 180, beginning at or immediately behind the middle of the
body, its rays much shorter than the anterior dorsal rays. Caudal
short, less than diameter of eye, rounded, confluent with dorsal
and anal ; 10 true caudal rays.

Nostrils : Both nostrils are situated dorso-laterally, divid-
ing the snout into almost equal thirds ; the posterior one is about
a third again as large as the anterior.

Pores and Lateral Line: On the head the pores are ar-
ranged in the following characteristic manner : Tip of snout, one
pair, very large, tubular, directed forwards or slightly upwards
(this was mistaken for a pair of nostrils in the type description
of Derichthys iselini) ; above each anterior nostril, two small
pores ; below same, one small pore ; on each side of snout profile,
above posterior nostrils, one pore, moderate; above each orbit,
three ; on each side of top of crown near level of posterior border
of eye, one; on each side of snout, behind anterior nostril, two;
below anterior corner of orbit, one; below posterior corner of
orbit, two; along mandible to end of gape, eight; continuing
this line posteriorly to level of lateral line origin, four, the fourth
above and slightly behind the third; between the first lateral
line pores of each side, extending dorsally along the boundary
between head and neck, three (one being median). In addition
to these pores, the head is also conspicuously marked with
equally characteristic groups of embossed striations. The lines
of each group are parallel, some groups being horizontal, some
vertical; usually there are six or seven lines in a group (Fig. 4).
The most posterior groups are located close behind the eye, with
the exception of an inconspicuous series a short distance be-
hind the pectorals. The striations are probably associated with

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 9

the lateral line system. The lateral line begins at the same level
as the “neck,” well in front of the pectoral. For a short distance
it runs near the dorsal profile, but soon descends to the mid-line,
which it subsequently follows. There are 80 to 90 pores, all
slightly tubular, stopping at a little more than a head’s length
in front of the caudal base. The course of the lateral line be-
tween the pores can be traced by a prominent ridge.

Myomeres and Vertebrae: 126 to 130.

Branchiostegals : 7.

Osteology : The general skeletal characters of Derichthys
may be summarized as follows: An expanded, dentigerous pre-
vomer united with the vomer by a narrow isthmus; frontals
fused or ankylosed; supraoccipital present; palato-pterygoid
slender, vestigial ; four pectoral radials ; ribs absent ; caudal ver-
tebrae without lateral transverse processes in addition to the
haemal arches. The following detailed description is derived
from a cleared and stained adult female, 268 mm. long, in the
Bermuda collection.

The entire skeleton is fairly well ossified with the excep-
tion of the posterior rays of the vertical fins and the major part
of the vertebral column. Only toward the tip of the tail do the
centra show more than faint traces of stain. These last verte-
brae, however, along with the jaws and components of the hyoid
and branchial arches, are the most .strongly ossified bones in the
body.

Skull: (Fig. 5). Although the skull has not taken up much
of the scarlet stain, still it is unquestionably well ossified, as it is
very firm. The most striking characteristic of the skull and of
the entire head as well is the prevalence of consolidated bones.
The frontals are ankylosed; the hyomandibular, quadrate and
preopercle are fused to form a single unit, and there is no hint of
a separate articular or angular.

Unlike Trewavas’s specimen from South Africa (Trewavas,
1932, p. 641), in which the frontals were united by suture, in
all three of the cleared Bermuda specimens they are firmly anky-
losed together, without a sign of division. They extend forward
almost as far as the anterior margin of the orbit, and then bend
downward behind the ethmoid, dying out just above the vomer
itself. The parietals are more than equal to the frontals in area,

10 Zoologica: N. Y. Zoological Society [XX; 1

Fig. 5. Derichthys serpentinus. Bones of the head of adult female, standard length 268 mm. Upper left, dorsal view of
skull; upper right, ventral view of upper jaw; lower, lateral view of head. (All x 5).

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 11

00

.2

£

o

fH

'V

cS

3

’o

>>

w

£

Q

.ii

E

12

Zoologica: N. Y . Zoological Society

[XX; 1

though the epiotics, posterior to these, are only slightly larger
than the tiny supraoccipital. The elongate pterotics, laterally
placed, extend the full length of the parietals and more than
half that of the frontals. Anterior to the pterotics are the tri-
angular orbito-sphenoids. Below, the posterior half of the para-
sphenoid, much broadened, forms the anterior part of the floor
of the brain-case, the unusual length of this bone and of the
pterotic being necessitated by the forward position of the eye in
connection with the elongation of the jaw. This also brings about
the unusual separation of the orbito-sphenoid and the sphenotic,
which is located slightly behind the junction of the frontal and
parietal.

Palato-pterygoid Arcade: (Fig. 5). The hyomandibular is
much fenestrated, and fused firmly with both the quadrate and
preopercle. The upper anterior arm is short, articulating only
a little behind the vertical from the anterior edge of the epio-
tics; a short projection from the ventral margin of the arm
articulates with the opercle ; the third arm, directed antero-ven-
trally, is fused to the inner face of the quadrate, only a rim of
the bone projecting above it; this arm extends almost to the jaw
angle. The quadrate, articulating with the undifferentiated
angular, is short, and broad posteriorly. The palato-pterygoids,
although well ossified, are of needle-like slenderness and seem to
serve very little practical use as they do not connect directly with
any bone.

Jaw Apparatus: (Fig. 5). The jaws and teeth are the most
strongly ossified elements in the body. The “praemaxillary re-
gion of the praemaxilla-ethmo- vomer” of Trewavas (1932 p.
641) may be termed more conveniently the pre-vomer. In her
specimen of 160 mm. from South Africa this bone is “united
with the ethmo-vomerine region by a narrow isthmus.” In Ber-
muda specimens both larger and smaller than hers, this isthmus,
though very distinct, is unstained. The pre-vomer forms the en-
tire front of the broad, obtuse snout and articulates with the
maxillaries in close-fitting joints, a pair of grooves in the dorsal
surface of the pre-vomer receiving a projection from each max-
illary. The latter show broad ventral surfaces anteriorly, hold-
ing several rows of teeth; posteriorly, however, they are much
attenuated. They reach well behind the posterior border of the

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 13

eye. The vomer proper is oblong, with the moderately broad,
flat parasphenoid arising from its excavated posterior border.
To its dorsal surface the forked end of the ethmoid is firmly
attached by suture, and not ankylosed as in Trewavas’s speci-
men. The ethmoid extends upward and backward, arching in its
posterior part to form a foramen with the frontal, which it
overlies throughout its length. The lower jaw is considerably
shorter than the upper and about twice as deep, and extends a
full third of its entire length behind the end of the maxillary.
There are no lines of demarcation into articular and angular.

Opercular Apparatus: (Fig. 5). The preopercle is com-
pletely separated from the rest of the series, being fused to the
ventral side of the hyomandibular by a large thin flange arising
from the antero-dorsal side of the typically tubular portion of
the bone. The interopercle is an elongate bar free from both
preopercle and subopercle. The latter is very slender and al-
most crescentic, lying immediately beneath the horizontally
placed opercle and following its curving outline. The opercle
reaches the vertical from the middle of the second vertebra.

Hyoid Arch: (Fig. 6). There is no trace of an interhyal,
and the epihyal and hypohyal elements are invisibly consolidated
in the ceratohyal. There are seven branchiostegals, all arising
from the posterior half of the ceratohyal, and all with more or
less swollen bases. The seventh is very slender, is not attached
directly to the ceratohyal, and reaches just beyond the opercle.
The ceratohyal is attached about midway of the length of the
glossohyal. The oblique posterior edge of the latter is joined by
suture to the first basibranchial. The urohyal is a slender, curv-
ing bone swollen both basally and distally. It extends to the level
of the second branchial arch.

Branchial Apparatus: (Fig. 6). There are three basibran-
chials, each separated from the preceding by more than its own
length. The first is large, half the size of the glossohyal; the
second is very small, slender; the third is again slightly longer,
though still slender. In the first two arches hypobranchials are
present. The first four ceratobranchials are all moderately
slender, the first curved. The fifth is shorter than the rest and
is attached to the inner anterior edge of the fourth, except at its
dorsal posterior end. It bears several rows of sharp teeth, about

14

Zoologica : N. Y. Zoological Society

[XX; 1

Fig. 7. Derichthys serpentinus. Vertebrae of adult female, standard length 268 mm.
Upper, 15th vertebra (5th behind pectoral origin) ; lower, 62nd vertebra (6th behind anal
origin). (Both x 8.8).

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 15

twenty-five in each row. There are four epibranchials of irregu-
lar shape. All are small, the fourth much larger than the others.
The three pharyngobranchials increase in size posteriorly, teeth
being present on the postero-ventral edge of the second and the
entire ventral surface of the third. The apparatus has no con-
nection with the vertebral column, and lies low in the branchial
cavity.

Pectoral Girdle: (Fig. 5). The supraclei thrum is located
at the level of the sixth vertebra, half of its length projecting
above the centrum. There is no connection with the column.
The cleithrum is broad and strong, the center edge straight, the
posterior convex. Its upper tip overlaps the lower half of the
supracleithrum and is placed close behind it. Coracoids are
absent. There are four radials, small but well ossified. The first
of the thirteen pectoral rays is slightly separated from the
others and has an enlarged base. The fin is far removed from
the pectoral girdle.

Vertical Fins and Supports: (Fig. 7). The dorsal fin

originates at the level of the 23rd vertebra, the anal at the 56th.
The basal thirds of the rays of both fins are ossified as far back
as the middle of the anal fin, but posterior to this neither dorsal
nor anal shows any bony deposit. Baseosts are well developed,
and show ossification more than a dozen rays behind the most
posterior finray that shows any stain. The last, unossified
baseosts have no definite terminations, each splaying out distally
and merging with the adjacent elements, while the bases of the
corresponding finrays are indistinguishably merged in the same
web of cartilaginous tissue. There are usually two baseosts and
rays to each vertebra. Tiny, horizontally placed radials also
show faint ossification as far back as the rays, and can be dif-
ferentiated slightly behind them.

Vertebral Column: (Fig. 7). The slight ossification of the
centra and neural arches contrasts with the strongly stained ver-
tebral appendages. Only the last 20 centra of the 130 vertebrae
in the large female under discussion show more than traces of
stain. The first vertebra is only half the size of the third, the
second intermediate. The fourth is very slightly longer than
the third, and this size is maintained until the origin of the anal

16 Zoologica: N. Y . Zoological Society [XX; 1

fin. From here to the tail there is the usual gradual decrease in
size to the urostyle.

Throughout the column the neural arches are very large,
equalling or exceeding the centra in height. Each arch inter-
locks with the one before by means of an anterior projection
which underlaps the preceding posterior edge. The neural spines
of the first nine vertebrae are diverse and specialized. The first
is split longitudinally, one half falling immediately behind the
other, and joined to it basally by suture. Thg pair is short, pos-
teriorly directed, and arises at the posterior edge of the arch.
The second spine is similar but single. The third, fourth and
fifth arise from the anterior half of the arches and are short
and forked, one prong behind the other. The sixth, at the level
of the pectoral girdle, is represented only by a minute bump.
The seventh is a small, posteriorly directed spine in the middle
of the arch; the eighth is equal in size, but anteriorly directed;
the ninth is low and again forked, arising from the posterior
half of the arch. From the tenth to the column’s end the spines
are well developed, unforked, backwardly directed, and situated
at the posterior end of the arch. They are longest in the region
of the anterior part of the anal fin, where the length of each is
more than twice that of neural arch and centrum combined.
Near the caudal base they are relatively much reduced in size,
but very strongly ossified. Epineurals, forked basally with only
the inner prong attached to the neural arch, are present on every
segment except the first and the specialized caudal vertebrae.
They are ossified, however, only slightly beyond the middle of
the anal fin, though their outlines are traceable almost to the
caudal base.

The parapophyses are strong, short spines directed obliquely
outward, their broad basal portions arising from the midst of
the ventral halves of the centra. There is no trace of ribs. The
first three haemal arches lie in front of the anus and lack all
trace of haemal spines. Behind the anus, however, they promptly
increase in length to equal that of the neural spines, the arches
arising from the anterior part of the parapophyses. The first
epipleural, an unattached sliver of bone, is found at the thirty-
ninth vertebra, slightly behind the middle of the abdominal
cavity. The succeeding ones increase in size posteriorly, the

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 17

Fig. 8. Derichthys serpentinus. End of vertebral column and base of caudal fin in
adult female, standard length 268 mm. (x 27.5).

tenth being the first with a forked base. At the origin of the
anal (fifty-sixth vertebra) they attain their maximum size, and
from here on correspond to the epineurals, ossification dying out
similarly behind mid-anal. Their outlines, however, are trace-
able, like those of the epineurals, almost as far back as the spe-
cialized caudal vertebrae.

End of Vertebral Column and Caudal Fin: (Fig. 8). Two
unusual characteristics of the tail structure of this eel are, first,
the almost complete absence of osseous tissue, and second, the
persistance of the neural arch and spines throughout the entire
dorsal length of the urostyle. The only caudal specialization of
centra, neural arches and spines is a gradual reduction in size.
There is a radical change posteriorly in the haemal arches and
spines. On the 125th or fifth pre-urostyle vertebra, the haemal
arch base, in typical fashion, extends almost the full length of
the centrum, narrowing in the center to form a well-marked
bay with the proximal portion of the spine. Posteriorly the back-
ward extension of the arch decreases and finally on the last ver-
tebra disappears. There is no open arch on the penultimate ver-
tebra, the two lateral elements being quite unjoined and very
unlike.

The urostyle extends back as a straight rod for a length

18

Zoologica: N. Y. Zoological Society

[XX; 1

greater than that of the preceding vertebra. On its dorsal sur-
face it supports two well developed neural arches. At its end,
and bounded above by the last neural spine, is the first hypural,
supporting five caudal rays. Below is a larger cartilaginous
area with an extensive central foramen, irregular but giving no
definite hint marking a division into separate elements. So we
must indicate the whole of this area as the second hypural, sup-
porting the succeeding four caudal rays. The haemal spine of
the last vertebra is extended backward into the long, slender
third hypural from which arises the last and tenth caudal ray.

Pore System: The pores of the snout are supported by tiny
bony tubules connected with unossified channels, while the lower
jaw has a perforation corresponding to each pore in that area.
Connection from both regions is made with the lateral line in
the usual manner, via the post-temporal canal and the pre-
opercle. The more posterior pores of the head and those of the
neck and lateral line have no bony support.

Fig. 9. Derichthys serpentinus. Digestive and reproductive systems in adult female,
standard length 268 mm. (x 1.2).

Digestive System: (Fig. 9). None of the digestive organs
is pigmented, although the lining of the coelom has a thin scat-
tering of chromatophores. The oesophagus opens directly into
the stomach, slightly in front of the posterior tip of the liver.
The stomach, very slender when not distended by food, barely
reaches the anus, lying to the left of the equally slender intestine.
The pyloric canal connecting the two organs has no caeca, extends
obliquely forward. The liver is single-lobed, the left half the
longer, and lies as usual ventral to the oesophagus and the oval
gall bladder. The bile duct is short, entering the swollen pyloric
region of the intestine at its most anterior point. The pan-
creatic tissue is practically indistinguishable from that of the
liver.

Reproductive System: (Fig. 9). The ovaries originate

at the level of about the middle of the liver and extend posteri-

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 19

orly against the dorsal wall of the coelom to a distance behind
the anus equal to two-thirds the length of the liver. The left
ovary is always slightly longer than the right. In the only speci-
men near breeding condition, there is a total of about 4,100 eggs,
each measuring about .75 mm. in diameter. In addition there is
at least an equal number of very minute eggs, each at most .14
mm. across. This circumstance makes it appear very probable
that these deep-sea eels breed more than once.

Development

Material: Adolescents and transitional adolescents pre-
dominate in the Bermuda collections; larvae and post-larvae are
absent :

Adolescents: 55 to 90 mm. — 8 specimens (Figs. 3A, 4A) .

Transitional Adolescents: 98 to 198 mm. — 9 specimens
(Figs. 3B, 4B) .

Adults (Females) : 255, 268 mm. — 2 specimens (Figs. 3C,

4C) .

Key to the Growth Stages :

A. Body more or less flattened, semi-leptocephaloid ; pigment

lacking Adolescent

AA. Body of adult form ; pigment present.

B. Gonads very inconspicuous, pigment incomplete or
pale; skeleton not fully ossified

Transitional Adolescent

BB. Gonads well developed; pigmentation complete; skele-
ton fully ossified Adult

Changes Occurring During Growth : The smallest speci-
mens (55 to 90 mm.) are typical eel adolescents, having no trace
of larval teeth, the fins complete and in practically their final
positions, and bodies, though somewhat flattened, well beyond
the leptocephalid stage and almost as slender, relatively, as in
the adults. The following differences are apparent, however,
when compared with transitional adolescents and adults: Pig-
ment is entirely lacking except for a line of minute chromato-
phores — doubtless remains of larval pigment — extending from

20

Zoologica: N. Y. Zoological Society

[XX; 1

the anus to the caudal in the mid-line; this is lacking in larger
adolescents. The body is otherwise perfectly white and opaque,
except for rosy iridescence on the head, opercles, and, irregu-
larly, along the sides; the abdomen is usually more brightly
iridescent, with blue and violet tints predominating. In the
largest adolescents there is a patch of pigment on the crown
beneath the epidermis. The head is slightly larger than in older
specimens; eye and snout, however, are of adult proportions.
The teeth are feeble, but the full number is present, with the
exception of the anterior cluster of vomerine teeth. The anal fin
may originate very slightly in advance of the middle of the
body, instead of at or behind this point. The finrays are more
easily countable at this stage than later on, as the membranes
are still thin. All of the pores, both cephalic and lateral line, are
present, fully formed on the head but rudimentary along the
lateral line. The striations are also developed, but are incon-
spicuous. The adolescent shows a moderate amount of ossifica-
tion, the jaws and teeth being very strongly stained, and the jaw
supports, hyoid and branchial arches, and pectoral girdle only
slightly less firmly ossified. The brain-case, opercles, basal pec-
toral rays, vertebral column and external cephalic canal bones
show moderate amounts of bony deposition in the larger adoles-
cents, but in no specimen of this stage do the vertical fins, their
baseosts or their radials show any trace of stain. In contrast to
corresponding bones in the adult, the seventh branchiostegal ray
is longest and strongest instead of shortest and weakest, while
the urohyal is perfectly straight, instead of deeply curved. The
digestive system differs from that of the adult only in the slightly
shorter stomach, which in the smaller specimens ends a full
snout’s length in front of the anus. The gonads are rudimentary.

In the transitional adolescents (98 to 198 mm.) the entire
skin is frequently tinged with warm pink, and true dark pig-
ment first appears, spreading from the top of the neck backward,
downward and forward. Specimens measuring 160 mm. and
over are completely covered with pigment, but these fish are
slightly paler than adults. There are small reductions in the
length of the head and in the depth. The skeleton gradually
becomes more strongly ossified. The digestive system does not

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 21

differ from that of the adult; the gonads are distinguishable, but
very slightly developed.

The two fully adult specimens of the collection differ in
proportions very little from the younger fish, as may be seen
from the following summary of measurements made on all the
specimens in the collection:

Growth Stage

Length

Length

Length

Head

Head

Length

Length

Depth

Head

Head

minus

Neck

Eye

Snout

Snout

to

Dorsal

Snout

to

Anal

Adolescent

18—

6.5—

10—

8.2—

5—

3.9—

2—

(55 to 90 mm.)

25.7

7

11.8

10.2

6.4

4.3

2.3

Transitional

16.5—

6.5—

10.4—

7.1—

5.4—

3.7—

1.8—

Adolescent
(98 to 198 mm.)

25

7.7

12.5

8.8

6.7

4.2

2

Adult

19—

8.1—

14.5—

8.2—

5—

3.8—

1.8

(255, 268 mm.)

22.5

8.4

15

8.8

6.2

4

Ecology

Seasonal Distribution : Eleven of the eighteen specimens
were taken in September, four in June, three in August.

Vertical Distribution : Derichthys occurred only between
500 and 1,000 fathoms, at an average depth of 755 fathoms. No
correlation is seen between season and depth.

Abundance : Derichthys is rare among the deep-sea fish of
Bermuda, only one occurring in every 55 nets drawn between
500 and 1,000 fathoms, the Bermuda limits of its vertical dis-
tribution.

Sociability: Not more than a single specimen was ever
taken in the same net.

Food: In five stomachs were traces of crustaceans, usually
unquestionably shrimps and recognizable in one case as Sergestes
sp. The latter measured 64 mm. in length, and had been swal-
lowed tail first by a 132 mm. Derichthys. Unrecognizable re-
mains of food were usually present in the intestines.

Enemies : A 198 mm. Derichthys had a number of parasitic
worms embedded in the stomach wall.

Viability : No Derichthys has ever been taken alive.

22 Zoologica: N. Y. Zoological Society [XX; 1

Study Material

The following list gives the catalogue number, net, depth in
fathoms, date, length and growth stage of each specimen of
Derichthys serpentinus taken by the Bermuda Oceanographic
Expeditions. All were caught in the cylinder of water off the
Bermuda coast described in Zoologica, Vol. XVI, No. 1, p. 5.
“Trans. Adol.” stands for “Transitional Adolescent.”

No. 10,297; Net 148; 700 F.; June 1, 1929; 85 mm.; Adolescent.

No. 10,450; Net 167; 800 F.; June 14, 1929; 178 mm.; Trans. Adol.

No. 10,534; Net 177; 600 F.; June 17, 1929; 80 mm.; Adolescent.

No. 10,953; Net 219; 600 F.; June 25, 1929; 104 mm.; Trans. Adol.

No. 13,518; Net 475; 800 F.; Sept. 13, 1929; 159 mm.; Trans. Adol.

No. 13,712; Net 495; 800 F.; Sept. 23, 1929; 102 mm.; Trans. Adol.

No. 17,501; Net 822; 600 F.; Sept. 1, 1930; 58 mm.; Adolescent.

No. 17,778; Net 837; 600 F.; Sept. 3, 1930; 55 mm.; Adolescent.

No. 18,611; Net 890; 1,000 F.; Sept. 15, 1930; 62 mm.; Adolescent.

No. 19,281; Net 941; 1,000 F.; Sept. 24, 1930; 198 mm.; Trans. Adol.
No. 19,451; Net 953; 1,000 F.; Sept. 26, 1930; 133 mm.; Trans. Adol.
No. 19,547; Net 964; 600 F.; Sept. 29, 1930; 116 mm.; Trans. Adol.
No. 21,884; Net 1,121; 500 F.; Aug. 3, 1931; 98 mm.; Trans. Adol.

No. 22,680; Net 1,209; 1,000 F.; Aug. 20, 1931; 268 mm.; Adult.

No. 22,975; Net 1,244; 800 F.; Aug. 31, 1931; 255 mm.; Adult.

No. 23,110; Net 1,261; 600 F. ; Sept. 4, 1931; 90 mm.; Adolescent.

No. 23,230; Net 1,278; 700 F.; Sept. 9, 1931; 79 mm.; Adolescent
No. 23,611; Net 1,317; 900 F.; Sept. 17, 1931; 104 mm.; Trans. Adol.

Synonymy and References

Derichthys serpentinus :

Gill, 1887, p. 433. (1 specimen; 8 in.; 1,022 fathoms; 39° 44'
30" N. Lat., 71° 04' W. Long.; off New Jersey; type speci-
men) .

Goode and Bean, 1895, p. 161, fig. 169. (Supplementary type
description) .

Parr, 1934, p. 32, fig. 10. (1 specimen; length not given;

1,05(1-1,100 metres; 25° 39' N. Lat., 77° 18' W. Long.;
Bahamas. Discussion of synonymy of D. iselini with D.
serpentinus.)

Derichthys iselini:

Borodin, 1929, p. 110. (1 specimen; 165 mm.; 1,000-0 fathoms;
50° 41' N. Lat., 27° 17' W. Long.; Middle North Atlantic,
three-fifths of distance between Newfoundland and Scilly
Isles) ,

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 23

Borodin, 1931, p. 75, pi. 3, figs. 4-6. (Supplementary type
description) .

Grammatocephalus kempi:

Norman, 1930, p. 339, fig. 34. (1 specimen; 160 mm.; 850-950
metres; 33° 50' to 34° 13' S. Lat., 16° 04' to 15° 49' E.
Long.; off Cape Town).

Derichthys kempi:

Trewavas, 1932, p. 641, text-fig. 2 (Supplementary descrip-
tion of the type specimen of Grammatocephalus kempi and
remarks on the relationships of the family.)

A bibliography will be found on p. 50 of the present volume.

Zoologica, Volume XX, Number 2

DEEP-SEA FISHES OF THE BERMUDA
OCEANOGRAPHIC EXPEDITIONS

Family NES30RHAMPHIDAE1
William Beebe

(Figs. 10-22 incl.)

Contents

Introduction p. 25

Taxonomy p. 25

Detailed Discussion : Nessorhamphus ingolfianus

Bermuda specimens: General trawling data p. 28

Specimens previously recorded p. 28

Description of adult p. 28

Development p. 29

Ecology p. 48

Study material p. 49

Synonymy and references p. 49

Bibliography p. 50

INTRODUCTION

All of the remarks in the Introduction to the preceding
paper on the family Derichthyidae apply also to the present
account. They will be found on p. 1 of this volume.

Family NESSORHAMPHIDAE Schmidt 1931

Body anguilliform, slender, more or less cylindrical anteri-
orly but with caudal region rather compressed; anus behind
middle of length ; scales absent ; lateral line with distinct pores ;
snout very long, flattened, spatulate, projecting far beyond the
narrow lower jaw, its tip enlarged and holding the olfactory sac ;
anterior nostril terminal, posterior slightly behind it, dorso-

1 Contribution No. 479, Department of Tropical Research, New York Zoological Society.

25

26

Zoologica: N. Y. Zoological Society

[XX; 2

lateral; neither nostril tubular; teeth conical, cardiform, in
bands on both jaws, on vomer proper and on intermaxillary
extension of vomer ; tongue scarcely or not at all free anteriorly ;
branchial apertures widely separated slits of moderate size, in-
serted immediately in front of the well developed pectorals;
dorsal beginning behind head, far in advance of anal; dorsal
and anal confluent with caudal, which is short but well de-
veloped. One genus.

Genus Nessorhamphus Schmidt 1930

With the characteristics of the family. One species de-
scribed; the Indo-Pacific forms may, however, prove to be dis-
tinct. General range: The warmer and saltier parts of the
Atlantic, Indian and Pacific Oceans.

Nessorhamphus ingolfianus Schmidt 1930 (Schmidt 1912)

Specimens Taken by the Bermuda Oceanographic
Expeditions

Twenty-one specimens; April to September, 1929 to 1931;
400 to 1,000 fathoms; from a cylinder of water eight miles in
diameter (five to thirteen miles south of Nonsuch Island, Ber-

A

B

Fig. 10. Stages in the development of N essorhamphus ingolfianus. A, egg, early stage ;
B, egg, late stage. (Both x about 16; after Taning, in Schmidt 1930).

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 27

i 1

H

Fig. 10 (continued). Stages in the development of N essorhamphus ingolfianus. C, Pre-
larva, 3 to 5 hours old ; D, larva, 28 mm. ; E, post-larva, 69 mm. ; F, adolescent, 81 mm. ;
G, transitional adolescent, 92 mm. ; H, transitional adolescent, 166 mm. The relative size
of the specimens is indicated by the straight lines. (C, after Taning, in Schmidt 1930 ;
D-H incl., from specimens taken by the Bermuda Oceanographic Expeditions).

28 Zoologica: N. Y . Zoological Society [XX; 2

muda), the center of which is at 32° 12' N. Lat., 64° 36' W.
Long.; standard lengths from 26 to 166 mm.

Specimens Previously Recorded

Many thousands, the majority young and immature forms,
were taken by the Danish research vessels within the past twenty
years. Dr. Schmidt only found time, however, to publish brief
preliminary descriptions. The eggs and smaller larvae were
found to be pelagic, and the older forms bathypelagic. Range:
Warmer parts of North Atlantic. Length: Egg to 248.5 mm.,
the type specimen.

Description of Adult

Color: Brownish, sides with bluish tinge.

Proportions: Depth in length 30 (small specimens) to 25
(large specimens) ; head in length 6 to 7.2; eye in head about
10.6; snout in head about 2.4; snout to dorsal origin in length
4.6 to 4.8; snout to anal origin in length about 1.8.

Teeth : Conical, cardiform ; maxillary and mandible with
narrow bands formed of from one to four irregular, broken rows
of teeth, the band being single-rowed only in the extreme pos-
terior part of the jaw, a maximum of 60 to 90 teeth in each row ;
a patch of similar teeth, about 35 or 40, on the intermaxillary
extension of the vomer, and, posterior to these after an interval,
an elongate patch on the vomer proper ; the latter series are the
last to develop; in the large, type specimen they number more
than 100 according to Schmidt’s figure (1930, pi. IV).

Fins : Pectoral rays 13, at least six times as long as eye in
perfect examples, but usually broken; dorsal rays 276 to 291,
the rays longest on the caudal peduncle, opposite the middle sec-
tion of the anal fin ; dorsal origin less than a snout’s length be-
hind pectoral base; anal rays 160 to 175, the rays much shorter
than those of the dorsal; caudal fin truncated, continuous with
dorsal and anal.

Pores and Lateral Line: Head with conspicuous mucous
pores; fine parallel ridges near tip of snout and before and be-
hind eye. About 132 pores in lateral line, which commences

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 29

above the mid-line, but coincides with it from about the middle
of the length posteriorly.

Branchiostegals : 6 to 7.

Myomeres and Vertebrae: 150 to 159 (72 to 73 pre-anal) .
Osteology; Digestive System: Descriptions of these as
they occur in slightly immature specimens commence on p. 37.

Development

Material: The Bermuda collection consists of all stages
from moderately young larvae to large transitional adolescents.
Eggs and pre-larvae were described in Dr. Schmidt’s paper by
Dr. Taning (1930, p. 275) and his description will be quoted
below, so that the various growth stages of this eel will be com-
pletely summarized in the present paper. We do not know, how-
ever, whether the 248.5 mm. type specimen was fully mature.
The Bermuda material is distributed as follows:

Larvae: 26, 28 mm.— 2 specimens (Figs. 10D, 11A).
Post-larvae: 68 to 72 mm. — 5 specimens (Figs. 10E, 11B).
Adolescents: 78, 81 mm. — 2 specimens (Figs. 10F, 11C).
Transitional Adolescents: 80 to 166 mm. — 12 specimens
(Figs. 10G, 10H, 11D, HE).

Key to the Growth Stages:

A. Body leptocephaloid (depth in length not more than 22; usu-
ally very much less) ; no general pigment on body.

B. Larval teeth present; snout scarcely or not at all pro-
longed beyond mandible; anal origin far back, at 118th

to 121st myomere Larva

BB. Larval teeth absent; snout showing characteristic shape
and prolongation.

C. Anal origin not in final position, but located be-
tween about 116th and 76th myomeres; depth in

length 7.4 to 9.6 Post-larva

CC. Anal origin in final position, between 74th and 72nd

myomeres; depth in length 18 to 20 Adolescent

A A. Body anguilliform (depth in length 29 or 30) ; general pig-
ment appearing on body Transitional Adolescent

30

Zoologica: N. Y. Zoological Society

[XX; 2

E

Fig. 11. Heads of N essorhamphus ingolfianus. A, larva, standard length 28 mm. ; B,
post-larva, standard length 69 mm. ; C, adolescent, standard length 81 mm. ; D, transitional
adolescent, standard length 92 mm. ; E, transitional adolescent, standard length 166 mm.
The relative size of the specimens is indicated by the straight lines.

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 31

(In fully pigmented, large specimens the body evidently, as is
usual, becomes secondarily deeper, the depth being contained ac-
cording to the type description only about twenty-five times in
the length.)

Egg and Pre-Larva (From Taning’s description, Schmidt,
1930, p. 275) : “The egg is a typical, highly transparent, floating
muraenoid-egg with a wide perivitelline space; vitellus nearly
colourless with a light tinge of yellow. The membrane of the
egg is for a deep-sea muraenoid-egg rather thick. Vitellus
vesicular with one large oil globule, which at a very early stage
in development — prior to advanced cleavage of the germinal
disc — is a cluster of small oil-globules. The size of the egg is:
Diameter 2.4-2.70 mm., vitellus 1.20-1.35 mm. and oil globule
0.42-0.48 mm. [see Figs. 10A, B]. The pre-larva, which leaves
the egg at an early stage (abt. 7 mm.) has the oil globule sit-
uated anteriorly in the yolk-sack; in this stage it has abt. 84
abdominal segments. Pigment not present.

“In the Sargasso Sea the eggs were found in large quan-
tities, especially numerous about May.

“The eggs resemble to a high degree those of the muraenoid
fish to which Leptocephalus anguilloides Schmidt 1916 belongs.
The eggs of this species are however smaller (less than 2.40
mm.) and have a larger vitellus (1.50-1.65 mm.), with accord-
ingly a smaller peri-vitelline space.’'

The following resume of the changes occurring during
growth is based only upon the Bermuda material; where com-
parison was possible from the short description of the larva and
from the series of photographs, the present specimens checked
perfectly with corresponding ones in the Danish collection. A
table of proportions will be found on p. 47.

Larva : The two larvae, 26 and 28 mm. in length, are typical
leptocephali. Pigment is present only on the tail, in an irregular
series of tiny chromatophores in and near the mid-line, and
sparsely along the posterior parts of the fin bases. The larva is
moderately deep ; the tail tip rather attenuated ; the head fairly
small ; eyes vertically elongate, larger than in succeeding stages ;
snout long, but no longer than the mandible; the latter already
shows its characteristic narrowness. The nostrils are located

32

Zoologica: N. Y. Zoological Society [XX; 2

oo°<?0>0

po

O 0 “ O o °o

o o

’ O <J

° ° o O Oo'
° o°« oo0

'Oo^ooV

o 0 O * o

Fig. 12. N essorhamphus ingolfianus. Dentition of upper jaw in transitional adolescent,
standard length 142 mm. (x 11).

just in front of the eye. The larval fangs number eight pairs in
the upper jaw and nine in the lower; in the maxillary the last
four teeth on each side are much smaller than the anterior ones ;
in the mandible the same is true of the last three pairs. The
pectoral has the usual fleshy base, fringed with raylets. The
bases of both dorsal and anal fins are elevated, the rays not yet
well enough developed to be counted with accuracy. The gut
ends and the anal commences far back, on the posterior tenth
of the body, lying between the 118th and 121st myomeres. The
dorsal is traceable almost as far forward as the middle of the
length. The caudal rays are well developed. Cephalic and lat-
eral pores are invisible. The gut, suspended below the body wall,
is entirely unpigmented; it is a simple tube with the anlage of
the liver and other organs barely appearing.

Osteology: (Figs. 13, 14, 15). These larvae of course show
no trace of ossification, but the principal cartilaginous elements
of the head are traceable even in the unsectioned, preserved
specimens, and are full of interest. The chief difference between
the heads of larva and adult lies in the proportions of the chon-
dro-cranium: the nostrils, as noted above, are situated not ter-
minally, but close in front of the orbits. The frontals, fused even
at this early stage, do not enter into either the roof of the snout
nor that of the skull proper, being confined to the interorbital
region. The future growth of the snout tissue will hence take
place chiefly in the now narrow area between nostril and orbit,
through the anterior projection of the frontals and posterior
growth of the ethmoid, so that eventually the nostrils will be
thrust far forward. As in the mature fish, the parietals are
large and well marked. A relatively large supraoccipital can
be traced. The otic region is not yet differentiated into epiotics
and pterotics, but occupies the entire posterior and lateral por-

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 33

tions of the brain-case. Viewed from above, the two epiotic re-
gions diverge posteriorly from the supraoccipital into large
wing-like processes.

All of the elements of the palato-pterygoid, jaw and oper-
cular apparati are discernible and differ from the correspond-
ing bones of the adult only in their typically immature lack of
shapeliness and their general tendency to robust stockiness. The
opercular apparatus in particular shows generalized characters,
the opercle being fan-shaped, bounded ventrally and posteriorly
by the broad subopercle. As in generalized fish also, the entire
apparatus does not extend behind the vertical from the posterior
margin of the brain-case, while in the adult it projects much
farther. In the larva, too, a well defined gill opening appears in
its normal place, beneath the gill flap formed by opercle and
subopercle. The hyoid apparatus likewise has the elements of
the adult, save that no branchiostegal rays are visible. The
ceratohyal arises relatively farther back along the glossohyal.

The branchial apparatus has the rudiments of the two hypo-
branchials, five ceratobranchials and four epibranchials of the
adult, but the pharyngobranchial elements are only question-
ably identifiable. The generalized shape of the gill arches is
apparent, all of them being vertical and consecutive, and all are
crowded beneath the broad main column of the hyomandibular
and the region immediately behind — far forward of their subse-
quent location in the “neck” of the metamorphosed eel.

Post-Larva: The post-larva is slightly more slender than
the larva, larval teeth are missing and permanent teeth appear-
ing, the snout assumes its characteristic configuration, with the
nostrils approaching ever closer to their final position, and —
an invariable character of this stage among the Apodes — the
anus and anal fin are in the act of moving forward, and lie
somewhere between about the 116th and the 76th myomeres.
The rays are easily countable at this stage, but are relatively
short, with their bases still elevated. The same applies to the
dorsal rays, which have also migrated, so that before the end
of the post-larval stage the fin originates in its final position back
of the head. The cephalic pores are now discernible. The gut is
almost completely enclosed within the body cavity.

34

Zoologica: N. Y. Zoological Society

[XX; 2

c

Fig. 13. Nessorhamphus ingolfianus. Skull of larva, standard length 28 mm.; dorsal view, (x 42)

soc

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 35

Fig. 14. Nessorhamphus ingolfianus. Skull of larva, standard length 28 mm.; lateral view, (x 42).

36 Zoologica: N. Y. Zoological Society [XX; 2

=r

i-

_Q m
W __Q

Fig. 15. Nessorhamphus ingolfianus. Hyoid and branchial apparatus of larva, standard length 28 mm. (x 100).

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 37

Adolescent: In the adolescent stage the fins are in the
same relative positions as in the transitional adolescent and
adult; the body is less than half as deep as in preceding stages
and a distinct thickening is apparent, although the general ap-
pearance is still distinctly leptocephaloid. All traces of larval
pigment have disappeared, and permanent coloration is appear-
ing on the tip of the snout, and, less densely, along its top and the
crown of the head, beneath the outer epidermis. As Schmidt has
remarked, in contrast to what is known in the case of many other
eels, a distinct increase in length is apparent throughout meta-
morphosis. In his series, metamorphosing stages (correspond-
ing to our post-larvae and adolescents) have a length of from
62 to 85 mm. ; the youngest post-larva and most advanced adoles-
cent in the Bermuda material measure respectively 70 and 81
mm.; a shrinkage of two or three millimetres at most seems to
be all that takes place at any point in the development.

Transitional Adolescents: The transitional adolescents,
the “glass eels” of Schmidt, have the rounded, slender form of
grown eels, but in this species pigment appears very gradually
and late, and the full number of teeth — especially along the
vomer proper — is only gradually attained. The largest Bermuda
specimen, measuring 166 mm., appears almost white to the un-
aided eye, although it has all other adult characters with the ex-
ception of complete vomerine dentition (Fig. 12) and mature
reproductive organs. Examination shows a dark spot on the
tip of the snout, and fine chromatophores scattered over the
head, jaws, cardiac and abdominal regions, the base of the anal
fin, the base of the posterior half of the dorsal fin, and the end
of the caudal peduncle; there is a fairly conspicuous, vertical
band of dark brown across the middle of the caudal rays.

Osteology : The skeleton of a specimen of 142 mm. has been
studied in detail, and compared both with younger examples of
the Bermuda collection and with the 128 mm. specimen described
and figured by Trewavas (1932). The latter differs from the
present specimens chiefly in smaller epiotics.

In the youngest of the transitional adolescents examined,
a specimen of 87 mm., the entire skeleton with the exception of
the vertical finrays showed faint ossification, strongest on the
jaws and weakest along the vertebral column. In the 148 mm.

38

Zoologica: N. Y. Zoological Society

[XX; 2

example, ossification is much stronger throughout, with the ex-
ception of the sides and base of the brain-case, the coracoids and
the radials, all of which show very faint stain, while the anterior
dorsal finrays and all of the anal rays are entirely unossified.

Skull: (Figs. 16, 17). The most interesting characteristic
of this region is the manner of the prolongation of the skull.
Unlike the snouts of the comparable Nemichthyds, the length-
ening has been brought about chiefly by the elongation of the
frontals. The brain-case is perfectly oval. The epiotics occupy
the posterior fourth of its dorsal surface. Anterior to these are
the oblong parietals, twice as long as the epiotics in our speci-
mens. The tiny supraoccipital is enclosed completely by the
epiotics and parietals. The elongate pterotics form the major
element of the sides of the brain-case. A sensory canal runs its
full length, overlapping the frontal. The frontals are completely
fused except anteriorly. About a third of their length is post-
orbital, forming the anterior fourth of the brain-case. At their
posterior border their combined width is about equal to the
length of this postorbital section. They narrow gradually, and
are extremely narrow above the eyes, then broaden slightly to
split at their junction with the ethmoid, the forked ends pro-
jecting into that bone laterally as far as the middle of the snout.
In the Bermuda specimens there is no ossification of the spindle-
shaped ethmoid anterior to the nostrils, although Trewavas
shows a strong, compact ethmovomer.

Palato-pterygoid Arcade: (Fig. 17). The hyomandibular in
this species is an irregular quadrilateral with excavated sides.
Its upper anterior angle articulates with the small, crescentic
sphenotic arising from the pterotic, beneath the postero-lateral
corner of the frontal. A broad wedge of the hyomandibular ex-
tends backward parallel with the pterotic, slightly beyond the
edge of the brain-case, its apex forming the second angle of the
quadrilateral. Diagonally opposite the sphenotic articulation a
shorter arm joins the opercle. The fourth angle is formed by
the typically hyomandibular arm which extends antero-ventrally
to join the quadrate. The latter, connecting with the angular, is
perfectly horizontal. The symplectic is large, equal to the quad-
rate in length and thickness ; it lies above the quadrate and pos-
teriorly overlaps the hyomandibular. The slender, rod-like pte-

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 39

rygoid extends between the posterior part of the quadrate and
the posterior portion of the vomer. There is no trace of a sepa-
rate palatine. The vomer lies against the ventral surface of the
ethmoid, broadening anteriorly and dipping ventrally to floor
the bulbous snout tip. Between vomer and brain-case extends
the strong and slender parasphenoid.

Jaw Apparatus: (Fig. 17) . The slender maxillary originates
at the posterior boundary of the swollen snout tip, ending close
to the vertical from the posterior margin of the eye. Posteriorly
it has a strong, downward curve. The strong dentary is notably
shorter than the snout, its tip being even with that of the maxil-
lary. It is slender anteriorly, with a slight expansion at the
symphysis. Posteriorly it broadens considerably. The articular
is of moderate size and clearly defined, although the angular
boundaries are indeterminable.

Opercular Bones: (Fig. 17). All of the opercular bones are
reduced and practically non-functional, as is usual in eels, lying
far forward of the gill opening. The preopercle, lying distinct
from and below the quadrate and hyomandibular, is slender and
tubular, forming part of the mucous canal system. Below this,
and extending beyond its posterior end, is the small interopercu-
lum, and behind this a wiry suboperculum, lying immediately
beneath the trapezoidal, oblong opercle, and curling up around
its posterior end.

Mucous Canal Bones: (Fig. 18) In this system the pterotic
canal dorsally and the preopercular and mandible ventrally,
serve as correlative units. Near the posterior margin of the
brain-case a small superficial tube links the preopercle with the
pterotic and lateral line. Similar superficial bones, only par-
tially ossified, surround the eye and extend dorsally and ventrally
throughout the length of the snout, extending to its extreme tip,
where a pair of small, terminal tubules lies between the ends of
the upper and lower sets.

Hyoid Arch: (Fig. 19). The glossohyal is long, and ossified
as far forward as the broadest portion of the dentary; poste-
riorly it almost reaches the vertical from the attachment of the
opercle. The ceratohyals arise from the middle of this ossified
portion. There is a hint in one of the Bermuda specimens of a
division into basihyals and ceratohyals, the lines of demarcation

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 41

, , Fi£- ,18 (Upper). Nessorhamphus ingolfianus. Cephalic canal system in transitional adolescent, standard length 142 mm.;

lalGiftl V16W. (X O.o).

i/(9 ^ Ness°rhamphus ingolfianus. Hyoid and branchial apparatus in transitional adolescent, standard length

14^ mm. (x 17.5) ,

42 Zoologica: N. Y. Zoological Society [XX; 2

falling between the first and second branchiostegal rays. There
is no trace of separate epi- or interhyals. Articulation with the
inner face of the hyomandibular is near the ventral border of
the latter, about midway between quadrate and opercle. There
are six or seven branchiostegals, the number sometimes varying
on the two sides of the same specimen. These bones are always
very slender, even basally, and only delicately ossified. Their
posterior ends are curved forward, lying free in the branchial
cavity. The last branchiostegal arises very close to the base of
the preceding ray, and is only indirectly attached to the bone.
A needle-like urohyal, its anterior end swollen, arises between
the bases of the ceratohyals and extends posteriorly as far as
the origin of the third gill arch.

Branchial Apparatus: (Fig. 19). The branchial arches lie
between the posterior end of the opercle and the cleithrum, the
bones lying almost horizontally, on a parallel with the upper
edge of the opercle and overlapping each other throughout.

Three ossified, widely separated basibranchials show plain-
ly in the Bermuda specimens. The first is triangular, the apex
anterior, and lies against the upper surface of the attenuated
end of the glossohyal. The second is slender and small, the third
very small. The latter was completely lacking in Trewavas’s
example.

Hypobranchials are present in the first two arches, the
second bone being half as long as the first. All five ceratobran-
chials are well developed, the first four of equal size — as long as
the ceratohyal, but more slender — the fifth only half as long.
The first two are deeply forked at their postero-dorsal tips, and
their whole central regions scarcely ossified; the third is closed
distally, but has an oblong terminal foramen; the fourth is
fairly well ossified throughout, while the small fifth cerato-
branchial is the strongest of all and bears an irregularly double
series of conical teeth. There are four epibranchials, the first
two minute and oval, the third long, and the fourth elongate
and altogether more than five times as large as the first. Two
pharyngobranchials are practically joined by suture, lying in-
ternal and immediately ventral to the third and fourth epiotics.
Their dentigerous surfaces are opposed to the corresponding
face of the fifth ceratobranchial. The anterior pharyngobran-

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 43

chial has only about half a dozen teeth, all confined to its postero-
ventral surface; the other bone, however, has one entire surface
covered with about forty conical teeth, arranged in roughly
diagonal rows of seven or eight teeth each.

Pectoral Girdle: (Fig. 17). The doubly-curving supra-
cleithrum is placed at the level of the anterior part of the eighth
vertebra, with which, however, it has no direct connection. The
cleithrum continues the girdle ventrally, curving forward in its
lower portion. Both bones, though slender, are strongly ossified.
Between the upper end of the cleithrum and the uppermost pec-
toral finray are two tiny disks of bone, only partially ossified,
the hyper- and hypocoracoids. Four minute radials are visible,
with only traces of ossification. Only the anterior, upper eleven
of the thirteen pectoral rays have bony deposits, and these
throughout only their proximal portions.

Vertical Fins and Supports: (Fig. 20) . The dorsal fin orig-
inates between the 17th and 18th vertebrae, the anal at the 75th.
Only the proximal halves of the dorsal rays are ossified, and
even here the stain dies out toward the end of the fin. The anal
rays are entirely unossified. Baseosts of both fins are well de-
veloped and stained, though less strongly in the case of the anal
supports. The same is true of the radials connecting the finrays.
These minute, horizontal bones must serve the same purpose, in
aiding the forward migration of the vertical fins during meta-
morphosis, as do the analagous elements in the quite unrelated
Idiacanthus (Beebe 1934, p. 213). There are usually two fin-
rays to each vertebra.

Vertebral Column : (Fig. 20). The count of 150 vertebrae
for the largest cleared-and-dyed specimen includes both the first
element, which is fused to the skull, and the urostyle. Ossifica-
tion is strong only toward the end of the column. The vertebrae
are all of similar length (about .8 mm.) as far back as the 135th
element. Posterior to this they decrease in size and the last
form the specialized support for the caudal fin (see p. 45) .

The dual character of the column — its formation around
the original neural tube and notochord — is very evident in this
species, and particularly in the immature specimen under dis-
cussion. Each centrum of typical, hour-glass, adult shape, is
attached by distinct suture to the exceedingly large, parallelo-

44

Zoologica: N. Y. Zoological Society

[XX; 2

grammatic neuropophysis, this element equalling or exceeding
the centrum in height.

The neural arches slope backwards and upwards in the
first sixteen or seventeen vertebrae, the neural spine being small
and near the anterior end of the arch. Posterior to this, the
direction of the slope is reversed, and the neural spine shifts
gradually to the posterior end of the arch, increasing progres-
sively in size backward.

The first vertebra gives rise to three epineurals on the left side
of the Bermuda specimen, and, possibly, on the right side also;
Trewavas shows two in her figure. These extra elements may
be vestiges of former vertebrae. Posterior to the pectoral fin

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 45

the epineurals are forked proximally. They die out at about the
124th vertebra. Throughout the length of the fish each epi-
neural measures about three times as long as a centrum.

The centrum of the first vertebra is very short and lacks
parapophyses. Between this and the pectorals the parapophyses
are small and widely separated by median lamellar projections
of irregular shape. Behind the pectorals the parapophyses be-
come longer and more ventral, and arise directly from the
lamellae, which are found in more lateral positions. From the
dorsal origin backward the parapophyses shift gradually to the
forward end of the centra, and each lamella extends the full
length of the centrum, forming anterior and posterior zygo-
pophyses toward the end of the caudal peduncle.

There are no pleural ribs.

Epipleurals are traceable from the 40th (37th in Trewavas’s
specimen) vertebra to the 124th, the same point at which the
epineurals vanish. They are of similar length, and are forked
proximally.

Posterior Part of Vertebral Column and Caudal Fin: (Fig.
21). There is remarkably little specialization of precaudal ver-
tebrae up to the very urostyle segment itself. What change
there is consists altogether of size, the individual vertebrae be-
coming gradually smaller from the 144th or seventh precaudal,
which measures .54 mm. in length, to the 150th or last before
the urostyle, which is only .24 mm. in length. The rounded ar-
ticulating anterior part of the urostyle segment gives rise at
once to a bony urostyle, longer than any of the precaudal ver-
tebrae. This supports a well-developed neural arch, closely ap-
proximating in size and shape that of the precaudal elements.

From the ventral side of the urostyle arise three bony
bases for the hypurals. The most superior one flares out into
a fan shape, whose truncate posterior aspect supports the upper
six of the twelve caudal rays. The second and third bases arise
as stout pillars, and fuse over their posterior ends enclosing a
large oval foramen. This entire plate supports four more of the
rays. A fourth hypural articulates at its base with the prolonged
haemal arch of the 150th vertebra, and an inferior facet of the
third hypural base. The slightly expanded tip of this hypural

46

Zoologica : N. Y. Zoological Society

[XX; 2

imphus ingolfianus. End of vertebral column and base of caudal fin of transitional adolescent, standard

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 47

oes

(L—

stom

Fig. 22. N essorhamphus ingolfianus. Alimentary canal of transitional adolescent, stand-
ard length 142 mm. (x .8).

supports the last two of the twelve caudal rays. All of these
rays are strongly ossified for a considerable distance.

Digestive System: (Fig. 22). This is of the simplest. The
blind sac stomach in our largest transitional adolescents is un-
pigmented and falls short of the anus. The pylorus is a snout’s
length behind the level of the pectoral base, at the posterior end
of the single-lobed liver. The anterior end of the liver lies only
several millimetres behind the pectoral base. The gall bladder
is small and is placed immediately anterior to the pylorus.

Urino-genital System: The kidneys are uniformly slender
and extend the full length of the coelomic cavity. They contain
pigmented matter similar to that commonly found in these
organs in certain families of this group.

The gonads are rudimentary, and appear as slender threads
along the roof of the coelom.

Relative Bodily Proportions : The following table shows
the relative proportions of N essorhamphus at different ages :

Growth Stage Length

Length

Head

Head

Length

Length

Pre-anal

Depth

Head

Eye

Snout

Snout to

Snout to

Myomeres

Dorsal

Anal

Larva 7. —

8.6—

6.5—

2.1—

1.8

1.1

120-121

(26, 28 mm.) 7.8

(Present Series)

9

5.8

2.2

Post-larvae 7.4 —

8.5—

8.2—

2.5—

4.4—

1.5—

89-78

(68 to 72 mm.) 9.6
(Present Series)

7.4

9.7

2.2

4.8

1.7

Adolescents 18. —

7.8—

7.4—

2.—

5.

1.8

72-73

(81, 78 mm.) 20.
(Present Series)

6.4

8.7

2.2

Transitional 29.5 —

6.2—

9.5—

2.6—

4.8—

1.9—

72-73

Adolescents 30.

6.

10.7

2.5

4.6

1.8

(80 to 166 mm.)
(Present Series)

Adult 25.

7.2

10.6

2.4

4.8

1.8

72

(248.5 mm.) (vertebrae)

(Type Specimen)

48

Zoologica: N. Y. Zoological Society
Ecology

[XX; 2

Seasonal and Vertical Distribution : According to
Schmidt, Nessorhamphus ingolfianus spawns at or near the sur-
face of the Sargassum Sea in spring and early summer, some-
what later in the east than in the west, and the metamorphosis
of the larva takes place in August, September and October at
lower levels. The present material lends corroboration. In the
spring only large transitional adolescents were captured off Ber-
muda ; no small larvae have been taken there, the only larvae of
any length being two half-grown leptocephali caught early in
September; finally, all of the metamorphosing specimens oc-
curred in August and September, in deep water.

Abundance: Nessorhamphus is rare among the deep-sea
fish captured off Bermuda, only once occurring in every 52.5 nets
drawn between 400 and 1,000 fathoms, the Bermuda limits of
its distribution.

Sociability : Only twice were more than a single specimen
taken in the same net; in each of these cases two transitional
adolescents of similar size came up together.

Food : The stomachs of six transitional adolescents, between
87 and 166 mm., were found to contain remains of schizopods or
shrimps. These were distributed as follows :

Length of
N essorhamphus

85 mm.

87 mm.

88 mm.
125 mm.
141 mm.
168 mm.

Length of

Food Food

1 Tkysanopoda 20 mm.

1 Tkysanopoda 11 mm.

1 shrimp or schizopod; much digested.

3 Tkysanopoda 17 to 23 mm.

3 Shrimps or schizopods; small; much digested
1 Sergestes 31 mm.

In every case the crustaceans had been swallowed tail first. An
80 mm. specimen had eaten radiolarians.

Enemies: Nessorhamphus has not yet been found in the
stomach of any other animal.

Viability: We have not taken Nessorhamphus alive in our
nets. Captain Hansen and Dr. Taning, however, both succeeded
in hatching the eggs on board ship, and in raising the pre-larvae
to a stage where comparison could be made with pre-larvae of
known identity caught swimming in the water.

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 49

Study Material

The following list gives the catalogue number, depth in
fathoms, date of capture, length and growth stage of each speci-
men of Nessorhamphus ingolfianus taken by the Bermuda
Oceanographic Expeditions. All were caught in the cylinder of
water off the Bermuda coast described in Zoologica, Vol. XVI,
No. 1, p. 5. “Trans. Adol.” stands for “Transitional Adolescent.”

No. 9,568; Net 24; 700 F.; April 15, 1929; 125, 135 mm.; Trans. Adol.
No. 9,751; Net 64; 600 F.; May 4, 1929; 141 mm.; Trans. Adol.

No. 12,499; Net 385; 600 F.; Aug. 17, 1929; 70 mm.; Post-larva
No. 13,561; Net 460; 700 F.; Sept. 11, 1929; 88 mm.; Trans. Adol.

No. 16,030; Net 707; 500 F.; June 16, 1930; 142 mm.; Trans. Adol.

No. 17,748; Net 834; 400 F.; Sept. 3, 1930; 85 mm.; Trans. Adol.

No. 17,819; Net 842; 600 F.; Sept. 4, 1930; 78 mm.; Adolescent
No. 18,024; Net 855; 700 F.; Sept. 6, 1930; 92 mm.; Trans. Adol.

No. 18,058; Net 859; 500 F.; Sept. 8, 1930; 69 mm.; Post-larva
No. 18,082; Net 861; 700 F.; Sept. 8, 1930; 68 mm.; Post-larva
No. 18,601; Net 890; 600 F.; Sept. 15, 1930; 80 mm.; Trans. Adol.

No. 19,274; Net 940; 1,000 F.; Sept. 24, 1930; 81 mm.; Adolescent
No. 23,041; Net 1,245; 1,000 F.; Aug. 31, 1931; 70 mm.; Post-larva
No. 23,126; Net 1,263; 800 F.; Sept. 4, 1931; 28 mm.; Larva
No. 23,292; Net 1,287; 1,000 F.; Sept. 10, 1931; 26 mm.; Larva
No. 23,550; Net 1,312; 400 F.; Sept. 16, 1931; 72 mm.; Post-larva
No. 23,565; Net 1,313; 500 F.; Sept. 17, 1931; 87, 87 mm.; Trans. Adol.
No. 23,600; Net 1,314; 600 F.; Sept. 17, 1931; 166 mm.; Trans. Adol.
No. 23,636; Net 1,324; 400 F.; Sept. 18, 1931; 84 mm.; Trans. Adol.

Synonymy and References

Leptocephalus ingolfianus :

Schmidt, 1912, p. 49, pi. Ill, fig. 8. (Description of larva).
Avocettina scapularostris :

Borodin, 1929, p. 109. (1 specimen; 153 mm.; 41° 29' N. Lat.,
47° 48' W. Long.; 800-0 fathoms).

Borodin, 1931, p. 74, pi. 3, figs. 1-3. (Supplementary descrip-
tion of preceding specimen).

Nessorhamphus ingolfianus :

Schmidt, 1930, p. 273, pis. IV-V. (Type description; resume
of development. Many thousand specimens from North At-
lantic, surface to deep water, in collection; eggs and fish
of all lengths up to at least 248.5 mm.).

Trewavas, 1932, p. 652, pi. IV, text figs. 7-9. (Osteology).
Parr, 1932, p. 19. (1 specimen; 125 mm.; 23° 24' 15" N. Lat.,
64° 29' W. Long. ; 10,000-foot wire) .

50

Zoologica: N. Y. Zoological Society

[XX; 2

Bibliography of References Consulted in the Two Preceding Papers
Beebe, W.

1931a Bermuda Oceanographic Expeditions 1929-1930. Introduction.
Zoologica, Vol. XIII, No. 1.

1931b Bermuda Oceanographic Expeditions 1929-1930. List of Nets
and Data. Zoologica, Vol. XIII, No. 2.

1932 Bermuda Oceanographic Expeditions 1931. Individual Nets and
Data. Zoologica, Vol. XIII, No. 3.

1933a Deep-sea Fishes of the Bermuda Oceanographic Expeditions.

Introduction. Zoologica, Vol. XVI, No. 1.

1933b Deep-sea Fishes of the Bermuda Oceanographic Expeditions.

Family Alepocephalidae. Zoologica, Vol. XVI, No. 2.

1933c Deep-sea Fishes of the Bermuda Oceanographic Expeditions.

Family Argentinidae. Zoologica, Vol. XVI, No. 3.

1934 Deep-sea Fishes of the Bermuda Oceanographic Expeditions.
Family Idiacanthidae. Zoologica, Vol. XVI, No. 4.

Borodin, N.

1929 Some New Deep Sea Fishes. Proc. New England Zool. Club,
Vol. X, pp. 109-111.

1931 Atlantic Deep-sea Fishes. Bull. Mus. Comp. Zool. Vol. LXXII,
No. 3.

Fowler, H. W.

1934 Descriptions of New Fishes Obtained 1907 to 1910, Chiefly in the
Philippine Islands and Adjacent Seas. Proc. Acad. Nat. Sci.
Phil. Vol. LXXXV, 1933. Published Jan. 20, 1934.

Gill, T. N.

1884 New Families of Fishes (Stephanoberycidae and Derichthyidae )
Recently Added to the Deep Sea Fauna. Amer. Naturalist, 1884,
Vol. XVIII, p. 433.

Goode, G. B., and Bean, T. H.

1895 Oceanic Ichthyology. A treatise on the deep-sea and pelagic
fishes of the world. Special Bull. U. S. Nat. Mus.

Meek, S. E. and Hildebrand, S. F.

1923 The Marine Fishes of Panama. Field Mus. of Nat. Hist. Publ.
No. 215. Zoological Series. Vol. XV.

Norman, J. R.

1930 Oceanic Fishes and Flatfishes Collected in 1925-1927. Discovery
Reports, Vol. II.

Parr, A. E.

1932 Deep Sea Eels, Exclusive of Larval Forms. Bull. Bing. Ocean.
Coll. New Haven. Vol. Ill, Art. 5.

1934 Report on Experimental Use of a Triangular Trawl for Bathy-
pelagic Collecting; with an account of the fishes obtained and a
revision of the family Cetomimidae. Bull. Bingham Ocean. Coll.
New Haven, Vol. IV, Art. 6.

1935] Beebe: Deep-Sea Fishes of the Bermuda Expeditions 51

Schmidt, J.

1912 Contributions to the Biology of some North American Species
of Eels. Vidensk. Medd. Dansk. Naturhist. Foren. Copenhagen,
Vol. 64.

1930 Nessorhamphus, a new cosmopolitan Genus of Oceanic Eels.
Vidensk. Medd. Dansk. Naturhist. Foren. Copenhagen, Vol. 90,
1930-1931.

Trewavas, E.

1932 A Contribution to the Classification of the Fishes of the Order
Apodes, based on the Osteology of some rare Eels. Proc. Zool.
Soc. London, Part 3, 1932.

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME XX. NUMBER 3

DEEP-SEA FISHES OF THE BERMUDA
OCEANOGRAPHIC EXPEDITIONS

No. 3— Family SERRIVOMERIDAE
William Beebe

&

Jocelyn Crane

PUBLISHED BY THE SOCIETY
THE ZOOLOGICAL PARK, NEW YORK

November 30, 1936

J?eto gorfe Zoological ikicietp

General Office : 101 Park Avenue, New York City

( ®flfera

President , Madison Grant

Vice-Presidents, W. Redmond Cross and Kermit Roosevelt
Chairman, Executive Committee, Madison Grant
Treasurer, Cornelius R. Agnew
Secretary, Henry Fairfield Osborn, Jr.

of

(Class uf IS37

George Bird Grinnell, Frederic C. Walcott, George C. Clark, W. Redmond
Cross, Henry Fairfield Osborn, Jr., George Gordon Battle,
Bayard Dominick, Robert Gordon McKay, Kermit
Roosevelt, John M. Schiff, Laurance S.

Rockfeller, F. Trubee Davison

(Class of 1338

Robert S. Brewster, Edward S. Harkness, Irving K. Taylor, Harry Payne
Bingham, Landon K. Thorne, J. Watson Webb, Oliver D. Filley,

De Forest Grant, George F. Baker, Robert L. Gerry,

Warren Kinney, William De Forest Manice

(Class nf 1333

Madison Grant, Lewis Rutherfurd Morris, Archer M. Huntington,
Cornelius R. Agnew, Harrison Williams, Marshall Field,

Ogden L. Mills, Vincent Astor, C. Suydam Cutting,

Childs Frick, Alfred Ely, Herbert L. Pratt

W. Reid Blair, Director of the Zoological Park
William T. Hornaday, Director Emeritus
Charles H. Townsend, Director of the Aquarium
C. M. Breder, Jr., Assistant Director , Aquarium
Raymond L. Ditmars, Curator of Mammals and Reptiles
William Beebe, Honorary Curator of Birds and Director of
Department of Tropical Research
Lee S. Crandall, Curator of Birds
H. C. Raven, Prosector
Charles V. Noback, Veterinarian
Claude W. Leister, Ass’t to the Director and Curator,
Educational Activities
Edward R. Osterndorff, Photographer
William Bridges, Editor and Curator of Publications

iEMtarial ©ammittTT

Madison Grant, Chairman

Charles H. Townsend
George Bird Grinnell

William Beebe
W. Reid Blair

William Bridges

Zoologica, Volume XX, Number 3

DEEP-SEA FISHES OF THE BERMUDA
OCEANOGRAPHIC EXPEDITIONS

Family SERRIVOMERIDAE
Part I : Genus Serrivomer1
William Beebe and Jocelyn Crane

Department of Tropical Research, New York Zoological Society

(Figs. 23-42 incl.)

Contents

Introduction P- 53

Important Points in the Following Study of Serrivomer p. 54

Genus Serrivomer, with a Taxonomic Revision of the Species p. 55

Serrivomer beanii Gill and Ryder 1883

Specimens Taken by the Bermuda Oceanographic Expeditions p. 65

Description of Adult p. 65

Development p. 80

Ecology p. 86

Study Material p. 91

Serrivomer brevidentatus Roule and Bertin 1929

Specimens Taken by the Bermuda Oceanographic Expeditions ... .p. 93

Description of Adult .t .p. 93

Development p. 97

Ecology p. 97

Study Material p. 100

Bibliography p. 101

Introduction

For detailed data in regard to nets, locality, dates, etc. con-
cerning the capture of the deep-sea eels treated in this mono-
graph, refer to Zoologica, Vol. XIII, Nos. 1, 2 and 3 and Vol.

1 Contribution No. 505, Department of Tropical Research, New York Zoological Society.

SEC 3 3936

54

Zoologica: N. Y. Zoological Society

[XX; B

XX, No. 1, pp. 1-2. For physical data, methods of measurement
and definitions of growth stages, see Zoologica, Vol. XVI, No. 1.

The drawings in the present papers are the work of George
Swanson.

Our thanks are due to the following persons for their co-
operation in enabling us to examine specimens of Serrivomer
deposited in various museums : Dr. George S. Myers of the U. S.
National Museum, Miss Erna W. Mohr of Ahrensburg, Germany ;
Dr. Albert Eide Parr and Y. H. Olsen of the Peabody Museum,
New Haven, and the summer custodian of the collections of the
Museum of Comparative Zoology in Cambridge.

Important Points in the Following Study of Serrivomer

Taxonomy and Distribution : The recorded specimens of
Serrivomer comprise three distinct species, S. beanii Gill and
Ryder 1883 and S. brevidentatus Roule and Bertin 1929, both
confined to the Atlantic Ocean, and S. sector Garman 1899, con-
fined to the Pacific and Indian Oceans.

Ecology and Development: S. beanii is the most com-
mon deep-sea eel found in our Bermuda trawling area, although
it is a rare fish compared with Cyclothone signata, Cyclothone
microdon, Sternoptyx diaphana and certain paralepids and
myctophids. S. sector, on the other hand, is one of the rarest
of all the fishes in the region.

The jaws of members of the genus Serrivomer are long and
strong with numerous fine, sharp teeth, while there is a long,
saw-like ridge of large, compressed teeth on the vomer; the
stomach is only moderately distensible; the digestive organs are
crowded into the anterior one-sixth of the fish; the principal
foods of the eels and elvers are shrimps and euphausids (chiefly
of one genus) , with occasional myctophids and cyclothones, while
the leptocephalids eat diatoms and minute crustaceans.

Bathysphere observations of Serrivomer showed that these
eels are brilliantly iridescent, fast swimmers, and travel alone,
in pairs and in groups of four or five; they were recognized
on six different occasions, between 125 and 250 fathoms. They
have been taken in the trawling nets off Bermuda between 50 and
1,000 fathoms and elsewhere down to 3,281 fathoms. No satis-

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 55

factory correlation has been found between depth and growth
stage, as large fish were seen from the Bathysphere much higher
than eels of equal size were caught in the nets, whereas larvae
were caught at very low, as well as at high levels. Off Bermuda
there seems to be a chief breeding season in late spring and
early summer.

The larva of Serrivomer is a typical leptocephalus, dis-
tinguishable by its moderate depth (7.8 to 10 in length), long
snout and characteristic number of myomeres (a total of 154
to 165, of which 89 to 97 are preanal) . It grows to about 63 mm.
in length before metamorphosis, which takes place rapidly.
Elvers of 100 mm. have all the general characters of the adults.
The great majority of the Bermuda specimens are immature
fish measuring between 100 and 300 mm. Probably no fully
adult fish is less than 500 mm. in length. The largest known
specimen is the type, 594 mm. long. It seems likely that in-
dividuals breed only once, as is the case with Anguilla. All of
the Bermuda specimens in which the sex can be determined are
females.

Genus Serrivomer, with a Taxonomic Revision of the Species

Generic Characters : Snout less than half length of head ;
gradually tapering, not sharply constricted, in front of eye;
vomerine teeth forming a high, serrated ridge along roof of
mouth. Trunk decreasing gradually in height and thickness
from shoulder to tail; no nuchal constriction; no caudal fila-
ment; jaws strong, the lower slightly the longer; maxillary
teeth strong and pointed, in several rows; mandibular teeth
similar; vomerine teeth large, flattened, triangular, forming the
above-mentioned serrated ridge along the roof of the mouth;
pectorals rudimentary; dorsal beginning well behind pectorals
and continuing to the tip of the tail; anal origin immediately
behind anus, at a distance about mid-way between pectoral base
and dorsal origin, and continuing to the end of the body; pos-
terior portions of dorsal and anal fins lax, not forming a pseudo-
caudal; about 30 anal rays in the last 3.5 per cent of the total
length of the fish, and about 20 rays occupying a space per ray
of less than one-tenth of one per cent, of the total length;
branchial clefts confluent; lateral line without pores, or with a

56 Zoological N. Y. Zoological Society [XX; 3

single series of small, inconspicuous pores immediately below it.
Three species.

Discussion and Taxonomic Revision of the Species:
Due both to the inadequacy of the early descriptions of the
species of Serrivomer and to the scattered locations of the vari-
ous collections, the validities of the described species have al-
ways been questionable. The present discussion should, how-
ever, clear up the major taxonomic issues for, in addition to
studying in detail our extensive Bermuda collection, we have
been able to examine the following specimens: the type of

S. beanii Gill and Ryder 1883 (U. S. National Museum, Wash-
ington, D. C.) ; the type of S. sector Garman 1899 and the other
Pacific specimens of the Albatross collection (Museum of Com-
parative Zoology, Cambridge, Mass.) ; four of the six Valdivia
specimens from the East Indies and the Indian Ocean (Berlin
Museum and the Museum of Leipzig) ; nine specimens taken by
the Bingham Oceanographic Expeditions from the Atlantic (Pea-
body Museum, New Haven, Conn.) ; four specimens of the Iselin
Expedition in the North Atlantic; and, finally, a number of un-
recorded specimens from the eastern Pacific and the Gulf of
California taken, respectively, on the Arcturus Oceanographic
Expedition (1925) and the Templeton Crocker Expedition
(1936), both trips having been undertaken by this department
of the New York Zoological Society. The only important col-
lection which has not been inspected is that of the Dana Expedi-
tion of 1920-1922, but since these specimens were reported upon
fully by Roule and Bertin in 1929, it has been found possible to
establish, in a general way, the taxonomic status of this collec-
tion as well.

From our study, it is apparent that there are three distinct
species of Serrivomer : S. beanii Gill and Ryder 1883 from the
Atlantic, S. brevidentatus Roule and Bertin 1929 (described as
S. sector type brevidentatus ), which is also from the Atlantic,
and S. sector Garman 1899, which is confined to the Pacific and
Indian Oceans. (Fig. 23).

In the differentiation of the species, four main characters
are involved, which may best be shown in the following tabular
form :

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 57

S. beanii
(Atlantic)

S. brevidentatus
(Atlantic)

S. sector
(Pacific)

Vomerine teeth

50-80, 3-4 times as
long as w’de.

20-30, twice as long
as wide.

like S. beanii.

Anterior tips of
first 5 brr. rays

project far beyond
hyoid element to
which they are at-
tached. (See Fig.
31).

project not at all.
(See Fig. 40).

like S. brevi-
dentatus.

Distance from
snout to anal ori-
gin contained in
length

4. 0-4. 5 times.

3.4-3.8 times.

like S', brevi-
dentatus.

No. of anal rays

124-140

159-173

150-160

The dental characteristics described by Roule and Bertin
(1929, p. 39) as distinguishing their two groups of specimens,
S. sector , type longidentatus (= S. beanii) and S. sector , type
brevidentatus (= S. brevidentatus) are fully corroborated by
Bermuda specimens measuring more than 300 mm. in length:
S. beanii differs from S. brevidentatus chiefly in having the
teeth of the vomerine ridge longer and more numerous and in
having those of the maxillary fewer and more scattered, there
being a distinct row of much longer, more pointed teeth. The
larger, teeth of the mandible also, although no fewer than in
S. brevidentatus , are longer and more slender in S. beanii. We
have found in addition that the dentition of true S. sector , from
the Pacific, is practically identical with that of S. beanii. The
details of the dentition in the two forms may be presented as
in the table on page 58.

Specimens of Serrivomer smaller than about 300 mm. are
indistinguishable specifically on the basis of the dentition, since
the teeth of all three species resemble those of adult S. brevi-
dentatus : the vomerine ridge teeth are relatively short and broad,
and few in number, while those of the maxillary and mandible
are comparatively homogeneous. Therefore, since it is chiefly by
means of the dentition that S. sector is distinct from S. brevi-

58

Zoologica: N. Y. Zoological Society

[XX; 3

Table Showing the Relative Dental Characteristics
of the Species of Serrivomer.

S. beanii and S. sector

S. brevidentatus

Vomer

50-80 teeth on ridge, each
tooth three to four times as
high as wide, alternating in
2 parallel rows forming a
ridge from anterior part of
eye to point 4/5 of distance
from eye to snout tip; eleva-
tion greatest in middle of
ridge; its greatest height al-
most equals vertical di-
ameter of eye. From end of
ridge to snout tip are num-
erous small teeth irregularly
arranged.

20-35 teeth on ridge, each
tooth twice as high as wide;
ridge extends from nostril to
point 3/5 of distance from eye
to snout tip, a little before
maxillary’s end; greatest
height of ridge equals 2/3
vertical diameter of eye ;
otherwise identical with S.
beanii and S. sector.

Maxillary

3 to 4 rows of teeth, increas-
ing in size from outer to in-
ner; the inner row is formed
of 20 to 25 very long, needle-
like teeth, sharply differen-
tiated from those of the other
rows. Teeth and bones end
about 4/5 of preorbital dis-
tance.

3 rows of teeth, increasing in
size from outer to inner; the
inner row is formed of 30 to
40 teeth, only moderately
larger than those of the other
rows, and triangular in
shape; otherwise identical
with S. beanii and S. sector.

Mandible

3 to 5 rows of teeth, increas-
ing in size from outermost to
next to innermost row ; inner-
most very small ; 35 to 45
teeth, slender like the long
teeth of the maxillary, in next
to innermost row, this row
ending about 3/4 of distance
from eye to tip of jaw; from
here on the arrangement is
irregular.

4 to 6 rows of teeth ; less
varied than in the other spe-
cies, the large teeth of the
next to the innermost row
being relatively shorter and
broader; and ending about
3/5 of distance from eye to
tip of jaw; otherwise identical
with S. beanii and S. sector.

dentatus, we cannot distinguish the young of these two species,
although there is no danger of confusing them since they in-
habit different oceans. The young of the two Atlantic species,
however, can be easily differentiated by the anal count and, more
easily, by the mode of attachment of the branchiostegal rays to
the hyoid arch. In S. brevidentatus, the first five rays are fast-
ened to the hyoid bones in the usual manner, by the extreme
anterior tips of the rays. In S. beanii , on the other hand, each

1936] Beebe & Crane: Fishes of the Bermuda Expeditions

59

• j- “it' The geo^raPllical ar*d vertical distribution of the genus Serrivomer. The shaded area in the Atlantic Ocean
indicates the range of the two species, S. beanii and S. breviientatus; the specimen marked by the circle at the tip of South
Africa is of questionable identity ; all others are S. sector. The relative numbers of specimens of Serrivomer taken at
different depths by the Bermuda Oceanographic Expeditions are shown diagrammatically at the left of the column which
gives the vertical range of the genus.

60 Zoologica: N. Y. Zoological Society [XX; 3

corresponding ray projects considerably in front of the hyoid
element to which it is attached. If the specimen is turned on
its back and the angles of the jaw are separated as far as pos-
sible with forceps, this character can almost always be deter-
mined without even rupturing the skin ; this character develops
earlier than any of the others of specific value, being determin-
able even in post-larvae. We have not, however, been able to find
a means by which the larvae (true leptocephali) of the Bermuda
collection can be separated. The only character of possible
value at this stage is the anal ray count, but as the full com-
plement of rays is found only in the oldest larvae, the count is
not of help in the majority. Since all of the 15 larvae in the
Bermuda collection are so similar, and since, in a total of 147
older specimens, all except seven are S. beanii, it is assumed in
this paper that the larvae are all of this species.

From a study of the skeletons of Bermuda specimens, the
following remaining differences were discovered between the two
Atlantic species, S. beanii and S. br evident atus : In S . beanii
the ethmo-vomer is more robust, in evident correlation with the
heavier burden of teeth; the preopercle is more extensive and
the opercle triangular rather than quadrilateral; there are dif-
ferences in the vertebrae, the most conspicuous being the great
length of the neural spine; and, finally, in all specimens exam-
ined, including the type of S . beanii, the anal fin has only 126
to 140 rays, instead of the 159 to 173 found in S. brevidentatus.
Roule and Bertin, however, state ( loc . cit.) that in their series
the dorsal and anal each numbered about 160 to 161 rays; it
seems possible that they did not happen to count the anal rays in
any of their “type brevidentatus ” group. No differences in
coloration, remarked upon by Roule and Bertin, are discernible
in Bermuda specimens of the two species.

In summary, S. beanii differs from S. sector in the greater
development of the bones throughout the skeleton, as evidenced
in the skull, teeth, opercular and hyoid apparatus, and in the
vertebral column. None of the differences can be attributed
either to age or sex, and as both forms occur in the same locality,
there can be no question of geographical race or subspecies.

Of the three species of Serrivomer, about 145 specimens
have previously been recorded from the Atlantic, Pacific and

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 61

Indian Oceans. They measured between 103 and 629 mm. and
were taken down to a depth of 3,281 fathoms (6,000 metres).
The following annotated synonymies of the three species, based
on the reexamination of the actual specimens or upon a study
of the published data, as indicated, are for convenience herewith
presented together:

1. Serri vomer beanii Gill and Ryder 1883
Serrivomer beanii:

Gill and Ryder, 1883, pp. 260-261. 1 specimen; 594 mm.; 41°
40' 30" N. Lat., 65° 28' 30" W. Long., about 200 miles east
of Nantucket; 855 fathoms. (Specimen reexamined; it is
typical in every way of the specimens which have been
taken more recently) .

Goode and Bean, 1895, p. 155, fig. 175. (Supplementary de-
scription of preceding specimen).

?Barnard, 1925, p. 200. 1 specimen; ? mm.; south of Agulhas
Bank, South Africa. (Since we were not able to examine
this specimen, and since it was taken on the boundary line
between the Atlantic and Indian Oceans — i.e. between the
known ranges of S. beanii and S. sector — the identity of
this specimen remains questionable.)

Nemichthys infans:

Vaillant, 1888, p. 93, pi. vii, Figs. 1, la. 1 specimen; 240 mm.;
2,995 metres; off Azores. (We refer this specimen with
little hesitation to S. beanii although we have not been able
to examine it, because, from Vaillant’s published measure-
ments, the distance from the snout to the anal origin is con-
tained 4.5 times in the length of the fish, this proportion
being typical of S. beanii and not of S. brevidentatus, the
other Atlantic species. Vaillant mentions that the anal fin
is badly damaged, and the small size of the specimen would
make the dentition of little value in specific determination.
The fourth character, the mode of attachment of the
branchiostegal rays, is unknown).

62

Zoological N. Y. Zoological Society

[XX; 3

Nemichthys richardi:

Vaillant, 1888, p. 385. Above specimen designated as new
species.

Serrivomer richardi:

Goode and Bean, 1895, p. 155. Above specimen referred to
Serrivomer.

Serrivomer sector:

Roule and Bertin, 1929, p. 35. Above specimen verified as a
Serrivomer , from a reexamination of the type by Roule and
Bertin.

ibid., pp. 39-47, part. An indeterminable number of specimens,
measuring between 79 and 629 mm. in length, taken be-
tween 150 and 6,000 metres in the North Atlantic (com-
prising all the specimens referred to S. sector type longiden-
tatus with the exception of those taken in the Gulf of Pan-
ama) .

Parr, 1932, pp. 2-3. 7 specimens; 186 to 291 mm.; 4,000 to
8,000 feet of wire; Bahamas and Bermuda. (Reexamined
by the present authors and found to be typical S. beanii) .

Gavialiceps microps:

Borodin, 1931, p. 74. 2 specimens; 216, 252 mm.; 1,500
metres; off Bermuda. (Reexamined by the present authors
and found to be typical S. beanii) .

Incertae sedis:

Parr, 1932, p. 3, part. 2 specimens; 126, 128 mm.; 10,000
feet of wire; off Bermuda. (Reexamined by the present
authors and found to be typical S. beanii) .

Serrivomer sp.

Beebe, 1933, p. 180, part. Preliminary note of specimens of
Serrivomer taken by the Bermuda Oceanographic Expedi-
tions.

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 63

Serrivomer sector longidentatus :

Parr, 1934, p. 6. 1 specimen; ? mm.; 1,050-1,100 metres;
Bahamas.

2. Serrivomer sector Garman 1899

Serrivomer sector:

Garman, 1899, pp. 320-323, pi. LXIII. 11 specimens; up to
570 mm.; 134 to 1,772 fathoms; 9 stations off Colombia,
from 325 miles west of Bahia del Choco to just south of
Cape Mala, Panama; 1 station in Gulf of California, 50
miles south of Guaymas. (Reexamined by the present
authors) .

Brauer, 1906, p. 132, pi. VIII, fig. 4. 6 specimens; 205 to
392 mm.; 1,213 to 2,400 metres; Indian Ocean: between
New Amsterdam and Sumatra, Binnen Bay on the west
coast of Sumatra, west of Chagos Archipelago, and off the
coast of Italian Somaliland. (4 of the specimens — -3
from the Sumatra and Sumatra-to-New Amsterdam re-
gions and 1 from off Somaliland — have been reexamined
by the present authors. In spite of the fact that the ex-
amples are not in good condition and that two of them,
including the one from Somaliland, measure only 205
and 210 mm. respectively, still enough characters remain
in the group as a whole to show that these individuals do
not differ specifically from those taken in the eastern
Pacific: in no specimen do the branchiostegal tips overlap
the hyoid elements as in S. beanii; the two largest [330 and
392 mm.] are old enough to show typical S. sector dentition ;
in the three specimens where the measurement could be
taken, the snout-to-anal distance is contained 3.6 to 3.8
times in the length ; and, finally, the anal count of the largest
example [the fins of all others are damaged] is 152).

Lloyd, 1909, p. 152. 1 specimen; ? mm.; 930 fathoms; Arabian
Sea, off Travancore.

Weber and Beaufort, 1916, pp. 331-332, text-fig. 160. Mention
of known Pacific specimens.

Townsend and Nichols, 1925, p. 12. 2 specimens; 485 and ?

64

Zoological N. Y. Zoological Society [XX; 3

mm.; 590 and 630 fathoms; Gulf of California, off Todos
Santos Bay and off Cape San Lucas, respectively.

Roule and Bertin, 1929, pp. 39-47, part . An indeterminable
number of specimens, measuring between 79 and 629 mm.
in length, taken, between 150 and 6,000 metres, in the Gulf
of Panama. (Probably these specimens were among those
referred to S. sector type longidentatus) .

Serrivomer beanii:

Gilbert, 1905, p. 386. 3 specimens; 311 to 1,067 fathoms;
Hawaiian Islands.

3. Serrivomer brevidentatus Roule and Bertin 1929

Serrivomer sector:

Roule and Bertin, 1929, part. An indeterminable number of
specimens, measuring between 79 and 629 mm. in length,
taken between 150 and 6,000 metres in the North Atlantic
(comprising all the specimens referred to S. sector type
brevidentatus) .

?Trewavas, 1932, p. 650, pi. III. 1 specimen; 160 mm.;
depth ?; locality ? Description of the skull. (The illustra-
tion shows that this specimen is unquestionably not S.
beanii. In placing it under S. brevidentatus instead of un-
der S. sector, we are assuming that it was captured in the
Atlantic rather than in the Pacific or Indian Ocean) .

Serrivomer sp.

Beebe, 1933, p. 180, part. Preliminary note of specimens of
Serrivomer taken by the Bermuda Oceanographic Expedi-
tions.

Note: The specimens of Serrivomer and Serrivomer- like
fish, which have been recorded by Beebe (1933a and 1934) as
observed from the Bathysphere, may obviously have belonged
to either or both of the Bermuda species of Serrivomer and pos-
sibly to the genus Platuronides as well, since all of these forms
are superficially much alike.

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 65

The specimens recorded by Borodin, 1931, p. 74, from near
Bermuda as Serrivomer sector proved upon reexamination by us
to be adolescent nemichthyids.

The Bermuda Oceanographic Expedition specimens of S.
beanii and S. brevidentatus will now be discussed in detail

Serrivomer beanii Gill and Ryder 1883

Specimens Taken by the Bermuda Oceanographic
Expeditions

155 specimens; April to October, 1929 to 1931; 50 to 1,000
fathoms; from a cylinder of water 8 miles in diameter (5 to 13
miles south of Nonsuch Island, Bermuda), the center of which
is at 32° 12' N. Lat., 64° 36' W. Long.; Standard lengths from 55
to 440 mm.

Description of Adult
(Figs. 24J, 25J)

Color: Black, with a fragile coating of silver skin which,
when fresh, has a high bronzy iridescence.

Proportions (from the three largest Bermuda specimens,
standard lengths 400, 405, 440 mm., and the type specimen,
standard length 596 mm.) : Depth in length 30 to 41; head in
length 5.3 to 6.4; eye (horizontal) in head 16 to 17.2; eye is
horizontally elongate; maxillary reaching well beyond vertical
from posterior margin of eye ; snout in head 2.4 to 2.8 ; snout to
dorsal in length 3.1 to 3.3 ; snout to anal in length 4 to 4.3.

Teeth : The dentition has been described in detail on p. 58.

Fins : Pectoral rays 6 or 7, very delicate, equal in length to
horizontal diameter of eye, inserted at upper angle of branchial
cleft. Dorsal rays 159 to 165, commencing well behind the anal
origin, at a distance 1.3 to 1.5 times the postorbital length of the
head ; the first dorsal ray is inserted above the eleventh to thir-
teenth anal ray. Anal rays 124 to 140, longer than those of
dorsal. The rays of both fins are longest, and the spaces between
successive rays greatest, in the anterior halves, behind the first
10 or 15 rays. Caudal rays 5 or 6, scarcely distinguishable from

66

Zoological N. Y. Zoological Society

[XX; 3

D

Fig. 24. Serrivomer beanii. A to D, incl., larvae, 17, 25, 36, and 43 mm., respectively.

1936] Beebe & Crane: Fishes of the Bermuda Expeditions

67

F

• — • - ■

■ ■ • • ■ ■■ • ■ - ' "

H

I

1

Fig. 24, cont. Serrivomer beanii. E, larva, 57 mm. ; F, post-larva, 59 mm. ; G and H,
adolescents, 59 and 61 mm. ; I, transitional adolescent, 127 mm. ; J, adult, 440 mm.

68

Zoologica: N. Y. Zoological Society

[XX; 3

B

G

1

J

Fig. 25. Serrivomer beanii. A to E, incl.,
larvae, 17, 25, 36, 43 and 57 mm., respectively
in standard length ; F, post-larva, 59 mm. ;
G and H, adolescents, 59 and 61 mm. ; I,
transitional adolescent, 127 mm.; J, adult,
440 mm.

D

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 69

those of dorsal and anal, with which the caudal fin is confluent.

Vertebrae: Bermuda specimens, 154-168; specimens de-
scribed by Roule and Bertin (1929, p. 41), 143 to 155.

Osteology: (Figs. 28-33). The following description is de-
rived from the largest (440 mm.) Bermuda specimen.

Skull: The skull and jaw from above are exceedingly elon-
gate and slender. There is no supraoccipital, unless its vestiges
are concealed beneath the epiotics, which are medially united by
suture. They form the posterior third of an oval space, the
anterior two-thirds of which are formed by the parietals. The
paired frontals are narrow and long, bent midway so abruptly
downward that they look like two pairs of bones. The sphenotics
extend laterally as two large, triangular wings with the anterior
face concave. Disregarding whatever part the vomer may play
ventrad, the ethmoid or dorsal element is relatively enormous.
Antero-posteriorly it occupies almost seven-eighths of the entire
length of the head, and equals the frontals in extreme width.
Posteriorly it narrows to a long, slender spear point which over-
lies the frontals to the vertical of the sphenotics.

Laterally the skull shows considerable depth. All the dorsal
elements are visible in a narrow, elongate line. The epiotics are
about twice the size of the supraoccipital and extend well back
of that bone. Articulating with the parietals and extending back
beneath and behind the epiotics are the pterotics, for once jus-
tifying their name. The alisphenoids are small and join the
frontals and sphenotics. The parasphenoid is a slender rod ex-
tending from beneath the braincase forwards and merging with
the dentigerous vomer somewhere beneath the covering palatines.

Palato-pterygoid Arcade: Each palatine shows an extended
base along the lateral ethmoid. The pterygoid is large, triangu-
lar, set in unossified tissue between the maxillary and ethmoid,
and sends a long, thin rod down and backward to the articulation
of the hyomandibular and quadrate with the jaw angle.

The hyomandibular has been drawn out into an almost inde-
scribable shape: A long rod extends almost to the articulation
of the lower jaw, then a stout anterior projection, matching but
not touching the posterior outline of the pterygoids ; posteriorly
another long rod underlies the pterotic, while to the lower face
is applied the short and wide quadrate.

70

Zoologica: N. Y. Zoological Society

s

s

|XX;

<N

E

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 71

Fig. 27. Serrivomer beanii. Cartilaginous elements of larval head, lateral view. Standard length

72

Zoologica: N. Y. Zoological Society

[XX ; 3

Jaiv Apparatus: The vomer, or from its intimate connec-
tion, the ethmo-vomer, is the bearer of a single row of large
palisaded teeth. These lessen in size anteriorly and posteriorly,
and at the level of the end of the maxillary are replaced by an
elongate area of minute irregular teeth, which occupy the whole
of the tip of the upper jaw. (See p. 58 for detailed description
of teeth) . The maxillae are long and slender, occupying about
three-fourths of the upper jaw, and are furnished with small,
irregular teeth. The lower jaw, similarly armed, is deep and
strong. No separate angular is distinguishable and the articular
is relatively small, and posteriorly restricted.

Opercular Apparatus: The opercle is large and triangular,
with its posterior point drawn out. The preopercle equals the
former in size, forming a wing-like extension which actually
overlies the elongate, bodkin-shaped interopercle. Curving
around the antero-inferior angle of the opercle is the boomerang-
shaped subopercle.

Hyoid Arch: The glossohyal in this eel is well developed
and not especially slender, of equal calibre throughout. The
urohyal is shield-shaped and curves down and back from its
articulation. Of the paired elements of the hyoid apparatus,
there are two linear bones, the hypohyal and the ceratohyal.
The latter is twice the length of the former, and the two elements
are perfectly distinct. Epihyal and interhyal are, however, in-
distinguishable. A second hypohyal lies outside the hyoid arch,
a thin rod closely applied to the first hypohyal and the proximal
half of the ceratohyal.

There are seven branchiostegals, of thread-like diameter,
all very long and curved gradually upward, except for the first
which is quite short. This short one arises from the distal end
of the first hypohyal, all the rest springing from the ceratohyal.
The first four are widened where they articulate with their sup-
ports and all six extend far forward beyond even the second
hypohyal as free rods, apparently a regular character of this
species. The seventh branchiostegal arises very close to its
predecessor and does not extend beyond its ceratohyal support.

Branchial Apparatus: The branchial arches are all slender,
but well ossified. There are two basibranchials, the second about
one-fourth the length of the first. Two hypobranchials, short

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 73

and fairly stout, are developed, subequal in size. All five cerato-
branchials are found, the first four long and slender, the fifth
strongly dentigerous and closely applied to the anterior face of
the fourth. Four epibranchials are followed by five pharyngo-
branchials, the end of the fourth and all of the fifth being den-
tigerous and half the length of the opposed and toothed cerato-
branchial.

Pectoral Girdle: The curving cleithrum is attached imme-
diately beneath the junction of the eighth and ninth vertebrae.
A third of its length from the top arises a small, oblong upper
coracoid. A much smaller lower one is ankylosed to its distal
end, and supports seven pectoral rays. The longest ray is equal
to the horizontal diameter of the eye. Radials are absent.

Vertical Fins and Supports: The dorsal fin commences at
the vertical from the end of the 27th vertebra, opposite the 12th
anal ray. Throughout the length of the fin the rays are very
short, never equalling more than half the depth of the body at
their points of attachment; they are always shorter and more
closely set than the corresponding anal rays. The latter com-
mence at the level of the end of the 21st vertebra. The first
dozen and the last 50 are set more closely together than the rest.
The rays are longest in the second third of the fin's length, where
they are fully three times as long as the corresponding dorsal
rays, and equal the depth of the fish in that region. Most of the
finrays of both fins have a flattened, expanded area somewhere
near the middle of their lengths, and in the caudal region the
rays become very short, flattened and wavy, though they are still
strongly ossified.

There is a radial with expanded ends occupying the entire
space between each two finrays of both fins. The anterior end
of each radial is slightly lower than the posterior in the dorsal
fin and higher than the posterior in the case of the anal. A
baseost is developed for every ray, all directed strongly back-
ward. In the anterior part of the anal fin, however, these ele-
ments lie almost parallel to the horizontal radials, out of the
way of the coelomic organs, which continue behind the anus.
Correlated with this is the absence of the haemal arch and spine
in the same region (see below).

Vertebral Column: In the large, cleared specimen under

74 Zoological N. Y. Zoological Society [XX; 3

Fig. 28. (upper left). Serrivomer beanii. Skull of adult, dorsal view; standard length 440 mm. (x 10).

Fig. 29. (upper right). Same, teeth of upper jaw and vomer, ventral view, (x 10).

Fig. 30. (lower). Same, bones of head, pectoral girdle and anterior part of vertebral column, lateral view, (x 10).

plwbr | harbr

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 75

Fig. 31. Serrivomer beanii. Hyoid and branchial apparatus of adult, standard length 440 mm. (x 3.4).

76

Zoological N. Y. Zoological Society

[XX; 3

Fig. 32. Serrivomer beanii. Upper, 25th vertebra and 10th and 11th anal rays; middle,
97th vertebra; lower, 111th vertebra. (All x 9.3).

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 77

discussion there were about 160 vertebrae (the exact number
indeterminable, as tail is missing). Excluding the atlas, which
is only half the maximum size, and the second and third, which
are slightly less than maximum, the centra remain of the same
length as far back as about the 97th, or almost to the posterior
fifth of the anal fin. From here backward there is a rapid re-
duction in size. The proportions remain about the same through-
out, the maximum height of each centrum being about two-fifths
of its length, and the minimum about one-tenth.

The pre-cleithral vertebrae — i.e. the first eight, counting
the atlas — are furnished with very large neurapophyses, which
more than equal the centra in height. Behind the girdle they
decrease rapidly in size and throughout most of the length of
the column are represented only by a pair of low ridges running
almost the full length of each centrum just beneath its dorsal
profile. Toward the end of the column, however — for about the
last sixty vertebrae — each neurapophysis takes the form of a
pair of wings, one in the front half, the other at the rear of the
centrum and equal to the maximum height of the centrum.

The second vertebra (first behind atlas) has two dissimilar
neural spines, posteriorly directed, near the posterior edge of
the neurapophysis. On the third vertebra they are median, min-
ute, nearly erect, and almost joined to form an arch. In the
fourth the arch is complete and anteriorly directed, furnished
distally with a short spine and a small antero-ventral projec-
tion which almost touches the pair of spines in the preceding
vertebra. The four succeeding arches— of the fifth through the
eighth vertebrae — are identical with that of the fourth, except
that the spine is longer and there is no antero-ventral projection.
The arch plus the spine in each case is half again as long as a
vertebra. At the ninth vertebra the direction of arch and spine
is abruptly reversed, extending obliquely backward instead of
forward. The arches become shorter and the spines increase in
relative length, reaching their maximum close behind the anus,
before the dorsal origin. Here the arch — vertical, nearly median,
and fairly short — gives rise to a spine which is almost twice as
long as the centrum and lies almost horizontally. From here
backward the spine is gradually reduced in size and vanishes at
about the 38th vertebra, the arch once more opening at the top

78

Zoologica: N. Y. Zoological Society

[XX; 3

to form a pair of spines, as in the second and third vertebrae.
In the most posterior part of the column, at about the 107th
vertebra, the arch is again completed, its apex almost touching
the dorsal baseost, and sends a short spine backward. Each of
the first 17 neural arches is strengthened by a cross bar placed
across the arch well below the apex.

Epineurals are distinguishable only on the first eleven ver-
tebrae after the atlas. These are well ossified and there is no
sign of epipleurals or of ribs anywhere along the column.

In the anterior part of the column haemapophyses corre-
spond exactly to the neurapophyses, taking the form of narrow
ridges close to the ventral profiles of the centra.

The first, incomplete, haemal arch — in the form of a pair
of median, vertical spines — does not appear until the 91st ver-
tebra, the 70th behind the anus. This is obviously connected
with the post-anal extension of the coelomic organs. Similar
structures are present on the 92nd, lacking on the 93rd and 94th,
present on the 95th, lacking on the 96th, and present on all sub-
sequent vertebrae. The first complete haemal arch appears on
the 107th vertebra, the apex of the arch almost touching the
anal baseost. Small, backwardly directed spines arise from the
arches of subsequent vertebrae.

End of Vertebral Column and Caudal Fin: The caudalwards
alteration of the vertebrae is so slow and gradual that it is dif-
ficult to say where it begins. The fourth from the urostyle is
very spool-like in lateral view, and shows a large, rounded neural
arch, topped with a short, irregularly backward directed spine.
The haemal arch is smaller with a correspondingly larger and
longer spine. The next three vertebrae are slightly smaller, but
have actually larger arches. In the ultimate vertebrae the neural
arch is slightly open, but the haemal is closed.

Although all the succeeding elements are fused, yet what
we must call the proximal part of the urostyle is intrinsically a
very distorted but entire vertebra. The widely separated arms
of the neural element wave irregularly in midair, and what was
a large, lateral foramen in the haemal arch of the preceding
vertebra has become enlarged and has broken the haemal arch
apart into two pairs of slender, bent processes. Posteriorly there
is a short, ventral pair of spines and the dorsal aspect is drawn

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 79

out into a long, slender projection, longer than the whole main
body. This supports a well developed extension of neural arch,
completely sheathing the spinal chord. About half way to the
end of the vertebral spine, this arch ends, and the naked chord
continues to the base of the caudal rays, then bends rather
abruptly upward, over the dorsalmost ray and actually out some
distance beyond the rounded homogeneous, cartilaginous caudal
plate.

The most dorsal, or first, hypural arises at the posterior
opening of the urostyle vertebra, extends backwards as a fairly
stout rod, and supports two caudal rays. The second hypural at
its base is indistinguishable from the first, but soon separates
and parallels it to the caudal tip where it, too, supports two rays.
The third hypural instead of as usual arising from the sepa-
rate, antipenult vertebra, originates in the same cartilaginous
mass as the two other hypurals. It extends downward and then
backward, leaving a large, rounded foramen along its dorsal
border. From its end arises the final or fifth caudal ray.

Coelomic Organs: Stomach extending slightly behind
anus, black; intestine straight; caeca absent; kidney spotted
throughout, ending above anus, suspended below anterior part

80

Zoological N. Y. Zoological Society

[XX; 3

of gonads ; gonads beginning at perpendicular of origin of stom-
ach, extending almost as far behind anus as anus is from tip of
snout.

Development

Material: All stages are represented in the collection, al-
though transitional adolescents predominate:

Larvae, 17 to 63 mm. : 15 specimens.

Post-larvae, 54, 59 mm. : 2 specimens.

Adolescents, 61, 90 mm. : 2 specimens.

Transitional Adolescents, 90 to 318 mm. : 133 specimens.

Adults, 400 to 440 mm. : 3 specimens.

Total: 155 specimens.

The division between transitional adolescents and adults
is unavoidably arbitrary, as none of our specimens is in breed-
ing condition, and it is unknown whether or not these deep-sea
eels breed more than once. In view of the fact that eggs when
distinguishable, are all of a size, it seems probable that each
individual spawns but a single time. Of the three specimens
over 400 mm., it is not likely that any has eggs far enough
developed to be deposited the same year. If this is the case, we
have of course no fully adult specimens. Nevertheless, it has
been found practicable to set off these three much larger and
more fully ossified specimens from transitional adolescents. Sex
is unquestionably determinable in about 18 specimens, all meas-
uring over 200 mm. ; these are without exception females.

Key to the Growth Stages:

A. Body more or less flattened laterally, its height much greater
than its thickness and contained not more than 20 times in
the length; translucent.

B. Depth in length 7.8 to 10; vertical fins incomplete; no
general pigment.

C. Larval fangs present ; anal origin between 89th and

97th myomere Larva (“Leptocephalus”)

CC. Larval fangs absent; anal origin between 80th and
25th myomere Post-larva (“Tilurid A”)

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 81

BB. Depth in length 15 to 20; vertical fins complete (anal
origin about 20th myomere) ; pigment appearing.

Adolescent (“Tilurid B”)

AA. Body rounded; slender (depth in length 30 to 80) ; opaque.

D. Gonads developing; pigmentation, ossification and den-
tition immature in earlier half of stage

Transitional Adolescent

DD. Specimens ready for breeding Adult

Brief Summary of Changes Occurring During Growth2:
Although the larval stage of Serrivomer, as in all eels, is so ex-
ceedingly unlike the adult in appearance, and unlike the larvae
of other fish both in appearance and in persistence of the stage,
yet in all salient points the leptocephalus is a typical larva, with
gut pendulous, finfolds present, eye vertically elongate, and the
alimentary canal a simple tube. But in two particulars it con-
trasts strongly with the larvae of other fish : first, the body is
compressed to more than flounder-like thinness and is corre-
spondingly deep; and, second, the larval stage persists until the
fish has reached a length of 55 or 65 mm. (in the present genus;
much more in some other eels), instead of having this period
confined to, say, a maximum of 10 mm. of length.

Following this unusual larval stage, however, development
is typical of an elongate deep-sea fish, such as a stomiatoid.
There is a hasty transformation, during which the most blatant
larval characters are swiftly modified into the general appear-
ance of those of the adult: the body becomes rounded and slen-
der, the head is lengthened (contrary to stomiatoid development,
where the head is reduced during this early stage), the eye is
reduced in size and horizontally instead of vertically elongate,
the jaws attenuated, and the fins assume their adult positions —
the dorsal and anal actually moving forward along the profiles
from their previous cramped location posteriorly, exactly as in
the isospondylous Idiacanthus. Also the usual brief period of al-
most complete toothlessness is present, and pigment develops
from adolescence onward. As in other fish, this transformation
takes place during the post-larval period and that of early

2 Detailed descriptions beginning on p. 82. See also table, p. 87.

82 Zoological N. Y. Zoological Society [XX; 3

adolescence. During this time growth almost ceases and an ac-
tual shrinkage of a few millimeters — not more in this species —
probably takes place. Growth is resumed in late adolescence,
when the body is losing the last traces of its leptocephalid com-
pression. This period merges into the usual long, transitional
adolescence of growth and osteological and gynecological devel-
opment, commencing when the fish reaches about 90 mm. Ser-
rivomer, with its slowly grown teeth and digestive organs, re-
sembles the isospondylous stomiatoids much more closely than
it does the digestively precocious Omosudis of the Iniomi. The
extreme forward development of the snout is, of course, unique,
as is the tardily appearing vomerine ridge, which is practically
indistinguishable until transitional adolescence, and yet is the
distinguishing mark of the genus.

Diagnostic Description of Growth Stages:

Larva: (Figs. 24, A to E inch; 25, A to E incl.) — Pigment:
This is found in two or three areas, the first two present through-
out the larval stage, the third tending to disappear in older lar-
vae: 1. An irregular series immediately below the midline, out-
lining the myomeres, usually several chromatophores to each
myomere where they occur, but they are not present on every
myomere. Pigment is especially scanty anteriorly; it may even
be completely lacking in the anterior two-fifths of the fish. 2. A
series of fine dots along the base of the anal fin. 3. A series of 6
larger spots, 3 above and 3 below the tip of the caudal. These
usually vanish when the fish reaches a length of 35 or 45 mm..

Proportions : Depth (measured to exclude intestine and fin-
folds) moderate, 7.1 to 9.6 in length (10.4% to 13%). Head in
length 8.5 to 4.2 (7% to 11.7%), largest in youngest specimens.
Eye 38 to 84 in length (2.6% to 1.2%), decreasing in relative
size with growth, vertically elongate in young specimens, be-
coming rounded in the older ones. Snout long, slender, almost
or more than equal to distance from posterior margin of eye
to pectoral origin; outline at first extremely concave while su-
praorbital distance is greater than the horizontal diameter of
the eye; with growth, the snout loses most of its concavity and
the supraorbital distance becomes much reduced, the highest

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 83

point of the head eventually falling well behind the eye; lower
jaw slightly projecting; jaw angle under posterior margin of
eye in young specimens, distinctly behind it in old; this is due,
however, to the relatively decreased size of the eye, not to a
relative increase in the length of the jaw.

Teeth: A pair of long, curved, fang-like larval teeth at tip
of both jaws, each tooth followed by from five to eleven slightly
smaller, less curved teeth. All of the teeth arise from the outer
sides of the jaw cartilages, not from their margins ; in the oldest
fish, due perhaps to consolidation of the jaw margins, the teeth
are closer to the rim than in the younger specimens. In these
larger ones, too, the upper teeth are in two groups, the posterior
three to five being much smaller than the others. The teeth in-
crease in number until the fish reaches a length of about 35
mm.; just before transformation commences they drop out, un-
til only several are left in each jaw in a transitional specimen
of 57 mm.

Fins: Pectoral large, fan-like, without true rays, base en-
larged, typically larvoid, becoming reduced in size with growth
of fish ; anal short, occupying last one-sixth to one-fourth of fish ;
dorsal considerably longer (up to half again as long) ; rays of
both fins undeveloped in smallest specimen; tail and surround-
ing fin distinctly pointed even in smallest; dorsal ftnfold per-
sisting throughout the larval stage, from the nape backwards,
the finrays elevated upon it; anal finfold similar but starting
considerably behind pectoral fin.

Osteology: Figs. 26 and 27 depict the cartilaginous struc-
ture of the larval head. At this time, remark will only be made
of the complete absence of a supraoccipital, its lack being an
important family character.

Coelomic Organs : The alimentary canal bends sharply
downward immediately behind the heart and from there back-
wards is depended from the myomeral body only by delicate
membranes and ligaments. Anlagen of the stomach and other
digestive organs are distinguishable in the older larvae. The
kidney is clearly visible and ends at about the 30th myomere,
the urethra proceeding from here to the anus. Gonad rudiments
indistinguishable.

Myomeral Count: 154 to 165 myomeres, 89 to 97 in front of

84

Zoologica: N. Y. Zoological Society

[XX; 3

anus (89 in smallest specimen; others with 93 to 97).

Shrinking: As the largest larva (63 mm.) has all its teeth
intact, while one of only 57 mm. shows but several in each jaw,
it is likely that a shrinkage of several millimetres takes place
immediately preceding metamorphosis.

Post-larva : (Figs. 24 F, 25 F). During the post-larval and
adolescent stages the actual transformation from leptocephalous
to anguilliform takes place : the post-larva differs from the larva
most obviously in the loss of larval teeth and in the forward
migration of the vertical fins, with a corresponding shortening
of the intestine and urethra ; the depth gradually decreases to
almost half that of the larva, while the body becomes slightly
but noticeably thickened. The only traces of larval pigment are
a few dots scattered along the anal fin; otherwise the fish is
completely colorless. The head is in form midway between that
of larva and adult: it is relatively longer; all of its parts are
lower and the eye smaller than in the larva, the jaws having
partially assumed their characteristic slenderness; the larval
teeth have entirely disappeared, but very rudimentary perma-
nent teeth are appearing as minute nodules in several rows on
maxillary, mandible and vomer. The pectoral has become fur-
ther reduced and true rays are forming. The dorsal and anal
fins move forward very rapidly although actual growth has prac-
tically ceased, and in the most advanced post-larvae these ver-
tical fins occupy almost their final positions. In the younger of
our two post-larvae (54 mm.) the anus is at the eighty-third
myomere, a forward migration of about 14 myomeres since the
larval stage; in the older (59 mm.) the anus falls still farther
forward, at the seventy-first. Traces only of finfold remain,
but the dorsal and anal are still noticeably elevated. The end
of the caudal peduncle is abruptly attenuated. The gut is still
pendulous, but lies close against the myomeral body. There is
no trace of ossification. The post-larval skull differs from that
of the larva principally in the reduction in its height and in the
extension of the postorbital and opercular regions. Except for
the change in dentition, the jaws show relatively little modi-
fication.

Adolescent : (Figs. 24 G, H and 25 G, H). In the adolescent
the form is semi-leptocephaloid, i.e. there is no mistaking the

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 85

identity of the fish at a glance, but the body is distinctly flattened
and though more slender than in the post-larva is still at least
twice as deep as in the adult. Pigment is developing, spreading
from the ventral portion of the fish upward, but the end of the
tail portion is quite colorless; traces of several larval pigment
spots may remain at the base of the anal. The head is almost of
adult aspect, relatively larger and lower, the eye decreased
nearly to adult proportional size and laterally elongate. The
teeth are developing, but those of the vomer are still in several
rows throughout, with the future vomerine ridge practically in-
distinguishable. All five nostrils are evident. The fins are com-
plete, the anus falling at about the twenty-second myomere.
Neither dorsal nor anal is any longer elevated on finfolds, but
arises directly from the myomeral body. All trace of finfolds
is absent. The intestine is completely enclosed. The viscera of
the adult are all plain — stomach, pylorus, intestine, liver (very
rudimentary), kidney, urethra and thread-like gonads — and all
are in their final positions. The stomach barely reaches the level
of the anus, however, and is scarcely pigmented. There is no
ossification.

Transitional Adolescent : (Figs. 24 I, 25 I) . As in most fishes
the transitional adolescence is a long period of growth previous
to arrival at full breeding condition. As in all eels, however, the
changes remaining to be made following adolescence are greater
than usual, probably because of the great extent of the metamor-
phosis, though they differ only in degree from the majority, not
in kind. In the early part of the stage the body is at its roundest
— even more so than in the full adult, where it becomes second-
arily comparatively flattened in the caudal region; pigment de-
velops steadily as well as a characteristic, iridescent, silver, outer
coat (fragile and almost absent in a number of specimens) . The
head remains of the same proportions as in the adolescent, al-
though the body cavity becomes relatively shorter due to the
increased length of the caudal peduncle, and the postorbital re-
gion develops at the expense of the snout. The teeth attain their
full development very late. This is especially true of those on
the vomer, as this bone retains extra, small, irregular teeth on
each side of the vomerine ridge until the fish has reached a length
of about 150 mm. The extra teeth disappear from the posterior

86 Zoological N. Y. Zoological Society [XX; 3

part of the vomer forward, the vomerine ridge becoming corre-
spondingly pronounced and high.

In a specimen of 132 mm. the teeth, jaws, ethmo-vomer,
hyomandibular and hyoid arch show moderate ossification; the
brain-case, first dozen vertebrae, opercles and pectoral girdle a
slight amount; and the remainder of the vertebral column, the
branchial apparatus, finrays and their supports, and the tail
none whatever. In a 250 mm. specimen all of the ossification is
much stronger, and only the vertical fins, their supports and
the end of the vertebral column are feebly or not at all stained.

The stomach becomes gradually black during this stage and
is extended slightly beyond the anus. Sex can be determined in
fishes of 200 mm. and over, but the gonads are minute.

Relative Proportions During Growth: The table on
page 87 shows the change of proportions and appearance of teeth
during the development of Serrivomer beanii.

Ecology

Seasonal Distribution : Both actually and theoretically
Serrivomer beanii is most numerous during mid-summer, al-
though it is fairly common throughout the trawling season. The
youngest larvae — i.e., several under 30 mm. in length — were
taken in June and July, making it probable that at least some
spawning takes place in late spring or early summer. Larger
larvae, however, were caught straight through from May to Sep-
tember. Successive stages show no correlation with the sea-
sons. (See Table A and B, page 88) .

Vertical Distribution : Serrivomer beanii was taken be-
tween 50 and 1,000 fathoms, at a mean depth of 606 fathoms.
When this average is computed on the theoretical basis of an
equal number of nets drawn at each level, however, the mean is
brought up to 485 fathoms, as Serrivomer was very common in
the comparatively few nets drawn above 500 fathoms. Only two
specimens, both larvae, were captured above 200 fathoms, but
the average depth (727 fathoms) at which all larvae were caught
is deeper even than the mean of the transitional adolescents
and adults (594 fathoms). The question of the number of these
leptocephali which entered the nets on their way to the surface

1936]

Beebe & Crane: Fishes of the Bermuda Expeditions 87

%
a H

l0t>t>0)00a!05O05H05 00H

^OjWhOOSHNNWOON

s

8.S «.S §

-M ^ +j 5h 05

ai a> 5 ^ 5

3 a 3 ft 0) g -5

H &

Snout-to-
Anal :
Length

%

CO CO CO c© asoooco co cd

00 00 00 C— t- t- CD <M (N W

Snout-to-
Dorsal ;
Length

%

CO CD CO N O t> 05 £

, t- t- t- co to oj <m co

Snout :
Length

%

tH 1© CO lO <N1 00 ^ ‘O

i© co co oicocot^ cd cd

Eye:

Length

%

CO ^ OJ <M CO C- CD 00

©1 (N t-H t— 1 t-H t— 1 tH

Head:

Length

%

11.7

9

7.8

7.2

7

8

7.4

16

15.5

16.6

Maximum
Depth :
Length

%

12.5
13
12.8

12.6
12
11.6
11.2
10.9

10.4
13
12
11.8
12.6

10.6

11.4

2.4

1.4
3.2

Post-anal

Myomeres

OsOe-T-HCOO-^tlDOJOSC^-CMas <D t- c- C~» C~‘

ZD CD CD CD r- CD ?D CD CD CDCD ZD Oi

P re-anal
Myomeres

Oi?DlOCOC^‘0>CDr-t-CDC--T^'rr (M M in

00O3O3O3©5O3OSO3O3O3''tf©5O3 05 t- <N (N (M

05

Length

(mm.)

!^i£l£^,;Cl(MCOCOC003<NCOCO C~ 03 O Tt< ©

^WCOWM^TjiTtTf^iaiocO 10 10 05 l-H ^

(M ^

* 9
Aw

c3 • .

* fl * e3

O • >

• -a : S

03c3a3e3^c3e3H3eGc3rtrtc3w^>-*

e3c3e3e3K5c363c3c3e3c3e3e3?-is3©

03 •

c c

.2 g

-t-> O

1 S -
S| s

T3

88

Zoologica: N. Y. Zoological Society

[XX; 8

is in this case especially important : according to the probability
averages, not more than a third of those caught in deep-level nets
were caught on the way up. Both small and large larvae were
taken at high and low levels. (See Table C). No correlation is
shown between depth and season.

Table A.

Relation

of Growth Stage

to Season

April

May

June

July

Aug.

Sept.

Oct.

Total

Larvae

1

5

2

2

5

15

Post-larvae

1

1

2

Adolescent 1

i

2

Trans. Adolescent 9

30

34

30

i.3

16

' i

133

Adult

1

2

3

Total 10

3i

41

33

is

24

i

155

Theoretical total3 68

. 58

57

49

27

24

26

Table B. Length of Larvae in Relation to Season

Length

May June

July

Aug.

Sept.

Total

17-25 mm.

2

2

26-35 mm.

*i

1

’i

3

36-45 mm.

1

i

l

3

46-55 mm.

1

l

2

4

56-63 mm.

2

1

3

Total

l

5

2

2

5

15

Table C .

Growth Stage

Larvae
Post-larvae
Adolescents
Transitional Adolescents
Adults

Average Depth
727 F.

600 F.

550 F.

589 F.

800 F.

If the two-foot Serrivomer- like fish observed from the
Bathysphere between 125 and 250 fathoms actually belonged to
this most abundant species, it is evident that the upper level
of the swimming range of the adults lies far above that at which
they are caught in the nets. In the latter only one small transi-
tional adolescent (140 mm.) was taken as high as 200 fathoms;
no specimen of 400 mm. and over was taken above 700 fathoms.

Abundance: Serrivomer beanii is fairly common among
the deep-sea fishes of Bermuda, and is taken only slightly more

3 Computed on basis of the same number of nets being drawn every month as during
September. See Zoologica, Vol. XVI, No. 1* p. 7.

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 89

often than other numerically comparable forms, such as Idia-
canthus fasciola , Bathytroctes rostratus , Bathylagus glacialis
and Omosudis lowii. It is far and away the most successful of
the deep-sea eels in this locality, its nearest numerical rival in
the collection being Gastrostomus , of which about 85 specimens
have been taken by the Bermuda expeditions (Serrivomer
beanii: 155 specimens) .

Sociability: Judging from the evidence of both Bathy-
sphere observations and trawling records, Serrivomer swims
both alone and in company with several other fish of about the
same stage of development. In the great majority of nets in
which it was taken, Serrivomer was solitary, but in 12 per cent,
(sixteen) from two to five specimens came up together. In all
but two of these cases the fish were transitional adolescents.
One of these exceptional nets held two larvae, the other a larva
and a transitional adolescent.

Food: Serrivomer is primarily an eater of fairly large
shrimps and euphausiids, although fish of equal size and small
crustaceans are occasionally found in the stomachs.

Roule (1934) has developed in detail the theory that deep-
sea fishes which seek relatively large prey undergo long rest-
ing periods between meals, as the majority of his specimens have
proved to have empty stomachs. If this supposition is correct,
Serrivomer would doubtless be included in the group : The stom-
achs and intestines of more than 135 transitional adolescents
and adults of the present collection were examined, but only 23
or about 16 per cent contained more than the faintest trace of
food. Roule draws his principal analogy, however, from the
habits of the great snakes on land — reptiles which actually re-
main quiescent during the “resting period.” Judging both from
the Bathysphere observations and from rationalization, however,
quiescence is impossible in the mid-depths of the ocean, where
it seems that the fish must maintain a certain degree of activity
all of the time. Judging again from Bathysphere observations,
“game” is plentiful, so that in the case of active fish at least,
long abstinence is unnecessary. Still again, some fish with the
greatest stomach capacity (e.g., Omosudis and Chiasmodon)
are very rarely taken without abundant food. And finally, plank-

90 Zoologica: N. Y. Zoological Society [XX; 3

ton feeders, such as Cyclothone and most young fish — which, as
Roule says, have a continuous supply of nourishment — are
taken with empty stomachs almost as often as are the large
predaceous forms. Therefore Roule’s conclusion seems very ques-
tionable; the answer to the puzzle is probably that digestion
takes place with extreme rapidity, at least immediately before
and after death, the process being perhaps stimulated by cap-
ture.

The food was distributed among the 23 Bermuda specimens

as follows:

Stomachs with shrimps or euphausiids 19

Stomach with copepods and Phronima -like hyper id 1

Stomach with 70 -mm, myctophid.... 1

Stomach with Cyclothone signata 1

Stomach with Cyclothone microdon 1

Those of the shrimps which were fairly well preserved
were all Pasiphaea- like; it is likely that all were either of the
same kind or very closely related. None of the remains was
definitely euphausian, but it was impossible to refer the poorly
preserved material definitely to shrimps. The fish and shrimps
were all large compared with the Serrivomer, each completely
filling, but not greatly distending, a stomach.

In the larval intestines remains of radiolarians were found.

Enemies: Serrivomer has not been found inside of any
animal.

Viability: No Bermuda Serrivomer has been brought up
alive. The only record I know of one of these eels being seen
alive is a Serrivomer sector which was caught at 500 fathoms
well up in the Gulf of California halfway between Guaymas and
Santa Inez Bay. This was on April 8, 1936, at Station 139 T-4
of the Zoological Society’s Templeton Crocker Expedition. The
eel was a large one, measuring 580 mm. (22 in.) in length. It
was very active and lived for more than an hour, snapping at
my fingers and swimming with equal ease backward and for-
ward in its aquarium. The iridescent, silvery epidermis was
very noticeable.

Habitat Observations: (See p. 54).

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 91

Study Material

The following list gives the catalogue number, net, depth in
fathoms, date, length and growth stage of each specimen of
Serrivomer beanii taken by the Bermuda Oceanographic Expedi-
tions. All were caught in the cylinder of water off the Bermuda
coast described in Zoologica, Vol. XVI, No. 1, p. 5, and VoL XX,
No. 1, p. 1. “Trans. Adol.,, stands for “Transitional Adolescent.”

No. 9,548; Net 32; 600 F.; April 24, 1929; 138 mm.; Trans. Adol.

No. 9,567; Net 34; 700 F.; April 24, 1929; 96, 106, 134 mm.; Trans. Adol.

No. 9,705; Net 36; 900 F.; April 24, 1929; 318 mm.; Trans. Adol.

No. 9,597; Net 39; 600 F.; April 25, 1929; 114 mm.; Trans. Adol.

No. 9,652; Net 41; 600 F.; April 25, 1929; 90 mm.; Adolescent
No. 9,697; Net 45; 500 F.; April 25, 1929; 144 mm.; Trans. Adol.

No. 9,656; Net 46; 600 F.; April 29, 1929; 127 mm.; Trans. Adol.

No. 9,648; Net 47; 600 F.; April 29, 1929; 98 mm.; Trans. Adol.

No. 9,878; Net 89; 600 F.; May 10, 1929; 120 mm.; Trans. Adol.

No. 9,928; Net 98; 400 F.; May 14, 1929; 115, 115 mm.; Trans. Adol.

No. 9,954; Net 101; 700 F.; May 14, 1929; 119 mm.; Trans. Adol.

No. 9,955; Net 102; 800 F.; May 14, 1929; 115 mm.; Trans. Adol.

No. 9,985; Net 111; 700 F.; May 16, 1929; 133 mm.; Trans. Adol.

No. 10,123; Net 118; 900 F.; May 18, 1929; 102 mm.; Trans'. Adol.

No. 10,138; Net 123; 800 F.; May 25, 1929; 110 mm.; Trans. Adol.

No. 10,152; Net 124; 900 F.; May 25, 1929; 34, 132 mm.; Larva, Trans. Adol
No. 10,175; Net 129; 400 F.; May 27, 1929; 105 mm.; Trans. Adol.

No. 10,183; Net 130; 500 F.; May 27, 1929; 142 mm.; Trans. Adol.

No. 10,319; Net 147; 600 F.; June 1, 1929; 135 mm.; Trans. Adol.

No. 10,350; Net 157; 1,000 F.; June 8, 1929; 120, 132 mm.; Trans. AdoL

No. 10,375; Net 158; 500 F.; June 12, 1929; 138, 141 mm.; Trans. Adol.

No. 10,420; Net 169; 1,000 F.; June 14, 1929; 115, 119 mm.; Trans. Adol.
No. 10,467; Net 170; 500 F.; June 15, 1929; 132 mm.; Trans. Adol.

No. 10,464; Net 174; 900 F.; June 15, 1929; 110 mm.; Trans. Adol.

No. 10,562; Net 184; 700 F.; June 18, 1929; 154 mm.; Trans. Adol.

No. 10,558; Net 185; 900 F.; June 18, 1929; 52, 56 mm.; Larvae

No. 10,559; Net 186; 1,000 F.; June 18, 1929; 42 mm.; Larva

No. 10,596; Net 187; 500 F.; June 19, 1929; 122, 136 mm.; Trans. Adol.

No. 10,683; Net 193; 500 F.; June 20, 1929; 125 mm.; Trans. Adol.

No. 10,753; Net 199; 500 F.; June 21, 1929; 110 mm.; Trans. Adol.

No. 10,754; Net 200; 600 F.; June 21, 1929; 160 mm.; Trans. Adol.

No. 10,757; Net 201; 700 F,; June 21, 1929; 123 mm.; Trans. Adol.

No. 10,755; Net 202; 800 F.; June 21, 1929; 142 mm.; Trans. Adol.

No. 10,875; Net 212; 600 F.; June 24, 1929; 144 mm.; Trans. Adol.

No. 10,876; Net 215; 900 F.; June 24, 1929; 63 mm.; Larva

No. 10,941; Net 217; 500 F.; June 25, 1929; 137 mm.; Trans. Adol.

No. 11,077; Net 233; 600 F.; June 28, 1929; 120 mm.; Trans. Adol.

No. 11,119; Net 240; 700 F.; June 29, 1929; 130 mm.; Trans. Adol.

No. 11,120; Net 241; 800 F.; June 29, 1929; 157 mm.; Trans. Adol.

No. 11,262; Net 254; 500 F.; July 5, 1929; 119 mm.; Trans. Adol.

No. 11,366; Net 271; 1,000 F.; July 8, 1929; 146 mm.; Trans. Adol.

No. 11,444; Net 279; 600 F.; July 10, 1929; 59 mm.; Post-larva
No. 11,705; Net 312; 700 F.; July 22, 1929; 154 mm.; Trans. Adol.

No. 11,913; Net 335; 700 F ; July 29, 1929; 140 mm.; Trans. Adol.

No. 11,984; Net 345; 500 F.; July 31, 1929; 25 mm.; Larva

No. 12,107; Net 355; 600 F.; Aug*. 8, 1929; 123, 152, 162 mm.; Trans. Adol.

92

Zoologica: N. Y. Zoological Society

[XX; 3

No. 12,167; Net 361; 500 F.; Aug. 10, 1929; 128, 129 mm.; Trans. Adol.

No. 12,174; Net 363; 1,000 F.; Aug. 10, 1929; 43 mm.; Larva
No. 12,284; Net 368; 700 F.; Aug. 14, 1929; 150 mm.; Trans. Adol.

No. 12,283; Net 371; 1,000 F.; Aug. 14, 1929; 200 mm.; Trans. Adol.

No. 12,500; Net 388; 900 F.; Aug. 17, 1929; 228 mm.; Trans. Adol.

No. 12,827; Net 399; 900 F.; Aug. 31, 1929; 49 mm.; Larva

No. 13,036; Net 416; 500 F.; Sept. 4, 1929; 138 mm.; Trans. Adol.

No. 13,073; Net 420; 900 F.; Sept. 4, 1929; 46 mm.; Larva
No. 13,134; Net 428; 1,000 F.; Sept. 5, 1929; 405 mm.; Adult
No. 13,163; Net 431; 600 F.; Sept. 6, 1929; 54 mm.; Post-larva
No. 13,237; Net 442; 1,000 F.; Sept. 7, 1929; 151 mm.; Trans. Adol.

No. 13,380; Net 453; 600 F.; Sept. 10, 1929; 123 mm.; Trans. Adol.

No. 14,418; Net 459; 400 F. ; Sept. 12, 1929; 149 mm.; Trans. Adol.

No. 14,741; Net 539; 600 F.; May 6, 1930; 111, 133 mm.; Trans. Adol.

No. 14,779; Net 546; 1,000 F.; May 7 1930; 119 mm.; Trans. Adol.

No. 14,871; Net 562; 500 F.; May 10, 1930; 122 mm.; Trans. Adol.

No. 14,844; Net 563; 600 F.; May 10, 1930; 117 mm.; Trans. Adol.

No. 14,886; Net 567; 700 F.; May 12, 1930; 126 mm.; Trans. Adol.

No. 14,967; Net 573; 400 F.; May 14, 1930; 182 mm.; Trans. Adol.

No. 15,030; Net 580; 400 F.; May 15, 1930; 150 mm.; Trans. Adol.

No. 15,002; Net 583; 700 F.; May 15, 1930; 167 mm.; Trans. Adol.

No. 15,040; Net 587; 500 F.; May 17, 1930; 127 mm.; Trans. Adol.

No. 15,056; Net 588; 600 F.; May 17, 1930; 132 mm.; Trans. Adol.

No. 15,101; Net 594; 400 F.; May 19, 1930; 125 mm.; Trans. Adol.

No. 15,132; Net 595; 500 F.; May 19, 1930; 130 mm.; Trans. Adol.

No. 15,160; Net 604; 400 F.; May 20, 1930; 140 mm.; Trans. Adol.

No. 15,283; Net 619; 500 F.; May 22, 1930; 126 mm.; Trans. Adol.

No. 15,341; Net 626; 500 F.; May 23, 1930; 115 mm.; Trans. Adol.

No. 15,403; Net 631; 400 F.; May 26, 1930; 132 mm.; Trans. Adol.

No. 15,470; Net 637; 500 F.; May 28, 1930; 119 mm.; Trans. Adol.

No. 15,458; Net 639; 700 F.; May 28, 1930; 250 mm.; Trans. Adol.

No. 15,618; Net 657; 700 F.; June 2, 1930; 120 mm.; Trans. Adol.

No. 15,773; Net 684 ;1, 000 F.; June 7, 1930; 130 mm.; Trans. Adol.

No. 15,823; Net 685; 700 F.; June 9, 1930; 125 mm.; Trans. Adol.

No. 16,101; Net 713; 700 F.; June 17, 1930; 400 mm.; Adult
No. 16,536; Net 762; 1,000 F.; July 2, 1930; 146 mm.; Trans. Adol.

No. 16,593; Net 765; 500 F.; July 3, 1930; 200 mm.; Trans. Adol.

No. 16,72 7; Net 776; 500 F.; July 5, 1930; 110, 140, 150 mm.; Trans. Adol.
No. 16,719 ; Net 784; 500 F.; July 7, 1930; 185 mm.; Trans. Adol.

No. 16,954; Net 796; 1,000 F.; July 9, 1930; 270 mm.; Trans. Adol.

No. 17,518; Net 823; 700 F.; Sept. 1, 1930; 440 mm.; Adult.

No. 17,760; Net 835; 500 F.; Sept. 3, 1930; 140 mm.; Trans. Adol.

No. 17,818; Net 842; 600 F.; Sept. 4, 1930; 148 mm.; Trans. Adol.

No. 19,246; Net 933; 600 F.; Sept. 23, 1930; 126 mm.; Trans. Adol.

No. 19,352; Net 944; 400 F.; Sept. 25, 1930; 135 mm.; Trans. Adol.

No. 19,566; Net 966; 400 F.; Sept. 30, 1930; 144 mm.; Trans. Adol.

No. 19,974; Net 967; 500 F.; Sept. 30, 1930; 202 mm.; Trans. Adol.

No. 20,515; Net 983; 500 F.; June 2, 1931; 61 mm.; Adolescent

No. 20,637; Net 994; 900 F.; June 4, 1931; 110 mm.; Trans. Adol.

No. 20,647; Net 996; 400 F.; June 5, 1931; 110 mm.; Trans. Adol.

No. 20,746; Net 1,005; 700 F.; June 6, 1931; 130 mm.; Trans. Adol.

No. 20,747; Net 1,006; 800 F.; June 6, 1931; 160 mm.; Trans. Adol.

No. 20,857; Net 1,014; 500 F.; June 13, 1931; 128 mm.; Trans. Adol.

No. 20,935; Net 1,030; 200 F.; June 22, 1931; 140 mm.; Trans. Adol.

No. 20,976; Net 1,037; 300 F.; June 25, 1931; 120 mm.; Trans. Adol.

No. 21,021; Net 1,042; 100 F.; June 26, 1931; 27 mm,; Larva

No. 21,018; Net 1,043; 300 F.; June 26, 1931; 115 mm.; Trans. Adol.

No. 21,115; Net 1,048; 300 F.; June 27, 1931; 214 mm.; Trans. Adol.

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 93

No. 21,116; Net 1,052; 300 F.; July 6, 1931; 120, 135, 135 mm.; Trans. Adol.
No. 21,153; Net 1,058; 300 F.; July 7, 1931; 110 mm.; Trans. Adol.

No. 21,222; Net 1,067; 300 F.; July 9, 1931; 120, 125 mm.; Trans. Adol.

No. 21,269; Net 1,072; 300 F.; July 10, 1931; 206 mm.; Trans. Adol.

No. 21,270; Net 1,073; 300 F.; July 10, 1931; 110, 120, 125, 130, 147 mm.;

Trans. Adol.

No. 21,325; Net 1,078; 300 F.; July 11, 1931; 120 mm.; Trans. Adol.

No. 22,114; Net 1,097; 700 F.; July 24, 1931; 17 mm.; Larva

No. 21,539; Net 1,101; 400 F.; July 25, 1931; 130, 136 mm.; Trans. AdoL

No. 21,549; Net 1,103; 600 F.; July 25, 1931; 147 mm.; Trans. Adol.

No. 21,625; Net 1,110; 600 F.; July 27, 1931; 216 mm.; Trans. Adol.

No. 21,634; Net 1,112; 900 F.; July 27, 1931; 136 mm.; Trans. Adol.

No. 21,693; Net 1,114; 400 F.; July 29, 1931; 150 mm.; Trans. Adol.

No. 21,983; Net 1,137; 600 F.; Aug. 6, 1931; 245 mm.; Trans. Adol.

No. 21,952; Net 1,139; 700 F.; Aug. 6, 1931; 125 mm.; Trans. Adol.

No. 22,498; Net 1,189; 700 F.; Aug. 17, 1931; 142 mm.; Trans. Adol.

No. 22,962; Net 1,241; 500 F.; Aug. 31, 1931; 127 mm.; Trans. Adol.

No. 23,038; Net 1,243; 700 F.; Aug. 31, 1931; 150 mm.; Trans. Adol.

No. 23,044; Net 1,248; 600 F.; Sept. 1, 1931; 195 mm.; Trans. Adol.

No. 23,079; Net 1,258; 900 F.; Sept. 3, 1931; 34 mm.; Larva
No. 23,140; Net 1,265; 1,000 F.; Sept. 4, 1931; 36 mm.; Larva
No. 23,209; Net 1,276; 500 F.; Sept. 9, 1931; 150 mm.; Trans. Adol.

No. 23,308; Net 1,291; 600 F.; Sept. 12, 1931; 125 mm.; Trans. Adol.

No. 23,397; Net 1,296; 600 F.; Sept. 14, 1931; 110 mm.; Trans. Adol.

No. 23,377; Net 1,300; 1,000 F.; Sept. 14, 1931; 155 mm.; Trans. Adol.

No. 23,455; Net 1,301; 50 F.; Sept. 15, 1931; 46 mm.; Larva

No. 23,549; Net 1,311; 300 F.; Sept. 16, 1931; 57 mm.; Larva

No. 23,577; Net 1,316; 800 F.; Sept. 17, 1931; 130 mm.; Trans. Adol.

No. 23,955; Net 1,339; 700 F.; Oct. 29, 1931; 177 mm.; Trans. Adol.

Serri vomer brevidentatus Roule and Bertin 1929

Specimens Taken by the Bermuda Oceanographic
Expeditions

7 specimens: May to October, 1930 and 1931; 500 to 800
fathoms; from a cylinder of water 8 miles in diameter (5 to 13
miles south of Nonsuch Island, Bermuda), the center of which
is at 32° 12' N. Lat., 64° 36' W. Long.; Standard lengths from
73 to 512 mm.

Description of Adult

(Figs. 34, 35 C, 36 C).

Color: Black, with a fragile coating of silver skin, which,
when fresh, gives off a high bronzy iridescence.

Proportions (from the 512 mm. Bermuda specimen) :
Depth in length 30; head in length 5.8; eye (horizontal) in head
17 ; eye is horizontally elongate ; maxillary reaching well beyond
vertical from posterior margin of eye; snout in head 2.9; snout
to dorsal in length 3 ; snout to anal in length 3.6.

94

Zoologica: N. Y. Zoological Society

[XX; 3

Fig. 34. Serrivomer brevidentatus. (x .7). The pores indicated by the tiny white dots were visible only in the fresh, adult specimen.

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 95

UfUjiU, U {QiWM'WMMt'Jtl'M ( lie{"

B

c

Fig. 35. Serrivomer brevidentatus. A, adolescent, 73 mm. ; B, transitional adolescent, 110
mm. ; C, adult, 512 mm.

Teeth : The dentition has been described in detail on p. 58.

Fins : Pectoral rays 6 or 7, very delicate, equal in length to
horizontal diameter of eye, inserted at upper angle of branchial
cleft. Dorsal rays 160 to 181, commencing well behind the anal
origin, above 13th anal rays, at a distance 1.3 times the post-
orbital length of the head. Anal rays 159 to 173, longer than
those of dorsal. The rays of both fins are longest, and the spaces
between successive rays greatest, in the anterior halves, behind
the first 10 or 15 rays. Caudal rays 5 or 6, scarcely distinguish-
able from those of dorsal and anal, with which the caudal fin is
confluent.

Vertebrae: Largest Bermuda specimen, 151; figures given
by Roule & Bertin, 1929, 143-155; figure given by Trewavas,
1932, 171.

Branchiostegals : 7 to 8.

Osteology: (Figs. 37-40). The skeleton of S. brevidentatus
is exactly similar to that of S. beanii except for the following
characteristics :

1. The ethmo-vomer is shorter and more slender.

2. The preopercle is less extensive.

3. The opercle is quadrilateral, with rounded corners, not tri-
angular, and the closely applied subopercle consequently

does not elbow sharply.

4. The proximal ends of the first five branchiostegal rays do

96

Zoologica: N. Y. Zoological Society

[XX ; 3

\ — l

C

Fig. 36. Serrivomer brevidentatus. A, adolescent, standard length 73 mm. ; B, transi-
tional adolescent, 110 mm. ; C, adult, 512 mm.

not project beyond the antero-dorsal margin of the hyoid
arch.

5. There is only one basibranchial, not two.

6. The coracoids are more rudimentary, showing no connection
with either cleithrum or finrays.

7. The dorsal origin is at the level of the 32nd to 34th (the
latter figure according to Trewavas, loc. cit., p. 651) ver-
tebra, not at about the 27th, and that of the anal at the 25th

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 97

to 26th (latter according to Trewavas) , not at about the
21st.

8. The neural spines of the anterior part of the vertebral col-
umn are shorter, and die out close behind the origin of the
anal fin, at about the 29th, not the 38th vertebra.

9. Epineurals are present on 14 vertebrae, not 11.

10. The neural arch of the last (urostyle) vertebra is more pos-
teriorly extended, sheathing more of the notochord.

11. The third (ventral) hypural is larger, with two instead of
only one foramina.

Minor differences, such as the increased anterior expanse
of the hyomandibular, may be attributed to the larger size (512
mm.) of the stained specimen of S. brevidentatus ; as the largest
S. beanii measures only 440 mm. To the same cause is probably
due the relatively larger ethmovomer and sphenotic of the Ber-
muda S. brevidentatus when compared with Trewavas’s 160 mm.
stained specimen of the same species ( loc . cit., PL III) .

Digestive and Reproductive Systems: (Figs. 41, 42).
These seem identical with those of S. beanii. The 512 mm.
stained specimen was a female with the eggs poorly developed.

Development

No differences are apparent between young specimens of
S. brevidentatus and corresponding examples of S. beanii , ex-
cept for the diagnostic character of the attachment of the
branchiostegal rays. (p. 58). At least three of the transitional
adolescents were females, as well as the single adult.

The seven specimens of S. brevidentatus are distributed as
follows among the growth stages:

Adolescent, 73 mm.: 1 specimen.
Transitional Adolescents, 110 to 280 mm. : 5 specimens.

Adult, 512 mm.: 1 specimen.

Ecology

Seasonal and Vertical Distribution : The present mate-
rial is too scanty for any general conclusions. It may be

98

Zoologica: N. Y. Zoological Society

[XX; 3

.by .by .by

E E E

37. (upper left). Serrivomer brevidentatus. Skull of adult, dorsal view; standard length 512 mm. (x 1.7).

38. (upper right). Same, teeth of upper jaw and vomer, ventral view, (x 1.7).

39. (lower). Same, bones of head, pectoral girdle and anterior part of vertebral column, lateral view, (x 1.7).

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 99

>-

_c

_o-~

or

U
• •

100

Zoologica: N. Y. Zoological Society

[XX; 3

Fig. 41. Serrivomer brevidentatus. Viscera of adolescent, standard length 73 mm.; oes :
oesophagus; stom : stomach; liv : liver; int. intestine; kid: kidney, (x 8.6).

remarked, however, that Serrivomer brevidentatus was taken
throughout the trawling season from May to October, between
500 and 800 fathoms, the average depth being 657 fathoms. The
single adolescent (73 mm.) was taken in late October.

Abundance : In contrast to Serrivomer beanii, which is by
far the most abundant of Bermuda deep-sea eels, Serrivomer
brevidentatus is one of the rarest of all Bermuda deep-sea fishes.

Food : Remains of crustaceans were found in several of the
stomachs.

Enemies, Viability, Habitat Observations: See S. beanii ,
p. 90.

Study Material

The following list gives the catalogue number, net, depth
in fathoms, date, length and growth stage of each specimen of
Serrivomer brevidentatus taken by the Bermuda Oceanographic
Expeditions. All were caught in the cylinder of water off the

Fig. 42. Serrivomer brevidentatus. Viscera of adult, standard length 512 mm.; phar :
pharynx ; other abbreviations as in Fig. 41. The kidneys and gonads were slightly damaged
so that it was impossible to trace the courses of their ducts, (x .5).

1936] Beebe & Crane: Fishes of the Bermuda Expeditions 101

Bermuda coast described in Zoologica, Vol. XVI, No. 1, p. 5.,
and Vol. XX, No. 1, p. 1. “Trans. Adol.” stands for “Transi-
tional Adolescent.”

No. 15,211; Net 612; 600 F.; May 21, 1930; 157 mm.; Trans. Adol.

No. 16,956; Net 778; 700 F.; July 5, 1930; 220 mm.; Trans. Adol.

No. 16,962; Net 793; 700 F.; July 9, 1930; 512 mm.; Adult

No. 18,071; Net 860; 600 F.; Sept. 8, 1930; 280 mm.; Trans. Adol.

No. 20,605; Net 992; 700 F.; June 4, 1931; 265 mm.; Trans. Adol.

No. 23,930; Net 1,336; 500 F.; Oct. 29, 1931; 110 mm.; Trans. Adol.

No. 23,962; Net 1,340; 800 F.; Oct. 29, 1931; 73 mm.; Adolescent.

Bibliography of References Consulted in the
Present Paper

Barnard, K. H.

1925 A Monograph of the Marine Fishes of South Africa. Ann.
South African Mus., Vol. 21, pt. I.

Borodin, N. A.

1931 Atlantic Deep-sea Fishes. Bull. Mus. Comp. Zoology at Harvard
College. Vol. LXXII, No. 3.

Beebe, W.

1931a Bermuda Oceanographic Expeditions 1929-1930. Introduction.
Zoologica, Vol. XIII, No. 1.

1931b Bermuda Oceanographic Expeditions 1929-1930. List of Nets
and Data. Zoologica, Vol. XIII, No. 2.

1932 Bermuda Oceanographic Expeditions 1931. Individual Nets and
Data. Zoologica, Vol. XIII, No. 3.

1933a Preliminary Account of Deep Sea Dives in the Bathysphere
with Especial Reference to one of 2200 Feet. Proc. Nat. Acad.
Sci., Vol. 19, No. 1, pp. 178-188.

1933b Deep-sea Fishes of the Bermuda Oceanographic Expeditions.

Introduction. Zoologica, Vol. XVI, No. 1.

1934 Half Mile Down. Harcourt, Brace and Company, New York.

pp. 125, 126, 164, 203, 268, 277, 320.

1935a Report of the Director, Department of Tropical Research.

Thirty-ninth Annual Report, New York Zoological Society, p. 76.
1935b Deep-sea Fishes of the Bermuda Oceanographic Expeditions.
Family Derichthyidae. Introduction. Zoologica, Vol. XX, No. 1.

Brauer, A.

1906 Die Tiefsee Fische, I. Systematischer Teil. Wiss. Ergebnisse
Deutsch. Tiefsee Exp. Valdivia, Vol. 15, Lief 1.

Garman, S.

1899 The Fishes. Report on an exploration ... by the U. S. Fish
Commission Steamer ‘‘Albatross,” during 1891. Mem. Mus.
Comp. Zool. Harvard Coll., Cambridge, Mass.

Gilbert, C. H.

1905 The Deep-Sea Fishes of the Hawaiian Islands. Bull. U. S. Fish
Comm., Vol. XXIII, Part II, Section II.

102

Zoologica: N. Y . Zoological Society

[XX; 3

Gill, T. and Ryder, J. A.

1883 Diagnosis of new genera of Nemichthyoid eels. Proc. U. S.
Nat. Museum, Washington, Vol. 6, pp. 260-262.

Goode, G. B., and Bean, T. H.

1895 Oceanic Ichthyology. A Treatise on the Deep-sea and Pelagic
Fishes of the World. Special Bull. U. S. Nat. Mus.

Lloyd, R. E.

1909 A description of the deep-sea fish caught by the R. I. M. S.
ship “Investigator” since the year 1900. Mem. Indian Mus.,
Calcutta, vol. 2, pp. 152.

Parr, A. E.

1932 Deep Sea Eels, Exclusive of Larval Forms. Bull. Bing. Ocean.

Coll. New Haven. Vol. Ill, Art. 5.

1934 Report on Experimental Use of a Triangular Trawl for Bathy-
pelagic Collecting; with an account of the fishes obtained and a
revision of the family Cetomimidae. Bull. Bingham Ocean. Coll.
New Haven. Vol. IV, Art. 6.

Roule, L., and Bertin, L.

1929 Les Poissons Apodes Appartenant au Sous-ordre des Nemich-
thydiformes. “Dana” Exped. 1920-1922, Oceanogr. Rep. 4.

Townsend, C. H. and Nichols, J. T.

1925 Deep Sea Fishes of the “Albatross” Lower California Expedi-
tion. Bull. Amer. Mus. Nat. Hist. New York. vol. 52, pp. 1-20.

Trewavas, E.

1932 A Contribution to the Classification of the Fishes of the Order
Apodes, based on the Osteology of some rare Eels. Proc. Zool.
Soc. London, Part 3.

Vaillant, L.

1888 Poissons, Exp. Scient. “Talisman et Travailleur.”

Weber, M. and Beaufort, L. F. de

1916 The Fishes of the Indo-Australian Archipelago, Vol. III.

3nhtx

Albatross collection, 56
Anguilla, 55

Arcturus Oceanographic Expedition, 56
Avocettina scapularostris, 49

Bathylagus glacialis, 89
Bathysphere, 55, 64, 88, 89
Bathytroctes rostratus, 89
Benthenchelys Fowler, 3
Bingham Oceanographic Expeditions, 56

Carencheli, 3
Chiasmodon, 89
Copepods, 90
Crane, Jocelyn, 2
Crustaceans, 21, 48, 54, 89
Cyclothone, 90

microdon, 54, 90
signata, 54, 90

Dana Expedition, 56

Derichthyidae, family, by William Beebe, 1-51
(Figs. 1-9 incl.)

Derichthys Gill, 3-23
iselini, 3, 8, 22
kempi, 5, 23
serpentinus 3, 5-23

(Figs. 3-9 incl.)
adolescents, 19-21
adult, 5-19
development, 19-21
ecology, 21
specimens, 5, 22

Euphausiids, 54, 89, 90

Gastrostomus, 89
Gavialiceps microps, 62
“Glass eels,” 37

Gorgasia Meek and Hildebrand, 3
Grammatocephalus kempi, 23

Hansen, Captain, 48
Hollister, Gloria, 2
Hyperid, 90

Idiacanthus, 43, 81
fasciola, 89
Incertae sedis, 62
Iniomi, 82

Iselin Expedition, 56

Leptocephalus anguilloides Schmidt, 31
ingolfianus, 49

Mohr, Erna W., 54
Muraenoid, 31
Myctophids, 54, 90
Myers, Dr. George S., 54

Nemichthyds, 38
Nemichthys infans, 61
richardi, 62

Nessorhamphidae, family, by William Beebe,
25-51

(Figs. 10-22 incl.)

N essorhamphus Schmidt, 26
ingolfianus, 26-51

(Figs. 10-22 incl.)
adolescent, 29, 37-47

adult, 28-29, 47
development, 29-47
ecology, 48
eggs, 31

larvae, 29, 31-33, 47
post-larvae, 29, 33, 47
specimens, 26, 28, 49
Nonsuch Island, Bermuda, 2, 5

Olsen, Y. H., 54
Omosudis, 82, 89
lowii, 89

Paralepids, 54
Parr, Dr. A. E., 3, 50, 54
Pasiphaea, 90
Phronima, 90
Platuronides, 64

Radiolarian, 48, 90
Roule, L., 89

and Bertin, L., 60, 69, 102
Schizopod, 48

Schmidt, J., 28, 29, 37, 48, 51
Sergestes sp. 21, 48
Serrivomer, 53-102

(Figs. 23-42 incl.)
bibliography, 101-102
distribution, 54 (Fig. 23)
ecology and development, 54-55
taxonomy, 54, 55-61
beanii, 54, 56, 57, 58, 61-63, 65-93
(Figs. 24-33 incl.)
adolescent, 80-82, 84-88
adult, 65-80, 87
development, 80-88
ecology, 86-90
larvae, 80-84, 87-88
post-larvae, 80, 84, 87-88
specimens, 65, 91-93

brevidentatus, 54, 56, 57, 58, 59, 60, 64-65,
93-101

(Figs. 34-42 incl.)
adolescent, 97
adult, 93-97
development, 97
ecology, 97-100
specimens, 93, 100-101
richardi, 62

sector, 54, 56, 57, 58, 60, 61, 62, 63-64,
65, 90

sector longidentatus, 62, 63
sp., 62, 64

Serrivomeridae, family ; Part I : Genus Serri-
vomer, by William Beebe and Jocelyn
Crane, 53-102

(Figs. 23-42 incl.)
for paged outline, see Serrivomer
Shrimp, 21, 48, 54, 89, 90
Sternoptyx diaphana, 54
Stomiatoids, isospondylous, 82
Swanson, George, 2

Taning, Dr., 29, 48
Tee- Van, John, 2

Templeton Crocker Expedition, 56, 90
Thysanopoda, 48

Trewavas, E„ 3, 9, 12, 13, 23, 37, 38, 42, 44,
45, 51

Valdivia specimens, 56

Jleto Horfe Zoological g>ocictp

Scientific Publications

A completely classified list of the subjects included
in each of the finished volumes of Zoologica, and all
other publications of the New York Zoological So-
ciety will be furnished on application.

Address

H. R. MITCHELL

Manager , Zoological Park

185th St. and Southern Boulevard, New York City

1

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS

OF THE

NEW YORK ZOOLOGICAL SOCIETY

VOLUME XXI
1936

Numbers 1-23

PUBLISHED BY THE SOCIETY
THE ZOOLOGICAL PARK, NEW YORK

jgeto gorfe Zoological Society

General Office: 101 Park Avenue, New York City

Officer#

President, Madison Grant

Vice-Presidents, W. Redmond Cross and Kermit Roosevelt
Chairman, Executive Committee, Madison Grant
Treasurer, Cornelius R. Agnew
Secretary, Henry Fairfield Osborne, Jr.

Scientific Staff

Zoological $ark

W. Reid Blair, Director
William T. Hornaday, Director Emeritus
Raymond L. Ditmars, Curator of Mammals and Reptiles
Lee S. Crandall, Curator of Birds
Charles V. Noback, Veterinarian

Claude W. Leister, Ass’t to the Director and Curator, Educational Activities
H. C. Raven, Prosector
Edward R. Osterndorff, Photographer
William Bridges, Editor and Curator of Publications

Aquarium

Charles H. Townsend, Director
C. M. Breder, Jr., Assistant Director

Bcpartmont of tropical &e3earcf)

William Beebe, Director and Honorary Curator of Birds
John Tee-Van, General Associate
Gloria Hollister, Research Associate
Jocelyn Crane, Technical Associate

Cbitorial Committee

Madison Grant, Chairman

Charles H. Townsend
George Bird Grinnell

William Bridges

W. Reid Blair
William Beebe

CONTENTS

PAGE

1. The Reproductive Habits of the North American Sunfishes

(Family Centrarchidae ) . By C. M. Breder, Jr. (Plates
I- VII; Text-figures 1-6) 1

2. Polychaetous Annelids from the Vicinity of Nonsuch

Island, Bermuda. By A. L. Treadwell. (Plates I-III) 49

3. Bermuda Oceanographic Expeditions. Individual Nets and

Data, 1932-1935. By William Beebe 69

4. Plankton of the Bermuda Oceanographic Expeditions. I.

By G. H. Wailes. (Introduction by William Beebe) 75

5. Plankton of the Bermuda Oceanographic Expeditions. II.

Notes on Protozoa. By G. H. Wailes. (Plates I & II) 81

6. Plankton of the Bermuda Oceanographic Expeditions. III.

Notes on Polychaeta. By Edith Berkeley 85

7. Plankton of the Bermuda Oceanographic Expeditions. IV.

Notes on Copepoda. By Charles Branch Wilson 89

8. Plankton of the Bermuda Oceanographic Expeditions. V.

Notes on Schizopoda. By W. M. Tattersall 95

9. Plankton of the Bermuda Oceanographic Expeditions. VI.
Bathypelagic Nemerteans Taken in the Years 1929, 1930
and 1931. By Wesley R. Coe. (Plates I-X; Text-figure

1) 97

10. Tissue Culture and Explantation in Nature: A Review of

Certain Experiments and Possibilities. By C. M. Breder,

Jr 115

11. Preliminary Note on the Nature of the Electrical Dis-

charges of the Electric Eel, Electrophorus electricus
(Linnaeus). By C. W. Coates & R. T. Cox. (Text-fig-
ure 1) 125

V

PAGE

12. The Morphology, Cytology and Life-history of Oodinium

ocellatum Brown, a Dinoflagellate Parasite on Marine
Fishes. By Ross F. Nigrelli. (Plates MX; Text-fig-
ures 1-5) 129

13. The Winter Movements of the Landlocked Alewife, Pomo-

lohus pseudoharengus (Wilson). By C. M. Breder, Jr.

& R. F. Nigrelli. (Text-figures 1-6) 165

14. Systematic Notes on Bermudian and West Indian Tunas of

the Genera Parathunnus and Neothunnus. By William
Beebe & John Tee-Van. (Plates I- VII) 177

15. Food of the Bermuda and West Indian Tunas of the

Genera Parathunnus and Neothunnus. By William
Beebe. (Plates I-III) 195

16. Notes on the Biology and Ecology of Giant Tuna, Phunnus

thynnus Linnaeus, Observed at Portland, Maine. By
Jocelyn Crane. (Plate I; Text-figure 1) 207

17. The Templeton Crocker Expedition. I. Six New Brachyu-

ran Crabs from the Gulf of California. By Steve A.
Glassell 213

18. Neoplastic Diseases in Small Tropical Fishes. By G. M.

Smith, C. W. Coates & L. C. Strong. (Plates I-III) 219

19. The Southwestern Desert Tortoise, Gopher us agassizii. By

Chapman Grant 225

20. Plankton of the Bermuda Oceanographic Expeditions. VII.

Siphonophora Taken During the Year 1931. By Cap-
tain A. K. Totton 231

21. The Female Bitterling as a Biologic Test Animal for Male

Hormone. By Israel S. Kleiner, Abner I. Weisman,
Daniel I. Mishkind & Christopher W. Coates.
(Plate I; Text-figure 1) 24 1

VI

PAGE

22. Some Tropical Fishes as Hosts for the Metacercaria of

Clinostomum complanatum (Rud. 1814) (= C. mar-
ginatum Rud. 1819). By Ross F. Nigrelli. (Plates I
& II) 251

23. Caudal Skeleton of Bermuda Shallow Water Fishes. I.

Order Isospondyli: Elopidae, Megalopidae, Albulidae,
Clupeidae, Dussumieriidae, Engraulidae. By Gloria

Hollister. (Text- figures 1-53) 257

Index to Volume XXI 291

Vll

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS

OF THE

NEW YORK ZOOLOGICAL SOCIETY

VOLUME XXI
Part I

Numbers 1 and 2

PUBLISHED BY THE SOCIETY
THE ZOOLOGICAL PARK, NEW YORK

April 8, 1936

CONTENTS

Page

1. The Reproductive Habits of the North American Sunfishes

(Family Centrarchidae) . By C. M. Breder, Jr. (Plates
I- VII; Text-figures 1-6) 1

2. Polychaetous Annelids from the Vicinity of Nonsuch Island,

Bermuda. By A. L. Treadwell. (Plates I-III) 49

ZOOLOGIC A

SCIENTIFIC CONTRIBUTIONS
OF THE

NEW YORK ZOOLOGICAL SOCIETY

1.

The Reproductive Habits of the North American Sunfishes (Family
Centrarchidae) . By C. M. Breder, Jr., New York Aquarium.

(Plates I-VII ; Text-figures 1-6)

Introduction.

The fresh water sunfishes, Centrarchidae, a large but rather uniform
family confined to North America, are well known even to those not con-
cerned with biological matters. Their nests, excavated in shallow places,
usually on sandy shores, are familiar objects in late spring and early sum-
mer. These have often been described and discussed in more or less detail
in both lay and technical publications. Notwithstanding, there are nu-
merous features of their reproduction that remain to be studied in order
to understand better the particular mode of life that these fishes have
adopted. The present contribution, therefore, has been prepared to extend
our knowledge of these fishes in greater length, with special reference to
the philosophical implications of certain features of reproductive behavior.
The habits of these fishes are so closely similar in many items that the paper
has been arranged by elements in the common reproductive pattern. A sys-
tematic list of species follows this, with explanations of particular dif-
ferences.

The studies referring to Eupomotis at Pines Lake, New Jersey, were
made possible through the courtesies of Mr. H. I. Hartshorn of that place.
His assistance in the field was of considerable value, especially in regard to
the preparation of the chart of nest distribution. All observations made on
Kensico Lake, Wampus Pond and Byram Lake were possible only through
the kind permission of Mr. Herman Forster, Deputy Commissioner of Water
Supply of New York City, as these waters are all part of the drainage
basins supplying that place.

The photographs of Ambloplites and Eupomotis in captivity were taken
at the New York Aquarium by Mr. S. C. Dunton of that institution.

The Reproductive Season.

All the sunfishes spawn in the late spring or early summer. Generally
the largest species, the so-called black basses, Aplites and Micropterus, are
the earliest and although some of the smallest may start early, their season
is apt to last the longest, e.g. Enneacanthus, Mesogonistius , in any one
region.

[1]

li W

2

Zoologica: New York Zoological Society

[XXI :1

If for one reason or another the spawning season is passed without
reproduction, a reabsorption of the eggs results without apparent injury to
the fish. This naturally happens frequently in aquaria where crowding may
make spawning impossible. Specimens of Ambloplites rupestris (Rafin-
esque) that have lived at the New York Aquarium for a period of eighteen
years at this writing and constitute one of the records of longevity of that
institution, Mellen (1919) and Flower (1925 and 1935), may be used as
an illustration. That this frequent passing of the season without spawn-
ing has not destroyed the genital tract is illustrated by the fact that
reproduction will ensue if a suitable environment, usually a reduction of
crowding, is provided. See Breder (1928), as well as the remarks on that
species, and illustrations, in this paper.

Lest it be thought that a long life in an aquarium is an essential part
of the failure to reproduce each season, it may be pointed out that such
occurrences may take place in a state of nature. For example, Mr. Herman
Forster caught a 19-inch Micropterus dolomieu in Kensico Lake on July
4, 1935. The ovaries of this fish were turgid with large eggs, although all
the black bass nests in this lake were but a memory at that date. Young
bass had been out of the influence of their parents for some time and the
schools had just broken up. Scattered remnants of schools composed of a
few specimens could be found, in all of which the fish were over 10 mm. in
length, and they — in all probability — were of the most backward ones. The
ovaries of the fish under discussion were hemorrhagic, the spleen was
grossly enlarged, the liver was mottled and the gall-bladder somewhat en-
larged. The fish externally appeared to be in vigorous health and harbored
no parasitic cysts in the ovaries, not uncommon in these fish, Moore (1925
and 1926) and Bangham (1927), or any other evident parasites. The in-
ference may be taken that at least some of the visceral peculiarities noted
were associated with the early stages of reabsorption of the post-season
eggs.

Apparent differences in the reproductive dates are probably to be
associated in most cases with the speed of the rise to the temperature at
which they spawn. This may be conditioned by the size and depth of the
pond or lake, the temperature and amounts of water in the inflowing streams,
rainfall and general weather conditions, since all influence the responsive-
ness of natural waters to seasonal change. It may also be mentioned that the
reaching of a certain absolute thermal value has relatively little to do with
the actual spawning. It is rather the antecedent temperatures, their dura-
tion and fluctuations, that determine to a considerable measure the metabolic
rate of the gonadal development in such poikilothermal animals.

True to their common name, as soon as the water warms sufficiently to
cause them to disperse from their winter hibernation, the sunfishes move
into shoal waters and bask in the early spring sunshine. The temperature
at which the hibernating habit breaks up is 10° C. for Lepomis auritis , as
shown by Breder and Nigrelli (1934). This species, which forms a winter-
ing school or hibernium, resembles Aplites salmoides to this extent, Town-
send (1916), and Acantharcus pomotis, and to a lesser extent Pomoxis
sparoides. Other species, such as Helioperca macrochira, Apomotis cyanellus,
Eupomotis gibbosus, Chaenobryttus gulosus, Ambloplites rupestris and
Enneacanthus gloriosus, while quiet below this temperature, did not show
such aggregating habits in the aquarium, at least. Passing from this pre-
spawning condition the more aggressive males make their way to the shore
line. The females either follow or remain in deeper water until some later
time.

There is nothing that can be considered a migration, under ordinary-
circumstances, other than the movement inshore from deeper waters. Meek
(1916), however, ascribes an anadromous migration to the Centrarchidae
for which there seems to be no evidence whatever.

1936]

Breder: Reproductive Habits of the Sunfishes

3

The Construction of the Nest.

The males alone are concerned with the construction of the nest. At
this time of year, before actual construction is commenced, there is a cen-
trifugal influence between fish and fish which becomes more intense in
shallow water. In Eupomotis gibbosus, Lepomis auritus and Aplites sal-
moides, for example, during the early spring several fishes may be found
resting quietly side by side in relatively deep water almost in the form of a
school. These make little sallies into shoal water. As soon as they reach the
area destined for nesting a dispersing influence sets in which scatters them.
There may be a little fighting but usually there is little more than a few
desultory chases. Presumably, the entering of the warmer shore stratum
of water increases activity practically instantly. Another factor would ap-
pear to be that of direct solar radiation. At least, sunfish are much more
active and aggressive on bright, sunny days. In fact, the sudden passage
of a cloud usually causes their retreat to deeper water. As soon as it passes
they immediately return to their interrupted task. This effect carries all
through the reproductive procedure, even interrupting the mating act.

Extensive observations on the effect of sunshine were made on Eupo-
motis gibbosus in several places, which agreed perfectly with fewer made
on Lepomis auritus , Aplites salmoides and Micropterus dolomieu. Regard-
ing the behavior of nesting Eupomotis in relation to sunlight, some rather
striking observations were made at Pines Lake. A well-formed colony was
being studied on a day on which fleecy clouds passed rapidly and more or less
regularly across the face of the sun. This colony is illustrated by PI. I, Fig. 1.
While the sun was bright and unclouded the fish remained on their nests
pursuing their usual activities of fanning eggs, clearing nests to receive
eggs and courting visiting females, as well as spawning. On the passage of
a cloud before the sun they all retreated from their nests and remained
headed toward them in a more or less open school that was quite quiescent,
with the exception of three that were inactively resting over their nests at
the time of the change. These remained where they were. As the sun
again made its appearance the others would rapidly return to their nests
and resume their activity. Readings on a Weston exposure meter used for
photographic work showed that the light reflected from the surface of the
water in sunshine reached or exceeded 1,000 on the scale of that instrument.
The above-described phenomenon occurred when the values fell below 700.
On dull days at Llewellyn Lake there was seen to be no activity, the fishes
remaining quietly on their nests. The same was true at night in both places.
A sunfish on its nest at night, photographed by flashlight, is shown by PI. II,
Fig. 2. The flashlight used to find them at night usually drove them off their
nests. If, however, the light was held steadily on a nest for a few minutes
the fish would return, and in one case such a male tried to mate with a
wandering female that had probably been attracted by the light. It would
consequently seem to follow that reproduction takes place only in bright
light and that other times are spent quietly resting over the eggs with only
a nominal fanning of them. In this connection it might be mentioned that
the water of Pines Lake was of a somewhat greater transparency than that
at Llewellyn Lake, which may also have some bearing on the depth at which
various sized nests were elaborated. Whether visible radiation or radiant
heat is responsible for the effect noted could not be determined, but might
well be made the subject of future experiment. There can be no doubt,
however, that the fishes were able to see each other in the dull periods at
Pines Lake. It is equally certain that there was no change in the water
temperature during these short periods.

The selection of the nesting site appears to be largely controlled by two
opposed influences: the centrifugal effect of breeding males, on the one
hand, and the limitation of suitable bottom, on the other, Breder (1935b).
Fish culturists have long known that numerous boxes open on one side per-

4

Zoological New York Zoological Society

[XXI :1

Text-figure 1.

Chart of a colony of Eupomotis gibbosus nests at Pines Lake, New Jersey. This is a different colony from that illus-
trated in PL I, Fig. 1. The two areas marked “False nests” represent sites on which nests were apparently started
and then abandoned. A few such spots are usually to be found in moderately large colonies.

1936]

Breder: Reproductive Habits of the Sunfishes

5

mit more nests of the black basses per given area. These really serve as
“blinders” from one nest to another. A third factor appears to be depth
of water. That is to say, the smaller the fish and the smaller the nest, the
more shallow the water standing over it will be. A number of nests were
measured in this regard at Pines Lake, a location where this was the only
species of sunfish to be found. These are used only because where mixed
species spawn they frequently appropriate each other’s nests, as will be
subsequently discussed. Although such probably occurs between . individ-
uals of the same species, our figures do not suggest it. Table I gives this
data and the actual nests are illustrated by Text-fig. 1, together with per-
tinent data. It will be noted that there is a considerable disposition to nest
close to objects, especially on the part of those nesting in shallow water.
None in water of less than seven inches nested free of some object protect-
ing the nest from one side. Patches of leafy bottom were left strictly alone
and those seeking the most shallow water permitted considerable crowding.
While these data are hardly enough to treat statistically, the second part of
Table I indicates clearly that under five inches 100% nested close to some
object. This percentage falls rapidly and evenly to zero for nests more than
fifteen inches deep. A consideration of Text-fig. 1 shows that this was not
because of an absence of objects to nest against in the deeper water. Note
especially the large stump, often a favorite site for fishes in shallow water.
More than one-third of these nests were found in water between five and
ten inches in depth. This we believe to be referable to the size of the fish—
an item to be discussed at another point. The increase in size with water
depth is also clearly indicated in Table I. The greater tolerance of nearby
nests in water of less than five inches, 85%, would seem to be referable
to a lesser amount of available space close to the shore line. The depth con-
tours in Text-fig. 1 give a measure of the amount of available space in the
various depth ranges, as expressed in Table I. The number of nests in each
depth range, however, is not proportional to the available area. If figures
were available it might be shown, however, that the numbers of various
sized fish compared with the suitable area for their size might determine
the extent of crowding.

Similar measurements were made in Llewellyn Lake. At this place the
fishes were found to be of a considerably larger size and some of the small
non-breeding fishes were found to be as big as the largest fish of Pines
Lake. The reasons, probably ecological, are not evident at this writing.
Numerous recently abandoned nests of Aplites salmoides, which spawns
earlier, were in evidence as well as some actively nesting Lepomis auritus.
In measuring and studying the nests of Eupomotis in this lake it soon be-
came apparent there was a frequent disparity in size between some nests and
their occupants, which was not found in Pines Lake. In both cases only such
nests were considered for study as were occupied by a defending male. This
feature, it was finally seen, was caused by the appropriation of an already
constructed nest by another species, Breder (1985c). Thus Lepomis was
seen to occupy old Aplites nests of sizes that could never have been built by
the occupant. Likewise Eupomotis occupied Lepomis nests and in two cases
even those of Aplites. The proof of this condition was fully established when
a Lepomis on a nest of an appropriate size was seen on the following day to
have been supplanted by a Eupomotis , of much smaller size. Also, a similar
case was seen in a single day between these two species, but over a nest
that had originally been constructed by Aplites. The typical defense re-
actions occurred on disturbance, but no eggs were removed from either.
Table II gives measurements of nests in this lake. The tabular arrangement
does not clearly express these differences, but set forth in graphic form, as
in Text-fig. 2, the relationship of nest size to species and nest size to depth
becomes evident. The smaller fish of Pines Lake form a rather uniform
group increasing in size with depth. Here there was no influence from other
species and the brooding fish were seen to “fit” their nests; that is, they

6

Zoologica: New York Zoological Society

[XXI :1

were approximately one-half the diameter of the nest they occupied. In
Llewellyn Lake nests of large size, nearly twice as large, were built in
water of the equivalent depth selected by the Pines Lake fishes. The fishes on

TABLE I.

Nesting data on Eupomotis at Pines Lake, N. J.

All measurements in inches.

Water

Depth

Diameter of
Nest Floor

Diameter of
Nest Rim

Adjacent to1
Another Nest

Adjacent to1
a Solid Object

4%

5

9

*

4%

Irregular

Irregular

*

*

4%

«<

n

*

*

4%

ll

tt

*

*

4%

a

a

*

*

4%

ii

tt

*

*

4%

u

n

*

*

7

a

<<

*

7

u

tt

*

7

a

n

*

*

8

6

11

*

8%

Irregular

Irregular

*

*

8%

5

9

*

9

6

11

*

9%

91/2

5

9

*

*

*

9%

*

9%

5

9

*

9%

5

9

*

9%

5

9

131/2

5y2

9

13%

6

9

141/2

51/2

9

*

141/2

5

8

*

141/2

5

8

*

15

5

8

*

15

5

8

15

*

16

51/2

8%

16

51/2

8%

17

6

9

*

17

6

9

*

17

6

9

18

51/2

8%

19

5

10

20

8

12

*

22

Average

8

5.6

15

9.4

*

Data according to depth ranges.

Depth Range

No. of
Nests

%of

Nests

Average
Floor dia.

Average
Rim dia.

% Adjacent to
Another Nest

% Adjacent to
a Solid Object

0-5

7

19

5

9

85

100

5-10

13

36

5+

9+

31

92

10-15

8

21

5+

8—

25

25

15-20

8

21

6 —

9+

37

0

over 20

1

3

8

15

100

0

1 An asterisk (*) indicates presence.

1936]

Breder: Reproductive Habits of the Sunfishes

7

these nests fitted them as to size. The other Eupomotis in Llewellyn Lake
did not equal the radius of the nests they occupied, and it is believed these
nests were constructed by other species; the circled ones by Lepomis and
those in triangles by Aplites. The latter is certainly the case and probably
most of the former. Some of these, however, may be cases of smaller
Eupomotis moving to deeper water and larger nests of the same species.
It is to be especially noted that those nests used for the construction of
Text-fig. 2 are by no means all that were considered. They are only those
on which it was possible to obtain sufficiently accurate measurements. Ir-
regularly shaped nests, those with vague outlines, due to the nature of the
bottom, etc., were rigorously excluded.

It is possible that angling may be a contributing factor to the sunfishes’
appropriation of other nests in Llewellyn Lake, for it is a rather small body

TABLE II.

Nesting data on Eupomotis and Lepomis at Llewellyn Lake, N. J.

All measurements in inches.

Water

Depth

Diameter of
Nest Floor

Diameter of
Nest Rim

Occupant

Probable

Builder

3

10

Eupomotis gibbosus

Occupant

4%

11

13

ii

ii

5

12

15

ii

ii

5

14

ii

ii

5%

10

12

ii

ii

5%

14

18

ii

ii

5%

9

ii

a

6

ii

13

ii

ii

6

19

21

ii

Lepomis

6

10

16

ii

Occupant

6%

10

12

ii

u

6%

12

ii

«

7

16

ii

“

7

11

16

ii

ii

7%

10

13

ii

“

8

17

ii

Lepomis

8

14

17

ii

ii

8

15

ii

ii

8%

15

18

it

ii

9

18

20

ii

ii

9%

11

ii

Occupant

9%

12

ii

it

10

13

ii

ii

10

12

ii

Lepomis

11

24

28

ii 1

ii

11

16

19

ii

ii

11%

11

ii

Occupant

12

15

18

ii

Lepomis

15

25

ii

ii

17%

24

27

it

ii

19

11

13

it

Occupant

28

27

30

it

Aplites

29

24

27

it

ii

13

22

24

Lepomis auritus

Occupant

14

24

27

ii

ii

22

36

40

a 2

Aplites

1 Occupied by a four-inch Eupomotis.

2 First seen occupied by a six-inch Eupomotis and later by a seven-inch Lepomis.

8

Zoologica: New York Zoological Society

[XXI :1

of water and angling by boys is carried on with considerable vigor. Since
brooding sunfish will frequently take the hook while on their nests, it is
not surprising that some would be left fatherless. In Pines Lake, on the
other hand, which is a much larger body of water, there is little activity of
this sort and the sunfish are mostly too small to take any but the smallest
hooks. Such angling as is indulged in at this place is mostly trolling with
artificial bait for pickerel and large bass that avoid the shallows used by
the sunfish. That there are other factors involved as well, however, has al-
ready been indicated by the finding of two nests occupied by both Eupomotis
and Lepomis over a period in which there was no angling. Similar observa-
tions concerning these three species were made in Kensico Lake in 1935.
Since there were no variations the above comments cover the situation at
this place equally well.

Text-figure 2.

Distribution of centrarchid nests at Pines and Llewellyn Lakes with
regard to size and depth of water. The appropriation of nests by
various species in the latter lake is indicated by symbols. Nests oc-
cupied by Eupomotis gibbosus, black spot; nests occupied by Lepomis
auritus, light square; nests apparently constructed by Lepomis , light
circle; nests apparently constructed by Aplites, light triangle; other-
wise apparently constructed by the occupant, Eupomotis. The largest
nest, built by Aplites, was occupied by both of the other species at
different times. The figures in the Pines Lake section indicate more
than one nest with identical measurements and water depth.

Small fish in these larger nests go through their customary activity, but
instead of the building of a smaller nest within the larger being the result,
the latter is simply swept clean, maintaining for most part its original con-
tour. This is because the nest has already been excavated to a relatively hard
bottom that the smaller fish could scarcely be expected to dislodge. On the

1936]

Breder: Reproductive Habits of the Sunfishes

9

other hand the light sediment that tends to settle in the nest is readily-
swept over its rim on the slightest activity of the occupant. Apparently
the only other mention of this kind of behavior is that of Hubbs (1919),
who notes that he saw a Eupomotis take over the nest of Helioperca macro -
chira.

More or less casual observation of the nests of these two species over
a period of many years has always revealed them to be located free from
any overhanging object and where sunlight may reach them for at least
part of the day, as of course is well known. Frequently they are adjacent
to some submerged object, as shown in Text-fig. 1. It is in part this tend-
ency that induces the basses to use the nesting boxes of fish culturists.
Although a similar condition was noted on the other side of the canoe dock
of Text-fig 1, absolutely none was constructed under its shelter. In another

locality in Pines Lake where a small foot bridge connects an island with
the mainland a similar condition obtained, but here a single nest was partly
within the shelter of the bridge. At Llewellyn Lake, however, three nests
were found to be constructed in a cavity formed by a large crack in a re-
taining wall. The occupants, just evident from above by their protruding
snouts, would retreat into the crevice on disturbance instead of swimming
some distance off. The condition here is illustrated in Text-fig. 3 in plan
view and section. This nesting site suggested nothing so much as a suit-
able place for Ameiurus to spawn. Adjacent to this location was a Eupo-
motis occupying what was most certainly an old Aplites nest. This is also
shown in Text-fig. 3.

It would seem likely that some part in the selection of nesting sites is
controlled by the social organization of the population of the area involved.

10

Zoologica: New York Zoological Society

[XXI :1

The study of such would involve the constant study of ponds with known
populations of a kind not at present available to the author. While the social
hierarchy of fowls has been long known, the recent note on that of the
lizard Anolis by Evans (1936) contains items that suggest that the terri-
torial holdings of nesting fishes may involve similar elements.

Due to environmental difficulties it is sometimes impossible for fishes
to find suitable places for their particular reproductive needs. Since the
centrarchids require a certain number of factors having well marked limits,
it is not unusual for them to find themselves in positions difficult for repro-
duction if indeed not impossible. These conditions are usually associated
with fishes in a new environment, in which either the latter has been modi-
fied or the fish introduced artificially or intruded naturally. Probably in
none of the illustrations which follow could the fish maintain a continuing
population. For example: Lepomis normally inhabits quiet waters and is
generally not found in rapidly moving water, at least during the spawning
season. A curious condition has somewhat modified this matter of habitat
preference, however, in the vicinity of New York City. Here the destruction
of many trout streams by the building of artificial lakes for real estate de-
velopment has occurred. The streams leaving these lakes have a summer
temperature too high for trouts and have been subsequently invaded by
pond fishes, although in many places they retain the physical characteristics
that are generally associated with trout streams. This condition has been
discussed in some detail by Breder and Redmond (1929). These warm
streams, with nevertheless a swift current, both Eupomotis and Lepomis
have successfully invaded. Their nest construction, however, is distinctly
different from those of individuals residing in quiet ponds. The nests are
widely scattered and each one is to be found on the downstream side of a
large boulder or other form of shelter. The shore line, as such, plays no part
in their location, the sole position-determining influence being shelter from
the current, since these streams are usually not too deep for sunfish nest-
ing at any distance from shore. The physical appearance of such a nest of
Lepomis auritus is shown in both elevation and plan view by Text-fig. 4.
This was prepared from a field sketch made from the banks of Waccabuck
Creek in Pond Ridge Reservation, Westchester County, New York, on June
11, 1933. The arrows indicate the general direction of current. The nu-
merous eddies and back flows at each irregularity of contour are not in-
dicated. Of interest in this connection is the position of the fish immediately
below the large rock protecting the nest. This will be discussed in another
place, as it possesses special significance. The value to reproduction in the
selection of such places in a current is evident, and in fact nesting in unpro-
tected sites would be physically impossible. Examined uncritically this
might well appear to be an expression of intelligence on the part of the nest-
building fish. Actually it can readily be explained on purely mechanical
grounds. Breder and Nigrelli (1934) have shown that this species can
maintain a stationary position in a slight flow with less physical exertion
than in absolutely still water, due to the effects of the exhalant water from
their gill clefts, and that their stationary position always points them into
a current, somewhat after the fashion of a weather vane. When living in
such a flowing stream Lepomis habitually seek sheltered spots, for in other
locations they can only maintain their position by continual active swimming
movements. Consequently, stationary specimens showing little swimming
or water-backing efforts act as positive indicators of the direction of flow.
Fishes below a rock, as in Text-fig. 4, always hold the position indicated,
for at about that point the current deflected to either side rejoins. Fishes
closer to an object than the one shown are found at about right angles to it
and face outward from a midline to meet the flow around the rock. Special
irregularities in certain cases have always been found to be explainable,
as indeed they must be for purely mechanical reasons. From this it follows

1936]

Breder: Reproductive Habits of the Sunfishes

11

that the nest construction is elaborated in a place in which the fish can find
a certain amount of rest — a point of minimum effort. The nest cannot be
excavated when lying parallel to the rock, on account of its nearness, leaving
the only possible position, shown in Text-fig. 4, the only one to be observed
in the field. Fowler (1923) also found this species nesting in streams but
makes no mention of the peculiarities here noted.

Such nests deviate from the circles found in really still water partly
because of the tendency of the current to carry the dislodged particles down-
stream, resulting in a cavity with its longer axis parallel with the flow, as
is shown. Even were it not for this effect of the flow the longer axis would
still have the position, for the fish pointing so much in a single direction
would excavate an oval space. Only in truly still water is a perfect circle
possible, because there alone is there no mechanical differential inducing
the fish to face mostly in one direction. In quiet ponds, where there is
nevertheless a slow “creep” of current along the shore, the fish are found to
mostly face into it. Under such conditions, the only slightly longer axis of
the nearly circular nest is parallel to the current. In other ponds where
there is a fairly steady on-shore wind, the fishes face the shore ; that is, with
the wind, and the long axis of the nest is parallel to it. This phenomenon,
at first difficult to understand, exists simply because the surface water mov-
ing with the wind curls under at the shore line and travels outward along
the bottom as a counter-current, to which the fishes react. This effect is
not common but is mentioned to indicate how slight an influence may bear
on this feature of the sunfish nest. The final check on the direction of water
flow was made in all cases mentioned above by referring to the observed
drift of small suspended particles.

In still water the normal direction of facing points to the source of
disturbance; e.g., another fish, the observer on shore, etc. This effect can
only be seen distinctly in very still water and must be considered secondary
to the mechanical force of the necessarily primary effect of water flow. The
depth of the nest seems to be determined solely hy the nature of the bot-
tom, the fish fanning and excavating according to a standard pattern of
behavior.

In ponds with very soft bottoms and no sandy or gravelly bars or
shores, centrarchids are usually absent. A very sufficient reason for this
will be seen in the following case. A small body of water, Wampus Pond,
was drained about half way down for water supply purposes in the late
winter. It lay in this condition through the 1935 spawning season. Before
this operation the pond had a common type of shore line; mixed between
rocky, sandy, gravelly and muddy and was well populated by three cen-
trarchids, Micropterus dolomieu , Eupomotis gibbosus and Lepomis auritus.
Being a very old natural pond its deeper portions were bottomed uniformly
with soft flocculent mud and the detritus of countless generations of various
aquatic plants. When the water was let down the new shore line left no
sandy or gravelly places suitable for the spawning of sunfishes. There were
some sheer vertical walls of smooth rock but the remainder of the shore
line was covered with a soft, pasty mud of unknown depth, an oar not reach-
ing bottom. Nevertheless, in season, the three species present attempted to
construct nests.

In their efforts to clear away the detritus and reach some solid surface
they succeeded simply in digging in deeper and deeper, so that eventually
they were at the bottom of deep chimney-like pits. In the case of the two
smaller species some of the holes were between a foot and two feet deep.
While they were actively excavating, all that could be seen was a cloud of
fine smoke-like silt pouring out over the lip of the excavation, that neverthe-
less managed to form a raised ridge around the hole, composed of the
heavier material. At other times, when resting, the back of the fish could
just barely be seen through a small cloud at the bottom of the pit, caused

12

Zoologica: New York Zoological Society

[XXI :1

Text-figure 4.

Semi-diagrammatic sketch of Lepomis auritus nesting in a flowing
stream. (Upper). Elevation. (Lower). Plan view. The arrows indi-
cate the direction of the main flow only. Pound Ridge Reservation,

New York.

by respiratory and other movements. The general appearance of these
efforts is shown diagrammatically in Text-fig. 5.

Apparently there was no spawning in these nests, which basically may

1936]

Breder: Reproductive Habits of the Sunfishes

13

have been caused by the male digging himself out of sight. Certainly in no
case could a successful hatch be expected. The extreme opposite of this would
appear to be a bottom so hard that nothing could be moved. What pass for
nests are constructed under such conditions and are quite successful. The
only requirement in this connection seems to be that the surface be sub-
stantially horizontal. Although sunfish eggs are slightly adhesive, appar-
ently no species will use an inclined or vertical surface. At least there is no
record of their ever having done so. The conditions at Wampus Pond could
have led very easily to this if there were any such disposition on the part
of the fish.

Helioperca macrochira was seen nesting on the floor of an ornamental
pond at Lake Mohawk, New Jersey, apparently with entire success. The only
thing here that the fish could brush away was a small amount of filamen-

Diagrammatic sketch of Lepomis auritus attempting to nest in a soft
flocculent mud of unknown depth. The heavy cloud about the fish indi-
cates the extent of turbidity caused by the resting fish. The upper
outline of a cloud indicates the extent of the turbidity caused by nest
building activity. At this depth the fish was unable to raise a cloud
sufficient to leave the pit, thus reaching the limit depth for this kind
of excavation. Wampus Pond, New York.

tous algae which then formed somewhat of a wall about the clean flat stone
flooring. Perhaps a better illustration is provided by Ambloplites spawning
in a tank at the New York Aquarium. This tank was floored with cement
in which were embedded small protruding pebbles. Excavating efforts re-
moved practically nothing but the slightest accumulation of fine material.
PL III, Figs. 3 and 4, shows this type of bottom.

The lack of importance to the centrarchids of a bottom suitable for ex-
cavation is indicated by Wiebe (1935) who wrote of bass in fish cultural
ponds: “Gravel for nest building is not necessary since the largemouth is
just as likely to build a nest on the bare bottom just beside a pile of gravel
as on the gravel.” To this he adds : “Clumps of turned over soil are not in-
frequently used as nests. In this case no excavation is made; the eggs are
simply scattered over the roots. In ponds with excessively soft and muddy
bottom, the largemouth will use the gravel and it would be wise to supply
the gravel in such cases. In ponds the northern smallmouth will spawn un-
der the same conditions as the largemouth except that they require gravel

14

Zoologica: New York Zoological Society

[XXI :1

for nest building.” Further study should be made to determine fully if this
is a real difference between these two species.

The actual mechanics of the nest construction is fairly simple. The
nests are typically circular excavations when it is possible for the fishes to
work on fairly uniform sand or gravel in the absence of current. When rocks
or other objects intervene the nests are correspondingly distorted from this
form. As has been pointed out by Leathers (1911) and others, the depres-
sion is made by vigorous fanning movements of the tail. As these movements
of themselves induce a forward resultant, it follows that the pectorals are
appropriately brought into play, for backing water, to offset this. The fishes
keep turning in all directions with the result that the diameter of the nest
is usually about equal to twice the length of the fish. This relationship is
somewhat modified by the nature of the gravel, the specific gravity of the
material moved, its average diameter, etc. It might be thought that the tail
of the fish points outward, pivoting on its nose as it were. Actually the
reverse is frequently the case, the tip of the tail remaining at the center of
the nest, roiling the sand immediately below the fish and pushing up the
opposite slope as shown in PI. V, Fig. 7. Large objects may be pulled away
by means of the mouth, but this usually happens only when an object is
introduced into a nest already constructed, for sites are generally selected on
fairly clean bottom. Such an object may be a twig or similar object, and
the nesting fish reacts to it in a manner similar to its reaction to an in-
truding fish or crayfish. It is suspected that this behavior is nothing but
the expression of the fighting reaction, as evidenced on an object that
neither flees nor defends itself. Franklin (1914) found that bits of string
and other objects dropped into the nest of Eupomotis gibbosus would be
carried to the edge of the nest by the mouth although no eggs were yet
present. Half a dozen bits caused the fish to retreat and apparently build a
new nest adjacent to the old one.

In addition to this behavior the nesting male on a sandy bottom more
or less persistently picks up mouthfuls of sand and expels them through the
gill-clefts. The function of this is not clear but it is notable that after the
eggs are laid the vigorous fanning is entirely given up, as obviously it must
be to preserve the integrity of the cavity and to prevent loss or burial of
the eggs. At such times the oral manipulation of the parts of the nest not
covered by eggs is more persistently indulged in. Not infrequently, as a
result, the exposed parts of the nest, if in very still water, reveal a pattern
of dimples showing where the fish has nosed into the sand.

Sex Recognition and Courtship.

At other than the spawning season the sexes of the various species of
Centrarchidae are usually somewhat difficult to differentiate. During the
spawning season they are not exactly easy to tell apart, as compared with
many other fishes, but with a little familiarity they usually can be picked
on a number of differential characters. Centrarchids are capable of some
color and pattern change and it so happens that the phase assumed by the
two sexes at this season differs, according to the species. The male assumes
his brightest livery and the female usually one that would, out of season, be
associated with extreme fright. In Eupomotis gibbosus, for example, the
female generally shows a number of dusky, vertical bands on the body.
This pattern is very unlike that of the resplendent breeding male, which is
shot with a variety of brilliant metallic spots and whose red opercular tip is
enlarged and intensified. The pattern of the female is very like that of
an immature male or a mature one that has been badly frightened. This
may be demonstrated by confining two males. If there is some fighting the
loser is at once evident by his “female” pattern that usually reaches its

1936] Breder: Reproductive Habits of the Sunfishes 15

height in three or four days. The philosophical implications of this will be
considered in the subsequent discussion.

In the breeding male various pigmented areas become intensified, as
already mentioned for Eupomotis gibbosus. In Lepomis auritus the reddish
tinge of the pectoral region becomes prominently brighter, the black edging
of the anal and ventrals and the red eye of Ambloplites rupestris become
intensified, and so on.

Although these differences are evident to the trained human eye, the
question of their significance to the fish is not easily settled. Apparently the
only observations attempting to illuminate the possible significance of these
differences to any centrarchid other than the present are those of Noble
(1934). He worked with Eupomotis gibbosus, and the present author chiefly
with Lepomis auritus, and although there is complete agreement as to the
behavior of the fishes the inferences based thereon are not, as has already
been indicated by Breder and Coates (1935a).

In Eupomotis, Noble (1934) found that males would attempt to mate
with a variety of objects, irrespective of their general appearance, pro-
vided they were so manipulated as somewhat to resemble the actions of a
female ready to spawn. In Lepomis auritus, as well as Eupomotis gibbosus,
similar results were obtained by the author at Llewellyn Lake. They may be
best presented by giving the definite experimental results.

A piece of ordinary heavy gray cardboard cut into an oval about the
size of the sunfish studied was attached to a long thin steel rod. With a little
practice this could be moved about under water in a quite lifelike manner.
The following experiments summarize the results obtained.

1. The model was moved toward a brooding sunfish on its nest in a
manner intended to resemble the approach of another fish. When within
about twelve inches of the nest the occupant would always rush toward the
object. If held still the fish would bite the cardboard about half the time,
and the remainder of the time would rush back to its nest, later returning — -
again and again — either to bite or simply to swim close to the intruding
cardboard.

2. Similar to (1) except that on the approach of the guarding sunfish
the model would be rapidly backed off. This resulted in a short chase of
perhaps as much as five feet, which greatly resembled the pursuit of other
fishes. No biting occurred, as the pursuer never caught up with the model.
If the latter was moved slowly the fish followed more slowly, apparently
satisfied to keep the intruder moving.

3. Similar to (1) except that the model was moved forward in the face
of the nester’s attack. In such cases the fish backed into his nest, sometimes
with further attack, but usually without. As nearly as such things can be
understood it seemed that this unaccustomed behavior on the part of the
model resulted in “confusion” on the part of the nest owner. The final result
was a backing off of the owner to a considerable distance where he rested
until the cardboard was removed or caused to “swim away.” This behavior
was apparently identical to that induced by fright, such as too much dis-
turbance on shore.

4. Similar to (3) except that the model was tilted over at an angle
and the hand holding the rod mildly shaken, thus transmitting a quivering
motion to the submerged cardboard. This was done with the model on the
nest, but before the male had retreated. As much as possible, the model was
made to “swim” in a circuit about the nest at the same time. This resulted
in the fish swimming up beside the quivering model and giving evidence of
trying to spawn with it.

These experiments concur completely with the views already set forth
by Noble (1934) to the effect that sex recognition on the part of the male
is based on the differential behavior of the female ready to spawn.

16

Zoologica: New York Zoological Society

[XXI :1

The male brilliance is considered in another connection in the section
discussing protection of the nest and young. Noble (1934) writes that since
the male sunfish are conspicuous and that “movement of bright objects ar-
rests attention” of laboratory fish, “it is probable that a true sexual selection
may occur in the sunfish since the females would presumably move into the
redds which attract their attention first.” Breder and Coates (1935) objected
to this on the basis that “such a condition would appear to be valid only in
the case of a large disparity between the number of males and females.
Thus, a relatively few females, if mating with the first available males (on
the average, most conspicuous) might become exhausted of roe before all
nests received a quota of eggs. Observation by one of us in a scattered va-
riety of places, over a number of years, leads to no such conclusion, however,
since what may be called ‘bachelor’ males have never been noted and the
proportion of the sexes is certainly not low on the female side.” Actual fig-
ures given under the heading “Sex Katio,” show the sexes to be near equal-
ity. Also, as pointed out previously, unspent females may be found at the
end of the season but as yet we know of no finding of unspent males at such
a time.

Fish culturists find that it is advisable to stock breeding ponds of the
black basses with an excess of females. Such results in a greater produc-
tion of fry, see Wiebe (1935). On the other hand, he writes “The use of
more males than females often leads to fighting among the males.” This is
understood to result from the greater activity induced by the continual
failure to find a fish that will enter their nests to spawn, which culminates in
more and more intensified activity of the normal sort on the part of the
unsuccessful males. From various observations it seems that a fish gyrating
over a depression is the signal for a female ready to shed her spawn. Such
places sex recognition, on the part of the female, also on a basis of behavior.
As discussed under “The Spawning Act” the unmated males become more
active in proportion to their failure to attract a female, a feature which
tends to insure a uniform distribution of eggs and surely overides any
possible effect due to small differences in visibility between one male and
another.

There can be no doubt that the females enter the nests of their own
volition. On watching various species it is hard to avoid the impression
that the male interferes with the entry rather than helps it. His attacks
and defensive opercular spreading in response to a variety of stimuli would
simply seem to be overridden by a female bent on spawning. Noble (1934)
has already emphasized the importantly active role the female plays. After
the entry has once been made, barring untoward events, the two fish circle
about together, which generally leads rapidly to spawning.

The Spawning Act.

In all the centrachids known, the female reclines to one side for the
purpose of spawning. A typical position for Ambloplites rupestris is shown
in PI. Ill, Fig. 4. Sometimes if the female is very small, about half the
size of the male, the former may be in quite a horizontal position and as
viewed from above may be nearly hidden from sight.

The minute details of the spawning of the above species have ap-
parently not been recorded. Consequently the following description of it
may serve as an example of centrarchid reproduction, as well as a record
of what is known of the reproductive habits of this species.

Observations and notes on the reproductive habits of Ambloplites have
been published by Wright and Allen (1913), Bean (1903), Jordan and Ever-
mann (1903), Tracy (1910), Hay (1894), Smith (1907), Bensley (1915),
Hankinson (1908), Adams and Hankinson (1928) and Evermann and

1936]

Breder: Reproductive Habits of the Sunfishes

17

Clark (1920). The present writer has not had the good fortune actually
to witness the details of spawning of this species in a feral state. The
Aquarium observations herein discussed agree very closely with the field
studies of others, especially with those of Bensley (1915) who wrote of
the nest as follows: “It is prepared by the male fish which usually works
most energetically, fanning out the sediment with his fins, thus making
a basin-like depression, clean of all debris, and of eight or ten inches in
diameter. The female is driven into the nest and is carefully guarded until
the deposition of eggs is accomplished. During the process of spawning and
fertilization, the two fish lie side by side in the nest. Only a few eggs at
a time are extruded, and at each period milt is extruded by the male. The
operation continues for an hour or more, and at the end of the period the
female leaves the nest and does not return. The eggs are carefully looked
after by the male, which takes up a position over the nest, and every now
and then sets up a fanning motion with the fins. In a few days after the
eggs are hatched, the fry gradually rise out of the nest and are soon left
by the male to shift for themselves.”

Bottoms covered with coarse gravel are usually resorted to by the rock
bass for spawning, but other items will be used when necessary. Hankinson
(1908) describes them spawning among bulrushes. Spawning occurs at
Ithaca, New York, from April to June, according to Wright and Allen
(1913), and from May 15 to June 15 in Lake Maxinkuckee, according to
Evermann and Clark (1920).

The description of Bensley (1915), quoted above, is close to our observa-
tions in aquaria and aside from a few remarks little need be added. Chief
among these is that in the process of spawning the female reclines on her
side and the male remains upright, as is shown in PI. Ill, Fig. 4. This
habit, as previously mentioned, is quite characteristic of the family and
has been described for Eupomotis gibbosus by Leathers (1911) and for
Micropterus dolomieu and Aplites salmoides by Reighard (1906). Other
species are described under specific heads. Probably the entire family
spawns in this type of position. The details of spawning are given below
in chronological order.

July 11, 1933. Two males of Ambloplites rupestris were noted to have
eggs in one of the large exhibition tanks of the New York Aquarium con-
taining about sixteen of the same species. One was near the back of the
aquarium and the other near the glass. The latter is shown in PI. Ill, Figs.
3 and 4, PI. IV, Fig. 5. In other years, when these same fish were in smaller
tanks or living in company with other and more numerous fishes, no repro-
ductive activity was noted. In 1927 (Breder, 1928) these fishes were no
more crowded than this year and then showed reproductive activity. The
floor of this aquarium is composed of concrete in which partly protruding
pebbles are embedded. Consequently not much of a “nest” could be con-
structed, but the males had removed all sediment and detritus, leaving an
area of nest size of a lighter color than the surrounding tank floor. To both
the pebbles and the concrete between them, eggs could be seen adhering.
Mr. H. E. Dixon, in charge of these fishes, reported spawning activity up
to about 9:00 A.M. Figs. 3 and 4, PI. Ill, were taken this day.

July 15. The males still fanning eggs. A female with a protruded
ovipositor attempted to mate with the male near the glass. The ovipositor
was distinctly red and very blunt, about as large in diameter as long. The
behavior was at first very like that of Aequidens described by Breder (1934) .
She quivered considerably and the two fish performed a peculiar rocking
motion in a head-to-tail position. This male did not pursue the female but
continued with little interruption to fan the eggs he was already guarding.
The male remained in the darkest phase and the female took on a light one
with faint vertical bands. Unfortunately, these color differences do not seem

18 Zoologica: New York Zoological Society [XXI :1

to be of photographic quality. The female dragged her ovipositor over th*e
smooth pebbles and stroked it with the tip of her ventral fins, but no eggs
appeared nor was there any spawning activity on the part of the male.
This activity was very like that described for Aequidens in the latter part
of its spawning, Breder (1934).

Such activity continued for several hours until about 3:00 P.M., when
spawning occurred. PI. Ill, Fig. 3, shows the fish just at the finish of one
of the “rocking” periods and PI. Ill, Fig. 4, the actual spawning a little
later. The photograph and earlier discussion of the mating position clearly
mark the most evident difference between centrarchid and cichlid spawning
acts. PI. IV, Fig. 5, shows the male, after the female had left, in a
characteristic “warning” attitude in response to a tap on the glass. The
earlier eggs, supposedly laid on July 11, could be observed hatching while
this spawning was in progress.

July 17. The male on this date was guarding two groups of eggs
alternately. These, though close together, were clearly separate. The young,
which were hatching on July 15, were not to be found. It is thus clearly
apparent that the males of this species are capable of spawning at least
three times over a short period (with as many females, or the same one?).
Also a male may guard two “nests.” This, however, may be a condition im-
posed by the hard bottom where no true nest cavity could be excavated.

July 20. The activity was all over and the males (both) relinquished
their guardianship. No young could be found. It may be noted in passing
that these males were very active in driving off a few large Cambarus kept
in the aquarium as scavengers. These crustaceans would repeatedly attempt
to reach the eggs, only to be driven off when they got within about an inch
of them. It seemed the fish recognized the power of their chelae and were
satisfied to rush at them to a point not quite within striking distance. This
was sufficient, however, to cause the crayfish to back up and put on a con-
siderable display of defense. As soon as the fish withdrew, the Cambarus
would again attempt to reach the eggs. This seesaw might go on for an
hour at a time before the latter would eventually withdraw.

In the summer of 1934 a pair was placed in a small laboratory aquarium.
Almost immediately the male excavated a nest that occupied practically
the entire floor of the tank. This is shown in PI. IV, Fig. 6. This illus-
tration seems to show the type of nest constructed when a uniform bottom
is provided and that it is essentially similar to those of other species, as
well as the fact that confinement in small space is not a deterrent to nesting,
although as previously stated crowding must be so considered. This figure
also illustrates the most prominent secondary sex characters ; the dark edged
ventrals of the male and the white edged ventrals of the female. The dark
edging of the male anal fin is not usually as clear cut a sexual difference as
this particular pair would suggest.

Noble (1934) thinks that the female is in control of the spawning
operation. With this we are in complete accord and indeed it may be con-
sidered as a general condition in fishes, since the expulsion of eggs is usually
dependent on the internal organs of a freely drifting object that may or
may not be able so to do on the stimulation of the male. Whether there is
a male orgasm or not is without biological significance unless the eggs are
present. While observations on sunfish have never been sufficiently close
range to be sure of such, it certainly happens in other fishes. For example:
in Ameiurus nebulosus, pair “B” of Breder (1935), in the summer of that
year, following publication, observations were made that such happened
repeatedly before the female was able to release her eggs.

After the eggs are deposited the female leaves at once. Many females
may visit one nest and one female may visit many nests. In fact, close
observation of Eupomotis gibbosus showed that nearly every nest in a small

1936]

Breder: Reproductive Habits of the Sunfishes

19

colony may be visited by one female. Since sunfish nests usually occur in
colonies and with the habits of the females above mentioned it follows that
the significance of sexual selection cannot be of any particular importance,
even if one especially bright male happens to attract females first. Further-
more, it has been repeatedly observed that the males become agitated on
the appearance of another fish and the increase in their activities certainly
makes them more conspicuous. A lone male on a nest adjacent to one con-
taining a spawning pair, because of his evident excitement is frequently
the most conspicuous object on the entire bed.

Under normal conditions the Centrarchidae seem always to mate in
pairs. Abnormal circumstances may lead to a simultaneous polygamy, how-
ever, as witness the following experiments: One male and three females,
Eupomotis , were placed in a large aquarium in the expectation of spawning.
The male constructed a typical nest, PI. VI, Fig. 9, and spawned repeatedly
with one or another of the females, as was to be expected, PI. VII, Fig. 11.
On one occasion three fish were found in the nest ; the male in the middle of
the group with one female on either side, reclined at nearly 90° and all headed
in one direction. Eggs were extruded and apparently all three took active
part in the spawning. It would seem that under these controlled conditions,
with a dearth of males, that two females becoming egg-burdened simultan-
eously accounts for the phenomenon. Probably if another nesting male had
been available such an occurrence would not have taken place. These four fish
were left in the aquarium throughout the following winter at laboratory
temperature (24°C.). On December 26 the male started excavating a nest
that soon took on the typical form. Pursuing of females was still going on at
this writing, January 30, but no spawning took place. It is thus evident that
the advent of the reproductive season may be considerably advanced by
the maintenance of a continued high temperature.

Protection of the Nest and Young.

After the female has deposited her eggs her role in the perpetuation of
the species may usually be considered finished. The value of incubation of
the male parent is relatively clear. The fanning motions of the pectoral
fins prevent the eggs from being suffocated under a layer of silt. They are
not necessary for aeration, as these eggs hatch excellently in aquaria of
standing water, Eupomotis, Lepomis and Enneacanthus all having been so
hatched by the author. The protection of the parent against predators is
certainly important, as described for Ambloplites in regard to crayfish.
Greeley (1934) describes the behavior of a male Lepomis auritus in defend-
ing its nest against others of its own species. By the simple expedient
of catching the male by hook and line he showed that this defense is neces-
sary, for on capture, he writes: “. . . the small sunfish quickly appropriated
the nest, mouthing the stones greedily to get the eggs.” From this he con-
cluded that “Nest defense of the male is clearly necessary to protection of
young in this and related species.”

This item of reproductive pattern is practically identical with that of
the cichlids (Breder, 1934). The male, however, is the only attendant on
the sunfish nest. In spite of the large literature on sunfish nests there are
few observations recorded in which both sexes may have taken part in pro-
tecting the eggs. Fowler (1923) states of Lepomis auritus that he found
a nest guarded against an active horde of minnows by “both parents, or at
least two adult fishes,” and that “This is the only case I know of where two
sunfish were actually seen guarding one nest.” In reference to Eupomotis
gibbosus he wrote: “Sometimes the female is said to assist the male in
the care of the nest, though I have never noticed this solicitude. Quite
likely the female seen on the nest with the male in most cases may be

20

Zoologica: New York Zoological Society

[XXI :1

incidental.” We may add to this the observation on both species that all
cases that had at first seemed to be this sort of behavior developed to be a
female looking for a nest in which to deposit her spawn, as was subsequently
established. Care by both parents has been described for Aplites salmoides
as exceptional behavior, by Smith (1907) and by Hankinson (1908).

The protection of the nest against larger fishes is a truly remarkable
performance. On one occasion a rather small Lepomis auritus was seen to
rout and chase for some distance a much larger Aplites salmoides. The
larger fish could easily have swallowed the defender in one gulp, but it
literally turned and fled when the sunfish dashed at it. The much larger
Aplites so completely outdistanced the Lepomis in a few strokes of its tail
that the “pursuit” immediately became a farce. The psychological attitude
of the two fishes under such circumstances is not too clear but it would
seem to have to do with territory and proprietorship, which appears again
and again under different guises. It is well known among aquarists that
a fish well established in an aquarium is almost certain to dominate subse-
quently introduced fishes even if the latter are larger and even when there
is no reproductive activity. Presumably, this was the cause of the attitude
observed by Nichols (1918) on the part of two porgys in which one “had
a bullying attitude and the other one a cringing attitude.” See Breder
(1934a) for remarks on a number of species regarding nesting territory
and Evans (1936) for similar data on a lizard.

Care of the young is not as strongly marked in these fishes as in various
other families, but there exists a certain amount of such activity for a
short time. It is probably best marked among Aplites and Micropterus and
has been recorded in the former by Reighard (1906), Smith (1907), Bean
(1903) and Richardson (1913). The author noticed extremely vigorous ac-
tivity in an Aplites protecting a well-formed school from other bass in
Byram Lake, New York. Reighard (1906) describes similar behavior on
the part of Micropterus dolomieu. Such activity as exists in the Centrar-
chidae seems to be very similar to that shown by some cichlids.

Continued and extensive yawning is an accompaniment of egg protec-
tion in numerous species of fishes. The yawn in Monocirrhus Coates (1933)
thought to be a defensive gesture due to the extreme change in appearance
that this movement produces in the species and which he figures. Breder
(1935a) suggested that in Ameiurus this activity might have to do with
aeration. It is also common in the Centrarchidae, see PI. V, Fig. 8. These
fish have neither an horrific appearance when yawning, due to their rela-
tively small mouth, nor can such a movement by any stretch of the imagina-
tion be construed as having an aerating function. It is true that usually
they can all be induced to do it on a not too frightening disturbance, and
it will be done as soon as nest construction is well under way, even before
eggs are laid. It would seem best at the present to consider the phenomenon
to be of some unknown physiological import that takes on other values
dependent on the morphology of the fish, the position of the eggs, their
needs, etc., so that in some forms it may be of “threatening” value and
in others of aerating significance to the eggs, as well as possibly other func-
tions in species as yet unstudied.

Sex Ratios and Natural Hybridization.

Since the Centrarchidae represent a rather closely related group of
fishes, with frequently several species living side by side and resorting to
the same places for reproduction at about the same season and even going
so far as to usurp each other’s nests, it is perhaps not surprising that
hybridization should occur naturally. Hubbs (1920) and Hubbs and Hubbs
(1931), (1932) and (1933), have established the fact of this phenomenon
and as a result have been able to reduce a number of nominal species to

1936]

Breder: Reproductive Habits of the Sunfishes

21

the status of hybrids. While a discussion of their work is outside the
province of the present paper, many of the items of reproductive behavior
herein mentioned give a background against which this extensive natural
hybridization rests. To state it another way, there is probably no group
of fishes, North American at least, in which there would seem to be a con-
catenation of reproductive and other events so well arranged as to lead to
extensive hybridizing ; i.e. the species are numerous ; there is less geographic-
al separation than usual; spawning occurs at about the same temperature
threshold; spawning sites are limited and similar for most species; nests
are exchanged among species. Thus far Hubbs has recorded hybrids be-
tween the following forms:

Chaenobryttus gulosus

X Apomotis cyanellus
X Helioperca macrochira
X Eupomotis gibbosus

Apomotis cyanellus

Lepomis auritus
Xenotis megalotis

Helioperca macrochira

X Xenotis megalotis
X Helioperca macrochira
X Allotis humilis
X Eupomotis gibbosus

X Eupomotis gibbosus

X Helioperca macrochira
X Eupomotis gibbosus

X Eupomotis gibbosus

The sex ratios of these hybrids, including laboratory reared material,
Hubbs and Hubbs found to be predominantly male, 81 to 95 per cent. In
such non-hybrid forms as have been examined the sex ratio is normally
about 1 to 1. Figures given by Hubbs and Hubbs (1933) and Hubbs and
Cooper (1935) follow:

Male

Female

% Male

Apomotis cyanellus

217

192

53

Xenotis megalotis

605

491

55

Eupomotis gibbosus

218

241

47

Helioperca macrochira

125

116

52

Even this slight bias to maleness may be due to collecting methods, which
Hubbs and Cooper suggest. Reid (1930) also found the sexes of Eupomotis
gibbosus to be in about equal numbers. Fish culturists find that the most
successful ponds for the breeding of the three black basses are those in
which the females are in excess.

In this connection it may be pointed out that these hybrids were always
found to be sterile, although males sometimes built nests and attempted
breeding. Hubbs and Hubbs (1933) wrote that an “indication of the un-
natural sexual behavior among Apomotis X Eupomotis hybrids was ex-
hibited by an aquarium-bred individual which played the part of a female
in an attempted mating, vigorously taking the initiative in the female
nuptial behavior. Yet a superficial and histological examination of the
gonads of this fish showed it to be a male.” Aside from the fact that this is
probably the first record of homosexuality in fishes it is of considerable
interest to note that this fish, in spite of its male gonads, imitated the
female act when in the appropriate “psychic” attitude. They continue as fol-
lows: “One of the few females of this same combination which we have
been able to keep under observation repeatedly dug nests and otherwise
played the part of the male in nesting and mating behavior. A male
Apomotis X Helioperca behaved as a female.”

22

[XXI :1

Zoologica: New York Zoological Society

Annotated Specific List of Habits.

The treatment of the reproductive features of the Centrarchidae in the
present paper has been that of considering the various items of behavior as
a whole and illustrating specific points with data on particular species,
mostly based on original observations. This method was pursued because
of the essential uniformity of the group. It remains now to consider and
compare the known data on each of the twenty-five species recognized today.

The classifications followed and species recognized are based on the
tentative opinions of Dr. C. L. Hubbs who kindly gave his views on the
taxonomy of the group, which he emphasized as “extremely tentative/’ in a
personal communication.

Micropterus dolomieu Lacepede.

The small-mouthed black bass, because of its interest to fish culturists
and anglers, has been discussed in regard to reproductive habits by a
large number of students. The author can add nothing to the data already
published by these numerous writers. Apparently throughout its range
Micropterus is the first to spawn with the coming of spring, Aplites being
the only other genus at all approaching it in time and temperature. Beeman
(1924) gives 64° F. as the spawning temperature and 60° as an inhibiting
level. Reighard (1906) gives 62° with the very beginning of nesting below
60°. Nash (1908) gives May to July as the spawning months in Ontario
and Tracy gives as early as March for New York State. The earliest date
noted by the author (Kensico Reservoir) is May 26, with a temperature of
64° F.

According to Wiebe (1935), “Most of the spawning by the smallmouth
occurs probably at 62° to 64°. ” Tester (1930) describes one nest with eggs
at a morning temperature of 55.4°.

According to all observers, nest building is typical for the group, Reig-
hard (1906), Beeman (1924), Adams and Hankinson (1928). The depth
of water according to Beeman is from two to twelve feet, Forbes and
Richardson (1909) about three feet, and Evermann and Clark (1920) about
six feet. The author found them nesting in Kensico Reservoir at depths of
about three feet, at Byram Lake in less than two feet, and in Lake Gilead
in depths considerably in excess of twelve feet. This latter lake is excep-
tionally clear and visibility to at least twenty feet is perfect.

The nest, as in other species, has a diameter roughly equal to twice the
length of the fish, Wright (1892). As previously indicated this varies con-
siderably with the type of the bottom. Beeman (1924) gives two to four
feet for the diameters he found. Forbes and Richardson (1909) write that
the construction may take from four to forty-eight hours. This may be
greatly prolonged, however, if cool or cloudy weather intervenes. Cheney
(1897) found that nest construction stopped if the temperature fell below
65° for long. Thus the breeding, as with most spring reproduction in fishes,
occurs earlier in the southern part of the range. After spending the winter
in a semi-dormant state in deep water, they approach shallow water and
there appear to be no other migrating movements, according to Bensley
(1915). The grounds resorted to are usually some gravelly spot along a lake
shore which may vary from two to twelve feet in depth with about three feet
as an average. Sometimes patches of vegetation are resorted to, according
to Beeman, which probably is associated with a lack of more suitable sites.

The male alone engages in the nest building. After a suitable spot is
selected the bottom is cleared of all loose material. This is done chiefly by
rooting and by means of mouth, aided by fanning motions of the fins, ac-
cording to Reighard (1906). When the nest is once constructed the male
awaits a female and makes short rushes at her when she appears. These

1936]

Breder: Reproductive Habits of the Sunfishes

23

may be found close by, seemingly awaiting the courtship procedure. Accord-
ing to Beeman (1924), this may happen repeatedly, the female remaining
on the nest a little longer each time. At the actual spawning sexual dimor-
phism is distinctly evident. The pattern of dark marks on the body of the
female becomes very distinct as the underlying ground color fades to a very
light tint, according to Adams and Hankinson (1928), quoting Reighard
(1906) : “When the female is finally ready to spawn, there is a marked
change in her appearance. The dark mottlings on her body become very
prominent, due to the ground color becoming much paler than usual. It is
only at the spawning time that there is a prominent sexual difference as
to colors. But close observation will show a red spot on the iris of the
male, which is not ordinarily present in the female.” Also, “during sexual
excitement the female may appear much darker than the male.”

The actual spawning may be quoted from Adams and Hankinson (1928)
as follows: “During these changes the female swims slowly in a circle or
floats motionless, and every two or three minutes rubs her belly against the
stones with a deliberate bending of the body to one side and then to the
other, and the male bites the female frequently though gently, on the opercle,
cheek and corner of the mouth. This act is interpreted by Reighard (L c.,
p. 20) as a stimulus for the emission of the eggs. During the emission of
the eggs, to quote Reighard: ‘The two fish lie side by side on the bottom.
The female is turned partly on her side so that her median plane forms an
angle of about 45° with the plane of the horizon. The male remains upright
with his head just back of the pectoral of the female or opposite it.’ The
male is quiet during the process while the female exhibits certain peculiar
fin movements. The eggs are emitted at periods when the female is with
the male in the nest. Reighard ( l . c. p. 12) noted four such periods occupying
from four to six seconds each and separated by periods of about 30 (22-45)
seconds. The female he observed remained two hours and twenty minutes
with the male in the nest, and when she departed the male pursued her,
but returned to care for the eggs, which meanwhile had become adherent
to the bottom stones of the nest . . .

“The male readily pairs with another female that may approach
the nest, the eggs being deposited with those already laid. Beeman ( l . c .)
noted that the time in which the male shows a disposition to spawn with
different females varies from 30 to 36 hours; and that he appears to be
able to fertilize the eggs of at least three females.

“A female may spawn in more than one nest (Reighard, ’06, p. 12).
Ordinarily a male spawns with but one female at a time, but Beeman (’24,
p. 99) describes a case of a male spawning with two females in the same
nest at the same time, with an alternation of the egg-laying*periods, and
both females leaving at about the same time after their eggs had been laid.

“Beeman (’24, p. 98) mentions males fighting over females, and such
fighting ensues generally when there are too few females to the number of
males in a breeding pond. Lydell (’04, p. 42) also notes fighting of male
fish especially when nests are close together, as they are likely to be in a
small body of water, and gives an instance where a male was killed and
its nest destroyed by the attack of ten or more other males.” See also
Lydell (1926) and Wiebe (1935) on numbers of pairs and sex ratios in
relation to size of hatchery ponds.

The incubation period varies from seven to sixteen days according to
Langlois (1932), varying chiefly with water temperature. At 59° to 60°,
according to Beeman (1924), incubation lasts twenty-one days. Lydell
(1904) gives six days at 60°. Tester (1930) found that if the tempera-
ture is raised from 61° to 73° just before the hatching time, the eggs will
not survive.

The spawning and parental care exhibited by the male is described by
Adams and Hankinson (1928), Bensley (1915), Cheney (1897), Evermann

24

Zoologica: New York Zoological Society

[XXI :1

and Clark (1920), Forbes and Richardson (1909), James (1930), Lydell
(1904), Moore (1926), Tester (1930) and Wright (1892).

The first mentioned write: “The male guards the eggs until they are
hatched. If another fish approaches too near he attacks it, and, according
to Reighard’s observations, the intruding fish will invariably flee.” This
agrees with the present author’s observations on this and other species,
in fact with practically all species of nest building fishes.

The newly hatched fish lie on the stones of the nest for some time but
rise with apparently one accord. After this time the male herds them to-
gether as they weave about in a dense school, very much after the fashion
of cichlid behavior analyzed by Breder (1934a). Probably the same meth-
ods of control are employed, even to the feature of causing the young to
descend to the bottom by means of suitable fin manipulation. Although the
author cannot claim to be certain of this, Howland (1931) indicates that
such is the case.

Micropterus pseudaplites Hubbs.

The reproduction of this species, according to Howland (1932), is
very like that of M. dolomieu. He mentions having observed a male remov-
ing pebbles for the nest by oral means. Also that in the spawning act
the male circles to the outside, which is probably true for the whole
family. Viosca (1931) writes: “The breeding habits of the southern small-
mouth seem to be essentially like those of its northern congener. It spawns
on gravel bars in the spring, but in South Louisiana, later than the large-
mouth.” Howland (1931), who had the species in fish cultural ponds, wrote
as follows concerning the parental care of three males: “The male bass
does not make moves to conduct the school of young fish. As soon as the
fish are hatched he moves off of the nest remaining in the immediate
vicinity, but not over the nest. He does not beat the young fish down by
the movements of his fins as does the small-mouth. The male was observed
to leave the fry at different times. In one case he left them when they were
about five days old, while in the other two cases he stayed with them until
they became approximately a half-inch in length.” Wiebe (1935) mentions
that in fish cultural practice twice as many females as males produce good
results.

Aplites salmoides (Lacepede).

Spawning occurs in spring about the time or even a little earlier than
that of M. dolomieu . Shallow waters are resorted to and the nest may be
constructed in from six inches of water, Richardson (1913), to about six
feet, Evermann and Clark (1920), but they tend to average probably about
two feet.

The water temperature at the time of nesting, so far as the author’s
records go, is 70° F. (Kensico Reservoir) which is in late May in shallow
bays warmer than the main body of water. Most observers seem to have
contented themselves with dates: Forbes and Richardson (1909), May

and June, Illinois; Richardson (1913), April 26 to May 18, Illinois;
Evermann and Clark (1920), May 15 to 30, Indiana; Tracy (1910), April
to May, Rhode Island; Bean (1903), April to July, New York; Bensley
(1915), early June, Georgian Bay; and Hankinson (1908), May 16, Michi-
gan. But as Adams and Hankinson (1928) point out, “Much has been writ-
ten on the life histories of black bass, but the two species are often treated
together, which is unfortunate since there are evidently important distinc-
tions between the two as to breeding.” Other data is given by Lambkin
(1901) and Langlois (1935).

Wiebe (1935) writes: “Wild bass from the Mississippi River, were
placed in a pond on May 16. These fish were seen on the nest the following

1936]

Breder: Reproductive Habits of the Sun fishes

25

day. On May 23rd, eggs and fry were removed from several nests. The
fry were not observed on the surface until May 30th. The temperature dur-
ing this period ranged from 65-73°F. at 8 a.m. and 69-82°F. at 5 p.m.
It is generally assumed that the large-mouth do not spawn at temperatures
much below 64 °F.”

The bottom selected may be gravelly, sandy or covered with dead leaves
or trash. The following description is quoted from Adams and Hankinson
(1928) : “The nest is a simple affair, usually difficult to locate, and many
times can be found only through the behavior of the fish guarding it. Reig-
hard (’06, p. 15) says: ‘They are much less conspicuous than the nests of
the Small-mouth Bass and are usually less excavated. Often the bottom is
covered with dead leaves, fallen from neighboring trees, and the fish has
merely swept away the thin layer of ooze from these and the eggs have been
laid upon them. In other cases the roots of low growing shoots of water
plants have been similarly cleaned. Sometimes an area of sandy gravel has
been swept clean, but has not been hollowed out nor has the sand been re-
moved from among the pebbles. All such nests are inconspicuous and are
usually found only by first observing the presence of the male bass. In but
one case have I seen a Large-mouthed Bass on a nest that was well hollowed
out and in which the sand had been removed from among the pebbles at
the center of the nest. This was, however, in a pond in which Small-mouthed
Bass were also present, so that the work may have been in part that of a
Small-mouthed Bass/ Evermann and Clark (’20, p. 417) describe the nests
as circular depressions filled in with pebbles from about the size of a hen’s
egg down, and the nests as about 2 y2 feet across. Hankinson (’08, p. 214)
describes the nests found at Walnut Lake as circular masses of blackened
bulrush roots. Bensley (’15, p. 41) says, ‘The fish construct nests, by fan-
ning out huge basins with the fins, sometimes three feet in diameter and a
foot into the bottom.’ Nash (’08, p. 89) also describes the nests as made by
scooping out sand and mud. Richardson (’13, p. 414) found the nests to
be well-excavated, nearly round (12-18 in. across) and with grass roots at
the bottom. Forbes and Richardson (’09, p. 268) say that the nests are
built by the males among fallen leaves or fibrous rootlets in sand or gravel.”

The inconspicuousness of these nests has not been noted by the present
author and the differences are believed referable to the types of bottom
available. In Kensico Reservoir the nests of Micropterus are generally
less evident than those of Aplites.

The description of the spawning act and parental care is quoted from
the same authorities: “Spawning has apparently not often been observed,
which may be due to its taking place at dusk, according to Reighard (’06,
p. 15) who gives an account of spawning in artificial ponds near Grand
Rapids, Michigan. The female in this case was somewhat darker colored
than the male and had a more distended abdomen. ‘The male was in the
nest or near it and repeatedly the female approached. The male circled to
her outer side and bit her flank and she then went away. Three or four
other bass, probably males, were seen ten or fifteen feet outside the nest.
I returned at 7 P.M. and found the same conditions. The female was seen
to approach the nest and to turn on her side with her head pointed obliquely
downward and to float thus, as though half dead. In this position she
entered the nest and the male followed and took up a similar position.
What happened in the nest could not be clearly seen. The tails of the two
fish could be seen and from their position it was clear that the fish lay
side by side on the bottom with their tails together and parallel. It could
also be seen that sometimes one and sometimes apparently the other fish lay
turned partly on its side. At this time no doubt the eggs were emitted.
After being in the nest for a short time the fish came out, and the female
was seen to be still floating, head downward. They then returned to the
nest and continued thus for half an hour, alternately lying on the bottom

26

Zoologica: New York Zoological Society

[XXI :1

within the nest and floating on its border. It was then too dark to make
further observations.

“ ‘That the male of the Large-mouthed Bass habitually receives more
than one female into his nest or receives the same female a second time after
a considerable interval is shown by the fact that in three nests in which the
eggs were examined in their earlier stages some were found that had been
recently laid and others that had been laid for forty-eight hours.’ Forbes
and Richardson (’09, p. 269) hold that the male seeks the female and that
the spawning is intermittent. The eggs are adhesive and several thousand
are laid by one fish (Smith, ’07, p. 247 ; Lydell, ’04, p. 40). They adhere to
roots, stones or other objects in the nest bottom. They hatch in 8 to 10
days, according to Forbes and Richardson (’09, p. 269), but Reighard (’03b,
p. 15) says the eggs are hatched usually at the end of three days. They are
guarded by the male and sometimes by both parents (Smith, ’07, p. 247 ;
Hankinson, ’08, p. 214). The young are also attended by the fish during the
time they are in and about the nest (Reighard, ’06, p. 16; Smith, ’07, p. 247).
The young may remain in the nest a week or ten days (Bean, ’03, p. 492).
After leaving it they swim in compact schools. Richardson (’13, p. 415)
noted 6000 young in two schools.”

This apparent spawning in weak light is certainly not typical of the
generality of Centrarchidae.

Chaenobryttus gulosus (Cuvier and Valenciennes).

The warmouth is reported as breeding in Illinois in June by Hubbs
(1919). The nest he described was fifteen feet from shore in about three
feet of water.

Apomotis cyanellus (Rafinesque).

A green sunfish, discussed by Hubbs (1919), is described as building
no nest at all but attaching its eggs to willow roots. It is hardly likely that
this is normal for such a fish but that it represents a badly built nest similar
to numerous others described herein. Hubbs and Cooper (1935) show that
the males are generally larger than the females and spawning takes place
from late June to August in Michigan.

Sclerotis punctatus (Cuvier and Valenciennes).

Apparently there is no data at all on this species.

Lepomis auritus (Linnaeus).

Considering the abundance and wide distribution of the redbreasted
sunfish it is surprising that there are not more references to it in the litera-
ture than appears to be the case. In the vicinity of New York its conspicu-
ous nests vie with those of Eupomotis in abundance. Nests have been found
from June 6 to as late as August 12 at temperatures ranging from 68° to
82° F. Fowler (1923) gives June 10 to June 25 in the vicinity of Phila-
delphia. Nests vary from about 10 inches to 24 inches in diameter and
are found in depths of from six inches to 18 inches in depth. Fowler (1923)
gives a diameter of twelve inches to sixteen inches and in a depth of water
from less than a foot to about twenty inches. For more comprehensive de-
tails on the nesting of this species see Table II and Table III. Text-figs. 4
and 5, together with the accompanying text, show the extremes to which
this species will go under favorable and unfavorable conditions, obviating
repetition here. See also Breder and Redmond (1929), who figure a typical
nest.

The nest building and courtship of this species form much of the basis
of the preceding sections and need not be repeated here, as in all cases of

1936] Breder: Reproductive Habits of the Sunfishes 27

details the species definitely studied is indicated. The frequent association
of this species with Eupomotis and Aplites leads to the complex interactions
of nesting already alluded to.

The pre-spawning behavior of this sunfish has been studied by Breder
and Nigrelli (1934). They indicate a primary urge to aggregate on the

TABLE III.

Nesting dates of Lepomis auritus in the vicinity of New York City.

Locality

Year

Pre-nesting

Period

Nesting Period

Post-nesting

Period

Temp.

°C.

Branch Brook Park,
Newark, N. J.

1913

June 13

1926

June 6

—

20

1927

May 11

—

—

Singac, N. J.

1919

June 30-July 12

—

Haskell, N. J.

1928

—

July 15-27

—

__

Pound Ridge, N. Y.

1933

. — .

June 11

—

—

1934

June 3

—

—

—

Llewellyn Lake, N. J.

1934

May 20

July 14

July 28

25.5

Lake Mohawk, N. J.

1934

May 20

July 14-29

Aug. 12

25.5

Kensico Reservoir

1935

June 9

June 14

Aug. 15

27.5

Byram Lake

1935

June 9

June 14

J uly 6

27.5

Range of dates ......

May 11-June 9

June 6- July 29

July 6- Aug. 15

20-27.5

part of fishes generally, except as inhibited by definite influences such as
reproductive urges, etc. The definite formation of schools was found to
occur, in adult material, only below 5° C.

Helioperca macrochira (Rafinesque).

The blue-gill, similar to the redbreast in habit, usually nests in similar
colonies. Hankinson (1908) mentions from nine to fifteen to a colony.
Spawning occurs in May in Illinois, as mentioned by Richardson (1913) ;
in May and June in New York, by Wright and Allen (1913) ; in May in
South Dakota by Churchill and Over (1933), and in June in Indiana by
Evermann and Clark (1920). Coggeshall (1923 and 1924) gives related
data. The present author has found them on nests from July 14 to 28 in
Lake Mohawk, northern New Jersey, with a temperature of 76° F. Of the
nesting of this species, Richardson (1913) remarks as follows: ‘‘The nests
were chiefly in bunches about the bases of the willows, in some cases as
many as a dozen about one tree, all in the shade, and many of them only two
or three feet apart. This fish seems particular to select about the same sort
of situation for all its nests — a rather hard bottom of sand and mud, with
little vegetation, but with some fine dead drift, grass, twigs, etc. The nests
are eight to twelve inches in diameter, usually quite round, and the excava-
tion of the bottom soil is always well marked — usually to a depth of half an
inch or an inch. . . . The males are much more shy than males of the war-
mouth bass, but they can easily be seen and identified on nests by approach-
ing quietly.”

The eggs of this species, experimentally immersed in sea water, were
found to sink rapidly even in water concentrated to a density of 1.035 sp. g.

Xenotis megalotis (Rafinesque).

The long-eared sunfish nests, according to Hankinson (1908), in Mich-
igan in July and in New York, according to Adams and Hankinson (1928),
during the same month, and from late June to August in Michigan according
to Hubbs and Cooper (1935). They describe the nests as follows: “Several
nests in about a foot of water were saucer-shaped depressions like the nests

28

Zoologica : New York Zoological Society

[XXI :1

of other sunfish. The bottom here was of fine gravel, of a character different
from any other bottom material in the lake for it had been hauled there for
some construction work, probably as a support for a pier. The eggs were on
the bottom stones. An adult male in gaudy breeding dress guarded each
nest, and small companies of females were moving about in the vicinity.
All of the nests were found on this patch of gravel, except one, which was in
Milton Point Bay, about a mile northeast of this place. This one exception
was situated close to the shore and was similar to the nests found at Brewer-
ton. A male was guarding it, but no eggs could be found. All of these nests
were found on July 25, 1916. The attending males were not quite four inches
in length; the females were decidedly smaller, nearer three inches long.”

According to Hankinson (1908), the eggs may be attached to the roots
of plants after the latter have been cleaned of bottom mud. Hubbs and
Cooper (1935) show that the males grow faster and generally reach a
larger size.

Allotis humilis Girard.

A detailed study of the orange-spotted sunfish is given by Barney and
Anson (1923) from which the following is drawn. Sexual differentiation
during the reproductive period is well marked, the males being even more
vividly colored, in a complicated pattern, than in most sunfishes.

The spawning season begins in Louisiana near the first of April and
extends to September and in Iowa from late May to August, when the tem-
perature reaches about 65°. The younger fish nest later and thus account
for the long spawning period, a condition probably true for the entire
family. The nests in soft bottom measured from 15 to 18 cm. in diameter
in a depth of from 12 to 36 inches of water. In Louisiana, with only a mud
bottom, apparently no distinct nest is formed. Females were found with
up to 4,200 eggs. Exceedingly large colonies were found by Barney and
Anson, one having 960 nests on a bank 365 feet long and of only a few feet
in width.

Lethogrammus symmetricus (Forbes).

Apparently there have been no data recorded on the reproduction of
this species.

Ewpomotis gibbosus (Linnaeus).

The common sunfish spawns in late spring or early summer. Abbott
(1884), Gill (1907), Hankinson (1908 and 1909), Forbes and Richardson
(1909) and Wright and Allen (1913) place nest building in May and
Bensley (1915) and Leathers (1911) up to August. Spawning has been
observed in June and July by Hankinson (1908), Leathers (1911) and
Embody (1915). As indicated in Table IV, nesting about New York City
extends from May to July. Activities in aquaria are described by Becker
(1908). Wild nests are figured by Breder (1926) and Breder and Redmond
(1928) and discussed by Krecker (1916) and Myers (1921).

The nest is usually constructed in less than two feet of water, generally
in quiet ponds, lakes or creeks. Extensive details are given in Tables I, II
and IV, Text-figs. 1, 2 and 3, Pis. I, II, Fig. 1, 2 and V-VII, Figs. 7-12, in-
clusive, with their accompanying discussion. The bottom may be of clay and
gravel or sand but shows considerable variation, as indicated in an earlier
part of the present study. The nesting and reproductive activity quoted
below is from Adams and Hankinson (1928) :

“The nest is typical of the other sunfish nests in being a more or less
circular bottom depression, made by a fanning movement of the tail; and
objects too large or heavy to be removed by this method are pulled away

1936]

Breder: Reproductive Habits of the Sunfishes

29

by means of the mouth (Leathers, ’ll, p. 252). The nests are usually as
nearly circular as bottom features will permit and in diameter are com-
monly about twice the length of the fish. A gravid female is brought to
the nest by the male, and in the spawning act the two fish apply their ven-
tral surfaces and move about in a circle, the eggs and sperm exuding.
Leathers (Z. c., p. 253) counted eleven circuits a minute made by spawning
individuals, and found that the male remains upright, the female horizontal.
Clouds of sperms intermixed with eggs could be seen emitted at intervals
and at such times the female would make quick tail movements, throwing
herself into an upright position.”

These observations are in essential accord with those of the author
except the bringing of a female to the nest by the male. The following de-
scription of difference of the sexes and the reproductive habits, also in sub-
stantial accord, are quoted from the same authorities:

“Reighard (’02, p. 575) notes that the male is brighter colored than
the female, with brighter vermicular cheek markings, and with black ventral
fins while those of the female are yellow ; and the dorsal and caudal fins in
the male a more brilliant hue. He also noted that the opercular flap in the
male is larger. In many observations made on spawning Common Sunfish,
Hankinson found the female usually smaller and decidedly lighter in color
and less brilliant, resembling the immature rather than the adult male.
Apparently it is only the male that constructs and attends the nest (Reig-
hard, ’02, p. 575; Bean, ’03, p. 485). He guards the eggs against other
fishes and other intruders. His boldness at this time is well known, and he
goes so far as to bite hands and fingers if held near the nest. The spread-
ing of the gill-covers and the displaying of colors appear to be instrumental
in driving away intruders (Reighard, ’02, p. 575) as well as in attracting
the female. It has been generally assumed that this sunfish and others guard
only the eggs and not the young. In this connection the observations of
Evermann and Clark (’20, Vol. 1, p. 408) are of interest, with regard to a
nest of Common Sunfish found July 7, 1901: ‘The young were quite minute,
transparent objects, the eyes being the most conspicuous part of them. They
hugged the bottom quite closely, but were pretty active. Now and then one
of them appeared to take a notion to leave the nest, and would swim up
toward the surface. Quick as a flash the parent fish would snap it up, and
it appeared at first glance as if it were devouring its young, but it was soon
discovered that each time it had taken in a young fish it immediately went
down to the bottom of the nest, head downward and spat the young out into
the nest near the ground.’ The eggs adhere to bottom objects such as soil
particles, small stones, roots and sticks.”

This habit of returning the young to the nest can only be considered as
feebly developed in the Centrarchidae, although it is typical of the Cichlidae.
Few students have seen sunfishes transporting their young by oral or other
means. The author has never seen anything even remotely resembling such
behavior.

Experimental studies in the sex recognition of this species have been
made by Noble (1934), the observations of which concur fully with those
of the present author but who does not find it possible to draw the same
inferences, Breder and Coates (1935a). These differences of interpretation
are considered in the discussion. The actual experimental results are men-
tioned in detail under the previous consideration of sex recognition. Other
items in connection with this feature of behavior are as follows: At the
site illustrated in Text-fig. 1 large numbers of females could be found under
the shelter of the canoe dock. These would cruise out in the vicinity of the
nests and be pursued by one or more males. Such attention usually drove
them back to the inactive group of females, but finally after considerable
play a female would enter a nest and spawning would ensue almost imme-

30

Zoologica: New York Zoological Society

[XXI :1

diately, the female inclining to one side in the typical centrarchid fashion.
Other males in nearby nests could be seen to pay the strictest attention, but
in no case was any attempt to interrupt the proceedings noted, once the
female had definitely entered a rival’s nest. The males were at their bright-
est and the females tended to display vertical bars, similar to those evident
in immature and badly frightened fish, see PI. VII, Fig. 11. If these dif-
ferences of pattern bear any social significance to these fishes it is not
evident from the present studies. On the nesting grounds, any wandering
fish even of other species, such as young Micropterus or adult Notemigonus,
may be pursued by males which sometimes leave their nests for considerable
distances. Flight is the customary reaction of the pursued, on which the
pursuer returns to his nest. The females usually perform in a similar man-
ner but when nearly ready to spawn do not retreat so far. They return
again and again and may seem to be driven into the nest by the male. The
direction of “driving” appears to be determined entirely by the direction
the pursued elects to follow. This view would give the role of spawning de-
termination to the female. That is to say, the males having established
themselves on a nest pursue practically anything, giving up the chase only
when it leads far away from the nest. This view fuses the “fighting” and
“courting” behavior into one, with the behavior of the female as the deter-
mining element. This is essentially the type of sex recognition already dis-
cussed for the cichlid Aequidens by Breder (1934a). Noble and Curtis
(1935) believe, however, they have detected selection of males on a color
basis by the females of a cichlid, but as the report is a short abstract, de-
tails are insufficient to interpret their findings. The trial and error method
in the present case rather presents a clearer cut picture than in the cichlid's
because of the mechanical circumstances involving the construction of a
nest prior to courtship.

Experiments were undertaken with sunfish in aquaria. Four fish were
used; three females and one male. The male constructed a large, well-
formed nest despite the fact that the aquarium was comparatively small —
3 feet long by 2 feet high by 1% feet wide. During the construction period
he prevented the others from approaching the center of the nest only by about
six inches. This short distance presumably was controlled by the small size
of the aquarium. The nest was started July 1 and worked on continually,
but no spawning was seen to be attempted before July 5. These attempts
were continued alternately with persistent nest cleaning but there were no
eggs until July 17. These were hatched on July 20 at a temperature of
28°C. The spawning of July 17 started at 5 P.M. and continued for an
hour. After a while a second female joined the original pair, and the
three attempted (or succeeded?) in spawning together. The male was
upright between the two females, one inclined on either side of him and
all headed in the same direction. This irregularity in behavior has already
been mentioned under the discussion of the spawning act. The young re-
sulting from this spawning had all disappeared by July 22. On July 24 the
male was again vigorously cleaning the nest and before August it was in
finished condition. On August 2 young fish were again found in the nest,
thus establishing the fact that these males may spawn at least twice in a
season in the same nest, with the same or with other females. Typical
poses of this male are given in Pis. V-VII, Figs. 7-11, in aquaria, and of
one in a state of nature, PI. VII, Fig. 12. This male again built a nest in
December, the details of which are discussed in the previous general section.

Abbott (1884) wrote as follows about this species: “Each fish, wher-
ever it may go, has some point which is recognized as the terminus of the
lane leading to the nest, and having found this it speeds up the narrow
pathway with incredible velocity and stops as suddenly just at or in the
nest.” Gill (1907) commented that “The actions noted by Abbott must be
manifested only under certain conditions. I have not noted analogous in-

1936]

Breder: Reproductive Habits of the Sunfishes

31

stances.” Among all the species studied in the field, observations were made
by the writer which in part confirmed Abbott’s views, with, however, numer-
ous exceptions. It would seem that the more or less immediate environment
is more familiar than territory farther off, which probably accounts for the
frequently observable bursts of speed near it; the “lane” leading to it
being merely an accustomed route, more or less forced by habit and obstruc-
tions on the bottom.

Abbott and early writers assumed that the female was the nest guardian,
an idea dispelled by Gill (1889) and adequately discussed by him (1907).
These earlier papers are not referred to here, as being unimportant in the
light of later corrections and at best of interest only in a historical sense.
See Dean (1916) for many such references.

TABLE IV.

Nesting dates of Eupomotis gibbosus in the vicinity of New York City.

Locality

Year

Pre-nesting

Period

Nesting Period

Post-nesting

Period

Temp.

°C.

Branch Brook Park,
Newark, N. J.

1913

June 13

1926

—

June 6

—

20

1927

May 11

— -

—

Elmdale, N. J.

1926

—

May 281

Haskell, N. J.

1928

—

July 15-172

—

Hackettstown, N. J.

1933

—

July 23

—

1934

June 25

—

—

—

Llewellyn Lake, N. J.

1934

May 20

July 14-29

Aug. 12

25.5

Sprain Lake, N. Y.

1934

—

—

Aug. 4

29

Pines Lake, N. J.

1934

June 15

July 6

Aug. 10

29

Lake Mohawk, N. J.

1934

May 20

July 14-29

Aug. 12

25.5

Kensico Reservoir

1935

June 9

June 14

Aug. 15

27.5

By ram Lake

1935

May 25

May 28

July 6

21

Pines Lake

1935

May 30

June 9- Aug. 14

—

22.5

Range of dates

May 11-June 25

May 28- July 29

July 6- Aug. 12

20-29

1 Eggs hatched June 1.

2 Young fish 48 to 53 mm. standard length on Aug. 1.

Eupomotis microlophus (Gunther).

There appears to be no specific data on this species, but presumably
it would be found to be closely similar to its congener, if not identical.

Enneacanthus gloriosus (Holbrook).

The blue-spotted sunfish seems to have been peculiarly neglected by
aquarists, although it is undoubtedly an attractive aquarium fish. Sawyer
(1920) gives the only specific account we were able to locate of this species
in aquaria but apparently he knew nothing of its reproduction. Bade (1932),
in his general book, states that this species builds nests in aquaria. Inci-
dentally, the several other species he describes are in good agreement with
those given herein and are not referred to in each case. Fowler (1923)
describes the nest as a “miniature sunfish nest in the moss with a diameter
of four or five inches. In the bottom were rootlets to which the eggs
adhere.” Breder and Redmond (1929), working on the development of the
eggs, were unable to find any nests, using stripped eggs for this purpose.
This irregularity is similar to that of Mesogonistius and we now know that
they may nest in the conventional sunfish manner or in plants, as is more
fully discussed under that species. Nests usually in beds of filamentous
algae are to be found in the New Jersey pine barrens, according to W. C.
Bennett of the New York Aquarium (May 12, 1935). These nests are usually
among lily pads and are not excavated all the way through the algae. Usually

32

Zoologica: New York Zoological Society

[XXI :1

the nests are nearly a foot in diameter, which larger size probably has to do
with the softness of the material. The water depth is usually about one
foot. See also under Mesogonistius.

Although the immature fish are conspicuously barred, the breeding
females are more drab, as figured by Breder and Redmond (1929). The
males are relatively brilliant but the pattern differs only slightly. They
wrote : “There must be considerable rivalry among the males as is evidenced
by the somewhat ragged condition of their fins, principally the caudal,
during the mating season. In other local sunfishes much time is spent in
nest building and while there is considerable rivalry it seldom seems to
come to actual combat.”

Enneacanthus obesus (Girard).

The species Enneacanthus obesus (Girard) may be distinct but at this
writing there is certainly no reason to introduce this problem into a con-
sideration of reproductive habits. As indicated by Breder and Redmond
(1929) the fishes known to the author are E. gloriosus. Presumably the
habits of these two, if they are eventually established as separate, should
be closely similar.

Mesogonistius chaetodon (Baird).

The black-banded sunfish, confined to the strongly acid waters of the
pine barrens of southern New Jersey and straggling southward in similar
environments to Maryland, appearing again in the Carolinas, will apparently
not thrive in waters of different and less acid content. The waters of their
New Jersey habitat average in their acidity from about 4.0 to 5.0 pH. In
aquaria with a range of about 6.5 pH upward we have not found them to
thrive satisfactorily for any great length of time, which is not the case with
the generality of centrarchids, or indeed with specimens of Enneacanthus
taken in company with them from the same bog waters. The latter, however,
have a much larger range and seem to accommodate easily a wide pH range.

Because of the attractive markings of the black-banded sunfish, aqua-
rists have spent considerable effort in attempts to establish them in aquaria.
Little success has been met in this direction and although some attempts
at reproduction have been recorded we have yet to hear of a second genera-
tion of aquarium fish.

It was reported that this species built its nest in plants in an aquarium,
Price (1915). He wrote that he observed “the male in the bunch of plants,
about six inches from the bottom, busily engaged in making a hollow, some-
what like a bird’s nest.” Merget (1918), repeated by Werner (1930), de-
scribes a similar condition but mentions a previously made hole in the
gravel below. Holbein (1926) and Wright (1928) both report nests in
aquariums in the sand at the bottom, similar to other sunfish. The German
aquarists, generally so successful in such matters, seem to be unable to
add anything to this, Walk (1921), Henzelmann (1930) and Anonymous
(1933). Abbott (1883) could not decide if nests were regularly constructed
or not ; at least he could not find anything of which he was sure.

These apparently contradictory statements are readily understandable
in the light of field observations made by Mr. W. C. Bennett of the New
York Aquarium staff, who is familiar with the haunts of these fishes and
has collected them for a long time. In May, in the territory occupied by
this species, many small dish-shaped depressions may be found in sandy
spots in a foot of water or less. Some of these are occupied by Mesogonistius
and some by Enneacanthus. Close examination divulges similarly sized de-
pressions in beds of filamentous algae also occupied by both species and
less frequently coarser plants will be pushed apart and such a sunfish found
in the space so created. It is thus apparent that there is considerable varia-

1936]

Breder: Reproductive Habits of the Sunfishes

33

tion in the selection of nesting sites. These nests average about four inches
in diameter and are usually in about one foot of water. The Aquarium
records, then, would seem simply to represent the spread of variation in
habit to be expected.

There is little differentiation of the sexes, but Price (1915) reported
that the male “became more transparent and pale, while the female grew
more intense in color.” On the basis that most female sunfish intensify
their bands while the males tend to lose them at the time of breeding, this
is perhaps to be expected. Males of other species generally have bright
metallic spots that intensify, but since this species has no such basic fea-
tures it is not surprising that they merely look pale on losing their usually
strong vertical bars. Regarding courtship and spawning, Price (1915)
wrote that when the nest was completed to the satisfaction of the male, “he
darted swiftly toward the female. Before her he quivered and spread his
fins, then swam back to the nest, repeating the performance several times.
At last the female yielded to his pleading and followed him into the nest.
Both trembled and vibrated, the pair in such a position that I was reminded
of the movements of the wings of a butterfly.” This latter statement de-
scribes well the typical centrarchid spawning position. The eggs hatched
in two days (June 16 to 18). On the following day (June 19) they spawned
again and the eggs hatched on June 21. There were two subsequent spawn-
ings from this pair but the dates are not given. Merget (1918) mentions
that the male guards the nest, which, of course is to be expected.

Archoplites interruptus (Girard).

There appears to be absolutely no data available on the reproductive
habits of this species.

Ambloplites rupestris (Rafinesque) .

The actual details of spawning of this species in captivity are given
on previous pages as an illustration of the reproductive behavior of the
Centrarchidae. See also Pis. Ill and IV, Figs. 3-6, and Table V for data
on reproduction at the New York Aquarium. According to Evermann and
Clark (1920), this species spawns earlier than other centrarchids in the
region of their report, some nesting as early as May 15. The following
bottoms have been noted for nesting sites: soil, Wright and Allen (1913) ;
gravel, Bean (1903), Jordan and Evermann (1903), Tracy (1910), Hay
(1894), Smith (1907); swampy places, Bensley (1915); marl, Hankinson
(1908). In the tanks of the New York Aquarium this species has spawned
in a depth of four feet, but this was the only depth available to them. The
depth of water may range from a few inches, Bensley (1915), to about a
foot, Hankinson (1908). Evermann and Clark (1920) say that the nest is
usually placed beside a rock, stick or similar object. The diameter of the
nests in aquaria was about two feet but the fishes were very large, old
specimens. In the cases, observed the male guarded the nest and young
and was successful until the young fish became too adventurous. The be-
havior of the rock bass in a state of nature has been described by Bensley
(1915) and is quoted in the preceding general section.

TABLE V.

Data on the breeding of Ambloplites rupestris in the New York Aquarium1.

Date

Hatch

Temperature °C

Time of Day

July 11, 1933
July 15, 1933
July 17, 1933

July 15
July 19?
July 20?

21.0-20.5

20.5-20.5

20.5-20.5

Before 9.00 A.M.
After 3.00 P.M.

1 The female may have been the same, or different, each time.

34

Zoologica: New York Zoological Society

[XXI :1

Ambloplites cavifrons Cope.

If this species is valid it can be expected to perform in the manner
described for the more widely distributed A. rupestris.

Acantharchus pomotis (Baird).

The nest of the mud sunfish has been recorded by Fowler (1923), who
writes that it “has been observed nesting near Willow Grove Lake, in
holes in a cranberry bog not far from Newfield, N. J., by R. 0. Van
Duesen. At this time, June 1st, the nest was guarded by the male. It
was about a foot in diameter and resembled somewhat the nest of the
long-eared sunfish. The depth of the water was about a foot, the bottom
sand, mud around the periphery, and surrounded by spatter-docks. In loca-
tion the nest was near the shore and partly shaded by trees.” Breder and
Redmond (1929) searched for the nests of this species in northern New
Jersey but were unable to find nests or ripe adults, although numerous
specimens were handled. Very young specimens were found in schools in
the New Jersey pine barrens near Lakehurst about the middle of June by
Mr. W. C. Bennett of the New York Aquarium staff. He noted that near
these specimens a larger fish was hovering. Although he is extremely
familiar with the region he never found a nest of the species. Apparently,,
for some reason not yet evident, this sunfish which is much more secretive
than other members of the family is likewise more secretive with its nests.

Abbott (1884) considered this species to be nocturnal in habit, which
would certainly account for some of the preceding remarks. He also
stated that it “has a well developed voice” and that it produces “a deep
grunting sound that cannot be mistaken.” On capture many fishes may
produce such a sound by grating their pharyngeals together or by vibrating
their air bladder. If such sounds have any significance in centrarchid repro-
duction we have no evidence to that effect and the group should no doubt
be considered voiceless as Gill (1907) suggested for Eupomotis. If Acan-
tharchus is truly nocturnal it is possible that the voice noted by Abbott may
have a real significance in this little-studied species, which would mark it
as a distinct departure from the other and diurnal members.

Centrarchus macropterus (Lacepede).

Apparently no data has been published on the reproductive habits of
this species in a state of nature. A short account of its courtship and
spawning in aquaria is given by Heinrich (1921). His remarks, so far as
they pertain to the present study, indicate agreement with the generality
of sunfishes.

Pomoxis annularis Rafinesque.

Due to its frequent confusion with Pomoxis sparoides, the exact habits
of this species, the white crappie, are uncertain but it is likely that they
are closely similar to those of that fish. Forbes and Richardson (1909) say
it apparently spawns in May in Illinois. Anonymous (1919) describes
spawning presumably of this species, in the Aquarium of the U. S. Bureau of
Fisheries, as occurring during the night of May 25, 1926. “The eggs were
attached to a dense growth of algae covering stones in the obliquely in-
clined back of the tank and some of them were practically at the surface.
The male fish jealously guarded the eggs and kept the water about them
in constant motion with his pectoral fins. Other fish were kept away and
objects that came near the eggs were savagely bitten. If a person placed
his hand within six inches of the surface of the water, the male fish would
leap clear of the water and strike the hand viciously.”

1936]

Breder: Reproductive Habits of the Sunfishes

35

Pomoxis sparoides (Lacepede).

Spawning of the black crappie or calico bass occurs in spring from
early May, Richardson (1913), in Illinois, to early July, Evermann and
Clark (1920), in Indiana, and in May in South Dakota, Churchill and
Over (1933). Circular nests are constructed in shallow water of from ten
inches, Richardson (1913), to two feet, Pearse (1919). The diameter may
be about eight or nine inches, according to Evermann and Clark (1920).
Spawning occurs when the water reaches about 68°F., Pearse (1919), but
Richardson gives about 64°. The bottom selected varies considerably but
apparently tends to be one of sand or fine gravel. A nest is described by
Richardson (1913) as follows:

‘‘It was hollowed out under the leaves of a water-parsnip, and sur-
rounded by smartweed and bog rush ( Juncus ). Some of the eggs were
adhering to fine roots in the bottom of the nest, but most of them were on
the leaves of the water-parsnip, at a level of two to four inches above the
bottom of the nest. The nest was guarded by a male, six inches long, who
was so gentle that we could reach out a hand to within three feet of
him before he moved away.” Evermann and Clark (1920) state that the
nests are usually at least five or six feet apart on fine gravel, coarse sand,
with a few empty snail shells at times and are generally surrounded by
Chara.

Pearse (1919) describes and figures the nest of this species along

clay banks, here repro-
duced as Text-fig. 6.
Pearse examined the
stomach contents of
nine males on nests and
found them well filled
with mostly aquatic in-
sects. Many fish cease
or nearly cease feeding
during the spawning
season but the sunfishes
so far as known to the
author all feed readily
all during the reproduc-
tive period both in cap-
tivity and in a state of
nature. Indeed, in most
states, game laws de-
signed to protect the
black basses, Microp-
terus and Aplites , give
a closed season during
the breeding period to
prevent anglers from
catching these fishes
from off their nests.

Elassoma zonatum Jordan.

The reproductive habits of the pigmy sunfish are described in aquaria
by Poyser (1919). In his studies spawning took place at 71° and 72° F.
Unfortunately he did not see the actual nest construction about which he
wrote: “While I did not see the operation, the rubbish was cleared and
heaped about the periphery, but not with the nicety of Eupomotis gibbosus,
as much flocculent matter was allowed to remain.” Of another male he
wrote: “One of the sites was amid a dense growth of algae and nothing
could be observed other than that the male was constantly there; certainly
there was no attempt to form a nest. This instance leads me to believe

Text-figure 6.

The nest of Pomoxis sparoides under a
bank. (After Pearse, 1919).

clay

36

Zoologica: New York Zoological Society

[XXI :1

that under certain unfavorable bottom conditions no attempt may be made
to clear a space if indeed this is not the normal method.”

Courtship was described as follows: “This stage reached, the male at
various times was observed making obvious efforts to attract a female,
indulging in most amusing gyrations for such a ‘stiff and usually sedate
fish. During these plays the body assumed the most intense coloring. . . .
The dorsal was flabby but erect, waving with movement while the action of
the caudal was quickened. A particular and conspicuous feature of the play
was the rapid, rhythmic, alternate backward and forward ‘clicking’ of the
ventrals, a feature I have not noticed in any other mating fish.” This latter
item is indulged in, as a matter of note, by other centrarchids, Eupomotis,
Lepomis, Aplites, Ambloplites and probably to some extent by all, as well
as in the Cichlidae, where it is notably conspicuous under various conditions
as indicated by Breder (1934).

The spawning operation is described as a violent trembling after the
“female approached quite unostentiously and without the slightest hesita-
tion.” Unfortunately no mention is made as to whether the female reclined
on one side or not. This spawning took place immediately over the nest but
another took place eleven inches above it at the surface. Another male in-
truded into this mating and both attempted to guard the subsequent eggs
for a time. The spawnings occured at 9 :30 A.M. and 8 A.M. respectively.
The eggs are described as non-adhesive, but this seems unlikely, since a
figure of an early egg shows some “peculiar process” which he was “unable
to explain” and which would seem to be some exudate.

Barney and Anson (1920), studying the ecology of this species, found
them spawning at 65° F., which is a week earlier than Allotis humilis in the
same locality, as the former breed in shallower water which warms to the
same temperature that much earlier. The habits so far as described are in
good agreement with those observed in aquaria.

Elassoma evergladei Jordan.

Three aquarium descriptions of the spawning of this species, Rachow
(1926), Mayer (1929) and Anonymous (1931), all agree in considering this
species a non-nest builder. The eggs are described as attached to plants
and other objects.

The descriptions of courtship are closely similar to those of the pre-
ceding species and the male is described as being decidedly black. Spawn-
ing occurred many times at intervals of about two weeks, with relatively
few eggs laid at a time.

Discussion.

The data presented in the preceding pages clearly show the Centrar-
chidae to be extremely uniform in their methods of reproduction. This is
in harmony with their restriction to the more quiet fresh waters of a
single continent. Confined, as they are, to a single basic type of habitat
it is perhaps not surprising that a reproductive method not uncommon to
such localities is used by all of them. Other fishes using methods more or
less similar, in similar localities, include such diverse forms as many of
the Siluridae, the Cichlidae, Amia and Heterotis. While it is true that
various species associated with the Centrarchidae spawn in diverse man-
ners, it is also true that such fishes all agree in spawning at an earlier
time, when the water is cooler; e.g. Perea , Morone, Stizostedion, Esox,
Umbra , Notemigonus , Fundulus, Pomolobus , etc. Presumably the tempera-
ture thresholds on which the various fishes spawn is largely a matter of
physiological differences. Consequently those forms that spawn late, on a
higher temperature, are forced to provide for the greater oxygen demand
that the temperature itself induces. Thus, while silting is of little importance

1936]

Breder: Reproductive Habits of the Sunfishes

37

to slow-breathing eggs of Esox and Perea in much colder water, the same
amount of silting on sunfish eggs might well be suffocating. Also, at this
later season, the temperature-controlled invertebrate predators are more
active. As has already been indicated, sunfish eggs hatch well without
parental attention in clear water protected from enemies. The function of
the parental activity would thus seem clearly associated with the tempera-
ture at the time of spawning and the effects incident to it.

If an attempt is made to trace the possible phylogeny of nesting habits
in the Centrarchidae, there is little to work on. The closest relatives, the
Kuhliidae, are in part marine, living mostly in open water, and according
to Meek (1916) “appear to spawn inshore, in brackish or fresh water.”
Some species live in fresh water in rivers, but not enough is known of
their reproductive habits to be useful in this connection. Since such cen-
trarchid eggs as have been tested are demersal in even excessively dense
sea water, it matters little whether we consider for present purposes the
centrarchid or the kuhliid method as primitive, since neither gives any
associative inferences. Perhaps both modes lead back to a common ances-
tral one, as both families are generally considered as evolved from some
central percoid type. In all the great aggregation of Percoidei there seems
to be no form of reproduction that could be used as a likely starting point
of centrarchid habits. The Cichlidae, Nandidae, Etheostomidae, Labridae
and the Pomacentridae all seem somewhat to resemble the centrarchids in
reproductive habit but certainly they are each removed too far to be little
more than chance resemblances. All of the rest show no parental care, or
it is so different in nature as surely to be an entirely independent develop-
ment.

The serranid, Roccus lineatus (Bloch), lays slightly heavy non-adhesive
eggs while those of Morone americana (Gmelin) are heavy and very ad-
hesive. Both species are anadromous where the latter is not landlocked ; both
spawn earlier than the centrarchids and there is no parental care. It is
conceivable that some such habit might lead to nesting and guarding by a
physiological need for a higher temperature to induce gonadal stimulation.
The two forms, mentioned as non-nesters, resort to special places that are
suitable for the eggs to develop in, which of course is the first step to nest
construction. All this is little more than speculation and it will not be
pushed further, as the inferences are obvious.

Considered from the standpoint of what this habit might lead to, one
is confronted with the strong probability that the centrarchids are terminal
along their line of reproductive specialization. Since it has been shown that
various nest building groups that use their mouths to a considerable extent
in nest construction or in manipulating eggs have members which practice
the curious habit of oral incubation, it might be expected that the centrar-
chids would give rise to such a habit. Three diverse groups — Cichlidae,
Siluridae and Labyrinthidae — all show such behavior, Breder (1933, 1934a
and b, 1935a) and there are at least two other groups falling within the
same category on which there is too little data for profitable speculation.
One is the Osteoglossidae and the other the Apogonidae, a member of the
Percoidei. We cannot, however, attempt to associate the apogonids with
the centrarchids, as the former are all marine and show evidences of being
a secondary invasion of reef environments from deep water, which is re-
flected in their bright red coloration and large eyes. So far as known they
all practice oral incubation.

With these remarks on the present inability to derive centrarchid habits
from others, or others from them, this part of the discussion may be left
with the comment that our present knowledge of the exceedingly uniform
reproductive behavior of this family completely fails to tie it in with those
of any of its possible relatives.

Examining the slight differences within the family Centrarchidae there

38

Zoologica: New York Zoological Society

[XXI :1

are some inferences discernible which are at least fairly suggestive. The
genera Micropterus and Aplites which Hubbs (in litt.) suggests recognizing
as a subfamily, Micropterinae, spawn earlier than the other genera asso-
ciated with them. Since these fishes have departed less from the central
percoid type than the rest, it may be that this slightly earlier spawning is
an indication of a still earlier one in the ancestral forms, without the con-
struction of a nest. Nest construction, nevertheless, is as well developed
in this group as in the rest.

Those Lepominae for which Hubbs (in litt.) suggests the tribe Lepo-
mini, including Chaenobryttus, Apomotis, Sclerotis , Lepomis, Helioperca,
Xenotis, Lethogrammus and Eupomotis, so far as their habits are known,
are quite similar in regard to reproduction. Such differences as are some-
times seemingly apparent would seem to refer to mechanical circumstances
of fish and environment, as is indicated in the specific considerations. It is
in this group that extensive natural hybridization has been found, a condi-
tion in itself giving some measure of reproductive similarities. These fish
all follow the Micropterinae in time of spawning, in the same waters.

The dwarf lepominid sunfishes, tribe Enneacanthini, according to
Hubbs’ suggestions including only Enneacanthus and Mesogonistius, are
relatively late spawners, overlapping the time of Lepomini. The frequent
failure of these fish to construct a fully formed nest is apparently a tendency
to a secondary loss of the habit. It would be difficult to think of them as
only now developing the habit from a non -nesting ancestor, since it is so
typical of other sunfish nests when made.

The tribe Ambloplitini, including Archoplites, Ambloplites and Acan-
tharcus, of which the nests of only the second two are known, is again
typical. They spawn relatively early as compared with the other Lepominae,
apparently deriving their reproductive habits from some form similar to
the Micropterinae with little change.

The subfamily Centrarchinae, including Centrarchus and Pomoxis, is
too scantily known in regard to reproductive habit to warrant much specu-
lation, but apparently differences are slight. Spawning in Pomoxis is rela-
tively early and would seem to be sometimes in deeper water than in other
forms.

The pigmy sunfishes, generally considered a separate family, the Elas-
somidae, would seem to be a subtropical specialization. The single genus,
Elassoma, diminutive as it is, shows the typical sunfish reproductive habits.
In aquaria, at least, it may sometimes forego nest building in a manner
analogous to that of the dwarfish Enneacanthini.

The temperature differences as here discussed are indicated in Table
VI. So far as these figures are reliable they show the primitive Microp-
terini as earliest, with the Centrarchinae, Elassomidae and Ambloplitini
next. The Lepomini have the highest mean value but a much greater
spread than either the Enneacanthini or Elassomidae.

TABLE VI.

Comparison of the nesting temperatures of the centrarchids.

Group

Temperature °C.

Minimum

Maximum

Mean

Centrarchidae

Micropterinae

15.5

21

18.25

Lepominae

Lepomini

20

29

24.5

Enneacanthini

21

23

22

Ambloplitini

20.5

21

20.75

Centrarchinae

18

20

19

Elassomidae

18.5

22

20.25

1936]

Breder: Reproductive Habits of the Sunfishes

39

Sex recognition, mating, parental care and other factors are so similar
or imperfectly known as to render no clues to the phylogeny of habit. For
the present, then, it is necessary to leave the habits of this group without
any very evident connections with others. They stand alone, as a compact
assemblage, with a highly specialized reproductive habitus in which the
traces of heritage are not evident.

The general tendency for immature fish, females and frightened males,
to show a series of dark vertical bars may be tentatively explained as rep-
resenting the basic color pattern that is overridden in the males by the effect
of the sex harmones. Just what significance, if any, this may have is cer-
tainly not clear at this time. Known behavior gives no clue to this.

There is a marked tendency for the males of the Centrarchidae to
grow to a larger size than the females, as has been indicated by the fol-
lowing: Eupomotis gibbosus , Creaser (1926) ; Micropterus dolomieu , Tester
(1932) ; Eupomotis gibbosus and Helioperca macrochira, Hubbs and Hubbs
(1932) ; Xenotis megalotis, Apomotis cyanellus and Ambloplites rupestris ,
Hubbs and Cooper (1935). The latter suggest that this inversion of the
usual size difference between the sexes in fish may have to do with the
active protecting role that the males assume.

Considered in a broad way the annual cycle of habit in the Centrar-
chidae presents an interesting series of items of behavior largely controlled
by temperature.

Starting in the winter the fishes are found in a semihibernating state.
They may be aggregated in compact masses or variously scattered about
individually. Micropterus dolomieu, Townsend (1916) ; Lepomis auritus,
Acantharchus pomotis, Pomoxis sparoides, Breder and Nigrelli (1934), all
have been noted to form aggregations on temperature reduction. Eupomotis
gibbosus, Chaenobryttus gulosus, Ambloplites rupestris and Enneacanthus
gloriosus were noted by the latter authors not to aggregate under similar
conditions. The factors involved in causing aggregation in the species
studied are numerous and not entirely clear. Whether the differences noted
are genetic or purely environmental in the various species is not fully evi-
dent at this time and there is nothing we can add to the remarks of Breder
and Nigrelli (1934) and Langlois (1936). At least in no species is fighting
known below a certain temperature. Consequently as the fishes naturally
seek optimum locations they are more or less drawn together on a purely
mechanical basis, without considering any social impulses. This also tends
to segregate the smaller, last season’s, fishes from the older, since on a size
basis alone their needs are somewhat different.

As the temperature rises in the spring feeding is the first evident reac-
tion. In this connection it may be mentioned that sunfishes feed continu-
ously until the cold weather of fall drives them into hibernation again. This
is somewhat unusual among fishes, since feeding is frequently interrupted
by breeding activities in many forms.

Shortly after feeding has well commenced, concomitantly with the devel-
oping gonads in the males, shallow water is sought out for nesting purposes.
That this is not purely a temperature factor encouraged by a seeking of
the warmer shallow marginal water is evident from the fact that the males
precede the females by an irregular but frequently not inconsiderable time.
It would seem that the increasing temperature and longer hours of day-
light (?) have a more immediate effect on the males than on the females.
Daylight is specifically mentioned in this connection partly because of the
demonstrated role it plays in other vertebrates, including fishes, Breder and
Coates (1935b), and because of the obvious importance sunlight plays in the
reproductive habits of sunfish, as already discussed. The males begin their
nest construction even in the absence of females, this activity being stimu-
lated by internal gonadal influences. At this time the males become hostile
toward one another and tend to scatter. Limitations of suitable bottom
somewhat counter this, causing colonies of nests frequently to be formed.

TABLE VII.

Comparison of the reproductive habits of the Centrarchidae with those of
the silurid Ameir us nebulo sus and the cichlid Aequidens latifrons.1

40

Zoologica: New York Zoological Society

[XXI :1

po-

CO

bJC

be

aera-

G

be

G

G

be

O

a>

<4-1

|

b£

S3

a> -
'“jj 4-3

O

>s

.5

*S

si

G

.2

G jj
o xn

o

a>

u

G

-4->

«H

O

0) c

° CO

co G
<D a>
1 — 1 S-i

II

TO 3

^ O
CO S

be o

O

a>

$4

P..-E

C 0 to

O O)

hJ G

a

Z

<P o
12 *43

Ph q.

O

£ g

be-£

w.E

G

O

CO

A*

UO

CD

00

as

1 Data on Ameiurus from Breder (1935a) and on Aequidens from Breder (1934a).

1936]

Breder: Reproductive Habits of the Sunfishes

41

The still more or less aggregated females begin to cruise about, even-
tually entering the nests and spawning. Their attention to family life is
only momentary and their aggregation usually does not fully or per-
manently disintegrate. The courtship activities of the male, which seem to
be indistinguishable from the fighting attitude, may be exhibited to other
fish or inert objects. The reaction of other males is generally one of flight,
rarely fighting, while that of females is one of attraction, the resulting
behavior thus being dependent on the activity of the visiting fish. Parental
care generally extends only to the time of hatching but there is some control
of the fry, especially in the Micropterinae, which is much like that of the
cichlids.

In the absence of suitable bottom nests may be constructed in which
reproduction is impossible, including sites with too rapidly flowing water or
mud that is too soft and flocculent. On a really hard bottom, such as a
cement floor, no nest is formed but the motions are made and the spawning
may be successful. The mechanics of nest making seem to be more or less
constant, the substratum determining the features of the nest. The in-
fluences determining the selection of a given spot seem to include chiefly
the existence of a prior depression, the proximity to a protecting object,
such as a rock, and the distance from the nearest other male. The same
general areas are used year after year in most cases, suggesting a certain
amount of habit formation. This would not be surprising when it is con-
sidered how common such a habit is in other fishes, becoming spectacular
only in forms undertaking long migrations. The movement inshore from
deep water can hardly be considered a migration in the ordinary sense of
the word but it is undoubtedly of the same nature.

After the reproductive period has passed, aggregating is again some-
times evident, but solitary feeding is the principal business and aggrega-
tions if appearing at all do not show up strongly until the winter ap-
proaches. This statement also applies to the young of the year after they
leave the nest.

A comparison of the features of the reproductive habits of two un-
related species with the sunfishes is given in Table VII. The data on the
silurid, Ameiurus nebulosus (Le Sueur), is modified from Breder (1935a)
and that on the cichlid, Aequidens latifrons (Steindachner), Breder (1934a).
Three of the nine items listed in the table — 1, 5 and 7 — may serve to de-
scribe broadly both the centrarchids and the cichlid, whereas none serves
mutually for the silurid and either of the others.

This may be interpreted in one of two ways. Since the centrarchids
and cichlids are both Percoidei it may be that the genetic relationship,
remote as it is, is expressing itself in these general resemblances. On the
other hand it may be that these are really parallelisms. In this connection
it may be recalled that the cichlids more or less replace the centrarchids
in Central and South America, or vice versa, considered in a general ecolog-
ical sense. The much more distantly related silurids also adopting nesting
habits have all the details differently arranged. Whether these differences
can be considered as of genetic difference of heritage or whether they are
of mechanical circumstance is not easily settled. The silurids are basically
olfactory types, whereas the percoids here considered are distinctly visual
types. How much of the difference can be associated with the predominance
of one receptor or another and its resulting influence on effectors can only
be speculated upon at present.

Summary.

1. The reproductive habits of the Centrarchidae are remarkably uni-
form.

42

Zoologica: New York Zoological Society

[XXI :1

2. In all known, the male constructs the nest and guards the eggs, the
female being only concerned with their deposition. In a few, nest con-
struction may sometimes be omitted.

3. Sex recognition is apparently based entirely on the differential
behavior, from all other fish, of females ready to spawn.

4. The position and form of the nest are controlled by a large number
of purely physical factors in the environment and include: temperature,
sunshine, depth of water, rate of flow, nature of bottom, proximity of pro-
tecting objects, etc.

5. The position of the nest in regard to other nests is controlled by a
centrifugal effect between the males and a centripetal effect induced by limi-
tations of suitable bottom and probably as well by some form of population
hierarchy.

6. Already-existant nests of other species are frequently made use of.

7. In all species studied the spawning positions are closely similar, the
male remaining in an upright position and the female reclining to one side.

8. Success in defense of the nest is apparently based more on the
fact of the owner being on familiar territory, rather than on any intrinsic
fighting ability.

9. Well made nests are marked in the Micropterinae, Lepomini and
Centrarchinae and least in the Enneacanthini, Ambloplitini and Elas-
somidae.

10. The extent of parental care follows the above, being most promi-
nent in the primitive Micropterinae.

11. The development of the centrarchid type of reproduction is not
clear phylogenetically, due in part to the relatively isolated position of the
group and to the lack of information on the nearest relatives.

Bibliography.

Abbott, C. C.

1883. On the habits of certain sunfish. Amer. Naturalist, 17: 1254-1255.

1884. A Naturalist’s Rambles About Home. New York, 1-485.

Adams, C. C. and Hankinson, T. L.

1928. The Ecology and Economics of Oneida Lake Fish. Roosevelt Wild
Life Annals, Bull. N. Y. State College Forestry, 1 (4a) : 236-548.

Anonymous

1919. Crappie Spawn in Washington Aquarium. Aquatic Life, 4 (11) : 137.
1931. Pigmy Sunfish. The Home Aquarium Bull., 1 (6) : 4-5.

1933. Der Scheidenbarsch. Das Aquarium, April: 67-68.

Bade, C.

[1932.] Das Susswasser-aquarium. 1-1212. F. Pfenningstorff, Berlin.
Bangham, R. V.

1927. Life history of bass cestode, Proteocephalus ambloplitis. Trans. Am.
Fish. Soc. 57: 206-208.

Barney, R. L. and Anson, B. J.

1920. Life history and ecology of the pigmy sunfish, Elassoma zonatum.
Ecology, 1 (4) : 241-256.

1923. Life history and ecology of the orange-spotted sunfish Lepomis
humilis. Appendix XV, Rept. U. S. Comm. Fish. (1922) : 1-16.

Bean, T. H.

1903. Catalogue of The Fishes of New York. Bull. N. Y. State Museum
(60) : 1-784.

1936]

Breder: Reproductive Habits of the Sunfishes

43

Becker, K.

1908. Der gemeine Sonnenfisch [Eupomotis aureus] Seine Brutpflege im
freien Wasser, sowie seine Haltung und Pflege im Zimmer aquarium.
Blatt. Aquar.-Terrar. Kunds. 19: 672.

Beeman, H. W.

1924. Habits and propagation of small mouth black bass. Trans. Amer.
Fish. Soc., 54: 92-107.

Bensley, B. A.

1915. The Fishes of Georgian Bay. Contributions to Canadian Biology.
Suppl. 47th Ann. Rept., Dept. Marine and Fishes. Fishing Branch:
1-51.

Breder, C. M., Jr.

1926. Sunfish Nests. Bull. N. Y. Z. S., 29 (6) : 179-180.

1928. Report of the Director of the Aquarium. 32nd Ann. Rept. of the
New York Zool. Soc. (1927) : 95-111.

1933. On the genesis of oral incubation in fishes. Anat. Rec., 57 (4) :
62 [Abstract 80].

1934a. An experimental study of the reproductive habits and life history of
the cichlid fish Aequidens latifrons (Steindachner) . Zoologica, 18 (1) :
1-42.

1934b. The reproductive habits of the painted Betta, a relative of the
Siamese Fighting Fish, new to aquaria. Bull. N. Y. Zool. Soc., 37
(5): 126-133.

1935a. The reproductive habits of the common catfish Ameirus nebulosus
(Le Sueur) with a discussion of their significance in ontogeny and
phylogeny. Zoologica, 19 (4) : 143-185.

1935b. The physical limitations of centrarchid nest construction. Anat.
Rec. 64 (1): 118. [Abstract 206].

1935c. Selection of nesting sites with intraspecific and extraspecific interac-
tions in nesting Centrarchidae. Anat. Rec. 64 (1) : 118 [Abstract 207].

Breder, C. M., Jr. and Coates, C. W.

1935a. Sex recognition in the guppy, Lebistes reticulatus Peters. Zoologica,
19 (5): 187-207.

1935b. Post-asymptotic studies in populations of Lebistes. Anat. Rec. 64
(1): 119. [Abstract 209].

Breder, C. M., Jr. and Nigrelli, R. F.

1934. The influence of temperature and other factors on the winter aggrega-
tions of the sunfish Lepomis auritus with critical remarks on the
social behavior of fishes. Ecology, 16 (1) : 33-47.

Breder, C. M., Jr. and Redmond, A. C.

1929. The blue-spotted sunfish; a contribution to the life history and habits
of Enneacanthus with notes on other Lepominae. Zoologica, 9 (10) :
379-401.

Cheney, A. N.

1897. Black bass and their distribution in the waters of the State of New
York. 2nd Ann. Rept. Comm. Fisheries, Game and Forests of New
York: 176-184.

Churchill, E. P. and Over, W. H.

1933. Fishes of South Dakota. The South Dakota Dept. Game and Fishes:
1-87.

Coates, C. W.

1933. Behavior of a pair of leaf fish, Monocirrhus polyacanthus Heckel.
Bull. N. Y. Zool. Soc. 36 (3) : 68-71.

44

Zoologica: New York Zoological Society

[XXI :1

COGGESHALL, L. T.

1923. A study of the productivity and breeding habits of the blue gill
Lepomis pallidus Mitchill [Abstract] Anat. Rec. 26: 382-383.

1924. A study of the productivity and breeding habits of the blue gill
Lepomis pallidus (Mitch.) Proc. Indiana Acad. Sci., 33: 315-320.

Creaser, C. W.

1926. The structure and growth of the scales of fishes in relation to the
interpretation of their life-history, with special reference to the
sunfish Eupomotis gibbosus. Misc. Pub. Mus. Zool. Univ. Mich. 17:
1-82.

Dean, B.

1916. A Bibliography of Fishes. Amer. Mus. Nat. Hist., New York, 3 Vols.

Embody, G. C.

1915. The Farm Fishpond. Cornel Reading Courses. Country Life Series
(3) : 213-252.

Evans, L. T.

1936 Social behavior of the normal and castrated lizard Anolis carolinensis.
Science, 83 (2144) : 104.

Evermann, B. W. and Clark, H. W.

1920. Lake Maxinkuckee, a Physical and Biological Survey. Dept, of
Conserv., Indiana. 2 Vols.

Flower, S. S.

1925. Contributions to our knowledge of the duration of life in vertebrate
animals — I, Fishes. Proc. Zool. Soc., London: 247-289.

1935. Further notes on the duration of life in animals — I, Fishes: as deter-
mined by otolith and scale readings and direct observations on living
individuals. Proc. Zool. Soc., London: 265-304.

Forbes, S. A. and Richardson, R. E.

1909. The Fishes of Illinois. Nat. Hist. Surv. of 111., 3: 1-357.

Fowler, H. W.

1923. Spawning habits of sunfishes, basses, etc. Fish Culturist, 2: 226-228.

Franklin, D.

1914. Note on a nesting sunfish. Copeia (11) : 1.

Gill, T. N.

1889. The “Hatchery” of the sun-fish. Nature, 1889, 40: 319.

1907. Parental Care Among Fresh-Water Fishes. Rept. Smithsonian Insti-
tution for 1905 (1688) : 403-581.

Greeley, J. R.

1934. Fishes of the Raquette Watershed with annotated list Supp. to 23rd
Ann. Rept. New York Conserv. Dept. (1933) 8 (2) : 53-108.

Hankinson, T. L.

1908. A Biological Survey of Walnut Lake, Michigan. Mich. State Board
Geol. Sur. Rept. for 1907 (1908) : 157-288.

1909. Field Problems on Stream Fishes for Secondary Classes. School Sci.
and Math. 9: 234-240.

Hay, O. P.

1894. The Lampreys and Fishes of Indiana, 19th Ann. Report. Dept. Geo. and
Nat. Resources. State of Indiana: 146-296.

Heinrich, O.

1921. Centrarchus macropterus, seine Zucht und Pflege im Zummer und
Frectundhechen Wochenschr. Aquar.-und Terrarienkunde, 18 (7) :

109-110.

1936]

Breder: Reproductive Habits of the Sun fishes

45

Henzelmann, E.

1930. Winke zur erfolgreichen Zucht des Scheibenbarsches. Das Aquarium
1930 (May) : 69-72. 1 fig.

Holbein, W. G.

1926. Breeding Black Banded Sunfish. Aquatic Life, 10 (2) : 17-18 and 30.

Howland, J. W.

1931. Studies on the Kentucky black bass ( Micropterus pseudaplites Hubbs) .
Trans. Amer. Fish. Soc. 61 : 89-94.

1932. The spotted or Kentucky black bass in Ohio. Ohio Bull, Bur. Sci.
Research Div. Cons., 1 (4) : 1-19.

Hubbs, C. L.

1919. The nesting habits of certain sunfishes as observed in a park lagoon in
Chicago. Aquatic Life, 4 (11) : 143-144.

1920. The hybridization of sunfish species in Nature. Aquatic Life 5:
101-103.

Hubbs, C. L. and Cooper, G. P.

1935. Age and Growth of the long-eared and the Green sunfishes in Mich-
igan. Pap. Mich. Acad. Sci. Arts and Letters. 20: 669-696.

Hubbs, C. L. and Hubbs, L. C.

1931. Increased growth in hybrid sunfishes. Pap. Mich. Acad. Sci. Arts and
Letters, 13: 291-301.

1932. Experimental verification of natural hybridization between distinct
genera of sunfishes. Ibid 15: 427-437.

1933. The increased growth, predominant maleness, and apparent infertility
of hybrid sunfishes. Ibid, 17: 613-641.

James, M. C.

1930. Spawning reactions of small mouthed bass. Trans. Amer. Fish Soc.,
60: 62-63.

Jordan, D. S. and Evermann, B. W.

1903. American Food and Game Fishes. 1-572. Doubleday Page & Co.

Krecker, F. H.

1916. Sunfish nests of Beimiller’s Cove. Contributions Zool. and Entomol.,
45: 125-134.

Lambkin, J. B.

1901. The spawning habits of the large mouth black bass in the South.
Trans. Amer. Fish. Soc., 1900 (1901) : 129-131.

Langlois, T. H.

1932. Problems of pond fish culture. Trans. Amer. Fish. Soc., 62: 156-166.

1934. The social behavior of bass in rearing ponds. Trans. Amer. Fish. Soc.,
164: 146-150.

1936. Survival value of aggregational behavior of bass under adverse con-
ditions. Ecology. 17 (1) : 177-178.

Leathers, A. L.

1911. A biological survey of the sand dune region on the south shore Sag-
inaw Bay, Michigan. Michigan Geo. and Biol. Surv. Pub. 4, Biol.
Series 2: 243-255.

Lydell, D.

1904. The habits and culture of the black bass. U. S. Fish. Comm. Bull.,
22 (1902) : 39-44.

1926. Small-mouthed black bass propagation. Trans. Amer. Fish. Soc. 56:
43-46.

Mayer, F.

1929. Elassoma evergladei Jordan. Aquatic Life, 13 (3) : 50-51 and 55.

46

Zoologica: New York Zoological Society

[XXI :1

Meek, A.

1916. The Migrations of Fish. 1-427. E. Arnold, London.

Mellen, I. M.

1919. Complete list of exhibits in the New York Aquarium. Appendix to
Rept. N. Y. Zool. Soc. for 1918: 35.

Merget, J.

1918. Chaetodons, their breeding. Aquar. Bull., Brooklyn Aquar. Soc., May:
128 and 132.

Moore, E.

1925. The inhibition of the spawning function of the small mouth black bass
by a parasitic flatworm. Trans. Amer. Fish. Soc., 55: 33-36.

1926. Further observations of the bass flatworm — Proteocephalus amblop-
litis. Trans. Amer. Fish. Soc., 56: 91-94.

Myers, G. S.

1921. The Common Sunfish. Aquatic life, 6 (2) : 20-21.

Nash, C. W.

1908. Check list of the vertebrates of Ontario and catalogue of specimens in
the biological section of the Provincial Museum: Fishes, Dept. Educa-
tion, Toronto: 1-122.

Nichols, J. T.

1918. The Fishes of the Vicinity of New York City. Amer. Mus. Nat. Hist.
Handbook Series. (7) : 1-118.

Noble, G. K.

1934. Sex recognition in the sunfish, Eupomotis gibbosus (Linne). Copeia
(4) : 151-155.

Noble, G. K. and Curtis, B.

1935. Sexual selection in fishes. Anat. Rec. 64 (1) : 84 [Abstract 126].

Pearse, A. S.

1919. Habits of the Black Crappie in Inland Lakes of Wisconsin. Appendix
III. Rept. U. S. Comm. Fisheries 1918: 1-16.

POYSER, W. A.

1919. Notes on the breeding habits of the pigmy sunfish. Aquatic Life, 4
(6) : 65-69.

Price, G. W.

1915. The Black-banded Sunfish. Aquatic Life, 1 (4) : 45-46.

Rachow, A.

1926. Elassoma evergladei Jordan. The Dwarf Perch. Aquatic Life, 10 (7) :
103-105.

Reid, H.

1930. A study of Eupomotis gibbosus (Linnaeus), as occurring in the Cham-
coak Lakes, New Brunswick. Contr. Canadia Bio. & Fish, 5 (16) :
457-466.

Reighard, J.

1902. The breeding habits of certain fishes. Science, n. s. 15 : 380-381.

Reighard, J. E.

1906. The breeding habits, development and propagation of the black bass.
16th Bien. Rept. Mich. State Board Fish Comm, (append.) : 1-73.

Richardson, R. E.

1913a. Observations on the breeding of the European Carp in the vicinity of
Havana, Illinois. Bull. 111. State Lab. Nat. Hist., 9: 387-404.

1913b. Observations on the breeding habits of fishes at Havana, 111., 1910 and
1911. Bull. 111. State Lab. Nat. Hist. 9, 1910-1913 (8) : 405-416.

1936]

Breder: Reproductive Habits of the Sunfishes

47

Sawyer, C. J.

1920. The Blue-Spotted Sunfish. Aquatic Life, 5: 65-66.

Smith, H. M.

1907. The Fishes of North Carolina. N. C. Geol. & Econ. Survey, 2: 1-453.
Tester, A. L.

1930. Spawning habits of the small-mouthed black bass in Ontario waters.
Trans. Amer. Fisheries Soc. 60: 53-61.

Townsend, C. H.

1916. Hibernating Fishes. Bull. N. Y. Zool. Soc., 19 (2) : 1345.

Tracy, H. C.

1910. Annotated list of the fishes known to inhabit the waters of Rhode
Island. 40th Ann. Rept. Comm, of Inland Waters of Rhode Island:
35-176.

Viosca, P., Jr.

1931. The Southern Small-Mouth Black Bass, Micropterus pseudaplites
Hubbs. Trans. Amer. Fish. Soc., 61: 95-98.

Walk. K. von

1921. Der Scheibenbarsch ( Mesogonistius chaetodon) . Wochenschrift Aquar.
und Terrarienkunde, 18 (4) : 53-54.

Wiebe, A. H.

1935. The Pond Culture of Black Bass. Bull. Game, Fish and Oyster Comm.,
Texas, Bull. (8) : 1-58.

Werner, B.

1930. Chaetodons, their breeding. Aquatic Life, 14 (4) : 86.

Wright, L. F.

1928. Mesogonistius chaetodon. Aquatic Life, 12 (6) : 110-111.

Wright, R.

1892. Preliminary report on the fish and fisheries of Ontario. Ontario
Game and Fish Comm. Rept. for 1892: 419-483.

Wright, A. H. and Allen, A. A.

1913. The Fauna of Ithaca, N. Y. : Fishes. Zoology Field Notebook, 4-6.
Ithaca, N. Y.

48

Zoologica: New York Zoological Society

[XXI :1

Fig. 1.

Fig. 2.

Fig. 3.
Fig. 4.

Fig. 5.
Fig. 6.

Fig. 7.
Fig. 8.

Fig. 9.
Fig. 10.

Fig. 11.
Fig. 12.

EXPLANATION OF THE PLATES.

Plate I.

Part of a typical colony of Eupomotis gibbosus nests on rough bottom at
Pines Lake, New Jersey. All but one nest was occupied at the moment
of exposure and a female is approaching the nest in the upper left-hand
corner.

Plate II.

A male Eupomotis gibbosus on its nest at night in the colony illustrated
in Fig. 1. Photographed by flashlight.

Plate III.

Ambloplites rupestris nesting on a concrete aquarium floor. The stones
are embedded in the cement. The positions are typical of those imme-
diately before spawning. The approaching female is at the right.

The spawning position of Ambloplites rupestris, showing the female re-
clined on her side. This photograph was taken shortly after that of
Fig. 3.

Plate IV.

A guarding male Ambloplites rupestris in a typical warning attitude.
Some of the eggs may be detected adherent to the stones under the head
of the fish. Photographed the day following that of PI. Ill, Figs. 3 and 4.

Ambloplites rupestris with their nest constructed in fine sand in a small
aquarium. The upper fish is the female.

Plate V.

Eupomotis gibbosus fanning out its nest. This is, perhaps, the most
typical position and posture of the nest building activity. This and the
following four photographs are of the same fish.

Eupomotis gibbosus in its typical “yawning” posture while guarding the
nest.

Plate VI.

Eupomotis gibbosus with a typical nest built in sand in an aquarium.
Note that a few empty snail shells have been left in place. To these
the eggs were attached.

Eupomotis gibbosus picking over the smooth sand of its nest.

Plate VII.

Eupomotis gibbosus during the spawning act. Note the prominent bars
on the female, which is in front of the male.

A typical solitary nest of Eupomotis gibbosus in fine sand. Note the
position in relation to the rocks as compared with those of the colony
shown in PI. I, Fig. 1, Pines Lake, New Jersey.

THE REPRODUCTIVE HABITS OF THE NORTH AMERICAN

BREDER.

PLATE I.

THE REPRODUCTIVE HABITS OF THE NORTH AMERICAN

SUNFISHES, FAMILY CENTRARCHIDAE.

PLATE III

[

BREDER.

FIG. 4.

THE REPRODUCTIVE HABITS OF THE NORTH AMERICAN
SUNFISHES, FAMILY CENTRARCHIDAE.

mm

BREDER.

PLATE IV.

FIG. 6.

THE REPRODUCTIVE HABITS OF THE NORTH AMERICAN
SUNFISHES, FAMILY CENTRARCHIDAE.

BREDER.

PLATE V.

FIG. 8.

THE REPRODUCTIVE HABITS OF THE NORTH AMERICAN
SUNFISHES, FAMILY CENTRARCHI DAE.

BREDER.

PLATE VI

FIG. 9.

FIG. 10.

THE REPRODUCTIVE HABITS OF THE NORTH AMERICAN
SUNFISHES, FAMILY CENTRARCHI DAE.

PLATE VII

FIG. It.

FIG. 12.

THE REPRODUCTIVE HABITS OF THE NORTH AMERICAN
SUNFISHES, FAMILY CENTRARCHI DAE.

1936]

Treadwell: Polychaetous Annelids, Nonsuch Island

49

2.

Polychaetous Annelids from the Vicinity of Nonsuch Island,
Bermuda1. By A. L. Treadwell, Vassar College.

(Plates Mil)

Introduction,

The following is a taxonomic account of some polychaetous annelids
collected in the vicinity of Nonsuch Island, Bermuda, in the years 1929'
1931, in connection with the work of the Department of Tropical Research
of the New York Zoological Society and submitted to me for study by Dr.
William Beebe. The greater number of individuals are pelagic species col-
lected during the deep-sea operations but a few are from shallow water and
shore localities, some of the latter collected by the writer while a guest at the
laboratory in August, 1931. For data relating to the Bermuda locality and
to the nets in which the pelagic species were taken, see Zoologica, Vol. XIII,
Nos. 1, 2 and 3.

A list of families represented and of new and old species occuring
in the collection follows:

Family Old species New species

Syllidae 1

Amphinomidae 2

Polynoidae 2 2

Sigalionidae 1

Chrysopetalidae 1

Glyceridae ? ?

Aricidae 1

Nereidae 3

Leodicidae 10

Tomopteridae 1

Alciopidae 1

Phyllodocidae 3

Opheliidae 2

Typhloscolecidae . 1

Cirratulidae 2

Sabellidae 1 1

27 8

Two papers dealing solely with the polychaetous annelids of Bermuda
have appeared, those of Webster (1884) and Verrill (1900), Webster’s was
a report on collections made by G. Brown Goode, while Verrill did extensive
collecting and study on the islands. Some annelids are described from
Bermuda localities in McIntosh’s (1885) Report on the annelids of the
Challenger Expedition and in a paper by Hoagland (1919). Bermuda was
grouped with the West Indian region in a monograph by the present writer

1 Contribution No. 481, Department of Tropical Research, New York Zoological Society.

50

Zoologica: New York Zoological Society

[XXI :2

(Treadwell, 1921) and while the main object of that paper was a study of
the Leodicidae, collections made in Bermuda in 1916 demonstrated that
in its essential features the annelid fauna of Bermuda is of the West
Indian type. This is noticeably the case in the Leodicidae, with the excep-
tion that the Atlantic “palolo,” Leodice fucata Ehlers, which is very abund-
ant farther south, has never been seen in Bermuda. A considerable number
of species listed by Ehlers (1887) occur in Bermuda and this monograph
should be available to any one studying the annelids of that locality.

Systematic Account.

Family Syllidae.

Typosyllis Langerhans.

Typosyllis corallicola Verrill.

Typosyllis corallicola Verrill, 1900, p. 603.

One specimen collected in tidepool, Nonsuch Island, Aug. 16, 1931
(No. 311,355).

Trypanosyllis Claparede.

Tetraglene phase.

With large numbers of Polyophthalmus (see p. 61, Nos. 311, 487 and 311-,
436), there were collected under electric light at the Nonsuch Island boat
landing a few individuals that I have identified as the tetraglene phase* of
an unknown species of Trypanosyllis. They average a length of 5 mm. and
a width of not more than 0.5 mm. The body is flattened and its most
characteristic features are the relatively enormous brown eyes and the
brown pigment patches on both dorsal and ventral surfaces near the bases
of the parapodia. There are no tentacular cirri or antennae. All cirri are
jointed, almost moniliform in character, the dorsal ones longer than the
body width, the ventral ones hardly longer than to the ends of the para-
podia. The anal cirri are from one-third to one-half as long as the body
and very large, each at its base being wider than half the width of the
pygidium. Notopodial setae are numerous, slender capillary in form and
are longer than the transverse diameter of the body. Neuropodial setae
are compound, the terminal joint varying in size, some being hardly longer
than their own basal width, others four times as long as this. Behind the
apical tooth is a small subapical one.

Family Amphinomidae.

Hermodice Kinberg.

Hermodice carunculata Kinberg.

Hermodice carunculata Kinberg, 1857, p. 14.

Collected in Castle Harbor, April 30, 1930 (No. 30,672) and Nonsuch
Island.

Eurythoe Kinberg.

Eurythoe pacifica Kinberg.

Eurythoe pacifica Kinberg, 1857, p. 14.

Collected in tidepool and coral rock, Oct. 22, 1930 (No. 301,360) ;
Sept. 12, 1930 (No. 301,675).

1936]

Treadwell: Polychaetous Annelids, Nonsuch Island

51

Family Polynoidae.

Polynoe Savigny.

Polynoe granulata Ehlers.

Polynoe granulata Ehlers, 1887, pp. 50-51; PI. 11, Figs. 2-7.

Collected in coral, Gurnet’s Rock, Bermuda, 35 feet deep, Sept. 29, 1930
(No. 301,308) ; and from reef in Castle Harbor, Aug. 14, 1931 (No. 311,298).

Harmothoe Kinberg.

Harmothoe fragment sp. ?

Collected in Net 928, 500 fathoms, Sept. 20, 1930 (No. 301,128).

Eunoe Malmgren.

Eunoe purpurea n. sp.

(Figs. 1—6).

The body of the type specimen is 18 mm. long, its greatest width 5 mm.
and it has 28 somites. The coloration varies somewhat in the different speci-
mens, the characteristic color and the one on which the specific name is
based being a purplish brown which is most pronounced on the anterior face
of the prostomium, the bases of the palps and the base of the proboscis. In
none is the proboscis more than very slightly protruded and it is not clear
how much of it shows this coloration. In one specimen, not much colored
elsewhere, the ventral surface, except for the mid-line, is dusted with this
pigment. In others the anterior and posterior thirds of the dorsal surface
are similarly dusted but in all cases the median dorsal third of the body
surface is uncolored. In one, the mid-ventral longitudinal line has a pearly
lustre and in another this is uncolored but there is on either side a grayish
green narrow band.

The width of the prostomium is roughly twice that of its length and all
angles are rounded (Fig. 1). The large heavily pigmented eyes occupy the
outer four corners of the prostomium, their colorless lenses being pro-
jected downward so as not to be visible from the dorsal surface. The peak on
either side is extended into a slender cirrus-like process and the anterior
margin is deeply incised for the insertion of the cirrophore of the median
tentacle, which is short and globular and has the appearance of a goblet
supported on a slender stalk. The style of the median tentacle is lost from
all specimens. The cirrophores of the lateral tentacles lie ventral to the
peaks. Their length is about equal to their width and they extend to about
the same distance as the median one. Their styles are short and inconspic-
uous. The palps are very long and heavy, fully eight times as long as the
prostomium. At their bases they are often pigmented, this pigment showing
a tendency to arrange itself in fine transverse lines. No dorsal cirri remain
and the tentacular cirri are too much mutilated for an accurate description,
but they are evidently much longer than the lateral tentacles. The ventral
cirri on the first two or three parapodia are very large, extending beyond
the seta tips, but in later parapodia they become progressively smaller. In
Fig. 1 the one shown at the left is a broken tentacular cirrus and the one on
the right is the first ventral cirrus.

The parapodia (Fig. 2, of the 16th), have a large neuropodium and a
very small notopodium, each having a large acicula which extends through
and beyond the tip of the lobe. In this parapodium (the 16th) , the ventral
cirrus extends considerably beyond the neuropodial tip. In the notopodium
there are a very few heavy setae (Fig. 3), with some much smaller. In the

52

Zoologica: New York Zoological Society

[XXI :2

parapodium figured the numbers were 2 and 3 respectively. The neuropodial
setae are of two kinds. The dorsalmost are the more slender and nearly
straight and carry two rows of sharp spines (Fig. 4). The ventralmost
ones vary in size, the smallest being near the ventral margin and in pro-
portion as they increase in size they acquire a deeper yellowish color. All
broaden toward the ends and then taper to a slender and slightly curved
apex. Along their broadened regions they carry transverse rows of toothed
plates. The heavy dorsal setae also have transverse plates but these are
very short and their margins only faintly toothed. (Fig. 5). Note the dif-
ference in the scales of magnification between Figs. 3 and 5.

Of the entire collection, only one elytron remains, the 1st or 2nd. It is
circular in outline and its surface is densely studded with sharp spines
(Fig. 6). These are larger near the middle of the elytron than around the
margins.

The type was collected in Net 838, 600 fathoms, Sept. 3, 1930 (No.
30,694), and is in the collections of the Department of Tropical Research of
the New York Zoological Society. Other specimens were taken in the fol-
lowing nets: Net 847, 500 fathoms, Sept. 4, 1930; Net 868, 900 fathoms,
Sept. 10, 1930, (No. 30,838) ; Net 902, 700 fathoms, Sept. 17, 1930 (No.
301,018) ; Net 942, 1,000 fathoms, Sept. 24, 1930 (No. 301,193) ; Net 1,000,
700 fathoms, June 5, 1931 (No. 31,126) ; and Net 1,004, 600 fathoms, June
6, 1931 (No. 31,155).

Drieschia Michaelsen.

Drieschia atlantica n.sp.

(Figs. 7-9).

There are three specimens in the collection, all more or less mutilated.
The type which is broken near the centre but evidently retains all of its
somites, is 12 mm. long and its greatest diameter, just back of the head, is
1 mm. The prostomium (Fig. 7) has rounded posterior angles and its lat-
eral margins are only slightly incurved to meet the cirrophores of the
lateral tentacles. These latter are not sharply separated from the prosto-
mium and are about one-quarter as long as it. There is no marked anterior
incision for the attachment of the cirrophore of the median tentacle, which
in the type is not sharply separated from the lateral ones. In other speci-
mens the distinction is sharper. The median cirrophore is about twice as
thick as the lateral and a trifle longer. The eyes are small but prominent,
the anterior pair lying near the lateral margins at about the middle of the
prostomium while the posterior pair are near the posterior margin and are
closer together than the anterior.

The style of the lateral tentacle is about twice as long as the cirro-
phore, slender and sharp pointed, and that of the median tentacle is more
than four times as long as the lateral and very slender in relation to its
length. The palps are also slender. In the type they extend to a distance
considerably beyond the tentacle tip but in another specimen (probably
injured), they are much shorter. The tentacular cirri are also slender and
as long as the palps.

The parapodia are uniramous, the notopodium being absent (Fig. 8).
The neuropodium is cylindrical and sharp pointed, having a single acicula
and a small ventral cirrus not reaching the apex of the parapodium. In
all cases the cirrophore of the dorsal cirrus is inflated to form a hollow
structure which in some anterior somites may be larger than the parapo-
dium and completely obscure it from a dorsal view. The cirrophores are
largest in the 5th and 7th parapodia and decrease in size posteriorly,
though their size relative to the parapodium is not so much less in this
region. Too many of the styles have been lost to determine the point with

1936]

Treadwell: Polychaetous Annelids , Nonsuch Island

53

accuracy but it seems as if, posterior to the 7th parapodium, the styles are
alternately long and short. They are all very slender. There are 2 slender
anal cirri.

Just dorsal to the apex of the parapodium is a tuft of two or three
long and slender setae and ventral to it a few much larger ones. These
latter (Fig. 9) broaden slightly toward the end and then narrow to a
blunt point, this terminal portion being very slightly spoon-shaped and
carrying a series of small toothed plates along the concave surface.

The type has 14 pairs of elytrophores but all of its elytra are lost. One
other specimen retains the most anterior and most posterior one on one
side. Apparently in life they cover the entire dorsal surface of the body.
Descriptions of elytra from other species of this genus speak of them as
“inflated, ” this together with the inflated character of the dorsal cirrophores
being correlated with their pelagic mode of life. Neither of the two above
mentioned elytra showed any sign of inflation but they very probably
were so when alive, for they look very much as a very thin-walled disk-
shaped sac might look if it were collapsed and its wall thrown into wrinkles.
The outline is nearly circular, the margin smooth and the surface covered
with numerous wrinkles running in all directions.

The type was taken in Net 953, 1,000 fathoms, Sept. 26, 1930 (No.
301,259) and is in the collections of the Department of Tropical Research
of the New York Zoological Society. The others were taken in Net 841, 500
fathoms, Sept. 4, 1930 (No. 30,719) and Net 939, 1,000 fathoms, Sept. 24,
1930 (No. 301,176).

Family Sigalionidae.

Eupholoe McIntosh.

Eupholoe nuda n. sp.

(Figs. 10-14).

A single specimen 30 mm. long and 3 mm. wide. The prostomium (Fig.
10) is hemispherical in outline but is truncated on its anterior margin for
the insertion of the single tentacle. The tentacle is small, its apex not
reaching to the ends of the underlying setae. There is a pair of small
eyes on the dorsal peristomial surface and a much larger pair on the ventral,
not visible from the dorsum. There are no tentacular cirri on the speci-
men. In Fig. 10 the prostomium is drawn as if the anterior dorsal margin
of the 3d somite, which when in place completely covers it, were drawn
back. The cirri figured at the sides are the dorsal cirri of the 3d somite.
The first two parapodia with their setae crowd together under the pro-
stomium and are in contact on the ventral line, lying between the pro-
stomium and the palps. The 2d parapodia extend forward of the prostomium
and as stated above, the 3d somite extends over the prostomium. The palps
are slender and extend only to a short distance beyond the ends of the
first setae. The first ventral cirrus is large, the next one smaller and
throughout most of the body they are very small.

In this genus the elytra should form an overlapping series along both
margins of the body, leaving the greater part of the surface uncovered.
In E. nuda no elytra are present, although oval or round areas looking like
the scars left when elytra are torn off occur in the appropriate somites.
The most reasonable conclusion would be that elytra were lost, but careful
examination of these structures fails to reveal any trace of torn or ragged
edges such as should appear if this had been the case. The dorsal body
surface is sprinkled with fine sand grains which extend down to the
parapodia and show no trace of disturbance in the places where elytra
should have been. It seems certain that the specimen is normally lacking in
elytra. These sand grains are quite uniformly scattered over the dorsal

54

Zoologica: New York Zoological Society

[XXI :2

surface, being generally a trifle larger in the mid-line than elsewhere but
otherwise show no differences in different body regions. The ventral sur-
face is thickly studded with short rounded papillae (Fig. 11) which give it
a “shagreen” appearance. On the dorsal surface of each parapodium, and
lying just under the above mentioned elytral “scar” is a single “branchia”
in the form of a blunt papilla whose length only slightly exceeds its breadth.

The parapodia (Fig. 11) have large neuro- and very much smaller
notopodia, the former truncated at the apex, giving it a triangular outline.
The notopodium is much smaller, looking like a mere outgrowth from the
neuropodium. It carries on its upper surface a small knob-like protuberance.
Each lobe has a single acicula. The ventral cirrus is small and does not
reach the end of the neuropodium. On the ventral and lateral surface of
the parapodia are numerous papillae.

The notopodial setae form a dense bunch radiating from a common
center on the dorsal face of the notopodium. They are slender and sharp
pointed and each carries two rows of toothed plates (Fig. 12). Two kinds
of setae occur in the neuropodium. The more dorsally placed ones are very
large and heavy, the terminal joint relatively rather short, and are without
any subterminal tooth (Fig. 13). Ventral to these is a tuft of more slender
setae in which the terminal joint is relatively longer and more slender and
has a single very slender subterminal tooth (Fig. 14). The relative sizes of
these two kinds of setae are shown in Figures 13 and 14, which are drawn
to the same scale.

The genus Eupholoe was established by McIntosh (1885, p. 157), for
E. philippinensis, on a single specimen dredged off Mindanao in the Philip-
pines (1885, pp. 157-159, PI. 22, Figs. 6, 7; PI. 24, Fig. 7; PI. 25, Fig. 10;
PI. 13A, Figs. 16, 17). The diagnosis is as follows: “Body elongated,
somewhat truncated in front and tapering posteriorly, the former end be-
ing covered with coarse and the latter with fine, sand grains. Elytra small,
confined to the lateral regions and furnished with peculiar processes which
like the other parts of the scales are covered with long cilia. A rudimentary
branchia (?) on each foot. Dorsal bristles slender with long spinous rows,
ventral with single short terminal processes beneath the hook of which is a
minute spine.” In the detailed description and figures of E. philippinensis
the prostomium has exactly the structure of E. nuda, but tentacular cirri
are absent from the latter (presumably by accident) and the two first
parapodia do not crowd together under the prostomium in E. philippinensis
as they do in E. nuda. A minor difference is that in E. nuda the sand
grains on the dorsal body surface are of uniform size throughout. Whether
the elytra in the latter species are really absent must be determined when
other material is available for study. McIntosh’s Fig. 16, PI. 13 A, is
exactly like Fig. 12, while his Fig. 17 is essentially like Fig. 13 except that
he represents a very small subterminal tooth. In his drawing this looks
more like an accidental irregularity than a tooth, and it may be that the
two are alike in structure. He does not figure the more slender ones like
Fig. 14.

The type is in the collections of the Department of Tropical Research
of the New York Zoological Society and was collected in coral at Gurnet’s
Rock, Bermuda, 35 feet deep, Sept. 29, 1930 (No. 301,308).

Family Chrysopetalidae.

Bhawania Schmarda.

Bhawania goodei Webster.

Bhawania goodei Webster, 1884, pp. 308, 309; PI. 7, Figs. 10-15.

Collected from reef in Castle Harbor, Aug. 16, 1931 (No. 311,350) and
from coral at Gurnet’s Rock, 35 feet deep, Aug. 19, 1931 (No. 311,483).

1936]

Treadwell: Polychaetous Annelids, Nonsuch Island

55

Family Glyceridae.

Glycera Savigny.

Two very young specimens of this genus, too immature for identifica-
tion, collected in Net 824, 800 fathoms, Sept. 1, 1930 (No. 30,672).

Family Aricidae.

Nainereis Blainville.

Nainereis setosa Verrill.

Aricia setosa Verrill, 1900, pp. 651-653.

Two specimens, collected on mud flat, St. David’s Island, Bermuda,
Aug. 17, 1931 (No. 311,397).

Family Nereidae.

Nereis Linnaeus.

Nereis bairdii Webster.

Nereis bairdii Webster, 1884, pp. 312-313; PI. 8, Figs. 22-28.
Collected at Gurnet’s Rock, 35 feet deep, Aug. 19, 1931 (No. 311,481) ;
tidepools, Nonsuch Island, Aug. 16, 1931 (No. 311,357) ; and mud flats, St.
David’s Island, Aug. 17, 1931 (No. 311,396). These were all in crevices in
coral rocks from shallow water.

Heteronereis phase.

(Figs. 15, 16).

Specimens collected in Net 817, 600 fathoms, Aug. 29, 1930 (No.
30,630) ; and Net 834, 400 fathoms, Sept. 3, 1930 (No. 30,680) were identi-
fied as N. bairdii from their jaw structure (Webster’s Figs. 22a and 23).
The prostomium carries very large eyes, those of the same side in contact
with one another and their lenses small. The peristomium is swollen so as
to be as wide as the 1st somite (compare Webster’s Fig. 22 for the atokous
phase). In one of the two specimens there are 14, in the other 18, somites
in the anterior region. This anterior region is widest at about the 6th
somite and from here narrows in either direction. The tentacles and palps
are little changed from the atokous condition but appear larger because of
the swollen prostomium. Neither specimen is entire, one retaining some
thirty somites in the posterior region, the other fewer. The tentacular
cirri retain the relative length of the atokous phase but their bases are
heavier and in the preserved material have a general “ram’s horn” effect.
In the anterior region dark brown pigment patches lie on the body wall just
dorsal to the parapodia, the usual arrangement being one nearer the an-
terior and one nearer the posterior border of the somite. These patches are
much darker in one specimen than in the other.

The anterior parapodia are not noticeably different from those of the
atokous phase except that the dorsal cirrus acquires a broader base and
becomes lanceolate in outline. The setae show no change (see Webster’s
Figs. 26 and 27) . The structure of the parapodia of the modified region can
best be understood by reference to Fig. 15, taken from the smaller of the
two specimens. Neither of the two contains sex products and from the
smooth character of the dorsal cirrus I assumed that this is a female. The
other specimen, however, is larger and in this, while the character of
the general modification is the same, the dorsal cirrus shows the lobings
characteristic of the male. It seems certain therefore that both are males
but that one has not yet fully reached the epitokous condition. The setae

56

Zoologica: New York Zoological Society

[XXI :2

are numerous, extend considerably beyond the parapodium lobes and have
the characteristic broad terminal joints.

Nereis glandulata Hoagland.

Nereis glandulata Hoagland, 1919, p. 575; PL 30, Figs. 1-6.

Heteronereis phase. (Fig. 17). Identified by a comparison of the
jaws with Hoagland's Figs. 2 and 3. This species also has very rounded
parapodial lobes and a pigment patch on the parapodium (Hoagland’s Fig.
4), which characters are retained in the anterior somites of the epitokous
phase. The prostomium, except for increase in the size of the eyes, is not
much changed from the atokous condition. The 1st somite is about equal to
the prostomium in width and from there there is a gradual increase in
width to the 6th somite. In this anterior region the only change is in
the character of the dorsal cirrus, which is larger in all somites and in the
middle of the region has the avicular character shown in Fig. 16. The setae
are unmodified (compare Hoagland's Figs. 5 and 6). In the posterior,
modified region the parapodia are as in Fig. 17. Since the dorsal cirrus is
lobulated I have assumed that both are males. The setae are numerous,
long and have the usual heteronereid form.

Collected in Net 767, 800 fathoms, July 3, 1930 (No. 30,382) ; and Net
775, 1,000 fathoms, July 4, 1930 (No. 30,396).

Nereis mirabilis Kinberg.

Nereis mirabilis Kinberg, 1865, p. 170.

Nereis gracilis Webster, 1884, pp. 313, 314; PI. 9, Figs. 29-35.

Name preoccupied. See Kinberg loc. cit., p. 170.

Collected at Gurnet's Rock, Bermuda, in coral, 35 feet deep, Sept. 29,
1930 (No. 301,308) ; Castle Harbor, reef, Aug. 15, 1931 (No. 311,334) ;
and tidepools, Nonsuch Island, Aug. 16, 1931 (No. 311,357). Found in
association with N. bairdii in decomposing coral rock.

A number of small heteronereids collected Aug. 18, 1931, swimming
at the surface under a light, are evidently N. mirabilis , because they show
the very unusual character of the prostomium of this species which is little
changed in the heteronereid. The first dorsal cirri are somewhat expanded
but the others show no modifications. The parapodia of the modified region
have the usual fin-like lobes.

Family Leodicidae.

Leodice Savigny.

Leodice mutilata Webster.

Eunice mutilata Webster, 1884, pp. 315-316; PI. 9, Figs. 36-40.

Leodice mutilata Treadwell, 1921, pp. 30-33; PI. 3, Figs. 5-8; Text-figs.

66-76.

In the collections sent me, this was recorded only from Gurnet's Rock,
35 feet deep, Sept. 29, 1930. It was abundant, however, in the collections I
studied on Nonsuch Island, and is the commonest Leodice of the region, oc-
curring in dead coral rock. From the twisted position it assumes among the
rock crevices, it is difficult to extract without breaking, a peculiarity which
probably accounted for Webster's specific name.

1936]

Treadwell : Polychaetous Annelids, Nonsuch Island
Leodice culebra Treadwell.

57

Leodice culebra Treadwell, 1921, pp. 49-51; PI. 2, Figs. 13-16; Text-figs.
144-153.

Collected in tidepools, Nonsuch Island, Aug. 16, 1931.

Leodice longicirrata Webster.

Eunice longicirrata Webster, 1884, pp. 318, 319; PI. 12, Figs. 75-80.
Leodice longicirrata Treadwell, 1921, p. 11-15; PI. 1, Figs. 1-4; Text-
figs. 3-12.

Collected at Gurnet’s Rock, 35 feet deep, Sept. 29, 1931 (No. 301,308)

Leodice denticulata Webster.

Eunice denticulata Webster, 1884, pp. 316-317 ; PI. 10, Figs. 41-45.
Leodice denticulata Treadwell, 1921, pp. 22-25; PI. 3, Figs. 1-4; Text-
figs. 41-53.

Collected at Gurnet’s Rock, 35 feet deep, Aug. 19, 1931 (No. 311,482).

Nicidion Kinberg.

Nicidion kinbergii Webster.

Nicidion kinbergii Webster, 1884, pp. 320, 321 ; PI. 12, Figs. 81-88.

Treadwell, 1921, pp. 91-93; PI. 6, Figs. 6-8; Text-
figs. 324-332.

One specimen, locality uncertain. It is generally found in the harder
portions of the dead coral rock near low water mark.

Marphysa Quatrefages.

Marphysa regalis Verrill.

Marphysa regalis Verrill, 1900, pp. 636, 637.

Marphysa fragilis Treadwell, 1911, pp. 2-5; Figs. 1-7.

Marphysa regalis Treadwell, 1921, pp. 66-69; pi. 5, Figs. 9-12; Text'
figs. 224-234.

Collected at Gurnet’s Rock, 35 feet deep, Sept 29, 1930 (No. 301,308) ;
tidepools, Nonsuch Island, Sept. 12, 1930 (No. 301,675). It is very common
in the soft beach rock.

Marphysa acicularum Webster.

Marphysa acicularum Webster, 1884, pp. 319-320; PI. 10, Figs. 50-53.

Treadwell, 1921, pp. 57-59; PI. 5, Figs. 1-4;
Text-figs. 184-193.

Collected on St. David’s Island, Aug. 17, 1931 (No. 311,399). It is com-
mon in muddy flats between tide levels.

Paramarphysa Ehlers.

Paramarphysa obtusa Verrill.

Paramarphysa obtusa Verrill, 1900, pp. 646, 647.

Treadwell, 1921, pp. 76, 77; Text-figs. 269-278.
Collected in coral, Gurnet’s Rock, 35 feet deep, Sept. 29, 1930 (No.
301,308).

58

[XXI :2

Zoologica : New York Zoological Society
Dorvillea Parfitt.

In most of the literature this genus is given as Staurocephalus. Verrill
(1900, pp. 647, 648) gave reasons for changing it to Stauronereis. Cham-
berlin (1919, p. 339), applying the laws of priority, showed that it should
be Dorvillea.

Dorvillea melanops Verrill.

Stauronereis melanops Verrill, 1900, pp. 647, 647.

Treadwell, 1921, pp. 125-127; Text-figs. 459-467.
A single small specimen was collected with numerous individuals of
Polyophthalmus (see p. 61) under electric light at the boat landing on Non-
such Island (No. 311,436) on the evening of Aug. 18, 1931. When adult,
it is not a pelagic species.

Dorvillea erythrops Verrill.

Stauronereis erythrops Verrill, 1900, pp. 649, 650.

One specimen, taken under electric light, Nonsuch boat landing, Aug.
19, 1931 (No. 311,485).

Family Tomopteridae.

Tomopteris Eschscholtz.

Tomopteris longisetis n. sp.

(Figs. 18-21).

Noticeably different from previously described species in that the sec-
ond cirrus with its seta is considerably longer than the body. The following
measurements were taken from four of the best preserved specimens.

Total body length Length of Second cirrus No. of parapodia

55 mm. 65 mm. 38.

23 mm. 31 mm. 25.

22 mm. 27 mm. ?

55 mm. 90 mm. 38.

The parapodial number given refers only to the well defined body
somites and does not include the rudimentary ones of the “tail.” In the third
of the above measurements the number was uncertain owing to imperfect
preservation. As has before been recorded in this genus, the somite number
varies with the total length.

In general body appearance T. longisetis differs from the usual in that
the body is broader in proportion to its length and the separation between
body and parapodia is much less distinct. The parapodia are short and
thick and are in contact at their bases, giving the whole body a much more
compact appearance than is usual. In other species the parapodia may be
as long as the body width; in this they are shorter.

In preserved material the body is opaque and in some has a pearly
white appearance, while others are brownish in tint. The tentacles (Fig.
18), are slender and extend to the 2nd parapodium. The 1st cirrus is
absent. The 2d cirrus has a triangular base whose apex extends about as
far as to the end of the tentacle. Seen from the ventral surface the
cerebral ganglion with commissures and ventral cord are prominent fea-
tures, being lighter in color and more opaque than the remainder of the
body. Anterior to the cerebral ganglion and visible on the ventral surface
is a pair of small eyes which look as if they are borne on narrow lobes

1936]

Treadwell: Polychaetous Annelids, Nonsuch Island

59

extending anteriorly from the cerebral ganglion. From the 1st to the 4th
parapodium there is a regular increase in length, the 4th being about twice
as long as the 1st. From the 4th backward there is a very slight increase in
length as far as the anterior third of the body and behind this a gradual
decrease to the narrow posterior end, the parapodia in all cases having a
definite size relation to that of the somite to which they are attached.
Their form is much the same throughout the body except for the “tail”
where they are cylindrical structures with a mere hint of a bifid apex.

The parapodia (Fig. 19) have short thick bases and bifid ends, the end
portions equal in size and each carry a much fluted fin extending all around
the terminal margin. The glandular structures are confusing in this ma-
terial. In the larger specimens where this fluting is most noticeable, no
trace of glands could be seen. In the smaller specimens the ventral lobe of
the parapodium carries a large gland-like structure, the row of these when
seen under low magnification looking like a series of suckers (Fig. 20).
These have no color. In the smallest specimens each apical branch of the
parapodium has on its outer margin a rosette gland with a dark sepia color
(Fig. 21). In some cases the rosette glands are not pigmented. In the un-
usual length of the second cirrus, T. longisetis agrees with T. nisseni Rosa,
as described by Southern (1910, p. 17, PI. 1, Figs. 1 and 2), but is funda-
mentally different in that T. nisseni has fewer parapodia.

Specimens were taken in 18 deep sea nets, drawn from 500 to 1,000
fathoms, during May 17-28, 1929, Sept. 1-24, 1930, and July 25 to Aug 18,
1931 (Nos. 29,475; 29,476; 3,050; 30,116; 30,537; 30,623; 30,757; 301,122;
301,124; 301,125; 301,129; 301,155; 301,675; 31,840; 311,078; 311,103;
311,164; 311,432).

Tomopteris sp ?

A single specimen 35 mm. long, with tail 15 mm. in length, thus the tail
relatively longer than in T. longisetis. The parapodia are much more dis-
tinct from the body than is the case in the latter species, are longer and
more slender and are without marginal frill. There is a single very large
opaque rosette gland on each ventral branch. On one side of the ventral
region is what seems to be a 1st cirrus and eyes are present, but the whole
is too much macerated for accurate description. The 2nd cirrus is not
more than one-half as long as the body. Because of gradual modification of
the parapodia until they seem to fade away into the general body surface,
it is not easy to determine their number in the tail region but there are at
least twenty. Taken in Net 792, 600 fathoms, July 9, 1930 (No. 30,486).

Family Alciopidae.

Vanadis Claparede.
Vanadis fusca-punctata Treadwell.

Vanadis fusca-punctata Treadwell, 1906, pp. 1159, 1160; Figs. 29-31.

The species was described from specimens collected in the Hawaiian
Islands. Others were taken at the Galapagos Islands by the Arcturus Ex-
pedition (Treadwell 1928, p. 462) and it is the most abundant species in
the Bermuda collections. It can easily be recognized by the form of the
head region and the rows of dark spots on the body.

Collected in 52 deep sea nets from 0 to 1,000 fathoms during the fol-
lowing dates: May 20 to June 17, 1930; Aug. 28 to Sept. 25, 1930, and June
3 to Aug. 14, 1931. (Nos. 29,471; 3,069; 30,121; 30,229; 30,234; 30,531;

fizinh • SO • 20

60

Zoologica: New York Zoological Society

[XXI :2

31,437; 31,438; 31,472; 31,496; 31,589; 31,594; 31,610; 31,630; 31,635;
31,661; 31,681; 31,706; 31,712; 31,756; 31,866; 31,927; 311,131; 311,184;
311,207; 311,254).

Fragments, probably of this species but too much injured for accurate
identification, were collected in 11 other nets towed at depths from 25 to
1,000 fathoms during June, 1930, and June 15 to Aug. 11, 1931 (Nos. 30,142 ;
30,841; 30,987; 30,990; 301,199; 31,217; 31,456; 31,681; 31,749; 31,758;
311,174).

Family Phyllodocidae.

Phyllodoce Savigny.

Phyllodoce oculata Ehlers.

Phyllodoce oculata Ehlers, 1887, pp. 135-140; PI. 40, Figs. 4-6.
Collected in coral, Gurnet’s Rock, 35 feet deep, Sept. 29, 1930 (No.
301,308) and August 19, 1931 (No. 311,484) ; Castle Harbor, Aug. 15, 1931
(Nos. 311,333; 311,336).

Lopadorhynchus Grube
Lopadorhynchus nans Chamberlin.

Lopadorrhynchus nans Chamberlin, 1919, pp. 116-119; PI. 17, Figs 1-5.

Two well marked varieties of this species are in the collection. Most
are slender, not more than 10 mm. in length, pearly white in alcohol and
must have been translucent when alive. The others are generally much
longer and have thick, opaque, yellowish brown bodies. Both agree with
Chamberlin’s description of the form of the head and anterior somites.
Chamberlin’s single specimen was taken between Easter Island and Peru in
17° 55' S. Lat. and 87° 42' W. Long. The Bermuda specimens were taken in
10 deep-sea nets, ranging from 50 to 1,000 fathoms, during September, 1930,
and from July 6 to Aug. 19, 1931. (Nos. 30,863; 301,138; 301,181; 31,453;
31,470; 31,494; 31,609; 31,633; 31,999; 311,476).

Lopadorhynchus uncinatus Fauvel.

Lopadorhynchus uncinatus Fauvel, 1916, pp. 57-61; PI. 1, Figs. 2, 3;

PL 4, Figs. 1-4.

Easily recognized by the excessive development of the first two para-
podia. Fauvel in his diagnosis of the species says that the eyes are black,
but in the later description he describes them as having a more or less cir-
cular margin of brown. In this respect and in the distribution of the chro-
matophores over the body surface the Bermuda specimens agree with Fan-
vel’s, which were collected in the vicinity of Monaco. The Bermuda speci-
mens were taken in the following nets: Net 945, 500 fathoms, Sept. 25, 1930
(No. 301,213), Net 947, 700 fathoms, Sept. 25, 1930 (No. 301,224) and Net
967, 500 fathoms, Sept. 30, 1930 (Nos. 301,314).

Family Opheliidae.

Ammotrypane Rathke.

Ammotrypane bermudiensis n. sp.

(Figs. 24-26).

The body is rounded dorso-laterally and flattened ventrally with a deep
longitudinal groove along the ventral surface. Since intersegmental constric-
tions do not appear, the number of somites can only be estimated by the
seta tufts of which there are 32 pairs, situated ventro-laterally. The type

1936]

Treadwell: Polychaetous Annelids , Nonsuch Island

61

is 32 mm. long, its greatest width 2.5 mm. From apex of prostomium to the
first seta tuft is 1.5 mm. The anal tube is 3 mm. long. At the posterior
end the somites (as indicated by the seta tufts), are very closely crowded
together. The prostomium is conical (Fig. 24), its basal diameter being
equal to one-half of its length and it carries a small mucron on the apex.
From the base of the prostomium there is a slight increase in width to the
middle of the body and from there a decrease posteriorly. At its base the
anal tube has the same width as the last somite, but it narrows to an apex
about one-half as wTide. It is cylindrical in form with transverse markings
in some cases clearly marked, in others more obscure. The anal opening is
dorso-terminal and is bounded laterally by two thin plates whose free edges
in some cases show simple scalloping, in others carry about six short papil-
lae (Fig. 25). What seems to be the normal condition of the anal cirri is
that there is one pair of short rounded ones, with a longer and more slender
one between them, located on the anal tube near its ventral surface. They
are sometimes covered over and obscured by the cirri of the last setigerous
somites.

Cirrus-like gills occur in each setigerous somite except the 1st. They
are long and slender and extend approximately to the mid-dorsal line. Lat-
erally, pigment spots first appear posterior to the 5th seta tuft on either side
and are continued posteriorly for at least 21 somites. The first two pairs
are very small, the next nine pairs are much larger and the remaining pairs
are small again. Associated with each of the two posterior pairs of these
pigment spots is a much larger spot, located ventrally and more deeply im-
bedded in the body tissue so that it is seen only obscurely from the surface.

The pharynx is only partially protruded in any specimen but apparently
is cylindrical. On either side of its base is a tuft of papillae which would
locate this species in Kinberg’s genus Terpsicore (1865, p. 257), but it lacks
the anal cirri given by Kinberg as diagnostic of that species.

The parapodia (Fig. 26) are very small, the neuropodium having a
prominent rounded end, and the seta tuft arises posterior to that. The noto-
podium has several rounded lobes and the dorsal seta tuft arises between
its base and that of the gill. The gills are very large relative to the para-
podium (Fig. 26). The ventral setae are much shorter than the dorsal and
are fewer in number, all being unilimbate, curving gently to an acute apex.
The dorsal setae are of varying lengths, some being nearly as long as the
gill, others much shorter, but in general are much longer than the ventral
ones. All are very narrowly limbate. The central axis is markedly striated.

The type was collected in sand near Nonsuch Island, 10 feet deep, June
25, 1929 (No. 29,470). Others were taken in sand, 35 feet deep, near Gur-
net's Rock, April 6, 1929 (No. 29,473), and in sand near Nonsuch Island,
10 feet deep, Oct. 1, 1931 (No. 312,233). A considerable number are re-
corded simply as from Castle Harbor in sand in association with Asymmet-
ron. There is a marked superficial resemblance between the annelid and the
cephalochordate which has been commented upon by various writers.

Polyophthalmus Quatrefages.

Polyophthalmus incertus n. sp.

(Figs. 27-29).

In material collected under electric light at the wharf on Nonsuch
Island on the evening of Aug. 18, 1931, was a single specimen of Polyoph-
thalmus which is the type of this new species. On the following evening a
large number of much smaller specimens of the same genus were collected
at this locality. The type is 8 mm. long and while the ones collected later
may reach a length of 5 mm., they are so slender that they give the im-
pression of being very much smaller. When alive in sea water their form

62

Zoologica: New York Zoological Society

[XXI :2

and movements give them very much the appearance of nematodes. The
type is well preserved but for some reason the preservation methods which
have given excellent results in other cases did not succeed with the others,
the cuticle being much swollen and thrown into folds so that the animals
look as if they are living in a definite “haus.” The rarity of the genus makes
it important that its occurrence should be recorded though the present scar-
city of material prevents a thorough species diagnosis. Furthermore, since
the species is not generally pelagic, it is probable that these are immature
stages. The following description must therefore be regarded as tentative,
to be corrected and expanded when better material is available.

The type is marked by 14 lateral “eye spots” of which the 1st is very
small, the 2nd about twice as large as the 1st and the 3rd more than twice
the size of the 2nd. The 4th to 9th inclusive are of the same size as the 3d,
while the next 5 become successively smaller. In the anterior part of the
body pigment patches occur in a scattered arrangement over the dorsal
surface but in about the region of the 1st lateral eye spot these are more
definitely localized in the mid-dorsal line and throughout the greater part
of the body they occur as transverse brown patches in the mid-dorsal sur-
face, presumably arranged somitically, though the somite limits are not easy
to determine. At the posterior end the body narrows into a tubular pygidial
region (Fig. 29) which carries a row of about twenty cirri around its pos-
terior margin. This pygidial region is marked on either side by a series of
7 pigmented rings.

The prostomium (Fig. 27) is broadly rounded and carries one pair of
prominent brown eyes. Behind the eyes is a transverse suture, presumably
indicating a nuchal organ. The mouth (Fig. 28) is an elongated slit, having
fleshy lateral lips.

Throughout most of the body the setae are very small and few in num-
ber so that they are difficult to see, but at the posterior end, where the body
is beginning to narrow into the pygidial region, 4 somites carry each a dor-
sal and a ventral tuft of very long setae, longer than the transverse di-
ameter of the body at this point. The setae are very slender and sharp
pointed.

In its body pigmentation this species apparently resembles P. pictus
Dujardin, but differs decidedly from that in the form of the head and tail
regions as figured by Fauvel (1914, pp. 247, 248, PI. 22, Figs. 8, 9).

The type was collected at an undersea lamp, Nonsuch Island, Aug. 18,
1931 (No. 311,437) and is in the collections of the Department of Tropical
Research of the New York Zoological Society. There is another specimen
from the same locality, Aug. 19, 1931 (No. 311,487).

Family Typhloscolecidae.

Travisiopsis Levinsen.

Travisiopsis atlantica n. sp.

(Figs. 30-33).

The type is 24 mm. long and has a body width of 4 mm. Thirty somites
carry the flattened cirri. The prostomium (Fig. 30) is bluntly conical but
has a short filamentous tip and its basal diameter is about equal to its length.
It is largely covered on either side by the overlapping 1st cirrus. On the
posterior border of the 1st somite is a pair of tentacular processes which
are about as long as the 2nd somite and in the preserved material extend
almost vertically. The mouth is a relatively large circular opening.

The first 3 pairs of cirri are notopodial but beginning with the 4th
they occur on both noto- and neuropodia and this arrangement is continued
throughout the body. The first pair are nearly circular in outline but later

1936]

Treadwell: Polychaetous Annelids, Nonsuch Island

63

ones become broadly ovate with flattened points of attachment and narrow
points on the outer margins (Fig. 31). A ring of 5 narrower cirri surround
the anal opening. These (Fig. 32) are carried on a prominent cirrophore
and are asymmetrically lanceolate in outline but have blunt apices. From the
base to the apex and nearer one margin is a band of much firmer texture
than elsewhere in the cirrus and must give it a considerable degree of
rigidity. A much smaller band occurs in the other cirri. When in position,
because they extend stiffly out from the posterior end while the other cirri
are apt to lie more closely appressed against the sides of the body, the anal
cirri appear to be much larger than the others. Figs. 31 and 32, drawn to
the same scale, show this to be erroneous. The cirrus from which Fig. 32
was drawn was cut from the type which was the only one in which these
anal cirri were not badly wrinkled. The setae in the smaller specimens are
more easily seen than in the larger. They are few in number in a single
tuft, stout and sharp pointed (Fig. 33).

The type was collected in Net 1,151, 600 fathoms, Aug. 8, 1931
(No. 311,116), and is in the collections of the Department of Tropical Re-
search of the New York Zoological Society. Others were taken in the fol-
lowing nets: Net 798, 600 fathoms, July 15, 1930 (No. 30,522) ; Net 860,
600 fathoms, Sept. 8, 1930 (NOi 30,793) ; Net 866, 700 fathoms, Sept. 10,
1930 (No. 30,830), and Net 929, 700 fathoms, Sept. 20, 1930 (No. 301,134).

Travisiopsis sp. ?

These are smaller than T. atlantica and their systematic position is
doubtful, though they may be the young of that species. Only a few cirri
are preserved. The setae are similar to those of T. atlantica but especially
in the posterior portion of the body are more prominent and extend to a
considerable distance from the body wall.

Collected in Net 693, 900 fathoms, June 12, 1930 (No. 30,182) ; Net
726, 800 fathoms, June 26 (No. 30,258) ; Net 893, 900 fathoms, June 15,
1930 (No. 301,058); Net 917, 600 fathoms, Sept. 19, 1930 (No. 30,964).
Fragments too much injured for any identification were taken in two other
nets (Nos. 30,964; 30,258).

Family Cirratulidae.

Cirratulus Lamarck.

Cirratulus multicirratus n. sp.

(Figs. 34, 35).

The type and only specimen is 60 mm. long and has a body width of
4 mm. in the widest portion. All except the first two somites are very short.
Cirri begin on the 1st setigerous somite and continue nearly to the extreme
posterior end. In the anterior region they occur on every somite but toward
the posterior end there are considerable gaps in the series, though consid-
ering the fact that they may appear in several consecutive somites and then
be lacking from another set, it is possible that their absence means a loss
rather than a normal condition. The prostomium (Fig. 34) is bluntly
conical and is so much fused with the 1st somite and this latter with the
2nd that it is difficult to distinguish the boundaries, except for faint lines
on the lateral surfaces. The first two somites together are about as long as
the prostomium, while the 3rd somite is hardly more than half as long as
the second. Posteriorly there is an increase in somite length. The body
narrows decidedly toward the posterior end and the pygidium is cylindrical
with an oval, dorsally directed anus. Setae begin on the 3rd somite as
capillary structures in both parapodial lobes. Hooks (Fig. 35) appear in
the ventral ramus by the end of the anterior quarter of the body and pos-

[XXI :2

64 Zoologica: New York Zoological Society

terior to the middle of the body they are the only ones represented in the
parapodia.

The first cirri are small but in the region from the 10th to the 20th
somites they are much longer, forming a noticeable bunch. Later they again
become smaller, but the length is always many times the body width. On
either side of the dorsal surface of the 5th setigerous somite is a tuft of
6-8 gills smaller than the neighboring cirri. They are bunched close to the
margins on either side, leaving the greater part of the dorsal surface un-
covered.

The type was collected on mud flats on St. David’s Island, Bermuda,
Aug. 17, 1931 (No. 311,398) and is in the collections of the Department of
Tropical Research of the New York Zoological Society.

Audouinea Quatrefages.

Audouinea pygidia n. sp.

(Figs. 22, 23)

Two specimens of which the larger is 22 mm. long and 2.5 mm. in great-
est width. The most characteristic external feature is that the posterior
end over a region of some twenty somites is entirely colorless, while imme-
diately in front of this an area of about the same number of somites is
colored black. If found in only one individual the colorless region might
reasonably be regarded as a regenerating portion, but since it is present in
both it is obviously a normal condition. The remainder of the body is light
brown in color, with a tendency toward a darker tone anteriorly, but the
extreme anterior end is lighter.

The prostomial width much exceeds its breadth and its margins merge
into those of the 1st somite (Fig. 22), the two making an area whose basal
width is about twice its length and except that the base is a trifle too wide,
is hemispherical in outline. The second somite is shorter at the margins
than on the dorsal surface where it expands to project posteriorly into the
anterior margin of the 3rd somite. The 3rd somite is twice as long as the
2nd. Later somites are all very short as compared with their width, but be-
hind the first eighth of the body this proportion slightly increases. Capillary
setae first appear in the 4th somite and hooks occur in the neuropodia after
about the 20th somite, while only capillary ones are found in the notopodia.

Apparently most of the dorsal cirri have been lost from both specimens
since their distribution is different in the two, but occur in scattering fash-
ion to the extreme posterior end. In the type there is a small one on the
3rd somite, a much larger one on the 6th and an irregular distribution far-
ther back. A very large dorsal gill was located on the 5th somite in the
type. This was broken away by an accident, but from the scars it seems
certain that originally there were two of these on either side of the body,
lying on the 5th and 6th somites.

The setae of anterior somites and of the notopodium in all somites are
capillary, numerous in each tuft, long, slender and sharp pointed. The hooks
are dark brown in color, curved to an acute point at the end and protrude
for a considerable distance from the body surface (Fig. 23).

The paratypes were collected from tidepools on Nonsuch Island, Sept.
12, 1930 (No. 301,675), and are in the collections of the Department of
Tropical Research of the New York Zoological Society.

Dodecaceria Oersted.

Dodecaceria sp. ?

Fragments too poorly preserved for identification, collected in tidepools,
Nonsuch Island, April 23, 1929 (No. 2,945) and Aug. 16, 1931 (No. 311,357).

1936]

Treadwell: Polychaetous Annelids, Nonsuch Island

65

Family Sabellidae.

Protulides Webster.

Protulides elegans Webster.

Protulides elegans Webster, 1884, pp. 325, 326 ; PI. 10, Figs. 63-74.
Collected from reef, Castle Harbor, Aug. 15, 1931 (No. 311,335). One
specimen.

V ermilia Lamarck.

Vermilia glandulata n. sp.

(Figs. 36-39).

The type has a total length of 13 mm. of which the thorax and branchiae
each make up 3 mm. The opercular stalk is as long as the branchiae and
carries a large globular operculum. There are 13 branchiae on the right and
12 plus the opercular stalk on the left. The opercular stalk (Fig. 36) is
divided into about twenty rings of which the terminal is the largest. The
operculum is a globular body which on its outer surface is drawn out into
a terminal portion which was covered by a white limestone deposit. This
terminal portion is brown in color and marked by circular brown lines
darker than the general surface. The remainder of the body is colorless.
The gill filaments are in two rows and are very short, hardly longer than
the diameter of the main branchial stem.

There are 7 somites in the thorax. The collar is in the form of a broad
wing on either side and on its anterior margin its breadth is equal to the
body width. Its lateral margins slope to meet the body wall at the level
of the 5th setigerous somite. The dorsal collar lobe is distinct from the
lateral ones and extends anteriorly to cover the bases of the branchiae.

The abdominal somites are very short and a noticeable feature is that
on the ventral surfaces of the posterior twenty are ventral shields which are
more prominent in one of the two specimens than in the other.

Setae of the simple type occur in the 1st setigerous somite and are of
two forms. The first (Fig. 37) are very slender, long and sharp pointed;
the second (Fig. 38) are fully four times as broad as the first and are
marked by fine diagonal striations. Similar setae are found in the other
somites throughout the body, but in the abdomen they are very short and
do not protrude far from the body surface. At the extreme posterior end
is a series of setae like those of Fig. 37 in general outline, but very long.
The uncini of thorax and of abdomen are alike, each (Fig. 39) having a
basal knob and 20 fine teeth.

In its annulated opercular stalk this species resembles Schmarda’s
V . annulata described originally from Jamaica, but the description revised
and enlarged by Ehlers (1887, p. 308, PI. 58, Figs. 12-16; PI. 59, Figs. 1-3),
but they differ in the form of the operculum, in the number of the branchiae
and in the shape of the uncini.

The tube has the form characteristic of this genus, a heavy shell
marked by longitudinal lines.

The type was collected at Gurnet’s Rock, 35 feet deep, Aug. 19, 1931
(No. 311,480), and is in the collections of the Department of Tropical Re-
search of the New York Zoological Society. Another specimen was taken
at Castle Harbor, reef, Aug. 16, 1931 (No. 311,347).

66

Zoologica: New York Zoological Society

Bibliography

[XXI :2

Chamberlin, R. V.

1919. The Annelida Polychaeta. Mem. Mus. Comp. Zool. Harvard College,
48, pp. 1-514, Pis. 1-80.

Ehlers, Ernst.

1887. Florida Anneliden. Mem. Mus. Comp. Zool. Harvard College, 15,
pp. i-vi and 1-328, Pis. 1-60.

Fauvel, Pierre.

1914. Resultats des Compagnes Scientifique accomplies sur son yacht par
Albert ler, Prince Souverain de Monaco. Fasc. 46, pp. 1-432, Pis. 1-31.

1916. Resultats des Compagnes Scientifique etc. Fasc. 48, pp. 1-152, Pis. 1-8.

Hoagland, R. A.

1919. Polychaetous Annelids from Porto Rico, the Florida Keys and Ber-
muda. Bull. Am. Mus. Nat. Hist. 41, pp. 571-591, Pis. 29-32.

Kinberg, J. G. H.

1857. Ofvers. af K. Vet. Akad. Forh. pp. 11-14.

1865. Ofvers. af K. Vet. Akad. Forh. pp. 239-258.

1865. Annulata nova. Ofvers. af K. Vet. Akad. Forh. No. 2.

McIntosh, Wm. C.

1885. Report on the Annelida Polychaeta collected by H.M.S. Challenger
during the years 1873-76. Report on the Scientific Results of the
Voyage of H.M.S. Challenger. Vol. 12, pp. i-ii and 1-554, Pis. 1-55,
1A-39A, 1 map, 91 Text-figs.

Southern, R.

1911. Polychaeta of the Coasts of Ireland. 1. The Alciopinae, Tomopterinae
and Typhloscolecidae. Fish. Ireland Sc. Invest. 3, 37 pp. 3 Pis. (Report
for 1910).

Treadwell, A. L.

1906. Polychaetous Annelids of the Hawaiian Islands collected by the
Steamer Albatross in 1902. Bull. U. S. Fish. Commission, for 1903,
pp. 1145-1181, Figs. 1-81.

1911. Polychaetous Annelids from the Dry Tortugas, Florida. Bull. Am.
Mus. Nat. Hist. 30, pp. 1-12, Figs. 1-29.

1921. Leodicidae of the West Indian Region. Pub. Dept. Marine Biol. Car-
negie Inst, of Washington, Vol. 15, pp. 1-131, Pis. 1-9, Text-figs. 1-467.

1928. Polychaetous Annelids from the Arcturus Oceanographic Expedition.
Zoologica, Vol. 8, No. 8, pp. 449-485, Figs. 1-64.

Verrill, A. E.

1900. Additions to the Turbellaria, Nemertina and Annelida of the Ber-
mudas, with Revisions of some New England Genera and Species.
Trans. Conn. Acad. Sciences. Vol. 10. Annelids, pp. 600-669, no Figs.

Webster, H. E.

1884. Annelida from Bermuda collected by G. Brown Goode. Bull. U. S. Nat.
Museum No. 25, pp. 307-327, Pis. 7-12.

1936]

Treadwell: Polychaetous Annelids, Nonsuch Island

67

EXPLANATION OF THE PLATES.
Plate I.

Eunoe purpurea

Fig. 1. Head x 7.

Fig. 2. Sixteenth parapodium x 14.

Fig. 3. Ventral seta x 65.

Fig. 4. Ventral seta x 65.

Fig. 5. Large dorsal seta x 85.

Fig. 6. Elytron x 10.

Drieschia atlantica

Fig. 7 Head x 10.

Fig. 8. Parapodium x 45.

Fig. 9. Ventral seta x 185.

Eupholoe nuda

Fig. 10. Head x 16.

Fig. 11. Parapodium x 45.

Fig. 12. Notopodial seta x 185.

Fig. 13. Neuropodial seta x 185.

Fig. 14. Second form of neuropodial seta x 185.

Plate II.

Nereis bairdii

Fig. 15. Parapodium of heteronereis phase x 45.

Nereis glandulata

Fig. 16. Parapodium from anterior portion of body in heteronereis phase.
Fig. 17. Parapodium from posterior portion in heteronereis phase.

Tomopteris longisetis
Fig. 18. Ventral view of head x 5.

Fig. 19. Parapodium x 5.

Fig. 20. Uncolored rosette x 45.

Fig. 21. Colored rosette x 68.

Audouinea pygidia

Fig. 22. Head x 10.

Fig. 23. Seta x 185.

Ammotrypane bermudiensis

Fig. 24. Anterior end x 5.

Fig. 25. Anal tube x 6.

Fig. 26. Parapodium x 85.

P oly ophthalmus incertus
Fig. 27. Dorsal view of head x 45.

Fig. 28. Ventral view of head x 45.

Fig. 29. Pygidium x 45.

Plate III.

Travisiopsis atlantica

Fig. 30. Head x 23.

Fig. 31. Cirrus from body somite x 45.

Fig. 32. Anal cirrus x 45.

Fig. 33. Seta x 68.

68

Zoologica : New York Zoological Society

[XXI :2

Cirratulus multicirratus

Fig. 34. Head x 10.

Fig. 35. Hook x 185.

Vermilia glandulata

Fig. 36. Operculum x 4.

Fig. 37. Slender seta x 185.

Fig. 38. Larger seta x 185.

Fig. 39. Uncinus x 250.

TREADWELL.

PLATE I.

POLYCHAETOUS ANNELIDS FROM THE VICINITY OF
NONSUCH ISLAND, BERMUDA.

TREADWELL.

PLATE II.

POLYCHAETOUS ANNELIDS FROM THE VICINITY OF

NONSUCH ISLAND, BERMUDA.

TREADWELL.

PLATE III.

POLYCHAETOUS ANNELIDS FROM THE VICINITY OF
NONSUCH ISLAND, BERMUDA.

Jleto gorfe Zoological ^ocictp

General Office: 101 Park Avenue, New York City

Officers

President , Madison Grant

Vice-Presidents , W. Redmond Cross and Kermit Roosevelt
Chairman, Executive Committee, Madison Grant

W. Reid Blair, Director
William T. Hornaday, Director Emeritus
Raymond L. Ditmars, Curator of Mammals and Reptiles
Lee S. Crandall, Curator of Birds
Charles V. Noback, Veterinarian

Claude W. Leister, Ass’t to the Director and Curator, Educational Activities

H. C. Raven, Prosector
Edward R. Osterndorff, Photographer
William Bridges, Editor and Curator of Publications

Aquarium

Charles H. Townsend, Director
C. M. Breder, Jr., Assistant Director

department of tropical &esearcf)

William Beebe, Director and Honorary Curator of Birds
John Tee-Van, General Associate
Gloria Hollister, Research Associate

Treasurer, Cornelius R. Agnew
Secretary, Henry Fairfield Osborn, Jr.

Scientific Staff

Zoological fjarfe

Ctntorial Committee

Madison Grant, Chairman

W. Reid Blair
William Beebe

Charles H. Townsend
George Bird Grinnell

William Bridges

For the information of persons or institutions

RECEIVING ZOOLOGICA:

Beginning with Volume XXI, Zoologica will be published in four quar-
terly Parts. Title page, Contents and Index will be found in Part IV of
each volume.

Beginning with Volume XXI, it will be assumed that each Part of
Zoologica is actually delivered to the person or institution to which it is
addressed. Consequently, no requests can be entertained for gratis copies
to complete files in which Parts are declared to be missing. As long as Parts
are in print, they will be supplied at the price fixed by the cost of printing.

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS
OF THE

NEW YORK ZOOLOGICAL SOCIETY

VOLUME XXI
Part 2

Numbers 3-11

PUBLISHED BY THE SOCIETY
THE ZOOLOGICAL PARK, NEW YORK

July 9, 1936

CONTENTS

Page

3. Bermuda Oceanographic Expeditions. Individual Nets

and Data, 1932-1935. By William Beebe 69

4. Plankton of the Bermuda Oceanographic Expeditions.

I. By G. H. Wailes. (Introduction by William Beebe) 75

5. Plankton of the Bermuda Oceanographic Expeditions.

II. Notes on Protozoa. By G. H. Wailes. (Plates I & II) 81

6. Plankton of the Bermuda Oceanographic Expeditions.

III. Notes on Polychaeta. By Edith Berkeley 85

7. Plankton of the Bermuda Oceanographic Expeditions.

IV. Notes on Copepoda. By Charles Branch Wilson 89

8. Plankton of the Bermuda Oceanographic Expeditions.

V. Notes on Schizopoda. By W. M. Tattersall .. 95

9. Plankton of the Bermuda Oceanographic Expeditions.

VI. Bathypelagic Nemerteans Taken in the Years 1929,
1930 and 1931. By Wesley R. Coe. (Plates I-X; Text-
figure 1) 97

10. Tissue Culture and Explanation in Nature: A Review of

Certain Experiments and Possibilities. By C. M.
Breder, Jr. § 115

11. Preliminary Note on the Nature of the Electrical Dis-

charges of the Electric Eel, Electrophorus elec trie us
(Linnaeus). By C. W. Coates & R. T. Cox. (Text-
figure 1) 125

Beebe: Individual Nets and Data, 1932-1935

3.

Bermuda Oceanographic Expeditions. Individual Nets
and Data, 1932-19351.

William Beebe.

Director, Department of Tropical Research,

New York Zoological Society.

The absence of nets during 1932 was due to constant and intensive
operation of the bathysphere dives.

The lists and data of preceding nets, from Nos. 1 to 1350, together with
an account of previous oceanographic investigations near Bermuda, descrip-
tion of the collecting apparatus, methods of trawling, and details of the
locality chosen for study, are to be found in Zoologica, Volume XIII, Num-
bers 1, 2 and 3.

All foot nets are diatom nets, of No. 20 standard bolting cloth. Other
nets are standard Michel Sars patterns, of 2XX bolting cloth.

Nets marked with an asterisk (*) were drawn in Castle Harbor, inside
of the outer reefs. All others were drawn out at sea.

INDIVIDUAL NETS AND DATA

Type

Depth

Date

Start

Duration

nf TTanl

Direc-

Weather

Wind

Sea

Net

of

tion

No.

Net

Fath-

oms

Me-

tres

1933

of Haul

Hrs.

Mins.

of

Haul

Direc-

tion

Force

1351

Metre

0

0

Sept.

4

8:05 p.m.

20

E

No moon

0

0

Calm

*1352

Foot

0

0

14

7:45

—

45

E

No moon

0

0

Calm

*1353

Foot

0

0

15

7:45

—

50

E

No moon

SE

4

Calm

*1354

Foot

0

0

17

10:00 a.m.

1

0

E

Overcast

SW

5

Choppy

*1355

Foot

0

0

17

7:35 p.m.

1

0

E

No moon

WSW

5

Choppy

*1356

Foot

0

0

17

7:35

1

0

E

No moon

WSW

5

Choppy

*1357

^-Metre

0

0

18

8:00

—

50

W

No moon

SW

3

Choppy

*1358

Foot

0

0

18

8:00

—

50

w

No moon

SW

3

Choppy

*1359

Foot

0

0

18

8:00

—

50

E

No moon

SW

3

Choppy

*1360

Foot

0

0

19

8:00

1

0

E

No moon

SW

5

Choppy

*1361

Foot

0

0

19

8:00

1

0

E

No moon

SW

5

Choppy

*1362

§-Metre

0

0

19

8:00

1

0

E

No moon

SW

5

Choppy

1363

Metre

0

0

20

8:00

1

0

N

No moon

SW

2

Choppy

1364

Foot

0

0

20

8:00

1

0

N

No moon

SW

2

Choppy

*1365

|-Metre

0

0

22

8:00

1

0

E

No moon

WSW

4

Choppy

*1366

Foot

0

0

22

8:00

1

0

E

No moon

WSW

4

Choppy

*136'7

PW

0

0

22

8:00

1

0

E

No moon

WSW

4

Choppy

1 Contribution No. 494, Department of Tropical Research, New York Zoological Society.

[69]

70

Zoologica: New York Zoological Society

[XXI :3

INDIVIDUAL NETS AND DATA ( continued )

Net

Type

Depth

Date

Start

Duration
nf Hcml

Direc-

Weather

Wind

Sea

of

tion

No.

Net

Fath-

oms

Me-

tres

1933

of Haul

Hrs.

Mins.

of

Haul

Direc-

tion

Force

*1368

Foot

0

0

Sept.

23

8:00

30

W

No moon

0

0

Calm

*1369

Foot

0

0

23

8:00

—

30

W

No moon

0

0

Calm

1370

1-Metre

0

0

24

5:00 a.m.

— 5

30

E

No moon

0

0

Calm

1371

Metre

0

0

24

5:00

1

0

E

No moon

0

0

Calm

1372

i-Metre

0

0

25

5:30

—

30

E

Clear

NE

1

Calm

1373

J-Metre

0

0

25

6:00

—

30

E

Clear

NE

1

Calm

1374

Metre

0

0

25

5:30

—

30

E

Clear

NE

1

Calm

1375

Metre

0

0

25

6:00

— |

30

E

Clear

NE

1

Calm

*1376

Foot

0

0

26

7:50

—

35

W

Moonlight

0

0

Calm

*1377

Foot

0

0

26

7:50

—

35

W

Moonlight

0

0

Calm

1378

Metre

0

0

27

4:45

—

20

E

No moon

0

0

Calm

1379

Metre

0

0

27

5:05

—

20

E

No moon

0

0

Calm

1380

Metre

0

0

27

5:25

—

20

E

No moon

0

0

Calm

1381

Metre

0

0

27

5:45

—

20

E

Clear

0

0

Calm

1382

i -Metre

0

0

27

4:45

—

20

E

No moon

0

0

Calm

1383

i-Metre

0

0

27

5:05

—

20

E

No moon

0

0

Calm

*1384

Foot

1

2

27

3:00 p.m.

—

15

W

Clear

0

0

Calm

1385

Metre

0

0

28

4:30 a.m.

— .

20

E

No moon

ESE

3

Choppy

1386

Metre

0

0

28

4:50

—

20

E

No moon

ESE

3

Choppy

1387

Metre

0

0

28

5:10

—

25

E

No moon

ESE

3

Choppy

1388

Metre

0

0

28

5:35

—

25

W

Clear

ESE

3

Choppy

1389

1-Metre

0

0

28

4:30'

—

20

E

No moon

ESE

3

Choppy

1390

|-Metre

0

0

28

4:50

—

20

E

No moon

ESE

3

Choppy

*1391

Foot

1

2

Oct.

2

3:30 p.m.

30

W

Clear

SE

3

Choppy

1392

Foot

0

0

4

7:25

—

20

N

Moonlight

SE

4

Choppy

1393

Foot

0

0

4

7:45

—

15

N

Moonlight

SE

4

Choppy

1394

Foot

0

0

4

8:00

—

25

S

Moonlight

SE

4

Choppy

1395

Foot

0

0

4

8:00

— '

25

S

Moonlight

SE

4

Choppy

*1396

Foot

0

0

7

7:30

—

45

w

No moon

sw

4

Choppy

*1397

Foot

0

0

7

7:30

—

45

w

No moon

sw

4

Choppy

1398

Metre

0

0

9

7:45

—

20

E

No moon

NE

3

Choppy

1399

Metre

0

0

9

8:00

—

20

W

No moon

NE

3

Choppy

1400

Foot

0

0

9

7:45

—

20

E

No moon

NE

3

Choppy

1401

Foot

0

0

9

8:10

—

20

W

No moon

NE

3

Choppy

1402

Metre

0

0

10

8:20

—

20

NxE

No moon

0

0

Calm

1403

Metre

0

0

10

8:40

—

20

NxE

No moon

0

0

Calm

1404

Metre

0

0

10

9:00

—

20

NxE

No moon

0

0

Calm

1405

|-Metre

0

0

10

8:20

—

20

NxE

No moon

0

0

Calm

1406

|-Metre

0

0

10

8:40

—

20

NxE

No moon

0

0

Calm

1407

|-Metre

0

0

10

9:00

—

20

NxE

No moon

0

0

Calm

1408

Metre

0

0

11

7:45

_

20

ExN

No moon

SE

3

Choppy

1409

Metre

0

0

11

8:05

—

20

ExN

No moon

SE

3

Choppy

1410

Metre

0

0

11

8:25

20

WxS

No moon

SE

3

Choppy

1411

1-Metre

0

0

11

7:45

—

20

ExN

No moon

SE

3

Choppy

1412

|-Metre

0

0

11

8:05

—

20

ExN

No moon

SE

3

Choppy

1413

Metre

0

0

12

8:15

—

20

E

No moon

SE

3

Choppy

1414

Metre

0

0

12

8:35

—

20

E

No moon

SE

3

Choppy

1415

Metre

0

0

12

8:55

—

20

W

No moon

SE

3

Choppy

1416

|-Metre

0

0

12

8:15

—

20

E

No moon

SE

3

Choppy

1417

^-Metre

0

0

12

8:35

—

20

E

No moon

SE

3

Choppy

1418

Metre

0

0

14

7:40

|H

20

E

No moon

NExN

2

Calm

1419

Metre

0

0

14

8:03

— .

20

E

No moon

NExN

2

Calm

1420

Metre

0

0

14

8:25

: -L-ii

20

W

No moon

NExN

2

Calm

1421

1-Metre

0

0

14

7:40

—

20

E

No moon

NExN

2

Calm

1422

|-Metre

0

0

14

8:03

—

20

E

No moon

NExN

2

Calm

1423

^-Metre

0

0

14

8:25

-M

20

W

No moon

NExN

2

Calm

1424

Metre

0

0

18

7:40

—

2n

EvN

No moon

NE

1

Calm

1936]

Beebe: Individual Nets and Data, 1932-1935

71

INDIVIDUAL NETS AND DATA ( continued )

Type

Depth

Duration

TTonl

Direc-

Wind

Net

of

Date

Start

tion

Weather

Sea

No.

Net

Fath-

oms

Me-

tres

1933

of Haul

Hrs.

Mins.

of

Haul

Direc-

tion

Force

1425

1-Metre

0

0

Oct.

18

7:40

20

ExN

No moon

NE

1

Calm

1426

Metre

0

0

18

8:05

—

20

WxS

No moon

NE

1

Calm

1427

f-Metre

0

0

18

8:05

—

20

WxS

No moon

NE

1

Calm

1428

Metre

0

0

18

8:30

—

20

WxS

No moon

NE

1

Calm

1429

|-Metre

0

0

18

8:30

—

20

WxS

No moon

NE

1

Calm

*1430

Metre

0

0

20

2:15

—

20

S

Clear

ExN

4

Choppy

*1431

Foot

0

0

20

2:15

—

20

S

Clear

ExN

4

Choppy

*1432

Metre

0

0

20

2:40

—

20

S

Clear

ExN

4

Choppy

*1433

Foot

0

0

20

2:40

—

20

S

Clear

ExN

4

Choppy

*1434

Metre

0

0

20

3:15

—

15

N

Clear

ExN

4

Choppy

*1435

Foot

0

0

20

3:15

—

15

N

Clear

ExN

4

Choppy

*1436

Foot

0

0

21

7:15

—

20

W

Overcast

ESE

5

Choppy

*1437

|-Metre

0

0

25

7:15

—

30

W

Moonlight

SW

2

Calm

*1438

1-Metre

0

0

27

8:00

—

30

w

Overcast

ESE

5

Choppy

1439

Metre

0

0

28

2:15

— .

20

E

Clear

SW

4

Choppy

1440

Foot

0

0

28

2:15

—

20

E

Clear

SW

4

Choppy

1441

Metre

0

0

28

2:40

—

20

W

Clear

SW

4

Choppy

1442

Foot

2i

28

2:40

—

20

W

Clear

SW

4

Choppy

1443

Metre

0

0

28

3:15

—

15

w

Clear

SW

4

Choppy

1444

Foot

21

28

3:15

—

15

w

Clear

SW

4

Choppy

*1445

Foot

0

0

Nov.

1

7:30

30

w

Moonlight

0

0

Calm

1446

Metre

0

0

2

4:45 a.m.

—

20

E

Moonlight

NNE

4

Choppy

1447

Foot

0

0

2

4:45

— ■

20

E

Moonlight

NNE

4

Choppy

1448

Metre

0

0

2

5:10

—

20

E

Moonlight

NNE

4

Choppy

1449

i-Metre

0

0

2

5:10

—

20

E

Moonlight

NNE

4

Choppy

1450

Metre

0

0

2

5:40

— .

20

W

Overcast

NNE

4

Choppy

1451

1-Metre

0

0

2

5:40

—

20

W

Overcast

NNE

4

Choppy

1452

Metre

0

0

11

2:45 p.m.

—

20

E

Overcast

ENE

5

Rough

1453

Metre

0

0

11

3:10

—

20

E

Overcast

ENE

5

Rough

1454

Metre

0

0

11

3:35

— •

20

S

Overcast

ENE

5

Rough

1455

§-Metre

1

2

11

2:45

—

20

E

Overcast

ENE

5

Rough

1456

Foot

1

2

11

3:10

—

20

E

Overcast

ENE

5

Rough

1457

^-Metre

1

2

11

3:35

—

20

S

Overcast

ENE

5

Rough

1458

Metre

0

0

21

7:45

—

25

E

Moonlight

0

0

Calm

1459

Metre

0

0

21

8:15

—

25

E

Moonlight

0

0

Calm

1460

Foot D.

0

0

21

7:45

— .

25

E

Moonlight

0

0

Calm

1461

f-Metre

0

0

21

8:15

—

25

E

Moonlight

0

0

Calm

*1462

VMetre

0

0

1934

May-

18

8:10 p.m.

20

E

Overcast

SE

5

Choppy

*1463

VMetre

0

0

20

9:10 a.m.

—

20

E

Clear

0

0

Calm

*1464

|-Metre

0

0

20

9:35

—

20

E

Clear

0

0

Calm

1465

i-Metre

0

0

21

11:45

—

15

S

Clear

0

0

Calm

1466

1-Metre

0

0

22

3:30 p.m.

—

10

S

Clear

—

—

Rough

*1467

|-Metre

0

0

26

9:15 a.m.

—

15

S

Clear

s

3

Choppy

1468

f-Metre

0

0

26

9:35

—

20

S

Clear

s

3

Choppy

*1469

|-Metre

0

0

31

8:50

—

20

S

Overcast

SW

3

Choppy

*1470

f-Metre

1

2

1

4

31

9:20

—

20

S

Overcast

SW

3

Choppy

*1471

Foot

1

2

31

8:50

—

20

S

Overcast

SW

3

Choppy

*1472

Foot

0

0

31

9:20

—

20

S

Overcast

SW

3

Choppy

1473

i-Metre

0

0

June

1

3:25 p.m.

20

E

Overcast

WxN

3

Swell

1474

1-Metre

1

2

1

3:50

—

20

W

Overcast

WxN

3

Swell

1475

Foot

1

2

1

3:25

—

20

E

Overcast

WxN

3

Swell

1476

Foot

0

0

1

3:50

—

20

W

Overcast

WxN

3

Swell

*1477

|-Metre

0

0

8

10:30 a.m.

—

40

s

Clear

SW

2

Choppy

72

Zoologica: New York Zoological Society

[XXI :3

INDIVIDUAL NETS AND DATA ( continued )

Type

Depth

Duration
of Haul

Direc-

Wind

-

Net

of

Date

Start

tion

Weather

Sea

No.

Net

Fath-

oms

Me-

tres

1934

of Haul

Hrs.

Mins.

of

Haul

Direc-

tion

Force

1478

Foot

0

0

July

2

3:00 p.m.

40

E

Clear

SW

4

Choppy

1479

^-Metre

0

0

2

3:00

—

40

E

Clear

sw

4

Choppy

1480

5-Metre

0

0

4

11:00 a.m.

—

20

S

Clear

0

0

Choppy

1481

|-Metre

0

0

12

3:00 p.m.

—

40

S

Clear

sw

3

Choppy

1482

Foot

0

0

12

3:00

—

40

s

Clear

sw

3

Choppy

1483

Metre

0

0

Aug.

20

10:30 a.m.

30

NW

Clear

0

0

Calm

1484

Metre

0

0

Sept.

7

9:30

20

E

Clear

0

0

Calm

1485

Metre

0

0

7

9:52

—

10

E

Clear

0

0

Calm

1486

Metre

0

0

9

2:40 p.m.

—

20

E

Clear

0

0

Calm

1487

Metre

0

0

9

3:01

—

15

E

Clear

0

0

Calm

1488

Metre

0

0

16

3:00

—

20

E

Clear

0

0

Calm

1489

Metre

0

0

17

2:20

—

20

E

Clear

0

0

Calm

1490

Metre

0

0

18

3:20

—

20

E

Clear

0

0

Calm

1491

Metre

0

0

18

3:44

—

20

E

Clear

0

0

Calm

1492

Metre

0

0

26

3:20

—

20

E

Clear

0

0

Calm

1493

Metre

0

0

26

3:40

—

20

E

Clear

0

0

Calm

1494

Metre

0

0

28

2:40

—

20

SE

Clear

ESE

5

Rough

*1495

Metre

1

2

i

4

Oct.

6

3:15

15

E

Overcast

4

Choppy

1496

Metre

0

0

9

2:45

—

20

SE

Clear

NE

3

Rough

1497

Metre

0

0

9

3:10

—

15

NW

Clear

NE

3

Rough

*1498

5-Metre

0

0

9

8:00

—

20

E

No moon

NE

3

Choppy

1499

Metre

0

0

13

3:30

—

15

E

Clear

SW

3

Rough

1500

Metre

0

0

14

3:25

—

20

E

Overcast

NE

3

Choppy

1501

Metre

400

732

July

25

9:40 a.m.

4

03

SSE

Clear

NNE

4

Choppy

1502

Metre

500

914

25

9:40

4

03

SSE

Clear

NNE

4

Choppy

1503

Metre

600

1097

25

9:40

4

03

SSE

Clear

■ NNE

4

Choppy

1504

Metre

700

1280

25

9:40

4

03

SSE

Clear

NNE

4

Choppy

1505

Metre

800

1463

25

9:40

4

03

SSE

Clear

NNE

4

Choppy

1506

Metre

900

1646

25

9:40

4

03

SSE

Clear

NNE

4

Choppy

1507

Metre

50

92

Aug.

14

12:16 p.m.

2

30

SSE

Clear

SW

2

Calm

1508

Metre

100

183

14

12:16

2

30

SSE

Clear

SW

2

Calm

1509

Metre

200

366

14

12:16

2

30

SSE

Clear

sw

2

Calm

1510

Metre

300

549

14

12:16

2

30

SSE

Clear

sw

2

Calm

1511

Metre

400

732

14

12:16

2

30

SSE

Clear

sw

2

Calm

1512

Metre

500

914

14

12:16

2

30

SSE

Clear

sw

2

Calm

1513

Tangle

1350

2470

14

9:30 a.m.

—

36

SSE

Clear

sw

2

Calm

1514

Metre

0

0

Oct.

15

4:00 p.m.

20

E

Clear

0

0

Calm

1515

Metre

0

0

15

3:35

—

20

E

Clear

0

0

Calm

1516

Metre

0

0

16

3:30

—

20

E

Overcast

0

0

Calm

1517

Metre

0

0

17

4:00

—

30

E

Overcast

NE

3

Rough

1518

Metre

0

0

1935

June

7

9:30 a.m.

44

E

Overcast

E

3

Choppy

1519

Metre

0

0

12

9:40

—

20

S

Clear

SW

3

Swell

1520

Metre

0

0

15

9:10

—

30

E

Clear

s

2

Choppy

1521

Metre

0

0

15

9:50

—

10

W

Clear

s

2

Choppy

*1522

Metre

0

0

19

2:40 p.m.

—

20

W

Clear

sw

3

Choppy

1523

Metre

0

0

24

2:20

—

50

E

Clear

SSE

3

Choppy

1524

Metre

0

0

29

July

3

9:40 a.m.

—

35

E

Clear

E

3

Swell

1525

Metre

0

0

9:10

—

40

E

Overcast

SE •

3

Rough

1936]

Beebe: Individual Nets and Data, 1932-1935

73

INDIVIDUAL NETS AND DATA ( continued )

Type

of

Depth

Duration
of Haul

Direc-

Wind

Net

Date

Start

tion

Weather

Sea

No.

Net

Fath-

oms

Me-

tres

1935

of Haul

Hrs.

Mins.

of

Haul

Direc-

tion

Force

1526

Metre

0

0

July

7

2:10 p.m.

20

S

Clear

SE

3

Choppy

—1527

Metre

0

0

9

8:20

—

25

E

Moonlight

SE

2

Calm

1528

|-Metre

0

0

12

11:15 a.m.

—

10

S

Clear

sw

3

Choppy

1529

^-Metre

100

183

15

2:36 p.m.

1

0

S

Clear

SE

3

Choppy

1530

Metre

0

0

15

4:00

—

10

s

Clear

SE

3

Choppy

1531

Metre

0

0

19

10:11 a.m.

—

15

s

Clear

SE

3

Rough

1532

|-Metre

125

228

20

9:19

1

0

SE

Clear

SE

2

Calm

1533

Metre

0

0

20

10:26

—

10

SE

Clear

SE

2

Calm

1534

1-Metre

225

410

22

2:45 p.m.

1

0

SExS

Clear

SE

1

Calm

1535

Metre

0

0

22

4:14

—

10

SExS

Clear

SE

1

Calm

1536

^-Metre

455

828

23

9:28 a.m.

1

30

SE

Clear

SE

1

Calm

1537

Metre

0

0

23

9:45

1

05

SE

Clear

SE

1

Calm

1538

|-Metre

300

549

24

2:25 p.m.

1

02

SE

Clear

SE

1

Calm

1539

Metre

0

0

24

4:04

—

10

SE

Clear

SE

1

Calm

1540

Metre

0

0

Aug.

4

2:50

20

E

Clear

SE

4

Choppy

1541

xMetre

0

0

7

2.00

—

25

E

Clear

SE

2

Calm

1542

Metre

0

0

24

2:30

—

24

S

Overcast

SW

3

Rough

1543

Metre

0

0

26

2:30

—

23

E

Overcast

SW

3

Choppy

1544

Metre

0

0

27

9:25 a.m.

—

30

E

Squally

SW

3

Choppy

1545

Metre

0

0

28

10:00

—

30

E

Clear

S

1

Calm

1546

Metre

0

0

31

10:00

—

27

E

Clear

S

4

Rough

1547

Metre

0

0

Sept.

1

5:00

24

E

Clear

SSE

2

Choppy

1548

Metre

0

0

1

5:32

—

12

E

Clear

SSE

2

Choppy

1549

Metre

0

0

3

9:00

—

30

E

Clear

s

2

Calm

1550

Metre

0

0

7

2:00 p.m.

— .

30

E

Clear

N

4

Rough

1551

Metre

0

0

9

9:10 a.m.

—

30

E

Clear

SW

4

Choppy

1552

Metre

0

0

13

9:00

—

20

E

Clear

SW

4

Rough

1553

Metre

0

0

13

9:45

—

10

W

Clear

sw

4

Choppy

1554

Metre

0

0

16

2:20 p.m.

—

30

S

Clear

sw

2

Choppy

*1555

Metre

0

0

16

3:10

—

10

N

Clear

sw

2

Calm

1556

Metre

0

0

17

9:10 a.m.

— .

30

E

Clear

NW

2

Choppy

1557

Metre

0

0

20

7:55 p.m.

—

30

E

No moon

SE

4

Choppy

1558

Metre

0

0

22

5:05 a.m.

v

30

E

Moonlight

0

0

Calm

1559

, Metre

0

0

28

9:27

—

10

W

Overcast

SW

2

Choppy

1560

Metre

1

2

28

9:15

—

10

E

Overcast

SW

2

Choppy

1561

Metre

0

0

28

9:00

—

10

E

Overcast

sw

2

Choppy

1562

Metre

0

0

Oct.

2

9:23

10

E

Overcast

ssw

4

Rough

1563

Metre .

1

2

2

9:10

—

10

W

Overcast

ssw

4

Rough

1564

Metre

0

0

6

2:05 p.m.

—

10

E

Overcast

sw

2

Swell

1565

Metre

0

0

6

2:17

—

10

E

Overcast

sw

2

Swell

1566

Metre

1

2

6

2:30

—

10

W

Overcast

sw

2

Swell

1567

1-Metre

0

0

12

11:21 a.m.

— 1

10

SxE

Clear

sw

2

Calm

1568

~2 -Metre

1.8

3

12

11:08

—

10

SxE

Clear

sw

2

Calm

1569

!-Metre

5

9.2

12

10:15

_

10

SxE

Clear

sw

2

Calm

1570

|-Metre

8.2

15.3

12

10:28

—

10

SxE

Clear

sw

2

Calm

1571

5-Metre

12.5

23

12

10:55

—

10

SxE

Clear

sw

2

Calm

1572

5-Metre

17

31

12

10:41

—

10

SxE

Clear

sw

2

Calm

1573

5-Metre

0

0

12

11:52

—

10

NNW

Clear

sw

2

Calm

*1574

5-Metre

0

0

12

12:20

—

10

N

Clear

sw

1

Calm

Wailes: Plankton of the Bermuda Oceanographic Expeditions

75

4.

Plankton of the Bermuda Oceanographic Expeditions1. I.
G. H. Wailes.

City Museum, Vancouver, B. C.

INTRODUCTION BY WILLIAM BEEBE.

The matter in this series of papers is a continuation of the oceano-
graphic work carried on by myself and my staff of the Department of
Tropical Research of the New York Zoological Society, off the island of
Nonsuch, Bermuda. Full details of this investigation, together with com-
plete data of hauls, may be found in Zoologic A, Volume XIII, Numbers 1,
2 and 3, and Volume XXI, Number 3. I have thought it worth while to add
a few paragraphs of the most pertinent data.

Locality.

The area in which the fifteen hundred-odd nets have been drawn is
roughly circular and eight miles in diameter. Observations by means of
the two light-houses, Gibb’s Hill and St. Davids, have made it possible to
get accurate sights at the beginning and end of each individual haul. To
give the location with more exactness : the eight-mile circle under considera-
tion has its center at 32° 12' N. Lat. and 64° 36' W. Long., which point
is 160 degrees by the compass, or south-south-east of Nonsuch. Its hori-
zontal boundaries are as follows:

Northern rim:
Southern rim:
Eastern rim:
Western rim:

32° 16' N. Lat.

32° 8' N. Lat.

64° 31' 20" W. Long.
64° 40' 40" W. Long.

At no place is its bottom less than 1,000 fathoms in depth. It slopes
rather rapidly in the northeastern corner to 1,357 fathoms, and along its
southern border is between 1,400 and 1,500 fathoms deep. My first deep
dive in the bathysphere, of 803 feet, was in the southwest sector, and the
later ones of 1,426, 2,200, 2,510 and 3,028 feet were all near the northern
rim.

Methods.

The deep-sea trawling has all been done from the aft deck of the tug
Gladisfen. When the rim of the imaginary cylinder has been reached the
weight at the end of the wire is put overboard and the cable gradually paid
out at a 30° angle. Six nets are usually attached, 1-metre nets of standard
oceanographic design, 20 feet in length, with the posterior portion of No. 2

1 Contribution No. 495, Department of Tropical Research, New York Zoological Society.

76

Zoologica: New York Zoological Society

[XXI :4

bolting silk and the upper part of No. 0 bolting silk. In addition to this there
is the usual collar of shrimp netting, with 10 mm. mesh.

These nets are placed along the cable at measured intervals so that at
a 30° angle they will haul 100 fathoms apart, usually from the surface to
500 fathoms, or from 500 to a depth of 1,000 fathoms. After as long a haul
as the time will permit the nets are pulled in and the contents taken ashore
to the laboratory for sorting and study.

The exact depth at which the nets are drawn is assured by the use
of a special bathygraph pressure gauge which registers the entire course of
the lowest net from surface to greatest depth and back to surface again,
together with the duration of the haul.

All of the deep-sea hauls are made with horizontal tow nets unprovided
with closing apparatus. After the first few hauls were made off Bermuda at
deep levels, it was found advisable, owing to the relatively small amount
of captured animal life, to extend the length of towing time for the nets
up to periods of four to six hours. Experience on the Arcturus and other
expeditions had taught us that closing mechanisms operating at great
depths were uncertain pieces of machinery at best, and quite useless for
gathering an adequate representation of the abyssal fauna. The necessity of
long horizontal hauls has completely precluded their use.

Two phases of the Bermuda Oceanographic Expedition hauls, seldom if
ever duplicated by other expeditions, are exceptionally interesting. The first
of these is the accomplishment of hundreds of hauls at various levels in one
location and over long periods of time, and the second is the long duration
of each individual haul at its level compared with the time spent in going
up and down. These two conditions make it relatively easy, with experience,
to determine the life zones in the sea to which each organism belongs.

Complete time records of all of the nets used during the Bermuda
Oceanographic Expeditions have been kept, and the following data show
the percentage of time spent by a net at the desired depth compared with
the total time the net was in the water :

100 -fathom nets : These nets towed at this depth from 86% to 96%
of their total towing time, their average effectiveness being 91.4%.

500 -fathom nets : The nets towed at this depth from 70% to 83% of
their total towing time, their average effectiveness being 79.4%.

1,000 -fathom nets: The nets towed at this depth from 59% to 72%
of their total towing time, their average effectiveness being 64.4%.

The exact determinations of the life zones of animals of the Bermuda
deep-sea fauna, have been restricted, so far, to fishes. These 'are now being
studied in the cases of the larger invertebrates and plankton.

Contents of a Typical Net Haul.

Net 779, Michael Sars metre net; Vessel, Tug Gladisfen; Date, July 5,
1930; Location, 5 m. SE to 15 m. SExS%S of Nonsuch Island, Bermuda;
Weather, overcast in A.M., rain squalls, clearing in P.M. ; Wind, SW, 2 to 3;
Sea, considerable swell, diminishing in P.M.; Time, beginning of haul 9:04
A.M., ending of haul 2:25 P.M., duration of haul, 5 hours, 21 minutes;
Depth of net, 800 fathoms; Length of cable, 2,925 metres; Angle of cable,
30°.

Contents of Net.

General Character : Mostly copepods, schizopods and sagitta.

Copepods, comprising a dozen or more species, mostly calanoids, with also
Corycaeus, Oithona, etc.

Schizopods, chiefly small species of Euphausia, with a dozen others belong-
ing to two or three genera.

1936] Wailes: Plankton of the Bermuda Oceanographic Expeditions 77

Shrimps, one specimen near Pandalus danae.

Ostracods, a few of one or two species.

Amphipods, few and small, a dozen individuals of four or five species.
Sagitta, apparently two or three species.

Polychaetes, one Tomopteris septentrionale.

Siphonophores, Diphys truncata.

Sponges, fair number of spicules of various kinds.

Radiolaria, large numbers of portions of a hexagonal framework and a few
small, conical specimens mostly incomplete ; numbers of perforated
spherical species and Astrophaeroidea.

Diatoms, almost entirely absent. One specimen of Asteromphalus heptactis,
also one cell of Melosira moniliformis and a Coscinodiscus.
Tintinnoinea, one Tintinnopsis cylindrica and one Parafavella near P. acuta.
Foraminifera, few, comprising two or three species.

Dinoflagellates, absent.

Fish larvae, three present, one belonging to a deep-sea specids, with very
large lower jaw and black spots on sides.

Fish, adolescent and adult:

6 Lampanyctus warmingi, 11 to 21 mm.

3 Myctophum benoiti, 11, 12 mm.

1 Lampadena chavesi.

13 Myctophum laternatum, 12 mm.

14 Cyclothone signata.

117 Cyclothone microdon.

1 Cyclothone pallida.

1 Bregmaceros macclellandii, 45 mm.

2 Omosudis lowi, 11, 38 mm.

1 Stomias ferox, 80 mm.

1 Lestidium intermedium , 87 mm.

In the preliminary work of this report, represented by Zoologica, Vol-
umes XXI, Number 4 to 8 inclusive, the labors of Mr. G. H. Wailes have
been very great, for he has patiently sorted out the various elements of
many typical plankton hauls made in the course of these expeditions, and
thus rendered the various phyla available for study by specialists.

This has resulted in a real contribution to the distribution and relative
abundance of marine life in the vicinity of Bermuda, but it is only a be-
ginning, for of the fifteen hundred-odd hauls made, from 1929 to the present
time, selections for the following reports have been made from only 44, or
about 3%.

Samples of marine plankton collected by Dr. William Beebe off Ber-
muda have been submitted to me and form the subject of the present re-
ports. Further reports will be issued as the material is identified and re-
ported upon.

I wish to. express my thanks to Dr. Beebe for the privilege of examin-
ing the material and to all those who have kindly identified specimens, among
whom I would especially mention the following for their courtesy in identi-
fying specimens in the groups on which they are acknowledged authorities :
Mrs. Edith Berkeley and Mr. C. C. A. Monro (Polychaetes), Dr. McLean
Fraser (Hydroids), Mr. Clarence R. Shoemaker (Amphipods), Dr. C. B.
Wilson (Copepods) and Dr. W. M. Tattersall (Schizopods) , also Dr. W. A.
Clemens for library facilities at the Departure Bay Station of the Bio-
logical Board of Canada.

78

[XXI :4

Zoologica: New York Zoological Society

Locality.

The samples numbered from 1 to 13 were obtained in 1930 from an
area eight miles in diameter, situated about eight miles south of Nonsuch
Island, the center of this area being in 32° 12' N. Lat. and 64° 36' W. Long.

Samples numbered from 14 to 24 were taken during the autumn of 1933
somewhat farther inshore.

Table I.
Data on hauls.

Sample

No.

Net

No.

Depth

Fath. Metres

Date

1930

Time of
Haul

Duration
of Haul

Type of
Net

1

779

800

1463

5

July

9.04 A.M.

5 hrs.

1

metre

21 min.

(870

100

183

11

Sept.

10.03 A.M.

2-27

66

66

Z

1876

100

183

12

Sept.

11.22 A.M.

3-00

“

66

(871

200

366

11

Sept.

10.03 A.M.

2-27

“

u

o

1877

200

366

12

Sept.

11.22 A.M.

3-00

u

u

(872

300

549

11

Sept.

10.03 A.M.

2-27

a

u

4

1878

300

549

12

Sept.

11.22 A.M. .

3-00

u

u

(873

400

732

11

Sept.

10.03 A.M.

2-27

u

u

0

1879

400

732

12

Sept.

11.22 A.M.

3-00

“

u

(874

500

914

11

Sept.

10.03 A.M.

2-27

u

u

o

1880

500

914

12

Sept.

11.22 A.M.

3-00

a

“

1875

600

1097

11

Sept.

10.03 A.M.

2-27

u

66

<

1881

600

1097

12

Sept.

11.22 A.M.

3-00

u

u

[891

700

1280

15

Sept.

9.20 A.M.

4-00

u

66

8

■] 896

700

1280

16

Sept.

9.29 A.M.

4-26

“

66

[902

700

1280

17

Sept.

9.10 A.M.

4-05

u

66

[892

800

1463

15

Sept.

9.20 A.M.

4-00

u

66

9

j 897

800

1463

16

Sept.

9.29 A.M.

4-26

“

66

[903

800

1463

17

Sept.

9.10 A.M.

4-00

u

66

[868

900

1646

10

Sept.

9.38 A.M.

2-52

u

66

10

\ 886

900

1646

13

Sept.

9.33 A.M.

4-27

u

66

[887

900

1646

13

Sept.

9.33 A.M.

4-27

u

66

[858

1000

1829

6

Sept.

9.07 A.M.

4-23

u

66

11

864

1000

1829

8

Sept.

9.29 A.M.

4-01

u

66

[869

1000

1829

10

Sept.

9.38 A.M.

2-52

u

66

12

900

0

0

16

Sept.

8.00 A.M.

1-00

u

66

13

976

0

0

12

Oct.

2.00 A.M.

2-00

u

66

Surface Hauls 1933

14

1358

0

0

17

Sept.

7.35 P.M.

1-00

1

ft. Diatom

15 1370

0

0

24

Sept.

5.00 A.M.

0-30

V2 -Metre

16

1378

0

0

27

Sept.

4.45 A.M.

0-20

1

Metre

17

1382

0

0

27

Sept.

4.45 A.M.

0-20

1

Metre

18 i

1380

0

0

27

Sept.

5.25 A.M.

0-20

1

Metre

18 1

1381

0

0

27

Sept.

5.45 A.M.

0-20

1

Metre

19

1408

0

0

11

Oct.

7.45 P.M.

0-20

1

Metre

20 1411

0

0

11

Oct.

7.45 P.M.

0-20

V2 -Metre Diatom

f:

1439

0

0

28

Oct.

2.15 P.M.

0-20

1

Metre

21 ]:

1441

0

0

28

Oct.

2.40 P.M.

0-20

1

Metre

l

1443

0

0

28

Oct.

3.15 P.M.

0-15

1

Metre

99 l

1442

1.6

3

28

Oct.

2.40 P.M.

0-20

1

ft. Diatom

1444

1.6

3

28

Oct.

3.15 P.M.

0-15

1

ft. Diatom

r

1446

0

0

2

Nov.

4.45 A.M.

1-15

1

Metre

90 ]

1448

0

0

2

Nov.

6.00 A.M.

1-15

1

Metre

Zo 'j

1449

0

0

2

Nov.

4.45 A.M.

1-15

V2 -Metre

l

1450

0

0

2

Nov.

6.00 A.M.

1-15

V2 -Metre

24

1457

1.6

3

11

Nov.

3.35 P.M.

0-20

V2 -Metre

1936] Wailes: Plankton of the Bermuda Oceanographic Expeditions 79

From Table I it will be seen that each sample represents the combined
results of from one to four separate net hauls, a total of 44 hauls being
included out of more than fifteen hundred made by Dr. Beebe up to the
present time.

General Composition of the Plankton.

Previous to the samples being received, in fact on the day of their col-
lection, the larger organisms such as fish, large crustaceans and medusae,
had been removed. The following remarks apply, therefore, to the re-
mainder.

Deep-water Hauls: These, comprising Samples 1 to 11, were large, each
consisting of from 400 to 500 cubic centimetres of organisms. All had mixed
with them considerable quantities of long hairs and gelatinous material of
unrecognized origin, much entangled with chaetognaths and crustaceans.

All the samples wei;e of a similar general character and consisted of
approximately 80% Crustacea, 10% Chaetognatha, 5% Coelenterata and 5%
of various other organisms and debris.

The crustacean portions were composed on an average of about (by
bulk) 40% schizopods, 35% shrimps and larvae, and 20-25% copepods, with
a small mixture of amphipods, ostracods, isopods, crab zoea and other
decapod larvae amounting to 1% or 2%.

Shallow-water Hauls: These were small in amount, varying from about
5 to 20 cubic centimetres each. The surface tows (Samples 12 and 13, Nets
900 and 976) were also small as they consisted only of portions of the hauls.

The surface hauls as represented by Samples 14 to 24 show great varia-
tion in the relative proportions of the various groups composing them, as
can be seen from Table II.

Table II.

Estimated percentages (by bulk) of the organisms comprising samples 14 to
24. (Those present but not equal to 1% are indicated by “x”).

Time

Pre-daylight Hauls,
4:45 to 6:00 A.M.

Daylight Hauls,
2:15 to 3:35 P.M.

Post-daylight
Hauls, 7 to
8:00 P.M.

Sample No

15

16

17

18

19

23

21

22

24

14

20

Copepoda

87

25

38

5

25

82

2

3

25

15

22

Shrimps and Larvae.. . .

10

72

50

60

37

10

92

96

68

70

2

Crab Zoea

2

2

10

2

2

1

2

1

2

X

1

Squilla Larvae

—

—

—

X

—

—

1

—

1

—

—

Amphipoda

—

X

1

30

1

1

—

—

—

X

—

Isopoda

1

—

—

1

X

1

—

—

—

—

—

Ostracoda

X

X

X

X

X

X

X

X

X

X

X

Polychaeta

X

—

—

■

X

—

X

—

—

10

—

Chaetognatha

X

—

1

2

30

1

—

— ■

2

5

70

Siphonophora

—

—

—

—

3

3

1

—

1

—

5

Medusae

—

—

X

1

1

X

X

—

— .

X

Tintinnoinea

X

X

—

—

—

—

— .

—

X

—

Debris, etc

—

1

—

—

1

X

2

—

1

—

X

These differences occur even in the case of hauls taken simultaneously,
as were Nos, 16 and 17, and are still more accentuated in No. 18 which was
taken immediately afterward. This latter sample was the only one in which
amphipods were present in larger quantities than 1%. In this instance the
species Synopia ultramarina formed 30% of the catch.

80

Zoologica: New York Zoological Society
General Results.

Nearly all of the identifications are new records for this region and
extend the known distributions more or less considerably.

Aside from the pelagic organisms which are generally distributed in
the oceans, the collection consists largely of species previously recorded from
the tropical and eastern Atlantic and the Mediterranean areas.

An unexpected result of the examination of the material was the almost
complete absence of diatoms and dinoflagellates. This may be due to Ber-
muda being situated in the Sargasso sea area where certain types of plank-
ton are less plentiful than elsewhere.

Wailes: Notes on Protozoa

81

5.

Plankton of the Bermuda Oceanographic Expeditions.
II. Notes on Protozoa1.

G. H. Wailes.

City Museum, Vancouver, B. C.

(Plates I & II).

This is one of a number of papers dealing with the planktonic contents
of a selected series of nets drawn at various levels off the coast of Bermuda
on the Bermuda Oceanographic Expeditions of the New York Zoological
Society under the direction of Dr. William Beebe. Full details of the nets,
locality, etc., will be found in Zoologica, Volume XIII, Numbers 1, 2 and 3,
and Volume XXI, Numbers 3 and 4.

The following 39 species of Protozoa were taken in these nets:

PHYLUM PROTOZOA.

Subphylum Mastigophora.

Class Phytomastigoda.

Order Chrysomanadida.

Family Silicoflagellidae.

Dictyocha fibula (Ehrenberg). Rare.

Order Dinoflagellida. (Dinoflagellates) .

Suborder Diniferina.

Family Noctilucidae.

Noctiluca scintillans (McCartney). Rare.

Family Peridinidae.

Goniodoma polyhedricum (Pouchet).
Gonyaulax digitate (Pouchet).
Peridiniopsis asymmetrica Mangin.
Peridinium cerasus Paulsen.

Peridinium claudicans Paulsen.
Peridinium conicum (Gran).

Peridinium grani Ostenfeld.

Peridinium oblongum (Aurivillius) .
Ceratium fusus (Ehrenberg).

Ceratium trichocerus (Ehrenberg).
Ceratium tripos var. atlantica Ostenfeld.
Ceratium karsteni Pavillard.

1 Contribution No. 496. Department of Tropical Research, New York Zoological Society.

82 Zoologica: New York Zoological Society [XXI :5

Suborder Adinina.

Family Prorocentridae.

Prorocentrum micans Ehrenberg.

The dinoflagellates live in the upper layers of the ocean; most of the
species recorded above were captured in surface hauls (Nets 1471, 1472,
1475 and 1476). They are all generally distributed in the North Atlantic
ocean, none being purely tropical species.

Subphylum Sarcodina.

Class Actinopoda.

Subclass Radiolaria.

The Radiolaria occurred very sparsely in the deep-water hauls and the
majority of individuals were mutilated or fragmentary. Probably nearly
twenty species are represented. The material awaits further study.

Subphylum Infusoria.

Class Ciliata.

Order Heterotricha.

Suborder Tintinnoinea.

Examination of Haul 779 made in 800 fathoms and upward, 8 miles
south of Bermuda, on July 5, 1930, afforded the following list of 22 species,
of Tintinnids, except Tintinnopsis bermudensis.

No species occurred numerously and all (except Parundella major)
have been recorded previously from either the tropical Atlantic, including
the Sargasso sea, North Atlantic or Mediterranean sea.

Parundella major was first described in 1925 from the Strait of Georgia,
British Columbia, and has also been recorded off San Francisco.

Tintinnopsis bermudensis occurred fairly numerously and was the only
species observed in Surface Hauls 1355 and 1370 taken on Sept. 17 and 24,
1933 (close to the shore of Bermuda over depths of five fathoms or less).
In 16 similar hauls taken from Sept. 27 to Nov. 11, 1933, no species of tin-
tinnid was seen.

In the accompanying plates figures are given of the species as here
recorded.

An examination of 24 hauls made during September, 1930, at depths
from 100 fathoms to 1,000 fathoms, 8 miles off Bermuda, disclosed no tin-
tinnids as present. Neither were any present in two surface hauls made at
the same time and place.

For information on the synonymy and distribution of the species, the
conspectus of this suborder by Kofoid and Campbell (University of Califor-
nia Press, Vol. 34, pp. 1-403, 697 Figs, in text, 1929) should be consulted.

Family Codonellidae Kofoid and Campbell.

Tintinnopsis Stein emended.

T. bermudensis Brandt. (Fig. 1).

T. cylindrica Daday. (Fig. 2). Generally distributed.

T. major Meunier. North Atlantic.

Codonella Haeckel emended.

C. amphorella Biedermann. (Fig. 5).

C. angusta Kofoid and Campbell. Sargasso sea.

C. apicata Kofoid and Campbell (Fig. 4).

C. nationalis Brandt. (Fig. 7). Atlantic ocean.

1936]

Wailes: Notes on Protozoa

83

C. oceanica Brandt emended. (Fig. 8). Gulf Stream.

C. rapa Kofoid and Campbell. (Fig. 6).

C. recta Kofoid and Campbell. Agulhas Current.

Family Codonellopsidae Kofoid and Campbell.

Stenosemella Jorgensen.

S. ventricosa (Claparede and Lachmann). (Fig. 3).
Codonellopsis Jorgensen.

C. longa Kofoid and Campbell. (Fig. 17).

C. tessellata (Brandt). (Fig. 16). Sargasso sea.

Family Cyttarocylidae Kofoid and Campbell.

Cyttarocylis Fol emended.

C. magna Brandt. (Fig. 15).

C. plagiostoma (Daday). (Fig. 18). Atlantic ocean.

Family Ptychocylidae Kofoid and Campbell.

Epiplocylis Jorgensen.

E. sargassensis (Brandt) emended. (Fig. 9).

Family Xystonellidae Kofoid and Campbell.

Parundella Jorgensen emended.

P. major Wailes. (Fig. 20). Off west coast of North America.

Family Undellidae Kofoid and Campbell.

Proplectella Kofoid and Campbell.

P. acuta Jorgensen. (Fig. 11). Mediterranean sea.

P. claparedei (Entz Sr.). (Fig. 10).

Family Bictyocystidae Haeckel emended.

Dictyocysta Ehrenberg emended.

D. dilatata Brandt. (Figs. 12, 13). Sargasso sea.

D. lata Kofoid and Campbell. (Fig. 14). Sargasso sea.

Family Tintinnidae Claparede and Lachmann emended.

Tintinnus Schrank emended.

T. macilentus Jorgensen emended. (Fig. 19). North Atlantic,

New Zealand.

Order Peritrichida.

Family Vorticellidae.

Cothurnia imberbis Ehrenberg.

Class Suctoria.

Family Acinetidae.

Acineta tuberosa Ehrenberg

The last two species were attached to floating alga. They are littoral
forms and generally distributed.

84

Zoologica: New York Zoological Society

EXPLANATION OF THE PLATES.

Note: All Figures magnified 375 times.

Plate I.

Fig. 1. Tintinnopsis bermudensis Brandt. Total length 84-94/*, greatest diam-
eter 60-68/*, diameter of collar 45/*.

Fig. 2. Tintinnopsis cylindrica Daday. Diameter 38-40/*, length 140-150/*.

Fig. 3. Stenosemella ventricosa (Claparede & Lachmann). Length 68-83/*,
diameter 55-68/*.

Fig. 4. Codonella apicata Kofoid & Campbell. Length 75/*, greatest diameter
59/*, diameter of collar 42/*, height of collar 16/*.

Fig. 5. Codonella amphorella Biedermann. Length 89-98/*, greatest diameter
55/*, length of horn, 18-26/*, length of collar 22-23/*, diameter of neck
36-40/*.

Fig. 6. Codonella rapa Kofoid & Campbell. Length 90-96/*, greatest diameter
49-50/*, length of horn 20/*, length of collar 23/*.

Fig. 7. Codonella nationalis Brandt. Length 74-84/*, greatest diameter 55-64/t,
length of collar 19-20/*.

Fig. 8. Codonella oceanica Brandt emended. Length 84/*, greatest diameter 66/*.

Fig. 9. Epiploeylis sargassensis Brandt emended. Length 132/*, greatest diam-
eter 70 /*.

Fig. 10. Proplectella claparedei (Entz Sr.). Length 58-65 /x, greatest diameter
40-42 /*, aperture 36/*.

Fig. 11. Proplectella acuta Jorgensen. Length 68/*, diameter 39/*, aperture 33/*.

Figs. 12, 13. Dictyocysta dilatata Brandt. Length 61-65/*, diameter 40-45/*,
diameter of collar 34-39/*, height of collar 16/*.

Fig. 14. Dictyocysta lata Kofoid & Campbell. Length 65 1±, diameter of bowl 53/*,
diameter of collar 50/*, height of collar 30/*.

Plate II.

Fig. 15. Cyttarocylis magna Brandt. Length 268 /*,, greatest diameter 130/*.

Fig. 16. Codonellopsis tessalata (Brandt). Length of bowl and neck 80/*, diam-
eter of bowl 68-70 ix, diameter of neck 42-52/*, length of horn 32-42/*,
total length up to about 225/*.

Fig. 17. Codonellopsis longa Kofoid & Campbell. Total length 235/*, greatest
diameter of bowl 65/*.

Fig. 18. Cyttarocylis plagiostoma (Daday). Length 106-123/*, diameter of aper-
ture 105-117/*.

Fig. 19. Tintinnus macilentus Jorgensen emended. Length 160/*, oral aperture
38ju, diameter, aboral aperture 24/* diameter.

Fig. 20. Parundella major Wailes. Diameter 30-32/*, length 130-136/*.

WAILivS.

PLATE I.

NOTES ON PROTOZOA ,

%

Y~7

NOTKS OI\ PROTOZOA

Berkeley: Notes on Protozoa

85

6.

Plankton of the Bermuda Oceanographic Expeditions.

III. Notes on Polychaeta1.

Edith Berkeley.

Pacific Biological Station, Nanaimo, B. C.

This is one of a number of papers dealing with the planktonic contents
of a selected series of nets drawn at various levels off the coast of Bermuda
on the Bermuda Oceanographic Expeditions of the New York Zoological
Society under the direction of Dr. William Beebe. Full details of the nets,
locality, etc., will be found in Zoologica, Volume XIII, Numbers 1, 2 and 3,
and Volume XXI, Numbers 3 and 4.

The identifications in the following list were made from specimens sub-
mitted to me by G. H. Wailes.

Certain members of Syllidae and Nereidae have pelagic swarming
forms, which accounts for their capture in the plankton. This is also true
of the members of the family Opheliidae included in the following list. The
members of the genera Lopadorhynchus, Vanadis, Alciopa, Corynocephalus,
Travisiopis, and Tomopteris are true pelagic forms. The genus Nectochaeta
is bathypelagic.

All the species here enumerated are generally distributed in the Atlantic
ocean.

PHYLUM ANNELIDA.

Class Chaetopoda.

Order Polychaeta.

Family Phyllodocidae Grube.

Lopadorhynchus Grube.

L. nationalis Reibisch.

L. uncinatus Fauvel.

Family Alciopidae Ehlers.

Vanadis Claparede.

V. longissima (?) (Levinsen).

V. formosa Claparede.

Alciopa Audouin et M. Edwards.

A. cantraini (Della Chiaje).

Corynocephalus Levinsen.

C. albo-maculatus Levinsen.

Family Typhloscolecidae Uljanin.

Travisiopsis Levinsen.

T. lobifera Levinsen.

1 Contribution No. 497, Department of Tropical Research, New York Zoological Society.

86

Zoologica: New York Zoological Society

[XXI :6

Family Tomopteridae Grube.

Tomopteris Eschscholtz.

T. apsteini Rosa.

T. nisseni Rosa.

Family Opheliidae Grube.

Polyophthalmus Quatrefages.

P. pictus (Dujardin).

Armandia Filippi.

A. poly ophthalmia Kukenthal.

Family Leodicidae Treadwell.

Nematonereis Schmarda.

iV. unicornis (?) (Grube). A very small
incomplete specimen.

Family Syllidae Grube.

Grubea Quatrefages.

G. clavata Claparede.

Family Polynoidae.

Nectochaeta Marenzeller.

N. caroli Fauvel.

From a portion of the material the following additional species were
identified by C. C. A. Monro (British Museum).

Family Nereidae Quatrefages.

Nereis Cuvier.

N. riisei Grube. Surface (Net 900).

Ceratonereis Kinberg.

C. mirabilis Kinberg. Surface (Net
900, September to November).

Perinereis Kinberg.

P. bairdii Webster. Surface (September
to November).

P. sp. Surface (Net 900).

Family Glyceridae Grube.

Glycera Savigny.

vG. tesselata Grube. Surface.

Family Polynoidae Kinberg.

Nectochaeta Marenzeller.

N. grimaldii Marenzeller. 1,000 fathoms and up.
Harmothoe Kinberg.

H. benthophila Ehlers. 1,000 fathoms and up.

Family Spionidae Sars.

Chaetosphaera Haecker.

C. sp. Haecker. September to October. Surface (Net
1370). Post-larvae.

Mr. Monro writes that all the forms in the above list identified by him
are epitocons, sexually modified forms, except the polynoids Nectochaeta
grimaldii and Harmothoe benthophila , which probably represent a phase in
the life of some well known benthic species. Chaetosphaera is a post-
larva of which the adult is not known.

1936] Berkeley: Notes on Protozoa 87

Selected Literature.

Apstein, C., Die Alciopiden und Tomopteriden der Plankton-Expedition.
(Kiel). 1900.

Reibisch, J., Die Pelagischen Phyllodicen und Typhloscoleciden der
Plankton-Expedition, Vol. 2. (Kiel). 1895.

Fauvel, P., Faune de France, Vol. 5 (Polychetes errantes), 1923, and
Vol. 16 (Polychetes sedentaires), 1927.

Wilson: Notes on Copepoda

89

7.

Plankton of the Bermuda Oceanographic Expeditions.

IV. Notes on Copepoda1.

Charles Branch Wilson.

State Teachers College, Westfield, Massachusetts.

This is one of a number of papers dealing with the planktonic contents
of a selected series of nets drawn at various levels off the coast of Bermuda
on the Bermuda Oceanographic Expeditions of the New York Zoological
Society under the direction of Dr. William Beebe. Full details of the nets,
locality, etc., will be found in Zoologica, Volume XIII, Numbers 1, 2 and 3,
and Volume XXI, Numbers 3 and 4.

The deep-water hauls provided large collections of copepods. From
these, 11 vials of specimens selected by G. H. Wailes were examined by me
and yielded 101 species ; other identifications brought the number of species
now recorded to a total of 112.

Of this number I have recorded 44 species (40%) from the Woods Hole
Region (U. S. Nat. Mus., Bull. 158, 1932) and all have previously been re-
corded from the Atlantic or Mediterranean areas.

The genus Copilia was represented by one or two individuals in nearly
every haul and specimens of the genus Sapphirina also occurred in several.
The genus Corycaeus was represented more or less numerously in all hauls.

One individual each of Gaetanus caudani and of Gaetanus latifrons were
obtained in Sample 9 (800 fathoms).

The surface hauls were characterized by the predominance of Acartia
clausi and Paracalanus parvus; other species noted in them were Corycaeus
spp., numerously, and Oncaea spp., sparsely, also Mecynocera clausi, Clauso-
calanus arcuicornis, Euchaeta acuta, Candacia aethiopica, Candacia simplex,
Pontellina plumata, Calanopia elliptica, Macrosetella gracilis, Mirada ef-
ferata, Oithona spinirostris, Lubbockia squillimana, Sapphirina angusta,
Sapphirina auronitens, Sapphirina metallina and a number of calanoids not
specifically identified.

A single specimen of Caligus curtus was observed in Sample 18.

Order Copepoda.

Suborder Calanoida.

Family Calanidae.

Genus Calanus Leach 1819.

C. propinquus Brady.

Genus Megacalanus Wolfenden 1904.

M. longicornis Sars.

*M. princeps (Brady) Sars.

M. sp. (immature).

1 Contribution No. 498, Department of Tropical Research, New York Zoological Society.

90

Zoologica: New York Zoological Society

[XXI :7

Genus Neocalanus Sars 1925.

*N. gracilis (Dana) Sars.

N. robustior (Giesbrecht) Sars.
Genus Undinula A. Scott 1909.

U. darwini (Lubbock) Scott.

*U. vulgaris (Dana) Scott.

Genus Eucalanus Dana 1852.

*E. attenuatus (Dana).

E. crassus Giesbrecht.

*E. elongatus (Dana).

E. mucronatus Giesbrecht.

E. pileatus Giesbrecht.

E. subtenuis Giesbrecht.

Genus Rhincalanus Dana 1852.

*R. cornutus Dana.

*R. nasutus Giesbrecht.

Genus Mecynocera I. C. Thompson 1888.

clausi I. C. Thompson.

Genus Bathycalanus G. 0. Sars 1905.

B. richiardi G. O. Sars.

B. rigidus G. 0. Sars.

Family Paracalanidae.

Genus Paracalanus Boeck 1865.

*P. parvus (Claus) Sars.

Genus Calocalanus Giesbrecht 1888.

*C. pavo (Dana) Giesbrecht.

Family Pseudocalanidae.

Genus Clausocalanus Giesbrecht 1888.

*C. arcuicornis (Dana).

Genus Pseudocalanus Boeck 1872.

*P. minutus (Kr0yer).

Family Aetideidae.

Genus Aetideus Brady 1883.

*A. armatus (Boeck).

Genus Chiridius Giesbrecht 1892.

C. poppei Giesbrecht.

Genus Chirundina Giesbrecht 1895.

*C. streetsii Giesbrecht.

Genus Gaetanus Giesbrecht 1888.

G. armiger Giesbrecht.

G. caudani Canu.

G. latifrons G. O. Sars.

*G. miles Giesbrecht.

Genus Gaidius Giesbrecht 1895.

*G. brevispinus (G. O. Sars)*.

*G. tenuispinus (G. O. Sars).
Genus Mesogaidius Wolfenden.

M. intermedius Wolfenden.
Genus Euchirella Giesbrecht 1888.

E. brevis (G. 0. Sars).

Genus Undeuchaeta Giesbrecht 1888.
*U. major Giesbrecht.

U. spectabilis (G. 0. Sars).

1936]

Wilson: Notes on Copepoda

91

Genus Pseudochirella G. 0. Sars 1920.

P. obesa G. 0. Sars.

P. obtusa (G. 0. Sars).

P. pustilifera (G. 0. Sars).

Genus Pseudaetideus Wolfenden 1904.

P. armatus (Boeck).

Family Euchaetidae.

Genus Euchaeta Philippi 1843.

E. acuta Giesbrecht.

*E. marina (Prestandrea) Giesbrecht.

Genus Paraeuchaeta A. Scott 1909.

P. bisinuata (G. 0. Sars).

P. hanseni (With.).

Genus Pseudeuchaeta.

P. norvegica (Boeck).

Family Phaennidae.

Genus Phaenna Claus 1863.

P. spinifera Claus.

Family Scolecithricidae.

Genus Scottocalanus G. 0. Sars 1905.

S. securifrons (T. Scott).

Genus Scolecithrix Brady 1883.

S. bradyi Giesbrecht.

*S. danae (Lubbock) Giesbrecht.

Genus Scolecithricella G. 0. Sars 1902.

S. abyssalis (Giesbrecht).

Family Centropagidae.

Genus Centropages Kr0yer 1849.

C. violaceus (Claus).

Family Metridiidae.

Genus Metridia Boeck 1864.

*Af. brevicauda Giesbrecht.

*M. longa (Lubbock).

*M. lucens Boeck.

M. normani Giesbrecht.

*M. princeps Giesbrecht.

M. venusta Giesbrecht.

Genus Pleuromamma Giesbrecht 1898.

*P. abdominalis (Lubbock) Gies. & Schmeil.
*P. gracilis (Claus) Giesbrecht.

P. quadrungulata (F. Dahl).

*P. robusta (F. Dahl) Sars.

*P. xiphias (Giesbrecht).

Family Lucicutiidae.

Genus Lucicutia Giesbrecht 1898.

L. clausi (Giesbrecht).

L. flavicornis (Claus).

L. longicornis Giesbrecht.

*L. magna Wolfenden.

L. maxima Steuer.

92

Zoologica: New York Zoological Society

[XXI :7

Family Heterorhabdidae.

Genus Heterorhabdus Giesbrecht 1898.

H. grimaldi (Richard).

*H. longicornis (Giesbrecht).

H. papilla ger (Claus).

H. spinifrons (Claus).

Genus Haloptilus Giesbrecht 1898.

H. ornatus (Giesbrecht).

H. longicornis (Claus).

Genus Euaugaptilus G. O. Sars 1920.

E. elongatus (G. 0. Sars).

Family Candaciidae.

Genus Candacia Dana 1846.

C. simplex (Giesbrecht).

C. aethiopica (Dana).

Family Pontellidae.

Genus Pontellina Dana 1853.

*P. plumata (Dana).

Genus Calanopia Dana 1852.

C. elliptica (Dana).

Family Acartiidae.

Genus Acartia Dana 1846.

*A. clausi Giesbrecht.

*A. danae Giesbrecht.

Suborder Harpacticoida.

Family Ectinosomidae.

Genus Microsetella Brady & Robertson 1873.
*M. norvegica (Boeck).

Family Macrosetellidae.

Genus Macrosetella A. Scott 1909.

M. gracilis (Dana).

Genus Miracia Dana 1846.

*M. efferata Dana.

Family Tachidiidae.

Genus Clytemnestra Dana 1847.

C. scutellata Dana.

Suborder Cyclopoida.

Family Oithonidae.

Genus Oithona Baird 1843.

0. attenuata Farran.

*0. similis Claus.

*0. spinirostris Claus.

Family Oncaeidae.

Genus Oncaea Philippi 1843.

O. conifera Giesbrecht.

0. curta G. 0. Sars.

0. media Giesbrecht.

0. mediterranea Claus.

1936]

Wilson: Notes on Copepoda

93

*0. minuta Giesbrecht.

0. tenella G. 0. Sars.

*0. venusta Philippi.

Genus Conaea Giesbrecht 1891.

C. rapax Giesbrecht.

Genus Lubbockia Claus 1863.

L. aculeata Giesbrecht.

L. squillimana Claus.

Family Corycaeidae.

Genus Corycaeus Dana 1845.

C. agilis Dana.

C. carinatus Giesbrecht.

C. catus M. Dahl.

C. crassiusculus Dana.

*C. elongatus Claus.

C. lautus Dana.

C. limbatus Brady.

*C. speciosus Dana.

C. typicus Kr0yer.

Genus Farranula Blake 1932.

F. carinata (Giesbrecht).

Genus Sapphirina J. V. Thompson 1829.

*S. angusta Dana.

*S. auronitens Claus.

S. metallina Dana.

Genus Copilia Dana 1849.

C. quadrata Dana.

C. vitrea Haeckel.

Suborder Caligoida.

Family Caligidae.

Genus Caligus Muller 1785.

*C. curtus Muller.

Note: The species marked have been recorded from the Woods Hole
region by Wilson.

Tatter sail: Notes on Schizopoda

95

8.

Plankton of the Bermuda Oceanographic Expeditions.

V. Notes on Schizopoda1.

W. M. Tattersall.

Professor of Zoology and Comparative Anatomy,

University College, Cardiff.

This is one of a number of papers dealing with the planktonic contents
of a selected series of nets drawn at various levels off the coast of Bermuda
on the Bermuda Oceanographic Expeditions of the New York Zoological
Society under the direction of Dr. William Beebe. Full details of the nets,
locality, etc., will be found in Zoologica, Volume XIII, Numbers 1, 2 and 3,
and Volume XXI, Numbers 3 and 4.

All of the species of schizopods here recorded are well known in the
North Atlantic and have been reported previously either by Hansen or
Tattersall from the eastern coast of North America. The only species which
is new to this area is the mysid Longithorax sp. nov.

PHYLUM ARTHROPODA.

Class Crustacea.

Order Euphausiacea.

Bentheuphausia amblyops Sars. 1,000-500 fathoms. Few.

Thysanopoda tricuspidata Milne-Edwards. 700 fathoms. One.

Thysanopoda aequalis Hansen. 1,000-300 fathoms. Very numerous.

Thysanopoda obtusifrons Sars. 1,000, 900, 400 fathoms. Few.

Euphausia brevis Hansen. 1,000-300 fathoms, mostly at 500 fathoms.
Numerous.

Euphausia mutica Hansen. 500 fathoms. Rare.

Euphausia americana Hansen. 900-300 fathoms. Few.

Euphausia tenera Hansen. 1,000-500 fathoms. Few.

Euphausia hemigibba Hansen. 1,000-300 fathoms. Numerous in 500
fathoms.

Euphausia gibboides Ortmann. 200 fathoms. One.

Thysanoessa parva Hansen. 1,000-400 fathoms. Numerous. .

Nematoscelis tenella Sars. 900-200 fathoms. Few.

N ematoscelis microps Sars. 1,000-200 fathoms. Numerous.

Stylocheiron carinatum Sars. 500 fathoms. Four seen.

Stylocheiron suhmi Sars. 200 fathoms. One.

Stylocheiron longicorne Sars. 900-300 fathoms. Numerous.

Stylocheiron abbreviatum Sars. 900, 800, 400 fathoms. One in each haul.

Stylocheiron elongatum Sars. 1,000-200 fathoms. Numerous.

1 Contribution No. 499, Department of Tropical Research, New York Zoological Society.

96

Zoologica: New York Zoological Society [XXI :8 — 1936]

Order Mysidacea.

Anchialina typica (Kr0yer). Surface, Net 900. One.

Euchaetomera tenuis Sars. 400 fathoms. One

Longithorax sp. nov. Surface, Net 900. Fourteen. Surface, Net 976.
Six. 400 fathoms, one; 500 fathoms, one.

Selected Literature.

Hansen, H. J., The Crustacea Euphausiacea of the United States Na-
tional Museum. Proc. U. S. Nat. Mus., Vol. 48, pp. 59-114, Pis. 1-4.
1915.

Tattersall, W. M., Crustaceans of the Orders Euphausiacea and Mysi-
dacea from the Western Atlantic. Proc. U. S. Nat. Mus., Vol. 69,
Art. 8, pp. 1-31, Pis. 1, 2. 1926.

Coe: Bathy pelagic Nemerteans

97

9.

Plankton of the Bermuda Oceanographic Expeditions. VI. Bathy-
pelagic Nemerteans Taken in the Years 1929, 1930 and 19311.

Wesley R. Coe.

Osborn Zoological Laboratory , Yale University.

(Plates I-X ; Text-figure 1).

The Bermuda Oceanographic Expeditions, under the direction of Dr.
William Beebe, were organized for the purpose of making an intensive
faunal survey of a closely limited area of the deep subtropical ocean. The
area chosen was a circle eight miles in diameter, with its center located at
32° 12' N. Lat., 64° 36' W. Long., a point about nine miles southeast of
Nonsuch Island, Bermuda. (Text-fig. 1). The depth at this locality is 1,000
to more than 1,400 fathoms, but most of the nets were so arranged as to
collect samples simultaneously at 100-fathom intervals from 500 to 1,000

1 Contribution No. 500, Department of Tropical Research, New York Zoological Society.

Text-figure 1.

Diagram showing location of Beebe eight-mile cylindrical trawling area and the
arrangement of collecting nets in which the bathypelagic nemerteans of this
report were obtained. (From Beebe).

98

Zoologica: New York Zoological Society

[XXI :9

fathoms inclusive. During the three summers, 1929 to 1931, a total of 1,042
nets one metre in diameter were drawn at these depths in all directions
across this imaginary cylinder of water.

From the enormous volume of water which passed through these nets a
total of 79 specimens of bathypelagic nemerteans was obtained. These rep-
resented 12 species, belonging to 10 genera, of which 6 species proved to be
new to science. Many of the others are of great interest, not only because
of the information which they furnish as to the geographical range of little-
known species but more particularly because their study has supplemented
our knowledge of the morphological peculiarities of several species of which
only one or two individuals, representing only one of the sexes, had been
previously recorded. No representatives of the group have been previously
reported from this region. One species, Pachynemertes obesa, was a mem-
ber of a new genus.

A list of the species secured by the expeditions, together with the num-
ber of specimens and the depth at which they were obtained, is as follows:

Species

Depth

No.

Specimens

1. Protopelagonemertes hubrechti Brink.

1,400-1,800 m.

3

2. Protopelagonemertes beebei sp. nov.

1,646 m.

1

3. Plotonemertes adhaerens Brink.

1,300-1,830 m.

27

4. Plotonemertes aurantiaca sp. nov.

1,463 m.

1

5. Crassonemertes robusta Brink.

1,100 m.

2

6. Pachynemertes obesa gen. nov. ; sp. nov.

1,646 m.

1

7. Paradinonemertes wheeleri sp. nov.

1,300-1,830 m.

6

8. Planonemertes labiata sp. nov.

1,800 m.

2

9. Phallonemertes murrayi Brink.

1,500-1,830 m.

2

10. Chuniella lanceolata Brink.

1,646 m.

1

11. N ectonemertes mirabilis Verrill.

1,300-1,830 m.

32

12. Balaenanemertes minor sp. nov.

549 m.

1

Total 79

When it is realized that a total of only about fifty species of these
bathypelagic nemerteans has previously been described from all the oceans
of the world it is surprising to find so many as here recorded for a single
locality. Only 6 of the 12 species were previously known to science and in
only 2 of these, Phallonemertes murrayi and N ectonemertes mirabilis , were
both sexes known.

The fact that in several cases only a single specimen of the species was
obtained in a total of 600 to 800 nets lowered to the water layer which the
species inhabits and drawn there horizontally for upwards of four hours
indicates how exceedingly sparse the population must be, for the worms are
so sluggish that they could not escape capture if the nets reached their
exact position. This makes it seem probable that even this cylindrical mass
of water, in spite of such intensive exploration, may still harbor additional
species.

It has been found that water layers of similar physical and chemical
characteristics extend for great distances in the oceans at depths of 1,000
metres or more. Consequently the bathypelagic species, even though they
be limited to a particular water layer, may have a wide geographical dis-
tribution. It is consequently not surprising to find in the Bermuda collec-
tions species hitherto recorded only from such remote seas as extend from
southern Greenland to the coasts of Great Britain.

At the Beebe eight-mile circle the temperature of the water at 1,200 m.
is about 5.29° C. and the salinity 35.08; at 1,600 m. to 2,500 m. the tem-
perature varies only from 3.96° to 3.21° C., the salinity from 35.04 to 34.96,
the density 27.81-27.89, the oxygen 5.9 to 6.0 per L., the pH 8.06-8.15 and

1936]

Coe: Bathypelagic Nemerteans

99

the P04 45-48. Such stable conditions allow a great vertical range below
1,200 m. and an almost unlimited geographical range. But since the bathy-
pelagic nemerteans are poorly adapted for active locomotion their rate of
dispersal would be extremely slow were it not for the ocean currents which
may carry them for very great distances.

With the exception of Planonemertes , which is known only from the
Pacific ocean, all the genera represented in the collections have been pre-
viously reported for some portion of either the North or South Atlantic
ocean. Curiously enough several genera, including Dinonemertes Plank-
tonemertes, Buergeriella, Amaueria and Natonemertes, which are known to
inhabit similar water layers in other parts of the North Atlantic, were not
taken at the Bermuda area.

The bathymetrical distribution of the species collected at Bermuda is
indicated in the following table (Table I).

TABLE I.

Bathymetrical Distribution of Bathypelagic Nemerteans
Collected by 1,042 Nets Drawn Across the Beebe Eight-mile
Cylindrical Trawling Area off Bermuda, Indicated by Number of
Specimens Obtained at each Depth. (Depths at which these
species have been previously obtained are indicated by “x.” The
figures representing temperature and salinity refer to the Ber-
muda area only).

Co

53

d

£

<53

33

.53

•<s>

553

eo

•<s»

<o

•<s>

5-

■ki

53

eo

<53

<53

53

53

'o

s

d

s-

<53

.O

<33

<53

53

£

53

s-

CO

52

rO

^3

^3

-Si

•o»

V.

S-

52

<53

33

52

-O

53

L-

+->

a

a

B

’3

52

-S£

<53

-O

*53

53

52

53

O

5-

-O

O

§

53

£

53

•<s>

£

03

CD

*03

*-^3

53

-si

d

•

Q

EH

Os

CL

CL

CL

O

o

CL

CL

CL

O

£

Cq

500

X

600

1L5

35.91

X

700

X

800

1CL3

35^35

X

X

900

X

1,000

6.8

35^07

X

x2

1,100

1

xl

1,200

5^3

35^08

x2

1,300

X

.... 2 ....

1

i

xl

1,400

1

.... 4 ....

x2

1,500

.... 6 1

1

x3

1,600

4.0

34^99

i

1 8 ...

1

3

X

x5

1,700

3

X ....

X

i

x2

1,800

l

.... 4 ....

’ 2

1 xl

x8

1,900

X

1 ....

x6

2,000

3.5

35.04

X

. .. x ....

X

X

X

Key to the Genera of Bathypelagic Nemerteans from the
Bermuda Area:

A. Body without tentacles in either sex B

B. Mouth and proboscis opening united; body slender, narrowed at
both ends; proboscis sheath nearly as long as body.

Protopelagonemertes

100

Zoologica: New York Zoological Society

[XXI :9

BB. Mouth more or less widely separated from proboscis opening C

C. Musculature of proboscis sheath of interlaced circular, spiral

and longitudinal fibers D

D. Body swollen and nearly cylindrical anteriorly, flattened
and sharply recurved in posterior portion; with pair of
specialized integumental glandular organs on ventral sur-
face; proboscis very large Plotonemertes

DD. Body broad and more or less flattened ; without specialized

glandular organs E

E. Body broad, oval, thick; caudal fin narrow and
sharply demarcated from body ; mouth and proboscis
opening not widely separated; intestinal diverticula
much branched, with well developed ventral branch.

Crassonemertes

EE. Body short, thick, rounded; caudal fin slightly dif-
ferentiated; intestinal diverticula without ventral

branches Pachynemertes

EEE. Body much flattened; caudal fin broad and not sharp-
ly demarcated from body; mouth and proboscis open-
ing widely separated; intestinal diverticula with

rudimentary ventral branch or none F

F. Body rather narrow, with parallel lateral mar-
gins, much flattened posteriorly and continuing
into flat caudal fin; mouth anterior to brain;
spermaries of adult males with strong muscular
walls and highly developed sperm ducts which
continue beyond body as long, tubular penes.

Phallonemertes

FF. Body oval and very flat; mouth on ventral sur-
face of head, slightly anterior, beneath or slight-
ly posterior to brain commissures, according to
state of contraction of cephalic tissues; sperma-
ries without external penes Paradinonemertes

CC. Circular and longitudinal muscles of proboscis sheath in separate

layers G

G. Proboscis sheath composed of three more or less separate

layers; body broad and very flat Planonemertes

GG. Proboscis sheath of two layers, outer circular and inner
longitudinal; body slender, rounded anteriorly, tapered to

narrow posterior extremity Chuniella

AA. Body with pair of lateral tentacles in one or both sexes ; caudal fin

highly developed H

H. Tentacles develop in adult males only ; body slender. N ectonemertes
HH. Tentacles in both sexes; body broad Balaenanemertes

Genus Protopelagonemertes Brinkmann.

1. Protopelagonemertes hubrechti Brinkmann.

Brinkmann, 1917, 1917a, page 178; Coe, 1926.

( Bathynemertes hubrechti Brinkmann, 1917, 1917a; Wheeler, 1934).
(Figs. 16, 18, 28, 29-31).

Two typical specimens of this species were contained in the Beebe col-
lections. Both were females, as was also the type specimen. One of these
measured 30 x 5 x 2 mm. and the other 40 x 5 x 2 mm. The type specimen

1936]

Coe: Bathy pelagic Nemerteans

101

measured 56 mm. in length and 10 mm. in greatest width. The body is
pointed at both ends and is much more slender and relatively thicker than in
most bathypelagic forms. A third specimen, also female, was evidently a
giant of the race, for its bulk exceeded by many times the largest of the
other individuals. It measured after preservation 78 mm. in length, 22 mm.
in width and about 10 mm. in thickness (Figs. 18, 28). Its massive appear-
ance seemed to indicate a distinct species, but careful study of the internal
anatomy showed a general conformity with other individuals except for the
vastly greater size of the organ systems.

Color in life red, scarlet or orange.

Mouth and proboscis opening united into a common atrium which may
disappear when proboscis is partially everted. Intestinal diverticula with
both dorsal and ventral lobulated branches. Lateral nerve cords with single
fibrous core.

The relatively large proboscis is provided with 27-29 well defined
nerves through most of the length of the anterior chamber, although some
of the sections in one specimen show only 26 and in another 30 ; posterior
to the nerve ring in the middle chamber the number of distinct nerves is
reduced from 29 to 13-18. Specimens from the South Atlantic, referred to
this species by Wheeler, had only 22, 24, 25 or 26 proboscidial nerves. The
basis is sharply curved and is provided with numerous sharply-pointed
conical stylets (Figs. 41, 42). About twenty accessory stylets of similar
form lie in six small pouches on the adjacent wall (Fig. 42). The proboscis
sheath reaches three-fourths to seven-eighths the length of the body. The
sheath is composed of closely interwoven fibers.

The female has thirty or more pairs of large, elongated ovaries, each
containing many ova. Each gonad is crescentic in section, with its tip near
the median line when fully mature and with an arch above the nerve cord
leading to the oviduct which opens lateroventrally. Many small ovocytes
and a few larger ova are present in each gonad. The largest specimen was
peculiar in having a single ovary in ventral side of body immediately
posterior to the brain, in position where spermaries are found in many
species. The male is as yet unknown.

Geographical distribution: The three specimens from the Beebe 8-mile
cylinder were taken at depths of about 1,400 to 1,800 metres. Reported by
Wheeler from the South Atlantic ocean west of South Africa. The type
specimen was obtained west of Ireland at a depth of about 2,000 metres,
indicating a wide distribution in both the North and South Atlantic oceans.

2. Protopelagonemertes beebei sp. nov.

(Figs. 25, 34).

The collection contained one specimen similar in general appearance to
medium sized individuals of P. hubrechti, except flatter, but differing in
having only 19-21 proboscidial nerves and in other morphological details as
noted below.

Body elongate oval, pointed at both ends, flattened except at anterior
end. Size of type specimen: 24 mm. long, 7 mm. wide and somewhat more
than 2 mm. thick. Body walls thicker than in most bathypelagic species,
both circular and longitudinal layers being well developed even on lateral
margins; cephalic musculature particularly thick, with very strong muscular
bands providing a firm proboscis insertion anterior to brain.

Color in life orange yellow, with paler lateral margins; color results
mainly from globules in intestinal epithelium, thus leaving a longitudinal
median dorsal band of creamy white; proboscis whitish.

Mouth and proboscis opening united; proboscis sheath about seven-
eighths as long, as body, composed of interlacing spiral and longitudinal
fibers. Proboscis longer than body, with 20 nerves in most sections of an-

102

Zoologica: New York Zoological Society

[XXI :9

terior chamber, but with apparently one more or one less at intervals, due
to variations in the interneural plexus ; nerves form continuous ring in basis
region ; retractor passes through dorsal wall of sheath to become anchored
in dorsal body wall. Basis of typical curved form and bears a dozen or
more conical, toothlike stylets. Accessory stylets were not found in this
specimen.

Esophagus slender, leading to remarkably voluminous stomach with
much convoluted walls posterior to brain commissures. Pylorus both long
and broad ; caecal diverticula and upwards of 40 pairs intestinal diverticula
have both dorsal and ventral branches, with lobes above and beneath the
nerve cords.

Dorsal vessel large, passing beneath proboscis sheath near posterior
end of pylorus to unite with lateral vessels at posterior end of body.

Brain large; nerve cords each with single fibrous core; dorsal nerve
conspicuous in all sections.

Reproductive organs: Female with upwards of 30 pairs of ovaries,
the most anterior pair situated on the ventral side of nerve cords imme-
diately posterior to brain. Anterior gonads small, with only one or two
large ova, but in middle of body gonads are very large, arching above nerve
cords to open ventroiaterally. Each ovary contains several large ova and
many small ovocytes, as well as a basal syncytium containing numerous
small nuclei. Male unknown.

Geographical distribution: Known only from the Beebe eight-mile

cylinder off Bermuda ; taken at a depth of 1,646 m.

The species is named in honor of Dr. William Beebe, Director of the
Bermuda Oceanographic Expeditions, whose ability and industry have made
the Bermuda trawling area the most thoroughly explored portion of all the
oceans.

Genus Plotonemertes Brinkmann.

3. Plotonemertes adhaerens Brinkmann.

Brinkmann, 1917, 1917a; Coe, 1926.

(Figs. 1, 7, 8, 14, 15, 22, 26, 27, 36, 37-42).

This species, of which only a single specimen has been hitherto re-
ported, was represented by no less than 27 individuals in the collections
studied. With the exception of N ectonemertes mirabilis it is evidently the
most abundant species in the Bermuda area. Males, females and sexually
immature individuals were represented.

The smallest specimen measured only 6 mm. in length and the largest
20 mm. The usually strongly recurved posterior extremity increases the
difficulty of accurate measurement. Some of the variations in size and
proportions are here indicated in millimeters: 6 x 1.5, 7 x 2, 7 x 2.5, 8 x 2.5,
10 x 3, 10 x 2, 11 x 2, 12 x 2, 12 x 3, 13 x 1.5, 13 x 2, 14 x 3, 15 x 2.5, 15
x 3, 16 x 5, 19 x 6, 20 x 5. The type specimen measured 30 x 9 mm. In
most specimens the body is elongated oval or club-shaped, tapering gradually
posteriorly, and usually with the flattened posterior end strongly recurved
dorsally or, less frequently, ventrally.

Mouth and proboscis opening separate, the former usually with pro-
truding circular lips due to partly everted stomach. Intestine with upwards
of 50 pairs of much lobed diverticula, each with distinct dorsal and ventral
branches; caecum with 6 pairs of similar diverticula. Proboscis sheath
three-fourths to seven-eighths as long as body, composed of interlacing longi-
tudinal and spiral fibers; proboscis relatively larger than in other bathy-
pelagic forms and fully twice as long as body; attached posteriorly near
end of dorsal wall of proboscis sheath. Number of proboscidial nerves
variable, usually 26-28, but in some individuals only 24 or 25 were found

1936]

Coe: Bathy pelagic Nemerteans

103

and in one specimen there were 30 distinct nerves anteriorly and only 25
farther back, while another had 28-34, of which several were smaller than
the others. The variation is evidently due to the irregular distribution of
nerves in the interneural plexus. Stylet basis curved at both ends, bearing
numerous conical stylets ; 6 small pouches each contain several more or less
perfectly formed accessory stylets (Figs. 41, 42).

Lateral nerve cords with main ventral core and small, imperfectly sepa-
rated, dorsal core ; in some specimens dorsal core not distinguishable except
near origin of dorsal peripheral nerves.

The glandular cutaneous organ found by Brinkmann on the ventral sur-
face of the male occurs in both sexes, but is more highly differentiated in
the male. This organ consists of deep convolutions of the surface epithelium,
the lateral pair of folds being deeper than the others. In certain states of
contraction this organ projects at an angle from the ventral surface when
the body is recurved dorsally and may serve as an organ for the adhesion
of the two sexes, as Brinkmann has suggested. The organ is not only highly
glandular but is doubtless sensory as well, since it receives large branches
from the adjacent lateral nerve cords.

Reproductive organs: Male provided with an irregular row of 7-11
spermaries opening ventrally on each side of body back of head ; female
with upwards of 30 pairs of narrow ovaries between intestinal diverticula
and dorsal to lateral nerves ; each ovary has several large ova and numerous
small ovocytes.

Geographical distribution: The specimens here recorded were taken at
depths of 1,300-1,800 m. on 16 occasions in the Beebe eight-mile area, while
the type specimen was found near the middle of the North Atlantic (47°
34' N. Lat., 43° 11' W. Long.) at a depth of 2,000 m.

4. Plotonemertes aurantiaca sp. nov.

(Figs. 19, 20, 33, 43).

This new species of Plotonemertes differs from P. adhaerens in con-
figuration of the body, in number of proboscidial nerves, in character of
proboscis armature, in length of proboscis sheath, in character of glandular
organ, in having but a single core in each of the lateral nerve cords and in
other details noted below.

Body elongate, about half as thick as wide, flattened ventrally and
rounded dorsally except in posterior sixth of body which is recurved behind
glandular organ and flattened into a distinct caudal fin (Figs. 19, 20). Length
of type specimen, 40 mm., width 11 mm., thickness in median line 5-7 mm.
near middle of body and 1 mm. or less in caudal fin.

Surface epithelium well preserved over entire body; basement layer
moderately thin and corrugated; muscular walls on lateral margins very
thin; dorso-ventral muscles between intestinal diverticula with some giant
fibers in addition to ordinary muscle cells.

Mouth separate from proboscis opening; proboscis sheath about two-
thirds to three-fourths as long as body; composed of interlacing longitu-
dinal and circular fibers except ventrally, where longitudinal fibers are few
in number; proboscis very large, and much longer than body; 21 distinct
proboscidial nerves. Epithelium with prominent papillae. Stylet basis pistol-
shaped, with curved, rounded base and nearly straight principal axis ; stylets
lost before examination, but the rounded pockets on face of basis indicate
that they were numerous (Fig. 33).

Color in life bright orange, with yellow margins and caudal fin and
deep yellow proboscis. Color well preserved after several months in alcohol.

Stomach close behind mouth, walls much folded and evidently capable
of extension as circular lips around mouth opening; pylorus wide but flat;

104

Zoologica: New York Zoological Society

[XXI :9

caecal and intestinal diverticula much lobulated, with ventral branches
beneath nerve cords; 40-50 pairs intestinal diverticula.

Cutaneous glandular organs widely separated from each other in the
female; epithelial convolutions with deep folds beneath basement membrane
(Fig. 43) ; provided with large nerves from both median and ventral sides of
lateral nerve cords.

Lateral nerve cords consist of but a single fibrous core ; with the usual
commissure on dorsal side of rectum and posterior to anastomosis of blood
vessels..

Dorsal vessel large, enters rhynchocoel near posterior end of brain re-
gion, passes beneath proboscis sheath anterior to intestinal region and
unites with lateral vessels at posterior end of body.

Parasites'. The rhynchocoel contained many large gregarines.

Reproductive organs : Type specimen was female with 26-29 ovaries on
each side (Fig. 43) situated close beside and above the nerve cords. Ova
large, 3-6 in each gonad ; oviducts open on ventral surface directly beneath
lateral nerve cords. Male unknown.

Geographical distribution : Known only from the Beebe 8-mile area;
depth 1,463 m.

Genus Crassonemertes Brinkmann.

5. Crassonemertes robusta Brinkmann.

Brinkmann, 1917, 1917a; Coe, 1926; Wheeler, 1934.

(Figs. 17, 21, 35).

The collections contained two specimens of this thick-bodied nemer-
tean; only two others have been previously recorded. One specimen meas-
ured 16 mm. in length, 7.5 mm. in width and 2 mm. in thickness after pre-
servation; the other was 23 x 12 x 2 mm.; the type specimen was 25 mm.
long, 10 mm. wide and 4.5 mm. thick.

Mouth and proboscis opening separate; proboscis sheath extends nearly
entire length of body ; composed of interlacing spiral and longitudinal muscle
fibers, with a tendency toward separate layers posteriorly; proboscis large,
armed with rather large, moderately curved basis, bearing many sharply
conical stylets and about a dozen shallow pouches of imperfectly formed re-
serve stylets; 20-23 proboscidial nerves; retractor passes through dorsal
wall of sheath to become interlaced with muscles of dorsal body wall.

Upwards of 40 pairs of intestinal diverticula and 5 pairs of caecal
diverticula profusely branched, both above and below lateral nerve cords.
Nerve cords with single fibrous core anteriorly, but with small and incom-
pletely separated dorsal core posterior to middle of body.

Reproductive organs : Female with 20-30 or more pairs of ovaries on
dorsal side of nerve cords, opening lateroventrally. Male unknown.

Geographical distribution: Evidently widely distributed in the North
Atlantic, the type specimen coming from northwest of Great Britain (57° 41'
N. Lat., 11° 48' W. Long.) at a depth of about 1,666 m. ; and the Bermuda
specimen from a depth of about 1,100 m. Wheeler reports this species from
off the west coast of Africa (6° 55' N. Lat., 15° 54' W. Long.) at a depth of
less than 800 m.

Genus Pachynemertes nov.

The collections contained two specimens which have a superficial re-
semblance to Crassonemertes robusta but differ so widely in internal anat-
omy as to require the establishment of a separate genus in the family
Planktonemertidae.

The genus Pachynemertes is diagnosed as follows: Body short, thick,

1936]

Coe: Bathy pelagic Nemerteans

105

rounded; with slightly differentiated caudal fin; mouth and proboscis open-
ing separate; intestinal diverticula without ventral branches; proboscis
sheath composed of interlaced fibers; lateral nerve cords separated from
ventral body wall by gelatinous tissue only.

6. Pachynemertes obesa sp. nov.

(Figs. 50, 51).

A single example of a short and thick-bodied nemertean bore the label
“Grenadine animal,” doubtless referring to the similarity of its shape to a
miniature hand grenade. This species differs from C. robusta in the number
of proboscidial nerves, in shape of stylet basis, in having less profusely
branched intestinal diverticula, in the shorter length of the proboscis sheath
and in other anatomical details.

Body after preservation short, oval, thick, with thin lateral margins
posteriorly, continuing into a slightly bilobed caudal fin (Fig. 50). Length
of type specimen 16 mm., width 8 mm., thickness 5 mm. Color in life un-
known ; body opaque and firm after preservation. Body walls thick on dorsal
and ventral surfaces, but thin laterally.

Proboscis sheath extends about three-fourths the length of the body;
circular, spiral and longitudinal fibers interlaced to form a single muscula-
ture. Proboscis slender; longer than body; provided with 14 distinct nerves
and, in some sections, two additional ones of smaller size; stylet basis
sharply curved, armed with upwards of 20 conical teeth (Fig. 51).

Mouth and proboscis opening well separated; mouth with protruding
lips (Fig. 50) ; stomach much convoluted ; slender pylorus of moderate
length ; about 35 pairs of intestinal diverticula, which are lobed, but not
distinctly branched; ventral branches rudimentary, allowing lateral nerve
cords to lie close to ventral body wall.

Dorsal vessel extends in rhynchocoel for a short distance, then passes
beneath the proboscis sheath and continues posteriorly to join lateral vessels
at posterior end of body.

Brain of moderate size; lateral nerve with single fibrous core.

Reproductive organs: The type specimen was an adult female with sev-
eral ova in each of upwards of 30 pairs of ovaries.

Geographical distribution: Known only from the Bermuda area, where
a single specimen was obtained at a depth of about 1,600 m.

Genus Paradinonemertes Brinkmann. (Emended).

According to the diagnostic characters of this genus as formulated by
Brinkmann (1917), the mouth is situated behind the brain. The six speci-
mens of a closely related species (P. wheeleri ) from this collection, un-
doubtedly belonging to the same genus, prove that the position of the mouth
relative to the brain is variable, depending on the state of contraction of
the anterior portion of the body. Consequently the generic diagnosis must
be emended to read: Body much flattened; mouth somewhat widely sepa-
rated from rhynchodeal opening, situated anterior, ventral or slightly pos-
terior to brain according to state of contraction of anterior portion of body ;
proboscis sheath extends into posterior third of body, its musculature com-
posed of interwoven fibers.

7. Paradinonemertes wheeleri sp. nov.

(Figs. 2, 3, 9, 10, 11, 23, 45-49).

The collections contained 6 excellently preserved specimens of a new
species of Paradinonemertes, the individuals of which differ from those

106

Zoologica: New York Zoological Society

[XXI :9

of P. drygalskii Brinkmann in many important respects, particularly in
length of proboscis sheath, armature of proboscis, number of spermaries,
character of lateral nerve cords and position of mouth.

Body translucent, broad, flat and thin, with broad caudal fin continuous
with thin lateral margins of body; about 3 to 4 times as long as wide;
measurements of several specimens were as follows: males, 14 x 5 x 2, 23 x
5.5 x 2; females, 11 x 3 x 2.5, 11 x 4 x 1.5, 13 x 4 x 2, 18 x 5 x 2, 42 x 15
x 2 (Figs. 2, 3, 9, 16, 23, 45, 46).

Mouth situated on ventral side of head ; widely separated from pro-
boscis opening but anterior to brain commissures except when cephalic
musculatures are abnormally contracted, with protruding lips in most pre-
served specimens, leading directly to convoluted walls of stomach. In speci-
mens having the proboscis fully retracted the mouth lies well anterior to
the brain ; the lips are widely protruded in the preserved specimens ; the
stomach walls have but few convolutions; 40-50 pairs of lobulated intestinal
diverticula, with small or rudimentary ventral branches not extending be-
neath nerve cords.

Proboscis sheath about two-thirds to three-fourths as long as body, with
interlaced muscular walls; proboscis about as long as body, with 12 or 13
distinct nerves ; basis strongly curved, with numerous rather sharply conical
stylets (Fig. 47). Proboscis sheath tapers to a point posteriorly and ter-
minates in the parenchyma between the intestinal diverticula and without
any fibrous connection with the dorsal body wall. Retractor muscles of pro-
boscis are attached to sheath only.

Cephalic blood lacunae rather small; dorsal vessel enters rhynchocoel
close behind ventral brain commissure, passes beneath the proboscis sheath
a short distance farther back and continues posteriorly to join the lateral
vessels at the end of the body.

Lateral nerve cord with small and imperfectly differentiated dorsal core.

Reproductive organs: Male with 3-5 pairs of spermaries opening on
ventral surface behind brain and between anterior caecal diverticula. Fe-
male with 30-36 or more pairs of ovaries, each containing, when mature, 1
to 5 large ova and several small ovocytes; situated dorsally and laterally to
nerve cords. (Figs. 45, 46).

Geographical distribution : Known only from the Beebe eight-mile area
off Bermuda, where it was obtained at depths of 1,400-1,800 m.

The species is named in honor of Dr. J. F. G. Wheeler, Director of the
Bermuda Station for Biological Research and a leading investigator of the
bathypelagic nemerteans.

Genus Planonemertes Coe.

8. Planonemertes labiata sp. nov.

(Fig. 52).

Two specimens of this new form were obtained. One of these was an
adult male measuring after preservation 21 mm. in length, 8 mm. in width
and 2 mm. in greatest thickness. The other was a female about 16 mm.
long and 6 mm. wide.

Body elongate oval, widest anterior to middle region, narrowing grad-
ually toward posterior end; rather thin even when strongly contracted, and
with very thin lateral margins continuous with thin caudal fin.

Color in life unknown; body translucent after preservation.

Body walls much thinner on dorsal surface than in many other forms
and very thin laterally. Much gelatinous tissue between the muscle bundles
and between the other organs in anterior half of body ; dorso-ventral muscles

1936]

Coe: Bathypelagic Nemerteans

107

well developed laterally and posteriorly. Ventral muscle plate several times
as thick as dorsal.

Proboscis sheath extends about four-fifths the length of the body ;
walls composed of three fairly distinct muscular layers, except for a short
distance posterior to proboscis attachment where longitudinal and spiral
fibers are interlaced. Farther back there is a well marked dilferentiation
into inner circular, longitudinal and outer circular or spiral musculatures,
although there are some spiral fibers extending between the inner and outer
layers; inner circular layer about as thick as two other layers combined.

Only the posterior portion of the proboscis was retained. This part
remained attached by the strong retractor to the wall of the sheath a short
distance anterior to the end of the rhynchocoel (Fig. 52). The longitudinal
fibers of the retractor were interlaced and in part continuous with the
longitudinal musculature of the proboscis sheath ; other fibers interlaced
with those of the two other muscular layers of the sheath. Posterior to the
retractor attachment the sheath becomes smaller and with thinner walls and
ends in the gelatinous tissue between the intestinal diverticula. No state-
ments can be made as to the armature of the proboscis or the number of
proboscidial nerves.

Mouth subterminal, separate from proboscis opening; with enormously
swollen lips in type specimen after preservation, due to strong contraction
of cephalic tissues and partial eversion of stomach during extrusion of pro-
boscis (Fig. 52). Caecal and intestinal diverticula much lobulated, but with-
out ventral branches. Caecal diverticula extend forward to brain region.
There are upwards of 20 pairs of intestinal diverticula.

Dorsal vessel enters rhynchocoel immediately behind brain, then passes
through the ventral wall of the sheath after a short distance and continues
posteriorly to join lateral vessels at posterior end of body.

Brain large; nerve cords with small but distinct dorsal fibrous core.
The nerve cords are situated about midway between median line and lateral
margins of body.

Reproductive organs: The type specimen was adult male with two clus-
ters of small spermaries on the ventral side of the head, immediately pos-
terior to the brain. (Fig. 52). The two clusters lie close to the median
line; each contains about 6 to 8 thin-walled spermaries with slender sperm
ducts, the ends of which protrude slightly from the body wall. The female
has the usual arrangement of paired ovaries between the intestinal diver-
ticula.

Geographical distribution: Known only from a depth of about 1,800 m.
at the Bermuda trawling area. The single previously described species of
the genus was taken in the Pacific ocean.

Genus Phallonemertes Brinkmann.

9. Phallonemertes murrayi Brinkmann.

Brinkmann, 1913, 1917, 1917a; Coe, 1926.

(Figs. 4, 24, 44).

Two specimens of this highly modified species were obtained by the
Bermuda expeditions. One of these was a large mature male measuring
40 mm. long, 10 mm. wide and about 2 mm thick; the second was a young
male 18 mm. long and 5 mm. wide.

Body elongated, with parallel sides, terminating posteriorly in a broad
caudal fin, usually recurved but not sharply demarcated from body (Figs.
4, 24) . Previously reported specimens varied from 34-61 mm. in length and
5-10 mm. in width. Color in life pink or red except caudal fin which is

108

Zoologica: New York Zoological Society

[XXI :9

translucent and colorless. Color due in part to numerous pigmented vacuoles
in alimentary canal.

Mouth and proboscis opening separate; when the proboscis is partially
everted the cephalic tissues may become so strongly contracted that the
mouth with its circular protruding lips may come to lie beneath the brain
and is thus widely separated from the proboscis opening. There are upwards
of 40-50 pairs of intestinal diverticula and 5 pairs of caecal diverticula,
somewhat lobed but without distinct ventral branches. Proboscis sheath
about three-fifths as long as body, with interlacing fibers; proboscis armed
with rather small, sharply curved basis bearing numerous conical stylets,
together with several shallow pouches of reserve stylets ; 13-17 proboscidial
nerves, with broad interneural plexus.

Lateral nerve cords with distinct dorsal core; dorso-lateral nerves rela-
tively large, with metameric connections with dorsal peripheral branches
of nerve cords.

Reproductive organs : Male with 4-7 pairs spermaries in single row on
each side of the body immediately posterior to the brain; each of these
when fully mature is provided with a long tubular spermatic duct projecting
as a penis far beyond the surface of the body (Fig. 44). Female with 20-30
or more pairs of ovaries, each of which produces upwards of a dozen ova
and opens on ventral surface lateral to nerve cord.

Geographical distribution: Depth 1,500-1,800 m.; previously reported
from various localities in the North Atlantic from Lat. 35° to near the
southern point of Greenland, at depths of 1,600-2,000 m.

Genus Chuniella Brinkmann.

10. Chuniella lanceolata Brinkmann.

Brinkmann, 1917, 1917a; Coe, 1926.

The single specimen of this species contained in the collections is of
particular interest because it represents the female of a form in which the
male only was previously known.

Body elongated, not much flattened, narrowed and pointed posteriorly,
without caudal fin; length of Bermuda specimen 5 mm., width 1 mm.; type
specimen 10 x 2.25 mm. Body walls with thick dorsal and ventral longitu-
dinal muscle plates, some of the muscular fibers being of comparatively gi-
gantic size; circular muscular layer very thin.

Mouth and proboscis opening separate ; upwards of 30 pairs unbranched
intestinal diverticula and 5 pairs of similar caecal diverticula. Proboscis
sheath nearly as long as body, walls of separate muscular layers; many of
the spiral fibers are remarkable for their relatively enormous size and dis-
tinct cross striations. Proboscis large, about twice as long as body, with
relatively large, curved basis and obtusely conical stylets; 21-23 probosci-
dial nerves.

Brain very large as compared with size of body; lateral nerve cords
close against ventro-lateral body wall, with single fibrous core except for
imperfectly differentiated dorsal core anteriorly.

Reproductive organs: Male with about a dozen spermaries in an irregu-
lar row on each side of body immediately posterior to brain. Female with
upwards of 20 pairs of ovaries, each with two or three large ova.

Geographical distribution: The specimen here recorded came from a
depth of 1,646 m. ; the type specimen was taken in the North Atlantic
southwest of Ireland at a depth of 1,000 m.

1936]

Coe: Bathypelagic Nemerteans

109

Genus N ectonemertes Verrill.

11. N ectonemertes mirabilis Verrill.

Verrill, 1892; Brinkmann, 1917, 1917a; Coe, 1926; Wheeler, 1934.

(Figs. 12, 13, 32).

Throughout the entire extent of the North Atlantic ocean this is by
far the commonest of all species of bathypelagic nemerteans. A total of
125 specimens has been previously reported. It was represented in the
Bermuda collections by 32 specimens, about equally divided between the two
sexes and including several young individuals, the smallest being only
8 mm. long. All conformed with previous descriptions of the growth stages
and adults. They were taken at various depths between 1,000 and 1,850 m.,
indicating a considerable vertical range in a single locality. Including one
large male collected off the coast of Cuba by the Bingham Expedition, and
not previously reported, the species is now known to extend through both
the North and South Atlantic oceans from the latitude of southern Green-
land, southward through the tropics to the latitude of South Africa.

Genus Balaenanemertes Burger.

12. Balaenanemertes minor sp. nov.

(Figs. 5, 6, 53).

The single individual of this species found in the collections measured
only 4 mm. in length, 2 mm. in width and 1 mm. in thickness, although it
was a fully mature female with ripe ova. If this specimen is truly repre-
sentative of the species, the individuals belonging to it are considerably
smaller than those of any other known bathypelagic form.

The species somewhat resembles B. lobata and B. musculo caudata but
differs in position and size of tentacles, in number of proboscidial nerves,
in number of intestinal and caecal diverticula and in other details.

Body small, oval, somewhat narrowed posteriorly with distinct, bilobed
caudal fin ; tentacles small, situated well back of head, with imperfectly de-
veloped musculatures. Body walls extremely thin, with thin basement layer ;
very little parenchyma between intestinal diverticula and body walls.

Proboscis sheath nearly as long as body, with delicate musculature ih
two layers, inner longitudinal and outer spiral, except on ventral side, where
longitudinal layer is lacking. Proboscis large, with 14 distinct nerves in
most sections, although one of these may be lost in the interneural plexus
in some sections and an additional one may be represented in others. Longi-
tudinal muscles of retractor interwoven with both longitudinal and spiral
layers of dorso-lateral wall of sheath at a point about four-fifths the dis-
tance from anterior to posterior end of body. Proboscis armed with curved
basis of typical form bearing numerous small conical stylets, also with at
least three shallow pouches of accessory stylets, of which as many as six
may be present in one pouch.

Mouth well separated from proboscis opening; esophagus wanting;
stomach simple, walls not convoluted in the type specimen; pylorus short;
caecum with median branch reaching to brain and one or two pairs of
voluminous, lobed diverticula, of which the anterior lobes reach the brain;
pylorus in this specimen opens into dorsal wall of intestine in such a position
as to make it uncertain whether the second pair of diverticula belong to
caecum or intestine; about 13 pairs lobed intestinal diverticula, without
ventral branch.

Lateral nerve cords lie close to latero-ventral body walls, with single
fibrous core ; with usual commissure • above rectum ; nerve cord muscles

small.

110

Zoologica: New York Zoological Society

[XXI :9

Dorsal vessel ends blindly after a short distance in rhynchocoel. Cephalic
lacunae large, with large corpuscles.

Reproductive organs: Female with 4-5 pairs of ovaries, each with single
enormous ovum when mature. Oviducts large, opening ventrally between
nerve cords and median line; ovarian wall much convulated after discharge
of ovum (Fig. 53). Male unknown.

Geographical distribution: Known only from Beebe eight-mile area off
Bermuda, where it was taken in a net drawn at a depth of about 549 m.
None of the other nemerteans in this area was found so near the surface,
but other species of the same genus have been collected at various localities
at similar depths as well as very much deeper.

Bibliography.

Brinkmann, August.

1913. Bathynectes murrayii n. gen. n. sp. Eine neue bathypelagische Nemer-
tine mit ausseren mannlichen Genitalien. Bergens Museums Aarbok,
1912, no. 9, p. 1-9, PI. 1.

1917. Pelagic nemerteans. Rept. Sci. Results Michael Sars North Atlantic
Deep Sea Exped., 1910, 3, p. 1-20, 2 Pis.

1917a. Die pelagischen Nemertinen (monographisch dargestellt) . Bergens
Museums Skrifter. 1917, ny raekke, 3, 8; p. 1-194, 16 Pis.

Coe, W. R.

1926. The pelagic nemerteans. Mem. Mus. Comp. Zook, Harvard Coll., 19:
p. 1-245, 30 Pis.

1935. Bathypelagic nemerteans collected within a 25 mile circle near Ber-
muda. Zool. Anzeiger, 111: p. 315-317.

Verrill, A. E.

1892. The marine nemerteans of New England and adjacent waters. Trans.
Conn. Acad., 8. p. 382-456, 7 Pis.

Wheeler, J. F. G.

1934. Nemerteans from the South Atlantic and Southern Oceans. Discovery
Repts., 9: p. 215-294, 2 Pis.

1936]

Coe: Bathy pelagic Nemerteans

111

EXPLANATION OF THE PLATES.

The lettering on the plates includes the following abbreviations:

a — anus

ac — anterior chamber of proboscis
br — brain

eg — cerebral ganglia
dv — dorsal blood vessel
ic — intestinal caeca
id — intestinal diverticula
go — glandular organ
In — lateral nerve cord
Iv — lateral blood vessel
m — mouth

ov — ovaries
p — proboscis

pa — proboscis attachment musculature

pc — posterior chamber of proboscis

ps — proboscis sheath

rc — rhynchocoel

ro — rhynchodeal opening

sb — stylet basis

sp — spermaries

st — stomach

t — tentacle

Plate I.

Figs. 1-15. Photographs of preserved specimens; all except Fig. 5 approximately
one-third larger than natural size.

Fig. 1. Plotonemertes adhaerens. Body recurved dorsally; proboscis everted;
lips protruding.

Figs. 2, 3. Paradinonemertes wheeleri. Male and female; body strongly con-
tracted and much flattened.

Fig. 4. Phallonemertes murrayi. Male, showing 6 pairs of spermaries.

Figs. 5, 6. Balaenanemertes minor. Female with large ovaries and everted pro-
boscis.

Figs. 7, 8, Plotonemertes adhaerens. Proboscis partially everted.

Figs. 9, 10. Paradinonemertes wheeleri. Female and young male.

Fig. 11. Same. Large individual with extruded proboscis.

Figs. 12, 13. N ectonemertes mirabilis. Young male and female, latter with ex-
truded proboscis.

Figs. 14, 15. Plotonemertes adhaerens. Females with proboscis retracted.

Plate II.

Figs. 16-27. Photographs of preserved specimens; all approximately one-third
larger than natural size.

Fig. 16. Protopelagonemertes hubrechti. Female with everted proboscis.

Fig. 17. Crassonemertes robusta. Female with partially everted proboscis.

Fig. 18. Protopelagonemertes hubrechti. Very large female. Posterior extremity
broken off.

Figs. 19, 20. Plotonemertes aurantiaca. Large female with everted proboscis.

Posterior end of body sharply recurved dorsally in position of glandular
organ.

Fig. 21. Crassonemertes robusta.

Fig. 22. Plotonemertes adhaerens.

Fig. 23. Paradinonemertes wheeleri. Large female showing intestinal diverticula.

Fig. 24. Phallonemertes murrayi. Male, showing six pairs of spermaries.

Fig. 25. Protopelagonemertes beebei. Body abnormally flattened; proboscis par-
tially everted.

Fig. 26. Plotonemertes adhaerens. Body compressed.

Fig. 27. Same. Small individual with body strongly recurved dorsally.

112

[XXI :9

Zoologica: New York Zoological Society

Plate III.

Fig. 28. Protopelagonemertes hubrechti. Very large female, showing dorsal and
lateral keels; proboscis slightly everted; A, dorsal surface; B, lateral
view; C, transverse section of body. Twice natural size.

Plate IV.

Fig. 29. Protopelagonemertes hubrechti. Portion of proboscis, showing portions
of anterior ( ac ) and posterior chambers (pc), stylet basis ( sb ) and posi-
tion of accessory stylet pouches.

Fig. 30. Same. Stylet basis more highly enlarged.

Fig. 31. Same. One of the six pouches of accessory stylets.

Fig. 32. Nectonemertes mirabilis. Anterior portion of body of young male, show-
ing small tentacles and the position of the spermaries ( sp ).

Fig. 33. Plotonemertes aurantiaca. Stylet basis.

Plate V.

Fig. 34. Protopelagonemertes beebei. Female, showing partially everted pro-
boscis (p), extent of proboscis sheath ( ps ) and position of ovaries (ov) ;
ic, caecal diverticula.

Fig. 35. Crassonemertes robusta. Female, showing partially everted proboscis
(p) , extent of proboscis sheath (ps), ovaries (ov), stomach (st), and
the profusely branched intestinal diverticula (id).

Plate VI.

Fig. 36. Plotonemertes adhaerens. Male with everted proboscis, showing extent
of proboscis sheath, position of spermaries (sp) and glandular organ
(go). Posterior extremity recurved.

Figs. 37, 38. Same. Ventral and lateral views of glandular organ (go).

Figs. 39, 40. Same. Configuration of glandular organ when posterior extremity
of body is sharply recurved dorsally.

Fig. 41. Same. Proboscis with stylet basis.

Fig. 42. Same. Stylet basis and the six pouches of accessory stylets.

Plate VII.

Fig. 43. Plotonemertes aurantiaca. Female from ventral surface, showing small
ovaries (ov) and the well-developed glandular organs (go).

Fig. 44. Phallonemertes murrayi. Male from ventral surface, showing extent of
proboscis sheath (ps), arrangement of intestinal diverticula (id), posi-
tion of mouth (m) and the six pairs of spermaries (sp), each with slen-
der, protruded penis.

Plate VIII.

Fig. 45. Paradinonemertes wheeleri. Male, showing extent of proboscis sheath
(ps) and the three pairs of spermaries (sp).

Fig. 46. Same. Female, showing position of ovaries (ov) ; m, mouth; eg, cerebral
ganglia; dv, dorsal vessel; sb, stylet basis.

Fig. 47. Same. Stylet basis and stylets.

Plate IX.

Fig. 48. Paradinonemertes wheeleri. Female; left half from dorsal surface and
right half from ventral surface, showing position of ovaries, oviducts
(od) and intestinal diverticula (id) ; and extent of proboscis sheath (ps) ;
eg, cerebral ganglia; dv, dorsal vessel; In, lateral nerve cord; Iv, lateral
vessel; sb, stylet basis in proboscis.

Fig. 49. Same. Stylet basis.

1936]

Coe: Bathy pelagic Nemerteans

113

Fig. 50. Pachynemertes obesa. Strongly contracted specimen, showing mouth
and everted proboscis.

Fig. 51. Same. Outlines of stylet basis and stylets.

Plate X.

Fig. 52. Planonemertes labiata. Outline of body of type specimen, showing
enormously swollen lips surrounding the mouth (m), the clusters of
cephalic spermaries (sp) , intestinal diverticula (id), rhynchocoel (rc),
and posterior end of proboscis ( p ).

Fig. 53. Balaenanemertes minor. Female with three very large, ripe ova; od,
preformed oviduct; ov', recently emptied ovary; id, intestinal diverticula;
m, mouth; pa, attachment of proboscis near posterior end proboscis
sheath (ps) ; sb, stylet basis in everted proboscis; t, minute tentacle.

COE

PLATE I.

BATHYPELAGiC NEMERTEANS TAKEN IN THE YEARS 1929. 1930 AND 1931

\

COE.

PLATE II.

BATHYPELAGIC NEMERTEANS TAKEN IN THE YEARS 1929. 1930 AND 19.T

COE.

PLATE III.

BATHYPELAGIC NEMERTEAN'S TAKEN IN THE YEARS 1929. 19:10 AND 1931

L .KRAUSE ofl.

BATH Y PELAGIC NEMERTEANS TAKEN IN THE YEARS 1929. 1930 AND 1931

COE.

ELATE V.

34

L.KRAUSE del.

_ .- ps

QV.

HATHYPELAGIC NEMERTEANS TAKEN IN THE YEARS 1929. 1930 AND 1931

COE.

IM.VTi: \ I.

sp.

BATHYPELAGIC NEMERTEANS TAKEN IN THE YEARS 1929. 1930 AND 1931 ,

COE.

PLATE VII.

BATHYPELAG1C NEMERTEANS TAKEN IN THE YEARS 1929. 1930 AND 1931

COE.

PLATE Mil.

l.KRAUSE

BATHYPELAGIC NEMERTEANS TAKEN IN THE YEARS 1929. 1930 AND 1931

COE.

PLATE IX

BATHYPELAGIC NEMERTEANS TAKEN IN THE YEARS 1929. 1930 AND 1931

COE.

PLATE X.

BATHYPELAGIC NEMERTEANS TAKEN IN THE YEARS 1929, 1930 AND 1931 .

Breder: Tissue Culture and Explantation in Nature

115

10.

Tissue Culture and Explantation in Nature: A Review of Certain
Experiments and Possibilities.

C. M. Breder, Jr.

New York Aquarium.

Introduction.

The possible methods of operation of organic evolution have for long
attracted the speculative faculties of biologists. Phenomena concerned
with the general aspects, as well as special phases, have occupied prominent
places in biological controversy. In spite of this great discussional activity
on the part of students, there is good reason to suppose that many possi-
bilities exist that have not been examined or even imagined. Until all such
conceivable methods have been appraised, the extent of the effect of those
whose reality is established must remain an open question.

The hypothesis subsequently set forth may serve as an example and
undoubtedly numerous others could be developed. All such hypotheses must
either be satisfactorily refuted or established in order to develop a thor-
oughly adequate and fully acceptable evolutionary scheme. The present dis-
cussion deals with a very special case, but, as with all such matters, its
probable limits of application cannot be easily anticipated at this time.

The argument discussed herewith is thoroughly documented with ref-
erences to the literature of tissue culture, but as nearly every paper on the
subject bears in some way on the present views, only a few have been
selected for definite mention. These have been chosen because of their
specific application to the points under discussion. See especially the biblio-
graphy of Lewis and Lewis (1924) and the Arch. Exper. Zellforsch, ed.
Erdmann, R. The writer has been given much assistance in assembling the
bibliographic data by Dr. J. N. Gowanloch, but is more especially indebted
to him for valued criticisms and suggestions.

Factors in Tissue Culture.

An abundant literature, as above noted, has established the fact that
animal and plant tissues may be readily cultured when removed from the
organism of which they form a part. Such explants, if placed in a suitable
medium, will perform their natural functions including those of reproduc-
tion. The following brief consideration of pertinent factors will^ serve to
establish the basic data necessary for the purposes of the present discussion.

1. Cellular immortality: The small confines of a hanging drop has
been sufficient for much tissue culture work, but due to the rapid loss of
nutrient value and the accumulation of toxic wastes, such are necessarily
limited to short periods unless frequent recourse is made to sub-culturing.
With careful renewal of media, explanted cells may be cultured indefinitely
and give every indication of being as capable of perpetuating their kind as

116

Zoologica: New York Zoological Society

[XXI :10

any natural organism. Tissue strains from chickens have already reached
ages far exceeding the life span of the deriving animal type and show no
abatement in vigor. Carrel (1912), Ebeling (1913 and 1922), Baker and
Carrel (1926a and b).

2. Media: A great variety of solutions is capable of supporting tissue
growth and there is a large range of permissible variation in tonicity, pH
value, and other important cytologic factors, Lewis and Lewis (1924). As
media are usually directly or indirectly imitative of plasma, with or without
the addition of nutrient substances, they necessarily tend also to approxi-
mate dilute ocean water. It has been shown that various dilutions of sea
water itself make satisfactory fluids for cultures in vitro, Lewis (1916),
Dederer (1921).

3. Influence of environment: Many papers discussing cultures of ex-
tirpated cells make mention of changes induced by modifications of the
culture media. Other environmental influences are responsible for appro-
priate modifications in morphology or behavior. Which of these can be con-
sidered as genetic, and which the direct impress of environment on each
individual cell, is still largely controversial, Uhlenhuth (1915 and 1916).
Experiments involving the use of X-rays, however, suggest the former,
as might be expected, Strangeways (1924).

4. Limits of space: One of the effects of small stagnant bodies of
media is outlined under “Cellular immortality.” Another is that of excessive
bacterial infection. Consequently, tissue culture is usually carried on under
sterile conditions in order to prevent the establishment of destructive bac-
terial colonies. Many kinds of bacteria are preeminently suited to rapid
proliferation in a nutrient hanging drop, whereas the explanted cells have
no such natural advantage. The same condition is presented in the main-
tenance of any aquatic organism in a laboratory jar. In any such case the
difficulty regarding offensive bacteria is directly proportional to crowding
and its attendant effects. It may be readily demonstrated that this is, in
oart. a function of the quantity of fluid in relation to the organic bulk. It
is, therefore, all the more remarkable that numerous students have been
able to grow both cells and bacteria in a common drop. Furthermore, they
have studied the ensuing interactions and in some cases phagocytosis was
observed to be of apparent nutrient value, Johnson (1915), Smyth (1916a
and c) . The presence of bacteria also apparently acts as a growth-promot-
ing stimulus. Smyth (1916b) wrote of his work, “These results seem to
indicate that many bacteria may be utilized by tissue cells as food for
growth, or may contain a substance or substances stimulating cell growth
or multiplication.”

If the preceding four items are considered with mutual reference it
becomes apparent that there is no evident reason to prevent the establish-
ment of a culture combining the factors that have already been established
separately by a large number of independent investigators. Such a culture
would require only some fluid medium of natural occurrence, such as dilute
sea water, of sufficient bulk and renewal to prevent an excessive bacterial
growth. In other words, the limitations of the tiny bodies of media ordi-
narily employed, with their consequent favoring of bacteria, must be cir-
cumvented. Suggested methods include the following: The fluid medium, in
proportion to the organic matter, must be of such a quantity that bacteria
have not enough dispersed food to form a dangerously rich culture. This
must be below some critical value. In this connection hints are to be found
in the maintenance of standing water aquaria in which the food substances
must remain below a certain point for similar reasons. An alternative
would be a circulation of the medium which might be worked out on a
modification of the technique of Burrows (1912). These concepts are ob-
viously derived directly from aquarium practice and it should now be ap-
parent that there is no trenchant difference between the fundamentals of
tissue culture and any form of animal or plant husbandry. In all, the

1936]

Breder: Tissue Culture and Explantation in Nature

117

desired organisms are retained in more or less restricted confines and must
be protected from enemies, fed, and freed of their own toxic wastes. The
smaller the environment in relation to the number of organisms the greater
the difficulty in maintaining a persistent culture. In the case under con-
sideration food might have to be introduced manually from time to time in
order to avoid nutrient fluids so especially beneficial to bacterial increase.

Experiments such as outlined above are now under way but it has
been deemed best to place the above facts and the hypothesis they support
on record at this time for the following reasons: The separate elements of
the experiment have already been established independently by others. The
only service the proposed experiments would perform would be to join
certain separate factors. To be of real value it would have to be an operation
involving a long period. Negative results might only be an indication of
lack of skill on the part of the operator, or some technical difficulty. Others,
better equipped than the writer, may be able to contribute more readily to
the establishment of proper technique. Some of the problems encountered
and the difficulties involved in the experiments thus far performed are dis-
cussed subsequently.

Inferences Based on Tissue Culture.

Since all of the conditions discussed are met with in a state of nature,
there is good reason to consider the possibility of cultures of this sort aris-
ing spontaneously. The natural occurrence of the specific requirements may
be considered for comparison with corresponding factors in laboratory cul-
tures.

1. Sources of materials: Any event which causes an animal or plant
to part with living cells supplies potential material. Such would include
destruction by predacious forms, fighting, accidents, physical malfunction-
ing, such as hemorrhage, and in certain cases normal functioning. Without
going into explicit detail it is sufficiently evident that wherever there is any
form of life, parts are being continually separated from individuals in rela-
tively immense quantities. Human foetal membranes and menstrual mucosa
were successfully cultured by Konrad (1928). Amoebocytes and other
cellular elements are normally shed in a living state by many organisms,
if not by all.

2. Media: Any natural water that is suitable as a culture medium is
potentially available for natural explants. For example, brackish sea water
has been shown experimentally to be such and is frequently closely similar
to plasma. The possibilities would thus vary, both specifically and geograph-
ically.

3. Influences of environment: Other environmental factors would act
favorably or not, according to the nature of the experiment ; e.g. explants
from poikilothermal animals would have little likelihood of being destroyed
by the temperature of the environment but homoiothermal ones might re-
quire water of approximately body temperature. That explanted cells, even
of homoiothermal forms, have a truly remarkable resistance to unfavorable
conditions of a most extreme kind has been repeatedly demonstrated. See,
for example, Rous and Jones (1916), Nageotte (1927) and Morosow (1928a
and b) .

4. Limits of space: In this regard, the natural culture would be much
more advantageously situated than the laboratory drop. It is the same
kind of difference that obtains between a small aquarium and a body of open
water. In a state of nature the food is automatically supplied and toxic
wastes are either washed away or rendered innocuous.

5. Enemies: Instead of being protected from enemies, natural explants
would have to take the chance of any organism invading a new environment.
Other things being equal, the dangers from bacteria would be vastly less
than in a laboratory culture. This limits successful establishment to rea-

118

Zoologica: New York Zoological Society

[XXI :10

sonably “clean” environments and rules out those that are characterized by
a large amount of decomposition. In some environments, consisting of long-
standing organic equilibria, bacteria seem to be at a disadvantage because
of the presence of a lysisime or bacteriophage. Such conditions not infre-
quently may be found in standing aquaria, Breder (1931). In this connec-
tion it is notable that the cells in their original locations within an animal
body are not in a sterile environment. The sterility of a culture medium
is chiefly a concession for the enforced disadvantages of the cramped quar-
ters, comparable to the omission of the natural enemies of fishes in an
aquarium. Considering such explants as invading organisms it should be
borne in mind that they would usually be reinforced continually by similar
cells from the original source. An example would be the case of a predacious
animal feeding chiefly on a single food animal (a type of specificity of
frequent natural occurrence) with the continued escape of its tissues on be-
ing crushed. Since embryological material is generally more suitable for
tissue culture, it may be noted that gravid females falling prey or the rob-
bing of nests would abundantly supply much material. A hypothetical case
is given below for illustrative purposes.

The arena might be a newly formed tropical swamp where the sea has
inundated a low fresh water bog, depressing the quantity of its micro-fauna
(protozoans, bacteria, et cetera), and in which the ocean water fauna (some-
what dilute) has not yet thoroughly established itself, but in which currents
have washed various areas clean. Crocodiles could be feeding rapaciously
on small animals, water fowl, et cetera, and so release — in some cases as
single cells — the various constituents of the blood and in a less quantity
epithelium, foetal material, et cetera.

It is experimentally demonstrable that an abundance of such explants
are actually being continually released wherever there is a struggle for
existence, that suitable natural media are clearly of worldwide occurrence
and that the potency of the chief primary enemies, bacteria, is generally
inversely proportional to the size of the fluid environment. Since the first
appearance of the earliest metazoans the proper constellation of factors
could easily have occurred many times. With this as a working hypothesis,
we may consider some of the more apparent philosophical implications.
Either such explants are capable of continued life, as is indicated by long-
time cultures, or they are not. If the latter is true they would all even-
tually die off, but in the interim would exist to confound protozoan sys-
tematics. Attention need hardly be called to the striking similarity in
morphology and behavior between free metazoan cells and protozoa, e.g.
Amoeba and phagocytes, Hogue (1922), or ciliates and the singly freed cells
of echinoderm eggs, Jenkinson (1909). Since numerous tissue culturists
have shown that various environmental factors influence the morphology
and behavior of their explants, natural cultures would certainly show com-
parable changes and probably bear little resemblance to their deriving
tissues.

It has been shown that protozoans do not necessarily require fertiliza-
tion to maintain vitality in subsequent generations. In eleven thousand
generations of Paramecium, Woodruff (1926) found perfectly normal in-
dividuals produced by binary fission, but he did find some activity within
single individuals involving a rearrangement of nuclear material for which
he created the term endomixis, Woodruff and Erdmann (1914), Wood-
ruff (1925), Calkins (1926). Behavior of nuclei of cultured tissues strangely
resembling endomixis has been repeatedly observed, Holmes (1914), Mack-
lin (1916) and Dederer (1921). There is no reason to believe that there
is any fundamental functional difference between such acts and the ob-
servations of Woodruff. With these facts at hand it is quite conceivable
that explanted cells under proper environmental stimulus might even show
other than binary fission. It would be interesting to know whether the
nuclear adjustments in tissue cultures are stimulated by explanting, or

1936]

Breder: Tissue Culture and Explantation in Nature

119

whether they represent normal activity. If the latter is true the connection
of the possible loss of such activity with senility and the relatively poor
behavior of adult material in vitro is suggested. It is to be noted in this
connection that while explanted cells usually retain their individual cell
characteristics they frequently show dedifferentiation pointing to a less spe-
cialized and somewhat embryonic condition, Lewis (1920), Fisher (1922).
These effects are clearly involved in the lack of endocrine and other control.
For example, there is evidence that plasma from old animals checks growth
in vitro, Baker and Carrel (1926 a and b).

Contemplating the other alternative, that natural explants are capable
of survival, there is nothing remaining to separate them from the protozoa,
and as they have been experimentally shown to be reactive to environmental
stimulus there is no reason to suppose that they lack the evolutionary possi-
bilities attributed to other forms of life. With such an assumption we
should have “protozoa” derived directly from metazoa but charged with
specific potentialities of various sorts dependent on the functions of their
immediate ancestors as parts of a metozoan unit. Geneticists may object to
this as a possibility on a basis of theory involving the continuity of germ
plasma as distinct from somatic structure. However, there appears to be
no theory or genetical experiments that have any bearing on the matter
and the distinction referred to cannot be extended to the present case, for
with self-sustaining and reproducing cells entirely free from the originating
metazoan unit they at once may be considered germinal and somatic in every
sense that a conventional protozoan is so considered.

A view set forth by Marchand (1935) would seem somewhat to approach
the present in that it attempts to derive solitary coelenterates from colonial
forms. Actually this view has little in common with the present, since the
argument runs to the effect that evolution from “single cells” to complex
forms was through colonial forms which subsequently separated, rather
than direct. Be that as it may, the present hypothesis concerns, itself solely
with the ability of organic fragments to survive and to evolve.

The consequences of the present view would be to upset the unity of
phylogeny and if shown tenable would introduce complications in a number
of disciplines. Just how far this might be carried is of course very uncer-
tain, but it could conceivably account for the lack of convergence in geologic
time of certain phyla. It is difficult to imagine the tracing of a major
phylogenetic tree with such a condition obtaining. As all the conventional
phyla extend a great distance in geologic time it would follow that an
origin from such a source could have occurred only in the remote past,
except possibly considering the protozoa as a mixed phylum. This would
tend to discourage the entire thought, were it not for the fact that there
are a large number of organisms that have no convincingly evident affinities
and have been buffeted about in a vain effort to find a place for them, as is
witnessed by any zoological text book with its numerous “appendices” or
“incertae sedis.” For example, Parker and Haswell (1910) list eleven phyla
and to these are added seven “appendix” forms. Without attempting to
press the speculation too far it is clear that this condition obviates any
necessity to allocate such an origin of forms to any particular geologic
period.

A consideration of the general geographical requirements of derivation
might be somewhat as follows : Tropical, homoiothermal vertebrates ; any
climate, poikilothermal vertebrates, invertebrates, plants.

Coupled with this list must go such factors of environment as suitable
chemical quality of the water, a proper osmotic pressure, a sufficient free-
dom from enemies, et cetera (i.e. a sufficiently suitable environment).

As it has been shown that a variety of transplants is possible from one
organism to another, for example blood transfusions, glandular transplants,
plant and animal grafting, et cetera, on the basis of the preceding, occasion-
ally natural transplants might well originate. The remarkable production

120

Zoologica: New York Zoological Society

[XXI :10

of complex plant monsters, Winkler (1914), and the composite Hydras of
Issayev (1924), both show the ability of very different cells to live side by
side. If recent explants are ingested or lodged in surface wounds the pos-
sibility of their survival arises. Certain carcinoma cells similar to their
associates except in growth rate come to mind, including even the possibility
of self-infection. It is noteworthy that malignant growths are most common
at first near the surface or in the alimentary or other tract open to the ex-
terior. Thus, a return cycle further complicating conditions would have
bearing on a large number of academic and practical matters. At least it
is certain the symbiotic relationships must have arisen in some such acci-
dental manner. Hydra viridis, Convoluta roscoffensis, the lichens, or even
man and his intestinal flora may serve as examples of various kinds of such
associations. Buchsbaum and Buchsbaum (1934) produced what they con-
sidered an artificial symbiosis between tissue culture cells and the algae,
Chlorella.

Passing beyond the limits of the indicated inferences of this hypothesis,
we might even imagine bacteria to be mitochondria released in a similar
manner. This would be inverse to the idea set forth by various students,
most recently Wallin (1927), which view is objected to by Cowdry (1924).
Viruses might be thought of as also being involved and derived from the
fluids of ruptured cells. Schultz (1930) gives a general view of the behavior
of viruses very suggestive in this connection. See also the discussion by
Riddle (1936). It is recognized that the above is pure speculation and it is
mentioned merely as indicative of the lines of thought engendered by a
consideration of certain facts in the technique of tissue culture.

Difficulties of Experimental Procedure.

The experimental establishment of explanted tissues in relatively “wild”
environments would place the entire suggestion on a relatively firm founda-
tion. As previously noted there are both experimental, and, in the case of the
writer, personal difficulties involved. Nevertheless, a considerable amount
of experimentation was undertaken. For verv substantial aid in this the
writer is indebted to Dr. R. F. Nigrelli who labored with most of the actual
physical material.

The following discussion of the results of this work is introduced
chiefly to point out that in addition to ordinary experimental difficulties
there are likewise theoretical obstacles to the establishment or destruction
of the present hypothesis by experimental means. The discussion of certain
experiments may serve to demonstrate the point.

The leucocytes of invertebrates were thought of as likely material for
such experimentation. For example, those of the common oyster normally
invade the mantle cavity and are frequently voided into the surrounding
water, especially under slightly suffocating conditions, Orton (1924),
Young (1928). This phenomenon, diapedeses. is apparently common to a
variety of animals. Leucocytes so voided by the American species, Ostrea
virginica, were found to live for as long as six days in a laboratory dish
with no attention whatsoever, Breder and Nigrelli (1933), while those of the
European oyster, O. edulis, lived for Orton up to four days. Such, then,
would seem to be ideal material.

However, it so happens that 0. virginica is infested with amoebic para-
sites (or commensals?), Vahlkampfia calkinsi, or V. patuxent or both. These
are usually found in the intestinal tract, are voided with the faeces, and
bear a strong superficial resemblance to the leucocytes. The latter likewise
invade the intestinal tract in the course of their functional activity. The
describer of these parasites, Hogue (1915, 1921 and 1922), was only able to
distinguish them from the leucocytes after fixation, when with suitable
staining the nuclear material was found to be differently arranged. Breder

1936]

Breder: Tissue Culture and Explantation in Nature

121

and Nigrelli (1933) found in their dishes that in addition to parasites
voided in the faeces, and leucocytes voided from the gill chamber, they also
had a free living Amoeba of the Umax type that normally lives on the ex-
terior of the shell. This latter form caused no confusion, however. In agree-
ment with Hogue they found (unpublished) that pure cultures of leucocytes
withdrawn from the heart would not survive on agar plates for any great
length of time, but that the parasites from the intestine could be so cultured.
Thus it follows that the differentiation of leucocytes from parasites is de-
pendent on (1) arrangement of nuclear material, for which the material
must be killed and stained, and (2) ability to grow on agar plates.

Leucocytes from the heart will not grow on plates but material from
the intestine will. Both parasites and leucocytes (as based on stained ma-
terial) are found there, but on old agar plates only the “parasite” type of
nucleus is found. It seems to the writer that there is just an even chance
that these “parasites” may be one phase of the normal oyster leucocyte, es-
pecially since they always seem to be present. If they could be shown to be a
phase or type of leucocyte occurring only in the gut, which has the possi-
bility of exterior survival, this could be used for considerable support of the
hypothesis. As it stands — parasite, commensal or leucocyte — it is clear that
any point of view can be argued and experiments may prove one or the
other, depending on the experimenter’s bias, with no present hope of ex-
perimentally further separating the material.

Giving this line up, earthworms were examined, since they void amoe-
boid cells with their casts. Without going into details it may be stated that
earthworms also harbor amoeboid parasites and a similar block to this line
checked an experiment that seemingly held promise and made us wonder if
all commensal or innocuous amoebic parasites were simply transformed
leucocytes.

The application of micro-dissection to the problem involves a further
philosophical consideration but points the way to the types of material that
may hold some hope for experimental verification. If certain types of ani-
mals are disassociated the pieces will reunite to re-form the originals.
Sponges, perhaps, represent the best known case of this sort. Hydra will
also show this phenomenon, Papenfuss (1934), but under certain conditions
will not. The fate of the individual cells under such conditions is not yet
certain. Obviously, freed cells that reunite to construct organisms cannot
be expected to be of much value for this kind of experiment. On the other
hand highly specialized and protected cellular elements could hardly be
expected to survive without the complex of conditions under which they nor-
mally exist. Consequently the type of tissue that presumably must be
sought after in this connection is something sufficiently unspecialized as to
be able to survive in a new environment and still without the ability to re-
construct an individual animal with its fellows. In this connection it may
be pointed out that it is sometimes surprising to note what ordinarily well
protected tissues may do in the way of survival under exposed conditions.
Nigrelli and Breder (1935) describe a prolapsed fish intestine, which while
fed with body juices was exposed to the ordinary standing water of an
aquarium. This pendant piece proliferated for several months and was
finally killed for study.

Conclusion.

The hypothesis that animal and plant cells when dislodged from their
original locations in situ by natural causes may continue living independ-
ently as distinct organic units rests on a large number of concrete experi-
mental demonstrations by independent investigators. These contributions
were all made with reference to special and distinctly diverse problems not
in the least connected with the present integrated interpretation of them.

122

Zoologica: New York Zoological Society

[XXI :10

That they are adequate and pertinent to this hypothesis can be sustained
by reference to the total literature of tissue culture and the absence therein
of any contrary findings. Experimental verification, however, must wait on
the development of a more satisfactory approach than is now available. The
author, at least, has thus far been unable to devise a practicable experiment,
the results of which can be interpreted in but one way. The continued pres-
ence of this duality of possible interpretation stands as an impediment to
experimental analysis of the problem. The conception of a critical experi-
ment must be realized before further progress can be expected. That tissue
culture has been possible and that a large variety of cells have been grown
and have perpetuated themselves for long periods in a considerable variety
of environments may be taken as good presumptive evidence in favor of this
hypothesis.

Bibliography.

Baker, L. E. and Carrel, A. 1926 A. Au Sujet du Pouvoir inhibiteur du Serum
pendant la Vieillesse. Compt. Rend. Soc. Biol ., 95, pp. 958-960.

........ 1926B. Action on fibroblasts of the protein fraction of embryonic tissue

extract. Journ. Exp. Med., 44, pp. 387-407.

Breder, C. M. Jr. 1931. On organic equilibria in aquaria. [Abstract.] Copeia,

(2), p. 66.

Breder, C.M. Jr. and Nigrelli, R. F. 1933. Lamellibranch leucocytes as living
material for class-room demonstration. Science, 78, (2015), p. 128.

Buchsbaum, R. and Buchsbaum, M. 1934. An artificial symbiosis. Science , 80,
(2079), pp. 408-409.

Burrows, M. T. 1912. A method of furnishing a continuous supply of new me-
dium to a tissue culture in vitro. Anat. Rec., 6, pp. 141-144.

Calkins, G. N. 1926. The Biology of the Protoza. Lea and Febriger, Philadel-
phia, pp. 1-623.

Carrel, A. 1912. On the permanent life of tissues outside of the organism.
Journ. Exp. Med., 15, pp. 516-528.

Cowdry, E. V. 1924. General Cytology. University Chicago Press, Chicago, pp.
1-754.

Dederer, P. H. 1921. The behavior of cells in tissue cultures of Fundulus hetero-
clitus, with special reference to the ectoderm. Biol. Bull., 41, (4), pp. 221-240.

Ebeling, A. H. 1913. The permanent life of connective tissue outside of the or-
ganism. Journ. Exp. Med., 17, pp. 273-285.

........ 1922. A ten year old strain of fibroblasts. Journ Exp. Med., 35, pp.

755-759.

Fischer, A. 1922. A pure strain of cartilage cells in vitro. Journ. Exp. Med.,
36, pp. 379-384.

Hogue, M. J. 1915. Studies in the life history of an amoeba of the Limax group,
Vahlkampfia calkinsi. Arch. f. Protist., 35, (2), pp. 154-163.

1921. Studies on the life history of Vahlkampfia patuxent n. sp. para-
site on the oyster, with experiments regarding its pathogenicity. Am. Journ.
Hyg., 1, pp. 321-345.

1922. A comparison of an amoeba, Vahlkampfia patuxent with tissue

culture cells. Journ. Exp. Zool., 35, (1), pp. 1-11.

Holmes, S. J. 1914. The behavior of the epidermis of amphibians when culti-
vated outside of the body. Journ. Exp. Zool., 17, pp. 281-295.

Issayev, V. 1924. Researches on animal chimeras. Journ. Genet., 14 (3), pp.
273-351.

1936]

Breder: Tissue Culture and Explantation in Nature

123

Jenkinson, J. W. 1909. Experimental Embryology. Clarendon Press, Oxford.
pp. 1-341.

Johnson, J. C. 1915. The cultivation of tissues from amphibians. Univ. Calif.
Pub. in Zool., 16, pp. 55-62.

Konrad, H. 1928. Lebens — und Wachstumsbeobachtungen an Menschlichen
Geweben und Geschwulsten im Explantationsversuche und ihre Bedeutung
fur klinische Fragen. Arch. Gynakol. , 134, (2) pp. 250-309.

Lewis, M. R. 1916. Sea water as a medium for tissue cultures. Anat. Rec., 10,
pp. -287-299.

1920. Muscular contraction in tissue cultures. Contrib. Embryol., 9,

Carnegie Inst. Wash. Pub. 272, pp. 191-212.

Lewis, W. H. and Lewis, M. R. 1924. In Cowdry, E. V. General Cytology, Univ.
Chicago Press., Chicago, pp. 383-447.

Macklin, C. C. 1916. Amitosis in cells growing in vitro. Biol. Bull., 30, pp.
445-466.

Marchand, W. 1935. Remarks on the evolution of animal phyla. [Abstract.]
Anat. Rec., 61, (2), pp. 3-4.

Morosow, B. D. 1928A. Explantationsversuche mit getrockneten Amphibien-
herzen. Arch. Exp. Zellforsch, 7, (2), pp. 213-220.

1928B. Explantationsversuche an getrockneten und wiederbelebten

Herzen der Menschen — und Hubnerembryonem. Munchener Med. Wochen-
schrift, 75, (40), p. 1713.

Nageotte, J. 1927. fiber die tiberpflanzung Von Abgetoteten Bindegewebsstucken
Erwiderung an Fr. Weidenreich und A. Busacca. Virchow’s Arch, fur Pa-
thologische anatomie und Physiologie, 263, (1), pp. 69-88.

Nigrelli, R. F. and Breder, C. M. Jr. 1935. Histological changes in the pro-
lapsed intestine of a fish, Mollienisia latipinna Le Sueur. Copeia, (3), pp.
68-72.

Orton, J. H. 1924. An account of investigation into the cause or causes of the
unusual mortality among oysters in English oyster beds during 1920 and
1921. Fisheries Invest. Ser. II, 6, (3).

Papenfuss, E. J. 1934. Reunition of pieces in Hydra, with special reference to
the role of the three layers and to the fate of differentiated parts. Biol. Bull.,
67, pp. 223-243.

Parker, J. T. and Haswell, W. A., 1910. A textbook of Zoology. 2nd ed. Mac-
millan Co., London. 2 Vol.

Riddle, O. 1936. The Confusion of Tongues. Science, 83 (2142), pp. 41-45.

Rous, P. and Jones, F. S. 1916. A method for obtaining suspensions of living
cells from fixed tissues and for the plating out of individual cells. Journ. Exp.
Med., 23, pp. 549-555.

Schultz, E. W. 1930. The ultrascopic viruses from the biological standpoint.
Scientific Monthly, November, pp. 422-433.

Smyth, H. F. 1916A. The reactions between bacteria and animal tissues under
conditions of artificial cultivation. II. Bactericidal action in tissue cultures.
Journ. Exp. Med., 23, pp. 265-274.

1916B. The reactions between bacteria and animal tissues under condi-
tions of artificial cultivation. III. The action of bacterial vacine on tissue
culture in vitro. Journ. Exped. Med., 23, pp. 275-281.

1916C. The reactions between bacteria and animal tissues under condi-
tions of artificial cultivation. IV. The cultivation of tubercle bacilli with
animal tissues in vitro. Journ. Exp. Med., 23, pp. 283-291.

124

Zoologica: New York Zoological Society

Strangeways, T. S. P. 1924. Tissue culture in relation to growth and differen-
tiation. W. Heffer & Sons, Cambridge, pp. 1-50.

Uhlenhuth, E. 1915. The form of the epithelial cell in cultures of frog skin and
its relation to the consistency of the medium. Jour. Exp. Med., 22, pp. 76-104.

1916. Changes in pigment epithelium cells and iris pigment cells of

Rana pipiens induced by changes in environmental conditions. Journ. Exp.
Med., 24, pp. 689-699.

Wallin, I. E. 1927. Symbionticism and the origin of species. Williams and Wil-
kins, Baltimore, pp. 1-171.

Winkler, H. 1914. Chimarenforschung als methodi der experimentelle Biologie.
Sitzungs Berichta Physik Med. Gesselschaft Jahrgeng 1913 (8) pp. 113-119.

Woodruff, L. L. 1925. The physiological significance of conjugation and endo-
mixis in the infusoria. Amer. Nat. 59, pp. 225-249.

1926. Eleven thousand generations of Paramecium. Quart. Rev. Biol.,

1 (3), pp. 436-438.

Woodruff, L. L. and Erdmann R. 1914. A normal periodic reorganization process
without cell fusion in Paramecium. Joum. Exp. Zool., 17, (4), pp. 425-502.

Yonge, C. M. 1928. The absorption of glucose by Ostrea edulis. Journ. Mar,
Biol Assoc., 15 (N.S.), pp. 643-653.

Coates & Cox: Electrical Discharges of the Electric Eel

125

ll.

Preliminary Note on the Nature of the Electrical Discharges of
the Electric Eel, Electrophorus elec trie us (Linnaeus).

C. W. Coates,

New York Aquarium,

&

R. T. Cox,

New York University.

(Text-figure 1).

Although the Electric Eel has been known to science since 1729 and
has been the subject of much speculation and an enormous literature in
respect to the nature of the electrical phenomena exhibited, there is still a
considerable discrepancy between the voltages observed by different investi-
gators and a great deal yet to be learned about the nature of the generation
and discharge of these voltages. Since the development of the Cathode Ray
Oscillograph, we think that more accurate observations may be made than
any heretofore, and present this short report of our findings to date.

Preliminary observations into the voltage developed were made by
means of a device of resistances and neon pips designed and built for us
by Mr. H. M. Ferree of the General Electric Vapor Lamp Company. This
indicated the voltages in steps ranging from 85 to 150 ; 150-170; 170-300;
200-450 and 450-600. The eels used in this preliminary investigation varied
in size from 11% inches to 7 feet 10 inches and readings were taken both
in and out of water. In no single instance was a voltage beyond the 170-300
range recorded. This is not in accord with voltages reported by Eilenfeld.1
Some of the fish were tested immediately upon arrival and some after they
had been in aquaria for several years, being fed, during that time, on a
variety of living fishes which necessitated the continued use of electrical
discharges on the part of the eel if it were to eat. Table I gives the re-
spective sizes of the various eels tested.

Small eels were found to be more suitable for investigations with the
oscillograph, which was a Radio Corporation of America Cathode Ray oscil-
lograph type T M V-122-B. In the most satisfactory observations the eel
under test was removed from the water, dried, and laid in an insulating
trough two inches wide, in which were set transverse tinned copper wires
one inch apart. By means of dial switches any two of these wires could be
connected to one pair of deflecting plates of the oscillograph. With the
timing circuit connected to the other pair of deflecting plates the variation
of the voltage in time could be observed.

Two types of discharge were clearly distinguished.2 In each type the
voltage between two points on the eel rose to a maximum and returned to
zero. The anterior part of the eel was always positive with respect to the
posterior and no reversal of voltage was observed. The curve showing the
variation of voltage with time appeared symmetrical and the shape sug-
gested a Gaussian errors curve. The duration of one discharge was of the

1 EILENFELD, Walter: Ueber den Reflexschlag von Gymnotus electricus nach Untersuchungen mit
dem Oszillographen Beitrage zur Physiologie, Berlin, 1927, Band 3, pp. 195-198.

2 This is in accord with those reported by Eilenfeld.

126

Zoologica: New York Zoological Society

[XXI :11

order of 10-3 second. The two types of discharge differed strikingly in poten-
tial variation along the eel and in the maximum voltage attained. Observa-
tions on both types made on an eel 11.5 inches long with 55 cc. displacement
in water are shown on the figure. The abscissa shows distances measured
along the eel from head to tail. The difference in ordinate between any two
points shows the peak voltage developed between the points on the eel cor-
responding to the associated values of the abscissa. It will be noticed that
in the principal discharge the potential is uniform over the anterior two
inches and the posterior two inches of the eel. In the less vigorous or
secondary discharge the potential is uniform over the anterior four inches
and the posterior two inches of the eel. The origins of the Large Electric
Organs and Hunter’s Organs were two inches posterior to the snout and the
origins of the Bundles of Sachs were about No. 7 on the figure.

The peak voltage observed in either type of discharge between any
two points was not repeated uniformly. In successive discharges of the
principal type deviations from the mean of 25 per cent, were observed. The
voltages developed also diminished as the eel was kept a longer time out of
water. The contrast between the two types of discharge, as shown in the
figure, is probably exaggerated by this latter cause, the observations on the
secondary discharge having been commenced after the observations on the
primary discharge were completed, and indeed after the maximum voltage
of the primary discharge had been observed to fall from its initial value
around 200 volts to about 135 volts.

In this respect it might be as well to note that there is a wide variation
in the rate of fatigue between different animals and the same animal in
different conditions. This has been observed but not recorded by one of us
over several years’ work with many eels of all sizes.

Observations of another eel of about the same size showed a principal
discharge of about 200 volts accompanied by a secondary discharge of about
20 volts, and observations of a third eel 15% inches long showed a principal
discharge of about 145 volts accompanied by a secondary discharge of about
30 volts.

Another difference observed between the principal and secondary dis-
charges was that while the secondary discharges follow one another ap-
parently at random intervals of time, the principal discharges commonly
occur in pairs with a rather regular interval which is several times the
duration of one discharge.

Some observations suggested a third type of discharge with a peak
voltage around one volt, but the disturbances to which the oscillograph is
subject with the high amplification required to show such a discharge render
its existence uncertain, as yet.

No satisfactory measurements were made of the power developed in
the discharges. Some rough observations on an eel about 12 inches long
give a value of the order of 3 watts for the power developed externally at the
peak of the principal discharge when the eel is in the water to which it is
accustomed.

To determine whether the change in potential begins simultaneously at
all points along the length of the eel or whether there is a progressive po-
tential pulse from head to tail, the timing circuit of the oscillograph was
cut off and one plate of each deflecting pair was connected to one point on
the eel somewhat behind the middle and the other plates were connected
near the head and tail respectively. If there were no time lag, the trace
on the oscillograph screen would be a straight line, since the two voltages
giving rise to the vertical and horizontal displacements of the beam of
cathode rays would have a ratio constant in time. The trace observed was,
on the contrary, a narrow loop, roughly elliptical. This seems to indicate
that the potential pulse is propagated along the eel in a time of the same
order as, or less than, the duration of the pulse between two points. Since

200

150

$

M>

&

1

100

I

50 “

O O^-CV'-O

4 O'

•

/

,0

/

Major f
/

/•

1

li

o

/

/

«

discharge

O

f

/

Minor discharge

°‘

L

- — O- — c — o

3 4 5

7 8 9 10 11

Electrode position in indies

Text-figure 1.

Graph showing mean peak voltages at 1-inch intervals on Eel No. 2.
The open circles on the major discharge curve represent voltage
readings taken from head to tail; the black dots represent readings
taken in the reverse order. In the minor discharge, no differences
were observed in readings from either direction.

[127]

128

Zoologica: New York Zoological Society

the eel was about one foot in length, the speed of the pulse would appear to
be of the order of 1,000 feet per second. Such a speed is higher than those
commonly found in the propagation of electrical impulses along nerves.

TABLE I.

Size of Eels on which measurements were made.

3 eels 7 feet long

3 “ 7 “ 3 inches

1 “ 7 “ 9

1 “ 7 “ 10

in

water

only.

1 “ 3 “ 4

1 “ 3 “ 1

1 “ 2 “ 10

1 “ 1 “ 3% “

2 “ 11% “

both

in

and

out

of

water.

TABLE II.

Eel No. 2. Length IIV2 inches.
Mean peak voltage of major discharge.

TEST No. 1

TEST

No. 2

Electrode

A

Electrode

B

Voltage

Electrode

B

Electrode

A

Voltage

Voltage reck-
oned from 0

0

2

X

11

0

196

0

3

37

1

196

0

4

86

2

196

0

5

135

3

172

24

6

159

4

122

74

7

172

5

86

110

8

196

6

49

147

9

196

7

25

171

10

196

8

12

184

11

196

9

6

190

10

X

196

TEST No. :

3

TEST No. 4

Mean peak voltage of minor discharge

Electrode

A

Electrode

B

Observed

voltage

Corrected

voltagef

Voltage
from curve

Electrode

A

Electrode

B

Voltage

Voltage reck-
oned from 0

0

11

135

196

196

0

11

10.5

0

0

10

135

196

196

1

10.5

0

1

10

135

196

196

2

10.5

0

1

9

135

196

196

3

10.5

0

2

9

135

196

196

4

10.5

0

2

8

122

177

189

5

6.6

3.9

3

8

110

160

159

6

5.3

5.2

3

7

98

142

145

7

4.0

6.5

3

6

74

107

125

8

2.7

7.8

4

6

49

71

73

9

X

10.5

4

5

37

54

42

196

f — Observed voltage x — — ■ to correct for fatigue.

135

x — voltage too small to read.

Tests were made consecutively with five minute rest intervals between each.

Jleto |9otfe Zoological Society

General Office: 101 Park Avenue, New York City

Officers;

President , Madison Grant

Vice-Presidents, W. Redmond Cross and Kermit Roosevelt
Chairman, Executive Committee, Madison Grant

W. Reid Blair, Director
William T. Hornaday, Director Emeritus
Raymond L. Ditmars, Curator of Mammals and Reptiles
Lee S. Crandall, Curator of Birds
Charles V. Noback, Veterinarian

Claude W. Leister, Ass’t to the Director and Curator, Educational Activities

H. C. Raven, Prosector
Edward R. Osterndorff, Photographer
William Bridges, Editor and Curator of Publications

Aquarium

Charles H. Townsend, Director
C. M. Breder, Jr., Assistant Director

department of tropical &esearcf)

William Beebe, Director and Honorary Curator of Birds
John Tee-Van, General Associate
Gloria Hollister, Research Associate

Treasurer, Cornelius R. Agnew
Secretary, Henry Fairfield Osborn, Jr.

Scientific Staff

Zoological $arfe

editorial Committee

Madison Grant, Chairman

W. Reid Blair
William Beebe

Charles H. Townsend
George Bird Grinnell

William Bridges

For the information of persons or institutions

RECEIVING ZOOLOGICA:

Beginning with Volume XXI, Zoologica will be published in four quar-
terly Parts. Title page, Contents and Index will be found in Part IV of
each volume.

Beginning with Volume XXI, it will be assumed that each Part of
Zoologica is actually delivered to the person or institution to which it is
addressed. Consequently, no requests can be entertained for gratis copies
to complete files in which Parts are declared to be missing. As long as Parts
are in print, they will be supplied at the price fixed by the cost of printing.

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS

OF THE

NEW YORK ZOOLOGICAL SOCIETY

VOLUME XXI
Part 3

Numbers 12-18

PUBLISHED BY THE
THE ZOOLOGICAL PARK,

October 15, 1936

SOCIETY
NEW YORK

CONTENTS

Page

12. The Morphology, Cytology and Life-history of Oodinium

ocellatum Brown, a Dinoflagellate Parasite on Marine
Fishes. By Ross F. Nigrelli. (Plates I-IX; Text-figures
1-5) - - - - ------ 129

13. The Winter Movements of the Landlocked Alewife,

Pomolobus pseudoharengus (Wilson). By C. M.
Breder, Jr. & R. F. Nigrelli. (Text-figures 1-6) 165

14. Systematic Notes on Bermudian and West Indian Tunas

of the Genera Parathunnus and Neothunnus . By Wil-
liam Beebe & John Tee-Van. (Plates I-VII) ....... 177

15. Food of the Bermuda and West Indian Tunas of the

Genera Parathunnus and Neothunnus. By William
Beebe. (Plates I-III) .195

16. Notes on the Biology and Ecology of Giant Tuna, Thun-

nus thynnus Linnaeus, Observed at Portland, Maine. By
Jocelyn Crane. (Plate I) 207

17. The Templeton Crocker Expedition. I. Six New Brach-

yuran Crabs from the Gulf of California. By Steve A.
Glassell . 2 1 3

18. Neoplastic Diseases in Small Tropical Fishes. By G. M.

Smith, C. W. Coates & L. C. Strong. (Plates I-III) . 219

Nigrelli: On Oodinium ocellatum Brown

129

12.

The Morphology, Cytology and Life-history of Oodinium ocella-
tum Brown, a Dinoflagellate Parasite on Marine Fishes1.

Ross F. Nigrelli.

Department of Biology, New York University,
and New York Aquarium.

(Plates I-IX; Text-figures 1-5).

Contents.

Introduction

Material and Methods

Life-history of Oodinium ocellatum

Parasitic Stage

Palmella Stages

The Dinospore Stage

Nuclear Division

Effects of Density of Sea-water on Oodinium. .

Taxonomy

Incidence of Infection in New York Aquarium

Transmission Experiment

Discussion

Summary

Literature Cited

Page
. 129
. 130
. 131
. 136
. 137
. 138
. 140
. 145
. 147
. 149
. 150
. 150
. 157
. 158

Introduction.

Parasitic dinoflagellates have been known for many years, but it is only
recently that a species has been described from fishes. Brown (1931), in a
preliminary note, reported a new species, Oodinium ocellatum , from the gills
and skin of marine fishes, as the cause of a high mortality in the Aquarium
of the Zoological Society of London. In a later paper (1934) she extended
her observations, but added little concerning the morphology and life-history
of the dinoflagellate.

According to Brown (1934), the parasite is found on marine fishes
collected from the East and West Indies. In the New York Aquarium, how-
ever, the infection has been centered in fishes taken from Sandy Hook Bay
and has spread to a few species from Key West, Florida. This is the first
record of the parasite from North American waters. None of the East
Indian forms present in the Aquarium were found infected.

A number of parasitic dinoflagellates have been described, and much of

1 Submitted in partial fulfillment of requirements for the degree of Doctor of Philosophy at
New York University.

130

Zoologica: New York Zoological Society

[XXI: 12

the literature has been reviewed in detail in the excellent monograph by
Chatton (1920). Kofoid and Swezy (1921) have pointed out the need for
more thorough investigation of these organisms, since many details are
lacking in the descriptions of their morphology and life-history. The recent
occurrence of Oodinium ocellatum in large numbers in the New York Aqua-
rium has afforded an opportunity for further investigation on this species.

The writer wishes to express his thanks to Professor R. P. Hall for the
many valuable suggestions and criticisms offered in the preparation of this
paper.

Material and Methods.

The parasites were removed for study from the gills of Chilomycterus
schoepfii and Spheroides maculatus, which were dying as a result of the
infection. Infected fish could be detected by their actions and also by a pink-
tinted mucous secretion on the surface of the body. This pink color is due
possibly to waste products of the parasites, much as “red water” is pro-
duced by free-swimming dinoflagellates (Meade, 1898; Kofoid and Swezy,
1921; Martin and Nelson, 1929).

Infected gills were fixed in 10% neutral formalin, Zenker’s fluid, cor-
rosive sublimate and Bouin’s solution. The material was sectioned and
after each fixative some sections were stained with Delafield’s hematoxylin
and others with iron-hematoxylin, eosin being used as counterstain in cer-
tain cases. Tissue fixed in Zenker’s fluid was also stained with Mallory’s
triple stain; other material fixed in corrosive sublimate and stained with
iron-hematoxylin was counterstained with Van Giesen’s fuchsin-picric acid
in an attempt to demonstrate fibrils in vegetative stages of the parasites.
Most of the hematoxylin-eosin material was destained for the study of
nuclear and cytoplasmic structures. In such cases, the rhizoid processes
of the attached stages were always completely destained. To overcome this
difficulty, a few of the sections were overstained; in this way the rhizoid
processes which penetrate the gills were demonstrated.

Parasites showing organelle of attachment in various stages of pro-
trusion were obtained by strongly shaking the gills and allowing the para-
sites to fall into Schaudinn’s and Bouin’s fluids. By this method many of
the flagellates were fixed before this peculiar organelle could be completely
retracted.

In order to obtain division stages a large number of the parasites
were washed in sterile sea-water, distributed to several petri dishes, and
then fixed at intervals.

The iodine-potassium method, such as employed by Hall and Nigrelli
(1931), was used to demonstrate the presence of starch, and in attempts to
determine the nature of the small cytoplasmic inclusions described as amy-
loid bodies by several investigators.

Living specimens were examined under the oil-immersion objective,
and many details in the morphology and life-history were observed which
would not have been evident in fixed and stained material. In order to check
the observations made on mass cultures, single flagellates were isolated in
sealed hanging drop preparations, and observed daily. In this way stages
in the life-history were traced in pure line cultures.

Experimental infections were attempted in two cases, in one instance
with the parasitic stage directly from the gills of an infected host, and
in the second with dinospores grown in the laboratory in petri dishes. The
results of these experiments are discussed later in the paper. The effects
of various temperatures and different specific gravities of sea-water on de-
velopment were also investigated. Cultures were kept at 12.5°C., 25°C., and

1936]

Nigrelli: On Oodinium ocellatum Brown

131

35°C. with the density kept constant at 1.028. In the density experiments,
sea-water having an initial specific gravity of 1.028 was gradually evapo-
rated to a density of 1.040. By dilution with normal sea-water, a range was
obtained from 1.040 to 1.028. A second series was started with sea-water
at a density of 1.028 and, by dilution with fresh water, the range was ex-
tended to 1.003. All density readings were taken at the standard tempera-
ture of 15°C.

Life-history of Oodinium ocellatum.

The parasitic stage of O. ocellatum is a pear-shaped organism (Text*
fig. 1, A-F; PI. I, Fig. 1) attached to the gill filaments of marine fishes by
means of fine rhizoids. When the organism attains a large size, this method

A-F. Camera lucida drawings of the living parasite, x 950. Various stages and
sizes of the parasitic form just after removal from the gills, st., stigma-neuro-
motor complex; pd., peduncle; rh., rhizoids; ft., flagellum; c. cw., cellulose mem-
brane; pp., periplast; ch., chromoplastids ; am., amyloid granules.

132

Zoologica: New York Zoological Society

[XXI: 12

of attachment apparently becomes mechanically inadequate and the para-
site drops off the gills. However, all the organisms, regardless of size, will
undergo division once they are removed from the gills. On settling to the
substratum, the dinoflagellate takes in water, possibly through the canal
present in the peduncle, and increases in volume by one-fourth or more of
the original size. All of the organelles, including a broad “flagellum” pres-
ent in this region, are gradually retracted within the body of the parasite
and a cellulose cap is secreted to seal the original opening (Text-fig. 2, G;
PI. I, Fig. 2). When this process is completed, division is initiated.

The cytoplasm, at the pole diametrically opposite the region of the
peduncle, recedes from the cellulose covering (Text-fig. 2, H; PI. I, Fig. 3).
This is the point at which the first fission will start ; therefore, the first
division is longitudinal. Succeeding divisions are more or less regular, and

Camera lucida drawings of the living parasite, x 950. G. After imbibition of
water a cellulose cap ( ca .) is secreted. Note the two stigma-neuromotor complex.
H. Recession of the cell proper from the cellulose wall at the anterior end. Note
absence of both erythrosomes and stigma. I. Stage in the retraction of the polar
structures. J. End of the first division; ery., erythrosomes.

1936]

Nigrelli: On Oodinium ocellatum Brown

133

at right angles to each other, giving rise to palmella stages of 2, 4, 8, 16, 32,
64, and 128 cells (Text-figs. 2, 3, 4, J-N; PL I, Figs. 4, 5). One more palm-
ella division occurs to form 256 minute dinospores. These become flagel-
lated, break through the cyst wall and for a short time are free-swimming
naked dinospores but without girdle or sulcus (Text-fig. 5, R, S; PI. I, Fig.
6). In many cases the dinospores emerge from the palmella before the final
division is completed (Text-fig. 5, P, Q). The dinospore then settles to the
bottom, secretes a new cellulose covering (Text-fig. 5, T; PI. I, Fig. 7) and

Camera lucida drawings of the living parasite, x 950. K. Beginning of the 2nd
cell division. L. End of the 2nd cell division; the erythrosomes have disappeared
and in each cell a red bar may be seen but without the “desmose.” M. 8 cell

stage (flattened) .

134

Zoological New York Zoological Society

[XXI: 12

Text-figure 4.

Camera lucida drawings of the living parasite, x 950. N. Surface cells of a 32 cell
stage, some of which show the alignment of the amyloid granules and the be-
ginning of dinospore formation. O. Later stage in dinospore formation.

1936]

Nigrelli: On Oodinium ocellatum Brown

135

by certain cytoplasmic changes (Text-fig. 5, S-V ; PI. I, Figs. 8, 9) becomes
transformed into a typical peridinian dinoflagellate (Text-fig. 5, W, X;
PI. I, Fig. 10). The latter possesses a well developed girdle with trans-
verse flagellum and a short sulcus from which the longitudinal flagellum
passes posteriorly. The metamorphosis of the peridinian into the parasitic
stage could not be followed completely. However, evidence shows that the
flagella are lost, while the sulcus widens out into a cone-shaped structure
(PI. IX, Fig. 75). From the latter, in all probability, arises the peduncle
with its rhizoid processes.

Text-figure 5.

Camera lucida drawings of the living parasite, x 950. P. “Binucleate” naked
flagellate stage. Q. Beginning of the final dinospore division. Note the separation of
the neuromotor apparatus and the bar-like stigma. R. Temporary free-living,
naked dinospore. S. Form in which cytoplasmic differentiation has occurred.
T. Rounding up and secretion of a new cellulose membrane. U. Note the orien-
tation of the neuromotor apparatus. V. Cytoplasmic differentiation showing girdle
and sulcus. The stigma-neuromotor complex has begun to move anteriorly. W, X.
Typical free-swimming Oodinium ocellatum.

136 Zoologica: New York Zoological Society [XXI :12

The Parasitic Stage.

Oodinium ocellatum, (PL II, Figs. 11-24) as observed on the gill fila-
ments of the host, measures from 12.4 x 9.9 to 103.7 x 80.5 microns. Aver-
age measurements for one hundred specimens taken at random are 61.1
microns in length and 50,1 microns in width. On the other hand, when the
organisms round up after removal from the gills, they increase in volume
and may measure as much as 150 microns in diameter.

The parasite is surrounded by a firm transparent membrane which,
according to Brown (1934), gives a positive reaction for cellulose (Text-
fig. 1, c. cw.). The nucleus may be round or oval in shape and in many of
the living cells moniliform threads typical of the dinoflagellate nucleus are
apparent. The endosome which can be seen in fixed and stained specimens,
is not evident in the living material. The cytoplasm contains numerous
round chromoplastids (Text-fig. 1, ch.) of varying sizes, usually a pale
green in color (PI. I) and giving the characteristic starch reaction with
iodine-potassium iodide solution. When large numbers of the parasites are
placed in a test-tube, the entire suspension will give the characteristic blue
color when treated with a few drops of the iodine solution. Also present
in the cytoplasm are numerous granules which show a purplish color on
treatment with iodine. Like Brown (1934), the writer interpreted these
as amyloid granules (Text-fig. 1, am.). ■

In fixed and stained specimens (PI. II), the cytoplasm appears vacuo-
lated. In Mallory preparations (PI. II, Figs. 21-24) there are a few red
small granules (microsomes?) which seem to be the same as those which
stain a light yellow with Van Giesen’s method (PI. II, Figs. 11-16). The
nature of these granules is not yet known ; they were not observed in living
material. The cellulose wall is stained yellow with Van Giesen’s, deep blue
with Mallory’s and a brownish color with iron-hematoxylin. The chromo-
plastids are usually aggregated around the nucleus as a result of centri-
fuging. After iron-hematoxylin and eosin, these plastids show small bars
or dumb-bell structures which retain the hematoxylin; otherwise they are
stained homogenously with eosin. With iron-hematoxylin and Van Giesen’s
stain (PI. II, Figs. 14-16), the bars are similarly colored while the rest of
each plastid is stained yellow. With Mallory’s, the chromoplastids are red
or orange with bluish margins while the amyloid granules are light blue
(PL II, Figs. 21-24). The latter were not evident in other preparations.

One of the interesting features of this dinoflagellate is the organelle
of attachment present at the posterior part of the cell (Text-fig. 1, A-F;
PL I, Fig. 1; PL II; PL III, Figs. 25-27). This organelle, staining a light
blue with Mallory’s and yellow with Van Giesen’s, is composed of a trans-
parent base or peduncle (Text-fig. 1, pd.) which originates within the cyto-
plasm proper, passes through a large opening in the cellulose wall and ter-
minates in a pseudopodial tip, sometimes rhizopodial and sometimes lobo-
podial in appearance, depending on the state of retraction (Text-fig. 1, A-F).
It is by means of these “pseudopodia” or rhizoids (Text-fig. 1, rh.) that the
parasite anchors itself to the gill tissue of the host. When viewed under oil
immersion, the details of this region are clearly visible. In sectioned mate-
rial (PL II), the rhizoids appear as fine threads passing among the cells of
the gill tissues. After removal from the gills, these rhizoids become thicker
and more blunt in appearance but undifferentiated. In individuals in which
a certain amount of retraction has taken place the cytoplasm at the base
of the peduncle appears granular. In some cases this granular zone ex-
tends anteriorly almost to the region of the nucleus. In Mallory’s triple
stain preparations this differentiated cytoplasm is stained blue, and in Van
Giesen’s, yellow. At the base of the peduncle of the attached forms there
is a ring which stains red with Mallory’s (PL II, Figs. 21-24) and yellow

1936]

Nigrelli: On Oodinium ocellatum Brown

137

with Van Giesen’s (PI. II, Figs. 11-16). The nature of this structure, which
was not observed in the living flagellate, is unknown.

Besides this organelle, there is also present a peculiar broad ribbon-
like “flagellum” (Text-fig. 1, fl.), which shows very slow sweeping movements.
The “flagellum” is as hyaline in appearance as the rhizoid processes and,
except for the movements in living specimens, it is difficult to distinguish
this organelle from one of the rhizoids. In fixed and stained specimens re-
moved from the gills both “flagellum” and rhizoid processes seem to arise
from the granular cytoplasm at the posterior end of the cell. The “flagel-
lum” apparently can be used in swimming. For example, several specimens
were placed in one side of a petri dish and on the opposite side were placed
pieces of non-infected gills taken from a killifish ( Fundulus heteroclitus) .
Within ten minutes, a dozen young parasites were found attached to the
gill filaments. Whether or not this stage of the dinoflagellate moves about
on the gills within the branchial cavity of the host has not been determined.
In one case the “flagellum” appeared continuous with the canal which extends
from the pedunclar region to a dense spherical mass of cytoplasm just poste-
rior to the nucleus. The canal (PI. II, Fig. 11; PI. Ill, Fig. 27) is hyaline
in appearance and takes a blue color with Mallory’s triple stain. In speci-
mens stained with iron-hematoxylin or Delafield’s alone, it appears as a
colorless structure, while in those counterstained with eosin, the canal takes
on a pink tint. Such a canal was reported by Brown (1934), who likewise
was unable to determine the exact relationships of the structure in either
living or fixed and stained preparations. However, she also described a
club-shaped vacuole (vesicle) connected with the canal in the pedunclar
region. In the present material no such vacuole was observed. The stigma
(Text-fig. 1, st.), characteristic of 0. ocellatum, is usually present in the
pedunclar region of the parasite and lateral in position. This organelle is
composed of a broad red and a thin black pigment bar, between which there
is a clear retractile area. Occasionally, there may be two such stigmas
(Text-fig. 2, G).

Palmella Stages.

From two to five minutes after the parasites have been removed from
the gill filaments, the rhizoids and the “flagellum” are gradually retracted.
During the process of retraction, the organism shows a distinct increase
in size, probably caused by imbibition of water, since brownian movement
of cytoplasmic granules, not previously noticeable, becomes quite evident.
After retraction is completed, the cell begins to secrete a layer of cellulose
to close the gap (Text-fig. 2, G; PI. I, Fig. 2). This freshly secreted sub-
stance also fixes the enlarged parasite to the substratum (e. g., petri dish,
bottom of the tank, coral, etc.),. Fission now begins. At room tempera-
ture (22° C.) and in sea-water with a specific gravity of 1.028, division is
more or less regular and always equal. In living material, fission apparently
begins in the cytoplasm diametrically opposite to the point of attachment.
In other words, the plane of fission passes through the former antero-
posterior axis. Subsequent divisions occur in more or less regular fashion
at intervals of about twelve hours, giving rise to 2, 4, 8, 16, 32, 64 and 128
cells. Just prior to the first division, the remnant of the retracted pedunclar
processes completely disappears. The ocellus may or may not disappear; as
observed in one case, the organelle appeared to be dividing just before the
first fission was completed (PI. I, Fig. 4). Scattered beneath the surface
of the cell may be found many red pigmented rodlets (“erythrosomes”) ,
which are constantly shifting position (Text-fig. 2, ery. ; PI. I, Fig. 2). At
the end of division these rodlets usually have disappeared (Text-fig. 3, K)
and in their place may be found a single red pigment bar near the surface

138

Zoological New York Zoological Society

[XXI: 12

on each daughter cell. In other cases both ocelli and “erythrosomes” may be
present at the same time. The origin and final disposition of these granules
was not determined. As the cell divides some of the chromoplastids and
amyloid granules pass to each daughter organism.

After the second division, a new cellulose cyst is secreted by each
daughter cell at the termination of each fission, so that, when the dinospore
stage is reached, each individual is enclosed in its own cyst from which it
eventually escapes.

It is interesting to note that if the cells fail to divide at any particular
stage within the palmella, certain changes occur in the daughter cells ; these
have been interpreted as degenerative. The chromoplastids take on a more
pronounced yellow color while the pigment granules gradually change from
a definite red to a reddish orange and finally to a definite yellow color. The
ocelli, however, remain unchanged. These changes appear to be associated
with a gradual dehydration of the organisms; indicated by the evaporation
of water from the slide.

The amount of fluid present is another important factor for develop-
ment. In the isolation experiments, if the organism is left in a hanging
drop, division may occur up to the 64 or 128 cell stages, depending on the
size of the parent cell and the size of the drop. In some cases where the
dinoflagellate was accidentally placed at the edge of the drop, division
ceased when the four cell stage was reached. However, if any of these or-
ganisms are placed in a larger body of medium or in a fresh hanging drop,
division will continue even to the formation of dinospores.

Several times there has been observed an interesting process in which,
after the first fission, one of the resulting daughter cells fails to undergo
further division, while the other gives rise eventually to free-swimming
dinoflagellates. This is the type of development (“Palisporogenesis”)
which occurs normally in the closely related genus Apodinium.

Dinospores.

As a rule, in the material under observation, division proceeded in a
fairly regular manner to the 128 cell stage and then metamorphosis into
dinospores occurred. After flagellation is completed one more division takes
place to produce 256 motile individuals. However, under certain conditions
(discussed below) dinospore formation by the smaller individuals may be in-
duced at the end of the 8, 16, 32, or 64 cell stages, although never as early as
the 2 or 4 cell stage. The factors inducing sporulation appear to be affected
by environmental conditions (density of the sea-water, temperature, crowd-
ing, etc.) and, contrary to the belief of Brown (1934), are not necessarily
related to the size of the original dividing cell.

The formation of the dinospores is an interesting process as observed
in living material. Just before the final division, the amyloid inclusions
become aligned in the plane of the coming fission (Text-fig. 4, N). Other
granules together with the chromoplastids are distributed around the peri-
phery of each cell. Near the surface, orange or red pigment granules of
varying shapes and sizes may or may not be present. A red pigment bar
with its companion black rod is present near one pole of the cell and in a
later stage of development there is at each end of this black rod a granule
from which a flagellum arises. In such stages this black rod appears as a
“desmose” between the two blepharoplasts (Text-fig. 5, Q; PI. I, Figs. 6-9).

As observed in permanent preparations (PL IX, Figs. 56-74), the
neuromotor apparatus arises from what appear to be the centrioles. Cer-
tain forms show diplosomes, still within the “centrosphere,” from which
the flagella are growing out. As will be noted, these centrioles are joined

1936]

Nigrelli: On Oodinium ocellatum Brown

139

by a minute fibril. In later stages the fibril increases in length and eventu-
ally comes to lie at the periphery of the cell as observed in the living mate-
rial. In this stage, the flagella appear to be equal in size (PI. I, Fig. 6).
Whether or not the fibril connecting the blepharoplast behave as a para-
desmose in fission has not been determined definitely.

By a constant whipping of the flagella, the dinospore frees itself from
the cyst wall. It moves about now and then, but it eventually becomes
quiet again and then secretes another membrane (Text-fig. 5, T; PL I, Fig.
7). In this stage the dinospore is spherical in shape and measures about
15 microns in diameter. The cytoplasm, except for a few chromoplastids
and amyloid granules, is clear and non-vacuolated. On the flagellar side,
between the cellulose membrane and the periplast (Text-fig. 1, pp.), there is
a large space within which lie the flagella (Text-fig. 5, T; PI. I, Fig. 7).
In fixed material only a few chromoplastids are seen in the granular cyto-
plasm. The nucleus is ovoid or spherical and shows comparatively short and
densely stained chromosomes.

Several stages in the transformation of the early dinospore into the
typical free-swimming dinoflagellate has been observed (Text-fig. 5, U-X;
PI. I, Figs. 7-10; PI. IX, Figs. 56-74). Many of the specimens show surface
depressions, which presumably will become the girdle and sulcus. These
rudiments appear as clear areas, usually on one side of the body. Other
flagellates have been found with transverse girdle and a very short sulcus
(Text-fig. 5, V; PI. I, Fig. 9), although the flagella had not yet assumed
their final position. In this stage the stigma-neuromotor complex has receded
some distance from the surface of the cell. In later stages one of the flagel-
lum comes to lie in the girdle while the other extends posteriorly in the space
between the periplast and the cellulose membrane (Text-fig. 5, U).

At the completion of metamorphosis (Text-fig. 5, W and X; PI. I, Fig.
10; PL IX, Figs. 72-74), the free-swimming dinoflagellates are small or-
ganisms, measuring about 12 microns in length and about 8 microns in
width. The epicone is slightly smaller than the hypocone. The nucleus is
approximately central in position and shows the structure typical of the
dinoflagellate nucleus. The ocellus lies to the right in the anterior hemi-
sphere. There are a few chromoplastids and amyloid granules, usually
arranged around the periphery of the cell. The transverse flagellum ex-
tends the full length of the girdle, while the longitudinal flagellum, held
more or less rigidly proximally, passes posteriorly along the sulcus. In
fixed preparations the fibril joining the blepharoplasts is seen paralleling
the longitudinal axis of the cell (PL IX, Figs. 72-74).

Although the organisms move about rapidly and are very difficult to
observe, it was noted that the transverse flagellum is active in swimming,
causing the flagellate to rotate to the right. Lashing movements of the
longitudinal flagellum drive the flagellates forward.

Transformation from Dinospore to Attached Stage.

Under natural conditions the free-swimming stage presumably invades
the branchial chamber of a fish, becomes attached to a gill filament and
metamorphoses into the parasitic type. This metamorphosis has not been
traced completely. However, the writer has observed certain changes which
appear to be stages in this transformation. In some cases both girdle and
sulcus are still present but the flagella have disappeared. Observations on
other stages indicate (PL IX, Fig. 75) that the rhizoids develop from the
sulcal area. In such specimens the girdle is still present, while in the
sulcus region there is a finely granular cone-shaped zone of cytoplasm, the
apex of which is slightly extruded through an opening in the cellulose wall.
Presumably this will eventually form the peduncle with the rhizoid processes.

140

Zoologica : New York Zoological Society

[XXI: 12

Nuclear Division.

1. Interphase.

The “resting” nucleus of the vegetative stage of O. ocellatum is oval
or rounded in shape, lying usually near the center of the body. In fifty
specimens taken at random, the average size of the ovoid type is 20 x 16
microns. The spherical nuclei average about 16 microns in diameter, the
smallest measuring about 12 microns and the largest about 30 microns. The
“resting” nuclei of the dinospores are much smaller, averaging about 8
microns.

In the “resting” nucleus of the attached parasites (PI. II, Figs. 11-24),
the chromatin is present in the form of very short, densely staining
“threads.” With the absorption of water in the early stages of the detached
parasites, the nuclei also increase in volume. In these forms, the chromatin
is again apparent as short “threads,” but staining very lightly with
hematoxylin. This staining reaction of the nuclear substance appears to be
correlated with this increase in size. Thus, after each division of the palmella
(during which stage no more water is absorbed) the nuclear material stains
more densely. Similar types of nuclei were reported by Calkins (1899)
and Chatton (1914 and 1920) in Noctiluca and Blastodinium respectively.
In both these forms the nucleus is tremendously large and superficially, at
least, appears vesiculated. These investigators also reported the fact that
the chromatin substance of the nucleus stains lightly with basic dyes.

There is some evidence that the nucleus in 0. ocellatum , during certain
phases of the parasitic stage, assumes an interphase in which chromatin
granules (PI. Ill, Fig. 25), rather than short “threads,” are evident. Chat-
ton (1920) reported this type of “resting” nucleus in young parasites and
dinospores of other species of Oodinium. Chromatin granules were also
reported by Entz (1921) for the interphase nucleus of the free-living dino-
flagellate, Ceratium hirudinella. Other investigators such as Borgert
(1910a) for Ceratium tripos, Jollos (1910) for C. tripos, C. fusus and
C. furca and Hall (1925 a and b) for C. hirudinella and Oxyrrhis marina
were unable to detect a “resting” stage in which scattered chromatin gran-
ules were present. However, no true interphase was observed in any other
stage in the life-history of O. ocellatum. Once division is initiated, the
nucleus assumes a typical prophase appearance in which the chromatin mate-
rial is present in the form of chromomeres composing beaded chromosomes.

2. Nuclear Membrane.

A definite nuclear membrane is present in the “resting” stages of O.
ocellatum. Chatton (1920) reported the absence of such a structure for
the parasitic form of O. poucheti, although it was found to be present in
O. fritillaria and O. amylaceum. In the vegetative stage of O. ocellatum
the membrane is thick, but after the absorption of water it becomes very
thin and plastic as is indicated by the indentations caused by the numerous
chromoplastids impinging upon it (PI. Ill, Fig. 25). During the early
stages of division, the membrane persists but eventually it disappears and
is reformed in the telophase of the first division cycle (PI. VI, Figs. 47, 48).
However, in the palmella division, the nuclear membrane is not apparent
until the telophase of the last division cycle or that division which gives
rise to the dinospores.

Calkins (1899) reported that in Noctiluca the nuclear membrane per-
sisted during mitosis and disappeared only at the stage when the nuclear
plate is formed and the chromosomes were ready for division. In this stage
it disappeared in the region between the nuclear plate and the central
spindle. Chatton (1914), however, finds that in Blastodinium the nuclear

1936]

Nigrelli: On Oodinium ocellatum Brown

141

membrane disappeared early in mitosis. On the other hand, Hall (1925
a and b) reported the persistence of the membrane throughout all the
stages of the division cycle of both Ceratium and Oxyrrhis.

3. The Achromatic Mass in the Resting Cell.

The achromatic mass or “archoplasm” is present in the sub-nuclear
area of the resting cell (PI. Ill, Figs. 26-28). It is the differentiated mass
of cytoplasm described above. This cytoplasm is stained light blue in Mal-
lory preparations (PI. II, Figs. 21-24), yellow with Van Giesen’s (PI. II,
Figs. 11-16) and brownish with iron-hematoxylin (PI. II, Fig. 20). In the
preparations stained by the first two methods no fibers were evident at this
stage. This achromatic mass is large and, in young attached parasites
(PI. II), appears to be continuous with or extends to the base of the peduncle.
In larger forms in which retraction of the polar processes has taken place,
the two areas, however, are well separated and are connected by the canal
described above (PI. Ill, Fig. 27).

In addition to the finely granular appearance of this differentiated cyto-
plasm, densely staining basophilic granules are present (PI. II). These
granules are not unlike the microsomes described by Calkins (1899) in the
spheres of Noctiluca and by Chatton (1914, 1920) for the spheres of Blas-
todinium. In the division stages of O. ocellatum similar granules were
found at the fork of the bifurcated strands passing out from the achromatic
mass (PI. VII, Fig. 51).

4. Mitosis.

The phenomena of nuclear division in Oodinium ocellatum are some-
what complicated, so that the following general summary will help to make
the details more clear.

Two kinds of nuclear activity are recognized, one taking place in the
first division of the cell, and the other in the palmelia, especially after the
8 cell stage, though not necessarily so. In the former, mitosis is not unlike
that described by Calkins (1899) for Noctiluca and Dogiel (1908) for
Haylozoon. Such a type is prevalent also in the sporozoans and in certain
radiolarians and designated by Belar (1926) as paramitosis. The latter
investigator reported such nuclear behavior for Aggregata eberthi and Col-
lozoum inerme.

In 0. ocellatum , as in the above species, the “sphere” which lies in the
sub-nuclear region elongates during the early stages of nuclear activity.
In these stages, the chromosomes are short, thin and stain lightly. Later
they appear long, thick and densely stained. The nuclear membrane disap-
pears and the chromosomes become oriented in parallel rows and at right
angles to the elongated spindle. The chromosomes split longitudinally
while in this stage and from each daughter chromosome mantle fibers pass
to both sides of the “sphere.” As the “sphere” divides, the chromosomes are
gradually drawn upon the central spindle formed and in a still later stage
assume a metaphase “plate” appearance. During the anaphase, the chromo-
somes are drawn towards opposite poles as a result of a further division of
the spindle. In the telophase, the chromosomes again become short.

In later palmelia stages, mitosis appears to be somewhat different
and correlated with the rapidity with which division occurs. In these
forms, no orientation of the chromosomes like that described above was
noted. However, in late prophase or early metaphase shorter V-shaped
chromosomes are present on an elongated spindle; the condition appearing
not unlike that found in the metaphase stage of Syndinium turbo (Chatton,
1921). In Oodinium, however, there is no evidence that the V-shaped chromo-

142

Zoologica: New York Zoological Society

[XXI :12

somes split as a unit, i. e., from the apex of the V and along the axes of the
“arms.” In later stages of division in Oodinium, the chromosomes again
appear as a “plate” and the migration of the daughter chromosomes to the
poles occurs as in the first mitotic cycle.

5. Prophase.

Although one or more endosomes are present (PI. Ill, Figs. 25, 26),
there is no evidence that this structure takes an active part in mitosis. In
large parasites, the endosomes vary in size and shape. They invariably stain
lightly and homogenously with hematoxylin.

It is very difficult to delimit the various phases of mitosis in O. ocel-
latum and it is with some hesitancy that the terms employed for the stages
of the nuclear cycle in metazoan cells are applied here. The behavior of the
nucleus during the early prophase is not completely understood as yet. In
Fig. 30 (PI. IV) the nucleus is elongated, while the chromosomes are still
in the shortened phase. The nuclear membrane is still present and within
the “sphere” mass the centriole may be seen. In Fig. 31 (PI. IV) the nuclear
membrane has disappeared from the side towards the “sphere.” In this
stage the short chromosomes of the vegetative nucleus are replaced by long,
thin and lightly stained ones. In this case and in others the mitotic figure
superficially resembles late anaphase or early telophase, with only one pole
of the divided nucleus showing. Thus, Fig. 31 (PI. IV) is comparable to
Calkins’ (1899) Fig. 39 (PI. 42) to which he refers as late anaphase. In
the present material, these stages have been interpreted as early phases of
mitosis in which the chromosomes have assumed a parallel arrangement
but as yet have not thickened. In Fig. 32 (PI. IV) the nucleus appears as
a bilobed structure but the chromosomes are still in the prophase stage.
Here, too, mantle fibers are present. Figs. 33 and 34 (PI. IV) might indicate
that the nucleus is forming a C-shaped structure and is beginning to sur-
round the elongated “sphere” (PI. IV, Fig. 34) somewhat like that reported
by Calkins (1899) for Noctiluca. However, in so far as could be determined,
such is not the case for both parts of each of the nuclei represented are
entirely separated.

In many instances the nucleus takes on a sheaf-like appearance (PI. V,
Figs. 38, 39) and although the nuclear membrane has entirely disappeared,
the chromosomes are still thin and lightly stained. In these forms, the
chromosomes show definite orientation towards the “sphere.”

6. Metaphase and Anaphase.

In the late prophase or early metaphase, of botn the initial and subse-
quent divisions (up to and including the palmella of the 4 cell stage), the
chromosomes are long, thick and rather densely stained (PI. IV, Fig. 35;
PI. V, Figs. 36, 37, 40). They are definitely arranged parallel to one an-
other and at right angles to the dividing spindle. At this and earlier
stages (PI. IV, Figs. 31, 32), fine mantle fibers pass from the chromosomes
to the “sphere.” The chromosomes have begun to move into the central
spindle formed as a result of the elongation of the “sphere;” during the
process some of the chromosomes, moving in opposite directions, pass each
other, so that a curious picture is produced. Presumably, as in other para-
mitotic divisions splitting has been completed by the time the chromosomes
begin to move onto the spindle. Once on the spindle, the chromosomes
appear as a metaphase “plate” (PL V, Fig. 41). There is no evidence, how-
ever, that this nuclear “plate” encircles the central spindle as was reported
by Calkins (1899) for Noctiluca and by Dogiel (1908) for Haplozoon. In
Oodinium ocellatum the separation of the daughter chromosomes is com-

1936] Nigrelli: On Oodinium ocellatum Brown 143

pleted by a transverse fission. This separation has just started in the
form represented in Fig. 42 (PI. V). During the anaphase the daughter
chromosomes are drawn to the opposite poles (PL VI, Figs. 44, 45; PL VII,
Figs. 50-51).

In later fission, after the 8 cell stage, this type of nuclear division was
not observed. In the early prophase no aligment of the chromosomes was
noticed. In later stages, radiating V-shaped chromosomes appear on the
spindle (PL VI, Fig. 49; PL VIII, Fig. 52). The writer interprets such
stages as metaphases and the chromosomes are doubling their number by
unipolar splitting. As the spindle begins to divide, the chromosomes are
straightened out to form the typical metaphase “plate.” Division of the
chromosomes is completed, possibly, by a transverse fission, much likte that
described by Hall (1925 a and b) for Ceratium and Oxyrrhis. Similar type
of radiating V-shaped chromosomes was reported by Chatton (1920, 1921)
for the parasitic dinoflagellate Syndinium turbo. In this form, the chromo-
some number is doubled by a splitting of the entire V.

7. Telophase.

In early telophase the chromosomes become condensed and for a short
time maintain their parallel arrangement (PL V, Fig. 43; PL VI, Figs. 46,
47; PL VIII, Fig. 53). In late telophase (PL VI, Fig. 47) a nuclear mem-
brane is reformed even before the spindle has been completely obliterated.
No evidence was obtained, however, to show that in this first division cycle
the reorganized nuclei assume the normal interphase appearance. In all
cases seen the chromosomes remained similar to those of the early prophase
(PL VI, Fig. 48). The nuclear membrane in such forms is thin as in the
prophase nucleus at the beginning.

After the 4 cell stage, on the other hand, division is very rapid and the
chromosome structure definitely is not altered in this phase but passes into
the prophase of the next division (PL VIII, Fig. 52). In the final dinospore
division, the chromosomes in the late telophase are short, thick and densely
stained (PL IX, Figs. 56-58).

7. Achromatic Figures.

The process involved in the separation of the chromosomes in 0. ocel-
latum is an intricate one. As was pointed out above, the achromatic mass,
or “sphere” as Calkins calls it, arises from the differentiated cytoplasm in
the region just posterior to the resting nucleus (PL II, Figs. 12-14; PL III,
Figs. 25-27). In certain forms, “centrospheres,” containing diplosomes de-
scribed above, are evident (PL IV, Figs. 30, 33; PL V, Fig. 40; PL VIII,
Fig. 52) and not unlike those present in Noctiluca (Calkins, 1899). From
each polar mass, fine strands pass to the periphery, bifurcate and end in
the periplast (PL VII, Fig. 51). At the fork of each bifurcation is often
found a granule (not unlike the microsomes found within the differentiated
cytoplasmic mass of the attached parasite), the significance of which is not
known. Fig. 32 (PL IV) shows what might be the beginning of the fine
protoplasmic strands, although the “sphere” has not yet begun to elongate.
At the termination of fission, especially at the end of the first nuclear cycle,
the achromatic mass is also reformed (PL VI, Fig. 48).

It has been shown above that from each of the divided chromosomes
minute fibrils (PL IV, Figs., 31, 32; PL V, Fig. 40) pass towards the center
of the “sphere.” When the latter elongates to form the central spindle these
fibrils converge towards each pole. These are in all probability the radial
fibers described by Ishikawa (1899) and the mantle fibers described by
Calkins (1899) for Noctiluca. The exact relationship of these fibers to those

144

Zoologica: New York Zoological Society

[XXI: 12

of the spindle was not determined for 0. ocellatum. Calkins, although not
certain, believed they were nuclear in origin. He showed that these fibers
are focused in the centrosome and connect with the chromosomes. Although
in 0. ocellatum fibers were seen passing from the chromosomes, just how
far they extended along the spindle could not be ascertained, for in many
of these forms no centrosomes were evident. In palmella divisions definite
centrioles within centrospheres were often present but no mantle fibers were
found. One detached parasite showed two centrospheres in the posterior
region of the body connected by minute fibers passing from the differentiated
cytoplasmic mass (PI. VIII, Fig. 55). This would indicate that these fibers
at least, are not nuclear in origin.

The centrioles arise as a result of the division of the granules and
desmose of the ocellus complex. In one case such a divided structure was
seen within a centrosphere-like structure (PI. VIII, Fig. 54). In the speci-
men mentioned above, two such centrospheres were present in the granular
zone of the posterior part of the cell (PI. VIII, Fig. 55). Unfortunately,
in most of these stages, the nucleus and the surrounding region are par-
tially masked by the large number of chromoplastids and the relationship of
this structure to the nucleus could not be determined. However, it is be-
lieved that these fibrils focusing towards the centrosphere are in direct
connection with the achromatic mass adjacent to the nucleus. Just how
they finally attain their final position at the poles of the spindle was not
determined.

As just stated, the centrioles arise as a result of the division of the
granules and desmose of the ocellus complex. It must be mentioned here,
however, that in the living parasite, no granules were observed at the ends
of the black pigment bar associated with the red pigment mass of this
organelle, although they were quite evident in the early living dinospore
stage. However, in all these non-flagellated forms, the red part of the ocel-
lus is always connected to the black portion by means of very minute
'‘fibrils.” It may well be that these connecting fibrils condense to form
the granules (blepharoplasts) present in the centrosphere. The evidence
for the origin of these centrioles from the ocellus complex is further sub-
stantiated by the fact that these granules give rise to the flagellar apparatus
while still within the centrosphere (PI. IX, Figs. 64-66).

No definite evidence was obtained as to the presence of a paradesmose
such as has been described for many of the free-living flagellates. Hall
(1925 a) was the first to demonstrate such a structure in a dinoflagellate.
He found that in Oxyrrhis marina a paradesmose was formed as a result of
a division of the centrosome. In this form the blepharoplasts disappear
during late prophase or early metaphase. Therefore in this species, at least,
the “desmose” is a centrosome-paradesmose. Kofoid and Swezy (1921) con-
sider the achromatic structure of Noctiluca as analogous to the centrosome-
paradesmose of other flagellates, since in this form the kinetic elements are
extra-nuclear throughout the entire process of mitosis. By the same rea-
soning, since the centrospheres and spindle are also extra-nuclear in origin,
the entire organelle in 0. ocellatum may be considered analogous to a para-
desmose. In this form, however, since it is definitely shown that the cen-
trioles are in fact the blepharoplasts which give rise to the flagella of the
dinospores, the desmose may be considered as analogous to a centroblepharo-
plast-paradesmose of the free-living flagellates.

It must be mentioned here, that in several cases a peculiar fiber was
noted, the ends of which terminated in small granules and from each of
which two small fibrils pass out to join each of the granules in the centro-
spheres (PI. VIII, Fig. 53). If this is a paradesmose, it is possible that
with the proper technique this structure can be demonstrated more defi-
nitely.

1936] Nigrelli: On Oodinium ocellatum Brown 145

8. The Mechanism of Mitosis in Oodinium ocellatum.

There are many theories concerning the mechanism of mitosis in gen-
eral. Calkins (1899) reported his concept of this process in Noctiluca. He
states “The nuclear membrane disappears and the mantle fibers connect the
ends of the chromosomes with the centrosomes in the spheres. The central-
spindle elongates, causing separation of the spheres ; the mantle fibers, re-
maining firm, move with the spheres, dragging the ends of the chromosomes
with them. As the central-spindle becomes longer, the chromosomes are
more and more separated, until finally the distal ends are separated and
the chromosome division is completed.” This process seems logical enough,
but the question may be asked, what causes the spindle to elongate? The
jvriter believes that the following may throw some light on this question.

In practically all the division stages of Oodinium ocellatum a seemingly
sol-gel reaction of the cytoplasm was noted. The “sol” phase manifested
itself in stained preparations as light and non-granular areas, usually at
the poles of the cell (PI. VII, Fig. 51). The fine strands passing out from
the “sphere” seem to be attached to the edges of such zones. A gradual
gelation occurs towards the poles of the cell. With this reversal of phase,
the strands of the spheres are “pulled” towards the poles, resulting in the
elongation and finally the division of the spindle and the separation of the
chromosomes on it. At the beginning of division these strands radiate out
in all directions. As division proceeds (and the cell elongates) the strands
begin to converge more and more towards the poles. When fission is com-
pleted, solation once again occurs, the cell rounds up and the protoplasmic
strands radiate out in all directions.

It is interesting to note that Calkins (1899) had seen similar strands
passing out from the “sphere” of Noctiluca but interpreted them as
analogous to the astral rays of metazoan cells. Their function in the
division processes of the dinoflagellate, however, was not discussed.

Effects of Density of Sea-water on Oodinium.

It was shown by Nigrelli (1935) that the monogenetic trematode
E'pihdella melleni MacCallum was unable to withstand sea-water of either
a high or low density. A similar experiment was carried out to determine
the optimum density necessary for complete development of these dino-
fiagellates and the effects of densities at either extreme.

Small quantities of modified sea-water with specific gravities ranging
from 1.040 to 1.003 (pH range from 8.4-7. 1) were made up as described
under material and methods. Aliquot portions were distributed to petri
dishes and to each dish were added parasites taken directly from the gills of
several spiny boxfish. Examination of each dish immediately after the
parasites were placed therein showed that all the flagellates were in the
vegetative stage (i.e., with the peduncle and rhizoid processes still pro-
truding). At the end of each twenty-four hours a differential count was
made of the various division stages present. This was continued for a
period of seven days.

The results showed that the optimal density for development to the
dinospore stages lies between 1.012 and 1.021. It is interesting to note that
within this range all forms may develop into dinospores at the end of the
second or third day. At a density of 1.040, division occurred very slowly
and dinospores were not formed during the period of our observations. In
the majority of specimens, at this density, development reached the 16 cell
stage, and only 2% were seen in the 32 cell stage. However, if the organisms
were transferred as late as the sixth day to sea-water of lower density
(1.028), division continued to the 128 cell stage and dinospores were formed.

146

Zoologica: New York Zoological Society

[XXI: 12

At a density of 1.036 division was slow and again a few non-motile dino-
spores were observed only at the end of the seventh day. In water with
specific gravities of 1.034 and 1.032, non-motile dinospores were seen on the
fourth and sixth days, and the majority became free-swimming on the
seventh day. In densities of 1.030, 1.029 and 1.028 non-motile dinospores
were observed* as early as the third and fourth days ; a few free-swimming
dinoflagellates were present on the fifth day and practically all were motile
on the sixth and seventh days. In a density of 1.024 many non-motile
dinospores were noted on the third day; these became active twenty-four
hours later. In the densities of 1.021, 1.018, 1.015 and 1.012 a few non-motile
dinospores were observed on the second day, but they did not become free-
swimming until the fourth day. In densities of 1.005 and 1.003, the or-
ganisms divided very slowly. At the end of the fifth, sixth and seventh
days the majority of forms were in the 8 and 16 cell stages and a few non-
motile (32 cell stage) dinospores were observed. In fresh water, develop-
ment continued to the 4 cell stage, and only 18% reached the 8 cell stage. At
the end of the experiment (seven days) these 4 and 8 cell stages were trans^
ferred to sea-water having a density of 1.009, but no further division oc-
curred. Development in water with a specific gravity of 1.003 was similar
to that in fresh water, except that on transfer of the palmellas to a density
of 1.009, division continued in a few instances and dinospores were formed.
Again, however, most of the palmella reached only the 16 cell stage.

These experiments were carried on at an average temperature of 22° C.
In another series, the temperature was varied while the specific gravity was
kept constant at 1.028. At 12.5° C., the development was very slow, a few
non-motile dinospores being formed at the end of seven days while the
majority (60%) were in the 128 cell stage. At this temperature, most of
the dinospores did not become free-swimming until the end of the tenth
day. At 25° C., the results were somewhat similar to those obtained at room
temperature, with a few non-motile dinospores being formed as early as
the third day and motile flagellates at the end of the fifth day. At 35° C.,
development was accelerated considerably at this density. Here, the rate of
division was somewhat equivalent to that which occurred in sea-water with
a density of 1.015 and at 22° C.

Similar results were obtained by Brown (1934), who found that the
organisms were inactive below 10° C. From 10-20° C., the flagellates
divided slowly; from 20-25° C., more rapidly, and at about 25° C., sporula-
tion was completed in three days.

These results indicate that the densities most suitable for development
lie between 1.012 and 1.028. This is approximately the range observed under
natural conditions. However, it is interesting to note that development
can occur (at room temperature) over a wide range of specific gravities
(1.005-1.036).

At the height of the epidemic in the Aquarium, analysis of the water
gave the following readings:

Density

pH

Temp.

Bound C02 Free C02
mM per liter

Bay Water

1.0120

7.5

22° C.

2.02

.20

Sea Water

1.0284

8.2

22° C.

2.50

.00

Conditions in the Aquarium were well within the range most suitable
for development of Oodinium ocellatum. The heaviest infection was found

1936]

Nigrelli: On Oodinium ocellatum Brown

147

in fishes in the closed circulation. This was no doubt due to the fact that
the infective stages were not washed to the sewer as would be the case in
the open circulation or bay water. Therefore, the results obtained in the
laboratory compare favorably with the conditions present in the tanks.

Taxonomy.

According to Chatton (1920), one of the first parasitic dinoflagellates
to be discovered was Gymnodinium pulvisculus, described by Pouchet (1884-
1885). Since the specific name had been previously applied by Klebs (1883)
to a fresh water type, Lemmermann (1899) renamed the parasite Gym-
nodinium poucheti. In view of the great variety of forms included in the
genus Gymnodinium, Chatton (1912) erected the new genus Oodinium with
0. poucheti (Lemmermann, 1899) as the type species. The genus is defined
by Chatton (1920) as follows: “Dinospores a hemisphere anterieur beau-
coup plus developpe que le posterieur. Pas de pigment Xantho-chlorophyllien,
mais un lipochrome. Formes vegetatives fixee par un tronc absorbant
fibrillaire. Parasitisme blastotrophe. Pas de scissiparite simple. Sporo-
genese intervenant apres liberation du parasite a produits, homodynames
epars.” Chatton gave this genus, as he did for many other genera of
parasitic dinoflagellates, family (Oodinidae) ranking under the sub-order
Gymnodinida. However, both Kofoid and Swezy (1921) and Calkins (1926)
relegated all the known parasitic genera to a single family, Blastodinidae
Chatton, 1906. In this family, they included the following genera: Schizo-
dinium Chatton (1912). Blastodinium Chatton (1906). Apodinium Chatton
(1907), Parapodinium Chatton (1920), Chrytriodinium Chatton (1912),
Paulsenella Chatton (1920), Haplozoon v.Dogiel (1906 a) (— Microtoeniella
Calkins, 1915), Oodinium Chatton (1912), Syndinium Chatton (1910 a),
and Trypanodinium Chatton (1920). However, Kofoid and Swezy (1921)
failed to include Haplozoon and, furthermore, they included as true dino-
flagellates the genera Ellobiopsis Caullery (1915) and Paradinium Chatton
(1910), two forms which, according to Chatton (1920), are perhaps not
dinoflagellates but were provisionally placed in the sub-order Cryptomona-
dinea because of their cryptomonad-like characteristics. More recently,
Reichenow (1930) classified all these forms under the family Gymnodiniidae
and added the genera Endodinium Hovasse (1922) and Merodinium Chatton
(1923) to the group of parasitic dinoflagellates.

The genus Oodinium, according to Chatton (1920), contains the follow-
ing species: O. poucheti, from the tunicate Oikopleura dioica; O. amylaceum
(Bargoni, 1894), occurring on Salpa mucronata and S. democratica; O.
fritillaria Chatton (1912) from Fritillaria pellucida, and O. appendicular iae
(Brooks and Kellner, 1908) from the acidian, Oikopleura tortugensis. Other
species of doubtful identity but temporarily placed by Chatton in the genus
Oodinium are forms described by Dogiel (1910) as Gymnodinium pulvisculus
from the annelid Alciope sp. and one form that Chatton (1920) has ob-
served on the pteropod Criseis acicula. Kofoid and Swezy (1921), in their
short discussion of parasitic forms, refer to a species as Oodinium parasiti-
cum (= Gymnodinium parasiticum Dogiel, 1906). According to Chatton,
G. parasiticum Dogiel is synonymous with Chytriodinium parasiticum
(Dogiel).

The life-history and morphology are not completely known for any of
the above species of Oodinium. In O. poucheti, according to Chatton (1920),
the parasitic forms are large in sizt (150-200 microns), ovoid or spherical,
without groove or flagella. The cytoplasm contains numerous minute yellow
lipochrome granules more or less evenly dispersed. The vesicular nucleus
is large. The organ of attachment is made up of a short robust peduncle,

148

Zoologica: New York Zoological Society

[XXI: 12

possessing fibrils and terminating in fine rhizoids, which, as in 0. ocellatum,
are capable of retraction. The entire body, including the rhizoids, is sur-
rounded by a cellulose membrane. Reproduction occurs by repeated and
equal division, resulting in the development of numerous free swimming,
naked dinospores, with a girdle but no sulcus. The details of fission and
metamorphosis of the dinoflagellate to the parasitic type are not known.

0. amylaceum was originally described by Bargoni (1894) as one of
the Foraminifera and it was Chatton who recognized its true relationships,
although palmella and dinospore stages have not yet been described. In
this form, the peduncle terminates in an extensive arborization of rhizoids.
The cytoplasm contains numerous amyloid granules. Caullery (1906) re-
discovered this parasite in the branchial cavity of Salpa africana but gave
no information on its life-history. However, from the figure he submitted
to Chatton (see fig. 4, Chatton, 1920) the rhizoids do not show the charac-
teristic arborization and in all probability it may be another species.

0. fritillaria measures 80 x 130 microns (115 microns in diameter for
the round forms). The nucleus in this species is very large, measuring 75
microns in diameter. There are a few yellow lipochrome granules in the
cytoplasm. This form differs from the other described species in that the
organ of attachment terminates in a broad basal disc measuring 60 microns
in diameter. Division and dinospore stages are not known.

O. appendicular iae was first described by Brooks and Kellner (1908)
as stages in the development of Oikopleura tortugensis and in the same
paper (page 93) reported certain forms as a new species of parasitic
Foraminifera ( Gromia appendicular iae) . It was Chatton (1920) who
pointed out that these attached forms were parasitic dinoflagellates, al-
though no other stages in the life-history are known.

Oodinium ocellatum agrees with the general generic description given
by Chatton ; since the morphology and life-history of the various species
are but imperfectly known, it is difficult, however, to determine specific
differences. According to Brown (1931), it “differs from all other members
of the genus in the possession of an eye-spot and in its somewhat smaller
size.” However, there are other differences, some of which are given in the
comparison with the type species, O. poucheti.

There is no doubt that the dinoflagellate found on fishes belongs to
the genus Oodinium. Since the generic characters are based on O. poucheti,
this species may be compared with 0. ocellatum. The parasitic stage of
O. ocellatum differs from that of 0. poucheti in the following characters:
(1) presence of chromoplastids and amyloid granules, (2) cellulose cell wall
surrounds the body of the parasite, the peduncle and rhizoids being naked,
(3) presence of one or more eye-spots, (4) the absence of yellow lipochrome
granules, but the presence of red or orange pigment rodlets or globules
during certain stages of development and (5) the presence of a peculiar
“flagellum” originating in the peduncle. The free-swimming dinoflagellate
differs from those of 0. poucheti in the following: (1) hypocone slightly
larger than epicone, (2) presence of a definite, although small sulcus, (3)
presence of a few small chromoplastids and amyloid granules in the cyto-
plasm, (4) red eye-spot and (5) presence of definite cellulose membrane
surrounding the entire organism. 0. ocellatum agrees with 0. poucheti in
the following: (1) the parasitic stage is usually pyriform-shaped organism
of rather large size with organelle of attachment composed of a peduncle
ending in fine rhizoids, and (2) division is usually equal, giving rise to a
palmella with subsequent formation of free-swimming dinospores.

Because the life history of the majority of the species of Oodinium is
not entirely known, the following tentative key is formulated on the basis
of certain characteristics found in the parasitic stage.

1936]

Nigrelli: On Oodinium ocellatum Brown

149

Key to the Species of Oodinium Chatton, 1912.

Peduncle terminating in few rhizoids

bj no eye-spot present 1. 0. poucheti (Lemmermann, 1899)

b2 eye-spot present 2. O. ocellatum (Brown, .1931)

A., Peduncle ending in a broad disc... 3. 0 fritillaria. (Chatton, 1912)

As Peduncle ending in an extensive arborization of rhizoids

4. 0. amylaceum (Bargoni, 1894)

Such forms as 0. appendiculariae (Brooks and Kellner, 1908), Oodinium
sp. (Dogiel, 1910) and Oodinium sp. Chatton (1920) are not included be-
cause structural details of the type to distinguish them from the species
listed in the key are not known. However, on the basis of the little that is
known about the parasitic stage, there can be but slight doubt that these
forms belong to the genus Oodinium .

Incidence of Infection in the New York Aquarium.

According to Brown (1934), the evidence shows that Oodinium ocel-
latum is “indigenous to the warm latitudes and is probably associated with
coral reef fishes in Bermuda and the East Indies.” The majority of fishes
in the New York Aquarium are collected from Key West, Florida, and from
Sandy Hook Bay for local species. An occasional specimen reaches the
Aquarium from Africa and the East Indies. As was pointed out above, the
center of the infection was localized in the spiny boxfish ( Chilomycterus
schoepfii ) and the northern or common puffer ( Spheroides maculatus ). The
former is commonly found in the more southern waters and only in late
summer or early fall does it visit the waters around New York, while the
latter is common about our coast from early spring to early winter. Where
they go in the late winter is not definitely known.

These species, together with the majority of others caught in local
waters, are first kept in bay water at seasonal temperatures and as cold
weather approaches they are gradually transferred to heated bay water
(22° C.). Since the parasites did not make their appearance in the closed
circulation until the early part of December, 1935, the question arose as to
what fish served as the original host of the dinoflagellate. Spiny boxfish
present in the floor pools (not connected with either the main or warm bay
water circulation) showed heavy infection and it is assumed that they
brought the parasites with them in their migration from warmer waters
and infected other fishes present in the Sandy Hook region at the time.

A list of infected hosts all collected from Sandy Hook Bay follows:
Order Acanthopteri (spiny rayed fishes), family Carangidae: (1) Caranx
hippos (Linn.), common jack, infection mild; (2) Caranx crysos (Mitchill),
hard-tailed jack, infection mild; (3) Trachinotus falcatus (Linn.), round
pompano, infection mild; (4) Naucrates ductor (Linn.), pilot fish, infection
mild. Family Pomatomidae: (5) Pomatomus saltatrix (Linn.), bluefish,
infection mild; (6) Roccus lineatus (Bloch), striped bass, infection mild;
(7) Centropristis striatus (Linn.), common sea bass, infection mild. Family
Sparidae: (8) Stentomus chrysops (Linn.), northern porgy, infection mild.
Family Sciaenidae: (9) Cynoscion regalis (Bloch and Schneider), weakfish,
infection mild; (10) Leiostomus xanthurus Lacepede, spot, infection mild;
(11) Menticirrhus saxatilis (Bloch and Schneider), northern kingfish, in-
fection mild. Family Tetradontidae : (12) Spheroides maculatus (Bloch and
Schneider), northern swellfish or common puffer, infection heavy. Family
Diodontidae: (13) Chilomycterus schoepfii (Walbaum), spiny boxfish, in-
fection heavy. Family Triglidae: (14) Prionotus carolinus (Linn.), Caro-

150

Zoologica: New York Zoological Society

[XXI :12

lina sea robin, infection mild; (15) Prionotus evolans (Linn.), striped
sea robin, infection mild. Two species collected from Florida were found
infected in the Aquarium. These were (16) Chaetodipterus faber (Brous-
set), spadefish, and Pomacanthus paru (Bloch), French angelfish. Both of
these belong to the family Ephippidae and the former was found to be
mildly infected while the latter species died as a result of a very heavy
infection.

A few of the species found infected by Brown (1934) are also present
in the New York Aquarium, but forms from the West Indies, such as
Angelichthys isabelita Jordan and Rutter, Chaetodon capistratus (Linn.)
and Holocentrus ascensionis Osbeck, have not as yet shown signs of infec-
tion. The cosmopolitan species, Mugil cephalus, also present in the
Aquarium, is likewise free of the parasites. The East Indian form, Amphi-
prion percula Lacepede, has never shown signs of infection in the Aquarium,
although Brown reported that this species and Psettus argentus Linn, are
always heavily infected and probably introduced the infection into the Lon-
don Aquarium, indicating that the parasite is present in their natural
locality.

It is interesting to mention at this time that swarms of dinoflagellates
have been reported from the New Jersey coast by Martin and Nelson (1929)
and since “red water” has been seen on many occasions in Sandy Hook Bay
by the writer, it is altogether possible that the source of the New York
Aquarium infection may be localized in this area.

Transmission Experiment.

Fundulus heteroclitus (Linn.), the common killifish, although present
in large numbers in the Aquarium, was never found infected. This is not
due to a natural resistance since infections have been induced under experi-
mental conditions. A number of these fish, acclimated to a temperature of
22° C., were placed in two-gallon tanks. In two of the tanks eight fish each
were introduced, while a third tank with four fish was used as control. In
tank I, a large number of dinospores collected from forms grown in petri
dishes were released. In tank II, the gills of a heavily infected spiny box-
fish were introduced. Examination of the gills of two killifish removed
from tank I on the second day gave positive results. This was to be ex-
pected because the infective stages had been introduced in large numbers.
Two fish from tank II, in which infected gills with adult parasites were
introduced, gave negative results on the second day, and similar results were
noted in a fish examined on the fourth day. On the sixth day, however, a
fish showed a mild infection. One individual found dead on the eighth
day showed a heavy infection of the gills and skin, probably the cause of its
death. Since all the parasites undergo a period of division when they are
once removed from the gills, it is not surprising to find that infecting the
fish in tank II was delayed until the sixth day.

Discussion.

Four definitely recognized species and three additional species of
doubtful validity were placed in the genus Oodinium by Chatton (1920).
Of these, the life-histories of but two species (0. poucheti and 0. ocellatum )
are known. The former is parasitic on the pelagic tunicates, Oikopleura
dioica and Oikopleura sp., while the latter occurs on the gills and skin
of marine fishes. Of the ten or more genera of parasitic dinoflagellates, the
following five are ectoparasitic : Apodinium Chatton (1907), Parapodinium
Chatton (1920), Chytriodinium Chatton (1912), Paulsenella Chatton (1920)
and Oodinium Chatton (1912). Paulsenella was described as parasitic on

1936]

Nigrelli: On Oodinium ocellatum Brown

,151

a diatom and Chytriodinium as ectoparasitic on copepod eggs. The genera
Parapodinium and Apodinium are found exclusively on pelagic tunicates,
while species of Oodinium have been reported from tunicates, pteropods,
siphonophores and annelids. Oodinium ocellatum is the first known dino-
flagellate parasite of vertebrates. Brown (1931, 1934) reported this species
from the gills and skin of marine fishes of the East and West Indies, while
the writer has found that in the New York Aquarium the parasite attacks
the spiny boxfish and the common puffer. The former is a warm water
species that migrates north on the Atlantic coast during the late summer
and early fall ; the latter, a local species ranging as far north as the coast of
Maine. Other North American species, cited above, have also shown the
infection.

In the Dinoflagellida, as in other groups of plant-like flagellates, many
species carry on photosynthesis and are thus holophytic, while others are pre-
dominantly saprozoic or holozoic in nutrition. As pointed out by Brown
(1934), Oodinium ocellatum is saprozoic during its parasitic stage and
growth continues until the organism severs its connection with the host
tissues. In addition to saprozoic nutrition, however, there is some evidence
that 0. ocellatum may carry on photosynthesis. Laboratory cultures were
carried successfully in filtered and sterilized sea-water. In addition, the
presence of chromoplastids and the demonstration of starch in the flagel-
late suggests the possible importance of holophytic nutrition in the unat-
tached stages of the life-history. Combined methods of nutrition have been
reported for other dinoflagellates, e. g., the holozoic-holophytic type in
Gymnodinium (Kofoid and Swezy, 1921). The chromoplastids, apparently
the structures referred to as “refringent granules” by Brown, are usually
a pale to a definite green in color and in degenerating cells take on an ochre
coloration. Such chromoplastids have never been recorded for Oodinium ,
although they have been reported for other parasitic dinoflagellates (e. g.
Paulsenella ) . Amyloid granules, similar to those observed in Oodinium ocel-
latum, have been described in all the recognized species of the genus.

The ocellus of the parasitic stage of O. ocellatum appears to be a true
eye-spot (as defined by Kofoid and Swezy), but of a primitive type, inter-
mediate between the simple stigma and the more complex ocellus of the
Pouchetidae. In O. ocellatum a black and a red pigment bar are associated,
but whether or not a definite hyaline lens is present between them could not
be determined, although the area between the two bars is highly retractile.
For the Pouchetidae, on the other hand, the ocellus is made up of two parts,
a retractile, hyaline lens and a pigment mass, the melanosome. The simplest
form of melanosome is one in which there is a loose aggregation of pigment
granules massed together on one side of the lens. In the more highly de-
veloped types, the core of the melanosome contains a red pigment.

The behavior of the more complex types of ocelli during division is en-
tirely unknown. As was mentioned above, the ocellus in Oodinium ocellatum
may disappear at the beginning of division only to re-appear at the end of
the process. In one case, however, the ocellus seemed to be dividing. It
was seen as an elongated structure near the surface between the dividing
cell and was devoid of the black pigment bar. This phenomenon may not be
so rare as the writer’s observations might indicate, but there is neverthe-
less no evidence that the ocellus usually divides in fission. The red pigment
rods (“erythrosomes”) here observed for the first time in Oodinium, usually
did not appear in the cytoplasm until the beginning of each division and
after the ocellus had disappeared. At the end of division, the red rodlets
disappeared, and the elongated ocelli were again formed at the periphery of
each daughter cell. This might suggest the breaking up of the red pigment
of the ocellus into small rods at the beginning of fission, such as was re-
ported by Hall and Jahn (1929) for the granules of the stigma of Euglena.
In other cases, however, both ocellus and erythrosomes were present at the

152

Zoological New York Zoological Society

[XXI: 12

same time. Evidence derived from degenerating cells also seems to indicate
that the erythrosomes and the ocelli are not of the same nature. Thus, just
prior to cytolysis the erythrosomes would change in color from red to
orange-red, orange and finally to yellow while the ocelli always remained red.
When cytolysis finally occurred the erythrosomes immediately disappeared
while the ocelli separated in two parts (black and red), persisted for some
time as such but eventually disappeared in the sea-water. The latter ob-
servations agree with the findings of Kofoid and Swezy (1921) for the
ocellus of Erythropsis extrudens.

The “canal” described by Brown for the parasitic form of 0. ocellatum
was also noted in the stained preparations of the writer’s material. This
“canal” which is stained lightly with eosin, extends from the peduncle to
the achromatic mass in the vicinity of the nucleus, and is perhaps involved
in the intake of fluid, as in certain free-living species (Kofoid, 1909).
Brown (1934) also regards this canal as homologous with the canal of the
free-living dinoflagellates, but the achromatic mass in which it terminates
is interpreted by her as the sac-pusule (associated with the canal in free-
living forms). The writer was unable to observe the small vesicle described
by Brown as emptying into the “canal” and interpreted as a collecting
pusule, such as described by Schiitt (1895). In the division stages or in
the free-swimming forms of 0. ocellatum , no “canal” or related structures
were observed either by Brown or by the writer.

Contractile fibers in the peduncle, such as reported by Brown, were
not observed in the present material. Fibers of this sort were reported by
Dogiel (1910) for Oodinium sp. and by Chatton (1920) for 0. fritillaria .
The latter investigator figured and described a complicated mass of fibers
passing from the broad basal disc of O. fritillaria and inserting in a
granular mass of cytoplasm adjacent to the nucleus. These fibers, according
to Chatton, are used in expanding or retracting the disc. Although similar
fibers are figured by Dogiel (1910) for Oodinium sp., he does not discuss
them in his text. On the other hand, he does refer to the “pseudopodial”
behavior of the rhizoids. In the present material, the rhizoids behave as
“pseudopodia” during retraction. Brown (1934) reported that the fibrils
were seen in the “stalk” of the attached parasites of Oodinium ocellatum.
In the writer’s material, no such fibers were noted in attached parasites
stained either with Mallory’s triple stain or by Van Giesen’s method.

In the genus Oodinium sporulation has been observed in three species,
0. poucheti , O. amylaceum and 0. ocellatum, and the present investigation
is the first in which most of the morphological changes have been observed.
The general features of sporulation in 0. ocellatum have been discussed by
Brown (1934), who noted that the parasitic form begins to divide, regard-
less of size, once it is detached from the gills of the host. This observation
has been verified in the present investigation. The writer has found that
just before division begins there is a sudden increase in size of the flagellate
as a result of imbibition of water; this is contrary to Brown’s statement
that no further increase in size occurred in water. The association of im-
bibition of water with the division cycle has been reported previously by
Entz (1931), who showed that Ceratium hirudinella increases in size in this
manner immediately following the division and that such a process takes
place only at this period in the life-cycle. Later increases are due to real
addition of living substance.

With the exception of the members of the two genera, Oodinium and
Paulsenella, the ectoparasitic dinoflagellates sporulate by a process desig-
nated by Chatton (1920) as “palisporogenese.” In this type of reproduction,
the first division, which is transverse, gives rise to two daughter cells.
One of these cells divides repeatedly, eventually giving rise to free-swim-
ming dinospores, while the other merely increases in size at first. However,

1936]

Nigrelli: On Oodinium ocellatum Brown

153

when sporulation is completed in the first line, the second of the original
cells divides into two cells, one of which gives rise to a second generation
of dinospores. As reported elsewhere in this paper, a somewhat similar
process has been occasionally observed in Oodinium ocellatum. In this case,
one of the two original daughter cells gave rise to dinospores at the end of
the 32 cell stage, while the second merely increased in size. However, it
was not noted if the latter cell also gave rise to dinospores.

Under certain experimental conditions (density and temperature of
sea-water) division of 0. ocellatum is more or less a regular process giving
rise to palmellas of 2, 4, 8, 16, 32, 64, 128 and 256 cells. The products
of the last division are flagellated dinospores. According to Brown (1934),
temperature is the important factor influencing sporulation, and she found
the optimum to lie between 23° C. and 27° C. The writer has shown that
density of the sea-water is another factor important in the development of
the flagellated stages. These results are also of interest in their bearing
on the adaptability of 0. ocellatum. Kofoid and Swezy (1921) have pointed
out the delicate nature of the majority of the unarmored dinoflagellates,
which are extremely sensitive to handling and to changes in salinity, tem-
perature and pressure. Oodinium, on the other hand, is a very hardy type
and is more nearly comparable with such forms as Oxyrrhis, Amphidinium
and Gymnodinium which Kirby (1934) described from the salt marshes in
salinities ranging from 3.5% to saturation.

The morphological changes involved in the transformation from the
parasitic type to the free-swimming dinospores have not been recorded pre-
viously for any of the species of Oodinium or for many of the parasitic dino-
flagellates. The movement and alignment of amyloid granules in the binu-
cleate stage of the dinospore, the development of the neuromotor apparatus
from the stigma complex, the secretion of a new cellulose membrane, the
development of the girdle and sulcus, the migration and orientation of the
neuromotor apparatus to its final position in the free-swimming stage are
observed here for the first time. Presumably, similar morphological changes
may be expected in comparable stages in the life-histories of other species.

The changes in the metamorphosis from the dinoflagellate stage to the
parasitic type are also described for the first time. As was mentioned above,
the flagellar apparatus is lost and the sulcus widens out to form a cone-
shaped structure, the tip of which becomes extruded and possibly gives rise
to the peduncle and all its processes.

The development of peculiar structures from the sulcal region has also
been reported in other dinoflagellates. The pseudopodia described by
Zacharias (1899) for Gymnodinium , the prod of Erythropsis (Hertwig,
1884), and the tentacles of other species are all developed from an extremely
plastic sulcal area. In nearly every case, with the development of these
specialized structures, there is a loss of one flagellum, usually the longitudinal
one. In Oodinium, however, both flagella are lost, and the girdle disap-
pears completely in growth of the more or less pear-shaped parasitic stage.

Nuclear division in Oodinium ocellatum is of the paramitotic type and
similar in some respects to the process described by Calkins (1899) and
Ishikawa (1899) for Noctiluca. As in Oodinium, Calkins noted the sphere
in the resting cells while Ishikawa had only observed this structure in later
stages of division. Chatton (1920) has figured (PI. 1, Fig. 10) a similar
differentiated mass of cytoplasm lying adjacent to the nucleus of O. fritil-
laria. However, the significance of this mass was not discussed. In the
parasitic stage of 0. ocellatum, this mass of cytoplasm extends to the region
of the peduncle and contains many basophilic granules (microsomes) . Ac-
cording to Calkins, these microsomes are first found in a peripheral zone
of the sphere in Noctiluca and only as division approaches do these granules
become concentrated within the sphere. The relation of these granules to

154

Zoologica: New York Zoological Society

[XXI: 12

mitosis is not known. In late stages of division of Oodinium similar granules
are found localized in the “forks” of the bifurcation of the strands passing
from the spheres. Chatton (1914, 1920) reported siderophilic granules
in the centrosphere of Blastodinium but gave no explanation as to what
they might be. In this species, the archoplasm or sphere is composed of a
granular central zone and a more or less thick, homogenously-staining peri-
pheral zone. Its appearance in the vegetative stage is not reported. How-
ever, in certain trophocytes (binucleate stages) centrospheres with astral
rays surrounding them are present at opposite poles of large nuclei. Each
nucleus contains several nucleoli and is traversed by filaments which he
terms plasmodendrites. These are the remains of the nuclear spindle fibers
formed by the division of the centrosphere. He points out further that
these peculiar structures are formed in sporocytes of all ages, but in the last
sporocyst division the centrospheres and achromatic figures disappear and
a simple type of “hapiomitosis” results. In both Noctiluca and Oodinium ,
however, “spheres” are noted throughout all the division stages, and, in
Oodinium at least, up to and including the last dinospore division. Dogiel
(1908) has also seen spheres with central spindle in Haplozoon armatum
but here again the early stages in the formation of this extra-nuclear struc-
ture were not observed.

The behavior of the spindle in 0. ocellatum during division is some-
what different from that of Noctiluca. In the latter, according to Calkins,
it elongates as the prophase chromosomes are being formed and at the end
of this nuclear phase it consists of two daughter-spheres connected by a
“central-spindle.” The nucleus elongates and bends to form a C-shaped
figure and the central spindle sinks into the depression. The spindle, there-
fore, lies in the secondary axis of the nucleus, which encircles it, the sphere
alone remaining outside. When the nuclear plate is formed, it is wrapped
around the spindle like a ring, the chromosomes lying midway between
the two poles. A similar arrangement of the nuclear plate was noted by
Dogiel (1908) for Haplozoon. In O. ocellatum no such behavior of the
nucleus was observed. In the early stages of division, the achromatic mass
elongates. The chromosomes elongate, thicken and lie in parallel rows at
right angles to the dividing spindle. As in Noctiluca, the chromosomes at
this stage are attached to the spindle by means of mantle fibers and as
further division of the spindle occurs they eventually form a “plate.” How-
ever, there is no evidence that this “plate” encircles the extra-nuclear spin-
dle as in Noctiluca and Haplozoon.

According to Calkins (1899), this behavior of the spindle and the
chromosome is a constant feature throughout all the division stages of
Noctiluca. However, in Oodinium ocellatum, it has been observed that, after
the 4 cell stage, the arrangement of the chromosomes, prior to forming
the “plate,” is different. In these forms no early prophases were noted.
Here division is more or less rapid and the chromosomes, instead of orient-
ing themselves in parallel rows and at right angles to the elongated spindle,
are present as thin, radiating, V-shaped structures within the “sphere.”
Presumably, these V-shaped chromosomes straighten out as the spindle
elongates to form a typical “plate.” It is further assumed that the division
of the chromosomes is completed by a transverse fission, much like the
condition reported by Hall (1925 a and b) for Ceratium and Oxyrrhis.
Chatton (1920, 1921) reported similar radiating V-shaped chromosomes for
the parasitic dinoflagellate Syndinium turbo. In this form, however, the
chromosomes are doubled by a longitudinal splitting of the entire V, starting
at the apex and continuing along both “arms.” There is no evidence that
such a process is present in the late palmella stages of Oodinium.

In the free-living dinoflagellates, “spheres” of the type found in Noc-
tiluca and Oodinium have not been definitely reported. According to Kofoid
and Swezy (1921), in certain of Borgert’s (1910 a) figures of Ceratium

1936]

Nigrelli: On Oodinium ocellatum Brown

155

tripos there are suggestions of an archoplasmic structure corresponding to
spindle and polar regions. Hall (1925 a) reported a similar condition in
Oxyrrhis. He states that “In one case (PL 28, Fig. 13), a noticeable dif-
ference in the two poles of the nucleus is seen; at the anterior end the
chromosomes have not yet converged, but seem to extend to a clear area of
the cytoplasm. This condition is quite similar to that at the ends of the
amphiaster of Noctiluca (Calkins, 1899, PI. 42, Fig. 31), the clear area of
the cytoplasm resembling a centrosphere of metazoan cells.” In the encysted
stage of Ceratium hirudinella, reported by Hall (1925 b), a closer similarity
to the condition found in Noctiluca and Oodinium is present. Thus, Hall
figures and describes for this form a U-shaped nucleus (pi. 8, figs. 35-36)
and in the subnuclear area (pi. 8, figs. 35, 37, 38 and text-fig. D, 1-6) may
be seen a differentiated mass of cytoplasm that strongly suggests the
“sphere” of the resting and early stages of division of Oodinium ocellatum.

The origin of the chromosomes is also different in O. ocellatum and
Noctiluca. In the resting cell of the attached parasite, the chromatin is
present in the form of very short, densely staining “threads.” When the
unattached individual takes in water, the nucleus and the cytoplasm both
increase in volume. In such forms, the short “threads” apparently lose
their ability to stain densely. In the prophase, long, thin and lightly-stain-
ing chromosomes are present. In later prophase and metaphase, the chromo-
somes are long, thick and more densely stained. In Noctiluca , according to
Calkins (1899) and Ishikawa (1899), the chromatin is contained in large
endosomes, which are referred to as chromatin reservoirs, each of which
breaks up into a mass of chromomeres. These collect in chromosome strings
or “chromospires.” However, the “resting” nucleus of 0. ocellatum is more
like the condition found in the majority of the free-living dinoflagellates.
In such forms as Ceratium and Oxyrrhis (see Lauterborn, 1895; Borgert,
1910 a and b; and Hall, 1925 a and b) the chromatin of the interphase
nucleus is present not in the form of disconnected granules (as reported by
Entz, 1921) but as chromomeres combined into distinct chromosomes. How-
ever, in certain of the recently detached stages of 0. ocellatum , following
the retraction of the polar processes, what appeared to be disconnected
granules were observed. Such an appearance was found only at this
stage. Chatton (1920), on the other hand, reported that in O. poucheti,
granular chromatin was present in young parasites and free-swimming
dinospores.

In 0. ocellatum , “centrospheres” containing diplosomes were often
noted, especially in certain division stages of the palmellas. Similar cen-
trioles were reported by Calkins for Noctiluca during metaphase and ana-
phase stages, and appeared to be the focal points of the mantle fibers. Ac-
cording to this investigator, “The centrosomes, possibly, come from the
nucleus, where, during the resting stages, a small, deeply staining granule
can be easily distinguished from the chromatin. This granule disappears
during the early stages of chromosome formation.” It later appears in
the “sphere.”

In 0. ocellatum there is some evidence which indicates that the cen-
trioles are not nuclear in origin, but rather arise from the black pigment
portion of the ocellus complex. This is supported mainly by the fact that in
living dinospores the flagellar apparatus is intimately associated with the
red pigment bar of the ocellus, and in stained specimens, the diplosomes
while still within the sphere were seen giving rise to the flagella. In early
stages (just after retraction of the polar processes) two elongated bars
with granules at each end were seen within a clear area in the posterior
region of the body and strongly suggesting the appearance of the diplosome
structure. In some cases, two diplosomes were found in the same region,
each of which seem to be connected by fibers coming from the “sphere ;”
unfortunately this area was masked by numerous chromatoplastids so that

156

Zoologica: New York Zoological Society

[XXI: 12

the details and relationships of these fibers to the sphere and the nucleus
could not be determined. However, it may be pointed out here, as Calkins
suggested, that it is very easy to mistake these centrioles for microsomes
present in the same region. Ishikawa (1899) reported a single large cen-
trosome for Noctiluca, much like the structure seen by Chatton (1920) for
Blastodinium.

Ever since the early work of Kofoid and his students (Kofoid and
Swezy, 1915 a and b, and Kofoid and Christiansen, 1915 a and b), the
neuromotor apparatus of flagellates and its behavior during mitosis has re-
ceived much attention. Jollos (1910) was one of the first to figure such an
organelle for dinoflagellates. He showed that in Gymnodinium fucorum the
apparatus consists of two blepharoplasts, two flagellar rhizoplasts passing
from the blepharoplasts to an extra-nuclear granule, and a rhizoplast con-
necting the latter structure with the endosome. However, the behavior of
this neuromotor system during mitosis was not traced. Chatton and Weil
(1924) also reported a neuromotor apparatus for Polykrikos schwartzi. In
this form two blepharoplasts were present, each of Which gave rise to two
unlike flagella. The blepharoplasts, in turn, were connected by rhizoplasts
to granules of a “desmose” found just outside of the nuclear membrane.
Here again, the complete behavior of this system of fibers was not followed.
Hall (1925 a and b) showed that in Oxyrrhis and Ceratium the neuromotor
apparatus was similar to that reported by Jollos, especially the one found
in Oxyrrhis. This investigator (Hall), however, was able to follow the
behavior of this system of fibers throughout division. He found that a
typical paradesmose (centrosome-paradesmose type) was formed between
two daughter centrosomes as they drew apart in the prophase and eventually
disappeared in the late anaphase. However, just before the end of the pro-
phase, the blepharoplasts disappear and each centrosome gives rise to new
flagella. Entz (1928), from observations on living and stained material,
found that in Gonyaulax poly gramma, the flagella ended in two blepharo-
plasts, but no connection was observed with the centrosphere lying near
the nucleus.

In the present material, no neuromotor apparatus was observed in
the parasitic stage, although, as discussed above, a broad “flagellum” is
present. During division “diplosome” centrioles within centrospheres are
often found. In later stages of development (dinospores) these centrioles
give rise to flagella. In living material the centrioles, together with their
“desmose” and flagella, are intimately associated with the ocellus. There-
fore, there is little doubt that the centrioles and the blepharoplasts are
identical.

Although no definite paradesmose was observed in 0. ocellatum during
the early formation of the palmella, a paradesmose-like structure was ob-
served in the late anaphase stage. Here, two sets of centrioles were present
on both sides of the massed chromosomes. These centrioles were connected
with each other by a long “desmose.” In other words, this apparatus
appears as a precociously developed paradesmose. Such a rapid division
of the achromatic figure, however, has been previously reported. Thus,
Ishikawa (1899) described and figured a division of the centrospheres in
Noctiluca miliaris while the chromosomes were still in the late metaphase.

The centrioles in the present material are of the centroblepharoplast
type and the central spindle can be considered analogous to a centroblepharo-
plast-paradesmose. Kofoid and Swezy (1921) consider the “sphere” of
Noctiluca as a structure analogous to the centrosome-paradesmose of other
flagellates. It is altogether possible, however, in view of the fact that the
centrioles in this form have not been traced through the final development
of the dinoflagellate, that this structure is similar to the one present in
0. ocellatum. Chatton (1920) reported a centroblepharoplast in an uncer-

1936]

Nigrelli: On Oodinium ocellatum Brown

157

tain form of dinoflagellate (an anhang to an uncertain genus, Atelodinium)
which he called “les spores a rostre.” In these spores definite centrosomes
were recognized at each pole of the mitotic figure. In later development,
these granules gave rise to an aciculated structure which Chatton called a
blepharoplast. However, one of his figures (pi. 17, fig. 193) shows a single
flagellum, not connected with this elongated “blepharoplast.” In 1921, Chat-
ton reported centrosomes which have rise to flagella in the parasitic dino-
flagellate Syndinium turbo.

Summary.

Oodinium ocellatum Brown is recorded for the first time from the
North American coast. The life-history has been redescribed and additional
details, previously unobserved, are recorded.

The life-history includes the following stages: (1) the parasitic stage,
a large pyriform organism anchored to the gill filaments of the host; (2)
palmella stages, in which fission occurs, develop from the parasitic stage
after it drops from the gill of the host; (3) flagellated dinospores; and (4)
a typical peridinian stage, with girdle, sulcus, transverse and longitudinal
flagella.

The parasitic stage is a large pyriform organism surrounded by a
cellulose membrane. It possesses chromoplastids, amyloid granules, ocellus
and an organelle of attachment composed of a peduncle, “flagellum,” and
fine rhizoid processes by which the parasite anchors itself to the gill fila-
ments of the host. The ocellus is a characteristic morphological feature of
this species. It is composed of a red or orange pigment bar connected to a
thinner black pigment bar by means of minute fibrils. The space between
the two is highly retractile, superficially appearing as a bar-shaped lens.

The parasite drops from the gill and immediately increases in size, as
a result of imbibition of water; the polar processes are retracted and the
opening closed by a secretion of a cap.

Fission is initiated at the pole opposite to the peduncle and the first
division is longitudinal. Just before each fission, erythrosomes or red pig-
ment rodlets usually appear near the surface of the organism and then dis-
appear at the end of division. The ocellus is usually lost when the erythro-
somes are present and reappear when the rodlets disappear.

Under certain densities of sea-water and temperatures, the organism
divides, giving rise to palmellas of 2, 4, 8, 16, 32, 64, 128 cells. The 128
cell stage divides once again to give rise to 256 free-swimming dinospores.
Experimentally, it has been shown that specific gravity is an important
factor concerned in the rate with which dinospores are formed. It also has
been shown that dinospores may develop at the end of 8, 16, 32 or 64 cell
stages and that this precocious development of flagellates is dependent in
some degree upon the specific gravity of the sea-water.

The dinospores, when first released, are without a cellulose covering.
In these forms, the neuromotor apparatus is part of the ocellus complex, each
flagellum arising from a blepharoplast at one end of the black pigment bar,
which acts as a “desmose.” Later, these naked dinospores settle to the bot-
tom, secrete a new cellulose covering and gradually metamorphose into a
typical peridinian dinoflagellate with a girdle and sulcus. This typical dino-
flagellate is the infective form.

In the transformation from the pelagic type to the sessile form found
on the gills, the flagella are lost and the sulcus hollows out to form a cone-
shaped area, the apex of which is later extruded. It is believed that the
polar processes together with the “flagellum” are developed from this highly
plastic sulcal region.

The vegetative nuclei of 0. ocellatum are spherical or oval structures,

158

Zoologica: New York Zoological Society

[XXI :12

containing chromatin in the form of short, densely staining “threads.” One
or more endosomes may be present, but these structures do not take part in
the division process.

When the parasites drop from the gills both the cytoplasm and nucleus
increase in volume as a result of imbibition of water. The chromatin in
these cells fails to stain densely.

A differentiated mass of cytoplasm is present, which in the attached
parasites extends to the region of the peduncle. In these forms, many
microsomes are found within this mass. In the unattached parasites, espe-
cially after rounding up has taken place, two granular cytoplasmic areas
become evident and are connected with each other by means of a “canal.”
One is localized in the subnuclear region and eventually gives rise to the
extra-nuclear spindle ; the other is formed as a result of the disintegration
of the polar processes after retraction.

During division, the achromatic mass or “sphere” elongates to form a
central spindle. An interpretation is reported which may throw some light
on the mechanism involved in the division of this spindle and the separation
of the chromosomes.

Mitosis is of the paramitotic type. Two kinds of chromosome behavior
were noted, one which takes place in the initial mitotic cycle and the other
in palmellas after the 4 cell stage.

In the early stages of the first mitotic cycle, the chromosomes are
present in the form of long, thin and lightly staining threads. The nuclear
membrane disappears, and in later stages the chromosomes become thicker
and more densely stained. At this stage they lie in parallel rows and at
right angles to the elongated spindle. From the ends of the divided chromo-
somes, mantle fibers pass into the spindle and as the latter structure con-
tinues to elongate, the chromosomes are gradually drawn on the spindle and
assume a metaphase “plate” appearance. There is no evidence that this
“plate” encircles the extra-nuclear spindle as in Noctiluca and Haylozoon.
The chromosomes continue to move to opposite poles and in early telophase
maintain their parallel arrangement for a short time but eventually pass
into the prophase of the next cycle. There is no evidence that a “resting”
stage occurs at the end of division. In later telophase a thin nuclear
membrane is reorganized, in some cases even before the spindle fibers are
completely obliterated.

In the later palmella division (after the 4 cell stage), chromosomes
typical of the early prophases of the initial cycle were not noted. Instead of
lining up at right angles to the spindle, the chromosomes appear within the
“sphere” as radiating V-shaped structures. It is assumed the V is the
result of a unipolar splitting, and as the “sphere” elongates the chromo-
somes become straightened out to form the metaphase “plate.” At this
stage, presumably, the daughter chromosomes are separated by a trans-
verse fission. In the telophase, no nuclear membrane is evident. In such
forms, the chromosomes pass directly into the prophase of the next division.

References.

Bargoni, E. (1894). Di un Foraminifero parassita nelle salpe ( Salpicola amy-
lacea, n. g., n. sp.) e considerazioni sui corpusouli dei Protozoi superiori.
Ricerche fatte nel laboratorio di Anatomia Normala della R. Universita de
Roma. 4: 43-64, Pis. 3-4.

Belar, Karl (1926). Der Formwechsel der Protistenkerne. Eine vergleichend-
morphologische Studie. Ergebnisse und Fortschritte der Zoologie. 6: 235-654,
263 figs, and 3 tables in text.

1936]

Nigrelli: On Oodinium ocellatum Brown

159

Bogert, A. (1910 a). Kern un Zelltheilung bei marinen Ceratium-Arten. Arch .
/. Protist. 20: 1-46, pis. 1-3.

(1910 b). Die Mitose bei marinen Ceratium-Arten. Zool. Anz. 35: 641-644,

3 figs, in text.

Brooks, W. and Kellner, C. (1908). On Oikopleura tortugensis, a new appen-
dicularian from the Tortugas, Florida, with notes on its embryology. (With
a note on a species of Gromia, G. appendiculariae) . Papers from the Tor-
tugas Lab. of the Carnegie Institution of Washington, D. C. Carnegie Inst.
Pub. (102), 189-194, 5 pis.

Brown, E. M. (1931). Note on a new species of dinoflagellate from the gills and
epidermis of marine fishes. Proc. Zool. Soc. London, Part 1, 345-346.

(1934). On Oodinium ocellatum Brown, a parasitic dinoflagellate causing
epidemic disease in marine fish. Proc. Zool. Soc. London. Part 3, 583-607,
pis. 1-3, 3 figs, in text.

Calkins, G. N. (1899). Mitosis in Noctiluca and its bearing on the nuclear rela-
tions of the protozoa and metazoa. Jour. Morph. 15: 711-772, pis. 40-42.
(1915). Microtoeniella clymenellae, a new genus and new species of colonial
gregarines. Bio. Bull. 29 : 46-49.

(1926). Biology of the protozoa. 623 pages. Lea and Febiger Publishers.
Philadelphia and New York.

Caullery, M. (1906). Personal communication with Chatton re O. amylaceum.
See Chatton (1920).

(1915). Ellobiopsis chattoni, n. g., n. sp. parasite de Calanus helg olandicus
Claus, appartenant probablement aus Peridiniens. Bull. Sc. France-Belgique.
44: 201-214, pi. 5.

Chatton, E. (1906). Les Blastodinides, order noveau de Dinoflagelles parasites.
C. R. Acad. Sci. 143: 981-983. 5 figs, in text.

(1907). Nouvel apergu surples blastodinides ( Apodinium mycetoides, n. g.,
n. sp.). C. R. Acad. Sci. 144: 282-285, 7 figs.

(1910 a). Sur Texistence de Dinoflagelles parasites Coelomiques. Les
Syndinium chez copepods pelagiques. C. R. Acad. Sci. 147 : 654-656.

(1910 b). Paradinium poucheti, n. g., sp., flagelle parasite d’Acartia clausi
Giesbrecht (copepod pelagique) C. R. Soc. Biol. 69: 341-343.

(1912). Diagnose preliminaires de Peridiniens parasites nouveaux. Bull.
Soc. Zool., France. 37 : 85-93, 7 figs, in text.

(1914). Transformation evolutives et cycliques de la structure peridiniens
chez certaines Dinoflagelles parasites. C. R. Acad. Sci., Paris. 158: 192-195,

4 figs, in text.

(1920). Les' Peridiniens parasites. Arch, de Zool. Exper. et Gen. 59: 1-475,
pis. 1-18, 160 figs, in text.

(1921). Sur un mecanisme cinetique nouveau: la mitose syndinienne chez
les Peridiniens parasites plasmodaux. C. R. Acad. Sci. 173: 859-862, 1 fig.
(1923). Les Peridiniens parasites des Radiolaires. C. R. Acad. Sci. 177:1246.

Chatton, E. and Weil, R. (1924). Sur l’appareil flagellaire des peridiniens et en
particulier du Polykrikos schwartzi. C. R. Soc. Biol. 91: 580-583, 6 figs.

Dogiel, V. (1906 a). Haplozoon armatum n. g., n. sp. der Vertreter einer neuen
Mesozoa-gruppe. Zool. Anz. 30: 895-899.

(1906 b). Beitrage zur Kenntniss der Peridineqn. Mitt, aus der Zool. Stat.
zu Neapel. 18: 1-45, pis. 1-2.

(1908). Catenata eine neue Mesozoen gruppe. Zeits. f. wiss. Zool. 89:
417-487, pis. 26-28.

(1910). Untersuchungen fiber einige neue Catenata. Zeits. f. wiss. Zool. 94:
400-446, pis. 13-14.

Entz, G. Jr. (1921). Uber die mitotische Teilung von Ceratium hirudinella. Arch,
f. Protist. 43: 416-431, pis. 13-14, 10 figs, in text.

(1928). Uber den Bau und fiber die Tatigkeit Geisseln der Peridineen. Ann.
Protistol. 1: 75-95, 29 figs.

160

Zoologica: New York Zoological Society

[XXI: 12

Entz, G. Jr. (continued)

(1931). Analyse des Wachstums und der Teilufig einer Population sowie
individums des Protisten Ceratium hirudinella unter den naturlichen ver-
haltnissen. Arch. f. Protist. 74: 310-361, 18 figs.

Hall, R. P. (1925a). Binary fission in Oxyrrhis marina Dujardin. Univ. Calif.
Publ. Zool. 26: 281-324, pis. 26-30, 7 figs, in text.

(1925 b). Mitosis in Ceratium hirudinella O. F. M., with notes on nuclear
phenomena in encysted forms and the question of sexual reproduction. Univ.
Calif. Publ. Zool. 28: 29-64, pis. 5-9, 5 figs, in text.

Hall, R. P. and T. L. Jahn (1929). Dispersed stages of the stigma in Euglena.
Science. 69: 522.

Hall, R. P. and R. F. Nigrelli (1931). The vacuome of the flagellate Chlamy-
domonas. Jour. Morp. 51: 527-541, pi. 1, 8 figs, in text.

Hertwig, R. (1884). Erythropsis agilis: Eine neue Protozoa. Morph. Jahrb.
10: 204-212, pi. 6.

Hovasse, R. (1922). Endodinium, chattoni nov. gen. et sp. C. R. Soc. Biol. 87: 845.

Ishikawa, C. (1899). Further observations on the nuclear division of Noctiluca.
Jour. College Sci. Imp. Univ. Japan. 6: 334-397, pis. 11-14.

Jollos, V. (1910). Dinoflagellatenstudien. Arch. f. Protist. 19: 178-206, pis. 7-10.

Kirby, H. Jr. (1934). Some ciliates from salt marshes in California. Arch. f.
Protist. 82: 114-133, 4 pis.

Klebs, G. 1(883). Ueber die Organization einiger Flagellatengruppen und ihre
Beziehungen zu Algen und Infusorien. Unters. Bot. Inst. Tiibigen, 1, 13 p.,
2 pis.

Kofoid, C. A. (1909). On Peridinium steinii Jorgensen, with a note on the nomen-
clature of the skeleton of the Peridinidae. Arch. f. Protist. 16: 25-47, pi. 2.

Kofoid, C. A. and E. B. Christiansen (1915 a). On the life-history of Giardia.
Proc. Nat. Acad. Sci. 1 : 547-552, 1 fig. in text.

Kofoid, C. A. and E. B. Christiansen (1915 b). On binary and multiple fission
in Giardia muris (Grassi). Univ. Calif. Publ. Zool. 16: 30-54, pis. 5-8, 1 fig.
in text.

Kofoid, C. A. and O. Swezy (1915 a). Mitosis in Trichomonas. Proc. Natl. Acad.
Sci. 1: 315-321, 9 figs, in text.

(1915 b). Mitosis and multiple fission in trichomonad flagellates. Proc. Am.
Acad. Arts and Sci. Boston. 51: 289-378, pis. 1-8, 7 figs, in text.

(1921). The free-living unarmored dinoflagellates. Memoirs of the Univ.
Calif. 5 : 1-562, 12 pis., 388 figs, in text.

Lauterborn, R. (1895). Protozoenstudien I. Kern- und Zellteilung von Ceratium
hirudinella (O. F. M.). Zeits, f. wiss. Zool. 59: 167-191, pis. 12-13.

Lemmermann, E. (1899). Planktonalgen. Ergebnisse einer Reise nach den Pacific
(H. Schauinsland, 1896-1897). Abhandl. Nat. ver Bremen. 16: 314-398, pis.
1-3.

Martin, G. W. and T. C. Nelson (1929). Swarming of dinoflagellates in Dela-
ware Bay, New Jersey. Bot. Gaz. 88: 218-224, 4 figs.

Meade, A. D. (1898). Peridinium and the “red water” in Narragansett Bay.
Science 8: 707-709.

Nigrelli, R. F. (1935). Experiments on the control of Epibdella melleni Mac-
Callum, a monogenetic trematode of marine fishes. Jour. Parasitol. 21 : 438,
abst. no. 40.

Pouchet, G. (1884). Sur un peridinien parasite. C. R. Acad. Sci. 98: 1345-1346.
(1885 a). Nouvelle contribution a l’histoire des Peridiniens marins. J. Anat.
Phys. 21 : 28-83, pis. 2-4.

1936]

Nigrelli: On Oodinium ocellatum Brown

161

(1885 b). Troisieme contribution a l’etude des Peridiniens. J. Anat. Phys. 21:
525-534, pi. 26.

Reichenow, E. (1930). Parasitische Peridinea (einschliesslich Ellobiopsidae) .
In. G. Grimpe u. E. Wagler. Die Tierwelt Nord-u-Ostsee. Lief 19, Teil II.
d3 85-100, 10 figs. Akad. Verglagsges. Leipzig.

Schutt, F. (1895). Peridineen der Plankton-Expedition der Humboldt-stiftung.
4: 1-170, 27 pis.

Zacharias, O. (1899). Ueber Pseudopodienbildung bei einem Dinoflagellaten.
Biol. Centrbl. 19: 141-144, 9 figs, in text.

162

Zoologica: New York Zoological Society
EXPLANATION OF THE PLATES.

[XXI: 12

Plate I.

Camera lucida drawings of various stages in the life-history of living Oodinium
ocellatum. x 950.

Parasite recently detached from the gills of a host.

Rounding up and the secretion of a cellulose cap. Note the erythrosomes.
Stage showing the recession of the cell from the outer membrane. This
is the anterior end and the point where fission will start.

First division of a smaller individual. Note the stigma undergoing division.
16 cell palmella stage. Certain cells show the alignment of the amyloid
granules prior to the formation of the dinospores.

Naked dinospore. These forms are devoid of a cellulose membrane. The
stigma-neuromotor complex lies at the posterior end of the cell.

A new cellulose membrane is secreted. There is a space between the cov-
ering and the periplast, within which the free ends of the flagella eventu-
ally come to lie.

Transformation of the dinospore to the typical dinoflagellate condition.
The stigma-neuromotor complex is oriented so that the transverse flagel-
lum lies in the girdle.

This form shows the differentiation of the cytoplasm of the anterior re-
gion into a girdle and sulcus. The flagellar apparatus has begun to move
anteriorly.

Typical free-living Oodinium ocellatum.

Plate II.

Camera lucida drawings of attached parasites, x 950.

Figs. 11-16. Corrosive sublimate fixation followed by iron-hematoxylin and Van
Giesen’s stain.

Fig. 11. Note the canal extending from the peduncle toward the center of the cell.

Figs. 12, 13. Note the ring in the peduncle. This structure stains yellow with Van
Giesen’s. Note the “microsomes” within the differentiated mass of cyto-
plasm.

Fig. 14. Section of a large parasite showing the chromoplastids. Note the densely
staining structures within the plastids some of which indicate division.

Fig. 15. Young parasite.

Fig. 16. Section of a parasite through the “sphere.” Note the plastids.

Figs. 17-20, Zenker’s fixation followed by iron-hematoxylin and eosin.

Fig. 18. Section at the level of the nucleus, showing the short interphase chromo-
somes.

Fig. 20. Individual showing the pseudo-flagellum. Note the radiating arrangement
of the chromomeres.

Figs. 21-24. Zenker’s fixation followed by Mallory’s triple stain. In these forms
the chromoplastids, ring and microsomes stain red; the cellulose wall
and the edge of the plastids a deep blue, while the amyloid granules take
on a lighter blue color.

Fig.

1.

Fig.

2.

Fig.

3.

Fig.

4.

Fig.

5.

Fig.

6.

Fig.

7.

Fig.

8.

Fig.

9.

Fig.

10.

Plate III.

Camera lucida drawings of detached parasites, x 950. Bouin’s fixation followed by
Delafield’s hematoxylin and counterstained with eosin.

Fig. 25. Detached parasite after a certain amount of swelling has occurred as a
result of imbibition of water. Note the chromatin in the form of distinct
granules. The peduncle and the rhizoids are slightly retracted.

Fig. 26. Note the elongation of the endosome. The structure, however, does not
take part in the nuclear division of this species. The “sphere” may be
seen just posterior to the nucleus.

1936]

Nigrelli : On Oodinium ocellatum Brown

163

Fig. 27. Typical detached parasite. Note the canal extending from the peduncle
to the “sphere” mass.

Fig. 28. Early prophase. The “sphere” mass in this individual is cup-shaped.

Fig. 29. Early prophase. The nuclear membrane has disappeared and the “sphere”
shows some elongation.

Plate IV.

Camera lucida drawings of mitosis in Oodinium ocellatus. x 950. Material fixed
in Zenker’s and stained with iron-hematoxylin.

Fig. 30. Early prophase. The nucleus has elongated, but the chromosomes are still
in the shortened phase. Note the “diplosome” in the center of the sphere
and the scattered microsomes.

Fig. 31. Early prophase. The chromosomes begin to show their parallel arrange-
ment. The nuclear membrane next to the sphere has disappeared and
distinct mantle fibres can be seen passing from the nucleus and converg-
ing towards the center of the sphere.

Figs. 32-34. Early prophases. These stages are not completely understood as yet.

In Fig. 32, the nucleus is a bilobed structure with both lobes extended
upwards; the chromosomes are parallel to each other and mantle fibres
can be distinguished. The sphere shows what appears to be the beginning
of the protoplasmic strands. In Fig. 33 the nucleus shows a still further
separation although the sphere has not begun to elaborate as yet. This
Figure and Fig. 34 might indicate that the nucleus is forming a C-
shaped structure somewhat like that in Noctiluca. However, no trace of
the part between the ends could be found.

Fig. 35. Late prophase. The chromosomes thicken and elongate considerably. The
sphere has begun to elongate. In this form a remnant of the nuclear
membrane is still present, although usually when the chromosomes have
reached this stage the nuclear membrane is entirely lacking.

Plate V.

Camera lucida drawings of mitosis in Oodinium ocellatum . x 950. Material fixed
in Zenker’s and stained with iron-hematoxylin.

Fig. 36. Late prophase. The chromosomes are being drawn upon the central spin-
dle which is formed as the sphere divides further and further.

Fig. 37. Slightly earlier stage than Fig. 36.

Figs. 38, 39. Prophase. Many forms were encountered with this sheaf-like forma-
tion of the nucleus.

Fig. 40. Late prophase. Superficially this figure appears as a metaphase stage but
actually the chrosomes have not begun to migrate onto the spindle. Note
the diplosome centrioles to which the mantle fibres can be seen converging.

Fig. 41. Metaphase. The chromosomes are aligned on the central spindle. The final
separation has just started in the form represented in Fig. 42.

Fig. 43. Telophase. The chromosomes have shortened and cell division has oc-
curred. The daughter nuclei are still connected by the remains of the
central spindle.

Plate VI.

Camera lucida drawings of mitosis. Zenker’s fixation followed by iron-hematoxylin
stain, x 950.

Figs. 44, 45. Early and late anaphase. Note the central spindle fibers.

Figs. 46, 47. Telophase. In Fig. 47 the daughter nuclei are still connected by the
remains of the central spindle and a thin nuclear membrane is reor-
ganized round each nucleus.

Fig. 48. Late telophase. Reorganization of the daughter nuclei has occurred and
the “sphere” mass once again reappears in the subnuclear region.

Fig. 49. 4 cell stage. Note centrosphere with diplosomes and V-shaped radiating
chromosomes in two of the cells. In the two cells to the right, the sep-
aration of the chromosomes has occurred, presumably after the V’s have
straightened out.

164

Zoologica: New York Zoological Society

Plate VII.

Camera lucida drawings of nuclear division.

Fig. 50. Early anaphase. 4 cell stage. After the chromosomes are arranged upon
the central spindle, a transverse separation occurs and the migration to
the opposite poles begins.

Fig. 51. Anaphase. Note the processes passing from the poles of the sphere to the
periphery. These processes bifurcate and at the forks of the bifurcations
may be seen a small densely staining granule.

Plate VIII.

Camera lucida drawings of mitosis in Oodinium ocellatum. x 950. Stages show-
ing the V-shaped radiating chromosomes. Note the centriole in Fig. 52.
Fig. 53. Early telophase. The figure to the right is an optical section showing a
paradesmose (?).

Fig. 54 shows a divided “desmose” in the posterior part of the cell.

Fig. 55 shows two diplosomes to which mantle fibers (?) converge.

Plate IX.

Camera lucida drawings of dinospores, x 950. Zenker’s fixation and iron-hema-
toxylin stain.

Figs. 56-74. Metamorphosis of Oodinium ocellatum to the free-swimming dino-
flagellate stage.

Figs. 56-62. Dinospore prior to the final division.

Figs. 56-58. Late telophase.

Fig. 59. Note peculiar spiral formation of the cellulose membrane.

Figs. 60-63. Flagellation.

Figs. 64-66. Note the flagella growing out from the diplosome.

Fig. 67. Orientation of the neuromotor apparatus. Note cytoplasmic differentia-
tion which eventually will form the girdle and the sulcus.

Figs. 68-70. Various forms showing different stages and appearance of the neuro-
motor apparatus.

Fig. 71. Neuromotor apparatus torn from the cell. Note the blepharoplasts and the
connecting desmose.

Figs. 72-74. Cytoplasmic differentiation resulting in the development of the girdle
and sulcus. Note the migration of the neuromotor apparatus to its final
point in the sucal-girdle junction.

Fig. 75. Stage in the transformation from the free-swimming dinoflagellate to
the parasitic type. In this form the sulcus has developed into a cone-
shaped structure, the tip of which protrudes through an opening in the
cellulose cell wall.

NIGRELLI.

PLATE I.

CAMERA LUCIDA DRAWINGS OF VARIOUS STAGES IN THE LIFE-HISTORY
OF LIVING OODINIUM OCELLATUM.

NIGRELLI .

PLATE II.

OODINIUM OCELLATUM. CAMERA LUCIDA DRAWINGS OF
ATTACHED PARASITES.

NIGRELLI

PLATE III

28

29

OODINIUM OCELLATUM. CAMERA LUCIDA DRAWINGS OF
DETACHED PARASITES.

CAMERA LUCIDA DRAWINGS OF MITOSIS IN OODINIUM OCELLATUM.
MATERIAL FIXED IN ZENKER’S AND STAINED WITH I RON -HEMATOXYLIN.

NIGRELLI.

PLATE IV.

NIGRELLI

PLATE V

CAMERA LUCIDA DRAWINGS OF MITOSIS IN OODINIUM OCELLATUM
MATERIAL FIXED IN ZENKER’S AND STAINED WITH IRON-HEMATOXYLIN

iff

NIGRELLI.

PLATE VI.

OODINIUM OCELLATUM. CAMERA LUC1DA DRAWINGS OF MITOSIS. ZENKER’S
FIXATION FOLLOWED BY I RON- HEMATOXYLIN STAIN.

NIGRELLI.

PLATE VII.

OODINIUM OCELLATUM. CAMERA LUCIDA DRAWINGS OF NUCLEAR DIVISION.

NIGRELLI

PLATE VIII

55

CAMERA LUCIDA DRAWINGS OF MITOSIS IN OODINIUM OCELLATUM.
STAGES SHOWING THE V-SHAPED RADIATING CHROMOSOMES.

1

NIGRELLI.

PLATE IX.

OODINIUM OCELLATUM. CAMERA LUCIDA DRAWINGS OF DINOSPORES.
ZENKER'S FIXATION AND IRON-HEMATOXYLIN STAIN.

Breder & Nigrelli: Winter Movements of the Alewife

165

13.

The Winter Movements of the Landlocked Alewife, Pomolobus
pseudoharengus (Wilson).

C. M. Breder, Jr.

&

R. F. Nigrelli.

New York Aquarium.

(Text-figures 1-6).

Complaints by householders in New York City of small fishes emerg-
ing from their faucets, led Mr. Herman Forster, Deputy Commissioner of
the Department of Water Supply, Gas and Electricity, to inquire at the
New York Aquarium concerning the possibility of control of these free but
unwelcome fishes, see Forster (1935). In order better to understand the
problem, a study was undertaken of the conditions reported by him. The
fishes were found to be typical landlocked specimens of Pomolobus pseudo-
harengus (Wilson), Text-fig. 1. The data subsequently accumulated un-
covered certain features of the behavior of these fishes not hitherto under-
stood and are presented herewith. This study was possible because of cer-
tain peculiar features of the environment, due to its being part of the
municipal water supply system, added to the cooperation of the Department.

Shortly after the work was undertaken a paper on this same species,
Odell (1934), reached our hands, which covered a considerable portion of
the ground we had originally planned to include. In view of this and the
fact that the New York State Conservation Department Biological Survey
was to cover our territory in its 1936 work, we abandoned all but such
parts of the original plan that seemed to be complementary to the work
of Odell. There is otherwise a marked dearth of papers on this species and
since Odell has indicated their contents it is not necessary to refer to them
here, especially since their bearing on present problems is slight. Specific-
ally, Odell had no data on the whereabouts of his fish during the winter
months and wrote: “In summer it is wide-ranging in habit, having been
taken at all depths down to 160 feet. The winter distribution is unknown.”
Fortunately it is exactly this latter time which is covered best by our data.

It is during this period that Pomolobus appears in the drinking water.
During certain years great quantities may be taken from the protective
screens which cover the outlet of Kensico Reservoir — the lake which forms
the habitat of these fishes. The outlets here referred to drain at a depth
of from 30 to 60 feet from the surface. No regular records of the fish on
the screens had been kept, which were suited to our studies, but it was a
relatively simple matter to have the file of complaints in the Water Depart-
ment office examined for a period of years. For this information and for
permission to use it, we are indebted to Mr. Forster. The data covering

166

Zoologica: New York Zoological Society

[XXI :13

eleven years are given by months in Table I. Actually these figures repre-
sent all complaints of fish in the water supply, but other species, such as
Perea flavescens and Anguilla rostrata, are so rare as to be negligible.
The mean for the eleven-year period shows a very definite peak in October,
see Text-fig. 2. Observations made on the screens in 1936 and the fragmen-
tary data of other years at that place show the peak of fish on the screens
to occur in February or March. This suggests that the small fish drop
down first, or at least scatter out sooner, for they do not appear on the
screens until the larger fish form a mat for them to rest against. Since
Pomolobus is in more or less evidence at the upper end of this lake from
about May to September and is not to be found near the outlet until the
end of that period, it would clearly seem to follow that there is an autumn
movement downstream of both large and small fishes.

Text-figure 1.

The landlocked Kensico Reservoir alewife, locally called sawbelly, Pomolobus
pseudoharengus (Wilson). A typical example of the year’s hatch, 51 mm. stand-
ard length, Aug. 23, 1935. Fishes this size and smaller sometimes are found in
the municipal water distribution system. (Drawing by Ralph Graeter).

If the number of complaints are plotted by years, as in Text-fig. 3, it
becomes evident that there is a wide divergence from year to year in their
quantity. Varying economic or political events may be expressing them-
selves here to some extent — a difficulty not inherent in the treatment of
Text-fig. 2, which uses the mean value for a number of years. However this
may be, it cannot invalidate the effect of the fishes themselves, the data of
which also agree with the memory of workers in the Water Department.
In other words there were many Pomolobus on the screens in 1926 and
greater numbers than ever before were noted on them in 1934 and 1935.
The period between these two peaks of abundance as thus measured comes
to eight years.

It was impossible to obtain non-selected samples from the screens as
to number and size. An unknown number of small fish slipped through
the meshes, so that it was not possible even to guess at the quantities or
the size composition of the Pomolobus that passed the %-inch mesh of the
screens. Text-fig. 4 gives some idea of the screens and the fishes removed
from them. The screens were removed and cleaned irregularly, according
to the quantity of fish collected on them, but the records were of such a
nature that statistical analysis was not possible.

1936]

Breder & Nigrelli: Winter Movements of the Alewife

167

TABLE I.

Written complaints to the New York City Water Department of fishes in
the water supply, by months for eleven years.

During the winter of 1935 a large number of samples was taken from
the screens and these were measured and sexed. While this gives a fair
idea of sizes in each group, it cannot be considered a measure of relative

Text-figure 2.

Complaints of fishes emerging from taps made to the New York Department of
Water Supply. Mean values by months for eleven years from 1925 to 1935 in-
clusive.

168

Zoological New York Zoological Society

[XXI : 13

quantities of old and immature fish. Although the samples were random
ones, these were already highly selected piles of fish as cleaned from the
screen. This does not apply to the sex ratios of the adults, however, as the
separation of males from females, in point of size, was so slight that a
mechanical selection on this basis was out of the question. Bearing this
out is the fact that distribution curves, not published, were found to be
quite symmetrical and that the much smaller young were taken regularly.
Table II gives these figures.

It will be noted that the sex ratios in the last column of Table II show
a remarkable amount of variation. Apparently this is not accidental, since
it follows a well-marked rise and fall as shown in Text-fig. 5. It would
seem that the schools being caught on the screen varied in their sex com-
position in some regular manner. The various interpretations that may be
placed on this phenomenon will be discussed subsequently, as will the nature
of the 1936 sample.

During the late summer and fall large quantities of young fish are to
be found at the upper end of Kensico Reservoir at a point where the flume
from the Catskill watershed enters. A measured sample of these showed a
mode at 4 cm., see Table II and Text-fig. 1. Scales of these specimens showed
no winter ring but were entirely uniform in structure, as would be expected.
The larger fish taken on the screens had been almost uniformly scaled by
the tremendous washing they received, making adequate age analysis by
this method impossible. However, such scales as could be found showed
that the mean position of the first winter ring exceeded the August mode of
4 cm., and fell below the early spring mode of 5 cm., see Table III and Text-

Text-figure 3.

Complaints of fishes emerging from taps made to the New York Department of
Water Supply. Total number of complaints per year for eleven years from

1925 to 1935 inclusive.

1936]

Breder & Nigrelli: Winter Movements of the Alewife

169

fig. 6. The interpretation of these data would seem to indicate the following :
The fishes, very much in evidence during the summer months, prin-
cipally in the upper half of the lake, disappear from there as the water cools.
At the same time some of the smaller ones, of the same year, emerge from
household taps, while later the larger sizes catch on the coarse protective
screens along with smaller sizes caught only because of the restricted
openings induced by the larger fish. The peak of the movement of the
year’s fishes measured by observations of the workmen who are charged
with the care of the screens, and the records of water consumer com-
plaints, is during October and reaches chiefly from September to December.
The larger and older individuals appear in their maximum numbers in Feb-
ruary or March.

TABLE II.

Sizes and states of Pomolobus pseudoharengus taken from the screen at
the outlet of Kensico Reservoir.

1935

Female

Male

Immature

Total

Sex

Ratio

No.

Max.

Mode1

Min.

No.

Max.

Mode

Min.

No.

Max.

Mode

Min.

Feb. 15-16

109

12.3

11

9.5

18

13.5

10

9.5

115

7.0

5

3.5

242

6.4

Feb. 22.. . .

70

12.7

10

9.5

4

10.5

10

9.0

100

6.8

5

3.7

174

17.5

March 1.. .

26

12.0

10

9.2

2

9.5

9

9.3

95

8.2

5

3.4

123

13.0

March 18..

7

12.0

11

10.5

5

11.0

10

9.0

2

5.5

5

5.0

14

1.4

April 8

39

13.5

11

10.0

24

11.5

11

9.5

1

7

64

1.6

April 22. . .

41

12.2

10

9.4

4

12.0

10

10.0

20

6.4

5

5^0

65

10.2

May 6

29

11.7

10

9.5

6

11.4

9

9.0

99

6.5

5

3.7

134

4.8

May 20.. . .

7

12.0

11

10.5

3

11.0

10

10.0

0

10

2.3

1936

Feb. 19.. . .

12

14.0

12

10.0

9

13.2

9

9.0

3

5.0

5

4.5

24

8.9

Total

340

75

435

850

Extremes. .

14.0

9.2

13.5

9.0

8.2

3.4

Means.. . . .

12.5

10+

9.6

11.5

10-

9.3

6.5

5+

4.1

4.5

1936

Fishes collected by seine in upper Kensico.

Aug. 17.. . .

300

6.0

4

2.6

300

Aug. 23.. . .

64

6.3

4

3.6

64

Total fish examined — . — 1214

i The modes as used throughout this paper were picked from distribution curves of 1 cm. intervals.
Extremes measured to the nearest mm. All measurements are in cm. and refer to standard
lengths.

If these fishes are to be considered a recently landlocked form of the
sea run Pomolobus pseudoharengus, it should seem that the fishes are
merely acting according to their normal anadromous nature. There are,
however, several reasons to question this close kinship between the two on
the basis of observed habits. Spawning of the sea run Pomolobus occurs
early in the spring in local waters. At Swimming River, New Jersey, for
example, the spawning fish arrive from the last week in March to the middle
of April. The old fish disappear about the middle of May. These fish have

170

Zoological New York Zoological Society

[XXI: 13

passed three or more winters. The males range from 22.2 to 27.3 cm.
standard length and have a mode at about 24.2, while the females range
from 24.2 to 27.3 cm. with a mode at about 26.0. This is based on a sample
of 50 fish taken in 1923, in part discussed by Nichols and Breder (1926).
In tributaries of the Hudson River at Troy, New York, Greeley (1935)
observed spawning fish in the middle of May at a temperature of 52° F.
This is considerably north of our locality and the slight difference in time
is to be expected. The sizes of these fish are not given. Since the adult
Kensico Lake fish do not appear on the screens until February there is
certainly no downstream migration until much later and, supposedly, spawn-
ing is much later, on a considerably warmer temperature. Odell, at Seneca
Lake, obtained his eggs from late May to mid-August. Of course this may
be only the immediate influence of environment, but one would hardly ex-
pect such a difference in temperature thresholds. The whirling splash of the
spawning fish described by Greeley is apparently identical to that of the
landlocked form.

Text-figure 4.

A typical mass of Pomolobus removed from a screen at the Kensico outlet,

March 2, 1935.

Concerning the wide differences in numbers coming from the lake
(Table II and Text-fig. 3), there is one item that may be used to question
this downstream migration. It has long been known that periodically there
are heavy mortalities among these fishes in certain lakes and at such times
windrows of dead fish may collect along the lake shores. Such a condition
has never been noted in Kensico and it may be possible that due to its
form, rate of flow, or other factors, fishes in a weakened and dying condition,
which might otherwise strew the shore line, gravitate to the outlet screens.
Since there was a period of eight years between outbreaks, it more or less
suggests the population cycles common to so many organisms.

Another way to look at this would be to assume that only on an occa-
sional year did a very successful spawning take place and that the heavy
coating of the screens with fishes in 1934 represented the peak of the death

1936]

Breder & Nigrelli: Winter Movements of the Alewife

171

rate of some earlier but peculiarly successful year class. Since our fish
showed modes between 9 and 11 cm., it may be that most of them had
passed through four to six winters, which would place the successful hatch
between 1929 and 1931, as suggested by Table III. Supporting this are the
data from the screens in 1936. The mode and range of the females exceed
anything in 1935, whereas the young and males are typical of the preceding
year. If these females represent the residue from the previous year of that
abnormally successful year class, such figures would be expected. Since the
males of such fishes frequently mature a year earlier than the females, it
would not be surprising to find few left. It may be noted that these 1936
males cover the full spread of variation found in the entire 1935 collection.
The relatively few fishes in the 1936 collection can scarcely be considered
as invalidating these views, since the 1935 material is remarkably uniform
and the mode of the 1936 collection closely approximates the extreme size of
females for the previous year — a type of selection almost impossible to
make artificially, especially in view of the constant size of the males.

TABLE III.

Comparison of Seneca Lake Pomolobus with growth from indications of
scale marks on Kensico Lake specimens.

Annulus

SENECA LAKE1

KENSICO LAKE

Female

Male

Female2

Male3

1

5.5

5.5

4.3

4.7

2

12.1

11.4

6.7

7.4

3

12.5

12.1

7.8

8.2

4

13.4

13.9

9.1

9.2

5

13.9

14.0

10.8

10.5'

6

11.1

7

12. 04

1 Data from Odell (1934) ; mean values of his Table I.

2 Mean of 5 females.

3 Mean of 3 males.

4 Scale edge in early spring = actual standard length. All measurements in cm. standard lengths.

As noted in another connection, scale examination from these screen-
caught fish was not satisfactory, but all examined appeared to be in their
fourth year or over. Adventitious and, perhaps, spawning rings were
found so confusing that on the basis of our scant material it would be
unwise to attempt a close analysis of these data. There are, however, two
pertinent points that emerge. One is that there seemed to be no appear-
ance of second or third year fish on either a basis of modes or scale ex-
aminations. In the latter the mean value for the second winter was 6.7 cm.
female and 7.3 cm. male. Compared with Table II it is clear that this is close
to the extreme maximum of size for immature fish and just where there
were exceedingly few individuals taken — the upper end of the distribution
curve; likewise the third ring with mean values at 7.8 cm. female and 8.2
cm. male. It is not until the fourth annulus is reached with mean values at
9.1 cm. female and 9.3 cm. male that the minimum values of the mature
fish of Table II are reached. Although these scale data are fragmentary,
it is clear from the measurements of the fish that some classes, which lie
between the young of the year and the smallest adults obtained, are prac-
tically or completely absent. If all classes had been present on the screens,
with the annuli falling where they do, no such clear separation, as indicated

172

Zoologica: New York Zoological Society

[XXI: 13

by Table II, could be possible. Furthermore, immature, male and female
groups showed sharp peaks to their very symmetrical distribution curves.
The immature fish from the screens (Table II) show a single annulus near
the edge of the scale, as of course they should.

The other evident item is the relatively slower growth of our fish as
compared with Odell's. No matter what age be ascribed to our fishes, mak-
ing allowance for the misinterpretation of adventitious or spawning marks,
none reached the sizes he obtained. Our tentative values as compared with
his are given in Table III and in Text-fig. 6. These fish do not reach the
modal sizes his did when laying down their third winter ring. The Kensico
fish, on this basis, have an extremely rapid growth before the first winter
and then slow down gradually. Odell's material, on the other hand, shows
materially faster growth, including the second season. Lest it be thought
that we merely missed the first annulus, it may be pointed out that the
scales from the fish collected by net in August were carefully examined and
showed no ring, whereas those slightly larger, taken in the screens in spring,
showed one most clearly. This agreed in position with the first ring on the
scales of the larger fish. It is possible that this difference in growth rate
between Seneca and Kensico fishes may be associated with a temperature

Text-figure 5.

Sex ratio by dates of samples of Pomolobus pseudoharengus caught on the screens
at Kensico Reservoir during 1935 and one for 1936. The figures in the vertical
index represent the number of females divided by the number of males present.

1936]

Breder & Nigrelli: Winter Movements of the Alewife

173

differential, but on this point we have no comparative data. The water de-
livered to Kensico Reservoir is treated with alum and soda ash before it
reaches there and sometimes copper sulphate is used in the lake. The alum
is precipitated in the upper part of Kensico. See Hale and Dowd (1917)
for a discussion of the chemical and physical conditions in this lake, includ-
ing a discussion of the thermocline which they found forming at between 16
to 23 feet. Later data indicate that it is apt to form at between 20 and 30
feet. This chemical treatment should have a depressing effect on the plank-
ton and it may be that at times when the species is numerous a starvation
dwarfing ensues. It is pointed out in this connection that the fishes taken
on the screens were entirely empty of identifiable remains. However, there
may be a complete cessation of feeding at this time of year. The fish of the
year taken by other means were found to contain microcrustacea only.
Although no effort was made to study this feature in detail, it agrees well
with the data of Odell, for in August, the month of capture of these fish,
he reported the stomach contents as 90% microcrustacea. He does not give
the size of his fishes, but in a personal communication he stated that there
was no feeding differential to be noted with different size groups.

Regarding the physical conditions, it may be mentioned that tempera-
tures in Kensico ranged in 1935 from a mean of 63° in August to 34° in
February. Thus the young fish began to appear in the faucets just after
the summer peak of temperature was passed (Text-fig. 2) and the adults
reached their maximum congestion on the screens about coincident with the
lowest winter temperature. According to the Water Department records,
the pH has been slowly dropping since 1930 when the average was 6.9,
whereas in 1935 it had reached 6.7, which figure it had shown since 1933.
We are unprepared to draw any inferences from these data, if indeed they
have any important bearing on the fish under consideration.

The variation in sex ratio expressed in Table II and Text-fig. 5 is diffi-
cult of interpretation, but probably has to do with a differential move-
ment of the sexes in regard to their spawning beds. The females markedly
exceed the males in number in late February and again in late April. If
this species possesses a 1 to 1 sex ratio it would seem that the males are
more generally successful in avoiding disaster on the screens, which in turn
would indicate less of a tendency to drop downstream. Just why this should
be most marked in February and again in April is not at all clear at this
time, unless the females descend to the screens at a faster rate after they
are once started, and remain longer after the males move upstream. Be-
cause of the virtual failure of the fish to appear on the screens in 1936, a
continuance of this study was impossible. We question the validity of the
apparently more rapid growth of the males, as based on scale examination,
preferring at this time to consider it as probably due to the small number
of fish involved. Likewise, corrections for Lea’s phenomenon, as given by
Odell, would be pointless in connection with our material, because of the
relative coarseness inherent in these figures based on so few fishes.

Young fish of the year showed a modal length of 4 cm. in the latter part
of August, which reached 5 cm. before the cold weather checked increment,
as indicated in Table II. If it is true that a successful spawning occurs rela-
tively rarely, it may be that intermediate classes drop out and all those of
the immature class measured by us were doomed to an early demise. At
least it is difficult to conceive how otherwise to account for the condition
and sizes of fish caught on the screen in 1936.

The items here discussed necessarily lead one to infer that these violent
epidemics of Pomolobus in Kensico Lake belong with that great group of
periodical fluctuations of animal populations and that, other conditions re-
maining static, a return of large numbers should not be expected for some
little time, perhaps for a period somewhat approximating that of the last

174

Zoologica: New York Zoological Society

[XXI: 13

&.s . g g

^ 2

Mg'* §51

3 q3 .g <D £ M S

Nn > ^ S

$-8 ilia*

• ISS'ss

-G £ 01 CS ®

Eh H cs -£ <u-G
^ fil3 M

32|5sg
e “ S § 1“

<» OS ft *H $ «!

V. <D $ 0)

J S g g «j g
§■£«£ g*".

S c$£ cs «

■f sg^-Si

ft q3h,h

® r2 ai bfl n3

&< 03 $ ft o3 G
<h'SfS~^5S

£ §«• j§ cj
te §9$ g
g+Sfi^C
^ ^ c2 P qj
0 oi rtiiW

1936] Breder & Nigrelli: Winter Movements of the Alewife 175

interval. This species is preyed upon to a large extent by Cristivomer, as
indicated by Eaton (1928) and Odell (1934). The latter is present in Ken-
sico in small numbers. Forster (1935) has considered the lake trout, as
well as other species, as factors in the control of this herring in Kensico
Reservoir, a not unimportant item in the maintenance of a satisfactory
water supply of a great city. The probable value of such control measures
can hardly be estimated at this time.

References.

Eaton, E H.

1928. The Finger Lakes Fish Problem. In “A Biological Survey of the Oswego
River System.” Sup. to 11th Ann. Rept. N. Y. State Conserv. Dept., 1927:
40-66.

Forster, H.

1935. Fish in the Faucets : Some unusual aspects of water supply control. Bull.
N. Y. Zool. Soc. 38 (4) : 139-142.

Greeley, J. R.

1935. Fishes of the Watershed. In “A Biological Survey of the Mohawk-Hudson
Watershed.” Sup. to 24 th Ann. Rept. N. Y. State Conserv. Dept., 1934:
63-101.

Hale, F. E. and Dowd, J. E.

1917. Thermocline Studies at Kensico Reservoir. Journ. Industrial and Engi-
neering Chemistry, 9 (4) : 370-382.

Nichols, J. T. and Breder, C. M. Jr.

1926. The Marine Fishes of New York and Southern New England. Zoologica ,
9 (1) : 1-192.

Odell, T. T.

1934. The life history and ecological relationships of the alewife ( Pomoiobus
pseudoharengus [Wilson]) in Seneca Lake, New York. Trans. Amer. Fish.
Soc. 64: 118-124.

Beebe & Tee-Van: Systematic Notes on Tunas

177

14.

Systematic Notes on Bermudian and West Indian Tunas of the

Genera Parathunnus and Neothunnus \

*

William Beebe

&

John Tee-Van

Department of Tropical Research,

New York Zoological Society.

(Plates I-VII).

Outline.

Page

Introduction 177

Parathunnus atlanticus, Black-finned Tuna

Synonomy 178

Taxonomic Notes; Comparison of Recent Specimens and Original De-
scriptions 178

Description of Bermuda and West Indian Specimens 181

Range 184

Neothunnus argentivittatus, Yellow-finned Tuna

Synonomy 184

Taxonomic Notes 184

Comparison of Specimens with High and Low Soft Dorsal and Anal

Fin Lobes 187

Description of Bermuda and West Indian Specimens 190

Range 192

References 192

Introduction.

During 1934 and 1935 the game fishermen of Bermuda became inter-
ested in fishing for tuna and the result during those two years was the
capture of at least 400 fish, ranging from 10 ounces to 60 pounds. Some
fifty-three of these fish were brought to the Nonsuch laboratory during 1935
by various interested persons and the following notes have been made from
investigations on these specimens. We are especially indebted to Mr. Childs
Frick, Mr. and Mrs. E. J. Weir, Captain Christiansen, Major Davis, Mr.
Leslie Howard, Mr. Robert Blackman, Mrs. Willett, Colonel Edwin Chance,
Commander Landman and Mr. Beecham.

In addition to these Bermuda specimens we have been able to examine

1 Contribution from the Bermuda Biological Station for Research, Inc.

Contribution No. 501, Department of Tropical Research, New York Zoological Society.

178

Zoologica: New York Zoological Society

[XXI: 14

critically twenty-eight examples of these fishes at various localities in the
West Indies, while we were the guests of Dr. and Mrs. Henry D. Lloyd on
board their yacht Hardi Biaou. All of the observations were made upon
fresh specimens.

The name “tuna” in Bermuda is applied to three fish — the common
tunny, Thunnus thynnus (Linnaeus), a rare fish at Bermuda; the yellow-
finned tuna, Neothunnus argentivittatus (Cuvier and Valenciennes), known
from Bermuda by two specimens recorded in this paper; and the black-
finned tuna, Parathunnus atlanticus (Lesson). The present paper is con-
cerned with the two latter species.

The taxonomy of the tuna and other mackerel-like fish is in great con-
fusion. This is due both to the inadequacy of early descriptions, definitions
and illustrations, to ignorance of the changes occurring during growth, as
well as the difficulty of preserving large specimens for detailed laboratory
examination. Many of the older descriptions are meager and lacking in nec-
essary detail, and the result is that anyone searching the literature soon
finds that he has at hand a mass of data that cannot be coordinated or cor-
related. For that reason the specimens in this paper have been described in
detail, so that comparisons with fishes of other localities can be made more
easily.

BLACK-FINNED TUNA
Parathunnus Kishinouye 1923.

Parathunnus atlanticus (Lesson) 1830.

(Plates I & II).

Synonomy.

Thynnus atlanticus Lesson, 1830, In Duperrey, L. I., Voyage autour du monde —
sur la corvette . . . “La Coquille,” pendant . . . 1822-25. Paris, 1830, II, p.165.
(Trinidad Island, South Atlantic).

Thynnus balteatus Cuvier and Valenciennes, 1831, Hist. Nat. Poiss., VIII, p. 98
(136). (Trinidad Island, South Atlantic).

Parathunnus obesus (not of Lowe), Beebe and Tee-Van, Zoologica , X, 1928, p. 100
(Haiti).

Parathunnus rosengarteni Fowler, 1934, Proc. Acad. Nat. Sci. Phila., LXXXVI,
p. 354, 356, figs. 3, 4, 5, (Florida).

Parathunnus ambiguus Mowbray, 1935, Description of the Bermuda Large-eyed
Tuna Parathunnus ambiguus, n. sp., by Louis L. Mowbray, Curator of Gov-
ernment Aquarium, Bermuda, May 1935. (Three page, unpaged, privately
printed pamphlet). (Bermuda; type not designated).

Parathunnus atlanticus, Beebe and Hollister, 1935, Zoologica, XIX, 6, pp. 213-214
(Union Island, Grenadines, B. W. I.) ; Beebe, 1936, Royal Gazette, Hamilton,
Bermuda, Jan. 20, 1936, p. 10 (Bermuda).

Taxonomic Notes.

The Black-finned Tuna of the Western Atlantic has been named twice
during the last three years, and investigation of the literature shows, as
indicated in the preceding synonomy and the following notes, that at least
two additional older names have been applied to this fish.

Cuvier and Valenciennes in 1831 gave a short description of a new
species, Thynnus balteatus , stating “Cette espece ne nous est connue que par
un dessin fait par M. Lesson, vis-a-vis la Trinite, du Bresil, par les 20° de
latitude australe, d’apres un individu de vingt-huit pouces.” The descrip-

1936]

Beebe & Tee-Van: Systematic Notes on Tunas

179

tion is of a tuna with complete scalation and with a pectoral fin intermediate
in length between the short pectoral of Thunnus thynnus and the exception-
ally long pectoral fin of Germo alalunga.

However, Lesson a year earlier, in 1830, under the name of Thynnus
atlanticus, had already described this same fish, evidently from the identical

drawing, as he states in his description “Cette bonite a 28 pouces ”

“Nous primes cette bonite, non loin des Martin-Was et de Tile de la Trinite

” Lesson’s description is considerably longer and better than that

given by Cuvier and Valenciennes. Evidently the latter authors saw only
Lesson’s drawing of the fish, which, apparently, was never published, and
were not aware of his previous description.

The subsequent history of these two names is that they have usually
been synonymized under the long-finned albacore, Germo alalunga. The
latter name, as reference to Jordan and Evermann’s “Fishes of North and
Middle America,” will show (I, p. 871), has been a catch-all for at least five
authentic but poorly known species.

The identification of atlanticus and its synonym balteatus with Germo
alalunga is quite wrong and the former species should be placed in the
genus Parathunnus. As far as the pectoral is concerned, Germo alalunga
possesses an exceptionally long fin, measuring some 2.3 to 3 times in the
length of the fish — the tip of the fin usually extending beyond the posterior
tip of the soft dorsal. Atlanticus , from the original description and from
the description of the nominal balteatus, has the same measurement, 4.6 and
4 respectively, in the total length. These two proportions must be reduced
somewhat as they are in the total length instead of the standard. How-
ever, even after being reduced, the figures show a fin considerably shorter in
atlanticus than in alalunga, a condition that is specifically mentioned in the
description of balteatus. In the original description of atlanticus and in our
entire recent series the pectoral fin measures from 1 to 1.17 times in the
length of the head.

These figures for fins and other characters given in the descriptions of
atlanticus and balteatus agree with the accounts of Parathunnus rosen-
garteni and Parathunnus ambiguus and with recent Bermuda and West In-
dian specimens of Parathunnus examined by the present authors. In order
to correlate the various accounts the following table is included, listing the
proportions and counts of the various nominal forms and comparing them
with our series of Bermuda and West Indian specimens and our previously
reported Haitian specimen.

In reviewing the proportions and counts given under the various head-
ings it will be noticed that the number of soft dorsal fin rays recorded for
atlanticus in Lesson’s original description is 8, this author making no dis-
tinction between rays and spines. This count is in disagreement with the
descriptions of Fowler’s rosengarteni (IV, 13), Mowbray’s ambiguus (I, 12
to 13) and our Bermuda material, the latter being topotypical with Mow-
bray’s ambiguus.2 This discrepancy can be assigned either to the fact that
Lesson, and Cuvier and Valenciennes described the species from an imper-
fect drawing or to the difficulty of counting the rays, a task that is espe-
cially troublesome in these fishes, as the fins are encased in rather thick
skin and the rays and spines can be accurately counted only by removing
the skin with a scalpel.3

2 The difference in fin count has already been ignored by previous authors when atlanticus was
placed in the synonomy of Germo alalunga. The latter species has a fin count comparable to that
of the later recorded specimens of atlanticus.

3 In this respect it is of interest to note that until very recently the soft dorsal and anal
fins of these fishes have been consistently described as possessing but a single anterior spine.
Fowler (1928, Fishes of Oceania, pp. 132, 133 ; 1934, Proc. Acad. Nat. Sci. Phila., LXXXVI, p. 354)
has found a multiple number of spines in the anterior portion of the soft fins in a number of
species, a condition that apparently holds true for most of the scombroids.

Parathunnus atlanticus (Lesson). Proportions and counts of original descriptions and of a series of Bermuda and West

Indian recent specimens.

180

Zoologica: New York Zoological Society

[XXI: 14

atlanticus, 1
specimen from
Haiti (recorded
as obesus ) ,
Beebe and
Tee-Van, 1928

CO CO PI CO 32 00

CT ^ COCO IOCO • ■ ^ »— T°° (—T 00 CO ■ ■ I>J

co 1 1 1 1

atlanticus, 1 speci-
men from
Grenada, B.W.I.,
Beebe and
Tee-Van, 1936

53, i- >o

lO ^ Cl CO lO Th-'^ O CT t*- tO lO CO PI

TtH rr CO CO IO CO <M ^ (— T00 h-P1" °° ’—l -.‘P PN<NCOr-irHNCOTf(

oo tn HH

atlanticus, 2 speci-
mens from St.
Lucia, B.W.I.,
Beebe and
Tee-Van, 1936

to (M*5i.tO Ht* rnN to

^iHdWtO ' 33 oo PI CO ,±j CO 00 N>OCOP

' ^ * • CO i * * • • • .

O cq CO CO to CO N CO .. CT -OO CO r-H «IOIOCI(NCOHH(N(NCO

CO PI •— 1 b>- PH l-H t-H

tO hH

atlanticus, series
of 22 specimens,
Tobago, B.W.I.,
Beebe and
Tee-Van, 1936

383 to 570

15 to 22.5

3.7 to 4.05
3.25 to 3.6

5.2 to 5.5
2.9 to 3.3
2 . 4 to 2 . 5
3.06 to 3.2
13 to 14

III, ll2
8 to 9

III, ll2
7 to 8
3 to 3.7

.9 to 1.08

I, 31-32
5 to 6
15 to 16
2.4 to 2.44

2.8 to 2.9
3 to 3 . 1

1.66 to 1.76
1.46 to 1.53
2.15 to 2.5

3.1 to3.6

3.2 to 4.2

atlanticus, series
of 52 Bermuda
specimens. Beebe
and Tee-Van,
1935

263 to 583

10.5 to 23
3.12 to 3.76

3.1 to3.4
4.35 to 6

3 to 3 . 5

2.45 to 2.5
3 to 3.1

13 to 14
III or IV,

10 to 122
7 to 8
II to III,

10 to 122
7 to 8

3 . 3 to 3 . 7

1 to 1.1(1.37 in
263 mm. fish)
I, 32-33
5 to 6
15 to 18

2.4 to 2.5
2.75 to 2.9
2.95 to 3.2
1.67 to 1.76

1.45 to 1.5

2.2 to 2.45
2.7 to 3$

2.85 to 3.4

ambiguus,' origi-
nal description
of Mowbray,
1935

mm ^3

ja^'^3-^© >0 CO m

■ — ' - ' — 1 - i— i cC to

\NtB\NmCOCO o

O ^ . .

^ - PICT) h»IN bfi-P PN

. § . § ^ co ^ ^ H- dg-5 • • -

-'o-’ococo HH l-H (7T S

J* , -U CO 02

e— e— CO

rosengarteni,
original de-
scription of
Fowler, 1934

co r- 10 to

co io oo pi to co £4 74 o pi to

co oococo^cooipico . -oo co • co PI PI PI

’ — 1 PI „

i>- i— ii— i • h- i

balteatus, origi-
nal description
of Cuviei and
Valenciennes,
1831

*

O OO Tt* • CT ^

i— 1 P| •••••■ rH •

• • •

!!ii

= 'g§8

*

* t- to t'-

tO# CO CO T-^ CO CO CO

o oo co 'r • • • ooct 0000 ^ i-H uo pi pi pi

N ...

1 .a

n ia is
a> bn bo
rrt S S3

M 'a a s

® .2 rcp

* 0, S

^2 -a

■&^§ ^ "S ^ §

rj Oh oj (D q (t!

■a S

.a g.M

rS3

"So ^ 33 §3

^ 2 g*J

O rj - -

« cSg£

i— h _ <D O

c3 o3 ^ ^

5 .3??

O O S O
<D m «

Ph Ph O O

.a

-y o

_ a s3^ a

5- era «J3 5P.a P ‘

1513 s

S3 *S.«p

| I'd 13

M .••. O P •— OT ?-l <~!

. QJ 'a3 Ira £3 P ^ 'P* 13

>. CSI «J -g pj g

,»0000’-<C^a3

33 hj -t-3 -t-a -u _c3 _c3 _sd

1936]

Beebe & Tee-Van: Systematic Notes on Tunas

181

As a further indication of the identity of these fishes the following
excerpts of color descriptions are cited : In every description of the various
forms assigned to atlanticus there is mention of a lateral golden, orange or
coppery band. Thus Lesson states, “Une large band de cuivre rouge dore
vient de l’oeil, suit un instant la courbe de la ligne laterale, et va se perdre
sur les cotes du corps au point ou il s’amincet.” Cuvier and Valenciennes
write of balteatus, “. . . depuis le maxillaire superieur jusqu’a la queue,
une bande de coleur de cuivre dore/’ Fowler’s description of rosengarteni
contains the following, . . From behind the eye rather dark golden band
crosses corselet and continues along below lateral line to caudal pedun-
cle . . ; while Mowbray in the description of ambiguus states : “Colour,
blue black above, a bright blue stripe, with one of yellow below it, separates
the upper colour from the lower, . .

From the foregoing paragraphs it becomes evident that in the western
Atlantic there exists a species of tuna with complete scalation, possessing
a pectoral fin intermediate in length between that of Thunnus thynnus and
Germo alalunga, distinctive gill-raker count, and a conspicuous type of colora-
tion with a lateral golden or coppery band, and that the specific name
atlanticus can be removed from the synonomy of Germo alalunga and be
established as the proper name for this fish.4.

Parathunnus obesus Lowe, described from Madeira in 1839, is very close
to the present species. From the literature, especially Frade’s redescrip-
tions (1929, 1931), one of the few definite characters separating the two
species is the shape of the air-bladder, which in atlanticus is simple, quite
short and broader than long — totally different from the relatively complex
air-bladder shown by Frade for obesus (1931, p. 121). The disparity in
size of Frade’s fish and the western Atlantic series makes it difficult to make
adequate comparisons in other characters.

The specimen recorded by Beebe and Tee-Van from Port-au-Prince
Bay, Haiti (1928, p. 100) as Parathunnus obesus has been reexamined; it
is definitely Parathunnus atlanticus.

Description of Bermuda and West Indian Specimens.

The following description is based on previously published accounts
plus notes made on the specimens that we have examined. In addition, the
Table on Page 180 should be consulted for measurements and proportions
not otherwise mentioned.

Parathunnus atlanticus.

Body thickset, spindle-shaped. Smaller individuals slightly more com-
pressed laterally than the larger. Depth 3.12 to 4.05 in the length, the great-
est depth of body about half way from base to tip of pectoral fin. Caudal
peduncle depressed, with a triangular, rather short dermal keel on each side,
the length of the keel being about one and one-half to two times the diameter
of the eye. On the base of the caudal peduncle above and below the large keel
is a very short, oblique dermal keel. A small keel on the body above the
upper edge of the pectoral fin which allows the upper edge of the fin to lie
flat with the contour of the body.

Body completely scaled, the scales small, compact, absent on head, small-
est below, especially anteriorly — those between the pectoral and ventral fins
minute. Scale from midside at vertical of the origin of the second dorsal

4 If, very improbably, the Trinidad Island, South Atlantic fish atlanticus should prove to be
different from the West Indian, Florida and Bermuda Black-finned Tuna, Fowler’s rosengarteni
will be the proper term for the northern form.

182 Zoologica: Neiu York Zoological Society [XXI : 14

fin in a 443 mm. fish measuring 2.75 mm. high and 2.6 mm. long. Corselet
small, inconspicuous, largely over the base of the pectoral.

Lateral line complete, slightly wavy, not very high anteriorly, becom-
ing median in position only on the posterior caudal peduncle.

Head conical, 3.15 to 3.6 in standard length, the lower profile slightly
more convex than the upper. Snout not especially sharp, 2.9 to 3.25 in the
head length. Eye large, obliquely set in head, 4.35 to 6 in head length (5.3
to 6 in fish over 380 mm.); adipose eyelid very small; interorbital space
convex, 2.6 to 3.6.

Posterior nostril an elongate vertical slit, its length slightly less than
one-half the eye diameter. Anterior nostril very small, situated a consider-
able distance anterior to the posterior nostril and at the level of the upper
edge of the latter.

Mouth oblique, the mouth opening convex when viewed from the side,
the convexity being upward. Maxillary 2.4 to 2.5 in the head, its posterior
margin extending from anterior margin of eye to anterior margin of pupil,
the width of the posterior expansion being about 2 in the diameter of the
eye.

Teeth moderate in jaws, uniserial, simple, conic, 33 to 40 above on each
side (27, Fowler), 32 to 41 on each side below (32, Fowler). Vomer, pala-
tines and a patch on the tongue with finely granular teeth.

Gill-rakers slender but strong, 4 to 6 plus 15 to 18 on the first gill-arch.
In 58 specimens in which the gill-rakers were counted, 13, or 22 percent.,
had asymmetrical counts on the arches of the right and left sides. The
seven combinations of the gill-raker counts in these thirteen specimens are
as follows:

Gill-Rakers

Number of
specimens

4 + 16 — 5 + 16

1

5+15 — 5 + 16

3

5 + 16 — 6 + 17

3

5+17 — 6+16

2

5 + 18 — 6 + 18

1

6+15 — 6+16

2

6 + 16 — 6 + 17

1

The remaining forty-five fish with symmetrical gill-raker counts dis-
tribute themselves as in the following table:

Rakers on upper limb
of first gill-arch

5

6

Rakers on lower limb of first gill-arch

15

16

17

5

13

1

15

11

Dorsal fin XIII to XIV — III or IV, 10 to 13 (last ray a connected fin-
let) — VII to IX. Second dorsal spine highest, the first almost as high as
the second, the spines after the second becoming progressively shorter, first
abruptly and then gradually. Soft dorsal low. Anal fin II to III, 10 to 12
(last ray a connected finlet) — VII to VIII. Anal lobe similar to dorsal in
shape and size.

Coloration: This species in life is exceedingly brilliant. A '555 mm.

1936]

Beebe & Tee-Van: Systematic Notes on Tunas

183

specimen whose colors were recorded before the fish was removed from
water was described as follows: Dorsal surface and inner side of pectoral
fins jet black, the former bordered laterally with bright blue. A lateral
band from snout to tail of brilliantly iridescent shining gold, very wide and
including the outer side of the pectoral fin. Lower sides shining silver, with
a large oval patch on the sides between the pectoral and pelvic fins silvery
iridescent. Sides and ventral parts with eleven vertical bars and an equal
number of bands of spots alternating with the bars. Second dorsal lobe
with a tinge of yellow, but all other vertical fins black with a narrow white
border, especially marked on the finlets.

Mowbray’s description of his specimens is as follows: “Colour, blue
black above, a bright blue stripe, with one of yellow below it, separates
the upper colour from the lower, which is a silvery gray; region of the
ventrals and the belly, milky white: the spinous dorsal is dusky, the mem-
brane lighter than the spines.

“The soft dorsal and anal are dusky with a silvery lustre, the finlets
are dusky, with a trace of yellow: this is more pronounced in some speci-
mens : the pectorals are black, the base outwardly is washed with silver.

“The ventrals are milky white outwardly when closed, and dusky in-
wardly with a metallic lustre when opened: the caudal is dusky, the sides
of the belly show white spots which appear as reticulations: this I believe
to be seasonal, as I noticed them only in the winter months ; I do not know
if they disappear in G. alalunga or not.”

There is considerable variation in color and in some specimens the
lower side of the pectoral fin is silvery, the fin not being included in the
golden lateral band. At death the golden band fades rapidly and usually
only traces of it remain.

The finlets and the vertical fins are occasionally greenish, as stated by
Mowbray, or even yellowish, but the fins are never as brilliantly colored
as in Neothunnus argentivittatus. These two species need not be confused
on this score, as the gill-rakers give an absolute differentiation in question-
able specimens.

The colors of the smallest known specimen of Parathunnus atlanticus
(263 mm.) were noted as follows: Bluish black above shading to grayish
silvery on the sides and whitish below. Middle of sides somewhat lead-
colored and with a series of five to six vertical lighter colored bands on the
lower sides, extending upward as far as the median line of the body. These
bands, posteriorily, are changed into series of horizontal, elongate spots
forming broken vertical bands (see PI. II, Fig. 4). Spinous dorsal, soft
dorsal and dorsal finlets dusky, the latter with a pale upper edge. Pectoral
blue-black above, grayish silvery below, darker toward the tip. Caudal fin
dusky. Lower surface of pelvic fin white, upper surface dusky with a yel-
lowish tinge, the rays brownish.

Mowbray states (1935) that “The white spots or reticulations sup-
posed to occur only in young specimens of Germo are not present in the
smaller specimens of this species.” Our smallest fish, however, as described
above, did have such spots and we have seen the pattern in other fish of
this species in the West Indies. (See PI. II, Fig. 4 of a Bermuda fish, and
PI. I, Fig. 2 of a fish from St. Lucia, British West Indies).

In other tunas this pattern is a juvenile one, and it has been previously
reported from Thunnus thynnus 5 6 7 8, Thunnus orientalist, Neothunnus argen-
tivittatus1, Neothunnus macroyterus 8 and Parathunnus mebachi9.

5 Frade, 1929, Bull. Soc. Portugaise des Sciences Nat., X, No. 20, 236.

6 Kishinouye, 1923, Journ. Coll, of Agriculture, Imp. Univ. of Toyko, VIII, No. 3, 438-439.

7 Cuvier and Valenciennes, 1831, Hist. Nat. Poiss., VIII, 98.

8 Kishinouye, 1923, Journ. Coll, of Agriculture, Imp. Univ. of Tokyo, VIII, No. 3, 447.

9 Kishinouye, 1923, Journ. Coll, of Agriculture, Imp. Univ. of Tokyo, VIII, No. 3, 444.

184

Zoologica: Neiv York Zoological Society

[XXI: 14

Size and Weight: This species, according to Mowbray, grows to a
weight of 60 pounds. The sizes and weights of a few individuals measured
by us are as follows: —

Length

Millimeters

Inches

Weight

Pounds

263

10.3

13 ounces

374

14.8

2.25

443

17.5

4.5

475

18.7

6.25

480

18.9

6.5

495

19.5

7.

530

21

6.75

560

22

11

583

23

12.5

660

26

15.5

Range.

Known from the coast of Florida, Bermuda, Haiti, Martinique, the fol-
lowing islands in the British West Indies: St. Lucia, Union, Grenada and.
Tobago, and from Trinidad Island, Brazil.

YELLOW-FINNED TUNA.

Neothunnus Kishinouye 1923.

Neothunnus argentivittatus (Cuvier and Valenciennes) 1831.
(Plates III-VII).

Synonomy.

? Scomber albacares, Bonnaterre, 1788, Tableau Encyc. Meth., Ichth., 140
(Madeira; after a drawing by Sloane, 1707).

? Scomber sloanei, Cuvier and Valenciennes, 1831, Hist. Nat. Poiss., VIII, 107
(Madeira; after a drawing by Sloane, 1707).

Thynnus argentivittatus , Cuvier and Valenciennes, 1831, Hist. Nat. Poiss., VIII,
97 (Atlantic and Pacific Oceans).

Thynnus albacora, Lowe, 1839, Proc. Zool. Soc. London, VII, 77 (Madeira).
Thunnus allisoni, Mowbray, 1920, Copeia, 78, Feb. 11, 1920, 9-10. fig. (Florida).
Neothunnus albacora, Frade, Rapp. Cons. Explor. Mer., 1931, 70, 123.

Neothunnus argentivittatus, Beebe, Royal Gazette, Hamilton, Bermuda, Jan. 20,
1936, 10 (Bermuda).

Taxonomic Notes.

In 1707 Hans Sloane in “A Voyage to the Islands Madera, Barbados,

Nieves, S. Christophers and Jamaica, with the Natural History

of those Islands,” Vol. I, p. 11, described a tuna from Madeira. His de-
scription of the fish is as follows :

“The Sea hereabout is very well provided with Albacores, or Thynni,
whose Description follows.

“The Fish was Five Foot long from the end of the Chaps to that of the
Tail, the Body was of the make and shape of a Mackarel, being roundish

1936]

Beebe & Tee-Van: Systematic Notes on Tunas

185

or torose, covered all over with small Scales, White in some places, and
Darker coulour’d in others, there was a Line run along each side. The
coverings of the Gills of each side were made of two large and broad Bones
covered with a shining Skin, the Jaws were about Six Inches long, having a
single row of short, strong, sharp Teeth in them, and were pointed. The
Eyes were large, and the Gills very numerous, behind which were a small
pair of Fins. Post anum was a Foot long Fin, about Three Inches broad at
bottom, and Tapering to the end. It had another on its Back answering
that on the Belly, and from these were, small Pinnulae at every Two Inches
distance to the Forked Tail, which was like a New Moon falcated, before
which on the Line of the two sides was a membranous thick horny Substance,
made up of the Fishes Skin, stood out about three-quarters of an Inch where
it was highest, something like a Fin. It was about Three Foot Circum-
ference a little beyond the head, where it was thickest. The Eye was about
an Inch and a half Diameter. The Figure of this Fish is here added, Tab. 1,
Fig. 1 taken from a dried Fish, where everything was perfect save the first
Fin on the Back, which I suppose was accidentally rub’d off.

“It is frequently taken by Sailers with Fisgigs or White Cloath, made
like Flying-fish, and put to a Hook and Line for a Bait; the Flesh is
coloured and Tasts as the Tunny of the Mediterranean, from whence I am
apt to believe it the same Fish. It is to be found not only about Spain,
and in the way to the West-Indies; but in the South-Seas about Guayaquil,
and between Japan and New-Spain every where.”

From this account and figure (See PI. VII, Fig. 13), Bonnaterre, in
1788, described Scomber albacares, and forty-three years later (1831) from
the same data Cuvier and Valenciennes described Scomber sloanei.

Bonnaterre’s description of albacares is as follows: “L’Albacore 9. S.
Albacares S. pinnulis plurimus : aristes duabus supra opercula, membrana
lucida tectis.

“Plusieurs fausses nageoires : deux aretes couvertes d’une peau luisante
au dessus des opercules.

“Le corps est rond & entierement couvert de petites ecailles: les
machoires, dont la longeur est d’environ six pouces, sont armees d’une seule
rangee de dents courtes & tres-aigues. La nageoires du dos correspond a
celles du ventre; elle est accompagnee de plusieurs fausses nageoires,
eloignees les unes des autres, d’environ deux pouces ; la nageoire de Fanus a
un pied de long, fui trois pouces de large: elle se termine en pointe; celle
de la queue est echancree en croissant; les parties laterales de la queue
forment, de part & d’autre, une saillie en carene, qui a trois quarts de pouce
d’elevation. Quelqus parties du corps sont blanches; les autres sont d’une
couleur foncee. Ce poisson a trois pieds de circonference dans sa plus grande
epoisseur, & cinq pieds de longueur. Sloane Hist. tho. Jamaic. vol. 2, p. 11.
La Jamaique.”

The following is Cuvier and Valenciennes’ description of Scomber
sloanei :

L’Auxide de Sloane.

( Scomber Sloanei, nob.)

“L 'albacore de Sloane, si Ton peut s’en rapporter a une figure grossiere,
comme toutes celles qu’a donnees cet auteur, semble devoir appartenir a
ces thons a dorsal ecartees.

“Son museau est court ; sa bouche, peu fendue, n’a que de petites dents.
Sa premiere dorsale parait avoir peu de rayons, et etre separee par un
grand intervalle de la seconde. Ses pectorales sont courtes. II a huit
fausses nageoires en dessus, et sept en dessous de la queue; mais ce qui

186

Zoologica: New York Zoological Society

[XXI :14

parait devoir lui former un caracteres specifique, c’est que sa seconde
dorsale et son anale sont plus hautes et plus pointues a proportion, que dans
aucune autre espece; elles ont en hauteur plus du cinquieme de la longueur
totalle. Nous n’avons rien vu qui ressemble a cette figure.”

The identity of Sloane’s fish and consequently of the names Scomber
albacares and Scomber sloanei, has always been questionable, and while
Sloane’s fish was, most probably, the long-finned, yellow-finned tuna here
discussed, we cannot be certain of this because of discrepancies between
modern specimens and Sloane’s description, especially those relating to the
length of the pectoral fins.

Sloane does not state the color of the fins of his fish. If he had, there
would be considerable less difficulty in assigning a position to his tuna. The
condition of the dorsal fin as shown in his figure, with a long space between
the spinous and soft dorsal, which caused Cuvier and Valenciennes to
assign a common name of “L’Auxide de Sloane” to this fish, may be ex-
plained by the fact that Sloane specifically states that the fin was damaged ;
this major discrepancy can thus be removed from the description of the
fish. However, the one character that stands out in marked contrast to
our knowledge of recent specimens of yellow-finned tuna is the length as
shown in Sloane’s figure of the pectoral fin. These fins are very much
shorter than those of recent specimens, and Sloane definitely states of these
fins “. . . and the Gills very numerous, behind which were a small pair of
Fins.”

We are thus left with the uncertainty as to the status of Sloane’s speci-
mens and the associated names, and while we might assume that the illus-
tration was carelessly drawn with fins of wrong length, yet Sloane’s definite
statement in the text as to the size of the fins prevents our using either
of the earlier names as the correct one for the yellow-finned tuna of the
Atlantic.

It is of interest that modern authors such as Jordan and Evermann
1896, Jordan and Evermann 1926, Jordan, Evermann and Clark 1930, and
Fowler 1928, have repeatedly given the type locality of Scomber albacares
Bonnaterre as Jamaica. This is an error and the fish, according to Sloane’s
volume, was originally taken at Madeira.

The earliest valid description of the Atlantic yellow-finned tuna is that
of Cuvier and Valenciennes, who in 1831, described Thynnus argentivit-
tatus from a drawing of an Atlantic specimen taken by Quoy and Gaimard.
The fish they described was a completely scaled tuna with pectoral fin inter-
mediate in length between that of the common tuna and the long-finned
albacore, Germo alalunga, and with a conspicuous coloration consisting of
bright yellow fins and a combination of alternate white vertical bands and
groups of spots on a slightly darker belly.

Lowe, in a brief description published in 1839, described Thynnus
albacora from Madeira. This form was forgotten, as far as the literature is
concerned, until the last few years, during which a number of observers
have described it from various localities in the eastern Atlantic, the prin-
cipal accounts being by Frade. Lowe recognized that the fish mentioned
by Sloane was similar to his as he synonymized Sloane’s account, and Cuvier
and Valenciennes’ Scomber Sloanei under Thynnus Albacora.

Mowbray in 1920 described Thunnus allisoni from Florida. This is the
typical long-finned and yellow-colored large adult form of the species.

Study of the original descriptions of argentivittatus Cuvier and Valen-
ciennes, albacora Lowe, the redescriptions of albacora by Frade, the original
description of allisoni Mowbray, a single specimen taken by us in Bermuda,
another Bermuda specimen taken in Bermuda by Mrs. W. B. Holler, and
five fish taken by us in the West Indies, lead us to believe that all of these

1936]

Beebe & Tee-Van: Systematic Notes on Tunas

187

relate to one species for which the name argentivittatus, Cuvier and Valen-
ciennes, 1831, has priority.

Cuvier and Valenciennes’ description of argentivittatus and the smaller
Bermuda and West Indian specimens agree in possessing low dorsal and
anal fin lobes, in contradistinction to the other original descriptions and our
larger West Indian specimens. This condition is discussed on later pages.

The detailed measurements and proportions of the various original de-
scriptions plus those of recent specimens, demonstrating the similarities,
are shown in the table on Page 188.

Comparison of Specimens with High and with Low
Soft Dorsal and Anal Fin Lobes.

The Atlantic species of Neothunnus , N. argentivittatus, (including
N. albacora ) and its Pacific congener Neothunnus macropterus, agree in
possessing forms with short soft dorsal and anal fin lobes and others in
which these fins are considerably produced, a fact that has caused discussion
as to the specific validity of these types.

The preponderance of opinion, so far, both as to the Atlantic and
Pacific species, has been that the high and the low-finned forms are the
same, although Fowler (Proc. Acad. Nat. Sci. Phila., LXXXV, 1933, p. 163)
evidently disagrees with this and has erected the genus Semathunnus for
those with high soft dorsal and anal fin lobes, leaving the low-finned fish in
Neothunnus.

Our recent observations of argentivittatus at Bermuda and in the West
Indies, and its Pacific relative macropterus in the Gulf of California, coupled
with examination of the literature has convinced us that the forms are the
same species, possessing short soft dorsal and anal fin lobes while small,
as Frade (1929) has already stated, the lobes becoming produced with
growth. In the larger specimens there is a tendency for the fins to vary in
length, some specimens having relatively long fins while in others they are
relatively shorter. The fins of the same fish are not always equal in length
(See Table II, last column) and Herre (1936) states of hundreds of Pacific
N. macropterus that he examined: “Sometimes the anal, more rarely the
dorsal, lobe would be elongated and the other remain comparatively short.”

As far as the literature is concerned the following quotations are per-
tinent. Cunningham (Proc. Zool. Soc. London, 1910, p. 106), writing of
albacora (which we here synonymize with argentivittatus ) and comparing
it with Parathunnus obesus and Germo alalunga states: “In one form,
namely the common albacore of the inhabitants of St. Helena — there are
considerable changes in the course of growth: but these changes do not
lead to any approximation to the other forms but rather to a greater de-
velopment of the special features: in a small specimen somewhat less than
3 feet in length the second dorsal and the first ventral [anal] fins were
scarcely higher than in the other two forms, while in the other specimens
the great vertical elongation of the fins is very characteristic.” (See PI. Ill,
Fig. 5 of a Cunningham specimen).

Frade in speaking of his Atlantic specimens says: “II est interessant
de signaler que, comme pour N. macropterus du Pacifique, il existe pour la
meme taille deux types de N. albacora; l’un a 2e dorsale et anales longues,
correspondant a N. macropterus forma itosibi et l’autre a 2e dorsale et anales
courtes, correspondant a N. macropterus forma macropterus .”

Kishinouye, in his excellent study of Japanese Neothunnus macropterus,
says: “. . . The second dorsal and anal are much elongated, especially in the
variety named itoshibi or gesunaga, the tips of these fins are whitish and
reach to the base of the caudal. So far as I have examined there is no

TABLE II.

Neothunnus argentivittatus (Cuvier and Valenciennes). Proportions and counts of original descriptions and of a series

of recent Bermuda and West Indian specimens.

188

Zoologica: New York Zoological Society

[XXI: 14

-a

^ cs *3 CQ

;S a©j

I’ImN’s

q ® pq e3

geo

hS

* g & -

§ P © a

^ <D 03

e “ P3 «

> .a a

;.S“8£

■S S =

Se»>,

Sjf'J

S o 2“

.s

m e <-<

•co.

. a 03

’§‘g43

•I 8 I

© o

™ a co

Sjs

e .2* g
si £

tO+3 ^ W

-2 o t, S

:1 J-£ §
1-al-l

©> b a

•g £

O as O ®

,-H CS OS t O OS

N5DIOM _ O „ O

• • • i— ( T3 ►— ( rO

<M <M HH O ►— t O

ONlO

OlMNiOM*

moo)

©* ©j <m’

§ ^
^ goo

Nnco-HHNIN

ia io

co ...

CO CO so

© O o CP
t^T^^goO'S-goO
®OChh 3 o 3
CO ©i ©i ’"V§ ^

H & > fe.

■CO lO OO lO in
■ 05 1— I !>• CO CO CO

3 "©

i— i -q ' —

.. © ~ ©

a t— i a

©. HH Q.

H3
© _
cs Os

03

3 -

«i3

l-3

s°

o lO 2 S

oo <n r2

§ ©

os I© >© CO .2 3
OS CO oo ■ © „

OS -CO .© ©

CO CIS gt»

<D

CO

S

-O ^ n3

llslis

■ g-§ © £ -© O

: 3 s-*3 fc©^

•-2 O.00^. ft00

ll . - ©

lO CO .-I QO
©i co ^ <H t-h

<NHHN«0

(Ncocot-Jh

: ©

a.

-S^o

* c<3 *-=<

•iO g °©

: -*-= .E*11© ^
^ -© S

■§*a-§

S3

*3 « 9

'-gg *■§£

§^§|>s§f“ifc‘

o KPS +3 '■ — «• © s — ' o

rjn

Ph

rq o t>0-

1.2 8

•S.S

3 3 jg £

Jill

-o

, 3

JL2-o

• <— i • ^3

:-&33-S

• fl bC+3 ©

; M g

• ’O TJ 3 Pi

• cs 2

. J3 Pi CO

_ -9 -9 .2 i&cc
§l|-g ^73
8 & & §«§

2 §•;=!£;£ -S3 O ! ®4« J

£ Jf g- £ &<& s§s1 s“ „

^_asas©o©oO’— i<Mcs©

S V V2 9 9 9 g bp bp bp ra |

SR '”■* • *2 2 2 2 2 *o *<d *c? § «-2
PlhOO^wmmmWWWp

-■SJ-il

gfusis

© © 03 _, -© 5

^ ~ 03 ^ ^ -©

^9-2 1^

‘m no

* Proportions marked with an asterisk are taken in the total length of the fish, consequently these proportions must be reduced somewhat to conform with
the other figures.

1936]

Beebe & Tee-Van: Systematic Notes on Tunas

189

marked difference in anatomical structure between the long-finned variety
and the ordinary form, except in the length of the second dorsal and anal
fins.”

Combining the facts in these quotations with our own observations
we come to the following conclusions: There is evident agreement that the
dorsal and anal fins are low in the smaller individuals and grow longer in the
larger : There is less agreement as to what happens in the fins of the fishes
of larger size. Both Frade and Kishinouye state that there are two forms,
a short-finned and a long-finned, and the former queries (1929, p. 238) as
to the possibilities of these being different sexes. It is our belief that in-
termediate forms will be found completely connecting the two groups.

Two references relating to the closely related Pacific Neothunnus
macropterus are in accord with this statement. Thus Herre (1936, pp. 106,
107) says: “I have examined many hundreds of this fine fish, alive, just
taken from the water, and after preservation in ice, in the Philippines, and
have also examined vast numbers of them in Japan. I agree with the late
Dr. Kishinouye that there is no difference between those with the anterior
dorsal and anal lobes of ordinary height and those in which the lobes, par-
ticularly the anal, extend to the base of the caudal or beyond. In a lot of
several hundred caught in a fish corral at one time, there would be every
graduation in the length of the dorsal and anal lobes. Sometimes, the anal,
more rarely the dorsal, lobe would be elongated and the other remain com-
paratively short. My observations have been made upon specimens rang-
ing in weight from 30 to 40 pounds up to those weighing nearly 400 pounds,
their lengths varying from two-thirds of a meter to nearly three meters.”

Walford (1936) writing of the Pacific yellow-finned tuna examined
by him in canneries says: “. . . Among several hundred specimens that I
examined there, the dorsal and anal fins were of all lengths, intergrading
to such an extent that it is impossible to separate them into two groups.
In general, the largest, consequently the oldest fish had the longest fins.
Unless there are some other characters to separate the two — and no one has
as yet been able to find any — it thus looks as if the Allison tunas of the
Pacific coast are merely old specimens of the yellowfin.”

The following table correlates the height of the dorsal and anal fin
lobes in a series of Atlantic yellow-finned tuna of progressively larger size.

Locality

Length, mm.

Percent of standard
length

2nd dorsal
lobe

Anal lobe

Bermuda, Dept. Trop. Res. No. 25,206 . .
St. Lucia, B.W.I., Dept. Trop. Res. No.

577

10.8%

10.1%

24,682.

St. Lucia, B.W.I., Dept. Trop. Res. No.

645

12.5%

broken

24,683.

690

13.3%

14.3%

St. Helena, Cunningham, 1910 (P.Z.S.)
(Photograph)

930 app.
Nearly three
feet long

12.5%

7

Bermuda, (Holler specimen)

St. Lucia, B.W.I., Dept. Trop. Res. No.

1,220

21%

22%

24,690

St. Lucia, B.W.I., Dept. Trop. Res. No.

1,360

38.5%

48%

24,689

1,450

36%

48%

Portugal, Frade, 1929

1,560

36%

39%

Canary Islands, Frade, 1931

990 to 1,740

36%

34.5%

Canary Islands, Frade, 1931

App. 1,600

35%

38%

190

Zoologica: New York Zoological Society

[XXI :14

It seems evident from the above tables and the preceding notes that the
various nominal forms of the yellow-finned tuna belong to the same species,
and that the forms typified by the large Allison’s tuna represent but large-
finned specimens of the smaller shorter-finned individuals. It is also evi-
dent that the genus Semathunnus, as far as Atlantic argentivittatus and
Pacific macropterus are concerned, is untenable.

Description of Bermuda and West Indian Specimens.

The following description is of the first specimen taken at Bermuda:
°\ No. 25,206, Department of Tropical Research, New York Zoological So-
ciety; taken one mile south of Nonsuch Island, Bermuda, with rod and
line by Mr. Beecham, July 27, 1935. Standard length 555 mm.; weight 8
pounds.

Field Characters : Robust, large, spindle-shaped, large-eyed fish with
a fleshy keel on each side of the caudal peduncle and a series of finlets fol-
lowing the dorsal and anal fins; pectoral fins rather long, extending almost
to the posterior end of the soft dorsal fin. Twenty-one gill-rakers on the
lower limb of each gill arch. Dark blue above, white below; a series of
alternate vertical narrow whitish bands and groups of spots on the lower
sides. Dorsal fin and dorsal and anal finlets bright yellow.

Description : Body spindle-shaped, the depth 3.6 in the length, some-
what compressed, the greatest width of the body being 5.55 in the length;
upper profile slightly more arched than the lower; caudal peduncle with a
large, rather wide, dermal lateral keel and with two very small keels at the
posterior end of the large keel — the latter is widest in its central portion
tapering equally anteriorly and posteriorly. A groove along the sides
into which fits the upper border of the pectoral fin.

Body covered with very small scales, smaller on the belly and larger
immediately above the pectoral fin; corselet indistinct.

Lateral line indistinct anteriorly and slightly undulating, descending
at the level of the posterior part of the spinous dorsal, posterior portion
straight.

Head 3.34 in length; opercle gently rounded posteriorly; preopercle
with its posterior limb not quite vertical and slightly concave.

Eye large, 5.73 in head and 1.8 in the snout, its vertical diameter
slightly greater than the horizontal one. Adipose eyelids very slightly de-
veloped.

Anterior nostril minute, situated slightly nearer eye than tip of snout ;
posterior nostril a vertical slit, half an eye diameter in length and placed
about one-third an eye diameter in front of eye.

Maxillary extending almost to anterior margin of pupil, its inferior mar-
gin rather abruptly downturned posteriorly. 31 to 34 teeth in each side
of each jaw; the teeth small, conical and turned slightly inward. A granular
patch of teeth on vomer and palatines.

Gill-rakers 10 or 11 plus 21 on first gill arch, the length of a median
raker of the lower half of the arch being about % the diameter of the eye.

Dorsal fin XIV-IV, 11-IX, folding into a groove; first spine highest,
the upper margin of the fin descending rather abruptly, the fifth spine only
one-half as high as the first, the last spine a mere nubbin of 3 mm. Space
between spinous and soft dorsal about equal to interspace between spines
so that the two fins are practically continuous. Soft dorsal fin rather low,
the height of the fin being 2.7 in the head and 2.6 in the pectoral length;
fin followed by 8 free finlets.

1936]

Beebe & Tee-Van: Systematic Notes on Tunas

191

Caudal fin with its posterior margin crescent-shaped, the expanse of
the fin being slightly greater than the length of the head.

Anal fin originating under the next to the last ray of the soft dorsal,
its lobe similar in shape to that of the dorsal; III, 11-IX. The first finlet of
the nine is connected to the end of the anal fin.

Pectoral fin originating under origin of the spinous dorsal, the tip of
the fin extending to the vertical of the 6th to 7th ray of the soft dorsal.
Upper and lower margins of the pectoral fin almost parallel for the anterior
half of the fin.

Pelvic fin originating under the anterior third of the pectoral fin base.

Measurements :

Total length : 577 mm.

Standard length : 555 mm.

Depth: 154 mm. (3.6).

Width of body: 100 mm.

Head: 166 mm. (3.34).

Eye: 29 mm. Greatest vertical diameter of eye: 32 mm.

Interorbital: 61 mm.

Snout: 52 mm.

Maxillary: 63 mm. Width of posterior expansion of maxillary: 13.5 mm.

Pectoral length: 161 mm.

Pelvic fin length: 60 mm.

Second dorsal fin lobe height: 62 mm.

Anal fin lobe height : 62 mm.

Height 1st dorsal spine: 63 mm.

Expanse of caudal fin : 185 mm.

Length of central caudal keel: 63 mm.

Greatest width across caudal keels : 49 mm.

Snout to middle of eye: 68 mm.

Snout to origin of 1st dorsal fin : 179 mm.

Snout to origin of pelvic fin: 188 mm.

Snout to origin of pectoral fin : 167 mm.

Snout to anus: 336 mm.

Counts :

Pectoral fin rays: 34.

Pelvic fin rays : 1,5.

Dorsal fin: XIV-IV, 11-IX.

Anal fin: III, 11-IX, first finlet connected to end of anal fin.

Gill-rakers: Right side 11 plus 21; left side 10 plus 21.

Teeth: Upper right 32; upper legt 33; lower right 31; lower left 34.

Coloration : Above dark metallic blue becoming silvery below, with
purple and lilac iridescence in an indefinite band on each side of the white
belly from the gills to the anal fin. Lower sides below the pectoral fin with
eight more or less vertical whitish bars, alternating with similar vertical
lines of small whitish spots, all on a dark grayish background. Pectoral
fin silvery below, dark blue above ; dorsal spines surrounded by narrow zone
of yellow, which is less bright on the membrane of the fin; second dorsal
orange-yellow, darker anteriorly, with a narrow black posterior edging and
a small white tip to the fin ; dorsal finlets bright yellow with dark margins ;
anal fin silvery at base, the remainder orange-yellow ; anal finlets similar to
dorsal but with considerable white posteriorly; pelvic fins dead white below;
the rays when spread open are orange.

Variation : This species except for fin-lengths, which are discussed on
preceding pages, does not differ much with age. The variations in propor-

192 Zoologica: New York Zoological Society [XXI :14

tions are shown on the table recording the various specimens and original
descriptions.

The white spots and bars on the lower sides of smaller fish are not
visible on the larger individuals that we have seen.

Frade in his 1929 paper describes the coloration of his large fish as
follows : “Coloration generale des specimens adultes : toute la region

dorsal au-dessus du bord superieur de l’oeil et de la ligne laterale, et la
queue, d’un bleu de Prusse opaque, plus clair vers la ligne laterale; regions
operculaire, mandibulaire et ventrale d’un blanc metallique, ligerement car-
mine, avec des tonalites verdatres surtout vers la region anale. Iere dorsale
avec les rayons cendres jaunatres et la membrane partageant de la meme
coleur mais plus foncee; 2eme dorsale et anale jaune fonce, s’attenuant vers
Textremitie, l’anale avec des tons argentes sur la base. Pectorale verdatre,
la pointe noir bleuatre; ventrale nacree; caudal noir bleuatre, jaunissant
vers les extremities. Pinnules jaune vif, liserees de noir.”

A 1,450 mm. fish taken by us at Castries, St. Lucia, B.W.I., was re-
corded as follows: Dorsal surface bluish black changing into iridescent
steel blue along the sides and silvery bluish white below; from anal fin to
caudal the body below is silvery white. First dorsal with spines and most
of the distal halves of the webs olive yellow. Second dorsal lobe yellow on
distal third and yellowish to base of each edge. Anal lobe same as second
dorsal. Caudal fin dark brown with yellow-white toward tips. Finlets bright
chrome yellow with strong narrow black margins, especially prominent on
the posterior edges.

In the smaller specimens the entire dorsal and anal fin lobes are bril-
liant yellow, and occasionally there is a golden band along the side of the
fish from the snout through the eye above and over the pectoral fin. The
smaller specimens also are characterized by the series of vertical alternate
bars and groups of spots.

Size and Weight: The sizes and weights of a series of individuals
of this species are as follows:

Millimeters

Length

Inches

Weight

Pounds

555

22

8

645

25.4

12

690

27

16

1220 (total)

48 (total)

45

1450

57

140

1750 (total)?

69 (total)?

143

Range.

Known from Scotland (?), Portugal, Angola, Canary Islands, Madeira,
Bermuda, Florida, Martinique, St. Lucia, and St. Helena.

References.

Beebe, W., 1936 Royal Gazette, Hamilton, Bermuda. Jan. 20, 1936, p. 10.

Beebe, W. and G. Hollister, 1935, The Fishes of Union Island, Grenadines,
British West Indies, Zoologica, XIX, 6, pp. 209-224.

Beebe, W. and J. Tee-Van, 1928, The Fishes of Port-au-Prince Bay, Haiti, Zoo-
logica, X, 1, 1-279.

Bonnaterre, J. P., 1788, Tableau encyc. meth., Ichthy. vi, 1-215.

1936]

Beebe & Tee-Van: Systematic Notes on Tunas

193

Cunningham, J. T., 1910, On the Marine Fishes and Invertebrates of St. Helena,
Proc. Zool. Soc. London , 1910, 1, 86-119.

Cuvier and Valenciennes, 1831, Hist. Nat. Poissons , VIII.

Fowler, H. W., 1928, The Fishes of Oceania, Mem. Bishop Mus., Honolulu, X,
1928, pp. 1-540.

Fowler, H. W., 1934, The Buckler Dory and Descriptions of Three New Fishes
from off New Jersey and Florida, Proc. Acad. Nat. Sci. Phila., LXXXVI,
353-361.

Frade, F., 1929, Sur Quelques Thons peu Connus de L’Atlantique, Bull. Soc.
Portugaise des Sciences Naturelles, X, No. 20, 227-243.

Frade, F., 1931, Donnees Biometriques sur trois especes de thon de I’Atlantique
Oriental. Rapp. Cons. Explor. Mer., Copenhagen, 70, 117-126.

Herre, A. W., 1936, Fishes of the Crane Pacific Expedition, Zool. Ser., Field
Mus. Nat. Hist., XXI, 1-472.

Jordan and Evermann, 1896, The Fishes of North and Middle America, I, 1-1240.

Jordan and Evermann, 1926, A Review of the Giant Mackerel-Like Fishes,
Tunnies, Spearfishes and Swordfishes. Occ. Pap. Cal. Acad. Sci., XII, 1-113.

Jordan, Evermann and Clark, 1930, Check List of the Fishes and Fishlike
Vertebrates of North and Middle America north of the boundary of Vene-
zuela and Colombia, Rep. U. S. Comm of Fisheries, for 1928, Part II, 1-670.

Kishinouye, K., Contributions to the Comparative Study of the So-called Scom-
broid Fishes. Journal Coll, of Agric., Imp, Univ. Tokyo, VIII, 295-475.

Lesson, R. P., 1830, Zoologie, In Duperrey, L. I. Voyage autour du monde —
sur la corvette . . . “La Coquille,” pendant . . . 1822-25. Paris.

Lowe, R. T. 1839, A supplement to a synopsis of the fishes of Madeira, Proc.
Zool. Soc. London, VII, 76-92.

Mowbray, L. L., 1920, Description of a Thunnus believed to be new, Copeia,
78, 9-10.

Mowbray, L. L. 1935, Description of the Bermuda Large-eyed Tuna Parathunnus
ambiguus, n. sp., Bermuda, May 1935. (Three page, unpaged, privately
printed pamphlet) . Government Aquarium, Bermuda.

Sloane, Hans, 1707, A Voyage to the Islands Madera, Barbados, Nieves, S.

Christopher and Jamaica, with the natural history of those Islands.

London, cliv, 1-264, plates.

Walford, L. A., 1936, Game Fishes of the Pacific Coast from Alaska to the Equator,
University of California Press.

194

Zoologica: New York Zoological Society

EXPLANATION OF THE PLATES.

Plate I.

Fig. 1. Parathunnus atlanticus (Lesson). Copy of the original illustration of the
type of Parathunnus rosengarteni Fowler, the drawing made from a
mounted specimen.

Fig. 2. Parathunnus atlanticus (Lesson). 560 mm. (23%-inch) specimen from
St. Lucia, British West Indies. Photograph by John Tee-Van.

Plate II.

Fig. 3. Parathunnus atlanticus (Lesson). 495 mm. (19%-inch) specimen from
Bermuda. Photograph by John Tee- Van.

Fig. 4. Parathunnus atlanticus (Lesson). 263 mm. (10-1/3-inch) specimen from
Bermuda. This is the smallest recorded specimen. Photograph by John
Tee-Van.

Plate III.

Fig. 5. Neothunnus argentivittatus (Cuvier and Valenciennes). Specimen from
St. Helena. Copy of plate labelled “Thynnus albacora,” from Cunning-
ham, Proc. Zool. Soc. London, 1910, p. 110, text-fig. 4.

Fig. 6. Neothunnus argentivittatus (Cuvier and Valenciennes). 555 mm. (22-
inch) specimen from Bermuda. Photograph by John Tee- Van.

Plate IV.

Fig. 7. Neothunnus argentivittatus (Cuvier and Valenciennes). 48-inch speci-
men weighing 45 pounds, taken at Bermuda by Mrs. W. B. Holler of
Birmingham, Michigan. Photograph by David Knudsen.

Plate V.

Fig. 8. Neothunnus argentivittatus (Cuvier and Valenciennes). 1,450 mm. (57-
inch) specimen from Castries, St. Lucia, British West Indies. Photograph
by John Tee- Van.

Fig. 9. Neothunnus argentivittatus (Cuvier and Valenciennes). Head of 1,450
mm. (57-inch) specimen from Castries, St. Lucia, British West Indies.
Photograph by John Tee- Van.

Plate VI.

Fig. 10. Neothunnus argentivittatus (Cuvier and Valenciennes). Swim-bladder of
1,450 mm. (57-inch) specimen from Castries, St. Lucia, British West
Indies. Photograph by John Tee- Van.

Fig. 11. Neothunnus argentivittatus (Cuvier and Valenciennes). Caudal keels
of 1,450 mm. (57-inch) specimen from Castries, St. Lucia, British West
Indies. Photograph by John Tee- Van.

Plate VII.

Fig. 12. Neothunnus argentivittatus (Cuvier and Valenciennes). 1,450 mm. (57-
inch) fish being brought on board yacht at Castries, St. Lucia, British
West Indies, showing type of native boat used for fishing for these
giant fish. Photograph by John Tee-Van.

Fig. 13. Copy of original figure of Hans Sloane of the Madeiran albacore. (Sloane,
Tab. 1, Fig. 1).

BEEBE & TEE-VAN

PLATE 1.

FIG. l.

FIG. 2.

SYSTEMATIC NOTES ON BERMUDIAN AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

PLATE II.

j

BEEBE & TEE-VAN.

FIG. 3.

W

FIG. 4.

SYSTEMATIC NOTES ON BERMUDIAN AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

BEEBE & TEE-VAN.

PLATE III.

FIG. 5.

FIG. 6.

SYSTEMATIC NOTES ON BERMUDIAN AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

FIG. 7.

SYSTEMATIC NOTES ON BERMUDIAN AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

PLATE IV.

BEEBE & TEE-VAN.

BEEBE & TEE-VAN.

PLATE V.

FIG. 9.

SYSTEMATIC NOTES ON BERMUDIAN AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

BEEBE & TEE-VAN.

PLATE VI.

FIG. 10.

FIG. 11.

SYSTEMATIC NOTES ON BERMUDIAN AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

BEEBE & TEE-VAN

PLATE VII.

FIG. 12.

SYSTEMATIC NOTES ON BERMUDIAN AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

Beebe: Food of Tunas

195

15.

Food of the Bermuda and West Indian Tunas of the Genera
Parathunnus and Neothunnusx.

William Beebe.

Director, Department of Tropical Research,

New York Zoological Society.

(Plates I-III).

Outline. ^

Page

Introduction 195

Pood of the Atlantic Black-finned Tuna, Parathunnus atlanticus :

Part I: Bermuda Notes 196

Part II: Florida Notes 203

Part III: St. Lucia and Tobago Notes 203

Food of the Atlantic Yellow-finned Tuna, Neothunnus Argentivittatus 203

Introduction.

Although these grand game fish are so widely distributed, and in certain
localities actually abundant, almost nothing of an accurate character has
been published in regard to the food of any species of tuna2. In paper after
paper we find that they are credited with feeding on “small fishes such as
herrings, sardines and anchovies.” Other authorities aver they subsist
almost entirely on Copepoda and similar constituents of pelagic plankton.
As a matter of fact, the above-mentioned fish enter hardly at all into their
diet, while Copepoda is almost the only general group of crustaceans which
we have never recorded from tuna stomachs.

This examination of the stomachs of tuna has developed varied and
unexpected side lines of interest. We have found rare or quite new species
of fish and invertebrates; from the food alone we have been able to learn
much of the fish fauna of certain areas, especially in depths of one hundred
to three hundred fathoms, off Bermuda, which are so obstructed with coral
reefs that neither net nor dredge can be used. For example, we have
found that Holocentrus meeki Bean (1906) is the immature stage of
H. ascensionis. We know by the food at what depths the tuna have fed,
and whether they have been swimming and feeding solitarily, in pairs or in
schools.

My thanks are due to Mr. Martin D. Burkenroad of the Peabody
Museum for the determination of decapods taken from the stomachs.

1 Contribution from the Bermuda Biological Station for Research, Inc.

Contribution No. 502, Department of Tropical Research, New York Zoological Society.

2 An exception is R. LeGendre’s paper on the food of Germo alalunga (Ann. Inst. Oceanogr.
Paris, 14, 1934, pp. 249 ff.).

196

Zoologica: New York Zoological Society

[XXI: 15

Food of the Atlantic Black-finned Tuna, Parathunnus
atlanticus (Lesson).

Part I.

Bermuda Notes.

Summarizing the food of 58 black-finned tunas taKen in Bermuda during
1935 and 1936, we find that seven phyla are represented, Coelenterata, Nema-
thelminthes, Platyhelminthes, Annelida, Mollusca, Arthropoda and Chordata.
The three general groups most abundantly and frequently represented are
fish, young crustaceans (especially squilla larvae) and squids. Twenty-nine
species of fish have been differentiated, not counting hopelessly mutilated
individuals, which, if they could be identified, would easily double the num-
ber of forms. Of the 29, 13 have to date been specifically identified. Of
the 58 stomachs examined, the majority contained both invertebrates and
fish. Seven, or 12%, were empty.

The following list shows the general types of food contained in the
stomachs :

Type of Food

Sargassum weed

Siphonophores

Worms

Gastropods

Pteropods

Cephalopods

Squilla young

Shrimps

Amphipods, crabs, etc.

Isospondyls

Non-Isospondyls

Unknown fish

No. of stomachs in
which it occurred.

4

1

6

3

5

31

16

20

25

2

54

17

We can hardly consider the tunas as herbivorous, yet in four stomachs
were such amounts of sargassum weed that there is no doubt it was swal-
lowed intentionally. The single siphonophore may very well be put down as
accidental, perhaps included when the squilla or the squid was swallowed.
Worms, too, are negligible, most of them being parasitic stomach nematodes.

The pteropods are an interesting although relatively unimportant item
of diet. Four species have been found, Cavolinia, Creseis, Cuvierina and
Limacina, one each in four occurrences, so this is obviously a casual element
of food. It illustrates, however, what we will find emphasized in other
groups, that organisms of remarkably small size are deliberately chosen.
The few, very small gastropods were young, free-swimming phases, not
yet settled down to a bottom life.

For relatively large size and frequency of occurrence, squids were far
ahead of any other group of invertebrates, and equalled only by the group
of fishes as a whole. Squids were found in 30 stomachs or more than 60%
of those which contained food, and they totalled 96 individuals. Some were
pelagic forms found in the upper layers of the ocean, but a considerable
percentage were luminous inhabitants of quarter-mile or greater depths.

The most universal food of tunas both of the Atlantic and the Pacific
is the young of squillas or stomatopods. There is not the slightest doubt
of the deliberate search for and choice of these organisms. They occur in
16 stomachs, nearly one-third of the whole, and to the number of 907
individuals. From these tuna records we may be certain that squilla larvae

1936]

Beebe: Food of Tunas

197

occur usually in swarms. In 5 out of the 16 the average was something
over 2 squillas to a fish, but in the remaining instances the average was over
80 squillas in each case. Many of these tunas had been feeding near the
surface, yet No. 24,513 which had undoubtedly been procuring its nutriment
a quarter-mile down, had devoured 60 squilla larvae.

The hyperid amphipods Oxycephalus and Phronima, the euphausiid
Thysanopoda, megalops and young swimming crabs were present, but in
small numbers and widely distributed throughout many stomachs. Other
hyperid amphipods occurred 10 times, one fish having swallowed 55. Single
specimens were taken of two peneids, one being the rare larva, Cerataspis
monstrosa, and the other an adult male Funchalia villosa, the third male of
its subgenus ever reported.

Carideans were usually rare, but a new species of Leptochela occurred
six times, in large numbers in every case. We counted 24, 34, 161, 144, 27
and 39, all of them being adult. The interesting thing is that the first four
lots were from tunas all taken in mid-September, 1935, from the same
locality and at the same time, so that there must have been a tremendous
swarming of the shrimp. When diving seven fathoms down near Gurnet’s
Rock, I have often seen great misty clouds of crustaceans, usually mysids.
In cases like this, it would hardly seem possible for the tuna to seize the
organisms one by one. The fish must swim through the swarm, snapping
right and left, and, as with all the other elements of the food, swallowing
the crustaceans whole. The two last lots of Leptochela were from fish
equally closely associated in time and space, but taken on September 28.
A species of Palaemonella, probably new, was present in two stomachs, and
undescribed caridean post-larvae (Hippolytid?) occurred in three.

The fishes of the tuna food may, for convenience, be divided into three
sections: first, the deep-sea element in tunas Nos. 24,512, 24,513 and 25,705;
second, the fish which from September 4 to 28, 1935, are closely interlocked
by their food ; and third, all the other fish food. Considering these in reverse
order, the third section includes scattered records of six fish used as food.
A tuna taken July 18 had eaten a three-inch red-tailed triggerfish, Xan-
thichthys ringens, and a six-inch creolefish, Paranthias furcifer. September
14, a tuna caught well offshore had eaten three pomfrets, Brama rail, meas-
uring little more than an inch each. October 9 we took an inchling trigger,
Monocanthus tuckeri, from another tuna.

A butterflyfish, Chaetodon, only half an inch long, was among the food
of a tuna taken in March, 1936. Four other tunas captured at this time, and
within a short time of each other, had been feeding on anchovies, Sardinella
anchovia. This quartet of tunas had eaten 55, 1, 2 and 16 respectively. The
anchovy, unlike its relations, the pilchards and green fry, is rare close in-
shore at Bermuda, but very abundant offshore in mid-depths.

On September 4, 1935, my good friend and very skilful angler, Mrs.
E. T. Weir, went out after tuna, and two miles southwest of Nonsuch Island
sighted a large school leaping and playing about at the surface. A conserva-
tive estimate was about three hundred and they all appeared to be in the
neighborhood of two feet long. Trawling back and forth, Mrs. Weir caught
three and I found the food of this trio intensely interesting. They were the
beginning of a series of seventeen, all interrelated through the similarity
of their food, all taken in the three and a half weeks’ time between Sep-
tember 4 and 28, 1935.

This general area from which they came is the almost unexplored zone
which lies between the shallow Bermuda shore waters (where, down to fifty
feet, we can walk about and observe by means of the diving helmet) and
the outer abyssal depths, where the bottom, being of soft, relatively level
ooze, is dredgeable. The prevalence of rough, jagged limestone covered
with living coral makes the mid-zone wholly impossible for either trawling

198

Zoologica: New Yoi'k Zoological Society

[XXI: 15

net or dredge. In the old Challenger Reports of sixty-three years ago, we
find that six soundings or dredges were made within this coral reef zone
directly south of Nonsuch Island where the tuna are feeding today. We
read of empty dredges or “in heaving in, the line carried away through
chafing on hard ground.” A short distance farther out, but still in the coral
zone, at Station 33, the dredge came up unharmed with a catch so rich that
the amazing fauna of this difficult ground was evident. The summary reads :
“Excluding Protozoa, over 80 specimens of invertebrates were obtained at
this station, belonging to about 70 species, of which 22 are new to science,
including representatives of 3 new genera ; 12 of the new species and 1 new
genus were not obtained elsewhere.” In addition to this list, 22 pteropods
and heteropods, 152 Foraminifera and 83 diatoms were found. So the total
for this one dredge haul was 337 species of organisms.

When I examined the stomach of the first of these tunas, I realized
that the contents were alien to the shallow waters of Bermuda along shore,
and yet had nothing in common with the fauna of the deeper, offshore areas.
And I will here anticipate another discovery which was emphasized again
and again, that these great fish had almost without exception been feeding
close to the bottom. Somehow, I had never visualized these swift, pelagic
beings as searching over, around and perhaps in the gorges and arches of
the eroded limestone. But for that matter I had never thought to find such
small, spiny organisms as squilla larvae dominant in their diet.

During the few nights preceding the first catch of these September
tunas, we had as usual netted many fish attracted to our New Nonsuch wharf
by the submerged and the ultra-violet lights. During this period we caught
a number of the rarest squirrelfish known to Bermuda. The type of Holo-
centrus meeki was taken only three miles from our wharf, near St. David’s
Island thirty years ago. On Nonsuch Island, we had once captured an
additional single specimen. Here were six taken in one evening, and in the
stomach of the first tuna were no fewer than sixty of the same squirrelfish,
in excellent condition. PI. Ill, Fig. 5, shows fifty of these, so little affected
by the gastric juices that even the colors were faintly evident. I was
amazed to see this unexpected mass of exceedingly rare fish.

We kept several of those taken at the wharf alive in aquariums, and
carefully watched them. Before our eyes, little by little, in the course of
several weeks they changed from Holocentrus meeki to the common squirrel-
fish, Holocentrus ascensionis. The soft dorsal and anal fins increased in
height, the depth of the body became greater, the greenish metallic tints
gave place to rose and scarlet, and the abrupt, unlovely profile of the caudal
peduncle slowly took form in our aquariums. There was no doubt about the
fact that meeki is the immature stage of ascensionis. Large, brilliantly
colored individuals of the latter species lived in cracks and crevices among
the rocks about our wharf, but we had never before seen the young, and had
no idea where they bred. Even in their habits and actions as well as in
physical attributes the younger ones were unlike the adults. They kept in
mid-water, well above the bottom, and were extremely wary, dashing here
and there, often pursued and driven off by our wharf’s habitues, their elder
brethren.

There is another less common species of squirrelfish, the black-barred,
Holocentrus vexillarius, slightly deeper in the body, darker in shade, with
fewer dorsal spines and rays and only about half as many gill-rakers as the
common form. There is also the very lovely, lesser butterflyfish, Chaetodon
sedentarius, which we count as a rare fish in Bermuda, in comparison with
three other species of the same genus. With this fore-knowledge of these
three forms in mind, let us prepare a table of the food of our seventeen
tunas, all taken within the period of two weeks, one to two miles off Non-
such, and see what we can learn from it (Table I).

1936]

Beebe: Food of Tunas

199

TABLE I.

Food of Bermuda specimens of Parathunnus atlanticus taken in September.

Tuna No.

17

18

19

23

24

25

26

27

34

35

36

37

38

39

40

41

42

Total

September

4

4

4

12

12

12

12

21

27

27

27

28

28

28

28

28

28

Sargassum weed

1

1

Holocentrus ascensionis

60

66

14

18

4

2

1

44

13

3

2

52

277

Holocentrus vexillarius

3

4

1

8

1

64

71

80

73

87

392

Chaetodon sedentarius

1

1

1

2

3

2

1

18

18

5

52

Leptochela sp. nov..

24

34

161

144

27

39

429

34

Squids

1

4

3

1

7

4

10

2

2

Squilla yg

33

12

210

65

71

391

Serranid

1

Parrotfish

2

Wrasse, incl. 1 Xyrichthys

4

Exonautes rubescens

. 1

Melanostomiatid-like fish

1

Flounder ( Etropusf )

1

Acanthurus sp

1

Creseis

1

Limaeina

1

Yg. Gastropods

2

Oxycephalus

1

1

Hyperids

5

1

Glaucothoe yg

1

Palaemonella sp. nov?

4

2

Caridean yg. (Hippolytid?) ....

2

1

1

Megalops.

1

2

2

Let us consider first the common squirrelfish, H. ascensionis (which for
a time we thought was H. meeki) . We find that it forms a large proportion
of the food of the first twelve tunas, extending over the entire period of
three and a half weeks. Although there are two tunaless intervals during
this time, one of a week and the other of ten days, yet the supply of young
squirrelfish never diminished. From a single individual to as many as 66
occur in a single stomach, the average eaten by each fish being 23. Here is
a total of 277 squirrelfish, all of exactly the same stage of growth, all
measuring two and a half inches in length (Fig. 5).

This paper purports to be concerned with consideration of the direct
elements of food of tunas, but before we leave this species of Holocentrus,
I want to carry them on to the indirect victualization of these fishes. I
select three young squirrelfish at random from among the 60 in the first
tuna’s stomach and, in turn, examine their stomachs.

Stomachs of H. ascensionis

A.

Candacia aethiopica 15

Other copepods 7

Ostracods 6

Euphausiids
Shrimps

Zoeas 8

B. C.

21 153

5 6

1

2

6 7

Here in three out of 60 ingested fish were 237 organisms in perfect
condition, giviner a still more intensive visualization of the life of the

200

Zoologica: New York Zoological Society

[XXI :15

isolated mid-zone. And in the stomachs of several of the Candacia were
radiolarians and diatoms, and so ad infinitum. This gives at least a more
vital conception as to what means we must resort if we are ever to get
farther in the study of these relatively inaccessible regions, than through
conventionally usual methods of direct observation.

The stomachal distribution of the black-barred squirrelfish, H. vexil-
larius, is enlightening. The first three tuna had caught 3, 4 and 1 respec-
tively, showing that these fish were about on September 4, after which not
an individual appeared until 23 days had passed, when a tuna dined off both
species. If we had had only this one fish, the significance of the reappear-
ance of vexillarius would not have been apparent, but the following day and
the same place yielded six tuna which were filled to repletion with vexil-
larius, while ascensionis, except for one fish, vanished forever from the diet,
although they kept coming in small numbers to the night lights at our wharf.
In the 5 tunas which had eaten only vexillarius were 375 individuals, an
average of 75 squirrelfish to a tuna.

Like vexillarius, one or two of the rare butterflyfish, Chaetodon seden-
tarius, were found and devoured on September 4, and then reappeared
on the 27th. On this and the following day, they must have been far from
rare around the deep coral reefs of the tunas’ feeding grounds, for 52 were
divided among 8 tunas, two of which had found and accounted for 18 each.
These butterflyfish will not take a hook, however small, and no trap can be
lowered and recovered in these waters, so were it not for the tuna gour-
mands we should never have known of their abundance anywhere near Ber-
muda. Chaetodonts are essentially reef fish, swimming in and out among the
branches and fans, and the young especially are never found far from
shelter. Yet there were a half hundred youngsters, none over an inch in
length, fallen victims to fish which, according to our preconceptions, should
find their living in swift pursuit of organisms in mid- and upper zones.

Additional evidence of the frequent bottom feeding of tunas is found
in one captured on September 27, which, in addition to both species of squir-
relfish and chaetodonts, had included 4 wrasse and 2 parrotfish in its diet.

I have already spoken in general of squids, squilla larvae and carideans,
and in the present consideration it is necessary only to look at the chart to
see how the relation of these organisms to other food elements holds good
for these 17 as for the entire 58 tunas. Thirty-four squids were eaten by
9 of the 17 tunas, with an average of less than 4 each; 391 squilla larvae
were found in the coral zone by 5 tunas, each averaging 78 ; and specimens
of the new Leptochela, as I have already mentioned, were present to the
number of 429. Other items of diet of the 17 fish were negligible, but the
food of tuna Number 27, taken on September 12, stands out by itself. The
fish was captured in company with four of its fellows and like them had
found H. ascensionis eminently to its taste, having pursued and captured 44,
which if laid end to end would extend a distance of ten feet! These were,
however, packed, sardine-wise, side by side within the tuna. With them
were 13 other organisms of especial significance. These included two species
of pteropods, an Oxyceyhalus, a small Melanostomiatid-like fish and a young
free-swimming flounder. These and a few others come to the surface of
the ocean at night, but in the day are found relatively deep and well offshore.
So this individual fish must have done part of its feeding by itself and
farther down and out at sea, or the night before. The fresh condition of the
food indicated the former course as the more probable one. One more link
with the habitat of the young squirrelfish and the butterflyfish was a single
surgeon fish in its food.

On August 7, 1935, Mr. Christianson brought us two tunas, caught al-
most simultaneously, three miles southeast of Nonsuch Island. This is be-

1936]

Beebe: Food of Tunas

201

yond the shallow, coral-lined waters of the coastal area, and over the abyssal
depths of the deep open sea itself. Tuna are so often seen and taken in
pairs that it appears reasonable to assume a definite relation between the
two. In this case the male weighed sixteen pounds and the female twelve.
They were typical black-finned tuna.

Although they were caught at the surface with white feather jigs, the
first glance at the food showed that very recently they had been feeding at
great depths. The food was quite fresh and the tuna must have come to
the surface with a rush and immediately seized the feather jigs, far above
their zone of feeding. Between them, these tuna had eaten thirty deep-sea
fish, some quite new, others exceedingly rare, and several of the world’s
record size. In addition they had swallowed a dozen luminous squids and
more than three score other creatures of the black abysses. The light organs
of all were plainly visible, together with the typical scarlet and black colora-
tion of inhabitants of the sunless depths. None of these hundred-odd or-
ganisms would ever, of its own free will, have risen even into the dimly-
lighted regions of mid-water. The average depth at which they thrive, and
must have been caught and devoured, is a half mile or more below the
surface.

I submit photographs of the contents of these two stomachs as the con-
tents were removed and roughly sorted, and opposite each of these two
figures are two others showing the accurate restorations of every item of
diet. In PI. I, Fig. 2, we see at the top 23 lanternfish, Diaphus effulgens ,
then an oval-bodied pomfret or Brama rail, and to the left an unknown, short-
toothed, gempylid-like fish. Below is a fierce-looking Alepisaurus or lancet-
fish with a gigantic sail-fin. Next come two sabre-toothed Omosudis lowii,
the larger a world’s record for size (180 mm.). Near the bottom are three
silver hatchetfish, Argyropelecus aculeatus, and scattered about we count 9
squid covered with luminous organs of various colors.

The second plate of reconstruction (PI. II, Fig. 4) shows what the
female tuna had for dinner on this particular seventh of August, and proves
that she must have eaten at the same level and time as her mate. There are
fewer lanternfish, in fact only three, and these are of a different species,
Diaphus rafinesquei, and we counted 60 larval squillas. This tuna had also
devoured exceedingly rare deep-sea fish. At the top is a new species,
gempylid-like, with saw teeth; then two most interesting members of the
family Trichiuridae or cutlassfish, a family wholly new to Bermuda which
we have never taken in any of our fifteen hundred nets. The name of this
fish, brought to us from a great depth by the tuna, is Benthodesmus atlan-
ticus.

Below these are five individuals of a probably new species of Paralepis
or ocean pikelet. Finally there are three luminous shrimps which were in
the stomach of this tuna. The relative sizes are indicated on the figures.
These deep-sea fish are larger than any fish which we have found in tunas
feeding at higher levels.

Table II shows, in diagrammatic form, the direct and indirect food of
these two deep-feeding tuna.

The stomach of the third and remaining tuna which showed indubitable
evidence of having fed, in part, at a considerable depth was sent to us
through the courtesy of Mr. Robert Blackman. It was caught near Bermuda
early in the spring of 1936, and contained the following organisms: one
170 mm. Avocettina infans, which is one of the rarest deep-sea eels; two
specimens of the lanternfish, Myctophum hygomi; one Sardinella anchovia ;
two squids, one of which had light organs ; and one squilla larva.

Direct and indirect food of two specimens of Parathunnus atlanticus.

202

Zoologica: New York Zoological Society

[XXI: 15

5 g

co

•<s>

§ |

O -3

a>

ft

>>

ffi

3

O'

CQ

m

3

O

3

g

J g,

S

©

ft

m

*£

3

o

m

T3

*3

3

ft

s **=»

©

’©

B

’£

<x>

ft

g

g

3

3

S

"So

Q ©

ft

>s

w

cr1

3

CO

O

CQ

w

O

o

CQ

CQ

CM

rH

CO

fc-

rH

rH

(N

rH

© CO

ft 5g

m

T3

© -hi

ft

o

ft ©

jjj

u

3

?> S

£

?H

•+J

r3

m

H v8

CQ

O

CO

T— 1

00

3

CQ

3

ft1

CQ

O

CO

>i <o

ft ft

s w ■

«Ss

CO

cq s

ft

©

e

©

§ »
^ CO

© ■

.e <§.

Q s

CO

1 Shrimp 1 Cyclothone sp. 2 Shrimp 1 Fish

3 Hyperids 1 Larval Fish 1 Squid 1 Squid

1 Gammarid 1 Squid

1936]

Beebe: Food of Tunas

203

Part II.

Florida Notes.

The data from this region is very scanty and refers to only two in-
dividual Parathunnus atlanticus, kindly sent me by Mr. Frederick Church.
Both were found in the stomach of a Blue Marlin, Makaira nigricans ampla
(Poey), which was more than ten feet long and weighed 304 pounds. The
tuna weighed respectively 7 and 4 pounds. The former had eaten a single
Sardinella of 55 mm. The second tuna ran true to form and though there
were only two small organisms in its stomach, they were both squilla larvae.

Part III.

St. Lucia and Tobago Notes.

A pair of tuna taken at St. Lucia and 20 at Tobago show such similar
stomach contents that I am considering them together. In our work at St.
Lucia we have to thank Major William Lambert for many courtesies.

A number of leaves of some tropical tree found in one fish is the sum
total of vegetable elements. Nine groups of invertebrates are present, of
which only three are dominant; megalops with a total of 114 in 15 tunas,
308 squilla larvae in 19 stomachs, and 43 squids eaten by 7 individuals.
Eleven genera of fish were identified in 11 tuna, besides many other small
ones not yet named. All occur singly or not more than 4 to a stomach except
Anchoviella (18), Polynemus virginicus (10), Monocanthus (12) and
puffers (8). The others are Harengula, Hemirhamphus, Fistularia serrata,
Jenkinsia, Decapterus macarellus, Selene vomer and Lactophrys. Except for
the Decapterus, which was 6 inches long, all are two inches or under, and
the majority are an inch or less. Their spiny character and almost wholly
bottom habitat are also evident.

We once recorded Rhomboplites aurorubens from the stomach of a
tuna taken near Haiti.

Food of the Atlantic Yellow-finned Tuna, Neothunnus
argentivittatus (Cuv. and Val.).

As far as I know there has been no attempt at recording any details
with regard to the food of this important commercial and game fish. The
appended table presents a list of the stomach contents of eight individuals,
one from Bermuda and seven from St. Lucia, the latter taken in the
Channel between that island and Martinique. The Bermuda yellow-fin had
swallowed only a few fresh fronds of Sargassum weed, and, strangely
enough, leaves and bark were found in the stomachs of two of the others.

Squids were found in five out of the seven St. Lucia specimens. Ten
species of fish have been identified and those which remain unknown would
about double this number. Oxyporhamphus is an interesting primitive
form of flyingfish with very small pectorals, and was unquestionably caught
near the surface, while Gonostoma, the only representative of a deep-sea
family found in the food of the yellow-fins, was probably swallowed several
hundred fathoms down.

St. Lucia yellow-finned tuna No. 24,689 had the most interesting food.
The invertebrate proportion was negligible, consisting of a single squid man-
dible, and 9 megalops. The fish totalled 58 individuals of at least 6 species,
and comprising 4 puffers, 1 leathery filefish, 1 red-tailed triggerfish, 9 uniden-
tified triggers and 42 flying gurnards. The average length of the puffers is
81 mm., of the triggerfish 31 mm., and of the two score gurnards 45 mm.

204

Zoological New York Zoological Society

[XXI :15

A general survey of the food of this individual reveals several unusual
facts which are characteristic of tunas of several species, both in the Atlan-
tic and Pacific Oceans. Here is a fish almost six feet in length (1,450 mm.)
with a correspondingly good-sized mouth, yet which has chosen to swallow 58
fish averaging barely two inches in length. Also it would be difficult to choose
a lot of fish less appetizing than these puffers, triggers, turbots and
gurnards. They epitomize spininess in respect to skin, opercles and fin ele-
ments, and correlated with this supposedly protective armor the actual flesh
and muscle necessary for swift movement are much less developed than in
more ordinary fish.

This choice must be deliberate when we consider the amazing speed
of these tunas, the ample size of their mouths and the abundance of smooth-
skinned, thick-fleshed fish of all sizes. The swift-moving white feather jig
would seem to have nothing in common with triggers, puffers and squilla
larvae, yet this white surface lure is pursued and seized with eagerness and
ease. Among the spiny organisms which dominated the tunas’ diet were
squillas, euphausiids, shrimps, megalops, Hyporhamphus, Oxyporhamphus,
Holocentrus, Cephalacanthus, Xanthichthys, Balistes, and Monacanthus,
while squids were the only smooth-skinned creatures of any numerical im-
portance (Table III).

TABLE III.

Food of yellow-finned tuna: Neothunnus argentivittatus.

Cat. No. of Tuna

24,682

24,683

24,(589

24,690

24,691

25,206

Place

St. Lucia

St. Lucia

St. Lucia

St. Lucia

St. Lucia

St. Lucia

St. Lucia

Bermuda

Length

645

690

1450

1360

555

Weight

12

16

140

....

8

Sex •

Male

Male

Male

Male

Male

Male

Sargasaum Weed

Few fronds

Leaf

1

Bark

i

Squid

Squilla larvae

"2

26

1

”5

2

5

"2

” 6

Euphausiids

3

. .

Shrimps

1

Megalops .

Sardinella anchovia

"9

55

1

*"i

Gonostoma sp

"2

Hyporhamphus unifasciatus.. .

1

Oxyporhamphus micropterus. .

1

Holocentrus ascensionis

1

"4

"io

Decapterus macerellus

1

" i

Cephalacanthus tolitans

"42

"2

Xanthichthys ringens

1

Balistes for cipat us

1

Monacanthus hispidus

1

Sphaeroides spengleri

4

Unknown flying fish

i

Unknown triggerfish

”9

" i

....

Unknown fish

"2

1

8

”2

1936]

Beebe: Food of Tunas

205

EXPLANATION OF THE PLATES.

Plate I.

Fig. 1. Contents of the stomach of a 16 pound male, black-finned tuna, Para-
thunnus atlanticus, Catalogue No. 24,512, taken at the surface south of
Nonsuch Island, Bermuda, August 7, 1935. Both the fish and the inver-
tebrates are deep-sea forms, relatively fresh and quite recognizable.

Fig. 2. Restoration of the deep-sea fish and invertebrate tuna food shown in
Fig. 1. There are 42 organisms including 23 Diaphus refulgens, 1 Brama
rail, 1 short-toothed gempylid, 1 Alepisaurus sp., 2 Omosudis lowii, 3
Argyropelecus aculeatus, 1 unidentified triangular-toothed fish, 1 hyperid
amphipod and 9 luminous squid. (Drawing by George Swanson).

Plate II.

Fig. 3. Contents of the stomach of a 12 pound female Parathunnus atlanticus,
Catalogue No. 24,513, taken at the same time and place as No. 24,512.

Fig. 4. Restoration of the deep-sea food of the female tuna shown in Fig. 3.
Sixty-seven abyssal organisms including 1 saw-toothed gempylid, 2 Ben-
thodesmus atlanticus, 2 short-toothed gempylids, 5 Paralepis sp., 3 Dia-
phus rafinesquei, 3 luminous squids, 51 squilla larvae. (Drawing by
George Swanson).

Plate III.

Fig. 5. Fifty out of 60 squirrelfish, Holocentrus ascensionis, taken from stomach
of black-finned tuna, Parathunnus atlanticus, Catalogue No. 24,558.

\

BEEBE.

PLATE I.

FIG. 2.

FOOD OF THE BERMUDA AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

BEEBE.

PLATE II.

FIG. 4.

FOOD OF THE BERMUDA AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

BEEBE.

PLATE III.

1 '

■I sit

FOOD OF THE BERMUDA AND WEST INDIAN TUNAS OF THE
GENERA PARATHUNNUS AND NEOTHUNNUS.

Crane: Biology and Ecology of Giant Tuna

207

16.

Notes on the Biology and Ecology of Giant Tuna, Thunnus
thynnus Linnaeus, Observed at Portland, Maine1.

Jocelyn Crane.

Technical Associate, Department of Tropical Research,

New York Zoological Society.

(Plate I; Text-figure 1).

Contents.

Introduction

Materials and Methods

Description of Thunnus thynnus, Based on Specimens Studied:

Colors

Measurements and Counts

Gill-raker Counts

Growth Characters

Ecology:

Size and Weight

Sex

Occurrence ,

Schooling

Parasites

Food

Page
. 207
. 207

208

209

209

210

210

210

210

210

211

211

Introduction.

The material for the following paper was gathered during the latter
part of July, 1936, at Portland, Maine. For their unfailing cooperation, I
wish to express my thanks to Mr. George L. Ratcliffe, President of the Port-
land Fish Co., to Dr. J. S. Jamieson of the same city, and to Mr. Walter H.
Rich, Portland agent of the U. S. Bureau of Fisheries.

My thanks are also due to Mr. John Tee-Van of the New York Zoo-
logical Society and to Mr. Martin D. Burkenroad of the Peabody Museum
for determinations of food samples, and to Dr. William Beebe for initiation
and supervision of the work.

Materials and Methods.

A total of 34 newly-caught tunas was examined, ranging from 41 inches
in standard length and 65 pounds in weight (undressed), to 97 inches and

1 Contribution No. 503, Department of Tropical Research, New York Zoological Society.

208

Zoologica: New York Zoological Society

[XXI :16

860 pounds. Gill-raker counts and food diagnoses were made in every case,
the sex was determined in all but four, and partial or complete measure-
ments and counts were secured from 10 specimens.

The methods of harpooning the tuna were observed by accompanying a
fishing boat to sea, and in this way color notes of two living specimens were
obtained and parasites preserved.

The giant tuna or horse mackerel are fairly common just outside Casco
Bay during the summer months, and their capture is becoming an industry
of increasing importance. Except for the few specimens which are caught
by sportsmen on rod and reel, all of the tuna taken are harpooned by the
native fishermen. The fishing launches, twenty to thirty-five feet in length,
are each rigged with a platform or “pulpit” in the bow, and from here the
harpooner makes his strike. Manned by a motorman and a single fisherman,
the boats leave the wharf early in the morning, run straight out to the
mouth of Casco Bay, and spend the day searching for tuna schools. While
the fish are often found along the very entrance to the Bay, they must some-
times be followed far out to sea, 15 or 20 miles from the dock. When fish
are sighted, the motorman approaches the school and follows it, manoeuver-
ing to give the harpooner in the bow a chance to strike with the 12- or 15-
foot “pike.” Sometimes he must cast the pole as much as 25 feet or more,
or submerge its full length under water. The largest fish taken during my
stay (860 pounds) was harpooned as it swam at least 15 feet in advance of
the bow and an equal distance under water. Small fish are usually brought
alongside with scarcely any struggle, while large specimens sometimes fight
several hours. Often no fish at all are caught by a given boat during the
day, while at other times eight or ten will be taken.

For a complete account of the tuna fisheries of Maine, as well as for
data on the numbers and weights of tuna caught during several recent sea-
sons, see “The Horse Mackerel (Tuna) Fishery of Maine;” Department of
Commerce, Bureau of Fisheries, Mem. S. 339, 1935, by Walter H. Rich.

Description of Thunnus thynnus L., Based on Specimens
Studied at Portland.

Colors: (From two specimens measuring 49 and 96.5 inches, standard
length [ca. 100 and 860 pounds, respectively], examined at sea when caught ;
the smaller fought scarcely five minutes, and was very active when lifted
into the boat, while the larger struggled nearly two hours and was almost
dead upon examination, yet there was no essential color difference between
the two. However, the color in both cases changed and faded entirely after
ten minutes in the air).

Body : Dorsal surface blackish-bronze as far down as the dorsal margin
of the eye or slightly lower; adjoining this a longitudinal stripe of bluish
bronze running entire length of body and extending as low as the level of
lower margin of pectoral fin base ; rest of body silver. At death, the dorsal
surface changed to greenish before losing color and becoming uniformly
dark; the lateral stripe became brownish, then inconspicuous; and the
ventral surface turned faintly violet, then dead white. Iris bronze.

Fins: Pectoral and caudal very dark, almost black; second dorsal silver
gray, darkening at death ; anal paler ; second dorsal lobe and anal lobe with
a narrow greenish-yellow stripe close to posterior margin and extending out
to tip of lobe. This stripe persisted hours after death, was identical in all
specimens examined on the wharf, and was invariably paler and greener
than the finlets. Finlets always bright golden yellow, the color persisting.

1936]

Crane: Biology and Ecology of Giant Tuna

209

Measurements and Counts : The data given below were compiled from
measurements and counts made upon ten specimens. Their standard lengths
and undressed weights were as follows:

No.

Standard Length

Weight

1.

1046 mm.

(41

in.)

65 lbs.

2.

1250

ii

(49

“ )

ca. 105 “

3.

1658

ii

(65

“)

244 “

4.

1691

ii

(66.3

")

189 “

5.

1900

ii

(74.5

“)

ca. 550 “

6.

2015

ii

(79

“)

ca. 600 “

7.

2142

ii

(84

“)

ca. 700 “

8.

2410

ii

(94.5

“)

ca. 700 “

9.

2460

ii

(96.5

“)

860 “

10.

2474

ii

(97

“>

772 “

External : Depth in standard length 2.8 to 3.3; head in length 3.2 to 3.9;
eye in head 7.7 to 10.7; snout in head 2.3 to 3; tip of snout to middle of
eye in head 2 to 2.5; maxillary in head 2.4 to 2.7 ; interorbital in head 1.9
to 2.8; snout to pectoral in length 3 to 3.5; snout to pelvic in length 3.1 to
3.3; snout to first dorsal in length 2.9 to 3.4; snout to second dorsal in
length 1.7 to 1.8; snout to anal in length 1.4 to 1.6; height of first dorsal
fin in head 2.3 to 2.6; height of second dorsal fin in head 1.6 to 2.4; height
of anal fin in head 1.9 to 2.4; pectoral fin length in head 1.4 to 1.9; end of
pectoral fin under 9th to 12th dorsal spine; caudal length in length 6.3 to
10.0; caudal spread in length 2.8 to 3.2. Pectoral rays I, 31 to I, 33; dorsal
count XIV + 1,11 + 9 to 11; anal count 1,10 to 11 + 8 to 10. Gill-rakers
11 — 25 to 14 — 27 (see below). In every case the anal origin was distinctly
behind the insertion of the last dorsal ray.

Internal : Length of stomach in length 3.9 to 5.6; length of longest
liver lobe in length 7.5 to 10.4 ; liver unspotted, of four major lobes and a
varying number of small lobes ; length of caeca sac in length 4.3 to 5.2 ;
length of testicle in length 2.4 to 2.8 (latter in small specimens).

Text-figure 1.

The alimentary canal of a young specimen of Thunnus thynnus (standard
length 41 in., weight 65 lbs.) taken at Portland, Maine. Oes., oesophagus; pylo.,
pylorus; stom., stomach; int., intestine, (x 1/5).

Gill-raker Counts: The gill arches of 34 specimens were examined,
of which 13 had both sides intact, the others having been damaged during
the dressing of the fish. Of these 13, only four specimens had the same
gill-raker count on both sides. As will be seen from the following table of
frequencies, the most usual counts were 13 — 25 and 13 — 26, while the ex-
tremes ranged from 11 — 25 and 12 — 24 to 14 — 27, giving a total of 36 to 41
gill-rakers on the first branchial arch. In the cases where the count was
unequal on the two sides, the largest number occurred indiscriminately on
either right or left side. Broken arches, in which only the upper or lower
count could be taken, are omitted from the table.

210

Zoologica: New York Zoological Society

[XXI :16

Gill-raker Count Frequency

11—25

1

12—24

2

12—25

2

12—26

2

(a pair)

13—24

2

.

13—25

7

(inch 1 pair)

13—26

11

(inch 3 pairs)

13—27

5

14—24

1

14—25

4

14—26

1

14—27

2

Growth Characters: There are a few consistent age differences ap-
parent in the series under consideration: In the smaller specimens, the

dorsal and anal lobes are relatively lower, the stomach is slightly shorter,
the liver has not as many small, sub-lobes and the reproductive organs are
poorly developed.

Ecology.

Size and Weight: Reports of Portland tuna smaller than 33 pounds
can not be verified, and even these relatively small fishes are exceedingly
rare. It is probable, however, that much smaller tuna are sometimes taken
in the mackerel nets ; John Doughty, an experienced tuna fisherman, took a
two and one-half pound fish from one of these nets last year, which he is
certain was a true tuna ; he states that the shape of the fish, the appearance
of the fins and the color and arrangement of the finlets all were typical
of a tuna.

The largest tuna harpooned are supposed to have reached 1,600 pounds,
but these weights are unofficial, having been taken from a newspaper
report. Recorded weights for Portland fish reach 967 pounds. The major-
ity weigh between 300 and 700 pounds.

When the tuna appear off Portland in the early summer (see below),
they are always much thinner than later on. However, large fish which are
caught on the same day in mid-summer and which are equal in length, may
differ almost a hundred pounds in weight (see measurement table, p. 209) ;
in the case observed, at least, sex was not a factor, as both of the largest
specimens listed were spent males.

Sex: Every one of the 30 specimens examined was a male. Of these,
five of the largest (weighing 700 pounds and over) had broadly distended
sperm ducts and spent testicles. In all of the others the testicles were far
from being in breeding condition, while the sperm ducts were extremely
slender.

Occurrence: Every year the first tuna appear in the latter part of
June, are common in July and August, and become scarce in September.
They vanish altogether early in October. Their season corresponds exactly
to that of the herring and mackerel. Usually the tuna remain outside of
Casco Bay in at least 35 or 40 fathoms of water and 12 or more miles from
the dock. This year, however, they have occasionally been seen further in-
shore. There is no report of the occurrence of any other species of tuna in
this area.

Schooling: Fish of approximately the same size form small schools
of which up to twenty or more may be visible at the surface at once, leap-
ing or swimming slowly along with the tips of their fins breaking water.
Small and large fish are never seen in the same school. As a rule, the smaller

1936] Crane: Biology and Ecology of Giant Tuna 211

the fish, the more individuals in the school, while the largest fish often seem
to be solitary.

Parasites : The 860-pound specimen mentioned earlier in the paper had
many dark red Caligus sp. on the bases of the anal fin, of the dorsal and
anal finlets and of the caudal fin. A few were found on the finlets and
the caudal fin. There were Distoma- like worms in about half a dozen of the
34 stomachs examined.

Food: The following food was present in the 34 stomachs examined:

Nos . of stom-

Food a chs in which

it occurred

Merluccius bilinearis (from 1 to 38 fish m a single stomach,
each measuring from 8 to 13 inches in length. In most of

the tunas the food consisted entirely of this species). 26

Seaweed (in stomachs containing little other food; only one or

two fronds were found in each stomach). 4

Squids (one or two in a stomach, alone or with shrimps). 3

Meganyctiphanes norvegica (numerous; all adults). 2

Clupeid, 215 mm. 1

Clupeids, different from above; three, ca. 75 mm 1

Sebastes mar inns; four, 53 to 117 mm. 1

Tylosaurus marinus; one, 135 mm. 1

Five of the tuna stomachs were completely empty. Almost all of the
food was in an advanced state of digestion.

Previous records of the food of Maine tunas are few and indefinite.
Bigelow (“Fishes of the Gulf of Maine,” 1925, p. 213) states that the prin-
cipal food consists of menhaden, mackerel and herring, with occasional dog-
fish, squid and the smaller schooling fishes. Rich (Joe. cit .) also writes
that herring, mackerel and menhaden are the most important elements of
diet. The fishermen almost never open the stomachs, except by accident,
and know only that tunas “often eat herring and mackerel.” Since four genera
of “herring” and five of “mackerel” occur in the Gulf of Maine, and since
my few days’ work yielded five kinds of organisms previously unrecorded
as tuna food, the need for detailed and sustained work on this interesting
subject is evident.

212 Zoologica: New York Zoological Society

EXPLANATION OF THE PLATE.

Plate I.

Fig. 1. A typical tuna-fishing boat of Portland, Maine, showing bow platform.
Photograph from Walter H. Rich.

Fig. 2. A fisherman in position for a strike. Photograph from Walter H. Rich.
Fig. 3. An 860-lb. specimen of Thunnus thynnus taken off Portland in July,
1936. Photograph by Dr. J. S. Jamieson.

CRANE.

PLATE. I.

FIG. 2.

NOTES ON THE BIOLOGY AND ECOLOGY OF GIANT TUNA. THUNNUS
THYNNUS LINNAEUS, OBSERVED AT PORTLAND. MAINE.

Glassell: Six New Brachyuran Crabs

213

17.

The Templeton Crocker Expedition. I. Six New Brachyuran
Crabs from the Gulf of California1.

Steve A. Glassell.

Research Associate in Crustacea ,

San Diego Society of Natural History.

[Introductory Note: This is the first of a series of papers dealing with
the specimens collected on the Twenty-fourth or Templeton Crocker Expedi-
tion of the Department of Tropical Research of the New York Zoological
Society; William Beebe, Director.]

Majidae.

Mithrax ( Mithrax ) mexicanus Glassell, sp. nov.

Type: Male, holotype; Cat. No. 36,712, Department of Tropical Re-
search of the New York Zoological Society; Station 136, Dredge 27; from
the Gulf of California, 23° 28' N. Lat., 109° 24' W. Long., 3 miles northeast
of Cape Pulmo, Lower California, Mexico; 50 fathoms; April 30, 1936;
4-foot Blake dredge; collected by William Beebe on Templeton Crocker’s
yacht Zaca.

Diagnosis: Horns long, divergent. Antennae long. Hepatic spine single,
most prominent of lateral spines. No paired spinules on mesogastric region.

Description of male: Carapace pyriform, tumid, setose, much longer
than wide; lateral margins without an angle. There are six median spines,
none very prominent, except the posterior marginal spine. Of the five
antero-lateral spines the hepatic is the most prominent, the other four
being small; a subhepatic spinule; a prominent postero-lateral spine, about
half the size of the hepatic spine, located above the level of the antero-lat-
eral spines ; a pair of epibranchial spines paraded the antero-lateral spines ;
a pair of spines on the metabranchial region parallel the antero-lateral
margin. The rostral horns are nearly *4 the body length, diverging, reg-
ularly tapering. The suborbital margin is armed with two spines, the
median large, the distal largest and prominent in dorsal view; the supra-
orbital margin has three spines, postorbital, preorbital and median, the
latter the smallest. The antennae are nearly as long as the width of the
carapace.

Chelipeds small, not as long as the first ambulatory leg; merus with a
spine at upper distal end, another just proximal to this; carpus with 3 or
4 small dull spines ; hands smooth, cylindrical ; fingers not gaping.

First ambulatory leg nearly as long as the carapace; legs tomentose
and setose; merus armed with a distal spine on upper carpal articulation.
Terminal segment of abdomen longer than wide, sides converging, tip
arcuate ; not as wide as penultimate segment. Buccal area widest anteriorly.

1 Contribution No. 504, Department of Tropical Research, New York Zoological Society.

214

Zoologica: New York Zoological Society

[XXI: 17

Measurements: Male holotype, length of carapace including rostral
horns 16.2 mm., width 9.4 mm.

Material examined: The male holotype.

Remarks: This proposed species is closely allied to M. ( M ). spinipes
(Bell), 1835 (1836), but differs from that species in that it lacks the small,
side by side spinules on the anterior mesogastric region; by the carapace
being pyriform, instead of having a lateral angle, in this respect it also
differs from M. (M) . acuticornis Stimpson, 1870. It differs from both of
these species in that the hepatic spine is the most prominent, the others
being very small, and by the shape of the ultimate segment of the abdomen
being subquadrilateral, instead of subtriangular.

Stenocionops beebei Glassell, sp. nov.

Type: Female, holotype; Cat. No. 36,714, Department of Tropical Re-
search of the New York Zoological Society; Station 136, Dredge 23; from
the Gulf of California, 23° 28' N. Lat., 109° 24' W. Long., iy2 miles north-
east of Cape Pulmo, Lower California, Mexico; 50 fathoms; May 1, 1936;
4-foot Blake dredge; collected by William Beebe on Templeton Crocker’s
yacht Zaca.

Diagnosis: Median spines 5; none on posterior margin. Lateral mar-
ginal spines 3. Rostral horns widely divergent, about 78°, nearly 1/5 length
of carapace.

Description: Carapace triangular-ovate; regions tumid, covered with
thick sponge-like pubescence and groups of curve-tipped setae. Median spines
5, the anterior very small, the rest large, stout, cylindrical, blunt; three
large, conical, upward-pointing, lateral spines, the hepatic joined to the
smaller subhepatic by a ridge; a pair of stout upward-pointing, meso-
branchial spines, one of which is opposite the cardiac region; with the
proximal antero-lateral spines and the cardiac spine, these form a trans-
verse line of five heavy spines across the carapace at this point; the other
branchial spines with the median lateral spines form a transverse row of
four spines across the carapace at the metagastric region.

Rostral horns widely diverging, at an angle of nearly 78°, regularly
tapering to slender, slightly incurving tips. Supraorbital spine heavy, up-
turned ; preorbital spine separated from supraorbital by a long narrow sinus.

Chelipeds almost as long as first ambulatory leg ; merus armed on upper
crest with 4 large spines and a distal lobe; carpus roughened with a few
low tubercules; manus long, cylindrical, tapering, armed with a single
proximally placed tubercule on upper margin ; fingers long, tapering, slightly
gaping proximally. Ambulatory legs stout, pubescent; dactyli pubescent,
curved.

Color in alcohol: Spines a bright pink. Fingers brown, tips white.

Measurements : Female holotype; length of carapace with horns 56 mm.,
without horns 48 mm., length of rostral horns 10 mm., width of carapace
with lateral spines 45 mm., without spines 38 mm., length of hand including
fingers 24 mm., width at base 4 mm.

Material examined: The female holotype, and a juvenile male para type
from the same dredge haul.

Remarks: This proposed species is allied to S. triangulata Rathbun,
1892, but differs from that species, not only by being larger, but, by having
5 median spines, instead of 9, and by having 3 large, lateral spines, instead
of 3 lateral spines and a hepatic.

This species is named for Dr. William Beebe, director of the Depart-
ment of Tropical Research of the New York Zoological Society, whose ex-

1936]

Glass ell: Six New Brachyuran Crabs

215

ploits and adventures in the field of natural history have cast a glamour
on his calling, and honor on his fraternity.

Xanthidae.

Actaea crockeri Glassell, sp. nov.

Type: Male, holotype; Cat. No. 36,731, Department of Tropical Research
of the New York Zoological Society; Station 136, Dredge 5; from the Gulf
of California, 23° 31' N. Lat., 109° *27' 30" W. Long., 2 miles south-southeast
of Punta Arena, Lower California, Mexico ; 33 fathoms ; April 19, 1936 ;
4-foot Blake dredge; collected by William Beebe on Templeton Crocker’s
yacht Zaca.

Diagnosis: Carapace with regional lobes spinuous anteriorly; tubercu-
late posteriorly; lateral lobes dentiform. Ambulatory legs crested with a
double row of lanceolate teeth on carpus and propodus. Chelipeds covered
with thickset forward-pointing spinules, coarser than on carapace.

Description: Carapace ovoid, nodulous; deeply furrowed, nude; the
nodules on the anterior half thickly covered with short, sharp-tipped, for-
ward-pointing spinules ; those of the posterior half, less thickly covered with
tubercules and granules. A granular ridge parallels the posterior border.
Cardiac region divided by a median sulcus. Front deflexed, a small median
sulcus, lobes arched, spinuous in dorsal view, sinuous in front view. Upper
ocular margin serrate with short, sharp-pointed spines ; lower margin tuber-
culate. Lateral dentate lobes four (besides the orbital), the first very small,
the second smaller than the third and fourth ; second and third, forward
and upward pointing, granulate on posterior face, spinate anteriorly; the
margins of 1 — 2 — 3 teeth are serrate with forward-pointing spinules.

Chelipeds unequal, the right the larger, covered with short, sharp
granules; the carpus with a distal transverse sulcus paralleling the margin,
a spine at the inner angle ; hands covered with thickset sharp spinules ;
fingers channeled, that of the major hand crested. Ambulatory legs short;
merus with a row of short spines on upper margin, granulous; carpus and
propodus armed with an irregular double row of lanceolate spines on the
flattened, channeled surface, also with spines and granules on the posterior
side; dactyli pubescent. Merus of outer maxillipeds with truncate, cylin-
drical tubercules.

Measurements : Male holotype, length of carapace 5.5 mm., width 8 mm.

Material examined: The male holotype.

Remarks: This proposed species is allied to A. angusta Rathbun, 1898,
but differs from that species by having the posterior half of the areolate
carapace granulate, the anterior areolations spinuous, instead of all the
areolations granulous, by having the lateral lobes covered with spinules,
instead of granulous, and by the legs being heavily spined, instead of
granulous.

This species is named for Mr. Templeton Crocker whose interest and
generosity made possible this expedition.

Pilumnus pelagius Glassell, sp. nov.

Type: Female, holotype; Cat. No. 36,731, Department of Tropical Re-
search of the New York Zoological Society; Station 136, Dredge 13; from
the Gulf of California, 23° 29' N. Lat., 109° 24' W. Long., 5 % miles south-
east of Punta Arena, Lower California, Mexico; 45 fathoms; April 20, 1936;
4-foot Blake dredge; collected by William Beebe on Templeton Crocker’s
yacht Zaca.

216

Zoologica: New York Zoological Society

[XXI: 17

Diagnosis: Antero-lateral spines 4, forming a ridge which leads down-
ward to buccal angle. Carapace covered with pubescence forming a ragged
pattern. Upper surface of carpus and manus of chelipeds flattened; outer
surface of hands rough and pubescent.

Description : Carapace about 2/3 as long as wide, uneven, lumpy, and
rough with scattered tubercules and granules amid the pubescence. Regions
margined by deep wide furrows, a deep median sulcus. Front deflexed,
bilobed, the outer margin of each lobe slightly concave. A spine at outer
angle of orbit, followed below by two spinules and a row of long hairs.
Antero-lateral spines forming a downward curve toward the buccal cavity
but joined at their bases in a smooth carina leading to the orbit; the two
posterior spines are much nearer than the remainder, and surmounted at
their apices with fine, sharp, forward-pointing spinules. A few spinules on
subhepatic region.

Chelipeds, except the fingers, densely pubescent, setaceous, concealing
the sharp granules or spinules interspersed; a stout, heavy, sharp-tipped
spine at inner angle of carpus; upper surface of carpus and manus flat-
tened; manus with a line of 5 or 6 spinules on flattened upper surface;
fingers nude, except for proximal upper crest of dactyli ; dactyl of major
hand armed with heavy, lobe-like teeth, the one at the gape the heaviest.

Ambulatory legs of moderate length, pubescent, and furnished with a
fringe of longer hairs, concealing spinules on upper surface. Abdomen
covered with pubescence and fringed with long hairs.

Sexual variation: None.

Color in alcohol : Anteriorly rather pink. Spines on ambulatory legs
a rose color. Fingers a dark brown. Pubescence a dirty brown.

Measurements: Female holotype, length of carapace 9 mm., width 14.3
mm. Male paratype, length 7.2 mm., width 9.2 mm.

Material examined: Two females and one male including the type; all
three specimens were taken in the same dredge haul.

Remarks: This proposed species is allied to P. limosus Smith, 1869.
but differs from that species in that the anterior half of the carapace is not
so densely tomentose, the antero-lateral spines are not in one plane, the
front without a deep, wide, median sulcus, the ambulatory legs with spines
and setae. The abdomen of the male is relatively wider, with the ultimate
segment subtruncate, instead of triangular as in P. limosus.

Goneplacidae.

Chasmocarcinus ferrugineus Glassell, sp. nov.

Type: Female, holotype; Cat. No. 36,735, Department of Tropical Re-
search of the New York Zoological Society; Station 136, Dredge 21; from
the Gulf of California, 23° 29' N. Lat., 109° 25' W. Long., 5 miles south-
east of Punta Arena, Lower California, Mexico ; 45 fathoms ; April 30, 1936 ;
4-foot Blake dredge; collected by William Beebe on Templeton Crocker’s
yacht Zaca.

Diagnosis: Orbits transverse. An antero-lateral marginal granulate
line. Sternum and abdomen of female lightly pubescent and punctate;
abdomen heavily fringed. Merus of ambulatory legs wide.

Description: Carapace nearly % as long as wide. Fronto-orbital dis-
tance less than one-half the width of carapace. Surface covered with
pubescence, punctate, granular toward the marginal angle and the posterior
border; sparse, clavate setae on the sides, antero-lateral shoulders and the
eye-stalks. Two deep longitudinal impressed lines in the center of the
carapace. A distinct, blunt, granular antero-lateral margin. Front trun-

1936]

Glass ell: Six New Brachyuran Crabs

217

cate, straight, entire. Eyes filling the orbits; orbits transverse in dorsal
view. Second and third joints of antennules very long.

Chelipeds lightly margined with hair, for the most part smooth; merus
with a transverse rounded lobe near distal end, a double margin parallels
the carpal articulation; ridges and lobes setose; carpus nearly square with
a setose lobe at inner angle; hands with fingers long, regularly tapering,
slightly arched, crossing at the tips, not gaping; pollex subhorizontal;
lower margin of hand slightly sinuous, with granulate, setose margin.

Ambulatory legs long, hairy; merus stout. Female abdomen lightly
pubescent and punctate, with heavy marginal fringe.

Color in alcohol: Pubescence and setae a dull brick red or rusty color.
Carapace under pubescence a bluish-gray.

Measurements: Female holotype, length of carapace 9.2 mm., width
13 mm. Male paratype, length 7.5 mm., width 10.5 mm.

Material examined: Two females, one male, including the holotype; all
three specimens from the same dredge haul.

Remarks: This proposed species is closely allied to C. latipes Rathbun,
1898, but differs from that species in that the orbits are transverse, instead
of oblique, the eyes fitting the orbits, instead of not fitting, the merus of
the chelipeds with an upper subdistal, transverse lobe or prominence, instead
of without this lobe.

It may be that this species is analogous to C. latipes. To establish these
differences with absolute certainty, however, the examination of a number
of specimens is necessary.

Cymopoliidae.

Cymopolia zacae Glassell, sp. nov.

Type: Male, holotype; Cat. No. 36,739, Department of Tropical Re-
search of the New York Zoological Society; Station 136, Dredge 26, from
the Gulf of California, 23° 27' N. Lat., 109° 24' W. Long., 2 miles northeast
of Cape Pulmo, Lower California, Mexico; 45 fathoms; May 1, 1936; 4-foot
Blake dredge; collected by William Beebe on Templeton Crocker’s yacht
Zaca.

Diagnosis: Five unequal antero-lateral teeth. Carapace half again as
wide as long. First ambulatory leg reaching nearly to distal end of carpus
of second.

Description: Carapace very broad and not very convex, with five antero-
lateral teeth besides orbital tooth ; the first two triangular, sub-acute, having
a longer base than the third and fourth ; these four teeth are upturned
and forward-pointing; the fifth tooth is lamellar, horizontal, on a lower
plane and occupies the lateral angle. Tubercules of carapace well marked,
high, trending forward ; intervening space filled with inconspicuous granules
and scant, short hairs. Two median frontal teeth rounded and separated
by a triangular sinus with a sharp base; lateral sinuses shallow, wider,
oblique; lateral teeth rounded, lobular. The first sinus of supraorbital
margin is narrow U-shaped; the second V-shaped; middle tooth broad,
obliquely truncate, next tooth narrower, rounded. Outer orbital tooth in-
clined forward and inward, outer margin straight. Ridge above posterior
margin crenulate with small tubercules.

Chelipeds slender, subequal, the right the major; fingers with tips
crossing. Sinuses of suborbital margin V-shaped, the inner the sharpest;
outer lobe oblique, bilobed ; inner lobe truncate, rounded posteriorly, ad-
vanced to half the height of outer lobe and spatulate pterygostomian lobe.

218

Zoologies: New York Zoological Society

First ambulatory leg slender, reaching nearly to distal end of carpus of
second ; dactyl slightly curved ; second and third legs nearly twice as long as
width of carapace; merus joints narrowing in distal half; four longitudinal
crests, anterior distal end forming a right angle; anterior lobes on carpus
obsolete, the margin forming a straight crest, the distal end concave with
a small tooth on propodal articulation; propodus enlarging little distally;
dactylus with sinuous posterior margin.

Measurements : Male holotype, length of carapace 8.5 mm., width 13.1
mm., length of second leg 25.5 mm., length of first leg 11.8 mm.

Material examined: The male holotype.

Remarks: This proposed species is closely allied to C. fragilis Rathbun,
1893, but differs from this species in having five antero-lateral spines, in-
stead of four, by having the first leg extending past the merus of the
second, instead of not extending past the merus, by the suborbital lobe
being truncate, instead of bilobed, and not equally advanced with the
pterygostomian lobe.

This species is named for Mr. Templeton Crocker’s yacht Zaca of San
Francisco, California, on board of which members of the New York
Zoological Society made an expedition into the Gulf of California during
1936 and collected the specimens described in the present paper.

Smith , Coates & Strong; Neoplasia in Tropical Fishes

219

18.

Neoplastic Diseases in Small Tropical Fishes1.

G. M. Smith,

Department of Anatomy, Yale University School of Medicine,

C. W. Coates,

New York Aquarium,

&

L. C. Strong,

Department of Anatomy, Yale University School of Medicine.

(Plates I-III).

Although many thousands of the smaller tropical fishes have been un-
der observation at the New York Aquarium during the past five years,
representing approximately 400 species, it is of interest to note that in only
five species has a tendency toward neoplasia been observed to date. These
species were the following:

1. Hybrids of the Mexican killifish, Xiphophorus helleri

Heckel and Platypoecilus maculatus Guenther.

2. Rasbora lateristriata (Bleeker).

3. Rasbora trilineata Steindacher.

4. Rasbora daniconius (Hamilton-Buchanan; .

5. Heterandria formosa Agassiz.

Before 1875 tumors in fishes were practically unknown. It was be-
lieved by many early pathologists that tumors were a characteristic of man
and the warm blooded animals in general, and that such new growths did not
occur in cold blooded creatures. It was Bugnion (1875) who is usually
credited with the earliest precise description of a fish tumor. This growth
was a giant cell sarcoma and occurred in the small European fish called the
Ellritze — Phoxinus laevis. With the end of the last century and the begin-
ning of the new, greater interest manifested itself in the study of tumors
in fishes, due largely to the widespread effort to learn about the origin of
cancer. As the result of the careful studies of Plehn (1906), Johnstone
(1915), Takahashi (1929), Thomas (1931), Haddow and Blake (1933) and
other investigators, various types of tumors in many species of fishes were
described. But even today our knowledge of fish tumors remains very in-
complete. There is little information regarding pathogenesis and course of
these hyperplastic lesions. We have no idea about the factors which, on
the one hand favor the growth of these tumors and, on the other, inhibit
their progress. One difficulty has been that relatively few tumors come
under observation, as it is well known that fishes succumb rapidly to any
form of disease and sick fishes are attacked and destroyed by other fishes.

Some peculiarities of fish tumors are now recognized. For example,

1 From the Department of Anatomy, Section Neuro-anatomy, Yale University School of Medi-
cine, and the Laboratory of the New York Aquarium.

220

Zoologica: New York Zoological Society

[XXI: 18

they remain very much localized and seem to be less infiltrative and de-
structive than mammalian and avian tumors. Transplantation of tumor
tissue from fish to fish, though frequently attempted, has been uniformly
unsuccessful, even when this has been undertaken with great care and ex-
acting technique. The occurrence of secondary growths or metatases, so
frequently seen in mammals and in birds, is rarely encountered in the case
of fishes. This has been thought to be due to the fact that the lymphatic
system of the fish is merely a diffuse arrangement of capillaries and spaces
possessing no organized lymphatic glands (Haddow and Blake, 1933). Sec-
ondary tumors in fishes seem to be the exception even with such growths as
the black pigmented melanoma, which in the human being is most malignant
and causes many widely scattered secondary growths followed by rapid
death. Owing to the numerous cutaneous pigment cells, containing granules
of different colors (black, red, yellow, etc.), the skin of fishes has proved
unique for the study of pigmented skin tumors. Some of these show a
marked difference from the skin tumors arising in terrestrial animals.
Thus, beside the commoner grayish-white fibromas of the skin, there are
known to be greenish-yellow tumors of the skin, red tumors or erythro-
phoromas, black tumors or melanomas, and even a silvery irridescent tumor
composed of cells containing irridescent crystals. The last tumor has been
described by the Japanese investigator Takahashi (1929), and he has named
this a guanophoroma. All these colored cutaneous tumors in fishes are
highly prized at the present time, because some of them occur in the ex-
tensively inbred small tropical fishes. They are furnishing an accurate
means of studying certain genetic principles underlying the causes of tumor
growth and cancer.

A neoplastic disease designated by Reed and Gordon (1931) as “Mela-
notic Over-growth,” is shown in Plate I, Fig. 1. It occurs in the hybrids
of the small Mexican killifish. Several instances of this condition have come
under our own observation. The disease has been studied extensively from
the genetic standpoint by the European investigators Haiissler (1928) and
Kosswig (1929) as well as by the Americans Reed and Gordon (1931). The
condition is characterized by a great multiplication of the black pigment
cells or melanophores which normally lie in small collections in the derma
directly under the transparent epidermis. With massing of these newly
formed melanophores, the affected part of the skin turns an intense black.
Very large collections of melanophores may actually form tumors which
penetrate into deeper lying muscles and destroy adjacent tissues. Thus, in
the later stages of the disease, the tail and the tissues at the base of the
tail may become invaded, destroyed and lost by a process of sloughing.
(PI. I, Fig. 2).

In another of the hybrids of Platypoecilus maculatus and Xiphophorus
helleri, which is highly pigmented in red, there arose a brick red tumor in
the region of the dorsal fin. (PL I, Fig. 3). This small mass had its origin
in a field of red pigmented cells (erythrophores) seen as an irregular red
patch in the mid dorsal region. The tumor invaded the skin and the tissues
of the dorsal fin, but not the deeper lying muscles. Histologically, the cells
composing the tumor are rather large, round, oval, or somewhat fusiform
(PI. II, Fig. 5). They show at times small protruding branches or dendrites.
The nucleus of the cell is small, and usually centrally located. The body of
the cells, prepared by the method of frozen sections, contained minute
granules of a dark red or orange color. These granules assumed a bright
red stain in the presence of Scharlach R., and were regarded as lipoidal
in character. In sections prepared by the paraffin method, the pigment
granules become dissolved, and the cytoplasm of the cells appear shrunken
and vacuolated. The red or orange pigment granules give the characteristic
red tint to the tumor during life. No black pigmented cells were noted in
any of the sections examined. The supporting tissue of the tumor cells

1936] Smith , Coates & Strong: Neoplasia in Tropical Fishes 221

are very delicate connective tissue cells which support a fine network of
small capillaries. Nerves were not noted. The overlying epithelium of the
skin consists of several layers of cells, but without any appreciable thick-
ening.

Erythrophoromas in fishes are distinctly rare. Thomas (1931) of
France described three such tumors in relatively large fishes. One occurred
in a tunny fish, and the other two were in the trout. More recently a
metastasizing cutaneous erythrophoroma was noted in a winter flounder
caught in the waters of Long Island Sound (Smith, 1934). Two instances
of erythrophoroma were produced experimentally by Kosswig (1929) by
cross breedings between Platypoecilus maculatus and Xiphophorus helleri.

Discussion on the Genetics of Fish Tumors.

(L. C. Strong).

Susceptibility to spontaneous tumor formation has been very completely
demonstrated, on a genetic basis, by Reed and Gordon (1931) in hybrids of
the Mexican killifish. They showed, in the first place, that the production
of macromelanophores (the histological unit involved in the formation of
the larger black areas of the fish) is inherited as a sex-linked dominant
factor; secondly, that another factor, an intensifier, found in Xiphophorus
helleri, is also inherited as a dominant; thirdly, that both factors in their
original location manifest themselves in an orderly manner; lastly, that if
both determiners are introduced into a hybrid fish, “macromelanophore in-
vasion results in a state of general melanosis in which there is a deteriora-
tion and final complete replacement of normal tissues by the invading cells.
A state is reached where there occur sharply delineated overgrowths, the
final tissue of the affected part is clearly a neoplasm.” Thus it was deter-
mined that hybridization, used for the production of new varieties by recom-
binations of variations in diverse species, had actually given a combination
of genetic determiners that led invariably to tumor formation. The very
important question arises whether these same factors in the original species,
not influenced by other determiners from other species, could, by themselves,
produce an abnormal growth such as a tumor. The problem presents itself
to a geneticist, ‘‘Has something gone wrong with the genetic factor or
factors which control the presence and physiological activity of the erythro-
phore (red pigmented cell) in this fish where an erythrophoroma was pro-
duced?” The question is an important one not only from the standpoint of
the fish fancier but also to the student of neoplasms in general. The prob-
lem needs further investigation. It could be worked out just as completely
as Reed and Gordon have done with the black tumors by the following pro-
cedure: 1. The histological examination of fishes with red tumors received
by a pathologist when the fishes are still alive, and (2) the statistical
analysis of the descendants of a fish known to have a red tumor at some
time during its life. It is hoped that fish fanciers may become interested
in the presence of tumors in their stocks and that they will make available
to other investigators such material as may come to their attention.

Contrasting with these neoplasms in hybrids of the Mexican killifish
above discussed was a small solitary fibrous growth seen in Rasbora dani-
conius (Hamilton-Buchanan) , (PL I, Fig. 4). This was a spontaneous
tumor in the region of the dorsal fin, sharply circumscribed, black in color
externally where the melanophores of the derma have aided in forming a
capsule around it. On sectioning the tumor, however, it was noticed that
the black color was strictly limited to the periphery of the growth, and
that the interior was composed of firm white fibrous tissue. Microscopically

222

Zoologica: New York Zoological Society

[XXI :18

the tumor is composed of loosely arranged interlacing bundles of elongated
connective tissue cells, in places hyaline in character (PI. II, Fig. 6). The
blood vessels are very few in number. The tumor appeared to belong to the
group of fibromas and was regarded as benign in nature. Melanophores
were irregularly scattered over the surface of the tumor, and the overlying
epidermis was destroyed. The melanophores were doubtless corial in their
origin and did not constitute a part of the tumor as such.

Epithelial and glandular tumors in this small series of neoplasms were
represented by two hyperplastic growths of the thyroid gland, the first
occurring in Rasbora lateristriata (PL III, Fig. 9) and the second in
Heterandria formosa Agassiz, (PI. Ill, Fig. 10). Microscopically, these
tumors resemble each other fairly closely in that they are composed of
masses of thyroid tissue, partly in a dense compact arrangement of cells
without acini or follicles, partly in closely grouped follicles distended with
varying amounts of colloid material (PI. II, Fig. 7). Both tumors in the
extension of their growth have encroached upon the structures of the gills,
so that the thyroid tissue replacing in part the epithelium of the gills
lies in closest contact with bone, cartilage, and muscle tissue. There is no
tendency to form a capsule.

To what extent the invasion of the gill structures can be regarded as
an evidence of malignancy in the case of thyroid tumors in fishes, is still a
matter of individual interpretation. It is not unlikely, however, that such
an extension beyond normal topographical limits on the part of a massed
growth of thyroid tissue encroaching upon bone, cartilage, muscle and
epithelium, may actually represent neoplasia of the thyroid gland of a
varying degree of malignancy. In both our fishes, the growth regarded
as adenocarcinoma, was a spontaneous one, forming an appreciable sized
tumor without metastases. Other fishes in these same tanks were not
affected in any way, strengthening the belief that the disease was not in-
fectious. Similar tumors occurring in other fishes, notably in the trout,
have been designated as adenocarcinoma. Widely studied in Europe and in
this country by Gaylord and Marsh (1914), Marine and Lehnhart (1910),
the disease has been attributed to an unknown agent in the water, perhaps
of an infectious nature, causing a disturbance in nutrition or metabolism.
The disease has not been transmitted by transplanting the diseased thyroid
tissue into other fishes of the same order.

The last tumor of the present series of tumors in small tropical fishes
occurred in Rasbora lateristriata (Bleeker). This fish developed a swelling
in the upper abdominal region near the liver. The intra-abdominal swelling
could be seen distinctly during life through the semitransparent abdominal
wall of the fish. After approximately two months of observation, the swell-
ing extended cephalically, and presently it was noticed that the right
operculum was pressed outward by a mass of whitish tissue involving the
gill on the right side (PI. Ill, Fig. 11). The health of the fish became
impaired and it was then sacrificed for the purpose of histological study.
Serial microscopic sections were prepared of the entire head and the ab-
domen.

The tumor tissue is composed of closely packed small lymphoid cells,
with deeply staining nuclei and scant cytoplasm. There are many small
necrotic areas where the determination of the structure of the tissue is
difficult. Here and there, intermingling with the masses of small lymphoid
cells, are larger paler vacuolated cells with vesicular nuclei, which suggest
structurally a larger type of lymphoid cell or a degenerating form of lym-
phoid cell. The exact site of origin of the tumor cannot be determined
by the microscopic study of the sections. The principal mass lies in the
abdomen ventral to the liver, spleen, kidney, gastro-intestinal tract and
ovarian tissue, yet in closest relation with all these organs. Plate II, Fig. 8

1936]

Smith, Coates & Strong: Neoplasia in Tropical Fishes

223

shows the tumor close to a part of the stomach. There is a distinct fusion
between tumor and the ventral peritoneum at one point, suggesting that
the neoplasm had its beginning in lymphatic tissue near the peritoneum.
The tumor was regarded as a lymphosarcoma with invasive malignant
properties. It extended as an irregular mass from the abdomen in a
cephalic direction, passing to the right of the heart to the roof of the oral
cavity, reaching outward and to the right to involve and destroy in large
part the tissues of the gill. There was also an extension of the tumor into
the cranial cavity, where it reached as far as the base of the brain and
laterally to the lower boundaries of the right and left saccule of the audi-
tory apparatus. The tumor encroached upon muscle, bone, and cartilage
wherever these structures were encountered in its course, causing in places
their destruction.

We found no parasites in tumor tissue, although several small encysted
parasitic larvae were noted in the abdomen at points remote from the
growth. The examination of the blood found in cross sections of the heart
and the larger blood vessels indicated no increase in the white cells, such
as might be expected in a leukaemic state if this were present.

Summary.

In this paper several neoplastic diseases are described occurring in
certain species of the small tropical fishes. These are (1) “Melanotic
Over-growths” and a red pigmented tumor (erythrophoroma) in hybrids of
Xiphophorus helleri Heckel and Platypoecilus maculatus Guenther; (2)
Fibroma in Rasbora daniconius (Hamilton-Buchanan) ; (3) Adenocarcinoma
of the thyroid in Rasbora trilineata Steindacher and Heterandria formosa
Agassiz; (4) an extensive lymphosarcoma in Rasbora lateristriata (Bleek-
er). The first group of pigmented cutaneous tumors in the hybrids of the
Mexican killifish are discussed from the genetic viewpoint.

Bibliography.

Bugnion,

1875. Deutsche Zeitschrift fur Tiermedizin und vergl. Pathologie, Bd. 1:132.
(Quoted by Plehn).

Gaylord, H. R. and Marsh, M. C.

1914. Carcinoma of the Thyroid in the Salmonoid Fishes. Publications from
State Institute for the study of malignant disease. Serial no. 99 issued
April 22, 1914. Page 363.

Haddow, A. and Blake, I.

1933. Neoplasms in Fish: a report of six cases with a summary of the lit-
erature. Jour. Pathology and Bacteriology, 36:41.

HAiissLER, G.

1928. Ueber Melanombildungen bei Bastarden Xiphophorus helleri und Plat-
ypoecilus maculatus var. rubra. Klin. Wochenschrift, 7:1561.

Johnstone, J.

1915. Diseased and abnormal conditions of marine fishes, Report on the
Lancashire Sea Fisheries Scientific Investigation for 1914. Trans.
Biol. Society of Liverpool, 29:18.

Kosswig, C.

1929. Zur Frage der Geschwiilstbildung bei Gattungbastarden der Zahn-
karpfen Xiphophorus und Platypoecilus. Zeitschrift f. lnduktion Ab-
stamm. und Vererbungslehre, 52:114.

224

Zoologica: New York Zoological Society

Marine, D. and Lehnhart, C. H.

1910. Observations and experiments on the so-called carcinoma of brook
trout (Salvelinus fontalis) and its relation to ordinary goitre. Jour.
Exp. Med., 12:311.

Plehn, M.

1906. Ueber Geschwfilste bei Kaltblfitern, Zeitschrift fiir Krebsforschung,
4:1.

Reed, H. D. and Gordon, M.

1931. The morphology of melanotic overgrowths in hybrids of the Mexican
killifish, American Journal of Cancer, 15:1524.

Smith, G. M.

1934. A cutaneous red pigmented tumor (erythrophoroma) with metastases
in a flat fish (Pseudopleuronectes americanus), American Journ. of
Cancer, 21:596.

Takahashi, K.

1929. Studien fiber die Fischgeschwfilste. Zeitschrift fiir Krebsforschung,
29:1.

Thomas, L.

1931. Les tumeurs des poissons (Etude anatomique et pathogenique) , Bulle-
tin de V Association Frangaise pour V etude du cancer, 20:703.

EXPLANATION OF THE PLATES.

Plate I.

Figs. 1 and 2. Melanotic overgrowth occurring in hybrids of the Mexican killifish.

Fig. 1 shows a moderately severe black pigmentation of parts of the skin,
while Fig. 2 shows an advanced lesion with a tumor-like mass near the
base of the tail which has been almost completely destroyed with the
progress of the disease.

Fig. 3. A red pigmented tumor (erythrophoroma) growing in a hybrid of the
Mexican killifish Platypoecilus maculatus and Xiphophorus helleri. The
growth involves the region of the dorsal fin. Drawing made from living
fish.

Fig. 4. Rasbora daniconius, with a small fibroma growing on the back of the
fish near the dorsal fin. The fibroma is surrounded with a capsule con-
taining black pigment cells. Drawing made after death of the fish.

Plate II.

Fig. 5. Photomicrograph of erythrophoroma occurring in fish represented in
Plate I, Fig. 3. Large, round, oval or fusiform cells with relatively
small nucleus. These cells in the fresh condition contained granules of
red or orange pigment, which gave the red tint to the tumor, x 230.

Fig. 6. Photomicrograph of tumor occurring in Rasbora daniconius, depicted in
Plate I, Fig. 4. The tumor is composed of interlacing bundles of elon-
gated connective tissue cells. A moderate number of black pigmented
cells, called melanophores M, are shown in the periphery of the growth,
x 80.

Fig. 7. Photomicrograph of thyroid tumor occurring in fish depicted in the pho-
tograph Fig. 9. A, compact thyroid growth; B, an area of the growth
showing thyroid tissue arranged in follicles; C, thyroid tissue infiltrat-
ing the structure of the gills, G. x 85.

Fig. 8. Photomicrograph of lymphosarcoma X in relation to area of the stomach,
S. x 200.

Plate III.

Fig. 9. Rasbora lateristriata showing neoplasm of the thyroid.

Fig. 10. Heterandria formosa with neoplasm of the thyroid.

Fig. 11. Photograph of the ventral side, Rasbora lateristriata. A, indicates mass
occupying the abdomen, and B, the extension of the growth to the gill
of the right side.

SMITH, COATES & STRONG.

PLATE I.

FIG. 1.

FIG. 2.

FIG. 3.

FIG. 4.

NEOPLASTIC DISEASES IN SMALL TROPICAL FISHES.

SMITH. COATES & STRONG.

PLATE II.

FIG. 6.

FIG. 7.

FIG. 8.

NEOPLASTIC DISEASES IN SMALL TROPICAL FISHES.

SMITH, COATES & STRONG.

PLATE HI.

FIG. 10.

FIG. 11.

NEOPLASTIC DISEASES IN SMALL TROPICAL FISHES.

jSetu JJorfe Zoological Society

General Office: 101 Park Avenue, New York City

Officers

President, Madison Grant

Vice-Presidents, W. Redmond Cross and Kermit Roosevelt
Chairman, Executive Committee, Madison Grant
Treasurer, Cornelius R. Agnew
Secretary, Henry Fairfield Osborn, Jr.

W. Reid Blair, Director
William T. Hornaday, Director Emeritus
Raymond L. Ditmars, Curator of Mammals and Reptiles
Lee S. Crandall, Curator of Birds
Charles V. Noback, Veterinarian

Claude W. Leister, Ass’t to the Director and Curator, Educational Activities

H. C. Raven, Prosector
Edward R. Osterndorff, Photographer
William Bridges, Editor and Curator of Publications

Aquarium

Charles H. Townsend, Director
C. M. Breder, Jr., Assistant Director

department of tropical &esiearcfj

William Beebe, Director and Honorary Curator of Birds
John Tee-Van, General Associate
Gloria Hollister, Research Associate

Scientific Staff

Zoological $arfe

Ctntorial Committee

Madison Grant, Chairman

W. Reid Blair
William Beebe

Charles H. Townsend
George Bird Grinnell

William Bridges

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS
OF THE

NEW YORK ZOOLOGICAL SOCIETY

VOLUME XXI
Part 4

Numbers 19-23

PUBLISHED BY THE
THE ZOOLOGICAL PARK,

December 31, 1936

SOCIETY
NEW YORK

CONTENTS

PAGE

19. The Southwestern Desert Tortoise, Gopher us agassizii. By

Chapman Grant . 225

20. Plankton of the Bermuda Oceanographic Expeditions. VII.

Siphonophora Taken During the Year 1931. By Cap-
tain A. K. Totton 231

21. The Female Bitterling as a Biologic Test Animal for Male

Hormone. By Israel S. Kleiner, Abner I. Weisman,
Daniel Mishkind & Christopher W. Coates. (Plate
I; Text-figure 1) 24l

22. Some Tropical Fishes as Hosts for the Metacercaria of

Clinostomum complanatum (Rud. 1814) (= C. mar-
ginatum Rud. 1819). By Ross F. Nigrelli. (Plates I

& II) 251

23. Caudal Skeleton of Bermuda Shallow Water Fishes. I.

Order Isospondyli: Elopidae, Megalopidae, Albulidae,
Clupeidae, Dussumieriidae, Engraulidae. By Gloria

Hollister. (Text-figures 1-53) 257

Index to Volume XXI 291

Grant : The Southwestern Desert Tortoise

225

19.

The Southwestern Desert Tortoise, Gopherus agasshii.

Chapman Grant.

Several enterprising youths in the vicinity of Hodge, San Bernardino
County, California, have started commercializing the desert tortoise,
Gopherus agassizii. May 8-11, 1935, was spent examining several small
collections and Robert Heckly’s two hundred specimens which were for
sale to tourists. On another trip, October 2-9 inclusive, one hundred more
specimens were examined. Most were in captivity, some were captured and
some shells of dead specimens were studied. The favorable hours during
which the tortoise ventures abroad were devoted to field work. Some ob-
servations that have not been seen in print are offered, together with addi-
tional data which may prove of interest.

Asymmetry.

Examination of plastrons showed that the suture between gulars is
offset to the right from one to five millimeters, making the left gular wider.
It does not protrude farther, but has a wider base and grows farther back
on the plastron. Among 366 specimens, 331, or 90%, had the suture on
the right, 23, or 6%, were median and 11 females and 1 male, or 3%, had
a left suture. The 23 with central sutures were about evenly divided be-
tween the sexes. Plate XXIV of Ditmars’ “The Reptile Book” (1907) illus-
trates the asymmetry. Additional evidence is shown in 15 plates by 0. P.
Hay in “The Fossil Turtles of North America.” The left gular is larger
in 7 plates and in no case is the right larger. Asymmetry of the gular scutes
is in contrast to the median suture of the symmetrical underlying epiplas-
tral bones. The gulars are either forked or worn chisel-shaped and re-
curved to a varying degree without apparent reference to age.

Dimorphism.

In the males secondary sexual dimorphism consists of larger size, longer
gulars and tail, smaller pelvic clearance measured from seam of anals to
edge of rear marginals, thickened anals and concave plastron, characters
which make their appearance as the juveniles exceed 12 or 14 cm., with the
difference in tail length showing first. The sexes have similar proportions
except that the male exceeds by 2% in length of gulars and 5% in tail
length, with a 2% smaller pelvic clearance. Among 366 specimens 19
females, 7 males and 4 young had major scute atavisms or abnormalities.
One female was nearly circular, having narrow costals; another was elon-
gated, having extra scutes, and a third was exceptionally high. The young
are light, but color apparently has no sexual significance. Two partial
albinos were seen with carapace, legs and toenails olive gray. The legs of

226 Zoologica: New York Zoological Society [XXI: 19

one large male were orange and black instead of the usual gray and black.
Varying recurvature of the marginals over the hind legs and the amount
of curving-under of the pygal as well as the considerable variation in num-
ber and arrangement of the inguinals are without apparent sexual signifi-
cance. The color of the iris is brassy or brown or these two colors mottled,
with about 80% brassy, 10% mottled and 10% brown without correlation to
sex or age.

Dimensions.

Table of dimensions in cm.

30 adult
females

Largest

female

30 adult
males

Largest

male

Carapace length ....

24.27

28.6

28.3

33.7

Carapace depth ....

10.4

11.5

12.

14.6

Greatest width

19.8

22.9

23.17

26.7

Plastral concavity. .

1.47

1.8

Gular length

4.17

' 5.3

5.47

7.2

Pelvic clearance ....

3.6

4.3

3.9

4.5

Tail length

2.27

2.8

4.00

5.8

Weight in gins

2,765.

5,809.

Description.

Occasionally an old female is found with the caudal worn flat or even
cut through, leaving a semicircular nick, as a result of being rubbed by
the anal plates of the male. Growth rings are usually visible on full grown
specimens, but disappear with age. Scutes become concave by thickening
along their edges when age is attained. Occasionally spines about as long
as the toenails, extending around the heel, give a starlike contour to the hind
foot. The retracted thighs are protected by pointed tubercles, usually in
the form of a circle around a large central spine.

Habits.

One old male was seen with tail hanging down and slowly swinging
from side to side with each elephantine step. A large male pushed himself
along the ground carrying Robert Heckly and one persisted in crawling-
under the rung of my chair and tilting me. The little Heckly girls staged
“roller skating” races, each standing on two tortoises.

Males utter grunting sounds while courting and recognize one another
immediately, possibly by scent which may come from glands beneath the
bulbs of the jaws. The long gulars are used by fighting males, which
charge one another, their heads retracted. When one manages to ram his
protruding gulars into the groin or anterior opening of the opponent’s shell,
he lifts and twists, often upsetting his antagonist. An overturned tortoise
flaps one front leg violently, stopping it suddenly at the most forward
position, the momentum jolting it slowly around so that any fixed object
within reach may be utilized to right itself. Males may perish if they
cannot turn over, since the hot sun is soon fatal. When asleep they sprawl
legs and necks in the most grotesque positions, appearing dead. They be-
come tame and inquisitive and seem to enjoy having their heads rubbed.

Fantastic stories of homing instincts may contain a grain of truth as
the tortoises usually return to the same burrow nightly. Accounts of treks

1936]

Grant: The Southwestern Desert Tortoise

227

or migrations and turtle towns are prevalent. I once saw what appeared
to be a migration of G. berlandieri in southern Texas and have noted that
tortoises are numerous only in restricted localities.

Robert Heckly furnished the following data: “In June, 1934, each of
three females laid five eggs and one laid two in holes dug by the hind feet.
It took an hour to dig the hole and fifteen minutes to lay the eggs. After
depositing an egg, a foot was thrust into the hole, seemingly to roll the
eggs from side to side or to pack them in. The hole was then filled and cov-
ered by the same process. Four from one female hatched in November, the
fifth hatched the following March. The rest were infertile. By October the
adults feed infrequently and by December have stopped entirely. They are
placed in the cellar under canvas until March when they are brought out
during the heat of the day to begin eating and mating, which is over by
the last of April. They lay the latter part of June. They have been known
to eat dried jackrabbit meat.”

Tortoises are eaten during lean years, but no one seemed enthusiastic
about the meat. I found the shells of three near the ashes of a small fire
in the desert with evidence that the largest carapace had been used for a
cooking vessel — an Indian custom. Escaped or liberated specimens are
found with a hole drilled in the edge of the shell or a portion of the edge
broken out.

Mr. Thomas Hallinan wrote an interesting article on G . polyphemus of
Florida. Contrasts between the two species are that the present species
makes no effort to escape by entering the burrow even if at the entrance,
whereas the Florida species scrambles to escape. The burrow is much
less steep, usually 12 degrees, starting against a small bank of wind-swept
sand at the base of a bush, and humidity is disliked. In Florida the bur-
rows are in damp ground and average fourteen feet long. In California
the four-foot burrows are much straighter, the occupant frequently being
visible. Never more than one tortoise and no associates, comensals or
parasites were found in any burrow in California, whereas the Florida
species has associates and parasites. Captive specimens at Thermal are
subject to very active ticks which attack the sutures of the carapace instead
of the skin and do not attach themselves.

Water Storage.

Persons in the desert say that the tortoise stores water in sacs in each
side of the body. The earliest reference noted is by E. T. Cox in the
American Naturalist of 1881. He stated that while preparing a museum
specimen he found on each side a large membranous sac filled with clear
water. Dr. R. D. Harwood of San Diego State College kindly dissected two
specimens for verification. He found the body cavity full of coelomic fluid
and a large, simple, urinary bladder of delicate membrane lying ventrally
across the plastron. The sides of the bladder, tucked up between the intes-
tines and marginals on each side, have apparently been mistaken for inde-
pendent sacs. Dr. Miller thinks that water is produced by metabolism.

Young.

A captive tortoise laid six eggs at Victorville on June 9, 1935. On
October 6, I purchased three of these eggs: one pipped, one nearly ready to
hatch and one infertile. The remaining eggs were not for sale, but were
sufficiently uncovered to see that they were fertile. The three eggs weighed
and measured as follows: pipped egg 27.2 gm., 42 x 36.5 mm; fertile egg
30.2 gm., 42 x 37.5 x 35.7 mm ; infertile egg with a large air bubble, 27.7 gm.,
42 x 36 mm. The shell is hard and unyielding and when blown weighs

228

Zoologica: New York Zoological Society

[XXI: 19

3.97 gm. Unfortunately the fertile egg was broken but the embryo was
preserved. It was not ready to hatch, having blood vessels surrounding
the white and the large yolk. Both this and the living specimen show great
scute abnormality. No evidence was found that the eggs are buried singly
as stated by Dr. Miller. The embryo lies partly curled and at right angles
to the long axis of the egg. It has an egg tooth at the center of the snout
with which it makes a ragged opening, parallel to the long axis. The egg
tooth apparently is not shed, but flattens out after several months. The
head is presented at 45 degrees, the eyes closed, but it opens its mouth
threateningly if touched. The hatchling is wider than long, but unrolls and
appears nearly normal in five days. A dozen captive babies were observed,
all showing pugnacity, bucking forward with open jaws and hissing if
touched on the carapace.

The baby hatched October 7, bearing a yolk sac and a small mass of
clear, thick, jellylike substance under the plastron. The jelly dried up and
the yolk was absorbed in two days. The baby measured in mm:

1935

1936

Oct. 7

Oct. 8

Oct. 11

June 18

Aug. 15

Length

36

41

44

44.5

48

Width

39

37.5

36.5

36.5

42.5

Weight

19.7

It drank frequently, but ate almost nothing until May. Efforts to feed
were clumsy at first, but by August it was eating tender vegetables, bananas,
bread, dry grass with a generous proportion of dirt and slacked lime. It
started to grow rapidly in July.

Mrs. Pajanew at Cinco had four that hatched on September 2. One had
been sold, but on October 4 the remaining three weighed 23, 23.5 and 25.5
gm. Mr. G. W. Hilton near Coachella had one that hatched on September 15
and his son, John, of Thermal, reported numerous hatchings about August
15. Young specimens are soft, but when 10 cm. in length, or about five years
old, seem able to resist carnivores, as several were seen with tooth marks
on their shells.

Range.

The range was plotted on a map, using records from scientific books
and articles, from the writer’s observations and from verbal reports where
three independent observers of apparent sincerity gave the same locality.

The tortoise does not occur in the Coachella-Thermal district where the
temperature is probably higher than in their habitat, but Mr. Hilton’s tor-
toises breed and reproduce well, showing that range restrictions are not
due to temperature or low elevation.

The range was found to extend northward into the Panamint Valley
and the Shoshone area in Inyo County. There are no records west of the
Coachella Valley, so it does not approach San Diego County. It is reliably
reported from the Beatty-Bunkerville line southward in Nevada and in
Washington County, Utah, and the Phoenix-Florence-Tucson area of Ari-
zona. There are a few Sonora records and one from Tiburon Island. This
is a vast area, but great stretches are uninhabited.

In determining the range, tracks, burrows, droppings and dead shells
are more convincing proof than a live animal encountered in an unexpected
place, as there is always the likelihood that it may have escaped. At least
600 specimens were collected within two miles of Hodge in 1934-1935 ac-

1936]

Grant: The Southwestern Desert Tortoise

229

cording to the census of the boys selling them. By driving to likely places,
making a hurried reconnaissance and hunting only where tracks abounded
the writer was able to collect ten specimens in about six hours — some in
burrows and some feeding. They emerge in the forenoon and afternoon
during spring and fall.

Literature Cited.

Cox, E. T. 1881. Notes on the Tortoises of Tucson. Am. Nat., XV : 1003.

Ditmars, Raymond L. 1907. The Reptile Book. PI. XXIV, Doubleday, Page &
Co., New York.

Hallinan, Thomas. 1823. Observations Made in Duval County, Northern
Florida, on the Gopher Tortoise (Gopherus polyphemus). Cojieia, 115:11-20.

Hay, 0. P. 1908. The Fossil Turtles of North America. Carnegie Institution
Publication No. 75. All plates.

Miller, Loye. 1932. Notes on the Desert Tortoise ( Testudo agassizii) . Trans.
San Diego Soc. Nat. Hist., VII : 187-208.

Totton: Siphonophora Taken During 1931

231

20.

Plankton of the Bermuda Oceanographic Expeditions.

VII. Siphonophora Taken During the Year 19311.

Captain A. K. Totton, M. C.

British Museum ( Natural History).

Introduction.

This is one of a number of papers dealing with the planktonic contents
of nets drawn at definite levels and in a definite area off the south coast
of Bermuda on the Bermuda Oceanographic Expeditions of the New York
Zoological Society. The area chosen was a circle eight miles in diameter,
with its center located at 32° 12' N. Lat., 64° 36' W. Long., a point nine
miles southeast of Nonsuch Island, Bermuda.2 The depth at this locality
is 1,000 to more than 1,400 fathoms. Further details concerning the nets,
position, etc., will be found in Zoologica, Volume XIII, Numbers 1, 2 and 3.

Captain Totton has kindly identified the hundred-odd vials of siphono-
phores which I sent him, and Dr. Henry B. Bigelow and Dr. Mary Sears
have furnished the faunistic notes. My own share has been only the col-
lecting, and the gathering of the scanty field notes.

William Beebe.

Suborder Calycophorae.

Family Prayidae.

Subfamily Amphicaryoninae.

1. Amphicaryon acaule Chun 1888.

Material: No. 3174; Net 988; 1,000 fathoms; June 2.

Color: Lemon yellow in life.

Distribution : This species is widespread in the tropical and sub-
tropical belts of the great oceans (for localities see Totton, 1932, p. 330).
It was first described by Chun (1888) from the Canaries, but it was not
taken again in that general region until 1910, when the Thor took it in the
Bay of Cadiz (Bigelow and Sears, in press), and in 1913, when the Hiron-
delle II took it in 38° 58' N. Lat., 44° 55' W. Long. (Leloup, 1933). There
is only one record of it in the Mediterranean, made by the Thor.

1 Contribution No. 514, Department of Tropical Research, New York Zoological Society.

2 For diagram of trawling area, see Zoologica, Vol. XXI, No. 9, page 97.

232

Zoologica: New York Zoological Society

[XXI : 20

Subfamily Nectopyraminae.

2. Nectopyramis sp. nov?

Material: No. 31942; Net 1120; 400 fathoms; Aug. 3.

No. 311087; Net 1146; 600 fathoms; Aug. 7.

No. 311652; Net 1239; 900 fathoms; Aug. 29.

No. 311767; Net 1259; 1,000 fathoms; Sept. 3.

Subfamily Prayinae.

3. IPraya dubia Quoy and Gaimard 1834.

Material: No. 31509; Net 1063; 300 fathoms; July 8.

No. 311617; Net 1226; 300 fathoms; Aug. 27.

COLOR: Transparent and colorless in life.

Distribution : Previous records of P. dubia are from Australian

waters, from the eastern tropical Pacific (Bigelow, 1911; 1931), and off
Valparaiso (Moser, 1925).

4. ? Rosacea cymbiformis Delle Chiaje 1842.

Material: No. 31540; Net 1064; 100 fathoms; July 9.

Distribution : Rosacea cymbiformis has been recorded from all three
oceans — in the eastern Atlantic from the tropics northward to the Bay of
Biscay (Chun, 1888; Haeckel, 1888; Leloup, 1933), in the South Atlantic
(Leloup, 1934; Hardy and Gunther, 1935), in the Philippines (Bigelow,
1919), and in the Indian Ocean (Huxley, 1859; Browne, 1926). There are
also frequent records from the Mediterranean whence it was first de-
scribed.

Family Hippopodiidae.

5. Hippopodius hippopus Forskal 1775.

Material: No.

No.

No.

No.

No. 311252

31305; Net 1036;
31445; Net 1050;
31480; Net 1055;
31733; Net 1089;
Net 1173;

No. 311301; Net 1185;

No. 311473; Net 1201;

No. 311499; Net 1210;

No. 311534; Net 1214;

No. 311618; Net 1227;

No. 311963; Net 1293;

No. 312066; Net 1308;
Color: Transparent and colorless.
Distribution : This is one of the more
the warmer oceans (for summary, see Moser,

200

25

50

50

400

900

800

1,000

1,000

400

800

100

fathoms ;
fathoms ;
fathoms ;
fathoms ;
fathoms ;
fathoms ;
fathoms ;
fathoms ;
fathoms ;
fathoms ;
fathoms ;
fathoms ;

June 25.
July 6.
July 7.
July 18.
Aug. 14.
Aug. 15.
Aug. 19.
Aug. 20.
Aug. 21.
Aug. 27.
Sept. 12.
Sept. 16.

common species found in all
1925).

6. Vogtia glabra Bigelow 1918.

Material: No. 311199; Net 1169; 700 fathoms; Aug. 12.

Distribution : This species was originally described from the Straits
of Florida (Bigelow, 1918) and has since been taken in the eastern side of
the temperate Atlantic between the Azores, Canaries and the coast of
Portugal and the Gulf of Gascony, as well as in the Mediterranean (Leloup,
1933; Bigelow and Sears, in press). It has never been reported from other
oceans, as have other members of this bathypelagic genus.

1936]

Totton: Siphonophora Taken During 1931

233

Family Diphyidae.

Subfamily Abylinae.

7. Abyla dentata Bigelow 1918.

Material: No. 311652; Net 1239; 900 fathoms; Aug. 29.

No. 311767; Net 1259; 1,000 fathoms; Sept. 3.

Distribution : This species was first described from the western At-
lantic between Bermuda and the American coast (Bigelow, 1918). It has
only been noted since then near the Cape Verdes (Moser, 1925, as “A.
quadrata”) .

8. Abylopsis eschscholtzii Huxley 1859.

Material: No. 311040; Net 1133; 700 fathoms; Aug. 5.

No. 311534; Net 1214; 1,000 fathoms; Aug. 21.

Color: Dull crimson in life.

Distribution : This species is widespread over the tropical Pacific and
Malayan region (Bedot, 1896; Lens and Van Riemsdijk, 1908; Bigelow,
1911, 1931; Totton, 1932), the Indian Ocean (Browne, 1926), in the South
Atlantic (Moser, 1925; Leloup, 1934), also in the tropical Atlantic (Chun,
1888; Mayer, 1900; Leloup, 1934). It has also been taken in the Mediter-
ranean (Leloup, 1933).

9. Abylopsis tetragona Otto 1823.

Material: No. 31883; Net 1107 ; 400 fathoms; July 27.

Distribution : This is one of the commoner species of siphonophores
and is found throughout the warmer zones of all oceans (see Moser, 1925).

Subfamily Ceratocymbinae.

10. Ceratocymba sagittata Quoy and Gaimard 1827.

Material :

No. 31517;
No. 31920;
No. 311250;
No. 311724;
No. 312031;
No. 312067;
No. 312071;
No. 312079;
No. 312169;
No. 312171;
No. 312182;

Net 1062;
Net 1117;
Net 1174;
Net 1248;
Net 1305;
Net 1308;
Net 1308;
Net 1309;
Net 1322;
Net 1322;
Net 1330;

300 fathoms;
1,000 fathoms ;
500 fathoms;
600 fathoms ;
500 fathoms;
100 fathoms;
100 fathoms;
100 fathoms;
300 fathoms;
300 fathoms ;
1,000 fathoms;

July 8.
July 29.
Aug. 14.
Sept. 1.
Sept. 15.
Sept. 16.
Sept. 16.
Sept. 16.
Sept. 18.
Sept. 18.
Sept. 19.

Color: Transparent anteriorly; milky white posteriorly.

Distribution : This species occurs in the warm parts of the North and
South Atlantic (Moser, 1925; Leloup, 1933; 1934), the eastern tropical
Pacific (Bigelow, 1911), the Indian Ocean (Browne, 1926), and the Malayan
region (Lens and Van Riemsdijk, 1908), and in the Mediterranean, where
it was taken for the first time by the Thor (Bigelow and Sears, -in press) .

Subfamily Diphyinae.

11. Diphyes dispar Chamisso and Eysenhardt 1821.

Material: No.

No.

No.

No.

31258; Net 1020; 14 fathoms; June 15.

31304; Net 1035; 100 fathoms; June 25.
31356; Net 1037; 300 fathoms; June 25.
31364; Net 1040; 25 fathoms; June 26.

234

Zoologica: New York Zoological Society

[XXI : 20

No. 31409; Net
No. 31435; Net
No. 31481; Net
No. 31507; Net
No. 31484; Net
No. 31517; Net
No. 31598; Net
No. 31621; Net
No. 31643 ; Net
No. 31650; Net
No. 31732 ; Net
No. 31787; Net
No. 311211; Net
No. 311496; Net
No. 311571; Net
No. 311625; Net
No. 312005; Net
No. 312030; Net

1045; 25 fathoms;

1050; 25 fathoms;

1055; 50 fathoms;

1055; 50 fathoms;

1056 ; !50 fathoms ;

1069 ; 50 fathoms ;

1069 ; 50 fathoms ;

1075 ; 50 fathoms ;

1075; 50 fathoms;

1079; 50 fathoms;

1089; 50 fathoms;

1099 ; 900 fathoms ;
1169; 700 fathoms;
1206; 800 fathoms;
1218; 700 fathoms;
1230 ; 900 fathoms ;
1285; 800 fathoms;
1301; 50 fathoms;

June 27.
July 6.
July 7.
July 7.
July 7.
July 10.
July 10.
July 11.
July 11.
July 14.
July 18.
July 24.
Aug. 12.
Aug. 20.
Aug. 24.
Aug. 27.
Sept. 10.
Sept. 15.

Color: All transparent and colorless except No.
scribed as “siphonophore with long yellow chain.”

31650, which is de-

Distribution : This species has been recorded from the warm zones
of all oceans.

12. Chelophyes appendiculata Eschscholtz 1829.

Material: No. 311122; Net 1154; 700 fathoms; Aug. 8.

No. 312071; Net 1308; 100 fathoms; Sept. 16.

Distribution : This is the commonest of all siphonophores, and is
found in all oceans.

13. Chelophyes contorta Lens and Van Riemsdijk 1908.

Material: No. 311534; Net 1214; 1,000 fathoms; Aug. 21.

Distribution : This species, first described from the Malayan region,
has subsequently been found in the Indian Ocean (Moser, 1925; Browne,
1926), on both sides of the Pacific, off the Great Barrier Reef of Australia
(Totton, 1932), in the China Sea (Bigelow, 1913), in the eastern tropical
Pacific (Bigelow, 1911), and also in the South Atlantic (Moser, 1925).

14. Eudoxoides mitra Huxley 1859.

Material: No. 311199; Net 1169; 700 fathoms; Aug. 12.
Distribution : This species, first described from the Indian Ocean, is
well known in the Pacific (Totton, 1932; Bigelow, 1911, 1913), and in the
Atlantic (Moser, 1925; Leloup, 1933, 1934; Bigelow, 1918).

15. Eudoxoides spiralis Bigelow 1911.

Material: No. 311534; Net 1214; 1,000 fathoms; Aug. 21.

Distribution: This species is cosmopolitan in warm seas; the records
include the eastern tropical Pacific (Bigelow, 1911), Japanese waters (Bige-
low, 1913), off the Great Barrier Reef of Australia (Totton, 1932), the
Indian Ocean (Browne, 1926), many localities in the South Atlantic, south
to Latitude 45° S. (Moser, 1925; Leloup, 1934); likewise the tropical and
subtropical North Atlantic, northward to the vicinity of the Cape Verdes
(Leloup, 1934) on the one side and in the region of Cape Hatteras, Ber-
muda and the Bahamas (Bigelow, 1918) on the other. It has also been
taken in the Mediterranean.

1936]

Totton: Siphonophora Taken During 1931

235

16. Lensia conoidea Keferstein and Ehlers 1860.

Material: No. 3174; Net 988; 1,000 fathoms; June 2.

No. 311816; Net 1263; 800 fathoms; Sept. 4.

Color: Lemon yellow in life.

Distribution : Since L. fowleri and L. subtiloides have often been
treated as synonyms of L. conoidea, it is impossible to learn its range from
published accounts. There are definite records of it from the North Pacific,
the Malayan region, the Gulf Stream, and the coast of Norway. It is likely
that it occurred among the specimens listed as Utruncata” from the South
Atlantic (Moser, 1925; Leloup, 1934; Hardy and Gunther, 1935), and in the
Mediterranean (Moser, 1925; Leloup, 1933). It was taken in the latter
sea in abundance by the Thor (Bigelow and Sears, in press).

17. Lensia multicristata Moser 1925.

Material: No. 311415; Net 1195; 800 fathoms; Aug. 17.

No. 311780; Net 1258; 900 fathoms; Sept. 3.

Distribution : Lensia multicristata Moser is widespread in the eastern
tropical Pacific (Bigelow, 1911), in the Indian Ocean (Browne, 1926), in
the South Atlantic (Moser, 1925; Leloup, 1934), and as far north in the
North Atlantic as the Bay of Biscay (Bigelow, 1911a), as well as in the
Mediterranean (Bigelow and Sears, in press).

18. Lensia profunda sp. nov.s

Material: No. 311377; Net 1190; 900 fathoms; Aug. 16.

No. 311601; Net 1217; 600 fathoms; Aug. 24.

19. Lensia sp.

Material: No. 311534; Net 1214; 1,000 fathoms; Aug. 21.

20. ?Chuniphyes multidentata Lens and Van Riemsdijk 1908.

Material: No. 311199; Net 1169; 700 fathoms; Aug. 12.

No. 311510; Net 1207; 900 fathoms; Aug. 20.

No. 311968; Net 1291; 600 fathoms; Sept. 12.

No. 312008; Net 1298; 800 fathoms; Sept. 14.

Color: Transparent and colorless in life.

Distribution : The bathypelagic species, C. multidentata , was first de-
scribed from Malayan waters. Later, it has been recorded from the eastern
tropical Pacific (Bigelow, 1911, 1931), the offing of California (Bigelow and
Leslie, 1930), the Eastern Sea between Japan and China (Bigelow, 1913),
near the Philippines (Bigelow, 1919), and on both sides of the Atlantic —
south to South Georgia, and north to the Bay of Biscay (Bigelow, 1911a;
Moser, 1925; Leloup, 1933, 1934; Hardy and Gunther, 1935). There is only
one record of it within the Mediterranean (Leloup, 1933).

Suborder Physophorae.

Family Forskalidae.

21. Forskalia sp.

Material: No. 3110; Net 980; Surface; May 17.

No. 311300; Net 1185; 900 fathoms; Aug. 15.

3 See MS. Discovery Report.

236

Zoologica: New York Zoological Society

[XXI : 20

No. 311301; Net 1185; 900 fathoms; Aug. 15.

No. 311511; Net 1207; 900 fathoms; Aug. 20.

Color: No. 3110 was described as “lemon-colored,” No. 311300 as

“orange-red,” and No. 311511 as “black-lined.”

Family Agalmidae.

22. Agalma elegans Sars 1846.

Material: No. 31492; Net 1054; 25 fathoms; July 7.

No. 312117; Net 1314; 600 fathoms; Sept. 7.

Color: Transparent and colorless.

Distribution : This species is well known from the Mediterranean,
from the coasts of Europe as far north as Norway and along the eastern
coast of North America from Cape Cod to the West Indies,, and in the east-
tern tropical Pacific (Bigelow, 1911), as well as in Malayan waters (Bedot,
1896) and the Indian Ocean (Browne, 1926).

23. Agalma okeni Eschscholtz 1825.

Material: No. 31210; Net 1018; 900 fathoms; June 15.

No. 31314; Net 1038; 300 fathoms; June 25.

No. 31976; Net 1128; 400 fathoms; Aug. 4.

No. 311616; Net 1226; 300 fathoms; Aug. 27.

No. 312068; Net 1308; 100 fathoms; Sept. 16.

No. 312071; Net 1308; 100 fathoms; Sept 16.

No. 312085; Net 1310; 300 fathoms; Sept. 16.

No. 312088; Net 1311; 300 fathoms; Sept. 16.

No. 312170; Net 1322; 300 fathoms; Sept. 18.

COLOR: Nectophores transparent, siphosomes coral and white in life.

Distribution : This species occurs in the warmer regions of all the
great oceans and has occasionally been taken in the Mediterranean and
Red Sea.

24. Stephanomia amphitridis Peron and Lesueur 1807.

Material: No. 3152; Net 984;

No. 3153; Net 984;
No. 311013; Net 1133;
No. 311040; Net 1133;
No. 311091; Net 1143;
No. 311092; Net 1143;
No. 311112; Net 1149;
No. 311534; Net 1214;
No. 311859; Net 1274;
No. 312196; Net 1326;

600 fathoms; June 2.
600 fathoms; June 2.
700 fathoms; Aug. 5.
700 fathoms; Aug. 5.
500 fathoms; Aug. 7.
500 fathoms; Aug. 7.
500 fathoms; Aug. 8.
1,000 fathoms; Aug. 21.
900 fathoms; Sept. 7.
600 fathoms; Sept. 19.

Color: Varying from crimson to pink and orange.

Distribution : This species, originally described from the Atlantic, has
also been recorded from the Pacific, the Malayan region and probably from
Ceylon.

Subfamily Nectaliinae.

25. Nectalia loligo Haeckel 1888.

Material: No. 311040; Net 1133; 700 fathoms; Aug. 5.

No. 311113; Net 1149; 500 fathoms; Aug. 8.

1936]

Totton: Siphonophora Taken During 1931

237

No. 311494; Net 1205; 700 fathoms; Aug. 20.

No. 311570; Net 1218; 700 fathoms; Aug. 24.

No. 311571; Net 1218; 700 fathoms; Aug. 24.

No. 311673; Net 1236; 600 fathoms; Aug. 29.

No. 311688; Net 1243; 700 fathoms; Aug. 31.

No. 311968; Net 1291; 600 fathoms; Sept. 12.

No. 311984; Net 1291; 600 fathoms; Sept. 12.

No. 311964; Net 1292; 700 fathoms; Sept. 12.

No. 311966; Net 1292; 700 fathoms; Sept. 12.

Color: Crimson and transparent.

Distribution : Few specimens of this species are on record. The type
came from the Canary Islands and subsequent records are two specimens
taken by the Plankton Expedition in 3° 6' N. Lat., 33° 2' W. Long., and the
northern border of the Gulf Stream south of Iceland, one from Orotava, one
from the eastern tropical Pacific, and from the South Atlantic (Moser,
1925).

Subfamily Anthophysidae.

26. Anthophysa formosa Fewkes 1882.

Material: No. 311187; Net 1162; 800 fathoms; Aug. 11.

No. 311281; Net 1178; 900 fathoms; Aug. 14.

No. 311367; Net 1188; 500 fathoms; Aug. 16.

No. 312205; Net 1325; 500 fathoms; Sept. 19.

Color : Lemon yellow and white.

Distribution : This is the member of this genus found in the Atlantic,
and may eventually prove to be identical with A. rosea Brandt. It is known
from the Gulf Stream in the western side of the North Atlantic, from
the Sargasso Sea, from the South Atlantic, where it was taken by the
Challenger, from the vicinity of the Canaries and from the Mediterranean.

27. Athorybia rosacea (Forskal 1775) Eschscholtz 1829.

Material: No. 311008; Net 1136; 1,000 fathoms; Aug. 5.

28. Athorybia sp.

Material: No. 312071; Net 1308; 100 fathoms; Sept. 16.

Suborder Rhizophysaliae.

Family Rhizophysidae.

29. Rhizophysa sp.

Material: No. 31719; Net 1084; 25 fathoms; July 15.

Color: Pink.

Family Physalidae.

30. Physalia physalis Linne 1758.

Material: Often taken at the surface.

Distribution : This is the common Portuguese-man-of-war of the
warmer parts of the Atlantic Ocean.

238

Zoological New York Zoological Society

[XXI : 20

Suborder Chondrophorae.

Family Porpitidae.

31. Porpita umbella 0. F. Muller 1776.

Material: Occasionally taken at the surface.

Distribution : This species is the one found in the warm waters of the
Atlantic Ocean. It may prove to be the same as P. porpita Linne from the
Indian Ocean, but it is probably distinct from P. pacifica Lesson.

Family Velellidae.

32. Velella velella Linne 1775.

Material: Occasionally taken at the surface.

Distribution : This species is found in the warmer waters of the
Atlantic and Pacific. On further study, it may prove identical with the
Velella of the Indian Ocean.

Bibliography.

Bedot, M.

1896. Les Siphonophores de la Baie d’Amboine. Etude suivie d’une revision
de la famille des Agalmides. Revue Suisse. Zool. Vol. 3.

Bigelow, H. B.

1911. The Siphonophorae. Rep. Scientif. Results Exped. to the Eastern
Tropical Pacific. U. S. Fish Comm. Steamer Albatross 1904/05. Mem.
Mus. Comp. Zool. Vol. 38.

1911a. The Siphonophorae. Biscayan Plankton Collected during a Cruise of
the H. M. S. Research, 1900. Trans. Linn. Soc. Zool. Vol. 10.

1913. Medusae and Siphonophorae Collected by the U. S. Fisheries Steamer
Albatross in the Northwestern Pacific, 1906. Proc. U. S. Nat. Mus.
Vol. 44.

1918. Some Medusae and Siphonophorae from the Western Atlantic. Bull.
Mus. Comp. Zool. Vol. 62.

1919. Contributions to the Biology of the Philippine Archipelago and adjacent
regions: Hydromedusae, Siphonophores and Ctenophores of the Alba-
tross Philippine Expedition. Bull. U. S. Nat. Mus., 100, 1.

1931. Siphonophorae from the Arcturus Oceanographic Expedition. Zoo-
logica, vol. 8.

Bigelow, H. B.. and Leslie, M.

1930. Reconnaissance of the Waters and Plankton of Monterey Bay, July
1928. Bull. Mus. Comp. Zool., vol. 70.

Bigelow, H. B., and Sears, M.

? ? Thor Siphonophora. (In press).

Browne, E. T.

1926. No. Ill: Siphonophorae from the Indian Ocean. Trans. Linn. Soc.
Zool. London, vol. 19, no. 1.

Chamisso, A., and Eysenhardt, D. G.

1821. De animalibus quibusdam e classe vermium, etc., fasc., 2. Acad. Caes.
Leop. Nova Acta, vol. 10.

Chun, C.

1888. Bericht fiber eine nach den Canarischen Inseln im Winter 1887-88
ausgefuhrte Reise. Sitz. Akad. Wiss. Berlin, vol. 44.

1936]

Totton: Siphonophora Taken During 1931

239

Delle Chiaje, S.

1842. Descrizione e Notomia degli animali invertebrati della Sicilia citeriore,
vol. 5.

Eschscholtz, D.

1825. Bericht iiber die Zoologische Ausbeute wahrend der Reise von Kron-
stadt bis St. Peter und Paul. Oken’s Isis, pp. 733-747.

1829. System der Acalephen. Berlin.

Fewkes, J. W.

1882. Notes on Acalephs from the Tortugas, Bull. Mus. Comp. Zool, vol. 9.

Forskal, P.

1775. Descriptiones animalium . . . quae in itinere orientali observavit, post
mortem edidit Carstem Niebuhr.

Haeckel, E.

1888. Report on the Siphonophorae. Rep. Scientif. Results Voy. Challenger,
Zool., vol. 28.

Hardy, A. C., and Gunther, E. R.

1935. The Plankton of the South Georgia Whaling Grounds and Adjacent
Waters, 1926-1927. Discovery Reports, vol. 11.

Huxley, T. H.

1859. The Oceanic Hydrozoa.

Keferstein, W., and Ehlers, E.

1861. Zoologische Beitrage Gesammelt im Winter 1859-1860 in Neapel und
Messina. I. Beobachtungen uber die Siphonophoren. Leipzig.

Leloup, E.

1933. Siphonophores Calycophorides Provenant des Campagnes du Prince
Albert Ier de Monaco. Res. Camp. Sci. Monaco, vol. 33.

1934. Siphonophores Calycophorides de l’Ocean Atlantique Tropical et Aus-
tral. Bull. Mus. Hist. Nat. Belgique, vol. 10, no. 6.

Lens, A. D., and Van Riemsdijk, T.

1908. The Siphonophorae of the Siboga Expedition. Monogr., vol. 38.

Linne, C. yon

1758. Systema naturae . . . Ed. 10, vol. 1.

Mayer, A. G.

1900. Some Medusae from the Tortugas, Florida. Bull. Mus. Comp. Zool.
Harvard Col., vol. 37.

Moser, F.

1925. Siphonophora. In Handbuch der Zoologie, vol. 1. Berlin and Leipzig,
1923-1925.

Muller, O. F.

1776. Beschreibung zweier Medusen. Beschaft. Berlin Gesell. Naturf.
Freunde, vol. 2.

Otto, A. W.

1823. Beschreibung einiger neuer Mullusken und Zoophyten. Nova Acta. Caes.
Leop. Carol., vol. 11.

Peron, F., and Lesueur, C. A.

1807. Voyage de Decouvertes aux Terres Australes execute sur les Corvettes
le Geographe, le Naturaliste, et la Goelette le Casuarina, pendant les
annees, 1800-1804.

240

Zoologica: New York Zoological Society

Quoy, J. R. C., and Gaimard, P.

1827. Observations faites a bord de V Astrolabe dans le detroit de Gibraltar.
Ann . sci. nat., vol. 10. Also Oken’s Isis, vol. 21.

Sars, M.

1846. Fauna littoralis Norwegiae, vol. 1. Christiania.

Totton, A. K.

1932. Siphonophora. Sci. Reports Great Barrier Reef Exp., 1928-1929. Vol.
4, No. 10.

1936?. Siphonophora. Discovery Reports. (In press).

Kleiner, Weisman, Mishkind & Coates: Male Hormone Test

241

21.

The Female Bitterling as a Biologic Test Animal
for Male Hormone1.

Israel S. Kleiner, Abner I. Weisman, Daniel I. Mishkind,

Department of Physiology and Biochemistry of the
New York Medical College and Flower Hospital,

&

Christopher W. Coates,

New York Aquarium.

(Plate 1; Text-figure 1).

Introduction.

A number of different methods of testing for male hormone have been
proposed (1) with the capon, the guinea pig, the rat, and the castrated male
bitterling (2) as test animals. None of these tests has been very satisfac-
tory. Recently Witschi (3) has described a change in the color of the bill
of the sparrow after injections of male hormone and has suggested this as
an indicator. The capon test is the one most commonly employed. Injections
of potent material cause a growth of comb and wattles in four or five days.
The measurement is not simple nor is it standardized (4). Injections do not
produce constant results and for each test six animals should be used (5).
Thus a considerable amount of expense, space, care, and attention are re-
quired for a single test. Castration must be complete and the bird can not
be used again after a positive reaction until the comb has regressed.

It therefore seems evident that a more convenient method of detecting
and measuring the male hormone would be of considerable value. The
ovipositor lengthening of the female bitterling appears to furnish a simple
biological reaction for this hormone, as the present experiments will indi-
cate. This lengthening of the ovipositor of the female bitterling (PI. I, Fig.
1) was first elicited by Ehrhardt and Kuhn (6) and, shortly after, in-
dependently by Fleischmann and Kann (7), but they ascribed this reaction
to follicular hormones. It was later suggested as of possible use in diagnos-
ing pregnancy by Kanter, Bauer, and Klawans (8). Although these authors
did not definitely state that it was a test for pregnancy, their paper inti-
mated this use so clearly that a number of laboratories accepted this view-
point. It was soon shown that it could not be so considered (9, 10, 11),
since urines from non-pregnant females, women in the post-climacteric
period, and even from men gave positive reactions. Occasionally a preg-
nant women’s urine would be negative. That this test is produced by male
hormone will be shown below (12).

1 Aided by a grant from the Lucius N. Littauer Foundation.

242 Zoologica: New York Zoological Society [XXI: 21

Experimental.

The female bitterling2 develops an ovipositor which depends from the
ventral margin of the body slightly anterior to the origin of the anal fin.
In the quiescent state the ovipositor is seldom visible, but in those indivi-
duals in which it is visible out of season it is very small and rarely reaches
5 mm. in length. During breeding activity the organ is prolonged until it
may reach 5 cm. in length (PI. I, Fig. 2) and at spawning it is inserted into
the inhalent siphon of a mussel, usually Unio or Anodonta, and the ova
extruded into the gill-folds. Fertilization is effected by sperm which is
liberated near and drawn into the inhalent siphon of the mussel and passed
over the embedded ova. Hatching occurs within the gill-folds and the fry
liberate themselves, in a post-larval state, two or three weeks after ovi-
position.

Description of Test.

To a small aquarium is added 4 liters of water, 2 liters from a stock
tank and 2 liters from the tap. Two female bitterlings are placed in the
tank and kept there 24 hours before introducing the material to be tested.
At the end of the 24-hour period readings are made of the size of the
ovipositor. The scale used is as follows: If the ovipositor is not visible the
reading is 0 ; if the length of the ovipositor equals the length of the first
ray of the anal fin the reading is 10 ; an ovipositor which reaches halfway
down the first ray is 5; etc. (That is, the length of the ovipositor is
compared to that of the first ray of the anal fin in equal units running from
0 to 10. (Text-fig. 1).) No fish with ovipositors exceeding 3 on the scale,
before the addition of any test material, are used.

Text-figure 1.

Diagrammatic representation of scale used in reading length of ovi-
positor of female bitterling.

Ovipositor readings are taken at 24, 48, and 72 hour intervals and
maximum growths under these conditions are usually observed at the
second reading.

We tentatively define a bitterling unit as the amount of material which,
when added to a tank containing 4 liters of water and 2 female bitterlings,

2 We are greatly indebted to Mr. C. M. Breder, Jr., of the New York Aquarium, for advice
and suggestions in respect to the fish.

1936] Kleiner, Weisman, Mishkind & Coates: Male Hormone Test 243

produces an increase in the length of the ovipositor of one or both fish of
7 or more on the scale within 48 hours.3 For assay it is suggested that a
series of dilutions may be set up and the lowest dilution giving a positive
reading may be considered to contain at least one unit, providing a positive
reaction is also given in the next higher dilution.

In this work the European bitterling ( Rhodens amarus) was used. Ran-
ter et al (8) used the Japanese bitterling ( Acheilognathus intermedium)
and suggest differences between the two species. However, comparative
initial experiments with pregnancy urines conducted under identical condi-
tions at the same time with both species by Coates (unpublished) did not
seem to indicate any substantial difference between them.

An important observation is that female bitterlings which have been
used throughout the year seem quite refractory during the late spring and
summer months. From about May 15 until August 15 they react very
weakly, as a rule, to preparations which at other times are definitely effec-
tive. This has been noted by other investigators. It may be a temperature
effect. Gottlieb (11) working in Quebec, got good results throughout the
breeding season, i. e., May-July. Whether temperature regulation or some
other means may be devised to overcome this difficulty remains to be seen.
In the meantime, this seasonal variation must be taken into account.

The preliminary experiments (9) already referred to, showed that
urines from non-pregnant females, pregnant females, women in post-cli-
macteric period, and men gave positive reactions in many instances. We
have since extended these observations and these indicate that most normal
urines from young adult males cause reactions if a large enough quan-
tity can be used without harm to the fish. This suggested that the re-
sponsible factor might not be the follicular hormones. Moreover, the fact
that urines from women 5-6 years after menopause still gave ovipositor
reactions in some cases gave further support to this hypothesis. We also
conducted a large number of experiments with commercial medicinal female
sex hormone preparations and obtained very irregular results. Up to this
point the results did not definitely indicate that the responsible factor is
hormonal in nature. It might possibly be some common urinary constituent,
such as creatinine, uric acid, or indican.

It was therefore seen to be necessary to determine what type of com-
pound present in urine caused the reaction. Urine was first dialysed with
the following results:

Experiment 1 : Urine was obtained from a healthy male subject whose
urines, previously tested, had invariably given positive results. 20 cc. was
dialysed in a cellophane membrane against 40 cc. of distilled water in the
refrigerator, for 24 hours. An undialysed portion was kept at the same
temperature for the same period. 16 cc. of the dialysed urine, 32 cc. of the
dialysate, and 16 cc. of the untreated urine were each added to 4 liters of
water containing 2 fish. Positive reactions were noted in 48 hours in the
tanks to which had been added the dialysed urine, as well as the control,
but not in that containing the dialysate.

Experiment 2: A similar experiment was done with pregnancy urine.
Similar results were obtained but it was noted that the dialysed urine was
distinctly ammoniacal. This led to the suspicion that an excess of am-
monium compounds might be the causative factor. An experiment was
therefore conducted in which ammonium hydroxide was added to the water
containing 2 fish. Results of this test proved to be negative. It thus seemed
evident that the ovipositor stimulating substance is not an ammonium salt
and is not dialysable. This also rules out many common urinary consti-
tuents, such as inorganic salts, creatinine, urea, uric acid, etc.

3 This differs slightly from our originally suggested definition (12), i.e., that the ovipositor
must reach the end of the fin. The present definition allows for differences in the initial length
of the ovipositor.

244

Zoologica: New York Zoological Society

[XXI: 21

The next step was to determine whether it belonged to any of the
classes of sex hormones. Even though we had shown its incapability of
being used as a pregnancy indicator, it was necessary to rule out definitely
the anterior-pituitary-like hormone. This was easily done by boiling some
“positive” urine and ascertaining that the resulting urine was still positive.
The anterior-pituitary-like hormone (“AP-L”) is heat labile.

The following experiment showed that the responsible factor is chloro-
form-soluble.

Experiment 3 : 500 cc. of mixed urine from 20 pregnant women was
treated with 20 cc. of cone. HC1 and was then extracted three times with
100 cc. portions of chloroform in a separatory funnel. An emulsion formed
which was cleared by the addition of 500 cc. of ether. The chloroform-
ether layer was drawn off, evaporated on a steam bath, and the brown
gummy residue dissolved as far as possible in 65 cc. of water containing
0.5 cc. of 10% NaOH. This was added to 7 liters of water containing 2
female bitterlings. In 16 hours both fish showed marked positive reactions.
35 cc. of the extracted urine from which the chloroform had been removed
gave no reaction under similar conditions in 72 hours. This experiment
led us to the conclusion that the effective factor is soluble in chloroform
and to the impression that it is similar to the ovarian follicular hormones.

It therefore seemed logical to expect that crystalline “theelin” or
“theelol,” or both, would produce this reaction. We were fortunately able to
obtain a small amount of each from Dr. Edward A. Doisy and they were
tested and showed very slight effects, if any, on the ovipositor.

As is well known, the chloroform and ether soluble hormones of urine
comprise the male as well as the female sex hormones, i. e., androsterone
and estrins. We consequently subjected urine to a rough separation of these
substances, using the method of Funk, Harrow, and Lejwa (13) with the
modifications of Butenandt and Tscherning (14) and Kochakian and Murlin
(15). A typical experiment is the following:

Experiment 4: 18 liters of mixed male urine was made acid to Congo
Red and 360 cc. cone. HC1 added. The mixture was concentrated on a
steam bath to 1,800 cc. 500 cc. chloroform was added and the mixture
refluxed for 12 hours on a steam bath. The aqueous fraction was discarded
and the chloroform extract evaporated to dryness. The gummy residue was
then dissolved in ether and shaken with 2N KOH until no further color ap-
peared in the aqueous phase. The washings were extracted with ether and
the ether solutions combined, and evaporated to dryness. The residue was
refluxed with 50 cc. 3N KOH in methyl alcohol for 2 hours. After cooling,
2 liters of water was added and extracted repeatedly with ether. The
aqueous fraction was saved. The extract was washed with water, dilute
acid, and finally again with water. This constitutes the male hormone
fraction.

The aqueous fraction was treated with HC1 until acid to Congo Red
and an additional 20 cc. HC1 were added. The mixture was heated on a
steam bath for 1 hour and extracted repeatedly with ether. The ether
extract was washed with water, dilute alkali, and finally with water.
This yields the female hormone fraction.

Each fraction was now tested for its ovipositor-lengthening effect on
the female bitterling, as follows :

Experiment 5 : Ether solutions equivalent to 64 cc. of urine were placed
in a mortar and the ether permitted to evaporate. The oily residue in each
case was emulsified with acacia and water and added to a tank of 2 bitter-
lings in the usual manner. After 18 hours positive reactions were seen to
have been produced by the male fraction and none by the female fraction.

That the male fraction actually contained male hormone was substan-
tiated by injection of a cotton-seed oil solution of it into a capon. A posi-

1936] Kleiner, Weisman, Mishkind & Coates: Male Hormone Test 245

tive result was seen, whereas a similar test with the female fraction was
negative. Both fractions, however, produced estrogenic effects when in-
jected into immature female mice. This harmonizes with the previous
experiences of many investigators (16) who found estrogenic effects with
male hormone preparations. Recent work indicates, however, that highly
purified or synthetic androsterone is non-estrogenic by the vaginal cornifi-
cation test (17). Hence, the estrogenic effect of our male fraction may
have been due to admixture of impurities.

The question now arose as to why pregnancy urines should be more
potent and more constant in their activity than urines from males and
non -pregnant females. We therefore repeated the above experiment, using
mixed pregnancy urines. Here again the bitterling test was positive with
the male fraction and negative with the female.

The facts thus indicate that the ovipositor-lengthening factor of urine is
present in the male fraction, but proof that one of the male hormones is
responsible was still lacking. This gap in the proof has recently been filled
by the use of synthetic preparations4. Despite the fact that our experiments
were performed during the refractory season we were successful in showing
the efficacy of these products. At the same time we can also report what
seems to be a more suitable solvent which is harmless to the fish and which
offers a much better menstruum for these sterols than the acacia which we
had formerly employed as an emulsifying agent. This is propylene glycol,
suggested to us by Dr. Warren M. Cox, Jr., of the Research Laboratory of
Mead Johnson & Co.

Experiment 6: 4 mg. of synthetic androsterone (Schering) and 3 mg.
of synthetic testosterone (Schering) were each dissolved in 5 cC. of propy-
lene glycol (Eastman) under the influence of slight heating. Aquaria with
two bitterlings in 4 liters of water had been set up the previous day. All had
extremely small ovipositors (0 to 2). One liter of water was removed from
the aquarium and the propylene glycol solution of the sterol added quickly
and shaken vigorously. It was then added to the aquarium from which the
water had been taken. A third aquarium with the same amount of
propylene glycol was observed as a control.

A positive result was noted in the bitterlings exposed to both the
crystalline androsterone and testosterone. Propylene glycol alone was nega-
tive. The fish exposed to androsterone were greatly weakened by this sub-
stance but the testosterone had very little systemic effect.

In the fall when the animals were found to be in a normally reactive
state both of these synthetic products were re-tested. It was found that 0.8
to 1.2 mg. of androsterone (18) produced positive reactions in 48-72 hours,
whereas small amounts were never effective and larger doses yielded vari-
able results. The larger doses seemed to have the same depressing effect
as was mentioned above. Crystalline testosterone also gave uniformly posi-
tive reactions at a certain dosage (0.6-0.8- mg.) (unpublished data)
whereas larger and smaller amounts were usually ineffective. Again it may
be stated that a positive reaction is one in which the lengthening of the
ovipositor totals 7 points or more on the scale. Many of the negative tests
at other dosages showed slight effects, i. e., a lengthening of less than 7. It
must also be pointed out that the results with these crystalline products ap-
peared more slowly than when urine had been used, presumably because
they are in a different physical state. From these experiments it is evident
that the ovipositor-lengthening phenomenon is due to male hormones. There
can be no suspicion of admixture of urinary impurities in these synthetic
products.

Confirmatory evidence has also been obtained by testing male hormone
concentrates kindly furnished by Dr. Benjamin Harrow of the College of

4 We wish to thank Schering and Company and Dr. Erwin Schwenk for the synthetic
androsterone and testosterone supplied.

246

Zoologica: New York Zoological Society

[XXI: 21

the City of New York. This material, an oily solution of the male hormone
fraction of male urine, has an activity of 10 capon units per cc., each cubic
centimeter representing 1,000 to 1,150 cc. of urine. As little as Oil cc. added
to an aquarium of female bitterlings gives a positive reaction in 48 hours.
The minimum effective dose has not yet been reached. ,

Technique for Assay of Urine.

In attempting to determine the amount of male hormone in normal male
urine, the following procedure was employed, using normal male medical
students as subjects: For each assay, 4 aquaria were set up, containing 4
liters of water and 2 female bitterlings in the usual manner. To each tank
was then added 10 cc., 25 cc., 50 cc., and 100 cc. portions of urine respec-
tively. 24 hour samples of urine were used. The smallest amount causing
a positive reaction was considered to contain one unit. The number of
units in a 24-hour sample was then calculated. It was noted that in many
instances the urine was toxic and even fatal in amounts from 50 to 100 cc.
Various methods have been tried in the attempt to detoxicate the urines
and finally it was found that dialysing is all that is necessary (19). By
this means dialysed urines representing as much as 200 cc. of the original
may be used without any ill effects whatever. Ordinarily the procedure
is the following: Measured amounts of urine, usually 200 cc., are dialysed
in membranes of cellophane (“plain transparent” not “moisture-proof”)
against running tap water for 18-24 hours. The volumes are then measured
and amounts equivalent to 10, 25, 50, and 100 cc., respectively, of the
original urine are added to 4 aquaria, each of which contains 2 female bit-
terlings. The ovipositors of the fish must have been read on the preceding
day and also just before addition of the dialysed urine. Subsequent read-
ings are made in 24, 48, and 72 hours, and the number of units determined
in the manner suggested above. The average excretion of male hormone is
about 35 bitterling units per day with a range of approximately 15 to 75
b.u. (see Table I). Further work to check and enlarge this series is in
progress.

TABLE I.

Daily excretion of male hormone by normal adult males.

Number of 24-hour
urine specimens

Number
added to

of cc. of urine which, when
4 liters of water, produce a
positive reaction

Bitterling units
excreted daily

5

10—15

75—120

15

25—35

24—45

10

75—100

9—22

Preliminary tests have indicated the unreliability of using casual speci-
mens for even rough quantitative work. We have noted that successive
samples obtained during the day from the same individuals have been ex-
ceedingly variable in the amounts of hormone excreted and some samples are
even entirely devoid of the hormone. Apparently 24-hour samples of urine
are needed in order to determine the output of this hormone with any
degree of accuracy.

Discussion.

Evidently neither theelin nor theelol is responsible for the reaction,
nor are various cholane derivatives which have thus far been tested, such
as cholesterol, ergosterol, and sodium taurocholate (20). Several of these

1936] Kleiner, Weisman, Mishkind & Coates: Male Hormone Test 247

give slight reactions, which suggest the possibility that other cholane
derivatives may be found which will react as well as the male hormone. Up
to the present the test seems to be specific for male hormone.

In support of this we may cite several references to the literature.
Glaser and Haempel (21) compared the effect of a follicular hormone prep-
aration with several other preparations, including a male hormone product,
“testosan forte.” The technique of the experiments included placing males
and females in the same aquarium. The results showed slight growth of
ovipositors in those females subjected to the male as well as the female
hormone, although the latter gave stronger results. The data given seem
inconclusive, however. Glaser and Haempel conclude that certain secondary
sex characteristics are influenced by both male and female hormones. Among
these are the lengthening of the ovipositor of the female bitterling and
the growth of the comb and wattles of the capon. This work was subjected
to criticism by Fleischmann and Kann (22). They confirmed Glaser and
Haempel’s contention that “testosan forte” gave positive results with the
female bitterling but they are of the opinion that this is due to an admix-
ure of follicular hormone. In support of this they state that this male hor-
mone preparation also gives the Allen-Doisy test on the mouse. A more con-
centrated male hormone preparation, “Proviron” (Schering) had the same
action, i.e., positive reactions with both the fish and mouse tests, whereas
crystalline “Proviron” reacted negatively with both. We suggest that this
negative result may have been a result of incomplete solution or improper
emulsification of the substance. The authors do not state the menstruum
or method employed for this purpose.

An interesting contribution has also been made by Ehrhardt and Kuhn
(23). In a long series of experiments they come to the conclusion that the
ovipositor-lengthening factor and the estrus hormones are not identical,
although both have a number of properties in common, i.e., solubility in
organic solvents, heat stability, absorption on charcoal, etc. Some points of
difference are the following : in the urine of pregnant mares are large quan-
tities of estrus hormone, but small quantities of the ovipositor-lengthening
factor. Non-pregnant women, on the other hand, excrete a urine with just
the opposite characteristics. Eight hundred units of technical (i.e., impure)
estrus hormone give a better ovipositor-lengthening result than several
thousand units of crystalline estrus hormone. The blood serum of the preg-
nant woman gives a strong Allen-Doisy test but a weak bitterling test.
Their preparations thus did not quantitatively show parallel results when
tested with the fish and with the mouse. Although they believe the two
phenomena to be due to two distinct substances they are of the opinion that
one is a transformation product of the other. Ehrhardt and Kuhn’s work
harmonize with the contention that the male hormone is the responsible
factor for this phenomenon.

We propose to assay urines and other biological fluids under various
physiological conditions and from pathological cases for their male hormone
content by this method. The method should also be of value in assaying
male hormone products to be used therapeutically.

Summary.

The lengthening of the ovipositor of the female bitterling produced by
the administration of human urine has been found to be due to the male
hormone present therein. Confirmation of this is shown by positive effects
caused by crystalline androsterone and testosterone. A bitterling unit is
provisionally defined and a method of measuring the ovipositor is described.

248

[XXI : 21

Zoologica: New York Zoological Society
Bibliography.

1. Harrow, B. and Sherwin, C. P.

1934. The Chemistry of Hormones. The Williams and Wilkins Co.,
Baltimore ; Chapter 6.

2. Glaser, E. and Haempel, O.

1932. Deutsche Med. Wochschr . 58:1247.

3. WlTSCHI, E.

1936. Proc. Soc. Exp. Biol. & Med., 33:484.

4. Koch, F. C. and Gallagher, T. F.

1932. Jour. Amer. Med. Assoc. 98:738

5. Funk, C., Harrow, B. and Lejwa, A.

1929. Am. J. Physiol. 92:440.

6. Ehrhardt, K. and Kuhn, K.

1933. Mschr. Geburtsh. 94:64.

7. Fleischmann, Walter, and Kann, Susanne.

1932. Arch. f. d. ges. Physiol. 230:662.

8. Kanter, A. E., Bauer, C. P. and Klawans, A. H.

1934 Jour. Amer. Med. Assoc. 103:2026.

9. Kleiner, I. S., Weisman, A. I. and Barowsky, H.

1935. Jour. Amer. Med. Assoc. 104:1318.

10. Personal communication from Rosalind L. Moses of the French Hospital,

New York, N. Y.

11. Gottlieb, R.

1936. Can. Med. Assoc. Jour. 34:431.

12. A preliminary announcement of this work was made at a meeting of the

New York Endocrinological Society, Jan. 29, 1936.

1936. Jour. Amer. Med. Assoc. 106:1643.

13. Funk, C., Harrow, B. and Lejwa, A.

1929. Proc. Soc. Exper. Biol. & Med. 26:569.

14. Butenandt, A. and Tscherning, K.

1934. Ztschr. f. physiol. Chem. 229:167.

15. Kochakian, C. D. and Murlin, J. R.

1935. J. Nutrition. 10:437.

16. Fellner, O. O.

1921. Arch. f. d. ges. Physiol. (Pfliiger’s) 189:199.

Carminati, V.

1927 . Endocrinol, e pat. costit. 2:337. (Dec.).

Laquer, E., Dingemanse, E., Hart, P. C. and de Jongh, S. E.

1927. VI Mitteilung, Klin. Wchnschr. 6:1859.

(Sept. 24).

17. Warren, F. L.

1935. Nature. 135:234.

Deanesly, R. and Parkes, A. S.

1936. British Med. Jour. 1:257.

18. Kleiner, I. S., Weisman, A. I. and Mishkind, D. I.

Proc. Soc. Exp. Biol, and Med.

(In press).

1936] Kleiner, Weismann, Mishkind & Coates: Male Hormone

249

19. Kleiner, I. S., Weisman, A. I. and Mishkind, D. I.

1936. Science. 84:142.

20. Kleiner, I. ,S., Weisman, A. I. and Mishkind, D. I.

1936. Proc. Soc. Exper. Biol. & Med. 34:367.

21. Glaser, E. and Haempel, O.

1932. Klin. Wochschr. 12:1491.

22. Fleischmann, Walter and Kann, Susanne.

1935. Klin. Wochschr. 14:644.

23. Ehrhardt, K. and Kuhn, K.

1934. Zentralb. f. Gyn. 58:2834.

250

Zoological New York Zoological Society

EXPLANATION OF THE PLATE.

Plate I.

Fig. 1. Female bitterling, showing elongation of ovipositor: considered a mini-
mum positive reaction.

Fig. 2. Female bitterling, showing elongation of the ovipositor in the process of
natural ovipositation.

KLEINER, WEISMAN, MISHKIND & COATES.

PLATE I.

FIG. 2

THE FEMALE BITTERLING AS A BIOLOGIC TEST ANIMAL FOR MALE HORMONE.

Nigrelli: Hosts for Clinostomum complanatum

251

22.

Some Tropical Fishes as Hosts for the Metacercaria of Clinostomum
complanatum (Rud. 1814) (= C. marginatum Rud. 1819).

Ross F. Nigrelli.

New York Aquarium.

(Plates I & II).

The following fishes received by the New York Aquarium were found
infected with the yellow grub stage of Clinostomum : Chriopeops goodei
(Jordan) from Florida; Mollienisia velifera (Regan) from Yucatan; Pia-
bucana sp. from the Amazon drainage ; Corynopoma riisei Gill from British
Guiana; Nannostomus trifasciatus Steindachner from the Tucantins, a
tributary of the Amazon; Lebistes reticulatus (Peters) from out-door
pools in Florida ; Hypopomus artedi Kaup from the Amazon drainage in
eastern Brazil; and Sternopygus macrurus (Bloch and Schneider) from the
Amazon drainage.

In some instances the fishes showed no infection when they arrived,
and only after they had been in the Aquarium for some time did the cysts
appear. Whether the fishes were originally infected and growth of the
parasite had taken place in the tanks, or whether the fishes acquired the
infection in the tanks is difficult to determine. Examination of snails and
water of the aquaria showed no fork-tailed cercariae.

In cases of severe infection the death of the host resulted. The
trematodes were usually encysted in the fish, and in Corynopoma riisei the
grub was embedded deep in the peculiar transparent muscle of the body
(PL I, Figs. 1 & 2). In the gymnotid, Sternopygus macrurus, 16 cysts were
found distributed along both sides of the base of the elongated ventral fin.

Observations on the ‘‘guppy’” showed that, since this species is not
influenced by seasonal changes, the cysts are not deserted as reported by
Ward (1918) for northern forms. The parasites were present on the fish
(PI. II, Fig. 3) for more than a year. In one of the specimens found dead,
moribund metacercariae were still in the cysts. Many of the temperate
zone fishes in the Aquarium were also found infected with the metacercariae
of Clinostomum. In the majority of instances the worms were encysted on
the membrane between the rays of the fins. These cysts appeared in the
spring and in winter were absent.

Examination of the parasites showed no morphological differences be-
tween the metacercariae found on tropical forms and those of temperate
zone species. In all instances the worms have been identified as C. com-
planatum (Rud. 1814) (— C. marginatum Rud. 1819) (PL II, Figs. 4 & 5).

The metacercaria of C. complanatum has been found on a large num-
ber of fish hosts throughout the world. As C. marginatum it has been
reported by several investigators from the following neotropical fishes (see
Baer 1933): Adnia ( Adinia ?) dugesi (Bean), Callichthys asper Quoy and

252

Zoologica: New York Zoological Society

[XXI: 22

Gaim., Poecilia vivipara Bloch and Schneider, Chaetosomus brachyurus
(Kner), Cynodon scomberoides (Cuv.), Satanoperca papatera (Heck.),
Chaetobranchus flavescens Heck., C. gulosus (Cuv. and Val.), Crenicichla
johanna Heck., Crenicichla saxatilis (Linn.).

These are definitely Central and South American species of fishes, but
because of the terminology employed, it is difficult to determine which form
is referred to, since in the present day usage some of the above genera
have been split up and others have been discarded. In North America the
metacercaria has been reported from a large number of fishes, several
species of amphibians ( Rana clamitans, R. pipiens, etc.), and, according to
Hopkins (1933), the parasite was found in a snake ( Thamnophis radix
(Baird) ) by van Cleave.

The following are some of the North American fishes found infected:
Perea flavescens (Mitchill)1 (yellow perch) ; Aphredoderus sayanus (Gil-
liams) (pirate perch) ; Micropterus dolomieu Lacepede1 (small mouthed
black bass) ; M. salmoides (Lac.)1 (large mouthed black bass) ; Ambloplites
rupestris (Raf.)1 (rock bass); Lepomis pallidus (Mitchill)1 (bluegilled
sunfish) ; L. auritus (Linn.)1 (red-breasted sunfish) ; L. cyanellus Raf.1
(green sunfish).; Eupomotis gibbosus (Linn.)1 (common sunfish); Cato-
stomus commersoni (Lac.)1 (common sucker) ; Pimephales promelas Raf.
(fathead minnow) ; Semotilus astromaculatus (Mitch.) (horned dace) ;
Ameiurus nebulosus (Le Sueur)1 (common bullhead) ; and Salvelinus fon-
tinalis (Mitch.) (brook trout). In Europe Ciurea (1911) reported the yel-
low grub of C. complanatum on Perea fluviatilis Linn., and Maccagno (1933)
found it on the loach, Cobitis taenia Linn. In Japan Yamaguti (1933) found
it under the skin and flesh of Carassius carassius Linn., Pseudogobio eso-
cinus, and Acheilognathis intermedia.

The following are other species of Clinostomum described from fishes:

1. C. heterostomum (Rud. 1809) from European fishes, and from an
unidentified species taken in the Vaal river, Africa (Monnig 1926).

2. C. dimorphum (Diesing 1850) from many of the tropical fishes men-
tioned above.

3. C. dictyotum (Monticelli 1893) ( — Clinostomatopsis reticulata

(Looss 1885) ) from an unidentified silurid of Costa Rica.

4. C. africanus Stoss, from the intestine of an unidentified fish of the
French Congo (see Galli-Valerio 1906).

5. C. piscidium Southwell and Prashad 1918 from Trichogaster fas-
ciatus and Nandus nandus of Ceylon, and from Trichogaster pec-
tor alis and T. trichopterus of Siam (Pearse 1933).

6. C. intermedialis Lamont 1920 from the silurid, Rhamdia quelen
Quoy and Gaimard.

7. C. chrysichthys Dubois 1930 from Chrysichthys kingselyi Gunth.,
a silurid of Angola.

8. C. clarias Dubois 1930 from Clarias angolensis Steind., a silurid of
Angola.

9. C. dalgi Tubangiu 1933 from the eye-socket and pericardium of
Ophiocephalus striatus Bloch., a Philippine fish.

Four orders of fishes are represented by the hosts of Clinostomum re-
ported in the literature. Arranged according to the number of species found
infected, these are: Ostariophysii, Acanthopterygii, Haplomi, and Isos-

pondyli. With the exception of the Isospondyli, all are basically pond fishes.
In the great majority of ponds the same relationship among these orders
probably exists; that is, the Ostariophysii includes the largest number of
species to be found in such a habitat, the Acanthopterygii the next largest,
etc.; and likewise as to number of individuals. The Isospondyli is repre-

1 Species present in the Aquarium on which Clinostomum was found.

1936]

Nigrelli: Hosts for Clinostomum complanatum

253

sented by one species, Salvelinus fontinalis (brook trout). It is altogether
possible that this form became infected during an excursion into still water,
which it sometimes makes.

It is not surprising to find that pond fishes show the highest incidence
of infection. The nature of the cercaria of Clinostomum is such that it
would be mechanically difficult to infect fishes in a fast stream. These
cercariae are fork-tailed organisms which float at the surface. The pond
fishes, moreover, are more sluggish than the stream forms, thus making it
more feasible for the cercariae to penetrate the skin.

The genus Clinostomum was erected by Leidy (1856) for a metacer-
caria found encysted in the skin of the common sunfish, Pomotis vulgaris
C. and V., Eupomotis gibbosus (Linn.), and in the intestine of Esox sp.
This species he called C. gracile and designated it as the type of the genus.
C. gracile Leidy 1856, however, is considered a synonym of C. complanatum
(Rud. 1814), an adult form taken from a heron ( Ardea cinerea ) by Rudolphi
and described by him as Distomum complanatum. Braun (1901) made
C. complanatum the type of the genus because Leidy’s description was
inadequate and because the original type was lost. There is no doubt, how-
ever, that the yellow grub, called C. marginatum by most North American
investigators and even now found on the type host and in the type locality,
is Leidy’s C. gracile. Braun (1899) considered C. marginatum (Rud. 1819)
a synonym of C. complanatum (Rud. 1814) because the two were morpho-
logically the same and were found as metacercaria on fishes and as adults
in the same genus of birds {Ardea). Baer (1933) expressed the same
opinion and added that the geographical distribution of the two was alone
not sufficient to warrent regarding them as distinct species.

According to Baer (1933) C. heterostomum (Rud. 1809) = Euclino-
stomum heterostomum (Rud. 1809), and C. dimorphum (Diesing 1850) =
I thy clinostomum dimorphum (Diesing 1850). In Euclinostomum Travassos
1928 the intestinal branches have long lateral diverticula, sometimes rami-
fying. In I thy clinostomum Witenberg 1926 the lateral diverticula are short
and never ramifying. This genus is further separated from Clinostomum on
the distribution of the vitelline glands. In the latter these glands extend
into the anterior part of the body.

The complete life history of C. complanatum was only recently worked
out. It was known for some time that the adult worms could be found in
the oral cavity of herons {Ardea) and related birds (Osborn 1911, 1912).
In this country Hunter and Hunter (1934, 1935) showed that when the
eggs of the parasites drop from the mouth cavity of the definitive host
into water, they hatch into miracidia which penetrate the liver of snails.
Two gastropods, Helisoma campanulatum (Say) and H. antrosum (Conrad),
were experimentally infected. In the snail each miracidium develops into a
sporocyst which produces rediae. The rediae in turn produce daughter
rediae which become fork-tailed cercariae. These emerge from the snail
and encyst as metacercariae, commonly called yellow grubs, on fishes. When
an infected fish is eaten by a bird, the larvae are released and grow to ma-
turity in the oral cavity.

Of the several species of Clinostomum described from fishes, the only
other forms in which the adults are definitely known are C. heterostomum
and C. dimorphum from ardeiform birds, and C. intermedialis from the
cormorant {Phalacrorax vigua) . Without doubt, however, a life-cycle sim-
ilar to that of C. complanatum may be expected in the entire genus.

The following are species of adult Clinostomum, described from various
definitive hosts, of which neither the first nor the second developmental
stages are known: C. dubium Leidy 1856, C. detruncatum Braun 1899, C.
foliiforme Braun 1899, C. heluans Braun 1899, C. lambitans Braun 1899,
C. sorbens Braun 1899, C. hornum Nicoll 1914, C. australiense Johnston

254

Zoologica: New York Zoological Society

[XXI : 22

1916, C. pusillum Lutz 1928, C. phalacrocoracis Dubois 1931, and C. lopho-
phallum Baer 1933.

Two species of Clinostomum have been described from amphibians.
Cort (1913) believed that the C. marginatum reported from batrachians
(Wright 1897 ; Osborn 1911, 1912) was a distinct species, which he called
C. attenuatum. His separation is based principally on certain morphological
characteristics. He designated the American bittern as the probable defini-
tive host of the metacercariae in amphibians, but Alvey and Stunkard
(1932) reported C. attenuatum from the great blue heron. Hunter and
Hunter (1934) demonstrated in feeding experiments that the great blue
heron could be infected with metacercariae taken from frogs, but that the
rate of infection was not as high as with C. marginatum from fish. They
further showed (1935 a, b) that when both sunfish and tadpoles were ex-
posed to the cercariae of C. marginatum, the sunfish became infected while
the tadpoles remained free of metacercariae. They concluded that C. at-
tenuatum is a distinct species. The other Clinostomum reported in amphibia
is C. pseudoheterostomum Tubangui 1933, which occurs in the thigh muscle
of a Philippine frog (Rana magna) . The adult of this species, however, is
unknown.

References.

Alvey, C. H. and H. W. Stunkard (1932). Clinostomum attenuatum from the
great blue Heron. J. Para. 19: 174.

Baer, J. G. (1933). Note sur un nouveau trematode Clinostomum lophophallum
sp. nov. avec quelques considerations generates sur la famille des Clinosto-
midae (Trematoda). Rev. Suisse Zool. 40: 317-342, 3 figs.

Braun, M. (1899). Ueber Clinostomum Leidy. Zool. Anz. 22: 484-493. (1901).

Die Arten der Gattung Clinostomum Leidy. Zool. Jahrb. Syst. 14: 1-48,
pis. 1, 2.

Ciurea, J. (1911). Eine europaische Clinostomum-lavvae. Zbl. Bakter. Paras.
Originate 60: 354-358.

Cort, W. W. (1913). Notes on the trematode genus Clinostomum. Trans. Amer.
Micr. Soc. 32: 169-182, pi. 9.

Dubois, G. (1930). Deux nouvelles Especes de Clinostomidae. Bull, de la Soe.
N euchdteloise des Science Naturelles. 54: 61-72, figs. 1, 2; pis. 1, 2.

(1931). Materiaux de la mission scientifique suisse en Angola. Trema-
toda. Bull, de la Soc. N euchdteloise des Science Naturelles. (1930) 55: 73-88,
15 figs.

Galli- Valerio, B. (1906). Michele Stossich und seine helminthologischen Ar-
beiten. Centralbl. Bakt. Parasit. Orig. 42: 49, fig. 2.

Hopkins, S. H. (1933). Note on the life-history of Clinostomum marginatum
(Trematoda). Trans. Amer. Micr. Soc. 52: 147-149, pi. 24.

Hunter, G. H. and Hunter, W. S. (1934). IX. Studies on Fish and Bird Para-
sites. Suppl. 23rd. Ann. Rep’t. N. Y. State Conserv. Dep’t., 1933, no. 8,
Rep’t. of a Biological Survey of the Raquette Watershed. 4 figs., 2 tables.

(1935 a). IX. Further studies on Fish and Bird Parasites. Suppl. 24 th
Ann. Rep’t. of the N. Y. State Conserv. Dep’t., 1934, no. 9, Rep’t. of a Bio-
logical Survey of the Mohawk-Hudson Watershed. 267-283, pis. A-C, 3 tables.

(1935 b). Studies on Clinostomum. IV. Notes on the penetration and
growth of the cerearia of Clinostomum marginatum. J. Para. 21 : 1, 2.

Johnston, S. J. (1916). On the Trematoda of Australian Birds. J. Proc. R. Soc.
N. S. Wales. 50: 187-261, figs. 1-10, pis. 9-19.

Lamont, M. E. (1920). A new species of Clinostomum. Occ. Papers Mus. Zool.
Univ. Michigan. (83) : 1-5, pi. 1.

1936]

Nigrelli: Hosts for Clinostomum complanatum

255

Looss, A. (1885). Beitrag zur Kenntnis der Trematode. Zeitschr. wiss. Zool.
41 : 390-446, pi. 23.

Lutz, Adolpho (1928). Estudios de Zoologica y parasitologia Venezulanos. 133
pp. 26 pis. Rio de Janeiro.

Maccagno, T. (1932). Clinostomum complanatum (Rud.) in Cobitis taenia Linn.
(Nota preliminare) . Boll. Zool. Napoli. 3: 285-290, 1 fig.

Monnig, H. O. (1926). Helminthological notes. Union So. Africa Dep’t. Agr.
Rep’ts Dir. Vet. Ed. and Res. 11/12 (1) : 221-226, 12 figs.

Nicoll, W. (1914). The trematode parasites of North Queensland. II. Parasites
of Birds. Parasitol. 7: 105-126. pis. 6, 7.

Osborn, H. L. (1911). On the distribution and mode of occurrence in the United
States and Canada of Clinostomum marginatum, a trematode parasitic in
fish, frogs and birds. Biol. Bull. 20: 350-356, 1 fig., pi. 1.

(1912). On the structure of Clinostomum marginatum, a trematode of
the frog, bass and heron. J. Morph. 23: 189-229, 1 fig., pis. 1-3.

Pearse, T. S. (1933). Parasites of Siamese fishes and crustaceans. J. Siam. Soc.
Nat. Hist. Suppl. 9: 179-191, 30 figs.

Southwell, T. and Prashad, B. (1918). Notes from the Bengal Fisheries Lab-
oratory. III. Some fish trematodes. Rec. Ind. Mus. 15: 348-350, pi. 12, fig. 1.

Travassos, L. (1928). Fauna helminthologica de Matto Grosso. Mem. Inst.
Oswaldo Cruz 21: 309-341, 13 pis.

Tubangui, M. A. (1933). Trematode parasites of Philippine Vertebrates, VI.
Description of new species and classification. Philippine J. Sc. 52: 167-197,
6 pis.

Ward, H. B. (1918). Fresh-water Biology. Ward and Whipple. John Wiley and
Sons. New York.

Witenberg, G. (1925). Versuch einer Monographic der Trematodenunter-familie
Harmostominae Braun. Zool. Jahrb. Syst. 51: 167-245, pis. 1, 2.

Wright, R. (1879). Contribution to American Helminthology No. 1. Proc. Cana-
dian Inst. 1: 54-75, pis. 1, 2.

Yamaguti, S. (1933). Studies on the Helminth Fauna of Japan. I. Trematodes
of Birds, Reptiles and Mammals. Jap. J. Zool. 5: 1-134, 57 figs.

References for Rudolphi (1814, 1819) and Diesing (1850), see Braun (1899).

256

Zoologica: New York Zoological Society

EXPLANATION OF THE PLATES.

Plate I.

Fig. 1. Corynopoma riisei Gill showing the yellow grub encysted in the trans-
parent muscle tissue. The worm appears as a dark spot behind the
visceral mass.

Fig. 2. Fish shown in Fig. 1, immediately after death. The cyst was broken
open and the living worm removed.

Plate II.

Fig. 3. Female ‘‘guppy”, Lebistes reticulatus (Peters), about IV2 inches long,
showing five or more Clinostomum cysts. Smaller males were also found
infected.

Fig. 4. Clinostomum complanatum (Rud. 1814) from the “guppy” shown in
Fig. 3. Enlarged about 22X.

Fig. 5. Clinostomum complanatum from the gymnotid, Sternopygus macrurus.
Enlarged about 22X.

NIGRELLI.

PLATE I.

FIG. 1

FIG. 2

SOME TROPICAL FISHES AS
COMPLANATUM (RUD.

HOSTS FOR METACERCARI A
. 1814) (=C. MARGINATUM

OF CLINOSTOMUM
RUD. 1819).

NIGRELLI.

PLATE II.

FIG. 3

FIG. 5

SOME TROPICAL FISHES AS HOSTS FOR METACERCARI A OF CLINOSTOMUM
COMPLANATUM (RUD. 1814) (=C. MARGINATUM RUD. 1819).

FIG. 4

Hollister: Caudal Skeleton of Bermuda Fishes

257

23.

Caudal Skeleton of Bermuda Shallow Water Fishes. I. Order
Isospondyli: Elopidae, Megalopidae, Albulidae, Clupeidae,
Dussumieriidae, Engraulidae.1

Gloria Hollister.

Department of Tropical Research.
(Text-figures 1 to 53).
Outline.

Introduction 257

Caudal Fin Terminology 258

Key op. 260

Bermuda Isospondyli :

Elopidae, Elops saurus 260

Megalopidae, Tarpon atlanticus 263

Albulidae, Albula vulpes . . 268

Dussumieriidae, Jenkinsia lamprotaenia 276

Engraulidae, Anchoviella choerostoma 280

Clupeidae, Harengula sp 282

Opisthonema oglinum 284

Sardinella anchovia 286

Summary 287

Bibliography 289

Introduction.

The following paper is a study of the caudal skeleton of the Bermuda
isospondylids. It deals principally with the adult fish but when young
specimens were available these have been included. The Isospondyli of Ber-
muda are represented by six families, eight genera, and eight species.

In the genus Harengula no specific name has been given to the material
examined. These fish have at various times been called H. macrophthalmus
or H. sardina, and their proper specific determination is a problem for
future study. j

This study was made from specimens which were cleared by potassium
hydroxide and stained by alizarin. Alizarin is a vital stain for bone and
the determination of the presence and position of bones is greatly facili-
tated by its use. The term KOH, which stands for potassium hydroxide,
the clearing chemical, was adopted in our field laboratory at Nonsuch, Ber-

1 Contribution No. 515, Department of Tropical Research, New York Zoological Society.
Contribution from the Bermuda Biological Station for Research. Inc.

258

Zoologica: New York Zoological Society

[XXI :23

muda, as a designation for specimens cleared and stained. They are num-
bered as such in a KOH catalogue. This term appears in the list of cata-
logue numbers and in general usage. For details of this clearing process
see “Clearing and Dyeing Fish for Bone Study,” Zoologica, Vol. XII, No.
10, and “Fish Magic,” in the Bulletin of the New York Zoological Society
for March-April, 1930. (Vol. XXXIII, No. 2).

The length of specimens in this paper is standard length, unless other-
wise stated.

I am especially indebted to Mr. and Mrs. George Arents, Jr., and Mr.
and Mrs. Bernard Baruch, Jr., for specimens of large Tarpon, and to the
American Museum of Natural History for a specimen of Elops. I thank
for their cooperation Dr. William Beebe and Mr. John Tee-Van of this de-
partment. The drawings are by Edward Delano, George Swanson, and
Helen Tee-Van.

Caudal Fin Terminology.

(Text-figs. 1, 2).

Caudal Ray Count: The dorsal and ventral counts of fin rays are considered
separately. The dorsal count begins anteriorly and continues around
the fin to the median division in the rays. The ventral count begins
anteriorly and continues around the fin to the median division. The
count is expressed as follows :

Dorsal raylets+rays=Total : 2+10 12

Ventral raylets + rays = Total : 2+10 12

Caudal Ray: A branched (usually) caudal element possessing one or more
transverse joints.

UN

Text-figure 1.

Typical tail of a Bermuda Isospondyli. C, centrum; DR, dorsal ray; EP, epural;
H, hypural; ML, median line; PB, prolonged base; SNP, specialized neural
process; UN, uroneural; UR, urostyle; VR, ventral ray.

1936]

Hollister : Caudal Skeleton of Bermuda Fishes

259

Caudal Raylet or Simple Ray: An unbranched caudal element possessing no
transverse joints. Raylets are always anterior to the rays.

Caudal Region: The vertebral column is divided into a trunk and caudal
region. The caudal region begins where the first or anterior haemal
process forms a closed haemal arch with a single haemal spine project-
ing. Ribs are absent in the caudal region.

Centrum: The central element of a vertebra on which the neural and
haemal processes are formed.

Epural: Any bone, or bones, that are dorsal to the urostyle and support
one or more caudal fin rays or spines. The bases are always unattached.

Haemal Arch: The arch on the ventral surface of a vertebral centrum
through which passes the haemal vessel.

Haemal Process: The haemal arch and haemal spine on the ventral surface
of a vertebral centrum.

Haemal Spine: The ventral projection below the haemal arch.

Hypural: Any bone that is ventral or posterior to the urostyle and supports
one or more caudal fin rays or raylets. An expanded haemal spine. The
hypural count is made from the anterior ventral part of the urostyle
around to its posterior and dorsal end. The anterior hypural is called
the first hypural.

Median Caudal Line: The median line is the natural median division seen
in the caudal rays. This line determines whether the rays are dorsal
or ventral.

Neural Arch: The arch on the dorsal surface of a vertebral centrum through
which passes the nerve cord.

Neural Process: The neural arch and the neural spine on the dorsal surface
of a vertebral centrum.

Neural Spine: The dorsal projection above the neural arch.

Uroneurals: The uroneurals are paired bones which are directed upward
and backward on the lateral and dorsal surfaces of the urostyle. They
probably represent specialized neural processes so developed to protect
the sharply upturned caudal. This term is adopted from Regan (1910.
2).

ANTERIOR HALMAL
ZVGAPOPHV5I5

ANTERIOR NEURAL
Z-VGAP0PMV5I5

Text-figure 2.
Typical caudal vertebra.

260

Zoologica: New York Zoological Society

[XXI :23

Urostyle: The posterior terminal vertebral segment or segments which

follow the last undoubted centrum. The urostyle is considered as one
in the total vertebral count.

Zyg apophysis: The dorsal and ventral articulating process of a centrum.

The inserted key illustrates the outstanding differences and similarities
found in the study of the caudal skeleton of the Bermuda isospondylids.
It will be noted that the three clupeids ( Harengula , Opisthonema, and Sar-
dinella ) are included under one caudal pattern. Further study with more
material is necessary in these species to determine the status of the three
forms as shown by the caudal skeleton.

Elopidae.

Elops saurus Linnaeus.

(Text-figs. 14, 15).

Diagnostic Characters :

9 hypurals.

2 reduced posterior neural processes on the two anterior urostyle
centra. Dagger-shaped.

4 distinct pairs of uroneurals. Origin of anterior pair on first or
anterior urostyle centrum.

Vertebral count:

49 + 26 = 75. Haiti specimen.

57 + 24 — 81. Gravesend Bay, N. Y., specimen.

43 -f 29 - 72. Jordan and Evermann, “Fishes of North
America.”

78-79. Elops saurus and affinis. Regan 1909.

This variation may, with further study, be correlated with geo-
graphical distribution.

Material Studied.

The following description is taken from two specimens : one caught in
Haiti, Cat. No. 7172, KOH Cat. No. 2030, length 258 mm., from which this
description has been made, and one caught at Gravesend Bay, which was
kindly supplied by the American Museum of Natural History, Cat. No. 669,
KOH Cat. No. 2136, length 280 mm. Elops saurus is included with the
group of Bermuda Isospondyli on the basis of a single record, that of E.
Linton in 1908, “Notes on Parasites of Bermuda Fishes,” Proc. U. S. Nat.
Mus., XXXIII, No. 1560.

Caudal Osteology.

Urostyle Centra: Four centra form the upturned caudal end of the ver-
tebral column. The anterior three are regularly shaped. The fourth (and
posterior) one is an irregular thin bony rod which curves upward pos-
teriorly and extends as far as the anterior dorsal edge of the 7th hypural.
This may represent three centra fused as it extends over the bases of three
hypurals. A cartilaginous notochord extends into the dorsal contour, 1 mm.
beyond the dorsalmost hypural, and is embedded between the bases of the
11th dorsal ray, above the median line (Text-fig. 15). There are about 35
horizontal striae in the cartilaginous notochord. Embedded between the
striae spool-shaped centra can be seen. This condition resembles that found
in the young rather than the adult Tarpon.

Uroneurals: This term is adopted from Regan (1910.2). There are
four pairs of uroneurals, which correspond to ancestral posterior neural

Key to Caudal Fin of Bermuda Shallow Water Isospondylid Fishes.

(Text-figs. 3-13).

1.

Group A

25 to 26 haemal arches.

(Total vertebral count 37 to 40).

Harengula sp.

(See Text-figs. 47 and 48).

Text-figure 12.

Group B

29 to 32 haemal arches.

(Total vertebral count 45 to 46) .

Opisthonema oglinum

(See Text-figs. 49, 50 and 51).

Sardinella anchovia

(See Text-figs. 52 and 53).

Text-figure 13.

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

261

processes. All four are elongated and crowded together on the lateral and
dorsal surfaces of the upturned centra.

The uroneural whose position is the most anterior and the most dorsal
is the longest (16 mm.) and the stoutest (1.5 mm. at the deepest point)
and extends from the first upturned centrum as far as the center of the
anterior base of the 9th (dorsalmost) hypural. Here it terminates in a
blunt tip. Anteriorly this uroneural is forked, the dorsal part commencing
on the first urostyle centrum which is the fourth from the last. According
to Regan (1910.1) “the forking indicates the compound nature of this bone,

Text-figure 14.

Elops saurus. Tail of a 258 mm. specimen (x 2.8).

and in some specimens the line of junction between the two component
elements can be clearly seen.” In our Haiti specimen of 258 mm. and Graves-
end Bay specimen of 280 mm., the line of junction is not evident. The
anterior pair of uroneurals is the only one that covers in part the dorsal,
as well as the lateral surfaces of the urostyle. The anterior half is entirely
lateral but above the posterior, or last, centrum the lateral, parts meet
dorsally, but do not fuse, and extend for a short distance covering both
the dorsal and upper lateral surfaces of the urostyle.

The bones of the second pair of uroneurals are spindle-shaped, 12.5 mm.
long and 1 mm. deep, unforked and extend along the ventral surfaces of
the first pair. The second pair arises anteriorly on the third centrum and
ends between the bases of the 11th caudal ray, extending 2 mm. beyond the
first pair.

The third pair resembles the second in shape but the bones are one-half
the size. The length is 6.43 mm. and the depth .43 mm. This pair arises at
the posterior third of the second pair and extends slightly beyond it, ending
posteriorly between the bases of the 11th ray.

The fourth pair is the smallest, length 3.86 mm. and depth .29 mm.
The bones extend obliquely across the 10th dorsal ray with the anterior tips

262

Zoologica: New York Zoological Society

[XXI :23

on the dorsal side and the posterior on the ventral side of the ray. The
direction does not parallel that of the other three pairs; instead there is a
decided posterior slant. The entire length of the two bones is free in the tis-
sue that covers the ray bases ; whereas the dorsal tips of the other three
uroneurals are between the bases of the 11th caudal ray.

None of these pairs are fused one with the other. The two bones of
each pair are distinctly individual.

Hypurals: There are nine hypurals on the urostyle, four below and
five above the median line. Two additional anterior long haemal spines
project into the ventral caudal contour with fin rays attached. The third
hypural (counting from the anterior to the posterior dorsal) is the largest.

The ninth and dorsalmost hypural is the smallest. The bases of the third
and fourth hypurals are ventral and adjacent to a single centrum. This
is also shown in Regan’s drawing (1910.1) and is present in Tarpon and
Albula and indicated in all the Bermuda isospondylids. None of the hy-
purals are fused. All of the bases are cup-shaped and the two dorsal ones
are more pronounced, almost bifid, the basal tips extending on either side
of the notochord. The Haiti specimen is more developed in this respect,
the bases being noticeably longer. (Text-fig. 17 for example in Tarpon).

Epurals: There are three epurals, all long flat bones expanded at the
ventral ends. The first and anterior epural is the longest, its ventral end
slightly over-lapping the tip of the anterior reduced neural process.

Specialized Neural Processes: There are two reduced neural processes
on the first and second upturned urostyle centra. Both are short, broad,
dagger-shaped bones, the anterior one being the larger. The anterior neural
process is on the first centrum that shows a tendency to turn upward. So
this centrum is considered the first of the urostyle series and the attached
haemal spine is called the first hypural in this paper.

The shape of the neural processes of the Gravesend Bay specimen is
different from the Haiti specimen, the tips being blunt and the same depth
as the rest of the process. There is a third smaller process, which is not
present in the Haiti specimen. This may be an individual irregularity and
not a common variation. But as there are only two specimens from which
to draw conclusions this point will be left open to be determined later. This
study is too new to base specific differences on characters such as the

Elops saurus. Tail of 258
mm. specimen with paired
uroneurals removed showing,
in black, the vertebral seg-
ments of the urostyle (x 3).

Text-figure 15.

1936] Hollister: Caudal Skeleton of Bermuda Fishes 263

neural processes without more specimens to establish the normal range of
variation.

Dorsal to the third urostyle centrum are two small round islands of
bone surrounded by a median plate of cartilage. In Regan’s illustration a
solid bone is shown filling this entire area below the three epurals. The
Gravesend Bay specimen has one round island of bone which is of interest
because of the smaller size of the fish. Regan’s specimen is undoubtedly
older than both of ours, although no length is given. In certain fishes
of other families where a series of specimens is available for study, ranging
in development from young to adult, this particular area is one of the last
to ossify. Our adult six-foot Tarpon is a striking example.

Caudal Fin Ray Count’.

2 —l— 16 • — 18

258 mm. x J 15 = 16 Haiti specimen. There are two dorsal and
one ventral anterior raylets which lack the characteristic cross bars of
the rays.

2 -I- 17 — 19

280 mm. \ _ y5 G Gravesend Bay specimen.

Specialized Ray-scales : Partly covering the first dorsal and ventral
raylets there is a thick elongate bony ray-scale. This was present in Lepto-
lepidae, and of the Bermuda isospondylids, is seen in a more reduced size in
Tarpon , Albula, and most of the clupeids.

Megalopidae.

Tarpon atlanticus (Cuvier & Valenciennes).

(Text-figs. 16-20).

Diagnostic Characters:

8 hypurals.

Small pointed reduced neural process on the anterior part of the
first urostyle centrum.

3 distinct pairs of uroneurals.

Vertebral count: 33 + 24 = 57.

In the closely related Pacific and Indian Ocean Megalops cypri-
noides, the vertebral count is 38 + 30 = 68. (Delsman, 1926).

Material Studied.

The following description is from one adult fish, weight one hundred
pounds, length six feet, or 1,800 mm. (Text-figs. 16, 17, 18). This speci-
men was taken in Florida, the gift of Mr. and Mrs. George Arents, Jr.,
KOH Cat. No. 2085. Any differences in younger stages are mentioned,
being described from a two-foot fish, 635 mm., caught in Florida, the gift
of Mr. and Mrs. Bernard Baruch, Jr., KOH Cat. No. 2083, and from three
specimens, (Text-figs. 19, 20), taken in Haiti, Cat. No. 7303, KOH Cat.
Nos. 2031 and 2033, lengths 140, 120, 115 mm. At the time of writing
(October, 1936), the 115 mm. specimen is the smallest Tarpon available for
study in the collection of the Department of Tropical Research and in all
other institutions with which I communicated. The two-foot specimen is
essentially like the six-foot fish and the drawing for the latter represents
both stages and all those in between.

In Bermuda Tarpon are rare. A single skin was seen by G. Brown
Goode, (Catalogue of the Fishes of the Bermudas, 1876), in the collection
of John T. Bartram of St. George. We have seen Tarpon only occasionally
while helmet-diving.

264

Zoologica: New York Zoological Society

[XXI :23

Caudal Osteology.

Urostyle Centra: In the six-foot fish two complete centra, one elongate
centrum and an additional reduced terminal centrum, form the urostyle.
The elongate element extends from the fourth hypural to the tip of the
seventh. This is heavier and larger than in the younger specimens. It is

Text-figure 16.

Tarpon atlanticus. Tail of 1,800 mm. specimen (x 4/5).

irregularly shaped throughout and in the center of its length on the ventral
side it resembles a thin keel which lies between the pseudoarches of the
fifth and sixth hypurals. The terminal bony segment, which is not present
in the smaller specimens, extends three-fourths of the length of the base
of the seventh hypural. It is open above but complete below and the ossifica-
tion is thin and delicate (Text-fig. 16). The notochord is seen extending
from this last ossified segment into the caudal contour 10 mm. beyond the
dorsal, or eighth hypural. Its end is embedded in the eleventh dorsal ray
above the median line. Enclosed in a tough fibrous sheath are 18 or more
separate irregular vertebral elements. All are slightly ossified. The ossifi-
cation of the tip end which extends beyond the hypurals is heavier on the
edges and the tip than in the center. Ventrally, this seems to be solid but
dorsally it appears cleft and may be two lateral plates in close proximity
(Text-fig. 18).

In the smallest specimens (Text-fig. 19), two complete centra and one
elongate terminal centrum form the urostyle. This posterior rod-like cen-
trum shows definite indication of fusion of two centra. As it extends over
three hypurals, the fourth, fifth, and sixth, similar to Elops, it may be a
composite of three centra. The cartilaginous notochord prolongation from
this last centrum, very similar to that of Elops, extends from the base of

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

265

the sixth hypural into the dorsal caudal contour, 3 mm. beyond the eighth
hypural (Text-fig. 20). As in Elops, it is embedded in the base of the 11th
dorsal ray, above the median line.

Uroneurals : There are three pairs of uroneurals, which correspond
to ancestral posterior neural processes. All three are elongated bones and
are close together on the dorsal and lateral surfaces of the upturned uro-

Text-figure 17.

Tarpon atlanticus. The eighth hypural showing the bifid, arch-like base in three-

quarter view (x 1).

style, covering the three centra and the cartilaginous notochord, all except
the extreme tip which extends beyond the hypurals. None of these pairs is
fused one with the other and the two elements of each pair remain distinct.

The first uroneural, anterior and dorsalmost in position, is the longest
(67 mm.) and the widest (8 mm.) at the deepest part, which is the anterior
end. (The measurements of the small specimens are 11 mm. by .8 mm.
wide). The bones of this pair extend from the anterior edge of the first
centrum of the urostyle, above, but on a vertical line with the center of the
dorsal side of the 8th hypural. They end in tapering pointed tips. The an-

Text-figure 18.
Tarpon atlanticus. Tail of
1,800 mm. specimen with
paired uroneurals removed
showing, in black, the os-
sified vertebral segments
(x 4/5).

terior is rounded and unforked and differs in this respect from Elops. This
is the only pair to cover in part the dorsal as well as the lateral surfaces of
the urostyle. The anterior half is entirely lateral and above the third uro-
style centrum the two lateral bones meet dorsally, but do not fuse.

The bones of the second pair of uroneurals are spindle-shaped with
rounded anterior ends. This differs from the three small specimens of 115,

266

Zoologica: New York Zoological Society

[XXI :23

120, and 140 mm. as the illustration shows (Text-figs. 16, 19). The length
of the uroneurals is 61 mm. and the width at the widest part in the center,
is 6 mm. They arise one-fourth of the distance from the anterior edge of
the second centrum of the urostyle and extend ventral and parallel to the
first pair. Distally they end between the bases of the 10th and 11th caudal
rays, above the median line, and extend 8 mm. beyond the first pair. (The
measurements of the three small specimens are 9.2 mm. by .4 mm. wide).

The third pair of uroneurals is very small, the left bone being longer
and more slender than the right and less closely associated with the under-
lying uroneurals. The length of the left bone is 25 mm. and the width 3 mm.
It arises, approximately, at the posterior third of the second pair above
the tip of the 8th hypural. The bones extend across the tips of the 2nd
uroneurals, bending obliquely upward and projecting across the notochord
beyond the epurals. (The measurements of the small specimens are 3 mm.
by .1 mm.).

Hypurals: There are eight hypurals on the urostyle, three below and
five above the median line. Three additional long haemal spines project
into the caudal contour with fin rays attached. As in Elops there are two

Tarpon atlanticus. Tail of 150 mm. specimen (x 4.75).

hypurals ventral and adjacent to a single centrum, the second. In Tarpon
the basal ends of the two hypurals are closely associated in the small
fish and appear as one in the two-foot and six-foot specimens. In our Elops
the basal ends are distinct. The association of two hypurals with the 2nd
centrum is also present in Albula vulpes where in the adult they are in close
proximity to each other and the centrum. In all of the Bermuda Isospondyli,
other than Elops , Tarpon, and Albula, the larger and ventral bone of the
two below the 2nd centrum is entirely free from the centrum but the smaller
and dorsal bone has become united with it.

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

267

All of the hypurals remain unfused and separate for their entire length.
As in many specimens of other families, there is a noticeable band of
cartilage on the extremity of all of the hypurals and in the big Tarpon
there is a trace, here and there, of ossification on the outer edge of this
band. As in Elops, the bases of the hypurals are cup-shaped and bifid
(Text-fig. 17).

Text-figure 20.

Tarpon atlanticus. Tail of
140 mm. specimen with
paired uroneurals removed
showing, in black, the os-
sified vertebral segments
of the urostyle (x 5.5).

Epurals: There are three epurals which are all long and rod-like. They
differ slightly in shape and length, the anterior epural being the longest
and the most slender and the posterior two being the shortest and the
stoutest.

In the three small specimens there is an unossified area between the
ventral tips of the epurals and the urostyle centra. There is a plate of
cartilage here which, in both the two-foot and six-foot specimens, is ossified.
In Elops this area is considerable smaller, being partly filled by the over-
lapping of the reduced neural processes and the tips of the epurals. In
Elops a cartilage plate is present with two small round ossified islands in
the larger and one center of ossification in the smaller.

Specialized Neural Processes: In the three small fishes there is one
reduced neural process which is located on the first upturned urostyle
centrum. This is a rounded hook-shaped bone which curves abruptly toward
the posterior. In the six-foot specimen this hook-shaped reduced neural
arch is present but in addition there is a second smaller one arising on
the same centrum and both are united by and appear embedded in a median
bony plate filling the area above this centrum. Possibly this is an individual
irregularity or malformation during growth of this particular specimen.

Caudal Fin Ray Count:

1,800 mm. fish. 16
13

635 mm. fish. 17
"14

140, 120, 115 mm. 3 + 13 = 16

1 or 2 -f 13 = 14 or 15

268

Zoologica: New York Zoological Society

[XXI :23

Specialized Ray-scales : Without the study of cleared and alizarin

stained specimens, the single dorsal and ventral ray-scale would be counted
as caudal rays, so perfectly do their extremities form part of the series
of graduated caudal raylets. But in the cleared specimens the bases are
seen to be quite unlike the rays in shape, origin, and position and resemble
more the structure of the ray-scales found in Elops and Albula, and some
of the clupeids, than that of raylets. Regan (1910.1) made the following
note: “In Elops, but not in Megalops, there is an oblong ray-scale above
and below, partly covering the first upper and lower rays.”

Albulidae.

Albula vulpes (Linnaeus).

(Text-figs. 21-39).

Diagnostic Characters :

7 hypurals.

No reduced neural processes as in Elops and Tarpon.

2 distinct pairs of uroneurals in the largest adult.

4 distinct pairs in the smallest adults of 47, 40 and 36 mm.

Vertebral count in large adults:

42 + 27 = 69. Bermuda specimens.

42 + 28 = 70. Jordan and Evermann, “Fishes of North
America.”

47 + 27 = 74. Delsman, Java Sea specimens.

Material Studied.

Group

Length Cat. No.

KOH Cat. No.

Text-fig.No.

(560 mm.

1134

21, 29

A

J 510 mm.

1084

21, 29

JA

] 403 mm.

815

21

[206 mm.

9647

670

21,28

10 Adults, ranging

B

i 87 mm.

25184

768

29

from 560 to 22 mm.

1 84 mm.

25183

767

25,26

f 47 mm.

25181

2057

22, 23, 24

C

40 mm.

25182

2060

22,23, 24

[ 36 mm.

25185

2081

22, 23, 24

D

22 mm.

25208

2120

34

1 Intermediate

26 mm.

25209

2119

33

f 30.8 mm

2115

32

Leptocephalus

. \ 45 mm.

2080

31

(3 out of 15)

[ 55 mm.

147

30

This paper dealing with isospondyls is concerned more particularly
with adult specimens. Albula is an exception in its series of fish which
range in size from young to adult. The following does not aim to be a com-
plete description of the leptocephalus phase but rather a summary of the
outstanding changes occurring in the different growth stages. During
seven seasons of work in Bermuda specimens of the above range in lengths
were collected and studied. The figure at the top of the column represents
the largest and oldest and the last figure, 55 mm., the youngest. It will be
seen by reading the column from the bottom to the top that the young in
growing first decrease in length ; at the same time they change from ribbon-
like creatures to the shape of the adult. Then growth continues by length-
ening.

Although specimens have been had in collections which show Albula

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

269

in both phases, it has not been known just when the change occurs and the
rapidity of this change. I definitely established these facts by observations
on a living Albula which grew and shrank from 55 mm. to 20 mm. in ten
days’ time. During this time it changed from the leptocephalus to the adult
in body shape.2

Caudal Osteology.

Urostyle Centra: The fully adult urostyle (560, 510, 403 and 206 mm.)
is composed of two complete centra, the posterior one being greatly reduced
in size (Text-figs. 21, 29). Almost completely hidden under the uroneurals
is a posterior terminal bony mass which probably represents several fused
centra (Text-fig. 28). In the three small specimens of 47, 40, and 36 mm.,
four centra can be distinguished (Text-figs. 22, 23, 24). The two posterior
ones are rod-shaped and very close together. The posterior terminal bone

Text-figure 21.

Albula vulpes. Tail of 560 mm. specimen (x 1.4).

is above the center of the base of the sixth hypural (third above the
median line). In the 87 and 84 mm. specimens there are only three centra,
the posterior two having fused (Text-figs. 25, 26, 27).

In dissecting the 206 mm. specimen (Text-fig. 28) the second urostyle
centrum was found noticeably reduced and the fourth hypural elongated
with its base in the position of the third centrum as seen in the 84 mm.
specimen (Text-figs. 25, 26). The terminal fused centrum is pushed out of
its youthful position, where it formed the upward curve in the urostyle,
and its anterior basal end is almost superimposed on the second urostyle
centrum. The notochord is seen extending posteriorly as in smaller speci-

2 See Bulletin, New York Zoological Society, May-June 1936, Vol. XXXIX, No. 3.

270

Zoologica: New York Zoological Society

[XXI :23

mens. Anterior to the base is a minute reduced bony element which looks
like a miniature arch base. This is median in position, lying between the
sides of the specialized neural process of the urostyle.

In the two largest specimens (560 and 510 mm.), no cartilaginous
notochordal prolongation can be found after dissecting away the heavy
uroneurals. But in all others, including the 206 mm. specimen, a delicate

Text-figure 22.

Albula vulpes. Tail representing 47, 40 and 36 mm. specimens (x 15.5).

notochord extends, within the uroneurals, into the caudal contour, ending
between the bases of the eleventh caudal ray above the median line. The
notochord extends to below the tip of the sixth dorsal raylet, counting from
the anterior (Text-fig. 28).

Uroneurals : In the four largest specimens (560, 510, 403, and 206
mm.) there are two pairs of heavy uroneurals which overlap each other
(Text-figs. 21, 29). The irregular forward edge of the anterior pair can
be traced very plainly, as it covers the greater part of the first urostyle
centrum. Ventrally, its long slender tip meets the dorsal edge of the cor-
responding hypural. This anterior pair extends dorsally and posteriorly
to about mid-length of the seventh hypural, which is the dorsalmost of the
series. The posterior half is completely covered by the second pair of
uroneurals, whose origin is in the center of the second urostyle centrum,
dorsal to the second and third hypurals. This pair extends almost to the
extremity of the seventh hypural.

In the smallest specimens (47, 40, and 36 mm.), there are four pairs
of uroneurals (Text-figs. 23, 24). The two pairs which correspond to those
of the adult elongate anteriorly with growth (Text-figs. 23, 29). In com-
parison with the 45 mm. leptocephalids (Text-figs. 36, 37), these bones
have almost doubled in length, and overlap each other. In the leptocephalids
only the distal and proximal ends meet. The uroneurals appear stained
in the 30.8 mm. leptocephalus. In 84, 47, and 40 mm. specimens the anterior
tips of the first uroneural is on the third urostyle centrum (Text-figs. 23, 25) .

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

271

In the 87 mm. and all larger specimens it is on the second urostyle centrum
(Text-figs. 27, 29). There is a third pair seen only in leptocephalus stages
and 87, 84, 47, 40, 36 and 22 mm. specimens. In the leptocephalus this pair is
the most dorsal and posterior (Text-fig. 36). These uroneurals do not
elongate with growth and remain identical in length from the leptocephalus
to the 87 mm. stage and become fused in the center of the first uroneural. In
specimens of 206 mm. and over, there is no trace of this third pair (Text-
fig. 29). A fourth pair is seen in the 47, 40, and 36 mm. specimens, ex-
tending in a dorsal direction posteriorly from the reduced neural process.
This is not to be found in the largest or fully adult specimen (Text-fig. 24).

Hypurals: There are seven hypurals ventral and posterior to the uro-
style, three below and four above the median line. Four additional haemal
spines project into the caudal contour. The anterior one has above it the
specialized ray-scale, and the three others have fin-rays attached (Text-
fig. 21).

As in Elops and Tarpon the two hypurals immediately ventral to the
median line arise from what appears to be a single centrum. In all other
Bermuda adult Isospondyli the larger hypural of the two is well separated
from any centrum attachment and is one of the largest hypurals. It re-
sembles an isolated triangular island of bone. In the very young stages
of several of the clupeids there is to be seen this same attachment to a

Text-figure 23.

Albula vulpes. Tail of 47, 40 and 36 mm. specimens
showing the position of the three pairs of uroneurals
in relation to the urostyle segments and hypurals
(x 20).

Text-figure 24.

Albula vulpes. Dissection of Text-fig. 23 showing-
four urostyle segments and a fourth pair of uro-
neurals which are dotted and lettered, being under
the anterior uroneurals. The smallest pair of uro-
neurals has been omitted in order to show the ex-
tent of the underlying tips of the anterior uroneurals
(x 20).

single centrum as in Elops, Tarpon, and Albula (Text-figs. 14, 16, 21). The
bases of the hypurals are bifid as in Elops and Tarpon (Text-fig. 17, Tar-
pon) .

Epurals: In the adult there are two epurals which are heavy irregular
bones and in such close proximity that their respective outlines are difficult
to trace (Text-fig. 21). In the young of 47, 40, and 36 mm. the two epurals
are distinct (Text-fig. 22). The area above the urostyle in the young and
adult is almost completely filled with the uroneurals and the specialized
neural process. It is interesting to note here that in the longest and conse-
quently the youngest of the leptocephali, where there is no evidence of

272

Zoologica: N civ York Zoological Society

[XXI :23

Text-figure 25.

Albula vulpes. Tail of 84 mm. specimen showing
the positions of the three pairs of uroneurals in
relation to the urostyle segments and hypurals. The
large pairs have lengthened with growth but the
small pair has remained the same (x 9).

Text-figure 26.

Albula vulpes. Dissected tail of 84 mm. specimen
showing three urostyle segments instead of four as
seen in the younger stages (x 7.8).

Text-figure 27.

Albula vulpes. Tail of 87 mm. specimens showing
the positions of the three pairs of uroneurals in
relation to the urostyle segments and hypurals. The
bones of the dorsalmost uroneural have lengthened
and the anterior tips are on the second urostyle seg-
ment instead of the third as in smaller specimens.
The length of the small pair has remained un-
changed (x 8).

Text-figure 28.

Albula vulpes. Dissected tail of 206 mm. specimen
showing the more consolidated urostyle and remains
of the fourth uroneurals of Text-fig. 24 (x 2.5).

segmentation in the notochord nor absorption of alizarin, the basal ends
of the two epurals are united (Text-fig. 36).

Specialized Neural Process: There is one reduced neural process on
the first centrum of the urostyle (Text-fig. 21). This is ossified and de-

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

273

veloped in the 36 mm. young-adults but cannot be detected, even in cartilage
form, in any of the younger specimens (Text-figs. 22, 36).

Caudal Fin Ray Count:

55 mm. 11

TT

45 mm. IT

TT

30 mm.

2

4-

10

=

12

2

+

10

=4

12

26 mm.

3

+

10

=

13

2

+

10

=

12

22 mm.

4

+

10

14

4

+

10

—

14

36 mm.
40 mm.
47 mm.

6

+

12

18

3

+

13

16

84 mm.

3

4-

15

-

18

2

+

14

=

16

87 mm.

3

4-

15

==

18

1

4-

15

=

16

206 mm.
403 mm.
510 mm.

1

4-

17

=

18

1

4

15

=

16

560 mm. 18

T6

Leptocephalus. Caudal rays
unstained (Text-fig. 36).

Caudal rays stained.

No trace of dorsal or ventral
caudal ray-scale or body scales.
(Text-fig. 37).

Caudal ray-scale present. Body
scales present. (Text-fig. 22).

(Text-fig. 21).

(Text-fig. 21).

Specialized Ray-scales : As in Elops and Tarpon, a thick, elongate, bony
ray-scale partly covers the first dorsal and ventral anterior raylet. In
Albula this structure is heavier than in any of the other isospondylids
(Text-figs. 21, 22). It is well developed in all specimens 36 mm. and longer

Text-figure 29.

Albula vulpes. Tail of largest adult showing uro-
neurals and segments of the urostyle reduced to two
elements respectively (x 1.5).

but not evident, even in cartilage form, in any of the younger stages or
leptocephalids.

Additional Characters Worthy of Note: All of the posterior neural
and haemal spines are very heavy and thick and by this character alone
adult Albula can be identified and distinguished from all other Bermuda
isospondylids (Text-fig. 21). It is interesting to note that in several speci-
mens there is a double neural spine structure (Text-figs. 21, 22). This

274

[XXI :23

Text-figures 30-39.

Albula vulpes. Summary of the caudal skeleton development and corresponding
change in body form from a 55 mm. leptocephalus to a 560 mm. adult. The
five upper figures in the left column are natural size. Text-figure 35 is x 1/9.

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

275

occurs in specimens of the following lengths: 510, 403, 87, 47, and 36 mm.
In two of these the double structure is on the next to the last, and in the
other specimens it is on the last vertebra. This is not correlated with
size, for in the other specimens of the series which are 560, 84, 40, and 22
mm., no double neural structure is present.

The comparative development of the skeleton is interesting in the
series at hand. In the longest leptocephalus (55 mm.) the notochord is not
yet segmented, nor has it taken up any alizarin (Text-fig. 36). In the
successively shorter leptocephali more and more vertical lines appear at
regular intervals which are the antecedents of the adult centra (Text-fig.
37). In all of the small specimens having the form of the adult, 22 mm. and
larger, there is vertebral differentiation which is less distinct in the~ small-
est specimen. Here the centra have not yet developed the shape of the
adult centra. The vertebrae in the smallest specimens are rectangular
and about twice as deep as wide (Text-figs. 21, 22). In all the small speci-
mens there is ossification in the head and caudal regions. In the smallest
fish of 22 mm. the only ossification of the notochord is in the caudal
region, on the dorsal and ventral surfaces, near the bases of the neural and
haemal spines.

It is consistent with the digging habits of Alhula that one of the first
areas of ossification should be the snout and head.

The first appearance of scales is in the 36 mm. specimen where they
appear heavily stained. No scales are apparent on the 22 mm. fish.

The first appearance of ossification in the notochord is in the 22 mm.
fish, and the urostyle segments are more definitely defined than in the
younger intermediate 26 mm. specimen.

The dorsal fin-fold is absent for the first time in the 22 mm. speci-
men (Text-fig. 34).

It is interesting to note the rapidity of development which occurs be-
tween the 30.8 mm. leptocephalus and the 20 mm. young. In the Alhula 3
that lived for ten days it took exactly six days to grow from 30 to 20 mm.
It is just here during the life span of Alhula that the unossified ribbon-like
leptocephalus changes into an ossified, compact fish. In the shortening of
the length the embryonic fin-folds disappear, the dorsal and anal fins and
anus move forward (Text-figs. 32, 33, 34).

In the 26 mm. specimen, which is here termed an intermediate stage,
change in the external shape of the body is more advanced than that of the
internal and caudal skeleton, (Text-fig. 33). This resembles the 30.8 mm,
leptocephalus more than the 22 mm. older form. In the 22 mm. fish the
external body form resembles that of the adult (Text-figs. 34, 35). Again
the development of the body is more advanced than that of the internal and
caudal skeleton. In this stage is the first appearance of ossification around
the notochord.

Table I.

Records of Extreme Lengths

Leptocephalus

Adult

From Literature

85 to 40 mm.

44 to 1220 mm.

From Bermuda Specimens

55 to 30.8 mm.

20s to 560 mm.

Intermediate

26 mm.

3 This specimen is not cleared and stained, as it is the fish reared from a 55 mm. leptocephalus
to a 20 mm. young Alhula, where the body resembled the adult. It will be seen from the figures
stated above in literature that this specimen probably is the smallest Alhula vulpes in any collection.
See the New York Zoological Society Bulletin, May-June 1936, Vol. XXXIX, No. 3.

276

Zoologica: New York Zoological Society

[XXI :23

Dussumieriidae.

Jenkinsia lampro taenia (Gosse).

(Text-figs. 40-44).

Diagnostic Characters:

7 hypurals.

1 reduced neural process. This is situated on the anterior part of
the urostyle and, unlike Elops and Tarpon, it is forwardly
directed.

3 distinct pairs of uroneurals.

Vertebral count: 27 + 16 = 43.

1 epural, the only isospondylid with one.

Prolonged bases of the two median caudal rays.

Total caudal ray count of 26, the smallest of the whole group.

Specialized neural processes unlike those of all other Bermuda
isospondylids.

Material Studied.

This description is taken from the five following specimens and com-

Jenkinsia lamprotaenia. Tail of 10 mm. specimen with unossified and unseg-
mented vertebral column. There are only five hypurals and one pair of uroneurals

ossified (x 140).

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

277

parative studies were made with fifty-five additional fish which range in
size from young to adult.

Length

KOH Cat. No.

Text-fig. No.

40 mm.

2096

26 mm.

343

44

18 mm.

657

43

15 mm.

657

42

11 mm.

657

41

10 mm.

657

40

Caudal Osteology.

Urostyle Centra: The adult urostyle appears as one bone. Anteriorly
it is shaped like a half centrum and posteriorly it has a slender upturned
end which is situated between the fourth and fifth hypurals, counting from
the anterior. The distal end has projecting ventrally a fan-shaped bone
that extends over the bases of the three dorsal hypurals, and covers an
unossified area between the hypural bases and the end of the urostyle.
In the young stages this bone is not present. It is first seen in an 18 mm.
fish (Text-fig. 43). This space does not exist before the uroneurals appear
and the urostyle has become fully ossified and reduced in size. In this
stage the distal end of the urostyle and the bases of the hypurals almost
meet (Text-fig. 43). With growth more and more space in this area ap-
pears. The urostyle is definitely divided into two parts in the young stages.
The line of junction shows plainly in the adult (Text-fig. 44).

Text-figure 41.

Jenkinsia lamprotaenia. Tail of 11 mm. specimen with unossified and unseg-
mented vertebral column. There are six hypurals and two pairs of uroneurals

ossified (x 108).

278 Zoologica: New York Zoological Society [XXI :23

Text-figure 42.

Jenkinsia lamprotaenia.. Tail of 15 mm. specimen with ossified and segmented
vertebral column. There are seven hypurals and three pairs of uroneurals os-
sified. The basal end of the second hypural is reduced (x 64).

Text-figure 43.

Jenkinsia lamprotaenia. Tail of 18 mm. specimen with ossified neural and haemal

processes (x 59).

1936] Hollister: Caudal Skeleton of Bermuda Fishes 279

Text-figure 44.

Jenkinsia lamprotaenia. Tail of 26 mm. specimen with single epural ossified

(x 39).

Uroneurals: There are three pairs of uroneurals in the adult. Each
pair is quite different in shape, size, and general position, as may be seen
in the accompanying drawings. The bones are numbered according to the
order of their appearance. In several young stages the first pair appear
as two separate bones which in the older stages fuse into a solid structure
which is directly dorsal to the urostyle. It extends from the tip of the
reduced adult urostyle into the caudal contour. In the older fish of 18 mm.
and larger there is a median wing-shaped bone on the anterior three-fourths
of the dorsal side. In specimens of 26 and 30 mm. this bone fills all the
center of the unossified area which is between the posterior neural spine,
the reduced neural process on the urostyle, and the epural.

The second pair of uroneurals first appears in the 11 mm. specimen.
Here the two short lateral bones extend along the cartilaginous tip of the
urostyle. In a 15 mm. specimen the bones are still separate but have
enlarged in length and depth. In specimens of 18, 26, and 30 mm. the two
bones are united. In each example the proximal ends overlap the tip of
the ossified urostyle. The two bones of the diminutive third pair remain
individual in all stages. They cannot be distinguished in a 11 mm. fish
but in a 15 mm. specimen they are present. Their position is between the
bases of the 10th and 11th dorsal caudal rays which are above the median
line. There is a noticeable space in specimens of all lengths and here the
cartilaginous prolongation of the notochard extends. The posterior ends
of these bones arise beyond the caudal contour of the hypurals and extend
obliquely forward and downward. (Text-figs. 41, 42).

The third pair of uroneurals is first seen in a 15 mm. specimen. This
is the smallest of the three pairs, and the direction of the bones is more
toward the anterior than that of the others. In the 15 mm. specimen the

280 Zoologica: New York Zoological Society [XXI :23

position is between the dorsal tips of the first and second uroneurals with
the anterior end of each bone crossing the end of the first uroneural. The
posterior ends extend beyond the distal edge of the hypurals and occupy the
area between the ray bases where the notochord extends. The accompanying
drawings show the position and lengthening with growth (Text-figs. 42,
43,44).

Hypurals : There are seven hypurals, three below and five above the
median line. Two additional long haemal spines project into the caudal
contour with fin-rays attached. In eleven young fish between the lengths of
10 and 13.44 mm., the second hypural is complete and the same length as
the other hypurals. In twenty specimens between the lengths of 13.58 mm.
and 40 mm., the base of the second hypural is reduced. It is entirely sepa-
rate and free from the other hypurals and appears as a triangular island.

Epurals: There is only one epural in Jenkinsia which is first seen ossi-
fied in the 26 mm. fish. This bone can be seen in the smaller specimen but
it is entirely unossified. Jenkinsia is the only Bermuda isospondyl having
but one epural.

Specialized Neural Processes: In the completely ossified specimens of
26 mm. and larger there is a hook-shaped neural process on the posterior
part of the urostyle. In the center of the urostyle this becomes very narrow
and again expands posteriorly. The proximal end of the first uroneural is
inserted between its tip ends. Dorsal to both is a thin median wing-like
bone which is first seen in the 26 mm. specimen (Text-fig. 44).

Caudal Fin Ray Count:

26 mm.

2

+

12 =

= 14

(Text-fig.

44).

1

+

11 =

= 12

18 mm.

3

+

11 =

= 14

(Text-fig.

43).

2

+

10 =

= 12

15 mm.

2

+

11 =

= 13

(Text-fig.

42).

1

+

10 =

= 11

11 mm.

10

(Text-figs

. 41, 40)

10 mm.

9

Additional Characters Worthy of Note: In the fully ossified specimens
of 26 mm. and larger the two median caudal rays have enlarged bases
which project anteriorly half the length of the hypurals (Text-fig. 44).
Among the Bermuda isospondyls this character is seen also in Anchoviella
and the clupeids.

Enlarged wings on the anterior margin of the proximal ends of the
haemal and neural spines are seen first in the 26 mm. specimen (Text-fig.
44).

In the youngest specimens vertebral ossification and individual ver-
tebrae are first seen in 15 mm. fish. Here the urostyle, hypurals and caudal
rays are well ossified. In the 10 mm. and 11 mm. specimens the vertebral
column is not constricted and there is no ossification anterior to the urostyle.

According to caudal pattern and characters Jenkinsia seems to stand
apart from the other Bermuda isospondyls. Conversely the three clupeids
and Anchoviella are closely associated by similar patterns and characters.

Engraulidae.

Anchoviella choerostoma (Goode).

(Text-figs. 45, 46).

Diagnostic Characters:

7 hypurals.

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

281

Text-figure 45.

Anchoviella choerostoma. Tail of 57 mm. specimen (x 11.5).

1 small reduced neural on the anterior base of the elongate neural
process of the urostyle.

Vertebral count:

20 + 20 = 40. 58 mm.

21 + 20 = 41. 50 mm.

21 + 21 = 42. 42 mm.

2 epurals and occasionally 2%.

Prolonged bases of the two median caudal rays.

Material Studied .

This description is based on twenty-nine specimens ranging in size
from 20 to '59 mm. The 20 mm. fish is the smallest Anchoviella in our col-
lection and is identical externally with the largest (59 mm.) fish. But the
internal caudal development is not the same in the young specimen, which

Text-figure 46.
Anchoviella choerostoma.
Tail of 20 mm. specimen
(x 28).

282

Zoologica: New York Zoological Society

[XXI :23

can be seen by the accompanying illustrations. The two specimens studied
in particular are:

Length KOHCat.No. Text-fig. No.

57 mm. 875 45

20 mm. 2095 46

Caudal Osteology.

Urostyle Centra: The interpretation of the adult urostyle is clarified
by first studying the structure of the youngest specimen at hand. Here
there are three distinct vertebral segments and the anterior part of a
fourth is partly visible. (Text-fig. 46). In the large specimen, judging by
the comparative position of the hypurals with that in the small fish, the two
anterior segments fuse and the two posterior segments fuse. The posterior
segment in the large specimen is further complicated by the presence of
a small arch-like bone on the ventral side. This surrounds the proximal
end of the fourth hypural. A similar structure is found in adults of the
three clupeids.

Uroneurals: There are three pairs of uroneurals in the adult and the
young specimen. The relative length, proximal and distal positions are
the same in both fish. The anterior end of the pair that arises on the first
urostyle centrum is not distinctly defined but there is little doubt that a
smaller specimen would show the definite outline of the end of this bone.

Epurals: There are two and sometimes two and a half epurals in

Anchoviella. The variation always occurs in the anterior bone which may
be split in the form of a Y or have a hole in the center. This variation
does not correlate with size. In fifteen specimens six have two epurals and
the others have either the split or a hole in the anterior bone.

Specialized Neural Process: In the large specimens a long dagger-shaped
bone extends from the anterior urostyle centrum to about the center of the
anterior epural. In the 20 mm. specimen this bone is shorter and does not
reach the basal end of the epural. This character, along with several others,
distinguishes Anchoviella from the clupeids.

Caudal Fin Ray Count:

57 mm. 5 + 13 = 18
5 + 12 = 17

In seven specimens out of fifteen, the caudal count is as stated above
and is found in fish from 28 to 59 mm. In eight specimens there is variation
in the total count from ffErt- In the specimens examined the count of the
dorsal and ventral is never it as in the clupeids.

Clupeidae.

1. Harengula sp.

(Text-figs. 47, 48).

Diagnostic Characters:

7 hypurals.

1 reduced neural process. This process is insignificant and re-
sembles the anterior neural zygapophysis.

3 pairs of uroneurals.

Vertebral Count: 12 to 14 + 25 to 26 = 37 to 40.

2 or 3 epurals.

Prolonged bases of two median caudal rays.

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

283

Text-figure 47.

Harengula sp. Tail of 165 mm. specimen (x 3.6).

Material Studied.

The KOH collection contains seventeen specimens ranging from 22 to
180 mm. The figure of the adult represents the tail structure of fish from
70 mm. to the largest. The illustration of the 22 mm. specimen, which is
the smallest in the department collection, shows slight differences in degree
of development. The two specimens studied in particular are the following:

Text-fig. No.

47

48

Text-figure 48.
Harengula sp. Tail of 22
mm. specimen (x 21).

Length KOH Cat. No.

165 mm. 847

22 mm. 656

284

Zoologicq: New York Zoological Society

[XXI :23

Caudal Osteology.

Urostyle Centra: As in Anchoviella, the interpretation of the adult
urostyle is best made by examining the youngest specimen (Text-fig. 48).
Here, as in Anchoviella, there are four centra or vertebral segments in the
urostyle. The anterior or first and second are separate but the third and
fourth appear fused. In the adult all but the anterior or first segment have
become comparatively reduced. The fourth cannot be seen under the heavy
uroneurals. As in Anchoviella the first hypural is attached to the anterior
urostyle segment and the third and fourth hypurals are attached to the
second and third urostyle segment.

Uroneurals : As in Anchoviella there are three pairs of uroneurals in
the adult and the young specimens. The illustrations show the relative size
and position of these bones.

Epurals: There are two or three epurals in Harengula. In seven speci-
mens ranging from 35 to 180 mm., I found two epural bones and in five fish
ranging from 22 to 173 mm. there were three epurals. The variation of a
split or perforated bone in Anchoviella has not been found in Harengula.

Specialized Neural Process: In the large specimen a stout dagger-
shaped bone extends from the anterior urostyle segment as far as the basal
end of the anterior epural. In the 22 mm. fish this bone is shorter (Text-
fig. 48). In Anchoviella this bone extends beyond the end of the epural.

Caudal Fin Ray Count:

165 mm. 1 + 18 = 19 (Text-fig. 47).

1 + 15 = 16
40 mm. 5 + 14 = 19

36 mm. 4 + 12 == 16

In the specimens counted the dorsal and ventral combination was con-
stant, 19/16, and never like that in Anchoviella.

Additional Characters Worthy of Note: The caudal pattern of Haren-
gula is very like the two other Bermuda clupeids. In the number of Haren-
gula specimens examined, only minor differences have been seen, such as the
variation in the number of the epurals. This variation occurs in all the
clupeids and in Anchoviella. Most Bermuda Harengula key to sardina as
described in the “Field Book of the Shore Fishes of Bermuda,” Beebe and
Tee-Van.

2. Opisthonema oglinum (Le Sueur).

(Text-figs. 49, 50, 51).

Diagnostic Characters :

7 hypurals.

Small reduced pointed neural on the anterior of the urostyle.

Vertebral Count: 16 + 29 = 45.

3 epurals.

Prolonged bases of two median caudal rays.

Material Studied.

This description is based on five specimens of the following lengths:

No. of specimens Length KOHCat.No. Text-fig. No.

2 110 mm. 850

3 75 mm. 849

49, 50, 51

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

285

Text-figure 49.

Opisthonema oglinum. Tail of 75 mm. specimen (x 4.3).

These lengths represent the smallest and largest Opisthonema captured
during our seven years in the field in Bermuda. The 75 mm. specimens are
identical in development with the 110 mm. fish.

Caudal Osteology.

TJrostyle Centra: This is very like Harengula , as the figures show.
The dissection made by removing the superimposed uroneurals shows the
adult elements of the urostyle. This condition is probably representative of
all the Bermuda clupeids.

Uroneurals : The bones of the second pair of uroneurals are stout and
noticeable heavier than those of Sardinella and Harengula. In other respects
there is similarity between the three genera.

Epurals: There are three epurals found in the five specimens examined.

Text-figure 50.

Opisthonema oglinum. Tail of 75 mm. specimen dis-
sected to show structure of the urostyle (x 3.7).

286 Zoologica: New York Zoological Society [XXI :23

Text-figure 51.

Opisthonema oglinum. Last ossified segment of urostyle with cartilaginous
notochord extending. A. View from the top; B. View from the side (x 8.5).

Specialized Neural Process: This stout neural bone is similar to that of
Sardinella and Harengula in relative size and position.

Caudal Fin Ray Count: The count is identical with that of Sardinella
and Harengula, 19/16.

Additional Characters Worthy of Note: In the specimens examined,
Opisthonema differs from the other clupeids in the noticeably heavier sec-
ond uroneurals and in the shape of the distal ventral surface of the fourth
hypural, which is more or less even instead of having a sharp projecting
point. The basal end of the second hypural is blunt instead of hammer-
shaped as in Sardinella.

3. Sardinella anchovia Cuvier and Valenciennes.

(Text-figs. 52, 53).

Diagnostic Characters :

7 hypurals.

Small reduced blunt neural on the anterior part of urostyle.

Vertebral Count: 14 or 15 + 31 or 32 = 46.

2% or 3 epurals.

Prolonged bases of the median caudal rays.

Material Studied.

Seven KOH specimens have been examined, which range in size from
32 to 135 mm. The caudal pattern is identical in specimens of 50 mm. and
larger.

Length

KOH. Cat. No.

Text-fig. No ,

60 mm. (
50 mm. \

671

52

34 mm. )
32 mm. \

2112

53

Caudal Osteology.

Urostyle Centra, Uroneurals: Almost identical with Opisthonema and
Harengula.

Epurals: There are three epurals in the 32 and 34 mm. specimens. In
the four larger specimens ranging from 50 to 135 mm., there are two and a
half, the anterior bone being dorsally bifid. This variation is seen occa-
sionally in Anchoviella.

Specialized Neural Process: This bone is almost the counterpart of that
seen in Opisthonema.

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

287

Caudal Fin Ray Count: As in Oyisthonema and Harengula, the caudal
count is 19/16.

Additional Characters Worthy of Note: The shape of the ventral distal
edge of the fourth hypural differs from Oyisthonema in that it has a sharp
point. The basal end of the second hypural is hammer-shaped and this, too,
differs from Oyisthonema.

Summary.

The following paragraphs correlate the salient similarities and dif-
ferences found in the study of the caudal Skeletons of the Bermuda Isos-
pondyli. According to caudal pattern and characters Jenkinsia seems to

Text-figure 53.
Sardinella anchovia. Tail
of 34 and 32 mm. speci-
mens (x 16.5).

288

Zoologicci: New York Zoological Society

[XXI :23

stand apart from the other Bermuda Isospondyli. Conversely, the three
species of clupeids and Anchoviella are closely associated by similar patterns
and characters.

Hypurals : In the first three families described, Elopidae, Megalopidae,
and Albulidae, there is no noticeable deviation of the general pattern of
the hypurals from that of Leptolepis dubius, which is the most primitive
ancestor of the Isospondyli. All the hypurals are long, expanded haemal
bones. But in the adults of all the other Bermuda Isospondyli the second
hypural lacks the basal part and is a triangular bone with a reduced and
free base. However in the few young specimens of Jenkinsia available for
study the primitive, ancestral, unreduced, second hypural is present and
resembles that of adult Elops, Tarpon, and Albula.

There are nine hypurals in Elops, eight in Tarpon, and seven in
Albula. In the first two, where the urostyle is turned up less abruptly and
the segments are less consolidated than in the other Bermuda Isospondyli,
there are more hypurals. It will be seen in the figures of the key that
Elops has one more hypural than Tarpon, which is numbered 0 (Text-figs. 3
& 4). Also that the dorsalmost hypural which is numbered 8 in Tarpon is
not present in Albula. Albula vulpes and the other species described in this
paper have seven hypurals. Hypurals with similar numbers correspond in
position in all the specimens. (Text-fig. 5).

The following table correlates the hypural count in the various genera:

Total hypurals

Elops

9

Tarpon

8

All other Bermuda
Albula Isospondyli
7 7

Dorsal hypurals

5

5

4

4

Ventral hypurals

4

3

3

3

Epurals : All Bermuda Isospondyli have more than one epural with the
exception of Jenkinsia which has only one. This is the last bone to become
fully ossified in Jenkinsia.

Caudal Ray Count : With growth and increase in size the dorsal and
ventral raylets change into rays, resulting in a shift of relative numbers of
the two elements.

In Albula the total caudal count of dorsal 18 and ventral 16 remains
constant in specimens ranging from 36 to 560 mm. But the dorsal and
ventral raylets diminish from 6 and 3 respectively to dorsal 1 and ventral 1,
with a corresponding increase of rays.

Jenkinsia has the smallest caudal count of the Bermuda Isospondyli,
which is a total of 26 as opposed to 30 or more in the other species.

The long functional caudal rays are present in Albula leptocephali and
also in very small Jenkinsia. With growth, additional smaller anterior rays
and raylets appear.

Prolonged Median Rays : In Elops, Tarpon and Albula there are no
prolonged bases of the two median caudal rays. All the other Bermuda Isos-
pondyli have these bases prolonged.

Ray-scale : The ray-scale which is prominent in Elops, Tarpon and
Albula is less obvious in the remaining Isospondyli.

General Observations : The study of a series of young Albula and
Jenkinsia show that development in the caudal region commences at the
posterior extremity and progresses toward the anterior. Segmentation and
ossification of the notochord begins in the urostyle region and the hypurals
are the first bones to appear. The neural and haemal processes increase
anteriorly with growth, the first to appear being near the urostyle. In

1936]

Hollister: Caudal Skeleton of Bermuda Fishes

289

Jenkinsia the anterior-ventral hypurals appear first and with increase in
size additional dorsal hypurals appear.

In Albula and Jenkinsia the caudal skeleton becomes ossified when the
fish is very small and from this stage the changes that occur during growth
are changes of degree and not of kind.

In Tarpon and Elops the last part to ossify is the area directly above
the urostyle. This is filled by a preformed cartilage plate. In our two
Elops specimens of 258 and 280 mm. there are two round islands of bone
within this cartilage which indicate centers of ossification. Regan’s speci-
men has a solid bone.

Bibliography.

I. References pertaining principally to the Isospondyli.

Bruch, C.

1861. Vergleichende Osteologie des Rheinlachses ( Salmo salar L.) mit be-
sonderer Beriicksichtigung der Myologie nebst einleitenden Bemerk-
ungen iiber die skelettbildenden Gewebe der Wirbelthiere.

Mainz, 1861, 22 p., 7 pis.

(A second edition was published at Mainz in 1875. 25 p., 7 col. pis.).
Delsman, H. C.

1926. Fish eggs and larvae from the Java Sea. Treubia, Batavia, vol. 8,
1926, pp. 389-412, figs.

(Page 408, vertebral counts of Megalops, Albula, and Elops).

Gill, T.

1905. The tarpon and lady-fish and their relatives. Smithson. Misc. Collect.,
1905, (1907), vol. 48, pt. 3, pp. 31-46, 4 pis.

(Morphology discussed, figures of skulls).

Kolliker, R. A.

1860. Untersuchungen iiber das Ende der Wirbelsaule der lebenden Ganoiden
und einiger Teleostier. Gratulationsschrift zur 400 jahr. Jubilaumsfeier
der Univ. Basel, Leipzig, 1860, 27 p., 4 pis., 4°.

(Excellent plates of Polypterus, Amia, Lepidosteus, Salmo, Cyprinus) .
MacBride, E. W.

1932. Recent work on the development of the vertebral column.

Biol. Rev., Cambridge, vol. VII, No. 2, April 1932, pp. 108-148, 48 figs.
(Page 121, drawing of herring tail).

Newton, E. T.

1882-84. On Fishes’ tails.

Journ. Quekett Micr. Club, 1882-84, vol. I.

(1 fig. of young Harenga sprattus) .

(1 fig. of Lepidosteus, after Kolliker).

Ramanujam, S. G. M.

1929. The study of the development of the vertebral column in Teleosts, as
shown in the life-history of the herring.

Proc. Zool. Soc. London, 1929, pp. 365-414, 28 figs.

Regan, C. T.

1909. A revision of the fishes of the genus Elops. Ann. Mag. Nat. Hist.,
1909, 8 ser., vol. Ill, pp. 37-40.

(Describes, on page 37, last three vertebrae. This revision is based
principally on vertebral counts) .

1910.1. The caudal fin of the Elopidae and of some other teleostean fishes.
Ann. Mag. Nat. Hist., 1910, 8 ser., vol. 5, pp. 354-358. 2 figs.

(Good drawing of Elops tail and upper caudal rays of Megalops cypri-
noides and Tarpon atlanticus) .

290

Zoologica: New York Zoological Society

1910.2. On the caudal fin of the Clupeidae, and on the teleostean urostyle.
Ann. Mag. Nat. Hist., 1910, 8 ser., vol. 5, pp. 531-533. 2 figs.

(Excellent drawings of young Clupea and of Chatoessus erebi) .

Whitehouse, R. H.

1910.1. The caudal fin of fishes (preliminary paper).

Proc. Roy. Soc. London, 1910, vol. 82 B, pp. 134-143, 4 figs.

(Figure of Clupea pilchardus) .

1910.2. The caudal fin of the Teleostomi.

Proc. Zool. Soc. London, 1910, pp. 590-629. 4 pis.

(Includes Clupea pilchardus).

1910.3. Some remarks on the teleostean caudal fin.

Ann. Mag. Nat. Hist., 1910, 8 ser., vol. 5, pp. 426-428.

(Discussion of Clupea tail and of Regan’s comments).

Woodward, A. S.

1889-1901. Catalogue of the fossil fishes in the British Museum Natural
History. 4 vols., London, 1889-1901, 79 pis. & 137 figs. 8°.

(In vol. Ill, pp. 500-530, in Family Leptolepidae, is the tail of Lep-
tolepis dubius, of the Jurassic period, most primitive of the isospon-
dylids).

II. General references of importance which, with their bibliographies, include all
the known literature on the caudal fin of fishes.

Agassiz, Alexander.

1878. On the young stages of some osseus fishes.

Proc. Amer. Acad. Arts and Sciences, 1878, vol. XIII, pp. 117-127,
2 plates, 32 figs.

Dean, B.

1895. Fishes, living and fossil; an outline of their forms and probable rela-
tionships. New York and London, 1895. XIV, 300 p., 344 figs.

(Pages 35 to 39 contain notes on the caudal fin).

Ryder, J. A.

1886. On the origin of heterocercy and the evolution of the fins and fin-rays
of fishes.

Rept. U. S. Fish Comm., 1884 (1886), vol. 12, pp. 981-1106, 11 pis.

Whitehouse, R. H.

1910. Caudal Fin of the Teleostomi.

Proc. Zool. Soc. London, 1910, Vol. II.

(Review of earlier literature on caudal fin. Terminology, and mor-
phology discussed. Bibliography containing 22 references).

1914. Evolution of the Caudal Fin of Fishes.

Report British Assoc. Advanc. Science, 1914, pp. 522-523.

Index to Parts 1-4

291

INDEX

Names in bold face indicate new species ; numbers in bold face indicate illustrations.

A

Abyla dentata, 233
Abylopsis eschscholtzii, 233
tetragona, 233
Acantharcus, 38
pomotis, 2, 34, 39
Acanthopterygii, 252
Acanthurus, 199
Acartia clausi, 89, 92
danae, 92

Acheilognathis intermedia, 252
Acheilognathus intermedium, 243
Acineta tuberosa, 83
Actaea angusta, 215
Actaea crockeri, 215
Adnia ( Adinia ?) dugesi, 251
Aequidens, 17, 18, 30
latifrons, 40, 41
Aetideus armatus, 90
Agalma elegans, 236
okeni, 236

Aggregata eberthi, 141

Albula, 262, 263, 266, 268, 271, 273, 275, 288
289

vulpes, opp. 260, 266, 268, 288 (Text-figs.
21-39)

Albulidae, 257, opp. 260, 288
Alciopa cantraini, 85
Alciope, 147
Alciopidae, 49, 59
Alepisaurus, 201, 202, 205
Alewife, landlocked, 165 (Text-figs. 1-6)
Allotis humilis, 21, 28, 36
Amaueria, 99
Ambloplites, 19, 36, 38
cavifrons, 34

rupestris, 2, 15, 16, 17, 33, 39, 48, 252
Ambloplitini, 38, 42
Ameiurus, 9, 20
nebulosus, 40, 252
Amia, 36

Ammotrypane bermudiensis, 60, 67
Amoeba, 121
Amphicaryon acaule, 231
Amphidinium, 153
Amphinomidae, 49, 50
Amphipoda, 77, 79, 196, 197
Amphiprion percula, 150
Anchialina typia, 96
AnchovieUa, 203

choerostoma, opp. 260, 280, (Text-figs. 45,
46)

Anchovies, 197
Angelfish, French, 150
Angelichthys isabelita, 150
Anguilla rostrata, 166
Annelids, polychaetous, 49 (PI. I-III)

Anolis, 10

Anthophysa formosa, 237
rosea, 237

Aphredoderus sayanus, 252
Aplites, 1, 7, 9, 20, 22, 25, 27, 35, 36, 38
salmoides, 2, 5, 17, 20, 24
Apodinium, 138, 147, 150, 151
Apogonidae, 37
Apomotis, 38

cyanellus, 2, 21, 26, 39
Archoplites, 38
inter ruptus, 33
Ardea cinerea, 253

Argyropelecus aculeatus, 201, 202, 205

Aricia setosa, 55

Aricidae, 49, 55

Armandia polyophthalmia, 86

Aster omphalus heptactis, 77

Astrophaeroidea, 77

Asymmetron, 61

Atelodinium, 157
Athorybia rosacea, 237
sp., 237

Audouinea pygidia, 64, 67

Avocettina infans, 201

B

Balaenanemertes, 100
lobata, 109
musculo caudata, 109

Balaenanemertes minor, 98, 99, 109, 111, 113
Balistes forcipatus, 204
Bass, black, 1, 5, 9, 21, 24
calico, 35

large-mouth, 13, 25, 26, 252
small-mouth, 13, 22, 24, 25, 252
common sea, 149
rock, 252
striped, 149

Bathycalanus richiardi, 90
rigidus, 90

Bentheuphausia amblyops, 95
Benthodesmus atlanticus, 201, 202, 205
Bhawania goodei, 54
Bitterling, 241 (PI. I; Text-fig. 1)
Blastodinium, 140, 141, 147, 154, 156
Bluefish, 149

Boxfish, spiny, 149, 150, 151
Brama raii, 197, 201, 202
Bregmaceros macclellandii, 77
Buergeriella, 99
Bullhead, common, 252
Butterfiyfish, 197, 198, 200

c

Calanoids, 76, 89
Calanopia elliptica, 89, 92
Calanus propinquus, 89
Caligus, 211
curtus, 89, 93
Callichthys asper, 251
Calocalanus pavo, 90
Cambarus, 18

Candacia aethiopica, 89, 92, 199, 200
simplex, 89, 92
Caranx crysos, 149
hippos, 149

Carassius carassius, 252
Carideans, 197, 199, 200
Catostomus commersoni, 252
Cavolinia, 196
Centrarchidae, 1
Centrarchinae, 38, 42
Centrarchus, 38
macropterus, 34
Centropages violaceus, 91
Centropristis striatus, 149
Cephalacanthus volitans, 204
Cephalochordate, 61
Cephalopods, 196
Cerataspis monstrosa, 197
Ceratium, 141, 143, 154, 156
furca, 140
fusus, 81, 140
hirudinella, 140, 152, 155
karsteni, 81
trichocerus, 81
tripos, 140, 154-5
tripos var. atlantica, 81
Ceratocymba sagittata, 233
Ceratonereis mirabilis, 86
Chaenobryttus, 38
gulosus, 2, 21, 26, 39
Chaetobranchus flavescens, 252
gulosus, 252

Chaetodipterus faber, 150

292 Zoologica: New

Chaetodon, 197
capistratus, 150
sedentarius, 199, 200
Chaetognatha, 79
Chaetosphaera, 86
Chaetosomus brachyurus, 252
Chasmocarcinus ferrugineus, 216
Chasmocarcinus latipes, 217
Chelophyes appendiculata, 234
contorta, 234

Chilomycterus schoepfii, 130, 149
Chiridius poppei, 90
Chirundina street six, 90
Chlorella, 120
Chriopeops goodei, 251
Chrysichthys kingselyi, 252
Chrysopetalidae, 49, 54
Chrytriodinium, 147, 150, 151
parasiticum, 147
Chuniella, 100

lanceolata, 98, 99, 108
? Chuniphyes multidentata, 235
Cichlidae, 19, 29, 30, 36, 37, 41
Cirratulidae, 49, 63
Cirratulus multicirratus, 63, 68
Clarias angolensis, 252
Clausocalanus arcuicornis, 89, 90
Clinostomum africanus, 252
attenuatum, 254
chrysichthys, 252
clarias, 252

complanatum, 251, (PI. I and II)

dalgi, 252

dictyotum, 252

gracile, 253

heterostomum, 252

intermediate, 252

marginatum, 253, 254

piscidium, 252

pseudoheterosternum, 254

sp., 253, 254

Clupeidae, 211, 257, opp. 260, 282
Clytemnestra scutellata, 92
Cobitis taenia, 252
Codonella amphorella, 82, 84
angusta, 82
apicata, 82, 84
nationalis, 82, 84
oceanica, 82, 84
rapa, 82, 84
recta, 82

Codonellopsis longa, 83, 84
tessellata, 83, 84
Collozoum inerme, 141
Conaea rapax, 93
Convoluta roscoffensis, 120
Copepoda, 76, 77, 79, 89, 195, 199
Copilia, 89, 93
quadrata, 93
vitrae, 93

Corycaeus, 76, 89, 93
agilis, 93
carinatus, 93
catus, 93
crassiusculus, 93
elongatus, 93
lautus, 93
limbatus, 93
speciosus, 93
typicus, 93
Corynocephalus, 85
albo-maculatus, 85
Corynopoma riisei, 251
Coscinodiscus, 77
Cothurnia imberbis, 83
Crab zoea, 79
Crabs, 196, 197
Brachyuran, 213
Crappie, black, 35
white, 34

Crassonemertes, 100

robTsta" 98, 99, 104, 105, 112
Crenicichla johanna, 252
saxatilis, 252
Creolefish, 197
Creseis, 196, 199
Criseis acicula, 147
Cristivomer, 175

York Zoological Society [XXI

Crustacea, 79
Cutlassfish, 201
Cuvierina, 196
Cyclothone microdon, 77
pallida, 77
signata, 77
sp., 202

Cymopolia fragilis, 218
Cymopolia zacae, 217
Cynodon scomberoides, 252
Cynoscion regalis, 149
Cyttarocyclis magna, 83, 84
plagiostoma, 83, 84

D

Dace, horned, 252
Decapterus macarellus, 203, 204
Decapod larvae, 79
Diaphus effulgens, 201, 202, 205
rafinesquei, 201, 202, 205
Diatoms, 77, 198, 200
Dictyocha fibula, 81
Dictyocysta dilatata, 83, 84
lata, 83, 84

Dinoflagellates, 82, 129 (PI. I-IX ; Text-figs.
1-5)

Dinonemertes, 99
Diphyes dispar, 233
Diphys truncata, 77
Distomum complanatum, 253
Dodecaceria, 64
Dogfish, 211
Dorvillea erythrops, 58
melanops, 58

Drieschia atlantica, 52, 67
Dussumieriidae, opp. 260, 276

E

Eel, deep-sea, 201

electric, 125, (Text-fig. 1), 127
Elassoma, 38
evergladei, 36
zona turn, 35
Elassomidae, 38, 42

Electrophorus electricus, 125, (Text-fig. 1), 127
Ellobiopsis, 147
Ellritze, 219

Elopidae, 257, 260, opp. 260, 288
Elops, 258, 264, 265, 266, 267, 268, 271, 273,
288, 289

saurus, 260, opp. 260, (Text-figs. 14, 1 5 )
Endodinium, 147
Engraulidae, 257, opp. 260, 280
Enneacanthini, 38, 42
Enneacanthus, 1, 19, 38
gloriosus, 2, 31, 39
obesus, 32

Epibdella melleni, 145
Epiplocylis sargassensis, 83, 84
Erythropsis, 153
extrudens, 152
Esox, 36, 37, 253
Etheostomidae, 37
Euaugaptilus elongatus, 92
Eucalanus attenuatus, 90
crassus, 90
elongatus, 90
mucronatus, 90
pileatus, 90
subtenuis, 90
Euchaeta acuta, 89, 91
marina, 91

Euchaetomera tenuis, '96 " " ' . “ : •

Euchirella brevis, 90 - - • .

Eudoxoides mitra, 234
spiralis, 234
Eugleng, 151
Eunoe purpurea, 51, 67
Euphausia, 76
americana, 95
brevis, 95
gibboides, 95
hemigibba, 95
mutica, 95
tenera, 95

Euphausiids, 199, 204
Eupholoe nuda, 53, 67
Eupholoe philippinensis, 54

1936]

Index to Paris 1-4

293

Eupomotis, 1, 5, 6, 7, 8, 9, 10, 19, 27, 36, 38
gibbosus, 2, 3, 4, 7, 8, 9, 11, 14, 15, 17, 18,
19, 21, 28, 35, 39, 48, 252, 253
microlophus, 31
Eurythoe pacifica, 50
Exonautes rubescens, 199

F

Farranula carinata, 93
Filefish, leathery, 203

Fishes, tropical, 219 (PI. I-III), 251 (PI. I
and II)

Fistularia serrata, 203
Flounder, 199, 200
Flyingfish, 203. 204
Foraminifera, 77, 198
Forskalia, 235
Fritillaria pettucida, 147
Fry, green, 197
Funchalia villosa, 197
Fundulus, 36

heteroclitus, 137, 150

G

Gaetanus armiger, 90
caudani, 89, 90
latifrons, 89, 90
miles, 90

Gaidius brevispinus, 90
tenuispinus, 90
Gammarid, 202
Gastropods, 196, 199
Gempylid, 201, 202

Germo alalunga, 179, 181, 183, 186, 187
Glaucothoe, 199
Glycera, 55
tesselata, 86
Glyceridae, 49, 55
Goniodoma polyhedricum, 81
Gonostoma, 203, 204
Gonyaulax digitale, 81
polygramma, 156
Gopherus agassizii, 225
berlandieri, 227
polyphemus, 227
Gromia appendiculariae, 148
Grubea clavata, 86
Guppy, 251
Gurnards, 203, 204
Gymnodium, 147, 151, 153
fucorum, 156
parasiticum, 147
poucheti, 147
pulvisculus, 147

H

Haloptilus longicornis, 92
ornatus, 92
Haplomi, 252

Haplozoon, 141, 142, 147, 154, 158
armatum, 154

Harengula, 203, 257, 260, opp. 260, 282, 286,
287 (Text-figs. 47, 48)

Harmothoe, 51
benthophila, 86
Hatchetfish, silver, 201
Helioperca, 38

macrochira, 2, 9, 13, 21, 27, 39
Heliosoma antrosum, 253
campanulatum, 253
Hemirhamphus, 203
Hermodiee carunculata, 50
Herring, 210, 211

Heterandria formosa, 219, 222, 223
Heteropods, 198
Heterorhabdus grimaldi, 92
longicornis, 92
papillager, 92
spinifrons, 92
Heterotis, 36

Iiippopodius hippopus, 232

Holocentrus ascensionis, 150, 195, 198, 199, 200,
204, 205

meeki, 195, 198, 199
vexillarius, 198, 199, 200
Hydra viridis, 120
Hydras, 120, 121

Hyperids, 199, 202
Hypopomus artedi, 251
Hyporhamphus unifasciatus, 204

I

Isopoda, 79

Isospondyli, 196, 252, 257, (Text-figs. 1-53)

Jack, common. 149
hard-tailed, 149
J enkinsia, 203

lamprotaenia, opp. 260, 276, 289 (Text-figs.
40-44)

K

Killifish, 137, 150

Mexican, 219, 220, 221
Kingfish, northern, 149
Kuhliidae, 37

L

Labridae, 37
Labyrinthidae, 37
Lactophrys, 203
Lampadena chavesi, 77
Lampanyctus ivarmingi, 77
Lancetfish, 201
Lanternfish, 201
Lebistes reticulatus, 251
Leiostomus xanthurus, 149
Lensia conoidea, 235
multicristata, 235
sp., 235

Lensia profunda, 235

Leodice culebra, 57
denticulata, 57
fucata, 50
longicirrata, 57
mutilata, 56
Leodicidae, 49, 50, 56
Lepominae, 38
Lepomini, 38, 42
Lepomis, 5, 10, 19, 36, 38

auritus, 2, 3, 5, 7, 8, 10, 11, 15, 19, 20, 21,
26, 27, 39, 252
cyanellus, 252
pallidus, 252

Leptochela sp., 197, 199, 200
Leptolepis dubius, 288
Lestidium intermedium, 77
Lethogrammus, 38
symmetricus, 28
Limacina, 196, 199
Loach, 252

Longithorax sp., 95, 96
Lopadorhynchus, 85
nans , 60
nationalis, 85
uncinatus, 60, 85
Lubbockia aculeata, 93
squillimana, 89, 93
Lucicutia clausi, 91
flavicornis, 91
longicornis, 91
magna, 91
maxima, 91

M

Mackerel, 210, 211
Macrosetella gracilis, 89, 92
Makaira nigricans ampla, 203
Man-of-war, Portuguese, 237 -

Marlin, blue, 203
Marphysa acicularum, 57
regalis, 57

Mecynocera clausi, 89, 90
Megacalanus longicornis , 89
princeps, 89
sp., 89

Megalopidae, 257, opp. 260, 263, 288
Megalops, 197, 199, 203, 204, 268
Meganyctiphanes norvegica, 211
Melosira moniliformis, 77
Menhaden, 211
Menticirrhus saxatilis, 149
Merluccius bilinearis, 211
Merodinium, 147

294

Zoologica: New York Zoological Society

[XXI

Mesogaidius intermedins, 90
Mesogonistius, 1, 31, 38
chaetodon, 32
Metacercaria, 251
Metridia brevicauda, 91
long a, 91
lucens, 91
normani, 91
princeps, 91
venusta, 91

Micropterinae, 38, 41, 42
Micropterus, 1, 20, 25, 30, 35, 38

dolomieu, 2, 3, 11, 17, 20, 22, 39, 252
pseudaplites, 24
salmoides, 252
Microsetella norvegica, 92
Microtoeniella, 147
Minnow, fathead, 252
Miracia efferata, 89, 92
Mithrax ( Mithrax ) acuticornis, 214
spinipes, 214
mexicanus, 213
Mollienisia velifera, 251
Monocanthus, 203
hispidus, 204
tucker i, 197
Monocirrhus, 20
Morone, 36
americana, 37
Mugil cephalus, 150
Myctophum benoiti, 77
hygomi, 201
laternatum, 77

N

Nainereis setosa, 55
Nandidae, 37
Nandus nandus, 252
Nannostomus trifasciatus, 251
Natonemertes, 99
Naucrates ductor, 149
Nectalia loligo, 236
Nectochaeta , 85
caroli, 86
grimaldii, 86
Nectonemertes, 100

mirabilis, 98, 99, 102, 109, 111, 112
Nectopyramis sp. nov. ? 232
Nematonereis unicornis (?), 86
Nematoscelis microps, 95
tenella, 95

Nemerteans, bathypelagic, 97 (PI. I-X; Text-
fig. 1)

Neocalanus gracilis, 90
robustior, 90
Necthunnus, 187

argentivittatus, 183, 184, (PI. III-VII), 203
macropterus, 183, 187, 189, 190
Nereidae, 49, 55
Nereis bairdii, 55, 67
glandulata, 56, 67
mirabilis, 56
riisei, 86

Nicidion kinbergii, 57

Noctiluca, 140. 141, 142, 144, 145, 153, 154, 156,
158

miliaris, 156
scintillans, 81
Non-isospondyls, 196
Notemigonus, 30, 36

o

Oikopleura dioica, 147, 150
tortugensis, 147, 148
sp., 150
Oithona, 76
attenuata, 92
similis, 92
spinirostris, 89, 92
Omosudis lowii, 77, 202
Oncaea, 89
conifera, 92
curta, 92
media, 92
mediterranea, 92
minuta, 93
tenella, 93
venusta, 93

Oodinium, 147, 149, 150, 152
amylaceum, 147, 148, 152
appendiculariae, 147, 148, 149
fritillaria, 147, 148, 149, 152, 153
ocellatum, 129 (PI. I-IX; Text-figs. 1-5)
parasiticum, 147

poucheti, 147, 148, 149, 150, 152, 155
sp., 149, 152
Opheliidae, 49, 60
Ophiocephalus striatus, 252
Opisthonema, 260

oglinum, opp. 260, 284 (Text-figs. 49-51)
Ostariophysii, 252
Osteoglossidae, 37
Ostracoda, 77, 79, 199, 202
Ostrea edulis, 120
virginica, 120
Oxycephalus, 197, 199, 200
Oxyporhamphus, 203
micropterus, 204

Oxyrrhis, 141, 143, 153, 154, 155, 156
amylaceum, 140
fritillaria, 140
marina, 140, 144
poucheti, 140
Oyster, 120

P

Pachynemertes, 100, 104
obesa, 98, 105, 113
Palaemonella sp. nov?, 197, 199
“Palolo,” 50
Pandalus danae, 77
Paracalanus parvus, 89, 90
Paradinium, 147
Paradinonemertes, 100, 105
drygalskii, 106

Paradinonemertes wheeleri, 98, 99, 105, 111, 112
Paraeuchaeta bisinuata, 91
hanseni, 91
Parafavella, 77
acuta, 77

Paralepis sp., 201, 202
Paramarphysa obtusa, 57
Paramecium, 118
Paranthias furcifer, 197
Parapodinium, 147, 150, 151
Parasites, 129
Parathunnus, 177
ambiguus, 179

atlanticus, 177 (PI. I-II), 195 (PI. I-III,
Tables I-II)
mebachi, 183
obesus, 181, 187
rosengarteni, 179
Parrotfish, 199
Parundella major, 82, 83, 84
Paulsenella, 147, 150, 151, 152
Perea, 36, 37

flavescens, 166, 252
fluviatilis, 252
Perch, pirate, 252
yellow, 252
Percoidei, 37, 41
Peridiniopsis asymmetrica, 81
Peridinium cerasus, 81
claudicans, 81
conicum, 81
grani, 81
oblongum, 81
Perinereis bairdii, 86
sp., 86

Phaenna spinifera, 91
Phalacrorax vigua, 253
Phallonemertes, 100

murrayi, 98, 99, 107, 111, 112
Phoxinus laevis, 219
Phronima, 197
Phyllodoce oculata, 60
Phyllodocidae, 49, 60
Physalia physalis, 237
Piabucana, 251
Pikelet, ocean, 201
Pilchard 197
Pilot fish, 149
Pilumnus limosus, 216
pelagius, 215

Pimephales promelas, 252

1936]

Index to Parts 1-4

295

Plankton, 75-113, 231-240
Plank tonemertes, 99
Planonemertes, 99, 100
Planonemertes labiata, 98, 99, 106, 113
Platypoecilus maculatus, 219, 220, 221, 223
Pleuromamma abdominalis, 91
gracilis, 91
quadrungulata, 91
robusta, 91
xiphias, 91
Plotonemertes, 100

adhaerens, 98, 99, 102, 111, 112
Plotonemertes aurantica, 98, 99, 103, 111, 112
Poecilia vivipara, 252
Polychaeta, 77, 85
Polylcrikos schwartzi, 156
Polynemus virginicus, 203
Polynoe granulata, 51
Polynoidae, 49, 51
Polyophthalmus, 50, 58
pictus, 62, 86

Polyophthalmus incertus, 61, 67

Pomacanthus paru, 150
Pomacentridae, 37
Pomatomus saltatrix, 149
Pomfret, 197, 201
Pomolobus, 36

pseudoharengus, 165 (Text-figs. 1-6)
Pomotis vulgaris, 253
Pomoxis, 38
annularis, 34
sparoides, 2, 35, 39
Pompano, round, 149
Pontellina plumata, 89, 92
Porgy, northern, 149
Porpita umbella, 238
Pouchetidae, 151
? Pray a dubia, 232
Prionotus carolinus, 149
evolans, 150

Proplectella acuta, 83, 84
claparedei, 83, 84
Prorocentrum micans, 82
Protopelagonemertes, 99

hubrechti, 98, 99, 100, 111, 112
Protopelagonemertes beebei, 98, 99, 101, 111, 112
Protozoa, 81 (PI. I-II)

Protulides elegans, 65
Psettus argentus, 150
Pseudaetideus armatus, 91
Pseudeuchaeta norvegica, 91
Pseudocalanus minutus, 90
Pseudochirella obesa, 91
obtusa, 91
pustilifera, 91
Pseudogobio esocinus, 252
Pteropods, 151, 196, 198

Puffer, northern (common), 149, 151, 203, 204

R

Radiolaria, 77, 82, 200
Rana clamitans, 252
magna, 254
pipiens, 252

Rasbora daniconius, 219, 221, 223
later istriata, 219, 222, 223
trilineata, 219, 223
Rhamdia quelen, 252
Rhincalanus cornutus, 90
nasutus, 90
Rhizophysa, 237
Rhodeus amarus, 243
Rhomboplites aurorubens, 203
Roccus lineatus, 37, 149
? Rosacea cymbiformis, 232

s

Sabellidae, 49, 65
Sagitta, 76, 77, 202
Salpa democratica, 147
mucronata, 147

Salvelinus fontinalis, 252, 253
Sapphirina, 89
angusta, 89, 93
auronitens, 89, 93
metallina, 89, 93

Sardinella, 203, 260, opp. 260, 285, 286

anchovia, 197, 201, 204, 286 (Text-figs.

52-53)

Satanoperca papatera, 252
Schizopoda, 76, 77, 79, 95
Schizodinium, 147
Sclerotis, 38
punctatus, 26

Scolecithricella abyssalis, 91
Scolecithrix bradyi, 91
danae, 91

Scottocalanus securifrons, 91
Sea robin, Carolina, 149
striped, 150
Sebastes marinus, 211
Selene vomer, 203
Semathunnus, 187, 190
Semotilus astromaculatus, 252
Shrimps, 77, 79, 196, 199, 202, 204
Sigalionidae, 49, 53
Siluridae, 36

Siphonophora, 77, 79, 151, 196, 231
Spadefish, 150

Spheroides maculatus, 130, 149
spengleri, 204
Spicules, 77
Sponges, 77, 121

Squids, 196, 199, 200, 201, 202, 203, 204, 211
Squilla, 79, 196 198, 199, 200, 201, 202, 203,
204

Squirrelfish, 198, 199
black-barred, 198, 200
Staurocephalus, 58
Stauronereis, 58
Stenocionops beebei, 214
Stenocionops triangulata, 214
Stenosemella ventricosa, 83, 84
Stentomus chrysops, 149
Stepanomia amphitridis, 236
Sternopygus macrurus, 251
Stizostedion, 36
Stomatopods, 196
Stomias ferox, 77
Stylocheiron abbreviatum, 95
carinatum, 95
elongatum, 95
longicorne, 95
suhmi, 95

Sucker, common, 252

Sunfishes, North American, 1-48 (PI. I-VTJ,
Text-figs. 1-6)
black-banded, 32
blue-gill, 27, 252
blue-spotted, 31
common, 28, 252, 253
dwarf lepominid, 38
green, 26, 252
long-eared, 27
mud, 34

orange-spotted, 28
pigmy, 35

redbreasted, 26, 252
Surgeonfish, 200
Swellfish, northern, 149
Syllidae, 49, 50
Syndinium, 147

turbo, 141, 143, 154, 157
Synopia ultramarina, 79

T

Tarpon, 258, 260, 262, 263, 266, 271, 273, 288,
289

atlanticus, opp. 260, 263, (Text-figs. 16-20)
Terpsicore, 61
Thamnophis radix, 252

Thunnus thynnus, 178, 179, 181, 183, 207 (PI.
I, Text-fig.l)

Thynnus argentivittatus, 186, 187
Thysanoessa parva, 95
Thysanopoda, 197
aequalis, 95
obtusifrons, 95
tricuspidata, 95
Tintinnoinea, 77, 79
Tintinnopsis bermudensis, 82, 84
cylindrica, 77, 82, 84
major, 82

Tintinnus macilentus, 83, 84
Tomopteridae, 49, 58

296

Tomopteris, 85
apsteini, 86
nisseni, 59, 86
septentrionale, 77
sp? 59

Tomopteris longisetis, 58, 67

Tortoise, southwestern desert, 225
Trachinotus falcatus, 149
Travisiopsis, 85
lobifera, 85
sp., 63

Travisiopsis atlantica, 62, 67

Trichogaster fasciatus, 252
pectoralis, 252
trichopterus, 252
Triggerfish, 204

red-tailed, 197, 203, 204
Trout, brook, 252, 253
Trypanodinium, 147
Trypanosyllis, 50

Tunas, Bermudian and West Indian, 177-194

(PI. I-YII) , 195-205 (PI. I-III)

Allison’s, 190

black-finned, 177, 178, 195
giant, 207

yellow-finned, 178, 184, 195, 203
Tunny, common, 178
Turbots, 204
Tylosaurus marinus, 211
Typhloscolecidae, 49, 62
Typosyllis corallicola, 50

[XXI

Undeuchaeta major, 90
spectabilis, 90
Undinula darwini, 90
vulgaris, 90

V

Vahlkampfia callcinsi, 120
patuxent, 120
Vanadis, 85
formosa, 85
fusca-punctata, 59
longissima (?) 85
Velella velella, 238
V ermilia annulata, 65
Vermilia glandulata, 65, 68
Vogtia glabia, 232

w

Warmouth, 26
Weakfish, 149
Worms, 196
Wrasse, 199

X

Xanthichthys ringens, 197, 204
Xenotis, 38

megalotis, 21, 27, 39
Xiphophorus helleri, 219, 220, 221, 223
Xyrichthys, 199

Zoologica: New York Zoological Society

Umbra, 36

u

Zoeas, 79, 199

z

Jteto gocfe Zoological £i>octetp

General Office: 101 Park Avenue, New York City

©{filers

President , Madison Grant

Vice-Presidents, W. Redmond Cross and Kermit Roosevelt
Chairman, Executive Committee, Madison Grant
Treasurer, CORNELIUS R. Agnew
Secretary, Henry Fairfield Osborne, Jr.

Scientific Staff

Hoologtcal $ark

W. Reid Blair, Director
William T. Hornaday, Director Emeritus
Raymond L. Ditmars, Curator of Mammals and Reptiles
Lee S. Crandall, Curator of Birds
Charles V. Noback, Veterinarian

Claude W. Leister, Ass’t to the Director and Curator, Educational Activities

H. C. Raven, Prosector
Edward R. Osterndorff, Photographer
William Bridges, Editor and Curator of Publications

!Ugaarium

Charles H. Townsend, Director
C. M. Breder, Jr., Assistant Director

department of tropical ilcsenrcl)

William Beebe, Director and Honorary Curator of Birds
John Tee- Van, General Associate
Gloria Hollister, Research Associate
Jocelyn Crane, Technical Associate

Cbttorial Committee

Madison Grant, Chairman

Charles H. Townsend
George Bird Grinnell

William Bridges

W. Reid Blair
Wtilliam Beebe

SMITHSONIAN INSTITUTION UBRARIES

3 9088 01405 8945