m^

■■*• 4<«

-»v' • ■• .je^

t.:c^i>e.<«.."^v :^'

' i ■

"-■fr

.J' )

V-

%'

%1

/•
(.

L

1

1^^ h

■M-

. /-

.ji-4..
*>

■^

.■*-

^^

r

\

■ u

-4

■■*"

X--''<

1 "■■

i

>^**^; ^^

1- »■ ,

. *.. ' I-.

:ipT' I-

r-^,

't-^r

:-?"'■

*:Ti..-fo'

i

DISCOVERY REPORTS

VOLUME VI

Cambridge University Press
Fetter Lane, London

Neiv Tork

Bombay, Calcutta, Madras

Toronto

MacmiUan

Tokyo
Maruzen Company, Ltd

All rights reserved

DISCOVERY REPORTS

Issued by the Discovery Committee
Colonial Office, London
on behalf of the Government of the Dependencies
of the Falkland Islands

VOLUME VI

CAMBRIDGE

AT THE UNIVERSITY PRESS
1932

PRINTED IN GREAT BRITAIN BY WALTER LEWIS, M.A., AT THE UNIVERSITY PRESS, CAMBRIDGE

CONTENTS

PYCNOGONIDA (published 7th December, 1932)
By Isabella Gordon, D.Sc, Ph.D.

Introduction

Notes on Occurrence and Distribution
Variation and Specific Characters

Systematic Account

The Central Nervous System in Decolopoda

Abnormalities

Encrusting Organisms

List of Literature

Index

page^

3

7

8

128

130

132

134
136

REPORT ON PENGUIN EMBRYOS COLLECTED DURING THE DISCOVERY
INVESTIGATIONS (published i6th December, 1932)
By C. W. Parsons, B.A.

Introduction page 141

Stages of Development .
Comparative Studies in Development

The Brain

The Feathers ....

The Cartilaginous Skeleton

The Heart

Discussion

Summary

List of Literature ....
Plates I to VI

142

ON

149

151

154

157
160

161

163

following page 164

THE DISTRIBUTION AND MOVEMENTS OF WHALES ON THE SOUTH
GEORGIA AND SOUTH SHETLAND WHALING GROUNDS (published 15th
December, 1932)
By S. Kemp, Sc.D., F.R.S., and A. G. Bennett

Introduction page 167

Collection and Analysis of Data

Discussion of Results

Plates VII to XLH

. 168

• 177
following page 190

ON THE DEVELOPMENT OF C£P//yiL0Z)/5C (75 (published i6th December, 1932)
By C. C. John, M.A., D.Sc, D.I.C.

Introduction

General Considerations

Development

Development of the Bud

Discussion: Larval and Bud Development

List of Literature

Plates XLIH and XLIV ....

page 193
194
196
202
203
204
following page 204

45969

\

CONTENTS

REPORT ON SOUNDINGS TAKEN DURING THE DISCOVERY INVESTIGA-
TIONS, 1926-1932 (published 30th December, 1932)

By H. F. P. Herdman, M.Sc.

Introduction page 207

Sounding Machines

The Scotia Arc

Detailed Analysis of the various Sectors of the Scotia Arc

Soundings in other localities

Plates XLV to XLVII .

Charts i to 7

. 208
214
219
229

following page 236
in separate chart case

SPONGES (published 30th December, 1932)
By Maurice Burton, M.Sc.

Introduction page 239

Systematic List of Species 239

List of Stations, with the names of species collected at each 243

Descriptions of Species 254

The Affinities of the Antarctic species of Tedania, with notes on the distribution of the

genus generally 342

The Evolution of the species of Iopiion, and its significance in relation to other Antarctic

species of Ectyoninae 348

Geographical Distribution 351

Embryonic Development and Breeding Seasons 359

Unattached Sponges

The Value of External Form in the Identification of Sponges .

The Abundance of Sponges in the Antarctic and its Geological Significance

List of Literature

Index

Plates XLVIII to LVII

• 371

• 375
378

■ ■ • 380

■ • ■ 385
following page 392

A LIST OF WORMS PARASITIC IN CETACEA (published 30th December, 1932)
By H. A. Baylis, M.A., D.Sc.

Introduction

Classified List of Parasites, with synonyms, references to literature and hosts

page 395

Trematodes

396

Cestodes

399

Nematodes

401

Acanthocephala

405

OST-LIST

407

BLIOGRAPHY

414

[Discovery Reports. Vol. VI, pp. 1-138, text-figs. 1-75, December, 1932.]

PYCNOGONIDA

By
ISABELLA GORDON, D.Sc, Ph.D.

Assistant-Keeper in the Department of Zoology,
British Museum (Natural History)

CONTENTS

Introduction page 3

Notes on Occurrence and Distribution 3

Variation and Specific Characters 7

Systematic Account 8

Decolopodidae 8

Colossendeidae 11

Nymphonidae 24

Phoxichilidae (Pallenidae) 82

Phoxichilidiidae 87

Endeidae 93

Ammotheidae 94

Pycnogonidae 125

The Central Nervous System in Decolopoda 128

Abnormalities 130

Encrusting Organisms 132

List of Literature 134-

Index 136

PYCNOGONIDA

By Isabella Gordon, D.Sc, Ph.D.

British Museum (Natural History)

(Text-figs. 1-75)

INTRODUCTION

THE Pycnogonida collected by the Royal Research Ships ' Discovery', ' Discovery II '
and ' William Scoresby ' during the years 1925-31 far exceed in number and variety
those obtained by any previous Antarctic Expedition.^ Sixty-five species are represented,
all, with two exceptions, from the Western Antarctic and sub-Antarctic area. Of these,
fifteen species are described as new, a proportion of one in four approximately. One
genus is new and two {Pallene and Nymphopsis) are new to the area in question. Many
of the specimens of Decolopoda, Colossendeis and Ammothea are of unusually large size.
One feature of the collection is the abundance of minute forms, testifying to the
excellence of the method of using fine nets in conjunction with trawl or dredge. None
of the new species is of outstanding morphological or systematic interest, and the study
does not add much of phylogenetic importance. It may well be that, while future
expeditions will add to the number of species and to our knowledge of their geographical
distribution, the present conception of the Antarctic pycnogonidan fauna (from shallow
water) will not be materially altered.

My best thanks are due to Prof. Ch. Gravier, of Paris, Dr A. Burr, of Strasbourg,
and Dr A. Panning, of Hamburg, who did all in their power to further my studies
during brief visits to the respective museums. A number of the commoner forms in
the earHer collection (1925-7) were identified by Dr W. T. Caiman, F.R.S., before
I began this study, but I am entirely responsible for the final determinations as well as
for all the conclusions arrived at in this report. I am under particular obligations to
Dr Caiman for his helpful criticism and encouragement.

A number of the drawings were prepared by Miss Joyce Townend ; the others are
outline sketches drawn with the camera lucida.

NOTES ON OCCURRENCE AND DISTRIBUTION

A list of the Antarctic and sub-Antarctic species represented is given, together with
other data, in the following table (pp. 4, 5).

As regards the numbers of individual specimens, out of a total exceeding 1800, at
least 1200 belong to the single genus Nymphon. Of these, three-fourths belong to the
two most common species A'^. aiistrale (500 +) and N. charcoti (300 +) ; two other forms

1 See Caiman, 1915, p. i.

^ «

f a v

2 « <«
."■ o ,^ 3 3 «J

a

flj "O

> c

O "m

;=cB

Pi oi oi

Oi Oi Di Pi

d d

!ii Pi

Pi

0^

X X

XXX

X : X ■ : X X X X

j3

X X ; ; X ; x

CO

u

j3 =^ y "o

3 C c «J si

O ^ M

« (A

u'

:x xx^^xxx x;;x

: X X '■ : X X X : X X

CO

-S 'So

c«,

O

XX x::x::x: :::xx;xxxx;ix;;;xii;x X

CO U

(L> {3

60 m

.in r^ to

5 u

■ X X ; X

X X X ^

It! c -c "a

C « > _U

O

3 (U

i-i (-1 i-H H-t ►-t N •-'

O 00 MD

O

o

O O u-1 o^ o o
O O O r^

O

T3

S c

to

V

o

V

CO S

s °

■s &^

u 2 o

CO j-i *^
"S 3

C
o

CA

bO
•T3
O

u

J3
6X)

1 1

•ii pa

? 3

c o
5 pa

Q

o
P-,
o
-I

M o CJ

o

^' CO

:-^

1^

c
o

. CO
to '-

- ,2 J
C .^ PH "n

-o 's "3 ■~

c 9 5

5:1

> to

3 -O
O O

CO

e

c

^

f=

ft

n

J

■f; -o; Z K ■> S S
-^ -T' -^^ r" ;t <ij —

. u
"2 3

c o

-pa
to

-Si -S

I

■a, V.

. w

«. 2

Cu 55

CO ^

■*-> ^

c

hJ

!" ^

.(=^ a.

C G

5 M

•f^ 'ji

r* O '^

1 S, So

n.sp.
?ttOT, Hoc

"^ ^ >S

CO

n.sp
on k

HILI

arita

n

^

-« U 60

4J - ■^'" £;

■S^-S

^

^x ^

aden
icula

may
vicai
rticu

S PL, ^

dS Si t

^ i^ -^

t^

^

S >>-2i

to a 0 <3

s ~ci -^j

-0

*^

^■|c5

[IH

o' o o

do

Pi Di

(^ Pi ^ _ a; oi oi Pi . p5

■ ■ o" d ^ ■ o"

"3

>

c

X X

X X

X X

■ X

X ■ X

X X

X X

XXX

• X

• X

: X

XXX

X • X

X : X

X X X X X : X X

• I
X

CO >-■ M ro

i/-jO roco u-iiyD o

^^

fe

s'-^

60

o

►J 1<

K

C P

-sT

1 1

mily Phoxichilidae
Austropallene corniger
brachyura (Bouvier

ouvier)

HILIDIID

j'/ora (He
(Hoek)

§1

o *^

<

lis (Hod

HEIDAE

r/a/a (M

n.sp.
odgson)
dgson)

cristata (B
mily Phoxic
Palknopsis p

patagonica

(J
u
Q
Z

w

1

Endeis austro
mily Ammoi
Anunothea st

1

.1

6«

longispina,
spinosa (H
minor (Ho

fe

(^

^i^

(^

Oh

^iJ

Ph «

2 :/3

til -*

Ph O

2 .S ,

C
O

CO

60

-a

o c ~ '
c - S §

c -a o ■^
U « - -c

^ <:^ « -Ci,

a
o

w

^ 60

5 !" "O

Ml) ^i^

si

•S .^rj

6
o

1-1

S .C Q

y 2

o

C/3 ^

^ >- 2 p^ p 5; jj

"^ -^'"^

;-o S^S

■a,-

a, O

^ o

.o, o

s; 2

"*= P-i

3 ■o, 5

•5; O

o

60
C
<

o

o

u

a.

<u

x:

M

t:

J3

o

c

C

o

rt

C

>-,

o

«

en

CI

ts

60
C

>

ifi

JS

C

a

o;

P

e

3

o

o

O

t*-.

G rH

60

c

i

60 Ji

>
O

« -o

«

OJ u

S OJ

g

^£

'o

u

+-- Lh

-G

-C ti

U c

*2 3

'G

T3kC

t S

1)

S '^

«5 "^

CO

1) c

o ji:

-3

V 3

" c;

Cl. -S

^ 3

O

1) £i
^ O

>,

^^

■«

60
3 C

G
o

O.-'Z

O c^

«

C o"

« 'u

o

(u'S

>;H

1-

a ==

^ '::r

1^

X -a

T3

>-•

2 'C

u

n

S f

ii ^

-o

< 60

T3 _3

3 J3
'C 60
OJ 3

C
en

l-H

oj <:.>

Ch O

% ^

m Q

^
t*

o ^

3

t/) "-*-.

t-l

o

_(L> O

IJ -

fl

^1

^ H

6

H e

T.t;

o

i> 'Z

C M

^t:

r3 4>

• 05

O tn

o f.

.2 3

^ J

^ 6

>^

^ £

fe j;

c P

^g

^eS

3H

* T3

■1 1 1

C. <.oo

a

4)

n

60

6 DISCOVERY REPORTS

are each represented by about lOO individuals, namely A^^. orcadense and A'^. brevi-
caudatiim. The remaining species of the genus are, as a rule, represented by single
specimens or by very small numbers. The family Ammotheidae is represented by over
300 specimens, but, with the exception of Achelia parvula, no species is represented
by more than twenty-five specimens. Twenty-four species were each obtained only at
a single station and all in small numbers. As against this Nymphon brevicaiidatum
occurred at twenty-three, and PaUenopsis patagonica at twenty-one stations.

The depths at which Pycnogonida were obtained rarely exceeded 500 and were
usually between 2 and 350 m.

With regard to geographical distribution, the majority of the species were collected
in the Glacial District of the Antarctic Zone^ (Palmer Archipelago to South Georgia,
the South Sandwich Islands and Bouvet Island) ; a considerable number occurred only
in the Magellan District of the sub-Antarctic Zone, and a few species were found in
both districts (see table, pp. 4, 5).

The ' Discovery ' has done very extensive collecting in the neighbourhood of South
Georgia, and the 'William Scoresby' concentrated on the neighbourhood of the
Falkland Islands (Magellan District). Although the former is not so very far south of
the latter group of islands, the pycnogonidan fauna apparently differs considerably.
Of the nineteen species from the Magellan District only six- are also represented from
South Georgia and one from the South Orkneys. One species, Colossendeis scoresbii,
may prove to be a northern form^ of C. frigida, which, though most commonly found
in the Glacial District, is also present in the Magellan District (pp. 4, 5).

Of the four species collected beyond the northern limits of the sub-Antarctic Zone,
two are very near to, if not identical with, sub-Antarctic forms {Tanystyliim pfejferi and
Pycnogomim mageUaniciim (?) from Gough Island and Tristan da Cunha).

Twenty-seven of the Antarctic and sub-Antarctic species have also been recorded
from the Ross Sea Area, ten from Kaiser Wilhelm Land and two from the Kerguelen
District (see table, pp. 4, 5). It is highly probable that the majority of these forms will
prove to be circumpolar in their distribution.

The decapodous species Pentapycnon charcoti, collected by the 'Pourquoi-Pas?' off
the South Shetlands, has not been rediscovered, although Dr S. Kemp was on the
look-out for it.

Breeding season. Ovigerous or larvigerous males of sixteen species were obtained.
As far as one can judge, in the absence of records covering all months of the year, the
breeding season (assuming that there is a definite season) differs in Antarctic and in
sub-Antarctic or Temperate waters. In the Antarctic Zone males were carrying ova or
larvae from October to April, usually from December to March (table, pp. 4, 5). In the
Magellan District all the ovigerous and larvigerous males were collected in July. In one

1 See map given by Regan (1914, fig- 2, p. 25).

2 Nymphon orcadense has also been recorded from South Georgia, in addition to those hated.

3 Similarly Nymphon biarticulatum may prove to be a more southern form of the common A'', brevi-
caudatum (p. 72).

VARIATION AND SPECIFIC CHARACTERS 7

instance only {Pallenopsis patagonica), ovigerous males were collected from both regions,
and here also there is a difference of three months (March and April in Antarctic waters,
July in the Magellan District). Antarctic specimens of Toiiystylum pfefferiwere carrying
ova from October to December, while an ovigerous male referred to the same species
was collected off Gough Island in June. Only a few records from South Temperate
and Tropical waters are available and these are for the period May to July.

VARIATION AND SPECIFIC CHARACTERS

Dr Caiman (1915, p. 6) divided the Pycnogonida into two categories distinguished
as follows. "Certain genera and families present large numbers of minutely separated
species, the distinguishing characters of which have at least the appearance of in-
constancy ; while other groups are composed of few species easily and sharply defined
by characters that are relatively invariable." The first category is by far the larger, and
I have found considerable variation also in two genera hitherto regarded as monotypic,
namely Pentanymphon and Amtrodeciis. On the other hand the five species of Ammothea
described as new each appear to be characterized by several distinctive, sharply defined
features, although these also may prove to be liable to considerable variation when
more material is available.

The larger genera belonging to the first category, e.g. Colossendeis, Polletiopsis and
especially Nymphon, present, as every student of the group is well aware, many systematic
difficulties and problems. Previous writers have sometimes taken considerable pains
to provide keys to facilitate the task of specific determination, but no one seems to have
undertaken a really comprehensive study of the specific characters. Since each
succeeding collection includes some new, or apparently new, forms, the multiplication
of species cannot profitably be continued until a thorough revision of each genus is first
undertaken. This would necessitate, in addition to a re-examination of all the available
types of known species, a study of the amount of variation within each species.
Frequently the available number of individual specimens is far too small to admit of this.
Some knowledge of the post-larval development would also prove most useful, but at
present very little is known even in the more common forms. There is some evidence
to show that, for example, the relative proportions of tarsus and propodus (often used
to distinguish between closely allied species) differ considerably in young and adult
specimens of certain species {Colossendeis scoresbii and Pentanymphon antarcticum).

The key to the determination of the " Longitarsal " species of the genus Colossendeis
(p. 11) is little more than an extended and slightly modified edition of that previously
given by Dr Caiman (191 5, pp. lo-ii). That drawn up for the determination of the
Antarctic species of the genus Ammothea (p. 95) is almost entirely new; as many of the
specific characters as possible have been included in the key, which thus incorporates
the results of what may be regarded as a preliminary revision of these species. Most of
the forms included in this report are now, it is hoped, adequately described and figured,
but many species not represented in the Discovery collections are still in need of revision.

8 DISCOVERY REPORTS

I found it impossible to be certain of the identification of many of the numerous
Nymphon species until I had re-examined the types of nearly every species previously
recorded from the Antarctic and sub-Antarctic Zones. This revision probably constitutes
the most important part of the paper. Species re-examined but not represented in the
present collection have been described and figured in a separate short paper.^ The
results of this study have been condensed into a synoptic key (p. 31) to the deter-
mination of the adult males, and a table (p. 28). Many of the species are known from
one or a very few specimens and several from the adult female only. Even when the
known species had been re-examined, the interpretation of the results was by no means
easy. The species appeared to fall into two fairly well-defined groups according to the
form of the oviger in the adult male. Each main group could again be subdivided, for
the most part, into smaller sections comprising 2-4 or more closely related forms,
although an occasional form was apparently quite isolated. When the whole genus can
be revised, the results of the present study may have to be considerably modified.
I am aware that another worker might possibly have arrived at a quite different inter-
pretation of the relationships of the species, but I hope the classification given here will
prove to be a step in the right direction.

SYSTEMATIC ACCOUNT
Family DECOLOPODIDAE
Genus Decolopoda, Eights
Decolopoda australis. Eights (Fig. 73 A).
Bouvier, 1913, p. 48, uhi bibl.

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock, 179-235 m. ; gy. M. Large otter trawl : i S (with Brachiopod
attached).

St. MS 68. 2. iii. 25. East Cumberland Bay, South Georgia, 17 miles S i E to 8i cables SE x E of
Sappho Point, 220-247 m. Large rectangular net: i c? (colour in life— bright orange on legs; body,
chelophores, palps and ovigers brown-red).

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point, to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl: 2 S6
(i with Brachiopod attached).

St. 142. 30. xii. 26. East Cumberland Bay, South Georgia, from 54° ii'3o"S, 36° 35' W to
54° 12' S, 36^ 29' 30" W, 88-273 m. Large otter trawl: 2 ??.

St. 148. 9. i. 27. Off Cape Saunders, South Georgia, from 54° 03' S, 36° 39' W to 54° 05' S,
36° 36' 30" W, 132-148 m.; gy. M. St. Large otter trawl: i S-

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76^° W
to 2-62 miles S 11° W of Merton Rock, 200-234 m. Large otter trawl: i ? (with small patches of
Polyzoa on limbs).

^ Gordon, 1932.

DECOLOPODIDAE 9

St. 154. 18. i. 27. Jason Harbour to Larsen Point, South Georgia, from 2-6 miles S 84° W to
51 cables S 26° E of Larsen Point, 60-160 m. Large otter trawl: 2 ?? (smooth legs— i with small
growth of encrusting Polyzoa on oviger).

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Shetlands, 62° 07' S,
58° 28' 30" W, 391 m.; M. St. Medium otter trawl: i $ (spiny legs).

Remarks. Bouvier (191 3, p. 49) states that in the female the trunk is narrower and
less discoidal than in the male. In the Discovery specimens the reverse is the case, the
ratio of width across the second lateral process to length of trunk being o-85-o-92 in the
male and 0-92-1 -oi in the female.

The accompanying table shows how this species differs from D. antarctica, Bouvier.

Table I
D. australis

W. across 2nd lateral processes _ (0-85-0-92 in <?
L. of trunk* [o'92-i'oi in ?

Palp 9-jointed

Shorter, thicker ist joint to chelophore:
, > L. of I St ioint
(«) L. of trunk -°-5°-°-65

{b)

L. of I St joint
W. of ist jointf

= 4-0-5-3

Tibiae of 4th leg somewhat shorter :
L. of tibia i

(«)
(*)

L. of femur
L. of tibia 2
L. of femur

•04

I -04-1 -13

Somewhat shorter proboscis, as a rule :

L. of proboscis , . . ,

-^-- f-T r~ = i-iS-i-37 (1-44 in I specimen)

L. of trunk

Palm relatively short ; fingers strongly arched

D. antarctica

= 0-96-1 -00 in cJ and ? (o-88 in one S)

8-io-jointed
Longer, more slender
{a) = 0-66-0-75

{b) = 6-0-8-0

Tibiae, especially the 2nd, somewhat longer:
{a) = I -05-1 -08

{b) = I -22-1 -33

Somewhat longer proboscis :

= I-38-I-54

Palm relatively long, fingers but slightly
arched

Ocular tubercle narrow, less than half width of cephalon | Ocular tubercle exceeding half width of

I cephalon

* To anterior end of abdomen. t Near base.

Distribution. The two species of the genus Decolopoda have been recorded only
from the western side of the Glacial District of the Antarctic Zone. It is interesting to
note that both forms appear to have nearly the same range, since both have been
recorded from South Georgia and the South Orkneys, D. australis from the South
Shetlands and the type of D. antarctica from Graham Land.

Decolopoda antarctica, Bouvier (Figs, i, 73 B, C and 74 i).
Caiman, 1920, p. 244, ubi bibl.
St. 39. 25. iii. 26. East Cumbedand Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock, 179-235 m.; gy. M. Large otter trawl: i ,J with all claws
much worn (Brachiopod attached, and small groups of an encrusting Polyzoon on ventral surface).

JO DISCOVERY REPORTS

St. 42. I. iv. 26. OfF mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of
Jason Light' to 4 miles N 39° E of Jason Light, 120-204 m. Large otter trawl: i ?, with some
encrusting Polyzoa and a Serpulid.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl: i <S
(whole ventral surface covered by a sponge with wide sockets left for movement of ovigers ; left chela
covered by an encrusting Polyzoon and a similar growth commencing on movable finger of right chela) .

St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 54° 02' S,
36°38'W to 54° 11' 30" S, 36° 29' W, 122-136 m.; gn. M. St. Large otter trawl: i? (various
encrusting Polyzoa on body and legs ; pyriform mass of Alcyonidiiim on two terminal segments of
right palp).

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76!° W
to 2-62 miles S 11° W of Merton Rock, 200-234 m. Large otter trawl: i <?, i ? (several encrusting
Polyzoa and a sponge on ventral surface of $).

St. 164. 18. ii. 27. E end of Normanna Strait, South Orkneys, near Cape Hansen, Coronation
Island, 24-36 m. Small beam trawl: i 3 (encrusting Polyzoa and several adherent Foraminifera).

Fig. I a-e. Decolopoda antarctica, Bouvier. To illustrate variation in the total
number of segments of the palp: x 8.

Remarks. Bouvier (1913, p. 50; 1906, pp. 27-9) has enumerated the differences
between D. australis and the holotype of D. antarctica, but now that more specimens of

COLOSSENDEIDAE n

the latter are available, it becomes evident that certain of these differences are no
longer valid.

In D. antorctica the number of segments in the palp is variable (8-10, see Table I,
Fig. i; Caiman, 1920, p. 245, figs. A and B). This extreme variabihty is most
unusual in Pycnogonida, where the palp has, as a rule, a constant number of segments
in each species. Since it is not yet certain that this is a quite normal phenomenon, it
has been dealt with under "Abnormalities" (p. 130).

According to Bouvier (1906, p. 28) the legs are considerably longer in D. antarctica,
being approximately twelve times the maximum width of the trunk instead of only
8-9 times as in D. australis. In the Discovery specimens the two species are ahke in
this respect, for while the ratio varies from 9-3 to i2-8, the average is 10-84 for
D. australis, 10-90 for D. antarctica.

In D. antarctica the two sexes do not differ from each other with regard to the relative
width of the trunk (Table I). The chief differences between the two species are
given in Table I ; the outstanding characteristics of D. antarctica are (i) the broad ocular
tubercle, (2) the slightly arched fingers, and (3) the long, slender first joint of the
chelophore.

Encrusting organisms, mostly Polyzoa and sponges, are found with far greater
frequency on specimens of D. antarctica, suggesting that it is probably of more sluggish
habits than D. australis (see p. 132). In the male from St. 123 the left chela is rendered
quite useless by a growth of Polyzoa, and in the specimen from St. 140 the ocular
tubercle is partially concealed.

Distribution. See p. 4.

Family COLOSSENDEIDAE, Hoek

Genus Colossendeis, Jarzynsky

Two species are described as new, although, when more material is available, one of
these may prove to be a sub-Antarctic form of a common Antarctic species. The other
species is remarkable for the lateral articulation between the seventh and eighth palpal

segments.

The key to the " Longitarsal " species of Colossendeis drawn up by Caiman (1915,
pp. lo-ii) is reprinted below with slight modifications, chiefly those required by the
incorporation of the two ntW'f species.

I. Sixth segment of palp more than three times as long as wide.
A Proboscis distinctly longer than trunk.

I. Trunk segmented C.articulata, hom'^n

II. Trunk non-segmented.

A. Lateral processes in contact.

1 . Seventh segment of palp longer than eighth; eyes absent. ..C. proboscidea (Sabine)

2. Seventh segment of palp shorter than eighth; eyes present C. scotti. Caiman

12

DISCOVERY REPORTS

B. Lateral processes separated.

1. Eighth segment of palp articulated laterally with seventh C. tortipalpis, n.sp.

2. Eighth segment of palp normally articulated with seventh.

\C. australis, Hodgson

a. Seventh segment of palp equal to eighth \ C. media, Hoek

[ C. brevipes, Hoek

b. Seventh segment of palp distinctly shorter than eighth,
i. Eyes absent.

a. Proboscis dilated distally C. orcadensis, Hodgson

{C. angusta, Sars
C. gracilis, Wotk

ii. Eyes present.

I C. megalonyx, Hoek^
«. Legs smooth [C.frigida, Hodgson

^. Legs spiny C. rugosa, Hodgson

B. Proboscis subequal to trunk [claw subequal to propodus] C. scoresbii, n.sp.

II. Sixth segment of palp not more than twice- as long as wide [proboscis, at most, hardly
longer than trunk].

A. Lateral processes in contact C. wihoni. Caiman

B. Lateral processes separated.

L Femur longer than second tibia.

A. Sixth segment of palp longer than seventh, eighth or ninth ...C. glacialis, Hodgson

(C. gracilipes, Bouvier)

B. Sixth segment of palp subequal to seventh, but shorter than either eighth or
j^jjjj^ji C. drakei. Caiman

n. Femur not longer than second tibia.

A. Lateral processes separated by their own diameter C. robusta, Hoek

B. Lateral processes separated by less than their own diameter C. lilUei, Caiman

Colossendeis scotti. Caiman.

Caiman, 1915, pp. 10 and 11, fig. i.

St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of Jason
Light to 4 miles N 39° E of Jason Light, 120-204 m; M. Large otter trawl: i c?, i ?.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point, to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl : i <?, i ?.

Distribution. Previously recorded from Ross Sea.

Colossendeis tortipalpis, n.sp. (Figs, zb-e, 36, ^ and 4 a).

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m.; R.
Large dredge: i ? (holotype), i S and i immature, soft specimen, with C. australis.

Diagnostic characters. Sixth palpal segment more than three times as long as
thick; eighth palpal segment articulated laterally (or termino-laterally) with seventh;
proboscis as a rule exceeding 1-5 times the length of the non-segmented trunk; lateral

1 C.frigida, C. megalonyx and C. rugosa may be co-specific; and C. scoresbii may be only a more northern
form of this species.

- The sixth palpal segment may be rather more than twice as long as wide in C. drakei.

COLOSSENDEIDAE

13

processes separated; eyes present; sub-chelate termination to tenth joint of oviger;
tarsus approximately equal to, claw shorter than propodus.

Description. Greatest width of trunk, across the first lateral processes, just under
two-thirds of its length ; lateral processes separated by approximately half their own
diameter. Ocular tubercle high, conical ; eyes, especially the anterior pair, well developed.

Proboscis decurved, over 1-5 times as long as trunk; narrow and cylindrical for the

Fig. 2. Palporfourterminalpalpalsegmentsof:a. Co/oMew</ma!«/ra/w, Hodgson: xii. b-e. C.tortipalpis,
n.sp. (6 and f, of male: xii and 20 ; </, of holotype : x 20; e, of immature specimen — slightly abnormal : X17).

first third of its length, expanding to nearly twice this width in the centre, then gradually
narrowing anteriorly; terminal just a little greater than basal width (Fig. 3 d).

Abdo?neri short, extending to within a short distance of the distal articulation of the
first coxa; narrow and cyHndrical proximally, expanding considerably at about two-
thirds of its length ; apex bifid.

Palp as represented in Fig. 2 b; length of second exceeding 1-5 times that of fourth
joint ; fifth and sixth joints approximately equal ; eighth joint laterally articulated with
seventh and somewhat shorter than the terminal one.

Oviger with fourth slightly longer than sixth segment; spines on the four terminal
segments in four rows; a large curved spine at the distal end of terminal segment
opposed to the claw (Fig. 4 a).

14

DISCOVERY REPORTS

Legs approximately six times as long as trunk ; several rows of minute spines— the
median dorsal row being most conspicuous— on femur and both tibias ; only an occasional
spine on tarsus and propodus (Fig. 3 b). The species, like C. australis, is characterized
by the presence of a semicircle of 4-8 spines placed ventrally at the distal extremity of

Fig. 3. Colossendeis australis, Hodgson: a. Terminal segments of third

leg: X 8. c. Proboscis, lateral view: x 4.

Colossendeis tortipalpis, n.sp. b. Terminal segments of third leg: x 10.

d. Proboscis, lateral view: x 3.

the second tibia and of the tarsus (see Fig. 3 a and b) ; a few smaller spines occur also
at the distal articulation of the first tibia.

The genital openings of the male are not very distinct and perhaps are not yet
perforate ; those of the female are very conspicuous.

Measurements (mm.)

Holotype ? 3

Length of proboscis 20-1 15-5

Greatest diameter of proboscis... 37 2-7

Length of trunk 11-4 9-0

Width across first lateral process 6-9 57

Length of abdomen 3-1 2-$

Third leg:

Coxa 6-s 5-3

Femur 20-0 13-6

First tibia 19-0 13-8

Second tibia lyo ii-2

Tarsus 67 4-4

Propodus ... ... ■■. 6-0 4'4

Claw SS 3-2

Immature
12-5

2-0

77
5-°

2-0

4-5
ii-o

II-2

9-4
3-8
3-8
3-1

COLOSSENDEIDAE

IS

Remarks. The subchelate termination of the oviger places this species near to
C. scotti, Caiman, and C. australis, Hodgson.^ It resembles the latter in having the
lateral processes separated, but differs from it, and from most previously described
species," in the lateral articulation of the eighth palpal joint. This method of articulation
is apparently normal, occurring in both palps of the holotype and of the male (Fig. 2 b,
c and d\ the seventh segment is longer in both the smaller specimens than in the
holotype). In the immature specimen the left palp is normal; but the right one is
abnormal (Fig. 2 e) in that the seventh segment appears to be absent, although there is
perhaps a hint of a suture near the distal extremity of the sixth joint.

C. tortipalpis differs further from C. australis in the following respects: (i) it is a
smoother, more slender form, (2) the proboscis is more tapering and graceful (cf. Fig. 3 c
and d), (3) the sub-terminal spine on last segment of oviger is more strongly developed
(cf. Fig. 4 a and b), and (4) the tarsus is much shorter, while the claw is longer (cf. Fig. 3 a
and b).

Fig. 4. Terminal segment of oviger with claw of: a. Colossendeis tortipalpis, n.sp.: x 36.

b. C. australis, Hodgson: xio.
Terminal segments of palp of: c. C. drakei. Caiman: x 20. d. C.glacialis, Hodgson: X15.

Colossendeis australis, Hodgson (Figs. 2 fl, 3 rt, c and 4 b).

Hodgson, 1907, p. 59, pi. ix, fig. i ; pi. x, figs, i and 2.
Bouvier, 1913, p. 63, text-figs. 20 and 21.
Caiman, 1915, pp. 10 and 14.

1 The position that the species occupies in the key, using different characters, points to the same relationship.
^ See Hoek, 1881, pi. ix, fig. 7, C. gracilis.

i6 DISCOVERY REPORTS

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m. ; R. Large
dredge: 7 specimens (2 soft), genital apertures very obscure.

St. 181. 12. iii. 27. SchoUaert Channel, Palmer Archipelago, 64° 20' S, 63° 01' W, 160-335 m-
Large otter trawl: i ? (very scabrous).

St. 371. 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large
otter trawl: i specimen (encrusted with Polyzoa).

St. 599. 7. i. 31. 67° 08' S, 69° 06J' W, 203 m. Large dredge: i specimen (very soft).
Distribution. Eastern and Western Antarctic Zone.

Colossendeis frigida, Hodgson (Figs. 5 «, b, 6 a, d and jc, d).

Hodgson, 1907, p. 63, pi. ix, fig. 3 ; pi. x, figs. 5 and 6.
Caiman, 1915, pp. 11 and 17.
Loman, 1923, p. 7.

St. 42. i.iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E to
4 miles N 39° E of Jason Light, 120-204 m.; M. Large otter trawl: i specimen, with Nymphon
charcoti.

St. 140. 23.xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 54" 02' S,
36° 38' W to 54° 11' 30" S, 36° 29' W, 122-136 m.; gn. M. St. Large otter trawl: i <^.

St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, from 54° 04' S, 36° 27' W to
53° 58' S, 36° 26' W, 155-178 m.; gn. M. S. Large otter trawl: i $.

St. 148. 9. i. 27. Off Cape Saunders, South Georgia, from 54° 03' S, 36^ 39' W to 54° 05' S,
36° 36' 30" W, 132-148 m.; gy. M. St. Large otter trawl: 2 ??, i c? (colour of the latter, "pale horn
colour throughout, with distal extremities of ovigerous legs and true legs tinged with reddish").

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76^° W
to 2-62 miles S 11° W of Merton Rock, 200-234 m.; M. Large otter trawl: i ?, i young.

St. 167. 20.ii. 27. Off Signy Island, South Orkneys, 60° 50' 30" S, 46° 15' W, 244-344 m.;
gn. M. Large otter trawl: i ?, probably young.

St. 180. II. iii. 27. 17 miles W of N point of Gand Island, SchoUaert Channel, Palmer Archi-
pelago, 160-330 m.; M. St. Large otter trawl: i specimen, probably ?.

St. 181. 12. iii. 27. SchoUaert Channel, Palmer Archipelago, 64° 20' S, 63° 01' W, 160-335 m.;
M. Large otter trawl: i ?, 2 SS-

St. 187. 18. iii. 27. Neumayer Channel, Palmer Archipelago. 64° 48' 30" S, 63° 31' 30" W,
259 m; M. Large dredge : 4 specimens — probably 2 ??, 2 S3-

St. 366. 6. iii. 30. 4 cables S of Cook Island, South Sandwich Islands, 77-152 m. Large otter
trawl: 15 specimens.

St. 371. 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large
otter trawl: 2 specimens.

St.WS245. 18. vii.28. 52°36' S, 63° 40' W, 304-290 m.;d.gn. S.Sh. Commercial otter trawl :
3 specimens.

St. WS 248. 20. vii. 28. 52° 40' S, 58° 30' W, 210-242 m. ; f. gn. S. Sh. Commercial otter trawl:
I specimen.

COLOSSENDEIDAE

17

Remarks. The specimens referred to C. frigida, Hodgson, are characterized as
follows : (i) they are of large size and slender build (the trunk is usually 9-12 mm. long
and the legs vary from nine to thirteen times the trunk length), (2) the second tibia is
usually at least four-fifths of the femur, the tarsus is considerably longer and the claw
shorter than the propodus (see Table II), (3) the proboscis in adult specimens is always
at least half as long again as the trunk. ^

The terminal claw of the oviger varies greatly in length (Fig. 5 a and b) and what
Hodgson (1907, p. 64) terms the "thin membranous fold" is not always present.- The
seventh palpal segment varies considerably in length, sometimes it is only as long,
sometimes more than twice as long, as wide.

Fig. 5. Terminal segment of oviger with claw of; a and b. Colossendeis frigida,
Hodgson: x 36. c. C.scoresbii, n.sp.: < 46.

The syntypes of C frigida, from the other side of the Antarctic, are slightly more
robust, with rather shorter legs (8-66-11 times the trunk length). The holotype of
C. megalonyx, Hoek, agrees closely with the syntypes of C. frigida (see Table II); some
of the other specimens, however, are intermediate between C. frigida and C. scoresbii
(p. 21 and Table II). As suggested by Loman (1923, p. 7) C. megalonyx, C. frigida
and C. rugosa may be co-specific, and C. scoresbii may prove to be just a smaller Northern
form of C. frigida (p. 21). In that case the species would have to be called by Hoek's
name — C. megalonyx.

'^ The proboscis was rather shorter than is typical in two small specimens of 6-8 and 7-6 mm. trunk
length respectively, being 1-45 and 1-27 times the trunk.

2 The claw may be much worn in adults where the special spines on segments 7-10 are reduced to
short stumps.

i8

DISCOVERY REPORTS

Table II

Syntypes of
C. rugosa

Discovery collection

Challenger collection
C. megalonyx

Syntypesof
C. frigida

Lengthfmm.)

Length (mm.)

8-51

9^8

C. fngida

C. scoresbii

IfO

8-0

7-3

7-0

L. of proboscis
L. of trunk

I •75-2-03

I-S9

2^I4

1-57 -2-39

o^94-ri9

1-82

1-69

f4o

1-30

L. of second tibia*

8-02-9-29

7-63

8-23

7-43t-9-i4

5 •90-7^00

7-8

8-0

6-58

6-90

L. of tarsus*

4-68-5-03

3-29

4-62

3-92 -^4-89

2-25-3-I8

4-51

4-66

3-2b

2-76

L. of propodus*

3 •62-4-00

2-75

3-27

3^20 -3-80

2-25-3^56

3-51

3-93

2-63

2^62

L. of claw*

i-78-2^27

2-33

2-31

1-72 -2-96

3-00-4-32

3-o8

3-b2

2-50

2-48

L. of tarsus

L. of propodus

f2o-i-37

1-20

1-41

i-i-jX-i-^6

0-94-1 -09

1-28

f20

1-24

i-i6

L. of claw

L. of propodus

o^49-o-58

0-85

0^71

0-54 -0^79

I-00-I-35

0-88

0-92

0-95

0-95

Distribution

Ant-

Antarctic—

Antarctic§

Sub-

Sub-Antarctic (Magellan

arctic —

Ross Sea

(S. Georgia

Antarctic

District)

Ross Sea

— to Palmer
Archi-
pelago)

(Magellan
District)

* Length of femur regarded as 10.
•j- Usually greater than 8.

J Usually greater than i-z.

§ A few sub- Antarctic (Magellan District).

Colossendeis scoresbii, n.sp. (Figs. 5 c, 6 6, c and "] a, b).
C. tnegalonyx, Hoek, 1881, p. 67, in part.

St.WS 91. 8. iv. 27. 52° 53' 45" S, 64° 37' 30" W, 191-205 m.; f. d. S. Sh. Commercial otter
trawl: 2 young specimens and i larva with chelophores.

St.WS 92. 8. iv. 27. 51° 58' 30" S, 65° 01' W, 143-145 m.; f. d. S. St. Commercial otter trawl:
I (J, I $, 6 young, of which 2 are soft, and 2 larvae with chelophores.

St.WS 93. 9. iv. 27. 7 miles S, 80^ W of Beaver Island, West Falkland Island, 130-133 m.;
gy. S. Commercial otter trawl: 2 $?.

St. WS 98. 18. iv. 27. 49° 54' 15" S, 60° 35' 30" W, 173-171 m. ; f. d. S. Commercial otter trawl:

1 $, holotype.

St. WS245. 18. vii. 28. 52° 36' S, 63° 40' W, 304-290 m.; d. gn. S. Sh. Commercial otter trawl :

2 specimens with C. frigida.

Diagnostic characters. Sixth palpal segment at least three times as long as wide,
seventh segment shorter than either eighth or ninth which are subequal; proboscis
subequal to trunk length; lateral processes separated; eyes present; terminal claw of
oviger short, less than one-third of tenth segment ; tarsus subequal to, claw as a rule
longer than, propodus.

Description of holotype. Trunk non-segmented ; greatest width across first lateral
processes two-thirds of its length ; second and third lateral processes separated by their
own diameter.

COLOSSENDEIDAE

19

Ocular tubercle high, distal half narrow and sharply conical ; anterior much larger
than posterior pair of eyes.

Proboscis scarcely longer than trunk; narrow and cylindrical in proximal third,
expanding in centre to 1-5 times, and again at apex to 1-25 times the proximal diameter.

[\

Fig. 6. Colossendeis frigida, Hodgson : a. Terminal segments of third leg of a

young specimen from St. 149. dznd d' . Terminal segments of palp.
C. scoresbii, n. sp. : b. Terminal segments of third leg of adult from St. WS 93.
b'. Same of chelophorous larva from St. WS 92. c and c' . Terminal segments
of palp. («, b and b' , x 11 ; c-d' , x 13.)

Abdomen reaching to distal articulation of first coxa.

Palp. Second more than half as long again as fourth segment ; terminal segments as
represented in Fig. 6 c and c' .

Oviger. Fourth and sixth segments subequal; seventh rather longer, tenth rather
shorter than either of the two intervening segments; terminal claw short; spines

3-2

20 DISCOVERY REPORTS

in principal row of seventh segment relatively wider than the corresponding ones in
C. frigida (Figs. 5 c and 7 a, b).

Legs six or seven times as long as trunk, slender and smooth ; proportions of the three
longest segments 5:4:3; tarsus subequal to, claw one-third as long again as propodus
(Fig. 6 b and Table II); genital openings conspicuous.

Larva. The chelophorous larva from WS 91 measures 7 mm. in total length (in-
cluding proboscis and abdomen). The legs possess a few longish slender spines along
the mid-dorsal surface of each of the three main segments. These spines are all quite long
in the immature specimens accompanying
the larva, but are greatly reduced in size
in the holotype. The tarsus is relatively
shorter^ than in adults but the terminal
claw is longer than the propodus.

The chelophore is rather shorter than
the propodus, with a two-jointed scape
and the palm longer than the claws. The
oviger is already well developed and the
palp has the full number of segments
although the terminal ones are relatively
shorter than in the adult. The proboscis
is equal in length to the trunk and of
almost uniform diameter throughout its
length. The lateral processes are less
widely separated than in the adult and
the abdomen reaches to the middle of
the second coxa.

Remarks. This species rather upsets
the balance of the key to the "Longi-
tarsal" species of Colossendeis given by
Caiman (1915, p. 10), because the pro-
boscis is subequal to the trunk as in the
second group, while the sixth palpal seg-
ment is relatively long and narrow as in
the first group.

The specimens referred to C. scoresbii
are readily distinguishable from all the
specimens referred to C. frigida, Hodgson. They are of smaller size, with shorter
walking legs ; the proboscis is much shorter ; the spines on the four terminal segments

1 In immature forms the tarsus is sometimes relatively short, the propodus relatively long; as growth
proceeds the former elongates more rapidly than the latter segment. See also Pentanymphon antarcticum,
p. 26.

Fig. 7. Spines from principal row of seventh segment
of oviger of: a and b. Colossendeis scoresbii, n.sp. c
and d. C. frigida, Hodgson. /. C. drakei, Caiman.
g. C. glacialis, Hodgson.

e. Spine from principal row of tenth segment of
oviger of C. drakei, Caiman.

COLOSSENDEIDAE 21

of the oviger are broader ; second tibia and tarsus are both shorter relative to the length
of the femur (see Table II) and the claw is much longer.

One of the Challenger specimens referred to C. megalonyx by Hoek undoubtedly
belongs to C. scoresbii. Of the remaining four specimens (Table II) two^ are very near
to, and are probably, as Loman (1923, p. 7) has suggested, identical with C. frigida.
The other two are intermediate between C. frigida and C. scoresbii (see Table II). No
intermediate forms were included in the Discovery collection, but should such forms
prove to be common in the Magellan District, it may yet be necessary to unite C.
megalonyx, C. frigida and C. scoresbii.'^ It is to be regretted that the present collection
does not contain any chelophorous larvae of C. frigida for comparison with those of
C. scoresbii. In the latter species the proboscis is subequal to the trunk in the larva
as well as in the adult ; it would be interesting to know whether in the larva of C. frigida,
the proboscis is short or long relative to the trunk, and whether the tarsus is longer
relative to the propodus than in C. scoresbii (Fig. 6 b').

Distribution. Sub-Antarctic, north and west of Falkland Islands (Magellan
District).

Colossendeis wilsoni. Caiman.

Caiman, 1915, pp. 11 and 18, fig. 2.

St. 170. 23.11. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m.; R.
Large dredge : i <J, with Scalpellum and ? pupa.

Remarks. This specimen agrees with the description of the holotype except in one
minor point, namely that the terminal segment of the palp is as long as the preceding
or seventh segment. The right palp is mutilated.

Distribution. Previously recorded from off Cape Adare.

Colossendeis glacialis, Hodgson (Figs. 4^, yg and Sa).

C. glacialis, Hodgson, 1907, p. 61, pi. ix, fig. 2; pi. x, figs. 3 and 4.
C. gracilipes, Bouvier, 191 1, p. 1137; 1913, pp. 58-63, figs. 12-19.

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock, 179-235 ni.;gy.M. Large otter trawl : i ?," pale buff with
legs banded with crimson on either side of each joint".

St. 42. I . iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of Jason
Light to 4 miles N 39° E of Jason Light, 1 20-204 m;M. Large otter trawl : i (J with small patch of
an encrusting Polyzoon.

St. 148. 9. i. 27. Off Cape Saunders, South Georgia, from 54° 03' S, 36° 39' W to 54° 05' S,
36° 36' 30" W, 132-148 m. ; gy. M. St. Large otter trawl : i ? with patches of an encrusting Polyzoon
on the body. The specimen is accompanied by the following note (Note loi) : " ' Harlequin' Pycnogon.
Horn colour throughout, with scarlet red markings as follows: A few faint streaks on either side of

1 The claw is longer, but otherwise the specimens agree with the syntypes of C. frigida.

^ Two possible explanations of the relationship of the forms from the American side of the Antarctic
and sub- Antarctic Zones suggest themselves. C. frigida and C. scoresbii may be northern and southern
forms of one species; or there may be two distinct species which may tend to inter-breed when the southern
{C. frigida) enters the locality of the more northern one (C. scoresbii).

22 DISCOVERY REPORTS

proboscis. A spot near distal end of proximal segment of palp. A lateral stripe on abdomen between
legs. Legs with dorsal patch on first 3 segments; two bands on 4th and 5th segments, one in basal
and one in distal third ; 6th and 7th segments except for small patch at distal end ; 8th segment with
proximal half red; claw horn-coloured".

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76^° W to
2-62 miles S i i°W of Merton Rock, 200-234 m.; M. Large otter trawl : 4 specimens (2 soft), probably
all $$. "Harlequin" Pycnogon. Small patches of an encrusting Polyzoon on the hard specimens.

St. 155. 18. i. 27. 4-1 miles S, 26J° E of Larsen Point, South Georgia, 260 m.; M. Large otter
trawl : i (J with small patches of encrusting Polyzoa on body and legs.

St. 156. 20.1.27. 53° 51' S, 36° 21' 30" W, 200-236 m.; R. Large dredge: i ?.

St. 158. 21.1.27. 53° 48' 30" S, 35° 57' W, 40i-4iim. ; R. Large dredge: i $ "Harlequin"
Pycnogon.

St. 160. 7. ii. 27. Near Shag Rocks, 53° 43' 40" S, 40° 57' W, 177 m.; gy. M. St. R. Large
dredge: i $ (genital apertures very prominent).

St. 170. 23.11. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m.; R.
Large dredge: 5 specimens, i with a small Lamellibranch, i with a very small Scalpellum, several
with encrusting Polyzoa.

St. 363. 26. ii. 30. 2-5 miles S 80° E of SE point of Zavodovski Island, South Sandwich Islands,
329-278 m.; Sc. Large dredge: 2 specimens.

St. 366. 6. iii. 30. 4 cables S of Cook Island, South Sandwich Islands, 77-152 m. Large otter
trawl: i specimen.

Remarks. The striking colour of this species in the Hving condition led the collectors
to give it the name of " Harlequin" Pycnogon (see St. 148). Traces of the red markings
are still apparent in a few of the specimens.

In small specimens, although the genital apertures are present, it is not easy to
distinguish the sexes. In adult females the genital opening is situated obliquely on a
round, blunt eminence that is very conspicuous in some specimens (e.g. from St. 160).

Colossendeis drakei, Caiman (Figs. 4 r, 7 f , / and 8 b).

Caiman, 1915, pp. 11 and 22, fig. 3.

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m.; R.
Large dredge : 2 specimens, probably $$, but genital apertures rather obscure.

St. WS245. i8.vii. 28. 52° 36' S, 63° 40' W, 304-290 m. ; d. gn. S. Sh. Commercial otter trawl :
I specimen.

Description. Trunk non-segmented, lateral processes separated by approximately
their own diameter. Ocular tubercle conical; eyes, especially anterior pair, well
developed.

Proboscis a trifle shorter than trunk, diameter at distal end approximately 1-5 times
that at base and rather greater than that in the centre.

Abdomen short, reaching to middle of first coxa, considerably expanded in centre.

Palp with second slightly longer than fourth segment ; fifth, sixth and seventh segments
subequal and each rather shorter than eighth or ninth.

Oviger with fourth joint equal to sixth; terminal claw half as long as tenth segment;
spines in principal row on each of the four distal segments relatively broader than in

COLOSSENDEIDAE

23

C. gladah's, and serrated distally like those of C. frigida, C. megalonyx and C. scoresbii,
n.sp. (Fig. 7 e and/).

Legs slender, without spines ; claw equal to or slightly shorter than propodus which
is only a little shorter than tarsus (see Fig. 8 b).

Surface of body and limbs destitute of spines or spinules but uniformly covered with
minute setae.

Fig. 8. Terminal segments of third leg of: a. Colosseiidets glacialis , Hodgson, paratype.
b. C. drakei, Caiman, holotype (both x 11).

Meastirements {mm.)

Length of proboscis

Greatest diameter of proboscis (at apex)

Length of trunk ...

Width across lateral processes

Length of abdomen

Third right leg;

Coxa

Femur ...

First tibia

Second tibia

Tarsus ...

Propodus

Claw

Remarks. These specimens are most nearly related to C glacialis, Hodgson, but
differ from that species in the following respects: (i) they are of more slender build
without the spinules on body and legs characteristic of C. glacialis, (2) the tarsus is
shorter, the claw longer and there are no spines on the legs (cf. Fig. 8 a and b), (3) the
terminal joints of the palp are narrower, the sixth segment is twice as long as wide, the
seventh nearly as long as and the two distal ones rather longer than the sixth, the setae
on the palp are much shorter (cf. Fig. 4 c and d) ; (4) the principal spines on the ovigers
are thin, broadened and serrate distally, recalling those of C. frigida and C. scoresbii
(cf. Fig. 7 a-/and^); (5) the abdomen is shorter and broader.

6-3

5-5

1-4

1-4

6-5

6-0

4-5

4-0

17

1-5

3-8

3-6

3-0

II-O

i-i

9-6

9-5

8-0

4-2

3-8

4-0

3-5

37

3-5

24 DISCOVERY REPORTS

The specimens agree well with the types of C. drakei; the ocular tubercle, however,
is considerably higher and more conical.

Distribution. Previously recorded from the Ross Sea area.

Family NYMPHONIDAE

Genus Pentanymphon, Hodgson

Pentanymphon antarcticum, Hodgson.

Hodgson, 1904, p. 459, pi. xiv; 1907, p. 36, pi. v; 1908, p. 177.
Bouvier, 1906, p. 30, text-figs. 3-6; 1913, p. 66, text-figs. 22-24.
Caiman, 1915, p. 27.
Loman, 1923, p. 9.

? St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61'" 25' 30" S, 53° 46' 00" W, 342 m.; R.
Large dredge : i S-

St. 181. 12. iii. 27. SchoUaert Channel, Palmer Archipelago, 64° 20' S, 63° 01' W, 160-335 m.;
M. Large otter trawl: i (J.

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Shetlands, 62° 07' oo" S,
58° 28' 30" W, 391 m. ; M. St. Medium otter trawl : 5 specimens.

St. 366. 6. iii. 30. 4 cables S of Cook Island, South Sandwich Islands, 77-152 m. Large otter
trawl : 4 specimens.

Remarks. Previous workers have commented on the variability of this species
(Bouvier, 1913; Caiman, 1915). The table of measurements opposite gives some in-
dication of the nature and extent of this variation. Two adult males, one ovigerous the
other larvigerous, were chosen at random from among the syntypes and measured. In
both the neck is relatively long, the cephaHc segment being 2-6-2-8 times as long as wide
anteriorly; the second coxa is long (1-63-1 -681 times the sum of the first and third) and
the tarsus is rather longer than the propodus. The two Terra Nova specimens measured
agree well with the syntypes ; the second coxa in the male is slightly longer and in both
specimens the tarsus is considerably longer than in the types.

Of the specimens in the Discovery collection, those from St. 366 agree well with the
syntypes (allowing for sexual difi'erences), but the neck is rather shorter. In the adult
female from St. 195 the neck is still shorter and, unusual in adults, the tarsus is shorter
than the propodus. The other specimens from the latter station are not quite mature,
but in each the tarsus is rather shorter than the propodus.

The male from St. 181 has a remarkably long neck so that the cephalic segment is
four times as long as wide anteriorly. Also the second coxa is unusually long and
slender ; the tarsus is nearly half as long again as the propodus (cf. the specimens in the
Terra Nova collection).

But most interesting of all is the specimen from St. 170. Although only 2-8 mm. long

1 Approximately 1-4 in the ?.

NYMPHONIDAE

25

it is apparently an adult male, for the genital pores are quite distinct. The neck is short,
and the tarsus is not quite as long as the propodus, so that in these respects it resembles
the specimens from St. 195. But the second coxa is very long, as in the specimen from
St. 181. Now, as a general rule, specimens of P. autarcticum do not appear to reach
sexual maturity until they are somewhere in the neighbourhood of 6 mm. in length. The
smallest syntypes examined measured 3 -4-3 -8 mm. in length and were obviously
immature. It is very probable that the specimen from St. 170 belongs to a much smaller
species than P. antarctictim. But until more specimens are forthcoming, and until more
is known of the changes that occur during the post-larval development of P. antarcticum,
this question must remain undecided.

Measurements {mm.)

I erra

Nova

Discovery collection

Syntypes

St.

333

St.

St. 366

St

19s

St.

Ovig.

S

Lar'g.

Ovig.

3

181

170

?

0

?

?

?

?

??

?c?

?

L. of proboscis

.. 2-2

1-0

3-2

2-6

2-0

2-8

3-65

2-0

2-S

2-0

1-9

0-8

I-o

Diameter of probosci

s 0-67

0-6

1-07

075

07

0-95

1-15

07

I-o

0-8

0-S

0-35

o-S

L. of trunk ...

.. 6-4

6-0

7-6

7-0

7-6

7-2

8-2

5-2

6-4

6-4

S-8

3-6

2-8

L. of cephalic segme

nt 2-8

2-6

3.6

3-0

3-6

3-0

3-8

2-3

2-5

2-4

2-4

1-4

1-2

W.acrossanteriorcep

ti-

alic lobes . . .

i-o

i-o

1-4

1-2

0-9

1-5

1-9

i-i

1-6

1-4

I-o

0-6

0-67

W. of neck . . .

.. 0-4

0-39

0-6

0-5

0-33

0-6

0-8

0-4

0-6

0-5

0-4

0-23

0-27

W. across second later

al

processes . . .

.. 3-6

3-2

4-0

3-8

3-S

3-8

4-5

3-1

3-S

3-2

2-8

1-8

1-47

L. of abdomen

.. 0-6

0-6

0-8

0-8

0-6

I-o

1-3

1-2

0-8

0-6

0-45

0-4

o-S

L. of cephalic segme
W. of cephalic lobes

"^ 2-8

2-6

2-57

2-5

4-0

2-0

2-0

2-1

1-56

171

2-4

2-33

1-79

W. of cephalic lobes*
W. of neck

2-5

2-56

2-33

2-4

273

2-S

2-38

27s

2-66

2-8

2-5

2-61

2-49

Third leg:

First coxa ...

i-o

0-93

1-2

I-o

I-o

1-2

1-4

i-o

I-o

0-9

0-8

0-45

0-5

Second coxa

.. 2-6

2-6

3-0

2-8

3-2

2-8

3-2

2-0

2-4

1-6

1-8

0-9

i-S

Third coxa

.. 0-6

0-6

0-93

0-6

0-6

0-7

I-o

0-6

0-8

0-6

0-6

0-3

0-2

Femur

.. 6-0

6-4

9-6

8-0

6-8

8-0

9-4

4-6

6-6

4-6

4-6

1-8

2-9

First tibia . . .

.. 7-2

7-2

10-2

8-8

8-2

9-2

lO-O

5-6

7-4

5-0

5-4

17

3-2

Second tibia

.. lO-O

10-4

15-2

14-0

117

13-6

15-0

8-6

10-4

7-4

7-8

2-85

4-8

Tarsus

.. 1-47

1-7

2-4

2-4

2-3

2-4

2-2

1-6

1-73

1-2

I-3S

0-5

0-93

Propodus . . .

•• 1-33

1-4

1-6

173

1-6

2-0

2-0

1-6

2-0

1-6

1-4

0-8

I-o

Claw

.. 0-6

0-6

0-7

07

07

0-8

0-9

0-62

0-8

0-6

0-5

0-3

0-25

L. of coxa 2

.. 1-68

1-63

1-41

175

2-0

1-47

1-33

1-25

1-33

1-07

1-29

1-2

2-14

L. of coxae i -f 3

L. of tarsus

L. of propodus

I-2I

I'll

i-S

1-39

1-45

1-2

i-i

I-o

0-87

0-7S

0-96

0-63

0-93

L. of leg
L. of trunk

■■ 472

5-21

S-8i

5-62

4-66

5-54

S-4

4-92

5-05

3-58

4-1

2-6

5-36

* See also Caiman, 1915, p. 28.

26 DISCOVERY REPORTS

An examination of the syntypes showed that in specimens measuring under 5 mm.
in trunk length the tarsus is either shorter than, or equal to, the propodus. Thus it
would appear that the tarsus is short in young specimens and, as growth proceeds, it
elongates more rapidly than does the propodus. Also, in immature specimens the femur
is much shorter than, and the third walking leg (claw excluded) is 2-5-4 times as long
as, the trunk. In adults the femur is subequal to or longer than, and the leg is 4-6-5-8
times as long as, the trunk (see measurements). In these respects the specimen from
St. 170 differs markedly from small specimens of P. antarcticum. This species of five-
legged Pycnogonida might prove of interest to the student of heterogonic growth in
Arthropoda.

Genus Nymphon, Fabricius, 1794

A revision of the large genus Nymphon, comprising some ninety species, although
a long and tedious business, is much needed. In addition to the fact that many species
are very imperfectly described, the characters that serve to distinguish one species from
another have never been comprehensively studied. It has not been possible to devote
time to a complete revision of the genus at present ; the study was therefore limited to
Antarctic forms.

While much work has been done on material from Antarctic and sub-Antarctic
waters during the last thirty years,i the relationships of the species are very imperfectly
known and the key to their determination given by Loman (1923, p. 14 — a modification
and extension of that of Bouvier, 1913, p. 72) I found to be far from satisfactory. In
order to identify the Discovery material with any degree of accuracy a revision- of all
the Antarctic species had to be undertaken.

In several instances the type specimen is immature (TV. frigidum, Hodgson ; A^. longi-
colliim, Hoek; A^. lanare, Hodgson and A'^. trideiitatiim, Hodgson), while not a few species
are known only from the female (see Table III). The species in which the adult male is
known fall into two main groups, each with several atypical or aberrant forms as listed
below (see also Table HI). The larger=^ (group I) includes most of the species referred
to the genus "Nymphon"^ by those authors who recognize two genera (e.g. Hodgson,
Bouvier, Loman); the smaller (group II) includes those species referred to the genus
" Choeto?ivmphon'' as well as several forms hitherto regarded as belonging to "Nym-
phon", viz. A^. capense, Hodgson, A^. cotnpoctum, Hoek, and A^. articidare, Hodgson.

The fact that the Antarctic species fall naturally into two main groups would seem to
be strong evidence in support of the validity of the genus Chaetonymphon, Sars. But
a careful study of all the characters, many of which are represented in tabular form

1 E.g. Hodgson, 1907, 1908; Bouvier, 1906, 1913; Caiman, 1915; Loman, 1923.

- The types of all the species included in Table III have been re-examined, with the exception of
N. subtile, Loman; the Discovery specimens, however, are from the same locality as, and agree with Loman's
description of, the type. (See also Gordon, 1932.)

^ Seven of the species known only from female or immature specimens (Table III B) probably belong
to group I ; A'^. villosum, Hodgson belongs to group II and is allied to .V. biwticulatum and A^. brevicaudatum.

* When the word Nymphon is enclosed in inverted commas in the text it means Nymphon scnsu strkto of
certain authors, i.e. group I and the species listed above now placed in group II.

NYMPHONIDAE

27

(Table III), shows that there is no distinctive character, or set of characters, that
holds good everywhere. Indeed, exceptions abound — these will be mentioned in the
discussion of each separate character — and as no practical advantage is to be gained by
retaining a genus that is not definite and distinctive, I have reunited Chaetonymphon
and Nymphon as Caiman (191 5, p. 28) and others have done.

Group I Group II

Typical

Atypical

Atypical

Typical

proceroides

charcoti

capense

orcadense

tenuipes

compactum

ariiculare

hiemale

loiigicoxa

australe

iteumayri

gracillimum

hamalum

brevicciudatwn

subtile

brachyrhynchiim

biartkiilalum

pfcfferi

darencei

moidosum

paucidens

proximtim

adareanum

bouvieri

The main characters on which I rely for the separation of these groups are found in
the male oviger (Table III). In the typical species of group I the fifth segment is long
and slender, of approximately uniform diameter throughout, and either straight (type I,
e.g. Fig. 10 a and c) or distinctly curved (type I a, Fig. 18 a). In A'^. charcoti the fifth
segment is relatively short and more robust, but slightly curved and narrowed proximally
(Fig. 10 b). The other four atypical species have the fifth segment long and slender,
but distinctly expanded at the distal end though not to the same extent as is typical
for group II (type I b, Fig. 23 b, cf. Fig. 27).

The majority of the species in group II have the fifth segment relatively short, usually
more massive than in type I, and clubbed distally (type II, Figs. 23 a, 26 a, c, and 27).
In the atypical species, the fifth segment is again short and rather robust, but certain parts
of segments 5 and 6 or 3, 4 and 6 are thin-walled, inflated and tend to collapse when
the animal is killed and fixed (type II a, Fig. 26 b, and Gordon, 1932, Figs. 10 b and 12 b).

None of the other characters is of such systematic importance as the male oviger,
closely allied species often differing markedly from each other with regard to one or
more of these; e.g. the chelophore of A^. paucidens, n.sp., is very difli'erent from that of
A^. pfefferi, which not only in its oviger but in most of its other characters closely
resembles it. The following survey of these characters follows the order in which they
are dealt with in the specific descriptions.

Trunk. In group I, the trunk is usually slender, non-setose and loosely built, i.e. the
lateral processes are separated by at least their own diameter. In A^. adareanum, however,
the lateral processes are separated by only half their own diameter.

In group II the body is usually compact and setose with the lateral processes in
contact or separated by narrow intervals — up to half their own diameter. Setae are
absent in three species — A^. capense, N. compactum and A'^. articulare ; in A^. australe the
lateral processes may be separated by their own diameter.

The neck is well defined and long, as a rule, in group I, very short in group II ; but
in several species of the former it is as short as in the latter group (e.g. A^. pfefferi and
A^. adareanum).

4-2

28

DISCOVERY REPORTS

Table III. Antarctic species of Nymphon.

J3
0

0)

Specific name

u

^

C

41 u

a u
ra be

">

o

Neck

#

c

s

3 0 ;

3'i

d S

^

c

0

■5. °

§ 0

(4-1

3

s

3

w t-

■a S

O

a

o

3

0 4i

i-i

0

^1

0 ">

z

A

proceroides

3-5

(E.) OT.C.l

10-13

21-24

tenuipes

2-5-3

(E.) d.L'

28-42

24-34

C

hietnale

6-S-7-2

(E.) d.l.'

27-40

36-46

o §■

gracillimum

4-8-6-8

(E.) rf./.'

22-33

33-34

i a I'

gracilipes

3-7-5-3

(E.) d.L'

14-22

54-59

2 _, M

subtile

I a

2-5

(E.) rf./.' +

14-22

26-34

^ D-'r

pfejferi

I a

2-2-2

s.

(E.) rf.c'

19-30

19-22

a O c'

paucidens

I a

2-5

m.

->s.

(E.) (i.c.i +

7-9

17-20

odareamtm

I a

I -6-1 -75

s.

(E.) d.c.'

6-8

5-9

"3 b t II charcoli

lo -i8-5

ni.

(E.) rf./.' -

65-90 and

40-51

% '^ ^ f longicoxa

■^ !!!! TTT ' hamatum

"o g brachyrhynchuni

-g 4 .2 L darencei

75-110

16

6-8

I.§-^m.

(E.) d.L' +

70 and 95

38-44

16

7-2-8-4

I.§-«-m.

(E.) ?ff.c.»

50-65

29-37

16

4-4-2

m.

^^s.

(E.) w.c.i

40-60

36-37

16

6-4-8-2

m.

-:-S.

(E.) rf./."

3°-45

35-4°

capense
conipactiini

II a

2-75

S.

(E.) </./.♦ +

60-65

48-50

II (7

6-7

s.

d.c' -

40-48

36-38

f fe^

- australe

II a

4-5-5-2

m.

^s.

d.c}

38-46

23-37

^ =

orcadense

4-8-6-4

s.

m.c}

35-46

30-36

^Sg

articidare

2-5

s.

d.c>

18-25

24-26

° &-a

neumayriW

4-4-4

s.

m.c."

6-9

22-27

•S 8 2

brevicaudatumW

3-5-5

s.

d.c} +

29 and 35

10-17

^o& v<

biarticulatum \\

S-5-S

s.

m.c.'

30-33 and

12-18

CO S

40-42

mendosumW

4-5

s.

m.c' +

20

11-12

proximumW

5

s.

d.c} +

9 and 14

15

'\

^ bouvieriW

3-6

s.

m.c}

12-16

IS-18

>. 0 '" "1 distensum (^

7-2

l-§

(E.) m.c>

38 and 55

30

I.

d.l} +

30-33 and

42

60-64

'o ? 5^ .£ signatum ^

d.l> +

35 and 62

37

i» o ■« oi midtidens

3-4

m

(E.) d.l.'

24-28

73

'o ^ g « tanare**

7-5

1. <-m

d.l.'' -

48 and 55

24

a '*" Si lungicolluiii**

4-2

1.

(E.) d.U

II and 12

25

" 1 S 1 frigidmn**

2-3

1.

dl."

15-16

21

cd%& B villosum\m

4-2

s.

d.c} +

iS and 22

16

* The neck is regarded as short when the base of the oviger occupies the whole or two-thirds of the space
between first lateral process and anterior cephalic lobe; of medium length when it occupies approximately-
half, and long when it only occupies one-third or less, of this space.

■j- Where the second is more than one-third longer than the first tibia the ratio tjti is given in brackets.

J See p. 42.

§ In these three species the base of the oviger is situated midway between first lateral process and cephalic lobe.

The base of the oviger is in contact with, or only a little in front of the first lateral
process as a general rule, in both groups ; in three exceptional cases it is situated much
farther forward, arising from the middle of the long neck — N. longicoxa, N. hamatum
and A^. procerum.

Abdomen. In group I the abdomen is elevated at an angle of 30-45° as a rule, and

NYMPHONIDAE

Table III (contd.)

29

Third leg

If)

+

Of
U
0^

■^J-

1

^X)

2
S

c

-a
CQ

C

^

1

1>

M

-f

V
CO

-(—

Vi

c>

S

J

i

Specific name

(0

0

2.S
1 S)

0
u

IS

•5

l-i

:3
-a

a

1

3

T3

a

1
•3

S

0

0

J

^.S

^^-t

g

"

2

2

0

0

&

0

i-i

c

f

0 0

0

8
0

t/i

M
C

0

0.

0

0

a
0

<

3
S
fa

1-3

9-12 T.

I-5-I-7

'1

I ±

\

5-3-5-7

proceroides -,

I

6-7

2

'-2

I +

I ±

—

8

teiniipes

I +

I2-l6t

I-8-I-95

«2(i-5)

I —

i-

+

8-5-9-5

hiemale

%
0

075-0-85

10 T4

2 —

t,(i-4)

I^li

i

+

gracillimum

13

13-16

1-5

'2(1-5)

1 +

\-

+

10-12

gracilipes

■I M

I

8-9

I -5-1 75

'2

I

i

+

6-5

subtile

.*"

I —

4 T.

I +

'2

\

i

+

4-S

pfejferi

Ji

I

4 T.

I +

'2

4

i

+

5-3

paucidens

C3

E

0-6 -07

2-3 T.

I +

^2

\

4±

+

3-5

adareamim

2 +

20

I +

'2

I ±

I ±

—

11-12

charcoti

II 1

0-65-07

?8

2 +

?,(i-6)

i

J

—

10

longicoxa

1

0-7-0-8

12-13 T.

1-4

'1

!+-§

4-

—

10

hamattim ,,, t! 1

2-S

?30

2

'2

i

+

5-4

brachyrhynchum

•§

1-3

10-12

2-175

«2(i-4)

5

i

+

5

clat encei

'£

?

I +

?

1-2

h

%

i

_

4-7

capense
conipacimn

0

I -2-1

20-27

I +

tl

2

-1 +

—

4-7

1"

I ±

16-18

I +

h

lA-ii

4 -

V.

3-5-4-S

australe

I ±

30

I +

h

ij-i

I -

+

3-8

orcadense

0-78

?

ti

I —

1-

+

3

articulare

U-i

0-75

?

h

I ±

*

+

2-5

neumayri \\

0

0-4-0-54

3-4 T.

h

%

4 +

+

2-3

brevicatidatum \\

<U

0-4-0-5

5-6 T.

h

li-i

i±

+

2-4

biarticulation ||

-V S

CO

0-35-0-44

4-5 T.

h

i +

i±

+

3-2

mendosum \\

0-42

5-7 T.

fl

j

J-

+

proximuniW
bouvieriW

0-26-0-28

3-4 T.

h

f

i-

+

2-2

I +

I +

h

i +

i-

8-4

proceruni § t^ -.

078

«2(i-46)

i

\-

+

distensum ^ ^;

0-96

'2

+

signatum f^ c

3 c ° 6

0-88

2

I +

'2(1-5)
'2

I +

i"-

+

6-6

multidens j
lanare**

0 0 "rt 4J

-ego.

0-83

2-86

k

I

■4-

—

longicollum**

1 '^ s

I

I

U

2

i-

+

frigidum ** ''

phon
from
matu

0-66

h

t

i +

+

3

villosum II ^ B

II Species with the last two body segments fused.

^ Species known only from adult female.

** Species known only from immature specimens.

Explanation of lettering used: — </. = to distal end of; ?«. = to middle of; /.'' = fourth lateral process; c' and c-
first and second coxae respectively; 1^ and t^ first and second tibia respectively; T. raised femoral gland
tubercles present; v. vestigial; (E.) elevated.

rarely extends beyond the middle of the first coxa. In group II it is not elevated above
10° except in A'^. capense (45°), and generally extends at least to the distal end of the first
coxa (see Table III for exceptions).

Chelophore. In group II the scape often bears a number of long, rather stout setae;

30 • DISCOVERY REPORTS

in several species, however, these setae are short and few in number, e.g. A'^. articulare,
N. capense, N. co^npachim. In group I, the setae, as a rule, are either few and fine or
wanting; in a few exceptional cases (e.g. N. adareanum, Gordon, 1932, p. 99, fig. i a)
a few strong setae are present.

With regard to the chela there is so much variation in each group that no generaliza-
tion ^ can be made. The palm may be shorter than wide (A^. proceroides and A^. neumayri,
n.sp. (Figs. 9 and 29 d), or as much as 3-4 times as long as wide (A^. biarticidatum,
N. longicoxa, N. procerum, Fig. 32 a and Gordon, 1932, figs. 7 and 9 a). The number
of spinules on each finger of the chela is very variable (6-1 10, see Table III); in a few
instances there are many more spinules on one finger than on the other, e.g. A'^. longicoxa
(70 and 90-1 10).

Palp. Segments 2-5 are long and subequal in a few exceptional cases (A^. charcoti,
Nymphon, sp.? (St. 181), A^. lanare and A^. hrachyrhynchiim — the last is not closely
related to the other species). In all other forms, segments 2-5 are of difl^erent lengths
and segment 2 is usually longest although the third may not be much shorter. As a

general rule the ratio ^-^ — , — ^ ^ — 5 ijgg between 0-7 c; and 1-3 ; in five species

length of segment 2 / j j r

of group II it is less than o-6 (usually less than 0-5, see Table III).

Number of spines on oviger. With one exception (A^. adoreamim, Hodgson, 1907,
pi. iii, fig. 3 b) each of the four terminal segments of the oviger bears a number of
denticulate spines. Loman (1923, p. 11) has shown that the total number of these
spines remains within certain definite limits for each species. This total may be less
than ten or as high as seventy in exceptional cases (see Table III); also it may be the
same in several unrelated species, e.g. 30-38 in A^. tenuipes, N. hamatum and A", brachy-
rhynchmn of group I and A'^. anstrale, N. compoctiim and A^. orcadense of group II.

Third leg. In all the species of group II and most of those of group I a (in which
segment 5 of the male oviger is long, slender and distinctly curved) and in A^. charcoti
the second coxa is approximately equal in length^ to the sum of the first and third.
In all the other species of group I the second coxa, especially in the male, is much longer.
The femoral gland openings in the male vary greatly in number; in several instances
they have not been detected. In the female, the femur is at least five times as long as
wide in all species of group I, A^. pfefferi and A'^. adareanum excepted. In group II it
is usually less than four times as long as wide, but is nearly five times in each of the
atypical forms.

The second is longer than the first tibia, as a rule, in group I, and the reverse holds
for most of the species in group II (Table III).

1 Loman (1923, p. 10) says that in " Nymplioii" the "hand" is not conspicuously broadened distally,
whereas in " Chaetonymphun" it is much broadened distally, but I am afraid I do not quite understand this
distinction. The palm in Figs. 32 and 29 d is not much broadened distally, yet these species are typical
" Chaetonymphom" . The only Antarctic species in which the palm is broadened distally is TV. rmdtidens, n.sp.

(Fig- 35 d).

^ The dorsal measurements are always given for the segments of the leg so that the third coxa is much
shorter than if measured ventrally.

NYMPHONIDAE 31

The auxiliary claws are usually well developed, but may be absent or, as in one case,
vestigial. The presence or absence of these claws has been used as the first division in
Key II, which is not concerned with possible relationships between the species, because
this character is (i) convenient and (2) is, so far as our knowledge goes, constant for
each species at all stages from immature to adult, and in both sexes.

The third leg is 5-7 times as long as the trunk, and either smooth or slightly setose
in group I. In group II it is 3-4-5 times as long as the trunk in all forms except A'^. capense
and A^. compactiim (6) ; while the legs are usually very setose, the two species just
mentioned and A'^. articiilare are again exceptions.

Synonymy. With regard to synonymy, A^. gmcillimum, Caiman, may prove to be the
same as A^. hiemale, Hodgson; A^. charcoti, Bouvier, with A'^. lanare, Hodgson, and
A^. biarticulatum, Hodgson, with A'^. brevicaudatum, Miers.

SYNOPTIC KEY TO THE DETERMINATION OF THE ADULT MALES
OF ANTARCTIC AND SUB-ANTARCTIC SPECIES OF NYMPHON^

I. Segment 5 of (J oviger long and narrow, straight or distinctly curved, of approximately
uniform diameter throughout.

A. Auxiliary claws absent [tarsus and propodus both elongated, subequal, without spines
on ventral margin; second coxa long, segment 5 of (J oviger straight (type I)].

1. Less than 20 (10-13) spinules on each finger of chela; 21-24 denticulate spines on
oviger; 9-12 low femoral gland-tubercles; first tibia the longest segment; claw
three-fifths of propodus N. proceroides, Bouvier

2. More than 20 (28-42) spinules on each finger of chela; 24-34 denticulate spines
on oviger; 6-7 femoral gland-openings, but no raised tubercles; second tibia the
longest segment; claw subequal to propodus A^. teniiipcs, Bouvier

B. Auxiliary claws well developed [second tibia the longest segment].
I. Segment 5 of (^ oviger straight (type I).

a. Number of denticulate spines less than 50 (30-46).

i. Two terminal palpal segments together equal to or slightly longer than
second segment; tarsus rather shorter than propodus; 12-16 femoral gland-
openings — low tubercles occasionally present N. hiemale, Hodgson

ii. Two terminal palpal segments together rather shorter than second segment;
tarsus considerably longer than propodus; 10 low femoral gland tubercles

N. graciUiimim , Caiman

b. Number of denticulate spines exceeding 50 [14-22 spinules on each finger of
chela; 13-16 small femoral gland-openings] N. gracilipes, Miers

1 Use in conjunction with Table III, p. 28; the large Roman numerals in the left-hand column of table
represent the five primary subdivisions of the key. Slight discrepancies in the numbers of spines on the
fingers of the chela or of the oviger may occur, but are not important — e.g. in Table III the number of
spines of each finger of the chela of a single specimen of A^. longicoxa is given, while in the key the lowest
and the highest number of spines on the movable finger of several adults is given. In Table III the ratios
represented by fractions are only approximate ; where decimals are employed a higher degree of accuracy
has been aimed at.

32 DISCOVERY REPORTS

2. Segment 5 of (^ oviger distinctly curved (type I a).

a. More than 10 spinules on each finger of chela (20-30 in adults).

i. 26-34 denticulate spines on oviger; second coxa at least half as long again as
the sum of the first and third ; 8-9 femoral gland-openings, but no raised
tubercles ; tarsus at least half as long as propodus N. subtile, Loman

ii. 19-22 denticulate spines on oviger; second coxa scarcely longer than the
sum of the first and third; 4 wide femoral gland-tubercles; tarsus only
one-third of propodus N. pfejferi, Loraan

b. Less than 10 (6-9) spinules on each finger of chela [2-4 low, wide femoral
gland-tubercles].

i. Number of denticulate spines on oviger 17-20; tarsus half of propodus;
ventral margin of the latter setose N.paiicidens, n.sp.

ii. Number of spines on oviger 5-9, not denticulate; tarsus one-fourth of pro-
podus; ventral margin of the latter spinose N. adareanuw, Hodgson

II. Segment 5 of ,^ oviger relatively short — less than twice as long as segment 6 — and stout,
narrowed and slightly curved in proximal third (Fig. 10 h) [palpal segments 2-5 sub-
equal, long; auxiliary claws absent; second tibia the longest segment; 20 femoral gland-
openings; 40-50 denticulate spines on oviger; 75-110 spinules on movable finger of chela
(10-20 less on immovable one); tarsus, propodus and claws subequal] ...A^. charcoti, Bouvier

III. Segment 5 of (J oviger long and slender; expanded somewhat at distal end (type I b)
[total number of denticulate spines on oviger 30-44].

A. Neck relatively long with oviger base inserted midway between first lateral process
and anterior cephalic lobe ; auxiliary claws absent [chela long and sickle shaped ; two
terminal palpal segments together |-| of second segment].

1. Second coxa more than twice the sum of the first and third; second tibia the
longest segment; 90-110 spinules on movable finger of chela; eyes present

A'^. longicoxa, Hoek

2. Second coxa less than twice the sum of the first and third (1-4 : i) ; first tibia the
longest segment ; 60-65 spinules on movable finger of chela ; eyes absent [spur on
distal end of femur; 12-13 femoral gland-tubercles] N. hamatum, Uoek

B. Neck relatively short, oviger base in contact with first lateral process; auxiliary claws
present [second tibia the longest segment; second coxa long].

1 . Palpal segments 3 and 4 subequal and longer than 2 and 5 which are again sub-
equal ; fingers of chela curved inwards towards scape distally, each armed with
40-60 spinules ; 30 femoral gland-openings N. brachyrhynchum, Hoek

2. Palpal segments 4 and 5 together rather longer than second segment, 2 and 3 sub-
equal ; fingers of chela not curved distally, each armed with 30-45 spinules; 10-12
femoral gland-openings N. clarencei, n.sp.

IV. Segment 5 of ,^ oviger relatively short, portions of segments 5 and 6 or 3, 4 and 6 thin-
walled, inflated, often partially collapsed in fixed material (type II a); neck short, body
rather compact; auxiliary claws absent or vestigial.

A. Second tibia the longest segment; 60-65 spinules on each finger of chela; 48-50
denticulate spines on oviger; tarsus shorter than propodus (4:5) [portions of
segments 3, 4 and 6 of cJ oviger thin-walled, segment 6 setose on the thin-walled area,
segment 5 short, slightly curved and widening gradually towards the distal end]

N. capense, Hodgson

NYMPHONIDAE 33

B. First tibia the longest segment; 38-48 spinules on each finger of chela; 23-38 denti-
culate spines on oviger ; tarsus longer than propodus.

1 . Auxiliary claws absent ; tarsus twice as long as propodus ; segment 5 of (J oviger
short, considerably expanded in distal half; 20-27 femoral gland-openings

N. compactimi, Hoek

2. Auxiliary claws vestigial; tarsus not more than half as long again as propodus;
segment 5 of (J oviger greatly inflated in distal half (thin-walled part usually
collapsed); 16-18 femoral gland-openings N. australe, Hodgson

V. Segment 5 of (^ oviger relatively short and distinctly clubbed at distal end (type II); neck
short and body compact ; auxiliary claws well developed.

A. Two terminal palpal segments together at least three-fourths of second segment; no
femoral gland-tubercles; number of denticulate spines on oviger exceeding 20 (22-36)
in adults.

1. Second tibia the longest segment; 30-36 denticulate spines on oviger; 35-46
spinules on each finger of chela ; two terminal palpal segments together subequal

to second segment; 30 femoral gland-openings A^. orcadense, Hodgson

2. First tibia the longest segment; 22-27 denticulate spines on oviger; less than 30
spinules on each finger of chela; two terminal palpal segments together three-
fourths of second segment ; femoral gland-openings not observed.

a. 18-25 short curved spinules on each finger of chela; palm approximately twice

as long as high N. articulare, Hodgson

b. 6-9 long spinules on each finger of chela ; palm only as long as high [trunk
segments 3 and 4 fused] N. tieiimayri, n.sp.

B . Two terminal palpal segments together approximately one-half to one-fourth of second
segment ; 3-7 wide femoral gland-tubercles ; number of denticulate spines on oviger
less than 20 (10-18) [first tibia the longest segment; trunk segments 3 and 4 fused].

1. Number of spinules on each finger of chela exceeding 25 (28-45).

a. Ocular tubercle of medium height; fingers of chela rather shorter than palm;
tarsus in adults considerably shorter than propodus (much shorter in immature
specimens) A'^. hrevicaudahim , Miers

b. Ocular tubercle high; fingers of chela equal to or rather longer than palm;
tarsus in adults rather longer than, in immature specimens subequal to,
propodus N. biarticidatum, Hodgson

2. Number of spinules on each finger of chela less than 25 (9-20).

a. Fingers longer than palm [each armed with approximately 20 spinules; 2-3
short, spinose projections on each lateral process and each first coxa ; tarsus and
propodus spinose] N. mendosum, Hodgson

b. Fingers shorter than palm.

i. Ocular tubercle only as high as wide ; two terminal palpal segments together
two-fifths of second segment ; setae on legs much shorter than the diameter
of the segments carrying them A'^. proximum, Caiman

ii. Ocular tubercle long and slender, 4-6 times as high as wide; two terminal
segments very short, together only one-fourth of second segment ; long setae
on scape, lateral processes and legs, equal to or longer than the diameter of
the segments carrying them N. bouvieri, n.sp.

34

DISCOVERY REPORTS

ARTIFICIAL KEY TO THE DETERMINATION OF ANTARCTIC
AND SUB-ANTARCTIC SPECIES OF NYMPHON

1 . Auxiliary claws absent or vestigial 2

Auxiliary' claws well developed 12

2. Four terminal palpal segments long, subequal 3

Four terminal palpal segments of different size; two terminal segments together from

§-i| of second segment 4

3. 40-51 denticulate spines in adults ; 90-1 10 spinules on movable finger of chela ; legs smooth ;

3 ovigeras in Fig. 10 b A^. clmiroti, Bouvier

30-31 denticulate spines in adults; 70-85 spinules on fingers of chela; legs setose, setae

very fine A^ charcoti var. ( ? = lanare, Hodgson)

4. Oviger base arises from middle of neck 5

Oviger base in contact with first lateral process 7

5. Eyes well developed [second tibia the longer segment; 90-100 spinules on movable finger

of chela which has the tip recurved] N. longicoxa, Hoek

Eyes absent ; ocular tubercle obsolete 6

6. Spur on distal end of femur ; first tibia rather longer than second ; 50-65 spinules on each

finger of chela ; tarsus and propodus subequal N. hamatum , Hoek

No spur on distal end of femur ; second tibia the longer segment ; 38 and 55 spinules on each

finger of chela; tarsus just over one-half of propodus N. procerum, Hoek

7. Neck long, lateral processes separated by their own diameter or more 8

Neck short, lateral processes separated by half their own diameter, or less, as a rule 10

8. Spinules on each finger of chela less than 20 (10-13) 9

Spinules on each finger of chela more than 20 (28-42) [palm long and narrow ; S oviger of

type I second tibia the longest segment ; 3-4 long setae at distal end of femur]

A'^. tsnuipes, Bouvier

9. Eyes well developed palm very short ; (^ oviger of type I ; first tibia the longer segment ;

second coxa half as long again as the sum of the other two A^ proceroides, Bouvier

Eyes absent, though ocular tubercle present; second tibia the longest segment; second coxa

nearly three times the sum of the other two N. longicollum, Hoek

10. First tibia the longer segment ; tarsus longer than propodus ; 38-48 spinules on each finger

of chela ; 23-38 denticulate spines on oviger 11

Second tibia the longer segment ; tarsus four-fifths of propodus ; 60-65 spinules on each

finger of chela; 48-50 denticulate spines on oviger A^. capense, Hodgson

11. Eyes and ocular tubercle absent; tarsus twice as long as propodus A', compacttim, Hoek

Eyes present, ocular tubercle high ; tarsus half as long again as propodus or rather less

N. australe, Hodgson

12. Spines on four terminal segments of oviger not denticulate and numbering less than

10 (5-9) [(? oviger of type I a; 6-9 spinules on each finger of chela; tarsus one-fourth
of propodus, the latter spinose on ventral margin; neck short] ...A'^. adareamim, Hodgson
Spines on four terminal segments of oviger denticulate 13

13. Four terminal segments of palp each long, third and fourth subequal and longer than

second and fifth which are again subequal [cj oviger of type I b ; 40-60 spinules on

each finger of chela, fingers slender and bent distally] N. brachyrhynchum , Hoek

Four terminal segments of palp of different size ; two terminal segments together vary from

equality with, to one-fourth of, second segment 14

NYMPHONIDAE 35

14. Second tibia the longer segment 15

First tibia the longer segment [cJ oviger of type II] 21

15. Total number of denticulate spines on oviger at least 70 ; chela with immovable finger very

short, only half of palm, 24-28 crowded spinules on each finger ; tarsus half of propodus.

N. nniltidens, n.sp.
Total number of denticulate spines on oviger 54-59 ; 14-22 spinules on each finger of chela ;

tarsus and propodus subequal {S oviger of type I] N. gracilipes, Miers

Total number of denticulate spines on oviger 46-30 16

Total number of denticulate spines on oviger less than 25 20

16. Approximately twice as many spinules on movable as on immovable finger of chela

iN. distensum, Mobius^
\A^ signaiiim, Mobius
An almost equal number of spinules on each fiinger of chela or diflterence not exceeding

2 5 per cent 1 7

17. Second coxa half as long again the sum of the first and third or longer i8

Second coxa short, subequal to the sum of the other two [neck short; ^ oviger of type II

and segment 7 termino-laterally articulated on segment 6; 40-46 spinules on each
finger of chela ; at least 30 femoral gland-openings in cj] N. orcadense, Hodgson

18. Tarsus and propodus subequal (f-|); 27-40 spinules on each finger of chela; S oviger of

type I; 10-16 femoral gland-openings in ^ iV. hiemale, Hodgson

(? = iV. gracilliinitm, Caiman)

Tarsus J— § of propodus 1 9

19. Neck long ; 14-24 spinules on each finger of chela ; oviger of type I a and 4 femoral gland-

tubercles in cj N. subtile, Loman

Neck short to medium ; 30-45 spinules on each finger of chela ; oviger of type I b and 10-12

femoral gland-openings in (J ^- clarenci, n.sp.

20. Less than 10 spinules on each finger of chela; tarsus half of propodus; oviger of type I a

and 4 femoral gland-tubercles in S N.pmicidens, n.sp.

[15-16 spinules on each finger of chela; neck long; tarsus half of propodus, the latter

spinose on ventral margin (type immature) N. frigidiim, Hodgson]

19-30 spinules on each finger of chela; tarsus one-third of propodus ; neck short; oviger of

type I a and 4 femoral gland-tubercles in c? N- PMeri, Loman

(= iV. antarcticum, Pfeffer)

21. Number of denticulate spines on oviger exceeding 20 [two terminal palpal segments

together three-fourths of second segment; femur much enlarged in $] 22

Number of denticulate spines on oviger less than 20 [and trunk segments 3 and 4 fused] ...23

22. 6-9 long spinules on each finger of chela, fingers curved, gaping ; palm only as long as high

[trunk segments 3 and 4 fused] N. rieumayri, n.sp.

18-25 short, curved spinules on each finger of chela; palm approximately twice as long

as high ^- articiilare, Hodgson

23. Two terminal palpal segments together two-thirds of second segment; 18-22 spinules on

each finger of chela N.villosum, Hodgson

Two terminal palpal segments together one-half of second segment or rather less 24

Two terminal palpal segments together one-fourth of second segment; ocular tubercle

very slender and 4-6 times as high as wide N. bouvieri, n.sp.

24. Number of spinules on each finger of chela exceeding 30 ; ocular tubercle higher than wide . . .25
Number of spinules on each finger of chela less than 20; ocular tubercle low and wide 26

1 These two species appear to differ somewhat, but each is imperfectly known.

5-2

36

DISCOVERY REPORTS

25. Ocular tubercle of medium height; fingers rather shorter than palm; tarsus somewhat

shorter than propodus in adults, much shorter in young specimens

A'^. brevicaudatiim, Miers
Ocular tubercle high ; fingers equal to or rather longer than palm ; tarsus scarcely shorter
than propodus in young, rather longer than the latter in adults

A'^. biarticulatum, Hodgson

26. Fingers longer than palm, spinules long and numbering about 20; 2 or 3 short spinose

projections on each lateral process and each first coxa N. mendosiim, Hodgson

Fingers shorter than palm, spinules short, numbering 9 on immovable, 14 on movable

finger; no spinose projections on lateral processes or first coxae N. proximum, Caiman

A. GROUP I

Nymphon proceroides, Bouvier (Figs. 9, 10 a and 75).

Bouvier, 1913, pp. 90-94, text-figs. 42-50.
Loman, 1923, p. 14 (in key).

St. 181. 12. iii. 27. SchoUaert Channel, Palmer Archipelago, 64° 20' S, 63- 01' W, 160-335 m.; M.
Large otter trawl: a dozen specimens with A'^. tenuipes and Nymphon sp.?

St. 182. 14. iii. 27. SchoUaert Channel, Palmer Archipelago, 64° 21' S, 62° 58' W, 278-500 m.;
M. Large otter trawl : 2 ??, 4 SS and 2 immature specimens.

St. 186. 16. iii. 27. Fournier Bay, Anvers Island, Palmer Archipelago, 64° 25' 30" S, 63" 02' W,
295 m. ; M. Large dredge: 3 S3-

St. 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 64° 48' 30" S, 63' 31' 30" W, 259 m. ;
M. Large dredge: a dozen specimens.

Description of male. Trunk of loose build, lateral processes widely separated — ^up
to twice their own diameter. Neck long, base of oviger in contact with first lateral
process. Ocular tubercle low, wide,
rounded ; eyes well developed.

Proboscis short, robust, sub-cylindri-
cal, much shorter than cephalon.

Abdomen reaching to middle of first
coxa, elevated at an angle of about 20''.

Chelophore. Scape rather longer than
proboscis, length 3-5-4 times the distal
width. Palm very short and high;
fingers twice as long as palm, each
armed with 10-13 curved spinules as
represented in Fig. 9 ; setose pad extending nearly half-way along immovable finger.

Palp. Third a little longer than, two terminal segments together longer than, second
segment (i-25-i-33 : i); segments 2-5 in the proportions 3-25 : 3-5 : 2-5 : 1-65.

Oviger (type I). Terminal claw approximately two-thirds of tenth segment, armed
with 4-7 spinules. Total number of denticulate spines 21-24. Segments 4-6 in the
proportions 2-14 : 2-86 : i (? 1-5:2:1); fifth segment nearly straight, long, and of
almost uniform diameter throughout (Fig. 10 a).

Fig. 9. Nymphon proceroides, Bouy'ier. Chela: x 60.

NYMPHONIDAE

37

Third leg. Second coxa long and slender, half as long again as the sum of the other
two (1-5-17 : i). Femur slightly shorter than second tibia; a row of 9-12 prominent
gland papillae on proximal, mid-ventral surface. First tibia the longest segment, half
as long again as femur. Tarsus and propodus both elongated, subequal; claw long,
about three-fifths of propodus ; auxiliaries absent.

Measurements {mm.)

?

Third leg:

?

0

1-2

First coxa ...

0-8

0-7

0-67

Second coxa

2-0

17

5-0

Third coxa ...

0-8

°-53

2-13

Femur

. ... 3-6

3-4

1-07

First tibia ...

5-47

4-6

0-4

Second tibia

4-0

37

2-6

Tarsus

1-55

1-47

0-93

Propodus

. ... 1-65

1-47

Claw

i-o

0-93

Length of proboscis ...
Diameter of proboscis
Length of trunk
Length of cephalic segment ...
Width of anterior cephalic lobes

Width of neck ...

Width across second lateral processes
Length of abdomen ...

Remarks. Bouvier has given a long and careful description of this species ; Fig. 75
represents an abnormal palp of a co-type in the Paris Museum collection. The Discovery
specimens agree very closely with the types.

This species is most closely allied to N. teniiipes, Bouvier (p. 39), but can readily be
distinguished from the latter by (i) the chela with its shorter palm and fewer spinules
on the fingers, and (2) the absence of very long setae on the walking legs.

Distribution. Previously recorded from the South Shetlands.

Nymphon tenuipes, Bouvier (Figs. 10 ^ and 11).

Bouvier, 1913, pp. 86-90, te.xt-figs. 35-41.
Loman, 1923, p. 14 (in key).

St. 177. 5. ill. 27. 27 miles SW of Deception Island, South Shetlands, 63° 17' 30" S, 61° 17' W,
1080 m. ; M. cs. St. Large dredge: i cJ.

St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 64° 20' S, 63° 01' W, 160-335 ni-!
M. Large otter trawl: i ? and 5 immature specimens.

Description of mature male. Trunk slender and loosely built; lateral processes
separated by one and a half, to more than twice, their own diameter. Neck long, base
of oviger in contact with first lateral process. Ocular tubercle low, broad, rounded at
apex; eyes well developed.

Proboscis sub-cylindrical, three-fourths of scape and of cephalon.

Abdomen short, reaching to distal end of fourth lateral process, elevated at an angle
of about 40-45°.

Chelopiiore slender. Scape with a few fine setae; length six times the distal width.
Chela subequal to scape, palm long and narrow, setose pad very short ; fingers armed
with about 39 and 42 spinules respectively (Fig. 11 b).

Palp slender. Segments 2-5 in the proportions 10 : 6-5 : 5 : 5.

38

DISCOVERY REPORTS

Oviger long and slender. Terminal claw a little shorter than tenth segment, armed
with ten long spinules. Total number of denticulate spines 34 (lo + 9 + 7 + 8). Segments
4-6 in the proportions 2 : 3-43 : i ; fifth
segment long, slender, nearly straight,
and of almost uniform diameter through-
out (Fig. 10 c).

Third leg long and slender. Second
coxa twice the sum of the other two.
Femur shorter than either tibia; three
very long, slender setae dorsally at distal
articulation; no distinct raised gland tu-
bercles on mid-ventral surface but six or
seven obscure gland openings present.
Both tibias bearing long slender setae
(Bouvier, 1913, fig. 41), second the long-
est segment. Tarsus, propodus and claw
elongated; claw shorter than propodus
in this specimen (but equal to or even
slightly longer than propodus in the other
specimens). Auxiliaries absent.

Female. One specimen from St. 181
is undoubtedly a female of this species,
but it difl^ers from the normal female in
one interesting respect. The two pos-
terior pairs of walking legs are quite
typical; each second coxa is enlarged distally and has a large genital pore; the
femur is considerably widened proximally and contains developing ova (Fig. 1 1 a).

Fig. 10. Segments 4-6 of male oviger of : a. Nymphon
proceroides, Bouvier: x 27. b. N. charcoli, Bouvier:
X 7. c. N. tenuipes, Bouvier: x 27.

Fig. 1 1 . Nymphon tenuipes, Bouvier. a. Femur and two distal coxae of third
leg of female: x 20. b. Chela x 60.

NYMPHONIDAE 39

The two anterior pairs of legs, however, are short,^ slender like those of immature

specimens, and have no genital pores.

The oviger is of the usual female type ; the fifth is scarcely longer than the fourth,

and twice as long as the sixth segment. The total number of denticulate spines is

32 (9 + 8 + 7 + 8).

Measurements {mm.)

St. 177 (J

St. 181

?

Length of proboscis

0-9

i-o

Diameter of proboscis

0-37

0-47

Length of trunk

2-8

2-95

Length of cephalic segment

1-23

1-35

Width of anterior cephaHc lobes

0-5

0-55

Width of neck

0-25

0-3

Width across second lateral processes

1-33

1-2

Length of abdomen

°-2S

0-5

Leg:

(3rd)

(4th)

(ist)

First coxa ...

0-4

0-4

0-4

Second coxa

1-4

1-6

i-i

Third coxa ...

0-3

0-35

0-3

Femur

3-33

3-4

2-13

First tibia ...

47

4-93

2-67

Second tibia

5-0

5-47

2-8

Tarsus

1-5

1-6

073

Propodus . . .

1-33

1-33

073

Claw

i-o

1-25

0-59

Remarks. This is a very small species ; Bouvier states that the holotype is immature,
but it is in reality an almost, if not quite, mature male. There are 32-33 denticulate
spines on the oviger and about twelve spinules on the terminal claw, although the latter
have been omitted in Fig. 38 (Bouvier, 1913, p. 88). The spinules on the fingers of the
chela number approximately 24 and 30-35 respectively.

A'^. tetmipes is most nearly related to N. proceroides, Bouvier, with which it was taken
at one station, but the chelophore is more slender, and the chela especially is very
different (cf. Figs. 9 and 11^); the long setae on the femur are very characteristic.

Distribution. Previously recorded from the South Shetland Islands.

Nymphon hiemale, Hodgson (Figs. 12a and 13c).

Nymp/io?i hiemale, Hodgson, 1907, p. 20, pi. iii, fig. i ; pi. x, fig. 8.

Nymphoji hiemale, Bouvier, 1913, p. 73 (in key).

Nymphon hiemale. Caiman, 1915, p. 32.

Nytiiphoii hiemale, Loman, 1923, pp. 14 and 16.

? Nymphon gracillimum, Caiman, 1915, p. 30, Fig. 5 A-D.

St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, 3-3 miles S 45° E of Jason Light,
no m.; M. R. Large dredge: 4 specimens, including i ovigerous (^.

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock, 179-235 m.; gy. M. Large otter trawl: 2 specimens,
I overgrown by a large compound Tunicate.

1 Compare the measurements of the fourth and the first leg; the third is incomplete.

40 DISCOVERY REPORTS

St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of
Jason Light to 4 miles N 39° E of Jason Light, 120-204 m. ; M. Large otter trawl : 3 specimens.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point, to i -2 miles S 62° W of Merton Rock, 230-250 m. ; gy. M. Large otter trawl : 4 speci-
mens, including i ovigerous (^.

St. 140. 23.. xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 54° 02' S,
36''38'W to 54" 11' 30" S, 36°29'W, 122-136 m.; gn. M. St. Large otter trawl: 5 specimens
including i ovigerous <J.

St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, from 54^ 04' S, 36" 27' W
to 53° 58' S, 36° 26' W, 155-178 m.; gn. M. S. Large otter trawl: i specimen.

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76^° W
to 2-62 miles S 11° W of Merton Rock, 200-234 m.; M. Large otter trawl: 5 adult (including
2 ovigerous and i larvigerous S3) and i immature specimens.

St. 159. 21. i. 27. 53° 52' 30" S, 36° 08' W, 160 m.; R. Large dredge: 2 specimens.

St. 160. 7. ii. 27. Near Shag Rocks, 53° 43' 40" S, 40° 57' W, 177 m.; gy. M. St. R. Large
dredge: 3 ,S<S (i ovigerous and i larvigerous). On the label there is the following note: "Body
testaceous. Proboscis with red flecks dorsally. Ill-defined red patches as follows: at either end of
basal segment of chelophore: on leg processes of body: on ist and 3rd leg segments, on 2nd and
4th at proximal end and on 2nd, 4th and 5th in distal third. All legs similarly coloured".

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m.; R.
Large dredge : i very small but ovigerous o .

Redescription of HOLOTYPE.i Trunk with lateral processes separated by approxi-
mately their own diameter. Neck long, the oviger base occupying one-third thereof
and in contact with first lateral process. Ocular tubercle as high as wide, rounded;
eyes well developed.

Proboscis equal to cephalic segment (cf. Caiman, 1915, p. 32), subequal to scape,
somewhat expanded in middle and again at apex.

Abdomen short, reaching to distal end of fourth lateral process ; elevated at an angle
of about 25°.

Chelophore. Scape rather longer than chela, length about five times the maximum
width. Fingers somewhat longer than palm; approximately thirty-five spinules on
movable, twenty-seven on immovable one ; setose pad extending nearly one-third along
the latter.

Polp with segments 2-5 in the proportions 8:7:4:5.

Oviger. Terminal claw rather more than half of tenth segment, armed with 10-12
spinules ; total number of denticulate spines, forty-one.

Third leg. Second coxa two-thirds as long again as the sum of the other two (1-65 : i,
measured dorsally). Femur a little shorter than, second half as long again as, first tibia.
Tarsus subequal to propodus (8 : 9), ventral margin of the latter spinose (Fig. 12 a).
Claw not quite half as long as propodus, auxiliaries well developed.

1 Female, selected by Caiman, 191 5, p. 32, in B.IM. collection.

NYMPHONIDAE

41

Male paratypes, measuring 6-8 and 7-1 mm.^ in length (trunk only), agree in most
respects with the holotype. There are forty-two and forty-three denticulate spines
respectively on the four terminal segments of the oviger. The second coxa is longer
(1-85 and 1-95 : i).

a

J-* ^ i>j> — >< ^

Fig. 12. Terminal segments of third leg of: a. Nymphon hiemale, Hodgson, paratype. b. N. gracillimum,

Caiman, holotype. c. N. gracilipes, Miers, holotype.

Table IV

L. of trunk (in mm.)

No. of spinules on fingers
of chela

L. of palpal segments 4+5
L. of segment 2

No. of denticulate spines
on oviger

Third leg :

L. of coxa 2

{b)

L. of coxae 1+3
L. of tarsus
L. of propodus

A'^. hiemale
paratypes

6-8 7-2

30-36
I I -06

42 43

1-64 1-95
0-89 i-o

Discovery collection

St. 27

6-0

24,29

0-87

31

0-83

St. 123

6-0

27.33
1-05

40

I -So
0-97

3
St. 160

6-0

33.38
i-o

38
1-90

St. 140

7-2

34.38
I -08

40

I -06

^1
St. 149

6-8
36,40

1-02
36

2-30
1-04

^2

St. 149

6-4

30,35
I-o

36

I -60
i-i6

^3
St. 149

6-8

27,33
o-8i

34

1-25

A'', gra-
cillimum

cJ
holotype

4-80

22,24
0-85

33

2-1

1-33

^ Measurements of holotype in Caiman, 1915, p. 32.

42

DISCOVERY REPORTS

On the proximal ventral margin of the femur are 12-16 minute gland openings, on a
level with the surface. Segments 4-6 of the oviger are in the proportions i-6 : 2-8 : i
(Fig. 13 f); the fifth segment is long, slender, slightly curved and narrowed proximally.

Remarks. The holotype of A'', gracillimim, Caiman, is very closely related to, and
may even be co-specific with, N. hiemale. In the Discovery collection there are many
specimens that show characters intermediate between the two species (see Table IV),
and one male from St. 149 that agrees very closely with the holotype of N. gracillimum
(Table IV, St. 149, S^- The femoral glands in the latter open on distinct tubercles;
in the specimens referred to N. hiemale the gland openings are usually flush with the
surface, but may be situated on faint, low tubercles.

Distribution. This species has been recorded from the Ross Sea area (Hodgson,
1907; Caiman, 191 5); it is quite abundant around South Georgia and occurs as far
south as Clarence Island.

Nymphon gracillimum. Caiman (Figs. 12 ^ and 13 a).
Caiman, 1915, p. 30, fig. 5 A-D.
St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76^° W to

2-62 miles S 11° W of Merton Rock, 200-234 m. ; M

The holotype differs from the types
of A'", hiemale, Hodgson, in the following
respects. The proboscis is shorter than
the cephalic segment ; the two terminal
segments of the palp measure 0-85 of
the length of the second segment.
There are twenty-two and twenty-four
spinules respectively on each finger of
the chela, and there is only a trace of
the setose pad at the proximal end of
the immovable finger. The tarsus is one
and a third times the length of the pro-
podus and the second coxa is 2-1 times
the sum of the other two coxal seg-
ments. (The second tibia, as in A^. hie-
male, is nearly 1-5 times as long as the
first.) There are thirty-three denticulate
spines on the oviger, and the femoral
gland openings are not flush with the
ventral surface but situated at the apices
of distinct low tubercles.

The difference in the number of
spines on the oviger may be due to the

Large otter trawl : i ovigerous <^.

Fig.

gracillimum

X15.

Segments 4-6 of male oviger of: a. Nymphon
Caiman: x 20. b. N. gracilipes, Miers:
c. N. hiemale, Hodgson: x 20.

smaller size of the specimens (4-8 mm. as against 6-8-7-2 mm. in the types of N. hiemale).

NYMPHONIDAE

43

In the Discovery collection there is one ovigerous male, measuring 6-8 mm. in length,
that agrees very closely with the type of A^. gracillinmm. There are only thirty-four
denticulate spines on the oviger; the two terminal palpal segments together are four-
fifths of the second segment (Table IV, St. 149, S^); the second coxa is 2-2 times the
sum of the other two coxal segments. The femoral gland tubercles, however, are not
quite so prominent as in the type; the tarsus is one-fourth as long again as the
propodus. The number of spinules on each finger of the chela is twenty-seven and
thirty-three respectively.

Measurements (mm.)

Length of proboscis ...

Diameter of proboscis

Length of trunk

Length of cephalic segment ...

Width of anterior cephaHc lobes

Width of neck

Width across second lateral processes

Length of abdomen ...

Third leg:

2-3

First coxa .

0-8

Second coxa

6-8

Third coxa .

3'2

Femur

1-2

First tibia

0-53

Second tibia

3-2

Tarsus

07

Propodus

Claw

Auxiliaries

I'O

4-0
0-8
8-0

1 0-0

H-5

2-0

1-6

0-8

0-2

Distribution. Previously recorded from McMurdo Sound.

Nymphon subtile, Loman (Figs. 14, 15 and 24 b).
Loman, 1923, pp. 14 and 19, fig. C.

St. 51. 4. V. 26. Off Eddystone Rock, East Falkland Island, from 7 miles N 50° E to 7-6 miles
N 63° E of Eddystone Rock, 105-115 m.; f. S. Large otter trawl: i ? from kelp root.

St. WS 229. I. vii. 28. 50° 35' S, 57° 20' W, 210-271 m.; f. gn. S. Commercial otter trawl:
2 ovigerous (JcJ, 2 young, with Pallene australis, n.sp.

St. WS 237. 7. vii. 28. 46° 00' S, 60° 05' W, 150-256 m.; c. br. S. Sh. Conmiercial otter trawl:
I 9, I ovigerous $.

St. WS245. 18. vii. 28. 52°36' 8,63° 40' W, 304-290 m.;d.gn.S. Sh. Commercial otter trawl :
I ?, I incomplete specimen.

Description of male. Trunk loosely built, lateral processes separated by wide
intervals (one and a half to two and a half times their own diameter). Cephalic segment
approximately equal to the sum of the three posterior segments ; neck long ; oviger base
small and in contact with first lateral process. Ocular tubercle wide, low and rounded;
eyes conspicuous. Setae absent (Fig. 15 a).

Proboscis short, sub-cylindrical; nearly equal to, though apparently much shorter
than, scape (from dorsal aspect); foreshortened in Fig, 15 a.

Abdotnen short, reaching to distal end of fourth lateral process or a little beyond ;
elevated at an angle of about 35°.

Chelophore. Scape shorter than chela, length four and a half times the distal width.
Fingers rather longer than palm each armed with eighteen and twenty-two spinules

respectively (Fig. 14 b).

6-2

44

DISCOVERY REPORTS

Palp. Third segment four-fifths of, two terminal segments together equal to, second
(Fig. 14 c).

Oviger. Terminal claw approximately half of tenth segment, armed with five spinules.
Total number of denticulate spines 32 (9 + 8 + 7 + 8). Segments 4-6 in the pro-
portions 1-5 :2-5 : i; fifth segment long, curved, and of almost uniform diameter
throughout (Fig. 14 a).

Third leg slender and sparsely setose. Second coxa longer than the sum of the other
two (dorsal measurements — i "5-1 "75 : i)- Femur shorter than either tibia; there appear
to be 8-9 minute gland openings on proximal two-thirds of mid-ventral surface but no
raised tubercles. Second tibia the longest segment. Tarsus half to two-thirds of pro-
podus; claw short, about one-third of propodus, auxiliaries well developed (half to
two-thirds of main claw). Numerous small spinose setae on ventral margin of tarsus
and of propodus (Fig. 24 b).

Female. The femur and second coxa are more robust and segments 4-6 of the oviger
are relatively short and almost straight, in the proportions i-88 : 2-25 : i. The second
coxa is never more than half as long again as the sum of the other two.

Measurements {mm.)

Length of proboscis ...

Diameter of proboscis

Length of trunk

Length of cephaUc segment

Width of cephalic lobes

Width of neck

Width across second lateral processes

Length of abdomen ...

Third leg:

First coxa ...

Second coxa

Third coxa ...

Femur

First tibia ...

Second tibia

Tarsus

Propodus ...

Claw

Auxiliaries ...

Remarks. These specimens agree closely with Loman's description (1923, p. 19) of
the holotype, an ovigerous male. The number of denticulate spines on the oviger varies
from twenty-four to thirty-six, the number of spinules on the fingers of the chela from
fourteen to twenty-four.

This species is most closely allied to N. pfejferi, Loman, and A'^. pancidens, n.sp.
(especially to the former), but differs from both in that the neck is longer and the
number of denticulate spines is higher. N. adareanum, Hodgson, has the same type of
male oviger (see Table III), but it diff'ers from all other Nymphon species in having
simple spines on the four terminal segments (see Gordon, 1932, p. 98).

S (WS 237)
0-6

? (WS 245)
0-85

0-4

2-4

0-47
2-6

1-2

0-6

1-25
0-6

0-25
1-6

0-27
1-4

0-4

0-35

0-4

04

1-3

i-o

0-35

0-3

2-5

2-4

2-9

27

375

3-4

07

°-5

i-i

i-o

0-35

0-34

0-25

0-2

Fig. 14. Nymphon subtile, Loman: a. Segments 4-6 of male oviger. b. Chela, c. Palp.

Fig. 15. Nymphon subtile, Loman: a. Dorsal view of body with chelophores
and palps — proboscis fore-shortened, b. Third leg of female from St. 51.

46

DISCOVERY REPORTS

Nymphon pfefferi (Pfeffer), Loman (Figs. i6, 176, d and 18 «, b, d).
N. antarcticum, Pfeffer, i?,^,Jahrb. Hamburg Wiss. Anst., vi, Heft 2, p. 42.
N. pfefferi, n.n., Loman, 1923, p. 17, fig. B.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point, to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl: 2
immature specimens.

St. 140. 23. xii. 26. Stronmess Harbour to Larsen Point, South Georgia, from 54° 02' S, 36° 38' W
to 54° 11' 30" S, 36° 29' W, 122-136 m.; gn. M. St. Large otter trawl: 6 specimens.

St. 141. 29. xii. 26. East Cumberland Bay, South Georgia, 200 yards from shore, under Mount
Duse, 17-27 m.; M. Small beam trawl: i (J.

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N y6l° W
to 2-62 miles S 11° W of Merton Rock, 200-234 "i-! M. Large otter trawl: i ?.

Fig. 16. Nymphon pfefferi, Loman: a. Dorsal view of body of male from St. 141,
with chelophores and palps: x 40. b. Third leg of female from St. 149.

Description of male. Trunk with lateral processes separated by their own diameter
as a rule. Neck short, almost entirely occupied laterally by the base of the oviger;
cephalon much expanded anteriorly. Ocular tubercle low, wide, rounded; eyes con-
spicuous. Setae absent (Fig. 16 a).

Proboscis short; subequal to, or shorter than scape, and half to three-fifths of
cephalic segment.

Abdomen long, slender; reaching to distal end of first coxa; elevated at an angle of

about 30°.

Chelophore. Scape short, length two and a half to three times the distal width. Chela
subequal to scape; fingers one-third as long again as palm armed each with 19-30
spinules; setose pad extending half-way along immovable finger (Fig. 17 b and d).

NYMPHONIDAE

47

Palp. Terminal longer than penultimate segment, both setose (Fig. i8 b). Two
terminal segments together equal to third, and a little shorter than second segment.

Oviger. Terminal claw equal to tenth segment and armed with 5-7 spinules. Total
number of denticulate spines 19-22. Segments 4-6 in the proportions 1-3 : 1-73 : i
(Fig. 18 a); fifth segment curved and of almost uniform diameter throughout.

Third leg. Second coxa a little longer than the sum of the other two. Femur slightly
shorter than first tibia; four low wide gland tubercles on proximal half of ventral
margin. Second tibia the longest segment. Tarsus short, approximately one-third of
propodus; claw short; auxiliaries long (|-f of claw). A number of rather long setae
on coxae, femur and first tibia, those on the distal segments shorter and more numerous.

Female. The femur is more robust, the coxae less setose and segments 4-6 of oviger
relatively short and straight (i'33 : 1-33 : i).

^T^T^TZ

Fig. 17. Nymphon pfejferi, Loman: b. Chelophore of young specimen (length = 1-4 mm.), d. Chela of

adult co-type.
A'^./)a««V/eHi, n.sp.: a. Chelophore of young specimen (length = i-8mm.). e. Chela of adult: x 60.
N. tridentaturn, Hodgson: c. Chela of holotype (young).

Measurements (mm.)

Length of proboscis

Diameter of proboscis
Length of trunk

(J (St. 141)
0-5
0-35

2-0

9 (St. 149)
0-87

0-45

2-2

Length of cephalic segment

Width of cephalic lobes

Width across second lateral processes

Length of abdomen ...

i-o
0-8

1-65

0-8

l-O

0-9

1-65
0-85

Third leg:

Coxae

1-8

1-8

Femur

1-95

2-1

First tibia

2-2

2-2

Second tibia

2-8

2-8

Tarsus

Propodus

°-45
1-35

0-45

1-37

Claw

0-4

o'5

Auxiliaries

0-3

0-3

DISCOVERY REPORTS
4»

Remarks. This small species has only once been figured, and that rather poorly, by
Loman (1923, fig. B). The Discovery specimens agree well with co-types in the Hamburg
Museum. The proboscis varies considerably in relative length (e.g. the trunk is four
times as long as the proboscis in the male, and only two and a half times in the female
measured), and may be rather longer or shorter than the scape.

The differences between this species and N. paiicidem are listed on p. 51, and

illustrated in Figs. 16-20.

Loman (1923, p. 18) suggested that N. tridentatum, Hodgson, may be a young
specimen of N. pfefferi. This is very probable; the chelophore is of the same type with
eleven spinules on the immovable, twelve or thirteen on the movable finger (Fig. 17 c) ;
the ratio of tarsus to propodus is about the same (Fig. 18 c) and the proportions of
the palpal segments are very similar. The specimen is very immature, the total ^ length
being only i mm. ; it has a longer neck and longer setae on the distal segments of the
legs than are typical of adult specimens of N. pfefferi.

Nymphon paucidens, n.sp. (Figs. 17 fl, e, 19 and 20).

St 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock, 179-235 ^■> EY- M. Large otter trawl: i 3, holotype.

St 123 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point, to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl: 2 33,
I 5, 2 immature specimens.

St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 54° 02' S,
36° 38' w'to 54° 1 1' 30" S, 36° 29' W, 122-136 m. ; gn. M. St. Large otter trawl : i ? and 3 immature
specimens.

St 149 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76!° W
to 2-62 miles S 11° W of Merton Rock, 200-234 m.; M. Large otter trawl: 2 33, one carrying two
egg- masses on the left oviger.

St. MS 74. 17. iii. 26. East Cumberland Bay, i cable SE x E of Hope Point to 3-1 miles SW of
Merton Rock, 22-40 m. Small beam trawl: i immature specimen.

Description of holotype (3). Trunk with lateral processes separated by their own
diameter or rather more. Cephalic segment approximately equal to the sum of the three
posterior segments; neck short, base of oviger in contact with first lateral process and
occupying posterior half of neck (Fig. 19 a). Ocular tubercle rather higher than wide,
rounded; situated a little in front of first lateral processes.

Proboscis sub-cylindrical, rounded at apex; two-thirds of cephalic segment (dorsal

aspect).

Abdomen long, narrow, pyriform; extending a little beyond distal end of first coxa;

elevated at an angle of about 40°.

Palp setose. Second segment somewhat longer than, two terminal segments together

equal to, third (Fig. 20 b).

1 I.e. including proboscis and abdomen.

NYMPHONIDAE

49

Chelophore. Scape longer than proboscis, length nearly six times the distal width.
Chela a little shorter than scape ; palm very short ; each finger armed with only 7-9 long
spinules (Fig. 17 a, e).

Fig. 18. Nymphon pfefferi, Loman: a. Oviger of male from St. 141: x 70. b. Palp of a rather

small specimen from St. 140: x 90. d. Terminal segments of third leg of a co-type.
A'^. tridentatum, Hodgson : c. Terminal segments of fourth leg of holotype.

Oviger. Terminal claw equal to tenth segment armed with five long, slender spinules.
Total number of denticulate spines 17-20. Segments 4-6 in the proportions 2-2 : 2-87 : i ,
fifth segment slender, curved, and of almost uniform diameter throughout.

Third leg slender, with numerous long, fine setae, especially on the three longest
segments. Second coxa a little longer than the sum of the other two. Femur shorter
than either tibia ; four low wide gland tubercles on proximal half of ventral surface.
Second tibia the longest segment. Tarsus half of propodus ; auxiliaries half as long as
main claw (Fig. 20, cf. 19 b).

The. female has the walking legs, especially as regards the femur, a little more robust

so

DISCOVERY REPORTS

(Fig. 19 b). Segments 4-6 of the oviger are straight and relatively short, in the pro-
portions 175 : 1-5 : I. Otherwise it agrees with the male.

Fig. 19,

Nymphon paucidens, n.sp. a. Dorsal view of body of holotype with chelophores:
X 20 — proboscis foreshortened, b. Third leg of female from St. 140.

Measurements {mm.)

Length of proboscis . . .
Diameter of proboscis
Length of trunk
Length of cephalic segment
Width of cephalic lobes
"Width across second lateral
Length of abdomen . . .
Third leg

Coxae

Femur

First tibia . . .

Second tibia

Tarsus

Propodus ...

Claw

Auxiliaries ...

processes

holotype ? (St. 140)

0-7

0-8

0-35

0-37

2-4

2-5

i-i

1-2

07
1-6

1-6

0-85

i-o

2-0

1-9

2-4

2-8

2-5

2-8

3-4
0-6

3-2
0-6

1-2

1-3

0-45

0-45

0-25

0-25

NYMPHONIDAE

51

Remarks. This small species may be distinguished from A'^. pfefferi, Loman,
with which it occurred in two
localities, as follows, (i) The
trunk is less compact and the
cephalon less expanded anteriorly
(cf. Figs. 16 a and 19 a). (2) The
legs are more slender and setose
(cf. Figs. 166 and 196) and the
tarsus is consistently longer in
proportion to the propodus (one-
half and one-third respectively).
(3) The scape of the chelophore
is longer and more slender (cf.
Fig. 17 a and b) ; the palm is much
shorter and there are fewer spin-
ules on the fingers of the chela
(cf. Fig. ij e and d). Even in very
small specimens these differences
are apparent, and when the trunk
length is only 1-4 mm. there are
already 15-16 spinules on the
fingers of the chela in A'^. pfefferi
(Fig. 17 «, cf. Fig. 176). (4) The
male oviger is more slender and
segment 6 is relatively short — half
as long as, instead of subequal to,
segment 4 (Figs. 20 a and 18 a).

Fig. 20. Nymphon paucidens, n.sp.: a. Oviger of holotype:
X 47. b. Palp of holotype: < 60. c. Terminal segments of
third leg of female : x 47.

Nymphon charcoti, Bouvier (Fig. 10 b).

N. charcoti, Bouvier, 1911, p. 1138; 1913, p. 81, text-figs. 32-34.
N. charcoti, Caiman, 191 5, p. 29.
A'', charcoti, Loman, 1923, p. 15.

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock, 179-235 m.; gy. M. Large otter trawl: 17 specimens,
including ovigerous and larvigerous $$, several with encrusting Polyzoa.

St. 42. I. iv. 26. OfT mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of Jason
Light, to 4 miles N 39° E of Jason Light, 120-204 m., M. Large otter trawl: 7 large (i with encrusting
Polyzoa) and many small specimens.

St. 45. 6. iv. 26. 27 miles S 85° E of Jason Light, South Georgia, 238-270 m.; gy. M. Large
otter trawl: many specimens of all sizes, some males larvigerous, a few with encrusting Polyzoa
(with A'^. australe). A number of the smaller specimens are soft.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl: i <^,
3 $$ and several immature specimens.

7-2

52 DISCOVERY REPORTS

St. 142. 30. xii. 26. East Cumberland Bay, South Georgia, from 54° 11' 30" S, 36° 35' W to
54° 12' S, 36" 29' 30" W, 88-273 rn-' ^- Large otter trawl: 3 specimens, i ovigerous, i overgrown
with Hydroids, the third very soft.

St. 143. 30. xii. 26. Off mouth of East Cumberland Bay, South Georgia, 54° 12' S,
36° 29' 30" W, 273 m. Large otter trawl: 2 $$, i cJ (ovigerous and with i large and several small
Isopods attached), i immature.

St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, from 54° 04' S,
36° 27' W to 53° 58' S, 36° 26' W, 155-178 m.; gn. M. S. Large otter trawl: i <S (soft).

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76!° W
to 2-62 miles S 11° W of Merton Rock, 200-234 m. ; M. Large otter traw4: i $ and 4 immature
specimens.

St. 154. 18. i. 27. Jason Harbour to Larsen Point, South Georgia, from 2-6 miles S 84° W to
5I cables S 26° E of Larsen Point, 60-160 m. ; M. Large dredge: several adult and immature
specimens, including i ovigerous cJ.

St. 156. 20. i. 27. 53° 51' S, 36° 21' 30" W, 200-236 m.; R. Large dredge: i (^, i $ (claw
short).

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m.; R.
Large dredge : 2 specimens, i very flabby (claw short).

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Shetlands, 62° 07' S,
58° 28' 30" W, 391 m. ; M. St. Medium otter trawl: at least a hundred specimens of all sizes.

St. WS 32. 21. xii. 26. Mouth of Drygalski Fjord, South Georgia, 91-225 m.; gy. M. Small
beam trawl : i ?.

St. MS 68. 2. iii. 25. East Cumberland Bay, South Georgia, 17 miles SJE to 8| cables SE x E of
Sappho Point, 220-247 m. Large rectangular net : many specimens of all sizes including ovigerous
and larvigerous (^(S, one is overgrown with a compound Tunicate and Hydroids, one has an Isopod
attached.

Description of male. Trunk with lateral processes separated by a considerable
interval (o-6-i-5 of their own diameter). Neck rather long, base of oviger large and in
contact with, or a little in front of, first lateral process. Ocular tubercle as wide as, or
even wider than high ; eyes conspicuous.

Proboscis expanded somewhat in middle and again at apex; subequal to scape and
slightly shorter than cephaHc segment.

Abdomen short, scarcely reaching to distal end of fourth lateral process, elevated at
an angle of about 30^.

Chelophore. Scape shorter than the curved chela, length four and a half times the
distal width, bearing a number of short, stout setae. Palm long, narrow, and provided
with a number of short setae along mid-dorsal surface. Fingers rather longer than, and
at an angle of about 45° to, palm; 65-90 crowded spinules on immovable, 75-110 on
movable finger.

Polp long and slender; segments 2-5 subequal; in the proportions 10 : 10 : 11 : 10
and 12 : 14 : 13 : 14-5.

Oviger. Terminal claw a little more than half of tenth segment and armed with 10-16
long spinules. Total number of denticulate spines in adults 40-51. Segments 4-6 in
the proportions i : 1-71 : i ; fifth segment slightly curved and contracted in proximal
third, sixth segment bent in a very open S curve (Fig. 73).

NYMPHONIDAE

53

Third leg. Second coxa a little longer than the sum of the other two. Femur shorter
than either tibia ; gland openings very minute, and numerous (exceeding twenty in the
larger males) on proximal mid- ventral surface; raised gland tubercles absent. Second
tibia the longest segment. Tarsus and propodus elongated, subequal; in adults the
tarsus is longer than, in immature forms a little shorter than the propodus (see Table V).
Claw long, subequal to propodus in immature specimens, appreciably shorter than,
occasionally less than half of, propodus in adults (see Table V) ; auxiliaries absent. The
legs, as a rule, appear smooth or very slightly setose to the eye.

Table V

Station

Sex

Relative lengths (

3f

Tarsus

Propodus

Claw

MS 68

S

1-38

o-8i

39

143

156

?
?

1-33
1-28
1-40

071

072
0-58

170
195

?

1-23
1-17

0-44
0-68

195
195

Immature*

1-30
1-03

074
0-88

195

,,

0-96

I -06

195

))

0-85

1-03

195

»»

0-84

1-03

* The immature forms decrease gradually in size, the last measures only 4-4 mm. from
front of cephalon to tip of abdomen, the first is approximately twice as long.

Remarks. This species is exceedingly common on the western side of the Antarctic
and is also one of the largest of the genus (trunk length up to 18-5 mm.). The Discovery
specimens agree in most respects with Bouvier's description (1913, p. 81) ; but the femur
of the male is similar to that of the female, not dilated (see Caiman, 19 15, p. 29).

Very few specimens have been collected on the eastern side. The three Terra Nova
specimens from St. 294 (Caiman, 1915, p. 29) differ from the Discovery specimens of
the same size (length of trunk 12-13 mm.) in the following respects: (i) the legs are
clothed with numerous fine setae, (2) there are only 30 or 31, instead of 41-44 denti-
culate spines on the oviger, and (3) the fingers of the chela are armed with 70-85 instead
of 90-110 spinules. These three specimens are closely related to, if not identical with,
the types of A^. lanare, Hodgson (see also Loman, 1923, p. 15). Unfortunately the type
specimens of Hodgson's species are immature, but they differ from immature N. charcoti
(South Georgia region) precisely in the three above-mentioned respects. These
differences seemed to point to the existence of a distinct variety or race peculiar to the
Ross Sea area, but the Terra Nova specimen from St. 349 proved to be identical with
the typical N. charcoti from the Western Antarctic region. Until more material is
available from the eastern region it seems advisable to refer the three Terra Nova
specimens (St. 249) to N. lanare, Hodgson, and to retain the name N. lanare for those
specimens that have 10-13 fewer denticulate spines on the oviger, 20-25 fewer spinules
on each finger of the chela, and more hairy legs than N. charcoti.

54

DISCOVERY REPORTS

Distribution. Previously recorded from South Shetlands, Southern Chili, Ross Sea
and McMurdo Sound ; South Georgia.

Nymphon clarencei, n.sp. (Figs. 21, 22 and 23 b).

St. 160. 7. ii. 27. Near Shag Rocks, 53° 43' 40" S, 40° 57' 00" W, 177 m.; gy. M. R. St. Large
dredge: i ovigerous S-

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' 00" W, 342 m.; R.
Large dredge: i 9, i c? (holotype), i ovigerous ^ with encrusting Polyzoa and i larvigerous S-

St. WS 27. 19. xii. 26. 53° 55' S, 38° 01' W, I metre horizontal tow-net, 106-9 m. (net touched
bottom): i <?, i ?, i immature specimen.

Description of holotype. Trunk with lateral processes separated by considerable
intervals (one-third of, to approximately their own diameter).
Neck rather short, posterior half bearing the large base of the
oviger; ocular tubercle as wide as high; eyes conspicuous
(Fig. 21).

Proboscis short, stout, sub-cylindrical; three-fourths of ce-
phalic segment, subequal to scape.

Abdomen short, reaching to distal end of fourth lateral pro-
cess ; elevated at an angle of about 30°.

Chelophore. Scape gradually widening distally ; length three
and a half times the maximum width. Chela equal to (or
slightly longer than) scape ; fingers slightly longer than palm,
armed with thirty-two and thirty-seven spinules respectively,
setose pad extending on to proximal fourth of immovable one
(Fig. 22 b).

Segments 2-5 oi palp in the proportions 7-66 : 8-33 : 4 : 6;
in another J 8 : 8 : 4 : 5 (Fig. 22 c).

Oviger. Terminal claw a little shorter than tenth segment
and armed with seven spinules. Total number of denticulate
spines 35(13 + 8 + 7 + 7). Segments 4-6 in the proportions ^

178 : 3-11 : i; fifth segment long, slender and somewhat with chelophores and palps:
clubbed distally; sixth segment slightly curved (Fig. 23 b). ><■ 8.

Third leg. Second coxa long, twice the sum of the other two. Femur slightly shorter
than first tibia, 10-11 gland openings, flush with surface, on the proximal mid-ventral
surface. Second tibia the longest segment, approximately half as long again as femur.
Tarsus two-thirds of propodus, 6-9 long spines on the ventral margin of the latter.
Claw nearly half of propodus ; auxiliaries well developed (Fig. 22 a).

Paratypes. The males all agree very closely with the holotype. In the female the
coxae, femur and first tibia of the walking legs are more robust ; the second coxa is
considerably shorter, less than half as long again as the sum of the other two

1 1-6 : 2-8 : I in the male from St. WS 27.

Fig. 21 . Nymphon clarencei,
n.sp. Dorsal view of body

NYMPHONIDAE 55

(1-24-1 -40 : i); segments 4-6 of the oviger are relatively short, hi the proportions
1-31 :i75:i.

Fig. 22. Nymphon claiencei,n.sp.: «. Terminal segments of third leg: >c 27.
b. Chela: x 27. c. Palp: x 20.

Measurements (mm.)

Holotype

Ovig. cJ

?

St. 170

St. 170

St. 170

Length of proboscis

2-4

2-2

2-4

Diameter of proboscis

1-2

1-2

1-2

Length of trunk

6'8

6-4

6-4

Length of cephalic segment

3-4

3-2

3-2

Width of anterior cephalic lobes

1-8

1-8

2-13

Width of neck

0-8

0-9

0-8

Width across second lateral processes

4-0

4-2

4-0

Length of abdomen ...

1-05

I-O

1-2

Third leg:

First coxa ...

1-2

1-2

1-3

Second coxa

4-0

4-0

2-6

Third coxa ...

0-8

0-8

0-8

Femur

8-4

7-8

8-0

First tibia

9-0

8-65

8-f)

Second tibia

13-2

II-4

II-6

Tarsus

17

1-4

1-6

Propodus

2-5

2-25

2-5

Claw

1-05

i-o

1-3

Auxiliaries

0-4

0-4

0-4

No. of denticulate spines on oviger..

35

40

37

Approximate no. of spinules on chela

32 and 37

30 and 35

35 and 43

S6

DISCOVERY REPORTS

Remarks. This species bears a strong superficial resemblance to N. hiemole, Hodgson,
but in addition to being a more robust form,
it differs from the latter in the following
respects: (i) the third palpal segment is equal
to or a little longer than the second, (2) the
tarsus never exceeds two-thirds of the pro-
podus, (3) the fifth segment of the male
oviger, though long, is expanded distally and
is very similar to that of A^. brachyrhynchum,
Hoek (Gordon, 1932, fig. 40). It is easily
distinguished from the latter as follows:
(i) the chela is subequal to, instead of half
as long again as, the scape; the fingers are
not bent distally and are armed with fewer
spinules (cf. Fig. 22 b with Gordon, 1932,
fig. 3); (2) the relative proportions of palpal
segments 2-5 differ markedly in the two
species (e.g. 7-66 : 8-33 : 4 : 6 and in A^.
brachyrhynchum 8 : 11 : 12 : 8); (3) the rela-
tive proportions ^ of segments 4-6 of the male
oviger are different.

Fig. 23. Segments 4-6 of male oviger of:
a. Nymphoii artkulare, Hodgson (sixth seg-
ment slightly twisted): >; 36. b. N. clarencei,
n.sp.: X 10.

Nymphon, sp.? (Fig. 24 e,f).

St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 64° 20' S, 63° 01' W, 160-335 m.;
M. Large otter trawl: 1 ^, incomplete.

Description of specimen. Trtuik slender, smooth and of very loose build ; lateral
processes separated by 2-3 times their own diameter. Neck long and slender; ocular
tubercle low, as wide as high ; eyes conspicuous.

Proboscis long, rather slender, slightly expanded in middle and again near the distal
end ; approximately two-thirds of cephalon and a little shorter than scape.

Abdomen very short, reaching only to about the middle of fourth lateral process;
elevated at an angle of about 45''.

Chelophore long and slender. Scape approximately seven times as long as wide
distally, bearing a very few short setae especially near the distal articulation. Chela
rather longer than scape (5 : 4) and slightly curved; fingers half as long again as palm,
each armed with about 50-52 crowded and uneven spinules (every 2nd-4th spinule
long, the intermediate ones of various lengths). Setose pad scarcely developed

(Fig. 24/).

Palp long and slender, segments 2-5 subequal in the proportions 4-25 : 4-5 : 4-5 : 475 ;
setae very minute.

1 In A', brachyrhynchum 2-5:4:1 and 2-33 : 3-66 : i.

NYMPHONIDAE

57

Oviger long and slender. Terminal claw three-fifths of tenth segment and armed with
7-8 spinules. Total number of denticulate spines 30 (12 + 7 + 5 + 6). Segments 4-6
in the proportions 1-33 : 2 : i ; segment 6 long and slightly curved (Fig. 24 e).

Leg. Only the two anterior pairs of legs are complete. The second coxa is long,
1-66 and 2 times the sum of the first and third (see measurements). The femur is shorter
than either tibia, with a row of about twenty-four very low, crowded gland tubercles
on the proximal two-thirds. Second tibia the longest segment. Tarsus propodus and
claw elongated and slender ; claw equal to or a little shorter than the propodus which
is approximately half of the tarsus. Auxiliaries absent.

a

b

Fig. 24. NympJwn subtile, Loman: b. Terminal segments of third leg.
M'Wj&/;o« sp.? St. 53: rt. Terminal segments of third leg: x 47. c. Palp: x 60. (/.Chela: ;< 47.
Nymphon sp..' St. 181 : e. Segments 2-6 of male oviger: ;< 16. /. Chela: x 16.

Measurements (mm.)

Length of proboscis ...
Diameter of proboscis
Length of trunk
Length of cephalic segment ...
Width of anterior cephalic lobes

Width of neck

Width across second lateral processes
Length of abdomen ...

Walking leg:

2-4

First coxa .

0-8

Second coxa

6-93

Third coxa .

3-4

Femur

1-22

First tibia .

0-43

Second tibia

3-6

Tarsus

0-65

Propodus .

Claw

(ist)

(2nd)

1-2

1-2

3-0

3-6

0-6

0-6

8-8

0-4

3-0

4-2

373

2-1

2-0

1-87

2-0

S8

DISCOVERY REPORTS

Remarks. This species belongs to group I, but the oviger is somewhat transitional
between type I b and type I since the fifth segment is gradually expanded towards the
distal end. On the whole, the specimen is more nearly related to A^. hngicoxa and
N. homohmi than to A^. proceroides and A-", tetiuipes (forms in which the auxiliary claws
are absent). The chela is long and slender, and the sixth segment of the oviger is curved,
as in the two first-named forms, but Nymphon sp. ? may be easily distinguished from both
by the palp which is very similar to that of A^. chorcoti; the second coxa and the tarsus
are both relatively much longer than in that species (cf. also the ovigers. Figs. lo b
and 24 e). The tarsus is longer than in any other species in group I.

Nymphon, sp.? (Fig. 24 a, c and d).

St. 53. 12. V. 26. Port Stanley, East Falkland Island. Hulk of 'Great Britain': i ?.

Description. Trunk rather compact; lateral processes separated by approximately
their own diameter. Neck short ; oviger base large, in contact with first lateral process.
Ocular tubercle rather higher than wide; eyes conspicuous. Setae absent.

Proboscis short, stout, sub-cylindrical, rounded at apex; two-thirds of cephalic
segment, and three-fourths of scape.

Abdomen reaching a short distance beyond fourth lateral process, elevated at an
angle of 45°.

Chelophore. Scape four times as long as wide distally, bearing a few very fine setae.
Chela subequal to scape ; movable finger subequal to palm armed with eleven or twelve
spinules; 15 spinules on immovable finger (Fig. 24 d).

Palp as represented in Fig. 24 c, two terminal segments together approximately four-
fifths of second segment.

Oviger. Terminal claw rather more than half of tenth segment, armed with about
seven spinules. Total number of denticulate spines 37 (11 + 10 + 7 + 9). Segments
4-6 short, in the proportions 1-5:2:1.

Third leg. Second coxa i -4 times the sum of the first and third. Femur nine times as
long as wide, subequal to first tibia. Second tibia the longest segment. Tarsus seven-
tenths of propodus, ventral margin of the latter spinose (Fig. 24 a). Claw rather more
than one-third of propodus, auxiliaries long, two-thirds of main claw. Setae few in
number and very fine. There are long irregular patches of a light brownish or purplish
colour on each of the three main segments.

Measurements {mm.)

Length of proboscis ...

Diameter of proboscis

Length of trunk

Length of cephalic segment ...

Width of anterior cephalic lobes

Width of neck

Width across second lateral processes

Length of abdomen ...

Third leg:

0-6

First coxa

0-35

0'33

Second coxa

i-i

1-8

Third coxa

0-42

0-9

Femur

3-0

0-58

First tibia

3-13

0-3

Second tibia

3-9

1-2

Tarsus

077

0-35

Propodus

i-oS

Claw

0-4

Auxiliaries

0-37

NYMPHONIDAE 59

Remarks. Although it is extremely small, this specimen is a mature female with
developing ova in the femur of each leg. The male is unknown, so that the affinities
of the species are uncertain, but it would appear to be allied to A'^. subtile, Loman.
It differs from the latter, however, in several respects: (i) the body is more compactly
built with a much shorter neck, (2) the legs are much longer relative to the size of the
specimen, being 7-6 times as long as the trunk (as against 5-4 and 4-6 in the male and
female respectively of A^. subtile — see measurements), (3) the tarsus is longer relative to
the propodus and the ventral margin of the latter is armed with five spines, (4) the two
terminal palpal segments are shorter, and (5) the legs are characterized by the presence
of long bands of a darkish colour.

This species shows affinities with some of the northern forms, e.g. N. brevitarse,
Kroyer (Sars, 1891, pi. v, fig. 3).

B. GROUP II

Nymphon australe, Hodgson (Figs. 25 d and zbb).

N. australe, Hodgson, 1902, p. 257, pi. xl.

Chaetonymphon altiuculum, Mobius, 1902, p. 181, pi. xxvi, figs. 1-6.

Ch. australe, Hodgson, 1907, p. 32, pi. x, fig. 14.

Ch. australe var. austrinorum, Hodgson, 1907, p. 35, pi. iv, fig. 4; pi. x, fig. 15.

Ch. assimile, Hodgson, 1908, p. 175, pi. i, figs, i, i a.

N. stylops, Bouvier, 1911, p. 1137; 1913, p. 73, text-figs. 25-31.

A'^. australe. Caiman, 1915, p. 36.

Ch. australe var. austrinorum, Loman, 1923, p. 21.

St. WS 32. 2i.xii. 26. Mouth of Drygalski Fjord, South Georgia, 91-225 m.; gy. M. Small
beam trawl: many specimens, including ovigerous and larvigerous i^^.

St. 42. i.iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of
Jason Light to 4 miles N 39° E of Jason Light, 120-204 m.; M. Large otter trawl: 20 specimens,
some brown, i with Polyzoa.

St. 45. 6. iv. 26. 2-7 miles S 85° E of Jason Light, South Georgia, 238-270 m.; gy. M. Large
otter trawl: hundreds of specimens, many with encrusting Polyzoa.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, 230-250 m.; gy. M. Large
otter trawl: 5 specimens, well chitinized, bearing Polyzoa; several smaller, very soft specimens.

St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m.; gn. M. St.
Large otter trawl: 12 specimens, several soft.

St. 144. 5.1.27. Off mouth of Stromness Harbour, South Georgia, 155-178 m.;gn. M. S. Large
otter trawl: many specimens, adult and immature.

St. 159. 21. i. 27. 53° 52' 30" S, 36° 08' W, 160 m.; R. Large dredge: 4 specimens.

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 342 m.; R. Large dredge: many specimens,
all very setose, including ovigerous and larvigerous <J^.

St. 172. 26. ii. 27. Off Deception Island, South Shetlands, 525 m.;R. Large dredge : 2 specimens
with N. bouvieri, n.sp.

St. 175. 2.iii. 27. Bransfield Strait, South Shetlands, 63° 17' 20" S, 59° 48' 15" W, 200 m.;
M. St. G. Large dredge: i specimen.

8-2

6o DISCOVERY REPORTS

St. i8o. 2. iii. 27. 17 miles W of N point of Gand Island, Schollaert Channel, Palmer Archi-
pelago, 160-330 m.; M. St. Large otter trawl: several specimens with Colosseiideis frigida, N. brevi-
caiidalum and Pallcnopsis pilosa.

St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 160-3350-1.; M. Large otter
trawl : numerous specimens.

St. 182. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 278-50001.; M. Large otter
trawl : several specimens ; i $ with whole dorsal surface covered by an encrusting Polyzoon.

St. 186. 16. iii. 27. Foamier Bay, Anvers Island, Palmer Archipelago, 295 m.; M. Large
dredge : 6 specimens.

St. 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 259 m.; M. Large dredge:
numerous specimens, the majority with body and legs banded with dark brown.

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Georgia, 391 m.; M. St. Large
dredge: 2 specimens with A^. charcoti and A^. biarticulatum.

St. 366. 6. iii. 30. 4 cables S of Cook Island, South Sandwich Islands, 155-322 m. Large dredge :
many specimens, including SS carrying eggs and advanced larvae.

St. 371. 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large otter
trawl: i ?, i immature specimen.

St. 456. 18. X. 30. I mile E of Bouvet Island, 40-45 m. Large dredge: i ?.

St. 458. 19. X. 30. 7 miles S 50° W of Cape Circumcision, Bouvet Island, 357-377 m. Large
dredge : many specimens including larvigerous ^S-

Description of male.^ Tnmk, as a rule, compact; lateral processes separated by a
distance varying from one-fourth of, to equality with, their own diameter. Neck short,
as a rule, but longer in some specimens than in others ; cephalic segment nearly as
wide anteriorly as long; base of oviger in contact with first lateral process. Ocular
tubercle high (2-5-3 times as high as wide), eyes sub-terminal. A few setae on each
lateral process and cephalic lobe.

Proboscis sub-cylindrical, approximately equal to cephalic segment and three-fourths

of scape.

Abdomen reaching to distal articulation of first coxa, pyriform, bearing a few slender

setae.

Chelophore. Scape 4-5 times as long as wide distally, setose (Bouvier, 1913, p. 76,
fig. 27). Fingers longer than palm, each armed with 38-46 crowded spinules ; setose
pad small, on proximal third of immovable finger.

Palp. For proportions of segments 2-5 see Table VI and Fig. 25 d.

Oviger as represented in Fig. zbb; terminal claw three-fourths of tenth segment,
armed with 7-9 spinules. Total number of denticulate spines 23-37. Segments 4-6 in
the proportions 1-36 : 2 : i ; segment 5 much inflated in distal half and the thin walled
portion {x) is often collapsed in fixed material; segment 6 also swollen and thin walled,
often caved in at x.

Third leg rather short and setose. Second coxa slightly longer than the sum of the
first and third. Femur shorter than either tibia ; numerous (16-18) small, obscure gland

1 This species has been described at considerable length by various authors (see synonymy), but a very
brief description of the male is given here for the sake of uniformity.

NYMPHONIDAE

6i

openings on proximal two-thirds of mid-ventral surface. First tibia the longest segment.
Tarsus and propodus both elongated, the former considerably longer than the latter.
Claw approximately half as long as propodus, auxiliaries vestigial.

The female has the oviger of the normal female type ; the femur is rather swollen
ventrally except in the distal third, the length being 3-5-4-5 times the maximum width.

Remarks. A re-examination of Hodgson's type specimens of A'^. assirnile left no doubt
as to their identity with A^. mistrale.

b

Fig. 25. Palp of: a. NympJion bouvieri, n.sp. b. N. brevicaudatum, Miers.
c. N. neumayri, n.sp. ci. N. australe, Hodgson. (All x 27.)

The Discovery specimens from St. 170 agree with the co-types of A^. australe; the
lateral processes are but little separated, the legs are relatively stout and bear numerous
long setae. The majority of the specimens, however, have the legs more slender and less
setose, the body less compact, and thus agree more closely with A^. australe var.
aiistriiiorum, Hodgson. Many of the specimens, especially those from St. WS 32 to
St. 144 (in the list of stations) are even less setose than the atistrinormn forms.

Loman (1923, p. 13) has emphasized the importance of the number of denticulate
spines on the four terminal segments of the oviger in the genus Nymphon. In nine
co-types of A^. australe examined the total number of denticulate spines varied from
twenty-three to thirty, and in six co-types of A^. australe var. austriiwnan^ from thirty
to thirty-seven. The greater number in the variety is probably due to the fact that the
specimens are of considerably larger size than the typical forms. At any rate, the
number varied from twenty-eight to thirty-four in all samples of the Discovery material

' Hodgson (1907, p. 35) states that in N. australe var. ausirinorum the "tarsus and propodus together are
distinctly shorter than the femur" and equal to it in typical specimens; but in the co-types measured, tarsus
and propodus together are nearly equal to the femur in A'^. australe as well as in the variety austrinorum.

62 DISCOVERY REPORTS

where the austrinorum forms are smaller, and the typical specimens (St. 170) somewhat
larger, than Hodgson's co-types.

Table VI

1

Relative proportions*

of palpal

Total length

Name of species in group II

segments 2-5

of segments

2

3

4

5

4 + 5

A^. capense

10 :

9-00

6-00

: 5-00

11-00

N. cotnpactiim

10

970

6-o8

5-80

11-88

N. compactum

10

9-10

5-46

5-46

10-92

N. aiistralc, co-type

10

9-20

5-60

4-80

10-40

A^. australe, co-type

10

9-30

5-56

4-44

10-00

A'^. australe, St. 172

10

7-00

5-00

4-50

9-50

A'^. australe var."!

, • [ co-types
austrinorum) ■' '^

10

ID

8-6o

8-20

6-00
5-30

570
470

11-70

10-00

N. orcadense, co-type

10

8-00

4-80

4-80

9-60

A^. orcadense, co-type

10

7-90

5-40

5-20

io-6o

A^. articulare

10

6-25

375

4-05

7-80

N. neumayri, holotype

10

7-80

4-80

3-00

7-80

A^. neumayri, paratype

10

6-66

4'io

3-33

7-43

A'', villosum, holotype

10

7-30

3-O0

370

6-70

A'', brevicaudatum, paratype

10

6-10

2-12

2-27

4-39

A'^. brevicaudatum, St. WS 42

10

6-66

2-00

2-00

4-00

A^. brevicaudatum, St. 159

10

6-10

2-6l

2-82

5-43

A'^. biarticuhitum, holotype

10

7-25

2 -GO

2-25

4-25

A'', biarticulatum, St. 182

ID

7-40

2-22

2-22

4'44

A^. biarticulatum, St. igo

10

6-66

2-00

2-20

4-20

N.proximum, holotype

10

6-66

2-II

2-II

4-22

A'^. mendosum, holotype

10

6-10

2-17

174

3-91

A'^. mendosum, co-type

10

5-60

2-00

1-50

3-50

A^. mendosum, T.N. 34of

10

6-00

2-22

2-20

4-20

A'^. ? biarticulatum, St. 167

10

6-86

1-86

171

3-57

A'^. .' biarticulatum

10

8-86

1-77

177

3-54

N. .' biarticulatum

10

6-66

1-83

1-50

3-33

N. bouvieri, paratype

10

5-5

1-82

0-91

273

N. bouvieri

10

6-40

1-82

I -00

2-82

N. bouvieri

10

6-00

I -60

I -00

2-6o

* In order the more readily to compare these proportions, the length of the second segment

is regarded as = 10.
f T.N. = Terra Nova collection.

N. aiistrole differs from the majority of the species in group II in two respects:
(i) the auxiliary claws are vestigial, and (2) the oviger in the male shows a considerable
departure from the norm, certain portions being much inflated and thin walled (cf.
Fig. 26 b with Figs. 26 a and c, and 27). Two other Antarctic species, not represented
in the Discovery collection, were also found to have certain segments inflated — viz.
A'', capense, Hodgson, and N. compoctiim, Hoek (see Gordon, 1932, pp. 1 15-120). In
both these species the auxiliaries are absent. The inflated condition is least developed
in N. compactum and most pronounced in N. australe. In all three species segment 5
is relatively short and stout so that they may be regarded as aberrant members of
group II. N. australe shows affinities with N. orcadense (see Table III), for the palp

NYMPHONIDAE

63

and chela are very similar in both and there is approximately the same number of
denticulate spines on the oviger.

Fig. 26. Male oviger of: a. Nymphon orcadense, Hodgson: X17. b. N. attstrale,

Hodgson: x 17. Type H a; segments 5 and 6 thin-walled and inflated at x.

c. Segments 5-7 of male oviger of A', brevicaudaium, Miers : < 27.

Distribution. Circumpolar ; this is the first record of the typical australe form from
the western side of the Antarctic (cf. Loman, 1923, p. 21).

Nymphon orcadense, Hodgson (Fig. 26 a).

Chaetonymphon orcadense, Hodgson, 1908, p. 173, pi. ii, figs. 2, 2 a.
Chaetonymphon orcadense, Loman, 1923, p. 22.
Nymphon orcadense. Caiman, 1920, p. 246.

St. 163. 17. ii. 27. Paul Harbour, Signy Island, South Orkneys, iS-27 m. Small beam trawl:
18 specimens, including ovigerous ^^, several with encrusting Polyzoa.

St. 164. 18. ii. 27. E end of Normanna Strait, South Orkneys, near Cape Hansen, Coronation
Island, 24-36 m. Small beam trawl: many specimens both adult and immature; several with en-
crusting Polyzoa.

Description of male. Trunk compact, lateral processes in contact proximally or
separated by very narrow intervals. Neck short; oviger base, as usual in the short-
necked forms, occupying the gap between anterior cephalic lobe and first lateral process.
Anterior width of cephalon seven-eighths of length. Height of ocular tubercle nearly
twice the width, which increases somewhat towards the apex where the eyes are situated.
A few setae near the distal articulation of each lateral process and on each anterior
cephalic lobe.

Proboscis equal to cephalic segment and a little shorter than scape, greatly expanded
in the middle.

64

DISCOVERY REPORTS

Abdomen rather short, reaching to middle of first coxa; bifid; not elevated.

CheJophore. Scape, which bears several long and numerous short, fine setae, nearly
four times as long as wide distally. Fingers of chela slightly longer than palm, each
armed with 35-45 spinules according to the size of the specimen; setose pad extending
half-way along immovable finger.

Palp. Two terminal segments together approximately equal to, and third four-fifths
of, second segment (Table VI).

Oviger. Terminal claw half as long as tenth segment, armed with 10-13 spinules;
total number of denticulate spines 30-36. Segments 4-6 in the proportions i-68 : 2-22 : i,
segment 5 short, rather curved and distinctly clubbed distally. The termino-lateral

C

Fig. 27. Segments 4-6 of male oviger of: a. Nymphon proximum. Caiman.

b. N. bouvieri, n.sp. c. N. mendosuni, Hodgson, d. N. biarticulatum, Hodgson.

e. N. neumayri, n.sp. (Type H — all x 20.)

species

Femur

articulation of the seventh on the sixth segment is characteristic of this
(Fig. 26 a).

Third leg. Second coxa a little longer, dorsally, than the sum of the other two.
shorter than either tibia, at least thirty crowded, small gland openings along mid-ventral
surface ; second tibia slightly longer than first. Tarsus and propodus both elongated and
armed ventrally with numerous fine setae ; the former always slightly longer than the
latter (1-05-1 -31 : i). Claw nearly half as long as propodus as a rule; auxiliaries from
one-fourth to one-third of main claw.

The legs are robust and setose; in addition to the fine, short setae, there are long
setae at the distal articulation of (i) the third coxa— ventrally ; (2) the femur— dorsally ;
(3) each tibia— ventrally ; and a number of strong setae along the lateral margins of
the latter.

NYMPHONIDAE 65

The female does not differ markedly from the male. The legs, especially as regards
the femur, are more robust and less setose. The oviger is of the usual type with the
seventh normally articulated to the sixth segment.

Remarks. The specimens in the Discovery collection agree well with the co-types
from Scotia Bay.

This species undoubtedly belongs to group II and yet it has (i) a rather shorter
abdomen, (2) a longer second tibia, (3) more numerous spinules^ on the fingers of the
chela, and (4) a higher number of denticulate spines on the oviger than any of the other
typical species (see Table III). In all these respects it approaches N. clarenci, n.sp., and
A'^. brachyrhynchiim, Hoek, especially the former in which the palp is of the same type.
Also the atypical^ species in group II are more closely allied to A^. orcadense than to any
of the other typical species.

Distribution. Previously recorded from Scotia Bay; Burdwood Bank, south of
Falkland Islands and South Georgia.

Nymphon articulare, Hodgson (Figs. 23 a, 28 and 29 a, b).

Hodgson, 1908, pp. 170-172, pi. i, figs. 4, 4 a.
Bouvier, 1913, p. 72 (in key).
Loman, 1923, p. 13 (in table).

St. 164. 18. ii. 27. E end of Normanna Strait, South Orkneys, near Cape Hansen, Coronation
Island, 24-36 m. Small beam trawl: i ?, i (^ and several immature specimens.

Description of male topotype.^ Trunk non-setose, rather compact ; lateral processes
separated by half of their own diameter or rather more. Neck very short, the posterior
two-thirds on either side occupied by the base of the oviger ; cephalic segment almost
as wide anteriorly as long. Ocular tubercle high and rather wide ; eyes small and set
near to the bifid apex.

Proboscis a little shorter than scape and equal to cephalic segment; stout, sub-
cyHndrical, contracted somewhat at base.

Abdomen pyriform, reaching to distal end of first coxa ; elevated at an angle of about 10°.

Chelophore. Scape rather longer than chela, with a few long, fine setae along the
dorsal surface; length 4-5 times the
greatest width. Chela as represented
in Fig. 28, fingers rather longer than
palm and each armed with short, ir-
regular, often recurved, spinules. The t — ^ — ^ — ^^^ \ vW '

number of spinules varies from eigh- pjg.^S. Nymphon articulare,\\oAg^orx. Chela of co-type:
teen to twenty-five in the five co-types, x 47. The palm may not be quite accurate as proximal
and there are 2-6 more on the movable P^'^ ^^^ damaged on slide.
than on the immovable finger. Setose pad extending half-way along the latter.

1 A'^. hrevicaudatum and A'^. biarticulatum have approximately the same number as .V. orcadense.

^ Especially A^ australe, see p. 62; and also Table VI.

^ Hodgson apparently did not examine this specimen, see p. 66.

66 DISCOVERY REPORTS

Palpal segments 2-5 in the proportions 4 : 2-5 : 1-5 : 1-62.

Oviger. Terminal claw half as long as tenth segment, armed with 7-10 small splnules.
Total number of denticulate spines 25 (7 + 6 + 5 + 7). Segments 4-6 in the pro-
portions 1-85 : 2-54 : I ; segment 5 short and clubbed distally (Fig. 24 a).

Third leg rather slender and setose ; a number of long, fine setae on the coxae and at
the distal end of the femur (maximum length equal to the diameter of the segment) ;
those on the distal segments sparse and very short. Coxae short and stout, dorsal length
of second equal to the sum of the first and third. Femur subequal to second, first
tibia the longest segment. Tarsus and propodus both elongated, the former nine-tenths
of the latter. Claw not quite half as long as propodus ; auxiliaries well developed, one-
third of main claw.

Measurements {mm.)

c?

?

Length of proboscis

I -08

I-I2

Diameter of proboscis

0-56

0-6

Length of trunk

2-4

2-6

Length of cephalic segment ...

I -06

I-I2

Width of cephalon

0-96

I-O

Width of neck

0-56

0-48

Width across second lateral processes

2-0

2-0

Length of abdomen

0-88

0-84

Length of scape

1-32

1-48

Third leg:

First coxa ...

0-6

0-6

Second coxa

0-9

0-8

Third coxa

0-3

0-25

Femur

1-8

1-8

First tibia

2-1

2 -08

Second tibia

1-9

175

Tarsus

0-9

0-9

Propodus

I -OS

i-o

Claw

0-44

0-48

Auxiliaries

o-i6

0'i6

The female is characterized by having the femur much enlarged in the proximal
two-thirds. The length is only 2-5 times the maximum width as against 4-5 times in the
male (Fig. 29 a and b).

Remarks. Hodgson (1908, p. 172) mentions only three adult females in his descrip-
tion, but five specimens, including one male and all undoubtedly belonging to this
species, were obtained on loan from the Royal Scottish Museum, Edinburgh. Hodgson
apparently did not have two of these specimens and referred the species to the genus
"A(ym/)/;o« ", while in reality it belongs to the " Chaetonymphon " group as is shown by the
form of the male oviger (Fig. 23 a). N. neumayri, n.sp., is closely related to A'^. articulare ;
the most outstanding differences between the two are listed on p. 69.

Distribution. South Orkneys.

Nymphon neumayri, n.sp. (Figs. 27 e, 29 c-e and 30).

St. 187. i8. iii. 27. Neumayr Channel, Palmer Archipelago, 64° 48' 30" S, 63° 31' 30" W, 259 m. ;
M. Large dredge : 3 ^^, 3 ?$ and 3 immature specimens, with A^. australe.

NYMPHONIDAE

67

'/F::=3

Description of holotype {^). Trunk compact, last two pairs of lateral processes in
contact proximally, the others separated by very short intervals. Cephalic segment as
w^ide anteriorly as long ; neck very short, base of
oviger occupying most of the space between
anterior cephalic lobe and first lateral process.
Ocular tubercle high (o-8 mm.), very wide at base
(o-6 mm.), tapering to about half that width at
the apex on which the four eyes are set. Setae
almost entirely absent (Fig. 30).

Proboscis stout, sub-cylindrical, rounded at
apex, somewhat contracted at base; slightly
shorter than either scape or cephalic segment.

Abdomen horizontal, fusiform, reaching to
middle of second coxa.

Scape of chelophore nearly four times as long
as wide, with a few very minute setae at the
distal end. Chela shorter than scape (see measure-
ments), palm short, ovoid, fingers half as long
again as palm, much curved, each armed with
seven or eight long spinules ; a few fine setae on
proximal half of immovable finger (Fig. 29 d).

Palp with two terminal segments together ap-
proximately equal to third, and three-fourths of
second segment (segments 2-5 in the proportions
10 : 7-7: 4-4: 3-2 and in female as 10 :']■% :4-8 13).

Oviger. Terminal claw subequal to tenth seg-
ment and armed with 4-5 spinules. Total number
of denticulate spines, 26.^ Segments 4-6 as
represented in Fig. 27 e, and in the proportions
1-82 :2-55 : i.

Third leg rather stout proximally, and tapering
gradually from the distal articulation of the
first tibia where the width is suddenly reduced to almost one-half (i.e. proximal
width of second is less than two-thirds distal width of first tibia, see Fig. 30 c, $).
Second coxa equal to the sum of the other two (dorsal measurements). Femur shorter
than either tibia, three times as long as wide and slightly inflated ventrally (Fig. 29 c) ;
no raised femoral gland tubercles are present and, indeed, no trace of gland openings
has been observed. First tibia the longest segment. Tarsus subequal to propodus ; claw
approximately half the length of the latter ; auxiliary claws very short, about one-fifth
of claw.

Fig. 29. Nymphon articularc, Hodgson: a.
Femurof third leg of female, b. Same of male.
Nymphon neumayri, n.sp. : c. Femur of male.
d. Chela, e. Terminal segments of third leg.
{a-c: X 20; (/ and e: x 27.)'

1 The number varies from 22-27 '" adults and the following are typical formulae 6 + 6 + 4 +

7 + 7 + 5 + 7;8 + 6 + 5 + 7;8 + 7-f-5 + 7;inan immature specimen 3 + 4 + 2 + 5.

9-2

6;

68

DISCOVERY REPORTS

Setae are small and few in number ; there are 4-6 spinose setae near the distal articula-
tion of the femur (Fig. 29 c) and a large spine, on the mid-ventral line, at the distal
end of the second tibia (Fig. 29 e).

The female shows the usual sexual differences of oviger and legs ; the femur (Fig. 30 c)
is much enlarged throughout three-fourths of its length; the genital pores are very
conspicuous.

Fig. 30. Nymphon neumayri, n.sp. : a. Holotype. Dorsal view of body with chelophores and palps.
b. Holotype. Lateral view of body, with chelophore and palp. c. Third leg of female.

Measurements {mm.)

Length of proboscis . . .

Diameter of proboscis

Length of trunk

Length of cephalic segment

Width of cephalic lobes

Width across second lateral processes

Length of abdomen

Length of scape

Length of chela

Third right leg:

First coxa ...

Second coxa

Third coxa ...

Femur

First tibia ...

Second tibia

Tarsus

Propodus ...

Claw

Holotype

?

1-4

1-8

i-o

0-9

4-0

4-4

1-8

2-0

1-8

2-3

3-3

3-4

2-1

2'0

2-2

2-2

1-6

1-6

0-8

o-S

1-2

1-2

0-4

0-4

3-01

3-2

4-2

4-4

3-6

3-8

1-8

1-9

1-75

1-8

0-8

0-9

NYMPHONIDAE 69

Remarks. This species is much less setose than is typical of the short-necked, com-
pactly built forms, there being no long setae on the body and very few on the legs. The
chela differs from that of all the allied forms ; it recalls that of A'', proceroides, Bouvier
(cf. Figs. 29 d and 9), but the fingers are much curved and the spinules are longer and
fewer in number.

A^. ueumoyri is most closely allied to A^. articulare, Hodgson, which also has the femur
in the female much enlarged throughout part of its length (Fig. 29 a). But it may be
readily distinguished from the latter by (i) its larger size and more compact build,
(2) the widely gaping fingers of the chela each armed with 6-9 instead of 18-25 spinules,
and (3) the marked difference in the outline of the femur in the male (cf. Fig. 29 b and c).

Nymphon brevicaudatum, Miers (Figs. 25 b, 26 c and 31 b, d).

N. brevicaudatum, Miers, 1875, p. 117.
N. brevicaudatum, Miers, 1879, p. 212, pi. xi, fig. 8.
Chaetonymphon brevicaudatum, Mobius, 1902, p. iSi, pi. xxv, fig. 9.
Ch. brevicaudatum, Loman, 1923, p. 21.

St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, no m.; M. R. Large dredge: 3 adult
and 4 immature specimens.

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, 179-235 m.; gy. M. Large otter trawl :
12 specimens.

St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, 120-204 m.; M. Large otter
trawl: many specimens, one carrying a Brachiopod.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, 230-250 m.; gy. M. Large
otter trawl: 12 specimens, i carrying a sponge.

St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m.; gn. M. St.
Large otter trawl: many specimens.

St. 141. 29. xii. 26. East Cumberland Bay, South Georgia, 200 yards from shore, under Mount
Duse, 17-27 m.; M. Small beam trawl: many specimens, several of the (JcJ carrying eggs or larvae,
two stages on one specimen.

St. 143. 30. xii. 26. Off mouth of East Cumberland Bay, South Georgia, 273 m.; M. Large
otter trawl: i larvigerous (?.

St. 144. 5.1.27. Off mouth of Stromness Harbour, South Georgia, 155-178 m.; gn. M. S. Large
otter trawl : several specimens.

St. 145. 7. i. 27. Stromness Harbour, South Georgia, between Grass Island and Tonsberg Point,
26-35 n^- Small beam trawl: several specimens.

St. 148. 9. i. 27. Off Cape Saunders, South Georgia, 132-148 m.; gy. M. St. Large otter trawl:
a dozen specimens, several ovigerous ,3^.

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, 200-234 m.; M. Large
otter trawl: 2 specimens.

St. 156. 20.1.27. 53° 51' S, 36° 21' 30" W, 200-236 m.; R. Large dredge : 5 specimens.

St. 159. 21. i. 27. 53° 52' 30" S, 36" 08' W, 160 m.; R. Large dredge: many specimens.

St. 170. 23. ii. 27. Off" Cape Bowles, Clarence Island, 342 m.; R. Large dredge: 3 specimens.

St. 175. 2.111.27. Bransfield Strait, South Shetlands, 63° 17' 20" S, 59° 48' 15" W, 200 m.;
M. St. G. Large dredge: 2 specimens with N. australe.

70 DISCOVERY REPORTS

St. 177. 5. iii. 27. 27 miles SW of Deception Island, South Shetlands, 1080 m. ; M. cS. St. Large
dredge: 2 specimens.

South Georgia:

St. WS 33. 21. xii. 26. 54" 59' S, 35° 24' W, 130 m.; gy. M. St. i m. horizontal tow-net:
I specimen.

St. WS 42. 7. i. 27. 54° 41' 45" S, 36^ 47' W, 198 m.; bottom, i m. horizontal tow-net:
I specimen.

St. MS 10. 14. ii. 25. East Cumberland Bay. J mile SE of Hope Point to I mile S of Government
Flagstaff, 18-26 m. Small beam trawl: 6 specimens.

St. MS 12. 17. ii. 25. East Cumberland Bay. i cable E to i mile S x E | E of Hobart Rock,
25-60 m. Small beam trawl: i specimen.

St. MS 25. 13. iv. 25. East Cumberland Bay. 4I cables NE to ij cables N x W of Hobart
Rock, 36 m. Small beam trawl: 2 specimens.

St. MS 33. I . V. 25. I cable E of Hobart Rock, East Cumberland Bay, 40 m. Small beam trawl :

1 specimen.

St. MS 63. 24. ii. 26. East Cumberland Bay. 1-3 miles S >: E to i-6 miles SE x S of Hope
Point, 23 m. Small beam trawl: 3 specimens.

Redescription of holotype (ovigerous ^).^ Trunk compact; lateral processes
separated by less than half their own diameter, last two pairs in contact proximally.
Neck short but distinct ; space between anterior cephalic lobes and first lateral process
entirely occupied by the oviger base ; anterior width of cephalon five-sixths of the length.
Ocular tubercle of medium height, cylindrical and rather stout ; eyes sub-terminal.

There are 4-6 long, and a few shorter, setae on the mid-dorsal surface of segments i ,

2 and 3 respectively, and two long setae on each cephalic lobe.- The lateral processes
each bear a number of setae of which 2-4 near the distal articulation are longest.

Proboscis short, stout, sub-cylindrical, rounded at apex; subequal to cephalic segment
and considerably shorter than scape.

Abdomen horizontal, reaching a short distance beyond first coxa; sub-cylindrical,
tapering distally ; three setae, of which the median is longest, at base and a few tiny
setae on the distal half.

Chelophore rather similar to that of A^. biarticiilatum, Miers. Scape setose ;3 length
nearly four times the distal width. Fingers equal to palm, each armed with twenty-nine
and thirty-five crowded spinules of unequal lengths.

Palp rather slender ; proportions of segments 2-5 as represented in Fig. 26 b (see also
Table V).

Oviger. Terminal claw subequal to tenth segment, armed with five spinules. Total
number of denticulate spines eleven (see Table III). Segments 4-6 in the proportions
1-35 : 2-3 : I ; segment 5 clubbed distally (Fig. 27 c).

1 The most complete of the type specimens has been selected as holotype; B.M. collection.

2 Three or four in all the paratypes.

^ 3-4 long setae longitudinally on the mid-dorsal surface; a few finer, shorter setae on proximal half of
outer margin ; a semicircle of setae, increasing gradually in length towards the inner edge, near the distal
articulation.

NYMPHONIDAE

Third leg relatively short and setose ; the length of the setae on each tibia and at the
distal end of femur is equal to the diameter of the segment. Second coxa as long as the
sum of the first and third (dorsal measurements). Three or four high wide gland
tubercles on mid-ventral surface of femur, which is equal to second tibia ; first tibia the
longest segment (see Table III). Tarsus shorter than propodus; claw less than half as
long as the latter; auxiliaries well developed, one-third of main claw.

\N.xL:^^..^^XAA

7-7-7-7

/ / ' ' ^

Fig. 31. Terminal segments of third leg of:

Nymphon biarticulatum , Hodgson: a, adult; c, young from St. 181 (length = 4 mm.).
A', brevicaiidatiim, Miers: b, adult co-type; d, young from St. 42 (length = 4 mm.).

Remarks. The majority of the specimens in the Discovery collection come from the
neighbourhood of South Georgia. A few specimens were collected off Clarence Island
and the South Shetlands, but none from Palmer Archipelago, where the bulk of the
specimens referred to N. biartictilatiim, Hodgson, occurred. A'^. brevicaudatiim and
N. biartindatiim are very closely allied and the latter may prove to be a more southern
variety of the former (see p. 72).

Distribution. Kerguelen, South Georgia and South Shetlands.

Nymphon biarticulatum, Hodgson (Figs. 27^, 310, c and 32^).

ChaetonytnpJwn biarticulatum, Hodgson, 1907, p. 28, pi. iv, fig. 2; pi. x, fig. 12.

St. 167. 20. ii. 27. Off Signy Island, South Orkneys, 244-344 m.; gi^- ^- Large otter trawl:
6 specimens.

St. 180. II. iii. 27. 17 miles W of N point of Gand Island, Schollaert Chaimel, Palmer Archi-
pelago, 160 m.; M. St. Large dredge: 3 S specimens.

^2 DISCOVERY REPORTS

St. i8i. i2.iii. 27. Schollaert Channel, Palmer Archipelago, 160-335 m.; M. Large otter trawl:
20 specimens.

St. 182. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 278-500 m.; M. Large otter trawl :
several specimens.

St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 315 m.; M. R. Large dredge: 4 speci-
mens (i immature).

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Shetlands, 391 m.; M. St. 2 adult
and 2 immature specimens with Nymplwn cliarcoti and A', aiistrale.

Redescription of holotype (9). Tnmk compact; lateral processes separated by
approximately half their own diameter, last two pairs almost in contact proximally.
Neck short ; base of oviger occupies the entire space between first lateral process and
anterior cephalic lobe. Ocular tubercle three times as high as wide ; eyes small, situated
near the apex. There is a pair of long setae ^ on the mid-dorsal surface of segments 1-3
and on each cephalic lobe.

Proboscis slightly longer than cephalic segment, a little shorter than scape ; stout and
sub-cyhndrical.

Abdomen reaching to middle of second coxa ; wide in middle and tapering abruptly
towards the apex; setae very minute. Not elevated.

Chelophore. Scape setose, length 4-5 times the distal width. Palm long and narrow,
equal to movable finger which is armed with 40-42 spinules ; immovable finger with
30-33 spinules and a setose pad on the proximal half (Fig. 32 o).

Palp very similar to that of N. brevicaudatum, Miers (for proportions of segments 2-5
see Table V).

Oviger. Terminal claw slightly longer than tenth segment, armed with eight long
spinules. Total number of denticulate spines 18 (6 + 5 + 3 + 4). Segments 4-6 rather
long, in the proportions 1-43 : 171 : i (Fig. 27 d).

Third leg. Second coxa equal to the sum of the other two (dorsal measurements).
Femur robust, 3-5 times as long as wide; first tibia the longest segment (Table III).
As in A'', brevicaiidatimi there are long setae on the legs, especially on the three principal
segments.

Tarsus and propodus slender, elongated, the former exceeding the latter in length
(Fig. 31 a). Claw approximately one-third of propodus, auxiliaries well developed.

Remarks. The holotype, which is now rather damaged and broken, is very closely
allied to A^. brevicaudatum, Miers, but the tarsus is longer, not shorter, than the propodus
(cf. Fig. 31a and b). A number of specimens in the Discovery collection have been
referred to A^. biartictdatum, Hodgson, because the tarsus is consistently longer relative
to the propodus than in A^. brevicaudatum. In adults the tarsus is equal to or longer than,
and in immature specimens is only a very little shorter than, the propodus (cf. Fig. 31 a,
c and b, d). The movable finger of the chela is rather longer than the palm.

This species is very closely related to, and may prove to be only a more southern

1 There are also a number of long setae on each lateral process, but they have been rubbed off.

NYMPHONIDAE

73

form of, A^. brevicaudatutn. In addition to the longer tarsus already mentioned, the
ocular tubercle is higher, the abdomen is more narrowed posteriorly and the fingers of
the chela are rather longer than in A^. brevicaudatutn.

Fig. 32. Chela of: a. Nymphon biarticulatum, Hodgson, holotype.
b. N. mendosum, Hodgson, c. N. bouvieri, n.sp. (."Ml :< 27.)

The specimens from St. 167 have been provisionally referred to this species. The
tarsus is nearly as long as the propodus — i.e. longer than is typical for A^^. brevicaudatutn
— but the two terminal segments of the palp are consistently shorter than in either
species (see Table V).

Distribution. The holotype was collected in the Ross Sea area; with the exception
of the specimens from Sts. 167 and 195 this form was collected beyond the southern
range for N. brevicaudatutn.

Nymphon bouvieri, n.sp. (Figs. 2^ a, 2jb, 32 c and 33).

St. 172. 26. ii. 27. Off Deception Island, South Shetlands, 525 m.; R. Large dredge : 4 specimens
2 (?(^ (length of holotype 5-3 mm.) and 2 $?.

Description of holotype (cJ). Trunk compact (Fig. 32«); lateral processes in
contact proximally, each with a few long setae; 2-3 long setae on mid-dorsal surface
near the posterior articulation of each segment. Neck very short ; cephalic segment
almost as wide anteriorly as long. Ocular tubercle high and slender (4-6 times as high
as wide); eyes small and sub-terminal.

Proboscis short, two-thirds of cephalic segment and of scape ; slightly expanded in
middle, rounded at apex.

74

DISCOVERY REPORTS

Abdomen pyriform, reaching to middle of second coxa ; not elevated.

Chelophore. Scape equal to cephalic segment, setose (Fig. 33 a). Fingers of chela
shorter than palm, abruptly bent distally, armed with twelve and fifteen spinules
respectively (Fig. 32 c).

Palp as represented in Fig. 25 a ; two terminal segments very short, together less than
one-third of second (this is true of all four specimens, see Table V).

Oviger. Terminal claw rather shorter than tenth segment, armed with five spinules.

Fig. 33. Nymphon bonvieii, n.sp. Holotype: a. Dorsal view of body
with chelophores : x 12 — proboscis foreshortened, b. Third leg.

Total number of denticulate spines 15-18. Segments 4-6 in the proportions
17 :2-5 : I (Fig. zy b).

Third leg short and setose (Fig. 33 b); setae, as a rule, longer than the diam.eter of
the segment to which they are attached. Second coxa equal to the sum of the other two.
Femur subequal to second tibia, with 3-4 high wide gland tubercles on mid-ventral
surface. First tibia the longest segment (see measurements). Tarsus approximately two-
thirds of propodus ; claw nearly half as long as the latter ; auxiliaries well developed
one-fourth of claw.

Meastirements {mm.)

Length of proboscis ...

Diameter of proboscis

Length of trunk

Width across second lateral processes

Length of cephalic segment ...

Width of cephalic lobes

Length of abdomen ...

Length of scape

Third leg:

I '4

First coxa ...

0-8

i-o

Second coxa

1-2

3-6

Third coxa ...

0-4

3-3

Femur

2-2

2-0

First tibia ...

3-25

1-8

Second tibia

2-4

2-0

Tarsus

I'D

2-0

Propodus

1-5

Claw

0-53

Auxiliaries ...

0-13

NYMPHONIDAE

75

The female shows the usual sexual differences ; the femur is more robust, the length
being just over twice the maximum width.

Remarks. Although it bears a strong superficial resemblance to A'^. brevicaudatum,
Miers, this species can readily be distinguished by (i) the extremely short terminal
palpal segments, (2) the longer and exceedingly slender ocular tubercle^which is even
more slender in the other specimens than in the holotype, and (3) the small number of
spinules on the fingers of the chela. In the latter respect it approaches N. proximum.
Caiman (1915, p. 34, fig- 6 D), but the palm is longer and more slender, the fingers
more curved distally and the spinules longer; the ocular tubercle is 4-6 times as high
as wide instead of only as high as wide ; the setae are much longer.

C. SPECIES OF UNCERTAIN SYSTEMATIC POSITION
Nymphon multidens, n.sp. (Figs. 34 and 35).

St. 456. 18. X. 30. I mile E of Bouvet Island, 40-45 m. Large dredge: i ? (holotype).

Description of holotype ($). Trunk rather compact ; lateral processes separated by
half their own diameter or rather more. Cephalic segment equal to the sum of the three
posterior segments ; neck rather long, base of oviger in contact
with first lateral process. Ocular tubercle as high as wide,
rounded; eyes well developed. Setae absent (Fig. 34).

Proboscis short, sub-cylindrical, rounded at apex; much
shorter than cephalic segment, and two-thirds of scape.

Abdomen short, sub-cylindrical, reaching to distal end of
fourth lateral process ; elevated at an angle of about 40^

Chelophore. Length of scape three times the distal width.
Chela somewhat shorter than scape and of a very unusual
type (Fig. 35 i), with immovable finger very short relative
to length of palm. There are twenty-four short, crowded
spinules on the movable, 26-28 on the immovable finger;
setose pad short but distinct.

Palpal segments 2-5 in the proportions 4 : 3-33 : 1-5:2

(Fig- 35 «)■

Oviger. Terminal claw half of tenth segment, armed with

8-10 short spinules. Total number of denticulate spines at
least seventy,^ but they are so crowded that it is difficult to
count the numbers exactly (formula approximately 23 + 16
+ 18 + 16 = 73); each spine is long and slender with nu-
merous short, crowded denticulations on each side. Seg-
ments 4-6 relatively short, in the proportions i -8 : 1 75 : i .

Third leg. First coxa very short, second approximately half as long again as the sum
of the other two (dorsal measurements). Femur equal to first tibia, not greatly distended

1 Hence the specific name.

Fig. 34. Nymphon multidens,
n.sp. Holotype. Dorsal view
of body with chelophores:
X 16.

76

DISCOVERY REPORTS

although containing developing ova (length six times the greatest width). Second tibia
the longest segment, half as long again as femur. Tarsus half as long as propodus,
ventral margin of each segment armed with numerous short spinose setae ; claw short,
auxiliaries relatively long (Fig. 35 c). Setae for the most part very minute.

J^ > ^ \ -, .>

Fig. 35. Nymplionmiiltidens,n.sp.: a.Filp. 6. Chelophore. c'
of third leg. (All x 33.) d. Chela: x 60.

. Terminal segments

Length of proboscis . . .

Diameter of proboscis

Length of trunk

Length of cephalic segment

Width of cephalic lobes

Width of neck

Width across second lateral processes

Length of abdomen ...

Measurements

{rmn.)

Third leg:

i-o

First coxa ..

0-6

Second coxa

3 '4

Third coxa ..

17

Femur

0'93

First tibia ..

0"5

Second tibia

2-3

Tarsus

0-6

Propodus

Claw

Auxiliaries ..

0-6

1-8

0-5

4-4

4-4
67

i-o

2-0

0-45
0-3

Remarks. This species differs from all other Antarctic and sub-Antarctic species as
regards the chela, which recalls that of, e.g., A'^. grossipes (Fabr.) from the North Atlantic
(Hoek, 1881, p. 44, pi. iii, figs. 9-12, pi. iv, fig. i). The northern species, however, has
the tarsus elongated and longer than the propodus; there are also three or four very
long spines on the proximal ventral margin of the latter and the claw is much longer.

The male is unknown, but the species probably belongs to group I.

Distribution. One mile off the east coast of Bouvet Island.

NYMPHONIDAE

77

D. TROPICAL SPECIES

Nymphon angolense, n.sp. (Figs. 36 and 37).

St. 272. 30. vii. 27. Off Elephant Bay, Angola, from 13° n' S, 12° 44' 45" E to 13' 09' 45" S,
12° 46' E, 73-91 m. ; gn. S. M. Large otter trawl : 30 specimens, including ovigerous and larvigerous
(JcJ; 2 specimens bear each a small bivalve mollusc.

Description of holotype (larvigerous 3). Trunk elongated and slender, lateral
processes separated by more than their own diameter (Fig. 36 «). Cephalic segment as
long as the remaining segments plus abdomen. Neck long and slender, less than half
as wide as the anterior cephalic lobes ; base of oviger small and rounded, in contact

Fig. 36. Nymphon angolense, n.sp.: a. Dorsal view of body of holotype with chelophores and palps.
b. Lateral view of body with chelophore and palp. c. Third leg.

with first lateral process. Ocular tubercle low and rounded ; a little higher than wide,
eyes conspicuous (Fig. 36 a and b).

Proboscis half as long as cephalic segment, cylindrical, rather rounded at apex.

Abdomen reaching a little beyond fourth lateral process, elevated at an angle of about
45°, approximately three times as long as broad.

Chelophore. Scape subequal to proboscis; length five times the distal width. Chela
slender, rather curved and somewhat longer than scape (Fig. 37 a) ; fingers a little longer
than palm, approximately forty-five crowded spinules on immovable, fifty-five on
movable one.

Palp long and slender, segments 2-5 in the proportions 10 : 6 : 3 : 3-3 (Fig. 37 b).

78 DISCOVERY REPORTS

Oviger long and slender. Terminal claw two-thirds of tenth segment, armed with
fourteen short spinules. Total number of denticulate spines 39 (13+10 + 8 + 8).
Segments 4-6 as represented in Fig. 37 J, in the proportions 2:2-3:1; segment 5
straight, of almost uniform diameter throughout, with a few hook-like spines near the
distal articulation (Fig. 37 d).

Third leg long and slender, sparsely beset with short, spinose setae on the three main
segments. Second coxa a little longer than the sum of the first and third (dorsal
measurements). Femur slender and, like first tibia, expanded in the distal third
(Fig. 36 c), more than half as long again as the three coxae together; gland openings very
inconspicuous, about ten in number. Second tibia the longest segment, approximately
five times the sum of tarsus and propodus. The latter slightly longer than the former
and armed ventrally with six or seven spines (Fig. 37 c). Claw rather more than half
as long as propodus ; auxiliaries very long — at least two-thirds of claw.

Measurements {mm.)

Holotype ?

Length of proboscis ...

Diameter of proboscis

Length of trunk

Length of cephalic segment

Width of cephaHc lobes

Length of neck

Width of neck

Width across second lateral processes

Length of abdomen ...

Third right leg:

First coxa ...

Second coxa

Third coxa ...

Femur

First tibia ...

Second tibia

Tarsus

Propodus

Claw

Auxiliaries ......

The female is very similar to the male, but segments 4-5 of the oviger are relatively
short. The femur is rather more robust and of almost uniform diameter throughout ;
the first tibia is expanded distally as in the male.

Remarks. The only species of the genus Ny77iphon previously reported from West
Africa is A^. graciUmum, Caiman (Loman, 1923 b, p. 5). N. angolense difl^ers from the
holotype of that species in several respects : (i) there are at least twice as many spinules
on the fingers of the chela, (2) there are a few hooked spines near the distal articulation
of the fifth segment of the male oviger, (3) the tarsus does not exceed, although it may
nearly equal the propodus, (4) the auxiliary claws are more highly developed, and
(5) the first tibia in both sexes, and the femur in the male, is expanded distally.

1-8

07

1-9

0-8

6-0
3-6

5-9
3-6

i-i

1-2

1-6

1-6

0-45

2-8

0-5

2-8

i-o

I-I

i-o

2-0

0-85

1-6

I-O

0-8

6-8

7-0

9-2

8-4

2-0

12-0

i-i

I-O

1-3

1-2

0-8

0-75

0-55

0-5

NYMPHONIDAE

79

A'^. angolense does not agree with any of the African species Hsted by Flynn (1928,
p. 6). ^

b

Fig. 37. Nymphon angolense, n.sp.: a. Chelophore. b. Palp. c. Terminal segments of third leg. (All x 27.)
d. Segments 1-7 of male oviger: >; 20, and distal end of fifth segment: x 60.

Genus Heteronymphon, n.g.

Diagnosis. Body smooth, rather compact, segmented, with well-defined lateral
processes; cephalic segment with short thick neck and low wide ocular tubercle
situated in front of neck and between the two anterior lobes of cephalon. Chelophores
long, slender; chela not exceeding half the length of the scape. Palp five-jointed.
Oviger ten-jointed with denticulate spines on the four terminal segments, but without
terminal claw; fifth segment in male relatively short and considerably expanded distally
(like that of group II in the genus Nymphon).

Remarks. In general appearance the specimens referred to this genus resemble those
belonging to the smaller species of Nymphon with short neck, e.g. Nymphon pfefferi.
But the anterior position of the ocular tubercle ^ and the absence of the terminal claw
on the oviger separates this form from all species of the genus Nymphon. These two
characters are also found in the genus Nymphonella, Ohshima (1927, p. 257), but the
latter would appear to belong to the Eurycididae rather than the Nymphonidae, since the
chelophore is reduced, the palp has nine segments, and the proboscis is ' ' directed ventrad ".
1 The ocular tubercle is never situated in front of the point of insertion of the oviger in Nymphon.

8o DISCOVERY REPORTS

The genus Heteronymphon serves to link the Nymphonidae to the Phoxichilidae
(Pallenidae) on the one hand and to the PhoxichiHdiidae on the other. In the former
family the terminal claw of the oviger is sometimes wanting or vestigial ; in the latter
family the ocular tubercle is situated near the anterior border of the cephalon in, e.g.,
Phoxichilidium mistrale.

Loman (1908, table facing p. 19) uses the oviger as a basis for his first division of the
order Pantopoda. Bouvier (191 3, p. 34) has pointed out that this division into a section
with and a section without a terminal claw to the oviger is not altogether satisfactory.
The genus Heteronymphon adds yet another exception to the first section, since the
terminal claw of the oviger is absent and several of the terminal denticulate spines are
enlarged, presumably to take the place of the claw (Fig. 39 d).

Heteronymphon kempi, n.sp. (Figs. 38 and 39).

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl: i speci-
men, probably cJ.

St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 54° 02' S,
36°38'W to 54° 11' 30" S, 36°29'W, 122-136 m.; gn. M. St. Large otter trawl: i ^, with
N. pfefferi.

St. MS 66. 28. ii. 26. East Cumberland Bay, 2J miles SE of King Edward Point Light to i\ cables
W X N of Macmahon Rock, 18 m. Small beam trawl: 1$ (length of trunk = 1-4 mm.) and
1 immature specimen.

St. MS 71. 9. iii. 26. East Cumberland Bay, 9J cables E x S to 1-2 miles E x S of Sappho Point,
60-1 10 m. Small beam trawl: 3 $? including holotype.

St. MS 74. 17. iii. 26. East Cumberland Bay, i cable SE x E of Hope Point to 3-1 miles SW of
Merton Rock, 22-40 m. Small beam trawl: i specimen (sex uncertain, probably immature), with
A^. brevicaudatum ; i ? with A*", pfefferi.

Description of holotype (9). Trunk rather compact, lateral processes separated by
approximately their own diameter, articulations between the segments well marked.
Cephalic segment less than half the length of the trunk, neck short and thick. Ocular
tubercle low and broad, situated in front of neck near the anterior border of cephalon,
eyes large (Fig. 38 a and b).

Proboscis slightly decurved, approximately as long as cephalic segment, sub-
cylindrical, rounded at apex.

Abdomen long, slender, very much elevated.

Chelophore slender ; scape rather longer than proboscis ; chela not exceeding half the
length of the scape, fingers scarcely longer than palm armed with fifteen and nineteen
spinules respectively (Fig. 39 h).

Palp five-jointed; third joint the longest, one-fourth as long again as the fifth and
nearly twice as long as second (Fig. 39 a).

Oviger ten-jointed; first three segments together somewhat longer than fourth,
which is only slightly shorter than fifth. Number of denticulate spines on each of the
four terminal segments five, five, four and seven (in female from MS 66 nine, six, six

NYMPHONIDAE

8i

and eight). There is no terminal claw, but the last three or four denticulate spines are
considerably longer than the others on segment lo (Fig. 39 d).

Leg with ova visible in femur and the two distal coxae. Second coxa equal to the

Fig. 38. Heteronymphon kempi, gen. et sp.n . : a. Holotype ; dorsal view of body with chelophores.
b. Holotype, lateral view of body with chelophores, palp and oviger. c. Third leg.

Other two together, with a distinct prominence on the dorsal surface (Fig. 38 c). Femur
and first tibia equal, second tibia the longest segment. Propodus nearly twice as long
as tarsus and more than twice as long as the main claw. Auxiliary claws absent.

Measurements (mm.)

Length of proboscis

Diameter of proboscis

Length of trunk ...

Length of cephalic segment

Width of cephalic lobes

Width across second lateral processes ...

Length of abdomen (animal placed in natural position)

Third leg:

First coxa

Second coxa

Third coxa ...

Femur ...

First tibia

Second tibia

Tarsus ...

Propodus

Claw

0-S5

0-20
1-25

0-54
0-30

I -GO

0-44

0-30
070
0-40
I -So
I -So

2-20
0-50

0-93
0-36

Remarks. Although the specimens are very small the genital pores are conspicuous
in all the females. The largest specimen is the adult from MS 66, and it is quite
probable that this female has the full number (29) of denticulate spines on the oviger
for the species.

II

82

DISCOVERY REPORTS

The single male specimen has both ovigers mutilated, but fortunately the fifth
segment is present on one side (Fig. 39 c) ; it is relatively short compared to the fourth

Fig. 39. HeteronymphoH kempi, gtn. etsp.n.: a. Palp. b. Chela, c. Segments 3-5 of male oviger — 6 incom-
plete, d. Segments 7-10 of female oviger with loth segment further enlarged — the denticulate spines are
not accurately represented in the smaller figure, {a and b: ■: 100; c and d: x 60 and circ. 240.)

segment and is considerably expanded or clubbed distally as in group II of the genus
Nymphon (cf. Figs. 39 c and 27).

The relative proportions of palpal segments 2-5 differ markedly from those of any
Antarctic species of the genus Nymphon.

Family PHOXICHILIDAE (PALLENIDAE)
Genus Pallene, Johnston

Pallene margarita, n.sp. (Figs. 40 and 41).

St. 42. I . iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of Jason
Light to 4 miles N 39° E of Jason Light, 120-204 m. ; M. Large otter trawl : 3 specimens.

St. 140. 23.xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 54° 02' S,
36°38'W to 54° ii'3o"S, 36" 29' W, 122-136 m.; gn. M. St. Large otter trawl: 3 specimens
including i S bearing ova and larvae.

St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, from 54° 04' S, 36° 27' W
to 53° 58' S, 36° 26' W, 155-178 m.; gn. M. S. Large otter trawl: i $.

St. WS 212. 30. V. 28. 49° 22' S, 60° 10' W, 242-249 m.; gn. S. M. Commercial otter trawl:
many specimens including ovigerous SS-

PHOXICHILIDAE

83

St. WS214. 31.V. 28. 48°25'S, 6o°4o'W, 208-219111.; f. d. S. Commercial otter trawl:
I ovigerous 3-

St. WS227. i2.vi. 28. 5i°o8'S, 56°5o'W, 320-298 m.; f. gn. S. Commercial otter trawl:
many specimens including ovigerous and larvigerous (JcJ.

St. WS 228. 30. vi. 28. 50° 50' S, 56° 58' W, 229-236 m.; Sh. c. w. S. Commercial otter trawl:
5 specimens, including larvigerous (J.

St. WS 229. I. vii. 28. 50° 35' S, 57° 20' W, 210-271 m.; f. gn. S. Commercial otter trawl : many
specimens, including ovigerous and larvigerous (J(^.

St. WS234. 5. vii. 28. 48°52'S, 60° 25' W, 195-207 m.; f. gn. S. Commercial otter trawl:
3 specimens.

St. WS 237. 7. vii. 28. 46° 00' S, 60° 05' W, 150-256 m.; c. br. S. Sh. Commercial otter trawl:
I larvigerous ^.

St. WS244. 18. vii. 28. 52° 00' S, 62^40' W, 253-247 m.; f.d.S.M. Commercial otter trawl:
I larvigerous i^.

St.WS245. 18. vii. 28. 52°36' S, 63° 40' W, 304-290 m.;d.gn.S. Sh. Commercial otter trawl :
I $, 3 ovigerous and larvigerous (J(J.

Description of holotype.^ Trunk segmented, of rather loose build; lateral pro-
cesses, with the exception of the last two pairs, separated by more than their own

Fig. 41. Pallene margaritti, n.sp.: a. Chela: x 90.
b. Segments 3-5 of male oviger: x 33.

Fig. 40. Pallene margarita, n.sp. Dorsal
view of body with chelophores: x 20.

diameter. Cephalic segment much longer than the sum of the three posterior segments,
and considerably expanded anteriorly; neck long. Ocular tubercle bluntly conical, a

1 A larvigerous male from St. WS 229.

84 DISCOVERY REPORTS

little higher than wide ; eyes well developed. Oviger base small and situated a short
distance in front of first lateral process (Fig. 40).

Proboscis short, stout, sub-cylindrical.

Abdovien short, elevated at an angle of 80-90° and reaching to middle of fourth lateral
process.

Chelophore short. Scape equal to proboscis and approximately two and a half times
as long as wide distally. Chela as represented in Fig. 41 (7; movable finger equal to
palm ; 7-8 minute teeth on immovable one.

Palp absent (as also in the female).

Oviger. Terminal claw absent; spines on the four terminal segments short, rather
broad and rounded distally, numbering 49^ (14+ 11+ 12+ 12). Fifth segment
relatively long and furnished with the usual distal lobe (Fig. 41 b).

Third leg. Second coxa long and slender, more than twice the sum of the first and
third (dorsal measurements). Femur subequal to first tibia and at least nine times as
long as wide. Second tibia the longest segment. Tarsus very small; 4-5 long spines
on proximal sixth of ventral margin of propodus. Claw and auxiliaries long and
slender.

In the female the femur is considerably dilated in the proximal two-thirds and the
fifth segment of the oviger is relatively short, without the distal lateral lobe.

Measurements (mm.)

$

Holotype

(St. WS 227)

Length of proboscis

0-6

07

Diameter of proboscis

04

0-5

Length of trunk

2-6

3-0

Length of cephaUc segment ...

1-67

2-0

Width of cephaHc lobes

0-65

075

Width of neck

0-22

0-3

Width across second lateral processes

1-2

1-3

Length of abdomen

0-3

0-25

Third leg:

First coxa ...

0-45

0-5

Second coxa

173

173

Third coxa

0-27

0-35

Femur

2-9

3-6

First tibia ...

2-9

3-53

Second tibia

3-6

4-4

Tarsus

o-i

Propodus

0-93

Claw

0-64

Auxiliaries ...

0-5

Remarks. This species appears to belong to the genus Pallene, but the two terminal
body segments are not fused as in the majority of the other species — a condition that is
also found, e.g., in P. emaciata, Dohrn.

The genus is seldom represented in Antarctic collections. Bouvier (1913, p- 20 and

^ 48-56 in adults.

PHOXICHILIDAE

8S

p. 96) mentions P. dimorpha,^ Hoek, as the only Antarctic species (from Kerguelen).
P. margarita differs from the latter chiefly in that the terminal claw of the oviger is
entirely suppressed and the palp is absent in both sexes.

Genus Austropallene, Hodgson

Austropallene cornigera (Mobius) (Figs. 42 and 43).

Caiman, 1915, p. 38, ubisynon. et lit.
Loman, 1923, p. 22.

Fig. 42. Austropallene cornigera (Mobius). Segments 5-10 of oviger of a, male; b, female.

St. 27. i5.iii. 26. West Cumberland Bay, South Georgia, 3-3 miles S45'E of Jason Light,
no m.; M. R. Large dredge: i ^.

St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89' E of
Jason Light to 4 miles N 39° E of Jason Light, 120-204 m. ; M. Large otter trawl : 3 ??.

St. 140. 23.xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 54" 02' S,
36°38'W to 54° 11' 30" S, 36°29'W, 122-136 m.; gn. M. St. Large otter trawl: i immature
specimen probably belongs to this species.

St. 156. 20. i. 27. 53° 51' S, 36° 21' 30" W, 200-236 m.; R. Large heavy dredge: i ?.

St. 159. 21. i. 27. 53° 52' 30" S, 36° 08' W, 160 m.; R. Large heavy dredge: i larvigerous ^.

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' 00" W, 342 m.; R.
Large heavy dredge: 16 specimens, including 5 ovigerous 3^, accompanied by the following note:
"White-spotted Pycnogon. Horn-coloured. Abdomen [trunk] opposite 2 middle pairs of legs whh
a white dorsal patch. All legs with a white spot at distal end of segments 3, 4, 5 and 6, with, in
addition, one spot at proximal end of segment 5".

St. 175. 2.iii. 27. Bransfield Strait, South Shetlands, 63° 17' 20" S, 59^ 48' 15" W, 200 m.;
m. St. G. Large dredge: i ovigerous c? and i young ? ?.

St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 64° 20' S, 63° W, 160-335 m.; M.
Large otter trawl: i ovigerous <?, rather slender, but apparently this species.

1 Schimkewitsch (1930) has established a new genus, Oropallene, for Hoek's species. According to Flynn
(1919 p 77) Pallene valida, Haswell, would also belong to Oropallene. Schimkewitsch appears to have
carried the multiplication of genera within the family Phoxichilidae (Pallenidae) to the utmost extreme and
characters such as (i) the presence or absence of auxiliary claws, (2) the fusion or non-fusion of two body
segments, (3) the presence or absence of a rudimentary palp, and (4) the presence or absence of a termmal
claw on the oviger assume far more importance in this than in almost any other family.

86

DISCOVERY REPORTS

St. 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 64^ 48' 30" S, 63° 31' 30" W,
259 m.; M. Large dredge: 2 cJ,^, i ?.

St. 195. 30. iii. 27. Admiralty Bay, King George Island,
South Shetlands, 62' 07' 00" S, 58^ 28' 30" W, 391 m.;
M. St. Medium otter trawl: i ovigerous 3*, with Nymphon
charcoti.

St. 363. 26. ii. 30. 2-5 miles S 80^ E of SE point of
Zavodovski Island, South Sandwich Islands, 329-278 m.; Sc.
Large dredge : i larvigerous ^, i ovigerous cJ.

Remarks. The white spots mentioned in the note
on the label (St. 170) are still quite clear in the pre-
served specimens from this station and are present,
though less well marked, in specimens from some of
the other stations also.

The spurs on the cephalon are longer and more
slender in the specimens from St. 42, 156 and 159 and
the short spinous projection on each lateral process is
more pronounced than is typical of the species.

Segments 5-10 of the male and female ovigers are
represented in Fig. 42 a and b. The two globular
bodies attached to the ventral surface of the specimen
from St. 187 (Fig. 43) appear to be egg cases of some
animal ; they can be detached without damaging the
Pycnogon and are apparently only glued on to the specimen

Fig. 43. AiistropaUetie contigera (M6-
bius). Ventral view of body showing
two (?) egg-cases of some animal ad-
hering to the epidermis.

Austropallene brachyura (Bouvier).

Caiman, 1915, p. 39, ubi synon. et bibl.

St. 140. 23.xii.26. Stromness Harbour to Larsen Point, South Georgia, from 54° 02' S,
36°38'W to 54° 11' 30" S, 36°29'W, 122-136 m.; gn. M. St. Large otter trawl: i immature
specimen probably belongs to this species, but the neck is somewhat longer than is typical.

St. 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 64° 48' 30" S, 63° 31' 30" W,
259 m.; M. Large dredge: i ?.

Austropallene cristata (Bouvier).

Pseudopallene cristata, Bouvier, 1911, p. 1138; 1913, p. 102, figs. 55-59.

St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, from 54° 04' S, 36° 27' W to
53° 58' S, 36° 26' W, 155-178 m.; gn. M. S. Large otter trawl: i immature specimen.

PHOXICHILIDIIDAE 87

Family PHOXICHILIDIIDAE

Genus Pallenopsis, Wilson
Pallenopsis pilosa (Hoek).

Plioxicliilidhan pilosiim, Hoek, 1881, p. 90, pi. xiii, figs. 10-13.

Pallenopsis pilosa, Hoek (1883), p. 9 (in list of species).

Pallenopsis pilosa, Hodgson, 1907, p. 15, pi. ii, fig. 2.

Pallenopsis pilosa, Bouvier, 1911, p. 1139; 1913, p- 107, figs. 60 and 61.

St. 170. 23 .ii. 27. OflF Cape Bowles, Clarence Island, 61° 25' 30" S, 53'46'W, 342 m.; R.
Large dredge: i (^, i ?? (genital pores not formed), i young specimen.

St. 177. 5. iii. 27. 27 miles SW of Deception Island, South Shetlands, 63"" 17' 30" S, 61° 17' W,
1080 m.; M. c. St. Large dredge: i immature specimen, with Nymphon brevicatidatum.

St. 180. II. iii. 27. 17 miles W of N point of Gand Island, Schollaert Channel, Palmer Archi-
pelago, 160-330 m.; M. St. Large otter trawl: i immature specimen with N. australe.

St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 64" 20' S, 63° 01' W, 160-335 ^■''
M. Large otter trawl : i c?, 4 immature specimens.

St. 182. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 64° 21' S, 62° 58' W, 278-500 m.;
M. Large otter trawl: i ovigerous (J, i ?, i immature specimen.

St. 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 64° 48' 30" S, 63° 31' 30" W, 259-
354 m.; M. Large otter trawl: i cJ, i 5, i immature specimen.

St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 64° 56' S, 65° 35' W, 315 m.; M. R.
Large dredge: i $ (with small Cirripede attached).

St. 458. 19. X. 30. 7 miles S 50° W of Cape Circumcision, Bouvet Island, 357-377 m. Large
dredge : i cJ with A'', australe.

Remarks. The majority of the specimens are rather more compactly buih and more
hairy than the types, with which they agree in other respects. Several of the specimens
are light brown, the others being of a pale yellowish colour.

Measurements (mm.)

St. 182,

Length of proboscis^ ...

Diameter of proboscis'^

Length of trunk

Width across second lateral processes

Length of cephalic segment

Length of abdomen
Length of scape
Length of chela
Third leg:

First coxa ...

Second coxa

Third coxa ...

Femur

First tibia . . .

Second tibia

Tarsus

Propodus

Claw

Auxiliaries ...

1 Measured laterally, not dorsally as in other species.

0 (ovig.)

St. 190 ?

3-4

4-0

1-4

1-65

6-2

7-2

4-8

5-2

3-6

4-0

3-6

4-4

2-0 i- 1-8

2-4 + 2-4

1-87

2-0

1-6

1-6

3-2

3-2

I -07

1-2

7-6

10-4

7-6

lO-O

9-2

I2-0

0-3

0-5

2-6

3-0

1-8

2-0

0-67

88 DISCOVERY REPORTS

Distribution. Eastern and Western Antarctic; from deep water at lat. 46'^ 16' S,
long. 48° 27' E and lat. 53° 55' S, long. 108° 35' E.

Pallenopsis patagonica (Hoek) (Fig. 44).

Phoxichilidium patagonicum, Hoek, 1881, p. 84, pi. xii, figs. 6-9.

Phoxichilidiiim patagonicum var. clegans, Hoek, 1881, p. 86, pi. xii, fig. 10.

Pallenopsis hicmolis, Hodgson, 1907, p. 17 in part — ? not the type specimen.

? Pallenopsis glabra, Mobius, 1902, p. 184, pi. xxvii, figs. 1-6. Bouvier, 1913, p. 109, figs. 62-65.

Pallenopsis glabra, Hodgson, 1907, p. 11 ; Caiman, 1915, p. 41.

St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, from 54° 02' S,
36° 38' W to 54° 11' 30" S, 36° 29' W, 122-136 m.; gn. M. St. Large otter trawl: i ,^, 2 ??.

St. 167. 20. ii. 27. Off Signy Island, South Orkneys, 60" 50' 30" S, 46" 15' W, 244-344 m.;
gn. M. Large otter trawl : i <^ with encrusting Polyzoa.

St. 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 64° 48' 30" S, 63" 31' 30" W, 259-
354 m.; M. Large otter trawl: i large ?.

St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 64° 56' S, 65° 35' W, 93-315 m.; M. R.
Large rectangular net : i $.

St. WS 83. 24. iii. 27. 14 miles S 64° W of George Island, East Falkland Island, from 52^ 28' S,
60° 06' W to 52° 30' S, 60° 09' 30" W, 137-129 m.; f. gn. S. Sh. Commercial otter trawl: 3 ??,
2 immature specimens.

St. WS 84. 24. iii. 27. 7J miles S 9° W of Sea Lion Island, East Falkland Island, from 52° 33' S,
59°o8'W to 52° 34' 30" S, 59" 11' W, 75-74 m.; c. S. Sh. St. i ovigerous ^, egg-mass simple
but attached to both ovigers: 6 small specimens, including 2 ^<S (i ovigerous) and i ? with genital
pores open, may belong to this species. Several of the specimens carry encrusting Polyzoa.

St. WS 85. 25. iii. 27. 8 miles S 66° E of Lively Island, East Falkland Island, from 52° 09' S,
58° 14' W to 52° 08' S, 59° 09' W, 79 m.; S. Sh. Commercial otter trawl: 7 specimens, very hairy;
I $ has the genital pores visible, the others are probably all immature.

St. WS 88. 6. iv. 27. 54° 00' S, 64° 57' 30" W, from 54° 00' S, 65° 00' W to 54° 00' S, 64° 55' W,
118 m.; S. Sh. St. Commercial otter trawl: i ? and ? i very young specimen with a long tapering
point to ocular tubercle.

St. WS 91. 8. iv. 27. 52° 53' 45" S, 64° 37' 30" W, from 52° 54' 30" S, 64° 39' W to 52° 53' S,
64° 36' W, 191-205 m.; f. d. S. Sh. Commercial otter trawl: i ?.

St. WS 92. 8. iv. 27. 51° 58' 30" S, 65^ 01' W, from 52° 00' S, 65° 00' W to 51° 57' S, 65° 02' W,
145-143 m.; f. d. S. St. Commercial otter trawl: 2 ^,S, i ovigerous with a single egg-mass carried
by both ovigers.

St. WS 93. 9. iv. 27. 7 miles S 80° W of Beaver Island, West Falkland Island, from 51° 51' S,
61° 30' W to 51° 54' S, 61° 30' W, 133-130 m.; gy. S. Commercial otter trawl: 4 small specimens,
including 2 (JcJ (i ovigerous), bearing adherent Foraminifera.

^St. WS 210. 29. v. 28. 50° 17' S, 60" 06' W, 161 m.; gn. S. Commercial otter trawl: i ?.

St. WS214. 31.V. 28. 48°25'S, 6o°4o'W, 208-219 m.; f. d. S. Commercial otter trawl:
I specimen.

St. WS215. 31.V. 28. 47°37'S, 6o°5o'W, 219-146 m.; f. gn. S. Commercial otter trawl:
I specimen.

1 The specimens from WS 210 to the end of the list probably belong to this species; they are mostly of
small size and compact build.

PHOXICHILIDIIDAE

50'' 20' S, 62° 30' W, 162-161 m.; gn. S. Sh.

48°52'S, 6o°25'W, 195-20701.; f. gn. S.

89

Commercial otter trawl:

Commercial otter trawl :

St. WS 225. 9. vi. 28.
I specimen.

St. WS 234. 5. vii.28.
I ovigerous (^.

St. WS 237. 7. vii. 28. 46^ 00' S, 60^ 05' W, 150-256 m.; c. br. S. Sh. Commercial otter trawl:
3 specimens (i was apparently undergoing a moult when captured).

St. WS243. 17. vii. 28. 5i^o6'S, 64^ 30' W, 1 44-141 m.; c. d. S. Commercial otter trawl:
I ovigerous cj.

St. WS 244. 18. vii. 28. 52" 00' S, 62° 40' W, 253-247 m.; f. d. S. M. Commercial otter trawl:

3 $$, 2 immature specimens.

St. WS 245. 18. vii. 28. 52" 36' S, 63° 40' W, 304-290 m.; d. gn. S. Sh. Commercial otter trawl:
5 ??, 3 (J(J (2 ovigerous) and several young specimens.

St. WS 246. 19. vii. 28. 52" 25' S, 61° 00' W, 267-208 m.; c. gn. S. P. Commercial otter trawl :

4 specimens, including i ovigerous cJ.

Remarks. Caiman, 191 5, p. 42, has suggested that Phoxichilidium patogonicum, Hoek,
is identical with Pallenopsis glabra, Mobius. All the syntypes in the Challenger collection
are females; the genital pores are quite distinct in all, although the oviger is not fully
developed in the smallest individual. Caiman has also commented on the compactness
of the body in the Challenger specimens as compared with the Terra Nova material
described as Pallenopsis glabra.

The material in the Discovery collection is, for the most part, of the compact type
and was also collected from the
Falkland Island area. The largest
specimens from this region, e.g.
the adult from WS 88, is very
similar to the largest Challenger
syntype. The small specimens
tend to have the lateral processes
but little separated and to have
rather long fine setae on the
walking legs. The smallest speci-
men from the more southern
region (that from St. 190, length
= 9-5 mm.) also has the lateral processes of the body but little separated and setose
legs. The largest specimens (Sts. 140 and 187) have the lateral processes separated by
a distance equal to their own diameter.

Although this species reaches a considerable size in the more southern, colder region
the males reach maturity early and carry ova when they have reached a length of
7-7-5 mm. (from ocular tubercle to tip of abdomen). The genital pores, also, are formed
in the females when they reach a length of about 8-9 mm., but the ovigers are not
fully developed until later. Even in adult females two of the four termmal segments
(7 and 8) are nearly always more or less fused together, as figured by Bouvier (1913,
p. Ill, fig. 64).

a

Fig. 44. Pallenopsis patagonica (Hoek). Femoral gland-tubercle
of male. «. Trunk length = 1 1-2 mm., St. WS 92. 6. A smaller
specimen from St. WS 84. c. Trunk length = 107 mm. from
St. WS92.

90

DISCOVERY REPORTS

The femoral gland is quite prominent in small males (Fig. 44); in the two males
from WS 92 it is situated on a raised mound ^ (Fig. 44 a, c), but is much shorter in
one specimen than in the other. In the male specimen of the first Discovery Expedition,
described by Hodgson under P. fiiemalis, 1907, p. 19, the femoral gland is still shorter,
and in the male from St. 140 it is not visible to the naked eye.

Table VII

Length of

Relative length

s of segments 2-6 of oviger

Specimen

Station

trunk plus

abdomen in mm.

2

3

4

5

6

<?

167

12-0

0-84

1-66

1-93

0-89

<S

140

15-0

078

1-38

1-38

070

<S

WS93

9-0

0-87

I -60

I -60

o-8o

(?

WS92

II-2

077

1-20

1-20

0-62

3

Valdivia

?

I 0-50

0-93

1-30

0-58

?

Valdivia

i6-o

I 0-93

1-19

1-19

073

?

187

17-3

o-8i

1-31

0-89

0-65

?

140

14-8

o-8o

1-32

i-i6

0-68

0

V

Challenger syntype

12-2

070

I -06

o-8i

0-56

Table VIII

Specimen

Station

Length of

trunk plus

abdomen

in mm.

Third right leg
(length of segments in mm.)

Femur

Tibia i

Tibia 2

?
<S
$
?

??

??

187
140
140
190

WS84
WS93

17-3

15-0

14-8

9-4

7-4

8-0

i8-8

17-4

19-6

9-6

67s
8-6

147

15-1

15-6

8-5

675

8-3

22-3
23-1
24-4
II-4

9-4

107

The type specimen of P. hiemaJis differs from all specimens of P. patagonica examined
(i) in the size and arrangement of the spines on the ventral surface of the propodus
(cf. Hodgson, 1907, pi. ii, figs. 3, 30), (2) in having a very well-developed tubercle
on each lateral process, and (3) in having relatively longer and narrower vestiges
of the palps. Most specimens of P. patagonica have 2-4, generally three, very long
spines at the proximal end of the propodus and the rest are very small ; whereas in
the type of P. hiemalis (9) the spines on the propodus are all of uniform size. The
specimens from St. 140, however, exhibit an intermediate condition, the proximal six
or more spines being relatively long, the others decreasing gradually in size.

Distribution. Eastern and Western Antarctic and Magellan District.

Pallenopsis spicata, Hodgson.

Hodgson, 1914-15, p. 146.
Caiman, 1915, p. 44, text-fig. 9.

St. 170. 23.ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53°46'W, 342 m.; R.
Large dredge: i immature specimen.

^ Also in males from WS 210 to the end of the list.

PHOXICHILIDIIDAE

91

Third leg:

2-8

First coxa

i-o

I'S

Second coxa

3-2

2-6

Third coxa ...

0-8

2-4

Femur

8-0

S'2

First tibia ...

7-2

5"°7

Second tibia

10-8

17

Tarsus

Propodus

Claw

0-27

2-4
1-2

Remarks. This is undoubtedly an immature specimen of P. spicata, Hodgson. It
agrees in most respects with the Terra Nova specimen ; the following differences are
doubtless mostly due to the fact that the specimen is not fully developed, (i) The
lateral processes are less widely separated, so that the cephalic segment is slightly longer
than the sum of the two succeeding segments. (2) The oviger is not fully developed, is
shorter than the proboscis and divided into only four segments. (3) The walking legs
are relatively short (67 instead of 9 times the trunk length), and the spinules are set
on higher tubercles.

Measurements {mm.)

Length of proboscis ...

Greatest width of proboscis

Length of cephalic segment

Width of anterior cephalic lobes
Length of trunk

Width across second lateral processes
Length of abdomen

Distribution. Previously recorded from Gauss winter quarters and McMurdo
Sound.

? Pallenopsis sp. (Fig. 45).

St. 256. 23.vi.27. 35° 14' S, 6° 49' E,
850-1100 (-0) m. Young fish trawl : i im-
mature specimen. Pelagic pycnogon !

Description. Trunk distinctly
segmented; lateral processes sepa-
rated proximally by narrow intervals,
distally by rather more than their
own diameter. Cephalic segment
equal to the sum of the three
posterior segments; cephalon little
produced over base of proboscis,
occupying about half the length of
cephalic segment, scarcely wider
than long. Ocular tubercle as high as
its basal diameter, recumbent, nearly
as wide as cephalon; eyes small,
rounded and prominent, apparently Fig. 45. ? Pallenopsis, sp. Young: a. Chela: x 100. b. Ter-

reduced to one pair (Fig. 45 c). -'-' ^TT:\ T^'"' '''^ ' '" " ''°"' '"'

. , ^ 1 J r U.-1 , body with left chelophore: X 27.

Proboscis long, slender, slightly ^

expanded in middle and again at apex, almost equal in length to scape.

Abdomen long, sub-cylindrical, blunt, almost vertical.

12-2

92

DISCOVERY REPORTS

Third leg:

1-07

First coxa

0-4

0-3

Second coxa

1-2

1-8

Third coxa

0-4

1-35

Femur

3-24

0-85

First tibia ...

3-5

0-47

Second tibia

2-87

0-5

Tarsus + propodus

07

1-2

Claw

0-35

Chelophore long and slender. ^ Scape distinctly two-jointed, the first nearly twice as
long as the second segment. Chela small; fingers longer than palm, gaping when
closed (Fig. 45 o).

Palp an elongated papilla between base of chelophore and oviger.

Oviger already ten-jointed, although segments 9 and 10 are not completely separated
off from each other; reaching to tip of proboscis when straightened out. The right
oviger is bifurcated near the tip (Fig. ']\c).

Third leg long and slender. Second coxa half as long again as the sum of the first and
third. Second tibia shorter than either of the two preceding segments; first tibia the
longest segment. Tarsus very short, propodus with two prominent spines proximally
on the ventral margin. Claw long, auxiliaries absent (Fig. 45 b). There are no setae on
the trunk and only a few spinose setae near the distal end of the leg.

Measiireme?its {mm.)

Length of proboscis ...

Diameter of proboscis near tip

Length of trunk

Width across second lateral processes

Length of cephalic segment ...

Width of cephalon

Length of abdomen ...

Length of scape

Remarks. This specimen is not quite mature, no genital pores can be distinguished.
It appears to belong to the genus Pallenopsis, but does not belong to any of the species
previously recorded from South African waters. There can be no doubt that the species
is pelagic, for the total depth at the locality in which this specimen was obtained exceeds
4000 m." No pelagic species of Pallenopsis appears to have been previously described,
but Flynn (1928, p. 25) has described a pelagic species of Anoplodactyliis from off
Port Natal.

The specimen probably belongs to a small species characterized by (i) the almost
complete absence of setae, (2) the complete suppression of the auxiliary claws on the
legs, and (3) the presence of but one pair of eyes.

Genus Phoxichilidium, Milne-Edwards

Phoxichilidium australe, Hodgson.

Hodgson, 1915, p. 145.

Caiman, 1915, p. 46, fig. 10 A-D.

St. 170. 23.ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m.; R.
Large dredge : 2 +$.

St. 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 64° 48' 30" S, 63° 31' 30" W, 259-
354 m.; M. Large otter trawl: i ^.

^ Length of scape eight to nine times the distal width.

^ The specimen could not have got into the net from dredged material obtained elsewhere: for nearly
two months prior to its capture plankton nets only were in use.

ENDEIDAE 93

St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 64° 56' S, 65° 35' W, 93-126 m.; R.
Large dredge : i $, i (^, i ovigerous 3.

Distribution. Previously recorded from the Gauss winter quarters and McMurdo

Sound.

Family ENDEIDAE

Genus Endeis, Philippi
Endeis australis (Hodgson).

Phoxichihis australis, Hodgson, 1907, p. 5, pi. i, fig. i.
Phoxichilus australis, Bouvier, 1913, p. 118, text-fig. 74.
Endeis australis. Caiman, 1915, p. 49, text-fig. 11.

St. 174. 28. ii-2. iii. 27. Deception Island, South Shetlands, outside entrance W of Light, 5-10 m.
In fish trap : i ovigerous ,^.

Remarks. This specimen has been referred to Endeis australis because, although
smaller and of more compact build with shorter stouter legs than is typical for the
species, it shows no distinctive differences that might be deemed of specific importance.

The oviger appears to have only seven segments both in the Discovery specimen and
in the Terra Nova specimen for which measurements are given below (cf. Bouvier,
1913, p. 119, fig. 74).

Caiman (1915, p. 49) states that the orifices of the cement glands could not be dis-
cerned in the male specimens, but in the male from Terra Nova St. 314 there are
30-35 small whitish papillae in linear series on the proximal three-fourths of the
posterior surface of ertc// walking leg (cf. Bouvier, 19 13, p. 119: " Les orifices des glandes
a cement sont situes sur de petites saillies qui forment une rangee longitudinale a la
face posterieure des pattes des trois dernieres paires"). These papillae are also present
on all four legs in the Discovery specimen, but are fewer in number (20-25 on each

femur).

Measurements {mm.)

Length of proboscis ...

Diameter of proboscis

Length of trunk

Width across second lateral processes

Length of cephalic segment ...

Length of trunk -;- length of proboscis

Third leg:

First coxa ...

Second coxa

Third coxa ...

Femur

First tibia ...

Second tibia

Tarsus and propodus

Claw

Auxiliaries

Length of second coxa -^ length

of first and third coxae ... i'75 '"33

Length of second coxa -f- greatest

width of second coxa 3'5 ^'4

Ferra Nova

Discovery

St. 314, 1

St. 174, 3

2-93

2-0

0-8

0-67

4-0

2-93

3-0

2-4

1-33

I-I

1-36

1-47

0-8

0-6

2-8

1-6

0-8

0-6

7-2

4-0

6-8

3-4

8-2

4-53

2-2

1-9

0-9

0-67

0-42

0-4

94 DISCOVERY REPORTS

Distribution. Antarctic: (i) Ross Sea area; (2) Margaret Bay, near Doumer Island,
Palmer Archipelago, and South Shetlands.

Family AMMOTHEIDAE

Genus Ammothea, Leach

Ammothea, Leach, 1814, p. 33. Bouvier, 1913, p. 123. Caiman, 1915, p. 49.
Leionymphon, Mobius, 1902, p. 183. Hodgson, 1907, p. 39.

The total number of specimens belonging to the genus Ammothea represents a small
percentage of the entire collection, yet no fewer than five new species, each with one or
two well-defined characters, are represented. Of these A. tetrapora is unique in having
four, instead of the usual two pairs of male genital openings. Two species have each
an eight-jointed palp as in the genus Achelia (see Bouvier, 1913, p. 122, key), but they
are of much larger size, and the high transverse body ridges characteristic of the genus
Ammothea are present. Two species, both of large size, retain the chelate chelophore in
the adult {A. gigantea and A. striata).

The proboscis varies greatly in relative length and in shape from species to species.
As a rule it is straight, though much curved in A. striata, and either pyriform or sub-
cylindrical. In three different forms {A. Joiigispina, Fig. 51, A. sp..? and A. stylirostris.
Fig. 56) it is widest at the base, which is surrounded by a distinct collar, and then tapers
gradually to a rather narrow point,^ forming a long piercing organ or probe. A similar
type of proboscis is found in the genus Austroraptus, where it is much more abruptly
narrowed in the distal half.

The definition of the genus given by Hodgson and earlier writers requires considerable
modification and enlargement. The chief characters of the genus are given below,
followed by a key to the determination of the Antarctic and sub-Antarctic species.

Trunk rather compact, with distinct transverse body-ridges often greatly elevated in
the centre; lateral processes in contact or separated by intervals up to their own
diameter. Cephalon broad, neck very short or wanting; ocular tubercle situated a
short distance behind the anterior border; eyes present.

Proboscis cylindrical, pyriform or gradually tapering distally (styliform).

Chelophore chelate in immature specimens, usually, but not always, with reduced
chela in adult ; scape single-jointed.

Palp eight- or nine-jointed.

Oviger ten-jointed, without terminal claw; sometimes an irregular series of special
spines on the last four segments. In the male the seventh segment is very short, setose,
and much wider than the three terminal segments.

Legs rather stout as a rule; second tibia usually the longest segment (the femur is
longest in A. gigantea, the first tibia in A. spinosa); tarsus very short; propodus usually
armed with a few stout spines on the proximal ventral margin ; auxiliaries present. The
genital apertures occur in the male on the two posterior pair of legs (on all four legs in
A. tetrapora); in the female on all the legs.

''- This form is designated "styHform" in the key.

AMMOTHEIDAE 95

KEY TO THE DETERMINATION OF THE ANTARCTIC AND
SUB-ANTARCTIC SPECIES OF AMMOTHEA

I. Chela large and perfect in the adult as well as in immature specimens [palp nine-ji)inted ;
transverse body ridges high; species of large size].

A. Proboscis sub-cylindrical, rather slender and much curved, equal to trunk; chela
slightly longer than scape ; palpal segments 2 and 4 subequal ; second tibia the longest
segment; no large spines on ventral margin of propodus A. striata (Mobius)

B. Proboscis straight, massive and somewhat pyriform, not quite three-fourths of trunk;
chela half as long as scape ; palpal segment 4 rather longer than 2 ; femur tiie hmgest
segment; two spines on proximal ventral margin of propodus, the larger just over
one-third of main claw A. gigantea, n.sp.

II. Chela large and perfect in immature specimens, imperfect and much reduced in adults.
A. Palp eight-jointed [transverse body ridges prominent, each rising in centre to an
acutely conical process; two spines on proximal ventral margin of propodus; second
tibia the longest segment] .

I. Proboscis not quite two-thirds of trunk, bluntly conical ; chelophore half as long as,
palp rather longer than, proboscis; cephalic segment approximately equal to the
sum of the three succeeding segments ; larger spine on propodus scarcely half as long

as main claw; male genital openings present on all four legs A. tetrapora, n.sp.

[II. Proboscis slightly shorter than trunk, styliform; chelophore one-third of, palp
rather shorter than, proboscis ; larger spine on propodus half of main claw ; cephalic

segment equal to the sum of the two succeeding segments ^. sp.? St. WS 216I]

III. Proboscis rather longer than trunk, styliform; chelophore scarcely one-third, palp
not more than two-thirds, of proboscis; larger spine on propodus four-fiftiis of
main claw; cephalic segment scarcely equal to the sum of the two succeeding

segments A. longispim, n.sp.

B. Palp nine-jointed [and longer^ than proboscis].

I. Transverse body ridges prominent, each, as a rule, rising in centre to at least the
height of the ocular tubercle.

A. First tibia the longest segment; propodi of first and second markedly different
from those of the third and fourth legs [claw almost two-thirds of propodus ;
proboscis scarcely as long as trunk, sub-cylindrical with a slight median
constriction] A. spinosa {Hodgson)

B. Second tibia the longest segment; propodus similar for all four legs.

1. Palpal segments 5-8 serrated ventrally [segments 2 and 4 subequal;
proboscis at least as long as trunk, pyriform; 4-6 small spines on proximal
half of ventral margin of propodus] A. minor (Hodgson)

{A. gracilipes, Bouvier)

2. Palpal segments 5-8 cylindrical, not serrated ventrally.
a. Proboscis not exceeding three-fourths of trunk.

i. Proboscis sub-cylindrical; chelophore two-thirds of proboscis [palp
longer than proboscis, segments 2 and 4 subequal; 3-4 spines on
proximal half of ventral margin of propodus, the larger just over

one-half of claw which is two-thirds of propodus] A. calmani, n.sp.

{A. striata} Caiman)

1 This form is inserted here as the palp is apparently not quite normal and might prove to be eight-jointed.

2 Palp not quite normal— may be longer or at least equal to proboscis in A. sp.? St. WS 216.

96 DISCOVERY REPORTS

ii. Proboscis styliform ; chelophore not exceeding one-third of proboscis
[fourth palpal segment the longest].

a. Palp longer than proboscis and segment 4 almost twice as long as 2 ;
two small proximal and one median spine on ventral margin of
propodus; claw less than half of propodus; cephalic segment
almost equal to the sum of the three succeeding segments

A. stylirostris , n.sp.

)3. Palp slightly shorter^ than proboscis and segment 4 almost half

as long again as 2; two proximal spines on ventral margin of

propodus, the longer at least half of main claw which again exceeds

half of the propodus ; cephalic segment equal to the sum of the two

succeeding segments A. sp.? St. WS 216

h. Proboscis equal to or longer than trunk [straight, pyriform or sub-
cylindrical].

i. Palpal segment 2 almost half as long again as 4; numerous long, fine
hairs on walking legs; chelophore at least half as long as proboscis

with long slender scape A. meridionalis, Hodgson

ii. Palpal segment 4 equal to or longer than 2 ; setules but no hairs on
walking legs; chelophore not exceeding one-fourth of proboscis in
adults, scape short and stout [2-4 spines on proximal ventral margin
of propodus, the longest not quite half of main claw].
a. Palpal segment 4 equal to, or rather longer than, 2 (1-1-45 • i)

A. glacialis (Hodgson)

j3. Palpal segment 4 almost twice as long as 2 A. caroli?iensis, Leach^

{A. graiidis, Pfeffer)
and A. gibbosa, Mobius
H. Transverse body ridges low, without median processes [proboscis longer than
trunk, pyriform ; chelophore not exceeding one-third of, palp longer than, proboscis ;
second tibia the longest segment ; auxiliaries two-thirds of main claw, which is half
of proboscis; species of small size].

A. Abdomen sub-vertical, in contact with posterior border of third segment;
3-4 small spines on proximal ventral margin of propodus, the longest rather
more than one-third of main claw A. claiisi, Pfeffer

B. Abdomen oblique, separated by a short distance from posterior border of third
segment ; 5 small spines on proximal ventral margin of propodus, subequal and
about one-fourth of main claw A. australis, Hodgson

Ammothea striata (Mobius) (Fig. 47 b).

Leionymphon striatum, Mobius, 1902, p. 183, pi. xxvi, figs. 7-12.

Ammothea striata, Bouvier, 1913, p. 124, figs. 75-77 (nee A. striata'^. Caiman, 1915, p. 55).

St. 167. 20. ii. 27. Off Signy Island, South Orkneys, 60° 50' 30" S, 46° 15' W, 244-344 '^■'■>
gn. M. Large otter trawl: i S, largely overgrown with Polyzoa and Alcyonidium.

Remarks. This species retains the chelate chelophore in the adult, as figured by
Bouvier (1913, p. 125, fig. 75).^ The scape is not quite half the length of the proboscis

^ Palp not quite normal — may be longer or at least equal to proboscis in A. sp.? St. WS 216.

" There is very little difference between A. carolinensis and A. gibbosa (see Bouvier, 1913, pp. 123, 129
and Caiman, 1915, p. 52). A. clausi and A. australis also are very similar; future investigations may result
in uniting the two forms in either or both of these cases.

^ See Caiman, 1915, p. 56, on this point; the specimen is mature.

AMMOTHEIDAE

97

and 2-33 times as long as wide distally. The chela is slightly longer than the scape, the
fingers curved and gaping when closed.

The proboscis is markedly curved and at least as long as the trunk if allowance is
made for the curvature.

A. striata is most nearly related to A. gigantea, n.sp., which also retains the chelate
chelophore in the adult. It is easily distinguished from the latter by (i) the rather
slender curved proboscis ; (2) the more massive chelophore, with chela even longer than
the scape; (3) the absence of large spines on the propodus; and (4) the longer second
tibia (see key, and table on p. 99).

Measurements (mm.)

Third leg:

Length of proboscis i5'° +

Diameter of proboscis ... ... 2-5

Length of trunk ... ... ... i5"0

Length of cephalic segment 7-4

Width of cephalic lobes ... ... 6-7

Width across second lateral processes 13-3
Length of abdomen ... ... ... 4"°

Distribution. South of Bouvet Island, South Shetlands and South Orkneys.

Ammothea gigantea, n.sp. (Figs. 46 and 47 a).

St. 371. 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large otter
trawl: i ovigerous <S (holotype); i c?, overgrown with Polyzoa, Tunicates and Hydroids, i ?.

Description of holotype. Trunk rather compact, lateral processes separated by
narrow intervals— not exceeding half of their own diameter. Transverse body-ridges

First coxa

2-7

Second coxa

8-0

Third coxa

3-0

Femur

26-6

First tibia

23-0

Second tibia

30-2

Tarsus and propodus

7-5

Claw

3.0

Auxiliaries

1-2

Fig. 46. Ammothea gigantea, n.sp. Lateral view of body with chelophore, palp and oviger: x 2.
prominent, rising in the middle Une above the level of the ocular tubercle to blunt,
rounded apices. Cephalic segment equal to the sum of the two succeedmg segments.
Ocular tubercle small, scarcely higher than wide, tapering abruptly to a pomt above the
level of the eyes which are of approximately equal size (Fig. 46).

13

98

DISCOVERY REPORTS

Proboscis very robust, almost three-fourths of trunk ; maximum width near the apex

(Fig. 46).

Abdomen reaching a short distance beyond fourth lateral process, slightly elevated.

Chelophore very well developed in all three adult specimens, almost as long as
proboscis. Scape long and rather stout, length approximately four times the distal
width. Chela short and thick set, length of palm twice the width ; fingers widely gaping
when closed.

Palp nine-jointed^ as in the majority of the species of Amniothea; fourth segment
rather longer than second (1-24 : i); seg-
ments 1-4 together as long as proboscis;
terminal segments rather setose.

Ovigers each carrying a large mass of de-
veloping ova and larvae; segments 6-10 of
the usual type ; segments 2-4 subequal and
each slightly shorter than segment 5.

Third leg long and siont; femur the longest
segment; second coxa slightly longer than
the sum of first and third. Terminal seg-
ments as represented in Fig. 47 a and
similar in all four limbs.

Setides arranged in distinct longitudinal
bands on the legs; they also occur on the
abdomen, the transverse trunk ridges, lateral
processes and chelophores.

The female shows the usual sexual differ-
ences of oviger and walking legs ; the femur is more robust, being scarcely more than
seven times as long as wide as against nine times in the male. The setules are also fewer

in number.

Measurements [mm.)

Fig. 47. Terminal segments of third leg of: a.
Ammothea gigantea, n.sp.: x 6. b. A. striata
(Mobius): x 8.

J (holotype)

?

Length of proboscis ...

167

19-2

Greatest width of proboscis ..

6-9

7-9

Length of trunk

21-8

23-0

Length of cephalic segment ..

9-2

10-4

Width of anterior cephalic lob

es ... 7-8

7-8

Width across second lateral processes 18-9

19-2

Length of abdomen ...

6-9

7-3

Third leg:

First coxa

5-0

44

Second coxa

10-6

ii-o

Third coxa

4-2

5-4

Femur

34-0

41-2

First tibia

29-4

32-6

Second tibia

32-3

36-8

Tarsus and propodus

8-8

9-2

Claw

S-2,

5-3

Auxiliaries ...

1-5

1-6

1 Measurements of segments of palp: 2-4; 5-8; 3-5; 7-2; 2-4; 2-4; 2; i-8; 2-4mm,

AMMOTHEIDAE

99

Remarks. This species is the largest that has been described up to the present.

It is most closely allied to Ammothea striata, Mobius ; in both forms the chelophore
is large and the chela is perfect right up to the adult condition. The chief differences
between the two species are listed in the following table.

Ammothea striata

Proboscis long, slender, curved; 6 times
as long as wide and as long as trunk

Scape subequal to chela and less than half

of proboscis

Second tibia the longest segment

No large spines on proximal ventral

margin of propodus, which is at least

twice as long as claw

Ammothea gigantea

Proboscis stout, straight, 2\ times as long
as wide and not exceeding four-fifths of
trunk

Scape twice as long as chela and three-
fourths of proboscis
Femur the longest segment
Two spines on proximal ventral margin
of propodus w-hich is only half as long
again as claw

Ammothea tetrapora, n.sp. (Figs. 48, 49 and c^oa).

St. 51. 4. V. 26. Off Eddystone Rock, East Falkland Island, from 7 miles N 50" E to 7-6 miles
N 63° E of Eddystone Rock, 105-115 m.; f. S. Large otter trawl: 2 $$, 2 ??.

St. WS244. iS.vii. 28. 52'00'S, 62° 40' W, 253-247 m.;f.d. S.M. Commercial otter trawl :
I larvigerous ^ (holotype), i ?.

Description of holotype. Trunk compact, lateral processes in contact or separated
by very narrow intervals proximally. Transverse body ridges prominent, each with a
high, narrow, conical median projection.
Cephalic segment considerably expanded
anteriorly, not quite as long as the sum of
the three posterior segments. Ocular tu-
bercle stout, scarcely higher than wide,
tapering to a point above the level of the
eyes ; anterior much larger than the pos-
terior eyes (Fig. 48).

Proboscis straight; directed obliquely
downwards; widest at the base, then
sub-cylindrical, narrowing distally to a
blunt rounded apex; slightly more than
half as long as trunk.^

Abdomen elevated at an angle of about 45^ and reaching to distal end ot fourth lateral

Droccss

Chelophore. Scape short, club-shaped; distal width half the length. Chela small;
fingers reduced and no longer capable of meeting distally.

Palp with only eight segments in all specimens; second approximately two-thirds
of fourth segment. The reduction in the number of segments may be due to fusion of

Fig. 48. Ammothea tetrapora, n.sp. Female. Lateral
view of body with chelophore and palp: x 6.

1 Measured dorsally.

13-2

loo DISCOVERY REPORTS

the fourth and fifth joints (Fig. 49 b), as the longest segment sometimes has a sUght
constriction near the distal end but never any trace of a suture.

Oviger. The ovigers are hidden almost completely by the two clusters of larvae ; one
from another male is represented in Fig. 49 a.

Third leg. Second coxa rather longer than the sum of the first and third. Femur
subequal to first and a little shorter than second tibia ; approximately six times as long
as wide. Terminal segments as represented in Fig. 49 a ; similar in all four legs. Setules

Fig. 49. Ammothea /efn/pora, n.sp.: a. Oviger of male. b. Palp.
(Both X 20.)

restricted almost entirely to the dorsal surface of the longer segments, with a single
row on each lateral and ventral surface. The genital openings are present on the second
coxae of oil four legs.

The female, apart from the usual sexual differences, agrees with the male. The femur
is only four times as long as wide.

Remarks. The smaller specimens from St. 5 1 (measuring 6-6-8 mm. in length) all
have conspicuous genital openings, but the fingers of the chelae are rather longer than
in the holotype and the legs are relatively shorter and stouter. There are four pairs of
genital openings in each male specimen ; in the other species of Ammothea the genital
openings are only present on the second coxa of the last two pairs of legs.

This species agrees with A. longispina, n.sp., in the number of palpal segments, but

AMMOTHEIDAE

it is a smaller form with a shorter, more cylindrical proboscis, a longer palp and much
shorter spines on the propodus (see key).

Measurements (mm.)

Length of proboscis .. .

Proximal width of proboscis

Length of trunk ...

Length of cephalic segment
Width of anterior cephalic lobes
Width across second lateral processes

Length of abdomen ...

Length of scape

Greatest width of scape

Third leg:

First coxa . . .

Second coxa

Third coxa

Femur

First tibia . . .

Second tibia

Tarsus and propodus

Claw

Auxiliaries ...

(holotype)

4-8

2-0

8-0
3-6

2-2

7-47
2-8

2-1
I-O

1-6

3-3
0-8

9-4
9-4

10-2
3-2

1-5
0-6

(St. 51)
3-6

2-0

6-8

3-4

2-43

6-0

2-0

I -8

0-9

1-47

2-5
i-o
6-4
6-4
6-8
2-75

1-2

0-53

?(St.

WS244)

4-8

2-0

8-0
4-0
2-4
6-8

2-75

2-2
1-0

1-6
3-2
1-0

9-2

8-8

lO-O
3-2

1-33

0-5

Ammothea longispina, n.sp. (Figs. 50 ft, 51 and 52).

St. 170. 23. ii. 27. Off Cape Bowles, Clarence
Island, 61" 25' 30" S, 53° 46' W, 342 m. ; R. Large
dredge: i $ (holotype).

Description of holotype. Trunk rather
compact; lateral processes separated by
narrow intervals, last two pairs by almost
their own diameter. Transverse body ridges
prominent, each with a narrow, conical
median projection. Cephalic segment with
a blunt rounded lobe external to the inser-
tion of each chelophore ; hardly as long as
the sum of the two succeeding segments.
Ocular tubercle rather higher than wide,
stout, rounded at apex; eyes set near apex,
of small size but the anterior rather larger
than the posterior pair (Fig. 51).

Proboscis longer than trunk, straight,
slender ; widest in proximal third narrowing
to almost half this width anteriorly; ap-
parently suited for probing.

Abdomen rather long but much elevated, so that it does not reach beyond the distal
end of the fourth lateral process.

Fig. 50. Terminal segments of third leg of: a. Am-
mothea tetrapora, n.sp. b. A. lonaispina, n.sp. (Both
X 20.)

DISCOVERY REPORTS

Chelophore almost one-third of proboscis ; scape twice as long as wide distally ; chela
small, fingers almost obsolete.

Palp much shorter than proboscis, with only eight segments; second subequal to
fourth segment (Figs. 51 and 52 a).

Ovi^er as represented in Fig. 52 b.

Fig

51. Ammoihea longispina, n.sp. Holotype. Lateral
view of body with chelophore and palp: •; 3.

Fig. 52. Ammothea longispina, n.sp. Holotype:
a. Palp: 15. 6. Oviger: xii.

Third leg rather short and stout. Second coxa approximately equal to the sum of the
first and third. Femur five times as long as wide, slightly shorter than second tibia.
Terminal segments as represented in Fig. 50 b ; distal spine on propodus unusually
large, the smaller spine wanting on the propodus of the first leg.

Setules on trunk restricted to the conical processes of each transverse ridge, the two
low cephalic protuberances and the distal border of each lateral process. On the legs
they are arranged in longitudinal rows on the longer segments ; on the dorsal border of
each tibia they are replaced by small, curved spines.

Measurements {mm.)

Third leg:

Length of proboscis ...

13-3

First coxa

2-8

Greatest width of proboscis ...

2-4

Second coxa

4-6

Length of trunk

11-6

Third coxa

2-0

Length of cephalic segment

4-9

Femur

14-0

Width of anterior border of cephalon

3'9

First tibia ...

12-2

Width across second lateral processes

lO-O

Second tibia

15-0

Length of abdomen ...

5-6

Tarsus and propodus

4-8

Length of chelophore

4-0

Claw

2-8

Auxiliaries

0-8

AMMOTHEIDAE

103

Remarks. Each palp seems to be quite perfect and there is no distinct evidence of
fusion of any two segments. This species is much larger than A. tetrapora and is
characterized by (i) the long, tapering proboscis, (2) the relatively short palp (in
A. tetrapora the palp is longer than the proboscis), (3) the ver}' large spine on the
propodus, and (4) the blunt antero-lateral cephaHc lobes.

Ammothea spinosa (Hodgson) (Fig. 53).

Leionymphon spinosum, Hodgson, 1907, p. 49, pi. vii, fig. 2.
Ammothea spinosa, Bouvier, 1913, p. 123 (in key).
Ammothea spinosa, Caiman, 191 5, p. 52.

St. WS215. 31.V. 28. 47° 37' S, 60° 50' W, 219-14601.; f.gn. S. Commercial otter trawl : I (J.
St. WS 245. 18. vii. 28. 52° 36' S, 63° 40' W, 304-290 m. ; d. gn. S. Sh. Commercial otter trawl :
I ovigerous S-

Remarks. There can be no doubt that these tvv^o specimens belong to A. spinosa
although they are more compactly built, with much smaller non-recurved "tubercular
processes" on each lateral process, than
in the holotype. Also the fine, silky hairs
on body and legs are very sparse.

One of the most outstanding charac-
teristics of the species, namely the varia-
tion in the terminal segments of the
walking legs, has not been adequately
described by Hodgson (1907, p. 50). In
all specimens examined the two anterior
pairs of legs each have the propodus rela-
tively short and stout, with numerous
spines on the ventral margin arranged in
the manner represented in Fig. 53 a. The
propodus of the third or fourth leg, on
the other hand, is relatively slender with
6-7 small spines on the proximal half of
the ventral margin and none on the distal

Fig. 53. Ammotliea spinosa (Hodgson). Holotype:
terminal segments of a, first leg; b, fourth leg: X15.

half (Fig. 53 b). The first tibia is longer than the second. In these two respects A. spinosa
differs from all the other Antarctic species of the genus.
Distribution. Ross Sea and Magellan District.

Ammothea minor (Hodgson).

Leionymphon minus, Hodgson, 1907, p. 44, pi. vi, fig. 2.
Ammothea minor, Bouvier, 1913, pp. 123, 131, figs. 83, 84.
Ammothea gracilipes, Bouvier, 1913, p. i32> figs- 85-87.
Ammothea minor. Caiman, 191 5, p. 52.
Ammothea {Leonymphon) minor, Loman, 1923, p. 23.

I04 DISCOVERY REPORTS

St. 27. 15. iii. 26. West Cumberland Bay, South Georgia; 3-3 miles S 44° E of Jason Light,
no m.; M. R. Large dredge: i c?-

St. 42. i.iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of
Jason Light to 4 miles N 39° E of Jason Light, 120-204 m.; M. Large otter trawl: i immature
specimen.

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54° E of
Larsen Point, to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl : i im-
mature specimen.

St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 64° 56' S, 65° 35' W, 93-130 m.;
St. M. R. Large dredge: 1 ? with a few Foraminifera and Polyzoa on legs.

St. 371. 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large
otter trawl: i immature specimen bearing some Polyzoa and Foraminifera.

Remarks. The specimens agree well with Bouvier's description of A. gracilipes.

Distribution. This species is recorded from both sides of the Antarctic — the Ross
Sea and the region around South Georgia.

Ammothea calmani, n.sp. (Figs. 54 and 55).

A. striata}. Caiman, 1915, p. 55.

St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 64° 20' S, 63° 01' W, 160-335 m.;
m. St. Large otter trawl: i immature specimen with large chelophores.

This immature specimen belongs to the same species as the large male referred, with
considerable doubt, to A. striata by Caiman (1915, p. 55). The aduh specimen may be
regarded as the holotype.

Description of immature specimen. Trimk with second to fourth lateral processes
separated by their own diameter. Transverse body ridges prominent but not greatly
elevated in the middle line. Cephalic segment equal to the sum of the two succeeding
segments. Ocular tubercle a little higher than wide, apex bluntly pointed and on a level
with the median transverse ridge. Anterior very much larger than the posterior pair
of eyes (Fig. 54).

Fig. 54. Ammothea calmani, n.sp. Young specimen.
Lateral view of body with chelophore and palp: x 5.

AMMOTHEIDAE

105

Proboscis short, sub-cylindrical; length almost three times the width and approxi-
mately equal to the sum of the two anterior segments of trunk.

Abdomen elevated at an angle of about 45°, reaching a little beyond fourth lateral
process.

Chelophore slightly longer than proboscis ; scape longer than chela and rather slender
(length 2-5 times the distal width).

Palp considerably longer than proboscis, with nine segments, of which the second
and fourth are subequal (Fig. 55 a).

Fig. 55. Ammothea calmani, n.sp. : a. Palp of young specimen, h. Terminal segments of oviger of holotype.
c. Terminal segments of third leg of holotype. ^. Same of young specimen, (a and*: xi5;candrf: x 10.)

Oviger small but with all ten segments distinct.

Third leg rather slender; femur 5-5 times as long as wide; second tibia the longest
segment. Propodus with two large and one small spine on the proximal ventral margm
(Fig. 55 d); the terminal segments are very similar in all four legs.

Measurements {mm.)

Length of proboscis ...

Width of proboscis

Length of trunk

Length of cephalic segment

Width of anterior border of cephalon
Width across second lateral processes

Length of abdomen

Length of scape
Length of chela

5-8

2-2
lO-S

4-2

3-1

7-9
3-6

4-0
3-0

Third leg :

First coxa

Second coxa
Third coxa ...
Femur

First tibia

Second tibia
Tarsus and propodus

Claw

Auxiliaries

1-8

4-0

1-8

ii-o

10-7

I2-0
2-4

0-8

14

io6 DISCOVERY REPORTS

Remarks. The adult male is not in a sufficiently good condition to give accurate
measurements. It agrees with the immature specimen in most respects. The transverse
body ridges on the second and third segments are more massive and are higher than the
ocular tubercle. The chela is greatly reduced so that the chelophore is shorter than the
proboscis. A fourth spine is present on the propodus of each walking leg (Fig. 55 c)
and the setules are arranged in distinct longitudinal bands on the three long segments.
Segments 2-6 of the oviger measure 4-4, 3-6, 4-0, 4-4 and 2-5 mm. respectively; the
terminal segments are represented in Fig. 55 Z). The right palp is sUghtly abnormal near
the tip ; segments 7 and 8 seem to be fused together.

This species appears to be related to A. stylirostris, from which it can easily be
distinguished by (i) the shape of the proboscis, (2) the longer claw and much larger
spines on the propodus, (3) the larger chelophore, and (4) the much shorter fourth palpal
segment.

Distribution. Palmer Archipelago and off Oates Land.

Ammothea stylirostris, n.sp. (Figs. 56 and 57).

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock, 179-235 m.; gy. M. Large otter trawl: i ? (holotype) and
I immature specimen with well-developed chelae.

Description of holotype. Trunk of rather compact build ; intervals between lateral
processes increasing posteriorly. Transverse body ridges rather prominent, each rising

Fig. 56. Ammothea stylirostris, n.sp. Holotype. Lateral view of body
with chelophore and palp: x 5.

in the centre to a low, rounded point. Cephalic segment nearly as long as the sum of
the three posterior segments, considerably expanded on a level with the ocular tubercle ;
a low rounded lobe projecting over the base of each scape. Ocular tubercle almost half
as high again as wide, tapering abruptly to a point distally. Anterior twice as large as
the posterior pair of eyes (Fig. 56).

Proboscis approximately two-thirds as long as trunk, straight, widest at base, which
is surrounded by a prominent collar, and gradually narrowing distally; presumably
adapted for probing.

AMMOTHEIDAE i°7

Abdomen elevated at an angle of about 45" and reaching to distal end of fourth lateral
process.

ChelopJiore short, less than one-third of proboscis ; scape short, clubbed ; chela small,
with fingers reduced to small blunt stumps.

Palp rather longer than proboscis, with nine segments (Fig. 57 b) ; fourth almost
twice as long as second segment.

Oviger as represented in Fig. 57 a.

Fig. 57. AmmotJiea stylirostns, n.sp. Holotype: rt. Oviger: ;<i5. 6. Palp: xis.
c. Terminal segments of third leg: x 11.

Third leg relatively short and stout. Second coxa equal to the sum of the first and
third. Femur almost five times as long as wide; second tibia the longest segment;
terminal segments as represented in Fig. 57 c. Setules arranged in longitudinal rows

on the longer segments.

Measurements {mm.)

Length of proboscis ...

Basal width of proboscis

Length of trunk

Length of cephalic segment ..

Greatest width of cephalon ..

Width of anterior border of cephalon

Width across second lateral processes

Length of abdomen ...

Length of chelophore

7-8
3-6

II-2

5-4
4-6

3-4
8-4
3-6

2-2

Third leg:

First coxa

Second coxa
Third coxa ...
Femur

First tibia

Second tibia
Tarsus and propodus

Claw

Auxiliaries ...

2-27

4-0

173

1 1 -2

10-6

12-8

4-6
1-8

0-75

Remarks. The immature specimen measures 8-5 mm. in length (trunk only) and is
very soft ; the chelophores are half as long as the trunk, with perfect chelae.

The proboscis is rather similar to that of A. longispina, n.sp., but is only two-thirds

» IA-2

14-2

io8

DISCOVERY REPORTS

as long as the trunk ; the palp, moreover, has nine instead of eight segments ; the spines
on the ventral margin of the propodus are much smaller and the claw is much shorter.
The specimen described as Amtnothea, sp. ? (WS 216) is nearly related to this species ;
but one palp is undergoing regeneration distally and the other does not appear to be
quite normal as regards the terminal segments (Fig. 58 b). The differences between the
two forms are given in the key.

Ammothea, sp..? (Fig. 58).

St. WS 216. I. vi. 28. 47° 37' S, 60° 50' W, 219-133 m.; f. S. Commercial otter trawl: i ?.

The single female specimen bears a strong superficial resemblance to the holotype of
Ajnmothea longispina but is of much smaller size. The proboscis is of the same type,
although it is not quite as long as the trunk. The right palp has nine segments ^ and the
fourth is distinctly longer than the second segment. Unfortunately the left palp has
been damaged and the tip is undergoing regeneration; the sixth segment is about to
bifurcate and a seventh segment is present on the inner antero-lateral angle. There are

Fig. 58. Ammothea, s^.} St. WS 216: «. Oviger: X15. 6. Right
palp: X15. f. Terminal segments of third leg: xii.

no prominent rounded lobes on the anterior margin of the cephalon and the anterior
spine on the propodus is much shorter than in A. longispina (Fig. 58 c, cf. Fig. 50 b).
The oviger is represented in Fig. 58 a.

The specimen appears to be mature, for the genital apertures are large and con-
spicuous. If the palp is normally nine-jointed this species would not be the same as
A. stylirostris. According to Bouvier's key it is nearest to A. glacialis (Hodgson), but
the proboscis is strikingly different in the two forms.

1 The terminal segments of this palp also may have been regenerated (see Fig. 58 b).

AMMOTHEIDAE

109

Length of proboscis ...
Basal width of proboscis
Length of trunk

Length of cephalic segment

Width of anterior border of cephalon
Width across second lateral processes
Length of abdomen ...
Length of chelophore

Measurements {mm.)

Third leg:
First coxa ..
Second coxa

6-8

2-0

7-6
3-2

2-53

6-4

3-07

2-2

Third coxa ...

Femur

First tibia ...

Second tibia

Tarsus and propodus

Claw

Auxiliaries ...

1-4

3-2

1-2

9-8

8-2

lO-O

3-2

1-55
0-6

Ammothea carolinensis, Leach (= Ammothea grandis, PfefFer).

A. carolinensis. Caiman, 1915 a, pp. 310-314, text-figs. 1-3 (references on p. 314, that of

Hodgson, 1907, excepted); 1915, p. 52.
Leionymphon grande, Hodgson, 1908, p. 179.

St. 141. 29. xii. 26. East Cumberland Bay, South Georgia, 200 yards from shore, under Mount
Duse, 17-27 m.; M. Small beam trawl: i immature specimen.

St. 145. 7. i. 27. Stromness Harbour, South Georgia, between Grass Island and Tonsberg Point,
26-35 m. Small beam trawl: i ovigerous ^.

S.S. Carl. i.iv. 25. Brought up on anchor in Maiviken, South Georgia: i specimen, soft,
probably a slightly immature ?.

Sappho Pt. 2. iv. 26. Grytviken, South Georgia, washed up on beach: i ?, overgrown with
Polyzoa.

Shore Collection. 25. ii. 26. Cumberland Bay, South Georgia, near Green Point: i ?, overgrown
with Polyzoa on proboscis and legs.

Distribution. Western Antarctic— South Georgia and farther south.

Ammothea clausi, Pfeffer.

Bouvier, 1913, p. 135, text-figs. 88, 89 (for references and synonymy).
Loman, 1923, p. 23.

St. 141. 29. xii. 26. East Cumberland Bay, South Georgia, 200 yards from shore, under Mount
Duse, 17-27 m.; M. Small beam trawl: i ?.

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W, 342 m.; R.
Large dredge : 2 ??.

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Shetlands, 62° 07' S, 58° 28' 30" \V,
391 m.; M. St. Medium otter trawl: 3 ?? (one with 4 egg capsules (?) attached to proboscis),
2 immature specimens.

WS25. 17. xii. 26. Undine Harbour (North), South Georgia, 18-27 m.; M. S. Small beam
trawl : 2 ?$.

WS 62. 19. i. 27. Wilson Harbour, South Georgia, 15-45 m. Small beam trawl: i ?.

MS 62. 24. ii. 26. East Cumberland Bay, i cable E to 3I cables S of Hobart Rock, 31-4° m-
Small beam trawl : i ?.

MS 67. 28. ii. 26. East Cumberland Bay, 3 cables NE of Hobart Rock to \ cable W of Hope
Point, 38 m. Small beam trawl : i immature specimen.

MS 68. 2. iii. 26. East Cumberland Bay, 17 miles S i E to 81 cables SE x E of Sappho Point,
220-247 m. Large rectangular net: 2 $$, I (J.

DISCOVERY REPORTS

Shore Collection. 2. iv. 26. South Georgia, Grytviken, Sappho Point, i ?, washed up on shore.
St. 371. 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large otter
trawl: i specimen.

Remarks. It is possible that Ammothea australis may prove to be only an eastern
form of A. clansi. At any rate the differences between the two are very slight.

Genus Achelia, Hodge
Hodge, 1864, p. 114.

Bouvier, 1913, p. 122 (in key), pp. 138-140.
Caiman, 1915, p. 56.
Loman, 1923, p. 24.

This genus differs from Ammothea in (i) the much smaller size of the species, (2) the
absence of transverse body ridges, and (3) in the type of the male oviger (cf. Figs. 49 a,
55 b and 59 a and b). The tubercles that bear the male genital openings and the openings
of the femoral cement glands are not always very prominent (e.g. in A. hoekii, Fig. 60 a).

Achelia hoekii (Pfeffer)
(Figs. 59 a, b and 60 a, c).

Ammothea Hoekii, Fiefier, 1889,

Jahrb. Hamburg. Wiss. Anst.,

VI, Heft 2, p. 46.
? Ammothea {Achelia) Hoekii,

Loman, 1923, p. 24.

St. 174. 28. ii.-i. iii. 27. Decep-
tion Island, South Shetlands, outside
entrance, W of Light, 5-10 m. Large
fish trap: 2 ??, i $.

Description. Trunk com-
pact, lateral processes in contact ;
sutures between the segments
distinct in two specimens — the
last suture is rather faint in one
female. Cephalic segment sub-
equal to the sum of the three
posterior segments ; cephalon
broad; ocular tubercle wider
than high, bluntly rounded dis-
tally.

Proboscis equal to trunk, wid-
est in the middle and bluntly
conical both proximally and

^ ■ Fig. 59. Achelia hoekii {PfeHer) : a. Segments 4-10 of male oviger : x 47

^Oaomew rather short, reach- and terminal segments still further enlarged, b. Palp of male: x 60.
ing to distal end of first coxa. A. A>mmums, Bouvier. c. Oviger of male: x 60.

AMMOTHEIDAE in

Chelophore scarcely half of proboscis ; scape twice as long as the greatly reduced chela.

Palp with the terminal segments distinctly serratiform ventrally in the male (Fig.
59 b), rather less so in the female ; segments 2 and 4 subequal.

Oviger in the male relatively more robust than that of Achelia communis, with the
terminal segments much more twisted (cf. Fig. 59 a and c). The prominent spine near
the proximal articulation of segment 6 is very characteristic (Fig. 59 a).

Third leg of male as represented in Fig. 60 a ; the male genital opening is situated
on a much lower tubercle than in A. communis and there is no prominent cement gland-
tubercle at the distal end of the femur (cf. Fig. 60 a with Bouvier, 1906, p. 48, fig. 32).

Fig. 60. Achelia hoekii {?ie.StT): a. Third leg of male: x 20. c. Terminal

segments of third leg: x 53.
A. communis Bouvier: b. Terminal segments of third leg: x 60.

The propodus is higher and more arched, with a distinct heel beset with a number of
short stout spines (cf. Fig. 60 b and c). There are no prominent conical tubercles on
the first coxae in the female, and just a hint of them in the male.

Remarks. The co-types of Achelia hoekii were examined in the Zoological Museum,
Hamburg. The Discovery specimens agree in all respects with sketches made of a male
and of a female co-type. Loman (1923, p. 24) states that A. communis, Bouvier, is
co-specific with A. hoekii. Specimens of the former, presented by Professor Bouvier
to the British Museum, differ in several respects from the types of A. hoekii. The
differences (described and figured above) in (i) oviger, (2) walking legs, seem to justify

112

DISCOVERY REPORTS

the separation of A. communis from A. hoekii. In the latter the palp is more serrated
distally and the antero-lateral processes on the cephalon are absent.
Distribution. South Georgia and South Shetlands.

Achelia intermedia, Caiman (Fig. 6i).
Caiman, 1915, p. 60, fig. 15 A-C.

St. 371. 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large otter
trawl: 6 $$, i $, with A. briicei.

? St. WS 27. 19. xii. 26. 53° 55' S, 38" 01' W, 107 m. i m. horizontal tow-net— net touched
bottom : i $.

Remarks. The specimens from St. 371 agree in most respects with the Terra Nova
specimens of Achelia intermedia. The walking legs, however, are relatively longer and
more slender (see Fig. 61 a and b, and the ratios given below).

Femur: length ~ maximum width
First tibia : length -I- maximum width
Second tibia: length ~ maximum width

Terra Nova

<? 9

4-S 3-3

5-0 475

6-83 6-66

Discovery

(? S

575 475

6-25 7-0

8-89 9-5

Fig. 61. Achelia intermedia, Caiman: a. Third leg of female — Discovery
collection. 6. Same — Terra Nova collection. (Both x 13.)

The specimen from WS 27 is rather different from all the others. The chelophore
is longer, exceeding the second palpal segment, but is not quite half of the proboscis^n
this respect the specimen approaches Achelia spicata, Hodgson (Caiman, 191 5, p. 57).
On the other hand the main claw is relatively shorter and only twice as long as the
auxiliaries — in this respect the specimen approaches Achelia brucei. The ocular tubercle,
however, is longer than wide, terminating in a long acute point. This specimen is another
aberrant or "abnormal " form, agreeing with the species in group a of Caiman's key in
most respects, but approaching those in group b as regards the auxiliary claws of the
walking legs (Caiman, 191 5, p. 57).

Distribution. Previously recorded from off Cape Adare.

AMMOTHEIDAE "3

Achelia brucei, Caiman.

Caiman, 1915, p. 61, fig. i6 A-C.

St. 141. 29. xii. 26. East Cumberland Bay, South Georgia, 200 yards from shore, under Mount
Duse, 17-27 m. ; M. Small beam trawl: i ?.

St. 371. 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large otter
trawl: 4 ?$ with A. intermedia.

Distribution. Previously recorded from off Cape Adare.

Achelia serratipalpis, Bouvier.

Bouvier, 1913, p. 140, figs. 90-95.
Loman, 1923, p. 25.

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Shetlands, 62° 07' S, 58^ 28' 30" W,
391 m. ; M. St. Medium otter trawl : 2 ??, 5 ,^S (3 ovigerous), with NympJion charcoti.

?St. 274. 4. viii. 27. Off St Paul de Loanda, Angola, from 8' 40' 15" S, 13° 13' 45" E to
8° 38' 15" S, 13° 13' E, 64-65 m.; gy. M. Large otter trawl: i incomplete ?.

Distribution. The incomplete female specimen from St. 274 undoubtedly belongs
to this well-defined species, which has been recorded only from Antarctic waters. There
appears to be little but the exoskeleton in this instance and it is probable that the
specimen was captured in the Antarctic but may have been left entangled in the net for
some time. At any rate, the material is not sufficient to warrant the supposition that
A. serratipalpis extends from Graham Land, the South Shetlands and South Georgia
northwards into tropical waters.

Achelia parvula, Loman (Fig. 62).
Loman, 1923 a, p. 2, fig. A.

Fig. 62. Achelia parvula,-Lomzn: fl.Palp: x 60. 6. Third leg of male: x 47.
c. Proximal segments of third leg of female: x 47.

15

114 DISCOVERY REPORTS

St. 53. 12. V. 26. Port Stanley, East Falkland Islands, hulk of "Great Britain", 0-2 m. Mussel
rake: (i) numerous specimens\ "sorted from washings from Hydroids and Mytilus clumps".
{b) 2 specimens, "from washings from Kelp root".

St. 58. 19. V. 26. Port Stanley, East Falkland Islands, 1-2 m. Mussel rake : 4 specimens.

Remarks. These specimens agree well with Loman's description of A. parviila. As
Loman's figures are rather poor the palp and third leg are represented again (Fig. 62).

Distribution. Previously recorded from Possession Bay, East Magellan Strait.

Genus Austroraptus, Hodgson
Hodgson, 1907, p. 54.
Caiman, 1915, p. 62.

Each of the three known species of this genus is represented in the present collection.
They may be distinguished from each other as follows :

I. Palp six-jointed in adult.

A. Chela imperfect in adult ; scape at least twice as long as wide, without dorsal tubercles ;
antero-lateral tubercles on cephalon obsolete A. poIaris, Hodgson

B. Chela perfect in adult ; scape not more than half as long again as wide, with 1-2 dorsal
tubercles; antero-lateral tubercles on cephalon present A.praecox, Caiman

II. Palp eight-jointed [chela imperfect in adult] A. juvenilis. Caiman

Austroraptus polaris, Hodgson.

Hodgson, 1907, p. 54, pi. viii, fig. 2.
Caiman, 1915, p. 62, fig. 17 A-C.
Loman, 1923, p. 30.

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53' 46' 00" W, 342 m.; R.
Large dredge : i $.

Distribution. This species has been recorded from the eastern and the western
side of the Antarctic Zone.

Austroraptus praecox, Caiman.

Caiman, 1915, p. 65, fig. 19 A-E.

St. 144. 5. i. 27. OflF mouth of Stromness Harbour, South Georgia, from 54° 04' S, 36" 27' W
to 53° 58' S, 36° 26' W, 155-178 m. ; gn. M. S. Large otter trawl: 2 SS-

Distribution. Cape Adare and South Georgia.

Austroraptus juvenilis, Caiman.

Caiman, 1915, p. 63, fig. 18 A-E.

St. 42. I . iv. 26. OflF mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of Jason
Light to 4 miles N 39° E of Jason Light, 120-204 m.; M. Large otter trawl: i 9-

1 None of the males are carrying ova or larvae.

AMMOTHEIDAE iiS

St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, 155-178 m.;gn. M. S. Large
otter trawl: 2 <$,S, with A. praecox.

St. 195. 30. iii. 27. Admiralty Bay, King George Island, South Shetlands, 62° 07' 00" S,
58° 28' 30" W, 391 m.; M. St. Medium otter trawl: i ?, i (J, with Nymphon charcoti.

St. 363. 26. ii. 30. 2-5 miles S 80' E of SE point of Zavodovski Island, South Sandwich Islands,
329-278 m.; Sc. Large dredge: i ^, bearing ? egg capsules or encysted Protozoa.

St. 371 . 14. iii. 30. I mile E of Montagu Island, South Sandwich Islands, 99-161 m. Large otter
trawl: i <^.

Remarks. In the male from St. 363 large numbers of minute oval bodies are adhering
to the appendages. These may be egg capsules or encysted Protozoa ; they are found,
in smaller numbers, attached to other Pycnogonida, e.g. Tanystylum pfejferi (Fig. 68 Z)).

Distribution. Previously recorded from off Cape Adare.

Genus Austrodecus, Hodgson
Austrodecus glaciale, Hodgson (Figs. 63 and 64).

Hodgson, 1907, p. 53, pi. viii, fig. i.
Bouvier, 1913, p. 147, figs. 96 and 97.
Caiman, 1915, p. 66, fig- 20.
Loman, 1923, p. 31.

St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, 3-3 miles S 44" E of Jason Light,
no m.; M. R. Large dredge: 7 ??, 15 i^.

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock, 179-235 m.; gy. M. Large otter trawl: i ?.

St. 42. i.iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of
Jason Light to 4 miles N 39° E of Jason Light, 120-204 m. ; M. Large otter trawl : 2 cjc?, i ?.

St. 53. 12. V. 26. Port Stanley, East Falkland Islands, hulk of 'Great Britain', 0-2 m. Mussel
rake : 2 ??, i S-

St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, from 4-1 miles N 54^ E of
Larsen Point to 1-2 miles S 62° W of Merton Rock, 230-250 m.; gy. M. Large otter trawl: 3 ^$,
8 adult and immature ??.

St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, from 54° 04' S, 36° 27' W
to 53°s8'S, 36°26'W, 155-178 m.; gn. M.S. Large otter trawl: i <?, 2?$ and 2 immature
specimens.

St. WS 33. 21. xii. 26. 54° 59' S, 35° 24' W, 130 m.; gy. M. St. i m. horizontal tow-net-net
touched bottom: 3 immature specimens.

St. WS 228. 30. vi. 28. 50° 50' S, 56° 58' W, 229-236 m.; Sh. c. w. S. Commercial otter trawl:

3 SS, 3 ??•

South Georgia:

St. MS 14. 17. ii. 25. From 1-5 miles SE x S to 1-5 miles S A W of Sappho Point, East Cumber-
land Bay, 190-iiom. Small dredge: i S-

St. MS 65. 28.ii.26. East Cumberland Bay, i-6 miles SE of Hobart Rock to i cable N of
Dartmouth Point, 39 m. Small beam trawl: i immature specimen.

St. MS 66. 28. ii. 26. East Cumberland Bay, 2\ miles SE of King Edward Point Light to i J cables
W X N of Macmahon Rock, 18 m. Small beam trawl: 2 immature specimens.

ii6

DISCOVERY REPORTS

St. MS 71 . 9. iii. 26. East Cumberland Bay, 9I cables E x S to 1-2 miles E x S of Sappho Point,
1 10-60 m. Small beam trawl : i (?, 6 ?$ and immature specimens.

St. MS 74. 17. iii. 26. East Cumberland Bay, i cable SE x E of Hope Point to 3-1 miles SW of
Merton Rock, 22-40 m. Small beam trawl : 3 (J^, i $.

Remarks. Only one species of this very distinctive genus is known. The holotype of
A. glociale, a female from the Ross Sea area, has four bluntly conical tubercles on the
mid-dorsal surface of the body and a pair of more acutely conical projections on the
first coxa of each leg, the first excepted as the anterior one is absent.

Fig. 63. Austrodecus gladale, Hodgson. Male: a. Palp: xioo.
b. Third leg— St. WS 228: :■, 60. c. Same— St. 53: :•; 100
(Group I)

Fig. 64. Austrodecus glaciale, Hodgson. Male: a.
Femur of third leg — Terra Nova collection (Groupl).
b. Third leg — St. 27 (Group II): < 60.

The numerous specimens in the Discovery collection fall into two groups. In the
first group, which is represented by a relatively small proportion of individuals, there
are either low conical, or high slender spinose projections on the mid-dorsal surface.
Those specimens with high spinose projections have a similar arrangement of the coxal
processes as in the holotype. Those specimens with lower, more robust, conical pro-
jections on the trunk have the coxal projections arranged as follows : a posterior one on
the first coxa of each of the two anterior pairs of legs and an anterior one on the first
coxa of each of the two posterior pairs of legs (in addition there is sometimes a very small
posterior projection on the coxa of each third leg).

The second group, which includes the bulk of the material, is characterized by (i) the
complete absence of any tubercles or processes on the dorsal surface of the trunk in

AMMOTHEIDAE ii7

adults and young, and (2) the presence of an anterior and a posterior conical projection
on the first coxa of each walking leg. Also, in the males, the femoral cement-gland opens
at the apex of a high, conical projection, whereas, in the first group, it is situated on a low
blunt eminence.

The specimens in the Terra Nova collection (Caiman, 191 5, p. 66), also from the
Ross Sea, agree very closely with the holotype ; in the male the femoral tubercle is high,
but more blunt and situated more proximally than is the acutely conical tubercle in
the males of our second group (cf. Fig. 64 a and b).

It is doubtful if these forms all belong to one variable species ; the walking legs show
considerable differences (Figs. 63 and 64). The main differences described above may
be represented briefly as follows:
Group I. Projections on mid-dorsal surface of trunk present; no anterior projection on first
coxa of first leg.
A. Projections on trunk low, rather robust, conical tubercles.

1 . Two projections on first coxa of legs 2-4 ; femoral gland in male opens at the
apex of a high, wide, blunt cone (Fig. 64 a)

Holotype and Terra Nova specimens

2. One projection on first coxa of all the legs [a small posterior one may also be
present on the third leg] ; femoral gland in male opens at apex of a very
low, blunt eminence (Fig. 63 b) Discovery, St. WS 228

B. Projections on trunk high, slender, spinose; two high projections on first coxa
of legs 2-4; femoral gland in male opens at apex of a low, blunt, rounded

eminence (Fig. 64 f) Discovery, St. 53, St. WS 33

Group II. Projections on mid-dorsal surface of trunk absent; anterior and posterior projections
on the first coxa of each leg; femoral gland in male opens at apex of a high, narrow
cone (Fig. 646) Discovery— all the remaining stations

The material described by Bouvier (1913, p. H?)- to judge from the arrangement of
the coxal processes, is similar to that from WS 228. The low femoral tubercles in the
male would be much more easily overlooked than the high conical ones, which are very
striking even in immature specimens (see Caiman, 1915, p. 67).

The males, in many cases, appear to be quite adult, although the genital openings have
not been observed with certainty, in several instances there seemed to be a small
opening on the second coxa of the last leg.

Distribution. Probably circumpolar; recorded from the Ross Sea area, Graham
Land region, South Georgia and the Magellan district.

Genus Tanystylum, Miers
A number of previous workers (e.g. Norman and Loman) regard the genus Chtenia,
Dohrn, as a synonym of Tanystylum, Miers. Bouvier (191 3- P- 45) retains Chtema, at
least provisionally, as distinct from Tanystylum, the former having only four, mstead
of six or seven segments in the palp. Bouvier, however, includes T. hoekianum, Schimk.,
in the genus Clotenia, although the palp is described and figured with six segments.

ii8 DISCOVERY REPORTS

Bouvier lists seven ^ species of Tanystyhmi, and two- other species have since been
described, namely T. oedinotwn, Loman (1923, p. 29), and T. ornatum, Flynn (1928,
p. 33). Several species are imperfectly known or known from a single specimen only
(e.g. T. hoekiimum, Schimk., T. chierchiae, Schimk., T. cakirostre, Schimk., and
T. oedinoiiim, Loman). Therefore it has not been thought advisable to give a key to the
determination of species not found in Antarctic and sub-Antarctic waters. The forms
with a four-jointed palp are not represented in the Discovery collection, so that the
question of the validity or non-vaUdity of the genus Clotenia, Dohrn, cannot be
discussed here.

Two Antarctic species, T. styligeriim, Miers, and T. pfefferi, Bouvier, have recently
been described elsewhere (Gordon, 1932 a, p. 87).

Key to the determination of the Antarctic and sub-Antarctic species :

Palp with seven segments :

Long setose spines on walking legs ; palpal segments 2 and 4 subequal, long; abdomen long,

extending beyond the second coxa of the last leg T. styligerum, Miers

No long setose spines on walking leg ; palpal segment 2 longer than 4 ; abdomen extending to
distal end of first coxa [a low, wide tubercle on dorsal surface of body, between ocular
tubercle and abdomen] T. oedinotum, Loman

Palp with six segments, or with seventh segment more or less fused with sixth.

Walking leg rather slender, second tibia nearly twice the sum of tarsus and propodus

(I7S-I-9 : i) T. pfefferi {typical form)

Walking leg relatively shorter and more robust; second tibia only about 1-25 times the sum

of tarsus and propodus T. pfefferi {Falkland I. var.)

Tanystylum styligerum, Miers (Fig. 65 b).

Nymphon styligerum, Miers, 1875, p. 76.
Tanystylum styligerum, Miers, 1879, p. 213, pi. xi, figs. 9-9 d.
Tanystylum kentrodes, Loman, 1923, p. 28, fig. E, 1-4.
Tattystyhmi longicaiidatum , Hodgson, 1907 fl, p. 13, figs. 4-6.
Tanystylum styligerum, Gordon, 1932 a, p. 88.

St. 53. 12. v. 26. Port Stanley, East Falkland Islands, hulk of 'Great Britain', 0-2 m. Mussel
rake: 14 specimens, adults and young, "sorted from washings from Hydroids and Mytilus clumps"
(with Achelia parvula and Austrodeeus glacials).

Remarks. The specimens agree well with the co-types of T. styligerum (Gordon,
1932 a, p. 88). Several specimens resemble T. kentrodes (Loman, 1923, p. 28) in
having a single long spine on the tubercle borne by each of the three anterior lateral
processes. In other specimens there are 1-3 spines on each tubercle. The anterior
border of the cephalon, measured to the outer rim of the sockets for the palps, is a little
less (c?) or a little more (?) than half the width across the second lateral processes.

The ocular tubercle is more slender and is set further forward than that represented

1 Excluding T. hoekiaman, Schimk.; T. longicaudatim , Hodgson, however, appears to be a synonym of
T. styligerutn, Miers.

2 T. kentrodes (Loman, 1923, p. 28) is also a synonym of T. styligerum.

AMMOTHEIDAE

lit;

in Loman's figure (1923, p. 28, fig. E, i). The fourth palpal segment is not curved and
is subequal to the second. The second may be a little shorter than the first tibia. The
male oviger is represented in Fig. 65 b; the specimens are apparently not quite adult,

Fig. 65. Tanystylum pfefferi, Bouvier: a. Male oviger — specimen not quite adult: x 100.
T. styligerum, Miers: b. Male oviger — St. 53: x 60 and 100.

as the genital pores are not yet present. The walking leg is considerably more slender
than in the female.

Distribution. Kerguelen, the Falkland Islands and Beagle Channel near Tierra
del Fuego.

Tanystylum pfefferi, Bouvier (Figs. 65 a, 66, 67 and 68).

Clotenia dohrnii, Pfeffer, 1889, p. 48.

Tanystylum pfefferi , Bouvier, 1913, pp. 5, 45 and 122.

Tanystylum pfefferi, Gordon, 1932, p. 90.

(rt) Typical forms from :

St. WS 25. 17. xii. 26. Undine Harbour (North), South Georgia, 18-27 m.; M. S. Small beam
trawl: i larvigerous c^.

St. 456. 18. X. 30. I mile E of Bouvet Island, 40-45 m. Large dredge: 1 ?, i larvigerous (^, with
Nymphon australe.

(b) Smaller forms from :

St. 51. 4. V. 26. Off Eddystone Rock, East Falkland Island, from 7 miles N 50° E to 7-6 miles
N 63° E of Eddystone Rock, 105-115 m.; f. S. Large otter trawl: i 6', i ?•

St. 53. 12. V. 26. Port Stanley, East Falkland Island, hulk of 'Great Britain', 0-2 m. Mussel rake:
many specimens, <?<?, ?? and young, "sorted from washings from Hydroids and Mytilus clumps";
one $ "from washings from Kelp root".

St. WS 124. 9. vi. 27. Gough Island, Penguin Island anchorage, 40° 16' S, 9° 58' W, 40-60 m.
Large dredge : i ?, i ovigerous S-

St. 4. 30.1.25. Tristan da Cunha, 36° 55' S, 12° 12' W, 40-46 m.; St. Large dredge : i specimen
from debris of stones with encrustations.

DISCOVERY REPORTS

Remarks. The specimens from WS 25 and St. 456 agree very closely with the female
specimen from the Hamburg Museum collection described elsewhere (Gordon, 1932 a,

a b

Fig. 66. Tanystylum pfefferi , Bouvier. Palp: a. St. 53, Falkland Islands: x 100.
b. St. WS 124, Gough Island: x 80. c. St. 456, Bouvet Island: ; 60.

p. 90). The almost vertical abdomen, the very short chelophores, the low, wide ocular
tubercle and the relative proportions of the palpal segments are characteristic of the
species.

Fig. 67. Ta7tystylumpfefferi, Bouvier. Male— St. 53 : a. Third leg. 6. Dorsal
view of body with chelophores and first coxae: : 47.

There is a distinct tendency for papal segments 6 and 7 to unite ; sometimes a suture
appears to be present, but when the palp is cleared and mounted on a slide, no actual
suture can be detected. Sometimes there is a distinct median emargination on one side

AMMOTHEIDAE

separating the setae into two groups (Gordon, 1932 «, p. 92, fig. 4^). Occasionally no
emargination occurs and the lateral setae are uniformly arranged, so that all trace of
fusion is obliterated and the palp has the appearance of a normal six-jointed appendage
(Fig. 66 c).

The egg masses are irregularly-shaped rings strung on each oviger like beads on a
thread; they number 1-4 on each oviger (the male oviger is represented in Fig. 65 a).
There are three pairs of male genital openings.

The specimens from the Falkland Islands differ considerably from typical T. pfefferi.
(i) The palp is six-jointed (without any trace of a seventh segment) and the terminal

Fig. 68. Tanystylum pfefferi, Bouvier. Third leg of male : «. Typical— St. WS 25. 6.St.WSi24;
a number of small capsules (? encysted Protozoa) attached to first coxa.

segment is shorter, not equal to or longer, than the second segment (Fig. 66 a). (2) The
walking leg is more robust and the propodus and tarsus are considerably longer relative

to the second tibia f l^^^I^h °f second tibk ^ 1:25 .^^^^^^ „f i75zi:9\ The
to the second tibia [j^^^^ ^f ^^j-sus + propodus i i /

other differences are very slight. (3) The specimens are of much smaller size, yet many
females and at least one male have genital apertures (see measurements). (4) There is
usually a narrow conical process on the ocular tubercle above the level of the eyes.
(5) There are as a rule no setae or tubercles on the lateral processes. (6) The abdomen
is more oblique. (7) The chelophores are rather longer (Fig. 67 a and b).

The specimens from Gough Island (WS 124) are much more like the typical form.
The abdomen is nearly vertical, the ocular tubercle and the chelophore are quite
typical, and the walking leg, though somewhat more robust, has a long second tibia
(Fig. 68 b: cf. Figs. 68 a and 67 a). There are one or even two small rounded tubercles

122 DISCOVERY REPORTS

on each lateral process. The palp is again six-jointed, the terminal segment being equal
to the second (Fig. 66 b). It is perhaps of interest to note that the egg mass in this
instance is single and carried on both ovigers.

For the present it is advisable to refer all these forms to a single species because of
the variation that occurs typically in the terminal segment (or segments) of the palp in
T. pfefferi. The specimens from Gough Island and that from Tristan da Cunha,
although found beyond the northernmost limits of the sub-Antarctic Zone, can hardly
be distinguished from the South Georgian specimens — except as regards the egg mass
and the palp. The specimens from the Magellan district include no ovigerous males,
but the walking legs of some of the females contain developing ova, so that it is highly
probable that adults do not exceed i mm. in trunk length. The smaller ratio between
the lengths of the second tibia and tarsus plus propodus can hardly be regarded as
of specific importance. Caiman (1915, p. 59, figs. 13 and 14) found a similar type
of variation in Achelia spicata, although in that case the specimens were (with two
exceptions) from one locality. The numerous specimens from the hulk of the ' Great
Britain' are a pure culture. In such a limited area close inbreeding doubtless occurs.

The breeding season appears to vary somewhat, males from the first two stations were
carrying larvae in October and December; the single male from Gough Island was
carrying eggs as early as July.

Measurements {mm.)

<s (St. cj a

Length of proboscis ...
Greatest width of proboscis
Length of trunk

Width across second lateral processes 1-75 0-9 i-i

Length of abdomen
Length of chelophore
Third leg:

Coxae

Femur

First tibia ...

Second tibia

Tarsus and propodus

Claw

Auxiliaries ...

WS 25) (St. 53) (St. 124)

1-2 0-55 0-82

0-55 0-3 0-4

i'65 0-8 i-i

07 0-33 0-4

0-2 0-13 0-13

1-6 07 0-95

1-8 073 i-o

2-05 079 1-2

2-4 08 1-3

1-2 0-6 075

0-5 0-3 0-25

0-35 0-15 0-2

Distribution. South Georgia, the Falkland Islands, Gough Island and Tristan

da Cunha.

Genus Rhynchothorax, Costa

Rhynchothorax australis, Hodgson.

Hodgson, 1907, p. 57, pi. viii, fig. 3; 1915, p. 148.
Caiman, 1915, p. 67, fig. 21.
Loman, 1923 a, p. 12.

St. WS 84. 24. iii. 27. 7I miles S 9° W of Sea Lion Island, East Falkland Islands, from
52° 33' S, 59° 08' W to 52° 34' 30" S, 59° 11' 00" W, 75-74 m.; c. S. Sh. St. Commercial otter
trawl: i immature specimen.

Distribution. Ross Sea area and the Magellan district.

AMMOTHEIDAE

123

Genus Nymphopsis, Haswell

Nymphopsis denticulata, n.sp. (Figs. 69 and 70).

St. WS 85. 25. iii. 27. 8 miles S 66° E of Lively Island, East Falkland Islands, from 52" 09' S,
58° 14' W to S2' 08' S, 58° 09' W, 79 m. ; S. Sh. Commercial otter trawl : i ?.

Description. Trunk compact ; lateral processes in contact, 1-3 each with an anterior
and a posterior spinose process, 4 with an anterior one only (Fig. 69). Cephalic segment
distinctly separated off by a suture from the succeeding segments, which are all com-
pletely fused. Two acute antero-lateral spinose processes on cephalon; ocular tubercle
high, tubular, situated mid-way between these on the anterior border of cephalon.
Eyes large, sub-terminal. Two high, slender, spinose projections on mid-dorsal surface,
between cephalic suture and abdomen (Fig. 69).

Fig. 69. Nymphopsis denticulata, n.sp. Holotype. Lateral view of body with chelophore: x 30.

Proboscis slightly longer than trunk, ovate, tapering anteriorly to a blunt point
(Fig. 69).

Abdomen, long, slender, sub-cylindrical, elevated in distal half and reaching a little
beyond the distal end of the second coxa; two small spinose projections on proximal
half (Fig. 69).

Chelophore approximately half as long as proboscis. Scape of almost uniform
diameter throughout and consisting of a single segment ; a short, stout, spinose process
on the mid-dorsal surface, a bifurcated process at the inner, and two smaller spines at
the outer distal angle (Fig. 70 a, lateral aspect). Chela small and imperfect.

Palp as represented in Fig. 70 b, nine-jointed and longer than proboscis.

Oviger ten-jointed, with one or two denticulate spines on the four terminal segments
as in some species of Achelia (Fig. 70 c: cf. Fig. 59 a, c); terminal claw absent.

Walking leg. The second walking leg, which was already detached from the specimen,
is represented in Fig. 70 d and e ; the spinose setae are simple, not compound and the
auxiliary claws are at least two-thirds as long as the main claw. Ova are present in the
two distal coxae and in the femur.

Measurements {mm.)

Length of proboscis ...

Greatest width of proboscis ...

Length of trunk

Length of cephalic segment ...

Width of anterior border of cephalon

Width across first lateral processes ...

Length of abdomen ...

Length of chelophore

1-35

Second leg:

0-6

Coxae

1-13

Femur

0-6

First tibia

0-67

Second tibia

1-2

Tarsus and propodus

0-6

Claw

0-68

Auxiliaries ...

1-2

1-4

14

1-47
0-87

0-35
0-25

16-2

124

DISCOVERY REPORTS

Remarks. This specimen differs from the four known species of the genus in having
(i) the cephaUc segment separated off from the remaining trunk segments, (2) simple
instead of compound spines on the legs, (3) the scape of approximately uniform
diameter throughout, and not hollowed out to form a cup or collar ^ around the proximal
part of the chela, and (4) denticulate spines on the terminal segments of the oviger.
These differences are not of generic importance.

Fig. 70. Nytnphopsis denikulata,n.sp. Holotype: «. Chelophore, lateral view. Zi. Palp.
c. Oviger and terminal segments, further enlarged, d. Second left leg. e. Terminal
segments of same, further enlarged, {a, b, c and e: >: 60.)

In three species — A^. annatus, Haswell (Flynn, 1919, p. 84, pi. xxi, fig. 18), A^.
korotnewi, Schimkewitsch, and A^. muscosus, Loman (1908, pp. 50-53, pi. xiii, figs.
175-188) — the chelophore exceeds the proboscis in length. In the remaining two
species— TV. abstrusiis, Loman, and N. denticulata— the chelophore is shorter than the
proboscis. The former differs from the latter species, however, in having a two-jointed
scape and three, instead of two, spinose projections on the dorsal surface of the trunk.

1 N. abstrusus, Loman, also seems to have no projecting collar round the chela, but the figures given are
very small (Loman, 1923 a, p. 8, fig. D i and 2).

PYCNOGONIDAE

125

Distribution. This is the first record of the genus from the western side of the
sub-Antarctic Zone.

Family PYCNOGONIDAE
Genus Pycnogonum, Briinnich
Pycnogonum rhinoceros, Loman (Fig. 71c).
Loman, 1923, p. 7, fig. A, 1-5.

St. 45. 6. iv. 26. 27 miles S 85^ E of Jason Light, South Georgia, 238-270 m.; gy. M. Large
otter trawl : i S-

Remarks. A single, large male undoubtedly belongs to this species. The type
specimens were not quite adult and the oviger (Loman, 1923, fig. A, 3, p. 8) possessed

Fig. 71. Pycnogonum platylophiim , Loman: a. Third leg of male. b. Same of female.

(Both: 20.)
P. rhinoceros, Loman : c. Oviger of male.

only seven segments. In the Discovery specimen the oviger is nine-jointed, with a short
terminal claw (Fig. 71 c). Loman states that there is "no trace of eyes"; in the
Discovery specimen, however, eyes are present and the anterior is larger than the
posterior pair.

Measurements {mm.)

Length of proboscis ...

Greatest width of proboscis .

Length of trunk

Length of cephalic segment

Width of cephalon

Width across second lateral processes

Length of abdomen ...

Distribution. South Georgia.

Third leg:

7-0

First coxa ...

1-6

37

Second co.xa

2-0

9-6

Third coxa

0-9

2-6

Femur

4-4

3-5

First tibia

... 4-8

7'3

Second tibia

4-0

2-4

Tarsus and propodus

2-8

Claw

i-o

126

DISCOVERY REPORTS

^

?

2-0

3-0

I-I7

1-6

3-6s

5-0

1-4

2-07

1-4

1-8

2-45

3-4

1-2

1-3

Pycnogonum platylophum, Loman (Fig. 71 «, b).
Loman, 1923 o, p. 10, fig. E, 1-5.

St. WS 88. 6. iv. 27. 54= 00' S, 64'' 57' 30" W, 118 m.; S. Sh. St. Commercial otter
trawl: i ^.

St. WS 93. 9. iv. 27. 7 miles S 80° W of Beaver Island, West Falkland Islands, from 51° 51' S.
61° 30' W to 51° 54' S, 61° 30' W, 133-130 m.; gy. S. Commercial otter trawl: i ?.

Measurements {mm.)

Length of proboscis ...

Basal width of proboscis

Length of trunk

Length of cephalic segment ...

Width of cephalon (anterior border)

Width across second lateral processes

Length of abdomen ...

Remarks. These specimens differ from the holotype described by Loman (1923 a,
p. 10, fig. E) in one minor point — namely, in having the lateral processes in contact
instead of separated by narrow intervals. Fig. yi a and b represents the third right leg
of the male and of the female respectively. The female is rather soft and of a light
brov^nish-yellow colour; the male is much darker brown in colour, with more slender
walking legs. The genital pores are present in each specimen.

Distribution. Sub-Antarctic, from the Magellan district.

Pycnogonum magellanicum, ? Hoek (Fig. 72).

P. magellanicum, Hoek, 1898, p. 296, pi. iii, figs. 20 and 21.

? P. magnirostre, Mobius, 1902, p. 194, pi. xxx [vii], figs. 12-14.

St. WS 228. 30. vi. 28. 50° 50' S, 56° 58' W, 229-236 m. Commercial otter trawl: 1,^,1$.

St. 399. 18. v. 30. I mile SE of SW point
of Gough Island, 141-102 m. Large dredge:
I $, I ovigerous (J.

Description of male (WS228). Trunk
compact, three distinct transverse body
ridges present, each with a median, round,
blunt tubercle ; a similar tubercle on each
lateral process. Lateral processes 2-4 each
in contact with the preceding body ridge.
Cephalic segment equal to the sum of
the two succeeding segments; ocular
tubercle situated on a broad slightly
raised ring near the anterior border of
cephalon and not quite as high as wide ;
eyes large. Width across first a little
greater than that across second lateral
processes.

Fig. 72. Pycnogonum magellanicum} ,\loek: a. Third leg
of male. Zi. Same of female — St.WS228. (Both: x 20.)

PYCNOGONIDAE

127

Proboscis a little shorter than cephalic segment, bluntly conical.

Abdomen somewhat clavate, reaching to distal end of second coxa of fourth leg.

Oviger short, with nine small segments and a terminal claw. The left oviger has been
entirely suppressed and there is no trace of the small process to which it ought to be
attached.

First coxae at least twice as wide as long, the anterior distal angle of each is drawn
out into a triangular lobe.

Third leg as represented in Fig. 72 a and at least half as long again as the trunk.

Female. The transverse body ridges and the anterior rim of the cephalon are more
pronounced than in the male. There are two small pointed tubercles a short distance
behind the ocular tubercle — these are much smaller in the male. The ovigers are, of
course, wanting, and there is a low, round tubercle dorsally on the second coxa of the
last leg.

The genital openings are apparently not yet present in the female, but they are quite
distinct in the male.

Measurements {mm)

St. ws

228

St.

399

Holotype

?

,

0"

V

0 (ovig.)

V

Length of proboscis ...

3-2

1-2

1-2

0-93

I-I

Basal width of proboscis

1-8

0-9

0-9

0-67

0-9

Length of trunk

5-73

37

3-8

2-5

3-5

Length of cephalic segment

2-2

1-6

1-4

1-0

1-3

Width of anterior border of cephalon

2-0

1-3

1-35

0-87

1-2

Width across second lateral processes

4-0

2-45

2-6

1-45

2-0

Length of abdomen

1-4

1-2

0-95

0-53

0-6

Third leg:

Coxae

1-8

1-0

i-i

0-6

073

Femur

17

1-3

I'l

0-5

0-63

First tibia

1-6

1-3

i-i

0-5

0-6

Second tibia

i-o

07

0-55

0-35

0-4

Tarsus and propodus

—

1-07

1-05

0-5

0-55

Claw

—

0-5

°-5

0-25

0-24

L. of trunk ~ w. across 2nd lateral processes .

1-43

1-51

1-46

172

175

L. of abdomen -^ 1. of trunk

0-24

0-32

0-25

0-21

0-17

L. of third leg -^ 1. of trunk

—

1-59

1-42

1-04

0-9

Remarks. The specimens from St. 399 differ from those just described in the
following respects, (i) The trunk is narrower, with lower transverse body ridges and
lower median tubercles. (2) The third leg is subequal to, instead of approximately
half as long again as the trunk length. (3) The abdomen also is shorter, although it
again reaches to the distal end of the second coxa (see ratios above). (4) The lateral
processes are separated by narrow intervals. The genital openings are quite distinct in
each specimen.

The Discovery specimens are referred provisionally to P. magellanicum, although they
do not agree with each other or with the holotype. Hoek's figure (1898, pi. iii, fig. 20)
does not show the transverse body ridges nor the low median tubercles of the holotype,
which in this respect agrees with the specimens from St. 399. In other respects the

,28 DISCOVERY REPORTS

holotype agrees more closely with the specimens from WS 228, e.g., as regards width
of trunk and length of abdomen (see ratios above). The third leg is incomplete ; but the
sum of the segments measured exceeds the trunk length, so that the complete leg would
probably be about half as long again as the trunk.

The holotype has a much longer proboscis than any of the Discovery specimens,
being almost as long as the sum of the first two segments. P. magnirostre, Mobius
(1902, p. 194, pi. XXX, figs. 12-14), from Kerguelen also has a very long proboscis and
may prove to be identical with P. magellanicum. It is possible that more than one
species of Pycnogoniim is represented, but the five specimens measured above are not
sufficient to enable one to settle this point satisfactorily.

Distribution. Off the Falkland Islands and Gough Island.

THE CENTRAL NERVOUS SYSTEM IN DECOLOPODA

The nervous system is very similar to that described for Colossendeis by Hoek (1881,
pp. 108-119, pi. xvii, fig. 2 and pi. xviii, fig. 4). The long circum-oesophageal com-
missures, considerably foreshortened in Fig. 73 A, enclose the oesophagus and a pair
of somewhat triangular muscles that work the proboscis.

The usual main nerves arise from the supra-oesophageal ganglia, namely: a, a median
azygous proboscidial nerve ; b, a pair of optic nerves each dividing into three branches
to innervate the anterior eye, the "sense organ "i and the posterior eye respectively
(see Fig. 73 C), and r, a pair of nerves to the chelophores. A pair of fine nerves, d,
probably correspond to those designated by Wiren (1918, pp. 73 and 79, figs. 10, 14
and 15) as nerves No. 3 in Nymphon grossipes-mixtum^. But, instead of pursuing a
course parallel to the azygous proboscidial nerve, as in Nymphon, they pass obliquely
outward and forward towards the base of the chelophores, innervating the muscles that
work these appendages. Near to these latter nerves, a pair of thin strands arise {e) and
run directly forwards to break up into a meshwork innervating the skin along the
anterior border of the cephalon. In D. aiitarctico this pair of nerves is well seen
(Fig. 73 B, e), but in the dissections of D. atistralis they were, as a rule, removed with
the skin. D. antarctica differs from D. atistralis in two respects: (i) the optic nerve is
much stouter, with shorter thicker branches; (2) nerve d arises from the optic nerve.
In the specimen from which Fig. 73 B was obtained, nerve d arose from the base of the
median branch of the optic nerve on the left side, as represented in the diagram, whereas
on the right side it arose from the main stalk of the optic nerve just below the forking.

From the first thoracic gangUonic mass three pairs of nerves arise (Fig. 73 A, i, 2, 3)
innervating the proboscis, palps and ovigers respectively. In D. antarctica from St. 42
the proboscidial nerves arise each from the circum-oesophageal commissure a short
distance from the thoracic ganglionic mass; in the male from St. 149 the nerve on the

1 Probably a vestigial eye (see Loman, 1924, p. 318).

- Only present in Nymphon (Wiren, 191 8, pp. 73 and 77).

NERVOUS SYSTEM OF DECOLOPODA 129

left arises similarly from the commissure, while that on the right arises from the
gangUonic mass (as in D. australis, Fig. 73 A).

Wiren (1918, p. 80, fig. 16) illustrates a nerve, designated as No. 5, arising from near

a

Fig. 73. Decolopoda australis. Eights: A. Central nervous system: x 13.
Decolopoda antarctka, Bouvier. B. Supra-oesophageal ganglia and their
associated nerves (diagram). C. Optic nerve and its branches below the
ocular tubercle : x 13

a. Median azygous proboscidial nerve, b. Optic nerve, c. Nerve to chelo-
phore. d. Nerve to muscle working chelophore. e. Superficial nerve to
anterior border of cephalon. i, 2 and 3. Nerves to proboscis, palp and
oviger respectively.

the base of the circum-oesophageal commissure in Colossendeis; this nerve (1918,
pp. 73-4) passes right forward to the base of the proboscis, near to the point of insertion
of the palp, to innervate muscles. In D. antarctka I have been unable to discover this

17

13°

DISCOVERY REPORTS

nerve, neither was it observed in the specimen of D. australis from St. 149 represented
in Fig. 73 A ; but a very slender nerve strand was present in one of the specimens
from St. 154.

As anticipated, the ventral nerve chain in Decolopoda possesses six, instead of the
usual five, ganglionic masses. The lateral nerves to palp, ovigers and each leg soon
bifurcate.

ABNORMALITIES

Abnormalities occur very infrequently in Pycnogonida ; at any rate, in the large
Discovery collection the following few examples are all that were noted, ^ apart from
the occasional loss and subsequent regeneration of a walking leg.

I. Palp. The tip of the palp is sometimes slightly abnormal, e.g. (i), the right palp
oi Colossendeis tortipalpis, n.sp.,rcpvtsentedm Fig.2e;(ii),the left palp oi A7nmothea,sp.}
described on p. 108 and (iii), the right palp of the holotype of Ammothea calmani, n.sp.
(Fig. 74 a), where segments 7 and 8 appear to be fused. In the last example the notch
X (Fig. 74 a) may represent the beginning of bifurcation or of segmentation.

Fig. 74. a. Ammothea calmani, n.sp. Holotype. Segments 4-9 of right
palp — segments 7 and 8 are apparently fused 1x15 and 20.

b. Decolopoda antarctica, Bouvier. Proximal portion of abdomen,
showing additional small clavate process on right side: x 20.

c. IPallenopsis sp. St. 256. Terminal segments of oviger showing
distinct bifurcation: xioo.

The most extreme case of abnormality is the tripartite palp of Nymphon proceroides,
Bouvier (Fig. 75). Here a is probably the original left palp, bearing the two extra palps
on the outer distal angle of the second segment. The long axes of palps a, b and c lie
in one plane ; ^ is a mirror image of a and c of b, according to the two rules enunciated
by Bateson (1894, p. 479).

1 Two of the abnormalities figured (Figs. 74 a, 75) occurred in specimens from other collections.

ABNORMALITIES Ui

These variations are sporadic and may be the result of some sHght injury; in one
species, however, sHght palpal abnormalities occur so frequently as to be characteristic
of the species. In Decolopoda antarctica, Bouvier (fig. i, a-e), the number of palpal
segments varies from 8 to lo (Table I). Until a large number of specimens are available
for study it is impossible to say whether eight or nine is the normal number of segments.
Three of the seven specimens examined have both palps eight-jointed as in Fig. i a;
one specimen has both palps nine-jointed. The numbers of palpal segments in each of
the seven specimens may be represented as follows: 8, 8; 8, 8; 8, 8; 9, 9; 8, 9;
8 + (fig. I J), 8 ; 8 + (fig. I 0' 9 + (fig- I ^)- The vestigial ninth or tenth segment
(Figs. I d and i e) may resemble a minute claw. Caiman (1920, p. 244) puts forward
the view that "the assumption of a claw-like form by the terminal segment may,
perhaps, be regarded as a case of homoeosis" since all appendages posterior to the palp
may end in claws. Occasionally the terminal segment is somewhat abnormal, as in
Fig. I b, suggesting regeneration of the apex after mutilation. Judging from the
abundance of encrusting organisms found on Decolopoda antarctica, it must be a very
sluggish species (see p. 132), and thus it may be unusually liable to have the palp nibbled
by an Isopod or some such animal.

2. Oviger. In one instance— Pallenopsis sp.?, St. 256— the tip of the oviger has
bifurcated (Fig. 74 c) ; in fact segment 9 has almost the appearance of trifurcating.

Quite a different type of abnormality occurred in a male of Pycnogomm magellanicim}
(WS 228), where the left oviger was entirely suppressed, that on the right side being
present.

3. Ocular tubercle. A specimen of Ausfrodecus glaciale, Hodgson, had had the greater
part of the ocular tubercle removed as the resuh of some accident. The wound had
healed over without any apparent attempt at regeneration. The long ocular tubercle is
so conspicuous a feature of this species that the low blunt stump attracted attention
at once. Perhaps, had the animal undergone another moult, the deficiency might have
been made good. But there is no experimental evidence on this point.

4. Abdomen. It is interesting to find, in a species showing frequent variation in the
distal segmentation of the palp, the following peculiarity. In the specimen of Decolopoda
antarctica from St. 164 there is, to the right of the abdomen, a short, ? movable, clavate
process projecting from the posterior end of the trunk (Fig. 74 b). A small circular area
at the posterior end is thin and membranous, but it is rather difficult to decide whether
it is perforate. Unfortunately the process was detached from the specimen after the
camera lucida drawing had been made; only a very small indication of the point of
attachment can now be observed. This process may be interpreted as an abortive second
abdomen.

5. Auxiliary claw. The third right leg of Pentanymphon antarcticum} , Hodgson,
from St. 170, has two auxiliary claws on the anterior side, and only one, as is usual,
on the posterior side of the main claw. There is no reversal of symmetry here and no

132

DISCOVERY REPORTS

indication of a median reversed claw between these two auxiliaries. The outer, or dorsal,
is probably the original auxiliary, since it is on a level with and subequal to the auxiliary
on the posterior side. The inner, additional auxiliary is very small.

6. The female of Nymphoii temdpes, Bouvier (p. 38), is interesting in that the two
anterior pairs of walking legs are shorter and
much more slender than the two posterior
pairs and do not contain developing ova. It
is possible that all four anterior legs were lost
and subsequently regenerated, although the
loss of so many legs at one and the same time
would be a rare occurrence. It is just possible,
of course, that the original specimen was cut
in two halves and that the whole anterior por-
tion of the body was regenerated. This is highly
improbable, however, for Loeb (1895, p. 251),
experimenting with Phoxichilidiiim maxillara,
found that "Bei den aboralen Halften konnte
ich dagegen nur eine Anschwellung am vor-
deren Endewahrnehmen". The trunk, chelo-

phores, palps and ovigers are all apparently quite normal and show the normal relative
proportions ; if the whole anterior half of the body had been regenerated it would very
likely have been small in proportion to the posterior half. Another possible explanation
may be advanced tentatively, namely, that in this species the female has normally only
two pairs of genital openings as in the male ; but until more females are available the
specimen may be regarded as somewhat abnormal.

Fig. 75. Nymphon procerot'des, BouYier. Co-type.
Ventral view of left palp.

ENCRUSTING ORGANISMS

Encrusting organisms are found on Pycnogonida with far greater frequency than on
Crustacea, for example. These are chiefly encrusting Polyzoa, adherent Foraminifera
and Hydroids ; but, in addition, an occasional Brachiopod, Sponge, Tunicate, Serpulid
and Cirripede {Scalpelhtm) may be attached. Occasionally, also, one or two Isopods
may be found clinging to a Pycnogonid.

From the abundance of encrusting Polyzoa found on Decolopoda antarctica, compared
to those on D. mistralis, it is evident that the former must be a more sluggish species
than the latter (see list below). Polyzoa are also found growing on five species of
Ammothea — e.g. the specimen of A. striata from St. 167 has very large patches of
Alcyonidium entirely covering a number of the long segments of the legs. Perhaps these
species are also rather sluggish, but the numbers of individual specimens are too small
to enable one to be at all dogmatic on this point. With the exception of Colossendeis
glacialis, the remaining species on the list are each represented by a large number of

ENCRUSTING ORGANISMS

133

Specimens ; one would therefore expect to find a certain number of Polyzoa on at least
some of the specimens in each case.

The Polyzoa and Tunicata listed below have been determined by Dr Anna B. Hastings
of the British Museum (Natural History).

POLYZOA

var.
var.

SchizoporcUa hyalina var. discreta
Ciisia sp.

Mcmbraiiipora galeata, Busk
*Schizopoyella hyalina (Linn.) var.
Osthimosia sp.
Figularia sp.

Schizoporella hyalina (Linn.) var.
Figularia sp.

Schizoporella hyalina (Linn.) var.
Schizoporella hyalina var. discreta
Schizoporella hyalina var. discreta
Schizoporella hyalina (Linn.) var.
Schizoporella hyalina var. discreta
Osthimosia sp.

Membranipora galeata. Busk
Alcyonidium sp.
Barentsia discreta (Busk)
Membranipora galeata. Busk
Schizoporella hyalina (Linn.)
Schizoporella hyalina (Linn.)
Cellar ia sp.

Schizoporella hyalina (Linn.) var.
Osthimosia sp.
Osthimosia sp.
Supercytis sp.

Schizoporella hyalina (Linn.) var.
Osthimosia sp.

Schizoporella hyalina (Linn.) var.
Tubulipora sp.
Amastigia nuda, Busk
Crisia sp.

Schizoporella hyalina (Linn.) var.
Schizoporella hyalina var. discreta
Schizoporella hyalina (Linn.) var.

Schizoporella hyalina var.
Schizoporella hyalina var.
Alcyonidium sp.
Osthimosia sp.
Schizoporella hyaliiia var.

(Busk)
"1

1

? /

. }

(Busk)
(Busk)

?

(Busk)

/
(Busk)

on Decolopoda antarctica St. 140

on Decolopoda antarctica St. 123

on Decolopoda antarctica St. 39

on
on

on
on

on

on

on

on
on

on
on

Decolopoda antarctica

Decolopoda antarctica
Decolopoda australis
Colossendeis glacialis
Colossendeis glacialis

on Nymphon charcoti

Nymphon charcoti
Nymphon charcoti

Nymphon australe

Nymphon australe

Nymphon australe

Pallenopsis patagonicum
Pallenopsis patagonicum

Pallenopsis patagonicum

Ammothea gigantea, n.sp.

Ammothea carolinensis
A?nmothea minor

Ammothea minor

Ammothea striata

Holotype of Ammothea
stylirostris, n.sp.

St. 149

St. 164 •
St. 154

Sts. 155 and 148
St. 149

St. 39

St. MS 68
St. 19s

St. 45

St. 182

St. 42

St. 84
St. 167

St. WS 84

St. 371

Cumberland Bay
Sappho Point and

St. 371
St. 190

St. 167

St. 39

TUNICATA

The Tunicates on Nymphon charcoti, St. MS 68, TV. hiemale, St. 39, and Amtnothea gigantea,
St. 371 all belong to the family Polycitoridae (Distomidae).

* Schizoporella hyalina var.? is very near the typical British form. By Schizoporella hyalina var.
(unqueried) I have indicated a distinct variety for which, at present, I have not found a name. [A. B. H.]

LIST OF LITERATURE

Bateson, W., 1894. Materiah for the Study of Variation. Macmillan and Co., London, pp. i-xii, 1-598,

text-figs. 1-209.
BouviER, E. L., 1906. Pycnogonides du ' Frarifais'. Expedition Antarctique Fran9aise (1903-5), pp. 1-69,

text-figs. 1-48, 3 pis.

191 1. Les Pycnogonides du ' Pour quoi- Pas?' C. R. Acad. Sci., Paris, clii, pp. 1136-42.

I9I3- Pycnogonides du ' Pourquoi-Pas?' Deuxieme Expedition Antarctique Fran9aise (1908-10),

pp. 1-169, text-figs. 1-109.
Calman, W. T., 1915. Pycnogonida. British Antarctic ('Terra Nova') Expedition, 1910, Nat. Hist. Report,

Zool., Ill, No. I, pp. 1-74, text-figs. 1-22.
1915 a. The Holotype 0/ Ammothea carolinensis. Leach {Pycnogonida). Ann. Mag. Nat. Hist. (8), xv,

pp. 310-14, text-figs. 1-3.
1920. On a Collection of Pycnogonida from the South Orkney Islands. Ann. Mag. Nat. Hist. (9), vi,

pp.. 244-47, I text-fig.
Flynn, T. T., 1919. A Re-examination of Professor HaswelVs Types of Australian Pycnogonida. Proc. Roy.

Soc. Tasmania, 1919 (24th Nov.), pp. 70-92, 5 pis.
• 1928. The Pycnogonida of the Marine Survey of South Africa. Fisheries and Marine Biological Survey,

No. 6, 1927-8, pp. 3-36, text-figs. 1-21.
Gordon, I., 1932. Re-description of some type-specimens of Pycnogonida of the genus Nymphon. Ann. Mag.

Nat. Hist. (10), IX, pp. 97-120, text-figs. 1-12.
1932 a. Re- descriptions of two species of Pycnogonida of the genus Tanystylum. Ann. Mag. Nat. Hist.

(10), X, pp. 87-93, text-figs. 1-4.
Hodge, G., 1864. List of British Pycnogonida, with descriptions of several new species. Ann. Mag. Nat. Hist.

(3). XIII, pp. 113-17, 2 pis.
Hodgson, T. V., 1902. Crustacea [and Pycnogonida]. Brit. Mus. Rep. Nat. Hist. 'Southern Cross', 1902,

pp. 228-61, pis. xxLx-xl.
1904. On a new Pycnogonid from the South Polar Regions. Ann. Mag. Nat. Hist. (7), xiv, pp. 458-62,

I pi.
■ 1907. Pycnogonida. National Antarctic Expedition, 1901-4, Nat. Hist., iii, pp. 1-72, 10 pis.

1907 ^- Pycnogoniden. Hamburg. Magalhaens. Sammelreise, pp. 3-20, text-figs. 1-6.

1908. The Pycnogonida of the Scottish National Antarctic Expedition. Trans. Roy. Soc. Edinb., XL VI,

Pt. I, pp. 159-88, 3 pis.
1915- The Pycnogonida collected by the ' Gauss' in the Antarctic Region. Ann. Mag. Nat. Hist. (8), xv,

pp. 141-9.
HoEK, P. P. C, 1881. Report of the Pycnogonida dredged by H.M.S. 'Challenger' during the years 1873-6.

'Challenger' Reports, Zool., in, pp. 1-167, 21 pis.
1883. The Pycnogonida dredged in the Faroe Channel during the cruise of H.M.S. ' Triton' (August 1882).

Trans. Roy. Soc. Edinb., xxxii, pp. i-io, i pi.

1898. OnfourPycnogonids, dredged during the Cruise of the ' Challenger' . Tijdschr. Nederl. dierk. Ver.,

v, pp. 290-300, pis. II and ill.
Leach, W. E., 1814. The Zoological Miscellany; being descriptions of new, or interesting Animals, i,

pp. 1-144, 60 pis.
LoEB, J., 1895. Bemerkimgen iiber Regeneration: (i. Vber die Regeneration des Rumpfes bei Pantopoden).

Arch. Entvv. Mechanik. Org., 11, pt. 2, pp. 250-3, text-figs. 1-3.
LoMAN, J. C. C, 1908. Die Pantopoden der Siboga-Expedition. Siboga-Expeditie, Monogr. xl, pp. 1-88,

15 pis.

1923. The Pycnogonida. Further zool. results Swedish Antarctic Expedition, 1901-3,1, No. 2, pp. 5-41,

text-figs. A-G.

1923 a. Subantarctic Pantopoda from the Stockholm Museum. Arkiv for Zool., K. svenska Vetens-

Akad., XV, No. 9, pp. 1-13, text-figs. A-E.

LIST OF LITERATURE I35

LoMAN, J. C. C, 1923 b. Pycnogoniden aus Sud-und West-Afrika. Goteborg Vetensksamh. Handl.,xxvi,

VI, pp. 1-7, 1922 (reprint dated 1923).
MiERS, E. J., 1875. Descriptions of nezv Species of Crustacea collected at Kerguelen's Island by the Rev. A. E.

Eaton. Ann. Mag. Nat. Hist. (4), xiv, pp. 73-6.
1879. Crustacea [and Pycnogonida] of the Transit of Venus Expedition, Kerguelen Island. Phil. Trans.

[Roy. Soc. London], CLXVIII, pp. 200-14, p'- ^^■
MOBIUS, K., 1902. Die Pantopoden der deutschen Tief see-Expedition, 1898-1899. Wiss. Ergebn. d. deutschen

Tiefsee-Expedition.. . .' Valdivia', 1898-9, in (6), pp. 179-96, pis. xxiv-xxx.
Ohshima, H., 1927. Nymphonella tapetis, n.g., n.sp., A Pycnogon parasitic in a Bivalve. Annot. zool.

Japon., II, No. 3, pp. 257-63, text-figs. 1-4.
Pfeffer, G., 1889. Zur Fauna von Sild-Georgien. Jahrb. Hamburg, wiss. Anst., vi, 2te Halfte, pp. 41-9

(Pycnogonida).
Regan, C. Tate, 1914. Fishes. British Antarctic ('Terra Nova') Expedition, 1910, Nat. Hist. Report,

Zool., I, No. I, pp. 1-54, text-figs. 1-8, pis. i-xiii.
Sars, G. O., 1891. Pycnogonida. The Norwegian North-Atlantic Expedition, 1876-8, x.x, 163 pp.,

15 pis., I map.
SCHIMKEWITSCH, W., 1929-30. Pantopoda. Faune de la Russia. Leningrad, 1930.
Wmfoi, E., 1918. Zur Morphologic tind Phylogenie der Pantopoden. Zool. Bidr. Uppsala, vi, pp. 41-181,

text-figs. 1-40, pis. ix-xvi.

INDEX

[Synonyms are indicated by italics. References to specific descriptions are in heavy type.]

abnormalities, 130

abstrusus, Nymphopsis, 124

Achelia, 94, no

adareanum, Nymphon, 28, 32, 34

altioculum, Chaelonymphon, 59

Ammothea, 3, 7, 94

Ammothea sp.?, 5, 95, 96, 108, 130

Ammotheidae, 94

angolense, Nymphon, 5, 77

angusta, Colossendeis, 12

antarctica, Decolopoda, 4, 9, 128, 131, 132

antarcticum, Nymphon, 35, 46

antarcticum, Pentanymphon, 4, 7, 24, 131

armatus, Nymphopsis, 124

articulate, Nymphon, 4, 28, 33, 35, 62, 65

articulata, Colossendeis, 11

assimik, Chaetonymphon, 59

australe, Chaetonymphon, 59

aiistrale, Chaetonymphon, var. aiistrinorum, 59

australe, Nymphon, 3, 4, 28, 33, 34, 59, 62

australe, Nymphon, var. austrinorum, 59, 61, 62

australe, Phoxichilidium, 5, 80, 92

australis, Ammothea, 96

australis, Colossendeis, 4, 12, 15

australis, Decolopoda, 4, 8, 128, 132

australis, Endeis, 5, 93

australis, Phoxichilus, 93

australis, Rhynchothorax, 5, 122

austrinorum, Chaetonymphon australe, var. 59

austrinorum, Nymphon australe, var. 59, 61, 62

Austrodecus, 7, 115

Austropallene, 85

Austroraptus, 94, 1 14

biarticulatum, Chaetonymphon, 71
biarticulatum, Nymphon, 4, 28, 33, 36, 62, 71
? biarticulatum, Nymphon, 62
bouvieri, Nymphon, 4, 28, 33, 35, 62, 73
brachyrhynchum, Nymphon, 28, 32, 34, 65
brachyura, Austropallene, 5, 86
breeding season, 6
brevicaiidatum, Chaetonymphon, 69
brevicaudatum, Nymphon, 4, 6, 28, 33, 36, 62, 69
brevipes, Colossendeis, 12
brucei, Achelia, 5, 113

calcirostre, Tanystylum, 118
calmani, Ammothea, 5, 95, 104, 130
capense, Nymphon, 28, 32, 34, 62
carolinensis, Ammothea, 5, 96, 109
Chaetonymphon, 26, 30
charcoti, Nymphon, 3, 4, 28, 32, 34, 51
charcoti, Pentapycnon, 6
chierchiae, Tanystylum, 118
clarencei, Nymphon, 4, 28, 32, 35, 54

clausi, Ammothea, 5, 96, 109
Clotenia, 117, 119
Colossendeidae, 11
Colossendeis, 3, 7, n, 128
compactum, Nymphon, 28, 33, 34, 62
cornigera, Austropallene, 5, 85
cristata, Austropallene, 5, 86
cr is tat a, Pseudopallene, 86

Decolopoda, 3, 8

Decolopodidae, 8

denticulata, Nymphopsis, 5, 123

dimorpha, Pallene, 85

distensum, Nymphon, 28, 35

distribution, 4, 5

dohrnii, Clotenia, 119

drakei, Colossendeis, 4, 12, 22

elegans, Phoxichilidium patagonicum, var., 88
emaciata, Pallene, 84
encrusting organisms, 132
Endeidae, 93
Endeis, 93

frigida, Colossendeis, 4, 12, 16, 18
frigidum, Nymphon, 28, 35

gibbosa, Ammothea, 96

gigantea, Ammothea, 5, 95, 97

glabra, Pallenopsis, 88

glaciale, Austrodecus, 5, 115, 131

glacialis, Ammothea, 96

glacialis, Colossendeis, 4, 12, 21, 132

gracilipes, Ammothea, 95, 103

gracilipes, Colossendeis, 12, 21

gracilipes, Nymphon, 28, 31, 35

gracilis, Colossendeis, 12, 15

gracillimum, Nymphon, 4, 28, 31, 35, 39, 42

grande, Leio>iymphon, 109

grandis, Ammothea, 96

grossipes, Nymphon, 76

grossipes-mixtum, Nymphon, 128

hamatum, Nymphon, 28, 32, 34
"Harlequin" Pycnogon, 22
Heteronymphon, 79
hiemale, Nymphon, 4, 28, 31, 35, 39
hiemalis, Pallenopsis, 88
hoekianum, Tanystylum, 117, 118
hoekii, Achelia, 5, 110
hoekii, Ammothea, no

intermedia, Achelia, 5, 112

juvenilis, Austroraptus, 5, 114

INDEX

137

kempi, Heteronymphon, 4, 80
kentrodes, Tanystylum, 118
korotnewi, Nymphopsis, 124

lanare, Nymphon, 28, 34
Leionymphon, 94
lilliei, Colossendeis, 12
longicaiidatiim, Tanystylum, 118
longicoUum, Nymphon, 28, 34
longicoxa, Nymphon, 28, 32, 34
longispina, Ammothea, 5, 95, loi

magellanicum ?, Pycnogonum, 5, 6, 126, 131
magiiirostre, Pycnogonum, 126
margarita, Pallene, 4, 82
maxillara, Phoxichihdium, 132
media, Colossendeis, 12
megalonyx, Colossendeis, 12, 17, 18
megalonyx, Colossendeis, 18
mendosum, Nymphon, 28, 33, 36, 62
meridionalis, Ammothea, 96
minor, Ammothea, 5, 95, 103
mi7ius, Leionymphon, 103
multidens, Nymphon, 4, 28, 35, 75
muscosus, Nymphopsis, 124

nervous system of Decolopoda, 128
neumayri, Nymphon, 4, 28, 33, 35, 66
Nymphon, 3, 7, 26
Nymphon sp. ?, 4, 56
Nymphon sp. ?, 4, 58
Nymphonella, 79
Nymphonidae, 24
Nymphopsis, 3, 123

oedinotum, Tanystylum, 118

orcadense, Chaetonymphon, 63

orcadense, Nymphon, 4, 6, 28, 33, 35, 62, 63

orcadensis, Colossendeis, 12

ornatum, Tanystylum, 118

Oropallene, 85

Pallene, 3, 8, 82

Pallenidae, 82

Pallenopsis, 7, 87

? Pallenopsis sp., 4, 91, 131

parvula, Achelia, 5, 6, 113

patagonica, Pallenopsis, 5, 6, 7, 88

patagonicum, Phoxichilidium, 88

patagoniciim, Phoxichilidium , var. elegans, 88

paucidens, Nymphon, 4, 28, 32, 35, 48

Pentanymphon, 7, 24

pfefferi, Nymphon, 4, 28, 32, 35, 46

pfefferi, Tanystylum, 5, 6, 7, 119

Phoxichilidae, 82

Phoxichilidiidae, 87

Phoxichilidium, 92

pilosa, Pallenopsis, 3, 5, 87

pilosum, Phoxichilidium, 87

platylophtim, Pycnogonum, 5, 126

polaris, Austroraptus, 5, 114

Polyzoa, list of, 133

praecox, Austroraptus, 5, 114

proboscidea, Colossendeis, 1 1

proceroides, Nymphon, 4, 28, 31, 34, 36, 130

procerum, Nymphon, 28, 34

proximum, Nymphon, 28, 33, 36, 62

Pycnogonidae, 125

Pycnogonum, 125

rhinoceros, Pycnogonum, 5, 125
Rhynchothorax, 122
robusta, Colossendeis, 12
rugosa, Colossendeis, 12, 17, 18

scoresbii, Colossendeis, 4, 6, 7, 12, 18
scotti, Colossendeis, 4, 11, 12
serratipalpis, Achelia, 5, 113
signatum, Nymphon, 28, 35
spicata, Achelia, 112
spicata, Pallenopsis, 5, 90
spinosa, Ammothea, 5, 95, 103
spinosum, Leionymphon, 103
striata, Ammothea, 5, 95, 96, 132
striata}, Ammothea, 95, 104
striatum, Leionymphon, 96
stvligerum, Nymplion, 118
styligerum, Tanystylum, 5, 118
stylirostris, Ammothea, 5, 96, 106
stylops, Nymphon, 59
subtile, Nymphon, 4, 28, 32, 35, 43
synonymy of Nymphon species, 31

Tanystylum, 1 17

tenuipes, Nymphon, 4, 28, 31, 34, 37, 132

tetrapora, Ammothea, 5, 95, 99

tortipalpis, Colossendeis, 4, 12, 130

tridentatum, Nymphon, 48

Tunicata, 133

valida, Pallene, 85

villosum, Nymphon, 28, 35, 62

wilsoni, Colossendeis, 4, 12, 21

[Discovery Reports. Vol VI, pp. 139-164, Plates I-VI, December, 1932.]

REPORT ON PENGUIN EMBRYOS

COLLECTED DURING THE

DISCOVERY INVESTIGATIONS

By

C. W. PARSONS, B.A.

Lecturer in Zoology in the University of Glasgow

CONTENTS

Introduction P^S<^ H'

Stages of development ^4^

Comparative Studies in Development

The Brain H9

The Feathers ^S^

The Cartilaginous Skeleton iS4

The Heart ^57

Discussion ^°°

Summary ^"^

Acknowledgments i"2

List of Literature i"3

Plates I to VI following page 164

REPORT ON PENGUIN EMBRYOS

COLLECTED DURING THE

DISCOVERY INVESTIGATIONS

By C. W. Parsons, B.A.

Lecturer in Zoology in the University of Glasgow
Text-figs. 1-9, Plates I-VI

INTRODUCTION

OUR knowledge of the development of birds is securely based on the detailed in-
vestigations that have been made of the embryology of the common fowl. Amongst
an exceptionally long list of modern workers who devoted their attention to the problems
involved some names stand out pre-eminently. F. M. Balfour was publishing on the
subject in 1873, Duval gave us his Atlas d'Embryologie in 1889, Keibel and Abraham
pubhshed the Normentofel in 1900, and in 1908 Lillie brought out his Development oj
the Chick.

The last-mentioned books give evidence of the massive research extending over
twenty years on chick embryology, and it is surprising that in this period comparatively
few other species of birds were the object of similar enquiries. The duck naturally
had a place in the work of embryologists — next to the fowl it is the most accessible
material — and of less common forms penguin embryos have been the subject of a certain
amount of published work. Material brought home by the Scottish National Antarctic
Expedition was worked out by Waterston and Geddes (1909), and other penguin embryos
collected by the Deuxieme Expedition Fran9aise were the basis of a joint report by
R. Anthony and L. Gain in 191 5. Their Contribution a V etude de Vembryologie des
Spheniscidae is devoted to three main points : (i) the development of the skeleton of the
wing, (ii) the development of the skeleton of the foot, (iii) the feathering ("pterylosis").

The embryos described by Waterston and Geddes were fixed in formalin and were
not in a satisfactory condition for microscopic examination in any great detail. The
limitations imposed by these conditions are of course very great, and the report, as far
as the embryos are concerned, deals chiefly with external features. There was therefore
an urgent need for new material, and during the Discovery investigations an extensive
collection of the embryos of two species of penguins was made in the course of the
1928-9 and 1 930-1 seasons.

The majority of the embryos— sixty-four— were of the Gentoo penguin, Pygoscelis
papua, from South Georgia. The remainder— fourteen— were of the Ring penguin,
Pygoscelis antarctica, from the South Shetlands. They were obtained and preserved by
Mr F. C. Eraser, the eggs being taken at random and the embryos dissected off the
blastoderms by him and then fixed. He used two fixatives. In the 1928-9 season at
South Georgia only the eggs of Gentoo penguins were taken and twenty-six embrj'os

142 DISCOVERY REPORTS

were fixed in saturated mercuric chloride and 5 per cent acetic acid solution. The excess
mercuric chloride was washed out in iodine alcohol before storage in 70 per cent spirit.
Thirty-eight Gentoo embryos were collected in South Georgia in the 1 930-1 season,
together with the Ring penguin embryos from the South Shetlands. These were
preserved in Bouin's fixative and were also received in 70 per cent spirit. Both batches
of material arrived in splendid condition ; the corrosive-fixed embryos proved the better
material for sectioning purposes, while the Bouin-fixed embryos were the better for
dissection.

Owing to the method of collection there is no information about the precise length of
time that any particular embryo has been incubated.

STAGES OF DEVELOPMENT

The embryos make up a remarkably complete series of stages, extending from a small
Gentoo embryo 3-8 mm. in length to a large Ring embryo on the point of hatching that
measures 150 mm. from the tip of its beak to the end of its tail.

The earlier stages, which are all of Gentoo embryos, bear so close a resemblance to the
corresponding stages in the fowl that it will be convenient to define them in terms of the
Normentafel of Keibel and Abraham.

1 . The smallest embryo (Plate I, fig. i) compares with a fowl of 24 hours' incubation.
NT (i.e. Normentafel), stage 9.

Length, measured from the anterior limit of the medullary folds to the posterior limit
of the primitive streak, 3-8 mm. Length of primitive streak loooju.

Medullary folds dilate anteriorly to an exceptional degree and widen posteriorly
around the headward end of the primitive streak. The folds are not continuous behind
the primitive streak as they are in Peripatiis and Lepidosiren (cf. Kerr, 19 19, pp. 496,
497). Number of mesoderm segments 6.

Notochord {utc), visible for a considerable distance through the floor of the neural
canal, terminates anteriorly in a Y-shaped split.

Fore-gut {fg) developed symmetrically. Extends from the headward end of the
division of the notochord nearly to the level of the first mesoderm segment.

Proamnion lies in front of the head fold, and precociously developed cavities of the
splanchnocoele containing fluid, at the sides of the fold, are conspicuous features of the
embryo at this stage.

2. The next stage is comparable with a fowl of 48 hours' incubation (Plate I, fig. 2).
NT 13. Length 5 mm.

Medullary folds closed anteriorly, head already bending towards the ventral side.
Optic rudiments established. Otocysts also clearly visible. Number of mesoderm seg-
ments 18.

Heart S-shaped. Aortae {ao) paired anteriorly, give off right and left vitelline arteries
posteriorly.

Embryonic excretory organ established in a rudimentary condition. It may be seen in

STAGES OF DEVELOPMENT 143

transverse sections cut behind the level of the seventh mesoderm segment. The organ
originates as a ridge of mesodermal cells between the mesoderm segments and the lateral
mesoderm. The latter splits into the splanchnic and somatic layers and, at this stage, is
united to the mesoderm segments by a narrow isthmus of cells. The excretory ridge
first makes its appearance just dorsal to this isthmus. Posteriorly it is a solid rod of cells,
but at its anterior end it acquires a cavity foreshadowing that of the archinephric duct.
The amnion has grown over the head and body of the embryo to the level of the
vitelline veins. The exceptional distance between these veins and the vitelline arteries in
the penguin embryo at this stage is one of the most remarkable diff'erences between it
and the fowl embryo of the same stage.

3. Stage comparable with a fowl of 67 hours' incubation (Plate II, fig. i). NT 18.
Length 10-4 mm., measured as though the embryo was straightened out.

Head flexed at right angles to the body, brain regions — hemispheres, optic lobes, and
medulla oblongata — clearly distinguishable. Choroid fissure open. Otocyst still in con-
tact with the surface ectoderm. Olfactory pit. Number of mesoderm segments 39.

Pharynx distorted by visceral clefts (c). Of these, numbers I and II perforate the
ectoderm, numbers III and IV lie close to the surface but do not penetrate it. The hind-
gut is completed.

Rudiments of (i) the liver, and (ii) the lung. The liver rudiments arise as two pouches,
one on either side of the main venous trunk as it enters the heart. The lung rudiment,
arising as a constriction on the floor of the pharynx, is already divided at its hinder end
into two small pouches. These come behind the last visceral cleft and extend nearly to
the level of the apex of the heart. As they are simply prolongations of the enteric wall
their cavities are lined with endoderm. Owing possibly to the position of the embryo,
with its left side towards the mass of yolk, the left lung rudiment is retarded in its growth
in comparison with the right. This inequality is of course only temporary, and in the
later stages there is no obvious diflference in size between the two lungs.

The heart lies within a closed pericardiac cavity. The ventricle and conus regions are
clearly demarcated on the right side.

Limb rudiments have made their appearance: (/) fore-limb, [h) hind-limb.

Amnion complete. At its hind end it is attached to the serous layer by a short con-
nective {a) which recalls the longer amniotic funnels figured by Schauinsland (1903) in
embryos of Piiffiiius, Phaeton, and Diornedea.

Allantois rudimentary, just visible outside the body {at).

4. Stage comparable with a fowl of 4 days 8 hours' incubation (Plate II, fig. 2). NT 25.
Length, measured along the dorsal surface from the top of the optic lobes to the tip

of the tail, 20 mm. This is about 5 mm. longer than the corresponding stage in the fowl.
Brain regions with the exception of the cerebellum are all established.
Retinal pigment appearing in the eye, separation of the lens from the ectoderm.
Otocyst deeply embedded and only indicated at the surface by a shallow depression.
Olfactory pits wide open and forming spacious cavities.

144 DISCOVERY REPORTS

Visceral clefts III and IV still visible at the surface, numbers I and II enclosed by the
overlapping hyoid arch (hy) which extends upwards close to the otocyst.

Umbilical stalk complete. Yolk duct opens into the intestine. Allantois well developed.

Liver bilobed : a conspicuous organ lying close to the base of the heart and occupying
a large share of the available coelomic space.

Lungs pvriform. They are still simple sacs with a wide central cavity forming the main
intrapulmonary bronchus (the mesobronchus). At this stage the mesobronchi are be-
ginning to bud off channels dorsally: the ectobronchi (Fig. i a).

Pancreas: three rudiments, one situated behind the stomach region, the other two
arising as pockets from the liver diverticula. It is interesting to note that the threefold
origin of this gland is not always clearly indicated by three ducts in the aduh penguin.
Watson (1883, p. 198) records considerable variation in the number of these ducts, for
example in Eiidyptes.

Maxillary ridge {mx). Mandibular ridge (m). Limb buds (e.g. h) begin to show dif-
ferentiation internally.

Excretory organ: opisthonephros well established. Its ducts have acquired openings
into the cloaca.

5. Stage comparable with a fowl of 8 days' incubation (Plate III, fig. i). NT 31.
Length 31 mm. Brain with very prominent optic lobes. Eyes socketed in epithelium,
through which lachrymal glands perforate later. Rudiment of nictitating membrane.

External auditory meatus. Olfactory organs enclosed, external nares, rudiment of
the beak.

Operculum {op) fully developed as a backgrowth from the hyoid arch.

Stomach large, intestine short and scarcely coiled. Bile duct greatly distended
forming a gall bladder of quite exceptional size.

Lungs growing rapidly and bearing at their hinder ends the rudiments of two of the
air sacs : (i) the posterior thoracic air sac, (ii) the abdominal air sac. Internally the meso-
bronchus of each lung has budded the rudiments of eight entobronchi on the ventral
side as well as the four dorsal ectobronchi. Its prolongation beyond the borders of the
lung forms the rudiment of the abdominal air sac, which conceals one of the entobronchi
beneath it in the example figured (Fig. i b).

Limb buds (/) are marked externally by the developing digits at their tips. The dif-
ference in shape between the fore- and hind-limbs is becoming clearly established.

Note. In the corresponding stage of the fowl embryo traces of feathers are apparent
on the rump, on the ventral body-wall, and on the hind-limbs. There are no feather
rudiments at this stage in penguins.

6. Stage comparable with a fowl of about 10 days' incubation (Plate III, fig. 2).

NT 34.

Length 58 mm., including beak. Altogether about 20 mm. longer than the fowl at

this stage.

STAGES OF DEVELOPMENT

145

Brain. Rapid growth of the hemispheres and the commencement of the growth of the
cerebellum combine to change the contour of the head to a dome-like shape.

Eye. The wall of the sclera bears a number of papillae strongly recalling the sclerotic
ossicles so commonly seen in the fossils of extinct aquatic reptiles and present in the
sclera of many birds. In the penguin their incidence is apparently only of short duration
at this stage in development. When the eyelids have grown they disappear (of. papillae
conjunctiva sclerae in the fowl, Lillie, 1908, p. 280).

•mb

'--.v.icl

DtK

Fig. I. The lung and air sacs, a, stage 25 ; b, stage 31 ; c, stage 37; d, stage 34. a, b, r, and d, diagrams of
the right lung drawn from the right side; e, the same lung as d, dorsal view.

ab, abdominal air sac ; «?//, anterior thoracic air sac ; ce, cervical air sac ; do, dorsal papilla ; ecb, ectobronchus ;
enb, entobronchus ; id, interclavicular air sac; mb, mesobronchus ; pa, pulmonary arterj^; pb, parabronchus ;
pth, posterior thoracic air sac; st, systemic artery.

The external auditory meatus takes on the adult appearance.

All traces of the visceral openings on the neck have vanished. The intestine is thrown
into two spirals completely enveloping the rudiments of the pancreas. The spleen de-
velops on the omphalo-mesenteric artery.

146 DISCOVERY REPORTS

The lung is now well developed with numerous passages radiating from the ecto-
bronchi and entobronchi. These are the first of the parabronchi (Fig. i d, pb). Besides
the rudiments of the posterior thoracic and abdominal air sacs, the lungs bear the pro-
cesses which afterwards develop into the other three air sacs of the hatched bird —
cervical, interclavicular, and anterior thoracic. In addition to these permanent air sacs,
the lung buds out five other papillae on the dorsal surface at this stage, which do not
persist (Fig. i e, do).

It has long been known that the lungs of certain of the Reptilia expand into air
sacs. They are very much smaller in the chameleon for example than they are in modern
birds, but they are also more numerous (Heilmann, 1926, p. 128). It is probable there-
fore that the additional outgrowths on the dorsal surface of the lung at this stage
represent a more primitive condition.

Limb buds. Fore-limb elongating with extraordinary rapidity. The contour and cha-
racter of the future flipper is plainly seen. The hind-limb is digitate with webbing
between the toes.

Feathers. The first rudiments of feathers make their appearance as papillae growing
in the skin of the tail region. The tongue (/) is visible intra-orally.

7. Stage comparable with a fowl of about 13 days' incubation (Plate IV, fig. i). It is
beyond the stages included in the Normentofel, and for convenience it is assigned here
to stage 37. Length 85 mm., including beak. Eye protected by well-developed lids.

Feather rudiments distributed all over the body. They appear last on the fore-limbs,
and the rudiments in this situation grow into feathers exactly like those that develop on
other parts of the body. There is no reason to suppose, therefore, that they have ever
previously been modified for the purposes of flight. In the water, indeed, the fore-
limb is used entirely as a swimming organ and it seems reasonable to regard it as
a flipper rather than as a wing. Its length at this stage of development in relation
to the rest of the body is quite out of the ordinary (Plate V, fig. 2). The hind-limb
is also disproportionately large. Webbing extends nearly to the ends of the digits and the
latter are thoroughly protected at their tips by thick cushions of integument ("neo-
nychia" — Agar, 1909, p. 374), which prevent the possibility of damage to the amnion
or other delicate membranes.

Viscera. These are illustrated by dissections (Plate IV, fig. 2 and Plate V, fig. i).
A few points call for comment :

(i) The alimentary canal. The crop region {cr) is not really separate from the stomach.
It is a characteristic of penguins that the whole of the anterior division of the canal is a
continuum which is enormously distensible (Wilson, 1907, p. 43). The capacity of the
fledglings for food, and the necessity the adults are under of landing sufficient for their
own needs as well as for the voracious young are well known facts (Wilson, 1907, p. 44 ;
Levick, 1910, p. 80, etc.). It is not surprising therefore to find, quite early in develop-
ment, that the anterior part of the alimentary canal is exceptionally wide.

(ii) The vascular system. The heart and main blood vessels are typically avian. In

STAGES OF DEVELOPMENT 147

the neck region the extraordinary size of the jugular vein is correlated with large di-
mensions of the head in comparison with the body. In the region of the abdomen the
dorsal aorta gives off two allantoic arteries. Of these the left artery is well developed
and the right is degenerating. A remarkable feature of the blood system at this stage is
that the primitive arrangement of a double aortic root is maintained. There is no trace
of the reduction in size, which takes place subsequently, of the left side of the root
(Plate V, fig. i). The duct of Botallus (db), i.e. the connective between the pulmonary
arch and the aortic roots, is also present on both sides.

(iii) The excretory organ. The posterior segments of the opisthonephros have enlarged
and grown in a headward direction dorsal to the main mass of the excretory organ. The
character of the organ is therefore undergoing a change towards the purely metanephric
type found in the hatched bird. The metanephric duct — the ureter — is already dis-
tinguishable (Plate V, fig. I x), but the division of the organ at the level of the allantoic
artery (ala) does not mean a separation between the mesonephric and metanephric
portions. The two parts are in organic continuity dorsally and the division probably
foreshadows the groove that marks off" the fully developed metanephros into two lobes.
In this respect the kidney of adult penguins difters from that of the majority of birds in
which there are three lobes (Watson, 1883, p. 217).

Up to this point the developmental stages have been covered by the bulk of the avail-
able material— the sixty-four Gentoo penguin embryos. For stages beyond it, right up
to the time of hatching, the additional fourteen Ring penguin embryos were at hand.
Fortunately the youngest of the latter coincided in development almost exactly with the
oldest of the Gentoo embryos. A comparison between the two of them after dissection
revealed no difl'erences in their anatomy. The extent of the permanent air sacs, for
example, in the smallest of the Ring penguin embryos, shown in Fig. i c, would serve
equally well to illustrate their condition in the Gentoo embryo (Plate IV, fig. i).

8. The anatomy of the oldest of the Ring embryos, fledged in the coat of silky down
feathers with which it hatches, is represented on Plate VI, figs, i and 2. The first of the
figures shows a superficial and the second a deep dissection. The embryo is large, about

150 mm. in length.

The most noticeable difference between this stage and the preceding one, apart from
the fact that it is nearly double the size and has a complete covering of feathers, is the
inclusion of the yolk sac within the body-wall of the embryo. On the hind-limb the
claws are moreover uncovered, and the neonychia (ne) are wearing away their pomts.
Preparation for using the claws to destroy the embryonic membranes is therefore nearly
complete, and the appearance of the limb is in this respect in remarkable contrast with
the protected limb of the previous stage (Fig. 2). The horny bill with its sharp-pointed
egg tooth (et) has also undergone a great change in readiness for the act of hatching. 1 he
horn is deposited beneath the surface layer of the epidermis, the so-called epitrichium,
from a specialized zone of cells. It is exposed after the epitrichium is shed. The details
of the process and the histological changes leading up to it have been described from a

148

DISCOVERY REPORTS

series of embryo Eudyptes by Max Lewin (1903) and they apply equally well to the Ring
penguin.

The cartilaginous skeleton at this stage is rapidly under-
going replacement by bone. Ossification of the chief ele-
ments of the axial skeleton is proceeding and to some extent
also replacement of the basic cartilages of the skull.

The brain, with the completion of the cerebellum, takes on
the aspect and proportions of that of a typical bird.

The eye, as shown in Plate VI, fig. 2, develops a large
pecten {pe) on its floor and the eyelids close. The pecten
in most birds is single, and it first arises in the penguin as a
simple lamellate process. By the time the embryo is ready to
hatch, however, this sheet of tissue is divided into ten units
(the teeth of the comb) which adhere to one another only by
their tips (Fig. 3 a).

With regard to the viscera, the stomach lies on the left side
of the body and the much-coiled intestine is thrust dorsally Fig. 2. Left hind-limb, Gentoo
and to the right side of the coelomic cavity by the yolk sac. penguin embryo, stage 37

. . . (Plate IV)

The rectum opens into the cloaca, which receives the meta-

nephric ducts (Plate VI, fig. 2, x), genital ducts (Plate VI, fig. 2,3'), and the bursa

Fabricii. At the earlier stage (Plate V, fig. i) the bursa {bf) is a sac of large dimensions.

It has dwindled in proportion very considerably by the time hatching is imminent.

Rudiments of both gonads, with their ducts, are present in these later stages of the

development, although in penguins, as in other birds, the right ovar}' and right Miil-

lerian duct are atrophied in the adult.

The large proportions of the gall bladder

in relation to the liver is illustrated in Plate

VI, fig. I, gh. It is rather puzzling to explain

the excessive secretion of bile that this indi-

It may be associated with the habit of

cates.

a b

Fig. 3. Pecten from eye of Ring penguin embryo
on the point of hatching (Plate VI). a, elevation ;
b, plan.

bringing large food supplies ashore in the
breeding season and of feeding the young
with partly digested chyme. To accomplish this the parents open their bills and the
young thrust in their heads as far as they can (Murphy, 1915, p. 120; Matthews, 1929,
p. 587). On the other hand, it may have a purely physiological significance. The food
taken is chiefly composed of Euphausians and Cephalopods. Both have a high per-
centage of fat, for the hydrolysis of which in large quantities during digestion an
abundant supply of bile salts may be necessary.

In Plate VI, fig. 2, the lung is indicated by a dotted line and the blood vessels on the
dorsal side of it are drawn as if they were seen through it. It is interesting to note that
even at this late stage both right and left aortic roots remain in full operation in the
supply of blood to the dorsal aorta (Plate VI, fig. 2, ar). The persistent left umbilical

THE BRAIN

149

vein (Lillie, igo8, p. 368) was dissected from the mid-ventral line and is drawn (Plate VI,
fig. I, uv) pulled over to the left side with a piece of the body- wall to which it was
attached. The latter is cup-shaped because it enveloped a protuberance on the ventral
side of the yolk sac.

COMPARATIVE STUDIES IN DEVELOPMENT

THE BRAIN
This commences in penguin embryos, as in most other vertebrates, with the formation
of longitudinal medullary folds. They are wider apart both anteriorly and posteriorly
than the corresponding stage in the development of the fowl. The medullary folds
quickly unite to form a neural tube out of which the compartments of the brain dif-
ferentiate. By the time stage 31 is reached, i.e. NT 31, the hemispheres, optic lobes, and

w

XL---

SI Y 'm

Fig. 4. Right side of the brain of Gentoo penguin embryos : a, stage 3 1 ; Z>, stage 34; c, stage 37, drawn from
the brain of a Ring penguin embryo.

ce, cerebellum; hm, hemisphere; o, optic lobe; me, medulla oblongata; pi, pineal body; I-XU, roots ol

cranial nerves.

ISO DISCOVERY REPORTS

medulla oblongata are established, and on the roof and floor respectively of the thala-
mencephalon the pineal and pituitary bodies can be clearly discerned. The roots of
most of the cranial nerves may also be found by dissection (Fig. 4 a).

Between stages 31 and 34, the hemispheres grow rapidly and encroach on the
region of the thalamencephalon. They develop in association with the olfactory sense
organs at the time when the beak itself is elongating most noticeably (Fig. 4 b). Mean-
while the cerebellum {ce) makes its appearance behind the optic lobes. At first it is little
more than a ridge, and it develops late, as might be expected of an organ particularly
connected with the control and perception of fine movements.

Further increase in the size of the hemispheres and of the cerebellum results in the
disposition of the main regions of the brain as illustrated in Fig. 4 c. The cerebellum,
now prominent, is divided by three sulci, the hindmost of which demarcates a broad
basal segment bearing a pair of lateral processes known as the flocculi. These are par-
ticularly clear at this stage (cf. the dorsal view of the brain shown in Fig. 5 b), and they

a b c

Fig. 5. Dorsal surface of the brain, a, of a Gentoo penguin embryo, stage 25 ; 6, of a Ring penguin embryo,
stage 37; c, of a Ring penguin embryo on the point of hatching (Plate VI). Scale: a x 6,b x ^i, c ,< zh

are protected in a special recess which develops on the inner wall of the auditory
capsule. The close association of processes of the cerebellum with the otocyst in develop-
ment may have some functional significance, in view of the future relations of both
organs to movement. It is also noteworthy that in Crocodilia, the only reptiles which
have the cerebellum at all well developed, two sulci may be distinguished ; and in them
again the basal segment of the cerebellum bears a pair of flocculi (Heilmann, 1926, p. 108).
At the time of hatching (Fig. 5 c) the cerebellum has grown to such an extent that it
forms with the hemispheres the main mass of the brain. The efl'ect of this growth is to
push aside the optic lobes permanently, and the surface of the cerebellum becomes
marked by numerous sulci. This surface sculpturing makes a strong contrast between
the cerebellum and the hemispheres. The latter remain perfectly smooth without any
trace of convolution, and wedged in between them and the cerebellum may be seen the
tip of the pineal body.

THE FEATHERS 151

The pineal body develops precociously in the penguin. It is a short process of the
roof of the thalamencephalon as early as stage 25, with a ring of nervous tissue surround-
ing it which forms the habenular ganglion. The appearance of the brain as a whole at this
stage (Fig. 5 a) recalls that of the poorly developed brain of the lamprey. The hemi-
spheres and the optic lobes in both cases are almost of equal size, and the pineal bodies
on the dorsal surface of the two brains strikingly resemble one another.

In the stage figured in Fig. 5 b, the glandular tip of the pineal body is shown free from
the surface of the brain. It has developed a large peduncle and its size is temporarily
disproportionate to the rest of the brain. In this way it anticipates the subsequent
growth of the hemispheres and cerebellum (Fig. 5 c) and retains its position on the
dorsal surface of the fully developed brain. In its later stages the organ undergoes con-
siderable degeneration and does not at any time show traces of eye structure.

THE FEATHERS

Feather papillae first show themselves in the skin of the tail (stage 34, Plate III,
fig. 2, rt). According to Anthony and Gain (1915, p. 22) the later papillae arise in
isolated feather tracks which spread and fuse with one another as development proceeds.
It is a remarkable fact that the last feather rudiments to appear are those which occur on

the wings.

In actual distribution the feathers of penguins are unique. They occur uniformly all
over the body in the aduU and only one apterium persists (Pycraft, 1907, p. 2). In the
embryo, naturally, this exceptionally complete coat of feathers is preceded by an equally
complete covering of feather papillae (cf. Plate IV, fig. i). But between the definitive
feathers and the papillae there intervene no fewer than two generations of nestling
down (Pycraft, 1907, p. 11). The embryo hatches with the first of these, and with their
development from the papillae the following observations are alone concerned.

A fragment of skin bearing feather papillae and filaments was dissected from the wing
of the youngest Ring penguin embryo. It will be remembered that this embryo was at
approximately the same stage of development as the Gentoo embryo figured on Plate IV,
fig. I. The piece of skin was drawn (Fig. 6 a) and sections were then cut longitudinally
and transversely through the papillae and also through the filaments. The sections
through the papillae (Fig. 6 b, c) show the thickened epidermis overlying a dermal layer
which is condensed in the region of each individual papilla to such a degree that it has
an almost granular appearance. In the epidermis Wohlauer (1902) made out three
different types of cell: (i) a surface layer of flattened cells forming the "epitrichium",
(ii) a layer of intermediate cells, and at the base (iii) a layer of cylindrical cells which may
or may not be double. The same sequence of epidermal layers is characteristic of the
early stages in the development of the normal reptilian scale (cf. Kerr, 19 19, p. 71), and
the view that feathers can be derived from conical papillae of this type is therefore well
founded (Boas, 193 1, p. 563). There are strong arguments, however, against the opinion
that the feathers are in fact to be regarded as modified scales. These are set out else-
where (e.g. Pycraft, 1907, p. 3 ; Ewart, 1921, p. 634, etc.), and the controversy remains

iSa DISCOVERY REPORTS

unsettled. It would seem sufficient, and certainly it would be in accord with the facts,
to regard the granular type of skin reproduced in the earliest feather papillae as being a
likely characteristic both of the ancestors of birds and of the scaly reptiles. Just such a
skin as this is still found in some of the more primitive lizards, e.g. chameleons.

The papillae give rise to filaments by a process of exaggerated growth (Fig. 6 d).
The dermal element retains its granular appearance peripherally but has now in addition
a central pulp cavity in which there are numerous blood islands (Fig. 6 e, pc).

For the next step in development the skin of an older embryo is necessary. A suitable
fragment is illustrated in Fig. 6 f.

When the longitudinal and transverse sections of one of these filaments are compared
(Fig. 6^, //), it is at once evident that the further increase in length of the feather filament
has been accompanied by changes both external and internal. The external changes
result in the formation of a horny sheath, the internal in the isolation of eight solid epi-
dermal masses which are the rudiments of the barbs of the down feather. The dermis
plays no part in the construction of the feather other than that of supplying the vascular
requirements of the early stages and a formative zone for the epidermis. The boundary
between the two layers, as Wohlauer showed, is set up in the later stages of development
by the cylindrical epidermal cells. The middle layer of intermediate cells form the barbs,
and the outer cells make up the sheath.

Just about the time when the embryo is due to hatch the sheath of the feather filament
disintegrates. Its loss releases eight barbs, which thereupon fluff out producing the
complete feather. In preparation for this event each of the barbs pushes out lateral
processes, or barbules, which therefore also arise in the intermediate cell layer. In
section they are seen intervening between the solid cell masses forming the barbs and
the surrounding sheath. Fig. 6 i was drawn from a fragment of the skin of the oldest of
the series of Ring penguin embryos. In this one of the feather filaments has lost the
outer sheath so as to free the barbs. Fig. 6j shows a longitudinal section through the
base of such a filament, and k a transverse section at the same stage cut far enough away
from the surface of the skin to show barbules as well as barbs.

A further point concerning feather development deserves attention. It will be seen
that the filaments become deeply embedded in the dermal tissue and reach down to the
level of the subcutaneous musculature. This ensures a considerable degree of muscular
control, an obvious necessity for a homoiothermic animal which relies largely on such
outgrowths of the epidermis for the conservation of heat. It will also be observed that
each filament has, on one side of it or the other, an attendant surface papilla (Fig. 6f,fp).
These have exactly the same composition as the original feather papillae, but though so
closely connected with the filament they are not included in its sheath. It seems probable
that these are the rudiments of the filoplumes (Ewart, 1921, p. 626), which are retarded
in development amongst the down feathers but are nevertheless known to be present
in the adult (Pycraft, 1907, p. i,).

In addition to these papillae each feather filament is encircled by still smaller papillae
of the same kind, which are less regularly distributed (Fig. 6/, pi). Exactly comparable

^^p
^""^m

m^m.

a

.■■■^.

•-'Tnu

%

J

Fig. 6. Development of down feathers: diagrammatic.
a, skin fragment, Ring penguin embryo, stage 37; b, longitudinal section of feather papillae; c, transverse
section of a feather papilla; d, longitudinal section of a feather filament at this stage; e, transverse section

of the same.

/, skin fragment, Ring penguin embryo, approximately half-way through the incubation period; §, longi-
tudinal section of a feather filament at this stage; h, transverse section of the same.

/, skin fragment. Ring penguin embrj^o, at the close of the incubation period ;;, longitudinal section of the
base of a feather; k, transverse section at a higher level, to show barbules.

ba, barb; bu, barbule; ds, dermis; co, blood corpuscle; ep, epidermis ;//>, filoplume; mu, muscle; /)f, pulp
cavity;M papillae; sc, "scale"; sh, sheath. (Transverse sections after Wohlauer.)

154 DISCOVERY REPORTS

Structures may be seen in the skin of an embryo duck of about fourteen days' incubation
(Lament, 1922, p. 234). They are stated to be the forerunners of the true down, while
the large feather filaments are said to give rise to the natal down which precedes the
development of the true feathers.

In the penguin the subsequent history of these small papillae is different. They are
no longer visible at the surface by the time the embryo is ready to hatch, and no traces
of them are to be found in sections of the skin. Even if they actually abort, their tem-
porary presence is of great interest in view of the fact that the epidermis, over which they
are scattered, splits up later into a regular pavement of soft scales between the feather
filaments (Fig. 6 /, sc). They clearly grow independently of the " scales ", since they pre-
cede them, and as their numbers are greater they are not to be regarded as the precursors
of the individual " scales ".

THE CARTILAGINOUS SKELETON

This was investigated in the early stages of development by means of the Victoria
blue staining method for cartilage (van Wijhe, 1902, 1922). Fixation in formalin is
necessary for the successful application of this stain, and therefore only the Bouin-fixed
material was suitable for its use. The result of using the method on four different stages
of the Gentoo embryos is illustrated in Fig. 7. The stages are: a, stage 22; b, stage 26;
c, stage 31 ; and e, stage 34. It will be seen that at stage 22 practically no cartilage is
laid down. The stain picked out, however, the track of the future vertebral column and
it must be presumed that the first cartilage is deposited there.

More is seen in stage 26 (Fig. y b). In the head region the mass of cartilage in the
vicinity of the otocyst, the rudiment of the quadrate and an already strongly developed
cartilage supporting the lower jaw (Meckel's cartilage), are all clearly made out. In the
body the first cartilages of the limbs make their appearance, namely those that form the
scapula, coracoid, humerus, radius and ulna of the fore-limb; and the femur, tibia and
fibula in the hind-limb.

Having once started the cartilaginous skeleton is rapidly built up ; for by the time
stage 31 (Fig. 7 c) is reached many additional elements have been laid down. With the
development of the beak the premaxilla can be distinguished from the mass of cartilage
at the front of the orbit, and continuous with it there runs posteriorly a bar of cartilage
out of which is built the interorbital septum — the mesethmoid — and the orbitosphenoid.
Immediately behind the orbit is the rudiment of the alisphenoid, a cartilage perforated
by the foramen of the fifth cranial nerve. It is temporarily in the shape of an inverted Y,
and between the two arms of the Y nearest the back of the head lies the otocyst already
firmly enclosed by peri otic cartilages. One arm of the alisphenoid passes under the
quadrate and makes contact with another similar cartilage, bearing a foramen, the
exoccipital.

In the body the most conspicuous change lies in the development of some of the digital
elements of the rapidly growing limbs. Their exact succession has been carefully de-
scribed by Anthony and Gain (19 15), and little need be added to the subject here. It is,

THE CARTILAGINOUS SKELETON

155

however, interesting to observe that digits 2 and 3 make their appearance first in the
fore-Hmb, and in the hind-hmb digits 3 and 4. Moreover in the latter, the digits have

Fig. 7. The cartilaginous skeleton, cartilages shown black, a, stage 22 ; b, stage 26; c, stage 31 ; e, stage 34;
d, pelvic girdle, stage 31. Scale: ■; 5.

each two cartilages at this stage and metatarsals 2, 3, and 4 are quite well developed. In
the fore-limb metacarpals 2 and 3 may be identified.

156 DISCOVERY REPORTS

The pectoral arch has added to it a sternum and the beginnings of sternal ribs. The
pelvic girdle is built up by the union of three separate cartilages, the rudiments of the
ilium, ischium, and pubis respectively. Owing to the position of the limb this point is
not clear in Fig. 7 c, and the girdle has therefore been drawn separately in Fig. 7 d. In
penguins, as in other birds, the pubis takes part in the formation of the acetabulum
(Mannich, 1903, p. 29), and it is clear that it rotates in development into a position
parallel with the ischium, in the manner discussed by Goodrich (1930, pp. 206-12).

Between stages 31 and 34 (Fig. 7 e) no great changes occur in the cartilages other than
growth. A few additions are made to their number. In the head, for example, the supra-
occipitals make their appearance just dorsal to the otocyst, and the articular separates
from Meckel's cartilage. Two hyoid cartilages are developed under the lower jaw close
beneath the articular. They are just visible in stage 31, but they keep pace with the
development of the tongue and are not prominent while the beak is small. The pre-
maxilla becomes shifted forwards during the active growth of the beak so as to be
separated from the mesethmoid. It takes the external nares with it, and in the space
developing between it and the mesethmoid a number of frail ethmoid cartilages are
laid down.

Most of the cartilages of the appendicular skeleton are complete by stage 34. In the
flipper the first digit is represented, and in the hind-foot a small two-jointed hallux
grows on the inner side. The metatarsals of digits 2, 3, and 4 are quite separate and,
owing to the position of the limb, the normal tendency of the fibula to cross over from
the outer to the inner side of the leg is accentuated.

Leaving the Victoria blue material one other preparation is figured with a view to
showing the cartilages, especially of the limbs. This specimen (Plate V, fig. 2) was the
last of the Gentoo embryos. The cartilages were exposed by peeling off the skin and
muscles, a task rendered easier because the tissues were rather brittle after fixation in
corrosive acetic. Two striking features of the cartilages at this stage are (i) the great
breadth of the scapula, and (ii) the firmness and extent of the completed pelvic girdle.

The broad scapula is especially characteristic of the genus Pygoscelis (Watson, 1883,
p . 25) but, in comparison with normal flying birds, all penguins have an exceptionally wide
shoulder blade. This of course is a well-known feature of the skeleton of many other
aquatic vertebrates and may almost be regarded as an indication of the use of the fore-
limbs as organs of propulsion through water. The powerfully developed pelvic girdle of
these birds is no doubt correlated with their upright gait on land. It is common know-
ledge that they prefer this slow method of progression to lying prone and moving fast
with the help of the wings as well as of the legs. Other points of interest about the
skeleton include : (i) the large size of the patella, (ii) the early development of the uncinate
processes of the ribs, (iii) the presence of two extra sesamoids, just indicated in the
swollen brachial tendon, (iv) the commencement of the replacement of the cartilages
by bone (indicated in the figure by the little granular areas in the centre of the long
cartilages), and (v) the advanced state of the development of the clavicle. This latter of
course is not preformed in cartilage at all: it is not a replacement bone, and it is re-

THE HEART 157

markable that while ossification in the cartilages is only beginning, the clavicle, a com-
paratively unimportant bone, should be so far advanced.

THE HEART

The early development of the heart of the penguin follows the usual course. It is laid
down as a simple tube of splanchnic mesoderm enclosing a thin endocardiac lining. At a
stage comparable with a fowl embryo of about 46 hours' incubation, stage 12, the
cardiac tube seen from the ventral side is a simple bladder-like vessel, narrowing an-
teriorly and slightly bent upon itself (Fig. 8 a). By stage 13 the tube has taken on the
familiar S-shaped bend of the heart of a forty-hour fowl (cf. Hochstetter, 1906, p. 38).
The flexure of the tube is due to its rapid increase in length within the confined space of
the pericardiac cavity (cf. Kerr, 1919, p. 370), and the increase in diameter of the hinder
end of the tube, where it receives the great vitelline veins, is reflected in the dominant
appearance of this part of the heart when it is seen from the ventral surface (Fig. 8 b).

1

k

KJ

ail

a b c d

Fig. 8. The heart, ventral view, a, stage 12; b, stage 13; c, stage 18; d, stage 24. In d the head is cut
away to the level of the optic lobes in order to expose the heart, au, atrium; co, conus arteriosus; v, ven-
tricle.

In side view, the flexure is plainer (Fig. 9 a). Sections of the heart at this stage show
that in their active proliferation the cells of the endocardium keep pace with the growth
of the myocardium. They are more active in growth in the localities where the heart is
least afl'ected by the flexure.

The next stage in the development of the heart, illustrated in Figs. 8 c and 9 b, was
drawn from an embryo at about stage 18. It compares therefore with the heart of a fowl
of 67 hours' incubation. In the intervening period between this stage and stage 13 the
heart has undergone profound modification. The centre of the cardiac tube has grown
extensively, forming the ventricle, and its fixed anterior end has come prominently into
view in the mid-ventral line and forms the conus arteriosus (Fig. 8 c, co). The other end
of the cardiac tube, the atrial part, remains on the dorsal side of the ventricle. Its
cavity is not divided until a later stage, but the opening into it from the vitelline veins is
guarded by a well-developed valve. The endocardium between the atrial and ventricular

3-2

158 DISCOVERY REPORTS

parts of the heart has thickened, and in the conus locaUzed swelhngs of this hning
indicate the presence of three separate endocardial ridges.

In stage 24, Fig. 8 d, the conus has undergone an apparently independent flexure.
This is not paralleled in the development of the heart of the fowl and it seems to be of
the same nature as the N-shaped flexure that develops temporarily in the conus of
reptilian embryos (cf. the heart of an embryo crocodile, Kerr, 1919, p. 391)- At this stage
the proportions of the conus indicate its importance as a chamber of the heart. At the
headward end its cavity is practically divided into two by the approximation of two of
the endocardiac ridges. At its base three endocardiac cushions continuous with the
ridges form the rudiments of pocket valves that later develop in that region. With regard
to other chambers of the heart the interauricular septum begins to form, and between
the atrium and ventricle a partial septum is laid down. The atrio-ventricular opening is
bounded by two thick endocardiac cushions which serve as atrio-ventricular valves.

Lastly Fig. 9 c shows the side view of the heart of a penguin embryo at stage 31 . This
is comparable with a fowl of 8 days' incubation and represents at least four full days'
advance beyond the heart last described.

In the interim the septa of the conus, the ventricle, and the atrium pass through the
chief phases of their development and the heart attains practically to its adult condition.

The conus septum isolates the pulmonary from the aortic cavities. At the same time
the muscular conus wall begins to be incised along the line of the internal septum and
the separation between the pulmonary and aortic roots commences. In the adult heart
the entrance to the conus from the ventricle is guarded by a circle of three pocket
valves, and a similar circle of three pocket valves also occurs at the entrance to the
pulmonary artery. The relation of these valves to the original three endocardiac ridges is
therefore a matter of considerable interest. In the fowl (LiUie, 1 908, p. 352) the aortic and
pulmonary trunks each receive one of the three primary ridges. The remaining primary
ridge splits : one half passes into the aortic trunk and the other half into the pulmonary
trunk. Provision is thus made for four of the future pocket valves. The other pair of
valves — the third valve in each trunk — arise from specially developed secondary endo-
cardiac ridges. In penguin embryos the course of development of the conus valves is
identical with that described for the fowl.

With regard to the atrial and ventricular septa, they are discontinuous from the very
beginning. The atrial septum originates in the myocardium, forming the inner surface
of the atrial wall, and the ventricular septum is derived from that of the ventricle. Thus
the latter develops in association with trabeculae and is at first a meshwork of muscular
fibres. As the septa grow in length the atrio-ventricular opening constricts into two and
they finally unite with the bridge of tissue that partitions this opening. In this way the
division of the heart into four quite separate compartments takes place, the auricles and
ventricles come into being, and the heart takes on the characteristic appearance seen in
stage 31, Fig. 9 c.

One further point remains to be mentioned. In the complete heart the auriculo-
ventricular openings are both provided with important valves. The pair on the right side

THE HEART

159

are muscular sleeve-like structures which develop from the endocardium of the original
atrio-ventricular opening, while on the other side of the interventricular septum three
valves are formed in continuity with the wall of the ventricle itself. They have thus the
appearance of cordae tendinae, with tasselated ends free and bent inwards towards the
cavity of the left ventricle.

Fig.g. The heart and associated organs, side view. «, stage 13; 6, stage 18; r, stage 3.. «», atr.um;
CO, conus arteriosus ; k, opisthonephros ; /, Hver ; /«, lung ; oe, oesophagus ; sv, sinus venosus ; vc, left anterior
vena cava.

i6o DISCOVERY REPORTS

DISCUSSION

If the main features of the development of these penguins are reviewed certain points
are outstanding. The first is the close similarity between the early stages of their develop-
ment and those of other birds. Compared with the fowl the only conspicuous difi^erences
are due to the greater absolute bulk of the penguin embryo in some of the stages inde-
pendent of the yolk.

The embryos are obviously penguins after probably one-third of the incubation
period. The fore-limb, the webbed hind-limb, the powerful beak and the development
of feather papillae all over the surface of the skin, are all characteristic.

There are not many special peculiarities in the development. The pineal body is de-
veloped in the early stages to an exceptional degree, and the thymus gland is dis-
proportionately large at the time when feather filaments first make their appearance.
The history of the feathers seems to show that they are of a very primitive type. The
skin itself is very scale-like, and the heart at one stage has a reptilian flexure. But in
more fundamental features the development follows the normal course and does not
provide evidence of the sort that will settle the debatable problem of the true position
of penguins in the class Aves. Are they specialized for their very isolated manner of
life, or are they a primitive stock? On this question it was hoped that the embryology
would supply a final answer. It has in fact failed to produce irrefutable evidence one
way or the other, but it has supplied some new and suggestive hints. The embryos fail
to produce any sign of teeth, for example, but they do develop a most peculiar pair of
fore-limbs. The buds when they first arise are not exceptional, but as soon as the
cartilages begin to appear in them they grow at a prodigious pace. Long before the
embryo is ready to hatch, before even it assumes its coat of pre-natal down, the limbs
have completed their development as far as the skeleton is concerned. It is surely not
an accident that this should occur. The appearance of the trunk, with the limbs hanging
vertically and the back arched, as in Plate V, fig. 2, is extraordinarily reminiscent of the
aquatic reptiles of the Mesozoic Age. The whole body of the embryo at this stage is that
of a creature specially adapted, with long paddles, for an aquatic habit of life, and if this
adaptation — so early in development — is not merely a coincidence, it may be regarded
as supporting the opinions of those who do not agree that it is reasonable to regard the
penguin as simply a flying bird that took to water.

The opposite view, that the avian stock actually arose from aquatic fish-eating reptiles,
and that penguins were probably an early oftshoot of them that never succeeded in
getting into the free air (Kerr, 1919, p. 453), has not as yet attracted many followers.
It is nevertheless an idea admirably suited to the record of the primitive characteristics
of both adult and embryo penguins. The adult characters of (i) a generalized type of
skeleton, especially in the hind-limbs, and (ii) an unvaried type of plumage, have been
discussed often enough. The suggestive points that have arisen in the course of the
study of their embryology have been mentioned. There is, however, one special feature
of the development that oflFers new light on the problem. It is concerned with the small

SUMMARY i6i

papillae which make a transitory appearance on the dorsal side of the lung at stage 34
(Fig. I e, do). They are undoubtedly incipient air sacs which do not proceed with tlicir
development ; and in considering the possible reasons for their disappearance the whole
question of the function of air sacs in relation to an aquatic life arises. The ability to
give additional buoyancy to their possessor would be of great value to a surface-swim-
ming animal such as an aquatic reptile. It would ensure rapid return to the respiratory
medium after fishing in moderate depths and, in addition, supply a necessary air reservoir
to enable the animal to stay beneath the surface for a considerable time. As soon as the
muscular effort of the pursuit of prey under water relaxed, the main lung — with the
considerable assistance of the air sacs — would operate in bringing the predator to the
surface. In doing this, however, any air sacs on the dorsal side of the lung would be
hemmed in between the distended lung and the unyielding dorsal wall of the coelomic
cavity, and in time — their usefulness being unexploited — they would be eliminated.
This, perhaps, is the meaning of the appearance of dorsal papillae on the lung of a
developing penguin for a brief time, and if so, they may be accepted as evidence that the
bird developing them is in fact a primitive form.

SUMMARY

The development of Pygoscelis has been described from a very complete series of
embryos of two species, Pygoscelis papiia and Pygoscelis antarctico. The material con-
sisted of a collection of 78 embryos with the following characteristics :

Number of
Species Characteristics specimens

Pygoscelis papua Open medullary folds 2

Incomplete amnion 19

Limb rudiments just apparent 8

Head flexure maximal 10

Beak rudimentary iS

Feather papillae restricted in distribution 6

Feather rudiments uniformly distributed I

Pygoscelis antarctica Feather rudiments uniformly distributed i

Fully fledged in pre-natal down 13

The developmental stages have been correlated with the standard stages of the fowl
as fixed by Keibel and Abraham's Normentafel.

The general anatomy of each of the stages has been examined and the more out-
standing features reported. Particular attention has been paid to the following topics:
the development of the lungs and air sacs ; the brain ; the feathers ; the cartilaginous
skeleton; the heart. Other special features to which shorter reference has been made
include: (i) eye structure with reference to the pecten, (ii) the skin and claws, (iii) the
digestive organs, (iv) the organs of excretion.

The close similarity between the course of development of the viscera in the penguin
as compared with the fowl has been demonstrated. Where differences exist they are
differences of detail only (e.g. the double aortic root is persistent for a longer time; the

i62 DISCOVERY REPORTS

gall bladder develops into an organ of much greater proportions than that of the fowl
at the same stage, etc.).

With regard to the configuration of the body as a whole, and the underlying details of
the skeleton, the very early adaptation shown for an aquatic habit has been stressed, and
the contribution that penguin embryology has made to the question of the position of
these birds in relation to the avian stock has been discussed.

ACKNOWLEDGMENTS

My most grateful thanks are due to Professor Graham Kerr, whose unfailing interest
in the progress of this work and whose helpful criticisms of it have been a constant
encouragement to me.

The work of the artists who have been responsible for the illustration of the report is
an eloquent testimony to my indebtedness to them. Six of the figures in the Plates and
one text-figure (Fig. 2) are reproduced from drawings by Mr A. Kirkpatrick Maxwell.
All the remaining Plates and text-figures have been made for me by Miss C. Brown
Kelly, to whose enthusiasm for work in a new medium I must pay special tribute.

Finally I have to thank the Discovery Committee for the opportunity of examining
their material, and Mr F. C. Fraser in particular for his skilful collection and pre-
servation of it.

i63

LIST OF LITERATURE

Agar, W. E., 1909. On an embryonic appendage of the clincs of the Amniota, probably of an adaptative nature.

Anat. Anz., xxxv, pp. 373-80.
Anthony, R. et Gain, L., 1915. Embryologie des Spheniscidae. Deuxieme Expedition Antarctique Fran^aise,

1908-10. Paris.
Balfour, F. M., 1873. (i) The development and growth of the layers of the blastoderm, (ii) On the disappearance

of the primitive groove in the embryo chick, (iii) The development of the blood vessels of the chick.

Quart. Journ. Micr. Sci., xiii, pp. 266-90.
Boas, J. E. v., 1931. Fedem. Handbuch der vergleichenden Anatomic der Wirbeltiere. L. Bolk, E.Goppert,

E. Kallius und W. Lubosch, i, pp. 565-84. Berlin.
Duval, Mathias, 1889. Atlas d'Embryologie. Paris.
EwART, J. CossAR, 1921. The nestling feathers of the mallard, with observations on the composition, origin, and

history of feathers. Proc. Zool. Soc. Lond., 1921, pp. 609-42.
Goodrich, E. S., 1930. Studies on the Structure and Development of Vertebrates. London.
Heilmann, Gerhard, 1926. The Origin of Birds. London.
HocHSTETTER, [F.], 1906. Die Entwickelung des Blutgefdsssy stems. Hertwig's Handbuch der vergleichenden

und experimentellen Entwickelungslehre der Wirbeltiere, iii, pp. 21-166, Jena.
Keibel, F. und Abraham, Karl, 1900. Normentafel zur Entwicklungsgeschichte des Huhnes (Gallus

domesticus). Jena.
Kerr, J. Graham, 1919. Text-book of Embryology, II. London.
Lamont, Augusta, 1922. On the development of the feathers of the duck during the incubation period. Trans.

Roy. Soc. Edinb., liii, pp. 231-40.
Levick, G. Murray, 1910. Natural history of the Adelie penguin. British Antarctic (Terra Nova) Expedition,

Zoology, I, pp. 55-83.
Lewin, Max, 1903. Ueber die Entwickelung des Schnabels von Eudyptes chrysocome. Jena. Zeit. f. Naturw.,

xxxvii, pp. 41-81.
Lillie, Frank R., 1908. The Development of the Chick. New York.
Mannich, Hermann, 1903. Beitrage zur Entwickelung der Wirbelsdule von Eudyptes chrysocome. Jena.

Zeit. f. Naturw., xxxvii, pp. 1-40.
Matthews, L. Harrison, 1929. The birds of South Georgia. Discovery Reports, i, pp. 561-92.
Murphy, Robert Cushman, 1915. The penguins of South Georgia. Mus. Brooklyn Inst. Arts and Sci., 11,

No. 5, pp. 103-33.
Pycraft, W. p., 1907. On some points in the anatomy of the Emperor and Adelie penguins. National Antarctic

Expedition 1901-4, Zoology, 11.
Schauinsland, H., 1903. Beitrage zur Entwicklungsgeschichte und Anatomic der Wirbeltiere. Zoologica, xvi.
VAN Wijhe, J. W., 1902. A new method for demonstrating cartilaginous mikroskeletons. Proc. Kon. Akad.

van Wetensch. te Amsterdam.
1922. Friihe Entwicklungsstadien des Kopf- und Rumpfskeletts von Acanthias vulgaris. Bijdr. tot de

Dierk., xxii, pp. 271-98.
Waterston, D. and Geddes, A. Campbell, 1909. Report upon the anatomy and embryology of the penguins

collected bv the Scottish National Antarctic Expedition. Trans. Roy. Soc. Edinb., XLVII, pp. 223-44.
Watson, Morrison, 1883. Report on the anatomy of the Spheniscidae. Challenger Reports, Zoology, vii,

pp. 1-244.
Wilson, Edmund A., 1907. Aves. National Antarctic Expedition, 1901-4, Zoology, 11.
Wohlauer, Ernst, 1902. Entwicklung des Embryonalgefwders von Eudyptes chrysocome. Zeit. f. Morph.

u. Anthrop , iv, pp. 149-78.

PLATE I

Pygoscelis papua.

Fig. I. Embryo, stage 9 (No. i). ap, area pellucida; av, area vasculosa;
fg, fore-gut ; mf, medullary fold ; ntc, notochord ; ps, primitive streak.

Fig. 2. Embryo, stage 13 (No. 2). ao, aorta; Ze, left eye; ms, mesoderm
segment ; nt, neural tube ; ot, otocyst ; re, right eye ; va, vitelline artery ;
■vv, vitelline vein.

DISCOVERY REPORTS, VOL. VI

PLATE L

L

o

CO

3
O

z

a.

PLATE II

Pygoscelis papua.

Fig. I. Embryo, stage i8 (No. 3). a, sero-amniotic connective;
al, allantois; c, visceral clefts;/, fore-limb rudiment; h, hind-limb
rudiment; p, pericardiac wall; v, ventricle (of heart).

Fig. 2. Embryo, stage 25 (No. 4). al, allantois; h, hind-limb bud;
hy, hyoid arch; m, mandibular ridge; mx, maxillary ridge ;3', yolk stalk.

DISCOVERY REPORTS, VOL. ^1.

PLATE n.

t/3

o

CO

O

Z
ui

0,

PLATE III

Pygoscelis papua.

Fig. I. Embrj'o, stage 31 (No. 5). b, beak ; /, limb buds; op, operculum.

Fig. 2. Embryo, stage 34 (No. 6). n, nictitating membrane; rt, feather
papillae; sc, papillae sclerae; t, tongue.

DISCOVERY REPORTS, VOL. VI.

PLATE ni.

CO

O

CQ

o
z

Oh

PLATE IV

Pygoscelis papua.

Fig. I. Embryo, stage 37 (No. 7). n, nictitating membrane; rm, feather
papillae of wing.

Fig. 2. Dissection right side (right lobe of liver and part of intestine
removed), al, allantois; ala, right allantoic artery; ca, carotid artery;
cd, coracoid ; d, clavicle ; da, coeliac artery ; cr, crop region ; go, gonad ;
gp, genital papilla; hm, hemispheres; i, intestine; j, jugular vein;
k, opisthonephros ; /, liver; lu, lung; nt, neural tube; 0, optic lobe;
p, pericardiac wall; pa, pulmonary artery; pn, pancreas; r, rectum;
s, systemic aorta; sa, subclavian artery; se, sternum; st, stomach;
tl, thalamencephalon ; th, thymus gland; tr, trachea; vc, vena cava;
vn, vagus nerve ;3', yolk duct.

DISCOVERY REPORTS, VOL. VI.

PLATE IV.

I

Hi

3
O

z

ui

PLATE V

Pygoscelis papua.

Fig. I . Embryo, stage 37 (No. 7), deep ventral dissection, ala, allantoic
artery; bf, bursa Fabricii; br, bronchus; da, coeliac artery; cr, crop
region; crd, thyroid rudiment; da, dorsal aorta; db, ductus Botalli;
go, gonad ; gp, genital papilla ; ina, innominate artery ; k, opisthonephros ;
lu, lung ; p, pericardiac boundary ; pa, pulmonary artery ; pv, pulmonary
vein; rea, renal artery; x, ureter; y. Wolffian duct; s, MuUerian duct;
1-5, orifices of air sacs.

Fig. 2. Skeletal elements, left side, cl, clavicle; cu, cuneiform;
fb, fibulare; fi, fibula; ha, hallux; pb, pubis; pt, patella; ra, radius;
s, scapula; sm, sesamoid; ti, tibia; u, ulna; mt IV, metatarsal IV; /, //,
///, digits.

DISCOVERY REPORTS, VOL. VI.

PLATE V.

c/5
O

>-

a:

cc

3

o
z

.^r

PLATE VI

Pygoscelis antarctica.

Fig. I. Embryo No. 8, ventral dissection of bodywith head and neck
turned to right, ca, carotid artery; gh, gall bladder; /, liver; ne, neony-
chium; rl, retractor linguae; sa, subclavian artery; st, stomach;
sv, subclavian vein; td, thyroid gland; uv, umbilical vein; vn, vagus
nerve; ys, yolk sac.

Fig. 2. Deep lateral dissection, left side, ama, anterior mesenteric
artery ; ar, aortic root ; bf, bursa Fabricii ; br, bronchus ; ce, cerebellum ;
da, coeliac artery; era, crural artery; et, egg tooth ;|'o, gonad; hm, hemi-
sphere; i, iris; j, jugular vein; k, metanephros; /, liver; In, lens;
me, medulla oblongata; mn, genital duct; o, optic lobe; oe, oesophagus;
ol, olfactory lobe; pa, pulmonary artery; pe, pecten; pn, pancreas;
pv, pulmonary vein; r, rectum; sea, subcutaneous cervical artery;
sp, spleen; spc, spinal cord; x, ureter; j, genital duct. / and IX cranial
nerves.

DISCOVERY REPORTS, VOL. VI.

PLATE VI.

CO

O
ea

o
z

a,

[Discovery Reports. Vol. VI, pp. 165-190, Plates VII-XLII, December, 1932.]

ON THE DISTRIBUTION AND MOVEMENTS

OF WHALES ON THE SOUTH GEORGIA AND

SOUTH SHETLAND WHALING GROUNDS

By

STANLEY KEMP, Sc.D., F.R.S.

AND

A. G. BENNETT

CONTENTS

Introduction poge 167

Collection and analysis of the data

Positional data 168

Directional data ly^

Discussion of Results

South Georgia whaling grounds lyy

South Shetland whaling grounds 184

Plates VII-XLII following page 190

ON THE DISTRIBUTION AND MOVEMENTS

OF WHALES ON THE SOUTH GEORGIA AND

SOUTH SHETLAND WHALING GROUNDS

By Stanley Kemp, Sc.D., f.r.s., and A. G. Bennett
(Plates VII-XLII; text-figs. i-6)

INTRODUCTION

IN the year 1922 the Government of the Falkland Islands issued to the whaling
companies operating under licence in the Dependencies a form designed to give
information on the distribution and movements of whales in the whaling areas. The
companies were asked to give the approximate position of each whale that was killed,
and if whales were seen to be moving, they were asked to report the direction of move-
ment. The vast majority of the completed forms^^ relate to Blue and Fin whales taken
either off South Georgia or the South Shetlands, and it is with the data concerning these
species from these two areas that the present paper is concerned.

The information obtained in this way presents a number of interesting features. It
demonstrates the particular regions within the areas where whales are accustomed to
concentrate, and it shows how these concentrations have shifted during the course of the
season. It indicates with some clearness the seasonal variations which have occurred,
and, in South Georgia, how greatly the whaling grounds have been extended in recent
years. Evidence bearing on the movements of whales is available both from alterations
from month to month in the centres of concentration and from the direct observations
recorded by the gunners. The results from these two sources are not always in agree-
ment and it would appear that reliable data can only be obtained by whale marking.
The first series of marking experiments has unfortunately proved unsuccessful : but
methods have been modified in accordance with experience and renewed attempts will
be made in the season 1932-3.

In the present state of our knowledge it is difficult, if not impossible, to explain the
seasonal variations which these data have brought to light ; but as information accumu-
lates on the plankton and hydrology of the areas some of the underlying causes may be

discovered.

Few attempts have hitherto been made to ascertain the distribution of whales in
whaling areas. Detailed records are usually not kept by the whaling companies and
unless arrangements are made beforehand particulars cannot be obtained. Hjort, Lie

1 The forms are stored in the offices of the Falkland Islands Government at Port Stanley.

i68 DISCOVERY REPORTS

and Ruud^ have, however, pubhshed a valuable summary of the catches made in the
Antarctic in 1929-30 and 1930-1 by Norwegian pelagic factories, illustrated by charts
which cover an area extending from the South Shetland Islands to the Ross Sea. In the
northern hemisphere D'Arcy Thompson in 1928- analysed data obtained at the Scottish
whaling stations and published charts, based on records for seven seasons, which show
for each month the positions where Fin and Sei whales were captured. In earlier
papers, published in 1918,''' the same author gave charts showing the positions off the
Scottish coast where Nordcapers, Sperm whales and Blue whales had been taken.

COLLECTION AND ANALYSIS OF THE DATA

The forms, as already mentioned, give information of two kinds, namely the positions
where whales were killed and observations on the directions in which whales were
moving. The reliability of the records, their limitations and the methods employed in
analysis may be conveniently discussed under these two heads.

POSITIONAL DATA

The positions recorded for each whale are invariably given by the whalers as bearing
and distance from the land, and since the whale-catchers are often out for several days,
frequently in thick weather and many miles from the coast, they can only be regarded as
rather rough approximations. We think, however, that this does not seriously affect
the accuracy of the charts on which the results are summarized, for errors from this
source tend to be eliminated by the methods we have employed. It is, moreover, only
the main concentrations of whales that show features of interest, and these, so far as
they were discovered by the whalers, we believe to be correctly indicated.

As will be seen from Plates XXIII and XL the areas within which whales have been
killed, during the seasons with which these results are concerned, are very large. Off the
coasts of South Georgia the whaling grounds are some 300 miles both in length and
breadth, embracing an area of over 50,000 square miles. At the South Shetlands the
grounds extend from east and south of Clarence Island south-west to the Biscoe
Archipelago: the length is some 570 miles, the breadth in the middle about 160 miles
and the area over 80,000 square miles. It is, however, only on exceptional occasions —
when whales are uncommonly scarce or, in the South Shetlands, at particular times of
the year— that the more distant parts of the area are visited, and as will be seen from the
monthly charts there is thus great variation in the extent of ground covered by the

1 Hjort, Lie and Ruud, Norwegian Pelagic Whaling in the Antairlic. Norsk Vidensk.-Akad. Oslo,
Hvalradets Skrifter, No. 3, 1932.

- D'Arcy Thompson, On Whales Landed at the Scottish Whaling Stations during the years 1908-14 and
1920-7. Fisheries, Scotland, Sci. Invest., in, 1928.

3 D'Arcy Thompson, On Whales Landed at the Scottish Whaling Stations, especially during the years
1908-14, Part I. Scottish Naturalist, 1918, p. 197. Parts II, III, ibid., p. 223.

COLLECTION AND ANALYSIS OF DATA

169

whalers. When whales are plentiful in inshore waters the catchers will naturally not go
farther afield, and we have evidently no means of knowing how abundant whales may
be in the unexplored parts. When, however, a large area is covered we may be confident
that within its limits no considerable concentration of whales existed without being
detected. In using the charts as evidence of distribution of whales this limitation must
constantly be borne in mind.

During the seasons with which we are dealing whaling in South Georgia was con-
ducted from five land stations and, until 1929-30, from one permanently moored floating
factory, all situated on the north-east coast of the island (Fig. i). The whale-catchers

Willis Is.

SOUTH GEORGIA

LZ

gerite Rks.

Fig. I. South Georgia, showing positions of whaling stations.

were thus working from the same base throughout the season. At the South Shetlands
conditions were different, for the land is heavily glaciated, and it is only at Deception
Island, where the ice is melted by lingering traces of volcanic activity, that the erection
of a shore station has been possible. In addition to this station there is a considerable
number of floating factories, and these though often moored at Deception Island, were
frequently shifted to other harbours. Those most often used were Admiralty Bay in
King George Island, Melchior Harbour in Schollaert Channel and Port Lockroy in
Neumayr Channel (Fig. 2). Other harbours in Trinity Island and in Gerlache Strait
were occasionally used, and in some seasons, when ice conditions were exceptionally
favourable, the factories went as far south as the Biscoe Archipelago. On approaching
the whaling grounds in the spring the factories were often delayed by pack-ice, and not
infrequently they worked for a period along the ice edge in the neighbourhood while
awaiting the opportunity to enter Bransfield Strait. Since the whale-catchers do not
ordinarily go more than 100 miles from the parent ship, it is evident that the position of

,-o DISCOVERY REPORTS

the latter will be reflected in our charts of positions. During most of the period under
review there were, however, enough factories employed to cover the whole ground
adequately; and since their movements were determined by the positions in which
whales were most abundant, we think the limitations to the data are much the same as
those on the South Georgia grounds — we believe that within the area examined no
large concentration of whales could have existed without being discovered. Heavy

Fig. 2. The South Shetland whaHng area.

pack-ice will naturally limit the movement of the factories, but it will equally limit the
movements of the whales ; for while whales— Blue whales in particular— often concen-
trate along its margin, they will not penetrate it unless sufficiently open to permit the
passage of a vessel.^

1 Captain Jorgen 0re, a veteran among South Shetland whalers, has told us that at the beginning of season
1914-15 he met impenetrable pack-ice, of extreme thickness and with many included icebergs, far to the
north of the islands. He sent his catchers long distances afield to look for an opening; but they could find
none, and for some time he worked along the ice-edge, taking Blue whales in good numbers. Then one day
the whales vanished : he realized at once that a passage to the south must have opened, and next day he found
a way through to Bransfield Strait.

COLLECTION AND ANALYSIS OF DATA 171

Difficulties have been met with in interpreting some of the forms. Sometimes they
have not been filled in with sufficient care, and sometimes, especially in those from the
South Shetlands, the gunners have used place-names of their own which are not to be
found on any chart. For one reason or another we have had to reject numerous forms
as unintelligible, with the result that the number of positions recorded for any season is
much below the total number of whales actually killed.

In working up the data the positions were in the first place plotted, with due allowance
for magnetic variation, on large-scale charts, separate sheets being used for Fin and
Blue whales and for each month of the season. These charts were then divided up into
squares of convenient size and the number of whales killed in each square was counted.
For the South Georgia grounds the area of each square was a quarter degree of latitude
and half a degree of longitude (15 x 17-4 miles = 261 sq. miles). In the South Shetland
area, which is much larger, each square was half a degree of latitude and a whole degree
of longitude (30 x 27-6 miles = 828 sq. miles). Contours have been drawn^ on the
assumption that the number of whales in any one square was concentrated at its centre,
and it is these charts which are here reproduced on a greatly reduced scale.^ By this
method it is easy to prepare charts which will represent the total monthly captures
throughout the series of seasons, or, indeed, to summarize the whole of the records for
one species on a single chart. Such combined charts may show interesting features, but
the variation from one season to another is sometimes so great at South Georgia that
monthly totals cannot be employed to advantage.

Owing to an error in the original instructions Blue whales were not distinguished from
Fin whales in the South Georgia returns for the first half of season 1922-3. The records
for this season have therefore been omitted, the data covering eight seasons, from 1923-4
to 1930-1 inclusive. For this period we have been able to chart 29,266 positions, and of
these 16,005 relate to Fin whales and 13,261 to Blue whales. The numbers for each
month are shown in Tables I and II, with the totals for each season and the total number
taken in each season by the combined South Georgia companies.

Data are available from the South Shetlands since 1922-3, but in recent years the
returns are not satisfactory owing to the new methods which the whalers have adopted.
Prior to 1927-8 the factories were moored throughout the season in one or other of the
harbours situated in the South Shetlands and Palmer Archipelago. If the fleet en-
countered heavy pack-ice when approaching the islands in the spring, it often worked
for a short time along the ice-edge in the neighbourhood, moving into Bransfield Strait,
and to Admiralty Bay or Deception Island, as soon as the ice permitted. In 1927-8,
however, more ambitious pelagic operations began, and on an increasingly large scale.

^ Charts for individual seasons are contoured by tens up to fifty; those for monthly totals over the whole
series of seasons by twenties up to one hundred; those for grand totals by hundreds up to five hundred.

^ Recent surveys, made during the course of the Discovery Investigations, have led to considerable
revision of the coast line of South Georgia and the South Shetlands, and the alterations are incorporated in
the latest Admiralty charts. The work of plotting the data presented in this paper began before these surveys
were completed; the earlier charts have thus been used and we have thought it best to reproduce these in
their original form, without the recent modifications.

173

DISCOVERY REPORTS

Table I

Numbers of Fin whales on the South Georgia grounds with recorded

positions of capture

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

1929-30

1930-1

September

24

—

—

—

43

88

—

October

38

162

122

56

123

509

219

166

November

88

190

144

158

238

202

349

289

December

134

190

485

106

287

231

566

222

January

154

378

1 196

202

279

766

770

231

February

116

305

1 188

157

219

541

209

148

March

lOI

179

683

234

89

332

184

22

April

71

184

850

108

—

98

35

—

May

—

71

476

—

—

—

'

~

Total

702

1683

5144

1021

1235

2722

2420

1078

Total taken by
South Georgia Stations

1375

2017

5713

1 147

1357

3124

3404

1416

Table II

Numbers of Blue zvhales on the South Georgia grounds with recorded

positions of capture

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

1929-30

1930-1

September

October

November

December

January

February

March

April

May

52
143
356
205
176

39
48

17

197

590
730
420

377
373
309
137

3150

31

139

75

89

188

406

397
230

74

260

515
813
841

492

344
48

312
662

556

272

68

17

52
219
282

413
60

82

IS

31
97
59
93

32

188

237

236

117

50

Total

1019

1629

3313

1887

1 123

312

828

Total taken by
South Georgia Stations

1926

3514

1845

3670

2125

1564

492

1084

In that year most of the factories spent the early months of the season along the ice-
edge near the South Orkney and the South Sandwich Islands, returning later to Brans-
field Strait, and in the two following seasons they spent still more time on the Weddell
Sea ice-edge and only returned to Bransfield Strait at the end of February. These
movements of the factories are reflected in the returns of positions. Since 1927-8 a very

COLLECTION AND ANALYSIS OF DATA

173

considerable part of the catch has been made to the east, far outside the area with which
this report is concerned, and though the land station at Deception remained working
throughout the period, returns from the Shetland area for the first half of the season
have greatly diminished. So few records of Blue whales for 1929-30 and of both species
for 1 930- 1 are available that they are insufiicient for analysis.

The figures for the South Shetland area are thus for eight seasons of Fin whales,
1922-3 to 1929-30, and for seven seasons of Blue whales, 1922-3 to 1928-9^. The
total number of positions recorded is 18,571, of which 11,849 relate to Fin whales
and 6722 to Blue whales. The figures are shown in Tables III and IV, together with

Table III

Numbers of Fin whales on the South Shetland grounds
with recorded positions of capture

1922-3

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

1929-30

November

December

January

February

March

April

61
232
237
378
164

126
298
460

213
127

181
496

227
365
314

320

593

332
500
172

20
260
342
396
556
677

22
106
196
428
469

27

29

250

481

444

51

85

355

571

288

Total

1072

1224

1583

1917

2251

I22I

1231

1350

Total taken by
South Shetland factories

1985

1559

1948

2390

3442

Table IV

Numbers of Blue zvhales on the South Shetland grounds

with recorded positions of capture

1922-3

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

October

November

December

January

February

March

April

May

82
342

340
112
108

57

161

495

325

91

34

20
181
229

343
260

133
142

243
588

318
85
95

490

114

121

67

72

152

36

113
189
62
46
83
31

53
62

45
92
95

15

Total

1041

1 106

1308

1819

562

524

362

Total taken by
South Shetland factories

1993

1369

1642

2157

127s

1 The Blue whale data for 1929-30 have, however, been incorporated in Plate XL.

,y4 DISCOVERY REPORTS

the total numbers taken by the whaHng vessels in the South Shetland area. The
official returns, on which the latter are based, give only the grand totals taken by the
South Shetland fleet. The areas in which the whales were taken are not discriminated,
and it is thus not possible to include in these tables the total numbers taken in South
Shetland waters for seasons 1927-8 and onwards.

DIRECTIONAL DATA

In addition to recording the positions in which whales were killed, gunners of the
whale-catchers were asked to supply information on the directions in which Blue and
Fin whales were seen to be travelling, and on this subject an even larger quantity of
data has been collected than of that dealing with positions.

The value of the evidence obtained in this way is not easily estimated. A trained
gunner can without question distinguish Blue and Fin whales accurately at a distance of
several miles; but to determine the direction of movement, at least to an untrained
observer, may be a matter of considerable difficulty. Blue and Fin whales normally
blow three or four times in succession, and at each blow are visible at the surface for
only a few seconds: they then sound and remain below water for some 10 or 15 minutes.
If, as often happens, several whales are in company, they will generally rise to the surface
and blow simultaneously. When whales are not plentiful it is possible to be reasonably
certain that the same school is being kept under observation; but when they are
abundant — and they are sometimes very abundant on Antarctic whaling grounds — the
difficulties are greatly increased, so much so that one feels that even gunners with life-
long experience may be at fault. On the other hand the whale-catchers can only visit a
comparatively small area in the few days they are away from the whaling station or
factory ship ; they will presumably become acquainted with the general trend of whale
movement in the area, and if this has been correctly stated in their returns the observa-
tions should have some value.

As might be expected whales are not always travelling. On reaching a locality where
food is plentiful they will merely cruise in the neighbourhood, and it is to be imagined
that after a full meal they will show little activity. Large numbers of whales have been
entered by the gunners as not moving ; but the numbers shown as moving are usually in
excess and sometimes greatly in excess.

These data on whale movement are of different quality to those which record positions
in which whales were killed, for the latter, though they may be rather rough approxima-
tions, are ascertained facts, while the former might be regarded merely as an expression
of the gunners' opinions. If, however, it were impossible to ascertain direction of
movement and the data supplied are mere guess-work, analysis of the returns should
demonstrate that the observations are unreliable. As explained below, resultant direc-
tions for each month have been worked out, and since these show some measure of
consistency they are included in this report. The results, none the less, show wide
variation and some of them present very great difficulty in interpretation. It is not
improbable that many of the reported movements are merely those of whales cruising
slowly about the whaling grounds and are thus of little significance.

COLLECTION AND ANALYSIS OF DATA

175

In recording direction of movement the whalers used the eight cardinal and sub-
cardinal points of the compass, noting the numbers seen travelling in each direction.
As will be understood the numbers recorded are much higher than for positions : many
more whales are sighted than are killed, and movements of the same whales arc doubtless
recorded on several occasions. With these records, as with those for positions, a number
of forms have had to be rejected as inadequate or unintelligible.

In analysing the data the total numbers for each direction and for each month of the
season were found. The directions for each month were then reduced from eight to four
by subtraction of those travelling in opposite directions, and thence from four to two
by finding the components along two lines at right angles to one another. A resultant
direction and number was then obtained by means of traverse tables. We are indebted
to Sub-Lieut. R. A. B.Ardley, R.N.R., and Lt.-Cmdr. J. M. Chaplin, R.N., for their
valuable assistance in obtaining these resultant values. The resultant directions are
shown in Figs. 3-6 (pp. 182-3, 188-9) ^y means of arrows, and the length of each arrow
represents the resultant number of whales travelling in this direction expressed as the
percentage of the total number recorded as moving. Records with long arrows thus
have greater validity than those with short arrows. Results for any one month are
evidently of little value if only a small number of whales was sighted, and if the numbers
travelling opposite ways are so evenly balanced that only a small percentage is found
moving in the resultant direction the record will also be unimportant. We have omitted
all monthly records in which the recorded number of moving whales falls below 50, and
we have also omitted those in which the resultant number is less than lo per cent, of
the total number moving.

The numbers of whales for which directions were recorded are shown in Tables
V-VIII. The data from South Georgia cover eight seasons, 1923-4 to 1930-1, and
comprise observations on 55,210 Fin whales and 17,706 Blue whales. Those for the
South Shetlands, for 53,573 Fin whales and 10,761 Blue whales, were obtained during

Table V

Numbers of Fin whales on the South Georgia grounds
with recorded direction of movement

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

1929-30

1930-1

September

October

November

December

January

February

March

April

May

147
246

305
465
472
416
230

64

616

1138

519
1105

865
505
449
208

267

325
1411
3862

4015
1644
1898
1249

141

527
393
957
621
(866)
414

373

946

1272

956
605

293

109
1907

769

897
3586
1823
1204

350

428
950

1340
2156
3170

(719)

561

93

537
1133

857
1037

719
80

Total

2281

5469

1467 1

3919

4445

10645

9417

4363

176

DISCOVERY REPORTS

Table VI

Numbers of Blue whales on the South Georgia grounds

with recorded direction of movement

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

1929-30

I 930-1

October

78

319

(13°)

413

473

63

62

272

November

124

712

74

745

794

393

148

352

December

407

862

88

1380

638

505

86

431

January

267

525

182

1383

325

624

158

212

February

(170)

382

37°

834

71

102

—

73

March

57

458

539

432

—

119

—

—

April

85

321

207

—

—

—

—

—

May

171

90

~

~

Total

1 188

3750

1680

5187

2301

1806

454

1340

Table VII

Numbers of Fin whales on the South Shetland grounds
zvith recorded direction of movement

1922-3

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

1929-30

1930-1

November

116

—

—

—

—

—

—

December

151

49°

884

1373

844

118

144

210

237

January

559

1793

2840

3756

1012

461

"3

536

(252)

February

1088

2863

1 140

1440

1541

1070

1309

1685

540

March

1037

932

1694

2111

2330

1754

2403

3110

562

April

(334)

417

1346

456

(1929)

1592

1734

1 194

73

Total

3169

6495

8020

9136

7656

4995

5703

6735

1664

Table VIII

Numbers of Blue whales on the South Shetland grounds
with recorded direction of movement

1922-3

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

1929-30

1930-1

November

99

198

408

330

(143)

181

—

54

December

324

713

449

944

178

307

126

95

124

January

457

425

609

560

62

118

122

107

64

February

189

113

488

(92)

(54)

—

78

74

(75)

March

155

—

211

(118)

167

144

153

—

III

April

118

■ —

231

675

—

50

194

—

—

May

—

—

—

—

—

—

74

—

—

Total

1342

1449

2396

2719

604

800

747

276

428

DISCUSSION OF RESULTS i77

the nine seasons 1922-3 to 1930-1. In these tables numbers which have yielded a
resultant value of less than 10 per cent and are thus not represented graphically in
Figs. 3-6 are enclosed in brackets.

DISCUSSION OF RESULTS
SOUTH GEORGIA WHALING GROUNDS

The whaling grounds at South Georgia are visited annually during the southern
summer by both Blue and Fin whales. Seasons vary greatly: in some years Fin whales
will be taken much more abundantly than Blue, in some years vice verso ; in some years
both species are plentiful and in some both are scarce. These marked fluctuations are no
doubt mainly dependent on alterations in the environment, and it appears probable that
the predominance of one species over another is determined, at least in some measure,
by the position of the ice-edge and the temperature of the water. Interesting evidence
bearing on this point has recently been published by Sir Sidney Harmer.^

Both species, however, have the same objective in visiting the island. They do so m
order to make use of the abundant food supplies that are to be found there. And since
the food of Blue and Fin whales is the same, and its presence appears to be largely
determined by a system of currents which varies but little, it is to be expected that no
differences will be found in the distribution of the two species on the whaling grounds.
That this is true is shown by Plate XXIII, in which all recorded positions of capture
over a series of eight years are incorporated. It will be seen from the charts that the two
species are taken over similar areas and that there is a remarkably close agreement in the
contours illustrating regions of concentration.

That the principal whaling grounds lie to the north-east of South Georgia is to be
explained on hydrological grounds. It is due to the system of currents which prevails in
the neighbourhood of the island. One of the currents, which comes from the Pacific,
flows through Drake Passage to the south of Cape Horn, and taking a north-easterly
direction passes to the west of South Georgia. As it progresses northwards it comes
more and more within the influence of the westerly winds and thus tends to take a more
easterly course. The second and more important current is of Weddell Sea origm. It
passes round the south-eastern end of South Georgia and then sweeps in a north-
westerly direction along the coast. Before, however, it reaches the northern end of the
land it meets the first current and the prevailing westerly winds and is reflected back-
wards to take a north-easterly course. The direction of the Weddell current is thus
largely due to the presence of South Georgia and to the position which the island occupies
athwart the region where the two streams converge.

Euphausia superba, the exclusive food of Blue and Fin whales in the Antarctic, is a
pelagic prawn which drifts northwards to South Georgia from higher latitudes, the

1 Harmer, Southern Whaling, Proc. Linn. Soc. London, 1931, p. 131.

178 DISCOVERY REPORTS

greater part coming from the Weddell area. The whaHng grounds he in the lee of
South Georgia, and it is here, in the shelter of the island, that the Euphausians are able
to congregate.

Turning now to the records for individual seasons, illustrated for Fin whales in
Plates VII to XIV and for Blue whales in Plates XV to XXII, it will be seen at once that
there is great variation in the positions of the centres of concentration. Some part of
this variation is no doubt to be attributed to irregularities in the arrival of different
schools of whales. It has been ascertained at the Marine Biological Station^ that there is
no steady flow of whales on to the whaling grounds during the course of the season ; on
the contrary, the whales tend to arrive in batches, usually rather closely aggregated, of
one species, and often with one or other sex or with immature predominating. As some
schools arrive others may be leaving, and it is probable that certain schools remain
longer on the grounds than others.

But when allowance has been made for these considerations, there still remain certain
obvious differences in the distribution of whales in the seasons under review, and it
appears that the most significant difference is to be found in the latter half of the season.
In some years, in January and February, there is a very definite movement to the south-
western side of the island, while in other years no trace of this movement is to be seen.
In seasons 1924-5 and 1925-6 (Plates VIII, IX, XVI, XVII) there is no trace of this
southward movement in either species, both of which show in the later months a
remarkably heavy concentration close inshore to the north-east of the island. In
1923-4 and 1927-8 (Plates VII, XI, XV, XIX) both species show a movement towards
the south-west. In 1926-7, when Blue whales preponderated, the concentrations of
this species (Plate XVIII) shift round the ends of the island in the later months, but the
movement is less marked in Fin whales (Plate X). In 1928-9 the abundance of the two
species was reversed: Fin whales, which were very numerous, moved to the south
(Plate XII), while this movement is not clearly shown by the smaller numbers of Blue
whales (Plate XX). The data for seasons 1929-30 and 1930-1 do not give conclusive
results : whales were scarcer and the area covered by the catchers had greatly increased ;
but except for Blue whales in 1929-30 there are indications that both species tended to
move to the south in February and March.

Summarizing the observations it will be seen that when either species occurs in
abundance it has moved to the south-western side of the island in the latter half of the
season, and that the only exceptions to this statement are the two outstanding seasons
1924-5 and 1925 6. The dense inshore concentrations in these two seasons are perhaps
to be explained in terms of hydrology and plankton; but unfortunately the research
ships began their work at the end of the 1925-6 season and full surveys of the conditions
on the South Georgia whaling grounds only date from 1926-7.

Sir Sidney Harmer, in his valuable account of " Southern Whaling",- has shown that

1 Mackintosh and Wheeler, Southern Blue and Fin Whales, Discovery Reports, I, pp. 458-60 (1929).
- Loc. cit., p. 131.

DISCUSSION OF RESULTS i79

there is a definite correlation between the mean temperature of the air at South Georgia
in September and the order in which Blue and Fin whales reach their maxima in the
following season. With low temperatures the Blue whale reaches its maximum first and
with high temperatures the Fin. Working on these data he has been able to classify the
seasons into three groups, but this grouping does not agree with that based on whale
movements. The two exceptional seasons as regards movement do not appear in the
same groups ; 1924-5, with low September temperature, falls in Group I, and 1925 with
high September temperature in Group III. This lack of agreement is perhaps to have
been anticipated, for the relation which Sir Sidney Harmer has discovered is between
events which occur in the early part of the season and it may well be that whale move-
ments in the latter part are due to other causes.

Another feature of interest in the South Georgia returns is the evidence of the great
extension of the whaling grounds in recent years. Since whaling began in 1904 the
whaling companies have made constant improvements in their whale-catchers, and with
increased size and horse-power and greater speed they have been able to work farther
afield.

If the South Georgia data for the eight seasons are divided into two four-year periods
they yield the results shown in Plate XXIV. Results for Fin and Blue whales are charted
separately, and as will be seen they show a very close correspondence. The area covered
by the whale-catchers during seasons 1923-4 to 1926-7 is indicated by a fine dotted
line, while that for seasons 1927-8 to 1930-1 is limited by a continuous line. The position
of the main concentration in each of the four-year periods is also shown.

It will be noticed at once that in each of the species the centre of concentration has
shifted some 40 or 50 miles to the east, but with the great variation between individual
seasons and the comparatively short periods involved, there must be some uncertainty
as to the significance of this movement. A change in the centres of concentration may
clearly be brought about by a change in the position of the shoals of Euphausians on
which the whales feed, and this in its turn will probably be dependent on changes in the
hydrology of the area. For the later seasons some information on these subjects will be
available when the surveys made by the research ships have been worked up. If,
however, average conditions in the environment were comparable in the two four-year
periods, it would appear that the alteration in the centres of concentration can only be
explained as a response to intensive whaling : the whales have been scared and no longer
venture so close to the island.

The considerable extension of the grounds during the recent four-year period, and
the fact that with the same number of whale-catchers fewer whales have been taken,
lends support to the generally held opinion that whales are now less abundant than
formerly. The data for the two periods may be summarized as shown in Table IX.

The figures indicate heavy depletion of the stock of Fin whales and still heavier
depletion of Blue whales; but in view of the short periods which are considered and the
great seasonal fluctuations in abundance, they cannot be regarded as reliable and calcula-
tions derived from them can only be misleading. The greatest catch of Fin whales ever

i8o

DISCOVERY REPORTS

made on the South Georgia grounds (5713) was in season 1925-6, and the greatest catch
of Blue whales (3670) in 1926-7. Both these maxima fall in the first of our four-year
periods. The average number of catchers is approximately the same in each period; but
in the second much more time was occupied in towing whales long distances to the
shore stations and in consequence less time was spent in hunting.

Table IX

Total number

Average
number of

Approximate

Average

of whales

area of

number of

taken at

whaling

whales

South Georgia

catchers per

grounds

killed in

stations

season

(sq. miles)

100 sq. miles

Fin whales

1923-4 to 1926-7

10,252

2475

27,500

37-3

1927-8 to 1930-1

9.301

25

50,500

i8-4

Blue whales

1923-4 to 1926-7

10.955

24-75

25,000

43-8

1927-8 to 1930-1

5.265

25

41,000

12-8

The results cannot be interpreted as giving more than a strong suggestion that there
has been a large reduction in the stock, and the indication is that Blue whales have
suffered more heavily than Fin. It is perhaps significant that the great development of
pelagic whaling along the Weddell Sea ice-edge falls within the second four-year period
and that in this area the incidence of capture falls more heavily on Blue whales.

The figures in the last column of Table IX give only the average density of whales
killed on the South Georgia grounds. It will be evident that they were taken in much
greater numbers in the centres of concentration and in much smaller numbers on the
fringe of the whaling grounds. The greatest concentration during the eight seasons was
of Fin whales in 1925-6. In April alone of this season 92 Fin whales per 100 square
miles were killed a short distance from the mouth of Cumberland Bay, and in the same
area the catches for the whole season amount to 2-8 Fin whales per square mile of

whaling ground.

Inspection of the charts for the different seasons gives the impression that the whales
generally pass across the grounds in an easterly direction. The first whales of the season
are usually taken to the north or north-west of the island. As they increase in number
their centres of concentration shift to the east and later, in most seasons, they move
towards the southern side of the island. At the end of the season there is considerable
variation in position, but the tendency is for whales to occur in the south-east. In some
of the charts (Fin whales, 1928-9, Plate XII, afltords the best example) the disappearance
of a school in the early part of the season and its replacement by later arrivals can be
traced.

DISCUSSION OF RESULTS

i8i

The impression of a general movement towards the east is borne out in some measure
by examination of the directional data, which are summarized, in the manner already
explained (p. 175), in Figs. 3 and 4. If the monthly records of direction shown in these
figures are resolved into their components along the four cardinal points of the compass
the results are as shown in Table X.

Table X

Number

of
monthly
records

Components

North East

South

West

Fin whales

Number
Percentage

57

39

34

43
31\

18
16

14

-'-2

Blue whales

Number
Percentage

47

21

22

43

46

26

28

4

It will be seen that in each species the highest number of components is in an easterly
direction. In Fin whales there is also a strong northerly component, whereas in Blue
whales the southerly component is more pronounced than the northerly.

In all the data hitherto discussed a very close correspondence has been found between
the two species of whale. They concentrate in the same areas, and their movements in
any season, as determined from monthly records of capture, are usually very much the
same. But in these directional data differences between the two species are apparent.
The most noteworthy difference is in the records for the earlier months of the season,
in which Fin whales were observed to be moving towards the north, while Blue whales
were moving east. In this respect the records for both species show a marked degree of
consistency.

For this unexpected result we are not able to offer any satisfactory explanation. It
might be thought that movements at the beginning of the season would afford an
indication of the direction from which whale stocks arrive on the South Georgia grounds.
The general inference from other data would be that they arrive from the west, and the
strong easterly movement of Blue whales in the early months is in agreement. But it is
very difficult to believe that Fin whales arrive from the south. The majority of them have
almost certainly passed the winter in warmer waters to the north. If they reach South
Georgia from the south, they must first have visited the South Orkney Islands : it seems
most improbable that they would take so circuitous a route, and if they had done so we
should expect to find that the whalers killed them on the south side of the island at the
beginning of the season. The strange difference in recorded movement between the two
species is thus unexplained, and it appears probable that the direction from which the
schools arrive on the whaling grounds is not to be ascertained from these records.

l82

DISCOVERY REPORTS

1923-4

1924-5

1925-6

1929-30

0 10

30

— t—

50

— t—

Fig ^ Recorded movements of Fin whales on the South Georgia grounds. The direction of each arrow is

the'resultant direction of observed whale movements; the length of the arrow indicates, as shown in the

scale the resultant number moving in this direction, expressed as the percentage of the total.

DISCUSSION OF RESULTS

183

u
o

01

0)

J5

n

E

F

OJ

«j

Q.

1923-4

1924-5

1925-6

1926-7

1927-8

1928-9

1929-30

1930-1

50

70%

Fig. 4. Recorded movements of Blue whales on the South Georgia grounds. The direction of each arrow is

the resultant direction of observed whale movements; the length of the arrow indicates, as shown m the

scale, the resultant number moving in this direction, expressed as the percentage of the total.

3-2

,8^ DISCOVERY REPORTS

During the middle months of the season the directional data, though giving a general
movement between north-east and south-east, show great variation, and frequently
there are marked differences between the two species. Seasons 1924-5 and 1925-6,
which were exceptional as regards distribution, do not differ conspicuously in recorded
directions, but in 1925-6 a regular north-easterly movement for both species is more
clearly shown than in any other season. In the last months of the season a reversal of the
general easterly trend is sometimes found. In March and April of the 1923-4 season
there is a marked westerly movement of both Blue and Fin whales. A similar change in
direction is also seen in the Fin whales returns for the later months of 1927-8, 1928-9
and 1930-1 ; for these months the Blue whale data are for the most part inadequate, and
it is not possible to compare the two species.

SOUTH SHETLAND WHALING GROUNDS

The area covered by the South Shetland whalers, as will be seen from Plate XL, is
very large, extending from the east and south of Clarence Island, through Bransfield
Strait and south-west to the Biscoe Archipelago. The two charts in this plate incorporate
all recorded positions for Blue and Fin whales over a series of eight seasons, and it will
be seen that they give the same result as those from the South Georgia grounds : the
distribution of the two species is closely similar and the centres of concentration occur in
the same places. The two chief concentrations are at the northern and southern ends of
Bransfield Strait, one east-south-east of King George Island and one south of Deception
Island. There are also smaller concentrations to the west of Smith Island and north-
west of Antwerp Island.

It is to be noted that the two main concentrations are not found throughout the
duration of the season. The concentration off King George Island is for the most part
found in the opening months of the season, while that near Deception Island usually
occurs at its end.

In discussing the South Georgia results we were able to show that the position of the
main concentration of whales may be explained by the hydrological conditions which
prevail in the neighbourhood of the island and the accumulations of whale food which
result from them. It is more difficult to account for the South Shetland concentrations
on these lines. The main current in this area runs in a north-easterly direction from the
Bellingshausen Sea through Drake Passage. One branch of this current enters Brans-
field Strait north of Brabant Island and flows with increased speed along the northern
side of the strait. On the southern side a weaker reverse current is found, and this
comes from the Weddell Sea and flows round Joinville Island to take a south-westerly
direction. Euphausians, both adult and young, have sometimes been found in abundance
in the strait and some of the shoals may be of Weddell Sea origin. Conditions in the
strait, both as regards plankton and hydrology, are very complex, and until the data
obtained by the research ships has been fully worked up, no complete explanation of the
problem can be attempted.

DISCUSSION OF RESULTS 185

Examination of the data for individual seasons, illustrated for Fin whales in Plates
XXV to XXXII and for Blue whales in Plates XXXIII to XXXIX, shows that there is
very great variation in the position of the main concentrations, and it is by no means
easy to ascertain the normal movements of the schools. Two points, however, seem
tolerably evident. It appears (i) that the whales first appear in the northern or north-
eastern parts of the area, and in the earliest months of the season are found either in the
neighbourhood of Smith Island or off King George Island, and (ii) that at the extreme
end of the season there is frequently a very close concentration in Bransfield Strait. In
the middle of the season the concentrations show great variation and they are sometimes
well to the south of these positions.

These facts, which appear on the whole to be warranted by the data, find their most
simple explanation in the theory that the whales arrive on the grounds from the north or
north-east, pass along the coast to the south-west and at the end of the season return
towards the north-east. The data bearing on this theory can best be examined by in-
corporating all the returns for each month in a single chart : this has been done for Fin
whales in Plate XLI and for Blue whales in Plate XLII.

Considering first the Blue whales, it will be seen that the species arrives on the grounds
in November, 1 increases in number in December, and that by January the concentra-
tions have shifted decisively towards the south-west. In February and March the whales
remain rather widely dispersed over the same area, but in April there is a heavy con-
centration in Bransfield Strait.

The monthly aggregates for Fin whales (Plate XLI) are similar to those for Blue; but
the species arrives later at the South Shetlands, not appearing in any numbers until
December, and the schools do not move so far south as those of the related species. In
April there is the same heavy concentration in Bransfield Strait.

The Blue whale is known to go farther south than the Fin ; it is most plentiful in the
neighbourhood of pack-ice and the distance that it travels on its southern migration is
no doubt determined by the latitude of the ice-edge — a factor which varies greatly from
season to season. In some years therefore, when the pack-ice is well to the north, Blue
whales may be expected to remain in the South Shetland area, whereas in others, with
the ice-edge far to the south, the majority of them may pass through the grounds and for
a time will be beyond the reach of the whalers. The movements of Fin whales are also
in all probability determined in some measure by the position of the ice-edge ; but since
the normal distribution of this species lies a little to the north of that of the Blue, the Fin
whales will less often penetrate beyond the limits of the whaling area. That such a
movement as this can actually occur appears to be indicated by events which occurred in
the 1925-6 season. The statistics of South Shetland whaling published by Harmer' show
that in this season the monthly catches of both Blue and Fin whales yield a conspicuously
bimodal graph, with peaks in December and April for Blue whales and in January and

1 In some seasons Blue whales have been taken in October, but the numbers of recorded positions are so
few that this month has not been included in Plate XLII.

- Harmer, Southern Whaling, Proc. Linn. Soc. London, 1931, pp. 114, 159, graph 4.

i86 DISCOVERY REPORTS

March for Fin whales. Blue whales were scarce in February and March and Fin whales
in February, facts which are also indicated in Plates XXVIII and XXXVI. In this
season at least there are grounds for believing that the whales travelled south beyond the
whaling area and again passed through it on their return to the north.

The year 1925-6, as Harmer has pointed out, was clearly an exceptional one: it is
only in this season that a bimodal graph is found in both the species. But there is some
reason for supposing that it is only in exceptional conditions that whale movements can
be indicated by the available data. Though we have little exact information on the
behaviour of whales which visit the South Shetlands, it may be assumed that they do not
diifer appreciably from those at South Georgia. The whales, no doubt, do not arrive in a
regular flow, but appear from time to time, in schools of greater or less concentration in
which one or other sex may predominate. Schools which arrive first will probably
travel farthest and in a normal year some of them, which have passed beyond the
whaling limits, will be replaced by later arrivals from the north. If this view is correct
the indications to be derived from monthly charts of distribution must necessarily be
blurred and indistinct ; and since the whalers pursue their quarry along several hundred
miles of coast-line, statistics will usually give no hint of movement. Only in an ex-
ceptional season, when the greater part of the stock is for a time inaccessible to the
whalers, can we hope to obtain precise information.

In the season 1925-6 the whalers penetrated south-west as far as the Biscoe Archi-
pelago, and here, in December, they obtained a few Blue whales. That they went so far
is evidence that the pack-ice lay well to the south and that whales were not abundant to
the north. During the period under review there is only one other season when the
whalers extended their operations so far to the south-west. This was in season 1923-4,
when both Blue and Fin whales were taken in some numbers at the extreme south-
western limit of the area (Plates XXVI, XXXIV). It is to be imagined that in this season
also the pack-ice lay far to the south and one would expect, therefore, that the monthly
catches would show early and late maxima as in 1925-6. The statistics given by Harmer
indicate that Blue whales reached their peak in December and Fin whales in January
and February, but in neither species did a second maximum occur. The absence of the
second peak is perhaps to be attributed to unusual whale movements at the end of the
season. In 1923-4 the customary autumn concentration in Bransfield Strait did not
occur, and it appears that the whales took a different route when leaving the grounds,
perhaps going due north from Smith Island.

The charts for the individual seasons indicate that the movements of the whale
schools may differ rather considerably. If the grounds are not approached directly from
the north (when whales first appear oflr Smith Island), they are generally taken in the
early months of the season at the northern end of Bransfield Strait and pass through the
strait on their passage south. This is to be seen for Fin whales in 1924-5 and 1925-6
(Plates XXVII, XXVIII), and for Blue whales in 1922-3 and 1925-6 (Plates XXXIII
and XXXVI). Occasionally, however, as in Fin whales, 1926-7 and 1927-8 (Plates
XXIX, XXX), and in Blue whales 1926-7 (Plate XXXVII), they appear to have avoided

DISCUSSION OF RESULTS 187

the Strait, passing outside the South Shetland Islands. Towards the close of the season a
concentration at the southern end of Bransfield Strait is the general rule, the whales
presumably passing up through it on their northward migration. Sometimes, however,
this concentration was absent. As already noted it was not found in 1923-4, when both
Fin and Blue whales appear to have gone directly towards the north, and in the following
season a considerable proportion of Fin whales seem also to have avoided the strait.

The south-western limits of the whaling grounds extend to the Biscoe Archipelago,
but this locality was only reached by the whalers in seasons 1923-4 and 1925-6. It is to
be presumed that in both these seasons the pack-ice in the Bellingshausen Sea lay
unusually far to the south. Season 1929-30 was an exceptionally open one in the Weddell
Sea. The vessels of the Weddell Sea fleet reached very high latitudes and those of the
South Shetland fleet — for the only time during the period under review — were able to
work east and south-east of Joinville Island. This is shown for Fin whales in Plate
XXXII. Blue whale returns for this season are so few that monthly charts cannot use-
fully be given, but the data are included in the totals shown in Plate XL in order that the
distribution of the two species can be compared.

Whales do not appear to occur in such dense concentrations in the South Shetlands as
in South Georgia. During the eight seasons reviewed the greatest concentration — as
shown by the number taken by the whalers — was of Fin whales in Bransfield Strait in
April 1927. In an area of 828 sq. miles 229 whales were killed during this month, or
277 whales per 100 sq. miles. This is less than one-third the maximum density found
on the South Georgia grounds (see p. 180), but the two figures are not strictly com-
parable.

The directional data from the South Shetlands have been summarized in the same
way as those from South Georgia and the results are shown in Figs. 5 and 6. They are,
if anything, less consistent than the South Georgia results, and their validity, as already
explained, is open to question. In Table XI the numbers of components along the four
cardinal points are given, and since a reverse movement is to be anticipated in the latter
half of the season, the numbers for each species are grouped in two series— the first series
running from November to January and the second from February to April or May.

It cannot be said that these data give any clear support to the theory that whales
normally travel south-west in the first part of the season and north-east in the latter
part. Such a resuh is, however, to be found in the Blue whale records for 1924-5 and
1925-6 and with less certainty in the Fin whale records for 1922-3 and 1924-5 (see
Figs. 5, 6). In all other seasons no agreement with this theory is to be found.

Table XI shows that in these data the predominant component throughout the season
is the westerly, the only exception being in Fin whales, where in the latter half of the
season the northerly component is in slight excess. If, however, the first half of the
season is compared with the second it will be seen that in the latter the numbers of
southerly and westerly components have diminished and the northerly and easterly have
increased. Only in this respect do the figures bear out the theoretical movement inferred
from the positional data.

DISCOVERY REPORTS

1922-3

1923-4

1926 -7

1927 - 8

1928 - 9

1929 - 30

1930- I

Fig. 5. Recorded movements of Fin whales on the South Shetland grounds. The direction of each arrow is

the resultant direction of observed whale movements; the length of the arrow indicates, as shown m the

scale, the resultant number moving in this direction, expressed as the percentage of the total.

DISCUSSION OF RESULTS

189

1922 -3

1923-4

1924-5

Fig. 6. Recorded movements of Blue whales on the South Shetland grounds. The direction of each arrow is

the resultant direction of observed whale movements ; the length of the arrow indicates, as shown in the

scale, the resultant number moving in this direction, expressed as the percentage of the total.

I90

DISCOVERY REPORTS

Table XI

Number

of
monthly
records

Components

North

East

South

West

Fin whales

November to January

Number

Percentage
February to April

Number

Percentage

i8

25

7
19

18
56

I
J

8

16

II
31

7
14

17

17
J4

Blue whales

November to January

Number

Percentage
February to May

Number

Percentage

24
17

3
9

2(,\

3

3

9

19

4J:
8

23\

21

46

4-r

If, in the middle of the season, some schools of whales were still moving south, while
others had commenced their return journey, we should expect the resultant values
during this period to be low, for the movements in opposite directions will partially or
wholly cancel one another. This is seen with some clearness in the Blue whale data for
1924-5 and 1925-6 (Fig. 6) and also in some other seasons, but with subsequent
directions which do not accord with our theory of movement.

The data for the first half of the season are more consistent than those for the latter
half. They indicate that, in general, Blue whales were moving south-westerly, whereas in
Fin whales the direction is more westerly, with the southerly component less strongly
marked. Similar but more striking directional differences between the two species were
found in the early part of the season at South Georgia, and in neither locality is it
possible to suggest an explanation.

In the present state of our knowledge little help is to be derived from these data.
While they give no clear support to the theory of movement based on the charts of
distribution, they do not point to the probability of any different theory. They tend
rather to imply that whale movements in the South Shetland area are haphazard and
irregular and that no ordered principle of migration is to be discovered ; and we think
this conclusion most improbable.

PLATES VII-XLII

DISCOVERY REPORTS, VOL. VI

ci:r:i'

II
I,'

/4 6 ,1' \ l(

OCTOBER

38

PLATE VII

^'1 I X

/\ 5y J 10 J i^__j.

'' ,^ / I

1,^ 4 / V^ 6 9 1

^;?_--

NOVEMBER ^'

JANUARY
154

MARCH

101

;4 iv^4 9 3 j: 8 ;

SCALE OF NAUTICAL MILES.
50 100 150

DECEMBER

134

'v 6 g4Xi hi

C ^/' V 6 (14) I V^l-_.1>

FEBRUARY

116

N 7 .1'

APRIL
71

Distribution of Fin whales taken on the South Georgia grounds in season 1923-4

DISCOVERY REPORTS, VOL. VI

PLATE VllI

')'r--z-4\

SEPTEMBER
2.4

OCTOBER
I6Z

f^.

NOVEMBER s".
190

A

JANUARY
378

\ 7 i' / j~-rir=zi>

MAY

71

SCALE OF NAUTICAL MILES.
50 100 150

I ' -^

200

Distribution of Fin whales taken on the South Georgia grounds in season 1924-5

DISCOVERY REPORTS, VOL. VI

{ 10 10 lo:^ (m\ ^^\

PLATE IX

APRIL
850

MARCH
683

SCALE OF NAUTICAL MILES.

lOO ISO

I I

P',

200

Distribution of Fin whales taken on the South Georgia grounds in season 1925-6

DISCOVERY REPORTS, VOL. VI

PLATE X

-'2 ,I0\2:-~U
^? Nl 12) 5)

/"Z "V,

'2^ 6 9 6 I X^
^0^ 3 5 I ^'1,

"-I 6 8 21

OCTOBER

56

/2 T 5 ^1,

V3 6 4 2 ~-r 4"^^^^

)3 2__3 8 9 7 "^$.^

'i: 3/ "■ ■ —

5"^

^,'2'

/ 10

I

\ 8
\ 8

/6 r-3\

4 2 2 4 2 V2D;4 !^

5 4 ^2__2__2_^2 I '^Iv^
3 lS^^ "V 7 4 \

10 6 7--^^\ 8 4 ^1

\^/@) 4 N

NOVEMBER ^

158

DECEMBER

106

> 4 V

/'\

JANUARY I
202 \

' (^ 4.

a/

/2-^ ; 8 \ (D

(7 2 4 6 (Tj) \
^/ 5 3^ 2 7 5 A

(2'^_l2:I)

FEBRUARY
157

r3\

'.I 7 ^^-^'l;--b

I K

MARCH
234

«H^

1 3 '~2_Ilb

\

1'+ *. S /- ~v

;\

i ^^ • V

/A

,'5

-% (g) 2' 10 2(
7 l^^'"'^3.^^^

\

/

N-.

-1'"

J/2 349 8 ^-^

,K 6^-^^^^ J 8 9 )

I

N-,_2.

APRIL
108

SCALE OF NAUTICAL MILES.
50 100 150

' I — *—

200

Distribution of Fin whales taken on the South Georgia grounds in season 1926-7

DISCOVERY REPORTS, VOL. VI

X 6 1\

PLATE XI

\, 3 7 3 l',^

OCTOBER

123

'k%

s^.

DECEMBER

287

^-''

NOVEMBER

;?38

' i'

ci:iii>-

y

C3\_^

f\ — f.^^ )2 4 "( 4 /

'^8 5 (I4)_3/'

JANUARY
279

SCALE OF NAUTICAL MILES.
50 lOO 150
I I 1—

200

FEBRUARY

219

MARCH

89

V2 Z^^

Distribution of Fin whales talven on the South Georgia grounds in season 1927-8

DISCOVERY REPORTS, VOL. VI

PLATE XII

'~1 1-"3-'"~2-

/4 ^l2'^IO^Sy^-Zk\4N

MARCH
33X

Distribution of Fin wliales taken on the South Georgia grounds in season 1928-9

DISCOVERY REPORTS, VOL. VI

/'6'\ .,

PLATE XIII

I?)-

,;-r::;/'' "^^^^ ^--c/ , . , ,

SCALE OF NAUTICAL MILES.
50 100 150

Distribution of Fin whales taken on the South Georgia grounds in season 1929-30

DISCOVERY REPORTS, VOL. VI

PLATE XIV

/J~ J 6 J 6
'V Cil.^— k^ 4 5

Distribution of Fin whales taken on the South Georgia grounds in season 1930-1

DISCOVERY REPORTS, VOL. VI

PLATE XV

/ '"v

/I y

r"l 4 ^L__3; { 5,

""■'"'^ '" ' OCTOBER

/X— i— 1-

/S 3,_2 6( 18 II (

/j 5/ XyVi^yi 2\

I' z

.—I — 2

/ 1 ■3 ii-^cisys il

^ 7 7
'\2 4 8

\

JANUARY

vr 6 3)
d— r''3 3 J'

MARCH
39

®

t>

(Z

^l 5 ® '1^') (i'

FEBRUARY

176

~ /

APRIL

48

SCALE OF NAUTICAL MILES.

100

"-L_:i.

200

Distribution of Blue whales taken on the South Georgia grounds in season 1923-4

niRCOVERY REPORTS, VOL. VI

PLATE XVI

SEPTEMBER

17

NOVEMBER ^-v^i "~~^ ,

590

DECEMBER
730

Ct)

JANUARY
420

-^^2^9 '^5_ 7n

710 7 I

FEBfiUARY ~\ U I

377 \^^^.2_,J'

MARCH

APRIL

309

SCALE OF NAUTICAL MILES.
0 100 150

200

Distribution of Blue whales taken on the South Georgia grounds in season 1924-5

DISCOVERY REPORTS, VOL. VI

PLATE XVII

SEPTEMBER 'i-'
31

■I-

\

I

•

(4

\

I
3'

OCTOBER

139

/I' 6 5 13 4^~l — ^1)
\l 3 Zi
\ 8 @ ^1

NOVEMBER
75

""3---' 4 )

DECEMBER
89

,d"---J'

JANUARY
188

MARCH '7)
397

^.

FEBRUARY ^»
406

APRIL

230

V. 3 /

0.. /"4

' 5 Z\

MAY
74

n

SCALE OF NAUTICAL MILES.
SO 100 150

100

Distribution of Blue whales taken on the South Georgia grounds in season 1925-6

DISCOVERY REPORTS, VOL. VI

PLATE XVIII

'\'^^A _9 6

26^18 ? 'V

ji-"J 8-8 ^\

OCTOBER

;260

_,^i— K

NOVEMBER

515

DECEMBER

813

JANUARY

841

FEBRUARY \

492

MARCH
344

APRIL f^'

48

SCALE OF NAUTICAL MILES
50 lOO

ISO

200

Distribution of Blue whales taken on the South Georgia grounds in season 1926-7

)ISCOVERY REPORTS, VOL. VI

PLATE XIX

6

4N.

/:5 6^^20-Z4-^N^2 ~~\^

OCTOBER

312

5 T 8
:i— I — V

NOVEMBER
66%

— 2._ 3 8

10^28^26) 13) 3 ~2x

^ 16 J^ )
% /3/ ■ C?)
9 2'

JANUARY

27;?

''Vi'^ FEBRUARY
I 68

SCALE OF NAUTICAL MILES.
50 100 ISO

MARCH
17

(A I'

V I /

Distribution of Blue whales taken on the South Georgia grounds in season 1927-8

DISCOVERY REPORTS, VOL. VI

PLATE XX

OCTOBER
52

n

, Z \\z\

NOVEMBER ^^'J^'"' ^^^^'

ra

,'3 2 4 2)

Ov5 6\I6 19 2-^1'

DECEMBER

'2-'3;

FEBRUARY
60

'2^ 7 ClO^

MARCH

--^3__3,,-l'

JANUARY
413

SCALE OF NAUTICAL MILES.
50 100 ISO

APRIL

15

(T;

Distribution of Blue whales taken on the South Georgia grounds in season 1928-9

DISCOVERY REPORTS, VOL. VI

PLATE XXI

y''% ~^K

C2"J"-~J--

OCTOBER
31

NOVEMBER
91

Z

2. ,i, 10

— \-"'■^^

4 !

I

/
I.'

^1 I,'

V-\-::iy

DECEMBER
59

JANUARY
9Z

(I -"3

SCALE OF NAUTICAL MILES.
50 lOO '50

I I I

200

Distribution of Blue whales taken on the South Georgia grounds in season 1929-30

DISCOVERY REPORTS, VOL. VI

v1>' \| 5 7 4 C-|-~^^^
'V' ''j2 (15^9 6 8 )

\ 7 6 5 _3\
e:- — — -■"" ^N X

PLATE XXII

®

OCTOBER
188

(!)

CI)

(D

H \y

JANUARY
117

I" 2"»
/I ^'x

' ^^ \._

y^ I 2:-i'

/4 4 dD '^-~

3 3 5 4 8 6 )

(n....:..f|---K

''I''

J-" 5 ^. r^\

C?::C ^ 3

DECEMBER
236

0 -7/

U ~l— ^-'3/
N^-3 6 3;

^'2v

\ 6 ( /5 5
M Z 9 2[

'IV

,1-H

^\

FEBRUARY
50

% 1 3l1>

SCALE OF NAUTICAL MILES
50 100 ISO

100

Distribution of Blue whales taken on the South Georgia grounds in season 1930-1

DISCOVERY REPORTS, VOL. VI

PLATE XXIII

^^-s

2 X 6 14 9

4\
7 2"

it:

7 \
I

<^\ ] 14 26 45 37 59 36 18 17 4 i^

" "3 J2 40 87^I05T3I\8I__4X 70 6 ^1 — 1-^

)\

89/113 145 149 133 I93\^

IJ7 I45/243~2C?99^\I74 106 25 6 {

i 10 45 73)I4I(

,29^::^5-5l6^444 les 198 I20\ ?.7 7 h-^^
^785 1045 688fe) 258^5 l65 1^ Xl 7 ^^,

i 7 27 82\I97 I66J7
<|^ % 11 Z% 35 80 1^7 34
\| 6 12 12 30 36 &Z

JV'\

6 12

209^38^.604^ I94"W4I W 7 4 /
^^64^217 125/78 36 13 23 J,i
■^115 109 6R 29 10 \C
1/178 86 85 72 3^^6^
13 20 57 113 II8_88 So 10 7 /

.i:>'-

15 56 37 50 33 5 4 2

^ 8 8 24 30 6 4 v^

\|/

FIN

6 26 60 13 7 SJ'J
12 9 12 2,-"''
'^' -J5 2'

\^

CD

I

2
2

\ "1-

2
5

5 ^r~2^^^^ yZ'

1 3 11 8'""3~'3 I
8 21 19 15 12 18 I

2\

;3 13 38 78 74 42 60 18 5
/s' 13 46/123. 2l4j:2l6z249\63 42 19

4 2_:i

■\2 25 38 /i6l/^87/'^8?6i8~~555V58 ^4 49 26

x\:

/a 45 Sd) I38(4£9J533 991 684W5$2z7l5 42 18 6^
•^ _ ... . J *^^^^79^296) 12l| 41 13

,l'^^_.2^ 7 18 AT. 17 8

^'

V

96\23j/ 111/96 14
\3 3 6 13 33 2lV^68 173/80 56 26 7^/
/2/" \ 7 .^ 16 64 tlii^SS :f4 15 5 , '

18 12 I
3 3x^

/

/ 7

/
/

U 5 10 18 3

12 20 44 28

8 6

2 4/"

I
6 I

V

^2^_3 6 25 12 3/

BLUE

{2

12
3

I

2'

2/

SCALE OF NAUTICAL MILES.
50 'OO 150

Distribution of Fin and Blue whales on the South Georgia grounds : based on all recorded positions of capture

during the eight seasons 1923-4 to 1930-1

DISCOVERY REPORTS, VOL. VI

PLATE XXIV

o

BLUE

SCALE OF NAUTICAL MILES.
50 100 150

200

t

Charts showing the recent extension of the South Georgia whaling grounds

limits of whaling area during seasons 1923-4 to 1926-7

„ „ „ „ .. 1927-8 to 1930-1

main concentration (300 contour) during seasons 1923-4 to 1926-7

„ „ (150 contour) „ „ 1927-8 to 1930-1

DISCOVERY REPORTS, VOL. VI

PLATE XXV

\4 5\

JANUARY

2:52.

DECEMBER

61

FEBRUARY

237

APRIL

164

SCALE OF NAUTICAL MILES.

MARCH

378

100 200

Distribution of Fin whales taken on the South Shetland grounds in season 1922-3

^

300

DISCOVERY REPORTS, VOL. VI

^.\

PLATE XXVI

as 4 /

lU

-r-^

DECEMBER

126

JANUARY
298

/4:X ^:i r^ijO V

'Mmi 6 .1.^ 6

300

Distribution of Fin whales taken on the South Shetland grounds in season 1923-4

DISCOVERY REPORTS, VOL. VI

PLATE XXVII

• 5 ^k.

♦r-f /

DECEMBER
181

JANUARY

'2 8

/4 |-"'l

10

Z 4

-r^

FEBRUARY

227

A — 'V.

f \-_/ , e,i|^,

jK^-^Cj

APRIL

314

SCALE OF NAUTICAL MILES.

MARCH

365

^

100 200

Distribution of Fin whales taken on the South Shetland grounds in season 1924-5

300

DISCOVERY REPORTS, VOL. VI

/
/

I I
• 1

PLATE XXVIII

4 ! i U25 \I5V!

l^'Q

2
J 3

/ 4 \,

^^ _^y

4 ^ <e>,©2

•I— -'

,-' 2~-s

DECEMBER

FEBRUARY

3 3^

APRIL

172

JANUARY

^r

MARCH

500

SCALE OF NAUTICAL MILES.

100

200

30O

Distribution of Fin whales taken on the South Shetland grounds in season 1925-6

DISCOVERY REPORTS, VOL. VI

'2'.

•4

PLATE XXIX

""I.

I
I

2 U6 19 \ ^ ^v^ >^K

**-'* \

.- -• I

I

8 2/
4 /''

NOVEMBER
SO

/?5X-^

DECEMBER

260

APRIl

MARCH
556

SCALE OF NAUTICAL MILES.

^

100 200 300

Distribution of Fin wliales taken on the South Shetland grounds in season 1926-7

DISCOVERY REPORTS, VOL. VI

PLATE XXX

. ,.j.

£>4^"""-'2;

DECEMBER

22

,'2.'

10 12) L_...|)

t., ^^ .^.

JANUARY

106

-i-l^ ■<

FEBRUARY

196

MARCH

428

SCALE OF NAUTICAL MILES.

AP'^IL 100 200

469

Distribution of Fin whales taken on the South Shetland grounds in season 1927-8

300

DISCOVERY REPORTS, VOL. VI

3

10

s>

c^

c? 2

\2! "^ C>

^

f^i^...

DECEMBER

PLATE XXXI

,T; ; <^ -i^

7 I

JANUARY

^9

~-£'1^"V' 6 ')

\.2 *

10 4/

FEBRUARY
250

^#

I .3/

b

/ 9 '^V^^^i^

APRIL

444

SCALE OF NAUTICAL MILES.

lOO 200

Distribution of Fin whales taken on the South Shetland grounds in season 1928-9

300

DISCOVERY REPORTS, VOL. VI

PLATE XXXII

&]^ 0^

o'

c^

DECEMBER

51

/' >...I?V

JANUARY
85

/ Ar\9

/' Ai^ \ ^ ',

,-'--N ( ® \ i

18

1 J /

FEBRUARY
355

100 200

Distribution of Fin whales taken on the South Shetland grounds in season 1929-30

300

DISCOVERY REPORTS, VOL. VI

NOVEMBER

82.

JANUARY

J40

^ 5 K^\i> ^'^ J^^'J)

MARCH
108

SCALE OF NAUTICAL MILES.

100

200

PLATE XXXIII

DECEMBER
342

FEBRUARY

112

^

O

dr"

300

Distribution of Blue whales on the South Shetland grounds in season 1922-3

DISCOVERY REPORTS, VOL. VI

PLATE XXXIV

Distribution of Blue whales on the South Shetland grounds in season 1923-4

DISCOVERY REPORTS, VOL. VI

,„o^|-.' -J'

PLATE XXXV

OCTOBER

zo

8 3»

^. ^ -'

NOVEMBER

81

,1-— I

--K ^ ^

t3 (i:_-:j>

' x<uf^ 4 5 4 4 1

\A

DECEMBER

229

100 200

Distribution of Blue whales on the South Shetland grounds in season 1924-5

300

DISCOVERY REPORTS, VOL. VI

PLATE XXXVI

l-r^-CS^^ 7 N

■^4^-.

JANUARY

DECEMBER

588

MARCH
95

\ 9
\

4 /
/

FEBRUARY
85

SCALE OF NAUTICAL MILES.

APRIL

490

lOO 200 30O

Distribution of Blue whales on the South Shetland grounds in season 1925-6

DISCOVERY REPORTS, VOL. VI

PLATE XXXVII

/ 5

NOVEMBER

114

^ '0-7

0K^

JANUARY
67

^^

2 2

MARCH

I /lO, 3 \

7

y

/ 5 6 1

/ /

/ /

(iD 4 \

/ 6 7 .'-r>^"7

/ 6

\ 5 3^1.

1\
\
I
I

DECEMBER

/3)

4 4/

I 4 5 / Vv^ ^ ''

1r^^<^

FEBRUARY

72

0

/

SCALE OF NAUTICAL MILES.

100 200

Distribution of Blue whales on the South Shetland grounds in season 1926-7

300

DISCOVERY REPORTS, VOL. VI

PLATE XXXVIII

NOVEMBER

113

(W

(f

/ , "lu£^,

--'--'--£>

'O/

0

0 ©

-r-«-'''^] G,_A

JANUARY
62.

MARCH

SCALE OF NAUTICAL MILES.

DECEMBER

189

2 ^-—'^

,'2 jv

'3 "'' ^

'^

FEBRUARY

46

^

lOO 200 300

Distribution of Blue whales on the South Shetland grounds in season 1927-8

DISCOVERY REPORTS, VOL. VI

PLATE XXXIX

/ 7_

/-#

o

<0^

^S^-'>

kj\^-^ (l^

DECEMBER
53

JANUARY

4 I

FEBRUARY

45

APRIL

95

SCALE OF NAUTICAL MILES.

100

200

l^

\1

'U

4 .c,<>

®

• '^

'^ '^^©^

5 5

300

MAY

15

Distribution of Blue whales on the South Shetland grounds in season 1928-9

DISCOVERY REPORTS, VOL. VI

PLATE XL

4 \

\

\

5 1

5 .-'

FIN

. 2, A "^'

/ \

I \

1 8 18 17 14 4 \

3 28 31 I5_ /|_49 113X89 38 |j /n , 1

29 53 .^^ 198^3^7)75 ^ ,g ^^ /

BLUE

30O

Distribution of Fin and Blue whales on the South Shetland grounds : based on all recorded positions of capture

during the eight seasons 1922-3 to 1929-30

DISCOVERY REPORTS, VOL. VI

PLATE XLI

•4 l'^ 5 ^|.^

5 5 4 r— |-Cf?"Q
47'^, V_.J^3«sA 10 /

MARCH

SCALE OF NAUTICAL MILES.

APRI

Monthly distribution of Fin whales on the South Shetland grounds : based on monthly totals for

the eight seasons 1922-3 to 1929-30

DISCOVERY REPORTS, VOL. VI

PLATE XLII

^^5,

NOVEMBER

.'21

V / - i.

/ 9

JANUARY

^

o

10 3

11 7

SCALE OF NAUTICAL MILES.

30O

■Oi-

— 1

'' 4

2)

9

5

^

"5-;.

9

II

Id

/'

/

10

IJ

y.^

DECEMBER

r.-c^c

,'2 4'

I 1,

FEBRUARY

APRIL

Monthly distribution of Blue whales on the South Shetland grounds: based on monthly totals for

the seven seasons 1922-3 to 1928-9

[Discovery Reports. Vol. VI, pp. 191-204, Plates XLIII, XLIV, December, 1932]

ON THE DEVELOPMENT OF
CEPHALODISCUS

By
C. C. JOHN, M.A., D.Sc, D.I.C.

CONTENTS

Introduction page ^91,

General considerations ^94

Development

Segmentation ^9"

Larval stages ^9°

Development of the bud 202

Discussion : Larval and bud development 203

List of literature 204

Plates XLIII, XLIV following page 204

ON THE DEVELOPMENT OF
CEPHALODISCUS

By C. C. John, M.A., d.sc, d.i.c.
(Plates XLIII, XLIV)

INTRODUCTION

THE observations which have hitherto been made on the development of Cephalo-
discus are mostly of a fragmentary nature owing to the difficulty of obtaining the
eggs and larvae. From the material procured by the Challenger expedition, Masterman
(1900) recorded a few segmenting eggs. Andersson (1903), in a preliminary note, showed
that the segmented embryos left the parent colony in the form of ciliated planulae.
Harmer (1905) observed that the eggs were heavily laden with yolk, that their segmenta-
tion was holoblastic and nearly equal, and suggested that they might give rise to solid
embryos in which the endoderm arises by delamination. Further, he showed that the
five body cavities of the adult arose at an early stage in development. Andersson (1907),
in his account of Cephalodisciis collected by the Swedish South Polar Expedition, de-
scribed a few more stages. He stated that there was a gastrula-like stage, in which there is
a centrally placed mass of yolk with a narrow lumen. He believed that this central mass
of yolk represented the endoderm, which originated by a process of invagination. He also
confirmed the observations of Harmer regarding the early appearance of the body
cavities. SchepotiefT (1909) also described a few stages, but his account added nothing
to the elucidation of the gastrula stage described by Andersson. He confirmed the exist-
ence of the planula-like larva, and described a later stage which was also free-swimming
and in which the five body cavities were visible. He regarded the central mass of yolk
in the embryo and the larva as representing the endoderm formed by invagination. The
account of Braem (191 1) does not add anything further to our knowledge of the develop-
ment of the group, but he drew attention to the resemblance between the free-swimming
larva of Cephalodisciis and that of certain Polyzoa.

The first connected account of the development was given by Gilchrist (1917). He
collected the eggs and larvae of C. gilchristi from the coast of South Africa and at-
tempted to rear them artificially. During the first year of his work he was able to obtain
all the early stages from the beginning of segmentation up to the free-swimming larval
stage ; but during the second year, when he repeated the trawling operations, he was not
able to get even the adult colonies, so that further observations were rendered im-
possible. He described the segmentation, the origin of yolk, endoderm, the coelomic
cavities, the pigment granules and the structure of the free-swimming larvae. Though
some of his observations are open to doubt, his account of the early segmentation is a
very valuable help for all further research.

194 DISCOVERY REPORTS

The material for the present paper was obtained in an unexpected manner from the
collections of the Terra Nova Expedition, which are now preserved in the Natural
History Museum. When I was engaged in reporting on the Cephalodiscus of the Dis-
covery Expedition I was kindly permitted by Miss Hastings to examine the Terra Nova
collections. On one occasion I was prompted to collect some of the sediment which had
settled at the bottom of jars containing colonies of C. tiigresce?is. I examined this in
October 1932 and discovered twenty-one larvae in different stages of development, and
two embryos. These I showed to Prof. E. W. MacBride, F.R.S., and at his kind
suggestion this work was undertaken.

As there were only a very limited number of specimens, whole mounts were not made,
and though every precaution was taken in preparing the sections the results were not
quite satisfactory, owing to the fact that the material was very old and imperfectly
preserved. The description of the genital organs and buds is based on Discovery material
of C. nigrescem.

I take this opportunity to record my deepest gratitude to Prof. E. W. MacBride,
F.R.S., for his kind help and encouragement and the facilities he gave me for carrying
out this work in the Huxley Laboratory of the Imperial College of Science and
Technology. I am also indebted to Miss Hastings, who is in charge of the Prochordate
section in the Natural History Museum, for kindly permitting me to examine the Terra
Nova collections.

GENERAL CONSIDERATIONS

The zooids of Cephalodiscus are either unisexual or hermaphrodite. The colonies of
C. dodecolophus, C. levmseni and C. gracilis have been found to contain only female
zooids in the specimens so far obtained, while the colonies of C. nigrescens, C. densus,
and C. kempi are formed of male, female and hermaphrodite zooids. In these species
the three different kinds of zooids are not differentiated by external characters, so that
it is impossible to identify them except by sectioning the adult zooids. On the other
hand the two species C. hodgsoni and C. sibogae show striking instances of dimorphism.
In C. hodgsoni the female zooids are crimson-brown and possess twelve arms, while the
male zooids are pale brown and possess ten or eleven arms. Some colonies are formed of
both male and female zooids, while others are entirely of one sex. In C. sibogae the
ordinary zooids are usually completely sterile or neuter individuals, in which the repro-
ductive organs are vestigial. The male zooids are entirely different in appearance and are
found in the basal jelly of the colony. They possess only a single pair of arms without
pinnules. The alimentary canal is vestigial and there is no trace of a glandular stomach.
It is believed that these male zooids are nourished through the vascular system by the
neuter individuals.

In the male zooids of C. hodgsoni, C. nigrescens and C. densus the testes are a pair of
organs symmetrically placed, one on either side of the median line, in the anterior part
of the trunk cavities between the pharynx and the anus. Each testis is a small ovoid body
with a narrow lumen in the centre. The testis therefore appears like a thick- walled
hollow sac, the lining epithelium of which gives rise to the spermatozoa. The lumen of

DEVELOPMENT OF CEPHALODISCUS i95

the testis is continued into a short narrow tube, the vas deferens, which opens externally.
I was not able to discover any trace of pigmentation in the vas deferens or in the ecto-
derm surrounding its external opening. In C. sibogae the two testes are large and fill the
greater part of the trunk cavities. Their structure is not clear from the account of
Harmer (1905), probably because the zooids were imperfectly preserved, but it would
appear from the description of his figures that the lining epithelium is comparatively
thin, and that it surrounds a large central cavity which is filled with ripe spermatozoa.

In the female zooids the ovaries are conspicuous organs and in a mature specimen of
C. fumosus they appear like swellings on either side of the anterior part of the trunk
(John, 1 93 1, pi. XXXV, fig. 3). Each ovary is an ovoid or flask-shaped body which is filled
with a varying number of immature eggs. In an adult individual one or two eggs, at the
end furthest from the external opening, are large and well developed and fill the greater
part of the ovary. In C. gilchristi each of these large eggs is lodged in a follicle of small
cells, while Schepotieff (1909) has described the large ova of certain species as lying be-
tween a central lumen, clothed with epithelium, and the wall of the oviduct, in a position
where they are surrounded by a large quantity of blood. The indication of a central
lumen in the ovary is also suggested by Harmer (1905), but no trace of any lumen has
been observed in any of the species collected by the Discovery Expedition.

The oviduct is a thick-walled pigmented tubular prolongation of the anterior end of
the ovary, and its external opening corresponds in position to the external opening of the
male duct. In some species, such as C. levinseni and C. dodecalophiis, the external
opening of the oviduct is situated at the dorsal end of an ectodermal recess.

Gilchrist suggests that the ova originate from the wall of the oviduct, but this view is
not supported by any convincing evidence. Very little is known about the origin of the
oviduct and its homology. From a comparative study of the developing buds and young
zooids it would appear that it is an ectodermal invagination, to the proximal end of
which is attached the primitive germ cell, surrounded by a thin mesodermal epithelium.
The oviduct is deeply pigmented like the red line on the ventral wall of the proboscis.
In addition to this red pigment the ectoderm of the trunk carries a black or brown pig-
ment in C. nigrescens and C. fumosus. In the developing buds and larvae, which will be
described later, pigment granules are confined only to the ectoderm. If, therefore, the
occurrence of the red pigment in the wall of the oviduct is taken as evidence for deciding
its homology, there can be no doubt of its ectodermal origin. The proximal end of the
pigmented tube is continued into a very short unpigmented part, which in turn opens
into the ovary. This latter part is more clearly seen in the figures given by Harmer
(1905, pi. xii, figs. 153, 154) than in my own sections. The continuation of this tube with
the lining epithelium of the ovary suggests that it is mesodermal in origin. The oviduct
is therefore really formed of two parts : a distal pigmented tube, which arises as an ecto-
dermal invagination, and a short proximal part which is probably mesodermal in origin.
According to this view the ectodermal recess in C. levinseni and C. dodecalophiis is the
anterior part of the invagination, which serves as a receptacle into which the egg is ex-
truded before it is finally liberated from the body of the parent.

ic,6 DISCOVERY REPORTS

It has already been stated that in the immature condition each ovary is filled with five
or six eggs, which are almost of the same size, but of these only one or two eggs reach
maturity, the others being probably used up as food by the developing eggs. Each egg is
composed of a large cytoplasmic part, which stains very lightly in haematoxylin, and a
conspicuous nucleus, which is situated basally. The nucleus is of a clear homogeneous
appearance and is spherical or oval in shape. As the present material was not fixed for
cytological purposes it is not possible to give details, but Gilchrist (191 7) has observed
that there are faint indications of chromatin elements and that a conspicuous deeply
staining nucleolus is always present. The protoplasmic part is filled with oval or rounded
bodies which ultimately give rise to yolk granules.

The extrusion of the egg and the mode of sperm transference in Cephalodiscus are not
known. Masterman suggested that the eggs are liberated at the death of the mother
zooids ; but in C. densus and C. fiimosiis the developing eggs have been found at the
bottom of some of the tubes, containing in their upper parts zooids which were pro-
bably not dead till the time of fixation.

DEVELOPMENT

SEGMENTATION
The early development of Cephalodiscus was first successfully worked out by Gilchrist
(1917), who was able to study the early stages with artificially reared eggs of C. gilchristi.
Segmentation is total and equal up to the four-celled stage ; but the succeeding divisions
are unequal, probably owing to the great amount of yolk present in the egg. The four-
celled stage gives rise to a six-celled embryo, which now assumes an oval shape and
shows the commencement of a blastocoeUc cavity. A perfect blastula is formed in the
nine-celled stage. The blastocoel is now a wide space filled with a clear homogeneous
substance. As the result of further division the blastocoel becomes wider, and at a stage
when the embryo consists of about fourteen cells in section, one of the cells at the
posterior end proliferates and the daughter cell thus formed sinks into the blastocoelic
cavity. During the subsequent stages the cells become more and more diflferentiated, so
that the anterior and posterior ends can be distinguished. The blastomeres, which were
till now spherical, become elongated in the anterior end, and this transformation slowly
extends towards the posterior end until all the cells assume the same shape. While these
changes are taking place in the structure of the embryo, a group of three cells situated at
the extreme posterior end continues to proliferate daughter cells into the blastocoel,
until the latter becomes completely filled with a solid mass of cells. The elongated cells of
the external layer are called columnar cells. These now begin to break up so that cell out-
lines are completely obliterated. The breaking up of the cell outlines of the external layer
of cells has been attributed to the abundance of yolk. The embryo now measures about
0-8 mm. in length and a longitudinal section of it is represented in Plate XLIII, fig. i.
The external layer is formed of deeply staining protoplasm, which carries numerous rod-
shaped bodies, the yolk columns, which are represented by the dark thick lines in the
figure. These are formed by the close juxtaposition of each of the multiplying nuclei with

DEVELOPMENT OF CEPHALODISCU S 197

a line of yolk granules after the cell outlines are broken up. The protoplasm also carries
numerous small pigmented bodies, which are probably excretor}' matter accumulated in
the tissue. They first appear as small black or brownish bodies, which increase in size
and fuse together into elongated large masses. It is probable that in C. nigrescens and
C. fumosus, in which the body is coloured brown or black, this pigment remains in the
tissue permanently, while in others it is finally extruded from the surface.

One of the sides of the embr^'o, probably the ventral, is comparatively thick and is
formed of elongated cells, which closely resemble those forming the thickened wall of
the proboscis. This has been termed the ventral thickening, and Harmer (1905) believed
that it somehow becomes converted into the proboscis of the adult. This comparison
is strongly emphasized in my own figures of the embryo and the later larval stages,
though I have not been able to show how the transformation takes place.

The inner part of the embryo, which represents a wide blastocoel, is filled with a solid
mass of yolk granules. Gilchrist (1917) maintains that below the external layer a few
cells appear, which become closely applied to the yolk granules. As the yolk granules in
this layer are used up, each of the cells sends out a long protoplasmic process towards the
adjacent ones on each side, so that ultimately they form a chain of cells which gives rise
to the endoderm. If any such cells are present on the outer surface of the yolk mass, they
must be clearly distinguishable from this mass ; but his own drawings do not throw much
light on this point, nor does he seem to be certain of their exact place of origin. According
to this view the endoderm originates by a process of unipolar proliferation of the cells
into the blastocoel. The cellular structure of these cells breaks down, and while some of
the nuclei with the associated protoplasm go to form the endoderm, others pass towards
the centre and become vitellophags. The same opinion was maintained by Harmer
(1905), but Andersson (1907) questioned the accuracy of this statement and said that at
least in his species the endoderm originates by typical invagination. Unfortunately he
was not able to carry out any study of the cellular structure. Owing to the imperfect
preservation of the material the ectoderm and endoderm appeared uniformly filled with
yolk granules, which were, however, absent from the immediate neighbourhood of the
central lumen. Schepotieff (1909), who found gastrula stages in C. indictis, also main-
tains that the endoderm is formed by invagination.

Between these divided opinions it is difficult to say which represents the correct mode
of origin of the endoderm. I have not been able to obtain any of the intermediate stages
between the solid blastula and the early larval stage which shows the origin of the
coelom ; but from a consideration of the structure of the latter I am more inclined to
agree with Andersson.

Judging from the quantity of yolk present inside the blastula one would infer that a
considerable time elapses before the commencement of further changes. During this
interval cilia develop on the outer surface of the body and the embryo leaves the parent
colony as a planula-like larva, which has been observed by Andersson (1903). During
this free-swimming planula stage the inner mass of yolk is slowly absorbed, leaving a
homogeneous detritus in which irregular cavities appear. The excretory matter accumu-

198 DISCOVERY REPORTS

lates in the external layer and a few refractive beads also appear in the dorsal wall, i.e.
the side opposite the ventral thickening.

The progressive diminution in the quantity of yolk causes the outer surface of the yolk
mass to shrink from the external layer, thus leaving a considerable space between the
two layers. At this stage (Harmer, 1905, pi. xiv, figs. 198, 199, 208), the yolk mass is at-
tached to the outer layer only by a short stalk-like process at one end, which is probably
the posterior ; so that absorption takes place only through the cells in this region. The
real coelomic cavities originate only at a later stage in development: the five body
cavities of the embryo described by Gilchrist (19 17) seem only to be parts of this space,
formed between the yolk mass and the external layer, which have been mistaken for
true coelomic pouches.

LARVAL STAGES

When the greater part of the yolk is absorbed, the blastocoelic cavity becomes an
empty space traversed only by a few protoplasmic strands, and the larva shrinks con-
siderably in size. An invagination now commences in the region where the stalk of the
yolk mass was originally attached. By this process a typical gastrula is formed. The
gastrula stage has been recorded by Andersson ; but its subsequent development, the
formation of the mouth and the transformation of the ventral thickening into the pro-
boscis region, is at present unknown. The youngest larvae which I have obtained re-
present a rather advanced stage in which the mouth is already formed and the proboscis
region is well marked (Plate XLIII, fig. 2). The outer wall of the proboscis is very thick
and is formed of large glandular cells, which stain very deeply in haematoxylin and in
which masses of brown pigment are very conspicuous. The collar region on one side is
very prominent and the ectoderm of the trunk and collar is formed of large columnar
cells, which carry numerous refractive beads. These are homogeneous transparent bodies
which are embedded in the cells of the ectoderm. In the adult zooid of C. hodgsoni and
C. dodecalophiis they are found at the tip of the arms, while in C. densus they are found in
the dorsal wall of the proboscis and in the dorsal wall of the arms. Both in the adult
zooids and in the larvae they are sometimes found partially extruded from the surface of
the ectoderm. Throughout the larval stages and in the young zooids which develop from
the larvae the refractive beads are found in large numbers in the ectoderm of the trunk ;
but when the young zooid settles down and begins to construct the coenoecium, they
disappear in the body wall and become restricted to the tips of the arms. It is therefore
possible that these refractive beads are the main defensive organs during the free-
swimming larval stages of development. The fact that they are not found on the trunk
or arms of the young buds constitutes one of the chief differences between the habits and
structure of the two.

The gut opens externally through a narrow mouth situated antero-ventrally, and
its walls are formed of cells which are somewhat shorter than those of the ecto-
derm. The mouth is roofed over by the inner wall of the proboscis, and its floor is
formed by a short swelling of the anterior end of the trunk, which represents the
future post-oral lamella. The mouth leads into a wide cavity, which becomes slightly

DEVELOPMENT OF CEPHALODISCUS 199

narrow as it approaches the posterior end, and ends blindly, the anus not being yet
formed.

The right and left coelomic pouches are seen on each side between the archenteron
and the ectoderm. The right coelom has completely separated itself into three parts. The
proboscis coelom (Plate XLIII, fig. 2, pc) appears like a soHd rod of cells, while in the
collar and trunk regions the beginnings of the respective cavities are visible. On the left
side the coelomic pouch is still connected with the wall of the archenteron and appears like
a single long pouch, the two walls of which are in close contact (Plate XLIII, fig. 2, cp).
At its anterior end it is slightly thicker, and in this region it is connected with the wall of
the archenteron. This condition of the left coelomic pouch throws considerable light on its
possible mode of origin. Though the cells lining the walls of the coelomic pouch appear
to be slightly diff"erent from those of the archenteron, in the region where the two are
connected they are almost similar. This single pouch severs its connection from the wall
of the archenteron, and later the anterior third of its length is cut off by a constriction,
thus giving rise to the collar and trunk cavities. This condition is seen on the right side,
which has developed more rapidly than the left.

It is difficult to say definitely from which side the proboscis coelom originates ; but
from the fact that the right side represents a more advanced stage in the development of
the coelom, it is probable that the mass of mesodermal cells in the proboscis is con-
stricted off from the anterior end of the right coelomic pouch.

The proboscis is situated at the anterior end, and its outer wall is thick. The larva now
elongates slightly and the anus is formed at the posterior end. Plate XLIII, fig. 3,
represents a larva at this stage which has been cut slightly obliquely, so that the proboscis
coelom and the mouth are not quite visible. It has now an almost worm-like appearance,
with a prominent swelling on the dorsal side which marks the commencement of the first
pair of arms (Plate XLIII, fig. 3, ra). The proboscis still occupies the anterior end, and
part of the proboscis coelom (pc) is seen towards one side. The masses of pigment which
were till now uniformly distributed in the wall of the proboscis are now accumulating
towards the outer surface, from which they will be extruded in due time. The mouth is
situated ventrally and the alimentary canal is an almost straight tube with the anus at
the posterior end. Near the posterior end there is a small bend in the aUmentary canal,
which marks the commencement of the U -shaped structure of the adult.

The body cavities are not very distinct, but part of the trunk coelom is seen on the
dorsal side of the alimentary canal. The dorsal part of the collar cavity extends as a wide
space into the rudiment of the arms, and the right and left halves extend round the
alimentary canal and meet on the ventral side in the small swelling, behind the mouth,
which represents the post-oral lamella. At this stage the larva resembles Balanoglossus.
It is worm-like and elongated in the antero-posterior axis. The proboscis is at the anterior
end and the anus at the posterior. The collar cavities surround the pharyngeal region
like a ring and the mouth opens ventrally; but this larva differs from the larva of
Balanoglossus in that the gill slits are not present and that the rudiments of the first
pair of arms have already appeared on the dorsal side of the collar.

200 DISCOVERY REPORTS

The larva has now reached its maximum development, and in the next stage it has all
the characters of a young Cephalodisais; the alimentary canal is U-shaped and the pro-
boscis shifts towards the ventral aspect. This transformation, since it is accompanied by
so many changes, can be regarded as a metamorphosis.

Plate XLIII, fig. 4, is a median longitudinal section of a young Cephalodisais which
shows almost all the characters of the adult. Though the alimentary canal is U-shaped, it
is a uniform tube, which is not differentiated into pharynx, stomach and intestine. The
postero-dorsal half of the tube is slightly wider and a lumen is clearly seen, but the
entire wall of the alimentary canal is still formed of cells which are similar in shape and
structure. Behind the mouth two diverticula are developed from the dorsal wall of the
pharynx. The first diverticulum, which is slightly longer than the other, tapers slightly,
and is directed forwards between the right and left collar cavities into the base of the
proboscis cavity. This gives rise to the notochord, though at this stage it is formed of
cells which are similar to those of the alimentary canal. The second diverticulum is
much smaller and is directed dorsalwards. It touches the dorsal body wall at the pos-
terior end of the brain and is called the pharyngeal diverticulum. As the section passes
through the median line the paired arms and the collar cavities are not seen. The pro-
boscis cavity is much compressed, but the ventral part of the trunk cavity is very
spacious and extends posteriorly into a small sac-like part which marks the origin of the
stolon.

Plate XLIV, figs, i , 2, are transverse sections of the collar and trunk regions of a young
individual slightly older than the one represented in the previous figure. These show
more details of structure, especially of the coelomic cavities. The notochord is situated
between the collar cavities and the proboscis. Its lumen is comparatively wide and is
filled with a deeply staining substance. Nuclei and cell limits are clearly visible in its wall.

The mesodermal cells lining the proboscis and collar cavities have multiplied con-
siderably, so that they not only form the thin lining epithelium of the coelom but fill the
greater part of the cavities. In the proboscis cavity they are aggregated into two
prominent clusters on either side of the notochord, and from these clusters protoplasmic
strands extend towards the ventral wall. These give rise to the bundles of transverse
muscles of the proboscis, which radiate from the septum between the proboscis and the
collar cavities (John, 1931, p. 250).

The ectoderm of the trunk is very thick and carries large numbers of refractive beads.
The cellular structure of the alimentary canal and the mesoderm is very distinct
(Plate XLIV, fig. 2). The stomach is a thick-walled tube with a narrow lumen. The
coelomic epithelium forms a continuous lining of the body cavities. On the dorsal side,
the right and left body cavities do not meet, as the intestine is pressed closely against the
dorsal body wall ; but ventrally the two cavities become slightly wider and meet along the
median line, giving rise to the ventral mesentery. The enlarged ventral part of the trunk
coelom, which has already been referred to, is filled with large numbers of mesodermal
cells. These give rise to the muscles of the stalk and the postero-ventral longitudinal
muscles.

DEVELOPMENT OF CEPHALODI SCU S 201

The heart with the pericardium is observed for the first time at this stage. It is
situated at the tip of the notochord, on tlie septum between the collar and proboscis
cavities. Though it is formed at an earlier stage, owing to the bad state of the material
it has not been possible to locate its position or trace its origin ; but from its relation to
the collar cavities and its structure, in which it closely resembles the heart of Echino-
derms, e.g. Asterias rubens and Echinus miliaris, it seems probable that it originates
exactly as in the latter group (Narasimhamurti, 1932).

The further development is not accompanied by any more changes. All the parts of
the adult zooid except the stalk are now visible. The anterior diverticulum of the pharynx,
which gives rise to the notochord, is still formed of cells which are undifferentiated from
the rest of the alimentary canal, and the muscle fibres are not fully formed. The meso-
dermal cells in the coelomic cavities are oval with distinct nuclei, but the protoplasm of
some of these cells has already commenced to elongate into strands. The stomach is a
thick-walled tube, which is not clearly marked off from the pharynx and intestine. The
gill slits and the vacuolated areas of the pharynx — the pleurochords — are not yet formed,
and there is only a single pair of arms. The brain is developed at the base of the dorsal
ectoderm of the collar and extends from the anterior end of the notochord to the tip of
the dorsal pharyngeal diverticulum ; but the various nerve tracts and the nerve plexus
in the ventral wall of the proboscis are still indistinguishable.

In the next stage the different parts of the body are much better emphasized, but the
young Cephalodisciis still appears to continue its free-swimming life, if one may judge by
the large number of refractive beads present in the wall of the trunk. Plate XLIV, fig. 3,
which is a frontal section of a young zooid, illustrates the very close resemblance between
this stage and the adult individual. The red line of pigment and the nerve plexus
{pni) in the ventral wall of the proboscis have begun to appear and the mesodermal cells
are completely transformed into muscle fibres. The stomach has begun to assume the
characteristic shape of the adult and the intestine appears as a long narrow tube ; but
from the sections it is not certain whether the pleurochords and gill slits are yet formed.

The structure and position of the heart is very clearly seen. It is situated on the
septum between the right and left collar cavities, slightly more towards the left side of
the median septum, and it projects into the proboscis cavity. The pericardium is an
empty sac, the posterior wall of which is invaginated to form the heart. The inner surface
of the wall of this invagination is thrown into numerous deeply staining ridges which
give rise to the glomeruH.

It is not necessary to give any more details of structure, as the young Cephalodisciis
now resembles the adult in all essential characters, except the arms and the stalk, which
are still in a rudimentary state. It seems probable that while the full complement of
arms with the characteristic pinnules are being formed, the refractive beads in the trunk
disappear, the young Cephalodiscus settles down on a suitable substratum and begins to
secrete the coenoecium as the starting-point of a new colony.

202 DISCOVERY REPORTS

DEVELOPMENT OF THE BUD

As the embryonic development of Cephalodiscus is very different from that of the buds
an interesting study is afforded by a comparison of the two. The buds are formed from
the sucker-hke proximal end of the stalk on either side of the median line. In its earliest
recognizable condition the bud is a spherical vesicle which originates as a diverticulum
from the interior of the parent stalk. The trunk coelom, which is continued into the
stolon, is completely filled with muscle fibres and connective tissue, and the rudiment
of the young bud is completely packed with this tissue. This tissue, which is mesodermal
in origin, soon arranges itself in such a way as to give rise to a cavity in the middle, which
communicates with the trunk coelom in the stolon. The bud emerges from the surface
of the stolon as a small knob which rapidly increases in length. The single cavity inside
it divides into two by the development of a median mesentery, and later a small spherical
vesicle is cut off from the distal end of one of the cavities : this has been regarded as the
pericardial sac. The proboscis cavity is believed to be cut off in a similar way from
the tip of the other cavity. The two posterior cavities, which are in communication
with the cavity of the stolon, give rise at a later stage to the paired trunk and collar
cavities.

As the bud continues to grow, it becomes slightly thicker at its distal end and the stalk
by which it is attached to the stolon continues to increase in length. The proboscis now
appears at the distal end of the bud. At first it is a small bulb-like structure marked off
from the rest of the body by a small constriction (Plate XLIV, fig. 5). It develops very
rapidly and assumes the flattened disc-like structure of the adult. The red line appears
on its ventral wall and the posterior lobe formed by the two layers of ectoderm without
an extension of the coelom is also differentiated at an early stage (Plate XLIV, fig. 4, rl
and pi). From the dorsal side of the collar region the arms begin to develop, and by this
process when all the parts of the body are unfolded the bud breaks oflt from the parent
stolon and becomes an independent zooid.

The number of buds produced by each zooid varies in the diflFerent species. In
C. densus and C. fumosus ten to fifteen buds are found, whereas in C. keftrpi two to five
seem to be the maximum number (John, 1931, pi. xxxv, figs, i, 2, 3).

According to Harmer (1905) and Masterman (1900) the ahmentary canal in the bud
makes its appearance as an ectodermal invagination, the orifice of which persists as the
mouth ; but the sections which I have obtained are not in agreement with this view.
Plate XLIV, fig. 4, is a highly magnified part of a sagittal section of a young bud of C.
?iigresce?ts. The trunk is in the form of a narrow tube only slightly thicker than the stalk.
The body wall is filled with aggregations of pigment granules, which give the characteristic
colour of the species. The proboscis is very large and the anterior part of its ventral wall
is thick and glandular. The arms and mouth are not yet formed, but the alimentary canal
is represented by a straight tube which extends downwards from the base of the heart.
Below the posterior end of the primitive gut there is an irregular mass of protoplasm,
which fills the greater part of the right and left trunk cavities. This is probably derived

DEVELOPMENT OF CEPHALODISCU S 203

from the mesodermal structures of the parent stolon, which at a later stage give rise to
the muscles and connective tissue of the trunk.

The heart with the pericardium is a large thick-walled vesicle, which at this stage is
slightly larger than the primitive gut. The invagination of the posterior wall of the
vesicle giving rise to the heart is clearly seen. The large space inside the heart is filled
with a clear fluid, which stains deeply in haematoxylin.

At a later stage, when the arms have begun to develop, the mouth is formed as an
ectodermal invagination which opens into the lumen of the primitive gut at about one-
third the distance from its anterior end. The region in front of the mouth opening be-
comes modified into the notochord, while the posterior part gives rise to the alimentary
canal. Fowler (1905), in describing the development of the buds in Rhabdopleura, states
that probably the greater part of the alimentary canal is derived from a thin-walled tube,
which originates from the mesentery of the stolon. He believed, however, that the
anterior part of the alimentary canal was derived from the ectoderm and that the notochord
was formed from the stomodaeal portion. Though the origin of the alimentary canal in the
buds of Cephalodiscus is to a great extent similar to that of Rhabdopleura, it can be
definitely shown that in the former group the notochord is a part of the primitive gut
and not a stomodaeal growth.

DISCUSSION
LARVAL AND BUD DEVELOPMENT

The differences between the development of the larva and bud of Cephalodiscus may
be summarized as follows :

(i) The bud is attached to the proximal end of the stolon by a narrow stalk, the length
of which increases with the growth of the bud. The trunk region of the bud is a con-
tinuation of the distal end of the stalk, there being no discernible difference between the
thickness of the two during the early stages.

(ii) The proboscis of the bud is much larger than the rest of the body and is cha-
racterized by the presence of the red line from a very early stage ; whereas in the larva
the red line is not visible and the proboscis is comparatively much smaller.

(iii) The bud breaks off from the stolon only after it has reached its maximum de-
velopment, but the larva is free-swimming or crawls about independently.

(iv) In the buds the refractive beads are found only in the tips of the arms or in the
dorsal wall of the proboscis ; but in the larva these are present in very large numbers in
the ectoderm of the trunk. If the view that these refractive beads are defensive organs
should prove to be correct, then it follows that their occurrence in such large numbers
in the body wall of the larva is sufficient evidence for the free-swimming or crawling
habits of the larva. These refractive beads are not present in the trunk wall of the bud
or of the adult zooid, as the bud is not exposed to danger and the trunk of the adult is
protected inside the coenoecium.

204 DISCOVERY REPORTS

(v) The alimentary canal in the bud originates from a rod-shaped primitive gut and
the coelomic cavities are derived from the cavity of the stolon ; but in the larva the
archenteron, which gives rise to the alimentary canal, is formed by invagination and the
coelom originates as pouches from the sides of the archenteron.

(vi) The stalk of the bud persists after the bud breaks away and gives rise to the
stolon of the adult, from the proximal end of which new buds are formed. In the larva
no trace of a stolon is present until the young individual settles down.

LIST OF LITERATURE

Andersson, K. a., 1903. Eine Wiederentdeckung von Cephalodiscus. Zool. Anz., xxvi, No. 697, pp. 368-9.
1907. Die Pterobranchier der Schwedischen Sildpolar-Expedition, 1901-1903. Wiss. Ergebn. Schwed.

Siidpolar-Exped., v, Stockholm, pp. 1-122.
Braem, F., 191 1. Pterobranchier und Bryozoen. Zool. Anz., xxxviii, p. 546.
Fowler, G. H., 1905. Notes on Rhabdopleura. Quart. J. Micr. Sci., xlviii, p. 23.
Gilchrist, J. D. F., 1915. Observations on the Cape Cephalodiscus (C. gilchristi, Ridewood) and some of its

early stages, with an appendix by S. F. Harmer. Ann. Mag. Nat. Hist., ser. 8, xvi, pp. 233-46.
1^17. On the Development of the Cape Cephalodiscus (C. gilchristi, Ridewood). Quart. J. Micr. Sci.,

LXii, pt. 2, pp. 189-21 1.
Harmer, S. F., 1905. The Pterobranchia of the Siboga Expedition, with an Account of other Species. Siboga-

Exp., XXII, monogr. xxvi bis, Leiden, 132 pp.
John, C. C, 1931. Cephalodiscus. Discovery Reports, iii, pp. 223-260.
Masterman, a. T., 1900. On the further Anatomy and Budding Process of Cephalodiscus dodecalophus.

Trans. Roy. Soc. Edinb., xxxiv, p. 507.
Narasimhamurti, N., 1932. The Development and Function of the Heart in Echinoderms. Proc. Roy. Soc,

B, 109, pp. 471-87.
SCHEPOTIEFF, A., 1909. Die Pterobranchier des Indischen Oseans. Zool. Jahrb., Abt. Syst., xxviii, p. 429.

PLATE XLIII

Cephalodiscus nigrescens, Lankester

Fig. I . Median longitudinal section of an embryo showing the ventral
thickening and the inner mass of yolk granules.

Fig. 2. Longitudinal section of a young larva showing the formation of
the coelomic cavities.

Fig. 3. Longitudinal section of a larva of a later stage, in which the
alimentary canal is a straight tube with the anus situated at the posterior
end.

Fig. 4. Longitudinal section of a very young C nigrescens.

a, proboscis; al, alimentary canal; an, anus; arch, archenteron; b, brain;
cc, collar cavity; cp, undivided coelomic pouch connected with the
archenteron; d, notochord; m, mouth; pc, proboscis cavity; pd, dorsal
pharyngeal diverticulum; pg, pigment granules; ra, rudiment of the
first pair of arms; rb, refractive beads; tc, trunk coelom; vp, ventral
prolongation of the trunk to give rise to the stolon ; vt, ventral thickening ;
yg, yolk granules.

DISCOVERY REPORTS, VOL. VI,

PLATE XLllL

rh-

m/

cp

tc

arch

Z.

P/J

3.

JohnbikSoM A LwukluenL'* L<d>1m>

DEVELOPMENT OF CEPHALODISCUS

PLATE XLIV

Ccphalodiscus nigrescens, Lankester

Figs. I, 2. Transverse sections of a young C. nigrescens passing re-
spectively through the collar and trunk-region.

Fig. 3. Frontal section of a young C. nigrescens slightly older than the
one represented in Figs. 1,2, showing the growth of paired arms and the
relations of the heart and pericardium.
Fig. 4. Sagittal section of a young bud.

Figs. 5, 6, 7. Drawings from whole mounts of three stages in the de-
velopment of the bud.

a, proboscis; b, main ganglion; cc, collar cavities; d, notochord;/,
stomach;^, intestine; h, pharynx; Ai, heart ;y, pericardium; mcp, meso-
dermal cells in the proboscis cavity forming muscular strands ; ms, mass
of mesoderm in the trunk cavity of the bud, derived from the stolon of
the mother zooid; n, primitive gut; pc, proboscis cavity; pg, pigment
granules ; pi, posterior lobe of the proboscis ; pni, nerve plexus in the
ventral wall of the proboscis; ra, first pair of arms; rl, red line in the
proboscis ; rp, rudiments of the paired arms ; st, stalk with which the bud
is attached to the stolon of the mother zooid ; tc, trunk cavity ; tr, trunk-
region of the bud; vp, ventral part of the trunk cavity with clusters of
mesodermal cells enlarging to form the stolon.

DISCOVERY REPORTS, VOL. VI.

PLATE XLIV.

ntcp

vp

prti.

"•^

v.-

r-' / ~-st

N

-.sf

5.

Jc^nB«k.5un*\D«jiirtua\L'*Lnndtin

DEVELOPMENT OF CEPHALODISCUS.

[Discovery Reports. Vol. VI, pp. 205-236, Plahs XLV-XLVII, Charts 1-7 (separately), December, 1932.]

REPORT ON SOUNDINGS TAKEN DURING
THE DISCOVERY INVESTIGATIONS,

1926-1932

By
H. F. P, HERDMAN, M.Sc.

CONTENTS

Introduction page 207

Sounding machines 208

Wire soundings 208

Echo soundings 208

Comparison of wire and echo soundings 212

Tiie Scotia Arc 214

Detailed analysis of the various sectors of the Scotia Arc 219

Tierra del Fuego to the Shag Rocks 219

South Georgia and the Shag Rocks 221

South Georgia to the South Sandwich Islands 224

South Sandwich Islands to the South Orkney Islands 225

South Orkney Islands to Clarence Island 227

Soundings in other localities 229

South Shetland Islands and adjacent waters 229

Bellingshausen Sea 232

Other sounding work 233

Echo-sounding table compiled from observations made in the South Sand-
wich Deep 236

Plates XLV-XLVII following page 236

Charts 1-7 ill separate chart-case

REPORT ON SOUNDINGS TAKEN DURING
THE DISCOVERY INVESTIGATIONS,

1926-1932

By H. F. P. Herdman, m.Sc.

(Plates XLV-XLVII; charts 1-7; text-figs. 1-5)

INTRODUCTION

THIS report covers all the oceanic soundings taken during the course of the Discovery
Investigations from 1926 until April, 1932, in the following areas:

(i) In the Scotia Sea and surrounding waters from 51° to 62^ S latitude between 20°
and 70° W longitude.

(ii) In Bransfield Strait, the South Shetland Islands and the Palmer Archipelago.

(iii) In the Bellingshausen Sea.

In the first area are included intensive series of soundings round South Georgia and
the Shag Rocks and some of the soundings taken during the running survey of the
South Sandwich Islands in 1930. The soundings have been obtained by the R.R.S.
'Discovery' in 1926 and 1927, by the R.R.S. 'William Scoresby' in the years 1926-31
and from 1930 onwards by the R.R.S. 'Discovery II'. The majority of the soundings
have been taken by the last-named ship and by the echo method. In the detailed discussion
of these soundings prominence has mainly been given to those taken in oceanic waters.
Inshore soundings taken round the South Sandwich Islands are included in the recently
published report ^ on that group of islands. Intensive soundings in the bays and harbours
of South Georgia have also been omitted, as they are included in the report on the survey
operations round South Georgia by Lt.-Cmdr. Chaplin.^

In the ' Discovery' and the ' William Scoresby' difiiculty was sometimes experienced
in obtaining good wire soundings owing to the impossibility of keeping the sounding
wires vertical in bad weather, and for this reason not all of their soundings can be
accepted. In the ' Discovery II ', however, it was possible to obtain reliable wire sound-
ings in moderately bad weather by keeping the ship head-to-wind and steaming slowly
ahead whenever necessary to correct the "stray" on the wire. Since echo soundings,
however, can be taken every few minutes, and usually without stopping, the wire method
was latterly abandoned. Approximately 10,000 echo soundings were obtained between
February 1930 and May 193 1, and of these the greater number were taken between
December 1930 and March 1931. During the season 1931-2 approximately 2000 more

1 Kemp and Nelson, The South Sandwich Islands, Discovery Reports, iii, pp. 133-98 ('930-

2 Chaplin, Narrative of Hydrographic Survey Operations in South Georgia and the South Shetland
Islands 1926-30, Discovery Reports, iii, pp. 297-344 (1932)-

2og DISCOVERY REPORTS

echo soundings have been obtained, and on the evidence provided by all these soundings
it has been possible to arrive at some definite conclusions as to the bottom relief in most
of the above-named areas.

The positions of all soundings obtained by the R.R.S. ' Discovery II ' were determined
by Lieut. A. L. Nelson, R.N.R., and the majority have been plotted by him on large-scale
charts. His careful and painstaking vi^ork has been of the greatest assistance in the
preparation of this report. Vice-Admiral H. P. Douglas, C.B., C.M.G., R.N. (retd.)
and Mr J. M. Wordie have been kind enough to examine this report in manuscript and
I am indebted to them for a number of valuable suggestions.

SOUNDING MACHINES

WIRE SOUNDINGS
For oceanic wire soundings the ships were fitted with the well-known Lucas sounding
machine,! manufactured by the Telegraph Construction and Maintenance Co., which
carries about 10,000 m. of 0-028 in. diameter single-strand bright steel piano wire in one
length. This wire had a breaking strain of 220 lb. The ordinary Baillie sounding rod, also
described in the Discovery Reports^ was used throughout for all soundings over 300 or
400 m., the weights being the usual cones and flats of about 22 lb. weight apiece. In
shallow water, and sometimes up to 500 m., the Admiralty 28 lb. sinker was used, as was
also a new pattern of bottom sampler designed by Mr F. E. C. Davies to bring up larger
quantities of material than were usually obtained by the ordinary methods. Certain
small defects which had been noted in the Lucas sounding machines fitted in the
'Discovery' and 'William Scoresby'^ were overcome in the later pattern fitted in the
' Discovery II ', which gave every satisfaction on the occasions when its use was necessary.
Motor-driven Kelvin sounding machines were fitted in the ' Discovery' and ' Discovery
ir and a hand-operated one in the 'William Scoresby'. In the 'Discovery H' this
machine was seldom used except for the purpose of checking the accuracy of the shallow-
water echo-sounding machine. The positions of the Lucas and Kelvin sounding machines
in the ' Discovery ' and ' William Scoresby' have already been discussed in the Discovery
Reports. In the ' Discovery II ' the Lucas sounding machine was placed on the port
side of the forecastle head, about 12 ft. aft of the stem, and the Kelvin machine on the
starboard side of the bridge deck close to the chartroom door, the wire being carried
outboard by a boom and traveller.

ECHO SOUNDINGS
The echo-sounding machines are of the British Admiralty pattern, and with the
exception of the experimental machine fitted in the ' Discovery ' for a year, were manu-
factured by Messrs Henry Hughes and Son of London. The mechanical details of the

1 Kemp, Hardy and Mackintosh, Objects, Methods and Equipment, Discovery Reports, I, pp. 168, 169,
190 (1929).

2 Kemp, Hardy and Mackintosh, loc. at. supra, pp. 210, 211.

3 Id., p. 168.

SOUNDING MACHINES 209

oceanic and the shallow-water machines differ considerably, but the principle is the
same : a blow from a hammer operating on the ship's hull produces a sound wave which
is reflected from the bottom and picked up by a very sensitive hydrophone, and from
the time interval between the transmission and the reception of the sound the depth of
water can be calculated. Observations made at various times on the salinity and tempera-
ture of sea water have been used for the calculation of the velocity of sound in that
medium.

Shallozc-water echo-sowiding machine. In this machine the hammer is electrically
operated and the speed of the transmitter switch is automatically adjusted by means of
a governor and moving resistance to ensure approximately 180 blows per minute. The
receiving gear, which is situated on the bridge, houses the transmitter switch and the
telephone gear; from it the depth can be read off directly in metres, the corrections
for salinity, temperature and pressure being negligible up to the extreme range of this
machine — 220 m. In practice this machine was seldom used for depths beyond 160 m. ;
for it was found more convenient to change over to the oceanic machine at this point in
order to avoid any doubt as to whether soundings were shoaling or increasing when the
limit of 220 m. was reached.

All the ships were fitted with the listening type, as distinct from the recording type, of
shallow-water echo-sounding machine, but only in the 'William Scoresby' and the
' Discovery II ' was it extensively used. In the former ship it proved very useful during
the trawling survey of the Falkland Islands in 1927, as by taking soundings every few
minutes when trawling, the length of warp paid out could be adjusted to any consider-
able change in depth. A later pattern of the same type of machine was fitted in the
'Discovery 11', which gave considerably better results than the earlier model, though
when in constant use for long periods, a considerable amount of attention was required
to keep it in adjustment. Apart from some slight delays caused by minor mechanical
defects, which were easily made good on board, and the leakage of water into the hydro-
phone through faulty assembling, the machine was in continuous use for weeks at a time
over a period of fifteen months and gave excellent results.

The hammer was fitted under the stokehold on the port side of the double bottom between the
oil-fuel tanks and the stokehold fresh-water tanks. Ahernative positions were provided for the
hydrophone, one at the after end of the double-bottom fresh-water tank on the starboard side of the
fore hold and the other about 6 ft. farther aft in the double-bottom between the fresh-water tank
and the starboard oil-fuel tank. After the hydrophone had been flooded in the fresh-water tank,
it was shifted to the alternative position, but although there was still a tendency towards several
false echoes, it worked quite satisfactorily. Neither position was entirely free from false echoes at
certain points on the scale, but with practice the operator could easily distinguish the true echo from
the false by the difference in note, and even when the true echo coincided on the scale with the false
the note was unmistakable. When the ' Discovery II ' was re-commissioned in 193 1, the hydrophone
was replaced in the fresh-water tank, where, with the exception of the recurrence of a trouble
previously experienced, it has worked quite satisfactorily.

When the hydrophone was being overhauled in February 193 1, while endeavouring to trace the
cause of very faint echoes, it was found that the stalloy plate which forms the base of the hydrophone
and on which the micro-button is mounted had been bent inwards towards the centre, thereby

2IO DISCOVERY REPORTS

throwing the button out of ahgnment. This was no doubt due to the freezing of the fresh water in the
chamber under the hammer, as it was against the hull and the temperature of the sea was about
— 1-5° C. A new stalloy plate was fitted, glycerine added to the fresh water in the chamber to
prevent freezing and no further trouble was experienced till 1932. When the hydrophone was replaced
in the fresh-water tank in September 1 93 1 , glycerine was again added, but there appears to have been
a small leak in the joint where the water chamber butts on the hull. This joint consists of a dermatite
ring, and the hydrophone and water chamber are braced down on it by a strongback clamp. When
the ' Discovery II ' was again in water of — i-o° C, the echoes, as before, became very faint, and on
examination of the hydrophone it was found that the stalloy plate was again bent inwards and that
all the glycerine had diffused into the fresh-water tank through the fault in the joint. This water
chamber, which is fitted under both hammer and hydrophone in the shallow-water machine, is
of great importance, as by pumping water into it till a pressure of about 20 lb. to the square inch is
reached, the chamber acts much in the same way as a lens does with rays of light : it concentrates
the sound waves and prevents the scattering of sound noticeable in the earlier models.

Oceanic echo-soimding tnachmes. The deep-water echo-sounding machine fitted in the
'Discovery' in 1925-6 was an Admiralty experimental model. Owing to the difficulty
of fitting this machine in a wooden ship, the hammer had to be slung outboard from a
davit, and it was found that the rolling of the ship caused excessive water noises in the
hydrophone. This, together with mechanical troubles due to the unusual position of
the hammer, prevented the machine from working properly, and it was removed in 1926.
The deep-water echo-sounding machine at present fitted in the 'Discovery II' is of
the listening type and has a much more powerful hammer than the shallow-water in-
strument. The hammer is operated by compressed air, and controlled electrically from
the receiving gear in the chartroom. Impulses can be sent out at three different rates,
one, nine, and ten in 11-25 ^ec, which is the time taken to complete one circuit of the
transmitter switch. Here, as in the shallow-water machine, the speed is controlled by
a governor in conjunction with a resistance. In the ' Discovery II ', however, this control
was not always satisfactory when the machine was running for long periods, and it was
necessary to check the speed of the transmitter switch at frequent intervals and correct
the soundings accordingly. It was found after several experiments that the temperature
inside the case of the receiving gear rose in 20 min. from 11-5" to 15° C, and that the
time for a revolution of the transmitter switch fell from 11 -22 to 10-93 sec. If the
machine is only running for 3 or 4 min. at a time the error from this cause is, however, prac-
tically negligible, and the bulk of the oceanic soundings were taken under these conditions.

The impulse rates are selected at will by the operator, and to obtain a sounding it is
usual to get the approximate time interval between hammer blow and echo on the one-
impulse circuit and then tune in more accurately on the nine- or ten-impulse circuit. The
reading obtained is in seconds and gives the time taken by the sound waves for their
double journey, i.e. from the hammer to the bottom and back to the hydrophone. A
skilled operator can easily determine the time correctly to o-oi sec, which roughly
corresponds to an error of 7 m., and the depth is quickly calculated after applying the
necessary corrections. A book of correction tables^ has been calculated by Mr D. J.

1 Tables of the velocity of sound in pure water and sea ivater for use in echo-sounding and sound-ranging,
H.D. 282, H.M. Stationery Office, London (1927).

SOUNDING MACHINES 211

Matthews of the Hydrographic Department of the Admirahy, in which the oceans of the
world are divided up, according to the sahnity and temperature of the water, into areas
in which the same corrections can be apphed. Instructions are also given for working
out new tables for any area from the hydrological observations. In this way tables were
worked out in the ' Discovery II ' for the South Georgia and Scotia Sea areas from the
average of a large number of salinity and temperature observations, but the results were
found to vary very little from those already determined by Matthews. A special table
(see p. 236) extending to 7250 m. was compiled from observations in the South Sand-
wich Deep and was used for the correction of all soundings in that area.

The proved effective range of this oceanic machine was from approximately 90 to
8100 m., and the maximum depth at which soundings have been obtained with the ship
under way was just over 5000 m. The limiting factors in obtaining soundings under way
are water noises in the hydrophone caused by wind and sea — factors which may vary
considerably. No definite statement can be made as to the best method of obtaining a
sounding in bad weather; only actual experience of a ship's behaviour in different
weather conditions will enable the operator to find the best way of manoeuvring the ship
in order to avoid excessive water noises in the hydrophone. In the area where we were
working heavy swells and high seas are unusually prevalent, but soundings up to 500 m.
could nearly always be obtained. In deeper water it was sometimes necessary to
manoeuvre the ship with regard to wind and sea. In our experience the time taken to
complete a 5500 m. sounding under average conditions rarely exceeded 4 min., and even
in very bad weather it was unusual to stop for more than 7 or 8 min. Only once during
our intensive sounding programme was the weather sufficiently bad to prevent us obtain-
ing any soundings and then conditions improved within a few hours.

The present instrument in the ' Discovery II ' has the hammer fitted in the well between the engine-
room and stokehold double-bottom tanks, on the port side. The hydrophone was fitted in the star-
board oil-fuel well just beside the alternative position for the shallow-water hydrophone. The ham-
mer is mounted on a water chamber in a manner similar to the arrangement of the shallow-water
gear, but the base of this chamber consists of a special circular steel casting which is let into one of
the bottom plates and consequently exposed to the sea. The hydrophone is enclosed in a sluice
valve and can be lowered into the sea or withdrawn into the hull at will. Although the recommended
position of the hydrophone for sounding was to lower it well clear of the hull we found that the best
position in the ' Discovery II ' was with the stalloy plate of the hydrophone flush with the hull, and
we had no trouble in obtaining our maximum sounding of 8102 m. with it in this position.

During the first season, 1929-30, the machine was not altogether satisfactory mechanically, and in
the middle of the South Sandwich Islands survey in March 1930, the hammer cylinder sustained
a fracture and broke in half. Fortunately it was a clean break, and our Chief Engineer was able to
devise a form of collar to hold the two parts together, enabling us to have the machine running again
in a day or so. This fracture, however, necessitated a new hammer being sent out from England and
fitted in Simonstown in July 1930. After this the machine worked extremely well, though like the
shallow-water instrument it required constant attention during periods of continuous use. Much of
its efficiency was due to the care and trouble expended on it by R. W. Mackay, the ship's technical
assistant, and the engineering staff. One of their problems was the refusal of the piston of the ham-
mer to move in very cold weather or its sudden stopping after it had been running for a few minutes.
This was found to be due to the compressed air which comes over from the compressor quite hot.

212 DISCOVERY REPORTS

This air gives up moisture on its sudden expansion in the cold hammer chamber causing a sticky
emulsion to form with the oil used for lubricating the piston. The piston then sticks and it is necessary
to take off the head of the hammer to remove the piston for cleaning at very frequent intervals.
Eventually, while the ship was in cold waters, the hammer was run practically without lubricant, the
piston being occasionally wiped with a rag which had been soaked in very thin oil. Even under these
conditions it was still necessary to remove the piston for cleaning every 8 to 12 hours when the
machine was in constant use. During the refit in London in 193 1 the hammer was surrounded by a
steam coil which has had the desired effect of preventing the condensation of the water from the
compressed air. Another matter which required the frequent attention of the engineers was the air
compressor itself, which was found to be inadequate for the long periods of continuous running
demanded from it by the sounding programme. It was only by careful and continuous attention that
the motor and bearings were prevented from wearing out completely, and as the ability of a compres-
sor to run continuously for several days is a very important factor in our sounding work, a new
compressor, with a much larger capacity and of heavier construction, was fitted in London in 193 1.
Trouble was also experienced with the reducing valve for regulating the pressure of air required to
reset the hammer, but a new valve constructed on board overcame this difficulty.

From 1926 to 1929, when only Lucas sounding machines were being used in oceanic
waters, soundings were usually taken when the 'Discovery' or the 'William Scoresby'
was stopped for observations on hydrology and plankton. In the ' Discovery II ' with an
efficient deep-water echo-sounding machine it was possible to take soundings at much
more frequent intervals, and on a number of occasions the machine was used day and
night without intermission for weeks at a time. For station work the Lucas machine was
superseded mainly on account of the great saving of time. A Lucas sounding to 5000 m.,
which is roughly the average depth of our oceanic stations, takes at least li hours to
complete. The ship requires to be constantly manoeuvred with regard to this one wire,
and as the forward hydrological machine on the forecastle head of the ' Discovery II '
is only 22 ft. aft of the Lucas sounding machine, it is impossible, except on very rare
occasions, to work deep water-bottles when the Lucas sounding wire is out. An echo
sounding on station does not take more than 5 min. to complete, and there is no hin-
drance to the deep water-bottles being lowered as soon as the ship is hove-to. In the
variable weather conditions which prevail in the Scotia Sea, where the majority of our
stations were worked, it was important to reduce the time taken for observations, and in
this respect great advantage was obtained from the use of the echo-sounding machines.
It is a testimony to their efficacy that at several hundred stations we relied wholly on
soundings obtained in this way, and that instruments were constantly lowered to within
a short distance of the bottom, even in the greatest depths, without ever coming in
contact with it.

COMPARISON OF WIRE AND ECHO SOUNDINGS

At various times, mainly during the early part of 1930, comparisons were made by the
' Discovery 11', when weather conditions permitted, between Lucas and echo soundings.
The accompanying table gives the data obtained, the average of the 21 comparisons
showing a percentage excess of Lucas depth over echo depth of 1-42. The greatest and
least errors, in soundings of over 2000 m., were respectively 3-35 and 0-28 per cent. In

COMPARISON OF WIRE AND ECHO SOUNDINGS

:i3

the table the velocity of the wind is given in knots, and it may be assumed that when
it falls below lo knots the ship was being allowed to drift, as the force of the wind
was not strong enough to enable us to adopt the usual practice of keeping the ship head-
to-wind. Later, when several lines of echo soundings had been run, it was found that the
crossing of a previous line gave a valuable check on the accuracy of the soundings, and
in all the cases where two oceanic lines of echo soundings cross the agreement has been
excellent.

Table of comparison between Lucas and echo soundings

Sounding (metres)

Wind

Excess of

Lucas
sounding

Date

Station

Lucas Echo

Direction

Velocity

/I i \

(knots)

0/

/o

20.1.30

301

1858

1839

NNW

II

1-02

29/30. i. 30

319

2409 2347

NNE

22

2-52

30.1.30

321

41 17 3994

NEby N

18

3-35

4.11.30

334

3250

3233

wsw

16

0-52

4.11.30

335

3416

3377

Wby S

18

I-I4

5.11.30

336

3193

311S

N

7

2-44

7.11.30

343

960

936

W

9

2-56

9. 11. 30

353

1167

1134

NW

18

2-83

9.11.30

355

3714

3641

NWby N

20

1-96

10. 11. 30

356

3272

3204

NNW

12

2-o8

10. 11. 30

357

3645

3616

NW

28

o-8o

II. 11. 30

358

3549

3539

WNW

28

0-28

19. 111. 30

373

2515

2493

Wby S

10

0-87

2. iv. 30

—

3213

3169

—

—

1-37

13/14- iv. 30

382

3652

3617

ssw

16

0-96

14. Iv. 30

384

3663

3632

SEby E

15

0-85

15. iv. 30

385

3637

3612

ENE

23

0-69

iS.iv. 30

386

4775

4734

Light airs

—

0-86

16. iv. 30

387

3102

3076

ENE

20

1-89

20. X. 30

460

3802

3730

ENE

22

0-84

30. X. 30

560

335

335

WSW

8

0-0

From exhaustive observations recently obtained by H.M. Surveying Ships in com-
paring echo and wire soundings at all depths up to 5000 m. it has been ascertained that
the echo method is by far the most accurate. This is principally due to the fact that
however carefully the ship may be manoeuvred there must be a divergence from the
vertical which the wire takes up. This is obvious from the above table where the depths
by the Lucas machine are all greater.

Up to the limit of the shallow-water echo-sounding machine soundings were usually
taken every i or 2 min., depending on the scale of the survey which was being carried out
and on the speed of the ship, but continuous soundings can be obtained if they are
necessary for navigational purposes. Under normal conditions the interval between
soundings was lengthened as the depth increased, till in oceanic waters it reached 30
or 60 min., depending largely on whether a stop of about 5 min. is required for the

214 DISCOVERY REPORTS

sounding. Certain oceanic work, however, such as the search for a reported rock or the
detailed examination of a submarine ridge may require soundings to be taken as often as
every 5 min., and up to a depth of about 800 m. soundings with the oceanic machine
have been taken, in favourable circumstances, at 2 min. intervals.

THE SCOTIA ARC

One of the most interesting features of the topography of the bottom in the area of
the Falkland Islands Dependencies is the supposition, much discussed by geologists,
that the Andes formerly joined Graham Land and the South Shetland Islands to form
a mountain arc or loop which extended eastwards into the South Atlantic. It is generally
held that if there is such a connexion the probable course must have been from Tierra
del Fuego through Staten Island, the Burdwood Bank, the Shag Rocks, South Georgia,
the Gierke Rocks, the South Sandwich Islands, the South Orkney Islands, Elephant and
Clarence Islands and the South Shetlands. This connexion, if it really exists, is not only
of great interest from a geological point of view, but it is also of very great importance in
the studies we are making of the hydrological conditions in the area, for if the missing
portions of the arc are represented by submarine ridges, they will certainly exert a
profound influence on the movements of the water masses in their vicinity.

Geological writers have generally referred to the supposed arc as the South Antillean
Arc. There are, however, no South Antilles, and in view of the grouping of the fragments
round the Scotia Sea, it seems preferable, as Mr Wordie has suggested, to speak of the
connecting loop as the Scotia Arc.

The first mention of a possible connexion appears to have been made by Bellings-
hausen, of whom Suess writes :^

" Second advance : Bellingshausen after a visit to the South Sandwich Islands in 18 19,
during which he witnessed a volcanic eruption in Sawadowskij [Zavodovski] , the most
northerly of these islands, wrote as follows : — ' The Sandwich Islands and the Traversey
Islands appear to form the summits of a mountain ridge which is connected through
Clerk's reef with South Georgia, and from there on by the Aurora reef with the Falkland
Islands'".

Since Bellingshausen's time and before Suess' examination of all previous data,
several other investigators, notably Barrow, Reiter, Arctowski, Nordenskiold and
Gunnar Andersson, had put forward similar theories.

Suess, after considering all the available evidence, regards it as established^ "that in
the interval we have just discussed the Pacific structure advances for the second time
into the Atlantic region''. It should be noted, however, that in Suess' opinion South
Georgia does not form a continuation of the Burdwood Bank, but that an arcuate con-
nexion through the Shag Rocks is more probable, and in his detailed comparisons^ he
observes that:

1 Suess, The Face of the Earth (authorized English Translation), iv, pp. 488, 489 (1909).
" Loc. cit., p. 496. ^ Loc. cit., p. 495.

THE SCOTIA ARC 215

"(e) No connexion can be shown to exist between the widely remote volcanic indica-
tions of Burdzvood bank, the volcanic arc of the South Sandivicli islands, and the
Bransfield volcauos (Bridgman, Deception, and others). The resemblance between the
South Sandwich Islands and the volcanos of the lesser Antilles has often been remarked
upon.

"Existing observations scarcely permit us to carry comparisons further. It must not
be overlooked that Staten Island and Cape Horn do not present a completely concordant
strike, and that Mount Oreille may mark perhaps the beginning of a new volcanic arc
resembling the South Sandwich Islands more closely than Burdwood bank. Arctowski
marks - 4040 meters a little to the south of Staten Island. The connexion of the other
groups, such as South Georgia, the South Orkneys and the South Shetland islands, must
remain an open question".

Suess also states that most modern investigators accept the hypothesis that the Andes
are seen again in Graham Land.

Since Suess discussed the question, a considerable amount of geological work has been
carried out in these regions, as a result of which Gregory, in his presidential address to
the Geological Society in 1929,^ strongly contested the theory of an arc along the line
generally indicated. His objections were based mainly on various observations and
collections made in South Georgia, by Ferguson in 1911-12'", and Douglas during the
visit of the 'Quest' in 1922.^ Gregory also supported the conclusions arrived at by
Kiihn in 1920,* who maintained that the supposed extension of the Andes through
South Georgia was supported neither by the geology of the island nor by the form of the
adjacent sea floor. The geological evidence in 1920 may have been sufficient to warrant
this statement about South Georgia, but at that time there were not, in my opinion,
sufficient oceanic soundings in this area from which any conclusions could be drawn as
to the form of the sea bottom round the island. As the result of all the evidence available
in 1929, Gregory considered that South Georgia appeared to be "a fragment of an
ancient South Atlantic land, and the south-western corner of the Flabellitesland of
Schwartz". In coming to this conclusion, Gregory ignored the work of Heim in 1911^
and Wordie in 1914,'' who both maintained that Ferguson's interpretation of South
Georgia was inadmissible.

Since then, however, HoltedahF has made a considerable number of observations in
South Georgia, the South Shetlands and some of the outlying rocks and islands, and

' Gregory, The Geological History of the Atlantic Ocean, Quail. Joiini. Geol. Soc. London, Lxxxv,
pp. cix-cxi (1929).

2 Ferguson, Tram. Roy. Soc. Edin., l, pp. 797-816 (1915).

^ Geol. Coll., Voyage of the 'Quest', pp. 51-54 (British Museum), 1930. See also Nature, cxxvi, p. 837.

* Kiihn, Der sogenannte " Siidantillen-Bogen" und seine Beziehungen, Zeit. Ges. Erdk. Berlin, pp. 249-
269, especially pp. 255, 261 (1920).

5 Heim, Zeit. Ges. Erdk. Berlin, No. 6, pp. 451-6 (1912).

" Wordie, Trans. Roy. Soc. Edin., liii, pp. 17-27 (1925).

' Holtedahl, On the Geology and Physiography of some Antarctic and sub-Antarctic islands, Norweg.
Antarct. Expeds., 1927-8 and 1928-9, No. 3, Oslo (1929).

2i6 DISCOVERY REPORTS

confirms Heim and Wordie's views on the structure of South Georgia. In discussing the
bearing of his observations on the problem of the South Antillean Arc, HoUedahl sees
no reason to beheve that South Georgia hes outside the main Andean fine of folding .^ He
has also endeavoured to show that there is a distinct geological likeness between South
Georgia and the Staten Island — Tierra del Fuego district. Concerning the remainder
of the arc he remarks: " the fact that various types of rocks, or even rocks of no doubt
different age, occur in various places, does not in any way contradict the assumption of
these land-masses being parts of the same old range", and with regard to the South
Orkneys he suggests that the divergence of the strike from the trend of the submarine
ridge may be purely local as "we know from many ranges that the strike may locally
diverge very considerably from the main trend of the folding belt".

Wilckens," in his criticism of Staub's theory'* of the non-existence of an arcuate con-
nexion between the Andes and Graham Land, supports Holtedahl's contention that
Tierra del Fuego and South Georgia are of a similar geological structure and, in general,
considers that the existence of a submarine arc has been proved.

The bathymetric chart of the Scotia Sea shown by Pratje,"* which is based mainly on
the echo-sounding profiles of the ' Meteor' during 1925 and 1926, supports the theory of
the island arc, though he shows a greater average depth between the Burdwood Bank and
the Shag Rocks and between South Georgia and the Shag Rocks than actually exists.
Comparison with our bathymetric chart of the same area (Plate XLV) will show that the
depth here is considerably less than that given by Pratje.

The principal evidence that we are able to bring forward in regard to the Scotia Arc
is derived from soundings, and, as will be seen from the charts and plates accompany-
ing this paper, the soundings indicate clearly that well-defined submarine ridges exist
between Graham Land and South America along the generally accepted line. The only
geological evidence obtained during the course of the Discovery investigations relates
to the South Sandwich Islands. During our running survey of the group in March
1930 some rock specimens were collected from Beach Point, Thule Island, by Dr E. H.
Marshall and Mr T. J. Hart. These have been carefully examined and analysed by
Tyrrell,^ who pronounces them to be of a definite Andean type.

Our soundings show very definitely that in many places there are concentric sub-
marine ridges which may in themselves be considered as favourable geological evidence
of an Andean loop. All the sections drawn across the various sectors of the ridge show
this peculiarity, though it is much more marked in the South Sandwich — South Orkneys
and the South Orkneys — Clarence Island sectors, as well as across the South Sandwich
Deep. Between the Burdwood Bank and the Shag Rocks the section (Fig. 3 b, p. 226),
drawn approximately down the 49th meridian, shows numerous rises; but between the

^ Loc. n't., pp. 104-17.

2 Wilckens, Natunv. Abt. Niederrheiii. Ges. Nat. Heilkunde, 1930-31, pp. 1-14, Bonn (1932).
^ Staub, Die Bewegungsmechanismus der Erde, Berlin.

* Pratje, Beitrage zur Bodengestaltung des Siidatlantischen Ozeans. Cent.f. Mill., Abt. B, p. 132 (1928).
"^ Kemp and Nelson, The South Sandwich Islands, with a report on rock specimens by G. W. Tyrrell,
Discovery Reports, in, pp. 191-7 (1931).

THE SCOTIA ARC 217

Shag Rocks and South Georgia the soundings indicate a connexion which is single for the
greater part of its length, a double ridge being as yet demonstrated for a few miles only
at the point where the section Fig. 2 b runs across. From South Georgia to the South
Sandwich Islands the section (Fig. 2 a) shows little evidence of a connexion, but, as will
be shown later, the influence of the South Sandwich Deep begins to be noticeable not far
from the Gierke Rocks. In the South Sandwich Group itself we have the most pronounced
example of ridging in the whole arc. Here there are two comparatively shallow sub-
marine ridges, with an average depth of water of about 2000 m., separated by a very
narrow deep with soundings of over 8000 m. It is of great interest to note that the
existence of this deep was forecast many years previously by Suess^ as follows:
" Immediately to the east of the South Sandwich islands no soundings have been made.
We might, perhaps, expect to find a foredeep there similar to that which occurs outside
a part of the Antilles".

The discovery of the South Sandwich Deep by the ' Meteor', in 1926, when running
their fifth profile eastwards along the 55th parallel of south latitude, confirmed this
theory of Suess. Although their line of soundings only crossed the Deep at its extreme
northern end, their conjecture that it extends right down the eastern side of the South
Sandwich Group has proved correct, at any rate in part. In November 1930, while the
' Discovery II ' was on passage from Cape Town to South Georgia via the ice-edge, we
were able to confirm the Meteor's observations. We were unable, however, to obtain as
deep an echo sounding as the 'Meteor' at this point, although we took a number of
soundings over a considerable area. The Meteor's deepest echo sounding in this area is
now given in their latest report ^ as 8264 m., the original echo distance of 8060 m. given
in an earlier publication => having apparently been corrected, using a velocity of sound in
sea water of 15 17 m. per second. The deepest sounding obtained by the ' Discovery IT
in this area was 7977 m., and the velocity of sound in sea water used for the correction of
the depth was 1515-7 m. per second. The table of velocities of sound (see p. 236) used in
the correction of all the echo soundings taken by the 'Discovery 11' in the South
Sandwich Deep and adjacent w^aters was compiled in accordance with the instructions
contained in the Admiralty tables of the velocity of sound in sea water. The hydrological
observations necessary for this work were made at St. 471 in 54° 57' S, 27° 59*' W,
when water samples and temperatures were obtained to a depth of 7250 m. Actually
this table varied very little from Table 23 in the Admiralty handbook as far as the
latter went; its great value was in providing velocities of sound below 5000 m. without
having to resort to extrapolation, which might have introduced errors.

In the latter half of February 193 1, two more lines of echo soundings were taken by
the ' Discovery 11' in the area of the South Sandwich Deep. The first crossed the line
of the islands in about 57° S, approximately midway between the Candlemas Islands and
Saunders Island. These islands lie about 45 miles apart, the greatest distance between

' Suess, loc. cit., supra, p. 496.

2 ^y„. Ergebn. Deutsch. Atlant. Exped. 'Meteor', i, pp. 183-4.

3 Spiess, Das Meteorfahric, p. 336, Berlin.

2i8 DISCOVERY REPORTS

adjacent islands in the group, and one of the main reasons for crossing here was the lack
of soundings between them. We had obtained a few soundings between the remaining
islands during the 1930 survey, although at that time the oceanic echo-sounding machine
was not giving the excellent results obtained in the following season. These soundings,
however, were sufficient to show that the remaining islands are almost certainly con-
nected by a submarine ridge, so that if there were any discontinuity, it could only be
found here. As will be seen from Fig 4 o, which is a section of the Scotia Sea from west
to east, this 193 1 line of soundings shows a gradual rise from the west towards the
connecting ridge between Saunders Island and the Candlemas Islands, the shallowest
sounding being 2185 m. On the eastern side of the group there is a very sharp drop to
8000 m. in the Deep. We continued this line of soundings along the 57th parallel of
latitude to approximately 23° W longitude. At this point we turned back and ran our
second line in a north-westerly direction, crossing the line of the islands again some
40 miles north of Zavodovski Island in about 28° W longitude. Both these traverses
confirmed the existence of the Deep as a very narrow trench, with a width of about
20 miles at the deepest part and having very steep sides below the 5000 m. level. While
carrying out this examination of the Deep in 193 1 we were unfortunate in having un-
suitable weather for echo sounding ; it was almost always necessary to stop for soundings
over 4000 m. and water noises made it difficult to hear the echo from over 7000 m. We
were able, however, to get thirty-eight soundings over 6000 m. on the two lines, and these
in conjunction with the echo soundings taken at the northern end of the Deep in 1930
gave us altogether forty-three soundings over 6000 m. Of these, twenty-seven were over
7000 m. and two over 8000 m., the deepest sounding being 8102 m. on the north-
westerly profile in a position approximately 56° 33 S, 24° 33' W (see Chart 4).

The northern end of the South Sandwich Deep can be traced by means of the 5000 m.
contour, to a position about 50 miles east of the Clerke Rocks. Several of the
soundings which determine this contour were taken by the 'Deutschland' in 1911-12
and the ' Meteor' in 1926. All the remaining soundings have been obtained by the ships
of the Discovery Committee and are mainly echo soundings from the ' Discovery 11'.
The southern end of the South Sandwich Deep, south of latitude 57', still remains to be
determined, but during the season 193 1-2 the ' Discovery IF obtained soundings of over
5000 m. east and south-east of the southern extremity of the South Sandwich Group,
and it is possible that these may represent the limits of the Deep in this direction.

It is of interest to note here that a preliminary comparison of a few of our results in
this area with those of the ' Meteor', has already been made by Stocks. ^ The ' Discovery'

1 Stocks, Zcif. Ges. Erdk. Berlin, pp. 299 ct seq. (1931).

Since this report was written, Stocks has published another brief paper (Der Sudantillen-Bogen im
Lichte neurerer Erkundungen, Zeit. Ges. Erdk. Berlin, 1932, Nr. 5/6) on the South Antillean Arc, which
appears to be based very largely on soundings taken by the vessels of the Discovery Committee, and recently
published in the Admiralty charts for the area. In this paper, which has perhaps been hastily written, the
following conclusions, which are not borne out by our recent work, are advanced:

(i) It is stated that soundings show that the probable continuation of the Patagonian Shelf towards South

TIERRA DEL FUEGO TO THE SHAG ROCKS 219

results are shown in the form of a contoured chart of the South Georgia — South Sand-
wich area, which gives the position of the soundings but no actual depths, and a section
drawn across the South Sandwich Deep. Stocks uses the uncorrected (unreduzierte
Werte) soundings of the 'Meteor' and assumes that the echo soundings of the
' Discovery II ' are also uncorrected. As already explained, however, all the ' Discovery
II ' echo soundings have been fully corrected. Stocks noted that the Deep extended to
a position farther west than the 'Meteor's' observations would tend to show, and
pointed out the bearing of the new facts on Suess' theory. A bathymetric chart of
the Antarctic and sub-Antarctic areas embodying these results was published by the
American Geographical Society in 1929.^ This chart shows the South Sandwich Deep,
and although on a very small scale, gives an accurate representation of the bathymetric
conditions in the Scotia Sea as then known.

DETAILED ANALYSIS OF THE VARIOUS SECTORS
OF THE SCOTIA ARC

TIERRA DEL FUEGO TO THE SHAG ROCKS

(Plate XLV, Charts 1,2,3)

It is now generally accepted that if the Andes are continued on towards South Georgia
the continuation must He through the Burdwood Bank, and our soundings are in definite
support of this theory. A distance of 500 miles separates the eastern end of the Burdwood
Bank from the Shag Rocks, and it will be seen from the bathymetric chart of the Scotia
Sea (Plate XLV) that a well-defined submarine ridge can be traced by means of the 2000
and the 3000 m. contours for nearly the whole distance. There appears to be a break of
approximately 60 miles from about 47° W to about 49° W, but further soundings to the
north of the line on which the present contours are based may show that the ridge is
continuous. The ridge is much more sharply defined to the west of this gap, there being
two well-marked areas where the soundings are considerably under 1000 m. The
minimum sounding in the largest of these areas was 368 m. in a position approximately
53° 55' S, 52° 25' W and in the other area, 594 m. in approximately 53° 28' S, 49' 51' W.

Georgia is through the Falkland Islands, whereas an examination of the contours in Plate XLV of this report
will show that the shelf is definitely continued through the eastern end of the Burdwood Bank.

(ii) In the bathymetric chart (Fig. 36) given by Stocks, a long ridge, parallel to the main arc, is shown as
extending from a point north of South Georgia to approximately 300 miles east of the South Sandwich
Islands. The depth of water on the ridge is shown as under 3000 m. The evidence for the existence of this
ridge appears to be very slender, for apart from the bank shown to the north of Cape Crewe, South Georgia,
only two other soundings are mentioned. One of these is a Lucas depth of 2744 m. obtained by the R.R.S.
'Discovery ' in February' 1926, in a position 51° 55' S, 32° 27!' W, the other is an echo sounding of 1800 m.
obtained by the ' Meteor ' in the same month to the east of the South Sandwich Deep. From our bathymetric
plate (Plate XLV) it will be seen that subsequent soundings leave no doubt as to the non-existence of this ridge.

(iii) North of the South Orkney Islands Stocks has shown another ridge, parallel to the main one, the
existence of which is not supported by the recent lines of echo soundings obtained by the 'Discovery II'.

1 Bathymetric map of the Antarctic (South Atlantic, Indian and Pacific Oceans) compiled by the Amencan
Geographical Society of New York, 1929.

220 DISCOVERY REPORTS

These depths were obtained on a zig-zag Hne of soundings run in March 193 1 from the
eastern end of the Burdwood Bank to the Shag Rocks. During this traverse fifty-three
soundings of under 2000 m. were obtained, and of these, four were under 500 m.,
nineteen under 1000 m., and forty under 1500 m. A subsequent Hne of echo soundings
during the season 193 1-2, running from south to north across this Hne in about 54° W,
gave six more soundings under 2000 m., one of which was under 1500 m. This hne
crossed the earHer Hne in about 54° S, where the soundings were in close agreement,
and continued north to 53° S. From here the Hne was run in a north-easterly direction
to approximately 51" 30' S, 49° 00' W, where the course was altered to south.

Fig. I. Key chart showing the areas covered by Charts 1-7, and positions
of sections in Figs. 2-4.

This southerly traverse along the 49th meridian crossed two of our earlier lines of
soundings, the first in about 53° 20' S, and the second in about 54° 30' S. Unfortunately
about 10 miles north of the former line the oceanic echo-sounding machine broke down
and soundings were not resumed for about 30 miles, so that the confirmation of the
depths recorded in the previous year was not possible. On the second line, however, the
agreement was good.

Fig. 3 b (p. 226) shows a section across the Scotia Sea which is based on this line of
soundings. From it can be seen very clearly both the Tierra del Fuego — Shag Rocks
ridge and also the connexion, double at this point, between the South Orkneys and
Clarence Island.

SOUTH GEORGIA AND THE SHAG ROCKS 221

From 47° W to the Shag Rocks many more soundings are necessary before the
contours can be drawn in accurately, but from the soundings now available it will be
seen that there can be no doubt as to the existence of a ridge.

Previous bathymetric charts of this area have shown the Rhine Bank in a position
approximately 55° 30' S, 53"^ 30' W with a depth of 121 m. In January' 1926 the 'Meteor'
searched unsuccessfully for the Bank,^ obtaining a sounding of 3640 m. very close to the
charted position, and in the course of the season 1931-2 the 'Discovery II' passed
between the Rhine Bank and the Burdwood Bank, obtaining soundings of 2500 3000 m.
within 25 miles of the former. In view of this recent evidence the Rhine Bank has been
omitted both from our sounding and bathymetric charts.

SOUTH GEORGIA AND THE SHAG ROCKS
(Plates XLV, XLVI, Charts 3, 5)

The chart of South Georgia and the Shag Rocks (No. 5) shows the majority of the
soundings obtained during the Discovery investigations since February 1925. They
number approximately 1600 and have mostly been taken by the echo method. Certain
inshore soundings taken in the bays and harbours during Lt.-Cmdr. Chaplin's survey^
have been omitted, as the scale of the chart is too small to insert more than a very small
proportion. The bathymetric chart (Plate XLVI) is based mainly on the 1600 soundings,
although account has been taken of all other available soundings, especially those of the
'Quest' and 'Meteor'. These and other soundings not obtained during the Discovery
investigations are shown on the bathymetric charts by circles as opposed to the black
dots which represent our soundings.

Nearly all these soundings round South Georgia and two of the lines between Willis
Island and the Shag Rocks were obtained during the various plankton surveys. During
these surveys lines of stations are worked out to a distance of 1 00- 1 20 miles from the land ,
starting from various points round the island. By coming back diagonally from the outer
end of one line to the shore end of the next and by changing the starting-points and the
angles of the lines with the land, it has been possible to cover a considerable amount of
ground. It was also usual when proceeding to and from South Georgia to set a slightly
different course on each occasion, which accounts for the greater number of soundings
obtained on the northern side of the island.

As a result of the survey work carried out between 1926 and 1930 by Lt.-Cmdr.
Chaplin, R.N., and the running surveys made by Lieut. Nelson, R.N.R., in 193 1, con-
siderable alterations have been made in the charted coast-line of the island, and as a result
of this work it has been necessary to omit as unreliable certain soundings which were fixed
by land bearings. With regard to the bathymetric chart, soundings from other sources
which are known to have been fixed from the land have been adjusted to the new outline
of the land as far as possible and doubtful ones have again been omitted.

^ Whs. Ergebn.Deutsdi.Atlant.Exped. 'Meteor ', I, p. 17s ■

2 Lt -Cmdr Chaplin, R.N. Narrative of Hydrographic Survey Operations in South Georgia and the
South Shetland Islands,' 1926-3°. Discovery Reports, iii, pp. 297-344. Pls- XL-XLIV, Charts 1-4 (1932)-

222 DISCOVERY REPORTS

From the bathymetricchart of South Georgia and the Shag Rocks (PlateXLVI)itwill be
seen that the coastal shelf of South Georgia has a considerable extent, the average distance
of the 500 m. contour from the coast being about 25 miles on the south-west side of the
island. This distance varies considerably, there being one point south-west of Cape
Disappointment where the 500 m. contour is only about 16 miles from the land and
another west-south-west of Annenkov Island where the shelf runs out for nearly
60 miles, and then falls away rapidly, the soundings increasing from 500 to 3000 m. in
about 15 miles. This sharp drop from the coastal shelf to oceanic depths appears to be
a characteristic feature of South Georgia, and, in conjunction with the numerous
soundings obtained up to the present, can be a valuable aid to navigation in thick weather
for any ship fitted for deep-water echo sounding.

It has already been stated in the general discussion on the Scotia Arc that in view of
its geological structure. South Georgia appears to be one of the weakest links in the
chain of islands and submarine ridges constituting the arc, and that some have
maintained that an arcuate connexion lying north or south of the island is more
probable. The numerous echo soundings round South Georgia do not bear out this
theory. On the northern side of the island a small bank has been found about 80 miles
north of Cape Crewe which has an average depth of water of about 1800 m., but it is
only a few square miles in extent and existing soundings do not show any trace of it
to the east or west. On the southern side of the island the concentration of soundings
is not so great, but there is sufficient evidence to prove that no submarine ridge parallel
to South Georgia exists on this side of the island.

Influenced, probably, by this theory, those who have drawn bathymetric charts or
maps of the Scotia Sea and surrounding areas have hitherto shown the Shag Rocks,
which lie about 130 miles east of Willis Island, as an isolated group apparently rising
quite sharply from the ocean bed. The depth of water usually shown between the Shag
Rocks and South Georgia is 3000-4000 m. In all except Pratje's contour map, where the
soundings on which the contours are based are not shown, there appears to be only one
sounding to support the theory of an isolated group. This is a sounding of 3380 m.
obtained by the ' Antarctic ' in a position slightly north of the direct line between Willis
Island and the Shag Rocks. With regard to the geological evidence there is no record of
any specimens having been obtained from the Shag Rocks, we therefore have to rely
entirely on soundings to prove the connexion with South Georgia.

In April 1929 the ' William Scoresby ' obtained a Lucas sounding of 232 m. on a rock
bottom at St. WS 427, almost exactly half-way between Willis Island and the Shag
Rocks, and in March 1930 at St. WS 525 and WS 526, two more wire soundings of
159 and 1545 m. respectively on approximately the same line. These soundings indi-
cated the probability of a connecting ridge, and between December 1930 and April 193 1
the 'Discovery II' was able to obtain four lines of echo soundings in the area. About
320 soundings were taken and of these about 50 were in the vicinity of the Shag
Rocks and a small outlying rock about 10 miles to the south-west of the main group.

The four lines of soundings have definitely established the existence of a connecting

SOUTH GEORGIA AND THE SHAG ROCKS

223

METRES
200D

300Q--

4000

METRES

1000-

2000

3000--

4000

--4000

5000

1 II I I I I I I I I I M I 1 I I I I I I I I I I I I I I I I M

Fig. 2.

a. Section from north-east to south-west between South Georgia and the South Sandwich Group :

based on 18 soundings made between 55" 05' S, 32° 26' W and 56° 01' S, 34° 13' W.

h. Section from north to south between the Shag Rocks and South Georgia : based on 8 soundings made

between 53^ 13' S, 39° 53' W and 54° 10' S, 40° 02' W.

c. Section from north to south between the South Orkneys and Clarence Island : based on 43 soundings

made between 59° 49' S, 49° 24' W and 61° ooJ' S, 49° 48' W.

Vertical scale of all three sections exaggerated 24 times.

3-2

224

DISCOVERY REPORTS

ridge with a maximum depth of water of under 2000 m . and an average depth of approxi-
mately 1300 m. Two further Hnes of echo soundings were taken to the north of the ridge
by the 'Discovery II' during the season 193 1-2, and these have been of considerable
value in determining the 4000 m. contour. This contour is of great importance in this
area, as it marks the return, on the northern side of the ridge, to the average depth of
this part of the South Atlantic.

The depth of water on the ridge gradually increases from South Georgia towards the
Shag Rocks till it reaches a maximum of 1650 m. about 55 miles from Willis Island. The
depth for the next 35 miles shows a certain amount of variation from the average, a
sounding of 997 m . occurring at a point a few miles due south of the ' William Scoresby 's '
original depth of 232 m. Farther on the bottom rises sharply from 1236 to 459 m. in less
than 4 miles, and for the remaining distance to the Shag Rocks the average depth of
water is about 150 m.

The shelf surrounding the Shag Rocks appears to extend for a considerable distance.
Thirty miles to the south the soundings at present available show a depth of iioo m.,
and in a northerly direction the depth only reaches 2000 m. some 35 miles from the
Rocks.

SOUTH GEORGIA TO THE SOUTH SANDWICH ISLANDS
(Plates XLV, XLVI, Charts 3, 4, 5)

The connexion between South Georgia and the Gierke Rocks, which lie about 40
miles east-south-east of the south-eastern point of South Georgia, is quite definitely
shown by the existing soundings, the average depth of water between the Rocks and
South Georgia being from 100 to 130 m. (Chart 5). According to Holtedahl,i the Gierke
Rocks are of considerable geological interest, as one of the specimens which he obtained
from there in 1928, and which is reported to occur as veins in the granite, is said to
represent a basic dyke. A detailed examination of this specimen has yet to be made. In
conclusion, Holtedahl says of the Gierke Rocks that "the find of an originally deep-
seated rock, cut by dykes, not very far from the intrusive masses of South Georgia, and
in the direction towards the volcanic ridge of the South Sandwich Islands is of consider-
able interest".

From the Gierke Rocks to the South Sandwich Islands, a distance of about 280 miles,
the connecting ridge is not very clearly defined at the moment. There is, however,
sufficient evidence from the soundings to show that a connexion can be traced. Fig. 2 a,
which is based on a line of echo soundings from north to south across the ridge, and
Fig. 4 a, which is a section crossing the ridge farther to the east and at a difi"erent angle,
both show that there is a gradual rise from the bottom of the Scotia Sea basin, followed
by a sharp drop into the north-western end of the South Sandwich Deep. The rise on
the southern side varies between 400 and 700 m. and the depth of water on the ridge
averages about 2700 m. The minimum sounding found was 1935 m. in approximately

' Holtedahl, loc. cif., supra, p. 102.

SOUTH SANDWICH ISLANDS TO SOUTH ORKNEY ISLANDS 225

55° 28' S, 31° 47' W. On the northern side the 4000 m. contour clearly shows that the
influence of the South Sandwich Deep can be traced to a position about 40 miles north-
north-east of the Gierke Rocks. Fig. 2 a shows a section taken across the ridge into this
part of the Deep and Fig. 4 a shows a section farther to the east, crossing the Deep at
an oblique angle.

From the trend of the looo and 2000 m. contours to the south-east of the Gierke Rocks
and to the north-west of Zavodovski Island it is probable that soundings taken along a
line about 10 or 15 miles south of the existing intensive lines of soundings will show that
the crest of the connecting ridge lies to the south of where it is now placed on the
sections and bathymetric chart. The data now available show the possibility of two
gaps in about 29° 30' W and 31° W.

The 'Meteor' ran a line of echo soundings across this ridge during February 1926,
starting from about 56° 30' S, 31° 00' W and proceeding in a north-easterly direction.
From these soundings evidence was obtained that there was a double ridge at this point,
the separate peaks having approximately the same depth of water of 2400 m.' This depth
is slightly less than the average found by the Discovery Investigations, but without more
detailed information as to the position of these soundings they cannot be compared with
our interpretation of the bottom relief. It seems probable, however, that they lie to the
south of our lines. Pratje, in his bathymetric map, does not show any indication of this
shoaling between South Georgia and the South Sandwich Islands, but his contours, like
ours, are drawn at 1000 m. intervals. With the data at present available it appears that
no intermediate contour will show the existence of a continuous connexion, and that
in this section of the arc the depth of water on the ridge may be only 400-700 m. less
than the average of that at the eastern end of the Scotia Sea.

SOUTH SANDWICH ISLANDS TO SOUTH ORKNEY ISLANDS

(Plate XLV, Gharts 3, 4)

The ridge connecting the various islands of the South Sandwich Group has already
been discussed during the description of the South Sandwich Deep, so that the next
section of the arc to be considered is from Southern Thule, in the above group, to the
South Orkney Islands. Here we have been unfortunate each season with regard to ice
conditions, and it was not till early in 1932 that the 'Discovery 11' was able to obtain
a line of echo soundings which crossed the ridge from south to north in about 36° W.
This line is within 40 miles of the line of soundings taken by the ' Deutschland ' in 1912,-
and examination of the contours will show that these observations have been confirmed.
The minimum sounding found by the 'Discovery 11' was 543m. in approximately
60" 20' S, 36° 03' W, and the section based on this line (Fig. 4 h) shows that the main
ridge is roughly 60 miles wide at this point with an average depth of water of about
1000 m. To the south a narrow trench over 4000 m. deep is found, followed by a

1 Whs. Ergebn. Deutsch. Atlant. Exped. 'Meteor', I, p. 183. See also Stocks, loc. cit., supra.

2 Brennecke, Arch. Deulsch. Seewaric, xxxix, Nr. i, tafel 2 (1921).

o

rri

o

o

o

u-1

lO

0

»

n

C/J

-o

n

•a

a

T3

on

c

t-t

n

o

-n

CN

c

-i-

n

aj

HH

C

00

N

«

O)

lO

c

C

c

■■M

^,

ro

-O

4-1

T-l

M
C

4)

G

o

fe i^

OT

ON

N

(I)

r^

a

§

o

-a

-T3

J3

u

>

ra

c/3

CO

CT]

Ph

o

«J

u

^

r/)

«

Q

-fi J3

3
O

-5 ^

o
c

c

e

o

n

o

c>

I)

1)

C/J

C«

SOUTH ORKNEY ISLANDS TO CLARENCE ISLAND 227

smaller but well-marked ridge where the soundings average about 3000 m. This in turn
is succeeded by a deeper trench with soundings over 5600 m. Comparison with the
Deutschland soundings shows that they also found a double ridge, their minimum
soundings being 1050 and 1012 m., with a deep trench of over 5000 m., lying to the
south. These deep soundings following a ridge show a great similarity to the conditions
prevailing in the neighbourhood of the South Sandwich Islands, and it seems possible
that the Deep extends round outside this sector of the arc.

About half-way between the Discover^' II line of soundings and Southern Thule the
British Admiralty chart^ shows three soundings under 2000 m. Two of these, 1079 and
1 158 m., were obtained by the ' Scotia ' between 1902 and 1904, the other of 1862 m. was
taken by the 'Deutschland' on her journey south to the Weddell Sea in 191 1. These
depths compare very closely with those taken to the westward, and may be regarded
as evidence in favour of a connecting ridge.

Between the line of Deutschland soundings in 37° W and the South Orkneys there
are also only a few soundings. These were taken by the ' Scotia' between 1902 and 1904
and show that the 500 m. contour extends to a point 75 miles east of the islands. The
extension of the coastal shelf so far to the east is strong evidence in favour of a Southern
Thule — South Orkney ridge. It is hoped to obtain more soundings in this area in
the near future.

SOUTH ORKNEY ISLANDS TO CLARENCE ISLAND
(Plate XLV, Chart 2)

Although geologically the South Orkneys do not appear to support the theory of an
arc owing to the apparent lack of any true Andean type of rock, it has been shown by
soundings that to the east there is a strong probability of a connexion between them and
the South Sandwich Group. On the western side of the South Orkneys the probable
existence of a submarine ridge was brought to our notice by two soundings obtained in
the 'Discovery' while working St. 169 in February 1927. Here a Lucas sounding of
2514 m. was obtained in a position 60° 48' 50" S, 51° 00' 20" W. After the station a deep
net was towed, but it fouled the bottom after a few miles and a second sounding gave
a depth of 1849 m. in a position approximately 5 miles to the west of the other. This
station was worked alongside a very large tabular berg, which, from our observations,
appeared to be aground.

From this date until the season 1930-1 no more soundings were obtained from this
area, but in this season the complete absence of pack-ice from these waters enabled us to
obtain numerous echo soundings. These have been considerably amplified during the
season 1931-2, and there can be no doubt that a submarine ridge with a depth of water
of under 2000 m . connects the South Orkneys and Clarence Island . The average sounding
is from 1600 to 1700 m., and it appears fairly certain that a double ridge is to be found
over practically the whole distance. The secondary' ridge, which lies to the north of the

1 British Admiralty Chart, 2202 b.

IT}

O

CO

c

o
o

cn

c

<u

c

3
O
c/l

ON

N

-d
1)

k "

3
o

-d
c

rsi

c«

c

■*-»

o

en

a
c

3
O

C
O

c

3
o

en

o

o

Q

^

3

rn

is
-d

j::

C

-1

en

o

en

3

C3 CO

en -^

. c

o "^

o ,-

u

ii O

lO

en -a

c ii

3 u

=« ss

tfl (^

m .5

JS ^

CO ^^

I— I rt

-d t

< ^

» -d

>. 3

o

§ 2

3
o
en

a,

' — tfi

(U CO

O I ^

o J3 a

en '^ u

■ 13 u

N

ca

-d P ^

iJ j=

o

en

3

'c

C

en

« N ~0 lo cj

SOUTH SHETLAND ISLANDS AND ADJACENT WATERS 229

main connecting ridge, is the more pronounced, having a depth of water in one place of
only 487 m. (60° 11' S, 50" 28' Wapprox.) and in another of 501 m. (60''-' 15' S, 51' 50' W
approx.). The main ridge is quite well defined, but the least depths over the correspon-
ding portions are 1501 m. in 60° 25' S, 48"^ 43' W, and 1523 m. in 60'' 37' S, 49 "" 40' W
(both positions approximate). Fig. 2 c, which is based on a line of echo soundings run
in December 1930 from north to south across the ridge in approximately 49° 40' W,
shows the double ridge very clearly. Fig. 3 b, which is based on a line of soundings across
the Scotia Sea in 1932, shows a section across the ridge in approximately 48 ' 20' W,
where the formation appears to be the same. This double ridge fades away in about
51° 40' W, from which point a broad shallow ridge continues to Clarence Island. The
depth of water between the ridges is about 3000 m., but at one point there is an abrupt
descent, apparently of about 45°, to 5000 m. at which depth a trench also occurs between
the eastern end of the secondary ridge and the South Orkneys.

Hydrological observations made in 1930 on each side of the supposed line of the ridge
gave valuable confirmation of the theory of a connexion. It was found that entirely
dissimilar conditions, both of temperature and salinity, existed on the two sides, and
that these conditions were such as could only have been caused by a submarine ridge of
considerable height interrupting the free interchange of water between the Weddell Sea
and Scotia Sea.

This section of the arc from the South Orkneys to Clarence Island may be termed the
last oceanic sector, as over the remaining distance from Elephant Island to the South
Shetlands comparatively shallow depths are to be found. These will be considered in the
remarks on the bottom relief of the Bransfield Strait and the waters adjacent to the
South Shetlands.

SOUNDINGS IN OTHER LOCALITIES

SOUTH SHETLAND ISLANDS AND ADJACENT WATERS

(Plate XLVII, Chart 6)

The Bransfield Strait, which lies between the South Shetlands and Trinity Peninsula,
and the area east of it as far as Elephant and Clarence Islands, have been extensively
sounded out. The total number of soundings on which our bathymetric chart (Plate
XLVII) is based is about 1500, of which number approximately 1400 were taken in the
' Discovery II ' by the echo method. The remainder consist of a line of echo soundings
taken by the ' Meteor' during her visit to Deception Island in January 1926,^ and a few

^ These soundings appear to have been calculated from a constant speed of sound in sea water for all
depths, i.e. 1490 m. per sec. {Wiss. Ergebn. Deutsch. Atlant. Exped. 'Meteor', i, Beilage XII), and by
this method, if the Admiralty tables of the speed of sound in sea water for this area are accepted, soundings
above 5000 m. will be too deep and those below this figure too shallow. In the bathymetric chart no attempt
has been made to reduce these soundings to the speeds given in the Admiralty tables, as the differences are
not sufficient to have any appreciable effect on the contours concerned.

230 DISCOVERY REPORTS

Lucas soundings by the 'Discovery', 'William Scoresby', 'Pourquoi PasP'^ and
' Antarctic ' .- The positions of soundings taken by the vessels of the Discovery Committee
are shown in black — those of other vessels as circles. The combined results give a good
general idea of the bottom relief of this area.

When considering the contours of the bathymetric chart (Plate XLVII) in detail, it will
be seen that there are two separate depressions, a small one with a maximum depth of
1371 m. lying south of Snow Island, and the main depression which stretches from
Deception Island almost to Clarence Island. The eastern end of the latter has a consider-
able area with a depth of over 2000 m., the maximum found by the ' Discovery II ' being
2879 m., but the western end is on the whole much shallower. Here there is only one
small area over 2000 m., with a maximum of 2085 m. found by the ' William Scoresby'
in February 1 929. The remainder of the depression has an average depth of 1 500-1 600 m.

The slope on the northern side of the main depression is fairly steep along the whole
length, but on the northern side of the South Shetlands and the Shetlands — Elephant
Island ridge the bottom falls away quite gently. This may be compared with the South
Sandwich Islands ridge and the Deep beyond it, which has already been described, though
the drop from the South Shetlands to the Bransfield depression is not so sharp and the
depression not nearly so deep. Geologically, however, it would appear that both may
have been formed in the same manner. Three lines of our echo soundings and one of the
'Meteor's' place the average distance of the 1000 in. contour about 40 miles from the
centre line of the South Shetlands on the northern side and 10 miles on the southern.

More soundings are required in the depression in the neighbourhood of Bridgman
Island in order to determine the position of the 1000 m. contour. Bridgman Island is
certainly volcanic and is regarded by Holtedahl ^ as " the remains of a much larger volcanic
mass which has been to a large extent planed down by marine abrasion". It seems
probable, therefore, that it was once part of the South Shetlands and that there is
comparatively shallow water between it and King George Island. There is only one
sounding in the vicinity, 620 m., taken by the ' Pourquoi Pas? ', and on this and on the
geological evidence we think it probable that the 1000 m. contour lies to the south of the
island. The position of Bridgman Island appears to be rather similar to that of Deception
Island, between which and Livingston Island a submarine plateau exists.

Unfortunately there are only two lines of echo soundings running north and south
between King George and Elephant Islands, but these, together with a Lucas sounding
midway between them, show the probable existence of a ridge. Ferguson, who was
unable to land on Elephant Island, has suggested* that some of the darker stratified rocks
may be bedded lavas. This would appear to be an error, and Wordie,^ who collected

' Deux. Exped. Fran. (1908-10), Cartes, No. xi (1912).
" Swed. Sild-Pol. Exped. (1901-3), i, Lief 2, Karte (1917).
^ Holtedahl, loc. cit., supra, pp. 97-8.
* Ferguson, Trans. Roy. Soc. Edin., Liii, p. 35 (1925).

'' Wordie, Shackleton Antarctic Expedition, 1914-17, Geological Observations in the Weddell Sea Area,
Trans. Roy. Soc. Edin., liii, p. 27 (1925).

SOUTH SHETLAND ISLANDS AND ADJACENT WATERS 231

specimens there after the loss of the ' Endurance ', has decided that the rocks, which were
critically examined by Tyrrell, are metamorphic and probably altered volcanic ashes, but
too metamorphosed to be properly identified. It would thus appear that although
the soundings indicate the presence of a ridge, the structure of these islands themselves
cannot be used as evidence of a former land connexion.

Hydrological observations tend to show that the oceanic warm deep-water layer in
Drake Passage, in which the maximum temperature is found at about 500-600 m., occurs
in the Bransfield Strait with the maximum temperature at an average depth of 300 m.
This difference of levels inside and outside the strait is accounted for by the rise of the
ocean bed towards the South Shetlands which forces the south-flowing warm deep water
upwards to a level where it can enter the strait. It is almost certain that it enters between
Smith and Snow Islands, where the soundings show a channel with an average depth of
600-700 m. It then passes across the smaller basin in a south-easterly direction and
enters the larger basin to the south of Deception Island. On the evidence of the present
soundings there appears to be a ridge between Deception and Tower Islands, but the
position of the 500 m. contour is a little uncertain, as it could also be drawn to run east
and west showing a channel with a depth of 600 m. and over. Such a channel would
allow the passage of the warm deep water. The characteristics of this layer gradually
grow weaker towards the eastern end of the main basin, disappearing altogether south
of Elephant Island. Hydrological observations in this area show that the chief water
movement is an inflow of Antarctic surface water from the Weddell Sea.

In the vicinity of Joinville Island and to the north-east of it, the only soundings
available are those taken by the 'Pourquoi Pas?' in 1908-10 and the 'Antarctic' in
1901-3. These, however, in conjunction with our zigzag Hne of echo soundings south
of Clarence Island show that Trinity Peninsula is continued as a shallow submarine
plateau towards the north-east and eventually joins up with the main Orkneys— Shet-
lands ridge. Here large icebergs from the Weddell Sea are frequently found stranded. It
is hoped to obtain more soundings south of Clarence Island to clear up some doubtful
points, one of which is the possible existence of a channel running east and west and
shown at the moment only by a very few soundings. The hydrological evidence here is
more against than for the existence of such a channel.

While the 'Discovery H' was on passage from Port Lockroy, Wiencke Island, to
Deception Island by way of the De Gerlache Strait, continuous echo soundings were
taken which showed that up the centre of the strait there was a channel with an average
depth of between 600 and 700 m. Near Cape Murray, at the eastern end of the Strait,
the soundings exceeded 1000 m. for several miles. Nine soundings of over 1000 m. were
obtained, the maximum being 1141 m. On coming out of the strait we continued up
towards Trinity Island, intending to pass through Orleans Channel, but owing to the
presence of pack ice we were compelled to turn north before entering it. On this north-
ward passage we discovered the existence of a reef to the west of Farewell Rock, Trinity
Island. We passed between this reef and the island, the soundings shoaling gradually
to 66 m. and falling away again quite sharply on the northern side.

4-2

232 DISCOVERY REPORTS

The lines of echo soundings taken by the ' Discovery II ' round the South Shetlands
and in the Bransfield Strait were all run between December 19, 1930 and March 10,
1931. Altogether 12 days were occupied in taking 1432 echo soundings, which is a
considerable advance on the number of wire soundings which could be taken in the
same period. Only on one line, from Deception Island to the South Orkneys via Clarence
Island, was it necessary to abandon echo soundings on account of weather conditions.
Here we were forced to give up sounding over a distance of 70 miles by a north-west gale
and a very lumpy sea, which made it impossible to manoeuvre the ship without causing
excessive water noises in the hydrophone.

BELLINGSHAUSEN SEA
(Chart 7)

Until the season 1 930-1 very few soundings had been recorded from the Bellings-
hausen Sea. Late in December 1930, the 'Discovery 11' left Deception Island in an
attempt to push as far south and west as possible. As will be seen from the sounding
chart (No. 7) the trend of Graham Land was followed as far down as Adelaide Island.
Here very heavy pack-ice was encountered and the ship had to strike north again for some
considerable distance before further progress to the west was possible. From this time
until the return to Adelaide Island some 12 days later the ship was either in pack-ice or
skirting the edge of it. Several attempts were made to get south, but owing to the very
heavy ice year it was not possible to get beyond 69° 49' S in approximately 101° W. This
was reached on January 6, 1931, and the echo sounding was 4131 m., about 100 m. less
than the average depth for that part of the Bellingshausen Sea. It was in this vicinity
that members of Cook's expedition were deceived by an appearance of land in January
1774. Forster, writing on Cook's voyage in the 'Resolution', says "we were amused
with the appearance of land, for after standing on towards it for some hours, it vanished
in clouds". Cook's position was then about 68° S and he stood on to 71° 10' S without
sighting land. In February 1930, the 'William Scoresby' obtained Lucas soundings of
574, 584 and 580 m., with a bottom of yellow mud farther to the east. These positions
are about 120 miles west of Charcot Island, which is the nearest known land.

On the return passage of the ' Discovery II ' from 101° W, it was found that although
some of the pack had moved away to the north-east it was still not possible to reach
Peter 1st Island, and we passed about 25 miles to the north of it. The soundings here
showed little change from the average for this part, that taken nearest the island being
3981 m.

The average depth of the ocean bed in the Bellingshausen Sea varies very little. The
general information gained from our soundings being that the depth lies approximately
between 3900 and 4400 m., the maximum found by the ' Discovery II ' being 4598 m. in
about 68° 05' S, 98° 20' W. In the area over which our soundings were taken, the greater
depths lie to the north and west of Peter 1st Island.

On our return we were able to reach Adelaide Island and three hues of soundings

BELLINGSHAUSEN SEA 233

were taken over the edge of the continental shelf between Alexander 1st Island and the
Biscoe Islands. These soundings show that the shelf extends some 80 miles from the
former and about 65-70 miles from the latter. The great breadth of the continental shelf
may presumably be taken as evidence of the antiquity of the land masses which lie to the
south. From the section (Fig. 4 c), which is based on a line of 32 soundings run on a
north-west bearing from Mt Gaudry, Adelaide Island, it will be seen that the edge of the
continental shelf is very steep, the depth increasing from 500 to 3000 m. in 15 miles.
A very similar slope occurs on the other two lines.

On these lines regular soundings over 4000 m. were not obtained until we were some
200-250 miles from the land, and it seems probable from this evidence and from the
extent of the continental shelf that if the same conditions prevail farther to the west, the
land must lie a considerable distance to the south of the highest latitudes reached by the
'William Scoresby' and 'Discovery II'.

Approximately 700 echo soundings were taken during the time spent in the Bellings-
hausen Sea and on the return trip to Port Lockroy, via the Biscoe Islands and the Pendle-
ton Strait. Only about 650 of these soundings are inserted on the chart, as the scale is
not sufficiently large to show those taken for navigational purposes when approaching
and leaving the Marin Darbel Islands. The majority of the soundings were taken with
the deep-water echo machine, and at times with considerable difficulty owing to the
noise made by the pack-ice on the hull, which completely drowned the echo in the
hydrophone. Another effect of pack-ice which we had anticipated did not, however,
occur. It had been thought that the close proximity of pack-ice might cause both the
hammer blow and the echo to be reflected from the undersides of the large floes, giving
in the first case a false echo and in the other a re-echo of the true depth. We have also
tried the effect of taking echo soundings in the vicinity of large icebergs, especially those
which would have a draft of several hundred feet, as the under-water portion might be
expected to reflect the sound waves from the hammer and cause false echoes. No
interference was ever observed, and it seems probable that the water chamber under the
hammer is effective in preventing the sound waves from scattering. On the other hand,
re-echoes have been obtained on several occasions, but only in waters completely free
from ice. One of these was recorded from the Orkney — Clarence Island ridge, where a
faint but quite distinguishable re-echo was picked up at exactly double the scale reading
of a sounding of 1344 m. These re-echoes were probably obtained on a clean rock
bottom, and to this they may be due.

OTHER SOUNDING WORK

During the course of our work in the south, opportunity has always been taken, as
conditions permitted, to search for various rocks or vigias which have been reported from
time to time from the South Atlantic and Southern Oceans. Most of them have been
reported, on single occasions only, as isolated rocks in mid-ocean, and we have not been
able to find any of those for which we have searched despite care and extensive echo

234

DISCOVERY REPORTS

sounding. It seems most likely in the light of our experience that small bergs of " black "
or "bottle-green" ice have been mistaken for rocks, for these bergs, which are not un-
common off the mouth of the Weddell Sea, are very confusing, especially in thick weather
or a bad light. Several of those sighted from the " Discovery II ' might easily be mis-
taken for a small rock awash, had it not been for the knowledge that the echo soundings
at that moment were well over 3000 m.

The most interesting search we carried out was for Thompson Island which was
supposed to lie to the north-eastward of Bouvet Island. This island was first reported

East Lorijjitude

Fig- 5-
Area covered in the search for Thompson Island and the Chimneys. Black dots are points at which ship's
position was accurately fixed. Shaded area represents the limit of visibility from the crow's nest. The figures
show soundings in metres.

by Captain Norris of the whaler ' Sprightly' in 1825 as being approximately 45 miles
north-east of Bouvet Island. Some rocks, called the Chimneys, were reported at the
same time to lie about 4I miles to the south-eastward of Thompson Island. From 1825
until 1893 no one appears to have sighted the island, but in the latter year Captain
Fuller of the ' Francis Allen ' is stated to have sighted it but to have been unable to fix its
position on account of wind, ice and fog. Since then it has been unsuccessfully searched

OTHER SOUNDING WORK 235

for by the 'Valdivia' in 1898, the 'Meteor' in 1926 and the 'Norvegia' in 1928, the
'Meteor' obtaining a sounding of 1580 m. on one of its supposed positions. The
'Norvegia', as we learn from Bjarne Aagard.i spent a number of days in searching for
the island both to the east and west of Bouvet Island, and she appears to have covered
a considerable part of the area which we ourselves examined: this, however, was not
known to us at the time. It is to be hoped that a chart showing the track of the 'Nor-
vegia', and the soundings obtained, will soon be published.

When the 'Discovery 11' searched unsuccessfully for the island in October 1930, we
were largely guided as to the area to be searched by the latest information supplied
by the Hydrographic Department of the Admiralty, which gave particulars of all
previous searches and the area which might still be regarded as unexamined. The
' Discovery II ' was very fortunate in having fine clear weather which made it possible
to fix the ship's position at frequent intervals, but a rough sea made echo sounding
difficult and only twenty-five soundings were obtained. The area examined was over
3000 square miles, as can be seen from Fig. 5. This figure also shows the limit of visibility
from the crow's nest, and the results indicate that although a definite plateau exists to
the north-eastward of Bouvet Island, there is no indication in this area of either Thomp-
son Island or the Chimneys. The soundings obtained near the charted position of the
island by the ' Discovery II ' were 1538 and 1629 m., which agree very closely with the
'Meteor's' sounding of 1580 m. As a resuh of the latest search these islands have been
removed from the British Admiralty Charts and Sailing Directions (see Admiralty
Notice to Mariners No. 406 of 193 1).

The following instance is an example of the means by which dangers and shoals come
to be reported without careful examination : When the ' Discovery II ' was in the neigh-
bourhood of the South Sandwich Deep in November 1930, a shoal of approximately
200 fathoms was reported by a whaling factory ship from a position where we had found
considerably deeper water. On asking for further details we were informed that this
depth had been obtained by one of their catchers after harpooning a whale, as the line
had run out vertically downwards for 200 fathoms and then had suddenly become
slack.

' Bjarne Aagard, Fangst og Forskning i Sydishavet, pp. 626-37, Oslo (1930).

236

DISCOVERY REPORTS

Echo-soiinding table compiled from observations made at St. 471 in the

South Sandwich Deep

Half-averag

e sounding

Range of deep-

velocity per second

Depth 1

water machine in
seconds

Fathoms

Metres

Fathoms

Metres

0-0-277

393-95

720-45

0-109

0-200

0-277-0-415

394-37

721-25

109-164

200-300

0-415-0-553

395-25

722-35

164-219

300-400

0-553-0-689

395-95

724-12

219-273

400-500

0-689-0-827

396-47

725-10

273-328

500-600

0-827-0-965

396-87

725-80

328-383

600-700

0-965-I-099

397-25

726-47

383-437

700-800

1-099-1-237

397-65

727-35

437-492

800-900

I-237-I-374

397-95

727-90

492-547

900-1000

I-374-I-7I3

398-42

728-63

547-683

1000-1250

I-7I3-2-053

399-07

729-80

683-820

1 250-1 500

2-053-2-393

399-70

730-95

820-957

1500-1750

2-393-2-731

400-30

731-80

957-1094

1750-2000

2-731-3-066

400-87

733-12

1094-1230

2000-2250

3-066-3-403

401-45

734-17

1230-1367

2250-2500

3-403-3-738

402-02

735-22

1367-1504

2500-2750

3-738-4-070

402-60

736-27

1 504- 1 640

2750-3000

4-070-4-404

403-20

737-32

I 640-1 777

3000-3250

4-404-4-737

403-75

738-40

1777-1914

3250-3500

4-737-5-068

404-30

739-47

1914-205 1

3500-3750

5-068-5-397

404-95

740-52

2051-2187

3750-4000

5-397-5-726

405-50

741-57

2187-2324

4000-4250

5-726-6-056

406-05

742-62

2324-2461

4250-4500

6-056-6-381

406-65

743-67

2461-2597

4500-4750

6-381-6-708

407-25

744-75

2597-2734

4750-5000

6-708-7-035

407-85

745-82

2734-2871

5000-5250

7-035-7-357

408-40

746-90

2871-3007

5250-5500

7-357-7-681

409-00

748-00

3007-3144

5500-5750

7-681-8-004

409-60

749-10

3 144-3281

5750-6000

8-004-8-324

410-20

750-20

3281-3417

6000-6250

8-324-8-645

410-80

751-30

3417-3554

6250-6500

8-645-8-964

411-42

752-42

3554-3691

6500-6750

8-964-9-282

412-07

753-60

3691-3828

6750-7000

9-282-9-586

412-65

754-65

3828-3964

7000-7250

9-586-9-917

413-20

755-67

3964-4101

7250-7500

9-917-10-234

413-80

756-75

4101-4238

7500-7750

10-234-10-472

414-40

757-85

4238-4374

7750-8000

DISCOVERY REPORTS, VOL- \

PLATE XLV

THE SCOTIA SEA.
Positions of soundings lulten during the Discovery Investigiil

DISCOVERY REPORTS, VOL. VI

PLATE XLVI

Sl.)l TH GKUKGIA AND TIIK SHAG RoCKS.
PosiUom of soundxngs fktn durins Ike Ducowry Invesligatwm are shown in black.

DISCOVERY REPORTS. VOL. VI

PLATE XLVII

BRANSFIELD STRAIT AND ADJACENT WATERS.
Positions of soundings taken during the Dticovery Investigations are shown in black.

[Discovery Reports. Vol. VI, pp. 237-392, Plates XLVIII-LVII, December, 1932 ]

SPONGES

By

MAURICE BURTON, M.Sc.

Assistant-Keeper, Department of Zoology,
British Museum (Nat. Hist.)

CONTENTS

Introduction page 239

Systematic list of species 239

List of stations, with the names of species collected at each 243

Descriptions of species 254

The affinities of the Antarctic species of Tedania, with notes on the

distribution of the genus generally 34^

The evolution of the species of loplion, and its significance in relation to

other Antarctic species of Ectyoninae 34^

Geographical distribution 35 ^

Embryonic development and breeding seasons 359

Unattached sponges 37^

The value of external form in the identification of sponges 375

The abundance of sponges in the Antarctic and its geological significance . 378

List of literature 3^0

Index 385

Plates XLVIII-LVII following page 392

T

SPONGES

By Maurice Burton, M.Sc.

Assistant-Keeper, Department of Zoology,
British Museum (Nat. Hist.)

(Plates XLVIII-LVII, text-figs. 1-56)

INTRODUCTION

HIS report is based on the study of material collected by the R.R.S. 'Discovery',
the R.S.S. 'William Scoresby' and the staff of the Marine Biological Station at
South Georgia during the years 1925-9. In many respects the collections obtained are
of unusual interest, including as they do specimens from the West African coast and
from the area around Tristan da Cunha, for our knowledge of the sponge faunas of these
two localities is as yet meagre. The rest of the material was obtained mainly from South
Georgia, South Shetlands, the Falkland Islands and the extreme southern part of the
South American continent.

The sponges in these collections are abundant and well preserved and include re-
presentatives of 168 species and varieties, of which only 35 are new: but the more im-
portant feature of the collection is the fact that many of the species are each represented
by numerous individuals. Five species of Tedania, for example, are represented
by nearly 200 specimens of all sizes. When abundant material is to hand for
comparative study, it is always found that the prevailing ideas concerning the species in
question require radical revision. It has been possible to show here that the variations
in T. massa, Ridley and Dendy, go far beyond what might a priori have been
expected. Moreover, the extremes of fluctuation in the dimensions of the spicules,
within a given species, are considerably wider apart than has hitherto been thought to
be the case. This is strikingly demonstrated in lophon chelifer, Ridley and Dendy, in
which species again considerable quantities of material are available.

In addition to the usual systematic studies, a certain amount of data is made
available from the identification of the Discovery collections concerning the develop-
ment and the breeding seasons of Antarctic species, and our knowledge of the distribu-
tion of species in the South Atlantic and Antarctic is considerably increased.

SYSTEMATIC LIST OF SPECIES

Sub-kingdom PARAZOA

Phylum NUDA

Order HEXACTINELLIDA

Family ROSSELLIDAE

Genus Rossella, Carter. R. villosa, Burton.

R. antarctica, var. intermedia, n. Genus Gymnorossella, Topsent.

R. nuda, Topsent. G. inermis, Topsent.
R. racovitzae, Topsent.

240 DISCOVERY REPORTS

Phylum GELATINOSA

Order CALCAREA
Family HOMOCOELIDAE

Genus Lencosoletiia, Bowerbank. L. discovereyi, Jenkin.

L. mackciyi (Lendenfeld).

Family LEUCASCIDAE

Genus Pericharax, PolejaefF. Genus Leucetta, Haeckel.

P. pyriformis, sp.n. L. leptoraphis (Jenkin).

L. macquariensis, Dendy.

Family LEUCALTIDAE

Genus Leucaltis, Haeckel. Genus Leucettusa, Haeckel.

L. gastrorhabdijera, sp.n. L. baeckcliana (Polejaeff).

L. sitnplicissima, sp.n.

Family GRANTHDAE

Genus Graniia, Fleming. G. cirrata, var. tenuipilosa, n.

G. cirrata, var. aurorae, Dendy.

Order TETRAXONIDA
Sub-order HOMOSCLEROPHORA

Family PLAKINIDAE

Genus Plakina, Schulze. P. trilopha, Schulze.

P. tnonolupha, Schulze.

Sub-order STREPTASTROSCLEROPHORA

Family PACHASTRELLIDAE

Genus Pachastrella, Schmidt. Genus PoecUlastra, Sollas.

P. mofiilifera, Schmidt. P. compressa (Bowerbank).

Sub-order ASTRO SCLEROPHORA
Family GEODHDAE
Genus Geodia, Lamarck. G. magellani (Sollas).

Family ERYLIDAE
Genus Erylus, Gray. E. discophorus (Schmidt).

Sub-order SIGMATO SCLEROPHORA

Family TETILLIDAE

Genus Telil/a, Schmidt. Genus Cinachyra, Sollas.

T. leptoderma (Sollas). C. antarctica (Carter).

C. barbata, Sollas.
C. coaclifera (Lendenfeld).

LIST OF SPECIES

241

Family HAPLOSCLERIDAE

Genus Haliclona, Grant.

//. nodosa (Thiele).

H. chilensis (Thiele).

//. variabilis (Thiele).

H. penicillata (Topsent).

H. gaussiana (Hentschel).

//. tiibiiloramosa (Dendy).

//. coiiica (Brondsted).

H. bilamellata, sp.n.

H. stephensi, sp.n.

H. petrosioides, sp.n.
Genus Haliclonissa, gen.n.

H. verrucosa, sp.n.

H. saccijormis, sp.n.
Genus Microxina, Topsent.

M. henedeni (Topsent).
Genus Hemigellius, gen.n.

H. rudis (Topsent).

H. pachydertna, sp.n.

Genus Adocia, Gray.

A. carduus (Ridley and Dendy).

A. glacialis (Ridley and Dendy).

A. siplionel/a (Thiele).

A. Lucurbitiformis (Kirkpatrick).

A. tremulus (Topsent).

A. tenellus (Topsent).
Genus Calyx, Vosmaer.

C. arciiarius (Topsent).

C. kergueletisis (Hentschel).

Genus Dasvchalina, Ridley and Dendy.

D. validissima (Thiele).

Genus Callyspoiigia, Duchaissang and Michelotti.

C. fort is (Ridley).

C.fusifera (Thiele).

C.flabellata, sp.n.
Genus Phloeodictyon, Carter.

P. cumitutn, Kirkpatrick.

Family DESMACIDONIDAE

Section ISODICTYEAE.

Genus Isodictya, Bowerbank.

/. kerguelensis (Ridley and Dendy).
/. kerguelensis, var. simillima (Hentschel).
/. setifer (Topsent).
/. delicata (Thiele).
/. antarctica (Kirkpatrick).
/. cactoides (Kirkpatrick).
/. microchela (Topsent).
/. erinacea (Topsent).
/. toxophila, sp.n.
Genus Guitarra, Carter.

G. fimbriata. Carter.
Genus Cercidochela, Kirkpatrick.

C. lankesteri, Kirkpatrick.
Genus Pliimocoliimella, Burton.
P. maeandrina (Kirkpatrick).
P. tnyxillioides, sp.n.
Section Mycaleae.
Genus Mycale, Gray.
M. magellanica (Ridley).
M. lapidiformis (Ridley and Dendy).
M. acerata, Kirkpatrick.
M. tridens, Hentschel.
M. macrochela, sp.n.
Genus Amphilectus, Vosmaer.
A.fucorum (Esper).

A. rugosa (Thiele).
A.flabellata, sp.n.

Genus Diemna, Gray.

B. chilensis, Thiele.

Genus Asbestopluma, Norman.

A. calyx, Hentschel.
Genus Arenochalina, Lendenfeld.

A. incrustans, sp.n.
Genus Acanthorhabdus, Burton.

A.fragilis, Burton.
Genus Sigtnotylotella, gen.n.

S. suberitoides, sp.n.
Section loPHONEAE.
Genus lophun, Gray.

/. radiatus, Topsent.

I.proximum (Ridley).
Section Tedanieae.
Genus Tedania, Gray.

T. lemiicapitata, Ridley.

T. massa, Ridley and Dendy.

T. spinata (Ridley).

T. charcoti, Topsent.

T. mtirdochi, Topsent.

T. oxeata, Topsent.
Section Myxilleae.

Genus Myxilla, Schmidt.

M. mollis, Ridley and Dendy.

M. aiislralis (Topsent).

M. asigmata, Topsent.

M. chilensis, Thiele.

M. elongata, Topsent.

M. basitnucronata, sp.n.

M. verrucosa, sp.n.

242

DISCOVERY REPORTS

Section Myxilleae [continued)
Genus Ectyodoiyx, Lundbeck.

E.paiipertas, subsp. nobile (Ridley and Dendy).

E. antarctica (Hentschel).

E.frotidosa, var. anacantha, Hentschel.

E. ramilobosa (Topsent).
Genus Anchiiioe, Gray.

A. latniiiciilioides (Ridley and Dendy).

A. areolata (Thiele).

A. leptochela (Hentschel).
Genus Steloduryx, Topsent.

S . pluridentata (Lundbeck).

S. discovereyi, sp.n.
Genus Kirkpatrickia, Topsent.

K. variolosa (Kirkpatrick).

K. coiilmani (Kirkpatrick).
Genus Acheliderma, Topsent.

A. topsenti, sp.n.
Genus Inflatella, Schmidt.

/. belli (Kirkpatrick).
Section Plocamieae.

Genus Plocamia, Schmidt.

P. gaussiana, Hentschel.
Section Clathrieae.

Genus Clathiia, Schmidt.

C. papulosa, Thiele.

C. toxipraedita, Topsent.

C. lipochela, sp.n.
Genus Rhaphidopliliis, Ehlers.

/?. paucispiculus , sp.n.
Genus Protoclathria, gen.n.

P. sitnplicissitna, sp.n.

Genus Ophlitaspongia, Bowerbank.

O. membranacea, Thiele.

O. thielei, sp.n.
Genus Artemisina, Vosmaer.

A. apollinis (Ridley and Dendy).

A. plumosa, Hentschel.
Genus Dictyociona, Topsent.

D. discreta (Thiele).

D. terrae-novae (Dendy).
Genus Axociella, Hallmann.

A. nidificata (Kirkpatrick).

A. flabellata (Topsent).
Genus Pseudanchinoe , Burton.

P. toxifera (Topsent).
Genus Eurypon, Gray.

E. mijiiaceum, Thiele.
Genus Stylotellopsis, Thiele.

S. amabilis, Thiele.
Genus Raspaxilla, Topsent.
R. phakellina, Topsent.
Section Hymedesmieae.

Genus Hymedesmia, Bowerbank.
H. irritans, Thiele.
H. laevis, Thiele.

//. simillima, var. antarctica, Hentschel.
H. longurioides , sp.n.
Section Crelleae.
Genus Crella, Gray.

C. crassa (Hentschel).

Genus Crellina, Hentschel.

C. tubifex, Hentschel.

Genus Hymeniacidon, Bowerbank.

H. cartmcula, Bowerbank.

H. sanguinea (Grant).

H.fernandezi, Thiele.

H. torqiiata, Topsent.

H. dubia, sp.n.
Genus Thieleia, gen.n.

T. rubiginosa (I'hiele).
Genus Axinella, Schmidt.

A. crinita, Thiele.
Genus Pseudaxinella, Thiele.

P. egregia (Ridley).
Genus Homaxinella, Topsent.

//. supratumesccns (Topsent).

Genus Suberites, Nardo.
S. montiniger, Carter.

Family AXINELLIDAE

Genus Rhizaxinella, Keller.

R. australiensis, Hentschel.
Genus Stylohalina, Kirk.

5. hirta (Topsent).
Genus Plicatellopsis , gen.n.

P. arborescens, sp.n.

P. flabellata, sp.n.
Genus Bubaris, Gray.

B. vermiculata (Bowerbank).

B. murrayi, Topsent.
Genus Ceratopsis, Thiele.

C. incrustans, sp.n.
Genus Eumastia, Schmidt.

E. attenuata, Topsent.

Family CLAVULIDAE

S. microstomus, var. stellatiis, Kirkpatrick.
S. papillatus, Kirkpatrick.

LIST OF STATIONS

243

Genus Pseudosiiherites, Topseiit.

P. hyulimis (Ridley and Dendy).

P. siilcatiis, Thiele.
Genus Tentorium, Vosmaer.

T. semisubeiites (Schmidt).
Genus Polymastia, Bowerbank.

P. isidis, Thiele.

P. invaginata, Kirkpatrick.
Genus Sphaerotylus, Topsent.

S. antarcticiis, Kirkpatrick.

Genus Stylocordyla, Thomson.

S. borealis, suhsp. aciiata, Kirkpatrick.
Genus Spirastrella, Schmidt.

S. purpurea (Lamarck).
Genus LatruncuUa, Bocage.

L. lendenfeldi, Hentsehel.
Genus Cliona, Grant.

C. aethiopicus, sp.n.

Order EUCERATOSA

Genus Halisarca, Dujardin.

H. dujarditn, var. ynagellaiiica, Topsent.
Genus Spongia, Linnaeus.

S. officinalis, Linnaeus.

S. magellanica, Thiele.

Genus Duseideia, Johnston.

D.fragilis (Montagu).

D. chilensis (Thiele).

D. tenuifibra, sp.n.
Genus Dendrilla, Lendenfeld.

D. membranosa (Pallas).

LIST OF STATIONS: WITH THE NAMES OF

SPECIES COLLECTED AT EACH

R.R.S. 'DISCOVERY'

St. I. i6. xi. 25. Clarence Bay, Ascension Island, 7^55' 15" S, 14° 25' W. Rectangular net,
coralline sand and shell, 16-27 ™-
Spongia officinalis, Linnaeus. Duseideia fragilis (Montagu).

St. 2. 17. ix. 25. Clarence Bay, Ascension Island, Catherine's Point and CoUyer Point. Shore
collecting.
Arenochalina incrustans, sp.n.

St. 6. i.ii. 25. Tristan da Cunha, 3 miles N 30° E of Settlement. Large dredge, rock, 80-140 m.

Leucosolenia macleayi (Lendenfeld). Sigmotylotella suheritoides, sp.n.

Leucaltis gastrorhabdifern, sp.n. lophon proximum (Ridley).

Plakina 7nonolopha, Schulze. Protoclathria simplicissima, sp.n.

Pachastrella monilifera, Schmidt. Hymedesmia irritans, Thiele.

Poecillastra compressa (Bowerbank). Hymeniacidon torquata, Topsent.

Eryhis discophorus (Schmidt). Pseudaxinella egregia (Ridley).

Halichna petrosioides, sp.n. Bubaris vermiculata (Bowerbank).

Adocia siphonella (Thiele). B. murrayi, Topsent.

Phloeodictyon eumitum, Kirkpatrick. Ceratopsis incrustans, sp.n.

Amphilecius fucorum (Esper). Pseudosuberites hyalinus (Ridley and Dendy).

A. rwosus (Thiele). Tentorium semisuberites (Schmidt).

St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, 3-3 miles S 44° E of Jason Light.
Large dredge, rock, no m.
Cinachyra harbata, SoUas. Plumocolumella maeandrina (Kirkpatrick).

244 DISCOVERY REPORTS

St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7° E of Macmahon Rock. Otter trawl, grey mud, 179-235 m.

Rossella niida, Topsent. Isodictya setifer (Topsent).

Leucetta leptoraphis (Jenkin). /. cactoides (Kirkpatrick).

Teiilla kptoderma (SoUas). /. toxophila, sp.n.

Cinachyra coactifera (Lendenfeld). Guitarra fimbriata, Carter.

Microxina hetiedeui (Topsent). loplion proximum (Ridley).

Hemigellius riidis (Topsent). Tedania charcoti, Topsent.
Calyx arcuarius (Topsent).

St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, from 6-3 miles N 89° E of
Jason Light to 4 miles N 39° E of Jason Light. Otter trawl, 120-204 m.

Rossella aiitarctica, var. intermedia, n. Tedania massa, Ridley and Dendy.

Leucetta leptoraphis (Jenkin). Myxilla mollis, Ridley and Dendy.

Cinachyra antarctica (Carter). M. basimucronata, sp.n.

C. barhata, SoIIas. Anchinoe leptochela (Thiele).

Haliclona gaiissiana (Hentschel). Kirkpatrickia coulmani (Kirkpatrick).

Hemigellius rudis (Topsent). Plocamia gaussiana, Hentschel.

Isodictya setifer (Topsent). Ophlitaspongia thielei, sp.n.

/. antarctica (Kirkpatrick). Artemisina apoUinis (Ridley and Dendy).

Plumocolumella maeandrina (Kirkpatrick). Polymastia invaginata, Kirkpatrick.

Mycale acerata, Kirkpatrick. Stylocordyla borealis (Loven), subsp. acuata, Kirk-

Amphilectus fiicorum (Johnston). patrick.

lophoti proximum (Ridley).

St. 45. 6. iv. 26. 27 miles S 85° E of Jason Light, South Georgia. Otter trawl, grey mud,
238-270 m.

Rossella niida, Topsent. Dendrilla membranosa (Pallas).

Mycale acerata, Kirkpatrick.

St. 51. 4. V. 26. Off Eddystone Rock, East Falkland Island, from 7 miles N 50° E to 7-6 miles
N 63° E of Eddystone Rock. Otter trawl, fine sand, 105-115 m.

Dasychalina validissima (Thiele). Tedania spinata (Ridley).

Amphikctus fucorum (Johnston). Clathria lipochela, sp.n.

lophon proximum (Ridley).

St. 53. 12. V. 26. Port Stanley, East Falkland Island, Hulk of 'Great Britain'. Mussel rake,
0-2 m.

Leucosolenia discovereyi, Jenkin. Dasychalina validissima (Thiele).

Leucetta macquariensis, Dendy. Mycale magellanica (Ridley).

Grantia cirrata, var. aurorae, Dendy. Tedania miirdochi, Topsent.

G. cirrata, var. tenuipilosa, n. Hymeniacidon fernandezi, Thiele.

Haliclona variabilis (Thiele). Eumastia attenuata, Topsent.

St. 55. 16. V. 26. Entrance to Port Stanley, East Falkland Island, 2 cables S 24° E of Navy Point.
Small beam trawl, 10-16 m.

Leucetta fnacquariensis, Dendy. Halisarca dujardini, var. magellanica, Topsent.

Mycale magellanica (Ridley).

St. 56. i6. V. 26. Sparrow Cove, Port William, East Falkland Island, i\ cables N 50° E of
Sparrow Point. Small beam trawl, 10J-16 m.

Leucetta macquariensis, Dendy.

LIST OF STATIONS 24s

St. 58. 19. V. 26. Port Stanley, East Falkland Island, on piles of Government Jetty.
Calyx kerguelensis (Hentschel). Pseudosuheritcs sulcatus, Thiele.

Myxilla chilensis, Thiele. Halisarca dujardini, var. magellanica, Topsent.

Hymedesmia laevis, Thiele.

St. 123. i5.xii. 26. Off mouth of Cumberland Bay, South Georgia. From 4-1 miles N 54° E
of Larsen Point to 1-2 miles S 62° W of Merton Rock. Large otter trawl, grey mud, 230-250 m.

Rossella niida, Topsent. lophon radiatus, Topsent.

Cinachym barbata, SoUas. Myxilla mollis, Ridley and Dendy.

Isodictya cactoides (Kirkpatrick). Stylocordyla borealis, subsp. aaiata (Kirkpatrick).

Plumocolumella tnacamliina (Kirkpatrick).

St. 140. 23.xii. 26. Stromness Harbour to Larsen Point, South Georgia. From 54° 02' S,
36° 38' W to 54° 1 1' 30" S, 36° 29' W. Large otter trawl, green mud and stones, 122-136 m.

Leucetta leptoraphis (Jenkin). Calyx kerguelensis (Hentschel).

Placina trilopha, Schulze. Dasychalina validissima (Thiele).

Tetilla leptoderma (SoUas). Tedania charcoti, Topsent.

Cinachyra barbata, Sollas. Axociella flabellata (Topsent).
Adocia glacialis (Ridley and Dendy).

St. 143. 30. xii. 26. Off mouth of East Cumberland Bay, South Georgia, 54° 12' S, 36° 29' 30" W.
Large otter trawl, mud, 273 m.
Isodictya setifer (Topsent).

St. 144. 5. i. 7. Off mouth of Stromness Harbour, South Georgia, from 54° 04' S, 36° 27' W to
53° 58' S, 36° 26' W. Large otter trawl, 4 mm. mesh, green mud and stones, 155-178 m.
Isodictya setifer (Topsent). Stylocordyla borealis, subsp. acuata, Kirkpatrick.

Clathria toxipraedita, Topsent.

St. 145. 7. i. 27. Stromness Harbour, South Georgia, between Grass Island and Tonsberg Point.
Small beam trawl, 26-35 m.

Leucetta macquariensis, Dendy. Tedania charcoti, Topsent.

Haliclona nodosa (Thiele). Myxilla asigmata, Topsent.

loplto}! radiatus, Topsent. Hymeniacidon fernandezi, Thiele.

St. 146. 8. i. 27. South Georgia, 53° 48' S, 35° 37' 30" W. Large dredge, 728 m.
Mycale magellanica (Ridley). Tedania charcoti, Topsent.

St. 148. 9. i. 27. Off Cape Saunders, South Georgia, from 54° 03' S, 36° 39' W to 54° 05' S,
36° 36' 30" W. Large otter trawl, grey mud and stones, 132-148 m.
Rossella niida, Topsent. Phcamia gaussiana, Hentschel.

Leucetta leptoraphis (Jenkin). Tedania massa, Ridley and Dendy.

Tetilla leptoderma (Sollas). T. charcoti, Topsent.

Calyx arcuarius (Topsent). Ophlitaspongia thielei, sp.n.

Isodictya setifer (Topsent). Axociella nidificata (Kirkpatrick).

/. toxohliila sp.n. Homaxinella supratumescens (Topsent).

Plumocolumella maeandrina (Kirkpatrick).

St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia, from 1-15 miles N 76!° W to
2-62 miles S 11° W of Merton Rock. Large otter trawl, mud, 200-234 m.
Microxina benedeni (Topsent). Plumocolumella maeandrina (Kirkpatrick).

Calyx arcuarius (Topsent). Kirkpatrickia variolosa (Kirkpatrick).

Isodictya setifer (Topsent). Axociella flabellata (Topsent).

246 DISCOVERY REPORTS

St. 152. 17. i. 27. South Georgia, 53° 51' 30" S, 36 ' 18' 30" W. Large dredge, rock, 245 m.

Rossella antarctica, var. intermedia, n. T. charcoti, Topsent.

Plakina trihpha, Schulze. Oph/itaspoiigia membranacea, Thiele.

Anchinoe areolata (Thiele). Pseudosuberites hyalinus (Ridley and Dendy).

Tedania massa, Ridley and Dendy. Suberites microstomus, var. stellatus, Kirkpatrick.

St. 153. 17. i. 27. South Georgia, 54° 08' 30" S, 36° 27' 30" W. Large dredge, rock, 106 m.
Cinachyra barbata, Sollas.

St. 155. iS. i. 27. South Georgia, 4-1 miles S 26^° E of Larsen Point. Large otter trawl, mud,
260 m.

Isodictya antarctica (Kirkpatrick).

St. 156. 20. i. 27. South Georgia, 53° 51' S, 36° 21' 30" W. Large dredge, rock, 200-236 m.

Rossella antarctica, var. intermedia, n. Eiirypon miniaceiim, Thiele.

Leucetta ynacqiiariensis (Dendy). Suberites papillatus (Kirkpatrick).

Asbestopluma calyx, Hentschel. Pseudosuberites sulcatus, Thiele.

Tedania tnassa, Ridley and Dendy. Polymastia invaginata, Kirkpatrick.

Anchinoe latnmculioides (Ridley and Dendy). Spongia magellanica, Thiele.
Ophlitaspongia thielei, sp.n.

St. 157. 20. i. 27. South Georgia, 53° 51' S, 36° n' 15" W. Large dredge, diatom ooze, stones
and fine sand, 970 m.

lophon proxinium (Ridley). Dictyociona terrae-novae (Dendy).

Ectyodoryxpaupertas,suhsp.fiobile{RidhyScDendy). Thieleia riibiginosa (Thiele).

Anchinoe areolata (Thiele). Pseudosuberites hyalinus (Ridley and Dendy).

St. 158. 21.1.27. South Georgia, 53° 48' 30" S, 35° 57' W. Large dredge, rock, 401-41 1 m.
Plumocolumella maeandrina (Kirkpatrick). Myxilla elongata, Topsent.

Tedania inassa, Ridley and Dendy. Ectyodoryxpaiipertas,s\ihsp.nobile{KxA\&yhT)tnAy).

T. charcoti, Topsent. Rhizaxinella australiensis, Hentschel.

St. 159. 21. i. 27. South Georgia, 53° 52' 30" S, 36° 08' W. Large dredge, rock, 160 m.

Calyx arcuarius (Topsent). Ectyodoryxpaupertas,s\ihsp.nobile{R\d\fiyk.Dtndy).

Isodictya setifer (Topsent). E. ramilobosa (Topsent).

Mycale acerata, Kirkpatrick. Anchinoe latrunculioides (Ridley and Dendy).

Tedania tnassa, Ridley and Dendy. A. leptochela (Thiele).

Myxilla mollis, Ridley and Dendy. Plocamia gaussiana, Hentschel.

St. 160. 7. ii. 27. Near Shag Rocks, 53° 43' 40" S,40° 57' W. Large dredge, grey mud, stones
and rock, 177 m.

Rossella racovitzae, Topsent. A. areolata (Thiele).

Pericharax pyriformis, sp.n. Acheliderma topsenti, sp.n.

Adocia carduus (Ridley and Dendy). Plocamia gaussiana, Hentschel.

Calyx arcuarius (Topsent). Rhaphidophlus paucispiculus, sp.n.

Isodictya setifer (Topsent). Pseudanchinoe toxifera (Topsent).

Amphilectus fucorum (Esper). Hymedesniia longurioides, sp.n.

Tedania massa, Ridley and Dendy. Hymeniacidon fernandesi, Thiele.

T. charcoti, Topsent. Suberites papillatus, Kirkpatrick.

Myxilla elongata, Topsent. Sphaerotylus antarcticus, Kirkpatrick.

Ectyodoryx antarctica (Hentschel). Stylocordyla borealis, subsp. acuata, Kirkpatrick.

Anchinoe latrunculioides (Ridley and Dendy).

LIST OF STATIONS 347

St. 163. 17. ii. 27. Paul Harbour, Signy Island, South Orkneys. Small beam trawl, 18-27 m.
Calyx arcuaiius (Topsent).

St. 164. 18. ii. 27. East end of Normanna Strait, South Orkneys, near Cape Hansen, Coronation
Island. Small beam trawl, 24-36 m.
Mycale acerata, Kirkpatrick.

St. 167. 20. ii. 27. Off Signy Island, South Orkneys, 60'' 50' 30" S, 46^ 15' W. Large otter
trawl, 7 mm. mesh, green mud, 244-344 m.
Polymastia iiivaginata, Kirkpatrick. Stylocordyhi borcalis, subsp. aaiata, Kirkpatrick.

St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W. Large dredge,
rock, 342 m.

Tetilla leptoderma (Sollas). AcanthorhabJus fragiUs, Burton.

Cinachyra harhata (Sollas). lophon radialus, Topsent.

Hemigellius rudis (Topsent). T. oxeata, Topsent.

Mycale tridens, Hentschel. Suberites papillatus, Kirkpatrick.

M. macrochela, sp.n.

St. 175. 2. iii. 27. Bransfield Strait, South Shetlands, 63° 17' 20" S, 59° 48' 15" W. Large
dredge, mud, stones and gravel, 200 m.

Rossella antarctica, var. intermedia, n. Tedania massa, Ridley and Dendy.

Cinachyra antarctica (Carter). Myxilla mollis, Ridley and Dendy.

C. barbata, Sollas. Kirkpatrickia variolosa (Kirkpatrick).

Adocia tremidus (Topsent). Plocamia gaussiana, Hentschel.

Calyx arcuarius (Topsent). Ophlitaspongia membranacea, Thiele.

Isodictya setifer (Topsent). Crella crassa, Hentschel.

/. toxophila, sp.n. Crellina tubifex, Hentschel.

Mycale magellanica (Ridley). Rhisaxinella australieiisis, Hentschel.

M. acerata, Kirkpatrick. Stylocordyla borealis, subsp. acuata, Kirkpatrick.

lophon radiatus, Topsent.

St. 177. 5. iii. 27. 27 miles SW of Deception Island, South Shetlands, 63° 17' 30" S, 61° 17' W.
Large dredge, mud, coarse sand and stones, 1080 m.

Adocia carduus (Ridley and Dendy). Polymastia invaginata, Kirkpatrick.

Biemna chilensis, Thiele. Stylocordyla borealis, subsp. acuata, Kirkpatrick.

St. 179. 10. iii. 27. Melchior Island, Schollaert Channel, Palmer Archipelago. In creek to SW
of anchorage. Small dredge, rock, 4-10 m.
lophon radiatus, Topsent.

St. 180. II. iii. 27. 17 miles W of N point of Gand Island, Schollaert Channel, Palmer Archi-
pelago. Large dredge, mud and stones, 160-330 m.

Rossella viUosa, Burton. . Adocia cucurbitiformis (Kirkpatrick).

Ilaliclonissa verrucosa, sp.n. Tedania massa, Ridley and Dendy.

St. 181. 12. iii. 27. Schollaert Channel Palmer Archipelago, 64° 20' S, 63°oi'W. Large
otter trawl, mud, 160-335 m.

Gymnorossella inermis, Topsent. Tedania massa, Ridley and Dendy.

Haliclona penicillata (l^opsent). Myxilla mollis, Ridley and Dendy.

St. 187. 18. iii. 27. NeumayrChannel.PalmerArchipelago, 64°48'3o''S, 63°3i'3o"W. Large
dredge, mud, 259 m.

Rossella antarctica, var. intermedia, n. Axociella nidificata (Kirkpatrick).

Cercidochela lankesteri, Kirkpatrick. Polymastia isidis, Thiele.

Biemna chilensis, Thiele.

248 DISCOVERY REPORTS

St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 64° 56' S, 63' 35' W. Large dredge,
rock or stones and mud, 93-130 m.

Rossella antarctica, var. intermedia, n. /. toxophila, sp.n.

R. nuda, Topsent. Myxilla australis, Topsent.

R. racovitzae, Topsent. Ectyodoryx frondosa, var. anacantha, Ilentschel.

R. villosa. Burton. Clathria toxipraedita, Topsent.

Gymnoiossella inermis, Topsent. Axociella nidificata (Kirkpatrick).

Plakina tiilopha, Schulze. Euiypon miniaceum, Thiele.

Microxina benedeni (Topsent). Rhisaxinella australiensis, Hentschel.

Adocia glacialis (Ridley and Dendy). Suherites papillatus, Kirkpatrick.

Isodictya setifer (Topsent). S. microstomus, var. stellatus, Kirkpatrick.

/. cactoides (Kirkpatrick). Stylocordyla borealis, subsp. acnata, Kirkpatrick.
/. erinacea (Topsent).

St. 222. 23. iv. 27. St Martin's Cove, Hermite Island, Cape Horn. Large rectangular net, 70-

Tedania spinata (Ridley). Halisarca dujardini, var. magellanica, Topsent.

(The specimens obtained from St . 222 appear to be all detached . Those of Tedania spinata are without
visible means of attachment and Halisarca dujardini, var. magellanica is seated on a piece o{Fiiciis(}).)

St. 271. 28-30. vii. 27. Elephant Bay, Angola. Shore collecting.
Hymeniacidon sanguinea (Grant).

St. 283. 14. viii. 27. Off Annobon, Gulf of Guinea, 075 to i mile N 12° E of Pyramid Rock,
Annobon. Large dredge, 18-30 m.

Stylohalina hirta (Topsent). Spirastrella purpurea (Lamarck).

Bubaris vermiculata (Bovverbank). Cliona aethiopicus, sp.n.

R.S.S. 'WILLIAM SCORESBY'

St. WS 25. 17. xii. 26. Undine Harbour (North), South Georgia. Small beam trawl, mud and
sand, 18-27 i^-
Ophlitaspongia thielei, sp.n.

St. WS 27. 19. xii. 26. South Georgia, 53° 55' S, 38° 01' W. i m. tow-net (touched bottom),
gravel, 106-109 m.

Tetilla leptoderma (Sollas). M. australis (Topsent).

Halic/ona bilawellata, sp.n. M. hasimucronata, sp.n.

Isodictya toxophila, sp.n. Kirkpatrickia variolosa (Kirkpatrick).

Plumohalichondria maeandrina (Kirkpatrick). Stylotellopsis amabilis, Thiele.

Tedania charcoti, Topsent. Hymeniacidon fernatuiezi, Thiele.
Myxilla jnoUis, Ridley and Dendy.

St. WS 33. 21. xii. 26. South Georgia 54° 59' S, 35° 24' W. i m. tow-net (touched bottom),
grey mud and stones, 130 m.

Grantia cirrata, var. aurorae, Dendy. lophon radiatus, Topsent.

Cinachyra barbata, Sollas. Clathria toxipraedita, Topsent.

PlumocolumcUa maeandrina (Kirkpatrick). Hymcdesmia simillima, var. antarctica, Hentschel.

St. WS 42. 7. i. 26. South Georgia, 54° 41' 45" S, 36° 47' W. i m. tow-net, grey mud and stones,
198 m.

Cinachyra barbata, Sollas. M. basimucronata, sp.n.

Tedania charcoti, Topsent. Plocamia gaussiana, Hentschel.

Myxilla mollis, Ridley and Dendy.

LIST OF STATIONS 249

St. WS 62. 19. i. 27. Wilson Harbour, South Georgia. Small beam trawl, 15-45 m.
Homaxinella supratimiescens (Topsent).

St. WS72. 5.iii.27. Falkland Islands, 5i°07'S, 57° 34' W. Commercial otter trawl, sand
and shell, 79 m.
Callyspongia fortis (Ridley). Tedania spinata (Ridley).

St. WS 73. 6. iii. 27. Falkland Islands, 51^ 01' S, 58° 54' W. Tow-net attached to trawl, fine
dark sand, i2i m.

Dasychalina validissima (Thiele). Mycale magellanka (Ridley).

Isodictya setifer (Topsent). Tedania teiiuicapitata, Ridley.

/. antarctica (Kirkpatrick).

St. WS 75. 10. iii. 27. Falkland Islands, 51° 01' 30" S, 60° 31' W. Commercial otter trawl,
72 m.
Tedatiia spinata (Ridley).

St. WS 76. II. iii. 27. Falkland Islands, 51° 00' S, 62° 02' 30" W. Commercial otter trawl,
fine dark sand, 205-207 m.

Dasychalina validissima (Thiele). T. spinata (Ridley).

Tedania tenuicapitata, Ridley. Myxilla mollis, Ridley and Dendy.

St. WS77. 12. iii. 27. Falkland Islands, 51° 01' S, 66° 31' 30" W. Commercial otter trawl,
coarse dark sand, 110-113 m.

Dasychalina validissima (Thiele). Isodictya setifer (Topsent).

Callyspotigia fusifera (Thiele). Tedania spinata (Ridley).

St. WS 79. 13. iii. 27. Falkland Islands, 51° 01' 30" S, 64° 59' 30" W. Commercial otter trawl,
fine dark sand, 131-132 m.

Isodictya kerguelensis (Ridley and Dendy). Myxilla mollis, Ridley and Dendy.

Guitarra fimbriafa, Carter. Extyodoiyxpaiipertas,suhsp.nobile{Rid\eySiDendy).

Plumocolumella myxillioides, sp.n. Dictyociona discreta (Thiele).

Tedajiia tenuicapitata, Ridley. Halisarca dujardini, var. magellanica, Topsent.
T. spinata (Ridley).

St. WS 80. 14. iii. 27. Falkland Islands, 50° 57' S, 63° 37' 30" W. Net of 7 mm. mesh
attached to trawl, fine dark sand, 152-156 m.

Plakina trilopha, Schulze. Tedania massa, Ridley and Dendy.

Isodictya setifer (Topsent). T. spinata (Ridley).

Plumocolumella myxillioides, sp.n.

St. WS 81. 19. iii. 27. 8 miles N 11° W of North Island, West Falkland Island. Commercial
otter trawl and net of 7 mm. mesh attached to trawl, sand, 81-82 m.
Geodia magellani (Sollas). Tedania massa, Ridley and Dendy.

Microxina benedeni (Topsent). Clathria papulosa, Thiele.

Adocia carduus (Ridley and Dendy). Raspaxilla pliakellina, Topsent.

Callyspongia fiahellata, sp.n. Polymastia isidis, Thiele.

Isodictya setifer (Topsent). Latrunculia kndenfcldi, Hentschel.

Plumocolumella myxillioides, sp.n.

St.WS82. 21. iii. 27. Falklandlslands, 5i°3o'S,6i° 15'W. Commercial otter trawl, 140-144 m.
Adocia glacialis (Ridley and Dendy) . Ectyodoryxpaupertas, subsp . nobile (Ridley & Dendy) .

Gellius fibulatus (Schmidt). Axinella crinita, Thiele.

Mycale magellanica (Ridley). Polymastia isidis, Thiele.

Tedania massa, Ridley and Dendy.

2SO DISCOVERY REPORTS

St. WS 83. 24. iii. 27. 14 miles S 64° W of George Island, East Falkland Island. Commercial
otter trawl, fine green sand and shell, 137-129 m.

Haliclona bilamellata, sp.n. Tedania temiicapitata, Ridley.

Adocia carduus (Ridley and Dendy). T. niassa, Ridley and Dendy.

A. glacialis (Ridley and Dendy). 7'. spinata (Ridley).

Dasychalina validissima (Thiele). Myxilla mollis, Ridley and Dendy.

Ctillyspongia fortis (Ridley). Ectyodoryxpaupeitas, siibsp . nobile (Ridley & Dendy).

Isodictya setifer (Topsent). Iiifiatella belli (Kirkpatrick).

/. antarctica (Kirkpatrick). Clathna papulosa, Thiele.

Plumocoliimella myxiUioides, sp.n. Rhaphidophlus paucispicuhis, sp.n.

Mycale ?iiagcllaitica (Ridley). Stylotdlopsis amabilis, Thiele.

Amphilectus fucorum (Johnston). Hymeniacidon dubia, sp.n.

lophon radiatus, Topsent. Plicatcllopsis arborescens, sp.n.

I. proximiiiii (Ridley). Latniiiculia leiidenfeldi, Hentschel.

St. WS 84. 24. iii. 27. 7J miles S 9° W of Sea Lion Island, East Falkland Island. Commercial
otter trawl, coarse sand, shells and stones, 75-74 m.

Leucetta leptorhaphis (Jenkin). Clathria papulosa, Thiele.

Leucetiiisa hacckeliana (Polejaeff). Rhaphidoplilus poucispiculus, sp.n.

L. simplicissinui , sp.n. Raspaxilla phakellina, Topsent.

Haliclona chilensis (Thiele). Hymeniacidon fernandesi, Thiele.

Dasychalina validissima (Thiele). Plicafellopsis fiabellata, sp.n.

Callyspongia fortis (Ridley). Latrunculia leiidenfeldi, Hentschel.

Myxilla mollis, Ridley and Dendy. Spojigia magellanica, Thiele.

Anchinoe latriinculioides (Ridley and Dendy). Duseideia chilensis (Thiele).

St. WS 85. 25. iii. 27. Falkland Islands, 52° 09' S, 58° 14' W. Commercial otter trawl with nets
of 4 and 7 mm. mesh attached, sand and shell, 79 m.

Grantia cinata, var. tenuipilosa, n. Tedania tenuicapitata, Ridley.

Amphilectus rugostis (Thiele). Hymeniacidon fernandesi, Thiele

lophon proximum (Ridley).

St. WS 86. 3. iv. 27. Falkland Islands, 53° 53' 30" S, 60° 34' 30" W. Commercial otter trawl,
sand, shell and stones, 151-147 m.

Haliclona bilamellata, sp.n. Plumocoliimella myxiUioides, sp.n.

Callyspongia fortis (Ridley). Ectyodoryx paupertas,svLhs^. nobile (R\d\e.y?iiT)e.ndy).

C.fusifera (Thiele). Pseudosuberites sulcatus, Thiele.

Isodictya microchela (Topsent). Polymastia isidis, Thiele.

Guitarra fimbriata. Carter.

St. WS87. 3.iv. 27. Falkland Islands, 54° 07' 30" S, 58° 16' W. Commercial otter trawl,
sand, shell and stones, 96-127 m.

Haliclona bilamellata, sp.n. Tedania massa, Ridley and Dendy.

Mycale magellanica (Ridley).

St. WS 88. 6. iv. 27. Falkland Islands, 54° 00' S, 64° 57' 30" W. Commercial otter trawl,
sand, shell and stones, 96-127 m.

Adocia tenelliis (Topsent). T. spinata (Ridley).

Isodictya setifer (Topsent). Myxilla mollis, Ridley and Dendy.

Plumoculumella myxiUioides, sp.n. Stelodoryx discovereyi, sp.n.

Amphilectus fiabellata, sp.n. Clathria papulosa, Thiele.

Tedania tenuicapitata, Ridley. LatruncuUa lendenfeldi, Hentschel.
T. 7nassa, Ridley and Dendy.

LIST OF STATIONS 251

St. WS 89. 7. iv. 27. Falkland Islands, 9 miles N 21° E of Arenas Point Light, Tierra del Fuego.
Commercial otter trawl, mud, gravel and stones, 23-21 m.
Tedania tenuicapitata, Ridley. Spongia magellanica, Thiele.

St. WS 90. 7. iv. 27. 13 miles N 83° E of Cape Virgins Light, Argentine Republic. Commercial
otter trawl, fine dark sand, 82-81 m.

Myaik magellanica (Ridley). Pseudosuberiles sulcatiis, Thiele.

Stelodoryx discovereyi , sp.n.

St. WS91. 8. iv. 27. Tierra del Fuego, 52° 53' 45" S, 64' 37' 30" W. Commercial otter trawl,
fine dark sand and shell. 191-205 m.
Tedania spinafa (Ridley] Pseudosuberiles sulcatus, Thiele.

St. WS93. 9.iv. 27. West Falkland Island, 51° 51' S, 61° 30' W. Commercial otter trawl,
grey sand, 133-130 m.
Tedania massa, Ridley and Dendy.

St. WS 95. 17. iv. 27. Falkland Islands, 48° 58' 15" S, 64° 45' W. Commercial otter trawl,
fine dark sand and stones, 108-109 ™-

Tedania spinata (Ridley). Dictyociona discreta (Thiele).

Clatliria papulosa, Thiele.

St. loi. 23. iv. 27. Falkland Islands, 50° 27' S, 62° 06' W. 70 cm. tow-net, 150-100 m.
Gymnorossella inermis, Topsent.

St. 102. 23. iv. 27. Falkland Islands, 50° 05' S, 62° 37' W. 70 cm. tow-net, 100-50 m.
Gymnorossella inermis, Topsent.

St. WS 99. 19. iv. 27. Falkland Islands, 49° 42' S, 59° 14' 30" W. Commercial otter trawl,
fine dark sand, 251-225 m.

Halidona bilamellata, sp.n. Tedania tenuicapitata, Ridley.

Microxina benedeni (Topsent). T. massa, Ridley and Dendy.

Isodictya setifer (Topsent). Myxilla mollis, Ridley and Dendy.

St. WS 108. 25. iv. 27. Falkland Islands, 48° 30' 45" S, 63° 33' 45" W. Commercial otter trawl,
fine dark sand, 118-120 m.

Halidona penicillata (Topsent). Pseudosuberiles sulcatus, Thiele.
Tedania spinata (Ridley).

St. WS 109. 26. iv. 27. FaOdand Islands, 50° 18' 48" S, 58° 28' 30" W. Commercial otter trawl,
fine dark sand, 145 m.

Dasychalina validissima (Thiele). Tedania massa, Ridley and Dendy.

Isodictya kerguelensis, var. simillima (Hentschel). T. spinata (Ridley).

I. setifer (Topsent). Rhaphidoplilus paucispiculus, sp.n.

/. (/e/;V«ta (Thiele). Artemisina plumosa,Yltntschd.

St. WS 128. 10. vi. 27. W side of Gough Island, inshore. Large dredge, 120-90 m.
lophon proximum (Ridley).

St.WSi77. 7.iii.28. South Georgia, 54=58' S, 35° 00' W. im. tow-net (touched bottom), 97 m.
Mycale magellanica (Ridley).

St. WS 210. 29. v. 28. Falkland Islands, 50° 17' S, 60° 06' W. Commercial otter trawl, green
sand, 161 m.

Hemigellius pachyderma, sp.n. Tedania spinata (Ridley).

Mycale magellanica (Ridley).

252 DISCOVERY REPORTS

St. WS 213. 30. V. 28. Falkland Islands, 49" 22' S, 60° 10' W. Commercial otter trawl, green
sand, mud and pebbles, 249-239 m.
Mycale magellanica (Ridley).

St. WS 216. I. vi. 28. N of Falkland Islands, 47° 37' S, 60° 50' W. Commercial otter trawl,
fine sand, 219-133 m.
Tedania massa, Ridley and Dandy.

St. WS 218. 2. vi. 28. N of Falkland Islands, 45° 45' S, 59° 35' W. Commercial otter trawl,
dark sand, 311-247 m.

Tetilla leptoderma (SoUas).

St. WS 222. 8. vi. 28. Falkland Islands, 48° 23' S, 65° 00' W. Commercial otter trawl, coarse
brown sand, 100-106 m.
Mycale magellanica (Ridley). Tedania spinata (Ridley).

St. WS 223. 8. vi. 28. NE of Falkland Islands, 49° 13' S, 64° 52' W. Commercial otter trawl,
coarse brown sand, 114 m.
Haliclonissa sacciformis, sp.n. Guitarra fimbriata, Carter.

St. WS 225. 9. vi. 28. Falkland Islands, 50° 20' S, 62° 30' W. Commercial otter trawl, green
sand, shells and pebbles, 162-161 m.

Rossella racovitsae, Topsent. T. spinata (Ridley).

Gymnorossella inermis, Topsent. Myxilla mollis, Ridley and Dendy.

Tetilla leptoderma (Sollas). Ectyodoryxpaupeitas, subsp. nobile (Ridley & Dendy).

Mycale magellanica (Ridley). Hymedesmia simillima, var. antarctica, Hentschel.

Tedania massa, Ridley and Dendy. Rhizaxinella australiensis, Hentschel.

St. WS 229. I. vii. 28. Falkland Islands, 50° 35' S, 57° 20' W. Net of 4 mm. mesh attached to
trawl, fine green sand, 210-271 m.

Amphilectus fucorum (Esper).

St. WS 231. 4. vii. 28. Falkland Islands, 50° 10' S, 58° 42' W. Commercial otter trawl, fine
green sand, 167-159 m.
Tedania massa, Ridley and Dendy.

St. WS 233. 5. vii. 28. Falkland Islands, 49° 25' S, 59° 45' W. Commercial otter trawl, fine
green sand, 185-175 m.
Dasychalina validissima (Thiele). Mycale magellanica (Ridley).

St. WS 237. 7. vii. 28. N of Falkland Islands, 46' 00' S, 60° 05' W. Commercial otter trawl,
coarse brown sand and shells, 150-256 m.
Dasychalina validissima (Thiele). Pseudanchinoe toxifera (Topsent).

St. WS239. 15. vii. 28. Falkland Islands, 51° 10' S, 62° 10' W. Commercial otter trawl,
coarse dark sand, 196-193 m.

Rossella nuda, Topsent. T. spinata (Ridley).

Isodictya setifer (Topsent). Myxilla mollis, Ridley and Dendy.

Mycale magellanica (Ridley). Ectyodoryx paupertas, swhs^). nobile {BAdXsiy&c'Dendy).

Tedania tenuicapitata, Ridley. Suberites montiniger (Carter).

LIST OF STATIONS 253

St. WS 243. 17. vii. 28. NE of Falkland Islands, 51° 06' S, 64° 30' W. Commercial otter trawl
and net of 4 mm. mesh attached, coarse dark sand, 144-141 m.

Leucetttisa haeckeliana (Polejaeff). T. spinata (Ridley).

Tctilla Icplodcrma (Sollas). Myxilla verrucosa, sp.n.

Ilaliclona tubuloramosa (Dendy). Ectyoduryx paupertas, subsp. iiobik (Ridley and

H. conica (Brondsted). Dendy).

H. hilamcllata , sp.n. Anchinoe latninculioides (Ridley and Dendy).

Dasychalina validissima (Thiele). Stelodoryx pluridcntata (Lundbeck).

Isodictya setifer (Topsent). Dictyociona discreta (Thiele).

Guitarra fimhriata, Carter. Cluthria toxipraedita, Top.sent.

Phimncolumella myxillioides, sp.n. Plicatellupsis arhorescens, sp.n.

Biemna chilensis, Thiele. Pseudosuberites hyalinus (Ridley and Dendy).

Tedania temiicapitata, Ridley. Latniuculia Icudenfeldi , Hentschel.

T. rnassa, Ridley and Dendy. Diiseideia teiiuifihra, sp.n.

St. WS 244. 18. vii. 28. Falkland Islands, 52° 00' S, 62° 40' W. Commercial otter trawl, fine dark
sand and mud, 253-247 m.

Mycale magellanica (Ridley). Myxilla mollis, Ridley and Dendy.
Tedania massa, Ridley and Dendy.

St. WS 246. 19. vii. 28. Falkland Islands, 52" 25' S, 61° 00' W. Commercial otter trawl, coarse
green sand and pebbles, 267-208 m.

Isodictya setifer (Topsent). T. massa, Ridley and Dendy.

Plumocolumella myxillioides, sp.n. Myxilla mollis, Ridley and Dendy.

Mycale magellanica (Ridley). Latninculia lendenfeldi, Hentschel.
Tedania tetiuicapitata, Ridley.

St. WS 247. 19. vii. 28. Falkland Islands, 52° 40' S, 60° 05' W. Large dredge, rock, 172 m.

Haliclona bilamellata, sp.n. Myxilla mollis, Ridley and Dendy.

Mycale magellanica (Ridley). Ectyodoryx paupertas, subsp. nohile (Ridley and

lophon proximum (Ridley). Dendy).

St. WS 248. 20. vii. 28. Falkland Islands, 52° 40' S, 58° 30' W. Commercial otter trawl, fine
green sand, pebbles and shells, 210-242 m.

Rossella racovitzae, Topsent. Tedania tcnuicapitata, Ridley.

Microxina benedeni (Topsent). T. massa, Ridley and Dendy.

Isodictya setifer (Topsent). • Inflatella belli (Kirkpatrick).

Plumocolumella myxillioides, sp.n. Polymastia isidis, Thiele.

Mycale magellanica (Ridley). Latrunculia lendenfeldi, Hentschel.
M. lapidiformis (Ridley and Dendy).

St. WS 249. 20. vii. 28. Falkland Islands, 52^^ 10' S, 57° 30' W. Large dredge, fine brown and
green sand, shells and stones, 166 m.

Mycale magellanica {KiAXty). Ectyodoryx paupertas, subsp. nobile (Ridley and

Teda7da spinata (Ridley). Dendy).

St. WS 250. 20. vii. 28. Falkland Islands, 51' 45' S, 57" 00' W. Commercial otter trawl, fine
green sand, 251-313 m.

Gymnorossella inermis, Topsent. T. massa, Ridley and Dendy.

Mycale magellanica (Ridley). Kr-^'H" '""//"- R'^ley =>"'! ^^ndy.

Tedania tcnuicapitata, Ridley. Polymastia ,s,d,s, Thiele.

DISCOVERY REPORTS

MARINE BIOLOGICAL STATION

St. MS 14. 17. ii. 25. From 1-5 miles SE by S to 1-5 miles S 1° W of Sappho Point, East Cumber-
land Bay. Small dredge, 190-110 m.
Homaxinella supratumescens (Topsent).

St. MS 68. 2. iii. 26. East Cumberland Bay, South Georgia, from 17 miles S r E to 8| cables
SE by E of Sappho Point. Large rectangular net, 220-247 m.
Rossella nuda, Topsent. hodictya cactoides (Kirkpatrick).

St. MS 71. 9. iii. 26. East Cumberland Bay, South Georgia, from 9I cables E by S to i-2 miles
E by S of Sappho Point. Small beam trawl, 120-60 m.

Cinachyra baibata, Sollas. Myxilla mollis, Ridley and Dendy.

Mycale frideiis, Hentschel.

Saldanha Bay, South-west Africa (beach).
Haliclona Stephens!, sp.n. Hymeniacidon carunada, Bowerbank.

DESCRIPTIONS OF SPECIES

Phylum NUDA

Order HEXACTINELLIDA

Family ROSSELLIDAE

Genus Rossella, Carter

Rossella antarctica, var. intermedia, n. (Fig. 3 b, p. 257).

Occurrence. St. 42: South Georgia, 120-204 m.; St. 152: South Georgia, 245 m.; St. 156: South
Georgia, 200-236 m.; St. 175: South Shetlands, 200 m.; St. 187: Palmer Archipelago, 259 m.;
St. 190: Palmer Archipelago, 93-130 m.

Remarks. The six specimens present are all immature forms obtained from South
Georgia and the South Shetlands. The smallest is about i cm. in diameter and the
largest no more than 7 cm. high. In most respects they are quite typical of the species,
but their method of attachment leads to a number of interesting considerations. Two of
the specimens have the normal tuft of basalia, a third is without the basal half of the
sponge, and the three remaining are without basalia but have anchored themselves to
pieces of black volcanic rock.

Two forms of R. antarctica have been known hitherto, var. antarctica from the
Antarctic and var. solida from the Kerguelen region. In the present group of specimens
we have a third variety which is intermediate between these two, suggesting that they
are no more than geographical varieties of a single species. The main differences between
the var. antarctica and the var. solida are that in the former there is invariably (?) a basal
tuft, that the velum is more strongly developed and that the rays of the pentactin pleuralia
are more or less cruciform ; whereas in the latter, the sponge is always attached directly
to the substratum, the velum is but feebly developed and the rays of the pentactin
pleuralia are confined to a sector of about 140° (Fig. 3 b). The present specimens are

HEXACTINELLIDA 2SS

completely intermediate between the two older varieties in these three distinguishing
features : the sponge may have a basal tuft or may be attached directly to the substratum,
the velum is moderately well developed (in one specimen it is present on one side of the
sponge and completely absent on the other, although the diactin pleuralia are equally well
developed all over), and the rays of the pentactin pleuralia are confined usually to a
sector of i8o'.

With regard to the question whether a species which normally has a basal tuft may be
represented by specimens attached directly to the substratum (cf. Kirkpatrick, 1907,
p. 6), the present specimens contribute some relevant data. The individuals of var.
intermedia seem to be able equally well to produce a basal tuft or to attach themselves
direct, and there is evidence that the same thing is true, though perhaps to a lesser
extent, in R. niida, Topsent, and R. racovitzae, Topsent {q.v.).

One of the specimens is a " twin", recalling the specimen of var. solida originally de-
scribed by Schulze (1887) from Kerguelen.

Among the specimens from St. 156 is a small bundle of pentactin pleuralia from an
individual of var. intermedia bearing a bud about 3 mm. in length. The presence of this
bud in a jar which included, together with non-Hexactinellid sponges, only a small
specimen of the present variety, suggests that it was found free and had not been
accidentally torn away from a larger, budding specimen. If correct, this surmise offers
an explanation as to how the buds of R. antarctica, which migrate along the shafts of the
pentactin pleuralia, get over the obstacle offered by the rays at the ends of the pleuralia.
Presumably the pentacts bearing the buds are shed and the buds thus released,

Rossella nuda, Topsent (Figs, i, 2).

(For synonymy see Burton, 1929, p. 409.)

Occurrence. St. 39: South Georgia, 179-235 m-l St. 45= South Georgia, 238-27001.; St. 123:
South Georgia, 230-25001.; St. 148: South Georgia, 132-14801.; St. 190: Palmer Archipelago,
93-130 m.; St. WS 239: Falkland Islands, 196-193 m.; St. MS 68: South Georgia, 220-247 m.

Remarks. The several specimens, only three of which are dry, range in size from a
few cm. to 77 cm. high, and show all the variations illustrated by me (1929, p. 406,
text-fig. i), including that of the hodgsoni form. Of the specimens from St. 190, one is
fragmentary and the other is a "twin" (cf. R. antarctica, var. intertnedia) 14 cm. high
and very like the holotype of " Aidorossella levis ". The specimens from Sts. 39 and MS 68
consist of papillae torn from the surfaces of large specimens of the c-iorm} The specimen
from St. 148 is sub-spherical, 10 cm. high, with surface almost apapillose, but with
strongly developed bundles of pleuralia projecting from the surface. There are only
vestiges of a root-tuft and the base is truncate with small pebbles, etc., buried in its outer
tissues. This specimen combines the sub-spherical shape of both the «-form and the
c-form, the surface of the i-form, and, in having no root-tuft, it approximates also to both
the e-Lrm and the /-form. The specimen from St. 123 is cyHndrical, 14 cm. high and
5 cm. in diameter, with slightly papillose surface beset with feeble bundles of pleuralia.

1 These letters correspond to the letters given in my Terra Nova Report, 1929, text-fig. 2.

3-2

256 DISCOVERY REPORTS

It is pedunculate, without root-tuft, and still bears at its base a portion of the black rock
on which it was growing.

Fig. I. Rossella 7iuda, Topsent. Base
of a specimen from St. 123, with basal
tuft and an incipient stalk, x f .

Fig. 2. Rossella nuda, Topsent. A second
specimen showing the same features, but
with the stalk placed laterally.

The next three specimens are dried. The first is cylindrical, 75 cm. high by 30 cm. in
diameter, with only a scanty root-tuft as in the ^-form. The lower two-thirds of the
sponge is covered with numerous well-developed papillae without projecting pleuralia,
but on the upper third the papillae are scarce. They are often as much as 5 cm. high and
may be bifid, or even trifid. In this sponge, then, are combined the characters of the
b-, c- and ^-forms. The second specimen, 22 cm. high and 9 cm. in diameter, is com-
pletely covered with small papillae devoid of projecting pleuralia, and with the basalia
segregated into sparse bundles as in the ^-form. The basalia are few in number and
feebly developed, recalling the condition found in the e- and /-forms. The chief interest
of this specimen is, however, the fact that the sponge was probably attached directly by
its base, and did not rely on its basal tuft although it is slightly developed, since this is
continued downwards as a short stalk (Fig. i). But, judging by the appearance of the
base, and by the presence of numerous small fragments of black rock embedded in the
outer tissues, it was not attached to a rock but buried in the mud. The third specimen is
very similar to the foregoing, is 30 cm. high and 17 cm. in diameter, completely covered
with papillae and bears sparse bundles of basalia. This sponge appears to have been
attached by a laterally placed basal plate (Fig. 2).

Distribution. South Georgia; Graham Land; Victoria Land; Wilhelm Land.

Rossella racovitzae, Topsent.

(For synonymy see Burton, 1929, p. 407.)

Occurrefice. St. 160: near Shag Rocks, 177 m.; St. 190: Palmer Archipelago, 93-130 m.;
St. WS 225: Falkland Islands, 162-161 m.; St. WS 248: Falkland Islands, 210-242 m.

HEXACTINELLIDA 257

Remarks. Of the two small specimens, the larger, 3 cm. high by 2 cm. in diameter,
has a distinct basal tuft; but the smaller, less than 5 mm. in diameter, has no recog-
nizable tuft and is fixed by the base to a Serpulid tube.

Distribution. South Orkneys; Graham Land; Victoria Land; Wilhelm Land.

Rossella villosa, Burton.

/1;. villosa. Burton, 1929, p. 412, pi. iii.
Occurrence. St. 180: Palmer Archipelago, 160-330 m.; St. 190: Palmer Archipelago, 93-130 m.

Remarks. There are two immature specimens, the largest 5 cm. high, which appear
to belong to this species ; but since the basalia are missing in each it is impossible to be
quite certain of the identification, especially as they resemble the young forms of
R. niida, Topsent, in many ways.

Distribution. Victoria Land.

Genus Gymnorossella, Topsent
Gymnorossella inermis, Topsent (Fig. 3 d-g).

G. inermis, Topsent, 1916, p. 164; id., 1917, p. 22, pi. i, fig. i, pi. v, fig. 4; Burton, 1929, p. 413.
Occurrence. St. 181 : Palmer Archipelago, 160-335 m.; St. 190: Palmer Archipelago, 93-130 m.;
St. WS loi : Falkland Islands, 100-150 m.; St. WS 102: Falkland Islands, 50-100 m.; St. WS 225:
Falkland Islands, 162-161 m.; St. WS 250: Falkland Islands, 251-313 m.

Remarks. The larger of the two specimens is 20 cm. high and both agree closely with
the holotype. The only point worth noting here is the variation in the angles of the rays

Fig. 3. Showing variation in rays of pentactin, pleuralia and basalia: a, Rossella antarctica solida;
b, R. antarctica intermedia; c, R. antarctica antarctica; d-g, Gymnorossella inermis.

of the pentactin basalia. In the larger specimen (from St. 181) these are all disposed
within an angle of less than 180° ; in the second specimen they vary a good deal, and
although in most of the basalia the rays are confined within an angle of less than 180°,
some are cruciform (Fig. 3 d-g) and there are all intermediates.

Distribution. Graham Land.

2s8 DISCOVERY REPORTS

Phylum GELATINOSA

Order CALCAREA

Family HOMOCOELIDAE

Genus Leucosolenia, Bowerbank

Leucosolenia macleayi (Lendenfeld).
(For synonymy see Burton, 1930, p. 14.)

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Remarks. Two small stipitate sponges, about 1 2 mm. in height and with a stalk 7 mm.
long, with the typical external form. In the spicules the basal ray measures 0-09 mm.
and the lateral rays 0-045 mm.

Distribution. Practically cosmopolitan.

Leucosolenia discovereyi, Jenkin.

L. discovereyi, Jenkin, 1908, p. 6, pi. xxviii, figs. 12, 13.
Occurrence. St. 53 : Falkland Islands, 0-2 m.

Remarks. Numerous small specimens growing among Hydroids on a Mytilus shell.
Distribution. Victoria Land.

Family LEUCASCIDAE

Genus Pericharax, Polejaeff

Pericharax pyriformis, sp.n. (Plate XLVIII, figs, i, 2).

Holotype. B.M. 28. 2. 15. 36.

Occurrence. St. 160: Shag Rocks, 177 m.

Diagnosis. Sponge pyriform, with single apical oscule ; skeleton a mixture of large
and small triradiates in chamber layer, with a dermal skeleton of triradiates, a gastral
skeleton of quadriradiates and oscular skeleton of quadriradiates.

Remarks. This is a typical Pericharax, the skeleton of which closely resembles those
of both P. heteroraphis and P. peziza, but which difl^ers from both in the external form.
The three specimens representing this species are all pyriform, with a single apical
oscule which leads out from a shallow cloaca bearing in its walls the openings of the main
exhalant canals. The surface is slightly uneven and rough to the touch. The colour, in
spirit, is a greyish white. The largest specimen, the holotype, measures 2 cm. high and
I -5 cm. across the broadest part. The diameter of the oscule is 2 mm.

Spicules, (i) Large triradiates of the main skeleton, equiangular and equiradiate,
rays straight but gradually and sharply pointed, measuring up to i -2 mm. long and
0-096 mm. thick at the base.

(2) Small triradiates of the main skeleton, equiangular and equiradiate, with straight
rays measuring up to 0-24 mm. long by 0-015 mm. thick at the base.

CALCAREA 259

(3) Triradiates of the dermal skeleton slightly irregular and arranged tangentially to
the surface, each ray about 0-24 mm. long.

(4) Quadriradiates of the oscular margin sagittal with facial rays 0-105 mm. long and
apical rays 0-12 mm. long.

(5) Quadriradiates of the gastral cortex and larger exhalant canals, of similar dimen-
sions to (4) but not sagittal.

Genus Leucetta, Haeckel
Leucetta leptoraphis (Jenkin).

Leiicandra primigenia, var. leptoraphis, Jenkin, 1908, p. 14, pi. xxix, figs. 33, 34; Leucetta
antarctica, Dendy, 1918, p. 8, pi. i, figs. 2-7; L. leptoraphis. Burton, 1929, p. 404, pi. v, figs. 1-4.

Occurrence. St. 39: South Georgia, 179-23501.; St. 42: South Georgia, 120-20401.; (St. 84:
32° S, 1° E, 75-74 m. ; there is evidently a doubt about this station which should probably be Palmer
Archipelago, St. 190); St. 140: South Georgia, 122-136 m.; St. 148: South Georgia, 132-148 m.;
St. WS 84: Falkland Islands, 75-74 m.

Remarks. Some ten specimens are present agreeing with the holotype in spiculation
and showing the same range in external form as that to which I have drawn attention
in my Terra Nova report (loc. at.).

Distribution. Victoria Land.

Leucetta macquariensis, Dendy.

L. macquariensis, Dendy, 1918, p. 9, pi. i, figs. 3, 8.

Occurrence. St. 53 : Falkland Islands, 0-2 m. ; St. 55 : Falkland Islands, 10-16 m. ; St. 56 : Falkland
Islands, 10I-16 m.; St. 145 : South Georgia, 26-35 m.; St. 156: South Georgia, 200-236 m.

Remarks. There are considerable differences between the individuals of this species
as regards the large oxea and the microxea. One or other of these types of spicules may
be sparingly present or even entirely absent, and but for the fact that they are never both
absent altogether, the task of identification would present some difficulty. Even so, it
is easy to imagine that single specimens examined alone might conceivably be placed
in a different species, owing to the absence of one or other of these spicules, and it is
quite possible that the case may yet arise in which an individual will be found to have
lost both sets of spicules entirely.
Distribution. Macquarie Island.

Family LEUCALTIDAE
Genus Leucaltis, Haeckel
Leucaltis gastrorhabdifera, sp.n. (Figs. 4, 5).
Holotype. B.M. 28. 2. 15. 833.
Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Diagnosis. Sponge erect, tubular, without oscular fringe ; surface very minutely and
sparingly hispid; skeleton composed of a dermal layer of triradiates, with large sagittal

26o DISCOVERY REPORTS

quadriradiates forming skeleton of chamber layer and a layer of large rhabds immediately
beneath gastral surface; rhabds simi-
lar to those of gastral layer found pro-
jecting, sparingly, from dermal surface ;
triradiates of dermal tangential layer,
equiangular and equiradiate, rays o-i8
by o-oi mm.; quadriradiates of cham-
ber layer, apical rays o-i8 by o-oi mm.
and basal rays 0-28 to 0-42 by
0-024 "^rn- ' rhabds 0-59 by 0-022 mm.
Remarks. The holotype is no more
than I cm. high and 2 mm. in diameter.
To have described adequately the spi-
cules of the oscular margin, according
to custom, would have meant the total

Fig. 4. Leucaltis gastrorhahdifera, sp.n. Section of body
wall to show arrangement of skeleton (semi-diagrammatic) .

destruction of the specimen, and since these spicules are not of diagnostic importance,
no attempt has been made to examine them.

Fig. 5. Leucaltis gastrorhahdifera, sp.n. a, rhabd; b, sagittal
quadriradiate; c, dermal triradiate. x 150.

The present species differs from the only other known species of the genus, L.
clathria, Haeckel, in external form, in the absence of small radiates from the gastral and
chamber layers, and in the presence of the large rhabds.

CALCAREA 261

Genus Leucettusa, Haeckel

Leucettusa haeckeliana (Polejaeff).

Leucetta haeckeliana, Polejaeff, 1883, p. 69, pi. ii, fig. 6, pi. viii, figs. 1-6 ; Leucettusa haeckeliana,
Dendy and Row, 1913, p. 739.

Occurrence. St. WS 84: Falkland Islands, 75-74 m.; St. WS 243: Falkland Islands, 144-141 m.

Distribution. Kerguelen.

Leucettusa simplicissima, sp.n. (Plate XL VIII, fig. 3).

Holotype. B.M. 28. 2. 15. 35.

Occurrence. St. WS 84 : Falkland Islands, 75-74 m.

Diagnosis. Sponge tubular ; skeleton of chamber layer composed of triradiates and
quadriradiates of fairly constant size; cortical skeleton composed of a dense mass of
large triradiates; without special gastral skeleton.

Remarks. The species is represented by a quantity of branching tubes, each one
being cylindrical and slightly flattened, with a single apical oscule. The tubes appear to
have grown erect but now there is nothing to show their method of attachment. The
largest are 3 cm. high and 7 mm. in diameter. There is no oscular fringe. The cortical
and medullary layers are well marked, the latter being the thicker. The flagellated
chambers are elongate, reaching from the gastral surface to just beneath the dermal
cortex.

The chamber layer contains small triradiates and quadriradiates of fairly constant size
scattered sparsely between the chambers. These are of the normal type and the rays of
a typical example measure 0-036 by o-oo6 mm. Occasionally these spicules may suff'er
reduction to two rays.

The cortical skeleton consists of a mass of large irregularly placed triradiates of fairly
constant size, several layers thick, the rays of which may reach as much as 0-21 mm. in
length and o-oi8 mm. in thickness.

There is no special gastral skeleton, although occasionally triradiates similar in all
respects to those of the chamber layer may be found lying tangentially in the wall of
the gastral cavity.

The characteristic feature of this species is the simplicity of its skeleton, and in this
it diflFers markedly from all the known species of Leucettusa. If anything is to be said of
its relations to other species it is that L. simplicissima is more nearly like L. tubulosa,
Dendy, from New Zealand, in external form, but approaches L. haeckeliana (Polejaeff),
from Kerguelen in spiculation. In fact, it differs from the latter mainly in the absence
of the large quadriradiates.

262 DISCOVERY REPORTS

Family GRANTIIDAE
Genus Grantia, Fleming
Grantia cirrata (Jenkin), var. aurorae, Dendy.

G. cirrata, var. aurorae, Dendy, 1918, p. 11, pi. I, figs, 4, 9.
Occurrence. St. 53 : Falkland Islands, 0-2 m.; St. WS 33 : South Georgia, 130 m.
Distribution. Off Gates Land.

Grantia cirrata (Jenkin), var. tenuipilosa, n.

Type of var. B.M. 28. 2. 15. 708.

Occurrence. St. 53 : Falkland Islands, 0-2 m.; St. WS 85: Falkland Islands, 79 m.

Remarks. The two specimens are very like the typical form in external appearance,
in the arrangement of the skeleton and in the size of the spicules. The only feature by
which this variety may be distinguished is the form of the long oxea echinating the sur-
face. These are usually 1-2 mm. in length but rarely exceed 0-012 mm. in diameter, and
are only slightly, though very irregularly, curved.

The typical form of the species, together with the var. aurorae and the present variety,
are all very much alike except for the form of the dermal oxea, and it is probable that a
distinction into varieties is unnecessary. But until we can see for certain that the dermal
oxea are subject to considerable variations in size, I prefer to consider the present
specimens as constituting a separate variety of the species.

Order TETRAXONIDA
Sub-order HOMOSCLEROPHORA
Family PLAKINIDAE

Genus Plakina, Schulze

Plakina monolopha, Schulze.

P. monolopha, Schulze, 1880, p. 407, pi. xx, figs. 1-7, pi. xxii, figs. 22-29; GraefFe, 1882, p. 319;
SoUas, 1888, p. 278; Topsent, 1890, p. 231 ; Lendenfeld, 1894, p. 96, pi. iii, fig. 46; Topsent,
1896, p. 549, pl.xxi,figs. 1,2, pi. xxii, fig. 12; Thiele, 1898, p. 28, pi. v, fig. 13, pi. vii, fig. 11 ;
Lendenfeld, 1903, p. 120; id., 1907, p. 336; Topsent, 1925, p. 629; Burton, 1929, p. 414.

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Distribution. North Atlantic ; Mediterranean ; Japan ; Antarctic.

Plakina trilopha, Schulze.

P. trilopha, Schulze, 1880, p. 407, pi. xx, figs. 1-7; Placina trilopha, Sollas, 1888, p. 279;
Topsent, 1890, p. 232; Lendenfeld, 1894, p. 98, pi. iii, fig. 45, pi. iv, fig. 54; Topsent, 1896,
p. 555, pi. xxi, figs. 4-6 ; Lendenfeld, 1903, p. 121 ; Plakina trilopha, subsp. aiitarctica, id., 1907,
p. 333, pi. XXV, figs. 30-43; Placina trilopha, Babic, 1921, p. 17; id., 1922, p. 291; Plakina
trilopha, Burton, 1929, p. 414.

TETRAXONIDA 263

Occurrence. St. 140: South Georgia, 122-136111.; St. 152: South Georgia, 245 m.; St. 190:
Palmer Archipelago, 93-130 m.; St. WS 80: Falkland Islands, 152-156 m.

Distribution. Mediterranean ; Wilhelm Land.

Sub-order STREPT ASTRO SCLEROPHORA

Family PACHASTRELLIDAE

Genus Pachastrella, Schmidt

Pachastrella monilifera, Schmidt.

(For synonymy see Burton, 1926, p. 9.)

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Distribution. Algeria; Mediterranean; Coast of Asturias; Cape Bojeador; Tristan
da Cunha ; Natal ; Agulhas Bank.

Genus Poecillastra, Sollas

Poecillastra compressa (Bowerbank).

(For synonymy see Burton, 1930, pp. 487, 498.)

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Remarks. In two of the specimens the calthrops have undergone a considerable re-
duction and distortion, one or more of the rays being frequently reduced and those
remaining being variously knobbed at the ends or ornamented with tubercles and
spines. The condition closely resembles that of Stelletta pathologica, Schmidt, or Pach-
astrella abyssi (cf. Sollas, 1888, pi. xi, figs. 15-19). The distortion has extended to the
microxea also, although not to the same extent, and these are thicker, less markedly
roughened, but more strongly tylote than usual.

Distribution. Arctic ; North Atlantic (east and west coasts) ; Mediterranean.

Sub-order ASTRO SCLEROPHORA
Family GEODIIDAE
Genus Geodia, Lamarck

Geodia magellani (Sollas).

Cydonium magellani, Sollas, 1886, p. 197 ; id., 1888, p. 221, pi. xxi, figs. 1-14; Geodia magellani,
Lendenfeld, 1903, p. 107; Thiele, 1905, p. 408.

Occurrence. St. WS 81 : Falkland Islands, 81-82 m.

Distribution. Tom Bay, Caibuco and Admiralty Straits (South America).

4-z

264 DISCOVERY REPORTS

Family ERYLIDAE
Genus Erylus, Gray
Erylus discophorus (Schmidt).

(For synonymy see Lendenfeld, 1903, p. 87.)

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Remarks. The present specimen forms a thin brown crust on the surface of a piece
of rock. The skeleton is composed as follows: amphistrongyles, 075 by o-oiamm. ;
dichotriaenes, shaft 0-4 mm. long, primary cladi, 0-15 mm. long, and secondary cladi,
o-i8mm. long; sterrasters, 0-15 by 0-09 by 0-012 mm. ; oxyasters, with 5-12 rays,
o-024-o-o6 mm. in diameter, the larger spicules having usually the smaller number of
rays ; centrotylote microxea, 0-045 by 0-002 mm. In spiculation therefore the specimen
agrees closely with the Mediterranean examples of the species except for the presence
of strongyles instead of oxea. On the other hand, I am strongly inclined to the view that
the spicules of the various species of Erylus may vary a great deal more than has hitherto
been suspected, judging by my observations on a number of specimens of this genus
from the Indo-Pacitic.

The following species from the Atlantic, and adjacent waters, E. discophorus, E. tnam-
millaris, E. transiens, E. jnnmmilifer, E. grafiuloris and E. chavesi, and E. oxyasier from
the Galapagos Islands, all agree in possessing oxea, or their derivatives, dichotriaenes,
sterrasters, oxyasters and microxea. The differences between them amount to no more
than small, and probably unimportant, differences in external form and spiculation, and
it is highly probable that they all represent nothing more than a single species.

Distribution. Mediterranean; Spain.

Sub-order SIGMATOSCLEROPHORA
Family TETILLIDAE

Genus Tetilla, Schmidt

Tetilla leptoderma (SoUas).

(For synonymy see Burton, 1929, p. 418.)

Occurrence. St. 39: South Georgia, 179-235 m.; St. 140: South Georgia, 122-136 m.; St. 148:
South Georgia, 132-148 m.; St. 170: Clarence Island, 342 m.; St. WS 27: South Georgia, 106-
109 m.; St. WS 218: north of Falkland Islands, 3 1 1-247 m.; St. WS 225: Falkland Islands,
162-161 m.; St. WS 243 : north-east of Falkland Islands, 144-141 m.

Distribution. Kerguelen; South America; Victoria Land.

Genus Cinachyra, Sollas
Cinachyra antarctica (Carter).

(For synonymy see Burton, 1929, p. 419.)
Occurrence. St. 42 : South Georgia, 120-204 m. ; St. 175 : South Shetlands, 200 m.
Distribution. Graham Land; Victoria Land; Wilhelm Land.

TETRAXONIDA 265

Cinachyra barbata, Sollas.

(For synonymy see Burton, 1929, p. 419.)

Occurrence. St. 27: South Georgia, no m.; St. 42: South Georgia, 120-204 m.; St. 123: South
Georgia, 230-25001.; St. 140: South Georgia, 122-136 m.; St. 153 : South Georgia, 106 m.; St. 170:
Clarence Island, 342 m.; St. 175: South Shetlands, 200 m.; St. WS 33: South Georgia, 130 m.;
St. WS 42: South Georgia, 198 m.; St. MS 71 : South Georgia, 120-60 m.

Distribution. Kerguelen ; Graham Land ; Victoria Land ; Wilhelm Land.

Cinachyra coactifera (Lendenfeld).

Tethya coactifera, Lendenfeld, 1906, p. 75, pi. ix, figs. 1-16, pi. x, figs, i-io; Cinachyra
coactifera, Burton, 1929, p. 396.

Occurrence. St. 39: South Georgia, 179-235 m.

Remarks. The eleven specimens range from 6 to 12 cm. in diameter. The majority
are without conspicuous surface pile or basal tuft, but in two of them both these
features are well developed. The porocalices in the dried sponge are inconspicuous and
might be easily mistaken for small and simple oscules, and it is probably on this account
that the species was first referred to Tethya (= Tetilla).

Distribution. Kerguelen.

Family HAPLOSCLERIDAE
Genus Haliclona, Grant
Haliclona nodosa (Thiele).

Reniera nodosa, Thiele, 1905, p. 461, figs. 7, 77.
Occurrence. St. 145 : South Georgia, 26-35 m.
Distribution. Calbuco, Chile.

Haliclona chilensis (Thiele).

Reniera chilensis, Thiele, 1905, p. 467, figs. 5, 84.

Occurrence. St. WS 84: Falkland Islands, 75-74 m.

Remarks. A small tubular specimen agreeing with the holotype in all respects, except
that it is dark brown in colour and possesses spicules measuring 0-12 by 0-009 "^"^- ^^
against 0-165 by 0-0 1 mm. in the type.

Distribution. Calbuco, Chile.

Haliclona variabilis (Thiele).

Acervochalina variabilis, Thiele, 1905, p. 477, figs. 17, 33, 96.

Occurrence. St. 53 : Falkland Islands, 0-2 m.

Remarks. Two of the specimens are almost identical in every way with the holotype
(Thiele, loc. cit. pi. xxviii, fig. 17). The four other specimens, which resemble the

266 DISCOVERY REPORTS

holotype in skeleton, colour, texture and appearance of the surface, are irregularly
pyriform with large oscules (Plate LI, figs, i, 2).
Distribution. Punta Arenas and Juan Fernandez.

Haliclona penicillata (Topsent).

Reniera pefiicillata, Topsent, 1908, p. 20, pi. ii, figs. 2, 3, pi. v, fig. 14; Chalina penicillata.
Burton, 1929, p. 420.
Occurrence. St. 181 : Palmer Archipelago, 160-335 m.; St. WS 108: Falkland Islands, 1 18-120 m.

Remarks. The two specimens agree closely with the holotype except that in one of
them the oxea are mostly centrotylote and a few are modified to styli.

Dlstribution. Graham Land; Victoria Land.

Haliclona gaussiana (Hentschel).

Siphonochalina gaussiana, Hentschel, 1914, p. 136, pi. iv, fig. 17, pi. viii, fig. 17.
Occurrence. St. 42: South Georgia, 120-204 m.

Remarks. The present specimen differs from the holotype only in the absence of
polyspicular fibres and in the dimensions of the spicules. These are: oxea 0-15 mm.
long, and toxa not exceeding 0-04 mm. in length.

The skeleton is a unispicular (renieroid) network with triangular mesh, and the outer-
most spicules hispidate the surface. There is no special dermal skeleton. The species is
therefore a typical Haliclona.

Distribution. Wilhelm Land.

Haliclona tubuloramosa (Dendy).

Gellius tubuloramosus, Dendy, 1924, p. 323.

Occurrence. St WS 243 : Falkland Islands, 144-141 m.

Remarks. A solitary example of this species is present. I have made a specially
careful comparison with the holotype, since species common to the Falklands and to
New Zealand are of rare occurrence.

Distribution. New Zealand.

Haliclona conica (Br0ndsted).

Pachychalina conica, Brondsted, 1924, p. 454, fig. 12.

Occurrence. St. WS 243 : Falkland Islands, 144-141 m.

Remarks. Through the kindness of Dr Br0ndsted I have been able to compare this
specimen with the holotype and find an exact agreement in every way.

Distribution. New Zealand.
Haliclona bilamellata, sp.n. (Plate XLVIII, figs. 5-9; Plate XLIX, figs. 1-3; Plate L,
fig. 2; Fig. 6).

Holotype. B.M. 28. 2. 15. 121.

TETRAXONIDA 267

Occurrence. St. WS 27 : South Georgia, 106-109 m. ; St. WS 83 : Falkland Islands, 137-129 m. ;
St. WS 86: Falkland Islands, 151-147 m.; St. WS 87: Falkland Islands, 96-127 m.; St. WS 99:
Falkland Islands, 251-225 m.; St. WS 243: Falkland Islands, 144-141 m.; St. WS 247: Falkland
Islands, 172 m.

Diagnosis. Sponge tubular, infundibular or bilamellate; texture elastic, very
soft, somewhat friable ; outer surface minutely hispid, shaggj^ often thrown into irregular
folds; inner surface even, minutely hispid, bearing numerous exhalant apertures
varying up to 3 mm. in diameter; colour, in spirit, greenish yellow or brown; skeleton
a regular isodictyal reticulation with bi- to trispicular primary fibres connected by trans-
verse single spicules at regular intervals; oxea measuring o-i6 by 0-007 ""'""'•

Remarks. The holotype (Plate XLVIII, figs. 5, 6) is an incomplete sponge,
apparently infundibular in life, bearing a strong resemblance to Calyx (Gellius)
imperialis (Dendy)i from New Zealand, with which at first sight it might be easily
confused. All the remaining specimens are damaged, though some only to a slight
extent, and they vary from tubular to bilamellate.

The characteristic feature of the species is the external form. In many respects this is
like that of H. {Reniera) Scotti (Kirkpatrick), but the sizes of the respective spicules differ
too much to allow of a specific identity at the moment.

There are a number of other specimens which, in spite of the differences in external
form, appear to belong to the same species. All have the same soft, friable texture, the
same appearance and a very similar type of skeleton. The diversity in external form
exhibited by them furnishes a good example of the extent to which the Chalinine sponges
may vary.

The external form of these specimens is as follows :

Specimen i. More or less pyriform with a deep cloaca, the walls of which are marked
with large exhalant apertures. Surface minutely conulose. Colour dark brown
(Fig. 6 a).

Specimen 2. Stipitate and tubular, with deep cloaca, the walls of which are marked
with large exhalant apertures. Surface markedly conulose, particularly near the top of
the sponge. Colour dark brown (Fig. 6 b).

Specimen 3. Sponge tubular in the lower two-thirds, infundibuliform above. Surface
of lower two-thirds with long, slender spinose processes. Colour dark brown

(Fig. 6 c).

Specimen 4. Sponge tubular. Surface covered with spinose processes (Plate XLIX,

fig. I).

Specimen 5. Sponge tubular. Surface covered with low, rounded, irregular ridges
(rather like Haliclona dancoi (Topsent) in appearance). Colour yellow (Plate XLVIII,

fig- 9)-

1 Gellius imperialis, Dendy, is a typical Calyx, differing from C. arcuarius (Topsent), described on page 277,
in its external form and in the substitution of sigmata for toxa. In all other respects the two species are
practically identical.

268

DISCOVERY REPORTS

Specimen 6. Infundibular, 15 cm. high, 9 cm. across. Outer surface thrown into
strongly developed rounded ridges. Colour dark brown (Plate XLVIII, figs. 5-8).

Specimen 7. Flabelliform, one surface pore-bearing with spinose processes : the other
bearing oscules. Colour yellow (Plate XLIX, figs. 2-3).

With the exception of specimen 7, all have a deep cloaca with oscules measuring up to
4 mm. in diameter. The series also shows that the species has a range of form very like

Fig. 6. Series showing external form in the smaller specimens oi Haliclona bilamellata, sp.n.
Dotted lines indicate shape and position of cloaca.

that of Haliclona dancoi {sensu Burton, 1929, p. 420), and differs from that species only
in texture and the size of the spicules.

While the external form is similar in every case, two types of skeleton are found.
In some specimens it is chalinoid and consists of a quadrangular network in which the
primary fibres contain a double row of spicules coated with a thin layer of spongin and
the secondary fibres contain one spicule only, the primary fibres being only one spicule
length apart. The oxea in these forms vary from 0-132 to o-i8 by 0-007 to o-oo8 mm.
In the remaining specimens, the skeleton is pachychalinoid and consists of a quadrangular
network of stout primary fibres, containing from eight to ten spicules serially arranged,
but with very little spongin, and secondary fibres containing three or four spicules. The
meshes of the network are from 0-3 to 0-4 mm. across and the spicules 0-240 by 0-0 1 mm.
This separation into two types is, however, more apparent than real, for in the second
group the skeleton may, in places, be identical with that of the first group, so that we

TETRAXONIDA

269

have here definite evidence of the transition from a chahnoid to a pachychalinoid
skeleton in one and the same species.

The species is very Hke Haliclona flaccida (Topsent), but this has spicules measuring
0-56 to 0-57 by o-oiy mm. As recorded above, however, there is considerable variation
in the size of the spicules of//, bi/amellata, and it is conceivable that the new species and
//. flaccida may even yet prove to be identical ; and possibly also synonymous with
H. dancoi.

There are also four specimens, from Sts. WS 27, WS 86 and WS 99, which have oxea
measuring 0-35 by 0-009 "^™- ^^^ ^ skeleton, composed of primary lines 8-12 spicules
thick and secondary lines 2-4 spicules thick, which gives a more fibrous appearance to
the sponge (cf. Plate XLIX, figs. 2, 3). In external form, however, they show the same
transitions from tubular to infundibular. One specimen in particular is worth attention,
from St. WS 86. This is sub-infundibular, rather like a cup with a cut extending from
rim to base (Fig. 6 d).

Haliclona stephensi, sp.n.

Holotype. B.M. 28. 2. 15. 102.

Occurrence. Saldanha Bay (beach).

Diagnosis. Sponge thinly encrusting (always ?) ; surface smooth, even, minutely
hispid ; texture friable ; colour, in spirit, yellow ; skeleton an irregular network of single
spicules, perhaps quadratic in places, with ill-defined polyspicular bundles usually
running towards surface of sponge; oxea slightly curved, abruptly pointed, o-i2 by
o-oo6 mm.

Remarks. The single specimen, an incrustation several square centimetres in area,
on a pebble, appears to agree closely with Reniera sp. from Reit's Bay described by
Stephens (1915, p. 464).

Haliclona petrosioides, sp.n. (Fig. 7).

Holotype. B.M. 28. 2. 15. 434.

Occurrence. St. 6: Tristan da Cunha,
80-140 m.

Diagnosis. Sponge encrusting to
massive and spreading ; surface uneven,
minutely hispid ; texture friable ; oscules
and pores not apparent; colour, in
spirit, white; skeleton a regular iso-
dictyal network with primary and
secondary lines five spicules thick,
and with numerous isolated spicules ,

scattered between meshes of main ^.^ ^ Haliclona petrosioides, sp.n. a, section at righ
skeleton; spongin not apparent; oxea angles to surface, showing structure of main skeleton

slightly curved, 0-154 by 0-007 "^'^- ^ ^°' *' ^" oxeote, x 400.

270

DISCOVERY REPORTS

Remarks. The species appears to differ from all other species of Haliclona in the
regularity and compactness of the skeleton, combined with the small size of the spicules.
It appears to be most closely related to several of the varieties of Petrosia similis, Ridley
and Dendy, from the Indian Ocean and the Kerguelen area.

Genus Haliclonissa, gen.n.

Genotype. H. verrucosa, sp.n.

Diagnosis. Haploscleridae with skeleton composed of fibres running vertically to
surface, often anastomosing to form an irregular network, with numerous isolated
spicules scattered between ; without special dermal skeleton.

Remarks. The genus appears to be intermediate between Haliclona and Microxina.

Haliclonissa verrucosa, sp.n. (Plate LI, fig. 3 ; Fig. 8).

Holotype. B.M. 28. 2. 15. 473."

Occurrence. St. 180: Palmer Archipelago, 160 m.

Diagnosis. Sponge composed of stout, irregularly cylindrical branches, bearing a
few conspicuous oscules 10 mm. in diameter ; oscules leading out from shallow, capacious
cloacae ; surface verrucose ; texture friable ; colour, in spirit, pale yellow ; skeleton com-
posed of diffuse fibres, 6-20 spicules thick, running vertically to surface and forming in

Fig. 8. Haliclonissa verrucosa, sp.n. Section at right angles to
surface, to show structure of skeleton, x 10.

places an irregular network, with numerous isolated spicules scattered between ; spongin
inconspicuous; oxea slightly curved, tapering gradually to a point at each end, 0-5 by
0-017 "-'i^-

Remarks. The holotype measures 13 cm. long by approximately 5 cm. in diameter.
In places it is almost regularly cylindrical, but is mainly somewhat flattened laterally.
It resembles in appearance some of the Antarctic species of Isodiclya and also, to a lesser
extent, HemigelUus rudis (Topsent), but the structure of the skeleton shows that it be-
longs to a different genus.

TETRAXONIDA 271

Haliclonissa sacciformis, sp.n. (Plate XLVIII, fig. 10; Fig. 9).

Holotype. B.M. 28. 2. 15. 621.

Occurrence. St. WS 223: Falkland Islands, 11401.

Diagnosis. Sponge sacciform, sessile, erect; surface soft to touch, minutely pilose,
having an appearance of fine velvet; exhalant
apertures opening into a deep central cloaca ; tex-
ture very soft and compressible ; colour, in spirit,
yellowish grey; skeleton of stout multispicular
primary fibres running vertically to surface, con-
nected by diffuse and irregular secondary fibres,
with numerous isolated spicules scattered between ;
spicules slender oxea measuring 0-32 by 0-007 mm.

Remarks. The single specimen is laterally
compressed, about 13 cm. high, 9 cm. across and
2-5 cm. wide, with a wide mouth at the apex
leading into a deep and capacious cloaca. It re-
sembles Chalina spoiigiosissima, Topsent closely
in texture, but differs in the external form and the
size of the spicules. The primary lines of the
skeleton, too, are multispicular instead of bi- or
trispicular.

Genus Microxina, Topsent

Microxina benedeni (Topsent) (Plate L, fig. i ; Fig. 10).

Gellius benedeni, Topsent, 1901, p. 6; GelUodes benedeni, id., loc. cit., p. 16, pi. ii, fig. 3, pi. iii,
fig. 5; Microxina charcoti, Topsent, 1916, p. 170; id., 1917, p. 72, pi. i, fig. 3, pi. ii, fig. 3, pi. vi,
fig. 17; GelUodes benedeni, \z.x . fortior , id., 1917, p. 75, pi. ii, fig. i, pi. vi, fig. 22; G. benedeni.
Burton, 1929, p. 423; Microxina charcoti, id., loc. cit., p. 423.

Occurrence. St. 39: South Georgia, 179-235 in.; St. 149: South Georgia, 200-234 m.; St. 190:
Palmer Archipelago, 315 m.; St. WS 81 : Falkland Islands, 81-82 m.; St. WS 99: Falkland Islands,
251-225 m.; St. WS 248: Falkland Islands, 210-242 m.

Remarks. The British Museum
now possesses a considerable
number of specimens of GelUodes
benedeni and Microxina charcoti,
and from these it is clear that the
two species are alike in external
form and spiculation, except that
the first possesses sigmata only,

with the addition of raphides in ^

the var.fortior, and the second has pig. 10. Microxina benedeni (Topsent), showing transition
microxea only. This being the from sigmata to microxea.

5-2

Fig. 9. Haliclonissa sacciformis, sp.n. Sec-
tion at right angles to surface, x 20.

272 DISCOVERY REPORTS

case, the suspicion that the two species are identical is unavoidable. The present speci-
mens offer abundant evidence that this is in fact the case.

The Discovery collection contains eight specimens ; five of these had been provision-
ally identified by me as Microxina charcoti and three as Gelliodes benedeni. On re-
examination, I found that several of the former contained isolated long, slender raphides,
and this was also the case in each specimen of the second species, but they were never
grouped in dragmata as in Gelliodes benedeni, var. fortior. Secondly, one specimen of
G. benedeni contained sigmata only, a second contained sigmata and an occasional
microxeote, and the third contained all transitions from sigmata to microxea (Fig. lo).
In addition, it is worth noting that the sigmata in Gelliodes benedeni are frequently
centrotylote, like the microxea of Microxifia charcoti. It is evident therefore that we
have to deal here with a single species, the microscleres of which may be sigmata and/or
microxea, to which raphides, either isolated or in dragmata, may be added.

Distribution. Graham Land ; Victoria Land ; Wilhelm Land.

Genus Hemigellius, gen.n.

Genotype. Gellius nidis, Topsent.

Diagnosis. Haploscleridae with skeleton composed of a coarse and irregular reticula-
tion of oxea formed by multispicular primary fibres, running vertically to surface, and
uni- or multispicular connecting fibres; special dermal skeleton formed by ends of
primary fibres spreading out beneath ectosome in a paniculate manner; microscleres,
when present, sigmata.

Remarks. The original description gives little information regarding the structure of
the skeleton in Gellius riidis (see Topsent, 1901, p. 14) ; but the surface is said to be finely
hispid, which at least indicates the absence of a tangential dermal skeleton. Kirk-
patrick's (1908) specimens assigned to this species resemble the holotype sufficiently,
however, to justify the framing of a diagnosis, as has been done here, on the characters
observed in his specimen.

The genus is closely allied to Haliclona.

Hemigellius rudis (Topsent).

Gellius rudis, Topsent, 1901, p. 6; id., 1901, p. 14, pi. i, fig. 9, pi. iii, fig. 4; id., 1907, p. 77,
pi. iii, fig. 2; G.fimbriatus, Kirkpatrick, 1907, p. 86; G. rudis, id., 1908, p. 45, pi. xvii, fig. i ;
G. fimbriatus, id., loc. cit., p. 46, pi. xvii, fig. 2, pi. xxiv, fig. 2; G. rudis, Hentschel, 1914,
p. 287; Burton, 1929, p. 422.

Occurrence. St. 39: South Georgia, 179-235 m.; St. 42: South Georgia, 120-204 m.; St. 170:
Clarence Island, 342 m.

Remarks. In a former work {loc. cit.) I included Gellius pilosus, Kirkpatrick, and
G. tenelltis, Topsent, as synonyms of this species. As shown on p. 276 the latter is
now included in the genus Adocia. As regards G. pilosus, Kirkpatrick, there can be

TETRAXONIDA 273

little doubt that it is congeneric with HetnigeUhis nidis (Topsent), but whether it is con-
specific is not so certain, and it is perhaps preferable to recognize it as a distinct species.
Distribution. Graham Land; Victoria Land; Wilhelm Land.

Hemigellius pachyderma, sp.n. (Plate XLVIII, fig. 4; Fig. 11).
Holotype. B.M. 28. 2. 15. 723.
Occurrence. St. WS 210: Falkland Islands, i6i m.

Diagnosis. Sponge consisting of massive cylindrical branches ; surface even, minutely
hispid; texture friable; oscules conspicuous, 3-4 mm. in diameter, scattered irregularly

Fig. II. Hemigellius pachyderma, sp.n. a, section at right angles to surface, showing
structure of main skeleton, x 30; b, oxeote, x 400.

over surface ; pores apparently evenly distributed ; subdermal cavities large ; colour, in
spirit, greyish yellow; skeleton a reticulation consisting of stout spicule bundles, 5-7
spicules thick, running towards surface and anastomosing slightly, connected by
isolated spicules set at varying angles; main vertical fibres ending immediately beneath
surface in a dense subdermal palisade ; oxea slightly curved, measuring 0-28 by o-oi4mm.

274 DISCOVERY REPORTS

Remarks. The holotype consists of two fragments, the one a small portion of a
cylindrical branch, and the other a cylindrical branch, 2 cm. in diameter, which bi-
furcates at one end into two short branches rounded at the ends. The second specimen
is in fact almost identical in appearance with the specimen figured by Kirkpatrick (1908,
pi. xvii, fig. i) under Gellius riidis, Topsent.

Hemigellius pachyderma, sp.n., diff"ers from H. rudis (Topsent) mainly in the size of
the spicules, in the absence of sigmata, and in the dense aggregation of spicules im-
mediately beneath the dermis. But all these characters are comparative, and it may
possibly be shown eventually that the two species are synonymous. In any case, the
dermal palisade may certainly be accepted as the homologue of the paniculate surface
brushes in H. rudis.

Genus Adocia, Gray

Adocia carduus (Ridley and Dendy).

Gellius carduus, Ridley and Dendy, 1886, p. 333 ; G. laevis, id., loc. cit., p. 333 ; G. carduus, id.,
1887, p. 39, pi. viii, fig. 3, pi. xiii, fig. 7; G. carduus, var. magellanica, id., loc. cit., p. 40,
pi. xiii, fig. 6; G. laevis, id., loc. cit., p. 40, pi. xiii, fig. 8.

Occurrence. St. 160: near Shag Rocks, 177 m.; St. 177: South Shetlands, 1080 m. ; St. WS 81 :
Falkland Islands, 81-82 m.; St. WS 83 : Falkland Islands, 137-129 m.

Remarks. Judging from the observations made on the new material, there is no
possibility of distinguishing between Gellius carduus and its variety magellanica and
G. laevis. The external form of the new specimens resembles, in each case, that of
G. carduus, var. magellanica. The megascleres, which vary from oxea to incipient
strongyla, measure from 0-3 to 0-45 mm. long, and the sigmata vary from 0-015 to
0-02 mm. chord.

Distribution. Crozet Island; Prince Edward Island; Marion Island; Magellan
Straits ; mouth of the Rio de la Plata.

Adocia glacialis (Ridley and Dendy).

Gellius glacialis, Ridley and Dendy, 1886, p. 333; id., 1887, p. 41, pi. viii, fig. u, pi. xiii, figs.
I, 15, 19; G. glacialis, var. nivea, id., loc. cit., p. 42, pi. viii, fig. 8, pi. xiii, figs. 4, 12; Kirk-
patrick, 1908, p. 49, pi. xvii, fig. 4; Burton, 1929, p. 397.

Occurrence. St. 140: South Georgia, 122-136 m.; St. 190: Palmer Archipelago, 93-130111.;
St. WS 82: Falkland Islands, 140-144 m.; St. WS 83 : Falkland Islands, 137-129 m.

Remarks. Comparison of the present specimens with those in the Swedish Antarctic
Collection and with the types of both the species and its variety has shown that the
var. nivea cannot be maintained. The question will be dealt with fully in my forth-
coming report on the sponges of the Swedish Antarctic Expedition.

The sigmata in one of the present specimens resemble those of A. {Gellius) flagellata
(Ridley and Dendy).

Distribution. Agulhas Bank ; Prince Edward Isle ; Victoria Land.

TETRAXONIDA 275

A comparison between A. carduus and A. glacialis
On comparing the types of Gellhis carduus, G. cardials, var. magellanica, G. laeris,
G. glacialis and G. glacialis, var. nivea, I find that there is no real difference between
them in external form or in the structure of the skeleton. The only distinction that can
be made rests on the differences in the dimensions of the spicules, especially of the
microscleres. I have already suggested that this group of three species and two varieties
represents no more than two species, so that the differences between these two species
may be expressed in the following way :

G. carduus: Oxea o-49-o-6 mm. long; sigmata 0-02 mm. chord.
G. glacialis: Oxea o-53-o-65 mm. long; sigmata up to 0-145 mm. chord.
It is evident from the examination of specimens which have come to hand since these
species were first described that the sigmata in G. glacialis rarely measure so much as
0-145 mm. Sigmata are always present measuring 0-02 mm. and, in addition, larger
sigmata measuring o-o6, 0-07, o-o8, up to 0-12 mm. or so. Further, the larger and
smaller sigmata are not differentiated into two distinct categories, but are connected by
numerous forms intermediate in size. I suspect therefore that the species called here
Adocia carduus is merely a form oi A. glacialis in which the larger sigmata have dropped
out. It may be necessary to regard them ultimately as conspecific, but I would like to
have more material for examination before taking this step.

There is, however, another point to be noted : that in the present collection there is
nearly a score of specimens from Sts. 160 and 177 which have the same general ap-
pearance as A. carduus and A. glacialis, the same arrangement of the skeleton, and the
same type of megascleres (i.e. oxea varying from abruptly pointed to sub-strong>-lote).
All these specimens are clearly conspecific and obviously belong either to A. carduus or
to A. glacialis, but to which it is quite impossible to say since none of them possesses
a single microsclere. It seems to me therefore that in this group of specimens from
Sts. 160 and 177, we have a number of individuals in which the reduction of the
microscleres, believed to have taken place in A. carduus, has been taken a step further
by the complete suppression of all microscleres.

Adocia siphonella (Thiele).

Reniera siphonella, Thiele, 1905, p. 460, figs. 75, 97-99.

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Remarks. The spicules in this specimen are usually strongyla, with a few oxea,
measuring 0-32 by 0-018 mm. Only a few fragments are to hand, but from these it
appears that, if my identification is correct, the species belongs to Adocia. Through the
kindness of Dr Arndt, of the Berlin Museum, I have been able to compare the present
specimen with one of the types, but as this also is in a fragmentary condition it does not
help in determining the generic position of the species.

Distribution. Calbuco, Chile.

276 DISCOVERY REPORTS

Adocia cucurbitiformis (Kirkpatrick).

Gellius cucurbitiformis, Kirkpatrick, 1907, p. 288; id., 1908, p. 48, pi. xvii, fig. 5, pi. xxiv, fig. 5.
Occurrence. St. 180: Palmer Archipelago, 160-330 m.
Distribution. Victoria Land.

Adocia tremulus (Topsent).

Gellius tremulus, Topsent, 1916, p. 171 ; /(/., 1917, p. 79, pi. i, fig. 7, pi. vi, fig. 20.
Occurrence. St. 175 : South Shetlands, 200 m.

Remarks. The specimen is quite typical except that no raphides are present.
Distribution. Graham Land.

Adocia tenellus (Topsent).

Gellius tenellus, Topsent, 1916, p. 171 ; id., 1917, p. 80, pi. vi, fig. 23.

Occurrence. St. WS 88: Falkland Islands, 96-127 m.

Remarks. There is a doubt about the locality in this case so that it would be in-
advisable to say much concerning the bearing that this specimen may have on the
value of the species or its distribution. The specimen itself is massive, with spicules of
the same shape and dimensions as those of the holotype of Gellius tenellus, and the
skeleton is a unispicular network throughout.

Distribution. Petermann Islands.

"Gellius fibulatus."

Occurrence. St. WS 82: Falkland Islands, 140-144 m.

Remarks. The specimen recorded under this title is a small massive sponge, of a very
friable texture, which reached me in a fragmentary and slightly macerated condition.
It appears to belong to Adocia, but of this it is impossible to be sure. The skeleton is a
unispicular, occasionally bispicular, network of oxea measuring 0-12 by o-oo8 mm.,
slightly curved and gradually and evenly pointed towards each end. The sigmata are of
the usual type and measure up to 0-05 mm. chord.

Renter a fibulatus, Schmidt (= Gellius fibulatus, Autt.), is a species of which we know
nothing with certainty. The holotype has not been seen since Schmidt first described
it, and the original description is hopelessly inadequate. All we know is that its skeleton
was composed of oxea and sigmata, but the sizes of these respective spicules are un-
known. Since Schinidt first established the species, it has been used by succeeding
authors for almost any sponge possessing small oxea and sigmata, so that, in the con-
ventional sense, the present sponge belongs to Schmidt's species.

The only reason for recording such a poorly preserved specimen is that it is from a
locality from which we have very few records and there is just the hope that, when the
so-called species " Gellius fibulatus" has been correctly split into its component parts,
we may be able to determine this particular specimen with certainty.

TETRAXONIDA 277

Genus Calyx, Vosmaer

Calyx arcuarius (Topsent).

Gellius arcuarius, Topsent, 1913, p. 638, pi. vi, fig. 11; Calyx stipitattis, id., 1916, p. 171;
id., 1917, p. Si, pi. iv, fig. 13, pi. vi, fig. 24; Burton, 1929, p. 422, pi. v, figs. 5, 6.

Occurrence. St. 39: South Georgia, 179-235 m.; St. 148: South Georgia, 132-148 m.; St. 149:
South Georgia, 200-234 m.; St. 159: South Georgia, 160 m.; St. 160: Shag Rocks, 177 m.; St. 163 :
South Orkneys, 18-27 "^-^ ^'^- '75 • South Shetlands, 200 m.

Remarks. There are several specimens which correspond closely with the original
description of Calyx stipitattis and with the specimens recorded by me under that name
in 1929. They are all thinly flabellate, the largest 40 cm. high and about the same across,
with evenly rounded margins, often with rounded outgrowths springing from the main
body. The surface is even, glabrous or only minutely hispid. The skeleton is composed
of stout bundles of oxea, usually 20-40 deep, running from base to margin, branching
and anastomosing, with a confused triangular network of single spicules filling the
spaces between the bundles. The dermal skeleton is a tangential unispicular network, in
which the spicules are often irregularly set so that the ends frequently project slightly at
the surface. In this way, although there is a special dermal skeleton comparable with
that found in Adocia, the surface is minutely hispid. The oxea vary from o- 18 to 0-28 mm.
long by o-oii to o-oi8 mm. thick.

The external form varies from branching, as in the specimen figured by me {loc. cit.,
pi. V, fig. 6), to flabellate, and the disposition of the oscules may be on one of two plans,
a few conspicuous oscules (cf. Burton, loc. cit., pi. v, figs. 5, 6) or numerous small
oscules scattered over one surface only (cf. Topsent, 1917, pi. iv, fig. 13). I was wrong
therefore in regarding the "orifices circulaires" (1929, p. 422) as due to commensal
cirripedes.

Calyx arcuarius, Topsent, is a species of varied form, with a dimorphic system of
exhalant apertures, and these different variations are well shown in the present series of
specimens. On the other hand, the main skeleton is constant in structure and forms a
character by which the species may be readily recognized. There can therefore be no
doubt that all the specimens before me at the moment are conspecific. In view of this,
the following observation is of particular interest ; namely, that some of these specimens
have abundant toxa, o-oy-o-i mm. long, some only a few, and some appear to be
entirely without them. It is obvious then that Gellius arcuarius, Topsent, and Calyx
stipitatus, Topsent, are synonymous, the holotype of the former being a fragment of a
large flabellate sponge having toxa in abundance, and the holotype of the latter being an
incomplete specimen having no toxa. We have to accept, as a consequence, the followmg
facts: that the members of the genus Calyx, a genus hitherto regarded as characterized
by the absence of microscleres, may possess toxa; that the absence of a particular
microsclere has, per se, no taxonomic value.

Distribution. South Orkneys ; Victoria Land.

278 DISCOVERY REPORTS

Calyx kerguelensis (Hentschel).

GeUiodes kerguelensis, Hentschel, 1914, p. 127, pi. viii, fig. 11.

Occurrence. St. 58: Falkland Islands, on piles of Government Jetty, St. 140: South Georgia,
122-136 m.

Remarks. This species, represented in the collection by a piece of a stout, cylindrical
sponge, has a skeleton very similar to that of C. arciiarius except that the dermal skeleton
is somewhat more diffuse.

In the present specimen, the sigmata measure, for the most part, up to 0-04 mm.
chord, but a few measure up to o-i mm.

Distribution. Kerguelen.

Genus Dasychalina, Ridley and Dendy.

Genotype. D.fragilis, Ridley and Dendy.

Diagnosis. Haploscleridae with dermal skeleton composed of an irregular multi-
spicular reticulation subdivided into primary and secondary meshes ; main skeleton of
stout multispicular fibres, or systems of anastomosing fibres, with numerous isolated
spicules scattered between ; spongin absent or present in very small quantities.

Remarks. This genus, estabhshed by Ridley and Dendy (1886, p. 329), was abandoned
by them (1887) because of a suspected identity with Pachychalina, Schmidt. The
characters of the latter are, however, unknown to us, and because Dosycholino frogilis
differs in the structure of its skeleton from other species of "Pachychalina^^ Autt., it is
proposed here to revive the use of the name. Dasychalina is intermediate between
Adocia and Callyspojigia (= "Pachychalina'", Autt.).

Dasychalina validissima (Thiele) (Plate L, figs. 3-7; Fig. 12).

Pachychalina validissima, Thiele, 1905, p. 473, figs. 16, 91 ; Halichondria magellanica, Dendy,
1924, p. 325.

Occurrence. St. 51 : Falkland Islands, 115 m.; St. 53: Falkland Islands, 0-2 m.; St. 140: South
Georgia, 122-136 m.; St. WS 73 : Falkland Islands, i2i m.; St. WS76: Falkland Islands, 205-
207 m.; St. WS77: Falkland Islands, iio-ii3m.; St. WS 83: Falkland Islands, 137-129 m.;
St. WS84: Falkland Islands, 75-74 m.; St. WS 109: Falkland Islands, 145 m.; St. WS 233 :
Falkland Islands, 185-175 m.; St. WS 237: Falkland Islands, 150-256 m.; St. WS 243: Falkland
Islands, 144-141 m.

Remarks. The specimens may be massive, as in the holotype, massively branched,
with branches cylindrical or flattened, or flabellate (PI. L, figs. 3-7), while the surface
may be smooth, tuberculate or spinose. The skeleton varies from that shown in
Fig. i2rt to that in Fig. 12 b. In spite of the difference in appearance, these two
types of skeleton structure are essentially the same. In the holotype (Fig. 12 a) the
skeleton consists of a reticulation of stout multispicular fibres, with numerous isolated

TETRAXONIDA

279

Spicules scattered between, but in most of the present specimens (Fig. 12 b) the
multispicular fibres are resolved into anastomosing systems of smaller fibres. There are,

Fig. 12. Dasychalina validissima (Thiele). a, b, sections at right angles to surface
(a is from holotype); c, dermal skeleton. All x 7J.

however, all gradations between these two extremes. The oxea measure o- 15-0-32 mm.
long.

Distribution. Calbuco ; Straits of Magellan.

Genus Callyspongia, Duchaissang and Michelotti.

Callyspongia fortis (Ridley) (Figs. 13, 14).

Siphonochalina fortis, Ridley, 1881, p. iii, pi. x, fig. 3.

Occurrence. St. WS72: Falkland Islands, 95 m.; St. WS 83 : Falkland Islands, 137-12901.;
St. WS 84: Falkland Islands, 75-74 m.; St. WS 86: Falkland Islands, 157-147 m.

Diagnosis, Sponge sub-ramose to lobose; surface minutely conulose; oscules large,
conspicuous, situated on crater-like projections or with margins level with surface, in
all cases leading into deep cloacae receiving large exhalant canals which penetrate body
of sponge ; pores apparently scattered in meshes of dermal skeleton ; skeleton differen-
tiated into main and dermal ; main skeleton consisting of a coarse-meshed network of
stout spongin fibres, often 0-3 mm. thick, containing an axial row of spicules, multi-
spicular in primary fibres, unispicular in secondary fibres; meshes of main skeleton
ranging from 0-4 to 1-2 mm. across; dermal skeleton a tangential network of fibres
differentiated into primary, secondary and tertiary fibres; primary fibres stout, of
similar dimensions to those of main skeleton; mesh usually triangular, but may be
quadrate or polygonal and divided into secondary and tertiary meshes; dimensions
of latter usually no more than a spicule length across ; spicules oxea, straight, o-o66 by
o-oo6 mm.

Remarks. The species is characterized by its external form and by the small size of
its spicules.

The inadequacy of Ceraochalina, as a genus for the reception for all Chalininae in
which the spicules are vestigial, is shown by one specimen in which the fibres of the

6-2

Fig. 13. C ally spongia for iis (Ridley). To show variation in form. All figures x f. For the sake of clearness
the oscules are marked by dotted areas (a and a' are oscules not yet opened).

Fig. 14. Cally spongia fortis {Kid\&y). Portion of tangential dermal skeleton, x 100.

TETRAXONIDA

281

main skeleton contain only feebly developed spicules, measuring 0-052 by 0-002 mm.,
while the dermal skeleton contains only normal spicules, measuring o-o6 by o-oo6 mm.

In the same way, the inadequacy of Siphonochaliua, as a genus for the reception of all
tubular Chalininae, is shown by the fact that, while the remainder of the specimens are
sub-ramose or lobate, one is erect and tubular, with an apical osculum and a deep cloaca
running the length of the body, and two accessory lateral oscula near the base.

Distribution. Chile.

Callyspongia fusifera (Thiele) (Plate LII, fig. i ; Figs. 15, 16).

Chalina fusifera, Thiele, 1905, p. 476, figs. 15, 32, 95.

Occurrence. St. WS 77: Falkland Islands, 1 10-113 ™-'^ St. WS 86: Falkland Islands, 157-147 m.

Remarks. The specimen assigned to this species is palmo-digitate (see Plate LII,
fig. i) and comparison of the figure given with that given for the holotype (Thiele, loc.
cit., fig. 15) shows that the two specimens have a similar appearance although the shape

a be

Fig. 15. Callyspongia fusifera (Thiele). Falkland Islands specimen. «, dermal skeleton, x 100;
b, main skeleton, x 100; c, oxeote, x 400; ec/i, echinating tufts of spicules seen in section.

is not identical. On the other hand, the arrangement of the dermal skeleton and the
shape of the spicules diff^er considerably in the two specimens. Through the kindness
of Dr W. Arndt, I have been able to examine a piece of the holotype and have found that
the dermal skeleton is a close-meshed reticulation showing an incipient differentia-
tion into primary and secondary meshes. The main skeleton is a coarse and large-meshed
reticulation with multispicular primary fibres and unispicular secondary fibres. The
spicules are o-o6 mm. long and rounded at the ends, although not forming pure
strongyla. The dermal skeleton of the present specimen does not show so marked a
differentiation into primary and secondary fibres even as the holotype. The main
skeleton is a close-meshed reticulation with bispicular primary fibres and unispicular
secondary fibres, and the dermal skeleton is echinated at the nodes by tufts of spicules
projecting from the ends of the primary fibres of the main skeleton. The spicules are

DISCOVERY REPORTS

oxea measuring 0-14 by o-oi mm. A comparison of Figs. 15 and 16 will show clearly the
obvious way in which the skeleton of the present specimen differs from that of the type.
In fact, there might be something to be said in favour of regarding the two specimens as
specifically different, but for the fact that the differences they exhibit can be shown by

Fig. 16. Callyspongia fusifera (Thiele). Holotype. a, dermal skeleton, x 66; b, main skeleton, x 66.

comparison with the known variations in other species of Callyspongia to be unim-
portant, and also by the fact that the co-types of the species are to an extent inter-
mediate between these two specimens. In the specimen from Calbuco, for example, the
spicules are almost identical in shape with those of the Discovery specimen, though
slightly smaller. Unfortunately, I have no information as to the structure of the skeleton
in the co-types, but it is probable that intermediate characters will be found in this too.
The co-type from Punta Arenas has oxea measuring 0-24 mm. long, considerably longer
than those of the present specimen.

Distribution. Calbuco ; Tumbes ; Punta Arenas.

Callyspongia flabellata, sp.n. (Plate XLIX, fig. 4; Fig. 17).
Holotype. B.M. 28. 2. 15. 359.
Occurrence. St. WS 81 : Falkland Islands, 81-82 m.

Diagnosis. Sponge flabellate, pedunculate; surface even, minutely hispid; oscules
numerous, 1-2 mm. in diameter, mainly disposed in an irregular linear series along
margins of sponge; pores apparently distributed generally over surface; colour, in
spirit, greyish yellow ; main skeleton irregularly isodictyal, uni- or multispicular ; dermal
skeleton a tangential network of fibres, similar to those of main skeleton, echinated at
nodes by tufts of spicules projecting at right angles ; spicules oxea, straight, smooth,
0-105 by 0-009 "^rn-

Remarks. The dermal skeleton is not easily detected. It cannot be readily removed,
as is usual in other Chalininae with a special dermal skeleton, and the fact that its nodes
are echinated by tufts of spicules makes it appear as though the sponge were a true

TETRAXONIDA

283

Haltclotia, devoid of special dermal skeleton and with the primary fibres of the main
skeleton continued outwards beyond the ectosome. Sections of the sponge, taken at

ilJMJJL

/I

\^1

a b

Fig. 17. Callyspongia flabellata, sp.n. a, section at right angles to surface, x 60;
b, dermal skeleton, x 60; c, oxeote, x 400.

right angles to the surface, show, however, even to the naked eye, that there is a definite
and unmistakable dermal skeleton. This is the more readily apparent since the interior
of the sponge is cavernous and the ectosome compact and firm.
The species is characterized mainly by its external form.

Genus Phloeodictyon, Carter

Phloeodictyon eumitum, Kirkpatrick.

P. eumatum {sic), Kirkpatrick, 1903, p. 253, pi. v, fig. 18, pi. vi, fig. 19 (the name is spelt
eumitum on p. 235, and in the explanation to the plates).

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Distribution. Natal.

Family DESMACIDONIDAE

Section ISODICTYEAE

Genus Isodictya, Bowerbank

Isodictya kerguelensis (Ridley and Dendy).

Homoeodictya kerguelensis, Ridley and Dendy, 1886, p. 2^6; id., 1887, p. no, pi. xxiii, fig. 3,
pi. xxiv, fig. 3; Desmacidon kerguelensis, Topsent, 1908, p. 24; Homoeodictya kerguelensis, id.,
1917, p. 67.

Occurrence. St. WS 79; Falkland Islands, 131-132 m.

Distribution. Kerguelen; Graham Land.

284 DISCOVERY REPORTS

Isodictya kerguelensis, var. simillima (Hentschel).

Homoeodictya kerguelensis, var. simillima, Hentschel, 1914, p. 80, pi. vi, fig. 7.
Occurrence. St. WS 109; Falkland Islands, 145 m.
Distribution. Wilhelm Land.

Isodictya setifer (Topsent).

Desmacidon setifer, Topsent, 1901, p. 6; id., 1901, p. 17, pi. i, fig. 3, pi. iii, fig. 6; Desmacidon
spinigera, Kirkpatrick, 1907, p. 283 ; id., 1908, p. 39, pi. xix, fig. 3, pi. xxiii, fig. 3 ; Desmacidon
setifer, Topsent, 1908, p. 26; Homoeodictya verrucosa, id., 1913, p. 636, pi. v, fig. i, pi. vi, fig. 13 ;
H. setifera, id., loc. cit., p. 637; H. obliquidens, Hentschel, 1914, p. 79, pi. iv, fig. 7, pi. vi, fig. 6;
H. setifera, id., loc. cit., p. 81 ; H. setifera, Topsent, 1917, p. 65, pi. vi, fig. 12; //. trigona, id.,
loc. cit., p. 66; Desmacidon spinigera, Brondsted, 1926, p. 3; Isodictya spinigera. Burton, 1929,
p. 424; /. verrucosa, id., loc. cit., p. 424.

Occurrence. St. 39: South Georgia, 179-235 m.; St. 42: South Georgia, 120-20401.; St. 143:
South Georgia, 273 m.; St. 144: South Georgia, 155-178 m.; St. 148: South Georgia, 132-148 m.;
St. 149: South Georgia, 200-234 m.; St. 159: South Georgia, 160 m.; St. 160: Shag Rocks, 177 m.;
St. 175: South Shetlands, 200 m.; St. 190: Palmer Archipelago, 93-130 m.; St. WS 73 : Falkland
Islands, 121m.; St. WS77: Falkland Islands, iio-ii3m.; St. WS 80: Falkland Islands, 152-
156 m.; St. WS81: Falkland Islands, 81-82 m.; St. WS 83 : Falkland Islands, 137-12901.; St.
WS 88 : Falkland Islands, 96-1 27 m. ; St. WS 99 : Falkland Islands, 251-225 m. ; St. WS 109 : Falkland
Islands, 145 m.; St. WS 239: Falkland Islands, 196-193 m.; St. WS 243: Falkland Islands, 144-
141 m.; St. WS 246: Falkland Islands, 267-208 m.; St. WS 248: Falkland Islands, 210-242 m.

Remarks. There are 32 specimens in the collection which belong, in my opinion,
to the same species, namely Isodictya setifer (Topsent). They all resemble each other
strongly externally, and have much the same appearance as that of the holotype of
Homoeodictya verrucosa, Topsent (19 13, pl.v, fig. i). The form is massive to lobose, often
stipitate. In spiculation they agree also in having a main skeleton of oxea and, for
microscleres, isochelae palmatae, which have the peculiarity that they are often contorted
and distorted and the palms are sometimes obliquely directed. This peculiarity of the
microscleres has been overlooked in many cases, as for example in Desmacidon spinigera,
Kirkpatrick, although Topsent and Hentschel drew attention to it in Desmacidon setifer
and Homoeodictya obliquidens respectively. The reason for this seems to be that the con-
tortion and distortion are most marked in the smaller spicules, that is in those least
easily seen, and will only be noticed if the microscleres are numerous and the eye
happens to catch one in an isolated position in the preparation lying in a particularly
favourable angle for examination. At all events, I have satisfied myself that such micro-
scleres are present in the type of Desmacidon spinigera, Kirkpatrick, and that they are
present in varying proportions in the specimens of the present collection. It often
happens, of course, that a particular individual may have a greater proportion of normal
chelae than abnormal, and were one not looking particularly for contorted or distorted
microscleres these would be completely overlooked. I have no doubt therefore that in
the types of Homoeodictya verrucosa, Topsent, and other species in the above list of
synonyms in which abnormal chelae have not been recorded, these same peculiar chelae
are present, though probably in small quantities.

TETRAXONIDA 285

With regard to the types of Desmacidon setifer, D. spmigera, Homoeodictya trigona,
H. verrucosa and H. obliqiddens we may say that in external appearance there is Hterally
nothing to choose between them. In the arrangement of the spicules in the main skeleton
there is the same similarity, and in the shape of the microscleres there are the same
peculiarities. It is only in the dimensions of the spicules, and more particularly of the
megascleres, that distinctions may be made between them. The dimensions of the
spicules are as follows :

Oxea (mm.) Chelae (mm.)

D. setifer o-88o-i-o o-023-o-03

D. spmigera 0'73i 0-024

H. trigona 0-740 0-060-0-07

H. verrucosa 0-52 0-027-0-037

H. obliquidens 0-528-0-672 0-038-0-055

The variation in the dimensions of the oxea in these five forms is from 0-52 to
i-o mm., and in the chelae from 0-023 to 0-07 mm. This, on the face of it, is quite a con-
siderable range. On the other hand, the five forms afford a convenient series of inter-
mediates from one extreme of measurement to the other, both in the oxea and in the
chelae, and it is obvious that if a large collection of specimens were examined and an
attempt made to divide them into species on the basis of spicule measurement, the
attempt would prove to be impossible. This is actually what has happened in the case
of the Discovery specimens, in which the average measurements of the spicules range as
follows: oxea o-35-o-7 by o-oi-o-025 mm., and chelae o-028-o-o6 mm. chord.

From direct observation and comparison of the 32 specimens of the present collection,
I have no doubt whatsoever that they are all conspecific : and since the range of variation
in the dimensions of their spicules is the same, or practically the same, as that found
in the five species tabulated above, especially in view of the strong resemblance these
five species have to each other in all other respects, I can see no alternative to regarding
them as synonymous, even though it means the admission that the oxea in this species
may have the surprisingly wide range of dimensions shown, from 0-35 to i-o mm., and
that the chelae may vary from 0-023 ^^ °'°7 ™^^-

Distribution. Wilhelm Land ; Victoria Land ; Graham Land ; Falkland Islands.

Isodictya delicata (Thiele).

Desmacidon delicata, Thiele, 1905, p. 435, figs, i, 55.
Occurrence. St. WS 109: Falkland Islands, 145 m.
Distribution. Admiralty Sound.

Isodictya antarctica (Kirkpatrick) (Plate LI, fig. 4).

Desmacidon kerguelensis , var. antarctica, Kirkpatrick, 1908, p. 37, pi. xix, fig. i, pi. xxiii, fig. i.

Occurrence. St. 42: South Georgia, 120-20401.; St. 155: South Georgia, 260 m.; St WS 73 :
Falkland Islands, 121 m.; St. WS 83 : Falkland Islands, 137-129 m.

Distribution. Victoria Land.

286 DISCOVERY REPORTS

Isodictya cactoides (Kirkpatrick).

Desmacidon kergiiekiisis, var. cactoides, Kirkpatrick, 1908, p. 38, pi. xix, fig. 2, pi. xxiii, fig. 2;
Isodictya cactoides. Burton, 1929, p. 424.

Occurrence. St. 39: South Georgia, 179-235 m.; St. 123: South Georgia, 230-25001.; St. 190:
Palmer Archipelago, 93-130 m.; St. MS 68: South Georgia, 220-247 m.

Distribution. Victoria Land.

Isodictya microchela (Topsent) (Plate LI, fig. 5).
Homoeodictya microchela, Topsent, 191 5, p. 37.
Occurretice. St. WS 86: Falkland Islands, 151-147 m.

Remarks. The original description of this species is unaccompanied by any illustra-
tion, so that it can only be identified with hesitation. The present specimen appears to
belong to it, but if the identification is correct then the external form of this species
is that shown in Plate LI, fig. 5.

Distribution. Burdwood Bank.

Isodictya erinacea (Topsent).

Homoeodictya erinacea, Topsent, 191 6, p. 169; //. kirkpatricki, id., loc. cit., p. 169; //. erinacea,
id., 1917, p. 68, pi. iii, fig. i, pi. vi, fig. 15 ; //. kirkpatricki, id., loc. cit., p. 70, pi. i, fig. 2, pi. vi,
fig. 14; Desmacidon doryphora, Brondsted, 1926, p. 2, figs, i, 2.

Occurrence. St. 190: Palmer Archipelago, 93-130 m.

Remarks. In view of the strong likeness in other respects, I cannot accept the
separation of Homoeodictya kirkpatricki and Desmacidon doryphora merely on account
of the difference in the sizes of their respective megascleres, for that is the only
character in which they diff'er. Under Isodictya setifer (see above), I have shown how
much the dimensions of the spicules may vary within a single species of Isodictya, and
the evidence from this source alone is sufficient to justify the identification of Br^nd-
sted's species with that of Topsent. With regard to Homoeodictya erinacea, there is so
much resemblance to the two other species that it is not possible to regard the mere
presence of raphides or of tiny bosses on the chelae as characters necessitating a specific
distinction.

Distribution. Graham Land; Victoria Land.

Isodictya toxophila, sp.n. (Plate LIT, figs. 2, 3; Plate LIII, figs, i, 2; Fig. 18).

Holotype. B.M. 28. 2. 15. 158.

Occurrence. St. 39: South Georgia, 179-235 m.; St. 148: South Georgia, 132-148 m.; St. 175:
South Shedands, 200 m.; St. 190: Palmer Archipelago, 93-130 m.; St. WS 27: South Georgia,
106-109 m.

TETRAXONIDA 287

Diagnosis. Sponge irregularly massive, with large oscules, or sub-infundibular with
one side oscular, bearing oscules up to 2 mm. diameter,
and other side poral, thrown in irregular folds ; surface his-
pid, often verrucose; texture firm, compressible; colour, in
spirit, yellow to brown ; skeleton sub-isodictyal, with primary
fibres 12-20 spicules thick, secondary fibres loose or re-
presented by irregularly scattered single spicules; oxea
0-42 by 0-02 mm. ; chelae o-o6 mm. chord ; toxa up to
0-175 ^^- long.

Remarks. The species, which is characterized mainly by

the possession of toxa in addition to isochelae, shows a

similar range in external form to that found in Haliclona

bilamellata (see p. 267). The microscleres are very variable

in their occurrence, the toxa are often very rare, and in one ^'8- '^- I^odictya toxoplnla,
1 ,, I 1 1 r 1 sp.n. Microscleres, x 300.

specimen no microscleres at all could be round.

Genus Guitarra, Carter
Guitarra fimbriata, Carter.

(For synonymy see Burton, 1929, p. 426.)

Occurrence. St. 39: South Georgia, 179-235 m.; St. WS79: Falkland Islands, I3i-i32m.;
St. WS86: Falkland Islands, i5i-i47m.; St. WS 223 : Falkland Islands, 114 m.; St. WS 243 :
Falkland Islands, 144-141 m.

Distribution. Arctic, Atlantic and Antarctic; Indian Ocean; New Zealand.

Genus Cercidochela, Kirkpatrick

Cercidochela lankesteri, Kirkpatrick.

C. lankesteri, Kirkpatrick, 1907, p. 284; id., 1908, p. 42, pi. xix, fig. 5, pi. xxiii, fig. 5 ; Hentschel,
1914, p. 74; Burton, 1929, p. 426, pi. v, fig. 8.

Occurrence. St. 187: Palmer Archipelago, 259 m.
Distribution. Victoria Land ; Wilhelm Land.

Genus Plumocolumella, Burton

Plumocolumella maeandrina (Kirkpatrick).

Desmacidon maeandrina, Kirkpatrick, 1907, p. 282 ; id., 1908, p. 40, pi. xix, fig. 4, pi. xxiii, fig.4 ;
Plumocolumella maeandrina, Burton, 1929, p. 425.

Occurrence. St. 27: South Georgia, no m.; St. 42: South Georgia, 120-204 m.; St. 123: South
Georgia, 160 m.; St. 148: South Georgia, 132-148 m.; St. 149: South Georgia, 200-234 m.; St. 158:
South Georgia, 401-411 m.; St. WS 27: South Georgia, 106-109 m.; St. WS 33: South Georgia,
130 m.

Distribution. Victoria Land.

7-2

288 DISCOVERY REPORTS

Plumocolumella myxillioides, sp.n. (Plate LIII, figs. 3, 4).

Holotype. B.M. 28. 2. 15. 321.

Occurrence. St. WS 79: Falkland Islands, 13 1-132 m.; St. WS 80: Falkland Islands, 152-156 m.;
St. WS 81 : Falkland Islands, 81-82 m. ; St. WS 83 : Falkland Islands, 137-129 m. ; St. WS 86 : Falk-
land Islands, 151-147 m.; St. WS 88: Falkland Islands, 96-127 m.; St. WS 243 : Falkland Islands,
144-141 m.; St. WS 246: Falkland Islands, 267-208 m.; St. WS 248: Falkland Islands, 210-242 m.

Diagnosis. Sponge erect, arborescent with flattened anastomosing branches or sub-
lamellate ; surface smooth and even, bearing a conspicuous veining due to large subdermal
canals showing through a transparent dermis; numerous circular depressions, scattered
singly or in groups generally over surface, which probably bear both inhalant and ex-
halant apertures; colour, in spirit, yellow; skeleton, composed of irregular wisp-like
bundles of tornota running longitudinally through branches, often forming a dense axial
core, and bending towards surface or send off lateral branches to end in dermal brushes ;
tornota of main skeleton usually straight and smooth, with ends evenly and gradually
pointed, frequently hastate or mucronate, measuring 0-3 by o-oo6 mm. ; isochelae
arcuatae (.''), 0-03 mm. chord.

Section MY GALEAE

Genus Mycale, Gray

Mycale magellanica (Ridley).

Esperia magellanica, Ridley, 1881, p. 117, pi. x, fig. 5; E. cunninghami, Carter, 1882, p. 300,
pi. xi, fig. 17 ; Esperella magellanica, Ridley and Dendy, 1887, p. 67 ; Mycale magellanica, Thiele,
1905, p. 442; Mycale sp., id., loc. cit., p. 443, fig. 61: Mycale sp., Kirkpatrick, 1908, p. 37;
M. magellanica, Topsent, 1913, p. 632, pi. iv, fig. 4, pi. vi, fig. io;M.pellita, id., loc. cit., p. 633,
pi. V, fig. 2; M. antarctica, Hentschel, 1914, p. 58, pi. v, fig. 7; M. rossi, id., loc. cit., p. 59,
pi. V, fig. 8; M. magellanica, Dendy, 1924, p. 336; M. lillei, id., loc. cit., p. 337.

Occurrence. St. 53 : Falkland Islands, 0-2 m. ; St. 55 : Falkland Islands, 10-16 m. ; St. 146 : South
Georgia, 728 m.; St. 175: South Shetlands, 200 m.; St. WS 73 : Falkland Islands, 121m.; St.
WS82: Falkland Islands, i40-i44m.; St. WS 83 : Falkland Islands, i37-i29m.; St.WS87:
Falkland Islands, 96-127 m.; St. WS 90: Tierra del Fuego, 82-81 m.; St. WS 177: South Georgia,
97 m.; St. WS 210: Falkland Islands, 161 m.; St. WS213: Falkland Islands, 249-239 m.; St.
WS 222: Falkland Islands, 100-106 m.; St. WS 225: Falkland Islands, 162-161 m.; St. WS 233,
Falkland Islands, 185-175 m.; St. WS 239: Falkland Islands, 196-193 m.; St. WS 244: Falkland
Islands, 253-247 m.; St. WS 246: Falkland Islands, 267-208 m.; St. WS 247: Falkland Islands,
172 m.; St. WS 248: Falkland Islands, 210-242 m.; St. WS 249: Falkland Islands, 166 m.; St. WS
250: Falkland Islands, 251-313 m.

Remarks. The synonymy of M. magellanica will be dealt with fully in a report, to be
published shortly, on the collections of the Swedish Antarctic Expedition. One point
only need be noted here in connection with this species, that a single specimen from
St. WS 222, otherwise typical, is entirely without microscleres.

Distribution. Southern extremity of South America; Graham Land; Victoria
Land ; Wilhelm Land.

TETRAXONIDA 289

Mycale lapidiformis (Ridley and Dendy).

Esperella lapidiformis, Ridley and Dendy, 1886, p. 338; id., 1887, p. 64, pi. xv, figs. 2, 10,
pi. xvi, fig. 2.

Occurrence. St. WS 248: Falkland Islands, 210-242 m.

Distribution. Mouth of Rio de la Plata.

Mycale acerata, Kirkpatrick.

M. acerata, Kirkpatrick, 1907, p. 280; id., 1908, p. 36, pi. xx, fig. i, pi. xxiv, fig. 10; .1/. acerata,
var. minor, Hentschel, 1914, p. 63; M. acerata, Topsent, 1917, p. 63; Burton, 1929, p. 430.

Occurrence. St. 42: South Georgia, 120-204 m.; St. 45: South Georgia, 238-27001.; St. 159:
South Georgia, 160 m.; St. 164: South Orkneys, 24-36 m.; St. 175: South Shetlands, 200 m.

Distribution. Graham Land ; Victoria Land ; Wilhelm Land.

Mycale tridens, Hentschel.

M. tridens, Hentschel, 1914, p. 56, pi. v, fig. 6; Topsent, 1917, p. 63 ; Burton, 1929, p. 430.

Occurrence, St. 170: Clarence Island, 342 m. ; St. MS 71 : East Cumberland Bay, South Georgia,
110-60 m.

Distribution. Graham Land ; Victoria Land ; Wilhelm Land.

Mycale macrochela, sp.n. (Plate LI, fig. 6).
Holotype. B.M. 28. 2. 15. 827.
Occurrence. St. 170: Clarence Island, 342 m.

Diagnosis. Sponge lamellar with large oscules, up to 8 mm. in diameter and with
conspicuous membranous collar, scattered over one surface; surface even, minutely
hispid; texture firm, friable; colour, in spirit, dark greenish grey; skeleton a loose
reticulation of bundles of subtylostyli, with brushes of subtylostyli at surface; sub-
tylostyli 0-53 by 0-017 mm.; anisochelae 0-07, 0-035 ^^'^ 0-022 mm. chord; sigmata
o-i mm. chord; trichodragmata 0-035 mi^- long.

Remarks. The species is remarkable for its plate-like external form and for the form
of the anisochelae. It resembles M. simonis, from South Africa, most closely in the form
of the anisochelae, but differs considerably from this species in all other respects.

Genus Amphilectus, Vosmaer

Amphilectus fucorum (Esper) (Plate LIV, figs. 1-4).

Spongia fucorum, Esper, 1794, p. 278, pi. xlix, figs, i, 2; S. parasitica, Montagu, i8i8, p. 114;
Gray, 1821, p. 360; Grant, 1826, p. 348; Halichotidria parasitica, Flemming, 1828, p. 521;
Halispongia parasitica, Blainville, 1834, p. 532; Halichondria parasitica, Bellamy, 1839, p. 268;
Thompson, 1840, p. 254; H. fucorum, Johnston, 1842, p. 112, pi. ix, pi. xii, fig. 2; Isodictya
normani, Bowerbank, 1866, p. 320 ; /. fucorum, id., loc. cit., p. 322 ; /. alderi, id., loc. cit., p. 323 ;
/. edwardii, id., loc. cit., p. 325 ; /. paupera, id., loc. cit., p. 328 ; /. uniformis, id., loc. cit., p. 329 ;
/. clarkei, id., loc. cit., p. 330; /. gracilis, id., loc. cit., p. 331 ; /. normani. Gray, 1867, p. 534;

ago DISCOVERY REPORTS

I.paiipcra, Parfitt, 1868, p. 8 ; /. iinijormis, Bowerbank, 1874, p. 139, pi. Iv, figs. 8-10 ; /. normani,
id., loc. cit., p. 141 , pi. Ivi, figs. 1-5 ; /. clarkei, id., he. cit., p. 142, pi. Ivi, figs. 1 1-15 ; I.fucorum,
id., loc. cit., p. 142, pi. Ivi, figs. 16-19; I- alderi, id., loc. cit., p. 143, pl. Ivi, figs. 20-26; /. ed-
wardii, id., loc. cit., p. 148, pl. Iviii, figs. 15-18 ; /. gracilis, id., loc. cit., p. 149, pl. Iviii, figs. 23-
26; /. imitata, id., he. cit., p. 223, pl. Ixxvi, figs. 3-6; /. invalida, id., loc. cit., p. 285, pl. Ixxxv,
figs. 8-10; /. dubia, id., he. cit., p. 323, pl. xc, figs. 4-7; Hymeniacidon ealhsus, Bowerbank,
1882, p. 86, pl. iv, figs. 6-8; Amphihetus gracilis, Vosmaer, 1880, p. iii; A. compressus,
id., he. cit., p. 112; A. dubius, id., he. cit., p. 112; A. imitatus, id., loc. cit., p. 116; A.fucorum,
id., loc. cit., p. 117; A. normani, id., he. cit., p. 117; A. clarkei, id., loc. cit., p. 119; Isodictya
gracilis, Bowerbank, 1882, p. 136 ; /. invalida, id., he. cit., p. 136 ; /. normani, id., loc. cit., p. 136 ;
/. hispida, id., loc. cit., p. 136, pl. xii, figs. 1-5; /. alderi, id., loc. cit., p. 138; /. edzvardii, id.,
loc. cit., p. 138; I.fucorum, id., loc. cit., p. 138; /. clarkei, id., loc. cit., p. 139; /. paupera, id.,
loc. cit., p. 139; /. uniformis, id., loc. cit., p. 139; /. dubia, id., he. cit., p. 140; /. imitata, id.,
loc. cit., p. 141 ; /. involuta, id., he. cit., p. 143, pl. x, figs. 1-4; /. pertenuis, id., he. cit., p. 144,
pl. xiii, figs. 1-4; /. seitula, id., loc. cit., p. 146, pl. iv, figs. 1-3, pl. ix, figs. 1-3 ; Esperia normani,
Fristedt, 1885, p. 42; Reniera fucorum , Topsent, 1888, p. 148; R. paupera, id., loc. cit., p. 148;
R. uniformis, id., loc. cit., p. 149; Stylinos uniformis, id., 1891, p. 535; Esperiopsis normani,
Lundbeck, 1905, p. 13, pl. viii, fig. 2; Esperiopsis sp., id., loc. eit., p. 15, pl. viii, fig. 3; E. ed-
zvardii, Thiele, 1905, p. 441 ; E. fucorum, Stephens, 1912, p. 36; E. edzvardii, Hentschel, 1914,
p. 69; E. informis, Stephens, 1915, p. 450, pl. xi, fig. 11; E. fucorum, Stephens, 1921, p. 17;
E. normani, Brondsted, 1923, p. 138, fig. 18; E. crasso-fibrosa, id., he. cit., p. 139, fig. 19;
E. edzvardii, Dendy, 1924, p. 341; E. normani, Brondsted, 1926, p. 5; Amphihetus gracilis.
Burton, 1929, p. 428; Axinosia incrustans. Burton, 1930, p. 333, fig. i.

Occurrence. St. 6: Tristan da Cunha, 80-140 m.; St. 42: South Georgia, 120-204 m.; St. 51:
Falkland Islands, 105-11501.; St. 160: Shag Rocks, 177 m.; St. WS 83 : Falkland Islands, 137-
129 m.; St. WS 229: Falkland Islands, 210-271 m.

Diagnosis. Sponge polymorphic, encrusting, massive, or branching ; surface even,
minutely hispid ; oscules inconspicuous in encrusting forms, small and scattered or large
and crateriform in massive forms, small, often arranged in linear series in branching
forms; skeleton isodictyal to sub-isodictyal ; no special dermal skeleton; megascleres
smooth, usually curved styli, o-i35-o-48 by o-oo3-o-oi9 mm. ; microscleres palmate
isochelae, o-oi4-o-028 mm.

Remarks. Bowerbank, in his Monograph of the British Spmigiadae (1866-82), de-
scribed twelve species of Isodictya possessing styli and palmate anisochelae. All agree so
very closely in spiculation that, judging from this aspect alone, there would be no
hesitation in regarding them all as synonyms of a single species. Certain differences in
their external appearance give rise, however, to a little hesitation in taking so drastic a
step, and it is necessary to compare them from this point of view.

/. normani is massive with large, crater-like oscules, or encrusting with small, scattered
oscules ; colour (alive) unknown, (dried) fawn-yellow.

/. fucorum is massive with comparatively large, scattered oscules; colour (alive)
"bright red or pink, to pale ash or brown".

I. alderi is encrusting, with inconspicuous oscules ; colour (alive) faint red.

/. edzvardii is encrusting, with inconspicuous oscules; colour (dried) "ochreous
yellow".

TETRAXONIDA 291

/. pauper a is encrusting, with inconspicuous oscules; colour (dried) "light ochreous
yellow".

/. gracilis is ramose and branching, with small oscules ; colour (alive) pale buff, (dried)
cream white.

/. imitata is encrusting, with no visible oscules ; colour (dried) " brown with tint of
green".

/. dubia is encrusting, with no conspicuous oscules; colour (dried) "light grey with
tint of yellow", (alive) orange.

/. hispida is encrusting, with no visible oscules ; colour (dried) dark grey.

/. nodosa is massive, with fistulous oscules; colour (dried) light brown.

/. involuta is encrusting, with no apparent oscules ; colour (dried) dark brown.

/. scittda is encrusting, with no apparent oscules; colour (dried) "light ochreous
yellow".

With the exception of /. gracilis, there is nothing in the shape of these sponges to
justify a segregation into different species. The colour records taken from the dried
state may be ignored. In my experience, the result of drying, so far as colour is con-
cerned, depends more on the manner in which the sponge is dried, than on the colour
in the living condition. The records taken from living specimens show considerable
variation in colour, but with a tendency to a reddish tint.

If any division is to be made it must take the form of separating the twelve species
into two forms of a single species, a branching form, including /. gracilis, and a massive
form including all the rest. Even so, it is probable that the two forms are purely eco-
logical and that in the course of future investigations, all gradations between the two
may be found.

The dimensions of the spicules of Amphilectus fucoriim, as exemplified by the holo-
types of the twelve species of Isodictya enumerated by Bowerbank, are: oxea 0-13 5-0-2
by o-oo3-o-oo8 mm., and isochelae o-oi4~o-oi8 mm.

Gray {loc. cit.), Vosmaer {loc. cit.) and Topsent {loc. cit.) add nothing new to our
knowledge of these forms. Ridley (1883) described two specimens from Scotland, under
Amphilectus edwardii, one of which was erect and branching and the other cylindrical.
At the same time, he doubts the identity of Isodictya gracilis with I. edwardii, although
his own specimens had the external form of the former. On the other hand, Levinsen
{loc. cit.), Lundbeck {loc. cit.) and Stephens (1921) leave no doubt that Esperiopsis
fucorum may be either encrusting or branched. Further, I have seen massive speci-
mens from Plymouth, with large oscules, in which the surfaces were beset with
numerous blunt processes, up to 2 cm. long, undoubtedly representing incipient
branches. We may now assume, I think, that there is nothing peculiar about Isodictya
gracilis and that it is a synonym of Amphilectus fucorum.

Numerous specimens of Amphilectus fucorum have been recorded, under various
names, from the southern hemisphere which agree in external form with those of the
northern hemisphere. They have this in common, however, that the styH are considerably
larger in the southern forms and the chelae slightly so. The recorded measurements of

2g2 DISCOVERY REPORTS

the spicules are: Thiele (Falkland Islands), styli o-225-o-26 by o-oi mm., chelae
o-o22mm.; Hentschel (Kerguelen), styU o-i85-o-276 mm., chelae o-oi9-o-022 mm.;
Br0ndsted (New Zealand), styh o-2-o-29 by o-ooy-o-ooS mm., chelae 0-025 mm. ; Dendy
(New Zealand), styli 0-3 by o-oi2 mm., chelae 0-028 mm. From this it is clear that
Stephens' Esperiopsis informis is also a synonym of Amphilectus fucorum. In addition to
those already referred to, I have found several small fragments of the same species
among the Terra Nova collections which came to my notice too late to be included in
my report on those collections. The distribution is thus extended to the Antarctic.

Of the eight specimens in the present collection, three are typical in form, the first
(Plate LI V, fig. 4) being encrusting, the others (Plate LIV, figs. 1-3) massive with incipient
branching processes. Of the remaining five, one is small and massive, the rest are
large and massive, quite unlike anything hitherto recorded for the species. At the same
time, the structure of the skeleton in all the specimens is the same, and is identical with
that 'of the European forms of A. fucorum, and the size of the spicules very similar.
The latter is in some cases greater than any hitherto recorded for the species. The
dimensions are: styli 0-27-0-48 by 0-009-0-019 mm., the chelae 0-021-0-028 mm., the
figures representing the range of average spicule size for the eight specimens.

There has been a tendency in the past to confuse Esperiopsis edwardii, var. amencana,
with the present species. I take the opportunity therefore to include a photograph of the
typical form of the variety (Plate LIV, fig. 5)- In all probability, E. rugosa and its var.
major are synonyms of this form, which I propose to regard as a separate species,
Amphilectus americanus (Ridley and Dendy), distinguished by its external form.

Distribution. Europe; South Africa; Falklands; New Zealand; Antarctic.

Amphilectus rugosus (Thiele).

Esperiopsis rugosa, Thiele, 1905, p. 440, H- 5o; E. rugosa, var. major, Hentschel, 1914, P- 68;
Amphilectus rugosa, var. major. Burton, 1929, p. 430.
Occurrence. St. 6 : Tristan da Cunha, 80^140 m. ; St. WS 85 : Falkland Islands, 79 m-
Distribution. Calbuco, Chile; Victoria and Wilhelm Lands, Antarctic.

Amphilectus flabellata, sp.n. (Plate LIII, fig. 5).

Holotype. B.M. 28. 2. 15. 409.

Occurrence. St. WS 88: Falkland Islands, 96-127 m.

Diagnosis. Sponge massively flabellate, erect; surface uneven, minutely conulose,
hispid; oscules and pores not apparent; colour, in spirit, dark brown; skeleton a sub-
isodictyal, usually triangular and unispicular, reticulation with occasional multispicular
bands running irregularly through sponge; no special dermal skeleton; little or no
spongin; megascleres smooth styli, usually slightly curved, 0-285 by 0-009 mm. ; micro-
scleres palmate isochelae of usual type, 0-015-0-024 mm. long.

Remarks. The holotype resembles Esperiopsis fucorum, and other allied species, in
the structure of the skeleton and in the size and shape of the spicules. It differs from

TETRAXONIDA 293

them all in external form. The specimen is massively flabellate, 12 cm. high, 14 cm.
across and 2 cm. thick.

Genus Biemna, Gray

Biemna chilensis, Thiele.

B. chilensis, Thiele, 1905, p. 434, fig. 54; B. macrorhaphis, Hentschel, 1914, p. 74, pi. vi, fig. 3.

Occurrence. St. 177 : South Shetlands, 1080 m. ; St. 187 : Palmer Archipelago, 259 m. ; St. WS 243 :
Falkland Islands, 144-141 m.

Distribution. Calbuco, Chile ; Wilhelm Land.

Genus Asbestopluma, Norman

Asbestopluma calyx, Hentschel.

A. calyx, Hentschel, 1914, p. 66, pi. iv, fig. 4, pi. v, fig. 11.
Occurrence. St. 156: South Georgia, 200-236 m.
Distribution. Wilhelm Land.

Genus Arenochalina, Lendenfeld

The genus, established by Lendenfeld (1887, p. 821) for a single species A. mirabilis,
possesses very slender subtylostyli, not oxea as originally stated, and according to
Hallmann (1912, p. 252 footnote), possibly anisochelae. There are several specimens in
the British Museum collection, identified by Lendenfeld, which if not actual type speci-
mens are sufficiently representative of the species to serve in their stead. From these it
is obvious that the species is a degenerate Mycale.

The second species which I propose to assign to the genus differs from the type
species in the structure of the skeleton, but agrees in the shape of its megascleres. The
two may possibly be unrelated, but they are sufficiently alike in some features to make
it convenient to include them provisionally in the same genus, and the diagnosis is
emended accordingly.

Diagnosis. Reduced Mycaleae with main skeleton composed of very slender sub-
tylostyli, having enlarged axial canals; skeleton either a sub-isodictyal network of
spongin fibres cored by spicules, or composed of loose wisps of spicules running
vertically to surface and branching and anastomosing en route ; special dermal skeleton
absent; microscleres, when present, palmate anisochelae, toxa or both.

Arenochalina incrustans, sp.n. (Fig. 19).

Hohtype. B.M. 28. 2. 15. 360.

Occurrence. St. 2: Ascension Island, attached to buoy.

Diagnosis. Sponge encrusting ; surface smooth, slightly conulose ; oscules and pores
not visible; colour, in spirit, white; skeleton composed of wisps of spicules running
more or less vertically to surface, branching and anastomosing as they go; outermost

294

DISCOVERY REPORTS

ends of wisps ending just below ectosome ; little or no spongin ; megascleres subtylo-
styli, occasionally almost stylote, 0-165 by 0-003 mm.; microscleres blunt-ended toxa,
0-07 by 0-002 mm.

0

Fig. 19. Arenochalina incrustans, sp.n. a, section at right angles to surface,
showing "nests" of toxa at centre, x 36; 6, subtylostyle, x 500; c, toxon,
X 500; d, base of subtylostyle, very much enlarged, to show relative size of
axial canal.

Remarks. The species is evidently a degenerate Mycale but whether it represents a
species in which the spiculation has become permanently degenerate, or whether it is
an individual, with skeleton pathologically affected, of a species with more normal
spiculation, it is impossible to say. At all events, the major part of a megasclere consists
of the axial canal (Fig. 19 d).

The toxa are not distributed evenly throughout the tissues, as is usual, but are found
in "nests" (Fig. 19 a).

Genus Acanthorhabdus, Burton
Acanthorhabdus fragilis, Burton.

A.fragilis, Burton, 1929, p. 432, pi. iv, fig. 2, text-fig. 5.
Occurrence. St. 170: Clarence Island, 342 m.
Distribution. Victoria Land.

TETRAXOXIDA

295

Genus Sigmotylotella, gen.n.
Genotype. S. suberitoides, sp.n.

Diagnosis. Mycaleae with main skeleton a confused reticulation of large tylostyli and
dermal skeleton a palisade of small tylostyli set more or less at right angles to surface ;
microscleres sigmata.

Remarks. The structure of the skeleton in Sigmotylotella resembles that of Suberites
very closely, but the shape of the tylostyli and the presence of the sigmata definitely
show it to be a member of the Mycaleae. It seems to be most nearly related to Tylodesma.

Sigmotylotella suberitoides, sp.n. (Fig. 20).
Holotype. B.M. 28. 2. 15. 400.
Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Diagnosis. Sponge encrusting; surface sparsely papillate, markedly hispid; oscules
situated at apices of papillae ; pores apparently distributed generally over surface ; in-
terior cavernous ; texture firm but friable ; colour, in spirit, black ; main skeleton a con-
fused reticulation of tylostyli of varying size arranged singly or in loose bundles ; dermal

o fa c d e

Fig. 20. Sigmotylotella suberitoides, sp.n. Tylostyli: a, normal; b, d, with
tuberculate base; c, with annulate shaft; e, developmental form.

skeleton a dense palisade of small tylostyli directed more or less at right angles to sur-
face; tylostyli of main skeleton with curved shaft and well-developed spherical head,
I -o by o-oi8 mm. ; tylostyli of dermal skeleton similar in shape to those of main skeleton,
0-6 by 0-OI2 mm.; microscleres simple contort sigmata, 0-028 mm. chord.

Remarks. A curious feature of the spicules is that, very often, the bases of the
tylostyli are tuberculate. This is more marked in the tylostyli of the dermal layer.
Further, it is significant that the bases of developing spicules are also tuberculate
(Fig. 20 e). Occasionally, a shaft may bear annular thickenings, and it is not improb-
able that this abnormality may be comparable with the tuberculations on the bases.

Section lOPHONEAE
Genus lophon. Gray

Alebion, Gray, 1867, p. 534 {nee Kroyer): preoccupied; lophon. Gray, 1867, p. 534 (genotype:
Halichondiia scandens, Bowerbank); Pocillon, Topsent, 1891, p. 539 (genotype: Halichondria
hyndmaiii, Bowerbank); lophonopsis, Dendy, 1924, p. 348 (genotype: Halichondria nigricans,
Bowerbank); Burtonella, de Laubenfels, 1928, p. 361 (genotype: Burtonella melanokhemia, de
Laubenfels).

8-a

,g6 DISCOVERY REPORTS

Remarks. The many attempts to establish a clear understanding of those species
of Halichondria, described by Bowerbank, in which the spiculation consists of
acanthostyli, tornota, anisochelae and bipocilla, have led to much confusion and a
multiplication of generic names. Gray {loc. cit.) divided them between two genera,
the first of which, Alebion, for Halichondria liytjdmam, Bowerbank, was preoccupied
{Alebion, Kroyer, 1863, Crustacea, fide Marschall, Nomeclator Zoologiciis, 1873). The
second, lophon, for Halichondria scandens, Bowerbank, is here retained. Later, Topsent
(loc. cit.) established the genus Pocillon for Halichondria hyndmani, which, as Dendy
{loc. cit.) pointed out, is probably a synonym of H. scandens. Pocillon becomes therefore
a synonym of lophon. At the same time, Dendy proposed to divide the species of lophon
into two groups, retaining the name lophon for the first and proposing the name lophon-
opsis for those species without echinating acanthostyli. My own investigations have
shown that some species are devoid of echinating acanthostyli during the early part of
their life, but possess them later, and that in those species in which they normally occur
these spicules may be only sparingly present or even absent from large tracts of the
sponge. As the presence or absence of echinating acanthostyli are the only features by
which a distinction between lophon and lophonopsis can be maintained, the use of the
latter name must be abandoned. Such a variable character obviously cannot be used
for purposes of generic distinction (cf. lophon chelifer). And finally, Burtonella is a
synonym of lophonopsis and is here included in the list of synonyms of lophon.

lophon radiatus, Topsent.

(For synonymy see Burton, 1929, p. 442, and 1931, pp. 512-518.)

Occurrence. St. 123: South Georgia, 230-250 m.; St. 145: South Georgia, 26-35 m.; St. 170:
Clarence Islands, 342 m. ; St. 175 : South Shetlands, 200 m. ; St. 179 : Palmer Archipelago, 4-10 m. ;
St. WS 33 : South Georgia, 130 m.; St. WS 83 : Falkland Islands, 137-129 m.

Distribution. Graham Land; Victoria Land; Wilhelm Land,

lophon proximum (Ridley) (Plate LVII, figs. 1-13; Figs. 21-24).

Alebion proximum, Ridley, 1881, p. 119, pi. x, fig. 8; lophon chelifer, Ridley and Dendy, 1886,
p. 349; /. pattersoni, id., 1887, p. 117; /. chelifer, id., loc. cit., p. 119, pi. xvi, fig. 3, pi. xvii,
figs. I, 3, 8; /. chelifer, Lambe, 1893, p. 30, pi. ii, fig. 7; id., 1896, p. 191; id., 1900, p. 23;
/. chelifer ostia-magna, Wilson, 1904, p. 143, pi. xx, figs. 2, 4, 10, 11, pi. xxiv, fig. i ; /. lamella,
id., loc. cit., p. 146, pi. XX, figs. 3, 7-9, 12, 13, pi. xxiv, figs. 2-4; /. lamella indivisus, id., loc.
cit., p. 149, pi. XX, figs. 14-16 ; /. chelifer, Thiele, 1905, p. 445, fig. 63 ; I.proximus {sic) reticularis,
Hentschel, 1914, p. 39.

Occurrence. St. 6: Tristan da Cunha, 80-140 m.; St. 39: South Georgia, 179-235 m.; St. 42:
South Georgia, 120-204 ni.; St. 51 : Falkland Islands, 105-115 m.; St. 157: South Georgia, 970 m.;
St. WS 83 : Falkland Islands, 137-129 m.; St. WS 85 : Falkland Islands, 79 m.; St. WS 128: Gough
Island, 120-90 m.; St. WS 247: Falkland Islands, 172 m.

Remarks. The Challenger specimens of /. chelifer exhibit a variability in spiculation
totally unsuspected from the description given by Ridley and Dendy. The lectotype.

«

'c

J^

>. M >.C

o
o
a

o -5 00 -a ^-5 u

oh

s

m

O J2 <u<.

o

gcSccE„2c
■ E^-OMpoo'aDo'OMST'^'' ^"5 •"

w «

<1>

«

c c ceo

O On OvO o O C

£- g^^i i" £"£■"§

o O -< - o o "
u u u u c

1/1 M M X o •!• O c.x M n "^ a o n

_g o " o - «i

;s -s -H, e

-"!• >-^ >-

0^ c
n u rl -n

n

= c c Si

O E^ 3

N O " J

I

I

CO..

£4g^

d ^ *^' C *"* P N N N r^, p r', G

-H O - y

£ I i= I I '

rj c| -i- -i- P •*-

;^

c

c

O

c

c

ri o

rl

£

r^

F

£
o

- o

Tf

c

o

t^

I--

X

X

X

o

8

8

r^

:^

o-

l^vO

^

t^

I^

't-

y

X

X

V y

X

X

X

O

•1-

o

8

O 0

o

0

O

o
o
rl

r^oo r^ r^

O "1 o o

rt O

W

T3

3 >H
« O

^ c ^.

c o c

^. — . o c o

4^ u-jf ly-jr- u-,^ lOC

«°So§°SoC

^-- ^ o " o

u C u

— 3 — '

O 3J

c o !=

c z. ^
C " t-

o ~-^

S32c
- I

O
1-1 c

^£

o

1> U-, 0;

4J " rt QJ

o " ^

z

o

c

O
in c

°E

o

^^ W U OJ

o

O O O o

u u o X.

^ r- a> U QJ

~ 1- u Ui

i o o o

zzz

■ O g

- c

■- o
s;Z

o

z

• =i o

0^ 00

X X

o o

ON -1-

X

■+ ■* 00
XXX

o o "-

O "

iri

ri

fl

rt

X X
o o

O "

CO rl

o

M

^0 -i-'^t-
iH M M M

N

X

X

X X X X

X

X

00

o

0 O O 0

CO

00

■^

E E

E £
£ E

fe "-.

£ E ^
EEE2

e

b

JO

o

o

a

r^

lU

X

u

l-r

F

rt

<)

1

8

0

o

c

J3

M

;i

;j *:: «

ar

M

M

or

c

C

C

c

3

3

3

3

u

O

u

o

c

C

C

c

U

W

W

W

ffl

q;

c

(J

>

c ^i

•£

E

is

>>.

1)

«

3

> 5b

«

(11

Ofi

^«

OJ

s.^

^

<s

OJ *i

!§.> -3

-D

3

CO

a»

>

rt

(d

T3 .5.5.5=2
M

C3

i» K 4) rt

u. u. I- q:;

a a a -

(fl Cfl C/5 V

^ S S S S SS ^

w 3 CO eg

J a;; CO « t/; en

^ t« C3 n ra CQ

;>^

M
C

n.

a

f:-

::■

>.

c

w

n

•a
2

^

S 3

•S '^

^

S is

^

CO

■a, u

^CD

c

#

(type

■§,

^■5

^

Q

^ q

en

M M en

en

CO ;=■

•a

c

■a
c

i

c

Q D

JO u

.t; .^ -c

en w en

£00 c r^ g. CD
« in « in '"^

i-i M C/3 L>

•O •'3 -CO

-00 -co ^-' K

■«m-

O

03

ci q

0^00,00,00

^4-.^ ^

Ci q

a

,— >•

I-

^^ (ft

c"

S^-s

^gs^

. tw^^

3 S c
.w ? 0

5 iJ2

00

Wi

00

en

rn

^

nell
elife
nell

^

CO

-2 "5 -3

en

1-

•o

i^

'M

^S-^""

CO

Q

>?^>^o

q

298 DISCOVERY REPORTS

B.M. 87. 5. 2. 116, takes the form of a number of dark brown fragments of a low-
growing sponge with the surface thrown into numerous folds, resembling the adult form
of /. pattersotii. The main acanthostyli have the same form as those of the
juvenile stages of the same species (Fig. 23 c), while the echinating acan-
thostyli have much the same form as those of the adult forms (Fig. 21).
The tornota are truncate at the ends where they are beset with a few
spines (Fig. 22 c). Chelae are rare, but the characteristic trifoliate, oc-
casionally bifoliate, bipocilla are present in abundance (Fig. 24 c). The
second specimen, B.M. 87. 5. 2. 158, from Prince Edward Island, of
which nothing remains except a microscopic preparation, has the same y^

skeleton as the foregoing except that the anisochelae are more numerous
and the acanthostyli slightly longer. The next specimen, B.M. 87. 5. 2.
157, the "young specimen" from the Cape of Good Hope (fide Ridley jX

and Dendy 1887, p. 120), is an extremely thin crust growing on a piece
of coral. In this the main acanthostyli are slightly longer than in the
lectotype and are strongly spined throughout their length, more particu-
larly at the base and at a point just below the apex (Fig. 23 b). Echinating pjg 2 1 . lophon
acanthostyli are present and the tornota have an oval head at each end proximum (Rid-
beset with numerous spines (Fig. 22 6). The latter, in addition to being ley): atypical
scattered among the meshes of the main skeleton, also form a tangential ^f ma ingacan-

° -11 thostyle, x 400.

skeleton in the dermis, where they are thickly packed but not arranged
in a definite reticulation. Anisochelae, measuring o-o 12-0-021 mm. long, are present, the
larger being arranged in rosettes. The bipocilla bifoliate, but occasionally trifoliate with
the smaller end multi-dentate, are sparingly present (Fig. 24 b).

Lambe's (1893) specimens from Vancouver appear to be almost identical with the
lectotype, while Thiele's (loc. cit.) diff'er apparently only in the smaller size of the
acanthostyli. In both cases, however, no echinating acanthostyli were recorded.
/. chelifer ostia-magna, Wilson, from the Galapagos Islands, difi^ers in the larger size of
the acanthostyli and tornota. Again no echinating acanthostyli are recorded. The ex-
ternal form, moreover, is lamellar. The same external form is found in a group of
specimens presented to the British Museum by the late Dr Gilchrist, B.M. 04. 12. 1. 109,
from the Natal coast, some of which are massive with low lamellate outgrowths and some
purely lamellate.

Among the Discovery sponges are a number which approximate to the type of
/. chelifer and must be considered in relation to it. In fact, despite obvious diff"erences,
it can be shown that they are actually conspecific with it and that these differences are
merely fluctuating variations, perhaps ecological in origin.

These specimens vary from encrusting to irregularly massive, massive and spreading,
sub-clathrate or flabellate (and erect). The surface in the massive specimens may be
smooth, ridged, meandrine, plicate or bearing small flabellate processes. The colour
is always dark brown and the texture friable. The skeleton is a sub-isodictyal
reticulation of acanthostyli, echinated at the nodes by smaller acanthostyli, and

TETRAXONIDA

299

the dermis is supported by a tangential layer of tornota. Anisochelae and bipocilla are
present.

In all respects, these specimens appear to be conspecific, despite obvious differences
in the shape and size of the acanthostyli and tornota and in the shape of the bipocilla ;
and this is even more clearly shown when they are compared with the specimens re-
corded under the various names included above as synonyms of /. chelifer. Nevertheless,
it must be confessed that the species is an unusually variable one. The variation in ex-
ternal form is not particularly marked, but in the spicules it is, although the series of
transitions which it is possible to construct leave no doubt as to the identity of all the

t

t'

w

Fig. 22. lophon proximum (Ridley): showing variation in ends of tornota. a, I. chelifer ostia-magna, Wilson
(type); b, I. chelifer, Ridley and Dendy (co-type: B.M. 87. 5. 2. 157); c, /. chelifer, Ridley and Dendy (type);
d, I. chelifer, Ridley and Dendy (Lambe's specimen); e, G iog\f, I. indentatus, Wilson (type); », /. lamella
indivisus, Wilson (type); /;, /. lamella, Wilson (type); i, I. pattersoni (Bowerbank), Ridley and Dendy
(Challenger Sts. 308 and 3ii);y, /. chelifer, Ridley and Dendy (Thiele's specimen); k, D 255; l-l" , D256;
m, D 251 ; n, D 739 ; 0, D 740 ; p, D 257 ; q, D 254, D 258 ; r, D431 ; s, D 252 ;<,/',£) 837 ; u, D 430 ; v, Alebion
proximum, Ridley (type); ro, lophon proximus (Ridley), var. reticularis, Hentschel. All figures x 1200.
Figs. a,f,g,h after Wilson ; fig. d after Lambe ; fig. j after Thiele. Fig. w from a preparation from the type.

specimens with a single species. Some of the variations can be correlated, but only
vaguely, with increase in size, but nowhere is there so clearly marked a sequence as I have
described (1930) in /. radio tus, Topsent. For example, with increase in size the specimen
tends to become flabellate, either erect or spreading over the substratum ; generally, the
average size of the main acanthostyli and the tornota tends to increase, but the tendency
is not well marked ; the dermal skeleton is regular in young encrusting individuals and
becomes more diffuse and irregular in the larger forms, but is never lost ; and the bi-
pocilla tend to become trifoliate (or bifoliate) as the specimen increases in size.

Other interesting features of the skeleton are: (i) the frequency with which the

P 9
m n r -I

Fig. 23. lophon proximum (Ridley), showing variation in size and shape of the main acanthostyH. a, I. chelifer
ostia-magna, Wilson (type); b, I. chelifer, Ridley and Dendy (co-type: B.M. 87. 5. 2. 157); c, I. chelifer,
Ridley and Dendy (type); d, I. chelifer, Ridley and Dendy (Lambe's specimen); e, Giog;f, I. indentatus,
Wilson (type) ; g. I. lamella indivisus, Wilson (type) ;//,/. lamella, Wilson (type) ; /, /. pattersoni (Bowerbank),
Ridley and Dendy (Challenger Sts. 308 and 311); 7, /• chelifer, Ridley and Dendy (Thiele's specimen);
k, D 255 ■,l,D256;m,D 251 ; n, D 739 ; 0, D 740 ; p, D 257 ; q, D 254, D 258 ; r, D 431 ; s, D 252 ; t, D 837 ;
u, D430; V, Alebion proximum, Ridley (type); w, lophon proximus (Ridley), var. reticularis, Hentschel.
All figures x 300. Figs. a,f, g, h after Wilson; fig. d after Lambe; fig.; after Thiele. Fig. w from a pre-
paration from the type.

■4j/

k '^li

Fig. 24. lophon proximum (Ridley), showing variation in size and
magna, Wilson (type) ; b, I. chelifer, Ridley and Dendy (co-type : B.M
met, but U'-b" represent the normal ; c, /. chelifer, Ridley and Dendy
(Lambe's specimen) ; /, /. indenlatus, Wilson (type) ; g, I. lamella
Wilson (type); /, /. pattersoni (Bowerbank), Ridley and Dendy, i is
the same, and i"-i"' are rarely met with; j, I. chelifer, Ridley and
/, D 236 ; m, D 251; n, D 739 ; o, D 740 ; p, D 257 ; q, D 254, D 258;
V, Alebion proximum, Ridley (type). All figures x 1800. Figs. a,f,
fig.y after Thiele.

shape of bipocilla. a, I. chelifer ostia-
. 87. 5. 2. 157), b and b' are occasionally
(type) ; d, I. chelifer, Ridley and Dendy
indivisiis, Wilson (type); h, I. lamella,
the normal spicule, i' is a side view of
Dendy (Thiele's specimen); k, D 233;
r, D 431; s, D 252; t,D837; u, D 430;
g, h after Wilson; fig. d after Lambe;

302 DISCOVERY REPORTS

echinating acanthostyli, persistently overlooked by previous authors, may occur (of.
Table, p. 297) ; (2) the variation in the occurrence and distribution of the anisochelae
and bipocilla.

Distribution. Kerguelen; Cape of Good Hope; Tristan da Cunha; South America
(Patagonia and Chile); Galapagos Islands; Canada (Pacific and Atlantic coasts).

Section TEDANIEAE

Genus Tedania, Gray

(For further notes on the Antarctic species of this genus see p. 342.)

Tedania tenuicapitata, Ridley (Fig. 25/).

T. tefitiicapitata, Ridley, 1881, p. 124, pi. xi, fig. i.

Occurrence. St. WS 73 : Falkland Islands, 121 m.; St. WS76: Falkland Islands, 205-207 m.;
St. WS 79: Falkland Islands, 131-132 m.; St. WS 83: Falkland Islands, 137-129 m.; St. WS 88:
Falkland Islands, 96-12701.; St. WS99: Falkland Islands, 251-225 m.; St. WS239: Falkland
Islands, 196-193 m.; St. WS 243 : Falkland Islands, 144-141 m.; St. WS246: Falkland Islands,
267-208 m.; St. WS 248: Falkland Islands, 210-242 m.; St. WS 250: Falkland Islands, 251-313 m.

Remarks. The holotype of T. tenuicapitata, Ridley is a small sub-pyramidal sponge,
from Chile, probably incomplete, and its external characters are thus somewhat obscure.
Its surface is even, minutely punctate, with several oscules, 2 mm. in diameter, on the
upper margin. The spicules are: styli 0-38 by 0-013 mm., tornota 0-28 by o-oo6 mm.,
and onychaeta o-o8-o-i05 mm. and 0-32 mm. respectively. The several specimens as-
signed to this species by Ridley and Dendy (1887) vary from small and irregularly
massive to flabellate, and form in themselves a well-graded series of transitions from the
holotype to the large flabellate specimen illustrated by these authors {loc. cit., pi. xi,
fig. 5). The surface may be almost smooth or thrown into meandrine folds, to an extent
far greater than suggested by Ridley and Dendy 's illustration ; and there are all stages
between these two extremes. Thus, in one specimen the surface is smooth and even, but
beneath the dermis, and visible through it, the underlying tissues have a meandrine
formation such as is seen in Artemisina diaiiae, Topsent (1908, pi. iii, fig. 4). In extreme
cases, the surface is folded like that of Tedania oxeata, Topsent (1917, pi. iv, fig. 14).
The spicules in the Challenger specimens measure: styli o-35-o-49 by o-oo7-o-oi4 mm.,
tornota 0-28-0-3 by o-oo4-o-oo6 mm., and onychaeta 0'07-o-28 mm., with no apparent
division into groups of diff'erent sizes.

The present specimens vary in form and appearance to a slightly greater extent than
the Challenger specimens, but are still massive to flabellate, with all intermediate forms
between these two, and the surface is even, with the underlying tissues thrown into
meandrine folds of varying size, or itself thrown into coarse and conspicuous meandrine
folds. The spicules measure: styli o-35-o-49 by o-oi-o-oi8 mm., tornota o-i8-o-39 by
0-004-0-0II mm., onychaeta, very variable, not always clearly differentiated into groups
of different sizes, but on the whole they may be represented by the measurements 0-07-
o-i8 mm. and o-28-o-43 mm.

Distribution. Chile.

TETRAXONIDA 303

Tedania massa, Ridley and Dendy (Fig. 25).

T. massa, Ridley and Dcndy, 1886, p. 335; id., loc. cit., i886, p. 53, pi. xi, fig. 4, pi. xxiii,
fig. 2; Paratedania tarantula. Burton, 1929, p. 441. (For further synonymy see Burton,
loc. cit.)

Occurrence. St. 42: South Georgia, 120-20401.; St. 148: South Georgia, 132-148 m. ; St. 152
South Georgia, 245 m.; St. 156: South Georgia, 200-236 m.; St. 158: South Georgia, 401-411 m.
St. 159: South Georgia, 160 m.; St. 160: Shag Rocks, 177 m.; St. 175: South Shetlands, 200 m.
St. 180: Palmer Archipelago, 160-330 m.; St. iSi : Palmer Archipelago, 160-335 m.; St. WS 80
Falkland Islands, 152-156 m.; St. WS 81 : Falkland Islands, 81-82 m.; St. WS82: Falkland
Islands, 140-144 m.; St. WS 83: Falkland Islands, 137-129 m. ; St. WS 87: Falkland Islands, 96-
127 m.; St. WS88: Falkland Islands, 96-127 m.; St. WS 93 : Falkland Islands, 133-130 m.;
St. WS 109: Falkland Islands, 145 m.; St. WS 216: Falkland Islands, 219-133 m.; St. WS 225:
Falkland Islands, 161-162 m.; St. WS 231 : Falkland Islands, 167-159 m.; St. WS 243: Falkland
Islands, 144-141 m.; St. WS244: Falkland Islands, 253-247 m. ; St. WS246: Falkland Islands,
267-208 m.; St. WS 248: Falkland Islands, 210-242 m.; St. WS 250: Falkland Islands, 251-
313 m.

Remarks. Topsent (1917, p. 59), in recording Tedania chorcoti for the third time,
included under this name four specimens having the same spiculation as the holotype of
the species but possessing a pecuHar external form combined with a chitinoid ectosome
and a dermal tangential reticulation of styli instead of the brushes of tornota set at right
angles to the surface. Later, I established the synonymy of these specimens with
Oceanapia tarantula (Kirkpatrick) and removed the species to Paratedania. More
specimens have come to hand in the present collection, and in the light of these it is
clear that Topsent's action in treating this form as a simple variety of T. charcoti
was nearer the truth than I had thought. In one jar, for example, there are five
specimens. The largest is sub-spherical, about 1 1 cm. in greatest diameter, and dilTers
in no obvious respect from the rounded specimens of T. charcoti described by Topsent
(1913, p. 630, pi. v, figs. 3, 7). The same may be said of the smallest. A third specimen
(Fig. 25 a) is fixed to a fragment of shell, and around the base of the sponge runs a
narrow belt of chitinoid ectosome, the rest of the sponge being no different apparently
from the first and second specimens. This chitinoid substance is also secreted around a
portion of the shell. In a fourth specimen the chitinoid belt is wider, and just above it is
an ill-defined, discontinuous pore area (Fig. 25 V). The upper part of the sponge, not
covered with chitinoid ectosome, is again indistinguishable from the spherical form of
T. charcoti. In the fifth (Fig. 25 c) specimen, the chitinoid covering is more extensive.
The only difl^erences from T. charcoti observed in these specimens, apart from the
chitinoid ectosome, are (i) that the ends of the tornata tend to be more irregular and
occasionally slightly spined; (2) the oscules tend to become papillate as the chitinoid
ectosome increases; (3) beneath this special ectosome the brushes of tornota are lost
and the dermal skeleton is composed of a tangential reticulation of styli (which may
sometimes be modified to strongyla) ; (4) some, at least, of the pores become grouped into
a meridional pore area ; (5) the onychaeta are more liable to bear a swelling near one end
(= tylorhaphides of Dendy). Typical specimens of Paratedania tarantula merely show

9-2

304

DISCOVERY REPORTS

an extreme form of this chitinoid development (Fig. 25 d, e). Despite the great dif-
ference between forms such as those shown in Figs. 25 a, b and 25 d, e, there is every
reason to suppose that the one may develop from the other and that they are conspecific.
Further, it is almost certain that the tangential skeleton of styli, the irregularity of the
tornota, etc., are correlated with the development of the chitinoid ectosome and may

Fig. 25 . a-e, Tedania massa, Ridley and Dendy, showing some of the variations in external form ; /, Tedania
tenuicapitata, Ridley, showing similarity of form due to the growth of an encrusting Haliclona around the
basal half.

possibly be induced by it. What gives rise to this ectosome is, however, still a mystery ;
but it is interesting to note that in some of the more typical individuals of T. massa
irregularly rounded patches of such an ectosome occasionally occur and that their
presence is always accompanied by some derangement of the outer layers of the skeleton,
although a tangential layer of styli may not have been formed. This cuticle is so con-

TETRAXONIDA 305

spicuous in several specimens that it is difficult to understand why Ridley and Dendy
should have failed to mention it.

(An interesting case is shown in Fig. 19/ where the base of a specimen of T. tenui-
capitata is covered with an encrusting Haliclona sp. and the sponge has the superficial
appearance of certain specimens of T. massa.)

Tedania vanhojfeni, Hentschel, has so much in common with certain specimens of
Paratedania tarantula that there can be little doubt as to their identity, but re-examination
of the type is necessary to establish this beyond doubt. Further, since the spiculations
of Tedania vanhoffeni and T. actiniiformis are so similar it is probable that the latter
may also be included eventually in the same species.

Paratedania tarantula (Kirkpatrick), originally described from Victoria Land, has the
following well-marked characteristics: Form carrot-shaped, with most of the surface
covered by a thick, light brown cuticle and having a well-marked, equatorial pore area
immediately above the upper limits of the cuticle ; spicules all of large size. According to
the Hmits of the species established by me (1929, p. 441), which further research has not
caused me to modify, the dimensions of the spicules are: styli o-437-o-504 by o-oi8-
0-04 mm., tornota o-344-o-504 by o-ooy-o-oia mm., onychaeta of two sorts o-o64-o-i62
and 0-4-076 mm. These measurements are, on the whole, greater than those of any other
species hitherto described from the Antarctic or adjacent localities, except Tedania massa,
Ridley and Dendy. Further, in eight specimens from the present collection, typical in
all other respects, the sizes of the spicules are: styli o-63-o-78 by o-o28-o-04 mm.,
tornota o-32-o-35 by o-ooy-o-oii mm., and onychaeta o-oy-o-zS and o-6-o-7 re-
spectively. The species seemed therefore to be characterized by the large size of its
spicules and by the peculiarities of external form.

There are, however, several specimens which differ somewhat in external form but
which would necessarily be included in the species. Thus, in six specimens from South
Georgia, the form varies from carrot-shaped to massive and sub-spherical, with
the cuticle restricted to the base of the sponge (cf. also p. 304, Fig. 25). In these,
the dimensions, and other characters, of the spicules conform to the measurements given
above as typical of the species. On the other hand, there are from the Falkland Islands
eleven specimens, varying in form from carrot-shaped to massive and sub-spherical, with
the cuticle restricted to the base of the sponge, in which the spicules have the following
dimensions: styH o-46-o-59 by o-oi-o-oiy mm., tornota, 0-35 by 0-009 mm., and
onychaeta 0-07-0-19 and 0-4-0-6 mm. respectively. Thus, the styli here are noticeably
more slender than in those recorded above.

(It may be noted here, that whereas most of the Falkland Islands specimens are
atypical in form, the onychaeta and tornota show the same characteristics, and the styli
the same tendency to modification to strongyla, though to a lesser degree, as the Ant-
arctic specimens.)

Following on this are three specimens from the Falkland Islands having the skeleton
typical of Paratedania tarantula, but having a more atypical shape than the preceding
specimens. The smallest is spherical, 10 cm. in diameter, and the largest sub-pyramidal

3o6 DISCOVERY REPORTS

and 20 cm. high. None is complete, the base being absent or damaged in each case, but
in only one is there any sign of a cuticle, and in that it is restricted to a small patch on
one side of the sponge.

The next batch of three, also from the Falklands, have a similar form, but more
irregular, and they are larger. The spiculation again is typical of P. tarantula. Here,
however, there is no sign of a cuticle ; but the specimens contain embryos, and these are
identical with the embryos found in the other groups of individuals referred to above as
belonging to this species. There can therefore be no doubt, on embryological evidence as
well as on spicular characters, that the carrot-shaped specimens of P. tarantula, the massive
specimens with a restricted cuticle, and the large spherical or pyramidal specimens with-
out cuticle are conspecific.

Another specimen, also from Falkland Islands, is unattached and discoid (cf. Fig. 56 a),
but has the same spiculation and embryology as the other specimens of P. tarantula.

Turning now to Tedania massa, Ridley and Dendy, with which species Paratedania
taranttda is most closely allied in spiculation, I find that the types of this species are as
follows : holotype and first co-type, large, massive or sub-pyramidal sponges, with ex-
tensive cuticle-covered areas at the base, resembling Fig. 25 b in general appearance but
much larger in size ; massive and without cuticle ; small and encrusting (on the back of a
crab), without cuticle^; and large and thinly flabellate, again without cuticle.^ More-
over, the embryos of Tedania massa are identical with those of Paratedania tarantula, so
that the species must be considered synonymous.

Finally, a group of a dozen specimens, also from the Falklands, belong here but vary
in form from massive, spherical and flabellate (all fixed ?) to discoid and unattached, but
in none of them can any trace of a cuticle be found.

Distribution. South America; Victoria Land; Wilhelm Land; Graham Land.

Tedania spinata (Ridley).

Trachytedania spinata, Ridley, 1881, p. 122, pi. x, fig. 10.

Occurrence. St. 51: Falkland Islands, 105-115 m.; St. 222: Cape Horn, 70-75 m.; St. WS 72:
Falkland Islands, 79 m.; St. WS 75: Falkland Islands, 72 m.; St. WS 76: Falkland Islands, 205-
207m.; St. WS77: Falkland Islands, 110-11301.; St. WS79: Falkland Islands, 131-13201.;
St. WS80: Falkland Islands, 152-156 m.; St. WS 83 : Falkland Islands, 137-12901.; St. WS 88:
Falkland Islands, 96-127 m.;St. WS 91 .-Falkland Islands, 191-205 m.;St.WS 95 : Falkland Islands,
108-10901.; St. WS 108: Falkland Islands, 118-12010.; St. WS 109: Falkland Islands, 14501.;
St. WS 210: Falkland Islands, i6i 01.; St. WS 222: Falkland Islands, 100-106 m. ; St. WS 225:
Falkland Islands, 162-161 01.; St. WS 239: Falkland Islands, 196-193 01.; St. WS 243: Falklaod
Islands, 144-141 ni.; St. WS 249: Falkland Islands, 166 01.

Remarks. The holotype of Trachytedania spinata, Ridley, forms a thin crust on the
valves of a Pecten. Its spiculation consists of smooth or basally spined styli o-i6-o-i9
by o-oo6 mm., tornota o-i8 by 0-004 mm., and onychaeta up to 0-15 mm. There are, in
the present collection, a number of specimens which appear to belong to this species.

' Specimen from St. 163 D (i.e. Australia). The station reference is queried in the original label.
^ The "encrusting" sponge of Ridley and Dendy.

TETRAXONIDA 3°?

In these the spicules are smooth styli, with occasional spines near the base, o-2i-o-28 by
0-005-0-0II mm., tornota o-i8-o-2i by o-oo4-o-oo7 mm., and onychaeta varying in
length from 0-052 to 0-21 mm., not divided into groups of varying size. The specimens
here assigned to Tedania spinata (Ridley) vary in form from sub-spherical (rarely),
irregularly massive (rarely), branched, with stout cylindrical branches, to flabellate (the
typical form) or discoid and unattached. The surface is typically smooth, even, punctured
by commensal Amphipods and the dermis is typically tough and readily separable from
the underlying tissues. Oscules, usually of small size, may be found scattered generally
over the surface or, in flabellate and discoid forms, around the margins.

The styli are, for the most part, smooth, but there is an occasional tendency to a slight
subtylostylote swelling to the base. In addition, spines may occur. The occurrence of
the spines is, however, not only rare but very sporadic. For example, in a single section
all the styli will be found to be smooth except for a small group situated at one point.
Or, again, the spicules may be smooth with perhaps two or three basally spined forms
found at rare intervals. Turning now to the holotype, the significance of this spining can
be readily appreciated. In sections taken at right angles to the substratum it is found
that the styli at the bases of the vertical columns are the most conspicuously spined and
those towards the surface are hardly spined at all. Moreover, small acanthostyli, re-
markably like those, to be described later, found in the embryos of the specimens here
assigned to T. spinata, are very occasionally to be found associated with the normal basally
spined styli set on the substratum. It seems probable therefore that the holotype of
this species represents nothing more than the immediate post-fixation form of the
species ; that there is retained in the adult a tendency to the formation of spines on the
styli, and that this tendency, although almost wholly suppressed in the adult, is yet found
in occasional spicules.

Distribution. Chile.

Tedania charcoti, Topsent.

T. charcoti, Topsent, 1908, p. 30, pi. i, fig. 3, pi. iii, fig. 3, pi. v, fig. 6; 1913, p. 630, pi. v,
figs. 3-7; nee Topsent, 1917.

Occurrence. St. 39: South Georgia, 179-235 m.; St. 140: South Georgia, 122-136 m.; St. 145:
South Georgia, 26-35 "i-; St. 146: South Georgia, 728 m.; St. 148: South Georgia, 132-148 m.;
St. 152: South Georgia, 245 m.; St. 158: South Georgia, 401-41 1 m.; St. 160: Shag Rocks, 177 m.;
St. WS 27: South Georgia, 106-109 m.; St. WS 42: South Georgia, 198 m.

The nineteen specimens range from a few centimetres in longest diameter to nearly
20 cm. From the structure of the skeleton it is possible to divide this group into two
smaller groups; the first, of five specimens, is characterized by a dense skeleton and
thick spicules (styli 0-36-0-49 by o-02-o-03i mm.), with the dermal skeleton composed
of a very dense palisade of stout tornota (measuring o-32-o-39 by o-oi4-o-o2i mm.).
The second group, of eleven specimens, is characterized by a loose skeleton composed
of few spicules, mainly slender (styli o-35-o-52 by o-oii-o-o2i mm.), with a diffuse
dermal skeleton of slender tornota (measuring o-24-o-34 by o-oo7-o-oo9 mm.). There

3o8 DISCOVERY REPORTS

are in addition three specimens with skeletons somewhat intermediate in character be-
tween those of these two groups. In external form the specimens of each group show the
same variations ; from massive with digitate or flabellate processes, to massive, with the
surface raised into low rounded prominences, and, in rare cases, sub-clathrate. Further,
all the five specimens of the first group are from St. WS 27 (South Georgia, 106-9 ™-
depth) and four of the second group are from the same locality. The remaining speci-
mens of group 2 are also from South Georgia, from depths varying between 132 and
411 m. The similarity in external form between the members of the two groups, the
presence of intermediates as regards skeleton structure, and the similarity between the
two groups in spiculation, except for size of spicules and numbers present, suggest that
there is no taxonomic significance in this sub-division. Similarly, since all the specimens
from the first, and a third of the specimens from the second group are from the same
station, an ecological significance is hardly a possibility. It is, however, remarkable that
seven specimens of group 2, out of a total of eleven, contain embryos while none of the
specimens of group i has any. Possibly the looseness of the skeleton may have something
to do with the development of the embryos, and we may therefore accept all nineteen as
conspecific.

Several of the specimens of this group are almost identical with the co-type of Tedania
charcoti, Topsent (1908, pi. iii, fig. 3). One or two of them approximate fairly closely
to the holotype {id., he. cit., pi. i, fig. 3), and there are several which difi"er slightly
from both. There can, however, be little doubt that all belong to T. charcoti. Here,
however, another point arises: Topsent (1908) records only two sorts of onychaeta in
his specimens, but in the present specimens exceptions are found to this. In the types
of this species, these spicules measure o-09-o-i2 and o-25-o-265 mm, long. In eleven
of the present specimens three sizes of onychaeta were found, the smallest measuring
from 0-05 to 0-105 mm., with an average of 0-08 mm., the largest, 0-24-0-32 mm.,
with an average of 0-29 mm. Intermediate sizes were found, admittedly rare,
varying from 0-105 to 0-24 mm., with an average of 0-163 "^rn. In six specimens,
no intermediates were found, but in two specimens four separate categories appeared
to be present. Moreover, in some specimens the onychaeta were rare, in others com-
monly present, and in rare cases so abundant that they formed the bulk of the skeleton.
It does not seem possible therefore to use the onychaeta for the determination of species.

Distribution. Graham Land; Burdwood Bank.

Tedania murdochi, Topsent.

T. murdochi, Topsent, 1913, p. 629, pi. v, fig. 5.
Occurrence. St, 53 : Falkland Islands, 0-2 m.

Remarks. The typical form of T. murdochi, Topsent (1913, pi. v, fig, 5), is massive,
with mammiform lobes bearing oscules at their summits. The surface is usually
wrinkled, and in the present specimens has somewhat the appearance presented by the
holotype of T. charcoti, Topsent (19 13, pi. i, fig. 3). The colour is deeper than in the

TETRAXONIDA 309

Other Antarctic species, being a deep olive-green or greenish brown. The spicules of
the holotype are: styli 0-235 by o-oo8 mm., tornotao-22 by 0-005 mm-> and onychaeta,
not differentiated into two groups, 0-04-0-175 mm. In the present specimens the sizes
are: styli 0-21-0-22 by 0-007 nim., tornota 0-18-0-2 by 0-004-0-005 mm., and onychaeta
0-07-0-21 mm.

Distribution. Falkland Islands.

Tedania oxeata, Topsent.

T. oxeata, Topsent, 1916, p. 169; 1917, p. 61, pi. iv, fig. 14, pi. vi, fig. 19; Burton, 1929, p. 441.
Occurrence. St. 170: Clarence Island, 342 m.
Distribution. Graham and Victoria Lands.

Section MYXILLEAE
Genus Myxilla, Schmidt

Myxilla mollis, Ridley and Dendy (Plate LV, figs. 1-4).

Myxilla mollis, Ridley and Dendy, 1886, p. 471 ; M. spongiosa, id., loc. cit., p. 471 ; M. mollis,
id., 1887, p. 133, pi. xxvii, fig. 4; M. spongiosa, id., loc. cit., p. 134, pi. xxvii, fig. 3 ; Kirkpatrick,
1908, p. 28; Hentschel, 1914, p. 97, pi. vii, fig. 5; M. magna, Topsent, 1916, p. 168; id., 1917,
p. 56, pi. iii, fig. 4, pi. vi, fig. 9.

Occurrence. St. 42: South Georgia, 120-20401.; St. 123: South Georgia, 230-25001.; St. 159:
South Georgia, 160 m. ; St. 175 : South Shetlands, 200 m. ; St. 181 : Palmer Archipelago, 160-335 "i- '>
St. WS 27: South Georgia, 106-109 "i-; St. WS 42: South Georgia, 198 m.; St. WS 76: Falkland
Islands, 205-207 m.; St. WS 79: Falkland Islands, 131-132 m.; St. WS 83 : Falkland Islands, 137-
129 m.; St. WS 84: Falkland Islands, 75-74 m.; St. WS 88: Falkland Islands, 118 m.; St. WS 99:
Falkland Islands, 251 m. ; St. WS 225 : Falkland Islands, 162-161 m. ; St. WS 239 : Falkland Islands,
196-193 m. ; St. WS 244 : Falkland Islands, 253-247 m. ; St. WS 246 : Falkland Islands, 267-208 m. ;
St. WS247: Falkland Islands, 172 m.; St. WS250: Falkland Islands, 251-313 m.; St. MS 71 :
South Georgia, 120-60 m.

Diagnosis. Young sponge cylindrical or flabellate, with smooth, even surface ; adult
sponge massive, sub-spherical or massively flabellate with surface more or less conulose,
or thrown into low, rounded tubercles ; main skeleton, at all times, ranging from regu-
larly isodictyal to sub-isodictyal with triangular meshes, often irregularly confused;
dermal skeleton a tangential layer of tornota in young sponges, which is gradually lost
with maturity, tornota being then arranged in dermal brushes or else almost absent from
dermis ; megascleres smooth, straight or curved styli or, occasionally, subtylostyli with
long oval heads, with one or a few spines at base in young forms ; microscleres ancorae
spatuliferae and sigmata, both forms being very variable in size or even, occasionally,
separated into two or more categories.

Remarks. The species is represented by twenty-one specimens which show that
the variations are of two kinds, due to (i) the fluctuating variations in size of spicule.

310 DISCOVERY REPORTS

arrangement of skeleton and external form, and (2) the changes which accompany
maturity.

The variation in the external form is not considerable and the present specimens agree
closely in this, and in the texture, with the holotype. The form ranges from irregularly
massive to sub-spherical or flabellate. The surface may be minutely conulose or thrown
into conspicuous rounded tubercles. The largest flabellate specimen measures 9 cm.
high, by 9 cm. across, by 4 cm. thick. The largest sub-spherical specimen is 7 cm. in
diameter. The smallest is sub-spherical and i cm. in diameter.

The skeleton varies between regularly isodictyal, with polyspicular primary fibres
joined at intervals by horizontally arranged single spicules, and sub-isodictyal, with
triangular meshes.

Distribution. Graham Land; Wilhelm Land; Victoria Land; east coast of South
America up to mouth of Rio de la Plata.

Myxilla australis (Topsent) (Plate LIV, fig. 6).

Detidoryx incrustans, var. australis, Topsent, 1901, p. 7; id., 1901, p. 17, pi. iii, fig. 11 ; Myxilla
australis, id., 1917, p. 53, pi. vi, fig. 10.

Occurrence. St. 190: Palmer Archipelago, 93-130 m.; St. WS 27: South Georgia, 106-109 ™^-

Diagnosis. Sponge encrusting or massive; surface smooth, uneven, minutely reti-
culate; oscules large, conspicuous, leading into deep cloacal cavities; pores scattered
generally over surface ; dermal membrane readily separable, translucent, showing large
sub-dermal cavities beneath; colour, in spirit, white; main skeleton a sub-isodictyal
reticulation of spined styli ; dermal skeleton composed of numerous loose bundles of
tornota, set at right angles to surface, with outer ends supporting a tangential reticulation
of similar spicules; styli of main skeleton, sparingly spined throughout length, o-44-o-6
by 0-0 1 5-0-0 1 8 mm. ; tornota, usually in form of straight, slender styli, bearing scattered
spines at each end, o-27-o-35 by o-oo8-o-oi mm.; tridentate isochelae of two sizes,
larger o-047-o-o6 mm., and smaller 0-024 ^^- long; sigmata 0-03-0-06 mm.

Remarks. The present specimen appears to agree closely with the holotype of the
species, except for the presence of the second type of chelae. Since these are small and
not numerous, it is possible that Topsent overlooked them in his original examinations.

Distribution. Graham Land.

Mj^illa asigmata, Topsent.

M. spongiosa, var. asigmata, Topsent, 1901, p. 18; Lissodendoryx spongiosa, var. asigmata, id.,
1908, p. 26, pi. iii, fig. I ; Myxilla spongiosa, var. asigmata, id., 1913, p. 625, pi. iii, fig. 3;
Hentschel, 1914, p. 99.

Occurrence. St. 145: ofi" South Georgia, 26-35 m.

Remarks. Of the specimens hitherto recorded the majority have been encrusting or
massive but spreading. That described by Topsent (1913) was, however, pyriform and
slightly pedunculate. In all the surface is slightly conulose. The present specimen

TETRAXONIDA 3"

therefore appears to correspond fairly closely with that described by Topsent {he. cit.),
but is more complete and better preserved than any yet recorded. It is pyriform and
pedunculate, 70 mm. high, and 60 mm. by 25 mm. across the top. Several small oscules
are situated on the upper surface. The general surface of the sponge is even, glistening
when removed from the alcohol, and very minutely conulose.

The dimensions of the spicules are the smallest on record. Those of the previously
described specimens range as follows: styli o-495-o-9 by o-02-o-028 mm., tornota
0-285-0-4 by o-ooy-o-oi mm., chelae o-04-o-o75 mm. Those of the present specimen
are: styh 0-3 by 0-009 mm., tornota o-i8 by o-oo6 mm., chelae o-03-o-o63 mm. The
most noticeable feature of this variation in spicule size is that although the two categories
of megascleres vary considerably in size, the chelae are almost constant. This holds true
for most species of sponges, that while the microscleres may be almost constant in size,
the megascleres may be subject to very large ranges of variation.

Distribution. Graham Land ; Victoria Land ; Wilhelm Land.

Myxilla chilensis, Thiele (Plate LIV, fig. 10).

M. chilensis, Thiele, 1905, p. 443, figs. 22, 62.
Occurrence. St. 58: Falkland Islands, 0-5 m.

Remarks. The single example is growing in intimate association with an alga, the
sponge tissues forming a thin coating on the branches of the alga. The skeleton is quite
typical except that the acanthostyh are smaller than those of the holotype and are more
tapering at the distal end, resembling more the acanthostyli of Hymedesmia laevis,
Thiele (loc. cit., pi. xxxi, fig. 69 b).

Distribution. Chile.

Myxilla elongata, Topsent.

M. elongata, Topsent, 1916, p. 168; 1917, p. 54, pi. iv, fig. 3, pi. vi, fig. 11; Burton, 1929
P- 435-
Occurrence. St. 158: South Georgia, 401-41 1 m.; St. 160: Shag Rocks, 177 m.

Remarks. The two specimens agree closely in all respects, the larger being irregularly
lobose and white in colour, with smooth but even surface. Both difl^er from the type in
size, the arrangement of the tornota, the more feeble spining of the tornota, and in the
presence of echinating acanthostyli (see Burton, loc. cit.).

Distribution. Graham Land; Victoria Land.

Myxilla basimucronata, sp.n. (Fig. 26).

Holotype. B.M. 28. 2. 15. 462.

Occurrence. St. 42: South Georgia, 120-204 m.; St. WS27: South Georgia, 106-109 m.; St.
WS 42: South Georgia, 198 m.

Diagnosis. Sponge of small size, sub-ramose with cylindrical branches or forming
dorso-ventrally flattened lobes; surface smooth, even; dermal membrane readily
separable; sub-dermal cavities spacious; pores and oscules not apparent; colour, in
spirit, pale brown ; main skeleton a regular reticulation of smooth styli, 0-47 by 0-017 mm..

3i:

DISCOVERY REPORTS

slightly curved and bearing a basal mucron; dermal skeleton a tangential layer of
tornota regularly crossing each other in all directions, supported by brushes of tornota
planted directly on distal ends of ascending fibres of main skeleton; tornota inequi-
ended, 0-25 by o-oo8 mm., with ends slightly swollen and bearing a terminal mucron
surrounded by numerous microspines ; microscleres sigmata of two sizes, 0-042 and
0-025 rnrn- chord, and ancorae spatuHferae of two sizes, 0-042 and 0-021 mm. chord.

Remarks. The species is very closely related to M. magna, Topsent, and it is indeed
possible that the two may eventually prove to be identical, but the present species differs
from Topsent's species in external form, in the dimensions of its spicules and in the
presence of a basal mucron on the styli.

A

n /

] (1

fX

Fig. 26. Myxilla basimucronaia, sp.n. a,
style, X 240; b, end of typical tornote,
X 500; c, sigmata, x 500; d, chelae, x 500.

Fig. 27. Myxilla verrucosa, sp.n. a, styli, and modifications,
X 300 ; b, tornota, showing variation in shape of ends, x 300 ;
chelae, x 500.

Myxilla verrucosa, sp.n. (Fig. 27).
Holotype. B.M. 28. 2. 15. 82S.
Occurrence. St. WS 243 : north-east of Falkland Islands, 144-141 m.

Diagnosis. Sponge small, irregularly massive; surface uneven, verrucose; texture
friable ; oscules and pores not apparent ; colour, in spirit, greenish grey ; main skeleton

TETRAXONIDA 313

a sub-isodictyal reticulation of styli (often modified to incipient acanthostyli) ; styli of
main skeleton, often bearing a few spines at the base, 0-24 by o-oo8-o-oi8 mm. ; dermal
tornota, straight, with ends strongylote to subtylote, often minutely spined, o-i6 by
0-007 iTim.; chelae, 0-035 "^"^- chord.

Remarks. The species may be distinguished from other Antarctic species of Myxilla
by the absence of sigmata.

Genus Ectyodoryx, Lundbeck

Ectyodoryx paupertas, subsp. nobile (Ridley and Dendy).

Myxilla nobilis, Ridley and Dendy, 1886, p. 473 ; M. nobilis, vnr.patagonica, id., loc. cit., p. 473 ;
M. nobilis, var. bacillifera, id., loc. cit., p. 473; M. nobilis, id., 1887, p. 140, pi. xxvii, fig. 15,
pi. XXX, fig. 2; M. ?iobilis, var. patagonica, id., loc. cit., p. 142, pi. xxvii, fig. 13 ; M. nobilis, var.
bacillifera, id., loc. cit., p. 142, pi. xxvii, fig. 14; Stylostichon nobile, var. patagonicum, Topsent,
1913, p. 622; Ectyodoryx nobilis. Burton, 1929, p. 399.

Occurrence. St. 157: South Georgia, 970 m. ; St. 158: South Georgia, 401-41 1 m.; St. 159: South
Georgia, 160 m.; St. WS79: Falkland Islands, 131-132 m.; St. WS82: Falkland Islands, 140-
144m.; St. WS83: Falkland Islands, 137-129 m.; St. WS86: Falkland Islands, i5i-i47m.;
St. WS 225 : Falkland Islands, 162-161 m. ; St. WS 239 : Falkland Islands, 196-193 m. ; St. WS 243 :
Falkland Islands, 144-141 m. ; St. WS 247 : Falkland Islands, 172 m. ; St. WS 249 : Falkland Islands,
166 m.

Remarks. The five specimens in the collection show clearly that it is unnecessary to
maintain the varieties patagonica and bacillifera. It is quite evident that the species is
variable in so far as the characters of the spicules are concerned. The bases of the styli,
for example, may be simply irregularly tuberculate, or may bear a few spines, or they
may be crowded with numerous small spines. The tornota may be simply strongylote at
each end, or the ends may be slightly inflated, and in both cases a few small spines may
or may not be present. An important feature which Ridley and Dendy overlooked is the
diff'erentiation of the chelae into two categories, the larger being approximately twice the
size of the smaller.

As has already been suggested by Ridley and Dendy (1887, p. 143) and Arnesen
(1920, p. 23), I propose to regard Myxilla nobilis and M. paupertas as forms of one
species, the former and its varieties constituting a subspecies distributed throughout
the South Atlantic and the immediately adjacent waters, and the latter as a subspecies
confined to the North Atlantic. The difference between these two subspecies consists
almost entirely in the larger size of the spicules and the slightly more profuse spining in
the styli of Ectyodoryx paupertas, subsp. typica, subsp.n. This subspecies includes, in
addition to the forms mentioned by Topsent (1904, p. 168), Anchinoe nobilis, Arnesen
(1920, p. 22).

Distribution. Crozet Islands; Patagonia; off the mouth of the Rio de la Plata;
Burdwood Bank.

314 DISCOVERY REPORTS

Ectyodoryx antarctica (Hentschel).

Lissodmdoryx antarctica, Hentschel, 1914, p. 102, pi. vii, fig. 9; Burton, 1929, p. 399.
Occurrence. St. 160: Shag Rocks, 177 m.

Remarks. The present specimen agrees with Hentschel's holotype from the Ant-
arctic in external form, in the shape of the acanthostyh and tornota, and, fairly closely,
in the dimensions of the spicules. It differs in the presence of small, echinating acantho-
styh, and in the somewhat broader shape of the chelae. The dimensions of its spicules
are: large acanthostyh 0-195 t>y 0-012 mm., echinating acanthostyh 0-09 by o-oo6mm.,
tornota o-i8 by 0-004 mm., larger chelae o-02i mm., smaller chelae 0-013 mm., sigmata
o-oi8-o-025 mm.

As regards the presence of the echinating acanthostyh, there is a specimen from
McMurdo Sound, in the British Museum collection, which I had identified as a per-
fectly typical example of L. antarctica. On re-examination, however, this was found to
have echinating acanthostyh also, but they were rather rare, and I had overlooked them
in my first examination. Possibly therefore the holotype of the species may also
possess echinating acanthostyles which were overlooked by Hentschel.

The dimensions of the spicules in the McMurdo Sound sponge are: large acantho-
styh 0-24 by 0-004 mm., echinating acanthostyh 0-15 by 0-009 mm., tornota 0-189 by
0-006 mm., larger chelae 0-026 mm., smaller chelae 0-018 mm., sigmata 0-02-0-03 mm.
This specimen thus approximates more closely to that described by Hentschel in the
dimensions of its spicules.

Assuming that the species rightly belongs to Ectyodoryx, we may say that it has con-
siderable resemblance to Ectyodoryx (Myxilla) compressa (Ridley and Dendy) and, to a
lesser extent, to Ectyomyxilla kerguelensis, Hentschel.

Distribution. Wilhelm Land ; Victoria Land.

Ectyodoryx frondosa (Ridley and Dendy), var. anacantha, Hentschel.

E.frondosa, var. anacantha, Hentschel, 1914, p. 107, pi. iv, fig. 11, pi. vii, fig. 12.
Occurrence. St. 190: Palmer Archipelago, 93-130 m.
Distribution. Wilhelm Land.

Ectyodoryx ramilobosa (Topsent).

Dendoryx ramilobosa,Topsent, igiG.p. 167; id., 1917, p. 47, pi. iii, fig. 3- P'- vi, fig- 6; Ectyo-
doryx ramilobosa. Burton, 1929, p. 399.

Occurrence. St. 159: South Georgia, 160 m.

Remarks. It will probably be found, when a comprehensive survey of the various
species of Myxilleae is made, that this species differs generically from the other species
of Ectyodoryx. The columns of smooth, basally spined styli, echinated abundantly by
similar but smaller styli, probably constitute a feature worthy of generic distinction.

TETRAXONIDA 315

Until it is possible to know more concerning the variations in skeletal structure found
in this group, however, it would be unwise to insist on such distinction.

Distribution. Victoria Land ; Graham Land.

Genus Anchinoe, Gray.

Anchinoe latrunculioides (Ridley and Dendy).

(For synonymy see Burton, 1929, p. 439.)

Occurrence. St. 156: South Georgia, 200-236 m. ; St. 159: South Georgia, 160 m.; St. 160: Shag
Rocks, 177 m.; St. WS 84: Falkland Islands, 75-74 m.; St. WS 243 : Falkland Islands, 144-141 m.

Distribution. Graham Land; Victoria Land; Wilhelm Land; east coast of South
America, up to mouth of Rio de la Plata ; South Africa.

Anchinoe areolata (Thiele).

Hytnedesmia areolata, Thiele, 1905, p. 452, figs. 23, 68 ; Kirkpatrick, 1908, p. 24, pi. xxii, fig. 3 ;
Burton, 1931, p. 519, pi. i, fig. i, text-fig. 3.

Occurrence. St. 152: South Georgia, 245 m.; St. 157: South Georgia, 970 m.; St. 160: Shag
Rocks, 177 m.

Distribution. Calbuco ; Victoria Land.

Anchinoe leptochela (Hentschel) (Plate LIV, fig. 11).

Hymedesmia leptochela, Hentschel, 1914, p. 115, pi. viii, fig. 2.
Occurrence. St. 42: South Georgia, 120-204 m. ; St. 159: South Georgia, i6o m.

Remarks. It is a natural suspicion that many of the species of Hymedesmia may
eventually prove to be the encrusting, possibly immediate post-fixation stages, of other
species of Myxilleae. This, at all events, appears to be the case with H. leptochela,
Hentschel. The holotype formed a thin encrustation on foreign bodies (Fremdkorper),
and had a smooth surface with a low papillate oscule. The skeleton showed the usual
Hymedesmia structure. The present specimens range from encrusting to massive, always
growing on fragments of rock, often incorporating much foreign matter in their sub-
stance, and bear a number of papillate oscules. The spicules are almost identical with
those of the holotype, but are arranged as in Anchifioe.

There is no obvious reason why the skeleton in the immediate post-fixation stage of
an Anchinoe should not consist of acanthostyli set at right angles to the substratum, with
their bases planted thereon, nor why, as growth proceeds, the acanthostyli should not
be rearranged to form the ascending fibres, cored by tornota and echinated by acantho-
styli, of a typical Anchinoe.

Distribution. Wilhelm Land.

3i6 DISCOVERY REPORTS

Genus Stelodoryx, Topsent.
Genotype. S. procera, Topsent, 1904, p. 174.

Remarks. The original diagnosis reserves this genus for stipitate " Dendoricinae "
with skeleton composed of longitudinal branching fibres, continuous with the axial
skeleton of the pedicel, connected at intervals by single spicules placed at right angles to
the main fibres. Whether the disposition of the fibres in the genotype constitutes a sound
generic character is an open question. Perhaps the more distinctive feature of its
spiculation is, however, the presence of polydentate isochelae, and while it is doubtful
whether this constitutes a sufficiently good character for generic distinction, a distinc-
tion based on this character alone is as good as that at present maintained between
Myxilla and Lissodendoryx, genera closely allied to Stelodoryx. For the present,
therefore, I propose to adopt the genus Stelodoryx as a convenient arbitrary division of
those Myxilleae in which the distinctive feature is the presence of polydentate isochelae.

Other species which fall naturally into Stelodoryx, as now understood, are Dendoryx
dentata, Topsent {loc. cit., p. 172), Myxilla diversiancorata, Lundbeck (1905, p. 150),
M. pluridentata, Lundbeck {loc. cit., p. 154), and S, discoveryi, sp.n.

Stelodoryx pluridentata (Lundbeck).

Myxilla pluridentata, Lundbeck, 1905, p. 150, pi. v, fig. 3, pi. xv, fig. za-i.
Occurrence. St. WS 243 : Falkland Islands, 144-141 m.

Remarks. Except that the spicules are slightly smaller it is impossible to find any
difli'erence between the present specimen and the type of Myxilla phiridentata. The
chelae, for example, seldom exceed 0-065 ^'n^- chord, as against o-oyi-o-ogy mm., the
tornota vary from 0-24 to 0-28 mm., as against o-226-o-32 mm., although the ends show
the same variability in each case, and the styli, although much the same size, are some-
times slightly tylote in the present specimen. Externally, the two specimens agree
closely except that the type is much smaller, the present specimen measuring 4 cm. by
4 cm. by 5 cm. high.

The specimen from Falkland Islands may be conspecific with the Arctic specimen
(i.e. the type), or the two may represent convergent species which are indistinguish-
able. At all events, the wide geographical separation is a point of interest.

Distribution. Iceland.

Stelodoryx discoveryi, sp.n. (Fig. 28).
Holotype. B.M. 28. 2. 15. 426.
Occurrence. St. WS 88: Falkland Islands, 96-127 m.; St. WS 90: Falkland Islands, 82-81 m.

Diagnosis. Sponge irregularly massive ; surface even and smooth to the naked eye,
very minutely hispid when viewed with a hand lens; oscules 1-2 mm. in diameter,

A

TETRAXONIDA 317

arranged in linear series on ridge-like portions of surface, or at apices of lobular
outgrowths; pores apparently arranged generally over surface; texture compressible
but elastic ; colour, in spirit, very dark brown ; main ^^ ^,<y^
skeleton an isodictyal reticulation of smooth styli | \ [
forming multispicular primary fibres, about six spi-
cules thick, running vertically to surface, connected
at intervals by irregularly arranged single spicules; Fig. 28. SteMoryx discovereyi, sp.n.
dermal skeleton composed of irregular brushes of Examples of variation in ends of tor-
tornota arranged more or less at right angles to sur- ""^^' ' '°°°'
face; megascleres smooth, often slightly curved, styli, 0-225 by o-oo6 mm., and straight
tornota, with ends mucronate, strongylote or truncate and sparingly spined, occasionally
asymmetrical, 0-165 by 0-005 "^i"- ! microscleres, polydentate isochelae, with usually five
teeth, 0-035-0-045 mm. long.

Remarks. The present species differs from the genotype in the absence of a pedicel
and the smaller size of the styli, from S. dcntata in having smooth styli, instead of
acanthostyli, from S. diversiancorata in having only one sort of isochela, and from
S. pliiridentata, which it most nearly resembles, in the smaller size of the isochelae.

Genus Kirkpatrickia, Topsent.

Kirkpatrickia variolosa (Kirkpatrick).

Tedania variolosa, Kirkpatrick, 1907, p. 279; /(/., 1908, p. 32, pi. xxi, fig. i, pi. xxv, fig. i;
Kirkpatrickia variolosa, Topsent, 1912, p. 3.

Occurrence. St. 149: South Georgia, 200-234 'ii-! St. 175: South Shetlands, 200 in.; St. WS 27:
South Georgia, 106-109 n^-

Remarks. The specimens in life were coloured bright crimson or deep magenta red.
Distribution. Victoria Land.

Kirkpatrickia coulmani (Kirkpatrick) (Plate LV, fig. 5).

Tedania coulmani, Kirkpatrick, 1907, p. 280; /(/., 1908, p. 33, pi. xxi, fig. 2, pi. xxv, fig. 2;
Kirkpatrickia coulmani, Topsent, 19 12, p. 3.

Occurrence. St. 42: South Georgia, 120-204 m.

Remarks. The present specimen agrees closely with the holotype except that the
styles are entirely smooth.

Distribution. Victoria Land.

Genus Acheliderma, Topsent.

The genus belongs obviously to the Myxilleae. Owing to Kallmann's (1920, p. 768)
diagnosis of Allocia, there appears, at first sight, to be some similarity between this
genus and Acheliderma. This is due to his having suggested the possibility of the
occurrence of toxa in forms, as yet undescribed, allied to the genotype of Allocia,
Spanioplon cheliferiim, Hentschel. It was probably the inclusion of the phrase " perhaps

3i8 DISCOVERY REPORTS

sometimes accompanied by toxa," in his diagnosis, that caused Kallmann to overlook
the obvious affinity of S. cheliferum with Ectyodoryx, which Hentschel (191 1, p. 363)
himself had suggested.

Acheliderma topsenti, sp.n. (Plate LIV, figs. 7-9).

Holotype. B.M. 28. 2. 15. 393.

Occurrence. St. 160: near Shag Rocks, 177 m.

Diagnosis. Sponge small, flabello-digitate ; surface uneven, hispid ; oscules and pores
not apparent; texture tough; colour, in spirit, yellow; skeleton a sub-isodictyal reticula-
tion of styli, with main fibres semi-plumose, echinated by acanthostyles ; a confused
layer of tornota occurs in dermis and tornota are found sparingly associated with main
skeleton ; main skeleton contains a small amount of spongin ; microscleres, isochelae
palmatae and toxa; styli of main skeleton, long, smooth, but bearing a few feeble spines
at base, slightly curved, 0-42 by 0-009 mm. ; acanthostyli, slightly curved, regularly and
sparingly spined throughout length, 0-15 by 0-007 mm. ; tornota, smooth, straight,
truncate, bearing a crown of a few spines at each end, 0-225 t»y 0-003 ^^^- > isochelae
palmatae, 0-021 mm. long; toxa long, smooth, ends not spined but ending in a sharp
point, 0-06-0-48 by 0-007 "fi"^-

Remarks. The present species difi^ers from the genotype, A. lemmscata, Topsent,
from the North Atlantic, in the longer toxa and in possessing isochelae instead of
raphides. Otherwise the two species agree very closely in spiculation.

Genus Inflatella, Schmidt

Inflatella belli (Kirkpatrick).

(For synonymy see Burton, 1929, p. 439.)

Occurrence. St. WS83: Falkland Islands, 137-129 m.; St. WS248: Falkland Islands, 210-
242 m.

Distribution. Victoria Land ; Wilhelm Land ; New Zealand.

Section PLOCAMIEAE

Genus Plocamia, Schmidt

Plocamia gaussiana, Hentschel.

Plocamia gaussiana, Hentschel, 1914, p. 120, pi. viii, fig. 5; Burton, 1929, p. 435.

Occurrence. St. 42: South Georgia, 120-204 m.; St. 148: South Georgia, 132-148 m.; St. 159:
South Georgia, 160 m.; St. 160: Shag Rocks, 177 m.; St. 175: South Shetlands, 200 m.; St. WS 42:
South Georgia, 198 m.

Remarks. Except that the chelae seldom exceed 0-03 mm. length, and that tricho-
dragmata, measuring 0-03 x o-oo6 mm., are present in small quantities, this specimen
is quite typical.

Distribution. Wilhelm Land; Victoria Land.

TETRAXONIDA

319

Section CLATHRIEAE
Genus Clathria, Schmidt
Clathria papillosa, Thiele.

C. papillosa, Thiele, 1905, p. 449, fig. 66.

Occurrence. St. WS 81 : Falkland Islands, 81-82 m.; St. WS 83 : Falkland Islands, 129-137 m.;
St. WS 84: Falkland Islands, 75-74 m.; St. WS 88: Falkland Islands, 96-127 m.; St. WS 95:
Falkland Islands, 108-109 m-

Remarks. Apart from the arrangement of the skeleton, all six specimens are typical,
except for negligible differences in the sizes of the spicules in two of them. The skeleton
is a regular network of triangular mesh, the sides of the meshes being one spicule length
only and composed of three to six spicules, with the acanthostyli confined to the nodes
of the network. In two of the specimens a slight departure from this arrangement is
found in that occasional fibres, formed of bundles of megascleres, are seen running
perpendicularly to the surface. The arrangement of the skeleton is therefore not only
more variable than is usual in the genus Clathria but is not very typical of that genus.

Distribution. Calbuco.

Clathria toxipraedita, Topsent.

C. toxipraedita, Topsent, 1913, p. 620, pi. v, fig. 4, pi. vi, fig. 12.

Occurrence. St. 144: South Georgia, 155-178 m.; St. 190: Palmer Archipelago, 93-13001.;
St. WS 33 : South Georgia, 130 m.

Remarks. The larger chelae, described by Topsent, are absent from this specimen,
which agrees with the holotype in all other respects.

Distribution. Burdwood Bank.

Clathria lipochela, sp.n. (Plate LV, figs. 6-7; Fig. 29).
Holotype. B.M. 28. 2. 15. 352.
Occurrence. St. 51 : Falkland Islands, 105-115 m.

Diagnosis. Sponge pedunculate, irregularly flabellate ; surface un-
even, porose; inhalant and exhalant openings not distinguishable as
such ; skeleton an irregular network of quadrate meshes in which stout
primary fibres may be recognized running vertically to surface ; fibres
containing much spongin, cored by principal megascleres and echin-
ated by acanthostyli; microscleres absent; principal megascleres
smooth styli, occasionally sparingly spined near base, 0-21 by
0-009 i''^'^-; echinating acanthostyli somewhat variable in size, with
spines evenly distributed throughout their length and slightly more , , j^
numerous at base, 0-09 by o-oo6 mm. ; auxiliary spicules subtylostyli, ^. clathria

smooth, straight, arranged in bundles at right angles to dermis or upochela, sp.n.
scattered singly among meshes of main skeleton, 0-15 by 0-003 "^i""- Spicules, x 300.

320

DISCOVERY REPORTS

Remarks. There are few species of Clathria in which microscleres are entirely absent,
and the present species is distinguished from all such hitherto described by the almost
complete absence of spines on the principal megascleres. In appearance it resembles
RaspaxiUa phakellimi, Topsent, but is distinguished from it by the absence of an axial
skeleton and plumose columns of spicules, as well as by the absence of the peculiar
acanthostyli in which the basal half is smooth.

Genus Rhaphidophlus, Ehlers
Rhaphidophlus paucispiculus, sp.n. (Plate LVI, fig. i ; Fig. 30).

Holotype. B.M. 28. 2. 15. 243.

Occurrence. St. 160: Shag Rocks, 177 m.; St. WS 83 : Falkland Islands, 137-
129 m.; St.WS 84: Falkland Islands, 75-74 m-"- St. WS 109: Falkland Islands,
145 m.

Diagnosis. Sponge massively flabellate, with tendency to give off
digitate outgrowths, to massively branching; surface uneven and
minutely conulose ; oscules and pores not apparent ; colour, in spirit,
light brown; main skeleton composed of irregular bundles of styli
running perpendicularly to surface or forming a confused reticulation
of spicules lying at all angles to each other; smaller styli, differing
from styli of main skeleton only in size, are scattered between fibres of
main skeleton and also form a dense dermal skeleton ; dermal skeleton
consisting of a tangential layer of smaller styli with brushes of similar
spicules situated immediately beneath and lying at right angles to it ;
styli of main skeleton smooth, slightly curved, o-54-o-7i by 0-024-
0-027 mm.; auxiliary styH, in dermal skeleton and scattered between
meshes of main skeleton, 0-24-0-31 by 0-006-0-009 mm.

Remarks. The species appears to be a typical Rhaphidophlus, but
differs from all other species of that genus in the absence of echinating
acanthostyli and microscleres. There appears to be Httle difference
between the styli of the dermal skeleton and the auxiliary styli
scattered between the meshes of the main skeleton, and for this
reason they have been treated in the diagnosis as one form. There may
possibly be a slight difference in size, but if so it is hard to appreciate
this difference. Further, it is worth noting that intermediates between
the larger and smaller styli also appear to be present, but whether they
are very young forms of the larger styli or half-grown forms of the
smaller spicules is again difficult to determine.

Fig. 30. Rhaphido-
phlus paucispicu-
lus, sp.n. Spicules
from a, main skel-
eton; b, dermis.
X 200.

Genus Protoclathria, gen.n.

Diagnosis. Clathrieae with skeleton composed of tylostyli and acanthostrongyla,
arranged in a triangular network; there is no dermal skeleton and microscleres are
absent.

TETRAXONIDA

321

Remarks. The genus, with its extremely simple skeleton, is not easy to classify. The
presence of acanthostrongyla suggests at once an affinity with the Ectyoninae, and, for
the time being at all events, I propose to regard it as a primitive Ectyonine, preferably
as one of the Clathrieae.

Protoclathria simplicissima, sp.n. (Plate LVI, fig. 2; Fig. 31).
Holotype. B.jVI. 28. 2. 15. 369.
Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Diagnosis. Sponge pyriform or massive, irregular; surface
even, hispid; oscules and pores not apparent; texture firm but
friable; colour, in spirit, dark grey; skeleton sub-isodictyal, con-
sisting of a densely knit triangular mesh, usually multispicular but
often unispicular in patches; no special dermal skeleton; little or
no spongin; microscleres of two kinds only, tylostyli and acan-
thostrongyla; tylostyli, forming bulk of skeleton, usually curved
or even strongly angulated at or near the centre, often reduced to
styli or, more rarely, to oxea, 0-26 by 0-015 mm. ; acanthostrongyla,
of very rare occurrence, not echinating the skeleton but occupying
same relative position as tylostyh, straight, o-i8 by 0-012 mm.

Remarks. In deciding the systematic position of this species,
the difficulty is increased by the distribution of the acanthostrongyla.
In the holotype they are numerous; in the first co-type they are
rare; and in the third specimen they appear to be absent altogether. Fig.^i.Protoclat/i-
In the last therefore the skeleton consists only of tylostyli, and the "" '''"Pl'''"'"'"'

• I 11 1 1 , r, , sp.n. a, substylo-

sponge might well have been placed among the Suberitidae had it style; h, acantho-

not been possible to make a comparison with the first two specimens. style. > 320.

u

b

Genus Ophlitaspongia, Bowerbank

Ophlitaspongia membranacea, Thiele.

O. membranacea, Thiele, 1905, p. 450, figs. 67, 105.

Occurrence. St. 152: South Georgia, 245 m.

Remarks. The two specimens show a marked resemblance to the holotype of O. mem-
branacea, Thiele, in the shape and disposition of the spicules, but differ in the external
form and in the dimensions of the spicules. The first specimen consists of an erect
digitate lobe bearing a small lateral branch, while the second is massive with incipient
digitate processes at the top. Both have, however, a strongly hirsute surface as in the
holotype. In the second specimen the smaller megascleres measure 0-5 by 0-024 rn"^-
and the larger 1-2 by 0-036 mm. The auxiliary subtylostyli ("diinnen Style" of Thiele,
he. cit., p. 45 1 ) measure 0-3 by 0-003 mm. In the first specimen, the dimensions of these
spicules are much the same except that the larger megascleres often attain a length of

322

DISCOVERY REPORTS

1-5 mm. The chelae are rare in the second specimen and do not exceed o-oo6 mm. in
length, and are apparently absent in the first. The toxa in both range from 07 to 0-07 mm.
long, the larger being o-oo6 mm. thick and strongly spined at the ends.

In spite of the differences from the holotype, I have little hesitation in identifying
these two specimens with Thiele's species. The diff'erences in the external form are
comparatively unimportant, and the difl'^erences in the measurements of the spicules are
no greater than those found in the various specimens assigned to O. thielei (see below),
a closely related species. Regarding the toxa, it is of interest to note that in O. thielei the
large toxa are often rare and difficult to find, and it is possible that large toxa are present
in the holotype of O. membranacea but in such small quantities as to be readily over-
looked. Even if they are actually absent, in view of the observations made on O. thielei
I should still feel justified in identifying the two present specimens with Thiele's species.

Table showing dimensions of spicules in O. membranacea.

Station No.

Styli

Subtylostyli

Toxa

Chelae

42

0-48 X 0-028

0-30 X 0-006

0-1-0-42,
scarce, especially the
larger forms

0-012

42

0-45 X 0-024

0-30 X o-oo6

0-03-0-22,
rare, especially the
larger forms

o-oi

148

0-38 X 0-013

0-23 X 0-006

0-12-0-3,
small toxa rare

0-012

156

0-36 :■: 0-015

0-24 X o-oo6

0-35-0-24,
all forms rare, but
more particularly the
larger

0-015

WS25

0-36 X 0-021

0-30 X o-oo6

0-065-0-48,
abundant

0-009

WS2S

0-51 X 0-018

0-32 X o-oo6

0-09 X 0-54,
small forms rare

0-013

The measurements of the spicules, which are averages, are given in mm.

Distribution. Juan Fernandez.

Ophlitaspongia thielei, sp.n. (Plate LV, fig. 8; Fig. 32).

Holotype. B.M. 28. 2. 15. 219.

Occurrence. St. 42: South Georgia, 120-20401.; St. 148: South Georgia, 132-148111.; St. 156:
South Georgia, 200-236 m.; St. WS 25: South Georgia, 18-27 "^•

Diagnosis. Sponge massive; surface hispid and produced into papillate processes or
meandrine ridges which may be so strongly developed as to give a clathrate appearance,
or may be reduced to a feeble furrowing; pores apparently small and simple, and
scattered evenly throughout dermis ; oscules circular, 1-2 mm. in diameter, with delicate
membranous velum, leading into cavernous sub-dermal crypts ; often there is a single

TETRAXONIDA

deep cloaca with a number of smaller oscules scattered over surface generally ;

firm and slightly compressible; skeleton a dense or

feebly knit reticulation of styli, often showing strongly

marked primary fibres which project slightly beyond

ectosome as surface brushes; auxiliary spicules, found

chiefly associated with dermal brushes, slender sub-

tylostyli; microscleres toxa and chelae; styli of main

skeleton slightly curved, o-36-o-5i by o-oi3-o-028 mm. ;

auxiliary subtylostyli, straight with crown of spines at

basal end and with distal end pointed or, occasionally,

blunted and spined, o-z^-o-^z by o-oo6 mm.; toxa of

varying size, smaller forms slender and strongly bent,

larger stout and strongly spined at ends, o-03-o-54 mm.

long; isochelae palmatae, o-oo9-o-oi5 mm. chord.

Remarks. The species is characterized by its external
form and by the character of the microscleres. In
many respects it resembles Artemisina dianae, Topsent,
and forms, in fact, a connecting Hnk between the two
genera Ophlitospoiigia and Artemisina, so that it becomes
a matter of doubt whether these two genera can be
maintained.

Genus Artemisina, Vosmaer
Artemisina apollinis (Ridley and Dendy).
(For synonymy see Burton, 1929, p. 431.)
Occurrence. St. 42: South Georgia, 120-204 m.

Fig. 32. Ophlitaspotigia thtelei,sp.n.
a, style of main skeleton ; b, auxiliary
subtylostyle ; c, c', toxa, showing ex-
tremes in size; d, chela. All x 200.

Distribution. Wilhelm Land; Victoria Land; Graham Land; Kerguelen; North
Atlantic.

Artemisina plumosa, Hentschel.

A. plumosa, Hentschel, 1914, p. 70, pi. iv, fig. 5, pi. vi, fig. i ; A. plumosa, var. lipochela, id.,
loc. cit., p. 72; A. strongyla, id., loc. cit., p. 72, pi. vi, fig. 2.

Occurrence. St. WS 109: Falkland Islands, 145 m.

Remarks. An irregularly branching and anastomosing specimen having the general
external characters of A. plumosa, var. lipochela, and a similar skeleton except that
microscleres appear to be quite absent.

Distribution. Wilhelm Land.

324 DISCOVERY REPORTS

Genus Dictyociona, Topsent

Dictyociona discreta (Thiele) (Plate LVI, figs. 3-4).

Microciona discreta, Thiele, 1905, p. 447, fig. 65; Dictyociona discreta, Topsent, 1913, p. 618,
pi. iii, fig. 5.

Occurrence. St. WS 79: Falkland Islands, 131-132 m.; St. WS 95: Falkland Islands, 108-109 m.;
St. WS 243: Falkland Islands, 144-141 m.

Remarks. On examining a piece of the holotype, I find that the smooth parts of the
acanthostyli are not so well defined as Thiele's drawings would suggest. Usually spines
are present on those parts, but they may be few and only feebly developed.

Distribution. Calbuco; Gough Island.

Dictyociona terrae-novae (Dendy).

Clathria terrae-novae, Dendy, 1924, p. 353, pi. xii, fig. 5, pi. xiv, figs. 9-13.
Occurrence. St. 157: South Georgia, 970 m.

Remarks. This species shows a strong likeness to Clathria toxipraedita, Topsent, and
C. mortemeni, Br0ndsted, and it is possible that all three may prove to be not only
congeneric but also conspecific.

In the present specimen, the most noteworthy feature is the variability in the main
megascleres, which show all stages between the large smooth styli and the acanthostyli.
The main skeleton consists of columns of styli running towards the surface and con-
nected at intervals by short fibres containing a few spicules only. The whole is echin-
ated with acanthostyU. Auxiliary spicules in the form of slender subtylostyli are also
present. The main megascleres are smooth styli in which the base may be entirely
tuberculate or bear only a few tuberculations ; or the spicule may be sparsely tuberculate
throughout. The apex also may be simple, or spined, or truncate and tuberculate. The
dimensions are 0-42 by 0-015 ^^- The echinating acanthostyli are entirely spined, with
the spines particularly concentrated at the base, and measure o- 18-0-27 t>y 0-003 rn"^-
The auxiliary subtylostyli are smooth and straight, with a crown of inconspicuous spines
at the base, and measure 0-105 by o-oo8 mm. The microscleres are palmate isochelae,
o-oii mm. chord, and toxa, with spiny ends, 0-065-0-09 mm. long.

Distribution. New Zealand.

Genus Axociella, Kallmann
Axociella nidificata (Kirkpatrick).

Ophlitaspongia nidificata, Kirkpatrick, 1907, p. 274; id., 1908, p. 25, pi. xxii, fig. 6; pi. xxvi,
fig. 5; Axociella nidificata. Burton, 1929, p. 433.

Occurrence. St. 148: South Georgia, 132-148 m.; St. 187: Palmer Archipelago, 259 m.; St. 190:
Palmer Archipelago, 93-130 m.

Distribution. Victoria Land.

TETRAXONIDA 325

Axociella flabellata (Topsent).

Ophlitaspongia flahellatci, Topsent, 1916, p. 167; id., 1917, p. 41, pi. i, fig. 4, pi. vi, fig. 2;
Axociella flabellata. Burton, 1929, p. 433.

Occurrence. St. 140: South Georgia, 122-136 m.; St. 149: South Georgia, 200-234 m.
Distribution. Graham Land ; Victoria Land.

Genus Pseudanchinoe, Burton
Pseudanchinoe toxifera (Topsent).

Stylostichon toxiferum, Topsent, 1913, p. 621, pi. iv, fig. 7, pi. vi, fig. 14; Anchime toxifertim,
subsp. antarctica, id., 1917, p. 43, pi. iv, fig. 5, pi. vi, fig. 5; Pseudanchinoe toxiferum. Burton,
1929, p. 434.

Occurrence. St. 160: Shag Rocks, 177 m.; St. WS 237: Falkland Islands, 150-256 m.
Distribution. Graham Land ; Victoria Land ; Gough Island.

Genus Eurypon, Gray
Eurypon miniaceum, Thiele.

E. miniaceum, Thiele, 1905, p. 446, fig. 64; Raspailia irregularis, Hentschel, 1914, p. 121,
pi. viii, fig. 6.

Occurrence. St. 91 : False Bay, South Africa, 35 m. ; St. 156: South Georgia, 200-236 m. ; St. 190:
Palmer Archipelago, 93-130 m.

Remarks. In external form, the two Antarctic specimens in the present collection are
practically identical and are similar to Raspailia irregularis ; but so far as the spiculation
is concerned the first is almost indistinguishable from R. irregularis and the second from
Eurypon miniaceum. The chief difference in spiculation between these two specimens is
that the " eigenthiimliche Acanthostyle " is absent in the one and present in the other,
although when present this spicule is not quite typical, since the base is tylote or subtylo-
stylote instead of stylote. Since the main difference between Raspailia irregularis and
Eurypon miniaceum consists in the absence and presence respectively of this curious
spicule, it seems almost certain in the light of the present observations that the two are
identical. It appears probable that, in addition to its being of variable occurrence, the
spicule is variable in form.

It is most surprising that a specimen of this species should have been found at
False Bay, South Africa. This specimen is sub-spherical, with no visible point of
attachment, and the surface is uneven, conulose and minutely hispid. The oscules and
pores are probably situated in certain deep but irregular depressions found on the surface
of the sponge. The spiculation appears to be typical except that all spicules are slightly
smaller than those of the holotype and the " eigenthiimliche Acanthostyle " is not to be
found.

Distribution. Calbuco, Chile.

326 DISCOVERY REPORTS

Genus Stylotellopsis, Thiele

Stylotellopsis amabilis, Thiele.

S. amabilis, Thiele, 1905, p. 456, fig. 72.
Occurrence. St. WS 27: South Georgia, 106-109 m.; St. WS 83 : Falkland Islands, 137-129 m.

Distribution. Punta Arenas.

Genus Raspaxilla, Topsent
Raspaxilla phakellina, Topsent.

R. phakellina, Topsent, 1913, p. 617, pi. i, fig. 4, pi. vi, fig. 15.
Occurrence. St. WS 81 : Falkland Islands, 81-82 m.; St. WS 84: Falkland Islands, 75-74 m.

Distribution. Burdwood Bank.

Section HYMEDESMIEAE
Genus Hymedesmia, Bowerbank
Hymedesmia irritans, Thiele.

Desmacidon ceratosa, Thiele, 1905, p. 435, fig. 56; Hymedesmia irritans, id., loc. cit., p. 455,
fig. 71. Nee Amphilectus ceratosits, Ridley and Dendy.

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Remarks. The present specimen has the form of a thin crust on a piece of coral. Its
structure is quite typical except that the sigmata and forceps are extremely rare. Also,
it may be noted that the acanthostyli are confined to those tissues immediately in con-
tact with the coral.

From the observations made on the Discovery example of this species, it seems pro-
bable that Desmacidon ceratosa (Ridley and Dendy), Thiele, belongs to the same species.
Both are encrusting, both possess amphitylota and chelae of the same size and shape.
The differences between them rest in the absence of acanthostyli, sigmata and forceps
in D. ceratosa. Since it has now been shown that these three forms of spicules may be
either very rare or very inconspicuous, it is probable that more careful examination may
reveal their presence in those forms identified by Thiele as D. ceratosa.

Hymedesmia, cf. laevis, Thiele.

//. laevis, Thiele, 1905, p. 453, fig. 69.
Occurrence. St. 58 : Falkland Islands, 1-2 m.

Remarks. The specimen agrees well with the holotype of Hymedesmia laevis, Thiele,
except that the tornota are mucronate at each end instead of simply strongylote.

Distribution. Calbuco, Chile.

TETRAXONIDA 327

Hymedesmia simillima, var. antarctica, Hentschel.
H. simillima, var. antarctica, Hentschel, 19 14, p. 112.
Occurrence. St. WS 33 : South Georgia, 130 m. ; St. WS 225 : Falkland Islands, 162-161 m.

Remarks. There is a slight difference from the holotype in the curvature of the shafts
of the chelae.

Distribution. Wilhelm Land.

Hymedesmia longurioides, sp.n. (Fig. 33).

Holotype. B.M. 28. 2. 15.468.

Occurrence. St. 160: Shag Rocks, 177 m.

Diagnosis. Sponge thinly encrusting; surface smooth, even;
oscules in form of low chimney-like chones; pores not
apparent; colour, in spirit, white; main skeleton composed of
acanthostyli of two sizes ; dermal skeleton an irregular tangential
layer of tornota; large acanthostyli almost entirely spined, with
swollen base, 0-35 by o-oi6 mm.; small acanthostyli entirely
spined, 0-105 by o-oo8 mm.; dermal tornota smooth, straight,
sharply but abruptly pointed at each end, 0-32 by 0-007 nfin^i-

Remarks. The acanthostyli are apparently divided into two
distinct categories with occasional intermediates. The surface
of the sponge bears a number of low chimney-like chones, but
whether these are poral or oscular, or both, it is difficult to say.

The species bears a resemblance to H. longurius, Lundbeck,
but differs in the characters of the tornota.

Section CRELLEAE

Genus Crella, Gray
Crella crassa (Hentschel).

Grayella crassa, Hentschel, 1914, p. 95, pi. vii, fig. 3.
Occurrence. St. 175: South Shetlands, 200 m.

Distribution. Wilhelm Land.

a

Fig. 33. Hymedesmia
longurioides, sp.n. a, der-
mal tornote; b and c,
acanthostyli. x 300.

Genus Crellina, Hentschel
Crellina tubifex, Hentschel.

C. tubifex, Hentschel, 1914, p. no, pi. iv, fig. 12, pi. vii, fig. 14.
Occurrence. St. 175: South Shetlands, 200 m.

Distribution. Wilhelm Land.

328 DISCOVERY REPORTS

Family AXINELLIDAE
Genus Hymeniacidon, Bowerbank
Genotype. Hymeniacidon carunciila, Bowerbank, 1858, p. 286 {fide Thiele, 1905, p. 421).
Diagnosis. Axinellidae possessing styli only ; main skeleton a confused reticulation ;
with special dermal skeleton in form of tangential layer of irregularly scattered
spicules.

Hymeniacidon caruncula, Bowerbank, Autt.

Occurrence. On the beach, Saldanha Bay.

Remarks. The two specimens appear to be identical with those described by
Stephens (191 5, p. 438) from the same locality.

Distribution. From the Arctic down the eastern shores of the Atlantic to Saldanha
Bay.

Hymeniacidon sanguinea (Grant), Autt.

Occurrence. St. 271 : Elephant Bay, Angola, 0-5 m.

Remarks. An orange-coloured sponge encrusting clumps of Ostrea.

Distribution. North Atlantic and Arctic Oceans.

Hymeniacidon fernandezi, Thiele.

H.fernandezi, Thiele, 1905, p. 422; Topsent, 1913, p. 615, pi. iii, fig. 6.

Occurrence. St. 53: Falkland Islands, 0-2 m.; St. 145: South Georgia, 26-35 "^-S ^t. 160: Shag
Rocks, 177 m.; St. WS 27: South Georgia, 106-109 m.; St. WS 84: Falkland Islands, 75-74 m.;
St. WS 85 : Falkland Islands, 79 m.

Remarks. One specimen is growing amidst the branches of an alga, in company with
lophon unicornis. In external appearance it resembles the specimen figured by Topsent
{loc. cit.), while its skeleton is very like that of the holotype, with which I have been able
to make a comparison. The styli measure 0-3 by o-oo8 mm.

A second specimen is massive and low-lying, encrusting fragments of shells. The
spicules measure 0-2 by 0-005 "^'^- ^^^ ^^^ sponge resembles very closely in appearance
and texture the specimens recorded above under H. carwiciila, Bowerbank.

There is little doubt, as Topsent has suggested, that the present species has a close
affinity with H. caruncula, and it is difficult to see how a distinction between the two
species can be made except on the grounds of small differences in the size of their re-
spective spicules. Until the variations in H. caruncula have been carefully studied,
however, it is better to regard the two species as distinct.

Distribution. Juan Fernandez; Falkland Islands.

Hymeniacidon torquata, Topsent.

//. torquata, Topsent, 1916, p. 166; id., 1917, p. 40.
Occurretice. St. 6: Tristan da Cunha, 80-140 m.

Remarks. The single specimen is doubtfully assigned to this species.
Distribution. Petermann Island.

TETRAXONIDA

Hymeniacidon dubia, sp.n. (Plate LVI, fig. 9; Fig. 34).

Holotype. B.M. 28. 2. 15. 364.

Occurrence. St. WS 83: Falkland Islands, 137-129 m.

Diagnosis. Sponge massive, clathrate ; surface minutely hispid and
beset with warty protuberances, as though sponge body were composed
of numerous thickly-set branches which had anastomosed, leaving
stunted ends exposed at surface; oscules not seen; skeleton a con-
fused reticulation of, usually, single spicules; spicules styli, usually
thickest at centre and tapering away at each end, 0-48 by o-ii mm.

Remarks. The species approximates more closely to Axinella verru-
cosa, Br0ndsted, from New Zealand, than to any other species of
Axinellidae. It diff'ers from that species in the absence of spicule fibres
and of strongyla and in the size of the spicules. The most important
feature they have in common is that the styli are thickest at the centre
and thence taper gently towards each end, in the manner of the
megascleres of Polymastio, instead of having the same diameter
throughout their length except at the distal end where they taper to a
point, as in the typical Axinellid style. In all respects, however, the
present species is a typical Hymeiiiacidoyi. Further, many of the
spicules, smaller than the rest and probably only young forms, are
regularly stylote, as in the typical Axinellid.

329

Fig. 34. Hytnenia-
cidon dubia, sp.n.
Style, X 200.

Genus Thieleia, gen.n.
Genotype. Hymeniacidon rubiginosa, Thiele.

Diagnosis. Main skeleton composed of delicate strands of spicules running vertically
to the surface and connected to each other by irregularly arranged masses of single
spicules. At the surface, the ends of the strands diverge to form dense brushes of spicules
the apices of which project slightly beyond the dermis. There is no special dermal
skeleton, unless the dense brushes formed from the ends of the strands of the main
skeleton can be regarded as such.

Thieleia rubiginosa (Thiele).

Hymeniacidon rubiginosa, Thiele, 1905, p. 421, fig. 44.
Occurrence. St. 157: South Georgia, 970 m.
Distribution. Iquique.

Analysis of the species of Hymeniacidon of the South Atlantic-Antarctic region
Of the numerous species hitherto assigned to the genus Hymeniacidon, the majority
have proved to be synonyms of H. caruncula and H. sanguinea, or are now known to
belong to other genera. Of the remainder, five are recorded from the South Atlantic-
Antarctic region.

330 DISCOVERY REPORTS

H. fernandezi, Thiele, 1905, p. 422. Skeleton arranged as in H. caruiicida.

H. kergiielensis, Hentschel, 1914, p. 123. Skeleton as in H. cariincula.

H. torquata, Topsent, 1917, p. 40. I have not seen a specimen of this species, but by
comparing Topsent 's description with a portion of the holotype of H. rubiginosa there
appears to be little doubt that the two species are synonymous. According to its author,
the spicules of H. torquata are " generalement marques pres de leur base d'un leger
bourrelet qui la renfle en base de subtylostyle". The same may be said of //. rubiginosa.

H. centrotyla, Hentschel, 1914, p. 125. Skeleton as in Thieleia rubiginosa (Thiele).

Genus Axinella, Schmidt
Axinella crinita, Thiele.

A. crinita, Thiele, 1905, p. 424, fig. 46.

Occurrence. St. WS 82: Falkland Islands, 140-144 m.

Distribution. Calbuco, Chile.

Genus Pseudaxinella, Thiele
Pseudaxinella egregia (Ridley).

Pliatiellia egregia, Ridley, 1881, p. 114, pi. x, fig. 6; Pseudaxinella egregia, Thiele, 1905, p. 426,
fig- 47-
Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Remarks. In his original preparations I find all the spicules Ridley has described
with the exception of the "Setaceous acuate". This I have been quite unable to find.
It must be assumed, therefore, that either this particular spicule is extremely rare or that
Ridley himself has made some mistake about it. The longest spicules in his prepara-
tions do not exceed 0-9 mm. Thiele also remarks that the spicules over o-8 mm. in
length were very rare in his Calbuco sponge.

The present specimens agree closely with the holotype except that the spicules are
more robust and the oxea much more numerous. One specimen is sub-cylindrical,
presumably erect, the other massive, almost pyriform.

Distribution. Calbuco and Sandy Point (South America).

Genus Homaxinella, Topsent

Homaxinella supratumescens (Topsent).

(For synonymy see Burton, 1929, p. 443.)

Occurrence. St. 148: South Georgia, 132-148111.; St. WS62: South Georgia, 15-45 m.; St.
MS 14: South Georgia, 190-110 m.

Remarks. The first specimen is quite typical in all respects. So also is the second,
except that it is attached at the base to the spine of an Echinodenn. The third specimen

TETRAXONIDA

331

consists of a group of three apparently very young specimens, which support Topsent's
(1908, p. 33) view that the species is characterized by rapid growth. The smallest is
100 mm. high and 1-5 mm. thick, the largest 180 mm. high and 2 mm. thick. The body
of the sponge in each case can only be described as incipient, consisting of an incon-
spicuous feathery structure occupying the upper third of the stalk, resembling in ap-
pearance that of Asbestopliima peniiatiila, Schmidt (see Lundbeck, 1905, pi. ii, fig. i)-
The stalk is evidently very quick growing, for it reaches a considerable height before the
body even begins to take form, a state of affairs very similar to that in Stylocordyla
borealis (Loven) (see Burton, 1928, p. 68).

Distribution. Graham Land ; Victoria Land ; Wilhelm Land.

Genus Rhizaxinella, Keller

Rhizaxinella australiensis, Hentschel.

R. aiistralimsis ,\lenXsche\,i^o^, p. 397, pi. xxii, figs. 4, 5; text-figs. 27, 28; Burton, 1929, p. 443.

Occurrence. St. 158: South Georgia, 401-411 m.; St. 175: South Shetlands, 200m.; St. 190:
Palmer Archipelago, 93-130 m.; St. WS 225 : Falkland Islands, 162-161 m.

Distribution. Australia; Victoria Land.

Genus Stylohalina, Kirk
Stylohalina hirta (Topsent) (Fig. 35).

Amorphina hirta, Topsent, 1889, p. 44, fig. 9 A.
Occurrence. St. 283 : Gulf of Guinea, 18-30 m.

Remarks. The larger of the two specimens is
2 cm. by 2 cm. across the base, and i cm. high,
massive with very uneven surface in the sense
that it is thrown into irregular ridges and depres-
sions, and with the surface further thrown into
minute tubercles and cerebriform ridges. The
skeleton consists only of styli, 078 by 0-014 mm.,
which are " remarquables par une courbure con-
stante bien accentuee a peu de distance de la
grosse extremite".

The external form of the present specimens
is very like that of the holotype, except that the
surface is not raised into "des projections greles
et hispides", although it is minutely hispid
throughout. Unfortunately, Topsent says nothing
about the structure of the skeleton, but as the two
specimens agree so closely in all other respects, I
feel sure that the present specimen is conspecific
with the holotype of Atnorphina hirta, despite

- e

Fig. 35. Stylohyalina hirta (Topsent). Section
at right angles to surface, to show the arrange-
ment of the skeleton, e, ectosome, 5, sub-
stratum.

332

DISCOVERY REPORTS

the fact that they were found at opposite sides of the Atlantic. The skeleton in the
present specimen consists of loose bundles of styli running from the substratum to the
surface (Fig. 29).

Distribution. West Indies.

Genus Plicatellopsis, gen.n.
Genotype. P. arborescetts , sp.n.

Diagnosis. Axinellidae with main skeleton a confused or regular reticulation of styli
or tylostyli from which ascending fibres pass outwards to dermis; dermal skeleton in
form of bundles of small styli or tylostyli; interstitial spicules, when present, long
slender tylostyli.

Remarks. The genus includes, in addition to the genotype, P. flabellata, sp.n. and
PlicateUa expama, Thiele. The use of Schmidt's genus Plicatella (1870, p. 45 : genotype,
Reniera labyriiithka, Schmidt, 1864, p. 39, pi. iv, fig. 9) by Thiele is unjustified, since
the only information we have is that the skeleton contains styli only. We know nothing
of the arrangement of these styli, and it is probable that Plicatella is more closely
related to Stylohalina than to any other genus. At the same time, there is a close relation
between Plicatella and Plicatellopsis; but until the type of the former is redescribed
it will be impossible to define it.

b . . . ^ .

Fig. 36. Showing arrangement of skeleton in a, Plicatellopsis expansa (Tliiele),
b, P. arborescens, sp.n., c, P. flabellata, sp.n.

TETRAXONIDA 333

Plicatellopsis arborescens, sp.n. (Plate LVI, fig. 5; Figs. 36 b, 37 h).

Holotype. B.M. 28. 2. 15. 388.

Occurrence. St. WS 83 : Falkland Islands, 137-129 m. ; St. WS 243 : Falkland Islands, 144-141 m.

Diagnosis. Sponge erect, dichotomously branched; surface even, minutely hispid;
texture firm, compressible, friable; oscules and pores not apparent; colour, in spirit,
brown; skeleton an irregular reticulation of large styli giving off fibres which run
vertically to surface ; special dermal skeleton of bundles of small styli associated with
outermost ends of ascending main fibres ; large styli, slightly curved, 0-65 by 0-025 "^'^^ ">
small styli, straight or slightly curved, 0-28 by 0-007 ^^•

Remarks. The skeleton is rather confused, its plan being obscured by the abundance
of isolated spicules scattered between the main fibres. The small styli are chiefly con-
fined to the dermis although a few, not in bundles, are to be seen scattered between
the meshes of the main skeleton. There is no visible spongin.

The species differs from P. expansa, Thiele (Figs. 36 a, 37 a) in the external form and
in the absence of any tylostyli.

Plicatellopsis flabellata, sp.n. (Plate LVI, fig. 6; Figs. 36 c, 37 c).

Holotype. B.M. 28. 2. 15. 726.

Occurrence. St. WS 84: Falkland Islands, 75-74 m.

Diagnosis. Sponge irregularly flabellate, somewhat plicated; surface minutely
tuberculate, markedly hispid; texture soft, friable; oscules and pores not apparent;
colour, in spirit, brown; internal skeleton a confused mass of tylostyli, crossing
each other at all angles, from which irregular fibres run vertically to surface; special
dermal spicules in form of small styli; with long, slender tylostyli lying between
ascending fibres of main skeleton; tylostyli of main skeleton usually much curved,
0-4 by 0-014 mm. ; small styli of dermal skeleton, 0-2 by 0-003 rnin-! long, slender,
flexuous tylostyH, 2-0 by o-oi mm.

Remarks. There is a marked relation between this species and P. expansa, Thiele
(cf. Figs. 34 a, 35 a), and P. arborescens, sp.n., but it differs from both in external form,
in the disposition of the skeleton and in having tylostyli in the main skeleton and small
styli for dermal spicules.

Genus Bubaris, Gray
Bubaris vermiculata (Bowerbank).

(For synonymy see Dendy, 1924, p. 351.)
Occurretice. St. 6: Tristan da Cunha, 80-140 m.; St. 283 : Gulf of Guinea, 18-30 m.
Distribution. Throughout Atlantic Ocean ; Kerguelen ; New Zealand.

D VI 13

b

Fig. 37. a, Plicatellopsis expansa
(Thiele), large style and small
tylostyle; b, P. arborescens, sp.n.,
large and small styli; c-e, P.
fiabellata (c = tylostyle of main
skeleton, c' small style, c", c'"
long flexuous tylostyle, d-d'"
bases of tylostyli of main skele-
ton, e showing shapes of main
tylostyli).

a'

TETRAXONIDA
Bubaris murrayi, Topsent.

B. murrayi, Topsent, 1913, p. 616, pi. iii, fig. i.
Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Remarks. It is probable that this species is synonymous with B. vermiculata.
Distribution. Gough Island.

Genus Ceratopsis, Thiele

Ceratopsis incrustans, sp.n. (Fig. 38).

Holotype. B.M. 28. 2. 15. 435.

Occurrence. St. 6: Tristan da Cunha, 80-140 m.

Diagnosis. Sponge encrusting; surface even, minutely
conulose; oscules and pores not apparent; texture soft, com-
pressible; colour, in spirit, grey; skeleton a multispicular sub-
isodictyal reticulation of styli of two sizes, larger 0-48 by
0-015 mm. and smaller 0-3 by o-oo6 mm.; microscleres,
toxiform microxea, o-i by 0-004 mm., scattered throughout
choanosome.

Remarks. The species differs from all previously known
members of the genus Ceratopsis in the dimensions of its
spicules and in its external form.

335

Genus Eumastia, Schmidt
Eumastia attenuata, Topsent.

E. attenuata, Topsent, 1915, p. 35, figs, i, 2.
Occurrence. St. 53 : Falkland Islands, 0-2 m.

Distribution. Falkland Islands.

\

Fig. 38. Ceratopsis incrus-
tans, sp.n. a, large style; b,
small style; c, c' , toxiform
microxeote. a, b and c,
X 200; c' , X 400.

Family CLAVULIDAE

Genus Suberites, Nardo

Suberites montiniger. Carter.

S. montiniger. Carter, 1880, p. 256; Vosmaer, 1882, p. 31, pi. i, fig. 26, pi. iv, figs. 137-139;
Topsent, 1892, p. 130; Lambe, 1894, p. 128, pi. iv, fig. 4; Swartschewsky, 1906, p. 319, pl.xiii,
fig. 4; Topsent, 191 5, p. 39; Pseudosuberites montiniger, Hentschel, 1916, p. 6; Suberella
topsenti, Burton, 1929, p. 446, pi. iv, fig. 5.

Occurrence. St. WS 239: Falkland Islands, 196-193 m.

Remarks. Since writing my Terra Nova Report, I have had occasion to examine a
number of specimens of S. montiniger, Carter, from the northern hemisphere and am

13-2

336 DISCOVERY REPORTS

now convinced that the sponges recorded by Topsent {loc. cit.) and myself, from the
Antarctic, are identical with Carter's type.

Distribution. Burdwood Bank; Victoria Land.

Suberites microstomus, Ridley and Dendy, var. stellatus, Kirkpatrick.
S. microstomus, var. stellatus, Kirkpatrick, 1908, p. 19, pi. xv, figs. 8-13.
Occurrence. St. 152: South Georgia, 245 m. ; St. 190: Palmer Archipelago, 315 m.
Distribution. Victoria Land.

Suberites papillatus, Kirkpatrick.

S. caminatus, var. papillatus, Kirkpatrick, 1908, p. 20, pi. xv, fig. 16, pi. xvi, figs. 11-14;
Tentorium papillatum, Topsent, 1917, p. 36, pi. iv, fig. 2; Suberites papillatus. Burton, 1929,
P- 445-

Occurrence. St. 156: South Georgia, 200-236 m.; St. 160: Shag Rocks, 177 m.; St. 170: Clarence
Islands, 342 m.; St. 190: Palmer Archipelago, 93-130 m.

Distribution. Graham Land ; Victoria Land.

Genus Pseudosuberites, Topsent

Pseudosuberites hyalinus (Ridley and Dendy).

(For synonymy see Burton, 1929, p. 445.)

Occurrence. St. 6: Tristan da Cunha, 80-140 m.; St. 152: South Georgia, 245 m.; St. 157: South
Georgia, 970 m. ; St. WS 243 : Falkland Islands, 144-141 m.

Distribution. Europe ; Red Sea ; Patagonia ; Graham Land ; Victoria Land.

Pseudosuberites sulcatus, Thiele.

(For synonymy see Burton, 1929, p. 334.)

Occurrence. St. 58: Falkland Islands, 1-2 m.; St. 156: South Georgia, 200-236 m.; St. WS 86:
Falkland Islands, 151-147 m.; St. WS90: Tierra del Fuego, 82-81 m.; St. WS91: Tierra del
Fuego, 191-205 m. ; St. WS 108: Falkland Islands, 1 18-120 m.

Remarks. The collection includes an encrusting form, comparable with Suberites
incriistans, Br0ndsted, two encrusting forms with convoluted surface, and one massive
specimen with papillate outgrowths resembling the holotype of the species. There are,
in addition to these, five large irregularly flabellate specimens, but as this particular
growth form is being dealt with in my forthcoming report on the sponges of the Swedish
Antarctic Expedition I refrain from further comment here.

Distribution. Tierra del Fuego; Gough Island; South Georgia; Victoria Land;
Auckland and Campbell Islands.

TETRAXONIDA 337

Genus Tentorium, Vosmaer.

Tentorium semisuberites (Schmidt).

Thecophora semisuberites, Schmidt, 1870, p. 50, pi. vi, fig. 2; T. ibla, Thompson, 1873, P- ^4^'
fig. 24; T. semisuberites, Whiteaves, 1874, p. 9; T. ibla, Verrill, 1874, p. 500, pi. viii; T. semi-
suberites, Vosmaer, 18S5, p. 18, pi. i, figs. 23, 24, pi. iii, figs. 22-26; Tentorium semisuberites, id.,
1885, p. 329, pi. ii, fig. 4, pi. xxi, fig. 19 ; Ridley and Dendy, 1886, p. 489 ; id., loc. cit., p. 221 ;
Thecophora semisuberites, Fristedt, 1887, p. 433; Tentorium semisuberites, Lambe, 1896, p. 198,
pi. iii, fig. 2; id., 1900, p. 25; Arndt, 1912, p. 113 ; Topsent, 1913, p. 25 ; Ferrer, 1914, p. 19.

Occurrence. St. 6: Tristan da Cunha, 80-140 m.
Distribution. Arctic and Atlantic Oceans.

Genus Polymastia, Bowerbank

Polymastia isidis, Thiele (Fig. 39).

P. isidis, Thiele, 1905, p. 414, figs. 25, 38 a~e; P. isidis, var. simplex, Hentschel, 1914, p. 47,
pi. V, fig. 3.

Occurrence. St. 187: Palmer Archipelago, 259 m.; St. WS81: Falkland Islands, 81-82 m.;
St.WS82: Falkland Islands, 140-144 m. ; St.WS86: Falkland Islands, 151-147 m.; St.WS248:
Falkland Islands, 210-242 m.; St. WS 250: Falkland Islands, 251-313 m.

Remarks. The species is represented by nearly a dozen specimens, all agreeing closely
with each other, but differing slightly from the holotype in external form and minor
details of the spiculation. They are all cushion-shaped, standing on a circular base, and
the upper surface bears a varying number of wart-shaped 1 1
papillae. The papillae are never elongate, as in the type, but
always low, not exceeding 2-3 mm. in height, and in some not
perceptibly raised above the level of the surrounding surface.
The chief difference in the spiculation is the presence of a
pecuUar annulation on many of the large tylostyles. The nor-
mal base of these tylostyles is as shown in Fig. ^q c, but more „. „ , . . .,.
■' ■' o oy J |.jg ^g Polymastia isidis,

than 50 per cent have a base such as that shown in Fig. 39 ^>, Thiele; showing variations
while Fig. 39 a represents the condition in about 10 per cent, in shape of bases of mega-
Many of these spicules are polytylote, as in the type, but scleres. x 300.
wherever this occurs the swellings on the shafts are of the kind shown in Fig. 39 a.

At first sight, it would appear that the differences between the specimens here de-
scribed and the type of the species are sufficient to justify the recognition of a new
variety, but the general similarity in te.xture, appearance of the surface, disposition of
the spicules in the skeleton, and the shape of the normal megascleres make it fairly
certain that the Discovery specimens represent at the most a tropus (sensii Vosmaer)
of P. isidis.

Making this more certain, there are a further nine specimens in the collection which
resemble closely in external appearance the dozen specimens referred to above. These
are mostly cushion-shaped, but one is irregularly massive and one sub-spherical. The

338 DISCOVERY REPORTS

surface of each is even and only minutely hispid ; and in most of them the papillae are
less than 3 mm. high, but in two they are 5 mm. or more high and agree closely with
those of the holotype. As regards the spiculation, the larger megascleres in most of
these nine specimens have the same type of base figured by Thiele {loc. cit.) for
the holotype, a few show traces of the annulation recorded above for the first group
of specimens, and in a few this annulation is fairly strongly marked. In some of the
specimens, also, there is no trace of the polytylote megascleres. We see in this group
therefore a series of typical specimens, together with all transitions between them and
the group of abnormal specimens described above.

The differences between the typical form and Hentschel's var. sifnplex are so slight,
especially when compared with those found in the present specimens, that I can see no
reason at all for maintaining this variety. In fact, if we compare the relation between the
holotype and Hentschel's specimen with the relations between the various individuals
recorded under P. jnammillaris, or any other species of Polymastia, there is not the
slightest justification for the introduction of this variety.

It is probable, to judge by the written description, that Brdndsted's (1923) P. gramdosa
from the Auckland and Campbell Islands will also prove to be synonymous with
P. isidis.

Distribution. Falkland Islands; Wilhelm Land.

Polymastia invaginata, Kirkpatrick.

P. invaginata, Kirkpatrick, 1907, p. 271; /(/., 1908, p. 15, pi. xii, fig. i, pi. xiv, figs. 5-15;
P. invaginata, var. gaiissi, Hentschel, 1914, p. 49, pi. v, fig. 4; P. invaginata. Burton, 1929,
p. 446.

Occurrence. St. 42: South Georgia, 120-204 m.; St. 156: South Georgia, 200-236111.; St. 167:
South Orkneys, 244-344111.; St. 177: South Shetlands, 1080 m.

Remarks. There are three specimens of this species which, although typical in all
essential respects, difi"er in appearance and in the disposition of the spicules ; and these
diff'erences seem to have a bearing on the problem of the extrusion of spicules. The
first specimen is cupola-shaped, 2 cm. across and i cm. high, with smooth minutely
hispid surface and normal spiculation. That is to say, the large megascleres are
arranged in regular radial bundles and the smaller spicules in a dermal palisade and in
stellate groups within the choanosome. There are no large spicules protruding at the
surface. The second specimen is sub-spherical, 2 cm. in diameter, and in this the surface
is smooth, or only minutely hispid, for the most part, but bears several patches of long
protruding spicules. The general appearance is, in fact, very like that of a long-haired
mammal in the last stages of shedding its coat. Moreover, the spicules composing these
shaggy patches are not firmly embedded in the tissues, but may be removed with the
greatest ease, and without the use of the slightest force, with a pair of forceps.
Internally, the spiculation is not quite normal. The radial bundles are normal, but the
characteristic stellate groups of small megascleres are not to be seen, these spicules being
irregularly scattered through the choanosome. The third specimen, incomplete but

TETRAXONIDA 339

larger than the two preceding specimens, is very conspicuously shaggy throughout the
whole surface. Internally, the spiculation is very disordered: the smaller megascleres
are most irregularly distributed and the larger megascleres are not only fewer in number
than is usual but do not form the usual radial bundles. Instead they are arranged in
a very confused manner.

I have (1931) suggested the possibility that the individuals of this species periodically
extrude their spicules and the present specimens appear to offer strong support to this
view. The first specimen is quite obviously in a normal condition. The third, equally
obviously, is not, but has every appearance of being in the process of extruding the bulk
of its larger megascleres at the surface. Specimen 2, on the other hand, appears equally
certainly to be in the final stages of a "moult," with the remaining patches of the ex-
truded spicules still loosely adhering to the surface, while internally the skeleton is
resuming its normal arrangement.

Distribution. Victoria Land ; Wilhelm Land.

Genus Sphaerotylus, Topsent
Sphaerotylus antarcticus, Kirkpatrick.
(For synonymy see Burton, 1929, p. 446.)
Occurrence. St. i6o: Shag Rocks, 177 m.
Distribution. Victoria Land; Wilhelm Land.

Genus Stylocordyla, Thomson

Stylocordyla borealis (Loven), subsp. acuata, Kirkpatrick.

S. borealis, var. acuata, Kirkpatrick, 1908, p. 22, pi. xvi, figs. 6-10; Hentschel, 1914, p. 34;
S. borealis, subsp. acuata, Burton, 1928, p. 66, fig. 6; S. borealis, var. acuata, id., 1929, p. 445.

Occurrence. St. 42: South Georgia, 120-204 m.; St. 123: South Georgia, 230 m.; St. 144: South
Georgia, 155-178 m.; St. 160: Shag Rocks (depth doubtful); St. 167: South Orkneys, 244-344 m-;
St. 175 : South Shetlands, 200 m. ; St. 177 : South Shetlands, 1080 m. ; St. 190 : Palmer Archipelago,
93-130 m.

Distribution. Victoria Land; Wilhelm Land.

Genus Spirastrella, Schmidt
Spirastrella purpurea (Lamarck) (Plate LVI, fig. 7).
(For synonymy see Vosmaer, 1911.)

Occurrence. St. 283 : ofT Annobon, Gulf of Guinea, 18-30 m.

Remarks. The specimen is similar to those figured by Vosmaer (loc. cit., pi. iii, fig. 2,
pi. iv, fig. 6) in appearance. The surface is obscured, except for papillate processes, by
calcareous debris and much debris is included throughout the tissues. The habitus
recalls that of Spirastrella aiirivilli forma excavans, Lindgren (1898, p. 41) and the
spicules are almost identical with those of S. congenera, Ridley (cf. Vosmaer, loc. cit.,
pi. xii, fig. i).

Distribution. Indo-AustraHan area ; West Indies.

DISCOVERY REPORTS

340

Genus Latrunculia, Bocage

Latrunculia lendenfeldi, Hentschel.

L lendenfeldi, Hentschel, 1914, p. 44- pl- v, fig- i-
Occ^.rence St WS 81 : Falkland Islands, 8i-8a m. ; St. WS 83 : Falkland Islands, ^37-^9 ™^ '.

Islands, 210-242 m.

Distribution. Wilhelm Land, Antarctic.

Genus Cliona, Grant
Cliona aethiopicus, sp.n. (Fig. 40).

Holotype. B.M. 28. 2. 15. 724-

Occurrence. St. 283 : Gulf of Guinea, 18-30 m.

Diagnosis. Tylostyli of usual form, 0-26 by 0-007 mm.;
spirasters with short stout shaft and a few conical spines,
0-028 mm. long.

Remarks. The single specimen is found in a fragment of
a thick {Ostraea ?) shell. It resembles very closely C. chilerms,
Thiele, but differs in the measurements of the spicules (in
Thiele's species these are: tylostyU 0-3^0-33 by 0-017 mm
and spirasters o-oi8 mm.). Had this specimen been found
in a locality adjacent to the type locality, I should have q
had no hesitation in identifying it with C. chUemis, but the Fig. 40. Cliona aetuopcus,
naa no ncMLduu j t, .^, ., ^ i;rr„„„„„„„ ;„ the sp.n. a-c, bases of tylostyli, a

geographical separation combined with the differences m the ^P_^^ ^^^ ^^^^ ^^^^^^^^^^

dimensions of the spicules have caused me to regard it as a ^^^^^. ^^ spiraster, x 700.
separate species.

Order EUCERATOSA
Genus Halisarca, Dujardin
Halisarca dujardini, Johnston, var. magellanica, Topsent.

H. dujca-dini, var. magellanica, Topsent, 1901, P- 44, pl- i. ^E- 2, pl- -, hgs. 11-14; Burton,

oZlence '1 55 : Falkland Islands, xo-16 m. ; St. 58 : Falkland Islands, 1-2 m. ; St. 222 : Cape
Horn, 3c^35 m.; St. WS 79: Falkland Islands, 131-132 m-
Distribution. Victoria Land; Londonderry Island.

Genus Spongia, Linnaeus

Spongia officinalis, Linnaeus, Autt.

Occurrence. St. i : Ascension Island, 16-27 m. ,. , • •

REMARKS. The Single specimen has the external form of Euspongia officinahs adnaUca

(Schulze, 1879, pl. xxxiv, fig. 2), but is black in colour.

Distribution. Mediterranean; West Indies; Indian Ocean; Australia.

b c

EUCERATOSA 341

Spongia magellanica, Thiele.

S. magellanica, Thiele, 1905, p. 483, figs. 18, 106.

Occurrence. St. 156: South Georgia, 200-23601.; St. WS84: Falkland Islands, 75-74 m.;
St. WS 89: Tierra del Fuego, 23-21 m.

Remarks. The fibres of the first specimen contain sand grains only, those of the
second specimen contain only fragments of sponge spicules. Otherwise the two
specimens are identical and both resemble the holotype closely.

Distribution. Punta Arenas and Calbuco.

Genus Duseideia, Johnston
Duseideia fragilis (Montagu), Autt.
Occurrence. St. i : Ascension Island, 16-27 ni-

Remarks. The specimen is thinly encrusting, but resembles the European forms of
the species very closely in structure.

A synonymy of this species will be given in my report on the Barrier Reef sponges.
Distribution. Practically cosmopolitan.

Duseideia chilensis (Thiele).

Spongelia chilensis, Thiele, 1905, p. 485, fig. 20.
Occurrence. St. WS 84 : Falkland Islands, 75-74 m.
Distribution. Chile.

c b

Fig. 41. Duseideia tenuifibra, sp.n. a, fibres near periphery of sponge; b, from near the base;
c, showing method of attachment of fibres to sand grains.

14

342 DISCOVERY REPORTS

Duseideia tenuifibra, sp.n. (Plate LVI, fig. 8; Fig. 41).

Holotype. B.M. 28. 2. 15. 840.

Occurrence. St. WS 243 : north-east of Falkland Islands, 144-141 m.

Diagnosis. Sponge lobate; surface minutely conulose; oscules 2-4 mm. diameter,
situated at or near ends of lobes ; texture soft, compressible ; colour, in spirit, greyish-
green; skeletal fibres very thin, o-oi4-o-02 mm. in diameter, forming an irregular net-
work of nearly quadratic mesh, with meshes subdivided by more slender fibres ; sand
grains not abundant, incorporated in fibres or lying loosely in meshes of skeleton.

Remarks. The species is characterized by the slenderness of its fibres and by the
external form. The manner in which the included sand grains are held together is un-
usual. Instead of being entirely enclosed by the fibre, the grains are connected together
by short strands of spongin the ends of which are attached to the surfaces of the sand
grains (see Fig. 41 c). Near the base of the sponge, and in the inner parts of the lobes,
the fibres are thicker and the meshes of the skeleton reduced in size (cf. Fig. 41 ^).

Genus Dendrilla, Lendenfeld

Dendrilla membranosa (Pallas).

(For synonynay see Burton, 1929, p. 448.)

Occurrence. St. 45 : South Georgia, 238-270 m.

Distribution. Wilhelm Land; Victoria Land; Graham Land; Australia; Indian
Ocean.

THE AFFINITIES OF THE ANTARCTIC SPECIES OF

TEDANIA, WITH NOTES ON THE DISTRIBUTION

OF THE GENUS GENERALLY

Of the six species of Tedania recorded on pp. 302-9 five resemble each other so closely
that their identification presents an unusual difficulty, but at the same time raises two
points of interest. The first of these points concerns the difficulty of classifying these
species ; and the second concerns the new importance of embryological observations in
systematic studies. Nearly 200 specimens of Tedcuiia were collected from the sub-
Antarctic and Antarctic and at first no satisfactory classification seemed possible. This
was largely due to the strong similarity in the shape of their spicules. Certainly, the
various specimens diff'ered in many ways and could be divided into five groups on the
basis of external form, but even then many intermediates could be found between the
typical forms of these five groups. Similarly, each group was characterized by a particular
range of spicule size, but here again intermediates could be found. It was, in fact, only
when the embryos were studied, and the data so obtained used as a starting-point for
the eventual classification, that it became possible to decide on the identification of
these specimens.

ANTARCTIC SPECIES OF TEDANIA

343

It is of some importance to emphasize that the close resemblance of all these specimens
of Tedania, and the fact that they present numerous intermediates would under other
circumstances have suggested that all were conspecific, had the embryology not indicated
otherwise. In fact, had the characters of the embryos been overlooked, we should have
had in Tedania a similar position to that presented by Vosmaer (191 1) in his study of
Spirastrella purpurea (Lamarck), and (1932 et seq.) in his studies of the Porifera
Incalcaria of the Bay of Naples. And while I believe that Vosmaer was mainly correct
in his conclusions regarding S. purpurea (there are points, of course, where he has
obviously gone wrong), it would be better
to endeavour to apply the embryological
test as far as possible, both to this and
all other species. This gives to the study
of embryology an importance, hitherto
unsuspected, as the deciding criterion in
systematic work.

The characters and close relationship
of the five species of Tedania here recorded
are given in the table on p. 308.

In dealing with ^o large a collection, it
was found possible after a while, and once
the main distinguishing features were
appreciated, to classify specimens with
considerable certainty even in the absence
of embryos. Nevertheless, so closely are
these five species related and interdigitated
in regard to their main characters that ^. ^. , . , . , .

1 1 1 r 11 'S- 42- Diagram showing relationships and over-

several specimens could only doubtfully ^^^^-^^ -^ ^^e Antarctic and sub-Antarctic species of
be assigned to a species. Where only a Tedania. The five main areas represent types of em-
few specimens are found in a collec- bryology: a A, T.tenuicapitata; bB, T.c/iarcoti;cC,T.
tion therefore, it is conceivable that this »"^ssa;dD,Tspinata;eE 7\ murdochi a^e represent

types of spiculation and A-D external forms of these

difficulty will be considerably increased, ^^^^^^^ respectively.
unless embryos can be found in them.

In order to stress the importance of the conclusions reached in this study of the
species of Tedania, it is considered worth while to express graphically the relationships
of these species, so that these may be contrasted and compared with the diagram given
by Vosmaer (191 1, p. 324) for Spirastrella purpurea (Lamarck). Vosmaer shows seven
groups representing typical forms of S. purpurea and indicates how these are connected
by intermediates. From this he concludes that the specimens comprising these groups
are all conspecific. In the present case, the corresponding groups are much more closely
connected and may be represented, not by detached circles, but by overlapping areas
(Fig. 42), and yet the specimens they represent actually belong to diff^erent species.
For example, in the area where T. niassa overlaps T. tenuicapitata may be included the

14-^

<u

"^

S
.to

>

""Si

s
o

o

to

e5

:2£

! C <u

o :o ;*.

o D.

C 3 m

, O <u

P tli'S;

-o

c •

t5 p

Is

CO p

CC Ov

O CI

1 o

o o

o -t

o o

J _, tc

n (L)

'l^'S

Pti

"m O

g M
o C

O 3 IH

O OJ

QJ *H 4-»

U - "

C " ts

S i, is

M O S

a." "

C

P

OJ o ^

u

■^"i

^

.2^3

p

C J3 O

.2S^

w

?,

5 3

V)

> Pt3

<u

P «^

o

Oj= c

n

OJ .- -

-Pi^.S

u

'a

lata
que
cer

>.

a

•sup

o

.■.^^_^

c
£

3

.SE- fe S

lapp

In

d m

ticu]

O

u

S-u"P 5

x;

s|^ °-

D Q U

c

TJ ?! 1-

p to P o

'^ S (U P

ation
he di
to b
and

'S tr-a £

y^

some V
point a
tyle ten
in othe

u

o

- -t- '^ c

C/D

ver
rup
the
tha

11

o

g ta O w

-= g s s

a

•-.St2 "

^

lere
ing
ical
om£

2

-*j -o a •"

n

■• P >■ c

ive specie
ved, and
massa the
, more so

p

o

OJ

o

respect

tly cur

InT.

ariable

.£•

•s

X. ^ >

;-l

t^ " p !"

p

3

0)

JP

typic
s sm
very
arnot

o

1)

•T3

s are
tyle i
t in e
the t

CO

s

p

C "' p •*-

en he
pical
arese
nds o

T3 Xl

S, 0 <u

aracters
ases, the
ese are n
ape of th

-"5

^ ^ J= JP

u P

* The

t In a

ase, but

1 The

Jp JJ

" oJ
3^

^

^

ANTARCTIC SPECIES OF TEDANIA 345

discoid specimen, with spicules of small size, which was to all intents a specimen of
T. teniiicapitata, but which was found to contain embryos typical of T. massa. Top-
sent's (1908, pi. i, fig. 3) T. charcoti might be
placed in the area where T. murdochi and T. char-
coti overlap. And so on. As will be seen from
the diagram, T. spinata and T. tenuicapitata,
T. mossa and T. spinata, and T. murdochi and
T. charcoti can only be confused in external form,
and T. spinata and T. murdochi resemble each
other only in spiculation. On the other hand, T.
temdspiculata, T. charcoti and T. massa are sub-
ject to considerable confusion.

The second point raised is that concerning the
specialized character of these species as compared
with the remaining species of Tedania}

Judging from the embryological evidence, and
the development of the external form and spicu-
lation, Tedania nigrescens (Schmidt) is the most
primitive species of the genus, and the Antarctic
group, including also T. mucosa, Thiele, T. ex-
cavata, Thiele, T. pectinicola, Thiele, T. fuegiensis,
Thiele, and T. infundibuliformis, Ridley and Dendy,
is the most specialized. And of these, the species
found in the Antarctic {sensu stricto) are the most
highly specialized of all. Thus, in T. oxeata the
styli are replaced by pseudoxea (cf. the life history
ollophon radiatus, Topsent (Burton, 1931, p. 514,
fig. 2e)); T. charcoti shows a higher degree of
development than the others, since the fully-
developed embryonic spicules are styli ; but in T. massa is found the most convincing
proof that the Antarctic species are more specialized than the non-iVntarctic. In

^ The section Tedanieae, of the Desmacidonidae, consists at present of the genera Tedania, Gray ; Trachy-
iedania, Ridley; Tedaniopsis, Dendy; Paratedaiiia, Burton; Tedaniella, Czerniavsky ; and Tedanione, Wilson.
As shown on p. 306, Trachytedatiia must be regarded merely as representing the post-larval stage of
Tedania spinata (Ridley) and perhaps others. Further, Tedaniopsis and Paratedaiiia are clearly synonyms
and, as shown on p. 306, the genotype of the latter is synonymous with Tedania massa, Ridley and Dendy.
Tedanione, on the other hand, appears to be synonymous with Rhaplioxya, one of the Axinellidae, and
Tedaniella, inadequately characterized, may be a synonym of Petrosia.

The species of Tedania, as here understood, appear to number 24: T. bispinata, Hentschel; T. charcoti,
Topsent; T. commixta, Ridley and Dendy; T. {Tedanione) connectens (Brondsted); T. coralUophila, Thiele
(with T. brez'ispiculata, Thiele; T. maeandrica, Thiele; and T. 7'eticulata, Thiele, us probable synonyms);
T. diraphis, Hentschel; T. diversirhaphidiophora, Brondsted; T. excavata, Thiele; T. fragilis, Lambe;
T. fuegiensis, Thiele ; T. infundibuliformis, Ridley and Dendy ; T. massa, Ridley and Dendy ; T. mucosa,
Thiele; T.inurdochi, Topsent; T. nigrescens {^tchmidt) (see Burton and Rao, 1932, p. 353, for full synonymy);

500n

400-

300

200-

lOO-l

BOO

700

600

500

400

300

200

Fig. 43. Showing the relative lengths of the
styli and tornota in: a, T. spinata; b, T. mur-
dochi; c, T. tenuicapitata; d, T. charcoti;
e, T. massa.

346

DISCOVERY REPORTS

this species the form varies from massive or flabellate, to the highly speciaUzed form
which I have called " carrot-shaped " (cf. fig. 25 d, p. 304). The specimens of this species

Fig. 44. Plan of distribution of the species of Tedania. T. nigrescens, Schmidt, is represented by the black
dots, the remaining species by crosses. The size of the crosses is proportionate to the degree of specialization
attained by the species they represent.

in the present collection, classified as follows, show how this specialization is increased
as the Antarctic continent is approached :

Locality

Massive to flabellate
(primitive)

"Carrot-shaped"
(specialized)

Falkland Islands
South Georgia
Antarctic continent

20

4

I

4
20

T. oxeata, Topsent; T.pectinkola, Thiele; T . phacellina , Topsent; T. placenlaeformis, Brondsted; T. scotiae,
Stephens; T. suctoria, Schmidt (with T. incresceris, Schmidt, and T. comdigera, Topsent, as synonyms);
T. tenuicapitata, Ridley and Dendy; T. {Tedaiiione) zvilsoni (Dendy); and T. {Tedaniopsis) lurbinata (Dendy).
[The following must be removed from the genus Tedania: T. crisla-gaUi , Dendy, to Axinellidae; T.
coubnani, Kirkpatrick, to Kirkpatrickia; T. variolosa, Kirkpatrick, to Kirkpatrickia; T. laevis, Kirkpatrick,
to ? Plocamicae; T. laxa, Lendenfeld, to Stylotella agminata, Ridley; T. tenuispina, Lendenfeld, to Stxlotella
agminata, Ridley; T. leptoderma, Topsent, to Lissodendoryx isodictyalis , Carter.]

ANTARCTIC SPECIES OF TEDANIA

347

The picture presented by the distribution of these species (Figs. 44, 45) is thus a
very interesting one. We have one species, T. nigrescem (Schmidt), the least speciaHzed
of the genus, variable in form, having a wide distribution, and all round the margins
of its area of distribution are found species of slightly more advanced development
and greatly restricted distribution. The culmination is found in the sub-Antarctic
and Antarctic species which appear to show a continued progression in development

Fig. 45. Plan of distribution of the species of loplwii. Tlie shaded area represents the distribution of
/. proximum, Ridley, the darker lines coinciding with localities at which the species has been actually recorded,
the thin lines where it may be assumed to occur. The size of the black dots, representing the remaining
species of the genus, are proportionate in size to the degree of specialization of the species.

as the Antarctic is approached, until, within the Antarctic itself, we have the three
most specialized species of all, T. charcoti, T. massa and T. oxeata. This picture, in
fact, suggests the possibility that this variable species, fluctuating considerably in its
main characters, has given rise to a number of localized and more highly developed
species at various points ; and that from these have arisen other species with, generally,
an increasingly higher degree of specialization as the polar waters are approached.

348 DISCOVERY REPORTS

THE EVOLUTION OF THE SPECIES OF lOPHON,

AND ITS SIGNIFICANCE IN RELATION TO OTHER

ANTARCTIC SPECIES OF ECTYONINAE

The genus loplion provides another striking plan of the distribution of species and,
in addition, a probable indication as to the generation of species. In the North Atlantic,
restricted to the coasts of the Mediterranean, and of Western Europe as far north as the
British Isles, are two species, closely allied and apparently the most primitive of the
genus. These are /. pattersoni (Bowerbank) and /. hyiidmaui (Bowerbank). It is a re-
markable fact that these two species are almost invariably found growing in association.
That is to say, if, in a single haul of the dredge, examples of one species are obtained,
examples of the second will almost certainly be present also. There is yet a third species,
/. scandens (Bowerbank), of rare occurrence, the characters of which are intermediate
between those of the other two species. It appears probable that future research may
show either: (i) that these three species represent different growth stages of a single
species; (ii) that /. pattersoni and 7. hyndmani are forms of a single dimorphic species,
and I. scandens an occasional intermediate; or (iii) that /. pattersoni and I. hyndmani
are distinct species, and possibly I. scandetis a hybrid. At all events, it will be convenient
to regard them here as representing a single primitive species.

The remaining six species of the genus are all more specialized than the foregoing in
many respects, especially in their external form and in the shape of the bipocilla. The
first of these is I. piceus^ (Vosmaer), found on the coasts of Norway and in the Arctic.
The second is I.proximum (Ridley), which is found throughout the sub-Antarctic, from
Kerguelen (and Natal) to South America, up the western coast of South America to
the Galapagos, on the we'^tern coast of North America and possibly round to the eastern
coast of Canada. In addition, it is found rarely in the Antarctic. The third species, 7.
laminalis, Ridley and Dendy, from Prince Edward Island, is closely related to, if not
synonymous with, 7. proximmn. The fourth, 7. omnivorns, Ridley and Dendy, from
Australia, is slightly degenerate, though still somewhat more specialized. The fifth,
7. laevistyliis," from New Zealand, is more specialized. The sixth, 7. melanokhemia (de
Laubenfels) from Puget Sound, is probably also a specialized offshoot from 7. proxi-
mim. The last, and most specialized, 7. radiatns, Topsent, though also found in the
Falkland Islands area of the sub-Antarctic, is most numerous in the Antarctic proper.
This species, in its development, passes at first through a stage recaUing the hyndmani-
pattersoni group, and, later, shows certain features associated with 7. proximum.

The picture (Fig. 45) presented by these species may be summarized as follows. The
group represented by 7. hyndmani and 7. pattersoni is a primitive one, and from it, to

1 There can be little doubt, in view of the variability of the spicules found in /. radiatus and /. proximum,
that I. piceus, Vosmaer, I.frigidus (Levinsen) and /. dubius (Hansen) are conspecific ; especially as they are so
alike in external form.

' lophon major and var. lemiis, Brondsted, and /. minor, Brondsted, are quite evidently conspecific with
/. laevistylus, Dendy, in view of what is now known concerning the variations in all species of lophon.

THE SPECIES OF lOPHON 349

the north, a more speciaHzed species has sprung (/. piceus). To the south, it has given
rise to /. proximum, again a sHghtly more speciaHzed species, but one in a state of extreme
fluctuation as regards its spicular characters (cf. Figs. 22-24). From this variable species
have arisen several more speciaHzed species, which have migrated in a southerly
direction: /. laminalis in the Southern Ocean, /. omnivonis in AustraHa, /. laevistyliis
in New Zealand, and /. radiatus in the Antarctic. At the northern limit of its range
of distribution, it has possibly given rise to another speciaHzed species, /. melanokhemia.

We have here an indication that the species of lophon have developed from a form
the prototype of which is located in the North Atlantic, and that its most specialized
species is found in the Antarctic. The same thing may be said of the species of Mycale.
In point of numbers, the species of the North Atlantic (i.e. on the European side) are
the most numerous. Elsewhere the species of this genus are found mainly in the
Antarctic and sub-Antarctic, and in the Indo-Australian area ; but whereas the Indo-
Australian species are either Httle more specialized than those of the North Atlantic, or,
more commonly, degenerate (cf. ArenochaUim mirabilis, Lendenfeld, from Australia, a
degenerate Mycale), the Antarctic species are somewhat more specialized than any of
the rest. Moreover it may be noted that the six Antarctic species of Mycale are found
in localities bordering on the area of distribution of M. magellanica (Ridley), a species
comparable with lophon proximum (Ridley) in the variability of its spiculation. Finally,
we find that the two most specialized species of Mycale, M. intermedia, Schmidt, and
M. acerata, Kirkpatrick, are found in the Arctic and Antarctic respectively.

The distribution of the genus Amphilectiis is also similar to that of lophon and
Mycale. The least specialized and most variable species, Amphilectus fiicorum (Johns-
ton), is distributed generally around the European coasts and in the Arctic, is found
on the west coast of Africa, around the southern extremity of South America, in the
Antarctic, and finally at New Zealand. Again we find specialized species, A. qiiatsi-
noensis (Lambe), on the west coast of Canada, and A. americana (Ridley and Dendy)
around the southern extremity of South America. In addition, there are Corybas lobata
(Johnston) in the North Atlantic and Arctic, and Brondstedia glaber (Br0ndsted) in
the Antarctic, both of which must be regarded as species closely related to Amphilectus
fucorum but slightly more specialized.

The genus Isodictya is confined practically to the Arctic, European coasts, south
coast of Africa, southern extremity of South America, the Antarctic and New Zealand.
And again we find the most primitive species in European waters, while around South
Africa the species of this genus show a rich development in numbers of species, size
of individuals and growth of spicules. This progressive tendency in the South African
species of Isodictya is carried still farther in the sub-Antarctic and Antarctic, where the
genus reaches its highest degree of development.

There is in this a striking parallel which cannot be ignored, and which suggests that
four of the most primitive sections of the Desmacidonidae — the Isodictyeae, the Myca-
leae, the Tedanieae and the lophoneae — have each arisen in an area approximating to the
European side of the North Atlantic, and from thence have spread northwards to the

15

3SO DISCOVERY REPORTS

Arctic in one direction, and southwards to the Antarctic, along the west coast of Africa,
in the other, with a lateral path of migration extending into the Indian Ocean. The
remaining sections of the Desmacidonidae, the Myxilleae and Clathrieae, are world-
wide in distribution ; but here again the more primitive species are found in the North
Atlantic, and so far as the Myxilleae are concerned, the more highly developed species
are found in the sub-Antarctic and Antarctic.

Nor can we ignore the coincidence that the widely distributed and primitive species
lophon proxhmmi, Tedania nigrescens and Amphilectiis fiicoriim, with a very variable
spiculation or external form, have a number of more highly developed species, which
appear to have developed from them, around the fringes of their areas of distribution.

With the line of progression from Europe to the Antarctic, down the west coast of
Africa, shown by the species oi Isodictya, Amphilectiis, Mycale, lophon and Tedania, may
possibly be correlated the curious, discontinuous distribution shown by the species of
Plakina (P. monolopha, P. dilopha, and P. trilopha, together with several others) which
are found only in the Mediterranean and the Antarctic. The full significance of this
cannot yet be assessed, but it is a matter which should not be overlooked. How this
particular distribution has been efi'ected is a question which raises numerous difficulties.

Br0ndsted (1926) has suggested that the distribution of New Zealand sponges may be
best explained on the Wegener hypothesis ; but whereas, in some parts of the world, the
Wegener hypothesis will explain the distribution of sponges as well as any other hypo-
thesis yet put forward, it is entirely insufficient in dealing with the sponges of the Atlantic
Ocean. The main features of the distribution of sponges in this area are on the whole
contrary to what might be expected on the Wegener hypothesis. They are as follows:

(i) The dissimilarity of the fauna of the eastern coast of the United States of
America and that of Europe.

(2) The similarity between the fauna of the West Indies and that of the Indo-
Australian area.

(3) The existence of species common to the European and West African coasts,
the sub-Antarctic (as represented by the Falkland Islands area) and the Antarctic, and
their absence elsewhere in the Atlantic.

(4) The presence in the Antarctic of several species found hitherto only in the
Mediterranean.

Nor can the "creeping" of species around the coasts explain more than a small
proportion of the facts of distribution of species. I have already suggested (1930) a
correlation between the main surface currents of the ocean and the distribution of
sponges in the North Atlantic, and it is now proposed to show that a similar correlation
exists in the South Atlantic and Indian Oceans.

35 1

GEOGRAPHICAL DISTRIBUTION

In the following Table, which represents the distribution of species in the Antarctic and
sub-Antarctic and in the adjacent areas, no mention is made of species newly described
in this report, of cosmopolitan or widely distributed species, or of species with a limited
distribution. The sole purpose of the Table is to indicate approximately the relations
between the various localities, indicated at the head of the Table, in so far as their
respective sponge faunas are concerned.

ji ,~^

"S

f-> a<

c

<t

A

-^

In

u, I-

CJ

HH

go

O

Pi;

a

J3

bjo

•TS d

bo

bO

"a

3

ia Land,
etc.)

Juding?!
d Elepha

a
CO
bjo

a
•3

(includin

e, Patago
o)

O

O

T3

C

1

•1 S

C/3

_3

a

•Si

Si "

j3
C
D

o

<

•a

a

»— 1

3
JO

■s

.2

^1

Si ^

a
-a

cd

o

>-' o

1-

IS

(3J

C

"3

3

13
a,

Q

O

o

3 3

<.^

.2

£?

2cfi

« CQ

•H

u

u

cd

3

<

O

CO

CO

fc

CO ^

H

M

u

U^

Rossella antarclica, var.

X

X

intermedia

R. nuda

X

X

X

X

R. racovitzae

X

X

X

X

R. villosa

X

X

Gymnorossella inermis

...

X

...

X

Leucosolenia discovereyi

X

...

X

Leucetta leptoraphis

X

X

...

X

L. macquariensis

X

X

...

...

Macquarie L

Leucettusa haeckeliana

X

X

Grantia cirrata, var.

aurorae

X

X

X

Plakina trilopha

X

X

X

X

Mediterranean

Geodia magel/ani

X

X

Tetilla leptoderma

X

X

X

X

X

X

Cinachyra antarctica

X

X

X

C. barbata

X

X

X

X

X

C. coactifera

...

X

...

X

Haliclona nodosa

.. .

X

...

X

H. chilemis

. .*

X

X

H. variabilis

...

...

X

X

H. penicillata

X

X

X

H. gaussiana

X

X

H. tubidoramosa

...

...

. . .

X

. • .

New Zealand

H. conica

...

X

...

...

New Zealand

Microxina benedeni

X

X

X

X

Hemigellius rudis

X

X

X

Adocia cardials

X

X

X

X

X

A. glacialis

...

X

X

X

...

X

15-2

352

DISCOVERY REPORTS

rArch.,
roup)

)ck)

Island

go

PS

.s

J3
bjo

-a n

bc

(50

'S

P

c

udingPi
Elephai

(3

a
'-3

c

Ph

o
O

■V

c

C

.2 u

't-i

11

.H s
S3

11

0)

a

o

-3

■&
O

o

a

kland Islands (
ardwood Bank)

America (Chile
ierra del Fuego

C

U
«

c
.J2

<

c

•T3

c

42
tn

l-H

a.

a

1

CO

Q

^^

5-

rn

c/i

tS"

ai^

1

^

Adocia siphonella

■ • >

...

...

X

X

A. cucurhitiformis

X

X

Calyx arcuariits

X

X

X

X

C. kerguelensis

...

...

X

X

...

X

Dasychalina validissima

...

...

X

X

X

Callyspongia forth

...

...

X

X

C. fusifera

...

...

X

X

Isodictya kerouelensis

...

X

...

...

X

...

X

I. kerguelensis , var.

X

...

X

simiUima

I. setifera

X

X

X

X

I. delicata

...

...

...

...

X

X

I. antarctic a

X

...

X

X

I. cactoides

X

X

...

X

I. erinacea

X

X

Cercidochela lankesteri

X

X

Plmnocolumella

X

...

X

fnaeandrina

Mycale magcUanica

X

X

X

X

X

M. lapidiformis

...

...

X

X

M. acerata

X

X

X

X

M. tridens

X

X

...

X

Amphilectus rugosus

X

...

X

X

X

Biemna chilensis

X

X

...

X

X

Asbestopluma calyx

X

X

Acanthorliabdus fragilis

X

X

loplion radiatus

X

X

...

X

i

Galapagos

I. proximum

...

...

X

X

X

X

X

-\

I., Canada

Tedania tenuicapitata

X

X

...

X

X

X

T. oxeata

X

X

T. massa

X

X

X

X

Myxilla mollis

X

X

X

X

X

M. asigmata

X

X

...

X

M. chilensis

...

...

...

X

X

M. elongata

X

X

Ectyodoryx paupertas.

...

X

X

X

X

subsp. nobile

E. antarctica

X

...

...

X

E. frondosa

X

X

E. raniilobosa

X

X

X

Anchinoe latnmcidioides

X

X

X

X

X

...

...

South Africa

A. areola ta

X

X

X

A. leptochela

X

...

X

Kirkpatrickia variolosa

X

X

X

GEOGRAPHICAL DISTRIBUTION

353

jj ^-^

-d

y o.

a

< §

S"

u t;

u

l-H

go

o

Pi

to

bo

.2"

"S

■a

3

.,

Cl, a

M

c

0

0

ia Land
etc.)

eluding'
d Eleph

j3
Cfi

B
■•3

(includi

e, Patag
0)

0

■a
c
•a

§

(Victor
Land,

CO

a (Chil
;1 Fueg

c
3
U

<

•v
a
«

l-H

3

1^

a —

1*

c

o

o

o

l-H o

11

B S

•a
c

■*-»
.2

c

6

(5
1-1

o^

CO

X

« CO

c-^

H

0

fS

Kirkpatrickia coulmani

X

Inflatella belli

X

X

New Zealand

Plocamia gaiissiana

X

X

X

Clatliria papulosa

X

X

C. toxipraedita

...

X

X

X

Ophlitaspongia

...

X

...

X

mcmbranacea

Artemisina apollinis

X

X

X

X

< > •

North Atlantic

A . plumosa

X

X

Dictyociona discreta

...

X

X

X

Axociella nidificata

X

X

X

A.flabellata

X

X

X

Pseudanchinoe toxifera

X

X

X

X

Eurypon miniaceiim

X

X

> • .

X

• * •

South Africa

Stylotellopsis amabilis

X

X

X

Hymedestnia irritans

...

. .•

X

X

H. laevis

...

X

X

H. sitnillima, var.

X

X

X

antarctica

Crella crassa

X

X

Crellina Uibifex

X

X

Hytnemacidon fernandesi

X

X

X

X

H. Inrquata

X

• • .

...

X

Thieleia ruhiginosa

...

X

X

Axinella crinita

X

X

Pseudaxinella egregia

...

X

X

Homaxinclla supiatumescens

X

X

X

Rhizaxinella atistraliensis

X

X

...

Australia

Siibeiites montiniger

X

X

S. miciostomus, var.

X

X

X

stellatus

S. papillatus

X

X

...

X

Pseudosiiberites hyalinus

X

X

X

• * .

X

X

...

Europe

P. sulcatus

X

> < •

X

X

X

X

...

X

Polymastia isidis

X

X

< . .

X

P. itwaginata

X

X

X

X

Sphaerotylus an tairlictis

X

X

Stylocordyla boreulis.

X

X

X

X

subsp. acuata

Latnmculia lendenfeldi

X

>*.

...

X

Halisarca dujardini, var.

X

X

...

...

X

X

tnagellaiiica

Spongia magellanica

> • •

. . •

X

X

Dendrilla membranosa

X

X

...

X

...

...

...

...

Australia ;
Indian Ocean

DISCOVERY REPORTS

354

The following analysis may be made from the table:

Of the 69 Antarctic species :

50 are found also off Graham Land.
^2 ,, South Georgia.

2Q „ Falkland Islands.

II ,, South America.

n ,, Kerguelen.

Of the 56 species from Graham Land:

50 are found also in the Antarctic.
32 „ off South Georgia.

23 ,, Falkland Islands.

II ,, South America.

^ „ Kerguelen.

Of the 57 species from South Georgia:

43 are found also in the Antarctic.
32 ,, off Graham Land.

iQ ,, Falkland Islands.

13 „ South America.

8 ,, Kerguelen.

Of the 51 species from Falkland Islands:

29 are found also in the Antarctic.

23 ,, off Graham Land.
ig „ South Georgia.

24 „ South America.
6 „ Kerguelen.

The list showing the species recorded from each station gives a clear insight into the
distribution of species within the Antarctic and sub-Antarctic zones. Taken m con-
iunction with my table of Antarctic species (1929, PP- 394-40i)> there appears to be no
really marked change in the fauna at Graham Land, where the species are dommantly
Antarctic. The same is true for the South Shetlands; but at South Georgia there^are,
in addition to numerous Antarctic species, a number common to that locality and the
Falkland Islands, and a few common to it and the mainland of South America. Around
the Falkland Islands there are still many Antarctic species to be found; but the
character of the sponge fauna is more that of the southern regions of the South American
continent (e.g. Patagonia and Chile). One Antarctic and South American species
is also recorded with this report (see p. 325) from the False Bay region of South

Takhia the records contained in this report in conjunction with my previous papers

(1029, Terra Nova Report, and 1930, Senckenbergiana) it becomes evident that there

s no real distinction to be made, so far as sponges are concerned, between the faunas

of the Antarctic or sub-Antarctic, nor is there any marked difference in the Magellan

GEOGRAPHICAL DISTRIBUTION 355

and Kerguelen regions (cf. Regan, 1914, Vol. i, British Antarctic ' Terra Nova' Exped.
Rep.). It appears, on the contrary, that there is a gradual diminution in the Antarctic
species as we travel outwards from the Antarctic Continent. There are, however, several
significant features of the distribution, which have an important bearing on the general
problem of the distribution of species in the South Atlantic. These are : (i) that one species
of Antarctic and one South American species are found around False Bay; and (ii) that
on the eastern shore of the South American Continent some Antarctic species appear
to have penetrated as far north as the mouth of the Rio de la Plata, but that beyond this
the character of the fauna changes entirely. To understand the full significance of this it
is necessary to consider the distribution of sponges in the South Atlantic in conjunction
with that in the Indian Ocean.

In his monograph of the horny sponges (1889), von Lendenfeld refers to a number of
species of Euceratose sponges as being common to both Australia and the West Indies,
and found only in those two areas. At first sight it seems unlikely that a species should
be restricted to two such widely separated areas, and unfortunately this author leaves
no record as to where the material on which he worked may be found, so that it is not
possible to verify his results by actual observation. Since, however, so many of his
identifications have been shown to be inaccurate, his conclusions are apt to be regarded
as unreliable. Moreover, the extreme difficulty of identifying species of the Euceratosa
with certainty only serves to increase the doubt felt on this point. Recent observations
have tended, however, to suggest that von Lendenfeld may have been correct in this
case. For example, the single commercial sponge found on the Great Barrier Reef
of Australia seems to be indistinguishable from some of the varieties of the Bahamas
grass sponge, and is, so far as I am aware, not found elsewhere unless it be in
the adjacent waters of the Indo-Pacific. The question remains therefore whether the
superficial resemblance between the Barrier Reef sponge and the West Indian grass
sponge may not be the result of convergence rather than of actual affinity. That this is
no case of convergence, and that the two have a common origin, is suggested by the fact
that at least five species of sponge, sufficiently well characterized to be identified without
hesitation, are common to both the West Indies and Australia. Moreover, in some cases
they have been found also in the Indian Ocean and on the west coast of Africa. This
seems to have an important bearing on the problems concerning the distribution of
sponges and on the systematic study of the group.

I have suggested (1930) that there might be some correlation between the distribution
of sponges and the main surface currents of the oceans, and that the study of these
currents might be of material benefit in interpreting the natural and well-marked areas of
distribution which undoubtedly exist. This arose from the study of the sponge fauna of
Norway, but the conclusions reached then concerning the North Atlantic appear to
hold true for the South Atlantic also. Assuming this to be the case, the chart of the
main oceanic currents is not only sufficient to show why the faunas of the Red Sea,
Indian Ocean and Indo-Pacific differ little from each other, but also why that of the
Barrier Reef of Australia should be composed almost exclusively of species belonging to

356 DISCOVERY REPORTS

those areas. This well-defined area of distribution, the Indian Ocean and Indo-Pacific
area, ends abruptly in the west in the region of the Agulhas current. On the west and
north it is bounded by the continents of Africa and Asia respectively, and in the south
by the cold waters of the West Wind Drift, which is to all intents an impassable barrier
to migration. The eastern boundary must be ignored for the present, as our knowledge
of the sponges of the Pacific is so imperfect, and to avoid unnecessary complications the
sponge fauna of Australia, exclusive of the Barrier Reef, must be left out of the reckoning.
The area thus delimited, which for the sake of brevity may be called the Indian Ocean
area, in addition to constituting a natural and well-marked faunal area, forms a closed
system of circulatory currents, and it is easy to imagine that these two conditions are
correlative. In the same way, that portion of the Atlantic south of a line approximating
to the Equator constitutes a second well-defined area and forms also a partially closed
system of circulatory currents. To the east and west it is bounded by the continents of
Africa and America and to the south, again, by the waters of the West Wind Drift. To
the north is a similar faunal area, with its circulatory system, with which the South
Atlantic system is put into communication mainly by the Gulf Stream ; and such species
as are common to both North and South Atlantic follow a line of distribution which
accords with the conclusions outlined in the paper already referred to. For the moment,
then, it will be convenient to ignore the North Atlantic system.

We have thus two closed systems of currents, each with its well-defined fauna, and
if we are to accept the hypothesis that the distribution of sponges is effected by surface
currents, we should expect to find a complete absence of species common to the two
areas. This is, as a general rule, true. A few species, not more than 2 per cent, have,
however, migrated from the Indian Ocean to the South Atlantic; but these excep-
tions are apparent only and serve to test the rule, since having once gained entrance into
the South Atlantic system, their distribution again follows the line of the main surface
currents. On the other hand, and this at first sight seems to be a much greater obstacle
to the development of the argument, while there is this migration from the Indian Ocean
to the South Atlantic, there is no evidence of a similar migration from the South Atlantic
to the South Pacific. This is the more remarkable since the gateway from the Indian
Ocean to the South Atlantic and that from the South Atlantic to the South Pacific are
so similar in character, both being barred by the meeting of a hot and a cold current,
with a considerable difference in temperature between the two. When we come to look
more deeply into the matter, however, we find that it is on the theory that the major
factor in migration is transportation by currents, rather than the limiting effects of
temperature, that this apparent anomaly can be explained.

The barrier between the Indian Ocean and the South Atlantic is situated along a line
approximately due south of the Cape of Good Hope, and lies along the junction of the
Benguela current passing northwards and the Agulhas current flowing southwards. The
general faunas east and west of this line are markedly different, and this is almost certainly
due to an abrupt drop in temperature in passing from the one current to the other ; but
the fact that some species do pass this line shows that the barrier is not complete. The

GEOGRAPHICAL DISTRIBUTION

357

second barrier, that between the South Atlantic and the South Pacific, is situated at a Hne
approximately level with Buenos Aires on the South American coast and passes in a
south-easterly direction to the region of the West Wind Drift. It is formed by the
junction of the Brazil current running down from the north and a cold current flowingup
from the region of the Magellan Straits. Here again there is a considerable difference in
temperature between the two currents, nearly twice as great as that between the South
African currents, and the fact that no South Atlantic species pass this barrier might be
taken to mean that temperature was the deciding factor. But to my mind, a more im-
portant factor than temperature is the relation of the currents forming the two barriers,
namely, that in the South African barrier the currents run parallel with each other, and
though in opposite directions, are not opposed, while in the South American barrier

Fig. 46. Illustrating the two closed circulatory systems of currents in the southern hemisphere, in relation to

the distribution of sponges in that area.

they are opposed. In the first case, the mixing of the waters at the line of junction
allows the migrating sponges to be carried from the southerly directed Agulhas current
into the northerly flowing Benguela current, and thence shortly to be restored to a third
current, the South Equatorial, in which the temperature is similar to that of the Indian
Ocean currents, whence they have come. Once in the South Equatorial current, their
passage to the West Indies and the coast of Brazil is only a matter of time, and we do,
in fact, find that one at least of the species common to the West Indies and Australia
is also found on the coast of Brazil.

Assuming that the distribution of sponges is the result of transport by ocean currents,
we should expect to find that the fauna north of Buenos Aires differs markedly from the
fauna south of that line, and so far as our limited knowledge of these regions can show,
this is actually the case. Unfortunately, we know very little of the faunas of these two
regions, and future research may conceivably prove the contrary, but from a rapid
examination made recently of a collection extending over these two areas this does not

D VI 16

:,58 DISCOVERY REPORTS

appear to be probable. What actually happens to make the South American barrier more
effective than that of South Africa is, I imagine, that, since the currents are opposed,
any mixing that may take place is nullified by the fact that the cold current from the
south encounters and passes under the Brazil current, continuing its journey northwards
beneath the surface. In this way, any floating bodies reaching the northern current will
be restored once again to the warm surface waters of the South Atlantic.

While it is true that many details have yet to be considered and the hypothesis tested
in other parts of the globe, it is reasonable to infer from the evidence afforded by the
sponges common to the West Indies and Brazil, on the one hand, and to Australia,
on the other, that while temperature does exercise considerable influence, their dis-
tribution will probably be found also to be correlated largely with ocean currents.

Theoretically, there is one great objection to this idea; that is the short duration of the
larval period, and it was this objection that doubtless gave rise to the view hitherto held
that extension of the range of distribution in sponges must always be explained by a
slow creeping action round the margins of the land masses or along the sea-bottom, and
that where a shallow- water species is represented in two areas widely separated by water,
the land masses of these areas must at some time in the earth's history have been closely
adjacent or actually connected. The free-swimming larval period of a sponge is normally
24 hours or less. Under abnormal conditions, as in marine aquaria, it may be extended
to 2 or even 3 weeks, but this is exceptional and, owing to technical difficulties, it is not
known whether such a larva will develop into a normal adult or not. The probability is
that it does not. For all practical purposes, therefore, we must assume that the larval
period is short, and that trans-oceanic migration on the part of the larva is either nor-
mally impossible or, if possible, that successful metamorphosis sufficient to establish
the species cannot be accomplished at the end of the journey. On the other hand, it is
not impossible that transport is effected after metamorphosis, and that young recently-
metamorphosed individuals may be carried by floating objects. In this respect it is
pertinent that the young sponge grows very little indeed for several weeks after meta-
morphosis, remaining usually as a thin inconspicuous incrustation, and in this condition
it would not only escape observation but could be carried on floating fragments of such
size and of such a nature as to be overlooked.

It is interesting to note that the species involved in this apparent migration from
Australia to the West Indies belong to two groups only, the order Euceratosa and to the
family Clavulidae of the order Tetraxonida. Whether it is that the larvae of these two
groups have certain habits and characteristics which fit them for such migration, or
whether the adult individuals are more adaptable and can become established in areas
where individuals of other species, carried thither by the same migratory sequence, will
perish, is a matter for future research. The important point is that some species of
sponges appear to be capable of transportation over long stretches of oceanic waters by
the medium of the prevailing currents.

The distance from Australia to the West Indies is very great, and from the presence
of species of commercial sponges on the Barrier Reef of Australia identical with, or

EMBRYONIC DEVELOPMENT 359

resembling extremely closely, sponges growing in the West Indies, we can only infer
either that their superficial resemblance is due to convergence, which is improbable, or
that the possible range of migration, of some species at least, is far greater than has been
hitherto believed possible; or, and this is the most probable, that these sponges have
also become established in favourable areas, perhaps extremely localized, in the Indian
Ocean and the South Atlantic, which have not so far been recorded in the scientific
literature. This is, to some extent, borne out by the record (St. i, p. 340) of Spongia
officinalis, Linnaeus, from Ascension Island.

EMBRYONIC DEVELOPMENT AND BREEDING SEASONS

It is probable that too little attention has been paid in the past to the value of describ-
ing embryos found in preserved material. From a purely embryological point of view
such data as may be obtained from them is insufficient, but to the systematist the know-
ledge gained may be of immense value. Sections cut for purposes of identification may
not ordinarily reveal the presence of embryos, which are often situated in the deeper
parts of the sponge. For example, hand sections made from a specimen of Tedania
temiicapitata, Ridley, in the present collection, revealed no trace of embryos, but these
were found in large numbers when the specimen was cut in two. Wherever practicable
examination should therefore be made for embryos and their state of development and
other cognate information recorded.

EMBRYONIC DEVELOPMENT

Since most information has been obtained regarding the development of the
Antarctic species of Tedania, as compared with species of other genera, it will be
convenient to deal with these first. Numerous embryos were found in some twenty or
thirty of the specimens of Tedania examined and from these a fairly complete record of
development in this group of species may be obtained. In most cases, the embryos
present in a given specimen represent only a limited part of the developmental sequence ;
but in a specimen of T. charcoti, Topsent, the whole sequence can be followed in a few
hand sections, from the unsegmented ovum up to the time at which the embryo escapes
from its capsule. Embryonic development appears to follow the same course, broadly
speaking, in all five of the Antarctic species, but differs considerably in such details as
the size, shape and behaviour of spicules and in small details of histology.

The embryos of a single species, as represented by samples from several individuals,
vary slightly in size, in the size and shape of the spicules, in the time at which the
spicules appear and in the subsequent behaviour of these spicules ; but such differences
are small and are for the present purpose negligible. The specimens examined were all in
the form of thin hand sections, cleared in clove oil, unstained or stained in borax
carmine. In these preparations, the embryos appeared as solid bodies, but in a few
instances specimens were found to have been satisfactorily sectioned by the razor. It

16-2

36o DISCOVERY REPORTS

was not possible, however, to study the histology in detail, this being reserved for a
future occasion. The illustrations given represent optical sections and are intended
mainly as accessory data for systematic study.

The embryonic development in T. charcoti, in which species a nearly complete series
of embryos was obtained, may be conveniently dealt with under three headings : (i) the
segmentation of the ovum and the differentiation of the cells so formed, (ii) nutrition,
and (iii) the development of the spicules.

Tedania charcoti, Topsent

Segmentation. The ovum is first seen, in the deeper parts of the choanosome, as a
semi-transparent body, charged with refringent granules, oval in shape, o-i by 0-07 mm.,
and lying in a capsule which is itself connected with the choanosomal tissues by radiating
suspensoria (Fig. 47 a). The nucleus is at first faintly visible, but with the accumulation
of food material is rapidly obscured. The first cleavage results in two cells of approxi-
mately equal size (Fig. 47 e), but the second gives rise to four cells of appreciably
different sizes (Fig. 47/). From this point onwards segmentation becomes very irregular
(Fig. 47 li), but results finally in a spherical mass of cells (Fig. 47 i) of approximately
equal size. Details of the intervening stages are missing, and the next stage in develop-
ment seen is that represented by Fig. 47 y, in which the embryo, now oval and measuring
0-24 mm. in longest diameter, consists of a dense central mass of small granular cells,
measuring 0-014 mm. in diameter, covered externally by a thin layer of much smaller,
flattened cells. It is at this point that the first spicules make their appearance. After this
the embryo increases considerably in size, becomes approximately spherical, 0-35 by
0-32 mm., and nearly transparent. A cortical layer of columnar cells becomes visible,
0-035 mm. thick, and the cells of the inner mass, now slightly smaller in size, are much
more diffusely distributed. The succeeding stages are concerned mainly with changes
in the distribution of the spicules (Fig. 47 /), and finally, a change in the outline of the
embryo itself (Fig. 47 m). This latter is probably concerned solely with the escape of the
embryo from the capsule, the walls of which have grown very thin (Fig. 47 k), and it
appears as if the embryos are liberated from the capsule and make their way to the
exterior by a series of quasi-euglenoid movements.

Nutrition. In the earliest stages, the embryonic capsule is thick-walled, and
entirely empty except for the ovum, but in its walls may be seen a number of granular
cells, spherical or oval in form, and measuring, on an average, o-oi mm. across. Similar
cells may also be seen in the suspensoria and in the adjacent choanosomal tissues. There
can be little doubt that these are nutritive cells. Immediately prior to the first cleavage
(Fig. 47 b) they become much more numerous and a few are seen migrating into the
capsular cavity (Fig. 47 c), and as segmentation proceeds the cavity becomes partially
filled with a granular mass (Fig. 47 d), derived presumably from the breaking down
of these cells. The ovum, too, becomes more opaque and granular and assumes a
golden yellow tint. After the second cleavage, the nutritive cells become appreciably
less in the capsular wall (Fig. 47/), the colour of the embryo deepening slightly, until

EMBRYONIC DEVELOPMENT 361

in the later stages of segmentation they are only sparsely present. At this stage, raphides
from the maternal tissues are carried in to support the wall of the capsule, which is from
then onwards progressively reduced in thickness (Figs. 47 h-j). Finally, the wall loses
its layer of raphides and becomes extremely thin (Fig. 47 k) in preparation for the escape
of the embryo. The appearance of the raphides in the capsular wall appears to mark
definitely the end of the nutritive period.

Development of Spicules. The first spicules to appear are extremely fine acan-
thostyU (Figs. 477 and 53 a), 0-087 by 0-003 mm., lying in (or immediately beneath?) the
outer layer of cells (cf. also T. teniiicapitata, Fig. 48 c). With the further diff'erentiation
of the embryonic tissues, as, for example, the formation of the outer cortical layer, the
spicules migrate inwards, and become converted into smooth styli, measuring 0-14 by
0-007 mm. At the same time, small raphides, 0-07 mm. long, appear, scattered generally
throughout the inner cell mass (Fig. 47 k). The further growth of the embryo is marked
by a segregation of the styli in a neat bundle at the aboral pole of the embryo and a rapid
multiplication in the numbers of the raphides (Fig. 47 / and m).

The raphides, the prototypes apparently of the adult onychaeta, resemble acanthostyli
in their early stages, and it is probable that they are modifications of this spicule form,
and not modifications of oxea, as is often inferred, and as appears to be the case in the
raphides of other groups.

T. tenuicapitata, Ridley
Only a few embryos were found in three specimens, but from these it is clear that the
development follows the same lines as that of T. charcoti. The earliest stage observed
consists of a spherical mass of polyhedral cells contained in a capsule, the walls of which
are strengthened with small onychaeta and contain a few nutritive cells (Fig. 48 a).
The next stage found (Fig. 48 b) is of interest because in this the formation of the outer
unicellular layer is seen to be in progress, but has not advanced sufficiently far for the
inner cell mass to be completely covered. In addition, the formation of spicules can be
seen to have begun in those parts of the outer layer which are well established. Fig 48 c
represents a section through an embryo at a slightly later stage, and is intended to show
the approximate position of the spicules in the early stages of their development. Fig. 48 d
represents the latest stage obtained. The first spicules to appear are small roughened styli,
o-i by 0-004 mm., followed by raphides, 0-07 mm. long.

T. nmrdochi, Topsent
Only the later stages were observed ; and although the histology appears to be similar
to that of T. charcoti, there are several points in which the embryos of the two species
difi'er. The earliest embryo of T. nmrdochi is spherical and measures 0-28 mm., and
the most remarkable feature about it is the manner in which the surface is thrown
into loose folds (Fig. 50 a, b). The spicules are roughened subtylostyli, 0-14 by 0-003 mm.,
and raphides, o-i mm. long, and it may be remarked that, in contradistinction to the
spicules of the embryos of other species, the subtylostyli of T. nmrdochi are arranged in
opposite directions, some with the base directed orally and some with the base directed

Fig. 47. Tedania charcoti (Topsent).
(In Fig. 47 h, r = residual nutritive matter.)

b c

Fig. 48. Tedania tenuicapitata, Ridley.

Embryology of Antarctic
Note: All embryos are contained within a capsule,

Fig. 49. Tedania spinata (Ridley).

Fig. 50. Tedania murdochi, Topsent.

a ^ c

Fig. 52. Tedania pectinicola, Tliiele.

U U U 0 u

ah c d e f

Fig. 53. Embryonic spicules: a~b, T.charcoli;
c, T.pectinicola; d, T. temiicapitala; e, T. mur-
dochi; f, T. spinata.

Species of Tedania

whether this structure is figured or not, except 49 f.

Fig. 51. Tedania massa, Ridley and Dendy.

364 DISCOVERY REPORTS

aborally (Fig. 50 b, c). After escape from the capsule the embryo assumes an elongated
form (Fig. 50 c) and the conspicuous folds of the surface disappear.

T. spinata (Ridley)

The first stage observed consists of a sub-spherical mass of spherical to polyhedral
cells of varying size, 0-03 5-0-07 mm. in diameter, the whole being 0-28 mm. long by
0-22 mm. across (Fig. 49 a). The subsequent histological changes resemble those of
T. charcoti and result in an oval embryo with a central mass of granular cells and an outer,
cortical layer of columnar cells (Fig. 49 b). The first spicules to appear are roughened
styli, 0-15 by 0-004 mm., followed by raphides measuring 0-07 mm. long. The sub-
sequent changes are similar to those found in T. charcoti (Figs. 49 c and d).

T. massa, Ridley and Dendy

A few large embryos (Fig. 51) were found in two specimens, all at about the same
stage of development, so that it is impossible to say much concerning the development
in this species. Two important features may, however, be noted: (i) the large size of the
embryos, and (ii) that the first spicules to appear are raphides, measuring 0-14 mm.
long.

T. pectinicola, Thiele

Although this species does not figure in the Discovery collections, drawings of embryos
found in the type specimen are included (Fig. 52 a-c) for the sake of comparison with
other species.

Other species in which embryos were found are :

Plakina monolopJia, Schulze. A specimen, collected on January 17, contains numerous
embryos in various stages of development.

P. Irilopha, Schulze. One specimen collected on December 23 contains a few
ova; the second, collected on March 14, is filled with embryos in all stages of develop-
ment and a section of it presents the same appearance as that figured by Schulze (1877,
pi. iv, fig. 20) for Oscarella lobuloris (Schmidt). This is the first time the embryos of
Plakina trilopha have been recorded, and although Schulze has figured the develop-
mental stages of P. monolopha, which so far as they go do not differ from those of
P. trilopha, he omitted to record the date on which reproduction takes place.

Haliclona bilamellata, sp.n. Two specimens, collected in April, contain scattered ova
or embryos in very early stages.

Callyspongia flabellata, sp.n., collected in March, contains numerous oval embryos,
0-18 mm. in longest diameter. A few are aspiculous, but most of them contain oxea,
0-039-0-06 by 0-002 mm., which appear to arise as a tangential layer in the epiblast
and then to migrate inwards to form, finally, a compact mass at the centre of the embryo

(Fig. 54)-

Isodictya setifer (Topsent). A specimen, collected on March 21, contains eggs (or
young embryos?), spherical and with granular appearance, 0-087 i^"^- ^^ diameter. A

EMBRYONIC DEVELOPMENT 365

second specimen, collected on March 14, contains numerous large embryos, spherical or
oval in shape, of a rich yellowish brown in colour which makes them very conspicuous,
and measuring up to i mm. in longest diameter. There is no sign of spicules or of a
marked differentiation into an inner cell mass and an outer layer of smaller cells.

Isodictya antarctica (Kirkpatrick). A specimen, collected on June 21, contains
numerous large embryos similar to those of /. setifer, but more golden yellow in colour
and measuring up to o-8 mm. in diameter.

Guitarra fijnbriata, Carter. A specimen, collected on July 17, contains a few aspicu-
lous embryos, situated just beneath the surface, and measuring o-i mm. in diameter.

Mycale magellanica (Ridley). A specimen, collected on September 20, contains a few
embryos in an early stage of development, 0-12 mm. in diameter.

Amphilectus fiicoriim (Johnston). A specimen, collected on March 24, contains
numerous aspiculous embryos, spherical or oval, measuring 0-28 mm. in diameter, and
a second specimen, collected on May i, has oval embryos measuring o-i6 mm. in
longest diameter.

Asbestopluma calyx, Hentschel. A specimen, collected on January 20, contains nu-
merous embryos in an advanced stage of development. The spicules contained are
clavate tylostyli (or styli), o-i by 0-007 mrn., and anisochelae, o-oii mm. chord. In the
earliest stages seen tylostyli only are present, forming a compact bundle at the centre
of the embryo. In others anisochelae have made an appearance and are scattered ir-
regularly around this bundle, and in the most advanced embryos the bundle of tylostyli
is surrounded at its upper (?) end by a dense corona of anisochelae (Fig. 54 k).

lophorr radiatiis Topsent. Specimens, collected from January 9 to June 21, all contain
embryos in an early stage of development. Those described by Kirkpatrick (1908),
which are at a later stage of development, were obtained in February.

Anchinoe latninculioides (Ridley and Dendy). A single specimen, collected on
June 17, contains large oval, aspiculous embryos, measuring 0-4 mm. in longest diameter.

Anchinoe leptochela (Hentschel). The single specimen, collected in April, contains
numerous embryos. Some are aspiculous, slightly flattened at one pole and measuring
0-32 mm. diameter. Others, of approximately the same size but regularly spherical,
contain numerous chelae of the same shape as those in the adult and, in some instances,
a few acanthostyli. The chelae measuring 0-021 mm., about two-thirds the size of those
of the adult, and the acanthostyli 0-092 mm. long.

Kirkpatrickia variolosa (Kirkpatrick). One specimen, collected on December 19, is
filled with spherical embryos, 0-036 mm. diameter, most of which are in a comparatively
early stage of development. The most advanced stage, and one rarely met with in this
specimen, shows a compact central mass of cells with a single epiblastic layer, and
internally a few incipient acanthostyli 0-035-0-12 mm. long.

Plocamia gaussiana, Hentschel (Fig. 546-?'). Several specimens contain embryos:
those taken in December contain only a few aspiculous embryos (i.e. early stages);
those taken in January contain numerous embryos in varying stages of development, but
all, or nearly all, possessing spicules. From these, the course of development may be

366

DISCOVERY REPORTS

Fig. 54. Embryonic development in certain species of Tetraxonid sponges.

a~d, Callyspongia flabeUata, sp.n.

«, embryo after first appearance of spicules; b, section of same, showing spicules in a tangential
layer; c, spicules migrating inwards; d, latest stage observed, x 200.

e-i, Plocamia gaussiana, Hentschel.

e, embryo after first appearance of spicules ; /, sector of same to show dermal epithelium and large
granular cells of inner mass; w, first appearance of acanthostyli; //, sector of an embryo at a slightly
later stage, showing columnar external layer; i, latest stage observed. e,g, h x 100,/, h x 250.

j, Dictyociona discreta (Thiele). The latest stage observed, with dense axial core of plumosely arranged
acanthostyli. x 200.

k, Ashestospluma calyx, Hentschel. Embryo, x 160.

EMBRYONIC DEVELOPMENT 367

followed. The aspiculous embryos are spherical and attain a maximum diameter of
0-36 mm. and show an inner mass of large irregular cells with a thin layer of flattened
cells on the outside. From this point a decrease in size takes place to 0-3 mm. diameter,
the central mass becoming more compact and the outer layer being clearly marked off
as a thick layer, 0-035 ^™- thick, of columnar cells. At first the outer layer is of uniform
thickness throughout. The first spicules to appear seem to be the chelae, which measure
0-026 mm. chord, as against 0-035 ii^"'^- i" the adult, but do not otherwise present any
peculiarities. They are scattered evenly through the central mass of cells. Small
acanthostyli, 0-07 mm. long and difl^ering considerably from those of the adult (cf.
Fig- 54 g), then appear, also scattered through the substance of the central mass. A re-
arrangement then takes place, the acanthostyli finally being found in a bundle at the
centre of the embryo and towards one pole, their apices centrifugally directed, with the
chelae in a diff'use corona around the inner end of the bundle. At the same time, the
outer layer of columnar cells becomes much thinner near the outwardly directed bases
of the acanthostyli (Fig. 54 /). This may, by analogy, be regarded as the aboral pole.

The specimens recorded by me (1929, p. 435) from McMurdo Sound, taken in
February, contain large aspiculous embryos.

Dictyociona discreta (Thiele) (Fig. 547). One specimen, collected in March, contains
aspiculous embryos only, but a second specimen, collected in April, has a few spherical
embryos, 0-32 mm. diameter, containing rare, almost indefinable, acanthostyli, and
numerous others 0-25 mm. diameter, with a layer of acanthostyli measuring 0-07 by
0-002 mm. The acanthostyli are all lying tangentially to and just beneath the surface of
the embryo. In a third specimen, collected in July, a single embryo was found con-
taining numerous acanthostyli, arranged in a dense bundle at the centre, and numerous
chelae of 0-008 mm. chord (cf. the oxea in Callyspongia flabellata, sp.n.. Fig. 54^).

Eurypon miniaceum, Thiele. Collected in October, contains ova and very young
embryos.

Hymedesmia irritans, Thiele. A single specimen, collected at Tristan da Cunha on
February i, contains spherical embryos, aspiculous, measuring o-i6 mm. in diameter.

Hymeniacidon sanguinea (Grant). One specimen, collected on July 28 from Angola,
contains numerous embryos up to 0-2 mm. in diameter. Most of the embryos are
aspiculous, but some contain styli, with faintly irregular surfaces, measuring 0-12 by
0-003 mm. These spicules are at first arranged tangentially to the surface, but later
migrate inwards to lie in a confused mass towards the aboral (.?) pole with their apices
directed towards the centre of the embryo.

Hymeniacidon diibia, sp.n., collected in March, contains ova and very young
embryos.

Plicatellopsis flabellata, sp.n. The specimen, collected on June 22, contains a few oval
embryos, 0-07-0-1 mm. in longest diameter, in early stages of development.

Stylocordyla borealis (Loven), subsp. acuata (Kirkpatrick). A single specimen, col-
lected on December 15, contains numerous embryos in an advanced stage of develop-
ment and a few that are entirely aspiculous. Embryos have been found in the Terra

17-2

368

DISCOVERY REPORTS

Nova specimens of this subspecies in January, and in the Challenger specimens of the
subspecies globosa, also in January.

Polymastia isidis, Thiele. One specimen, collected on July 17, contains numerous
sperm morulae, 0-045 n^"^- ^^ diameter, in the inner tissues, and indications near the
surface that budding was about to begin. Around the outer ends of the bundles of the
main skeleton are gathered masses of cells of a peculiarly yellow and granular appearance,
and if we may judge from Eichenauer's (1915) work on Tethya maza, Selenka, these are
groups of amoebocytes (? archaeocytes) which will eventually form the bulk of the
substance of the buds.

Sphaerotyliis antarcticus, Kirkpatrick. Very young specimens are recorded by Kirk-
patrick (1908, p. 17) which were collected in January, June and September respectively.

o b b'

Fig. 55. Embryonal megascleres of a, Plocamia gaiissiana, Hentschel; b, Dictyociona discreta (Thiele);
c, Anchiiwe hptochela (Hentschel); d, Kirkpatrickia variolosa (Kirkpatrick); e, Ashestopluma calyx,
Hentschel. All x 500, except e, which is x 400 {b' and c' are developmental stages).

and as these young specimens were approximately of the same size, it is probable that
some form of reproduction goes on in this species all the year round. This is probably
an asexual budding. S. schnemis, Sollas, was also recorded by the same author to be
actively budding in January.

Spongia magellaiiica, Thiele. A specimen, collected on March 24, contains a number
of spherical embryos, 0-07 mm. in diameter, showing a marked similarity to those of
Hircinia variabilis , war. flavescens, figured by Schulze (1879, pi. iii, fig. i), and especially
to the one in the bottom left-hand corner of the figure. A clear area, spherical in outline,
is readily discernible at one pole of the embryo. The embryos are found chiefly on or
near the principal fibres of the skeleton.

POST-LARVAL DEVELOPMENT

Tetilla leptoderma (Sollas). There is a single small specimen from St. WS 27 which
measures 3 mm. in diameter. The skeleton consists of radiating bundles of oxea, pro-
and anatriaenes, which start from an excentrically placed nucleus. The form of the

BREEDING SEASONS

369

spicules is the same as in the adult, but the size is considerably less. The interesting
feature about this sponge is that the small cortical oxea, instead of being set more or less
at right angles to the surface, as in the adult, are arranged in a tangential layer at the
surface.

BREEDING SEASONS

In the table given below, the evidence obtained from the finding of embryos in situ,
based on the records of previous authors and the results of the present study, is brought

Table showing the approximate breeding periods of Antarctic species of
sponges, as deduced front the finding of ova or embryos.

Species

Calcarea

Acliramorp/ia nivalis, Jenkin
Tenthrenodes scotli, Jenkin
Megapogon raripilis, Jenkin
Leucetta leptoraplns (PolejaefF)
Tetraxonida
Oscarella lobularis (Schmidt)
Plakina monolopha, Schulze
P. trilopha, Schulze
Craniella leploderma (SoUas)
Cinachyra barbata, Sollas
C. antarctica (Carter)
Haliclona bilamellata , sp.n.
Petrosia fistidala, Kirkpatrick
P. depeUens, Topsent
Adocia tremulus (Topsent)
Hemigellius riidis (Topsent)
Callyspongia flabcllata, sp.n.
Isodictya antarctica (Kirkpatrick)
/. setifer (Topsent)
Guitarra fimbriata. Carter
Mycale magellanica (Ridley)
Amphilectus fucuriim (Johnston)
Asbestopluma calyx, Hentschel
Tedania charcoti, Topsent
T. massa, Ridley and Dendy
T. spinata (Ridley)
7'. nnirdochi, Topsent
T. tenuicapitata, Ridley
lophon radiatus (Topsent)
Myxilla australis (Topsent)
Lissodetidoryx imiominata. Burton
Anchinoe latriincidinidcs (Ridley and Dendy)
A. leptochela (Hentschel)
Kirkpatrickia variolosa (Kirkpatrick)
Plocamia gaussiana, Hentschel
Crelta crassa (Hentschel)
Dictyociona discreta (Thiele)
Eurypon miniaceum, Thiele
Hymedesmia irritans, Thiele
Dolichacantha macrodon, Hentschel
Hymeniacidon dubia, sp.n.
Plicatellopsis flabellaia, sp.n.
Stylocordyla borealis acuata (Kirkpatrick)
Polymastia isidis, Thiele
Sphaerotylus antarcticus, Kirkpatrick

Euceratosa
Halisarca dujardini , var. magellanica, TopsenX
Spongia magellanica, Thiele

Jan.

Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec

Where a definite record for a given month has been obtained, a black line is used: a dotted line signifies the appro,\imate
period, based on the stage of development reached in the embryos.

37°

DISCOVERY REPORTS

together in an endeavour to determine the breeding seasons for sponges in the Antarctic.
It is clear that the records are by no means sufficiently complete to be able to draw
definite conclusions from such a table; but the tentative deductions are interesting. In
the first place, there does not appear to be any part of the year during which a majority
of species are in a state of reproduction. The table shows that so far as can be told from
the available evidence some species are breeding at all times of the year ; but, as none of
the specimens investigated were littoral, there is perhaps little surprising in this. Secondly,
there is an indication that all or most species of Calcarea breed throughout the year, for
although the number of specimens examined is relatively small, the number of records of
ova and embryos is comparatively high. This might be said of the Euceratosa also.

Comparison with the breeding seasons in the northern hemisphere is difficult, partly
again because of the paucity of observations (many authors have described embryos but
omitted to give the date when they were seen), and partly because the records here are
mainly from littoral or shallow-water specimens, whereas the Antarctic records are
mainly from deeper water (down to 500 m.). Nevertheless, we again find that the
Euceratosa are in a state of reproduction throughout the year, and again a suggestion
that the same may be true of the Calcarea (cf. Sycon raphamis, Schmidt ; see table given
below). Further, there is a marked breeding period from approximately May to October.

Table showing the approximate breeding seasons of the European
and North Atlajitic species of sponges.

Species

Calcarea

Sycon ciliatum (Fabricius)

5. raphanus, Schmidt

Grantia compressn (Carter)

Leuconia aspera (Schmidt)
Tetraxonida

Oscarella lobiilaris (Schmidt)

Halicluna ocitlata (Pallas)

Mycale contarenii (Bowerbank)

M. minima (Waller)

Myxilla inctustans (Johnston)

Hymedesmia diijardini, Johnston

Halichondria panicea (Pallas)

H. bozcerbanki, Burton

Hyyneniacidon caruncula, Bowerbank

Tethya aurantiiim, Pallas

Cliona celata, Grant
Euceratosa

Spongia officinalis, Linnaeus

Hippospongia equina, Schmidt

Hircinia variabilis, Schulze

Halisarca diijardini, Johnston

Jan.

Feb.

Mar

Apr.

May

June

July

Aug.

Sept.

Oct.

Nov.

In the one case, Oscarella lobularis (Schmidt), of a species common to the North
Atlantic and the Antarctic (in which the breeding season is known) it is found that the
same season is occupied in each locality. Thus, in the northern hemisphere, this
species is, presumably, reproductively active from June to November, and in the
Antarctic, from August to December.

Dec.

371

UNATTACHED SPONGES

In discussing the possible means by which the distribution of sponges is effected
(see p. 358), the assumption is usually made that there is no such thing as a free-living
sponge capable of being transported in the adult stage by currents. Certain specimens
of Tedania teniiicapitota, Ridley, in the present collections indicate the possibility that
even adult sponges, often of large size, may pass their life floating just above the sea-
floor, or at least resting unattached on the sea-bottom and capable of being lifted by
currents of moderate force. This^ may help to explain a number of anomalies pre-
viously observed in other species ; and, in view of the importance of such observations as
possibly affecting geographical distribution and the bionomics of sponges as a whole,
it will be of interest to deal fully with the subject.

The sole means of observing sponges in the living state is confined to observation
between tide marks, in aquaria, and in exceptionally favourable circumstances, with
a diving helmet. In the first two, it is obvious that nothing in the nature of floating
sponges can be observed, and in the third, the results of such investigations have
never been recorded. It is necessary therefore to rely on indirect observation of
preserved material.

Tedania temiicapitata, Ridley, is represented in the present collections by some 200
specimens. Most of these are damaged or at least incomplete at the base, presumably
where the trawl has torn them from the substratum. A few are, however, entire, and some
of these show no visible sign of having been attached either to the sea-floor or to any such
object as a hydroid or polyzoon. For the most part, such specimens are sub-spherical
and of small size. There are, in addition, three discoid specimens which seem to offer
convincing evidence in support of the suggestion that some sponges may not be
permanently fixed to the substratum and, under favourable circumstances, slightly
buoyant.

The first of these specimens (Fig. 56 a) is discoid, 16 cm. across, i2 cm. high and
3 cm. at the thickest point. The margin is even, and a cross-section in any direction
shows a fairly regular ellipse (Fig. 56 b). At the base, the sponge is attached to the concave
inner surface of a bivalve shell, and between the shell and the adjacent sponge tissue
is a small quantity of sand. The margins of the sponge which encroach on the outer
surface of the shell are, however, entire, without sand inclusions, and show no sign
whatsoever of having at any time been attached to the substratum, rock, stone or hy-
droid, or of having been in contact with, or buried in sand. Further, there are several
hydroids growing on the shell, and in each case these point downwards. The only
possible assumption is that the sponge was free-living and, since it was obtained in a
trawl, floating just above the floor of the sea.

1 The idea is not novel, for Bowerbank (i866,p.97) szys of Halknetnia patera : "The habit of the sponge of
including a small pebble in its centre, apparently as ballast, is very remarkable, and this is the only case in
which there is an indication of a natural tendency to locomotion belonging to the adult species that I have ever
observed among the Spongiadae".

372 DISCOVERY REPORTS

The second specimen is even more convincing. This is more regularly discoid
(Fig. 56 c), 21 cm. high by 17 cm. across by 3 cm. thick at the centre. The margin is
entire throughout, with no sign of attachment, and the only inclusions are a few patches
of sand at the centre and a small pebble embedded in the lower third (Fig. 56 d). Some
hydroids are growing out from the surface of the sponge, and the direction in which they
are pointing suggests that the sponge floated vertically (as figured) with the portion
containing the pebble lowermost.

The third specimen has much the same shape and size as the second, but is entirely
without large inclusions, shell or pebble. The margin is entire throughout, but at several
widely separated points on it there is a peculiar roughening of the surface, suggesting
incipient rooting processes, as though the sponge, although unattached, had at times
come to rest on the sea-floor sufliciently long for rooting processes to start growing.

That there can be no question of these sponges having been erect, with their bases
permanently embedded in sand,^ can be proved from the following:

(i) Although members of the species readily take up sand (see below) these specimens
are particularly free from it.
(ii) The tissues and skeleton are normal throughout the entire margin,
(iii) Flagellated chambers are present just beneath the surface throughout the margin.
(Had these sponges been embedded in sand, it is unlikely that the flagellated chambers
would have persisted in the submerged portions.)

With regard to the first point, the ease with which individuals of Tedania teniii-
capitata will absorb sand grains is shown by the majority of the Discovery speci-
mens. In some cases, the inclusions are confined to the base, which was presumably
buried in the sand, but in most cases sand is found throughout the specimen (Fig. 56^).
Fig. 56 e shows a section of a flabellate specimen with a sandy base, and Fig. 56/ shows
an irregularly "table-shaped" sponge (spreading horizontally) with several "legs"
which were buried in the sand at the base. In view of this therefore it is improbable
that the margins of the discoid specimens could have been in contact with the "coarse
brown sand" of the sea-floor without showing some trace of its inclusion.

It is noteworthy that both the discoid specimens, when immersed in 80 per cent,
alcohol, only just sink. That is to say, they will sink, but the slightest touch on the under
surface, when the specimens are standing erect, will cause them to rise in the spirit. This
feature is more strongly marked than in any other preserved sponge experimented with.
Since the specific gravity of sea water is greater than that of alcohol, it is reasonable to
suppose that these particular sponges would be unusually buoyant.

Other examples of sponges which may have floated above the sea-floor may be easily
found. The young of Cinachyra barbata, Sollas, for example, are sub-spherical (cf.
Kirkpatrick, igo8, Plate ix, fig. 3), completely covered with a surface pile of spicules and
showing no sign of having been attached ; and the same may be said of adult specimens
up to 10 cm. in diameter. Many specimens of Thenea show no evident point of attach-

1 Coarse brown sand formed the substratum at Sts. WS 222 and WS 96 where these sponges were found.

UNATTACHED SPONGES 373

ment, while Cinachyra antarctica, Carter, only rarely shows a basal tuft. In that species,
the usual form is sub-spherical, with projecting tufts of spicules and numerous
porocalices distributed evenly all over the surface (see Lendenfeld, 1907, Plate xxii,
figs. 22-33, under Cinachyra vertex). Were such a sponge to lie on the sea-floor, some
of the porocalices would be rendered useless, and would probably disappear as a con-
sequence, and the ends of the spicule bundles would almost certainly be damaged. And
further, the projecting spicules in the, supposedly, free sponges, as opposed to the
rare examples of this species in which a basal tuft is present, are more markedly spiral
in arrangement, suggesting that the sponge has been subject during development to a
gentle rotary movement.

It is difficult to imagine Disyringa dissimilis, Sollas (Fig. 56 i), leading anything but
a free existence, and it is interesting to compare this with the three known species of an
allied genus, Monosyrwga. The first of these, M. longispinum (Lendenfeld), is spherical
with an oscular tube (Fig. 56 k) and the body is covered with delicate sucker-like pro-
cesses. In this species there is no sign of attachment and the processes are all free, but
in M. mortenseni, Br^ndsted and M. brondstedi. Burton, the known individuals, identical
in external form with M. longispinum (Lendenfeld), are covered with pebbles and
calcareous debris, each pebble or fragment being attached to the sponge body by one or
more of the sucker-like processes (Fig. 56 /). Presumably, as there is no evidence of
other point of attachment, these species may be free-living or may attach themselves to
numerous small bodies which serve as anchors, or rather ballast.

In all probability, the species described by Ridley and Dendy (1887) as Cladorhiza
longipinna,^ C. siniilis, C. inversa and Axoniderma mirabile are also free-living forms. It
is, at least, questionable whether the radiating processes would serve "to support the
sponge in the soft mud on which it lies " {loc. cit., p. 94). Rather would it appear that the
shape here is an adaptation to a free-living existence, especially by analogy with Crino-
rhiza and certain Echinoderm larvae. The sponges of this genus are probably figured
upside-down by Ridley and Dendy (cf. Fig. 567).

Another species which appears occasionally to be free-living is Polymastia invaginata,
Kirkpatrick. In most cases, the individuals of this species are attached to stones of
sufficient size to anchor them to the substratum, but in others there is reason to believe
that, although a pebble is embedded in the base, the sponge actually floated above the
level of the sea-floor. In a few individuals in the Terra Nova collections, there is no sign
at all of attachment (Fig. 56 n), in others there is likewise no sign of attachment but a
small pebble is present in the centre of the under surface (Fig. 56 m). This suggests
that the larva originally settled on a pebble, but from the shape of the adult it seems
unlikely that the sponge could have rested permanently on the bottom. The most
illuminating specimen of this species is, however, one recorded by Kirkpatrick (1908),

1 Topsent {C.R. Acad. Sci. Paris, cxxxiv, 1902, pp. 58-60) has advanced anatomical arguments for
believing that the habitual poise of these species is contrary to that suggested by Ridley and Dendy. I, too,
have found that no matter how the specimens be thrown into a liquid they quickly and invariably assume
the orientation shown in Fig. 56 y.

D VI i8

Fig. 56. Sponges with little or no attachment to the substratum.

a~h, Tedania sphiata (Ridley).

a, a discoid specimen, with hydroids growing vertically downwards from the basal shell and small
patches of included sand; h, section of same; c, a second discoid, with included pebble, hydroids
growing out from the surface and included patch of sand at the centre ; d, section of same ; e, a
flabellate specimen, showing method of attachment ; /, a " table-shaped "specimen, showing method
of attachment ; jC-, section through a flabellate specimen, showing manner in which sand is taken up
by the sponge ; A, a sub-spherical specimen, bearing hydroids and a cirripede, from St. 222, taken in
a tow-net. a-g x ^, h x h

i, Disyringa dissimilis, iioMis. x i.

j, Cladorhiza longipinna, Ridley and Dendy, showing orientation, x i.

k, Monosyringa longispintim, Lendenfeld, showing suckers.

/, Monosyringa brondstedi, Burton, showing suckers with attached pebbles, etc. x i.

tn-p, Polymastia invaginata, Kirkpatrick.

OT, a (floating ?) specimen with a small pebble embedded in the base; n, a "free" specimen; o, a
specimen seated on a rock; p, section through a young specimen, showing choanosome {cli),
collenchymatous base (col) and included calcareous fragment [cf). m-o x \,p x 2.

EXTERNAL FORM AS A SPECIFIC CRITERION 375

I cm. in diameter, in which the lower ( ?) half of the sponge consists only of collen-
chymatous (gelatinoid) tissue completely enclosing fragments of calcareous debris
(Fig. 56/)). The collenchymatous portion naturally suggests something in the nature of
a float, with the fragments of debris acting as balancers.

Except for the species of Cladorhiza and Axoniderma, which may be habitually free-
living, Disyritiga and Motiosyringa , which may be occasionally free-living, it is probable
that the larvae first settle on small stones or pebbles, perhaps even large sand grains.
If these be sufliciently heavy the sponge becomes attached for life ; but if small, the
growing sponge will gradually lift the stone or pebble on which it has settled, as their
combined specific gravity decreases, and even, eventually, become grown round the
stone or pebble, entirely enclosing it. In this connection it must be remembered that the
specific gravity will vary from species to species, or from one individual to another of
a single species, accordingly as the skeleton is slightly built or densely constructed, the
relative quantities of spicules in individuals of a given species being subject to consider-
able fluctuation. It may, on the other hand, be argued that no organism with tissues
charged with silica and devoid of appendages or special organs of flotation could
possibly do anything else than remain permanently on the bottom. Yet, since the
tissues of many sponges are densely charged with fat globules, and the spicules are
not solid silex but contain an axial canal, it is possible that the specific gravity of
sponges generally is less than is usually assumed.^ The only real test worth applying
would be specific gravity readings on the (apparently) unattached forms while still
alive. One thing, at least, is clear, that many species of sponges are definitely free-
living, and there is every indication that some of these at least may not rest permanently
on the floor of the sea.

THE VALUE OF EXTERNAL FORM IN THE IDENTIFICA-
TION OF SPONGES: WITH NOTES ON VARIATIONS IN
THE SIZE AND SHAPE OF SPICULES

In the early days of the systematic study of sponges, external form was the only thing
considered worthy of description. It was, in fact, not until after i860, when Bowerbank
and Schmidt published their well-known monographs, that the spicules were seriously
considered as diagnostic features. In a very short time, the skeleton became the chief
object for study by the systematist. To-day, considerably more attention is paid to
the skeleton in the description of a new species than to the external form, and the
tendency is rather towards omitting illustrations of the external form. As a result of
my own experience, I am coming to the conclusion that, for the identification of species,

1 I am indebted to members of the Discovery Staff, Dr Mackintosh and Mr Eraser, for the information
that many specimens of Tedania, notably T. massa, emit large quantities of mucus when taken from the
dredge. The presence within the tissues of large quantities of mucus would possibly result in lowering
totally the specific gravity of the sponge as a whole, and it is conceivable that the texture and volume of a
preserved specimen of T. massa differ appreciably from those of the living sponge.

18-2

376 DISCOVERY REPORTS

external form is a more reliable feature than the skeleton ; that it is more constant than
the shape of the spicules, and far less variable, under normal conditions, than is usually
assumed.^ In addition far too little attention has been paid to the construction of the
skeleton. In my opinion, the surest guide for the diagnosis of families is found in the
types of spicules present ; for genera, their arrangement and disposition in the skeleton,
with minor variations in their shape; for species, the external form, with the minor
variations in the arrangement of the spicules, and, less important, the categories of
spicules present. In other words, the best method of identification lies between that
adopted prior to 1860-70 and that which has been in use since.

Another practice which is to be deplored is that of founding species on badly damaged
specimens, fragments of specimens and obviously pathological individuals (including
those which occasionally replace their proper skeletons by sand or foreign spicules).
Such objects, used as holotypes, give infinite trouble and rarely can their true characters
be ascertained.

If, however, the skeleton is to be used in so important a role, it is at least time that
some systematic effort were made to determine the extent of variation (i) in size, (ii) in
shape, and (iii) in arrangement. Also, to collect data on the extent to which categories
of spicules may disappear in individuals of a species, or be differentiated into two or more
sizes. In studying the sponges collected by the Discovery expeditions, the following
observations have been made and appear worthy of record :

Variation in the size of spicules

The first important observation on this point concerns Calyx arciiarius (Topsent). The
oxea in the type measure 0-35 mm. long, in the specimens recorded by me (1929, p. 422),
0-27-0-35 mm., not 0-2-0-3 mm. as stated, in the present specimens of the Discovery,
0-22-0-24 mm., and in a specimen collected at South Georgia by Mr P. Stammwitz,
0-15 mm. long. The spicules of the last specimen are less than half the size of the maxi-
mum recorded.

In Axociella flabellata (Topsent) the large styli and dermal subtylostyli have a maxi-
mum of i-o and 0-650 mm. respectively, in the Terra Nova specimens and, in one of
the present specimens, of i-2 and 0-640 mm. In the other Discovery specimen and
a specimen collected by the Shackleton-Rowett expedition, 0-64 and 0-35 mm. respec-
tively. Again, the maxima recorded are twice the recorded minima.

In lophon proximum (Ridley), the acanthostyli vary from 0-14 to 0-44 mm. (taking
average sizes in different individuals), and the tornota from 0-14 to 0-315 mm.

It is clear from these examples alone that such phrases as " this species differs from . . .
in the greater size of its spicules", are useless. Ceteris paribus, spicule size can have no

1 Strong objection is here taken to the all too prevalent use of the term "amorphous" in describing the
external form of sponges. The term has been used to include encrusting, massive, lobate, sub-digitate,
clathrate and plicate sponges, and a number of combinations of two or more of these. The type of one of
Lendenfeld's species of Australian Chalininae is a beach-worn sponge which has been, presumably, trodden
under foot. This was described as "amorphous" and is the only truly amorphous sponge I have ever seen.

EXTERNAL FORM AS A SPECIFIC CRITERION 377

significance for the systematist, and it is only necessary to count the number of varieties,
subspecies and species described in the Hterature devoted to Antarctic sponges, and
founded on spicule size alone, to realize this.

Variation in the shape of spicules

The shape, too, of spicules is not so constant as is commonly supposed. It will vary
more in one species than another, it is true, and from one stage in the life of an individual
to another (cf. Burton, 193 1), but in all species allowance must be made for these minor
differences, instead, as has been the practice, of investing them with specific importance.
A good example of this is given under lopJion proximiim (Ridley) (see pp. 296-30 1 ), but the
most striking case, which I hope to deal with in a forthcoming report, concerns Clathria
toxipraedita, Topsent, in which the ornamentation of the acanthostyli, styli and dermal
subtylostyli is very variable, while the general shape of these spicules also varies con-
siderably.

The arrangement of the spicules in the skeleton is more constant than either their
shape or size, but even here variations are found which can be very misleading. In
appreciating the unreliability of the skeleton for systematic study, we are driven therefore
to accept the external form of the sponge as the final and deciding criterion in its
identification.

The suppression of microscleres

The dropping out of categories of spicules is a common phenomenon, but one which
is not sufficiently appreciated. This seems to be particularly evident among the Mycaleae,
but all types of sponges are liable to it. It is probable that sufficiently long searching
would show that the apparently absent spicules will be found to exist, even though in
minute quantities, in some parts of a sponge. That may or may not be true, but in any
case the practical difficulty still remains. Apropos this point, there is a specimen of
Mycale magellanica in the present collection in which no spicules other than tylostyli
could be found ; yet the external form showed it clearly to belong to that species. I have
myself described (1930, p. 333) as Axiiiosia incrustans, sp.n., a specimen which proved
eventually to be an Amphilectus fucoriim (Johnston) with very rare chelae. Most of the
species of Retiiera described by Dendy (1921) from the Indian Ocean contain rare
sigmata, which may be found only after prolonged search. Numerous examples of this
could be related (see also Mycale magellonica, p. 288), yet to suggest that micro-
scleres are an unreliable criterion for taxonomic purposes is to make an unwelcome
proposal. The fact is that they are secondary in importance to the structure of the main
skeleton.

Differentiation within a single category of spicules

Thiele (1898, pp. 13-14) describes two species of Stelletta from Japan, S. validissima
and S. orientalis, the former having oxea, dichotriaenes, protriaenes and two sorts of
anatriaenes, the latter having oxea, dichotriaenes, protriaenes and only one sort of
anatriaene. Both are from adjoining localities, have the same external appearance, and

378 DISCOVERY REPORTS

are, indeed, similar in all respects but the possession of two sorts of anatriaenes by the
one species, and one sort by the second species. It seems unreasonable to doubt that the
two species are identical, but for this it must be presupposed that any category of
spicule is capable of differentiation into two sizes. A striking example of this is found in
Craniella disigma, Topsent (1904, p. 100), which differs from Tetilla cranium (Miiller)
Autt., in one respect only, that it has two sizes of sigmata instead of one. Here, again,
it seemed probable that Craniella disigma was identical with Tetilla cranium. To test
this I took a jar from the Porcupine collection, containing some thirty specimens
identified by Carter as Cranilla cranium (Miiller). Only one had been previously cut.
The specimens were all alike externally and proved, on being sectioned, to be all alike
internally; one of them, however, had two sizes of sigmata, exactly as in C disigma,
Topsent. Further, Clathria frondifera, var. dicliela, Hentschel (1912), is probably a
specimen of C. frondifera (Bowerbank) in which a similar differentiation has taken place.
Taking, now, all the variations to which the skeleton may be subject, it is not unreason-
able to suggest that the external form may conceivably offer a more reliable means of
identification, if only as the final arbiter.

THE ABUNDANCE OF SPONGES IN THE ANTARCTIC
AND ITS GEOLOGICAL SIGNIFICANCE

According to Prestwich {Geology: Chetnical, Physical and Stratigraphical, 1886, p. 125),
"in their geological relations the forms and distribution of the sponges are of the
highest interest". This being so it would be as well to be sure that the statements
upon which geological arguments are based are correct. Merely a superficial acquaint-
ance with geological literature is sufficient to show that this is not always the case.
There is, for example, a strong belief among geologists, as shown by the literature,
that the growth of sponges is most prolific in warm seas, and that, a priori, a deposit
rich in sponge remains must have been laid down in a warm sea. The abundance
of sponges in the Antarctic, as shown by the present collections, definitely proves
that in recent times at all events this is not the case. Judging by the collections and
by the verbal accounts of collectors, there is every reason to believe that sponges
are at least as abundant in the Antarctic as in, say, the West Indies, Australia or
the Indian Ocean. And the probability is that they are considerably more abundant.
Certainly, the average size of the specimens exceeds that observed in any other part of
the world. And finally, what is even more important from the geological point of view,
whereas a large percentage of the sponges in warmer seas are keratose or pseudo-
keratose, the remains of which are unlikely to be preserved in a fossil state, those of
the colder waters are almost exclusively siliceous, and the spongin, when present, does
not form so great a percentage of the skeleton as in corresponding species in the warmer
seas.

There is, further, the belief that sponges reached their "maximum development in

GEOLOGICAL SIGNIFICANCE OF SPONGES 379

the Cretaceous period". Again there is no evidence to support this. Neither in the
complexity of the skeleton, the size of the spicules, the size of individual sponges, nor
in the diversity of species, even when allowance be made for the fact that many species
living in Cretaceous times are as yet unrepresented in our collections, can the Cretaceous
fauna be said to have reached a degree of development greater than that represented by
the sponges of the Antarctic to-day, or, for that matter, any other quarter of the globe.

And, finally, there is the question of the segregation of silica in the sea, in relation to
the formation of chert and flint. Some geologists find it difficult to believe that the vast
quantities of silica contributed annually to the sea by rivers (320,000,000 metric tons)
can all be accounted for by the silica deposited as animal skeletons,^ even allowing for a
small percentage known to be deposited in a colloidal state in the normal course of
sedimentation. Judging by the sponge fauna of the European or American coasts, there
is doubtless some ground for such doubt, but if the Antarctic be considered too the
case assumes a difl^erent aspect. Here, in a vast expanse of ocean, with the erosive forces
calculated to contribute silica to the sea reduced to a minimum, the growth of siliceous
sponges is greater than in any other part of the world : and it can only be supposed that
the supply of silica in this area is drawn from other quarters. Even more important,
however, is the suggestion that, when it is considered how many siliceous sponges were
collected by the Discovery expeditions at a relatively few pin-points of the Antarctic
expanse, these 3 20,000,000 metric tons are probably barely sufficient to meet the demands
of all silica-secreting organisms, sponges, radiolaria, etc. In fact, the assumption is
more probable that, in addition, a fair proportion of the silica from the skeletons of dead
sponges, and other organisms, may enter once again into circulation in the sea.

There is another point, from which no definite conclusions can be drawn, but the
mention of which may be of interest. Among all the Antarctic collections represented
in the British Museum, numerous examples may be found of accumulations of spicules
into irregularly rounded masses, or balls (see Plate LVI I , figs. 14-16). These are formed of
the larger spicules of siliceous sponges, matted and felted together in compact masses up
to 20 cm. in diameter : and their occurrence suggests that, indirectly, sponges may be the
agents for a considerable amount of the segregation of silica necessary for the formation
of chert and flint, in a manner hitherto unsuspected. There can be little doubt that these
" spicule balls " are formed in much the same way as the balls of Posidonia fibres, by the
action of waves, and it is probable that consideration of their presence in large numbers
on the sea-bottom, taken in conjunction with the suggested continuous extrusion of
sponge spicules during life (cf. Burton, Proc. Zool. Soc. London, 193 1, p. 524) may be
of importance in the study of cherts and their origin.

1 See Tarr, Univ. Missouri Studies, 1926, Vol. i, No. 2, p. 2.

LIST OF LITERATURE

Arndt, W., 1912. Zoologische Ergebnisse der ersten Lehr-Expedition der Dr P. Schottldnderschm Jubildums-

Stiftung. Jahresber. Schles. Ges. vaterl. Cultur, pp. 110-36.
Arnesen,E., 1920. Spongia. Rep. Sci. Res. 'Michael Sars', North Atlantic Deep Sea Exped., 1910. Bergen,

in, 2, Zool., pp. 1-28, 2 charts, 5 pis.
Babic, K., 1921. Monactinellida und Tetractinellida der Adria. Glasnik Zagreb, xx.\iii, pp. 77-93.

1922. MonacUnellida und Tetractinellida der Adria. Zool. Jahrb., Abt. Syst., XLVI, pp. 217-302, 2 pis.,

50 figs.
Bellamy, J. C, 1839. The Natural History of Soiit/i Devon. Plymouth.

Blainville, M. H. D. de, 1834. Manuel d'Actinologie ou de Zoophytologie . Paris, viii + 695 pp.
BowERBANK, J. S., 1858. On the Anatomy and Physiology of the Spongiadae. Parti. On the spicula. Phil.

Trans. Roy. Soc. London, CXLVIII, pp. 279-332, pis. xxiii-xxvi.

1866. A Monograph of the British Spongiadae. London, 11, 388 pp.

1874. A Monograph of the British Spongiadae. London, iii, 376 pp., 92 pis.

1882. A Monograph of the British Spongiadae. London, iv, 250 pp., 17 pis.

BR0NDSTED, H. v., 1923. sponges from the Auckland and Campbell Islands. Vidensk. Medd. Kjobenhavn,

LXXV, pp. 117-67, 36 figs.

• 1924. sponges from New Zealand. Part L {Papers from Dr Th. Morfe?isen's Pacific Expedition, i^i^-

16.) Vidensk. Medd. Kjobenhavn, LXXVII, pp. 435-83, 34 figs.

1926. sponges from New Zealand. Part IL Vidensk. Medd. Kjobenhavn, Lxxxi, pp. 295-331, 13 figs.

. 1527. Antarctic and subantarctic Sponges collected by S. Wallin, 1924. Ark. Zool. Stockholm, 19 A, 6,

pp. 1-6, figs. 1-4.
Burton, M., 1926. Description of South African Sponges collected in the South African Marine Survey.
Part L Myxospongida and Astrotetraxonida. Fish. Mar. Biol. Survey, Cape Town, Rep. ix, 1925,

PP- 1-29. 3 pis-

1929. A comparative study of the Characteristics of shallow-water and deep-sea Sponges, with notes on

their external form and reproduction. Journ. Quekett Microsc. Club, London, xvi, pp. 49-70, i pi.,

7 figs.

1929. Porifera. Part I L Antarctic Sponges. British Antarctic ('Terra Nova') Exped., 1910. Natural

History Report, Zool., vi, 4, pp. 393-458, 5 pis., 9 figs., London.
. 1Q30. The Porifera of the Siboga Expedition. lU. Calcarea. Siboga-Expeditie, Leiden, VI A, pp. 1-7,

8 figs.

1930. Norzvegian Sponges from the Norman Collection. Proc. Zool. Soc. London, pp. 487-546, 2 pis.,

8 figs.
i(j30. Report on a collection of Sponges from South Georgia and from Campbell Island, South Pacific,

obtained by Dr Kohl-Larsen. Senckenbergiana, Frankfurt, xii, pp. 331-5, i fig-

1930. Additions to the Sponge fauna of the Gidf of Manaar. Ann. Mag. Nat. Hist., v, pp. 665-76, 3 figs.

1931. The interpretation of the embryonic and post-larval Characters of certain Tetraxonid Sponges. Proc.

Zool. Soc. London, pp. 511-25, i pi., 6 figs.
Burton, M., and Srinivasa Rao, H., 1932. Report on the shalloiv-water Marine Sponges in the collection of

the Indian Museum. Rec. Ind. Mus. Calcutta, xxxiv, pp. 299-356, i pi., 16 figs.
Carter, H. J., 1S80. The Zoology of Barents Sea. Ann. Mag. Nat. Hist., vi, pp. 253-7.
__ 1882. Some Sponges from the West Indies and Acapulco in the Liverpool Free Museum described, with

general and classificatory Remarks. Ann. Mag. Nat. Hist., ix, pp. 266-301.
Dendy, A., 1918. Calcareous Sponges. Australian Antarctic Exped., 1911-14, Series C, vi. Part i, pp. 1-17,

I pi.

1921. Report on the Sigmatotetraxonida collected by H.M.S. ' Sealark\ Trans. Lmn. Soc, Zool.,

London, xviii, pp. 1-164, pis. i-xviii.

- 1924. Porifera. Part L British Antarctic ('Terra Nova') Exped., 1910. Natural History Report,

Zool., VI, 3, pp. 269-392, pis. i-xv. British Museum, London.

LIST OF LITERATURE 381

Dendy, a., and Row, R. W. H., 1913. The Classification and Phylogeny of the Calcareous Sponges, with

a reference list of all the described species, systematically arranged. Proc. Zool. Soc. London,

pp. 704-813.
EiCHENAUER, E., 1915. Die Knospenentwickhmg von Donatia ingalli und Donatia maza. Zool. Anz., xli,

pp. 271-84, 4 figs.
EsPER, E. J. C, 1794. Die Pflanzenthiere . Niirnberg, 410, 303 pp., 49 pis.
Ferrer Hernandez, F., 1914. Esponjas del Cantabrico. Parte Segunda. Ill, Myxospongida. IV, Tetra-

xonida. V, Triaxonida. Trab. Mus. Nacion. Cienc. Xat. Madrid, Ser. Zool., xvii, 1914

(46 pp.).
Fleming, J., 1828. A history of British Animals. Edinburgh: London, 8vo.
Fristedt, K., 1885. Bidrag till kdnnedomen om de vid Sveriges vestra Kiist lefvande Spongiae. Vet.-x\kad.

Handl., Stockholm, xxi, pp. 1-56.
1887. Sponges from the Atlantic and Arctic Oceans and the Behring Sea. ' Vega ' Exped. Vetensk. lakttag.,

Stockholm, IV, pp. 401-71, pis. xxii-xxxi.
Graeffe, E., 1882. Uebersicht der Seethierfaiina des Golf es von Triest. ii. Coelenteraten,Spongiariae. Arb.

Zool. Inst. Wien, iv, pp. 313-21.
Grant, R. E., 1826. Observations on the Structure of some Siliceous Sponges. Edinburgh New Phil. Journ.,

I, pp. 341-51. Transl. Froriep's Notizen, xvi, Coll. 113-21.
Gray, J. E., 1867. Notes on the arrangement of Sponges, with the descriptions of some new genera. Proc. Zool.

Soc. London, pp. 492-558.
Gray, S. F., 1821. A Natural Arrangement of British Plants. Vol. I. London, xxviii + 824 pp., 21 pis.

(The description of the sponges is by J. E. Gray.)
Kallmann, E. F., 191 2. Report on the Sponges obtained by the F.I.S. 'Endeavour' on the Coasts of New

South Wales, Victoria, South Australia, Queensland and Tasmania. Part I. Zool. Results Fish.

Experim. 'Endeavour', Part II, pp. 117-300, pis. xxi-xxxvi.

1920. New Genera of Mona.xonid sponges related to the genus Clathria. Sydney Proc. Linn. Soc.

N.S.W., XLiv, 176, pp. 767-92, pis. xxxvi-xl.
Hentschel, E., 1909. Tetraxonida. I. Die Fauna Siidwest-Australiens, hrsg. von W. Michaelsen und

R. Hartmeyer, 11, 21, Jena, pp. 347-402, 29 figs., 2 pis.
• 1911. Tetraxonida. II. Die Fauna Siidwest-Australiens, hrsg. von W. Michaelsen und R. Hartmeyer,

III, pp. 279-393, 54 figs.

1912. Kiesel- und Hornschwdmme der Aru- und Kci-Inseln. Abh. Senckenb. Ges.,xxxiv, pp. 295-448,

pis. xiii-xxi.

1914- Monaxone Kieselschwdmme und Hornschwdmme der deutschen Siidpolar-Expedition, 1901-3.

Deutsch. Siidpol. -Exped., xv. Zoology, 7, pp. 35-141, pis. iv-viii.

1916. Zoologische Ergebnisse der schzvedischen Expedition nach Spitsbergen, 1908. Vet. -Akad. Handl.

Stockholm, Liv, 3, pp. 1-18.
Jenkin, C. F., 1908. Calcarea. National Antarctic Exped., Natural History, iv, 3, 50 pp., 12 pis.
Johnston, G., 1842. A history of British Sponges and Lithophytes. London: Edinburgh: Dublin, 8vo,

264 pp., 25 pis.
Kirkpatrick, R., 1907. Preliminary Report on the Monaxonellida of the National Antarctic Expedition. Ann.

Mag. Nat. Hist., London, XX, pp. 271-91.

1907- Porifera, Hexactinellida. National Antarctic Exped., Natural History, III, i, 25 pp., 7 pis.

1908. Tetraxonida. National Antarctic Exped., Natural History, IV, 2, 56 pp., 15 pis.

Lambe, L. M., 1894. Sponges from the Pacific Coast of Canada. Proc. and Trans. Roy. Soc. Canada,
XI (iv), pp. 25-43, pis. ii-iv.

1896. Sponges from the Atlantic Coast of Canada. Trans. Roy. Soc. Canada, ir, pp. 181-211,

pis. i-iii.

1900. Sponges from the Coasts of North-eastern Canada and Greenland. Trans. Roy. Soc. Canada (2),

VI, pp. 19-49, pis. i-vi.
Lendenfeld, R. von, 1887. Die Chalineen des australischen Gebietes. Zool. Jahrb., 11, pp. 723-828, 10 pis.

1889. A monograph of the Horny Sponges. London, 936 pp., 50 pis.

D VI 19

382 DISCOVERY REPORTS

Lendenfeld, R. von, 1894. Die TetractineUiden der Adria {init einem Anhange iiber die Lithistiden) .
Denkschr. Mathem.-naturw. CI. K. Akad. Wiss. Wien, LXI, pp. 91-204, pis. i-viii.

1903- Tetraxonia. Das Tierreich. XIX. Porifera. Berlin. 8vo, 168 pp.

1906. Die Tetraxonida. Wiss. Ergebn. Deutsch. ' Valdivia' Tiefsee-Exped., xi, 2, Jena, pp. 59-373,

pis. ix-xlvi.

I907- Tetraxonia der deutschen Siidpolar -Expedition 1901-3. Deutsch. Siidpol. Exped. Berlin, ix,

Zoologie, I, pp. 303-42, pis. xxi-xxv.

LuNDBECK, W., 1905. Pmifera. Part II. Desmacidonidae (pars). Danish Ingolf Exped., Copenhagen,
VI, 2, 219 pp., pis. i-xx.

Montagu, G., 1818. An Essay on Sponges, etc. Mem. Werner. Soc, 11, pp. 67-122.

Parfitt, E., 1868. On the Marine and Freshwater Sponges of Devonshire . Trans. Devonshire Assoc, pp. 443-
62.

PoL^jAEFF, N., 1883. Report on the Cakarea dredged by H.M.S. 'Challenger'. Rep. Sci. Res. Voy. 'Chal-
lenger', Zool., viii, pp. 1-76, 9 pis.

Ridley, S. O., 1881. Spongida {Account of t/ie Zoological Collections made during the survey of H.M.S.
'Alert'). Proc. Zool. Soc. London, pp. 107-37, 140^1- 2 pis.

1883. Notes on Zoophytes and Sponges obtained by Mr F. Day, off the east coast of Scotland. Joiirn.

Linn. Soc. London, xvii, pp. 106-8.
Ridley, S. O., and Dendy, A., 1886. Preliminary Report on the Monaxonida collected by H .M .S . 'Challenger'.
Ann. Mag. Nat. Hist., xviii, pp. 325-51, 470-93.

1887. Monaxofiida. Rep. Sci. Res. Voy. H.M.S. ' Challenger', Zool., xx, 275 pp., pis. i-li.

Schmidt, O., 1864. Supple?nent der Spongien des adriatischen Meeres. Leipzig, 4to, 48 pp., 4 pis.

1870. Grundziigc einer Spongien-Fauna des atlantischen Gebietes. Leipzig, Fol., 88 pp., 6 pis.

ScHULZE, F. E., 1877. Untersuchungen iiber den Bau unddie Entwicklung der Spongien. Die Gattung Halisarca.

Zeitschr. wiss. Zool., xxix, pp. 87-122, 2 pis.

1879- Ufttersuchungen iiber den Bau mid die Entivicklimg der Spongien. Zeitschr. wiss. Zool., xxxii,

PP- 593-660, 5 pis.; xxxiii, pp. 1-38, 4 pis.

1880. Untersuchungen iiber deti Bau iind die Entwicklung der Spongien. IX. Die Plakiniden. Zeitschr.

wiss. Zool., XXXIV, pp. 407-51, 3 pis.
1887. Hexactinellida, 1876. Rep. Sci. Res. Voy. H.M.S. 'Challenger', Zool., xxi, 53, 514 pp., pis. i-

civ, and a map.
SoLLAS, W. J., 1886. Preliminary account of the Tetractinellid Sponges dredged by H.M.S. 'Challenger',

1872-6. Part I. The Clioristida. Proc. Dublin Soc, v, pp. 177-99.

1888. Tetractinellida. Rep. Sci. Res. Voy. H.M.S. 'Challenger', Zool., xxv, pp. 1-458, pis. i-xliv.

Stephens, J., 1912. Clare Island Survey. Part 59. Marine Porifera. Proc. Roy. Irish Acad., xxxi, 42 pp.

1915- Atlantic Sponges collected by the Scottish National Antarctic Expedition. Trans. Roy. Soc.

Edinburgh, l, pp. 423-67, pis. xxxviii-xl.

1921. Sponges of tlie Coasts of Ireland. II. The Tetraxonida (concluded). Fisheries, Ireland, Sci.

Invest., London, 11, 75 pp., 6 pis.
SwARTSCHEWSKY, B., 1906. Beitrage zur Spongien-Fauna des Weissen Meeres. Mem. Soc. Nat. Kiew, xx,

PP- 307-71 > pis. x-.xvi.
Thiele, J., 1898. Studien iiber pazifische Spongien. Zoologica, xxiv, pp. 1-72, 8 pis.

1905- Die Kiesel- und Hornschwdmme der Sammlung Plate. Zool. Jahrb., Jena, Suppl. VI (Fauna

Chilensis), pp. 407-96, pis. xxvii-xxxiii.
Thompson, W., 1840. Additions to the Fauna of Ireland. Ann. Mag. Nat. Hist., v, pp. 245-57.
Topsent, E., 1888. Contributions a I'etude des Clionides. Arch. Zool. Exper. v, suppl., pp. 1-165, pis. i-vii.

1889. Quelques Spongiaires du Banc de Campeche et de la Point e-a-Pitre. Mem. Soc. Zool. Fr., II,

PP- 30-52-
■ 1890. Sur la distribution geographique de quelques Microsclerophora. Bull. Soc. Zool. Fr., xv, pp. 231-3.

1891. Essai sur la Fauna des Spongiaires de Roscoff. Arch. Zool. Exper., ix, pp. 523-54, i pi.

1892. Contribution a I'etude des Spongiaires de I'Atlantique Nord. Res. Camp. Sci. Albert de

Monaco, 11, 165 pp., 11 pis.

1892. Sponges de la Mer Rouge. Mem. Soc. Zool. Fr., v, pp. 21-9, i pi.

LIST OF LITERATURE 383

ToPSENT, E., 1896. litude monographique des Spongiaires de France. II. Caniosa. Arch. Zool. Exper., iii,
pp. 493-590, pis. xxi-xxiii.

1901. Spongiaires. Exped. antarct. Beige. Res. Voy. S.Y. 'Belgica', 1897-1899, 54 pp., pis. i-vi.

1901. Notice preliminaire sur les eponges recueillies par I'expedition antarctique beige. Arch. Zool.

Exper. IX, Notes, pp. 5-16 (sep. copy pp. i-ii).

1904- Spongiaires dcs Afores. Res. Camp. Sci. Albert de Monaco, xxv, 280 pp., 18 pis.

1907- Poccilosclerides nouvelles recueillies par le ' Franfais' dans V Antarctique . Bull. Mus. Hist, nat.,

Paris, pp. 69-76.

1908. Spongiaires. Expedition antarctique fran9aise (1903-5) (Expedition Charcot). Paris, 4to,

37 PP-. pis- i-v.

1912. Sur une grande Tedania abyssale des Azores (Tedania phacellina, n.sp.). Bull. Inst. Oceanogr.

Monaco, 252, pp. 1-5, 2 figs.

1913- Spongiaires provenant des campagnes scientifiques de la ' Princesse Alice' dans les mers du Nord.

Res. Camp. Sci. Albert de Monaco, XLV, pp. 1-67, pis. i-v.

1913- Spongiaires de I'expedition antarctique nationale ecossaise. Trans. Roy. Soc. Edinburgh, XLix, 3,

9. PP- 579-643- pis- i-vi-

1915- Spongiaires recueillies par la 'Scotia' dans l' Antarctique (1903-4). Supplement. Trans. Roy.

Soc. Edinburgh, Li, pp. 35-43.

1916. Diagnoses d'eponges recueillies dans V Antarctique par la ' Pourquoi Pas?' Bull. Mus. Paris, pp.

163-72.

1917- Spongiaires. Deuxieme expedition antarctique fran9aise (1908-10) commandee par le Dr Jean

Charcot. Sciences Physiques: Documents Scientifiques. Paris, 4to, 88 pp., pis. i-vi.

1925. l£tude de Spongiaires du Golfe de Naples. Arch. zool. Paris, i.xiii, pp. 623-725, pi. viii.

Verrill, a. E., 1874. Brief Contributions to Zoology from the Museum of Yale College. . . .Results of recent

Dredging Expeditions on the Coast of New England. Amer. Journ. Sci., vii, pp. 38-46, 131-38,
405-14, 498-505, pis. iv-viii.
VosMAER, G. C. J., 1880. Sponges of the Leyden Museum. I. The family of the Desmacidinae . Notes from
the Leyden Museum, 11, pp. 99-164.

1882. Report on the Sponges dredged up in the Arctic Sea by the ' William Barents.' Niederl.

Arch. Zool., Suppl. I, pp. 1-58, pis. i-iv.

1885. The Sponges of the ' William Barents' Expedition. Bijdr. Dierk. Amsterdam, xii, pp. 1-47.

1885. Porifera. Bronn, H. G., Die Klassen und Ordnungen des Thierreichs, 11, pp. 177-368, pis. xix-

XX V.

191 1. The Porifera of the Siboga-Expedition . II. The genus Spirastrella. Siboga-Expeditie, Leiden,

VI A, pp. 1-69, 14 pis.

1932. Porifera Incalcaria of the Bay of Naples. Capita Zoologica (in press).

Whiteaves, J. F., 1874. Report of deep-sea dredging operations in the Gulf of St Lawrence, p. 9 {fide
Lambe, 1896, supra).

ig-2

INDEX

[Synonyms are indicated by italics.]

Acanthorhabdus, 294
Acanthorhabdus fragilis, 294
acerata, Mycale, 289
acerata, var. minor, Mycale, 289
Acervochalina variabilis, 265
Acheliderma, 317
Acheliderma topsenti, 318
actiniiformis , Tedania, 305
Adocia, 274
Adocia carduus, 274, 275

cucurbitiformis, 276

glacialis, 274, 275

siphonella, 275

tenellus, 276

tremulus, 276

aethiopicus, Cliona, 340
alderi, Isodictya, 289, 290
Alebion, 295

Alebion proximum , 296-301
amabilis, Stylotellopsis, 326
americanus, Amphilectus, 292
Amorphina hirta, 331
Amphilectus, 289
Amphilectus americanus, 292

clarkei, 290

compressus, 290

dubius, 290

flabellata, 292

fucorum, 289-90

gracilis, 290

imitatus, 290

normani, 290

rugosus, 292

rugosus, var. major, 292

Anchinoe, 315
Anchinoe areolata, 315

latrunculioides, 315

leptochela, 315

nobilis, 313

ioxiferum, var. antarctica, 325

antarctica, Cinachyra, 264

antarctica, Ectyodoryx, 314

antarctica, Isodictya, 285

antarctica, Leucetta, 259

antarctica, Lissodendoryx, 314

antarctica, Mycale, 288

antarctica, Rossella, 254, 255

antarctica, var. intermedia, n., Rossella, 254, 255

antarcticus, Sphaerotylus, 339

apollinis, Artemisina, 323

arborescens, Plicatellopsis, 332-3

arcuarius. Calyx, 277

arcuarius, Gellius, 277

Arenochalina, 293

Arenochalina incrustans, 293

areolata, Anchinoe, 315
areolata, Hymedesmia, 315
Artemisina, 323
Artemisina apollinis, 323

plumosa, 323

plumosa, var. lipochela, 323

strongyla, 323

Asbestopluma calyx, 293
asigmata, Myxilla, 310
australiensis, Rhizaxinella, 331
australis, Myxilla, 310
Axinella, 320
Axinella crinita, 330
Axinosia incrustans, 290
Axociella, 324
Axociella flabellata, 325
nidificata, 324

basimucronata, Myxilla, 311
belli, Inflatella, 318
benedeni, wir. fortior, Gelliodes, 271
benedeni, Gelliodes, 271
benedeni, Gellius, 271
benedeni, Microxina, z-ji
barbata, Cinachyra, 265
Biemna, 293
Biemna chilensis, 293

macrorhaphis, 293

bilamellata, Haliclona, 267

borealis, subsp. acuata, Stylocordyla, 339

Bubaris, 333

Bubaris murrayi, 335

vermiculata, 333

Burtonella, 295

Burtonella melanokhemia, 295, 296

cactoides, Isodictya, 286
Callyspongia, 279
Callyspongia flabellata, 282

fortis, 279

fusifera, 281

Calyx, 277

Calyx arcuarius, 277

kerguelensis, 278

stipitatus, 277

calyx, Asbestopluma, 293

callosus, Hymeniacidon, 290

caminatiis, var. papillatus, Suherites, 336

carduus, Adocia, 274, 275

carduus, Gellius, 274, 275

carduus, var. magellanica, Gellius, 274, 275

caruncula, Hymeniacidon, 328

Ceraochalina, 279

ceratosa, Desmacidon, 326

Cercidochela, 287

386

Cercidochela lankesteri, 287
centrotyla, Hymeniacidon, 330
Ceratopsis incrustans, 335
Chalina fusifera, 281

peniciUala, 266

charcoti, Microxina, 271
charcoti, Tedania, 303, 307
chelifer, lophon, 296-301
chelifer ostia-magna, lophon, 296-301
clieliferum, Spanioplon, 317
chilensis, Biemna, 293
chilensis, Cliona, 340
chilensis, Haliclona, 265
chilensis, Myxilla, 311
chilensis, Reniera, 265
chilensis, Spongeliii, 341
Cinachyra antarctica, 264

barbata, 265

coactifera, 265

cirrata, var. aurorae, Grantia, 262
cirrata, var. tenuipilosa, Grantia, 262
clarkei, Amphileclus, 290
clarkei, Isodiclya, 289, 290
Clathria, 319
Clathria lipochela, 319

mortenseni, 324

papillosa, 319

ierrae-novae, 324

toxipraedita, 319, 324

Cliona aethiopicus, 340

chilensis, 340

coactifera, Cinachyra, 265
coactifera, Tethya, 265
compressa, Poecillastra, 263
compressus, Amphileclus, 290
conica, Haliclona, 266
conica, Pachychalina, 266
coulmani, Kirkpatrickia, 317
coulmani, Tedania, 317
crassa, Crella, 327
crassa, Gray el la, 327
crasso-fibrosa, Esperiopsis, 290
Crella, 327
Crella crassa, 327
Crellina, 327
Crellina tubifex, 327
crinita, Axinella, 330
cucurbitiformis, Adocia, 276
cucurbitiformis, Gellius, 276
cunninghami, Esperia, 288
Cydonium magellani, 263

dancoi, Haliclona, 267, 268, 269
Dasychalina, 278
Dasychalina validissima, 278
delicata, Desmacidon, 285
delicata, Isodictya, 285
Dendoryx deniala, 316

incrustans, var. australis, 310

ramilobosa, 314

INDEX

Dendrilla, 342
Dendrilla membranosa, 342
dentata, Dendoryx, 316
Desmacidon ceratosa, 326

delicata, 285

doryphora, 286

kerguelensis , 283

kerguelensis, var. antarctica, 285

kerguelensis, var. cactoides, 286

maeandrina, 287

setifer, 284

spinigera, 284

Dictyociona, 324
Dictyociona discreta, 324
terrae-novae, 324

discophorus, Erylus, 264

discovereyi, Leucosolenia, 258

discovereyi, Stelodoryx, 316

discreta, Dictyociona, 324

discreta, Microciona, 324

diversiancorata, Myxilla, 316

doryphora, Desmacidon, 286

dubia, Hymeniacidon, 329

dubia, Isodictya, 289, 291

dubius, Amphilectus, 290

dubiiis, lophon, 348

dujardini, var. magellanica, Halisarca, 340

Duseideia, 341

Duseideia fragilis, 341

tenuifibra, 342

Ectyodoryx, 313
Ectyodoryx antarctica, 314

frondosa, var. anacantha, 314

nobilis, 313

paupertas, subsp. nobile, 313

paupertas, subsp. typica, 313

ramilobosa, 314

edwardii, var. aniericana, Esperiopsis, 292

edwardii, Esperiopsis, 290

edwardii, Isodictya, 289, 290

egregia, Phakellia, 330

egregia, Pseudaxinella, 330

elongata, Myxilla, 311

embryology in:

Amphilectus fucorum, 365

Anchinoe latrunculioides, 365

leptochela, 365

Asbestopluma calyx, 365

Callyspongia flabellata, 364

Dictyociona discreta, 367

Eurypon miniaceum, 367

Guitarra finibriata, 365

Haliclona bilamellata, 364

Hymedesmia irritans, 367

Hymeniacidon dubia, 367

sanguinea, 367

lophon radiatus, 364

Isodictya antarctica, 365

setifer, 364

INDEX

387

embryology in:

Kirkpatrickia variolosa, 365

Mycale magellanica, 365

Plakina monolopha, 364

trilopha, 364

Plocamia gaussiana, 365

Polymastia isidis, 368

Sphaerotylus antarcticus, 368

Spongia magellanica, 368

Stylocordyla borealis, subsp. acuata, 367
erinacea, Homoeodictya, 286
erinacea, Isodictya, 286
Erylus, 364

Erylus discophorus, 264
Esperella lapidiformis, 289

magellanica, 288

Esperia cuiminghami, 288

magellanica, 288

Esperiopsis crasso-fibrosa, 290

edwardii, 290

edwardii, var. americana, 292

fucorum, 290

informis, 290

normani, 290

rugosa, 292

rugosa, var. major, 292

Eumastia, 335

Eumastia sitiens, 335

eumitum, Phloeodictyon, 283

Eurypon, 325

Eurypon miniaceum, 325

fernandezi, Hymeniacidon, 328-9
fibidata, Reniera, 276

fibulatus, Gellius, 276

fimbriata, Guitarra, 287
fimbriatus, Gellius, 272

flabellata, Amphilectus, 292

flabellata, Axociella, 325

flabellata, Callyspongia, 282

flabellata, Ophlitaspongia, 325

flabellata, Plicatellopsis, 332-3

fortis, Callyspongia, 279
fortis, Siphonochalina, 279

flaccida, Haliclona, 269

fragilis, Acanthorhabdus, 294

fragilis, Duseideia, 341
frigidus, lophon, 348

frondosa, var. anacantha, Ectyodoryx, 314

fucorum, Amphilectus, 289, 290
fucorum, Esperiopsis, 290
fucorum, Halichondria, 289
fucorum, Isodictya, 289, 290
fucorum, Reniera, 290
fucorum, Spongia, 289
fusifera, Callyspongia, 281
fusifera, Chalina, 281

gastrorhabdifera, Leucaltis, 259
gaussiana, Haliclona, 266
gaussiana, Plocamia, 318

gaussiana, Siphonochalina, 266
Gelliodes bencdcni, 271

benedeni, vsx . fortior , 271

kerguelensis, 278

Gellius arcuarius, 277

benedeni, 271

carduus, 274, 275

carduus, var. magellanica, 274, 275

cucurbitiformis, 276

fibulatus, 276

fimbriatus, 272

glacialis, 274, 275

glacialis, var. nivea, 274, 275

imperial is, 267

laevis, 274, 275

pilosus, 272

rudis, 272

tenellus, 272, 276

tremulus, 276

tubuloramosus, 266

genotype of:

Allocia, 317

Acheliderma, 318

Arenochalina, 293

Burtonella, 295

Dasychalina, 278

Hemigellius, 272

Hymeniacidon, 328

lophon, 295

lophonopsis, 295

Plicatella, 332

Plicatellopsis, 332

Pocillon, 295

Sigmotylotella, 295

Stelodoryx, 316
Geodia, 263
Geodia magellani, 263
glacialis, Adocia, 274, 275
glacialis, Gellius, 274, 275
glacialis, var. nivea, Gellius, 274, 275
gracilis, Amphilectus, 290
gracilis, Isodictya, 289, 290, 291
Grantia, 262
Grantia cirrata, var. aurorae, 262

cirrata, var. tenuipilosa, 262

Grayella, 327
Grayella crassa, 327
Guitarra, 287
Guitarra fimbriata, 287
Gymnorossella, 257
Gymnorossella inermis, 257

haeckeliana, Leucetta, 261
haeckeliana, Leucettusa, 261
Halichondria fucorum, 289

hyndmaui, 295, 296

magellanica, 278

nigricans, 295, 296

parasitica, 289

scandens, 295, 296

388

INDEX

Haliclona, 267
Haliclona bilamellata, 267

chilensis, 265

conica, 266

dancoi, 267, 268, 269

flaccida, 269

gaussiana, 266

nodosa, 265

penicillata, 266

petrosioides, 269

Stephens!, 269

■ tubuloramosa, 266

variabilis, 265

Haliclonissa, 270
Haliclonissa sacciformis, 271

verrucosa, 271

Halisarca, 340

Halisarca dujardini, var. magellanica, 340

Halispongia parasitica, 289

Hemigellius, 272

Hemigellius pachyderma, sp.n., 273

rudis, 272

hirta, Amorpliiiia, 331

hirta, Stylohalina, 331

hispida, Isodictya, 290, 291

Homaxinella, 330

Homaxinella supratumescens, 330

Homoeodictya erinacea, 286

kerguelensis, 283

kerguelensis, var. similUma, 284

kirkpatricki, 286

microcheJa, 286

obliqtiidens , 284

setifer, 284

trigona, 284

verrucosa, 284

hyalinus, Pseudosuberites, 336
Hymedesmia, 315
Hymedesmia areolata, 315

irritans, 326

laevis, 326

leptochela, 3 1 5

longurioides, 327

similliina, var. antarctica, 327

Hymeniacidon, 328
Hymeniacidon callosus, 290

caruncula, 328

centrotyla, 330

dubia, 329

fernandezi, 328, 329

kerguelensis, 329

sanguinea, 328

torquata, 328, 329

hyndmani, Halichondria, 295, 296

ibla, Thecophora, 337
imitata, Isodictya, 290, 291
imitatus, Amphilectus, 290
imperialis, Gellius, 267
incrustans, Arenochalina, 293

incrustans, var. australis, Dendoryx, 310

incrustans, Axinosia, 290

incrustans, Ceratopsis, 335

incrustans, Suberites, 336

inermis, Gymnorossella, 257

Inflatella, 318

Inflatella belli, 318

infonnis, Esperiopsis , 290

invaginata, var. gaussi. Polymastia, 338

invaginata, Polymastia, 338

invalida, Isodictya, 290

involuta, Isodictya, 290, 291

lophon, 295

lophon, evolution of spp., 348

lophon chelifer, 296-301

chelifer ostia-magna, 296-301

dubius, 348

frigidus, 348

lamella, 296-301

lamella indivisus, 296-301

major, 348

major, var. tenuis, 348

minor, 348

pattersoni, 296-301

piceus, 348

proximum, 296-301

proximus reticularis, 296-301

radiatus, 296

lophotwpsis, 295

irregularis, Raspailia, 325

irritans, Hymedesmia, 326

isidis, Polymastia, 337

isidis, var. simplex, Polymastia, 337

Isodictya, 283

Isodictya alderi, 289, 290

antarctica, 285

cactoides, 286

clarkei, 289, 290

delicata, 285

dubia, 289, 291

edwardii, 289, 290

erinacea, 286

fucorum, 289, 290

gracilis, 289, 290, 291

hispida, 290, 291

imitata, 290, 291

invalida, 290

involuta, 290, 291

kerguelensis, 283

kerguelensis, var. simillima, 284

microchela, 286

normani, 289

paupera, 289, 290, 291

pertemds, 290

scitula, 290, 291

setifer, 284

spinigera, 284

toxophila, 287

uniformis, 289, 290

verrucosa, 284

INDEX

389

kerguelensis, Calyx, 278

kerguelensis, Desmacidon, 283

kerguelensis, Gelliodes, 278

kerguelensis, Homoeodictya, 283

kerguelensis, Hymeniacidon, 329

kerguelensis, Isodictya, 283

kerguelensis, var. antarctica, Desmacidon, 285

kerguelensis, var. cactoides, Desmacidon, 286

kerguelensis, var. simillima, Homoeodictya, 284

kerguelensis, var. simillima, Isodictya, 284

kirkpatricki, Homoeodictya, 286

Kirkpatrickia, 317

Kirkpatrickia coulmani, 317

variolosa, 317

labyrinthica, Reniera, 332

laevis, Gellius, 274, 275

laevis, Hymedesinia, 326

lamella, lophon, 296-301

lamella indivisus, lophon, 296-301

lankesteri, Cercidochela, 287

lapidiformis, Esperella, 289

lapidiformis, Mycale, 289

Latrunculia, 340

Latrunculia lendenfeldi, 340

latrunculioides, Anchinoe, 315

lendenfeldi, Latrunculia, 340

leptochela, Anchinoe, 315

leptochela, Hymedesmia, 315

leptoderma, Tetilla, 264

leptoraphis, Leucetta, 259

Leucaltis, 259

Leucaltis gastrorhabdifera, 259

Leucandra primigenia, var. leptoraphis, 259

Leucetta, 259

Leucetta antarctica, 259

haeckeliana, 261

leptoraphis, 259

macquariensis, 259

Leucettusa, 261
Leucettusa haeckeliana, 261

simplicissima, 261

Leucosolenia, 258
Leucosolenia discovereyi, 258

macleayi, 258

lillei, Mycale, 288
lipochela, Clathria, 319
Lissodetidoryx antarctica, 314

spongiosa, var. asigmata, 311

longurioides, Hymedesmia, 327

macleayi, Leucosolenia, 258
macquariensis, Leucetta, 259
macrorhaphis, Biemna, 293
maeandrina, Desmacidon, 287
maeandrina, Plumocolumella, 287
magellani, Cydonium, 263
magellani, Geodia, 263
magellanica, Esperella, 288

magellanica, Esperia, 288

magellanica, Halichondria, 278

magellanica, Mycale, 288

magellanica, Spongia, 341

magna, Myxilla, 309

major, lophon, 348

major, var. tenuis, lophon, 348

massa, Tedania, 303-6

melanokhemia, Burtonella, 295, 296

membranosa, Dendrilla, 342

membranacea, Ophlitaspongia, 321

Microciona discreta, 324

microstomus, var. stellatus, Suberites, 336

Microxina, 271

Microxina benedeni, 271

charcoti, 271

microchela, Homoeodictya, 286
microchela, Isodictya, 286
minor, lophon, 348
miniaceum, Eurypon, 325
mollis, Myxilla, 309
monilifera, Pachastrella, 263
monolopha, Plakina, 262
fnontiniger, Pseudosuberites, 335
montiniger, Suberites, 335
mortenseni, Clathria, 324
murdochi, Tedania, 308
murrayi, Bubaris, 335
Mycale, 288
Mycale acerata, 289

acerata, var. minor, 289

antarctica, 288

lapidiformis, 289

lillei, 288

magellanica, 288

pellita, 288

rossi, 288

tridens, 289

Myxilla, 309
Myxilla asigmata, 310

australis, 310

basimucronata, 311

chilensis, 311

diversiancorata, 316

elongata, 311

magna, 309

mollis, 309

nobilis, 313

nobilis, var. bacilli f era, 313

nobilis, var. patagonica, 313

pluridentata, 316

spongiosa, 309

spongiosa, var. asigmata, 310

verrucosa, 312

myxillioides, Plumocolumella, 288

nidificata, Axociella, 324
nidificata, Ophlitaspongia, 324
nigricans, Halichondria, 295, 296
nobile, var. patagonicum, Stylostichon, 313

39°

nobilis, Anchime, 313

nobilis, var. hacillifera, Myxilla, 313

nobilis, Ectyodoryx, 313

nobilis, Myxilla, 313

nobilis, var. patagonica, Myxilla, 313

nodosa, Haliclona, 265

nodosa, Rcniera, 265

normani, Amphilectiis, 290

normaui, Espeiiopsis, 290

normatii, Isodictya, 289

nuda, Rossella, 255

obliquidens, Homoeodictya, 284
officinalis, Spongia, 340
Ophlitaspongia, 321
Ophlitaspongiaflabellata, 325

membranacea, 321

nidificata, 324

thielei, 322

oxeata, Tedania, 309

Pachastrella, 263
Pachastrella monilifera, 263
Pachychalina conica, 266

validissima, 278

pachyderina, Hemigellius, 273
papillatiim, Tentorium, 336
papillatus, Suberites, 336
papillosa, Clathria, 319
parasitica, Halichondria, 289
parasitica, Halispongia, 289
parasitica, Spongia, 289

Paratedania, 345

Paratedania tarantula, 303-6
pattersoni, lopbon, 296-301

paucispiculus, Rhaphidophlus, 320

paupera, Isodictya, 289, 290, 291

paupera, Reniera, 290

paupertas, subsp. nobile, Ectyodoryx, 313

paupertas, subsp. typica, Ectyodoryx, 313

pellita, Mycale, 288

penicillata, Chalina, 266

penicillata, Haliclona, 266

penicillata, Reniera, 266

Pericharax, 258

Pericharax pyriformis, 258

pertenuis, Isodictya, 290

petrosioides, Haliclona, 269

Phakellia egregia, 330

phakellina, Raspaxilla, 326

Phloeodictyon, 283

Phloeodictyon eumitum, 283

piceus, lophon, 348

Plakina, 262

Plakina monolopha, 262

trilopha, 262

Plicatellopsis, 332
Plicatellopsis arborescens, 332-3

flabellata, 332-3

pilosus, Gellius, 272

INDEX

Plocamia, 318
Plocamia gaussiana, 318
Plumocolumella, 287
Plumocolumella maeandrina, 287

mjrxillioides, 288

plumosa, Artemisina, 323

plumosa, var. lipochela, Artemisina, 323

plurideiitata, Myxilla, 316

pluridentata, Stelodoryx, 316

Pocillon, 295

Poecillastra, 263

Poecillastra compressa, 263

Polymastia, 337

Polymastia invaginata, 338

invaginata, var. gaussi, 338

isidis, 337

isidis, var. simplex, 337

primigenia, var. leptoraphis, Leucandra, 259
Protoclathria, 320
Protoclathria simplicissima, 321
proximum, Alebion, 296-301
proximum, lophon, 296-301
proximus reticularis, lophon, 296-301
Pseudanchinoe toxifera(um), 325
Pseudaxinella egregia, 330
Pseudosuberites, 335
Pseudosuberites hyalinus, 336

montiniger, 335

sulcatus, 336

purpurea, Spirastrella, 339
pyriformis, Pericharax, 258

racovitzae, Rossella, 256
radiatus, lophon, 296
ramilobosa, Dendoryx, 314
ramilobosa, Ectyodoryx, 314
Raspailia irregularis, 325
Raspaxilla, 326
Raspaxilla phakellina, 326
Reniera chiknsis, 265

fibulata, 276

fucorum, 290

labyrinthica , 332

nodosa, 265

paupera, 290

penicillata, 266

scotti, 267

siphonella, 275

uniformis, 290

Rhaphidophlus paucispiculus, 320
Rhaphoxya, 345

Rhizaxinella, 331

Rhizaxinella australiensis, 331

Rossella, 254

Rossella antarctica, var. intermedia, n., 254, 255

antarctica solida, 254, 255

nuda, 255

racovitzae, 256

villosa, 257

rossi, Mycale,\2S8

INDEX

391

rubiginosa, Thieleia, 329

rudis, Gellius, 272

rudis, Hemigellius, 272

rugosa, Esperiupsis, 292

rugosa, var. major, Esperiopsis, 292

rugosus, Amphilectus, 292

rugosus, var. major, Amphilectus, 292

sacciformis, Haliclonissa, 271

sanguinea, Hymeniacidon, 328

scandens, Halichondria, 295, 296

scitula, Isodictya, 290, 291

scotti, Reniera, 267

semisuberites, Tentorium, 337

semisuberites, Thecophora, 337

setifer, Desmacidon, 284

setifer, Homoeodictya, 284

setifer, Isodictya, 284

Sigmotylotella, 295

Sigmotylotella suberitoides, 295

simillima, var. antarctica, Hymedesmia, 327

simplicissima, Leucettusa, 261

simplicissima, Protoclathria, 321

siphonella, Adocia, 275

siphonella, Reniera, 275

Siphonochalina fortis, z-jq

gaussiana, 266

sitiens, Eumastia, 335
Spanioplon cheliferum, 317
Sphaerotylus antarcticus, 339
spinata, Tedania, 306
spinata, Trachytedania, 306
spinigera, Desmacidon, 284
spinigera, Isodictya, 284
Spirastrella, 339
Spirastrella purpurea, 339
Spongelia chilensis, 341
Spongia, 340
Spongia fucorum, 289

magellanica, 341

officinalis, 340

parasitica, 289

spongiosa, M3rxilla, 309

spongiosa, var. asigmata, Lissodendorvx, 311

spongiosa, var. asigmata, Myxilla, 310

Stelodoryx, 316

Stelodoryx discovereyi, 316

pluridentata, 316

stephensi, Haliclona, 269

stipitatus. Calyx, 277

strongyla, Artemisina, 323

Stylinos uniformis, 290

Stylocordyla, 339

Stylocordyla borealis, subsp. acuata, 339

Stylohalina, 331

Stylohalina hirta, 331

Stylostichon nobile, \nr. patagonicum, 313

— toxifenim, 325

Stylotellopsis, 326
Stylotellopsis amabilis, 326

Suberella topsenti, 335

Suberites, 335

Stiberites caminatus, var. papillatus, 336

incrustans, 336

tnicrostomus, var. stellatus, 336

montiniger, 335

papillatus, 336

suberitoides, Sigmotylotella, 295
sulcatus, Pseudosuberites, 336
supratumescens, Homaxinella, 330

tarantula, Paratedania, 303-6

Tedania, 345

Tedania, aifinities of Antarctic spp., 342

embryology of Antarctic spp. 360-4

revision of spp., 345-6

Tedania actiniiformis, 305

charcoti, 303, 307

coulmani, 317

massa, 303-6

murdochi, 308

o.xeata, 309

spinata, 306

tenuicapitata, 302

vanhojfeni, 305

variolosa, 317

Tedaniella, 345
Tedanione, 345
Tedaniopsis, 304, 345
tenellus, Adocia, 276
tenellus, Gellius, 272, 276
Tentorium, 336
Tentorium papillatum, 336

semisuberites, 337

tenuicapitata, Tedania, 302
tenuifibra, Duseideia, 342
terrae-novae, Clathria, 324
terrae-novae, Dictyociona, 324
Tethya coactifera, 265
Tetilla, 264

Tetilla leptoderma, 264
Thecophora ibla, 337

semisuberites, 337

thielei, Ophlitaspongia, 322

Thieleia, 329

Thieleia rubiginosa, 329

topsenti, Acheliderma, 318

topsenti, Suberella, 335

torquata, Hymeniacidon, 328, 329

toxifera{um), Pseudanchinoe , 325

toxiferum, var. antarctica, Anchinoe, 325

toxiferum, Stylostichon, 325

toxipraedita, Clathria, 319-24

toxophila, Isodictya, 287

Trachytedania, 345

Trachytedania spinata, 306

tremulus, Adocia, 276

tremulus, Gellius, 276

tridens, A-Iycale, 289

trigona, Homoeodictya, 284

392

trilopha, Plakina, 262
tubifex, Crellina, 327
tuhdoramosus, Gellius, 266
tubuloramosus, Haliclona, 266

uniformis, Isodictya, 289, 290
uniformis, Reniera, 290
uniformis, Stylinos, 290

validissima, Dasychalina, 278
validissima, Pachychalina, 278

INDEX

vanhqffeni, Tedania, 305
variabilis, Acervochalina, 265
variabilis, Haliclona, 265
variolosa, Kirkpatrickia, 317
variolosa, Tedania, 317
vermiculata, Bubaris, 333
verrucosa, Haliclonissa, 271
verrucosa, Homoeodictya, 284
verrucosa, Isodictya, 284
verrucosa, Myxilla, 312
villosa, Rossella, 257

PLATE XLVIII

Fig. I. Pericharax pyriformis,sp.n. Holotype. Nat. size.

Fig. 2. Perkharax pyr!fonms,sp.n. Co-type. Nat. size.

Fig. 3. Leucettusa simplicissima, sp.n. Holotype. Nat. size.

Fig. 4. Hemigellhis pachyderma, sp.n. Holotype. x |.

Fig. 5. Haliclona bUamellata, sp.n. Holotype: showing outer surface.

X 3. _

Fig. 6. Haliclona hilamellata, sp.n. Holotype: showing inner surface.

Fig. 7 . Haliclona bilamellata, sp .n . Portion of another specimen showing

outer surface, x f .
Fig. 8. Haliclona bilamellata, sp.n. The same, from the inner surface.

X 5.

Fig. 9. Haliclona bilamellata, sp.n. A tubular specimen, x f.
Fig. 10. Haliclonissa sacciformis, sp.n. Holotype. x J.

DISCOVERY REPORTS, VOL. VI.

PLATE XL VIII.

!>>^iJ

lO.

■ ■iV'fc*:::

;in.;^M

Jtilir. BMt ^ons 4 Diuuf liian. L** LcaOa.

ANTARCTIC SPONGES.

XLIX.

PLATE XLIX

Fig. I. Halidona bilamellata,sp.n. A tubular specimen, x f .

Fig. 2. Halidona bilamellata, sp.n. A flabellate specimen showing

oscular surface, x J.
Fig. 3. Halidona bilamellata, sp.n. The same, showing poral surface.

X

A 3.

Fig. 4. Callyspongtaflabellata, sp.n. Holotype. x f .

DISCOVERY REPORTS, VOL. VI.

PLATE XLIX.

■^•y>G£c:4^u^A^

JohnbiJESofi: 4 Ci-nj^liiin U' Lor

ANTARCTIC SPONGES

TE L.

f

PLATE L

Fig. I. Mia-oxina bejiedent (Topseni). x J.

Fig. 2. Haliclona bilamellata, sp.n. A macerated specimen, showing the

fibrous appearance imparted to the skeleton when the spicules

are large, x f .
Fig. 3. Dasychalina validissima (Thiele). Portion of a sub-flabellate

specimen, x f .
Fig. 4. Dasychalina validissima (Thiele). A massive specimen, x §.
Fig. 5. Dasychalina validissima (Thiele). A massive specimen, x |.
Fig. 6. Dasychalina validissima (Thiele). A flabellate specimen, x |.
Fig. 7. Dasychalina validissima (Thiele). A flabellate specimen, x |.

DISCOVERY REPORTS. VOL. V]

PLATE L.

^^:m

John Bak Sons 4. Danialiwn L'^ London

ANTARCTIC SPONGES

E L]

6.

>nBjtJi:Sana& UnniolsiimL" LktwIdd

PLATE LI

Fig. I. Haliclona variabilis (Thiele). Slightly less than natural size.

Fig. 2. Haliclona variabilis (Thiele). Natural size.

Fig. 3. Haliclonissa verrucosa, sp.n. Holotype. x f .

Fig. 4. Isodictya antarctica (Kirkpatrick). x \.

Fig. 5. Isodictya microchela (Topsent). x f .

Fig. 6. Mycale macrochela, sp.n. Holotype. x f .

DISCOVERY REPORTS, VOL. VI

PLATE LI.

^ . ^ -^ •'♦

■'.,>-.

2.

John ilojc Sons & Diuueluon. L*^ Lnndan

ANTARCTIC SPONGES

^TE L

PLATE LII

Fig. I. Callyspongia fusif era (Thiele). x |.
Fig. 2. Isodictya toxophila, sp.n. x §.
Fig. 3. Isodictya toxophila, sp.n. x |.

DISCOVERY REPORTS, VOL. VI.

PLATE LII.

John B«lc Sons 4 t«JUoLi»oi\ l-*^ London

ANTARCTIC SPONGES

Lm.

PLATE LIII

Fig. I. Isodictya toxophila, sp.n. x \.

Fig. 2. Isodictya toxophila, sp.n. x \.

Fig. 3. Plumocolumella myxillioides, sp.n. Holotype. x |.

Fig. 4. Plumocolumella myxillioides, sp.n. x f .

Fig. 5. Amphilectus flabellata, sp.n. x f .

DISCOVERY REPORTS, VOL. VI.

PLATE LIII.

'wk..^

John Bale Sons * tfonmliaoivl.'* Ltoi

ANTARCTIC SPONGES

.TE L

PLATE LIV

Fig. I. Amphilectus fucorum Qohnston). x j.

Fig. 2. Amphilectus fucorum (Johnston), x |.

Fig. 3. Amphilectiis fucorum Qohnston). x |.

Fig. 4. Amphilectus fucorum Qohnston). x \.

Fig. 5. Amphilectus edwardi, var. americana (Ridley and Dendy).

Fig. 6. Myxilla australis (Topsent). x f .

Fig. 7. Acheliderma topsenti, sp.n. x i|.

Fig. 8. Acheliderma topsenti, sp.n. x i\.

Fig. 9. Acheliderma topsenti, sp.n. x ij.

Fig. 10. Myxilla chilensis, Thiele. x §.

Fig. II. Anchinoe leptochela (Hentschel). x j.

DISCOVERY REPORTS, VOL. VI

PLATE LIV

^'*.S^

John Bale Sons A. DaiutLuon. I-'* london

ANTARCTIC SPONGES

E L

urlum V* Londor

PLATE LV

X f.
x|.
x|.
x|.

Fig. I. Myxilla mollis (Ridley and Dendy).
Fig. 2. Myxilla mollis (Ridley and Dendy).
Fig. 3. Myxilla mollis (Ridley and Dendy).
Fig. 4. Myxilla mollis (Ridley and Dendy).
Fig. 5. Kirkpatrickia coulmani {KxrhpaXxKk). x|
Fig. 6. Clathria lipochela, sp.n. x |.
Fig. 7. Clathria lipochela, sp.n. x f .
Fig. 8. OphlitaspotJgia thielei, sp.n. x \.

DISCOVERY TfEPORTS, VOL. VI.

PLATE LV.

JohnB<k Saa & DuiHlumL.*' LwirJoo

ANTARCTIC SPONGES

Lv;

^

PLATE LVI

Fig. I. Rfiaphidophlus paucispiculus,sp.n. Holotypc. x 5.

Fig. 2. Protodathria simpHcisshna, sp.n. Holotype. x \.

Fig. 3. Dictyociona discreta (TWide). A macerated specimen.

Fig. 4. Dictyociona discreia (Thiele). x |.

Fig. 5. Plicatellopsis arboresceiis, gen. et sp.n. Holotype. x |.

Fig. 6. Plicatellopsis flabellata, gen. et. sp.n. Holotype. x J.

Fig. 7. Spirastrella purpurea (Ldmarck). x [.

Fig. 8. Duseideia tenuifibra, sp.n. Holotype. x {.

Fig. 9. Hymeniacidon dubia, sp.n. Holotype. x §.

DISCOVERY REPORTS, VOL. VI.

PLATE LVI.

^'•"i

JohnBnJi-Sons A DAnivIuonL'' Lontkn

ANTARCTIC SPONGES

PLATE I

"TS,

i» & OimklMon.L*'' Lot

PLATE LVII

Figs. 1-13. lophon proximum (Ridley), i = D 257; 2-6 =D 251; 7-

9 = D 740; 10 = D 258; II = D 255; 12-13 = D 430.
Figs. 14-16. Spicule balls.

DISCOVERY REPORTS, VOL. VI.

PLATE LVII.

John BaJe ^Qiis \ DamglsMivL'^^ Lnnfton

ANTARCTIC SPONGES

',-11

[Discovery Reports. Vol VT /.^

^- ^^'PP- 393-418, December, 1932]

A LIST OF WORMS PARASITIC
IN CETACEA

By

H.A.BAYLlS,M.A.,D.Sc.

Department ofZooIogy, British Museum

(Natural History)

CONTENTS

. . page 39S

Introduction .••■••• literature, and hosts

Classified list of parasites,. ith synonyms, references ^^,

Trematodes .••••■■' _ .... 399

Cestodes .••••■"' _ _ . . 401

Nematodes .••••'■ . . 405

Acanthocephala .••••■ 407

Host-list ..•••■■' . . 414
Bibliography. ••••■■■

A LIST OF WORMS PARASITIC
IN CETACEA

By H. A, Baylis, m.a., d.sc.

Department of Zoology, British Museum (Natural History)

INTRODUCTION

THE compilation of a catalogue of the worm parasites of Cetacea was undertaken
some time ago at the request of Dr Stanley Kemp, F.R.S., and of the Discovery
Committee. The hst now presented is, it is hoped, as complete as it is possible at the
present time to make it, though our knowledge of many of the forms mentioned is very
scanty, and the systematic position of some of them remains uncertain.
The catalogue is divided into three sections :

(i) A list of the parasites, in which are given their synonyms, references to the more
important literature upon them, and the hosts from which they have been recorded.
The parasites are here arranged, as far as possible, in systematic order, and are classified
under their families, and these again under the headings of the four principal groups,
Trematodes, Cestodes, Nematodes and Acanthocephala. No attempt has been made
to give a complete bibliography for every species mentioned, but the works referred to
include those in which descriptions of the species are given or their systematic position
discussed, or in which references to further literature may be found. The date of
publication follows the author's name, difl^erent publications in the same year being
distinguished by the letters «, b, etc., following the date. Where the date is followed by
a number, this is a page reference. The full references will be found by turning to the
bibliography in the third section.

(2) A list of the hosts, with the parasites recorded from each. In this section the
names of the hosts are arranged alphabetically, and their principal synonyms and ver-
nacular names, when known, are given. The names of the parasites are again arranged
systematically.

(3) A bibliography, arranged in the alphabetical order of the authors' names.

The writer is greatly indebted to Sir Sidney F. Harmer, F.R.S., for very kindly
checking the names and synonyms of the hosts. The synonymy of many of the Cetacea
is extremely involved, and it is sometimes impossible to be certain what species of whale
is meant by authors who have recorded parasites from these animals.

396 DISCOVERY REPORTS

CLASSIFIED LIST OF PARASITES, WITH SYNONYMS,
REFERENCES TO LITERATURE, AND HOSTS

TREMATODES

Family HETEROPHYIDAE

Galactosomum erinaceum (Poirier, 1886), Nicoll, 1923.

Synonyms: Distovmm erinaceum, Poirier, 1886; D. (Dicrocoelium) erinaceum, Stossich, 1892;
Astiotrema erinacea, Stossich, 1904.

References: Poirier, 1886, 37; Stossich, 1892, 22; 1904, 2; Looss, 1899, 590; Nicoll, 1923,
240, 244; Price, 1932, 40.
Recorded from the intestine oi DelpJmnis delphis (in cysts).

Family OPISTHORCHIDAE
Opisthorchis tenuicollis (Rud., 1819), Stiles and Hassall, 1896.
Reference: Price, 1932, 26 {q.v. for synonyms).
Recorded, according to Price, from the bile ducts of Phocaena phocaena.

Amphimerus lancea (Dies., 1850), Barker, 191 1.

Synonyms: Distomimi lancea, Diesing, 1850; D. {Dicrocoelium) lancea, Stossich, 1892; Opis-
thorchis lancea, Braun, 1901.

References: Diesing, 1850, 334; 1855, 64; v. Beneden, 1870, 362; Cobbold, 1876, 35; 1879,
416; Stossich, 1892, 26; Weski, 1900, 579; Braun, 1901, 897; Price, 1932, 29.
Recorded from the liver of Orcella brevirostris, Sotalia tuciixi.

Cyclorchis campula (Cobbold, 1876), Liihe, 1908.

Synonyms: Distoma campula, Cobbold, 1876; Campula oblonga, Cobbold, 1876, nee Cobbold,
1858; Opisthorchis campula, Looss, 1899; Metorchis campula, Looss, 1899.
References: Cobbold, 1876, 40; Looss, 1899, 559, 565; Luhe, 1908, 432; Price, 1932, 28.
Recorded from the bile ducts of Platanista gangetica.

Family TROGLOTREMATIDAE

Pholeter gastrophilus (Kossack, 1910), Odhner, 1914.
Synonym: Distomiim gastrophilum, Kossack, 1910.
References: Kossack, 1910, 118; Odhner, 1914, 233 ; Price, 1932, 24.
Recorded from the stomach of Phocaena phocaena (in cysts).

Family FASCIOLIDAE, sub-family CAMPULINAE
(= CAMPULIDAE, Fuhrm.)

Campula oblonga, Cobbold, 1858, of Braun, 1900.

Synonyms: Campula oblonga, Cobbold, 1858, nee Cobbold, 1876; Distomum ohlongum, Braun,
1891; D. (Brachylaimus) oblongum, Stossich, 1892; Distomum tenuicoUe, Olsson, 1893; Opis-
thorchis oblonga, Kowalevsky, 1898; Brachycladium oblongum, Looss, 1902.

WORMS PARASITIC IN CETACEA 397

References: Cobbold, 1858, 168; 1879, 419; Stossich, 1892, 16; Monticelli, 1893, 44; Stiles,
1895, 219; Looss, 1899, 558, etc.; 1902, 716; Braun, 1900, 249; Price, 1932, 7.

Recorded from the bile ducts of Phocaena phocaena.

Campula delphini (Poirier, 1886), Bittner and Sprehn, 1928.

Synonyms: Distomum delphini, Poirier, 1886; Cladocoelium delphini, Stossich, iSgz; Brachy-
cladium delphini, Looss, 1899.

References: Poirier, 1886, 34; Stossich, 1892, 10; MonticelH, 1893, 44 et passim; Stiles, 1895,
219; Looss, 1899, 558; Odhner, 1905, 348.

Recorded from the liv'er of Delphinus delphis.

Campula palliata (Looss, 1885), Looss, 1901.

Synonyms: Distomum palliatiim, Looss, 1885; Cladocoelium palliatum, Stossich, 1892; Brachy-
cladium palliatum, Looss, 1899.

References: Looss, 1885, 390; 1899, 556; 1901, 208; Stossich, 1892, 10; Monticelli, 1893, 27
et passim; Stiles, 1895, 219; Odhner, 1905, 341, etc.; Price, 1932, 9.

Recorded from the bile ducts of Delphinus delphis.

Campula rochebruni (Poirier, 1886), Bittner and Sprehn, 1928.

Synonyms : Distomum rochebruni, Poirier, 1886 ; Cladocoelium rochebruni, Stossich, 1892 ; Brachy-
cladium rochebruni, Looss, 1899.

References: Poirier, 1886, 36; Stossich, 1892, 11 ; MonticeUi, 1893, 44, etc.; Stiles, 1895, 219;
Looss, 1899, 558; Odhner, 1905, 344, etc.; Price, 1932, 11.

Recorded from the liver of Delphinus delphis.

Lecithodesmus goliath (v. Beneden, 1858), Braun, 1902.
Synonym: Distomum goliath, v. Beneden, 1858.

References: v. Beneden, 1858, 95; Diesing, 1858, 336; Stossich, 1892, 34; Braun, 1902, 800;
Odhner, 1905, 344; Price, 1932, 12.

Recorded from the liver of Balaena mysticetiis, Balaenoptera borealis, B. acutorostrata.

Synthesium tursionis (Marchi, 1873), Stunkard and Alvey, 1930.

Synonyms: Distomum tursionis, Marchi, 1873; D. longissimum, Poirier, 1886; D. (Dicrocoelium)
longissimutn, Stossich, 1892; D. {Dicrocoelium) tursionis, Parona, 1896; Fasciolopsis (?) tursionis,
Nicoll, 1923.

References: Marchi, 1873, 304; Poirier, 1886, 29; Stossich, 1892, 37; Nicoll, 1923, 238;
Stunkard and Alvey, 1930, 332; Price, 1932, 16.

Recorded from the intestine of Tiirsiops trimcatus.

Hadwenius seymouri, Price, 1932.
Reference: Price, 1932, 18.
Recorded from the intestine of Delphinapterus leucas.

398 DISCOVERY REPORTS

Family NOTOCOTYLIDAE

Ogmogaster plicatus (Creplin, 1829), Jagerskiold, 189 1.
Synonym: Monostomum plicatiiin, Creplin, 1829.

References: Creplin, 1829, 878; Diesing, 1850, 324; v. Beneden, 1870, 357; Cobbold, 1879,
421 ; Jagerskiold, 1891 a; 1891 b, 129; Brandes, 1892, 509; Monticelli, 1892 b, 696, etc.; Odhner,
1905, 366; Leiper and Atkinson, 1915, 37; Hamilton, 1916, 131; Price, 1932, 45.

Recorded from the intestine of Balaenoptera acutorostrata, B. borealis, B. miisadus,
B. pJiy sains.

UNCLASSIFIED FORMS

Distoma andersoni, Cobbold, 1876.

Synonym: Distomum (Brachylaimus) andersoni, Stossich, 1892.

References: Cobbold, 1876, 46; 1879, 420; Stossich, 1892, 19; Price, 1932, 56.

Recorded from the small intestine of Platam'stn gangetica.

Distomum pallasii, Poirier, 1885.

Synonym: D. {Dicrocoelium) pallasii, Stossich, 1892.
References: Poirier, 1885, 477; Stossich, 1892, 27.

Recorded from the stomach of Phocaena phocaena.

Distomum philocholum, Creplin, 1845.
Reference: Creplin, 1845, 330.
Recorded from the liver of Phocaena phocaena (see Nicoll, 1923, 243. According to
Price, 1932, 56, the host is Delphimis delphis.)

Distomum validum, v. Linstow, 1886.

Synonym: D. [Brachylaimus) validum, Stossich, 1892.
References: v. Linstow, i886, 124; Stossich, 1892, 15.

Recorded from the stomach of Delphinus sp.

Monostomum delphini, Diesing, 1850.

Synonyms: Monostomum blainvillei, Cobbold, i860; Monostomulum delphini, Brandes, 1892;
Agamodistomum delphini, Price, 1932.

References: de Blainville, 1825, 141; Diesing, 1850, 330; Cobbold, i860, 39; 1879, 421;
V. Beneden, 1870, 358; Monticelli, 1892 b, 711 ; Brandes, 1892, 510; Price, 1932, 57.

Recorded from the blubber of "Delphinus dalei spec, affin.''^ (in cyst).

1 From information kindly supplied by Sir Sidney Harmer, F.R.S., it appears that the whale here
referred to was probably either Hypcroudon rostratus or Mesoplodun bidens.

WORMS PARASITIC IN CETACEA 399

CESTODES

Family DIPHYLLOBOTHRIIDAE

Diphyllobothrium stemmacephalum, Cobbold, 1858.

Synonyms: Dibothrium stemmacephalum, Diesing, 1863 ; Bothriocephalus stemmacephalus, Braun,

1883.

References: Cobbold, 1S58, 167; 1879, 105, 422; Diesing, 1863, 239; Stiles and Hassall, 1899,

170; Scott, 1909, 88; Baer, 1932, 214.

Recorded from the intestine of Delphinus delp/iis, Phocaena phocaena.

Diplogonoporus balaenopterae, Lonnberg, 1892.

References: Lonnberg, 1892, 4; Liihe, 1899, 50; Ariola, 1900, 385; Baer, 1932, 214.
Recorded from the intestine of Balaenoptera borealis.

Family PHYLLOBOTHRIIDAE

Phyllobothrium delphini (Bosc, 1802), Gervais, 1885.

Synonyms: Hydatis delphinii, Bosc, 1802; Cysticercus delphini, Laennec, 1804; Cysticercus
delphini, V^udo\'p\\\, 1810; Vermis delphini delphis, Rud., 1810; Cephalocotyleum delphini delphis,
Diesing, 1850; } Cysticercus physeteris. Dies., 1863 (vide Phyllobothrium physeteris, infra).
References: Bosc, 1802, i, 324; Laennec, 1804, 136; Rudolphi, 1810, 236, 265; Diesing, 1850,
493; 1863 b, 423; V. Beneden, 1870, 360; Cobbold, 1879, 422; Moniez, 1880, 117; Gervais, in
Carus, 1885, 114; Linton, 1905, 820; Stiles and Hassall, 1912, 171; Southwell and Prashad,
1920, 2; Baer, 1932, 219.

Recorded from the blubber of Delphinus delphis, Grampus griseus, Tursiops truncatiis.

Phyllobothrium inchoatum, Leidy, 1891.

(Possibly identical with P. lactuca, v. Beneden, 1850.)

References: Leidy, 1891, 418; Southwell and Prashad, 1920, 2; Southwell, 1925, 152.
Recorded from the blubber of Mesoplodon bidens.

Phyllobothrium physeteris (Dies., 1863), Meggitt, 1924.
(Possibly identical with P. delphini (Bosc).)

Synonyms: Cysticercus balaenae mysticeti, Diesing, 1850; Cysticercus physeteris, Diesing, 1863
References: Diesing, 1850, 493; 1863 &, 423; v. Beneden, 1870, 348; Cobbold, 1879, 421;
Moniez, 1880, 117; Meggitt, 1924, 156.

Recorded from the blubber and skin of Balaena tnysticetus, Physeter catodon.

Phyllobothrium sp. "(Carnot, 1822)," Meggitt, 1924.
References: Cobbold, 1879, 421; Meggitt, 1924, 157.
Recorded from the nasal sinus of Tursiops truncatiis.

Phyllobothrium, sp., Linton, 1905.
Reference: Linton, 1905, 819.
Recorded from the mesentery of Lagenorhynchus acutus.

400 DISCOVERY REPORTS

Monorygma chamissonii (Linton, 1905), Meggitt, 1924.
(Possibly identical with M. grimaldii, vide infra.)

Synonyms: Taenia chamissonii, Linton, 1905; ? Cysticercus delphini, Rudolphi, 18 19, nee 18 10.
References: Linton, 1905, 819; Meggitt, 1924, 152; Baylis, 1926 b, 667.
Recorded from the blubber and mesentery of Lagenorhynchus acutus.

Monorygma delphini (Gervais, 1870), Meggitt, 1924.
Synonym: Stenotaenia delphini, Gervais, 1870.

References: Gervais, 1870, 779; Baylis, 1919, 417; igzb b, 667; Meggitt, 1924, 152.
Recorded from the diaphragm of Delphimis delphis.

Monorygma grimaldii (Moniez, 1889), Meggitt, 1924.

Synonyms: Taenia Grimaldii, Moniez, 1889; Cysticercus Taeniae Grimaldii, Moniez, 1889;
Cysticercus grimaldii, Braun, 1898.

References: Moniez, 1889, 825; Braun, 1898, 1558; Baylis, 1919, 417; 19266, 665; Meggitt,
1924, 152.

Recorded from the blubber of Lagenorhynchus acutus, Kogia breviceps, {})Delphinus
delphis.

Scolex delphini, Stossich, 1898.
Reference: Stossich, 1898, 8.
Recorded from the rectum of Grampus griseus.

Family TETRABOTHRIIDAE

Tetrabothrius affinis (Lonnberg, 1891), Lonnberg, 1892.

Synonyms: Diplobothrium affine, Lonnberg, 1891 ; Tetrabothriwn {Diplobothrium) affine,
Lonnberg, 1892.

References: Lonnberg, in Jagerskiold, 1891 b, 130, etc.; Lonnberg, 1892, 16; Baylis, 1926 a,
163; Baer, 1932, 196.

Recorded from the intestine of Balaenoptera borealis, B. fnuscidus.
Tetrabothrius forsteri (Krefft, 1871), Johnston, 191 2.

Synonyms: Taenia forsteri, Krefft, 1871 ; Prosthecocotyle forsteri, Monticelli, 1892; Tetra-
bothrium triangtdare, Leidy, 1892, nee Diesing, 1850.

References: Krefft, 1871, 218; Cobbold, 1879, 422; Monticelli, 1892, 6; Fuhrmann, 1898,
386, etc.; 1899 a, 181 ; 1899 b, 865, 869; Johnston, 1912, 13; Linton, 1923, 52; Baylis, 1926 a,
169; Baer, 1932, 197.

Recorded from the intestine of Delphimis delphis, Mesoplodon bidens, Steno rostratus.
Tetrabothrius ruudi, Nybehn, 1928.
Reference: Nybelin, 1928, 312.
Recorded from the intestine of Balaenoptera physalus.
Tetrabothrius wilsoni (Leiper and Atkinson, 1914), BayHs, 1926.
Synonym: Oriana wilsoni, Leiper and Atkinson, 1914.

References: Leiper and Atkinson, 1914, 225; 1915, 46; Baylis, 1926(7, 168; Baer, 1932, 198.
Recorded from the intestine of Balaenoptera borealis.

WORMS PARASITIC IN CETACEA 401

Priapocephalus grandis, Nybelin, 1922.

References: Nybelin, 1922, 194; Baylis, 1926 a, 161 ; Baer, 1932, 205.
Recorded from the intestine of Balaenoptera borealis, B. musculus, " Balaena ant-
arctica".

Priapocephalus minor, NybeHn, 1928.
Reference: Nybelin, 1928, 309.
Recorded from the intestine of Balaenoptera borealis.
Strobilocephalus triangularis (Dies., 1850), Baer, 1932.

Synonyms: Tetrabotluium triangulare, Diesing, 1850; Pwsthecocotyle triangidaie, Fuhrmann,
1899; Tetrabothriits triangularis, Fuhrmann, 1903.

References: Diesing, 1850, 601; 1863 «, 255; v. Beneden, 1870, 362; Cobbold, 1879, 422;
Leidy, 1891, 417; Fuhrmann, 1899 rt, 1S3; 18996, 865, etc.; Linton, 1923, 52; Joyeux and
Dollfus, 1931, 112; Baer, 1932, 202.
Recorded from the intestine of Delphimis sp., Hyperoodon rostratus, Lagenorhynchus
acutiis.
Trigonocotyle monticellii (Linton, 1923), Baer, 1932.

Synonym: Prosthecocotyle monticellii, Linton, 1923, nee Fuhrmann, 1899.
References: Linton, 1923, 52; Baylis, 1926 a, 169, 170; Baer, 1932, 199.
Recorded from the intestine of Globicephala melaena.

NEMATODES

Family ASCARIDAE

Anisakis catodontis, Baylis, 1929.

Reference: BayHs, 1929, 544.
Recorded from the stomach of Physeter catodon.
Anisakis dussumierii (v. Beneden, 1870), BayHs, 1920.

Synonyms: Ascaris {Anisakis) simplex of Dujardin, 1845, nee Rudolphi, 1809 ; Ascaris dussumierii,

van Beneden, 1870.

References: Dujardin, 1845, 220; v. Beneden, 1870, 362; Cobbold, 1879, 426; Stiles and

Hassall, 1899, 161 et passim; Stossich, 1896, 17; Baylis, igzo b, 260; 1923 a, 216.

Recorded from the intestine of Delphmus sp.

Anisakis insignis (Dies., 1851), BayHs, 1920.

Synonyms: Peritrachelius insignis, Diesing, 1851 ; Ascaris insignis, Jagerskiold, 1893.
References: Diesing, 1851, 210; 1855, 181 ; Jagerskiold. 1893, 7; 1894, 543; Stiles and Hassall,
1899, 107, 138; Baylis, 1920 b, 260; 1923 a, 216.
Recorded from the stomach of Ma geojfrensis [Delphimis amazonicus] .

Anisakis kiikenthalii (Cobb, 1888), BayHs, 1920.
(Possibly a synonym of A . simplex.)
Synonym: Ascaris kiikenthalii, Cobb, 1888.

402 DISCOVERY REPORTS

References: Cobb, 1888, 4, etc.; Stossich, 1896, 53; Stiles and Hassall, 1899, 144 et passim;
Baylis, 1920 b, 260; 1923 a, 216.

Recorded from the stomach of Delphiuapterus leiicas.

Anisakis physeteris, BayHs, 1923.

References: Baylis, 1923 rt, 211 ; 1929, 543.

Recorded from the stomach of Physeter catodon.

Anisakis simplex (Rud., 1804), BayHs, 1920.

Synonyms: Ascaris simplex, Rudolphi, 1804; (not Ascaris (Anisakis) simplex, Dujardin, 1845);
Ascaris angiilivalvis, Creplin, 1851.

References: Rudolphi, 1804, 94 {nomen nudum); 1809, 170; Dujardin, 1845, 220; Creplin, 1851,
158; Diesing, 1851, 155; 1861, 656; v. Beneden, 1870, 363; Krabbe, 1878, 47; Cobbold, 1879,
426; Monticelli, 1889, 69; Leidy, 1891, 411 ; Jagerskiold, 1891 b, 131, 132; 1893, 27; 1894, 474;
1898, 738 et passim; Stossich, 1896, 17; Stiles and Hassall, 1899, 120 et passim; Baylis, 1920 b,
260; 1923 a, 216.

Recorded from the stomach of Boloenoplera acutorostrata , B. borealis, B. musadtis,
Delphiiiapleriis leiicas, Delphimis sp., Hyperoodon rosfratiis, Lagenorhyiichus olbirostris,
Mesoplodon bideiis, Monodon monoceros, Phocaena phocaena, {})Platanista gangetica.

Anisakis typica (Dies., 1861), BayHs, 1920.

Synonyms: Conocephalus typicus, Diesing, i86i; Ascaris typica, Jagerskiold, 1894; Ascaris
[Peritrachelius) typica, Jagerskiold, 1894; Peritrachelius typicus, Jagerskiold, 1894.
References: Diesing, 1861, 669; Jagerskiold, 1894, 450, 453; Stiles and Hassall, 1899, 127 et
passim; Stossich, 1902, 6 (66); Baylis, 1920 b, 260; 1923 a, 216; 1929, 543.

Recorded from the stomach of Delphimis delp/iis, Globicephala melaena, Lageno-
rhyiichus obscurus, Phocaena phocaena, Prodelphinus sp.

Anisakis sp. (larval).

Reference: Baylis, 1929, 544.
Recorded from [the stomach of ?] Megaptera nodosa.

Contracaecum lobulatum (Schneider, 1866), BayHs, 1920.

Synonyms: Ascaris lobulata, Schneider, 1866; } Ascaris delphini, Rudolphi, 1819 {vide infra);
} Ascaris simplex of Diesing, 1851 (part).

References: Schneider, 1866, 44, 193; Krabbe, 1878, 47; Jagerskiold, 1893, 20; 1894, 467;
Stossich, 1896, 43 ; Stiles and Hassall, 1899, 159 et passim; v. Linstow, 1907, 37; Baylis, 1920 b,
260; Baylis and Daubney, 1923, 557.

Recorded from the mouth, oesophagus, stomach and intestine of Plataiiisla gangetica.

Contracaecum sp. (larval).
Reference: Baylis, 1929, 548.
Recorded from the stomach of Balaenoptera physalus.

"Ascaris" delphini, Rud., 18 19 (sp. inq.).
(? synonym of Contracaecum lobulatum).

References: Rudolphi, 1819, 54, 296; Dujardin, 1845, 221; v. Beneden, 1870, 359; Cobbold,
1879, 426; Stossich, 1896, 17; Stiles and Hassall, 1899, 162 et passim.

Recorded from the mouth and stomach of PlaUviista gangetica.

WORMS PARASITIC IN CETACEA 403

Family ONCHOLAIMIDAE

Odontobius ceti, Roussel de Vauzeme, 1834.

References: Roussel de Vauzeme, 1834, 326; Dujardin, 1845, 292; Cobbold, 1879, 427;
Baylis, 1923, 617.

Recorded from the baleen-plates of Balaena australis, Balaenoptera iniiscnliis, B. p/iy-
salus, Megaptera nodosa.

Family METASTRONGYLIDAE

Pseudalius inflexus (Rud., 1809), Schneider, 1866.

Synonyms: Stwiigylus inflexus, Rudolphi, 1809; S. inflexus major, Raspail, 1829; S. major,
Raspail, 1829; Pseudalius filum, Dujardin, 1845; Prosllwcosacter inflexus, Diesing, 1851;
? Prosthecosacta convolutus, Cobbold, 1864; ? Prosthecosacter convolutus, Cobbold, 1879. (Not
Stenurus inflexus, Dujardin, 1845.)

References: Rudolphi, 1809, 227; Kuhn, 1829 rt, 150; 18296, 139; 1829 c, 363, 367; Raspail,
1829, 250, 251; Dujardin, 1845, 135; Diesing, 1851, 323; Cobbold, 1864, PI. vi, fig. 3; 1879,
424; Schneider, 1866, 173; Baylis and Daubney, 1925, 204.

Recorded from the bronchi, blood-vessels and heart of Phocaena phocaena.

Stenurus arcticus (Cobb, 1888), Baylis and Daubney, 1925.

Synonyms: Strongylus pallasii, v. Beneden, 1870 (tiomen nudum); Sirongyhis arcticus, Cobb,
1888; Pseudalius arcticus, v. Linstow, 1906.

References: v. Beneden, 1870, 366; Cobb, 1888, 64; 1889, 150; v. Linstow, 1906, 114;
Baylis and Daubney, 1925, 208.

Recorded from the "hearing organs" (probably tympanic cavity) of Delphinapterus
leucas.

Stenurus globicephalae, Bayhs and Daubney, 1925.
Reference: Baylis and Daubney, 1925, 206.
Recorded from the base of the blow-hole of Globicephala melaena.

Stenurus minor (Kuhn, 1829), Baylis and Daubney, 1925.

Synonyms: Strongylus jninor, Kuhn, 1829; Stenurus inflexus, Dujardin, 1845 {nee Strongylus
inflexus, Rudolphi, 1809) ; Prosthecosacter minor, Diesing, 1851 ; Prosthecosacter (Stenurus) minor,
Diesing, 1861 ; Pseudalius minor, Schneider, 1866; Pharurus minor, Cobbold, 1879.
References: Kuhn, 1829 a, 152; 18296, 139; 1829 f, 363; Raspail, 1829, 246; Dujardin, 1845,
266; Diesing, 1851, 323; 1861, 719; Schneider, 1866, 174; Cobbold, 1879, 423; v. Linstow,
1888 a, 235; Baylis and Daubney, 1925, 205.

Recorded from the bronchi, tympanic cavity and blood-vessels (especially the venous
sinuses of the head) of Phocaena phocaena.

Stenurus ovatus (v. Linst., 1910), Baylis and Daubney, 1925.

Synonym: Pseudalius ovatus, v. Linstow, 1910.

References: v. Linstow, 1910, 133; Baylis and Daubney, 1925, 208; Baylis, 1928, 464.
Recorded from the oesophagus (?), stomach (?), blow-hole, larynx and trachea of
Tursiops truncatus.

404 DISCOVERY REPORTS

"Strongylus" [PStenurus] alatus, Leuckart, 1848.

Synonyms: Prosthecosacter alatiis, Diesing, 1851 ; Strongylus {Pharurus) alatus, Diesing, 1851 ;
Pseudalius alatus, v. Linstow, 1888. (Not Strongylus alatus, v. Linstow, 1879.)
References: Leuckart, 1848, 26; Diesing, 1851, 324, 325; v. Linstow, 1888 Z), 15; Baylis and
Daubney, 1925, 208.

Recorded from the pharyngeal cavities, mouth, Eustachian tubes and "cranial
cavity" (probably in blood-sinus) of Monodon monoceros.

Torynurus convolutus (Kuhn, 1829), Baylis and Daubney, 1925.

Synonyms: Strongylus convolutus, Kuhn, 1829; Prosthecosacter convolutus, Diesing, 1851;
Prosthecosacta convolutus , Cobbold, 1864; Pseudalius convolutus, Schneider, 1866; } Pseudalius
bicostatus, v. Linstow, 1906.

References: Kuhn, 1829 <", 365; Diesing, 1851, 324; Cobbold, 1864, 91; 1879, 423; Schneider,
1866, 174; V. Linstow, 1906, 114; Baylis and Daubney, 1925, 209.

Recorded from the bronchi and pulmonary blood-vessels of Globicephala melaena,
Phocaena phocaena.

Halocercus delphini, Baylis and Daubney, 1925.
Reference: Baylis and Daubney, 1925, 210.
Recorded from the bronchi of Delphimis delphis.

Halocercus infiexocaudatus (v. Sieb., 1842), Baylis and Daubney, 1925.

Synonyms: Filaria inflexocaudata, v. Siebold, 1842; Pseudalius tumidus, Schneider, 1866;
? Strongylus vagans, Eschricht, 1S41 ; ? Strongylus invaginatus, Quekett, 1842.
References: v. Siebold, 1842, 348; Schneider, 1866, 174; Baylis and Daubney, 1925, 213, 215.
Recorded from the lungs of Phocaena phocaena (in nodules).

Halocercus lagenorhynchi, Baylis and Daubney, 1925.
Reference: Baylis and Daubney, 1925, 211.
Recorded from the bronchi of Lagenorhyuchiis albirostris.

Halocercus pingi, Wu, 1929.
Reference: Wu, 1929, 276.
Recorded from the lungs of Neomeris phocaenoides (in cavities in the tissue).

Family SPIRURIDAE

Crassicauda bennetti, Spaul, 1926.
{= Crassicauda sp., Baylis, 1920).
References: Baylis, 19200, 417; Spaul, 1926, 581.
Recorded from the kidney oi Hyperoodon sp. (probably ^/flHZ/irowjr).

Crassicauda boopis, Baylis, 1920.

Synonym: C. crassicauda of Leiper and Atkinson, 1914.

References: Leiper and Atkinson, 1914, 226; 1915, 29; Baylis, 1916, 144; 1920, 417.
Recorded from the renal tubules and stomach (wall ?) of Megaptera nodosa, and
doubtfully from the renal tubules (.?) of Ziphhis cavirostris.

WORMS PARASITIC IN CETACEA 405

Crassicauda crassicauda (CrepHn, 1829), Leiper and Atkinson, 191 5.

Synonym: Filaria crassicauda, Creplin, 1829.

References: Creplin, 1829, 874; Dujardin, 1845, 50; Diesing, 1851, 264; Molin, 1858, 374;
V. Beneden, 1870, 356; Cobbold, 1879, 425; Jagerskiold, 1891 b, 129; Hamilton, 1916, 132;
Baylis, 19200,410; 1922,9; 1929, 550; Joyeux and Baer, 1931, 198.

Recorded from the renal tubules, urethra and other parts of the urino-genital system
oi Balaenoptera physalus, B. musctdus, Tursiops truncatus, {})Balaenoptera borealis.

Crassicauda fuelleborni (Hoeppli and Hsii, 1929).'

Synonym: Onchocerca fuelleborni, Hoeppli and Hsii, 1929.
Reference: Hoeppli and Hsii, in Hoeppli, Hsii and Wu, 1929, 33.
Recorded from nodules in the musculature of the vagina of Neomeris phocaenoides .

ACANTHOCEPHALA

Bolbosoma aurantiacum (Risso, 1826), Porta, 1908.

Synonyms: Echinorhymchiis aurantiacus, Risso, 1826; E. anmdatus, Molin, 1858; E. bifasciatus,
Liihe, 1904; ? E. pellucidus, Leuck., 1828; E. serrani, Linton, 1901 ; Bolborhynclius aurantiacus.
Porta, 1907.

References: Risso, 1826, 261; Leidy, 1891, 413; Porta, 1905, 167, 181; 1908, 273; 1909, 250;
Van Cleave, 1924, 293.

Recorded from the intestine of Delphirms delphis, Mesoplodon bidens.

Bolbosoma brevicolle (Malm, 1867), Porta, 1908.

Synonyms: Echinorhynchus brevicollis. Malm, 1867; Bolborhynclius brevicollis, Porta, 1908.
References: Malm, 1867, 95; v. Beneden, 1870, 357; Cobbold, 1879, 428; Borgstrom, 1892;
V. Linstow, 1896, 20; Shipley, 1899, 262; Porta, 1906, 269; 1908, 274; 1909, 251 ; Baylis, 1929,
555; Meyer, 193 1, 13.

Recorded from the intestine of Balaenoptera acutorostrata, B. borealis, B. miisculus,
B. physalus, Physeter catodon.

Bolbosoma capitatum (v. Linst., 1880), Porta, 1908.

Synonyms: Echinorhynchus capitatus, v. Linstow, 1880; Bolborhynclius capitatus. Porta, 1908;
} Echinorhynchus vasculosus , Rudolphi, 1819.-

References: Rudolphi, 1819, 75, 334, 581 ; v. Linstow, 1880, 49; Shipley, 1899, 265 et passim;
Porta, 1906,235; 1908,273; 1909, 254; Hamilton, 1916, 132; Baylis, 1929, 556; Meyer, 193 1, 14.

Recorded from the intestine of Globicephala melaena, Physeter catodon, Pseudorca
crassidens, Steno ros trains.

^ From the description of this worm, and its position in nodules of connective tissue in the musculature
of the vagina of its host, there can be little doubt that it belongs to the genus Crassicauda. The description,
however, which was based on fragmentary material, is not sufficient to determine whether the species is
identical with any of those already known. Since the male appears to be without spicules, it seems probable
that it is distinct at least from C. crassicauda.

^ A larval form found in various fishes, according to Meyer (193 1) possibly the larva of B. capitatum.

4o6 DISCOVERY REPORTS

Bolbosoma hamiltoni, Baylis, 1929.

Reference: Baylis, 1929, 556.
Recorded from the intestine of Bolaenoptera musculus, B. physalus.
Bolbosoma pellucidum (Leuckart, 1828), Porta, 1908.

(? Synonym of B. aitrantiacum.)

Synonyms: Echinorhynchus pellucidiis , Leuckart, 1828; [Bolborhynchus pellucidus]. Porta, 1906.

References: Leuckart, 1828, 23; Diesing, 1851, 44; Cobbold, 1879, 428; Leidy, 1891, 413;

Shipley, 1899, 264 et passim; Porta, 1906, 267; 1908, 273; 1909, 250.

Recorded from the intestine of Delphi/ins delphis, Mesoplodon bideiis.
Bolbosoma porrigens (Rud., 1814), Porta, 1908.

Synonyms: Echinorhynchus balaenae, Gmelin, 1790, e.p.; Echinorhynchus porrigens, Rudolphi,
1814; Bolborhynchus porrigens , Porta, 1906.

References: Rudolphi, 1814,98; 1819, 71, 325, etc.; Westrumb, 1821, 28; Cobbold, 1879,427;
Jagerskiold, 1891 b, 128, etc.; Borgstrom, 1892, i ; v. Linstow, 1896, 20; Shipley, 1899, 262 et
passim ; Liihe, 1904-5, 154, 180, etc. ; Porta, 1906, 235 et passim ; 1908, 276 ; 1909, 255 ; Hamilton,
1916, 132; Wesenberg-Lund, 1926, 148; Johnston and Deland, 1929, 147; Meyer, 1931, 13
{B. balaenae).

Recorded from the intestine of Balaena viysticehis, Balaenoptera aciitorostrata,
B. borealis, B. ttiusailus, Megaptera nodosa.

Bolbosoma turbinella (Dies., 1851), Porta, 1908.

Synonyms: Echinorhynchus turbinella, Diesing, 1851; E. ruber, Collett, 1886; E. porrigens of
Kaiser, 1893; Bolborhynchus turbinella, Porta, 1908; Pomporhynchus turbinella, Leiper and
Atkinson, 1915.

References: Diesing, 1851, 54; Cobbold, 1879, 427; Collett, 1886, 258; Jagerskiold, 1891, 128,
etc. ; Borgstrom, 1892 ; Kaiser, i%()t,, passim ; v. Linstow, 1896, 20 ; Shipley, 1899, 262 et passim;
Liihe, 1904-5, 284, etc. ; Porta, 1906, 235 et passim ; 1908, 274; 1909, 253 ; Leiper and Atkinson,
1915, 31 ; Wesenberg-Lund, 1926, 147; Baylis, 1929, 555; Meyer, 193 1, 14.

Recorded from the intestine of Balaeuopiera borealis, B. musculus, Hyperoodon
rostra tus, Megaptera nodosa.

Bolbosoma sp.

Reference: Van Cleave, 1925, 155.
Recorded from the intestine of Balaeuopiera velifera {} = B . physalus) .
Corynosoma sp., Johnston and Deland, 1929.
Reference: Johnston and Deland, 1929, 147.
Recorded from [the intestine of ?] Delphinus delphis.

WORMS PARASITIC IN CETACEA 407

HOST-LIST

The names of the hosts are arranged alphabetically.

Balaena australis (Southern right whale).
Nematodes.
Odontobiiis ceti, Roussel de Vauzeme, 1834.

" Balaena antarctica."
[} B. australis i\
Cestodes.

Priapocephahis grandis, Nybelin, 1922 [fide Baer, 1932, 224].

Balaena mysticetus (Arctic right whale, Bowhead).

Synonym : B. groenhndica.
Trematodes.

Lecithodesmus goliath (v. Ben., 1858).
Cestodes.

Phyllobolhrhim physeieris (Dies., 1863) {^=P. delpliitii (Rud.)).
Acanthocephala.

Bolbosoma porrigens (Rud., 18 14).

Balaenoptera acutorostrata (Lesser rorqual. Lesser finner, Little piked whale).

Synonym: B. irjstrata, auctt.
Trematodes.

Lecithodesmus goliath (v. Ben., 1858).

Ogmogaster plicatus (Crepl., 1829).
Nematodes.

Anisakis simplex (Rud., 1S09).
Acanthocephala.

Bolbosoma brericolle (Malm, 1867).

Bolbosoma porrigens (Rud., 18 14).

Balaenoptera borealis (Sei whale, Pollack whale, Rudolphi's whale).

Synonym: Balaena rostrata, Rud., 1S22, ncc MuIIer, 1776.
Trematodes.

Lecithodesmus goliath (v. Ben., 1858).

Ogmogaster plicatus (Crepl., 1829).
Cestodes.

DiplogoHoportis balaeuopterae, Lonnbg., 1892.

Tetrabothriiis ajfinis (Lonnbg., 1892).

Tetrabothrius u-ilsoni (Leiper and Atk., 1914).

Priapocephahis grandis, Nybelin, 1922.

Priapocephahis minor, Nybelin, 1928.

4o8 DISCOVERY REPORTS

Nematodes.

? Ascaris angulivalvts, Crepl., 185 1.

? Crassicauda crassicauda (Crepl., 1829).
Acanthocephala.

Bolbosoma brevicoUe (Malm, 1867).

Bolbosoma porrigens (Rud., 1814).

Bolbosotna tiirbinella (Diesing, 185 1).

Balaenoptera musculus (Blue whale. Sulphur-bottom, Sibbald's rorqual).

Synonyms: B. sibbaldii, B. intermedia.
Trematodes.

Ogmogaster plicatus (Crepl., 1829).
Cestodes.

Priapocephalus grandis, Nybelin, 1922.

Tetrabotliriiis affinis (Lonnbg., 1892).
Nematodes.

Ascaris angulivalvis, Crepl., 185 1.

Odontobttis ceti, Roussel de Vauzeme, 1834.

Crassicauda crassicauda (Crepl., 1829).
Acanthocephala.

Bolbosoma brevicoUe (Malm, 1867).

Bolbosoma hamiltoni, Baylis, 1929.

Bolbosoma porrigens (Rud., 18 14).

Bolbosoma tiirbinella (Dies., 185 1).

Balaenoptera physalus (Fin whale. Finback, Finner, Common rorqual).

Synonym: B. niuscidiis, auctt.
Trematodes.

Ogmogaster plicatus (Crepl., 1829).
Cestodes.

Tetrabothrius rmidi, Nybelin, 1928.
Nematodes.

Contracaecmn sp., Baylis, 1929.

Odontobius ceti, Roussel de Vauzeme, 1834.

Crassicauda crassicauda (Crepl., 1829).
Acanthocephala.

Bolbosoma brevicoUe (Malm, 1867).

Bolbosoma hamiltoni, Baylis, 1929.

Balaenoptera velifera.
[Probably = B. physalus.]
Acanthocephala.

Bolbosotna sp.. Van Cleave, 1925.

WORMS PARASITIC IN CETACEA 409

Delphinapterus leucas (White whale).
Synonym : Beluga leucas.
Nematodes.

Anisakis kiikenthalii (Cobb, 1888).
Anisakis simplex (Rud., 1809).
Stemiriis arcticus (Cobb, 1888).

" Delphinus dalei."

[Probably =//_y/)erooJo?i rostratus or Mesoplodon bidens.]
Trematodes.

Monostomum delphini, Dies., 1850.

Delphinus delphis (Common dolphin).
Trematodes.

Galactosomum erinoceiim (Poirier, 1886).

Campula delphini (Poirier, 1886).

Campida palliata (Looss, 1885).

Campula rochebruni (Poirier, 1886).

} Distomum philochohim, Crepl., 1845.
Cestodes.

Diphyllobothrium stemmocephahim , Cobbold, 1858.

Phyllobothriiini delphini (Rud., 18 19) {}= P. physeteris).

Monorygma delphini (Gervais, 1847).

} Monorygma grimaldii (Moniez, 1889), Meggitt, 1924. (Cysticercus Taeniae
Grimaldii, Moniez, 1889.)

Tetrabothrius forsteri (Krefft, 1871).
Nematodes.

Anisakis typica (Dies., i860).

Halocercus delphini, Baylis and Daubney, 1925.
Acanthocephala.

Bolbosoma aurantiaciim (Risso, 1826).

Bolbosoma pelhicidum (Leuck., 1828).

Corynosoma sp., Johnston and Deland, 1929.

Delphinus sp.
Nematodes.

Anisakis simplex (Rud., 1809).

Delphinus sp.
Nematodes.

Anisakis dussumieri (v. Ben., 1870).

Delphinus sp.
Trematodes.

Distomum validum, v. Linst., 1886.

4IO DISCOVERY REPORTS

Delphinus sp.

Cestodes.

Strobilocephaliis triangularis (Dies., 1850).

Globicephala melaena (Pilot whale, Caa'ing whale, Blackfish).
Synonyms : Globicephalus melas, G. globiceps, G. svineval.

Cestodes.

Trigonocotyle monticellii (Linton, 1923).
Nematodes.

Anisakis typica (Dies., i860).

Stemirus globicephalae , Baylis and Daubney, 1925.

Toryminis convoliitus (Kuhn, 1829).
Acanthocephala. /

Bolbosoma capitatiim (v. Linst., 1880).

Grampus griseus (Risso's dolphin).
Synonym : G. cumeri.
Cestodes.
Phyllobothriiini delphini (Rud., 18 19) {}= P. pliyseteris).
Scolex delphini, Stossich, 1898.

Hyperoodon rostratus (Bottle-nosed whale). [See also "Delphinus dalei."]

Cestodes.

Strobilocephaliis triangularis (Dies., 1850).
Nematodes.

Anisakis simplex (Rud., 1809).
Acanthocephala.

Bolbosoma tiirbinella (Dies., 1851).

Hyperoodon sp. (probably H. planifrons).
Nematodes.

Crassicauda bennetti, Spaul, 1926.

Inia geoffrensis ("Bouto" (native name)).

Synonyms : Delphinus amazotiicus, D. geofjroyi, Inia boliviensis.

Nematodes.

Anisakis insignis (Dies., 185 1).

Kogia breviceps (Pygmy Sperm whale).
Cestodes.

Monorygma grimaldii (Moniez, 1889), Meggitt, 1924. (Cysticercus Taeniae
Grimaldii, Moniez, 1889.)

WORMS PARASITIC IN CETACEA 411

Lagenorhynchus acutus (White-sided dolphin).
Cestodes.

Phyllobothrium sp., Linton, 1905.

Monorygma chamissonii (Linton, 1905).

Monorygma grimaldii (Moniez, 1889), Meggitt, 1924. {Cysticercus Taeniae

Grimaldii, Moniez, 1889.)
Strobilocephaliis triangularis (Dies., 1850).

Lagenorhynchus albirostris (White-beaked dolphin).
Nematodes.

Anisakis sitnplex (Rud., 1809).

Halocerciis lagenorhynchi, Baylis and Daubney, 1925.

Lagenorhynchus obscurus.

Nematodes.

Anisakis typica (Dies., i860).

Megaptera nodosa (Humpback).
Synonyms : M. longimana, M. boops.
Nematodes.

Anisakis sp., Baylis, 1929.
Odontobius ceti, Roussel de Vauzeme, 1834.
Crassicauda boopis, Baylis, 1920.
Acanthocephala.

Bolbosoma porrigens (Rud., 1814).
Bolbosoma turbinella (Dies., 185 1).

Mesoplodon bidens (Sowerby's whale). [See also '' Delphinus dalei".]
Synonym : M. sowerbtensis.
Cestodes.

Phyllobothrium inchoatum, Leidy, 1891 (probably=P. lacttica, v. Ben., 1850).

Tetrabothrius forsteri (Krefft, 1871).
Nematodes.

Anisakis sitnplex (Rud., 1809).
Acanthocephala.

Bolbosoma aiirantiacum (Risso, 1826).

Bolbosoma pelliicidum (Leuck., 1828).

Monodon monoceros (Narwhal).
Nematodes.

Anisakis simplex (Rud., 1809).

Steniiriis{}) alatus (Leuck., 1848).

3-2

412 DISCOVERY REPORTS

Neomeris phocaenoides ("Common Chinese porpoise").

Nematodes.

Crassicaiida fuelleborni (Hoeppli and Hsii, in Hoeppli, Hsii and Wu, 1929).
Halocerciis pingi, Wu, 1929.

Orcella brevirostris.
Trematodes.

Amphimeriis lancea (Dies., 1850).

Phocaena phocaena (Common porpoise).
Synonym : P. communis.

Trematodes.

Opislhorchis tenuicolUs (Rud., 1819).

Pholeter gastrophilus (Kossack, 1910).

Campida oblonga (Cobbold, 1858).

Distonmm paUasii,Vo\r'\er, 1885.

? Distoiniim philocholum, Crepl., 1845.
Cestodes.

Diphyllobothrium stemmacephahim, Cobbold, 1858.

Nematodes.
Anisakis simplex (Rud., 1809).
Anisakis typica (Dies., i860).
Pseudalhis inflexus (Rud., 1809).
Stemirus minor (Kuhn, 1829).
Torymiriis convohitus (Kuhn, 1829).
Halocerciis inHexocaiidatus (v. Sieb., 1842).

Physeter catodon (Sperm whale, Cachalot).
Synonym : P. macrocephalus.

Cestodes.
Phyllobothrium physeteris (Dies., 1863) (? = -P. delphini (Rud.)).

Nematodes.

Anisakis catodontis, Baylis, 1929.

Anisakis physeteris, Baylis, 1923.
Acanthocephala.

Bolbosoma brevicolle (Malm, 1867).

Bolbosoma capitatum (v. Linst., 1880).

Platanista gangetica (Gangetic dolphin).

Trematodes.

Cyclorchis campnla (Cobbold, 1876).
Distomum andersoni, Cobbold, 1876.

WORMS PARASITIC IN CETACEA 413

Nematodes.

? Anisakis simplex (Rud., 1809).
Contracaeciim lobulatum (Schneider, 1866).
Ascaris delphini, Rud., 18 19.

Prodelphinus sp.

Nematodes.

Anisakis typica (Dies., i860).

Pseudorca crassidens (False Killer).
Acanthocephala.

Bolbosoma capitatiim (v. Linst., 1880).

Sotalia tucuxi.
Trematodes.

Aniphimerus laiicea (Dies., 1850).

Steno rostratus.

Synonym : Delphinus rostratus.

Cestodes.

Tetrabothrius forsteri (KrefFt, 187 1).
Nematodes.

Porrocaecum or Anisakis larvae [Baylis, 1929, 546].
Acanthocephala.

Bolbosoma capitatiim (v. Linst., 1880).

Tursiops truncatus (Bottle-nosed dolphin).
Synonym: T. tiirsio.
Trematodes.

Synthesium tursionis (Marchi, 1873).
Cestodes.

Phyllobothrium delpliini {Rud., 1819) {} = P . physeteris) .

? Phyllobothrium sp. (" Cysticercus ") (Carnot, 1822).

? Monorygma delpliini (Gervais, 1847).
Nematodes.

Stenunis ovatiis (v. Linst., 1910).

Crassicauda crassicauda (Crepl., 1829).

Ziphius cavirostris (Cuvier's whale).
Nematodes.

? Crassicauda boopis, Baylis, 1920.

BIBLIOGRAPHY

Ariola, v., 1900. Rivisione della famiglia Bothriocephalidae s.str. Arch. Parasitol., Paris, iii, 369-484,

pis. viii-x.
Baer, J. G., 1932. Contribution a Vetiide des Cestodes de Cetaces. Rev. Suisse Zool., xxxix, 195-228.
Bayl'is, H. a., 1916. On Crassicauda crassicauda {Ciepl.) [Nematoda] and its hosts. Ann. Mag. Nat. Hist.

(8), XVII, 144-8.

igig. A remarkable Cysticerciis from a rare dolphin. Ann. Mag. Nat. Hist. (9), in, 417-24-

1920 a. Observations on the genus Crassicauda. Ann. Mag. Nat. Hist. (9), v, 410-19.

ig20 b. On the classification of the Ascaridae. I. The systematic value of certain characters of the

alimentary canal. Parasitol., xii, 3, 253-64.

1922. Note on the habitat and structure of Crassicauda [Nematoda]. Parasitol., xiv, i, 9-12.

1923 a. An Ascaridfrom the Sperm-whale. Ann. Mag. Nat. Hist. (9), xi, 211-17.

1923 b. On Odontobius ceti, Roussel de Vauzeme, a nematode living on the baleen of whales. Ann. Mag.

Nat. Hist. (9), XII, 617-23.
1926 a. Some Tetrabothriid Cestodes from whales of the genus Balaenoptera. Journ. Linn. Soc. Lond.,

Zool., xxxvi, 161-72.
19266. Note on the occurrence 0/ " Cysticercus Taeniae Grimaldii" in a new host. Ann. Mag. Nat.

Hist. (9), xviii, 665-7.
1928. Note on Stenurus ovatus {v. Linstow), a little-known lung-worm of Cetacea. Ann. Mag. Nat.

Hist. (10), II, 464-6.

1929. Parasitic Nematoda and Acanthocephala collected in 1925-1927. Discovery Reports, I, 541-59.

Baylis, H. a., and Daubney, R., 1923. A further report on parasitic Nematodes in the collection of the

Zoological Survey of India. Rec. Ind. Mus., xxv, vi, 551-78.

1925. A revision of the lung-worms of Cetacea. Parasitol., xvii, 2, 201-16.

Beneden, p. J. VAN, 1858. Note sur une nouvelle espece de Distome, le geant de safamille, habitant le foie d'une

Baleine, nommee Distoma goliath v. Ben. Bull. Acad. roy. Sci. Belg. (2), v, 95-7, i pi.

1870. Les Cetach, leurs commensaux et leurs parasites. Bull. Acad. roy. Sci. Belg. (2), xxix, 347-68.

Bittner, H., and Sprehn, C, 1928. Trematodes: in Schulze, P., Biologic derTiere Deutschlands, Berlin,

Lief. 27,Teil 5, 1-133.
de Blainville, M. H. D., 1825. Note sur un Cetace echoue au Havre et sur un ver trouve dans sa graisse.

N. Bull. d. sc. Soc. philom., Paris, Sept., 139-40.
Borgstrom, E., 1892. Ueber Echinorhynchus turbinella, brevicoUis und porrigens. Bihang t. k. svenska

Vet.-Akad. Handl., Stockholm, xvii, 4, 60 pp., pis. i-v.
Bosc, L. A. G., 1802. Histoire naturelle des Vers, contenant lew Description et leurs Moeurs. [In Buffon,

Histoire Naturelle, ed. R. R. Castel (Suites)], 3 Tom., Paris.
Brandes, G., 1892. Revision der Monostomiden. Centralbl. f. Bakt. u. Parasitenk., xii, pp. 504-11.
Braun, M., 1898. Vermes. In Bronn, Klassen und Ordnungen d. Thier-Reichs, Leipzig, iv, Abt. i b.

Lief. 56-8, pp. 1535-614, pi. lix.

1900. Ueber Campula oblonga Cobb. Centralbl. f. Bakt. u. Parasitenk., Abt. I, xxviii, 249-54.

1901. Zur Revision der Trematoden der Vogel. II. Centralbl. f. Bakt. u. Parasitenk., Abt. i, xxix,

895-7-
1902. Ueber Distoma goliath P. jf. v. Ben., 1858. Centralbl. f. Bakt. u. Parasitenk., Abt. I, xxxii,

800-3, I pi.
Carus, L. v., 1885. Prodromus faunae Mediterraneae, etc. Vermes, i, 112-282. Stuttgart.
Cobb, N. A., 1888. Beitrdge sur Anatomic und Ontogenie der Nematoden. Jena. Zeitschr., xxiii, 41-76,

pis. iii-v.

■ 1889. Neue parasitische Nematoden. Arch. f. Naturg., LV, i, 149-59, pi. vii.

CoBBOLD, T. S., 1858. Observations on Entozoa, etc. Trans. Linn. Soc. Lond., xxii, 155-72, pis. xxxi-x.xxiii.
i860. Synopsis of the Distomidae. J. Proc. Linn. Soc. Lond., Zool., V, 1-56.

BIBLIOGRAPHY 41S

CoBBOLD, T. S., 1864. Entosoa: An introduction to the study of Helminthology . London.

1876. Tretnatode parasites from the dolphins of the Ganges, Platanista gangetica and Orcella brevirostris.

Journ. Linn. Soc. Lend., Zool., xiii, 35-46, pi. x.

1879. Parasites : A treatise on the Entozoa of man and animals . London.

CoLLETT, R., 1886. On the external characters of Rudolphi's Rorqual (Balaenoptera borealis). Proc. Zool.

Soc. Lond., 225-60.
Creplin, F. C. H., 1829. [Descriptions of nezv species of Filaria and Monostomum found in " Balaena

rostrata".] Verb. d. k. Leop.-Carol. Ak. d. Naturf. (Bonn), xiv, 2 Abt., 871-82, pi. Hi.

1845. Nachtrage zu Gurlt's Verzeichniss der Thiere, bei zvelchen Entozoen gefunden leorden sind. Arch.

f. Naturg., XI, i, 325-36.
1851. Ascaris angulivalvis, eine neue Spulwurmart, aus dem Schnabelwalfische. Arch. f. Naturg.,

XVII, i, 158-60.
DiESiNG, K. M., 1850. Systema Helminthum I. Wien.

• 1 85 1. Systema Helminthum 11. Wien.

. 1855. Sechzehn Gattungen von Binnemviirmern und ihre Arten. Denkschr. k. Akad. Wiss., Wien,

IX, I, 171-85, pis. i-vi.

1858. Revision der Myzhelminthen . Abtheilung: Trematoden. Sitz. k. Akad. Wiss., Wien, xxxii,

207-390, pis. i-ii.

1861. Revision der Nematoden. Sitz. k. Akad. Wiss., Wien, XLii, 595-736, i pi.

1863 a. Revision der Cephalocotyleen, Abth. Paramecocotyleen. Sitz. k. Akad. Wiss., Wien, XLVIII,

Abt. I, 200-345.
■ 1863 b. Revision der Cephalocotyleen. Abtheilung: Cyclocotyleen. Sitz. k. Akad. Wiss., Wien, xlix,

Abt. I, 357-430.
DujARDiN, F., 1845. Histoire naturelle des Helminthes. Paris.
FuHRMANN, O., 1898. tjber die Genera Prosthecocotyle Monticelli und Bothridiotaenia Lonnberg. {Vorldufige

Mittheilung.) Zool. Anz., xxi, 385-88.

1899 a. Z)ai Ge«;w Prosthecocotyle. Zool. Anz., xxii, 180-3.

• 1899 b. Das Genus Prosthecocotyle. Centralbl. f. Bakt. u. Parasitenk., Abt. i, xxv, 863-77.

Gervais, H., 1870. Sur les Entozoaires des Dauphins. Compt. rend. Acad. Sci., Paris, lxxi, 779.
Hamilton, J. E., 1914. [Report on Belmullet Whaling Station.] Brit. Assoc. Rep. (1913), 125-61.

1916. [Report on Belmullet Whaling Station.] Brit. Assoc. Rep. (1915), 124-46.

HoEPPLi, R., Hsu, H. F., and Wu, H. W., 1929. Helminthologische Beitrage aus Fukien und Chekiang.

Arch. f. Schiffs- u. Tropen-Hyg., xxxiii, Beiheft i, 43 pp., pis. i~xii.
Jagerskiold, L. a., 1891 a. Ueber den Bau des Ogmogaster plicatus, Creplin (Monostomum plicatum,

Creplin). K. Svensk. Vetensk. Ak. Handl., xxiv, 7, 32 pp., 2 pis.

1891 b. [Lonnherg, in] Einiges iiber die Schmarotzer der tiordatlantischem Balaenopteriden. Biol. Foren.

Forhandl., Stockholm, iii, 7, 127-34.

1893. Bidrag till Kcinnedojnen om Nematoderna. Diss., Stockholm. 86 pp., 5 pis.

1894. Beitrage ztir Kenntnis der Nematoden. Zool. Jahrb., Anat., vii, 449-532, pis. xxiv-xxviii.

Johnston, T. H., 1912. On a re-examination of the types of Krefffs species of Cestoda in the Australian

Museum, Sydney. Part i. Rec. Austral. Mus., ix, i, 1-35, pis. i-vi.
Johnston, T. H., and Deland, E. W., 1929. Australian Acanthocephala, No. i. Census of recorded hosts

and parasites. Trans. Roy. Soc. S. Australia, liii, 146-54.
JOYEUX, C, and Baer, J. G., 1931. Sur la presence du Nematode, Crassicauda crassicauda {Creplin, 1829)

chez un Dauphin des cotes de la France. Bull. Soc. Path, exot., xxiv, 198-203.
JoYEUX, C, and Dollfus, R. Ph., 193 i. Sur quelques Cestodes de la Collection du Musee de Miinich. Zool.

Jahrb., Syst., LXII, 109-18.
Kaiser, J. E., 1891-3. Die Acanthocephalen und ihre Entwickelung. In Leuckart and Chun, Bibliotheca

Zoologica, Cassel and Stuttgart, vii, 136 pp.
KosSACK, W., 1910. Neue Distomen. Centralbl. f. Bakt. u. Parasitenk., Abt. i, LVI, 114-20.
Krabbe, H., 1878. Saelernes og Tandhvalernes Spolorme. K. Dansk. Vidensk. Selsk. Forh., 43-51, pi. i.
Krefft, G., 1871. On Australian Entozoa, with descriptions of nezo species. Trans. Entom. Soc. N.S. Wales,

Sydney, 11, 206-32, pis. i-iii.

4i6

DISCOVERY REPORTS

KUHN, J., 1829 a. Description d'uiic nouvelle espece tie Strangle trouvee dans le Marsouin. Bull. Sci. Nat. et

Geol., XVII, 150-3.
1829 b. Anatomic comparee dc deux especes de Strongylus, qui vivent dans le Marsouin; par M. Raspail.

Bull. Sci. Nat. et Geol., xix, 139-43.
1829 c [1830]. Description d'un nouveau genre de I'ordre des Douves, et de deux especes de Strongles.

Mem. Mus. Hist. Nat., Paris, xviii, 357-68, pi. xvii bis, issued with vol. xix.
Laennec, R. T. H., 1804. Menwire sur les vers vesiculaires, et principalonent sur ceux qui se trouvent dans le

corps liumain. 176 pp., 3 pis. [n.p.].
Leidy, J., 1891. Notices of Entosoa. Proc. Acad. Nat. Sci. Phila. (3), xx, 410-18.
Leiper, R. T., and Atkinson, E. L., 1914. Hehninthes of the British Antarctic Expedition, 1910-1913. Proc.

Zool. See. Lond., 222-6.

1915. Parasitic Worms: British Antarctic (" Terra Nova ") Exp., 1910. Natural History Report, Zoology,

II, 3, 19-60, pis. i-v.
Leuckart, F. S., 1828. Breves animalium quorundam tnaxima ex parte marinorum descriptiunes, etc. Heidel-

bergae.
Leuckart, K. G. F. R., 1848. Beschreibung sweier neuen Helminthen. Arch. f. Naturg., xiv, i, 26-9, pi. ii.
LiNSTOW, O. VON, 1880. Hehninthologische Untersiichungen. Arch. f. Naturg., xlvi, i, 41-54, pi. iii.

1888 a [? 18S9]. Helminthologisches. Arch. f. Naturg., Liv, 235-46, pi. xvi.

1888 b. On Pseudalius alatus Leuckart, collected by Mr Robert Gray in the Arctic Seas, and other

species of the genus. Proc. Roy. Soc. Edinburgh, xvi, 15-17.

1896. Nemathelminthcn. Hamburg. Magalhaensische Sammelreise, Hamburg, 21 pp., i pi.

1906. Ostpreussische Nematoden. Schrift. Physik.-Okon. Ges., Konigsberg, XLVii, 111-14, i pi.

1907. Ascaris lobulata, Schneider, ein Parasit des Darms von Platanista gangetica. Journ. and Proc.

Asiat. Soc. Bengal, in, 37-8.

1910. Pseudalius ovatus n.sp. Centralbl. f. Bakt. u. Parasitenk., Abt. i, LV, 133-5.

Linton, E., 1905. Notes on Cestode cysts. Taenia chamissonii, new species, from a porpoise. Proc. U.S. Nat.
Mus., XXVIII, 819-22, pi. XXXV.

1923. A new Cetacean Cestode (Prosthecocotyle monticellii, sp.nov.), with a note on the genus Tetra-

bothrius Rudolphi. Journ. Parasitol., x, i, 51-5, pi. 5.
LONNBERG, E., 1892. Anatomische Studien iiber Skandinarische Cestoden, II. Zwei Parasiten aiis Walfischen

iind swei aus Lamna cornubica. Kongl. Svensk. Vetensk.-Akad. Handl., xxiv, 16, 1-30, i pi.
Looss, A., 1885. Beitrdge sur Kenntnis der Trematodcn, etc. Zeitschr. f. wiss. Zool., xli, 390-446,

pi. xxiii.

1899. Weitere Beitrdge sur Kemitniss der Trematoden-Eauna Aegyptens, etc. Zool. Jahrb., Syst.,Xli,

521-784, pis. xxiv-xxxii.

1901. Natura doceri, eine Erkldnmg und Begriindung einiger Grundsdtse, etc. Centralbl. f. Bakt. u.

Parasitenk., Abt. i, xxix, 191-210.
1902. Ueber neue und bekannte Trematoden aus Seeschildkroten. Nebst Erorterungen sur Systematik und

Nomenclatur. Zool. Jahrb., Syst., xvi, 411-894, pis. xxi-xxxii.
Luhe, M., 1899. -^"'' Anatomic und Systematik der Bothriocephaliden. Verb. d. deutsch. Zool. Ges., Leipzig,

30-55-

1904-5. Geschichte und Ergebnisse der Echinorhynchen-Forschung bis auf Westrumb (1821), etc. Zool.

Ann., Wiirzb., i, 139-353.
1908. Zur Systematik und Faunistik der Distomen. I. Die Gattung Metorchis Looss, nebst Bemer-

kungen iiber die Familic Opisthorchiidae. Centralbl. f. Bakt. u. Parasitenk., Abt. I, XLViii, 428-36.
Malm, A. W., 1867. Monographic iUustree du Baleinoptere trouve le 29 Octobre 1865 sur la cote occidentale de

Suede. Stockholm, no pp., 20 pis.
Marchi, p., 1873. Sopra una specie nuova di Distomum trovata nelle intestina del Delphinus tursio. Atti

Soc. ital. di Sc. nat., Milano, xv, 304, pi. v, fig. B.
Meggitt, F. J., 1924. The Cestodes of Mammals. London.

Meyer, A., 193 1. Die Acanthocephalen des Arktischen Gebietes. Fauna Arctica, vi, i, 9-20.
Molin, R., 1858. Versuch einer Monographic der Filarien. Sitz. K. Akad. Wiss., Wien, xxviii, 365-461,

pis. i-ii.

BIBLIOGRAPHY 417

MONIEZ, R., 1880. Essai monographiquc sur les Cysticerques . Trav. de I'lnst. Zool. tie Lille, Paris, in,

Fasc. I, 1-190, pis. i-iii.
• 1889. Sur la lurvc dii Taenia Grimaldii, tiov. sp., parasite dit Daiipliirt. Compt. rend. Acad. Sci.,

Paris, cix, 825.
MONTICELLI, F. S., 1889. Elenco degli elminti raccoUi dal Capitaiio G. Chierchia durante il viaggio di

circumnavigasioiw del/a r. corvetta " I'ettor Pisani". Boll. Soc. Nat. Napoli (i), iii, 67-71.

1892 a. Nota intorno a due forme di Ces/odi. Boll. Mus. Zool. Torino, vii, 127, 1-9, i pi.

1892 b. Studii sui Trematodi endoparassiti Monostonuim cymbiuni Diesing. Mem. r. Accad. Sci.

Torino (2), XLll, 683-727, i pi.

1893. Studii sui Trematodi endoparassiti: Prima contributo di osservasioni sui Distomidi. Zool. Jahrb.,

Suppl. 3, 229 pp., pis. i-viii.
NicoLL, W., 1923. A reference list of the Tremaiode parasites of British Mammals. Parasitol., xv, 3, 236-52.
Nybelin, O., 1922. Anatomisch-Systematische Studieniiher Pseudophyllideen. Kungl. Vetensk.- och Vitterh.-

Samh. Handl., Goteborg (4), xxvi, i, pp. iv, 1-228.

• 1928. Zicei neue Cestoden aus Bartenwalen. Zool. Anz., Lxxviii, 11/12, 309-14.

Odhner, T., 1905. Die Trematoden des arktischcn Gebietes. In Romer and Schaudinn, Fauna Arctica, iv,

2, 291-372, pis. ii-iv.

1914- ^"^ Verwandtschaftsbeziehungen der Trematodeiigatiung Paragonimus Brn. Zool. Bidrag fran

Uppsala, III, 231-46.
PoiRiER, J., 1885. Contribution a Vhistoire des Trematodcs. Arch. Zool. exp. et gen. (2), ill, 465-624,

pis. xxiii-xxxiv.

■ 1886. Trematodes nouveaux ou peu contnis. Bull. Soc. Philomat. Paris (7), x, 20-40, pis. i-iv.

Porta, A., 1905. Gli echinorinchi dei pesci. Arch. Zool., Napoli, 11, 2, 149-214, pis. x-xii.

— — 1906. Ricerche anatotniche sulV Echinorhynchus capitatus v. Linst., e note sulla sistematica degli

echinorinchi dei Cetacei. Zool. Anz., xxx (8-9), 235-71.

■ 1908. Gli Acantocefuli dei Mammiferi. Arch. Parasitol., Paris, xii, 2, 268-82.

■ — — 1909. Gli Acantocifali dei Mammiferi. Arch. Zool., Napoli, iv, 239-85, pi. v.

Price, E. W., 1932. The Trematode parasites of tnarine Mammals. Proc. U.S. Nat. Mus. lxxxi. Art. 13,

1-68, pis. i-xii.
Raspail [F.-V.], 1829. Anatomic comparee de deux especes de Strongylus quivivent dans le Delphinus phocena.

Ann. Sci. d'Obs., Paris, 11, 244-54, P'^- vii-viii.
Risso, A., 1826. Histoire naturelle des principales productions de VEurope meridionale, etc., v. Paris.
RoussEL de Vauz^me, 1834. Note sur /'Odontobius ceti de I'ordre des intestinaux cavitaires. Ann. Sci. Nat.,

Paris (2), I, 326-33, pi. ix.
RuDOLPHi, C. A., 1809. Entosoorum sive Vermium Intestinalium Ilistoria Naturalis, 11, i, 457 pp., pis.

vii-xii. Amstelaedami.

1810. Entozoorum sive Vermium Intestinalium Historia Naturalis, 11, 2, xii+386 pp. Amstelaedami.

1814. Erster Nachtrag zn meiner Naturgeschichte der Eingezveidewiirmer. Mag. f. d. n. Entdeck. in d.

ges. Naturk., Berlin, vi, 2, 83-113.

18 19. Entozoorum Synopsis. Berlin.

Schneider, A., 1866. Monographic der Nematoden. Berlin.

Scott, T., 1909. Some notes on fish parasites. Ann. Rep. Fishery Board for Scotland, xxvi. Part ill, 73-92,

pis. iii-vii.
Shipley, A. E., 1899. Notes on the species of Echinorhynchus parasitic in the Cetacea. Arch. Parasitol.,

Paris, II, 2, 262-9.
Siebold, C. von, 1842. Bericht tiber die Leistungen im Gebiete der Helminthologie ivahrend des Jahres 1841.

Arch. f. Naturg., viii, ii, 338-72.
Southwell, T., 1925. A monograph on the Tetraphyllidea, with notes on related Cestodes. Liverpool School

of Tropical Medicine Memoir (New Series), No. 2, pp. i-xv, 1-368.
Southwell, T., and Prashad, B., 1920. A revision of the Indian species of the ge?ius Phyllobothrium. Rec.

Ind. Mus., XIX, I, 1-8, pi. i.
Spaul, E. a., 1926. Crassicauda bennetti, sp.n., a new Nematode parasite from the Bottle-nosed whale

(Hyperoodon). Ann. Mag. Nat. Hist. (9), xvii, 581-5.

4i8

DISCOVERY REPORTS

Stiles, C. W., 1895. The tinatomy of the large American Fluke (Fasciola magna), etc. J. Comp. M. and Vet.

Arch., N.Y., xvi, pp. 139-47, pis. v-vi, pp. 213-22, pis. vii-viii, pp. 277-82.
Stiles, C. W., and Hassall, A., 1899. Internal parasites of the Fur seal. (In "The Fur seals and Fur seal

Islands of the North Pacific Ocean.") Report on Fur seal Investigations, Washington, Pt. 3,

99-177-
1908. Inde.x-catalogue of medical and veterinary zoology. Subjects: Trematoda and Trcmatode

diseases. U.S. Pub. Health and Mar.-Hosp. Serv., Hyg. Lab. Bull., 37, 398 pp.

1912. Index-catalogue of medical and veterinary zoology. Subjects: Cestoda ami Cestodaria.

U.S. Pub. Health and Mar.-Hosp. Serv., Hyg. Lab. Bull., 85, 467 pp.

1920. Index-catalogue of medical and veterinary zoology. Subjects: Round-worms. U.S. Pub.

Health Serv., Hyg. Lab. Bull., 114, S81 pp.
Stossich, M., 1892. Distomi dei Mammiferi. Program, d. civ. scuola r. sup., Trieste, 42 pp.

1896. // Genere Ascaris Linne. Lavoro Monografico. Boll. Soc. Adriat. Sci. Nat., Trieste, xvii, 9-120.

1898. Note parassitologiche. Boll. Soc. Adriat. Sci. Nat., Trieste, xviii, i-io, pis. i-ii.

1Q02. Sopra alcuni Nematodi della CoUezione elmintologica del Prof. Dott. Corrado Parona. Boll. Mus.

Zool., Geneva, 16 pp., pis. iii-v. (Also Atti Soc. Ligust., Genova, xiii, 61-76, pis. iii-v.)
1904. Alcuni distomi della CoUezione elmintologica del Museo Zoologico di Napoli. Ann. Mus. Zool.

Napoli, U.S., I, 1-14, pi. ii.
Stunkard, H. W., and Alvey, C. H., 1930. The inorphology o/Zalophotrema hepaticum, with a reviezv of the

Trematode family Fasciolidae. Parasitol., xxn, 326-33, pi. xxxiii.
Van Cleave, H. J., 1924. A critical study of the Acanthocephala described and identified by Joseph Leidy.

Proc. Acad. Nat. Sci. Phila., lxxvi, 279-334, pis. xiv-xviii.

1925. Acanthocephala from Japan. Parasitol., xvii, 149-56.

Wesenberg-Lund, E., 1926. Acanthocephaler . Meddelelser om Gronland, Copenhagen, Pt 23, Suppl. xxi,

143-55- .
Weski, O., 1900. Mitteilungen iiber Distomuni lancea Dies. Centralbl. t. Bakt. u. Parasitenk., Abt. i,

XXVII, 579-83.
Westrumb, a. H. L., 1821. De helminthibus acanthocephalis [etc]. Hanoverae.
Wu, Hsien Wen, 1929. On Halocercus pingi n.sp. a Lung-worm from the porpoise, Neomeris phocoenoides.

Journ. Parasitol., xv, 4, 276-9.

DISCOVERY
REPORTS

Issued by the Discovery Committee, Colonial Ojjice, London
on behalf of the Qovemment of the Dependencies of the Falkland Islands

Vol. VI, pp. i-vi

TITLE-PAGE AND LIST OF CONTENTS

CAMBRIDGE

AT THE UNIVERSITY PRESS
1932

Price ninepence net

^

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

MacmiUan

Tokyo
Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VI, pp. 1-138, text-figs. 1-75

Issued by the Discovery Committee, Colonial Office, London
on hehalf of the Qovemment of the Dependencies of the Falkland Islands

PYCNOGONIDA

Isabella Gordon, D.Sc, Ph.D.

CAMBRIDGE

AT THE UNIVERSITY PRESS

1932

Price fifteen shillings net

'4

^

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

WALTER

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VI, pp. 139-164, plates I-VI

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the FalUand Islands

REPORT ON PENGUIN EMBRYOS

COLLECTED DURING THE

DISCOVERY INVESTIGATIONS

C. W. Parsons, B.A.

CAMBRIDGE

AT THE UNIVERSITY PRESS

1932

Price nine shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

MacmiUan

Tokyo
Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

WALTER ttia^tSy LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VI, pp. 165-190, plates VII-XLIl

Issued by the Discovery Committee, Colonial Office, London
behalf of the Qovemment of the Dependencies of the Falldand Islands

on

ON THE DISTRIBUTION AND MOVEMENTS

OF WHALES ON THE SOUTH GEORGIA AND

SOUTH SHETLAND WHALING GROUNDS

h

Stanley Kemp, Sc.D., F.R.S., and A. G. Bennett

CAMBRIDGE

AT THE UNIVERSITY PRESS

1932
Price eight shillings net

Cambridge University Press

Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

WALTER

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VI, pp. 191-204, plates XLIII, XLIV

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the Falkland Islands

ON THE DEVELOPMENT OF
CEPHALODISCUS

C. C. John, M.A., D.Sc, D.I.C.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1932

Price four shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

■;

t

PRINTED

I N GREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VI, pp. 205-236, plates XLV-XLVII

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the Falkland Islands

REPORT ON SOUNDINGS TAKEN DURING
THE DISCOVERY INVESTIGATIONS,

1926-1932

H. F. P. Herdman, M.Sc.
{with 7 charts in separate chart-case)

CAMBRIDGE

AT THE UNIVERSITY PRESS
1932

Price fifteen shillings net

Cambridge University Press
Fetter Lane, London

Neuj York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo
Maruzen Company, Ltd

All rights reserved

PRINTED

IN GREAT BRITAIN

BY

WALTER

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VI, pp. 237-392, plates XLVIII-LVII

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the Falkland Islands

SPONGES

Maurice Burton, M.Sc.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1932
Price fwenty shillings net

y^

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

«

PRI NTED

I N GREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VI, pp. 393-418

Issued Iry the Discovery Committee, Colonial Office, London
hehalf of the Qovemment of the Dependencies of the Falkland Islands

on

A LIST OF WORMS PARASITIC
IN CETACEA

H. A. Baylis, M.A., D.Sc.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1932

Price two shillings and sixpence net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRINTED

I N GREAT BRITAIN

BY

WALTER

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

'^-^4

V^%

J-r .■' t

'T "^ 'V.'*'*

.. ^'f^^v

^•'"r ^

W -^N^^-'

/V-'^'V

^f%- 'C

V 'V/

■\ t t^

l»i*^'^l

M^r-'^

■•4..>-f'i-

:.<,Xt , .