■ □

WHO

m

LO

a

D

S^^

□

D

a

1

DISCOVERY REPORTS

VOLUME VII

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

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 VII

CAMBRIDGE

AT THE UNIVERSITY PRESS
1933

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

\A/0C50S
MC3L-E.

CONTENTS

FOSSIL FORAMINIFERA FROM THE BURDWOOD BANK AND THEIR
GEOLOGICAL SIGNIFICANCE (published 27th February, 1933)
By W. A. Macfadyen, M.C., M.A., Ph.D., F.G.S.

Description of Samples P'^S'^ 3

Previous records of Fossil Foraminifera from the Patagonia-Graham Land Region

Age of the Burdwood Bank Fauna

Correlation of the Regional Stratigraphy

Conclusion

Summary

5

7

13

H

15

FAECAL PELLETS FROM MARINE DEPOSITS (published 8th March, 1933)
By Hilary B. Moore, B.Sc

FORAMINIFERA. PART II, SOUTH GEORGIA (published 27th June, 1933)
By Arthur Earland, F.R.M.S.

Introductory Note

Previous Work in the Area

Material Examined

List of Stations

List of New Genera, Species and Varieties

Systematic Account

Appendix

Supplementary Bibliography

Index

page 17

page 29
29

30

32

44

45

132

133

134

Plates I-VII following page it,^

ON VERTICAL CIRCULATION IN THE OCEAN DUE TO THE ACTION OF
THE WIND WITH APPLICATION TO CONDITIONS WITHIN THE
ANTARCTIC CIRCUMPOLAR CURRENT (published 14th November, 1933)
By H. U. Sverdrup - page i39

A GENERAL ACCOUNT OF THE HYDROLOGY OF THE SOUTH ATLANTIC
OCEAN (published 23rd November, 1933)
By G. E. R. Deacon, B.Sc.

Introduction page 173

The Surface Waters of the South Atlantic Ocean

Antarctic Surface Water .....

Sub-Antarctic Water

Sub-Tropical and Tropical Waters
The Deep Waters of the South Atlantic Ocean

Antarctic Intermediate Water ....

Warm Deep Water ......

Antarctic Bottom Water .....

Distribution of Phosphate and Nitrate .
Appendix. The Winds of the Atlantic Ocean South of 40° S, by Lieut. R

R.N.R

List of Literature

Plates VIII-X

. A. B. Ardley,

235

237

folloioing page 238

173
206

216

221
226
231

233

46108

CONTENTS

WHALING IN THE DOMINION OF NEW ZEALAND (published 1 8th December,
1933)
By F. D. Ommanney, A.R.C.S., B.Sc.

Introduction page 241

History

Modern Whaling in New Zealand
List of Literature . . . .
Plates XI-XIII . . . .

241

. 247

• 252

following page 252

ISOPOD CRUSTACEA. PART I, THE FAMILY SEROLIDAE (pubHshed 21st Decem-
ber, 1933)
By Edith M. Sheppard, M.Sc.

Introduction page 255

List of Species 256

Station List 256

Geographical Distribution 264

Classification 267

General Morphology 271

Key to all Known Species 278

Description of Specie? 282

• List of References 360

Plate XIV ............. following page T^bz

SOME ASPECTS OF RESPIRATION IN BLUE AND FIN WHALES (published
2 1 St December, 1933)
By Alec H. Laurie, M.A.

Introduction page 365

Southern Whales and their Environment

Gas Analyses .

Summary .

Literature Cited

Appendix .

Plate XV

365

374

399
400

400

following page 406

[Discovery Reports, Vol. VII, pp. 1-16, February, 1933.]

FOSSIL FORAMINIFERA FROM THE

BURDWOOD BANK AND THEIR

GEOLOGICAL SIGNIFICANCE

By

W. A. MACFADYEN, M.C., M.A., Ph.D., F.G.S.
Sedgwick Museum, Cambridge

CONTENTS

Description of samples page 3

Previous records of fossil Foraminifera from the Patagonia-Graham Land

region S

Age of the Burdwood Bank fauna 7

Correlation of the Regional Stratigraphy 13

Conclusion 14

Summary iS

FOSSIL FORAMINIFERA FROM THE

BURDWOOD BANK AND THEIR

GEOLOGICAL SIGNIFICANCE

By W. A. Macfadyen, M.C, M.A., Ph.D., f.g.S.
Sedgwick Museum, Cambridge

(Text-figs. 1,2.)

FOSSIL Foraminifera were recognized in bottom deposits from several stations of
the R.R.S. 'Discovery 11' and the R.R.S. ' WiUiam Scoresby'\ and in the present
investigation samples dredged from three stations on the northern part of the
Burdwood Bank were examined.

DESCRIPTION OF SAMPLES

I. A dredging at St. WS 87 in lat. 54° yi' S, long. 58° 16' W, dated 3. iv. 1927,
came from a depth of 96 m. It yielded a bottom sample comprising sand, with recent
shells and Foraminifera, etc., together with many fragments of shale, and loose fossil
Foraminifera.

The shale all consisted of rounded pebbles, the largest of which measured
7x6x3 cm. The larger pebbles were partly stained by a brown surface coloration
and were pierced by borings, up to 6 mm. in diameter, of present-day marine organisms ;
several recent specimens of Polyzoa and Foraminifera were adherent to their surface.
The shale was of two kinds, both being rather hard, but showing no sign of crushing
or thermal metamorphism. Most was of a green-grey colour, highly glauconitic (best
seen at the eroded surface where the glauconite grains were left prominently exposed),
and very finely sandy. Traces of the original bedding were visible in some of the pebbles.
The less common type consisted of slightly sandy micaceous shale, light grey when
dried, and of exceedingly fine texture, without glauconite. There was a single pebble,
2-5 cm. long, of rather soft grey argillaceous limestone.

The loose Foraminifera examined numbered about no and were in many cases
badly crushed and distorted, while some were preserved as calcite casts. Haplophrag-
tnoides subglobosiis and species of Cyclammina were the commonest forms. Mr Earland
found a number of the specimens specifically and some generically indeterminable, a
conclusion I fully share. Thirteen, all of arenaceous species, had traces of green finely
sandy shale adhering to them.

The two kinds of shale were washed separately for Foraminifera. Nearly all of those
obtained came from the fine light grey shale, with Spiroplectammina spectabilis and

1 See Heron-Allen, E., and Earland, A., 1932, Foraminifera. Part I, The Ice-free Area of the Falkland
Islands and Adjacent Seas, Discovery Reports, iv, pp. 297-8, Cambridge.

DISCOVERY REPORTS

Globigerina biilloides as the most frequent forms. The glauconitic shale was practically
barren. The specimens, sixty-eight of which were mounted, were in general not crushed,
and many were filled with clear calcite. Of the following list, therefore, which is
largely Mr Earland's work, many species are quite well preserved, whilst others are yet
nameable, though sometimes doubtfully.

Fossil Foramiiiifera idejitified

Bathysiphon sp.

Ammodisats incertus (d'Orbigny), A and B forms

Glomospira charoides (Jones and Parker)

G. gordialis (Jones and Parker)

Hormosina globidifera , Brady

Haplophragmoides acutidorsatus (Hantken)?

H. aff. crassimargo (Norman)

H. stibglobosiis (Sars)

H. coronatus (Brady)

Trochammina squamata, Jones and Parker

T. globigerinifuimis (Parker and Jones)

Cyclammina cancellata, Brady

C. orbicularis, Brady

C. bradyi, Cushman

C. elegam, Cushman, A and B forms

Rzehakina epigona (Rzehak)

Spiropkctammina spectabilis (Grzybowski),A and ?B forms

Pseudotextidaria globidosa (Ehrenberg)

Bolivina punctata, d'Orbigny

Bulimina pupoides, d'Orbigny

B. ovata, d'Orbigny

Allomorphina cretacea, Reuss?

Nodosaria ambigua, Neugeboren

A^. linibata, d'Orbigny

N. communis (d'Orbigny)?

Cristellaria rotulata (Lamarck)?

Globulina gibba, var. globosa, Miinster

EUipsopleurostomeUa sp.

ValvuUneria allomorpliinoides (Reuss)

Gyroidina nitida (Reuss)

Globigerina buUoides, d'Orbigny

G. cretacea, d'Orbigny

Pulletiia sphaeroides (d'Orbigny)

Anomalitia ammonoides (Reuss)

Geological

range

'

■ ^

St.WS

87

St. 719

St. 720

3
0

cd
*->

u

.2

u

i)

60

a

L

S

L

s

L

S

>

4

I

21

...

I

X

X

X

4

I

X

X

X

I

X

X

X

2

X

X

X

2

...

X

X

...

I

• • •

. .>

X

28

I

...

...

X

X

X

3

...

...

X

X

X

I

...

X

X

I

.. .

I

...

X

X

X

10

16

...

I

...

X

X

12
I

4
6

...

...

X

X
X

I

...

36

...

X
X

'?

...

I
I

I

I

• . .

4

X
X

?

X

X

I

...

...

...

...

X

X

2

...

...

...

X

X

X

I

...

...

X

...

...

I

...

...

...

X

X

I

4

I

I

X
X
X

X
X
X

X
X
X

I
...

I

3
18?

...

...

...

X
X
X

X

X

4

4

I

::;

X
X
X

X
X
X

X

X
X

L = found loose; S = washed from shale. The figures in the table indicate the number of specimens
mounted.

The specimens are preserved in the Heron-Allen and Earland Collection in the British Museum
(Natural History), slides No. TS 525 (i)-TS 525 (5).

FOSSIL FORAMINIFERA FROM THE BURDWOOD BANK 5

In view of the records of Radiolaria from Cretaceous strata of this region, noted
below, it is of interest and importance to record finding in the residue from the grey
shale four specimens of a radiolarian species. In shape it suggests a slightly spinose
form of the foraminifer Orbulina.

2. A dredging at St. 719 in lat. 54° 00' S, long. 60° 00' W, dated 13. xi. 1931, came
from a depth of 108 m. It consisted of greenish grey sand with recent shell frag-
ments, Foraminifera, etc. ; many fragments of greenish argillaceous limestone up to
9x6x5 cm., and bored by organisms; some smaller fragments of green-grey shale,
brownish when dry; and many well-rounded pebbles^ up to 6x5x3 cm., and to
which recent organisms were adherent.

The shale was separated and washed, but yielded very few fossil Foraminifera, and
five specimens of Radiolaria. The loose sand, however, yielded many fossil Foraminifera,
some 125 of which were mounted: the most frequent forms included species of
Cyclammina, particularly C. elegans, up to 3-1 mm. in diameter, Ammodtscus, and
Bathysiphon fragments.

3. A dredging at St. 720 in lat. 53° 58' S, long. 61° io\' W, dated 13. xi. 1931, came
from a depth of 140 m. It consisted of greenish grey sand with many recent shell
fragments, amongst which those of brachiopods were common, echinoid spines,
Foraminifera, etc. ; a few small fragments of bored greenish argillaceous limestone up
to 2-5 X 2 X I cm. ; some greenish shale, in part very glauconitic; and a small quantity
of well-rounded pebbles up to 2 x 2 x 1-5 cm.

The shale, separated and washed, yielded hardly any fossil Foraminifera. Four
specimens of Radiolaria were picked out. The loose sand likewise yielded only a very
few rather small fossil Foraminifera.

From the abundance of the smaller fragments of shale in all three samples, and of
limestone, particularly at St. 719, it seems clear that the beds must outcrop on the sea-
bottom at or close to the stations, and that no adventitious origin will explain the
occurrence. The well-rounded igneous pebbles appear to have been washed out of
a pebble bed, since such rounding of this material is not to be explained by movement
on the sea-bottom, but is perhaps the result of beach conditions. No trace of macro-
scopic fossils was found in any of the three samples.

PREVIOUS RECORDS OF FOSSIL FORAMINIFERA FROM
THE PATAGONIA-GRAHAM LAND REGION

Hyades, writing in 1887,'^ records (p. 124) numerous Foraminifera without further
identification in sections of a "schiste argileux" from the coast of Cape Webley,
Ponsonby Bay (between Tierra del Fuego and Cape Horn).^ On p. 130 he records from
Button Island, Ponsonby Bay, "schistes bleuatres feuilletes, semblables a des schistes
ardoisiers ". He continues : " Mais ce qui rend cette roche interessante, c'est I'abondance

1 See note on p. 16 below.

2 1887, Miss. Set. du Cap Horn, 1882-3, iv, Geologie. Paris.
^ See Fig. 2 below (p. 13).

1-3

6 DISCOVERY REPORTS

des squelettes quartzeux de foraminiferes qu'elle renferme en tres grande quantlte.
La structure intime de ces foraminiferes a ete effacee dans I'acte de la fossilisation.
Cependant les contours sont assez nettement accuses pour qu'on puisse y reconnaitre
des formes anciennes (carboniferes ou permiennes) rappelant celles des Textularia ou
des ClimacMnmina. Nous ne pensons pas qu'une determination paleontologique aussi
incertaine permette d'arriver a des conclusions nettes relativement a I'age de ces roches ".
This record has apparently often been referred to, as Richter notes, but without
Hyades's final cautionary sentence. Its value for dating the strata is, however,
negligible.

M. Richter,! records (p. 535) from Staten Island and New Year Island a few Fora-
minifera from Inoceranms-htds of somewhat metamorphosed shales and limestones rich
in Radiolaria. He figures " Wligosteghia laevigata, Kaufmann" and Globigerina sp.
from Cape Conway, Staten Island. From the Argentine-Chile border in the middle of
Tierra del Fuego he records (p. 537) a reddish grey dirty limestone rich in Foraminifera,
whose chambers are mostly filled with light green glauconite. With them are a few
Radiolaria and many macroscopic fossils. " Wligostegina laevigata, Kaufmann" is
figured from here. From the same locality he also records (p. 542) Cristellaria rotulata
(Lmk.), and notes as seen in section Cristellaria, abundant Globigerina, and also Textu-
laria, Nodosaria and " Wligostegina laevigata, Kaufmann". Richter considers the age
of these beds to be Albian on the ground of the occurrence in them of Aucellina.
Specimens from Mount Tarn on the Brunswick Peninsula, Patagonia, consisted of a
finely sandy grey limestone with pyrites, glauconite and mica (p. 537), and with
Foraminifera, particularly ''Globigerina cf. cretacea, d'Orb." which is figured. The
chambers of the tests are filled, so he says, with amorphous silica. He concludes (p. 564)
that in South Patagonia and Tierra del Fuego there have been found fossils possibly of
Upper Jurassic age ; certainly of Lower Cretaceous age ; and of the Upper Cretaceous,
Albian, and particularly Upper Senonian, the Inocerarmis steinmamii-heds.

His Globigerina cf. cretacea, appears from the figures to be the true G. cretacea, but
his Oligostegina laevigata is, as was Kaufmann's original form, too doubtful to
identify.

R. Holland^ had material from the islands east of Trinity Peninsula, Graham Land.
From the Senonian of Seymour Island (or Snow Hill Island, there is doubt as to precise
locality but not, apparently, as to age) he had four specimens, which he described and
figured as two new species, Ammodiscus grandis (one specimen) and Trochammina
cretacea (three specimens). The latter, however, appears to be a Haplophragmoides.

From the ? Pliocene Pec/ew-conglomerate of Cockburn Island Holland records and
figures eleven species. Of the 304 specimens submitted to him five-sixths were Cassi-
dulina crassa. The full list he identified as follows :

1 1925, Beitriige zur Kenntnis der Kreide in Feuerland, Neuesjahrb. Min., hu, Beil-Bd., Abt. B, pp. 524-
68, pis. vi-ix.

2 1910, The Fossil Foraminifera, pp. i-ii, pis. i, ii, Wiss. Ergeb. Schwedisch. Siidpolar-Exped. 1901-3,
ni, Lieferung 9, 410. Stockholm.

FOSSIL FORAMINIFERA FROM THE BURDWOOD BANK

Biloculina ringens (Lamarck)
B. elongata, d'Orbigny
Miliolina grata (Terquem)
CassiduUna crassa, d'Orbigny
Lagena globosa (Montagu)
Cristellaria gibba, d'Orbigny

Polymorph ina gutta, d'Orbigny
Truncatulina refulgens (Montfort)
T. lobatula (Walker)
T. ungeriana (d'Orbigny)
Rotalia beccarii (Linne)

AGE OF THE BURDWOOD BANK FAUNA

Of the fossil Foraminifera under discussion many have no value by themselves as age
markers, since they range from the Cretaceous to the present day. Four species, how-
ever, Rzehakina epigona, Spiroplectammma spectabilis, Pseudotextidaria globulosa and

E, T. Talbot del.

m

Fig. I.

a, b. Rzehakina epigona (Rzehak), / 37; St. WS 87, loose.

c, d. Spiroplectammina spectabilis (Grzybowski), form ?B, x 37; St. WS 87, shale. The early part of the
test is difficult to see clearly, and the initial chamber may have been over-emphasized in the figure.
Alternatively the specimen may be the Aj form.
e. Nodosaria lijnbata, d'Orbigny, x 37; St. WS 87, loose.
f,g,h. Gyroidina nitida (Reuss), x 37; St. WS 87, shale.

ij. Spiroplectammina spectabilis (Grzybowski), form A, x 37; St. WS 87, shale.
k, I. Pseudotextidaria globulosa (Ehrenberg), x 37; St. 87, loose.
?n, n. Cyclammina elegans, Cushman, form B, x 18; St. 719, loose.
The age of these specimens is probably Upper Cretaceous (Senonian).

8 DISCOVERY REPORTS

Cyclammina elegans, stand out, according to the literature, as a definite indication of an
uppermost Cretaceous age. In the case of Rzehakina and Pseiidotextularia (or Gum-
belina, which name I take as a synonym) the genus alone would appear adequate for this
age determination. Nodosaria limbata and Gyroidina nitida are also typically Upper
Cretaceous species, but it is not quite clear that they are confined to strata of this age.

Pseiidotexttdaria globulosa (Ehrenberg) was described from the Upper Cretaceous of
various countries.^ It has been recorded from Tertiary, Quaternary and even present
day material, but in such cases I believe that it has generally occurred as a derived
Cretaceous fossil, a point which has frequently not been made clear. It is often very
abundant in Upper Cretaceous strata, and Cushman has frequently stated that in
the American occurrences so far known it has not been found above the top of the
Cretaceous.^

Spiroplectammina spectabilis (Grzybowski) was described from the Upper Cretaceous
of Krosna, Poland, under the generic name Spiroplecta.^ It has page priority over the
megalospheric form,* described separately under the name Spiroplecta brevis, Grzy-
bowski,^ which is thus a synonym. Later, a rather broader microspheric form of
what appears to be the same species was described under the name Spiroplecta
clotho, Grzybowski,^ also from the Upper Cretaceous of Poland. I am indebted
to Mr Earland, who had a translation made from the Polish text, for the information
that Grzybowski states that all three of the above species are "siliceous with a smooth
surface".

The Burdwood Bank specimens are entirely siliceous ; the wall of the test is fairly
smooth, but is clearly composed of fine sand grains set in a large excess of cement.

Galloway and Morrey describe and figure a form under the name Spiroplectammina
rosula (Ehrenberg) from the Upper Cretaceous of Mexico," which seems to be the same
as the megalospheric form of the Burdwood Bank species. They state that " The wall is
now entirely siliceous, the surface is rough and granular, and there is little doubt that
the wall was originally arenaceous".

The species does not appear to be recorded from other than Upper Cretaceous
strata.

Cushman and others have described and figured forms from the Upper Cretaceous
of Mexico and Trinidad under the name Spiroplectoides clotho (Grzybowski), which

1 1838, Abh. k. Ak. Wiss. Berlin, p. 135, pi. iv, lig. ^, frequens.

^ E.g. J. A. Cushman, 1926, Contrib. Cushman Lab. Foram. Research, 11, p. 16; 193 1, op. cit., vii, p. 39;
1927, Amer.J. ScL, xiii, p. 324. Since writing the above, a new species, Giimbelina wilcoxensis, has been
described by Cushman and Ponton from the Wilcox formation, or upper part of the Lower Eocene, of
Alabama, U.S.A. (1932, Contrib. Cushman Lab. Foram. Research, viii, p. 66, pi. viii, figs. 16, 17).

^ 1898, Rospr. Ak. Lhniej. [Mal.-Przyrod.) Krakow, ser. 2, xiii, p. 293, pi. xii, fig. 12.

^ There seems a possibility that trimorphism may be shown by this species, in which case the Aj form
may be Grzybowski 's S. brevis; Aj, S. spectabilis; and B, S. clotho.

= Loc. cit., p. 293, pi. xii, fig. 13.

^ 1901, Rozpr. Ak. Umiej. (Mat.-Przyrod.) Krakozv, ser. 3, I, dz. B, p. 283, pi. vii, fig. 18.

' 1931, jfourn. Pal., v, p. 335, pi. xxxvii, fig. 10.

FOSSIL FORAMINIFERA FROM THE BURDWOOD BANK 9

agree closely in shape with the present specimens.^ Cushman's genus Spiroplectoides,
with genotype Spiroplecta rosula, Ehrenberg, 1854," was erected, however, for forms
whose wall is calcareous and finely perforate, and it is therefore doubtful whether these
records really refer to Grzybowski's species.

Rzehakina epigona (Rzehak) was described from a single specimen^ from a deposit
stated to be of " Alttertiar " age. Rzehak's appear to be the only records in the literature
of the genus from alleged Tertiary strata. The species is recorded from the Velasco of
Mexico* and a variety (var. lata) from the Upper Cretaceous of Trinidad.'' What seems
from the figure to be the same genus is described by Grzybowski^ from the Upper
Cretaceous of Galicia under the name Spiroloctdina inclusa.

The paper in which Rzehak describes R. epigona gives no indication of the evidence
on which he based the ' ' Alttertiar ' ' age of the strata in which it was found. The question
is of importance, since other species, certainly one of which [Pseiidotextiilaria varians)
is now well known as an Upper Cretaceous indicator, were described from these beds.
The question is earlier discussed at some length by Rzehak himself,^ where he records,
unfortunately without figures, and with many nomina niida, 181 species and varieties of
Foraminifera, from a number of specimens of strata collected by E. Kittl from the
"Alttertiar" in the neighbourhood of Bruderndorf, Lower Austria. Rzehak concludes
on balance that " the fauna " is of Tertiary age on the ground of a few poorly preserved
orbitoids and a nummulite which he identifies with confidence as Niimmtilites boiicheri,
de la Harpe. He discusses the strong Cretaceous elements on pp. 7-9, and states that
the fauna appears to be of the same age as the Leitzersdorf bei Stockerau Foraminifera,
described by Karrer^ definitely as Upper Cretaceous on the ground of the foraminiferal
fauna. Rzehak admits that several of the Bruderndorf species are typical Upper Cre-
taceous forms, and that others differ very slightly from Upper Cretaceous species. Such
are Bolivina draco, Marsson, Cristellaria rotulata var. macrodisca, Reuss, Marginulina
solnta, Reuss, and forms closely allied to Frondicularia reticidata (Reuss), F. interpiinc-
tata (Reuss) {sic), F. lanceolata, Reuss, F. angulosa, d'Orbigny, Vaginulina angustissima,
Reuss, Cristellaria gosae, Reuss, C. bacillum, Reuss, C. ntida, Reuss. To this list must
be added Pseiidotextidaria varians, Rzehak (recorded in this paper under the name
Cuneolina elegans), now well known only as an Upper Cretaceous fossil. To-day I think
such species would be accepted as definitely dating the beds in which they occurred as

^ E.g. igz"], Joum. Pal., I, p. 159, pi. xxviii, fig. 6; 1929, op. cit., in, p. 32, pi. iv, fig. 5; 1932, Proc. U.S.
Nat. Mus., Lxxx, Art. 14, p. 43, pi. xiii, figs. 5, 6.

" 1927, Contrib. Cushman Lab. Foram. Research, n, p. 77.

^ From Zdaunek in Moravia, 1895, Ueber einige merkwiirdige Foraminiferen aus dem osterreichischen
Tertiar, Ann. k. k. Nat. Hofmus. Wien, x, p. 214, pi. vi, fig. i.

* 1927, Cushman, Journ. Pal., i, p. 150, pi. xxiii, fig. 4; 1928, White, Jowni. Pal., n, p. 186, pi. xxvii,
fig. 6.

^ 1928, Cushman and Jarvis, Contrib. Cushman Lab. Foram. Research, iv, p. 93, pi. xiii, figs. 11 a, b;
1932, Cushman and Jarvis, Proc. U.S. Nat. Mus., lxxx. Art. 14, p. 20, pi. vi, figs, i a, b.
'' 1901, Rozpr. Ak. Umiej. {Mat.-Prsyrod.) Krakow, ser. 3, I, dz. B, p. 260, pi. vii, fig. 20.
' 1891, Ann. k. k. Nat. Hofmus. Wien, vi, pp. 1-12.

* i%-]o,Jahrb. k. k. Geol. Reichsanst., xx, pp. 157-84, pis. x, xi.

10 DISCOVERY REPORTS

Upper Cretaceous. It seems probable that more than one fauna is represented in
Rzehak's collection, the nummulite and orbitoids at least (noted by Karrer from near-by
Eocene) having been accidentally introduced.

A paper by A. Liebus^ describes a fauna largely similar to that recorded by Rzehak in
his paper of 1891, but without the nummulite or orbitoids. As indicated in the title
Liebus, following Rzehak, assigns it to the Eocene on grounds that do not appear to be
adequate. He discusses the age on pp. 341-2, where he admits the very large and typic-
ally Upper Cretaceous element of the fauna, but relies on other species, of which he
quotes three, to prove the Tertiary age. Of these three, Glomospira gordialis (Parker and
Jones) has been recorded and figured from strata of most ages from the Carboniferous to
the present day, including the Cretaceous. The other two species he records as Vaginulina
briickentholi, Neugeboren, and Clavuliiia szaboi (Hantken). The fauna (or faunas, for
two are recorded) seems to be at least mainly of Upper Cretaceous age, and since Prof.
Liebus did not collect the material himself it may perhaps not be quite free from
suspicion of containing some admixed Tertiary forms.

An important paper by O. Kiihn^ gives the full fossil evidence for the presence of
Danian strata, which are mapped as a number of small isolated outcrops in the Brudern-
dorf area. His map shows much more complex geological conditions than the older
map indicated. A foraminiferal fauna of forty-one species is listed by Ozawa (pp. 550-
2); while it contains some typical Cretaceous forms, species of Pseudotextularia are
not included. The evidence of this paper increases the probability that part of Rzehak's
mixed fauna is of Upper Cretaceous age.

It may be recalled that the Esna Shales of Egypt were once held to be of Eocene age,
though they are now considered to be definitely Upper Senonian (Danian).^ I have a
considerable fauna of Foraminifera from them, hitherto unrecorded, including such
forms as Pseudotextularia varians, Rzehak, and P. globidosa (Ehrenberg). The Esna
Shales are thus probably equivalent to the Austrian strata noted above.

According to the literature, therefore, the forms Pseudotextidaria globidosa, Rzehakina
epigona (of which two, however, only a single specimen each was found in the
Burdwood Bank material), and Spiroplectammina spectabilis may be taken as definitely
indicating an uppermost Cretaceous age. This appears to be valid in both Europe and
North and Central America. Cyclammina elegans, Cushman and Jarvis,** has recently
been described from the Upper Cretaceous of Trinidad, and so strengthens the evidence
for the presence of strata of this age on the Burdwood Bank. I have, however, lately
examined three samples of material from the South American region, with results
that cast doubts on the reliability of the first two species as Cretaceous indicators

1 1927, Neue Beitrage zur Kenntnis der Eozanfauna des Krappfeldes im Karnten, Jahrb. Geo!.
Biindcsanst., Wien, lxxvii, pp. 333-92, pis. xii-xiv.

^ 1930, Das Danien der ausseren Klippenzone bei Wien, Geol. Pal. AbJi., Wien, xxi, Heft 5, pp. 492-576,
pis. xxvi, xxvii.

^ See W. F. Hume, 191 1, Quart. Journ. Geol. Soc, lxvii, pp. 124-8.

* 1932, Proc. U.S. Nat. Mus., lxxx, Art. 14, p. 13, pi. iii, figs. 6 a, b.

FOSSIL FORAMINIFERA FROM THE BURDWOOD BANK ii

in that area. Unfortunately in two cases the history of the samples is not known
sufficiently well to exclude any question of contamination, so that while there may be
suspicion on the point proof is yet lacking.

Mr R. Wright Barker kindly allowed me to examine a sample of Foraminifera from
the Clay Pebble Bed of Ancon, Ecuador, which is, so far as is known, of Eocene age, and
no fossiliferous Cretaceous rocks are known or suspected, I believe, in the vicinity. In
the sample I found a single specimen of Rzehakina epigona. The other Foraminifera did
not suggest to me any definite age.

Some samples from the "Clay of Payta", North-west Peru, were sent by the late
Dr Jose J. Bravo, Director del Cuerpo de Ingenieros de Minas at Lima, and are
preserved in the Sedgwick Museum. They may have been collected from the Lobitos
formation (Eocene), but in default of more than a locality label this cannot be stated as a
fact. One tube contains an abundance of Pseudotextiilaria globulosa together with a number
of other forms in the same state of preservation and of Tertiary aspect, such as Nodosaria
spinosa, Berry, 1928, {non Neugeboren), described from the Eocene of Peru, Dyocibicides
sp., Ceratobidimbia sp., Uvigerina sp., and three forms oi Plectofrondiadaria, etc.

A third sample, from another locality, yielded many Rzehakina and abundant Pseiido-
textularia globulosa, together with some well-known forms described from the Eocene
of the Mexican region, in the same state of preservation.

In addition to the four characteristic species noted above from the Burdwood Bank,
the following may form part of the same probably Upper Cretaceous fauna ; they have
been recorded from the Upper Cretaceous particularly in the Mexican and West Indian
region, and one, Globigerina cretacea, is often rather characteristic of it. Since the species,
or such closely allied forms that it is not possible to separate them in the literature,
nearly all occur in Tertiary deposits, and most are living at the present day, they are by
themselves not of definite value as age markers :

Hormosina globulifera
Trochammina globigeriniformis
Haplophragmoides subglobosiis
H. coronatus
Glomospira charoides
G. gordialis
Nodosaria limbata
N. communis

Cristellaria rotulata ?
Allomorphina cretacea ?
Valvidineria allomorphinoides
Gyroidina nitida
Globigerina cretacea
PuUenia sphaeroides
Anomalina ammonoides

The similarity of foraminiferal faunas in facies and species between localities so far
apart as the Burdwood Bank and the Mexican-West Indian region (some 4000 miles
direct), is remarkable; particularly when the great difference in latitude is considered,
54° S as compared with 10-25° •'^- ^^^ many species of Cretaceous Foraminifera are
known to be of very wide range ; they are identical in America and Europe according
to Cushman,^ and appear also to be the same in Australasia according to Chapman.^

1 Cf. i()2i, Journ. Pal., v, p. 298.

^ Cf. 1926, Pal. Bull., XI, Geol. Surv. New Zealand.

13 DISCOVERY REPORTS

The similarity of the Burdwood Bank and Trinidad faunas is more readily understood
when the rather deep-water facies of both is taken into account.

The widespread similarity of the Foraminifera is strikingly in keeping with the re-
sults obtained from a study of the Senonian ammonites of the South Patagonia-Graham
Land region as noted below, while the Upper Senonian MoUusca are said to be similar
to those of New Zealand.^ The Tethys, according to Gregory,^ had, in the Upper
Cretaceous, perhaps its widest extension. It ranged during the Middle Senonian
" from Kansas to England, Algeria, southern India, and to the Gingin Chalk of Western
AustraUa, On the other side of the Pacific the European fauna reached northern Chile
(23° S), doubtless through the West Indies . . . ". Since the south Atlantic region appears
to have been largely occupied by land at this period^ the connection between the Senonian
sea over the Mexican and South Patagonia-Graham Land regions must have lain to the
west of the present South America.*

Of the Burdwood Bank fossil Foraminifera other conspicuous species are Cyclammina
conceUata, C. orbicularis and Ammodisciis incertus. These three, with seven other species
of the complete list above, have been recorded by Nuttall from the Tertiary (Upper
Eocene to Middle Miocene) of Trinidad.^ Nuttall's material is preserved in the Sedg-
wick Museum in Cambridge, and I have checked the agreement of his specimens with
those from the Burdwood Bank. Certain of these species, particularly of Cyclammina,
are common forms of the Trinidad tertiaries, and, taken into consideration with the
regional stratigraphical development, may point to strata of similar age exposed on the
Burdwood Bank. These arenaceous species can, however, be regarded only as evidence
of similar facies, and not of age. Of a total of some 300 specimens of Foraminifera
mounted there are at least eighty-six specimens referred to Cyclammina spp., so that
it is a common genus, only approached in numbers in the collection by Haplophrag-
moides spp., of which there are at least thirty-five specimens, and Ammodisciis with
twenty-seven. It may be noted, however, that these are all rather large forms found
loose on the sea-bottom, and it is likely that they have been preserved owing to their
size and stout build when smaller forms have been destroyed or removed.

The crushed state of many of the specimens, particularly of arenaceous species, is a
feature which has been remarked by Cushman and Jarvis in the Trinidad Cretaceous
(1928, loc. cit., p. 85), and by Nuttall in the Trinidad Tertiary (1928, loc. cit., p. 70).
I have observed a severely crushed foraminiferal fauna from the Tertiary of Ecuador
(Clay Pebble Bed of Ancon), and there are many badly crushed specimens in the fauna
from the Clay of Payta, North-west Peru.

^ Cf. J. W. Gregory, 1930, Proc. Geo!. Soc, p. xcviii.
^ Loc. cit., p. xciv.

^ See Gregory, 1929, Proc. Geol. Soc, p. cxviii, etc.; and von Ihring, 1931, Quart, jfoiirn. Geol. Soc,
Lxxxvii, p. 386, etc.

* See also A. Windhausen, 1932, Zeitschr. Ges. Erdkuiide, Berlin, p. 28, text-fig. 2.
^ 1928, Quart. Journ. Geol. Soc, lxxxiv, pp. 57-115, pis. iii-viii.

FOSSIL FORAMINIFERA FROM THE BURDWOOD BANK

13

CORRELATION OF THE REGIONAL STRATIGRAPHY

Marine Upper Cretaceous (Senonian) strata are known in Southern Patagonia and
Tierra del Fuego/ where, following Wilckens, they are termed the San Jorge Series, and
possibly extend up into the Danian. In Graham Land similar strata form part of the
Snow Hill Beds and Seymour Island Beds.- In each area they contain an abundant
fauna of ammonites ; one that is common to both gives its name to beds in the Pata-
gonian area, Lahillia luisa, Wilckens. This ammonite fauna is said to resemble those of
the same age so far afield as Natal, Southern India, Japan and Vancouver. In addition
the fauna includes corals, Mollusca, etc.

Fig. 2.

Through the kindness of Dr G. W. Tyrrell, material for comparison from the
Cumberland Bay Series of South Georgia was available from the Ferguson Collection
in the Hunterian Museum, University of Glasgow. From this series an ammonite was
collected by Dr Konig of the German Antarctic Expedition under Lieut. Filchner, and
is possibly of Cretaceous age; Radiolaria believed to be Mesozoic were found in the
material in the Ferguson Collection.^ The material now examined consisted of hard,

1 Cf. O. Wilckens, 1906, Neues Jahrb. Min., Beil-Bd. xxi, pp. 98-195.

^ Cf. T. G. Andersson, 1906, Bull. Geol. Inst. Univ. Upsala, vii (1904-5), pp. 19-71, 2 maps,

^ In J. W. Gregory, 1914, Geol. Mag., p. 64.

14 DISCOVERY REPORTS

altered and crushed sediments, which might well have compared, in their original state,
with the Burdwood Bank shale. No Foraminifera were, according to Dr Tyrrell, ob-
served in the sections in which the Radiolaria were found. Treatment of the material by
disintegrating it in an iron mortar likewise unfortunately yielded me no trace of
Foraminifera.

Tertiary strata are found following the Senonian in both Patagonia (cf. Wilckens,
loc. cit.) and in Graham Land (cf. Andersson, loc. cit.). They often contain marine
fossils and are said to extend in age from the Eocene to the Pliocene. In Patagonia they
are known as the Pyrotherium-Notostylops Beds (Eocene and OUgocene, largely ter-
restrial but partly marine), Patagonian Molasse (Lower Miocene, marine), Santa Cruz
Beds (Middle and Upper Miocene, largely terrestrial), and Parana Beds (Pliocene,
marine). In Graham Land the tertiaries appear conformably to follow the Upper
Cretaceous in Seymour Island, and contain what is recorded as an Upper Oligocene
or Lower Miocene assemblage of MoUusca ; many of the forms are the same as those of
the Patagonian Molasse. Von I bring, however {loc. cit., p. 387), states that the flora
recorded from there by Oliver as Oligocene is really of Palaeocene age. The ? Pliocene
Pecten-congXomtraXQ is a much later deposit which rests unconformably on the older
strata; Andersson {loc. cit., p. 52) compares it with the Patagonian Parana Beds.

The whole succession of Upper Cretaceous and Tertiary strata in the South Patagonia-
Graham Land region may thus be compared, for instance, with that of Trinidad.

CONCLUSION

Hitherto the geological evidence from the Burdwood Bank appears to have been
confined to the two soundings by Ross referred to by Suess,^ which yielded volcanic
rocks in lat. 54° 18' S, long. 60° W, and in lat. 54° 41' S, long. 55° 12' W. The present
evidence of sedimentary strata is therefore of considerable geological interest.

Although there appears to be some doubt, as shown above, on the degree of reliance
that may here be placed on certain species of Foraminifera generally taken to be precise
Upper Cretaceous markers, in my opinion there is httle doubt that, on the ground of
the foraminiferal evidence, the Burdwood Bank beds include Upper Cretaceous (Sen-
onian) strata, together with representatives of the Lower Tertiary succession. In the
circumstances, and on account of the rather deep-water facies of the Foraminifera,
it is perhaps not a matter for surprise that faunas of the two ages cannot be clearly
separated.

The Burdwood Bank beds thus appear clearly as a direct continuation of those of
Staten Island and Tierra del Fuego, as was believed by Suess. The latter beds are seen
to have taken part in the continuation of the Andean folding, which has been discussed
by authors under the ill-named title of the "South Antillean Arc"; Mr Wordie has
agreed that it may better be referred to as the "Scotia Arc", since it surrounds the
newly named Scotia Sea. This new evidence appears to strengthen the hypothesis that

1 The Face of the Earth (English Trans.), 1909, iv, p. 490.

FOSSIL FORAMINIFERA FROM THE BURDWOOD BANK 15

South Georgia with its folded Cumberland Bay Series, probably in part at least of
Cretaceous age, is part of the same arc of folding.^

The very numerous echo-soundings taken by the Discovery Expedition have been
studied by H. F. P. Herdman,^ and the results are of great importance from the geo-
logical standpoint. It appears now to be clear for the first time that the contour of the
sea-bottom definitely fits in with the line of the Scotia Arc from Tierra del Fuego,
through Burdwood Bank, Shag Rocks, South Georgia, Gierke Rocks, South Sandwich
Islands, South Orkney Islands, to the South Shetland Islands and the Trinity Peninsula
of Graham Land.^ This trend is clearly marked on the new charts by the ridge,
frequently irregular, which rises from the sea floor and, it seems, must reflect the
geological fold structure. An equally important point is that there appears to be no
room for any alternative arc, since the available soundings are set too close to allow of
any appreciable ridges across the Scotia Sea farther west having been overlooked.

What may have been two abortive attempts to shorten the arc by closure farther west
are perhaps indicated by, firstly, the south-eastward projection of the eastern end of
the Burdwood Bank, which has no mapped counterpart on the south of the arc;
and secondly, by the position and trend of South Georgia, which has a counter-
part on the south of the arc indicated by the northerly double projection of the
3000 m. line. In neither case, however, can the ridges be regarded as more than a
tentative eff'ort of the folding, which never arrived at completion. In the former case,
indeed, the projection is towards an area of greater than the usual depth in the Scotia
Sea.

For suggestions and criticism of this paper I am much indebted to Dr Stanley Kemp,
F.R.S., to Mr J. M. Wordie, and to Mr A. G. Brighton; and for much help in the
determination of some of the Foraminifera I am greatly indebted to Mr A. Earland.

SUMMARY

Three bottom samples dredged by the Discovery Expedition yielded fossil Fora-
minifera of rather deep-water facies; some were obtained loose on the sea-floor and
others were washed in the laboratory from fragments of green-grey shales, which also
contained a few Radiolaria. Thirty-four species are recorded and six are figured, none
is new; they include some considered to be of Upper Cretaceous (Senonian) age, and
others probably from the Lower Tertiary. Previous records of the fossil Foraminifera

1 For recent discussions of this see O. Holtedahl, 1929, On the Geology and Physiography of some
Antarctic and Sub-Antarctic Islands, Set. Res. Norwegian Ant. Exps., 1927-8, 1928-9, No. 3, pp. 104-17.
8vo, Oslo; and also O. Wilckens, 1932, Der Bogen der Siidlichen Antillen (Antarktis), Sitz. Naturw. Abt.
Nieder-rhein. Ges. Naliir- iind Heilkunde, 1 930-1, herausg. v. d. Naturhist. Ver. preuss. Rheinlande und
Westfalens, pp. 1-14 (separate), Bonn.

^ H. F. P. Herdman, 1932, Report on Soundings taken during the Discovery Investigations, 1926-32,
Discovery Reports, vi, pp. 205-36, pis. xlv-xlvii, Cambridge.

^ Cf. T. Stocks, 1932, Zeitschr. Ges. Erdkunde, Berlin, pp. 198-208.

i6 DISCOVERY REPORTS

of the area are reviewed, and the regional stratigraphy is discussed. It is considered that
the Burdwood Bank beds are clearly shown to be the continuation of those exposed on
Tierra del Fuego and Staten Island, and a part of the (renamed) Scotia Arc of folding,
which is continued on a trend precisely determined by the soundings to lie on the line
of the Shag Rocks, South Georgia, Gierke Rocks, South Sandwich Islands, South
Orkney Islands to the South Shetlands and Graham Land.

Note. Pebbles from Sts. 719 and 720 were submitted to Mr W. Campbell Smith
for his opinion, and he kindly reports that in the sample from St. 719 he found a
large number of dark brown (some almost black) pebbles rich in phosphate and
containing fragments of Radiolaria. Most of these are slightly calcareous as well as
phosphatic, and are rather like the phosphate nodules of Agulhas Bank preserved in
the British Museum (Natural History).

In the sample from St. 720 he found similar phosphatic nodules, and also rounded
pebbles of greywacke, quartz-plagioclase-porphyry (quartz-porphyrite of some authors),
quartz-diorite, quartz-gabbro, hornblende-granulite, and a slaty rock with bands rich
in clastic felspar. Such rocks must have been derived from a land mass. The "slate",
"greywacke", quartz-diorite, and quartz-plagioclase-porphyry can be matched almost
exactly with rocks from points on the west of Tierra del Fuego, e.g. a similar quartz-
diorite occurs on Hermite Island. He found no fragments of volcanic rocks in either
sample.

[Discozwry Reports. Vol. VII, pp. 17-26, March, 1933.]

FAECAL PELLETS FROM MARINE

DEPOSITS

By
HILARY B. MOORE, B.Sc.

FAECAL PELLETS FROM MARINE

DEPOSITS

By Hilary B. Moore, b.Sc.
(Text-fig. i)

IN many accounts of marine deposits, mention is made of small ovoid bodies, of more
or less the same consistency as the matrix in which they are found, and forming up to
30 per cent and in rarer cases as much as 100 per cent of the deposit.

Since these bodies are found in both recent and fossil marine deposits, and are of
world-wide occurrence, and since various theories have been suggested as to their
origin, it is thought desirable to bring together here the facts which are known as to their
nature and occurrence.

Buchanan (1890), in an account of some mud taken off Arran in the Firth of Clyde in
1878, describes elongated pellets which could be separated from the ground-mass of the
mud by elutriation. He rightly ascribes to these pellets an animal origin, but intro-
duces an error which has been followed by several subsequent authors in saying that
they are the faeces of ophiuroids. The error arose from the fact that he took the mud by
means of an iron bucket which skimmed the surface of the mud. This collected at the
same time both the surface mud with its contained pellets, and a large number of
ophiuroids, probably A7nphmra chiajii. He therefore drew the conclusion that these
were responsible for the production of the pellets, which, he says, were formed by the
trituration of sand grains in their stomachs. In the same paragraph, however, he states
that in those particular muds there is practically no sandy material. In any case, ex-
amination of Amphmra chiajii shows that it does not form pellets of either this, or any
other regular shape, and that its excreta are in very much larger masses than the mud
pellets. Further, it has since been demonstrated (Moore, 193 1) that the pellets in these
particular muds are formed by Maldanid worms. And the worms themselves are
rarely taken in any quantity by an instrument which merely skims the surface of the
mud.

Buchanan (1890) further records a similar mud taken in 1879 in the Sound of Rasay
in 155 fathoms, and again he says: " Sticking to the outside of the bag were many large
ophiuroids, which will account for the coprolites (Extract from the Deck Book of the
Steam Yacht ' Mallard ', 1879)." He continues: " Later, in the year 1886, when in charge
of the expedition to survey the Gulf of Guinea, in the steamship ' Buccaneer', I found
the same thing almost universal all along the African coast, and developed in the most
remarkable manner on the coast flat within a considerable radius of the mouth of the
river Congo. Here it was necessary to introduce new designation for muds, and in this
district, the most frequent entries in the deck book as to the nature of the bottom are

20 DISCOVERY REPORTS

' cop.m. ', meaning coprolitic mud. These so-called coprolites were almost jet-black, and
of the size of mouse droppings, and they were covered with the same substance m
flocculent form, or were free from it, according to the scour of the tide in the locality.
It was best developed in comparatively shallow water, in a depth of 50 fathoms, when
the large ash bucket, to the use of which I found it convenient to revert, came up full of
these coprolites, without any flocculent matter whatever. All along the coast the mud of
the locality was moulded in a similar way, though it was not so striking. When the course
of the cruise took us across the open ocean to Ascension, and thence northwards, we
were able to trace the transition of the more earthy shore coprolites into the mineralised
and glauconitic pelagic ones". I think the last sentence means that pellets found in the
deeper deposits were formed from material of pelagic origin, and not by pelagic animals.
Murray and Renard (1891), in describing the bottom samples of the Challenger
Expedition, say: " In examining the samples of Blue Muds, and especially those near the
mouths of rivers, many oval-shaped bodies, about 0-5 mm. in length, were observed.
These were described by some observers as Foraminifera. Mr Murray, after numerous
observations, came to the conclusion that they were mostly the excreta of echinoderms,
principally of holothurians. When these pellets are voided by the animal they are
covered by a slimy substance ; many of them may indeed be united in a chain. In some
deposits this dung is exceedingly abundant, but as a rule it is impossible to recognize
these oval bodies in any of the organic oozes, and in the Red Clays only some doubtful
specimens have been met with. They appear to fall asunder when the deposit is granular,
like a globigerina ooze, or when long exposed without being covered up, as in the case of
red clays".

There are two arguments against the supposition that these are holothurian pellets.
In the first place the theory entails the existence in considerable numbers and with a
world-wide distribution of a small holothurian, about 2 cm. long. That size is arrived at
from the diameters of known holothurian pellets, a Holothuria nigra 20 cm. long having
a pellet 4 mm. in diameter, and a Cucumaria hyndmani 5 cm. long having pellets with a
diameter of 075 mm. Such a holothurian is at any rate not present in the Clyde, where
these pellets are very abundant (Buchanan, 1890; Moore, 193 1). In the second place,
all the holothurian pellets which I have seen are either in the form of rods, or else of
rods constricted at more or less regular intervals by deep clefts, the rods tending to
break into short lengths at these constrictions. Also they are characterized by their
extreme fragihty, being much more friable than the ovoid pellets of the mud, which
remain firm after a hundred years (Moore, 1931) and may even fossilize. Murray him-
self confirms their friability. Further, the pellets into which the holothurian faecal rods
may break are not ovoid, as are the pellets found in these deep-water deposits, but rather
cylindrical, with slightly rounded ends. This shape seems to be fairly constant for the
holothurian pellets, as is the ovoid shape for the pellets from the deposits, and it has
been shown (Moore, 193 1 a, 193 1 b) that such differences of shape are, when constant,
of definite specific importance.

Murray and Philippi (1908) describe and figure a sample from off the mouth of the

FAECAL PELLETS FROM MARINE DEPOSITS 21

Congo, the description being as follows: "Viele ovale gerundete Korper, warschein-
lich die Exkremente von Echinodermen. Bei einigen von ihnen lasst sich deutlich
Glaukonitsierung beobachten. Sie variieren in der Lange von o-4-o-8 mm., in der
Breite von o-2-o-6 mm." They figure also a pteropod ooze in which there are some
similar coprolites. On p. 103 they say: " Den wichtigsten und interessantesten Teil
dieser Probe bilden die ovalen Korperchen welche von Sir John Murray fiir Exkre-
mente von Echinodermen gehalten werden. Sie wechseln in der Farbung von grau zu
braun und dunkelgriin. Sie wurden bereits von der Challenger Expedition erwahnt, und
sind seitdem besonders durch Buchanan aus dem Golf von Guinea und von der
Kongomiindung bekannt geworden. Buchanan konstatierte, dass mancherorts in einer
Tiefe von 50 Faden die ganze Anlagerung aus diesen Exkrementen besteht, und ge-
brauchte dafiir die Bezeichnung 'Coprolitic Mud'. Diese Exkremente fanden sich
bisher nur in tonigen Ablagerungen in der Nahe des Landes, besonders an der Miin-
dung grosser Strome".

Vaughan (1924) has figured and compared similar pellets from the Bahamas, and
oolites from the same region. He says (p. 327): "For some time I thought the ellip-
soidal aggregates in the fine-grained muds were to be considered oolite grains, for in
external features they are very similar, but the grains in the mud do not exhibit the
concentric structure of the ooHte grains. . . . However, the cores of the oolite grains are
similar to the grains in the muds, and it may yet be shown that the grains in the mud
represent a stage in the formation of oolitic limestone. . . . The origin of the ellipsoidal
grains is a puzzle. They resemble in size, form and general structure, glauconite and
greenalite grains, and it is probable that when their formation is explained, the ex-
planation will be of wide appUcation". In his photographs of sections of pellets from
the Bahamas there is, as stated, no trace of concentric structure, nor of localization of
any particular material, but there appears to be a darker region round the outside of the
pellet, presumably corresponding to a region of slightly finer material, possibly to some
extent cemented with mucus or some other material. If this is the case they are similar
in structure to the pellets described from the Clyde (Moore, 193 1, 193 1 a; GaUiher,
1932, fig. 2).

Takahashi and Yagi (1929) describe the distribution of similar pellets in various
localities in Japan. With the exception of Kasumiga-Ura, which is a nearly fresh-water
lake, the pellets were found only in marine deposits, and not in fresh water. The pellets
formed in some cases as much as 5 per cent of the deposits, and are described as elongated
ellipsoidal, with a diameter of 0-5-1 -4 mm. They were rather soft and friable, and varied
in colour from Hght grey to dark green, the apparent density varying from i-6 in the
grey to 2-2 in the green ones. In section they appeared, with rare exceptions, to be
homogeneous. Takahashi and Yagi compare these pellets with very similar structures
from Tertiary deposits, and trace the probable course of glauconization in the recent
pellets. With regard to the origin of these pellets they suggest that these are probably
the faeces of some mud-eating animal. It is interesting to note that some few of their
pellets, although ovoid in outline, had heHcoidal cores. Although I have not so far met

22 DISCOVERY REPORTS

with the pellets of any animals which quite correspond with this type, those of the
mollusc Aplysia punctata have a similar outer layer, enclosing an inner pellet of different
shape, while those of many molluscs have spiral groovings on the surface, and in extreme
forms the pellet may be in the form of a spirally twisted rod. It is easy to imagine the
conjunction of these two types giving rise to an ovoid pellet with a spiral core, as
described above.

Galliher (1931) has described, under the name of "Sporbo", structures which are
apparently fossilized pellets of this same ovoid shape, from Miocene oil shales in
America, and in a later paper (1932) he compares these with similar recent pellets from
the Clyde. With the exception of hardening, and in many cases pyritization of the
former, they appear to be very similar.

Thorp (1931) describes three stations off the coast of Panama, on the Atlantic side,
in depths of 3595, 2131 and 3445 m. respectively, in which pellets form 26-1, 38-1 and
44-1 per cent of the mud ; and of thirty- two stations described by him from the western
North Atlantic and Caribbean Sea, sixteen are recorded as containing such pellets. With
regard to the origin of these, however, he is very doubtful. He says : " The size and shape
of the individual ovoid suggest an organic excrement but the preservation for any
length of time of such an object appears doubtful. Bacteria would find nourishment in
the unassimilated food of higher organisms. As a result most of the remains would be
converted into gaseous and water-soluble products. These in turn would be dispersed
by the sea water. It appears more probable that chemical and physical aggregation is
responsible for the formation of these objects".

In the case of the pellets found in the Clyde it has been definitely shown that they are
animal products (Moore, 193 1), and furthermore that they are able to retain their form
as well as their differentiation from the surrounding mud for periods of at least a
hundred years. At the end of this period they are quite as definite as they were when
they were first formed, and there is no reason to suppose that any new agencies come
into play to cause their breakdown after that time. At any rate bacteria are known to be
abundant throughout the whole of the time, and at all depths in the mud (Lloyd, 193 1).
In the Discovery material there was a type of pellet which was very abundant in
some of the plankton nettings. Samples were examined from Stations 549, and WS 399,
and the pellets in them were almost certainly those of Euphausia siiperba, which is very
abundant in the plankton there. They agree in form with those in the gut of the animal,
as well as with those described from other euphausids from the Clyde (Moore, 193 1 a).
No pellets of this type were observed in any of the bottom deposits, which was probably
due to their quick breakdown as was the case in the Clyde, where pellets of euphausids
and of Calamis finmarchiciis could be seen abundantly on the extreme surface of the
mud, but were never recognizable below the surface.

The pellets from the bottom deposits fall into two classes, rod-shaped and ovoid, of
which the latter were numerically by far the most abundant. They occur at various
stations and at all depths, and though their size and shape vary slightly, they are all of
the same general type. The details of these are given in the table on page 24. Since

FAECAL PELLETS FROM MARINE DEPOSITS

23

this ovoid shape of pellet is common to very many kinds of animals, it is probable
that the ones found in the deposits represent the faeces of several different species.

Smooth rod-shaped pellets with a diameter of up to 0-15 mm. were found abundantly
at St. WS 144, and in smaller numbers at St. WS 502. A single specimen of a simply
sculptured rod was found at St. 375. It consisted of a rod, circular in section, with a
single groove along one side (Fig. i a). In transverse section it appears to be composed
largely of diatom remains, more so in fact than the mud from which it came, or the ovoid
pellets found there, so that it must be the pellet of a selective feeder. I would suggest
further that it is probably the pellet of a mollusc, although none are at present known of
quite this type.

Fig. I. Transverse sections of pellets from marine deposits.

a, Pellet of a mollusc ?, from St. 375. b, Pellet of Niiciila sp.,
from St. WS 144. c, Pellet of Niicida tenuis, from the Clyde.
d, Pellet of Niicula moorei, from the Clyde.

A second and more complexly sculptured type of rod-shaped pellet was found
abundantly at St. WS 144, and consisted of a rod up to o-i6 mm. in diameter with a
series of longitudinal grooves on the surface. In transverse section (Fig. i b) this also
appeared to be formed almost entirely from diatom remains, and may also therefore be
classed as a probable selective feeder. So few pellets have as yet been described that it
would be unreasonable to attempt to identify an unknown pellet with any given animal
without a considerable knowledge of the fauna of the locality. But this pellet in all
respects resembles those at present known for the genus Nucula, and it is at least
reasonable to advance the possibility of its belonging to this genus. Nucida pellets are
all rods with longitudinal grooves ; and all the species which I have so far examined may

24

DISCOVERY REPORTS

be differentiated by specific differences in the number and arrangement of these grooves
(Moore, 193 1 a). The pellets from WS 144 have five grooves in common with N. tenuis
(Mont.) and N. moorei, Winckworth, which are shown in transverse section in Fig.
I c, d for comparison. The grooves are fairly, but not extremely, deeply cut, and in this

%

Station

369

372
373

375

Position

S9°i7i'S,26^57'W

57° 57' S, 29° 53' W
58° GO'S, 33° 44' W

57' 47' S, 40° 49' w

417 38°09'S, i7°45rE
425

WS 144

WS240
WS248
WS374

WS 394
WS403
WS 474
WS 501
WS502

34°S0'S,26°4ii'E
54° 08' 8,36° 10' W

51° 55'
52° 40'

55° 09'
62° 51'

59° 40'
61° 03'

64° 52'
69° 43'

S,65°
S,s8°
8,40°
S, 60°
3,64°
8,56°
3,63°
S,99°

10' W
30' w
00' w
40' w

35' W
42' w
58' w
38' w

WS506 70° 31' S, 81° 36' w

WSSI9

WS52I

52°09i'S,53°2irW

52° 41' S, 47° 14' w

Depth :
m.

1766

292
2515

3665

4778
4107

279

141

210

3226

274
3721
2813

583
4224

584

2270
3780

Amount of

material
examined :

CO.

7-0

0-65

0-42

4-95

2-25

2-25

7-0

4-'5

5-25

i-o

0-4

1-5

3-15

Number of pellets per cc.

Ovoid

7-4

o
1-5

52

o
o
o
o
o
o
o

148

175

3-3

Simple
rods

o
o

179

o
o
o
o
o
o
o

98

Sculptured
rods

o
o

79

o
o
o
o
o
o
o
o

Notes on abundance and size
of pellets

A few large and some smaller ovoid
type present. Diam. 0-55 mm.,
length 0-95 mm., length/breadth
I -75 ; smaller sizes down to
0-25 mm. long.

No pellets found.

A very few ovoid type present.
Diam. o-i2mm., length 0-26 mm.,
length/breadth 2-2.

Ovoid type abundant, but no
count made. A single specimen
found of a rod-shaped type with
one longitudinal groove.

No pellets found.

Ovoid type abundant, but no
count made.

Only a few ovoid type present, but
plain and sculptured rods abun-
dant. Ovoids, diam. 0-28 mm.,
length 0-46 mm., length/breadth
1-64; plain rods, diam. 0-12 mm.;
sculptured rods, diam. 0-26 mm.

No pellets found.

No pellets found.

No pellets found.

No pellets found.

No pellets found.

No pellets found.

No pellets found.

A few large ovoid type, and a few
smooth rods. Rods diam. 0-05-
0-15 mm., and in lengths of 3 or
4 times the diam.

A lot of ovoid type present, but so
broken that the count might easily
represent i/io only of the true
number. Lengths up to 0-5 mm.

A few large ovoid type present.
Diam. 0-24 mm., length 0-47 mm.,
length/breadth 1-95.

No pellets found.

they resemble those of N. moorei. But the ridges are all of about the same size, and the
mid-ventral ridge is very prominent, as in A*", tenuis. It does not belong therefore to
either of these species, and if it does belong to a Nucitla it probably belongs to one whose
pellets have not yet been described.

FAECAL PELLETS FROM MARINE DEPOSITS 25

The estimates of the numbers of pellets per cubic centimetre, given in the table,
must be taken as only very rough approximations. The volumes of mud on which they
are based are the volumes of a portion of the deposit, shaken up in distilled water, and
allowed to settle for 24 hours at about 15'^ C. The highest number of the Nucula type of
pellet found was 79 per cc, which is about three times the highest value found in the
Clyde (Moore, 193 1, p. 355); but the highest value for the ovoid type — 179 per cc. — is
far lower than the value of 3400 per cc, from Loch Striven.

The facts known with regard to the formation of these oval pellets may be summarized
as follows. They are of world-wide distribution, occurring in recent deposits in depths
of from o to over 4000 m., and in conditions varying from almost fresh to salt water, and
also in certain fossil deposits. Although more than one type is found (and E. M. Thorp
[personal communication] states that there are two common and easily distinguished
forms in the West Indian region), yet there is a simple oval type which occurs abundantly
in a number of localities. In structure they are composed of the same material as the
mud in which they are found, although naturally limited to the smaller particles of that
mud. They show no trace of concentric structure as would be expected if they were
oolite-like aggregates. Further, their occurrence as central cores for some oolites
suggests an entirely different origin from that of the outer layers of the oolites. The
possibility of their production by the roUing together of the surface layers of the mud
under the influence of currents is negatived by their comparative firmness, and also by
the fact that they may form as much as 100 per cent of the mud, with no interstitial fine
material at all. It is hard to see how such rolling could proceed so far as to remove the
whole of the fine material. It might, however, produce such a deposit composed en-
tirely of pellets by washing out all the fine material from between the heavier pellets.
Such a process would not account though for the presence in the mud of patches, such
as were occasionally found in the Clyde, where the whole of the mud had been converted
into pellets, these patches being quite small in extent, and sometimes surrounding the
mouth of the burrow of a worm (Moore, 1931, p. 354, fig. n). In relation to the latter
observation it is interesting to note that Galliher also (1931, p- 265) states that the fossil
"sporbo" occurs "in lenses parallel with the bedding and in small oval pockets".

In favour of the animal origin of these pellets is the fact that the pellets from the
different localities appear to be similar, and that those from the Clyde have been defi-
nitely shown to be of animal origin, and that Maldanid worms, which are abundant in
these Clyde muds, shed a pellet identical in nature with the pellets found in the mud.

Although the pellets of a number of littoral and shallow- water animals are now known,
nothing is as yet known of the pellets of any very deep-water animals (except for one or
two Anomura). The nature of the ovoid pellets suggests that they are those of either
polychaetes or molluscs ; they are not likely to be crustacean in origin (see Moore, 1932,
for general description of crustacean pellets). It remains, therefore, to obtain specimens
of the fauna of the muds in which these pellets are most abundant, and to examine
their gut contents to see whether similar pellets can be found in them.

26 DISCOVERY REPORTS

REFERENCES

Buchanan, J. Y., 1890. On the Occurrence of Sulphur in Marine Muds and Nodules, and its bearing on their

tnode of formation. Proc. Roy. Soc. Edin., xviii.
Galliher, E. W., 1931. Collophanc from Miocene Brown Shales of California. Bull. Amer. Assoc, of Petroleum

Geologists, XV, No. 3.

1932. Organic Structures in Sediments. Journ. Sedimentary Petrology, li. No. i, pp. 46-47, April 1932.

Lloyd, B., 1931. Muds of the Clyde Sea Area. II. Bacterial Content. Journ. Mar. Biol. Assoc, N.S., XVII,

No. 3.
Moore, H. B., 1931. The Muds of the Clyde Sea Area. III. Chemical and Physical Conditions; Rate and

Nature of Sedimentation; and Fauna. Journ. Mar. Biol. Assoc, N.S., xvil, No. 2.

193 1 a. The Specific Identification of Faecal Pellets. Journ. Mar. Biol. Assoc, N.S., xvii, No. 2.

193 1 b. The Systematic Value of a Study of Mollusc an Faeces. Proc. Malac. Soc, xix. Part 6.

1932. The Faecal Pellets of the Anomura. Proc. Roy. Soc. Edin., Lii, Part 3, No. 14.

Murray, J. & Philippi, E., 1908. Die Grundproben der Deutschen Tiefsee-Expedition. Wissensch. Ergebn. d.

Deutschen Tiefsee-Expedition 1898-1899 auf dem Dampfer ' Valdivia'.
Murray, J. & Renard, A. F., 1891. Deep sea deposits. Rep. Sci. Results H.M.S. ' Challenger".
Takahashi, J. & Yagi, T., 1929. Peculiar Mud-grains and their relation to the origin of Glauconite. Econ. Geol.,

xxiv. No. 8.
Thorp, E. M., 193 1 . Descriptions of deep-sea bottom samples from the Western North Atlantic and the Caribbean

Sea. Bull. Scripps Inst. Oceanog., Tech. Ser., ill, No. i.
Vaughan, T. W., 1924. Oceanography in its relation to other Earth Sciences. Journ. Wash. Acad. Sci., xiv.

No. 14.

[Discovery Reports. Vol. VII, pp. 27-138, P/ates I-VII,Jime 1933]

FORAMINIFERA

PART II. SOUTH GEORGIA

By
ARTHUR EARLAND, F.R.M.S.

CONTENTS

Introductory Note P^ge 2()

Previous Work in the Area ^9

Material Examined 3°

List of Stations 32

List of New Genera, Species and Varieties 44

Systematic Account 45

Appendix ^3^

Supplementary Bibliography '33

Index . . • 134

Plates I-VII following page 138

NOTE

Owing to illness my colleague Edward Heron-Allen was unable to take as
large a share as usual in the preparation of this Report. At his own request,
and against my wish, his name is omitted from the authorship.— A. E.

T

FORAMINIFERA

PART II. SOUTH GEORGIA

By Arthur Earland, f.r.m.s.

(Plates I-VII, text-figs. 1-3)

INTRODUCTORY NOTE

HE first part of this Report dealt with the bottom deposits from the Falkland
Islands and the adjacent area. The present Report deals with the island of South
Georgia and the outlying Shag Rocks, situated some 800 miles to the eastward of the
Falklands in the Southern Ocean.

Although there is no great diff"erence in latitude between the position of the Falkland
Islands (5i°-52° 30' S) and South Georgia (54°-55° S), it would be difficult to find two
areas so nearly in the same latitude presenting greater contrasts. The Falklands, lying on
the Continental Shelf of South America, are surrounded by a wide area of shallow water
with generally sandy bottom deposits, and, owing to the influence of the warm Pacific
water coming through the Drake Strait, are entirely free from ice and present a fauna of
a distinctly cool, temperate type. South Georgia, on the other hand, lies outside the
influence of the Pacific warm water and, surrounded by the cold Antarctic current
flowing northwards, is within the region of pack-ice. The island rises more or less
abruptly from deep water, so that the loo-fathom line lies quite near the coast. The land
area is mountainous and ice-covered, and there are many glaciers. These influence the
formation of the coastal deposits which, in contrast with the sandy deposits of the
Falkland area, are almost universally composed of a tenacious blue mud of which
Diatoms, so abundant in the surface waters of the Antarctic seas, form a notable con-
stituent. As a general result the bottom faunas of the two areas are very diff^erent. The
majority of the species which are dominant in the Falkland area are rare, and sometimes
absent in South Georgia. The local fauna in coastal waters is of a distinct type including
many species new to science, while in the deeper water it is more or less identical with
that found at similar depths in all seas. Following the general rule that the Arenacea
favour cold water, this group plays a larger part than usual even in the shallow coastal

areas.

PREVIOUS WORK IN THE AREA

South Georgia represents practically virgin ground so far as the Foraminifera are
concerned. It was not touched by the Challenger, Scotia or Terra Nova Expeditions,
and, although the members of the Quest Expedition did much work on the island, the
paper on "Deep Sea Deposits and Dredgings" by Miss A. Vibert Douglas^ deals only

1 Report on the Geological Collections made during the voyage of the "Quest" on the Shackleton-Rowett
Expedition to the South Atlantic and Weddell Sea in 1921-1922. British Museum (Natural History), 1930,
pp. 145-56.

30 DISCOVERY REPORTS

with the physical characters of the deposits. Four species of Globigerina only are men-
tioned (G. dubia, rubra, bulloides, sacadifera), the last of which is almost certainly mis-
identified.

MATERIAL EXAMINED

Bottom samples were received from about ninety stations, over fifty of them being
fairly evenly distributed in the shallower coastal waters, while the others consist of lines
of soundings run out in various directions into the very deep water surrounding the island.
The vast majority of the samples consisted of sounding materials only, the quantity
rarely exceeding 25 cc, and in most cases much less. Almost without exception these
soundings consisted of tenacious mud of shades varying from grey and blue to almost
black. They were treated by washing on a silk sieve of 200 meshes per linear inch, and the
residues were generally very small, seldom more than a few cc. The muds, which were
preserved in spirit, were frequently refractory, in which case the method followed
was to dry thoroughly after a preliminary washing, break down the dried residue with
boiling water and wash again. In a few instances only it was found necessary to use soda to
break down the deposit.

The chief difficulty in the cleaning process lay in the abundance of Diatoms. Apart
from the smaller forms which passed readily through the sieve, larger species, notably
Fragilaria antarctica, Cocconeis imperatrix and Cestodisciis gemmifer, were often present
in such numbers as to form a felt in the sieve. It was not possible to separate them from
the smaller and lighter species of Foraminifera, such as Virgulina, which were often
equally abundant, and their presence added to the difficulty and monotony of the ex-
amination of the residues. Another Diatom, presumably Nitzschia sp., which was
equally abundant and the principal cause of the felting in the sieves, could be more or
less eliminated after cleaning, as it separated from the dried material when shaken in a
tube, forming masses like cotton-wool.

No great variety of species can ever be expected in shallow-water soundings. As the
Foraminifera exist mainly in the surface layer, the area sampled by the sounding tube is
not large enough to ensure much of a catch. The large arenaceous species which, from
other sources, are known to be present, are not in sufficient numbers to ensure capture,
and when taken were generally fragmentary.

The examination of these shallow-water soundings therefore proved a long and
monotonous task. Station after station yielded approximately 20-30 species, mostly
identical. Now and then the occurrence of a species not previously observed gave some
encouragement to proceed with the work.

The typical species inhabiting the coastal waters occur in nearly every sample,
material from each station as a rule furnishing one or two other species of less constant
distribution. Broadly speaking, the typical species are as follows (those marked * may
be regarded as being peculiarly characteristic and constant) :

Ammodiscus incertus *Cassidulina suhglobosa

Cassidulina crassa *Ehretibergina crassa

*Cassididina parkeriana *Glohigeriiia dutertrei

MATERIAL EXAMINED 31

*Globigerina pachyderma Pulvinulina karsteni

Glomospira gordialis Puhimilina peruviana

Haplopbragmoides canariensis *Reophax siibfiisiformis
*Lagena biancae Texttdaria tenuissima

*Miliammina obliqua Tholosina vesicularis

*Miliatnmina oblonga *Trochammina malovensis

*Nonion depressuliim *Uvigeriiia angidosa
*Nomonella iridea Virgulina bradyi

PuUenia siibcarinata * Virgulina schrcihersiana
Psammosphaera fitsca

Very fortunately I received additional material from other nets, frequently from
stations at which soundings had also been taken. This corrected the impression, which
might otherwise have been formed, that the South Georgian fauna was limited and
scanty. The additional material consisted of five dredgings, four samples of trawl
residues, and fourteen gatherings taken from nets which had touched bottom or had
been attached for other purposes to the trawl. This is not a novel method of collecting
Foraminifera ; it has been used with great success on muddy bottoms by both the Scottish
and Irish Fisheries Boards. The trawl boards throw up the surface mud which passes
through the coarse mesh of the net, leaving behind the larger Foraminifera and sufficient
mud to supply the smaller species also.

These net and trawl residues yielded a surprisingly long list, including several new
genera and species. The evidence obtained from dredgings and soundings indicates the
sea-bottom in the coastal areas of South Georgia to be a surface of tenacious mud,
loaded with Diatoms derived from the surface waters, and having an abundant fauna of
small Foraminifera limited in species, and a more sparsely distributed but characteristic
fauna of larger species. Many of these are mud-eaters, and find an abundant food supply
in the Diatoms, masses of which are found inside their tests. Conditions are so similar
everywhere that there is less variation observable than is usually the case.

Owing to the fact that the area has not been previously investigated there is quite a
formidable list of new forms, including four distinctive new genera Gordiospira,
Pelosphaera, Armorella and Hippocrepinella, besides Miliammina, which we described
from South Georgia in 1930. As regards the last, it is now known that its range extends
both southwards into the Antarctic and northwards to Tristan d'Acunha (St. 6) and
South African coastal waters (St. WS 4), but I am not at present able to state whether
any of the other genera, except Armorella, extend outside the South Georgian area.

The gatherings from depths outside the coastal waters do not call for many comments.
They include Globigerina ooze, diatomaceous ooze, and a few soundings characterized by
abundant Radiolaria, though not in sufficient numbers to constitute a true radiolarian
ooze. Their foraminiferal faunas are varied but present no special characteristics. It is
perhaps noteworthy that, even in the deepest of the soundings, pebbles and sand grains
of all sizes are usually present, probably derived from floating ice.

In conclusion, it may be stated that South Georgia appears, on the evidence at present
available, to possess a somewhat characteristic foraminiferal fauna of its own. Dis-

32 DISCOVERY REPORTS

missing the numerous species which have a world-wide distribution under similar
conditions of depth and temperature, it possesses many forms unknown or com-
paratively rare elsewhere. It has little apparent connection with the Falkland fauna;
apart from widely distributed species common to both areas, very few of the Falkland
species proper are found in South Georgia, and those few very rarely. How far the
examination of the material received from the Antarctic Islands will modify this view of
the isolation of the South Georgia area, I cannot yet say, but from the present state of
the work it would seem that the two regions are not closely connected, although many
South Georgian species occur in the Antarctic area, as might be expected from the
similarity of physical conditions.

A list of the stations worked over is as follows :

STATIONS MADE BY THE R.R.S. 'DISCOVERY'

13. TS 589.1 (See Fig. z.f

3. iii. 26. 5-7 miles N 490" E of Jason Light. Sounding rod, 143 m.

About 15 CO. of tenacious dark slate-coloured mud. Residue of pebbles and dark sand grains of
all sizes. Diatoms, spicules and very few Foraminifera, all of common species.

14. TS 544. D II.3

3. iii. 26. 15-4 miles N 44° 30' E of Jason Light. Sounding rod, 260 m.

About 15 cc. of dark greenish brown mud yielding a residue of dark sub-angular sand grains, a
few sponge spicules, Radiolaria and scanty Foraminifera; Miliammina oblonga, M. obliqiia and
Amniobaciilites americanus being the only species occurring in any numbers.

15. TS 547. D L

3. iii. 26. 25 miles N 45° 30' E of Jason Light. Sounding rod, 191 m.

A very small sample of grey mud, mainly diatomaceous. Sixteen species of Foraminifera were
recorded, all of which, except Buliminu fusiformis, were represented by one or at most two specimens.

16. TS 549. E I.

3. iii. 26. 36-5 miles N 46" E of Jason Light. Sounding rod, 727 m.

A few cc. of slate-coloured mud yielding a residue of angular sand grains. Diatoms, Radiolaria
and a few Foraminifera, including a specimen of Lagena spumosa.

17. TS550. EI.

4. iii. 26. 46 miles N 46° E of Jason Light. Sounding rod, 1900-1950 m.
A few pebbles and some fine sand with very few Foraminifera.

20. TS 539-540. DI.

4. iii. 26. 26-5 miles N 54° E of Jason Light. Sounding rod, 210 m.

Two samples :

(i) About 12 cc. tenacious slate-brown mud.

(2) About ID cc. slate-green mud.

The residues were similar, angular sand grains with Diatoms and few Foraminifera, Miliammina
and Virgulina being dominant.

1 These numbers refer to the station sUdes in the Heron-Allen and Earland collection in the Natural
History Museum.

- This reference is to Fig. 2, p. 35, Chart of Cumberland Bay, South Georgia.

^ These numbers afford reference to the positions of the stations as shown in Fig. i, p. 33.

34 DISCOVERY REPORTS

23. TS 592. (See Fig. 2.)

14. iii. 26. 5-3 miles N 44° E of Merton Rock. Sounding rod, 228 m.

About 10 cc. of tenacious slate-coloured mud. Residue mainly Diatoms and sponge spicules,
with numerous Foraminifera of a limited number of species. CassiduUna parkeriana was very com-
mon, Miliammi?ia oblonga, M. obliqua and Virgulina sclireibersiana common, the few other species
mostly very rare.

27. TS 596-8. (See Fig. 2.)

15. iii. 26. 3-3 miles S 44° E of Jason Light, West Cumberland Bay. Dredge, no m.

A quantity of coarse black angular sand with Polyzoan and Echinoderm debris. Abundant
Foraminifera of the larger species: Ammohaciilites americamis very common; Tholosina hullo, T.
vesicularis, Ilaplophragmoidcs canarietisis, all common; Hyperammina suhnodosa and V anhoeffenclla
gaiissi, frequent; ArmoreUa sphaerica, Pelosphaera con/iita, rare.

28. TS 599-600 A. (See Fig. 2.)

16. iii. 26. 3-3 miles S 45° W of Jason Light, West Cumberland Bay. Sounding rod, 168 m.
About 14 cc. of tenacious slate-coloured mud yielding nothing but Diatoms and some Miliam-

minae. Another sounding (STB) of about 8 cc. of similar mud from 65 m. yielded sand. Diatoms,
many Miliammmae and a few specimens representing five other species of Foraminifera.

A sample of 7 cc. of coarse black sand from a dredging at 168 m. contained a few Foraminifera,
including HippocrepineUa hiriidinea.

29. TS 600 B. (See Fig. 2.)

16. iii. 26. 5-9 miles S 51° W of Jason Light, West Cumberland Bay. Sounding rod, 23 m.

About 13 cc. of tenacious dark slate-coloured mud yielded a residue of fine sand and mud
pellets, very few Diatoms, and fewer Foraminifera, Psammospliaera parva and a single specimen only
of Bulhtiijia marginata being recorded.

30. TS 601-2. (See Fig. 2.)

16. iii. 26. 2-8 miles S 24° W of Jason Light, West Cumberland Bay. Sounding rod, 251 m.
Fifteen cc. of tenacious dark blue mud, yielding only a fractional residue of Diatoms, spicules

and Foraminifera, including Tcxtularia temiissima. A quantity of muddy washings from dredge
residues, largely worm tubes, yielded about forty species of the smaller forms, including Ehrenbergina
pupa very rare, E. crassa frequent, and a single Tubinella funalis.

31. TSS55. DIL

17. iii. 26. 13-5 miles N 89° E of Jason Light. Sounding rod, 220 m.

About ID cc. of tenacious slate-coloured mud. Residue of dark sand grains, many otoliths and
Diatoms, some Radiolaria and a few Foraminifera of the commoner species. Miliammina oblonga,
M. obliqua, Virgulina sclireibersiana and Nonionella iridea all common, Tholosina vesicularis frequent,
all other species rare.

41. TS557. DIL

28. iii. 26. i6| miles N 39° E of Barff Point. Sounding rod, 272 m.

About 15 cc. of tenacious olive-brown mud. Residue mainly of Diatoms and fine sand grains.
A few Foraminifera, of three species only : Ammobaculites rostratus, Ehrenbergina crassa, Miliammina
obliqua.

42. TS 588. (See Fig. 2.)

i.iv. 26. Off the mouth of Cumberland Bay. Sounding rod, 204 m.

About 12 cc. of dark slate-coloured mud. Residue of sand grains, Diatoms and a few Fora-
minifera of the commoner species.

45. TS 593-4. (See Fig. 2.)

6. iv. 26. 2-7 miles S 85° E of Jason Light. Nets on trawl, 238-270 m.

Muddy trawl residues, largely consisting of sponge spicules in a felted mass with abundant
Diatoms and many Foraminifera. Biloculinae of several species, Pelosina, HippocrepineUa hirudinea

MATERIAL EXAMINED

35

among the larger species were all very common ; Hyperammina subnodosa, Storthosphaera elongata,
Ammohaculites bargmanni common; a long list of the usual smaller species of varying degrees of
rarity, Virgidina schreibersiana and V. bradyi being very common.

36 40'

ae'sc

36°20'

23

Scale of Sea Miles

o 1 E 3

ae-^o'

36°30'

36°20'

Fig. 2. Cumberland Bay, South Georgia, and its approaches.

123. TS 603. (See Fig. 2.)

15. xii. 26. Off the mouth of Cumberland Bay. Sounding rod, 250 m.

Three and a half cc. of dark slate-coloured mud with a residue of Diatoms and a iew Miliamminae.

Muddy trawl residues, mainly Crustacean and Echinoderm debris with coarse sand, yielded a
large number of species, including Hyperammina subnodosa, Ammobaculites americanus, Hippo-
crepinella hiriiditiea, Tiibinclla funalis, Armorella sphaerica and all the commoner forms.

36 DISCOVERY REPORTS

126. TS 529. D I.

19. xii. 26. 53° 58' 30" S, 37° 08' W. Net touched bottom at 100 m.

About 70 cc. of dark angular sand and pebbles. Miliamtjwia ohlo?iga, M. ohliqiia and Ehren-
herqina crassa very common. Many other interesting Arenacea, including Wcbbinclla limosa, Pilo-
sphacra cornuta, HippocrepineUa hirudinea and H. alba. Many large specimens of Hyperammina
siibnodosa were received preserved in spirit, having been selected on the ship.

129. TS531. DI.

19. xii. 26. 53° 28' 30" S, 37° 08' W. Sounding rod, looi m.

A few cc. of dark mud. Residue of a pebble, and fine angular sand with many Diatoms and
Radiolaria but few Foraminifera, eight species only being represented.

131. TS 545-6. D I.

20. xii. 26. 53° 59' 30" S, 36° 11' W. Sounding rod, 240 m.
Two samples :

(i) About 12 cc. of dark greenish brown mud, yielding as residue a pebble, dark angular sand
grains. Diatoms, sponge spicules and Radiolaria. Foraminifera very rare except Miliammina and

Virgulina.

(2) About 20 cc. of black sandy mud, yielding similar residue. Although Foraminifera were not
numerous, a good many species were listed, but none of particular interest.

133. No TS.^

20-21. xii. 26. 53° 45' 30" S, 35° 46' 30" W.

A very small pelagic sample from a vertical haul of 270-1 00 m. (N 70 V) yielded a few Globigerinae ,
mostly G. dutertrei. G. bulloides, G. triloba and G. conglomerata were present in lesser numbers.

136. TS 560. E II.

21. xii. 26. 54° 22' S, 35° 21' W. Sounding rod, 246 m.

About 15 cc. of dark brown mud. The residue included some angular fragments of a schistose
rock with sessile specimens of Placopsilina cenomana, Tholosina vesicularis and Tolypammina vagans.
The finer residue consisted of angular sand with scanty Foraminifera, Uvigerina angiilosa dominant,
Globigerina pachyderma, Virgulina schreibersiana and Cassidulina parkeriana all common.

138. No TS.

22. xii. 26. 54° 17' S, 34° 47' W.

Small samples of pelagic material from two vertical hauls (N 70 V). The first, between 750 and
500 m., contained a few Globigerinae only, mostly G. conglomerata and G. dutertrei, the latter being
iess frequent. The second, between 500 and 250 m., consisted largely of Copepoda and Diatoms but
yielded many Globigerinae, G. conglomerata, G. dutertrei, G. bulloides, G. pachyderma, their relative
abundance being in the same order as listed. A few specimens of Globorotalia scitula were also seen.

139. No TS.

22-23. xii. 26. 53° 30' 15" S, 35" 50' 45" W.

A very small pelagic sample from a vertical haul (N 70 V), between 250 and 150 m., contained
only a few specimens of Globigerina conglomerata and G. dutertrei.

140. TS534. DII.

23. xii. 26. Stromness Harbour to Larsen Point, from 54° 02' S, 36° 38' W to 54° 11' 30" S,
36° 29' W. Trawl, 122-136 m.

A small quantity of trawl refuse, mainly Echinoderm and sponge debris. Many interesting
species of the larger Arenacea, Pelosphaera, HippocrepineUa, Vanhoeffenella, etc. Also many small
species including Patellina corrugata and Discorbis chasteri.

143. TS 595. (See Fig. 2.)

30. xii. 26. 54°i2'S, 36°29'3o"W, oflF the mouth of East Cumberland Bay. Net on trawl, 273 m.

1 Where no TS (Station Slide) was prepared the Station has generally been omitted from Figs. 1,2.

MATERIAL EXAMINED 37

Five cc. of organic debris yielding a few Foraminifera including Pelosina, Miliammina and
Hippocrepinella.

144. TS 536-8. D I, II.

5. i. 27. From 54° 04' S, 36° 27' W to 53° 58' S, 36' 26' W, off the mouth of Stromness Har-
bour. Nets on trawl, 155-178 m.

Fine washings from trawl, mud and organic debris. Residue largely of sponge spicules, Diatoms
and dark sand. Many Foraminifera : Tholosina bulla and T. vesicularis abundant among the larger
forms, and Miliammina oblonga, M. obliqua, Virgulina schreibersiana, Cassidiilina crassa, C. siib-
globosa, Haplophragmoides canariensis, Pulvinulina peruviana and P. karsteni equally abundant among
the smaller species.

145. TS 590-1. (See Fig. 2.)

7. i. 27. Between Grass Island and Tonsberg Point in Stromness Harbour. Trawl, 26-35 m.

Washings from trawl, organic debris of all kinds. Diatoms abundant but Foraminifera very few,
though including such interesting species as Turritellella slioneana, Cornuspira selseyensis, Gordiospira
fragilis and Spirillina obconica.

148. TS533. DII.

9. i. 27. From 54° 03' S, 36° 39' W to 54" 05' S, 36" 30' W, off Cape Saunders. Nets on trawl,
132-148 m.

About 40 cc. of coarse organic debris. Hyperammina subnodosa, Ammobaculites americanus and
Reophax subfusiformis very common. Many other large species, notably Bilocidinae and Arenacea, but
very few of the smaller South Georgian types, these presumably having been washed out. Armorella
sphaerica, Thurammina protea, Astrorhiza triangularis, Hippocrepinella and Pelosphaera occur.

149. TS 604. (See Fig. 2.)

ID. i. 27. Mouth of East Cumberland Bay. Nets on trawl, 200-234 m.

A quantity of muddy debris, mainly Annelid and Crustacean, yielded a long list of 104 species,
no less than twenty-one of which were not recorded elsewhere in South Georgia. The majority were
species of world-wide distribution and might reasonably have been expected in the area, but several
forms being normally of warm water habitat, it was decided to reject the whole of the twenty-one
species for fear that the gathering might have been inadvertently fouled by admixture with foreign
material. Armorella sphaerica and Pelosina fusiformis were common.

151. TS551. EI.

16. i. 27. 53° 25' S, 35° 15' W. Sounding rod, 3200 m.

About 5 cc. of grey mud with darker spots yielding a residue of dark sand, felted Diatoms and
Radiolaria. The Foraminifera were few, but interesting, including Cyclammina cancellata, Glomospira
charoides, Ammochilostoma pauciloculata and Ammobaculites foliaccus.

157. TS 541. D I.

20. i. 27. 53" 51' S, 36° 11' 15" W. Sounding rod, 970 m.

About 3 cc. of dark mud yielding a residue of angular sand grains. Diatoms, Radiolaria and a few
of the common species of Foraminifera.

660. TS 609. (See Fig. 2.)

0-6 miles E of Hope Point, East Cumberland Bay. Dredge, 216 m.

A quantity of slate-blue mud received dry in lumps. Difficult to clean and yielding very few
Foraminifera. Virgulina bradyi and V. schreibersiana were the only species occurring in any numbers,
but Ehrenbergina crassa and Miliammina oblonga were common. On the whole perhaps a typical
South Georgian mud.
No station no. TS 554. E I.

19. ii. 26. 53° 00' S, 34° 22' 30" W. Sounding rod, 2472 m.

About 12 cc. of pale grey mud — a Diatom ooze. Residue consisting of a few sand grains,
Diatoms, Radiolaria, veiy few Foraminifera, nearly all arenaceous.

38 DISCOVERY REPORTS

STATIONS MADE BY THE R.R.S. 'WILLIAM SCORESBY'

WS 18. TS 542. D II.

26. xi. 26. 54° 07' S, 36° 23' W. Sounding rod, 113 m.

About I cc. of refractory slate-blue mud, which resisted disintegration. Treated with soda it
yielded a residue of sand grains, mud pellets. Diatoms, sponge spicules and ten specimens of
Foraminifera representing seven species, including one specimen of Patellina conugata.

WS 20. TS 548. E I.

28. xi. 26. 53° 52' 30" 3,36° 00' W.

A few grains of dark angular sand, containing specimens of six common Foraminifera only,
brought up in a net lowered to 500 m., hauled to 250 m., then closed. The Lucas Sounding Machine
had previously given bottom at 535 m. — rock.

WS 25. TS 586-7. D II.

17. xii. 26. Undine Harbour (North). Small beam trawl, 18-27 "i-

Residues and mud from small beam trawl, containing many Foraminifera, Discorbis ghbidaris
being extremely common and Cassidiiliita crassa, Ehrenbergina crassa and Trochammiiia malovemis
very common. Glohigerinae frequent but very pauperate. Most other species very rare.

WS 26. No TS.

18. xii. 26. 53° 33' 15" S, 37° 45' 15" W. Sounding rod, 1 180 m.

A few cc. of slate-grey mud, yielding a residue of Diatoms, Radiolaria and fine sand grains.

WS 27. TS 526. C I.

19. xii. 26. 53° 55' S, 38° 01' W. Residue from nets, 106-9 "''•

A small quantity of dark muddy sand with organic debris and a few pebbles. Many Foramini-
fera. Cassidiilina crassa and Ehrenbergina crassa were very common. Many species, but few specimens
of Lagena.

WS28. TS 527-8. CI.

19. xii. 26. 53° 48' 15" S, 38° 13' W. Sounding rod, 150-346 m.
Two samples :

(i) About a cubic inch of black sand from net which touched bottom at 145 m. Ilaplophrag-
tnoides canarieiisis very common, other Arenacea frequent, including Pelosina fiisiformis, Hippocre-

pinella.

(2) From a sounding of 346 m., a small quantity of tenacious grey mud with very few Fora-
minifera, Miliammina only being common.

WS 31. TS 566. E II.

20. xii. 26. 54° 52' S, 35° 36' W. Sounding rod, 77 m.

Two stones and a very small quantity of sand. The stones had a few sessile specimens of
Psammosphaera fiisca, Tolypammina vagans and Glomospira gordialis, and the sand yielded many
specimens of shallow-water types representing six species.

WS32. TS565B. EII.

21. xii. 26. Mouth of Drygalski Fjord. Sounding rod, 225 m.

About 25 cc. of light grey mud, which left only 2 cc. on a 200-mesh sieve. Diatoms, fine sand
and abundant minute Foraminifera, Virgulina schreibersiana dominant, Textularia tenidssima common.

WS 33. TS 568 A-B, 569. E II.

21. xii. 26. 54° 59' S, 35° 24' W. Sounding rod, 135 m.

(i) Bottom sample — about 15 cc. of slate-coloured sandy mud with few Foraminifera. Ehrcn-

MATERIAL EXAMINED 39

bergina crassa was the commonest species. No less than fifteen species of Lagena, mostly represented
by one or two specimens.

(2) Debris from net which touched bottom at 130 m., sponge and zoophyte fragments, pebbles
and black sand with numerous Foraminifera, including Astrorhiza limicola, Planispirina bucculenta
var. placentiforwis, Turritellella laevigata, Nouria harrisii, Hippocrepina indivisa and other rare forms.

WS 36. No TS.

22.xii.26. 55° 20' IS" 8,34^46' 30" W.

A very small sample of pelagic material from a vertical haul (N 70 V) of 500-250 m. contained
many Globigerinae ; G. dutertrei most frequent, then G. coiiglomerata ; G. bulloides and G. pachyderma
present, but rare. Specimens rather thin- walled.

Another sample from 250 to 100 m. contained many Copepoda and Diatoms, but few Globi-
gerinae of the species G. dutertrei, conglomerata and bulloides.

WS 37. TS 563. E II.

22. xii. 26. 54° 45' S, 35° 11' W. Sounding rod, 318 m.

About 15 cc. of tenacious dark grey mud. Very little residue, Diatoms, sponge spicules, angular
sand grains. Foraminifera numerous and well developed, but limited in species. Many Lagenae.

WS 38. No TS.

22-23. ^"- 26. 54° 01' S, 35° 14' w.

A very small sample of pelagic material from a vertical haul (N 70 V), 250-100 m., yielded many
specimens of Globigerina dutertrei.

WS 39. No TS.

22. xii. 26. 54° 08' S, 35' 43' W. Sounding rod, 237 m.

About 13 cc. of slate-green mud yielding hardly any residue on a 200-mesh sieve. Diatoms and
spicules but no Foraminifera.

WS40. TS571. EIII.

7. i. 27. 55° 09' S, 35° 58' W. Sounding rod, 183 m.

About 12 cc. of dark slate-coloured mud. Residue of pebbles and sand grains of all sizes,
Diatoms, a few Radiolaria and Foraminifera, the latter well-developed specimens of the commoner
species.

A vertical haul (N70 V), between 175 and 100 m., yielded a very small sample of pelagic material,
principally Globigerina dutertrei; G. bulloides very rare. All the specimens very thin walled.

WS41. TS573. DII.

7- i- 27- 54° 32' 45" S, 36° 43' 45" W. Sounding rod, 140 m.

About 25 cc. of dark grey mud which yielded very little residue on the 200-mesh sieve, fine
sand. Diatoms and many Foraminifera of a few species only, Virgulina schreibersiana dominant.

WS42. TS 574-5. DII.

7. i. 27. 54° 41' 45" S, 36° 47' W. (i) Sounding rod, 175 m. ; (2) Net touched bottom, 198 m.

(1) About 15 cc. of tenacious grey mud. Residue of stones, sand. Diatoms and a few
Foraminifera.

(2) About 6 cc. of black sand with organic debris. Not many species of Foraminifera but some
interesting forms, including Hippocrepina indivisa (not found in Sample i), Hippocrcpinella hirudinea,
and Textularia tenuissima. Ehrenbergina crassa and other South Georgian species were common in
both samples.

WS 43. TS 576. D II.

7-8. i. 27. 54'' 54' S, 36° 50' W. Sounding rod, 200 m.

About 10 cc. of dark slate-coloured mud, difficult to wash. Residue of sub-angular sand grains,
Diatoms and a few Foraminifera of the commoner species.

40 DISCOVERY REPORTS

WS 44. No TS.

8.i.27. 55° 06' S, 36° 57' W.

A small quantity of pelagic material taken in a vertical haul (N 70 V) between 250 and 100 m.
yielded abundant Globigerinae. G. dutertrei was dominant, associated with G. conglomerata frequent,
and G. pachyderma rare. Only a few specimens of G. biilloides were present.

WS 45. TS 577. D II.

8. i. 27. 54° 38' 30" S, 37° 30' 55" W. Sounding rod, 180 m.

About 15 cc. of tenacious dark grey mud. Residue of stones and angular sand grains of all
sizes, Diatoms, scanty Foraminifera of common species.

WS46. TS578. DII.

8. i. 27. 54° 20' 15" S, 37° 32' 30" W. Sounding rod, 194 m.

About 12 cc. of slate-coloured mud. Residue of pebbles and sand grains of all sizes, Diatoms,
very few Foraminifera, mostly common species, but including a specimen of Lageua spiimosa.

WS 47. TS 579. D II.

9. i. 27. 54° 22' S, 37" 50' W. Sounding rod, 160 m.

About 15 cc. of very tenacious blue-grey mud, yielding 2 cc. of residue, consisting of angular
black sand with abundant Diatoms, but very few Foraminifera, which, however, included some
interesting forms : Proteonina tiibulata, Textidaria temtissima, and Coniuspira seheyeusis.

WS 48. TS 580. C II.

9. i. 27. 54° 24' S, 38° 09' W. Sounding rod, 224 m.

About 15 cc. of dark slate-coloured mud. Residue of pebbles and angular sand grains. Diatoms,
very few Foraminifera of fourteen species only, six of which were Lageua.

WS 49. No TS. C II.

9. i. 27. 54'^ 28' S, 38° 22' 15" W. Sounding rod, 223 m.

About 15 cc. of tenacious dark grey mud. Very little residue, sand grains, spicules. Diatoms.
Only a few Foraminifera of the commoner species.

WS50. TS581. CII.

9. i. 27. 54° 30' 30" S, 38° 40' 30" W. Sounding rod, 230 m.

About 10 cc. of tenacious dark slate-coloured mud. Residue principally of a felted mass of
Diatoms with many Foraminifera of the usual common species, Virgulina schreibersiana dominant.

WS51. TS582. CII.

9.1.27. 54° 34' S, 38° 57' W. Sounding rod, 210 m.

A few small sub-angular pebbles and sand grains, many of the pebbles covered with sessile
Foraminifera, including a specimen of VanhoeffencUa gaiissi loosely attached by extruded proto-
plasm. The few species obtained from the sand included Milianimina lata and Ehrenbergina hystrix
var. glabra, suggesting an area rich in species.

WS52. TS585. CII.

ID. i. 27. 54° 03' 30" S, 38° 35' W. Sounding rod, 84 m.

About 15 cc. of tenacious dark grey mud. Residue of angular black sand. Diatoms, sponge
spicules, Radiolaria, and twelve species of Foraminifera, represented by one or two specimens only,
except Miliammina oblonga and M. obliqua, which were frequent.

WS61. NoTS.

18. i. 27. 53° 37' 30" S, 37" 06' 30" W. Sounding rod, 1893 m.

A few small pebbles, attached to which were four species of sessile Foraminifera. Also some
sand grains, many glauconitic, some Radiolaria and Diatoms, but no other species of Foraminifera.
The sessile specimens were ToJypammina vagans common, Placopsiliiia cenomana many, Glomospira
gordialis several, and Psammosphaera fusca one only.

MATERIAL EXAMINED 41

WS63. TS583A. CII.

20. i. 27. 54° 36' S, 39° 14' W. Sounding rod, 1752 m.

About 12 cc. of dark grey mud. Residue of small pebbles and angular sand grains of all sizes,
Radiolaria, Diatoms, very few Foraminifera. Globigerina pachyderma dominant.

Between WS 63 and 64. TS 583 B. C I, II.

21. i. 27. 54° 36' S, 39^ 14' W to 53° 48' 45" S, 38' 34' W. Sounding rod, 251 m.

About 12 cc. of dark slate-coloured mud. Residue of a felted mass of Diatoms with a few
Foraminifera. Virgulina schreibersiana and Globigerina pachyderma were extremely common, the few
other species rare or very rare.

WS66. TS525E. A I.

18. ii. 27. 53° 31' 15" S, 42° 03' 30" W. Sounding rod, 150 m.

A very small sample of dark angular mineral grains, broken shells and many Foraminifera,
including Cassidiiliiia laevigata var. tiimida, and Ehrenbergina hystrix var. glabra. Dominant species,
Cassidulina crassa, C. subglobosa and Globigerina dutertrei.

WSllO. NoTS.

26. V. 27. 53° 46' S, 35° 47' W. Sounding rod, about 1000 m.

Three stones with sessile Foraminifera, mostly damaged. There were many specimens of
Tholosina vesiadaris and one each of Crithionina mamilla and Truncatidina lobatiila.

WS113. TS543. DII.

28. v. 27. 54° 07' S, 36° 24' W. Sounding rod, 155 m.

Three stones coated with tenacious dark blue mud. Residue mainly of Diatoms, the few Fora-
minifera being pauperate. Miliammina oblonga, M. obliqtia and Trochanmmia malovensis were the
only species represented by many specimens. Very fine specimens of Sorosphaera dcpressa on one of
the stones.

WS 154. TS 535. D I.

26. ii. 28. 54° 00' S, 36° 52' W. Net touched bottom, 160 m.

About two ounces of mud and organic debris, including a large worm tube ( ?) covered with
Rhizammina algaeformis and several other sessile species. Hyperammina subnodosa, Vanhoejfenella,
Hippocrepinella, Pelosina, Armorella and Tholosina occurred.

WS 177. TS 564. E II.

7. iii. 28. 54° 58' S, 35" 00' W. Net touched bottom, 97 m.

A shell, stones, and a worm tube, with a few grains of finer sand, which yielded many species of
Foraminifera, suggest that the locality possesses a rich fauna.

WS 314. TS 525 F. B I.

I. xii. 28. 53° 36' S, 41° 05' W. Sounding rod, 137 m.

A very small sample ; angular mineral grains and shell fragments with abundant Foraminifera,
including Patellina corrugata and Polyinorphina inlliamsoni. Dominant species, Cassidulina crassa,
C. subglobosa, Globigerina dutertrei, G. pachyderma.

WS 322. No TS. C I.

16. xii. 28. 53° 45' 30" S, 38° 23' W. Sounding rod, 258 m.

A few large sand grains without sessile organisms or Foraminifera.

WS 329. No TS. D I.

27. xii. 28. 53° 56' 30" S, 36° 06' W. Sounding rod, 165 m.

About \ cc. of dark sand with only a few specimens representing three common species of
Foraminifera.

42 DISCOVERY REPORTS

WS 334. TS 552. E I.

30. xii. 28. 53° 19' S, 35" 10' 30" W. Sounding rod, 3705 m.

About 15 cc. of grey ooze with dari^ specks. Residue of pebbles and angular sand grains of all
sizes. Radiolaria, abundant Diatoms and many arenaceous Foraminifera, often fragmentary.
Marsipella cylindrica, Proteouwa diffltigiformis, Haphphragmoides subglohosiis, Psamniospliaera fiisca
were all frequent. The smaller species were poorly represented but included Ammochilostoma galeata
and Textularia nitens.

WS 336. TS 553. E I.

30. xii. 28. 53° 06' S, 34° 44' W. Sounding rod, 3647 m.

About 8 cc. of pale grey mud with darker specks. Residue of pebbles and sand grains of all
sizes, Radiolaria and Diatoms. Very few Foraminifera, all arenaceous except Verneuilina bradyi and
ClavuUna communis.

WS 340. No TS.

8. i. 29. 53° 32' S, 37° 12' 30" W. Sounding rod, 740 m.

A few sand grains only, the finer material probably washed out on the way up. No organisms.

WS 343. TS 532. D I.

8. i. 29. 53° 02' S, 37^ 06' W. Sounding rod, 2856 m.

About 17 cc. of greenish grey mud which left only 2 cc. on the 200-mesh silk sieve. Residue of
Diatoms, Radiolaria and fine angular sand. Foraminifera very starved, and extremely rare, only six
species being recorded, among which were Textularia tenuissima and T. nitens.

WS 344. No TS.

9.1.29. 52° 50' S, 37° 01' W. Sounding rod, 2215 m.

A few tiny pebbles, one of which bore a specimen of Tolypammina vagans, the only organism
seen.

WS 345. No TS.

9.1.29. 52° 41' S, 37° 06' W. Sounding rod, 2174 m.
A single tiny pebble without sessile organisms.

WS348. TS558. EII.

II. i. 29. 54° 23' 10" S, 35° 52' W. Sounding rod, 135 m.

About 55 cc. of tenacious slate-coloured mud. Residue largely of Diatoms, sponge spicules and
fine sand. Very few Foraminifera, all of common types.

WS 349. TS 559. E II.

ii.i. 29. 54° 23' S, 35° 32' 30" W. Sounding rod, 267 m.

About 35 cc. of fine grey mud with hardly any coarse residue, fine sand, Diatoms, spicules,
Radiolaria. Not many Foraminifera, all very small and pauperate. Virgulina schreibersiana and
Globigerina pachyderma common.

WS351. TS561. EII.

II. i. 29. 54° 21' 30" S, 34° 59' W. Sounding rod, 1170 m.

About 15 cc. of blue-grey mud. Residue of pebbles and sand grains of all sizes, Diatoms,
Radiolaria and very few Foraminifera. These included Cyclammina canceUata, Marsipella elongata,
Proteonina tiibulata, Lagetia fimbriata and Bolivina cincta, six species of Globigerina and three of
Globorotalia.

WS 353. TS 562. E II.

12.1.29. 54° 18' S, 34^ 25' W. Sounding rod, 4041 m.

About 2^ cc. of light grey ooze. Residue consisting of a flocculent mass of Diatoms, Radiolaria
and angular sand grains in about equal proportions. Very few Foraminifera and, except for a few
large fragmentary Arenacea, all were small and pauperate. No Globigerinidae or Rotaliidae.

MATERIAL EXAMINED 43

WS 357. TS 567. E II.

13- i- 29- 54^' 55' 30" S, 35 30' W. Sounding rod, 264 m.

About 25 cc. of dark blue mud. Residue of pebbles, sub-angular sand grains of varying sizes,
Diatoms, spicules, and a very few Foraminifera, Uzigerina aiigulosa dominant.
WS361. NoTS. EIII.

14. i. 29. 55" 24' S, 34° 42' W. Sounding rod, 1444 m.

About 0-5 cc. of coarse sand, with three sessile arenaceous species attached to the larger grains,
and a few specimens of Globigcrina pachyderma among the finer portion.
WS 363. No TS. E III.

14. i. 29. 55° 38' S, 34° 14' W. Sounding rod, 1332 m.

A small pebble and a little sand. Tolypammina vagaiis and Tltohsina vesicularis sessile on the
pebble ; no other organisms.

WS 365. TS 570. F III.

H-i-29- 55° 52' 10" S, 33° 53' W. Sounding rod, 3219 m.

About 20 cc. of light grey mud containing pebbles and sand grains, many Radiolaria and
Diatoms but very few Foraminifera. Only six species were identified in all.
WS 373. TS 556. E II.

22. i. 29. From 54° 10' S, 35° 40' W to 54^ 27' S, 34° 58' W. Sounding rod, 1540 m.

About 4 cc. of yellowish grey mud with dark specks. Residue of angular sand grains of all sizes,
Radiolaria, Diatoms and very few Foraminifera.

WS 418. TS 572. D III.

10. iv. 29. 55° 02' 30" S, 36° 31' W. Sounding rod, 227 m.

About 2 cc. of dark sandy mud. Residue of pebbles and sand grains with scanty Foraminifera.
Uvigerina angulosa and Glohigerina pachyderma dominant, Miliammina curiously rare.

WS 425. No TS. C II.

15. iv. 29. 54° 37' 10" S, 39° 07' 30" W. Sounding rod, 228 m.
One small black pebble without sessile organisms.

WS426. TSS84. CII.

15. iv. 29. 54° 38' 35" S, 39^ 22' W. 1500-2000 m.

Some pebbles and a few grains of coarse sand obtained from a reversing water-bottle, which
touched bottom between these depths. Many sessile Foraminifera on the pebbles and seven species
obtained from the sand, including Cyclammina cancellata, Ammodisats incertus and Glomospira
gordialis, suggest a rich local fauna. A bottom-sample of pebbles only from 2433 m. furnished the
same sessile species.

WS428. TS525D. A I.

29. iv. 29. 53° 07' S, 42'' 30' W. Sounding rod, 1966 m.

A very small quantity of dark grey sand, with much glauconite, some Radiolaria, Diatoms and a
few Foraminifera. Globigerina dutertrei was the only species occurring with any frequency.
WS 429. TS 525 B.i

30. iv. 29. 53° 02' 30" S, 45° 28' W (between South Georgia and the Falkland Islands).
Sounding rod, 2549 m.

Fine grey mud yielding a residue of Radiolaria, Diatoms, fine angular mineral grains and many
Foraminifera, including Proteonina tubidata, Reophax spiculifer, Textiilaria >ntens, and many species of
Lagena. Dominant species, Globigerina pachyderma, G. dutertrei, Globorotalia crassa, NonioneUa iridea.
WS521. TS52SA.

28. ii. 30. 52° 41' S, 49" 14' W (between South Georgia and the Falkland Islands). Sounding
rod, 3780 m.

1 This Station (and the next three) are outside the range of West Longitude of Fig. i.

44 DISCOVERY REPORTS

Light grey mud yielding a residue of mineral grains of all sizes, Globigerina ooze and Radiolaria
in about equal proportions. Globigeritiae of several species dominant. Many interesting and rare
species including Bolivina cincta, B. decussata, Ehrenbergina bradyi, Lagena formosa.

WS 522. TS 525 G.

28. ii. 30. 52" 56' S, 47° 14' W (between South Georgia and the Falkland Islands). Sounding
rod, 2550 m.

Light grey Globigerina ooze with angular pebbles, sand grains and glauconite. Globigerina and
Globorotalia of various species formed quite 98 per cent of the organic remains. A long list of other
Foraminifera mostly represented by single specimens, including many Lagenae.

WS 523. TS 525 C.

2. iii. 30. 53° 07' S, 45° 00' W (between South Georgia and the Falkland Islands). Sounding
rod, 1697 m.

Grey mud, yielding a residue of Radiolaria, Diatoms and fine mineral grains with a few Fora-
minifera, including Hormosina globuUfcra, Rcophax spiculifer and R. robustus. Globigerina pachy-
dcrma dominant, forming the bulk of the organic residue.

No station no. TS 565 A. E II.

21. xii. 26. Drygalski Fjord (to NW of Station WS 32). Sounding rod, 178 m.

About I cc. of tenacious blue mud and a pebble. Residue of sand grains and mud pellets with
ten species of Foraminifera. Pelosina rotundata, P. variabilis were common, Miliammina frequent,
Spiroplectaniniina bifortnis and Tcxhdaria tenuissima very rare.

STATIONS MADE BY THE STAFF OF THE MARINE BIOLOGICAL STATION

MS 14. TS 605. (See Fig. 2.)

17. ii. 25. From 1-5 miles SE x S to 1-5 miles S \° W, of Sappho Point, East Cumberland Bay.
Small dredge 109-180 m.

A small quantity of coarse black angular sand with organic debris (residues from small dredge),
yielded many species of Foraminifera, mostly rare, except Milianiniina oblonga and Ehrenbergina
crassa, which were common.

MS 68. TS 606-8. (See Fig. 2.)

2. iii. 26. 1-7 miles S J° E to 8| cables SE x E of Sappho Point, East Cumberland Bay.
Rectangular net 220-247 m.

Several gatherings of mud, sand and sponge debris from nets, yielded a large number of species,
including many rarities, in spite of the unsatisfactory nature of the material, which was very hard to
clean. The area would appear to be very rich in Foraminifera if sufficient material could be collected.
Hippocrepinella hirudinea, H. alba, Hippocrepina oviformis, Textularia tenuissima, Miliammina lata,
etc.

LIST OF NEW GENERA, SPECIES AND VARIETIES

Flintia sohita *Pclospliaera, gen.n.

*Gordiospira, gen.n. *Pelosphaera cornuta

*Gordiospira fragilis Proteonina decor ata

Astrorhiza triangularis Webbinella liniosa

Vanhoejfenclla ocidus *Arniorclla, gen.n.

Pelosina fusiformis *Armorella sphaerica

Pelosina variabilis var.n. constricta Tliiirammina protea

Stortliosphaera elongata var.n. impudica *Hippocrepinella, gen.n.

*Sorosphaera depressa * Hippocrepinella hirudinea

MILIOLIDAE

45

*Hippocrepinella hinidinea var.n. crassa
*Hippocrepinella alba

Reophax subfusiformis

Reopluix distaiis var.n. gracilis

AmmobacuUtes bargmanni
*Animobaculites rostraius

Tiirritcllella laevigata
*Miliammina, gen.n.
*Miliammina cribrosa nom.n. (p. 90)
*Miliamntina lata
*AIiliainniina obliqua

*Miliammina oblonga
Textularia nitens
Textularia temiissima nom.n.
Textularia wiestwi
Bigenerina minutissima

*Elireiibergina crassa
Lagena formosa var.n. costata
Lageiia hartiana
Lagcfia herdmani
Lagena mackintoshiana
Discorhis maraaritaceus

* These genera and species were figured and described by E. Heron-Allen and A. Earland in the Journal
of the Royal Microscopical Society in 1929-32. The descriptions and plates are repeated in this Report by
the courtesy of the Council of the Society.

SYSTEMATIC ACCOUNT

Note. To economize space, no synonyms are given for species which have already been
described in the Report on the Falkland area. For purposes of reference, the Falkland
No. is printed in brackets after the specific name: e.g. Biloculina murrhyna, Schwager

(F3)-

Order FORAMINIFERA

Family MILIOLIDAE

Sub-family MILIOLININAE

Genus Biloculina, d'Orbigny, 1826

Note. The BiloaiUnoe have been determined from external characters only, as time
has not allowed for the preparation of sections. The genus is difficult to diagnose with
certainty, owing to the lack of superficial distinctions.

1. Biloculina murrhyna, Schwager (F 3).
Three stations: WS 429, 521, 522.

Rare everywhere and the specimens are small and pauperate, the best being at WS
522. All the stations are in deep water, between 2000 and 4000 m.

2. Biloculina serrata, Bailey (F 4).
One station: WS 523.

A single weak and broken specimen. The serration is confined to the aboral edge.

3. Biloculina bradyi, Schlumberger.

Biloculina ringens, Brady, 1884, FC, p. 142, pi. ii, fig. 7.

Biloculina bradyi, Schlumberger, 1891, BGF, p. 557 (in the reprints p. 170), text-figs. 15-19,

pi. X, figs. 63-71.

Five stations: 27, 45, 126; WS 154, 177.

Common at several of the stations; the best specimens at WS 154 and WS 177.

46 DISCOVERY REPORTS

4. Biloculina vespertilio, Schlumberger (F iC^).
Five stations: 27, 45, 123, 144, 148.

Confined to the Cumberland Bay area, where it is frequent and attains large dimen-
sions, notably at St. 45.

5. Biloculina milne-edwardsi, Schlumberger.

Bilocidina milne-edioardsi, Schlumberger, 1891, BGF, p. 567 (in the reprints p. 180), text-figs.
29, 30, pi. xi, figs. 79, 80.

Five stations: 27, 45, 126, 144, 148.

This species appears to be almost confined to the Cumberland Bay area, where it is
common and attains large dimensions, particularly at Sts. 45 and 144.

6. Biloculina elongata, d'Orbigny (F 6).

Twelve stations: 27, 30, 45, 123, 126, 144, 148, 149; WS 33, 42, 113; MS 14.

Never very common. Most frequent at St. 144, where the specimens were of a small
regular type, very Uke our British form. This small form is the usual type, but the
species attains a large size at some stations, notably Sts. 144 and 148. Occasional large
specimens are found at stations where the small type prevails. From St. 144 a large
specimen was received preserved in spirit, with a branching mass of protoplasm and mud,
extending from the orifice in lobose processes, its bulk largely exceeding that of the test.
Protoplasmic masses, extracted from spirit specimens obtained at this station and
stained, were found to be loaded with Diatoms taken in as food, but there was no mud in
this internal protoplasm.

y. Biloculina patagonica, d'Orbigny (F 7).
Four stations: 27, 126, 140, 144.
Very rare always, the best specimens at St. 126.

8. Biloculina anomala, Schlumberger (F 10).

Five stations: 27, 45, 149; WS 25, 33.

A few specimens appear to be referable to this species, so far as external character-
istics are concerned, but Schlumberger himself admits the difficulty of identification.
They are always uncommon.

9. Biloculina pisum, Schlumberger.

Biloculina pisum, Schlumberger, 1891, BGF, p. 569 (in the reprints p. 182), text-fig. 31, pi. xi,
figs. 81-3.

Nine stations: 27, 123, 126, 140, 144; WS 66, 177, 314, 522.

MILIOLIDAE 47

Not very common, except at WS 177, where many excellent and typical specimens
were obtained. Larger, but less typical, and rare at Sts. 126, 144.

10. Biloculina globulus, Bornemann (F 11).
Three stations: 27; WS 27, 33.

Very rare, mostly single specimens.

Genus Flintia, Schubert, 191 1

11. Flintia soluta, sp.n. (Plate I, figs. 1-4).
Two stations: 27, 144.

Test large, porcellanous, smooth and highly polished; the last two chambers are
extremely inflated and separated by a space within which the ante-penultimate chamber
is exposed to a variable extent. Aperture large, with slightly recurved lip, furnished
with a broad incurved tooth having rounded extremities. Length up to 2-0 mm. In the
best specimen, the greatest width of the final chamber was 0-64 mm., of the penultimate
chamber 0-9 mm., while the width of the median space or ante-penultimate chamber
was only 0-24 mm.

Very rare, a single specimen at St. 144, several at St. 27. No attempt has been made
to cut sections owing to paucity of material, but the internal structure is doubtless
regularly biloculine, the separation being confined to the final chambers.

This is apparently the most rudimentary form of the genus, and its approximation to
Biloculina is so close that the specimens were at first regarded as mere abnormalities,
there being considerable difl^erences in the extent of exposure of the earher chambers,
even among the few specimens which have been found. The extremely inflated contour
shows some resemblance to Biloculina isobelleaua, d'Orbigny (F 9), but the aperture is
ver}' diff'erent. In this feature, and in the curvature of the edge of the chambers, there is
a close resemblance to Biloculina peruviana, d'Orbigny (F 8), from which it is probably
derived.

Genus Miliolina, WiHiamson, 1858

12. Miliolina seminulum (Linne) (F 12).

Ten stations: 27, 123, 149, 660; WS 27, 28, 43, 177, 314, 522.

Singularly rare, seldom more than one or two specimens at each station.

13. Miliolina vulgaris (d'Orbigny) (F 14).

Four stations: 27, 45 ; WS 27, 66.

A single large specimen at each station, except at WS 66, where the example is very
small.

14. Miliolina oblonga (Montagu) (F 15).
Four stations: 45, 144, 149; WS 27.

Not uncommon at the three Discovery stations, where it is represented by a small thin-
shelled form. At WS 27, it is represented by a single large and normal specimen.

48 DISCOVERY REPORTS

15. Miliolina bosciana (d'Orbigny) (F 16).
Four stations: 140, 144, 149; WS 33.

Rare and very small at WS 33. More frequent at the other stations. All the specimens
are very thin-walled.

16. Miliolina subrotunda (Montagu) (F 18).

Six stations: 45, 144, 145, 157; WS 25, 27.

Typical but verj-- rare, except at WS 25, where it is fairly frequent.

17. Miliolina lamarckiana (d'Orbigny) (F 20).
One station : WS 50.

A single small specimen.

18. Miliolina pygmaea (Reuss) (F 25).
Three stations: 149; WS 32, 33.

Very rare, the best specimens at WS 32 ; very small and pauperate at St. 149.

19. Miliolina venusta (Karrer) (F 26).
One station: WS 2^.

A single small specimen.

20. Miliolina tricarinata (d'Orbigny) (F 28) (Fig. 3).
Eight stations: 45, 144, 148; WS 33, 43, 429, 522, 523.

/ m 'a n

Fig. 3. Variations in the aperture of Miliolina tricarinata.

Usually rare and very small at WS 429, 522. Larger at WS 33. At St. 144 the species
is common and attains a very large size up to 2-0 mm. in length. The aperture at this
station varies with the size of the individual, being typically Milioline in the smallest,
cruciform in the medium, and irregularly stellate in the largest specimens. The medium

MILIOLIDAE 49

specimens are identical with the figure of Cnicilociilina triangularis, d'Orbigny, recorded
by him from the Falkland area, but which was not discovered in the Falkland material.
The accompanying figure illustrates some of the range of variation in the aperture.

Wiesner has recorded as a new variety, M. tricarinata var. crucioralis (W. 193 1,
FDSE, p. 105) what is apparently a typical CniciloaiJina triangularis, d'Orbigny.
Wiesner does not figure his specimens, but the description is of an identical organism.

21. Miliolina circularis (Bornemann) (F 29).

Six stations: 123, 126, 144, 145; WS 27, 28.

Except at WS 27, where small specimens are common, the species is very rare. Seldom
more than a single specimen, which is often a large one, at each station. Very large at
Sts. 126 and 144.

Genus Sigmoilina, Schlumberger, 1887

22. Sigmoilina obesa, Heron-Allen and Earland (F 38).
One station : WS 27.

A single specimen only.

23. Sigmoilina tenuis (Czjzek) (F 40).
Four stations: WS 334, 353, 429, 521.

Very rare everywhere, but exhibiting the same range of convexity as in the Falkland
area.

Sub-family HAUERININAE
Genus Tubinella, Rhumbler, 1906

24. Tubinella funalis (Brady) (F 41).
Six stations: 30, 123, 149; WS 25, 27, 33.

Rare everywhere, but large and typical specimens at Sts. 123, 149, WS 27. Wiesner
(FDSE, i93i,pp. 67, i09,pl. i,fig. 6; pl.xv, fig. 183 ;pl.xvi,figs. 184-5) has made this
species the type of a new genus TubineUi)w, on the grounds of a suspected relationship
with Tiihinella perforata, Rhumbler (FLC, 1906, p. 27, pi. ii, fig. 5). But the similarity
appears to be external only. Rhumbler 's species is distinctly perforate, and if really a
foraminifer, would be widely separated from T. funalis, which exhibits no sign of
perforation of the wall. Rhumbler would appear to have assigned his specimen to
Tubinella, akin to Articulina, largely because he supposed the specimen figured by
Millett under the name Articulina funalis var. inornata (Millett, 1898, etc., FM, 1898,
P- 5^3' pl- xii, fig. 1 1) to indicate signs of perforation. Millett's type specimens are in the
H.-A. and E. collection and I have subjected them to an examination under a high power.
Their nature appears to be rather uncertain, but they are not perforate. The markings
shown in Millett's drawing are foreign bodies, some being food particles.

so DISCOVERY REPORTS

Genus Planispirina, Seguenza, 1880

25. Planispirina irregularis (d'Orbigny) (F 43).

Seven stations: 140, 144; WS 27, 33, 154, 177, 314.

A good series in all stages of growth at St. 140, and WS 33 and 154, and single
excellent specimens at the other stations.

26. Planispirina sphaera (d'Orbigny) (F 44).

Three stations: 27, 144, 148.

Many large and typical specimens at St. 144, single good individuals at the other
stations.

27. Planispirina bucculenta (Brady) (F 45).
Two stations: 27, 144.

A single large specimen at St. 27. Many in all stages of growth at St. 144.

28. Planispirina bucculenta var. placentiformis (Brady).

Miliolina bucculenta var. placentiformis, Brady, 1884, FC, p. 171, pi. iv, figs, i, 2.
Planispirina bucculenta var. placentiformis, Chapman, 1914, EDRS, p. 43, pi. v, fig. 5.

One station : WS 33 .

Many specimens in all stages of growth.

Sub-family PENEROPLIDINAE
Genus Cornuspira, Schultze, 1854

29. Cornuspira involvens (Reuss) (F 46).

Nine stations: 30, 123, 144, 149; WS 25, 27, 33, 113, 348.

Always rare or very rare except at St. 144, where it was frequent in both megalo-
spheric and microspheric forms, and St. 149, where the microspheric form was frequent
and larger than the average, the South Georgian specimens being usually small. The
megalospheric form is as usual predominant.

30. Cornuspira selseyensis, Heron-Allen and Earland (F 48).
Seven stations: 27, 45, 123, 144, 145, 149; WS 47.

Always very rare, but good specimens especially at Sts. 45, 144 and WS 47.

31. Cornuspira foliacea (Philippi) (F 50).
Five stations: 45, 144, 148, 149; MS 68.

Always very rare but good specimens, notably at St. 144, where a perfect individual
over 3 mm. in diameter was found. Fragments of specimens almost as large were seen at
MS 68.

MILIOLIDAE 51

32. Cornuspira diffusa, Heron-Allen and Earland (Plate I, figs. 5-7).

Cornuspira foliacea, Brady, 1884, FC, pi. xi, fig. 7 (monstrous specimen, no ref. in text).

Cornuspira foliacea, Rhumbler, 1903, ZRR, p. 287, fig. 141 b.

Cornuspira diffusa, Heron-Allen and Earland, 1912, etc., NSG, 1913, pp. 272-6, pi. xii; CI,

1913. P- 37; FWS, 1916, p. 217.

Cornuspirella diffusa, Cushman, 1918, etc., FAO, 1929, p. 85, pi. xxi, figs. 6, 7.

Two stations: 149; MS 68.

Many characteristic fragments at each station.

Genus Gordiospira, Heron-Allen and Earland, 1932.

Test free, porcellanous, very thin-walled and fragile, approximately circular in shape,
consisting of a proloculum around which a non-septate tubular chamber forms several
coils in different planes, finally becoming planospiral, and involute for several con-
volutions. In the planospiral stage the tube rapidly expands in width and thickness. The
umbilical area is depressed and exhibits the edges of some of the earliest convolutions.
Aperture large and terminal.

Gordiospira is isomorphous with Glomospiro, Rzehak (1885), but the irregular con-
volutions of the initial coil are less visible externally. They become very evident in
transparent preparations.

33. Gordiospira fragilis, Heron-Allen and Earland (Plate VI, figs. 10-15).
Gordiospira fragilis, Heron-Allen and Earland, 1929, etc., FSA, 1932, p. 254, pi. i, figs. 1-6.

Six stations: 45, 144, 145, 149; WS 33; MS 68.

Test free, porcellanous, oval when young, becoming circular with full growth, very
thin and fragile, papery white or translucent, surface often irregular and marked with
recurved lines of growth. Viewed as an opaque object, it exhibits 2-3 planospiral and
embracing whorls of a tube, which increases in diameter and thickness so rapidly that
the final convolution forms the bulk of the entire test. The central portion of the test is
depressed, and shows one or two transverse tubes. The aperture is terminal, very
large, the outer margin projecting, the inner margins recurved to join the previous
whorl.

Viewed as a transparent object, Gordiospira fragilis is seen to consist of a proloculum
around which an unseptate tube is irregularly coiled in 3-5 convolutions set in different
planes. Subsequently, the tube becomes planospiral, forming 2-3 convolutions rapidly
increasing in size and thickness. These later convolutions are involute to some extent,
each concealing at least half of the previous convolution.

The surface of the tubes is often rather irregular and always exhibits faint recurved
lines of growth.

Both megalospheric and microspheric forms have been identified, the latter being the
larger, as usual. The megalospheric proloculum is about 0-02 mm. in diameter, the
microspheric too small to be measured with certainty. The size of the test ranges up to
1-5 mm. or rather more, in diameter, but the general average is under i-o mm.

S3 DISCOVERY REPORTS

Small specimens are usually oval in contour, owing to the change in shape when the
tube assumes the planospiral condition. After the first planospiral convolution, it
rapidly assumes a more circular contour.

Specimens taken direct from spirit, stained and mounted in balsam show that the
protoplasmic body is voluminous, almost filling the tube from end to end. The proto-
plasm is finely granular and filled with food bodies, including Diatoms and spicules.

Gordiospira fragilis was observed at six stations in South Georgia, and is frequent at
Sts. 45, 149, MS 68, at all of which a series in all stages was obtained. In depth its
range extends between 26 and 270 m.

Family ASTRORHIZIDAE
Sub-family ASTRORHIZINAE
Genus Astrorhiza, Sandahl, 1857

34. Astrorhiza limicola, Sandahl (Plate I, fig. 32).

Astrorhiza limicola, Sandahl, 1857, Ofvers. K. Vet. Ak. Fork., xiv, p. 299, pi. iii, figs. 5, 6.
Astrorhiza limicola, Brady, 1884, FC, p. 231, pi. xix, figs. 1-4.

Three stations: 27, 144; WS 33.

The rarity of this species in what might have been regarded as favourable surroundings
is rather surprising. A single specimen only was found at each station, and every example
is small compared with the individuals dredged off our own coasts. The specimen from
WS 33 is noteworthy as incorporating many long sponge spicules in its material. These
spicules project irregularly, and their presence is probably quite fortuitous, though they
may serve as supports in the ooze. I do not recall having previously seen spicules so
used in this genus.

35. Astrorhiza crassatina, Brady.

Astrorhiza crassatina, Brady, 1879, etc., RFC, 1881, p. 47; 1884, FC, p. 233, pi. xx, figs. 1-9.
Astrorhiza crassatina. Goes, 1894, ASF, p. 13, pi. ii, figs. 11-15.
Astrorhiza crassatina, Flint, 1899, RFA, p. 265, pi. ii.

One station: 45.

A single recognizable fragment.

36. Astrorhiza triangularis, sp.n. (Plate I, figs. 8, 9).

Two stations: 144, 148.

Test triangular, compressed and cushion-shaped, consisting of a thin shell of sand
grains of varying sizes firmly cemented together and enclosing the single chamber,
which occupies the whole interior of the test. An aperture at each corner, flush or
provided with a short external tube. Surface rough, owing to the projecting sand grains.
Average length of side, excluding oral tubes, about 1-3 mm. Tube up to 0-3 mm. in
length.

ASTRORHIZIDAE 53

Very rare at the two stations at which it was observed, but eight specimens in all were
found.

It differs from A. augulosa, Brady, in the character of the interior chamber, which is
large and occupies the whole body of the test, the apertures opening directly into it. In
Brady's species, the walls, built of fine sand only, are very thick and the central chamber
is not much more than a junction, formed by the three tubes diverging to the orifices at
the angular points.

Genus Iridia, Heron- Allen and Earland, 19 14

37. Iridia diaphana, Heron-Allen and Earland (F 52).
Two stations: 30; WS 25.

Several very good specimens were found at WS 25, and a single, less typical, at St. 30.

Genus Vanhoeffenella, Rhumbler, 1905

38. Vanhoeffenella gaussi, Rhumbler (Plate I, figs. 16-21).

Vanhoeffenella gaiissi, Rhumbler, 1905, MF, p. 105, fig. 9; 1909, FPE, p. 216, fig. 57.
Vanhoeffenella gaussii, Heron-Allen and Earland, 1922, TN, p. 76, pi. i, figs. 14, 15.

Nine stations: 27, 45, 126, 140, 144, 148; WS 51, 154, 334.

Frequent at Sts. 27, 140, 144, 148, rare or very rare elsewhere.

Vanhoeffenella has hitherto been known only from Rhumbler 's original description ;
our own note on two specimens found in the Terra Nova material (exact locality
unknown), and recently from a record by Wiesner (W 193 1, FDSE, p. 78, pi. iii,
figs. 21-6) whose specimens came from the Antarctic, near the coast of Kaiser Wilhelm
Land, depth 70-385 m. Its occurrence, in considerable numbers, in some of the South
Georgian material greatly enlarges our knowledge of the genus, and enables me to correct
the observations made on the Terra Nova specimens. We then described "the angular
framework supporting the characteristic chitinous membrane which forms the two faces
of Vanhoeffenella" as "a hollow tube with labyrinthic interior, constructed of minute
Diatom and mineral debris".

The present examination of a long series of specimens proves that this tubular
appearance is only assumed by collapsed and pauperate individuals, and that the struc-
ture is really much more simple.

Vanhoeffenella may be compared to a tambourine, the parchment sides of which are
represented by the two chitinous faces. These are separated, and in life kept widely
apart, by the rim, which is a flat belt of chitin agglutinated with mud, fine sand and
Diatoms. At intervals this belt widens out and forms tubular extensions. These vary in
number from one or two in the smallest individuals to as many as six or seven in large
specimens. They are formed by the extension of one of the chitinous faces together with
one edge of the sandy belt. The chitin forms the greater part of one side of the tube, the
sandy rim folding over on itself to complete the other side, and the end. The tubes are
formed from either edge of the belt ; the sandy side of a tube may be either uppermost or
underneath, as the individual lies on one chitinous face. Most specimens show tubes

4-2

54 DISCOVERY REPORTS

formed from either edge. The development of the first few tubes in many specimens
exhibits great regularity, the tubes being spaced evenly round the rim. This regularity
seldom persists beyond the formation of the third tube, specimens with five to seven
tubes often having them very irregularly disposed.

After death there is a tendency for the chitinous sides of the tambourine to collapse
and come together, with the result that the flat sandy belt folds inwards and forms an

encircling tube.

The earHest stage of Vanhoeffenella observed is oval with a tube at each extremity.
The smallest specimen seen measured only 0-45 mm. in greatest diameter, including the
tubes. It was sessile on a pebble at WS 51, attached by protoplasm extending from the
two tubes. Development proceeds at first by an expansion of one side of the oval from
which a third tube is extended, and later by constant enlargement of the rim with the
formation of additional tubes. The greatest diameter of the adult ranges between i -o and
2-0 mm. (including tubes). Average diameter about i-o mm.

Specimens apparently multiplying by fission have been found. In one case this
proceeds by contraction of the specimen, the rims being pinched in towards each other.
In another much larger specimen, a sandy rim has commenced to grow across the
chitinous membrane.

Vanhoejfenella is not a mud-feeder. The protoplasmic body is large and opaque but
not loaded with mud and Diatoms like many of its relatives.

The distribution of the type round South Georgia is confined to shallow water,
100-270 m. There is one remarkable exception which appears to be distinct.

39. Vanhoeffenella oculus, sp.n. (Plate I, fig. 22).

One station : WS 334.

A single specimen found at this station at the abnormal depth of 3705 m. presents
differences which appear to be worthy of specific distinction. In form the test is
an almost perfect oval and the tubular extensions, two in number, are small, purely
chitinous and of the frailest kind. They might easily be entirely overlooked. The
specimen is presumably a dead individual, as there is no trace of protoplasm within the
central diaphragms, and the sandy rim which is constructed of the finest material has
collapsed to form a tube. Similar specimens occur in deep water in the Antarctic area.
Greatest width o-6 mm., least width 0-5 mm., exclusive of the rudimentary tubes.

Genus Pelosina, Brady, 1879

40. Pelosina rotundata, Brady.

Pelosina rotundata, Brady, 1879, etc., RRC, 1879, p. 31, pi. iii, figs. 4. 51 1884, FC, p. 236,

pi. XXV, figs. 18-20.

Pelosina rotundata, Cushman, 1918, etc., FAO, 1918, p. 55, pi. xxi, figs. 4-6.

Ten stations: 27, 30, 140, i43> i44> 148; WS 25, 154; Drygalski Fjord; MS 68.

Frequent to common at nearly all stations and quite typical, the test being constructed

ASTRORHIZIDAE 55

of mud. At MS 68 many of the specimens, though otherwise perfect, were flattened or
collapsed, apparently owing to the thinness of the mud wall.

41. Pelosina fusiformis, sp.n. (Plate I, figs. 10-12).
Seven stations: 45, 126, 144, 148, 149; WS 28, 154.

Test free, fusiform, one end usually more prolonged than the other and provided with
an extended but usually collapsed neck. Wall, firm and thick, and composed of mud with
or without admixture of sand grains, smooth and neatly finished. Internal cavity rather
small, the two walls accounting for more than half the maximum diameter of the test.
The inner wall of the cavity generally exhibits a number of funnel-shaped depressions
through which no doubt the protoplasm extrudes, although they do not extend to the
outer surface of the test. The cavity usually contains a sub-spherical solid mass, which
is the protoplasmic body containing a food mass of Diatoms enclosed in the chitinous
lining of the test. Colour grey.

Length up to 2-0 mm. breadth up to 1-2 mm.

Very common at Sts. 45, 148, frequent elsewhere. There is considerable variation in
the amount of sand incorporated ; at WS 28 a good deal of black sand is used, whereas at
St. 45 none at all. The aperture is seldom visible except in the few instances where the
protruding neck has resisted disintegration.

P. fusiformis does not occur in company with P. rotundata except at Sts. 144, 148 and
WS 154. Fusiform specimens of Pelosina rotundata have been recorded by Egger and
Millett (E. 1893, FG, p. 254, pi. xi, fig. 60; M. 1898, etc., FM, 1899, p. 249, pi. iv,
figs. I a, b), and although their figures are not identical with the South Georgian speci-
mens, it is probable that they represent the same organism, which I regard as speci-
fically separable from P. rotundata.

42. Pelosina variabilis, Brady.

Pelosina variabilis, Brady, 1879, etc., RRC, 1879, p. 30, pi. ill, figs. 1-3; 1884, FC, p. 235,

pi. xxvi, figs. 7-9.

Pelosina variabilis, Flint, 1899, RFA, p. 266, pi. iv, fig. i.

Thirteen stations: 27, 28, 45, 140, 143, 144, 148, 149; WS 27, 47, 154; Drygalski Fjord; MS 68.

Frequent to common at most of the stations where it was recorded, although perfect
specimens are not very frequent. It often attains a large size, notably at Sts. 27, 45 and
148. At several stations the specimens are very narrow and elongate and the walls are
more smoothly finished than usual. Such specimens are evidently closely allied to
others which I am separating under the varietal name constricta. They are found at
Sts. 28, 143, 144 and 149, sometimes in company with the normal type.

43. Pelosina variabilis var. constricta, var.n. (Plate I, figs. 13-15).
Five stations: 45, 126, 140, 144; MS 68.

Test long, narrow, tapering to each extremity, broadest near the oral aperture which is
usually furnished with a short, distinct tube. More or less constricted at irregular
intervals, as many as five constrictions have been observed. There is no evidence that the

56 DISCOVERY REPORTS

constriction indicates septation, but it probably marks a stage of growth. The test is
constructed of very fine mud, neatly finished and devoid of larger incorporated particles.

Length up to 6-o mm. ; breadth up to (about) o-8o mm.

Frequent at MS 68, rare at the other stations, the best specimens at St. 45. In the
Antarctic gatherings it reaches far greater dimensions than in material from South
Georgia. Specimens have been found measuring up to 4 cm. in length.

Genus Storthosphaera, F. E. Schulze, 1875

44. Storthosphaera elongata, Cushman.

Storthosphaera elongata, Cushman, 1918 etc., FAO, 1918, p. 40, pi. xviii, figs. 1,2; pi. xix, fig. i.
Two stations: 45, 144.

At St. 144 a single specimen, agreeing exactly with Cushman's description and figure,
and with British specimens from West Scotland, 205-1600 m., and South-west
Ireland, 523-680 fathoms, where it is not uncommon. Cushman's types were from the
North Atlantic region about 40° 16' N, 67° W.

At St. 45 a form is common, which I think is probably only a local form of S.
elongata. It is smaller, rarely more than half the usual dimensions, and very variable in
shape, ranging between sub-spherical, compressed, and cylindro-elongate, sometimes
four times as long as wide. The character of the test agrees with Cushman's description :
"wall comparatively thin, composed of a felted mass of fine amorphous material and a
large percentage of acerose sponge spicules with little or no cement; aperture not
developed, surface smooth, colour greyish-white".

It would be difficult to separate the sub-spherical and compressed varieties from
Crithionina, but for the thin test and large central cavity, facts to which Cushman has
drawn attention. Nor would it seem possible to assign these varieties to the same species
as the very elongate forms but for the presence of a complete series of connecting links.

45. Storthosphaera elongata var. impudica, var.n. (Plate I, figs. 23, 24).
Two stations: 45, 144.

In general construction the test agrees with the type, the wall being thin and the
cavity large and entire. But at about half its length the test is sharply constricted and
continues growth in cylindrical form of about half the previous width. The entire
organism has a grotesque phallus shape.

Length up to 2-4 mm. Breadth 1-2 mm. at broader, 0-7 mm. at narrower, extremity.

Genus Crithionina, Goes, 1894

46. Crithionina granum. Goes (F 54).
Two stations: 144; WS 33.

Frequent but poorly developed at WS 33, large but rare at St. 144.

ASTRORHIZIDAE 57

47. Crithionina mamilla, Goes (F 55).

Two stations: 144; WS no.

A single specimen, sessile on Hyperammina siibnodosa at St. 144, and another on a
stone at WS no, are the only certain records. A few doubtful specimens were seen
elsewhere, not worth recording.

48. Crithionina pisum, Goes (F 56).
Four stations: 144; WS 27, 33; MS 14.

Singularly rare. The only specimens recorded without hesitation as belonging to this
species were from St. 144, where several good examples were obtained. Spicules are
largely employed, with the usual fine sand, and it is not easy to state when the specimens
should be attributed to the type, and when to var. hispida. Two specimens from this
station which are otherwise typical are transfixed by very long spicules, which no doubt
serve the same purpose as the similar variation in Psammosphaera parva, viz. to support
the organism in the surface film of ooze. At the other stations the specimens are small
and obscure.

49. Crithionina pisum var. hispida, Flint.

Technitclla melo de Folin {non Norman), 1895, SRR, p. 13, pi. O, fig. 3.

Crithionina pisum var. hispida, Flint, 1899, RFA, p. 267, pi. vi, fig. 2.

Crithionina abyssorum, Kiaer, 1899, NNAE, p. 7, pi. i, figs. 1-4.

Crithionina pisum var. hispida, Heron-Allen and Earland, 1909, TNS, p. 410, pi. xxxiv, fig. 7.

Two stations: 144; WS 27.

Good specimens are frequent at WS 27, less typical at St. 144.

Genus Dendronina, Heron-Allen and Earland, 1922

50. Dendronina papillata (Heron-Allen and Earland) (F 59).
One station: 126.

Two specimens in the youngest or "basal pad" stage.

Sub-family PILULININAE

Genus Bathysiphon, M. Sars, 1872

51. Bathysiphon capillare, de Folin (Plate I, fig. 26).

Bathysiphon capillare, de Folin, 1887, B, p. 276, pi. v, fig. 2 a-e; 1887, RR, p. 114, fig. 7 b.
Eight stations: 27, 45, 140, 144, 148; WS 33 ; Drygalski Fjord; MS 14.

Fragments are not uncommon at most of the stations and very common at WS 33, but
there isonlyone approximately perfect specimen, at St. 45. This is regularly tapering and
curved and the apertural end rounded off, with a central opening. Total length 3-4 mm.
Maximum width 0-14 mm. The texture of the tube varies considerably, in some
specimens fine sand grains are liberally used, in others only fine mud. The wall is un-
pohshed, thin and frequently collapsed.

58 DISCOVERY REPORTS

The specimens are attributed with some hesitation to B. capiUare, as they appear to
agree better with de FoHn's figure and description than with any other species. At the
same time there appears to be Httle difference between this species and B. filiformis
except size, and the duller surface of the tube.

52. Bathysiphon rufescens, Cushman (Plate I, fig. 25).

Bathysiphon rufescens, Cushman, 1917, NFP, p. 651 ; 1921, FP, p. 43, pi. ii, fig. 3.
Three stations: 30, 45; MS 14.

Fragments only, the best at St. 45. In spite of their small size, the largest fragment
being only 2-15 mm. in length, they appear to agree generally with Cushman's species,
which attains a length of 15 mm. in deep water off the Philippines.

Cushman describes his species as very slightly curved or tapering, slender, the wall
marked by annular rings, surface rough, very light yellowish or reddish brown, dull,
differing from B. nifum in the rough granular surface and straight test.

The South Georgian specimens have these characters. They are rusty brown in colour.

53. Bathysiphon rufum, de Folin.

Bathysiphon rufum, de Folin, 1887, B, p. 283, pi. vi, figs. 8 a-c.

Bathysiphon rufus, Cushman, 1918, etc., FAO, 1918, p. 29; 1921, FP, p. 42, pi. ii, fig. 2.

Two stations: 53° 00' S, 34° 22' W; WS 336.

A single specimen at each station. That from WS 336 is a small curved individual
regularly tapering to a point and marked by many superficial constrictions. At the other
station only a fragment of a large specimen was found. Both are of a yellowish colour,
in which feature they would agree better with B. flavidwn, de Folin ; but that species is
stated to be sub-cyhndrical and only very slightly tapering. There is, however, con-
siderable range in the colour of B. nifum within our experience, and so I assign the
South Georgian specimens to that species. There appears to be some likelihood that de
Folin's two species are really identical, his figure of B. flavidiim does not suggest a
perfect specimen.

Sub-family SACCAMMININAE

Genus Sorosphaera, Brady, 1879

54. Sorosphaera confusa, Brady.

Sorosphaera confusa, Brady, 1879, etc., RRC, 1879, p. 28, pi. iv, figs. 18, 19; 1884, FC, p. 251,

pi. xviii, figs. 9, ID.

Sorosphaera confusa, Rhumbler, 1903, ZRR, p. 235, fig. 63.

Sorosphaera confusa, Cushman, 1918, etc., FAO, 1918, p. 39, pi. xv, figs. 4, 5.

Three stations: 144; WS 33, 154.

Extremely rare. One excellent free specimen at St. 144 and a good sessile one at
WS 154. A very doubtful specimen at WS 33.

ASTRORHIZIDAE 59

55. Sorosphaera depressa, Heron-Allen and Earland (Plate V, figs. 20, 21).
Sorosphaera depressa, Heron-Allen and Earland, 1929, etc., FSA, 1929, p. 102, pi. i, figs, i, 2.

Eight stations: 17, 27, 136, 144, 148; WS 33, 113, 365.

Test attached, light to dark grey in colour according to the constituent particles,
consisting of a variable number of irregularly shaped chambers spreading without
definite plan over a stone or other surface of attachment, and often following a crack or
superficial depression for protection. The walls are thin, firmly and smoothly constructed
of fine sand grains and cement, but they are dull and unpolished. The outer surface is
rougher than the interior wall of the chambers, which have a chitinous lining.

Each chamber forms a distinct and separate entity enclosed within its own walls and
base. There is no sign of any aperture or means of communication between the chambers.
Communication with the external medium can be only through minute interstitial
openings in the wall of the test. The openings shown in Plate V, fig. 21, are only inci-
dental fractures of the wall.

When a colony is attached to a stone, as in Plate V, fig. 21, the test forms quite a solid
object capable of sustaining considerable stress, but when growing on a flexible base, the
chambers are readily separable without fracture of the walls of the test (see Plate V,
fig. 20).

Monothalamous Arenacea, when sessile, are usually more or less semi-globular in
plan, but in Sorosphaera depressa the chambers are distinctly polygonal. This in some
cases is due to the irregularities of the surface of attachment, but specimens have been
observed (see Plate V, fig. 20) which suggest that the organisms first form a quadrate
colony, and thereafter spread irregularly in the line of greatest protection or least
resistance.

The dimensions of the chambers vary considerably from 0-3 to o-8 mm. in diameter.
A colony may cover a space of 0-5 square centimetre. The thickness of the wall is only
about 0-02 mm.

Always very rare, but some good specimens at St. 27 and WS 113, where it was more
frequent than at any other station. Both sessile and free individuals were found, the
latter always showing evidence in their flat base of having become detached.

Genus Psammosphaera, F. E. Schulze, 1875

56. Psammosphaera fusca, Schulze (F 60).

Thirty-five stations: 17, 30, 42, 123, 126, 140, 144, 148, 151; 53° 00' S, 34° 22' W; WS 27, 28,
31, 32, 33, 40, 42, 51, 61, 63, 154, 334, 336, 348, 349, 351, 353, 365, 373, 429, 521, 522, 523;
DrygalskiFjord;MS68.

Generally distributed and sometimes frequent, notably at WS 32 and 154. Excep-
tionally large single specimens up to 3-0 mm. in diameter were found at St. 151, WS 63
and 521 in company with normal individuals. There is the usual wide range of variation,
but the general average for the area is a rather small and neatly constructed form. At
those stations where the bottom deposit contains black sand, e.g. St. 148, WS 28 and

6o DISCOVERY REPORTS

154, it has a rather striking appearance owing to the contrast of the black sand with the
copious grey cement used for building purposes. At St. 140 and WS 33 spicules were
utilized as well as sand grains, a rather unusual occurrence, those at the latter station
being left projecting, perhaps to act as supports, though their irregular disposition does
not show much evidence of selection.

57. Psammosphaera parva (Flint).

Psammosphaera fiisca, de Folin, 1895, SRR, p. 16, pi. O, figs. 4, 5.

Psammosphaera parva, Flint, 1899, RFA, p. 268, pi. ix, fig. i.

Psammosphaera parva, Heron-Allen and Earland, 1912, etc., NSG, 1913, p. 17, pi. ii, figs. 7, 8.

Psammosphaera parva, Cushman, 1919, RFNZ, p. 594, pi. ixxv, fig. 3.

Eight stations: 20, 28, 29, 42, 144, 149; WS 33, 46.

Rare or very rare at all the stations and seldom as neatly constructed as usual. Only
a single specimen of the selective form, which is transfixed by a sponge spicule, was
found at St. 144.

58. Psammosphaera rustica, Heron-Allen and Earland (Plate I, fig. 27).

Psammosphaera rustica, Heron-Allen and Earland, 1912, etc., NSG, 1912, p. 383, pi. v, figs. 3,

4; pi. vi, figs. 2-4.

Psammosphaera rustica, Cushman, 1918, etc., FAO, p. 37, pi. i.x, figs. 3, 4; pi. x, figs. 2-4.

One station: 144.

A single large specimen. It is not so highly selective as the types from the North
Sea, the body of the test being constructed of fine sand and cement, but it possesses the
characteristic projecting spicules.

Genus Pelosphaera, Heron-Allen and Earland, 1932

Test large, free, roughly spherical, constructed of large and small irregularly shaped
mineral grains joined firmly together with copious cement, which externally is soft and
friable, but internally firm and smooth. Furnished externally with two or more pro-
jecting processes, conical in shape, hollow, formed of fine sand grains and loosely
aggregated mud and cement, similar in appearance to the external cement between the
sand grains of the test. There is no visible external aperture to either the test or the
processes, but the processes extend from, and conceal, large apertures in the test, which
are clearly seen from the inside, when the sphere is laid open.

This is a very distinctive form in the perfect condition, but the conical processes are
so friable that few specimens retain them throughout the cleaning process. Devoid of
processes, the specimens, except for their abnormal size, would pass for Psammosphaera
fiisca, the apertures being usually concealed by mud. Only a few young individuals were
found. These bear two processes only, sometimes almost equalling in length the
diameter of the sphere, and usually, but not always, built of smaller sand grains than
those employed by the adult organism.

ASTRORHIZIDAE 6i

Pelosphaera is no doubt closely allied to Psammospliaera, but even without its t}-pical
processes would be distinguishable by its relatively enormous size and the friable nature
of the external cement.

59. Pelosphaera cornuta, Heron-Allen and Earland (Plate VII, figs. 24-7).

Pelosphaera cornuta, Heron-Allen and Earland, 1929, etc., FSA, 1932, p. 255, pi. ii, figs. 12-15.
Five stations: 17, 27, 126, 144, 148.

The characteristic features of the genotype have been given under the description of
the genus. The best specimens, both large and small, were obtained at Sts. 126, 144 and
148, where the material was obtained from nets attached to the trawl, and the specimens
had been subjected to less friction than at St. 27. At this station, dredged material had
been passed through sieves with the result that few specimens retained the character-
istic processes. It is of frequent occurrence at Sts. 27, 126 and 148 — rare elsewhere.
At St. 17 the only specimen found is abnormal, both in size and shape. It is roughly
triangular in outline, about 6-o mm. in greatest diameter, and constructed of relatively
enormous sand grains. A specimen from St. 148, which was laid open, contained a large,
orange-coloured sphere almost filling the central cavity, which is probably the proto-
plasmic body in a chitinous envelope. Similar enclosures have been found in Pelosina.

Dimensions range between 3-0 and 5-0 mm. in diameter.

Genus Saccammina, M. Sars, 1868

60. Saccammina sphaerica, M. Sars.

Saccammina sphaerica, M. Sars, 1868, LUHD, p. 248.

Saccammina sphaerica, Brady, 1884, FC, p. 253, pi. xviii, figs. 11-17.

Saccammina sphaerica, Heron-Allen and Earland, 1912, etc., NSG, 1913, p. 1, pi. i and pi. ii,

figs. I, 2.

One station: 151.

A single rather small specimen.

61. Saccammina minuta, Rhumbler.

Saccammina minuta, Rhumbler, 1909, etc., FPE, 1913, p. 375, pi. i, figs. 8, 9.
Two stations: WS 343, 349.
Very rare, never more than one or two specimens at a station.

Genus Proteonina, Williamson, 1858

62. Proteonina difflugiformis (Brady) (F 61).

Twenty stations: 13, 30, 42, 131, 136, 144, 151; 53° 00' S, 34° 22' W; WS 33, 42, 48, 334, 336,
343> 349. 353. 428, 429. 522, 523-

Frequent at WS 429, but rare or very rare elsewhere. Considerable variation in
construction is exhibited by specimens from diflFerent stations, but the most common
form is a long oval flask of quartz grains showing no attempt at neatness of construction,

5-2

62 DISCOVERY REPORTS

the exterior being rough and irregular. Such specimens are usually small, but at St. 42,
WS 353, 428 and 523 they are very much larger than average size. The last three
stations are in deep water between 1697 and 4041 m. A neatly constructed variety, in
which the sand grains are laid so as to form a smooth exterior, occurs in company with
the rough form at Sts. 144 and 151, and by itself at WS 33 and 349. At the former
station, the sand grains in two of the three specimens are too small to be distinguishable.
A very abnormal specimen was noted at WS 522, nearly globular and ferruginous,
with a large slightly projecting oral extremity. It may possibly be a proloculum of
Reophax 7wdulosiis, but this species was not recorded at the station.

63. Proteonina decorata, sp.n. (Plate I, figs. 28, 29).
Three stations: WS 28, 32, 41.

Test free, monothalamous, pyriform or fusiform, devoid of a produced neck; the
aperture, which is small, being situated at the narrower extremity of the test. Wall
rather thick, rough externally and internally, constructed of large irregular sand grains,
embedded in very fine sand and cement, the cement predominating, so that the sand
grains appear to be scattered over the surface like plums in a cake. Length o-6 mm.,
breadth 0-30 mm.

This is rather a striking form in appearance, owing to the marked contrast between
the dark sand grains and the white cement which forms the bulk of the test.

It is rare at all the stations, the best specimens at WS 28.

64. Proteonina tubulata (Rhumbler) (Plate I, figs. 30, 31).

Saccammina tubulata, Rhumbler, n.sp., Wiesner, 193 1, FDSE, p. 82, pi. xxiii. Stereo-fig. a.
Six stations: 45, 136; WS 47, 351, 353, 429.

Test more or less globular and roughly constructed of comparatively large mineral
grains embedded in cement. Exterior rough owing to the projecting edges of the sand
grains. Aperture furnished with a projecting neck built of very minute sand grains
neatly cemented together. Owing to its fragility the neck is seldom, if ever, perfect, but
at WS 351 its length is about three-quarters that of the body of the test.

Length of test without neck o-3-o-4 mm., breadth about the same; length of neck up
to 0-2 mm., width 0-02 mm.

This is a very distinctive form owing to the contrast between the roughly finished
shell and the delicately constructed neck. Though very rare, seldom more than a single
specimen at a station, its distribution is sufficiently wide to justify specific distinction.
In depth it ranges between 160 and 2549 m. It has recently been described and
figured by Wiesner {nt supra) from two stations of the German South Polar Expedition,
depths 385 m. and 3410 m. under what is apparently a MS. name of Rhumbler 's. I
cannot agree with Rhumbler 's attribution of the species to Saccammina. The long neck,
in itself, seems to forbid such an association.

ASTRORHIZIDAE 63

Genus Webbinella, Rhumbler, 1903

65. Webbinella depressa, Heron-Allen and Earland (F 64).
Two stations: 27, 145.

A single specimen only at St. 27 and two excellent specimens at St. 145.

66. Webbinella limosa, sp.n. (Plate II, figs, i, 2).
Two stations: 27, 126.

Test sessile, monothalamous, irregular in form but usually roughly circular, more or
less highly convex. Wall rather thick and without visible aperture, composed of mud
and very fine sand grains. Friable but sufficiently firm to allow the convex test to be
detached as a whole from its surface of attachment. The test is then seen to consist of a
single irregularly shaped cavity with thin basal floor. The cavity is nearly always filled
with a mass of Diatoms and mud ingested as food. Colour yellowish grey.

Size from i-o to 2-0 mm. in diameter.

Not uncommon at the two stations where it occurs sessile on small pebbles. Most of
the specimens have been more or less damaged in the cleaning process, the convex shape
and friable wall causing it to be easily worn away.

W. limosa is no doubt closely allied to W. depressa, Heron- Allen and Earland, but is
readily distinguishable owing to its greater convexity, thicker but more friable wall and
yellowish colour.

Genus Tholosina, Rhumbler, 1895

67. Tholosina bulla (Brady) (F 65).

Seventeen stations: 27, 45, 123, 140, 144, 145, 148, 149; WS 27, 33, 40, 42, 154, 334, 31^3;
MS 14, 68.

Less abundant than T. vesiciilaris but common at Sts. 27, 144 and 149 and very
common at WS 33. Mostly rare elsewhere. There is considerable range of variation in
the convexity of the test: very highly convex specimens at St. 149 and MS 68. At Sts.
45 and 123 small specimens attached to sponge spicules and zoophyte stems are almost
globular in shape. They appear to be very similar to Wiesner's figures of Tholosina
laevis, Rhumbler (W. 1931, FDSE, p. 86, pi. vii, figs. 80-2). If they are identical there
seems little excuse for the formation of a new species.

The material employed varies at different stations, at some only fine sand is used, at
others coarse sand, the latter being distinguishable from T. vesiciilaris only by the
absence of the apertural tubes.

68. Tholosina protea, Heron-Allen and Earland (F 66).

Six stations: 27, 144, 145; WS 25, 27, 154.

Detached specimens only were observed, except at St. 145. They are frequent, both
large and small, at the three WS stations, rare at the others. Nearly all the various
eccentric shapes observed in the Falkland material occur, but many of the specimens

64 DISCOVERY REPORTS

appear to have been attached to sponge spicules, judging by the shape of the scar of
attachment.

69. Tholosina vesicularis (Brady) (F 67).

Eighteen stations: 17, 20, 27, 31, 126, 136, 144, 148, 149; 53' 00' S, 34'' 22' W; WS 33, 40, 42, 51,
110, 154, 363, 418.

Very common indeed at St. 144 and common at Sts. 27, 126, 148, WS 33 and no,
frequent to rare at the remaining stations. The specimens, with few exceptions, are quite
typical. At St. 144, in addition to the type, specimens were seen sessile on sponge
spicules, the two apertural tubes being extended along the spicule in each direction. At
St. 27 a single specimen had long collapsed chitinous tubes extending freely from the
ends of the usual attached tubes.

70. Tholosina vesicularis var. erecta, Heron-Allen and Earland (F 68).

Recorded with hesitation, no perfect specimen having been seen. Fragmentary tubes,
which may belong to this variety, were noticed at Sts. 30, 129, WS 25 and 33.

Genus Armorella,^ Heron-Allen and Earland, 1932

Test free, monothalamous, approximately spherical, furnished with a variable number
of extended tubes of different length, with an aperture at the end of each tube. Wall
firm, but very thin, constructed of fine sand. Diatoms and sponge spicules incorporated
with much cement, occasional larger sand grains and spicules projecting from the other-
wise smooth and rather shining surface. Interior surface similarly smooth. Colour
light grey.

This is a very distinctive form, closely allied to Thnrommiiio and Tholosina, its
affinities probably lying with the latter genus. Small specimens furnished with short
tubes, or remains of broken tubes, are very like TJinramnii?ia papillata in their spherical
form, but a series of specimens links them up with the large and multitubular individuals,
which have no resemblance to that species. Moreover the broken tube ends are very
unlike the aperture of Thiirarnmina.

Small sponge spicules are often employed to a considerable extent as building material,
being smoothly incorporated in the wall. In specimens from St. 144, they play a larger
part than usual in construction, the sphere in some cases being built round a bundle of
spicules, the ends of which may project to an extent equal to the diameter of the test.
This spicular construction to some extent also modifies the shape of the test, which
tends to become polyhedral instead of spherical. Such tests are probably not evidence of
selective powers, or only in a limited degree comparable with the use of spicules in
Psammosphaera rustica. But these projecting spicules would undoubtedly serve a useful
purpose in supporting the organism in the surface layer of mud, and this would be of
value to the animal, which is not one of the mud-eaters. The protoplasmic body is large,
but not loaded with mud and Diatoms as in many Arenacea.

1 In memory of Armorel Daphne Heron-Allen, who died July 3, 1930, aged 22.

ASTRORHIZIDAE 65

On the other hand, a specimen found at St. 45 exhibits a definite instance of selection,
similar to Psammosplwera parva, the spherical test, which is very neatly built of fine
sand only, being transfixed by a very long spicule (Fig. 21).

These projecting spicules are sometimes used as supports for the tubes which are
attached to them. But there is no general practice, and frequently a tube is seen growing
out quite close to a projecting spicule, but unattached.

Armorella has probably a wide distribution in deep or cold waters. A similar organism,
though specifically distinct, has been found in several dredgings round the British Isles,
but always of rare occurrence. Several of the figures attributed to Thiirammina in
Haeusler's papers would appear to be referable to our genus, in which case its record
extends back to Jurassic times.

71. Armorella sphaerica, Heron-Allen and Earland (Plate VII, figs. 16-23).

Armorella sphaerica, Heron-Allen and Earland, 1929, etc., FSA, 1932, p. 257, pi. ii, figs. 4-1 1.
Twelve stations: 27, 31, 45, 123, 140, 144, 148, 149; WS 33, 154, 334; MS 68.

The description of the genus is sufficient for the species, which is very common at
St. 144, common at Sts. 148, 149, frequent at Sts. 45 and 140, very rare at the remaining
stations. The range of depth lies between 1 10 and 270 m., except for a single specimen
at WS 334 in 3705 m. There is a considerable range of size, the specimens reaching
1-2 mm. in diameter without tubes. An average size is about i-omm. in diameter.
Tubes average up to 0-3 mm. in length. There can be no doubt that the small individuals,
which represent the species at those stations where it is rare, are merely young or
pauperate individuals.

The external texture of the test varies to a lesser extent. In general the sphere is
smooth externally, owing to its homogeneous construction, but occasionally the animal
incorporates sand grains which, being larger than the thickness of the wall, project and
give an unfinished appearance to the test.

The tubes vary enormously both in size and number. It is difficult to give a maximum,
as a broken tube may leave little trace. Specimens with four tubes are common. The
length of the tube has no relation to the size of the sphere : many large specimens have
very short tubes and vice versa.

Genus Thurammina, Brady, 1879

In 1917 we published a paper "On Thurammina papillata, Brady; a Study in Varia-
tion" (H.-A. and E. 1912, etc., NSG, No. 5, 1917) in which we expressed the opinion
that "all hitherto recorded species of the genus Thurammina, including Thuramminopsis
canaliailata, Haeusler, are referable to a single specific type, Thurammina papillata,
Brady", and that while "for taxonomical reasons numerous varietal names must be
employed. . .they have no biological significance".

The experience of the intervening years has not caused me to vary from the opinion
then expressed. On the contrary, I believe that a similar intensive study of other

66 DISCOVERY REPORTS

genera will in time result in an enormous reduction of so-called species. Nevertheless
in a report of this character it seems more convenient to minimize the number of
varietal names employed, and I am therefore reverting to the use of specific names
instead of the varietal names used in our paper.

72. Thurammina papillata, Brady.

Tluirammina papillata, Brady, 1879, ^t^., RRC, 1879, p. 45, pi. v, figs. 4-8; 1884, FC, p. 321,
pi. xxxvi, figs. 7-8.

Thurammina popillata, Heron-Allen and Earland, 1912, etc., NSG, 1917, p. 543, pi. xxvi,
figs. 1-13; pi. xxvii, figs. 9-13.

Two stations: 123; WS 33.

The typical sandy sphere is almost absent from our material. Apart from a small
individual at St. 123, the only specimen which can be referred to the type is a chitinous
individual characterized by an abnormal number of papillae, which are scattered as
closely as possible all over the partially collapsed sphere. A sponge spicule transfixes
one side of the sphere.

73. Thurammina haeusleri (Heron- Allen and Earland).

T/iiirafiwiina papillata, Haeusler, 1883, JVT, pp. 262-6, pi. viii, figs. 5-8, 11, 13-24; 1890,
FST, pp. 46 et seq., pi. vi, figs. 14, 18.

Thurammina papillata, Brady, 1884, FC, p. 321, pi. xxxvi, figs. 13, 14 (only).
Thurammina papillata var. haeusleri, Heron-Allen and Earland, 1912, etc., NSG, 1917, p. 547,
pi. xxviii, figs. 1-12; pi. xxix, fig. 16; pi. xxx, fig. 8.
Three stations: 27, 123, 144.

Rare everywhere, most numerous at St. 123 where the specimens were small. A
single large individual at Sts. 27 and 144.

74. Thurammina parallela (Heron-Allen and Earland).

Thurammina papillata var. parallela, Heron-Allen and Earland, 1912, etc., NSG, 1917, p. 546,
pi. xxvii, figs. 14-17.

One station: 144.

Only two specimens were observed, one of which was very small, the other large and
typical. Occasionally T. protea takes on this habit, but never attains such length as to
justify separation from its type, the length of the specimens seldom exceeding twice the
breadth.

75. Thurammina albicans, Brady.

Thurammina albicans, Brady, 1879, etc., RRC, 1879, p. 46; 1884, FC, p. 323, pi. xxxvii, figs.

2-7.

Thurammina papillata var. albicans, Heron-Allen and Earland, 1912, etc., NSG, 1917, p. 550,

pi. xxix, figs. 12-15.

One station: WS 365.

A single specimen from 3219 m. at this station.

ASTRORHIZIDAE 67

76. Thurammina protea, sp.n. (Plate II, figs. 3-10).
Nine stations: 20, 27, 123, 140, 144, 148; WS 154, 348; MS 14.

Test free or sessile, typically monothalamous but sometimes forming an aggregation
of two or more individuals, without direct communication with each other beyond the
apertures in the wall, which is thin but firm, constructed of fine sand and ferruginous
cement on a chitinous membrane. Apertures in the form of small nipple-like protu-
berances, very variable in number and prominence, either scattered over the wall of the
test or more frequently confined to salient edges. Shape protean, cushion-like or
hemispherical when sessile, but when free irregular, angular and polyhedral. Colour
rusty brown. Size very variable, ranging up to i-6 mm. in diameter. Average i-o mm.

Although not generally distributed T. protea is one of the most striking and character-
istic of the South Georgian Foraminifera and by far the most abundant representative
of its genus, all the other species being comparatively rare. It is common at St. 27 and
WS 154, frequent to rare at the other stations.

It is more frequently found in the sessile condition than any other form of Thiiram-
niino, and in this condition appears to be closely related to T. hemisphaerica, Haeusler
(H. 1883, ALB, p. 60, pi. iv, figs. 14, 14 (?; and H. 1890, FST, p. 47, pi. vii, figs. 10, 11).
Young specimens in the sessile condition are circular and cushion-shaped, almost
devoid of nipples, or having them only round the margin. Later the specimens become
irregular in outline and form, and develop nipples over the surface.

There is no doubt that the extraordinary shapes assumed by the organism are largely
due to the conditions under which it lives, either attached to other bodies, or in crevices.
In many cases it takes up its abode inside the empty tubes oi Hyperammina stibnodosa, and
probably inside other tube-forming organisms. This leads to the formation of a cylin-
drical test, examples of which were found at St. 144 {in situ in H. stibnodosa) and WS
154. The specimen from the latter station is an aggregate of three individuals in line.

The shape of the species is particularly protean at WS 154, the irregularity being
due to the formation of the test in the crevices of other organisms, or between pebbles.
The specimens have taken the exact impress of the cavity, and the apertures are confined
to the exposed marginal edge, the remainder of the test being quite smooth.

At St. 27 a very curious specimen was found which had grown around two large
tetractinellid sponge spicules, possibly while the spicules were still projecting from the
sponge. Another instance of incorporated spicules was observed at WS 348, but such
inclusions, whether of spicules or large sand grains, are very rare.

T. protea is evidently closely related to T. haeusler i, but appears to be quite a distinctive
local form.

77. Thurammina tuberosa, Haeusler.

Thurainmina tuberosa, Haeusler, 1890, FST, p. 49, pi. vi, fig. 24; pi. vii, figs. 6-9.
Thurammina popillata var. tuberosa, Heron- Allen and Earland, 1912, etc., NSG, 1917, p. 548,
pi. xxviii, figs. 13-16.

Three stations: 123, 144, 148.

68 DISCOVERY REPORTS

Frequent at St. 144, where the best specimens were found. Very rare at the remaining
stations. The specimens are not highly complex aggregations, such as were figured by
Haeusler, but simpler aggregations such as we figured, especially in pi. xxviii, fig. 13
{ut supra).

Sub-family RHABDAMMININAE

Genus Jaculella, Brady, 1879

78. Jaculella obtusa, Brady (F 70) (Plate II, fig. 11).

Four stations: 16; WS 40, 353, 429.

Very rare, only single specimens at each station and those very small. The specimen
from WS 40 is rather interesting, being built almost entirely of broken sponge spicules,
selected of approximately proportionate length according to their position in the tube.
They are cemented together at right angles to the axis of the specimen and so give a
quadrate shape to the oral end. Length i-6 mm.

Genus Hippocrepina, Parker, 1870

79. Hippocrepina indivisa, Parker (Plate IV, figs. 31-4).

Hippocrepina indivisa, Parker, 1870, GSTL, p. 176, fig. 2.

Hippocrepina indivisa, Brady, 1881, HNPE, p. 100, pi. ii, figs. 3,4; 1884, FC, p. 325, pi. xxvi,

figs. 10-14.

Five stations: 144; WS 33, 40, 42, 154.

Single specimens at St. 144, WS 40, 154, five at WS 33 and eight at WS 42, which is
the deepest station, in 198 m. All the specimens are fine and typical except one at WS 42,
in which the aperture is large, occupying nearly the whole of the oral extremity, and
surrounded by an irregular reverted collar. It is probable that such specimens represent
an individual in the act of enlarging its test, the apertural end being absorbed prior to the
addition of fresh material lengthening the test. It may be remarked that all Brady's
figures are drawn from specimens in this condition.

80. Hippocrepina oviformis, Heron-Allen and Earland.

Hippocrepina oviformis, Heron-Allen and Earland, 1914, etc., FKA, 1915, p. 617, pi. xlvi,
figs. 23, 24.

One station : MS 68.

A single specimen, which agrees with the original figure of the type from the Kerimba
Archipelago, East Africa. The type specimen is unfortunately not available for com-
parison, damp having caused deterioration.

81. Hippocrepina flexibilis (Wiesner) (Plate II, figs. 12-15).
Technitella flexihilis, Wiesner, 193 1, FDSE, p. 85, pi. vii, fig. 75.

Four stations: 144, 149; WS 28, 32.

ASTRORHIZIDAE 69

Test monothalamous, pear-shaped with a well-defined aperture at the narrow end,
aboral end rounded. Wall thin, very smooth but unpohshed, constructed of very minute
particles without visible cement. Colour light grey. Size about 0-5 mm. long, 0-25 mm.
broad.

Frequent at WS 28 and 32, ver}^ rare at the other stations. The shape of this little
organism agrees with that of Proteo?ima diffli/gifor?nis, but the character of the test is
quite distinctive and indicative of relationship to Hippocrepina oviformis, Heron-Allen
and Earland. In life, the test is probably flexible, several of our specimens being
collapsed without fracture. There can be little doubt that the South Georgia specimens
are identical with Technitella flexibilis, Wiesner. He describes his species as snow-white,
oval, very flexible when wet but collapsing when dried. The dry shell is said to be rigid
and comparatively strong, built of the finest possible mineral fragments. The aperture is
small and circular on a produced neck. Wiesner's figures agree with the South Georgian
specimens, but I do not agree with his attribution of the species to Technitella, the shell
structure being quite different and showing no selective tendency.

Genus Hippocrepinella, Heron-Allen and Earland, 1932

Test free, monothalamous, irregularly cylindrical and sometimes curved, furnished
with two terminal apertures. Wall thin compared with the size of the large central
cavity, constructed of extremely fine sand and mud with little cement, and generally
without inclusion of larger particles ; smoothly and neatly finished, but often exhibiting
numerous fine transverse wrinkles. It is probably flexible during life, but dry specimens
are rigid and fragile. Colour varying from white to very dark grey.

Although widely distributed round the coast-line of South Georgia, Hippocrepinella
is mainly characteristic of the Cumberland Bay area, the majority of the records being
from stations in or near that bay. It favours the tenacious mud found in that area,
although a few specimens have been recorded elsewhere on sandy bottoms.

Hippocrepinella appears to be closely related to Hippocrepina. Indeed, but for the
existence of the secondary aperture, we should have had no hesitation in referring the
specimens to that genus, as the wall of the test is very similar in character though more
delicate. Also, owing to the finer materials employed in its construction, the surface of
the test is smoother and more polished.

Of the two apertures, one, which may be regarded as the principal oral opening, is
always well defined and sometimes quite large, while the secondary or basal opening is
usually inconspicuous, and sometimes only to be detected with difficulty.

There is little doubt that the test is flexible and extensible in life. The apertural ends
probably expand for the absorption of food and contract for digestion, opening again
for the rejection of the empty Diatom shells, which form the food of the animal. Diatom
valves have been observed inside the cavity, of dimensions larger than the aperture. The
flexibility of the living test would also account for the curvature of some specimens and
the transverse wrinkles observed in others.

6-z

70 DISCOVERY REPORTS

82. Hippocrepinella hirudinea, Heron-Allen and Earland (Plate VII, figs. 1-9).

Hippocrepitiella hirudinea, Heron-Allen and Earland, 1929, etc., FSA, 1932, p. 258, pi. i, figs.
7-15-
Thirteen stations: 27, 28, 45, 123, 126, 140, 143, 144, 148, 149; WS 28, 42; MS 68.

Test free, monothalamous, irregularly cylindrical, occasionally curved, rounded at
the extremities which are sometimes slightly clavate, sometimes tapered off". Wall thin,
smooth and neatly finished, shining or "matt", often covered with fine transverse
wrinkles. Apertures, central and terminal, usually varying in size, one being more
pronounced than the other. Colour varying from light to dark grey. Size up to 2-0 mm.
in length, 0-5 mm. in width.

This species, which is the genotype, is very variable in size and general appearance,
while very constant in its specific features. Although many specimens are to be found in
perfect condition, the majority exhibit compression, distortion or shrinkage in varying
degrees. The explanation is to be found in the condition of the interior of the test. We,
ourselves, have laid open many tests, and Mr J. T. Holder, F.R.M.S., has been so good
as to cut serial longitudinal sections of others. The cavity is found to be more or less
compactly filled with an ingested mass of food-stuft's, principally Diatoms (Figs. 2-3),
and it depends upon the compactness of this mass, whether or not the test preserves its
outline after death. Mr Holder's sections have also been useful, in demonstrating the
fineness of the material used in construction, and the almost total absence of larger
particles of sand.

Occasional specimens noticed at several stations, notably Sts. 140 and 144, exhibit a
number of irregularly placed pustular openings in the walls of the test, the origin of
which is obscure. From the nature of the openings, they are clearly not due to external
agencies, but originate inside the test. They may be subsidiary openings for the emission
of young individuals, but it seems more probable that they are made by minute or-
ganisms, perhaps Nemertine worms, which have been ingested with the mud-mass as
food, and have successfully eaten their way through the wall of their captor.

Hippocrepinella hirudinea is very common at St. 45, common at Sts. 144 and 148, all
three of which are in or ofl^ Cumberland Bay. At the other stations it is rare or very rare.
In depth the range extends between 100 and 346 m.

An abnormal specimen found at St. 45 is bifurcate at one extremity, each of the arms
bearing the usual aperture (Plate VII, fig. i).

83. Hippocrepinella hirudinea var. crassa, Heron-Allen and Earland (Plate VII, figs.

13-15)-

Hippocrepinella hirudinea var. crassa, Heron-Allen and Earland, 1929, etc., FSA, 1932, p. 259,
pi. ii, figs. 1-3.

Two stations: 660; WS 32.

General characteristics as in the species, but the test is much broader in proportion
to its length, of an elongate oval or fusiform shape, round in section or compressed.

ASTRORHIZIDAE 71

Walls thicker and composed of coarser material, rough in texture, apertures incon-
spicuous. Length 1-2 mm., breadth 0-5 mm.

The genotype Hippocrepinella hirudmea was not recorded at either St. 660 or WS 32.
Its place appears to be taken by a forin which we prefer to regard as a variety, rather
than as a separate species, although its appearance is very distinctive, especially in the
case of the specimens from St. 660, which is in Cumberland Bay. The specimens from
WS 32 are less rough. The organism is rare at both stations.

84. Hippocrepinella alba, Heron-Allen and Earland (Plate VII, figs. 10-12).

Hippocrepinella alba, Heron-Allen and Earland, 1929, etc., FSA, 1932, p. 259, pi. i, figs. 16-18.
Seven stations: 27, 45, 126, 144; WS 33, 154; MS 68.

Test monothalamous, cylindrical or fusiform, furnished with a large principal aperture
on a produced neck, with or without a collar ; a secondary basal aperture may be present ;
wall very smooth and of paper-hke thinness, constructed of very minute particles
without visible cement. Inner cavity enormous compared with the thickness of the wall.
Colour uniformly dead white.

Size very variable, the largest specimen being 0-30 mm. broad and 2-80 mm. long,
and the smallest 0-52 mm. long and 0-09 mm. broad.

The above account is an attempt to describe an organism which, owing to its rarity
and fragility, is represented by very few entire specimens, hardly any of which agree in
all details, though all conform in the nature of the test.

The wall of the test is extremely thin in comparison with the size of the organism, and,
owing to the absence of cement and the uniformly minute size of the particles employed
in its construction (apparently fragmentary Diatoms), is, when dry, fragile to the last
degree. In life it is almost certainly flexible and distensible, but nearly all our specimens
are more or less collapsed and broken.

The great variation of specimens in size probably represents stages of growth only,
although there is an equally remarkable range of form between broadly fusiform and
elongate cylindrical.

The most striking point of difference in the specimens lies in the form of the aboral
extremity. The principal aperture is always large and conspicuous on its more or less
produced neck and is sometimes furnished with a thickened collar. The secondary or
basal aperture hardly exists, as such, at all. In many specimens the basal end is pro-
duced into a pronounced nipple, which may or may not be pierced ; in other specimens,
it presents an unbroken rounded extremity.

These points of difi^erence, especially the last mentioned, raise the questions whether
(i) the specimens represent more than one species, and (2) whether they are proper to
Hippocrepinella. We think the second point must be left for final decision when more
material is available, but having regard to the identical nature of the wall in all the
specimens and its probable plastic nature when living, we attach little importance to the
variations in size and shape, or even to the apparent suppression of the basal aperture.

72 DISCOVERY REPORTS

Hippocrepinella alba is very rare everywhere, but a good many specimens, more or less
fragmentary, have been obtained, the best at Sts. 45, 144, WS 154 and MS 68. Its
food consists of ingested Diatom-mud, as in the case of Hippocrepinella hiriidinea.

Its range in deptli extends between 100 and 270 m.

Genus Hyperammina, Brady, 1878

85. Hyperammina elongata, Brady (F 72).

Four stations: 151 ; WS 353, 373, 523.

Very pauperate specimens at WS 523, recognizable fragments elsewhere. Always
very rare.

86. Hyperammina laevigata, J. Wright (F 73).
One station: WS 33.

Some recognizable fragments only.

87. Hyperammina novae-zealandiae, Heron-Allen and Earland (F 75).

Two stations: 27, 144.

A very large specimen and a small one at St. 27, also a good specimen at St. 144: all
microspheric and typical.

88. Hyperammina subnodosa, Brady.

Hyperammina subnodosa, Brady, 1884, FC, p. 259, pi. xxiii, figs. 11-14.
Hyperammina subnodosa, Cushman, 1920, CAE, p. 5, pi. i, figs, i, 2.

Thirteen stations: 20, 27, 45, 123, 126, 131, 140, 144, 148; WS 33, 42, 154, 348.

Frequent at many of the stations, where its large size makes it a very conspicuous
object. The finest series was obtained at Sts. 20, 126 and 144, where specimens up to an
inch in length occur. These largest specimens are nearly always much curved, and
exhibit a tendency to become narrow at the oral extremity, which is always narrower
than the aboral end next to the proloculum. The proloculum varies in diff"erent speci-
mens, being sometimes quite inconspicuous but usually a pronounced knob. These
differences may represent the microspheric and megalospheric forms. The characteristic
constrictions of the tube are extremely marked at St. 144, where some of the specimens
might almost be described as jointed. The construction is the same everywhere, fine
sand grains and sponge spicules are used in about equal proportions, the spicules pre-
dominating towards the apertural end. The tubes are often covered with other sessile
Foraminifera, belonging to many genera and even the interior of dead specimens forms
a retreat for sessile forms.

ASTRORHIZIDAE 73

Genus Saccorhiza, Eimer and Fickert, 1899

89. Saccorhiza ramosa (Brady) (F 56A).

Saccorhiza ramosa, Cushman, 1910, etc., FNP, 1910, p. 65, fig. 81.
Hyperammina ramosa, Heron-Allen and Earland, 1922, TN, p. 86, pi. i, fig. 13.

Seven stations: WS 63, 334, 336, 353, 429, 522, 523.

Fragments are not uncommon at these stations all of which are in deep water, the
depths ranging between 1697 and 4041 m.

At WS 522 a large and very fine specimen occurs, ramifying over the surface of a stone
to which the branching tubes are very lightly attached.

By an oversight this species was placed out of order (No. 56 a) in the Falkland Report.
It should have been No. 76 a.

Genus Psammatodendron, Norman, 1881

90. Psammatodendron indivisum, Heron-Allen and Earland (F 77).
Two stations: 123 ; WS 33.

A few detached tubes which appear to belong to this species, the best at WS 33.

Genus Marsipella, Norman, 1878

91. Marsipella elongata, Norman.

Marsipella elongata, Norman, 1878, GH, p. 281, pi. xvi, fig. 7 (3 on plate).

Marsipella elongata, Brady, 1884, FC, p. 264, pi. xxiv, figs. 10-19.

Marsipella elongata, Heron-Allen and Eariand, 1922, TN, p. 90, pi. iii, figs. 10-12.

One station: WS 351.

A single specimen only.

92. Marsipella cylindrica, Brady (F 78).
Four stations: WS 33, 334, 353, 429.

Fragments of varying sizes are not uncommon, especially at WS 33, which was the
best station. Here all the specimens used spicules, principally or entirely, for building,
and one specimen bears the terminal crown, which we figured in 1912. At WS 429 some
specimens used spicules and others mineral grains only, as also did the specimens at
WS 334 and 353.

Genus Rhabdammina, M. Sars, 1869

93. Rhabdammina discreta, Brady (F 80).
One station : WS 365.

A few fragments from this station, depth 3219 m.

74 DISCOVERY REPORTS

Genus Rhizammina, Brady, 1879

94. Rhizammina algaeformis, Brady.

Rhizanmiina algaeforwis, Brady, 1879, etc., RRC, 1879, p. 39, pi. iv, figs. 16, 17; 1884, FC,

p. 274, pi. xxviii, figs. i-ii.

Rhizammina algaeformis, Flint, 1899, RFA, p. 272, pi. xv, fig. i.

Seven stations: 20, 45 ; WS 50, 154, 336, 349; MS 68.

Fragments doubtfully referable to RMzommhia occur at these stations, but only at
MS 68 was a branching fragment seen which could with any certainty be referred to
R. algaeformis. The others are flat, unbranching, ribbon-like organisms of variable but
regular width throughout, very similar to the figures of "chitinous Rhizopod-tubes^
probably related to Rhizammina'", which Brady figures (B. 1884, FC, pi. xxix, figs. 1-3).
They are more or less coated with mud. Diatoms, etc., and their size, though very
variable, is nearly always much less than that of R. iudivisa, Brady. They possibly repre-
sent one or more species still to be described. At WS 154 a large worrn tube was found
festooned with such organisms.

Family LITUOLIDAE

Sub-family LITUOLINAE
Genus Reophax, Montfort, 1808

95. Reophax scorpiurus, Montfort (F 82).

Seventeen stations : 16,20, 131, 136, 140, 144, 151, 157; 53° 00' S, 34° 22' W; WS 28, 66, 334, 351,
353> 373. 429. 522.

Except at WS 334, where it is frequent, this species is uniformly rare, seldom more
than one or two specimens at a station. Very good specimens, however, were found at
WS 66 and 353.

96. Reophax subfusiformis, sp.n. (Plate II, figs. 16-19).

Reophax scorpiurus {pars), Goes, 1894, ASF, p. 25, pi. vi, figs. 166, 167.

Twenty-six stations: 14, 20, 23, 27, 30, 42, 45, 126, 131, 136, 144, 148, 149, 660; WS 18, 32, 33,
37, 42, 43, 45, 46, 154, 348, 349; Drygalski Fjord.

Test large, usually composed of four chambers only, though specimens have been
observed up to six chambers. Chambers increasing rapidly in size, the last one forming
the bulk of the entire test, sometimes as much as four-fifths of the whole. The chambers
are turgid with sutural lines deeply depressed and are arranged on a more or less
strongly curved axis, the apertures being situated near the outer edge of the curve. The
final chamber is fusiform and tapering to the apertural end, which carries a prolonged
neck with large round aperture. The wall is thin and smoothly finished externally, built
of sand grains of varying sizes, often including some very large grains, embedded in
cement. Inner surface of wall very rough and irregular. Colour grey to nearly black

LITUOLIDAE 75

according to the minerals employed for building. Size varies up to 2-2 mm. in length,
0-8 mm. in greatest breadth.

This is by far the most abundant representative of the genus in the South Georgia
area; very common at St. 148, common at Sts. 27, 126, 144, frequent to very rare at
the remaining stations.

Goes {lit supra) gives a long series of figures ascribed to R. scorpiiirus. Cushman has
already separated figs. 160-3 as the types of R. curtus, which differs from R. subfusiformis
in the lesser number of chambers, typically three, and the absence of a produced neck.
This produced neck, which is so pronounced a feature of R. subfusiformis , is also typical
of R. denlaliniformis, Brady, but the chambering of my species is much nearer R.
scorpiiirus, and indeed R. subfusiformis might be regarded as occupying a position inter-
mediate between these species and combining the most prominent features of each.

The name is derived from Goes, who describes his fig. 167 as "subfusiformis, e sinu
Gullmaren Bahusiae, profund. 140 met."

97. Reophax pilulifer, Brady (F 82 a).

Three stations: 30; 53° 00' S, 34° 22' W; WS 523.

Extremely rare, single specimens and fragments only at these stations, the last two of
which are in deep water.

98. Reophax robustus, Pearcey.

Reophax robustus, Pearcey, 1914, SNA, p. 1006, pi. i, figs. 6-10.
One station: WS 523.
A single young specimen with two chambers only.

99. Reophax fusiformis (Williamson) (F 83).

Ten stations: 14, 23, 31, 131 ; WS 28, 33, 42, 113, 348; MS 68.

Very rare everywhere. The best specimens at Sts. 23, 131 and WS 113.

100. Reophax spiculifer, Brady (Plate II, fig. 20).

Reophax spicuHfera, Brady, 1879, etc., RRC, 1879, p. 54, pi. iv, figs. 10, 11; 1884, FC, p. 295,

pi. xxxi, figs. 16, 17.

Reophax spiculifera. Chapman, 1914, FORS, p. 62, pi. iii, fig. 16.

Two stations: WS 429, 523.

Five specimens at WS 429 and two at WS 523, all perfectly typical. This highly

selective species has a wide distribution.

ID I. Reophax dentaliniformis, Brady (F 84).

Three stations: 53° 00' S, 34° 22' W; WS 42, 429.

Frequent at WS 429 where the specimens are built up of rather coarse sand grains
with little cement. Single specimens only at the other stations, but more neatly con-
structed and conforming to Brady's figure.

76 DISCOVERY REPORTS

102. Reophax flexibilis, Schlumberger (F 86).
One station: WS 32.

A single fine specimen with sixteen chambers.

103. Reophax nodulosus, Brady (F 84 a).

Six stations: 151 ; WS 33, 50, 334, 336, 521.

Very rare everywhere. Fragments of typical specimens of medium size at St. 151
and WS 334. Small but fairly typical at WS 50 and 336. At WS 33 and 521 fragments of
a form occur which is assigned with some hesitation to R. nodulosus. The chambers
though very variable are long as compared with their breadth, sharply constricted at the
sutures and thin-walled, cement predominating over the minerals employed.

104. Reophax distans, Brady.

Reophax distans, Brady, 1879, etc., RRC, 1881, p. 50; 1884, FC, p. 296, pi. xxxi, figs. 18-22.
Reophax distafis, Faure-Fremiet, 1913-14, FMAF, 1913, p. 260, fig. i ; 1914, p. 2, pi. O, fig. 2.
Reophax distans, Cushman, 1918, etc., FAO, 1920, p. 12, pi. iii, figs. 5, 6.

Three stations: 53° 00' S, 34° 22' W; WS 63, 336.

A few fragments at each station. There can be little doubt that at WS 63 and 336 they
are parts of large specimens of Brady's type, coarsely built. The fragments from
53° 00' S, 34° 22' W are more doubtful. They may be terminal chambers either of a
more delicately constructed specimen of Brady's type, or perhaps of another and un-
described species.

105. Reophax distans var. gracilis, var.n. (Plate II, fig. 21).
Seven stations: 151 ; 53° 00' S, 34° 22' W; WS 334, 336, 353, 429, 522.

Test elongate, straight or perhaps slightly curved, consisting of pyriform or fusiform
chambers connected by more or less elongate stolon tubes. Usually fragmentary, no
specimen seen with more than two connected chambers. Wall thin, constructed of fine
sand grains embedded in cement. Exterior somewhat rough. Colour pale yellow, darker
at the stolons.

The description is based both on a single specimen (0-45 mm. in length) of two
chambers, one of which is apparently the proloculum, and on numerous individual
chambers of varyingsizes and proportions. In the two-chambered specimen, the breadth
of the chambers is roughly half the length (final chamber 0-2 mm. long, o-i mm. broad),
but apparently the length increases in proportion to the size of the chamber and in the
largest isolated chambers is three or four times the breadth (separate chambers measure
up to 0-7 mm. long, 0-2 mm. broad).

This very fragile little organism is frequent at WS 334, and 353, rarer at the other
stations, all of which are in very deep water. It is probably nearly related to R. distans,
Brady, but is sufficiently distinctive to merit varietal rank. Even at the same station there
is considerable variation in the size of the sand grains employed, with corresponding
differences in the external appearance. Sometimes the sand grains are so small as to be

LITUOLIDAE 77

hardly distinguishable, the surface then being neatly and smoothly finished. Otherwise
they are large and the surface is rough and angular.

Fracture appears to occur always at or near the base of the chamber and not as might
be expected in the middle of the stolon tube. The separate chambers thus resemble a
Lagena with tapering neck, and bear considerable resemblance to the figures of Lagen-
ammina laguncula, Rhumbler (R. 191 1, FPE, p. 92, pi. i, fig. 4). Indeed, until the dis-
covery of the specimen with two chambers, these isolated chambers were assumed to
belong to that species, or to some allied form of Proteonina. I have no personal know-
ledge of Rhumbler's species, so cannot say how closely the specimens agree with his
types, which were from the North Atlantic at depths of 1524-2400 m.

106. Reophax sabulosus, Brady.

Reophax rudis, Brady {non Costa), 1879, etc., RRC, p. 49.

Reophax sabulosa, Brady, 1882, FKE, p. 715; 1884, FC, p. 298, pi. xxxn, figs. :;, 6.

One station: WS 334.

One specimen from a depth of 3705 m. It is not perfect, but the characteristic thick
outer coating of sand grains is well preserved.

107. Reophax aduncus, Brady.

Reophax aduiica, Brady, 1S82, FKE, xi, p. 715; 1884, FC, p. 296, pi. xxxi, figs. 23-6.
Reophax aduncus, Cushman, 1918, etc., FAO, 1920, p. iq, pi. v, fig. i.

Two stations: 131 ; WS 33.

Only a single small specimen at each station.

Genus Hormosina, Brady, 1879

108. Hormosina globulifera, Brady (F 89).
One station: WS 523.

A few specimens only, mostly megalospheric single-chambered individuals.

Genus Haplophragmoides, Cushman, 1910

109. Haplophragmoides canariensis (d'Orbigny) (F 90) (Plate III, fig. 13).

Thirty-nine stations: 13, 14, 15, 20, 27, 30, 31, 42, 45, 123, 126, 131, 140, 144, 148, 149, 157,
660; WS 25, 27, 28, 31, 32, 33, 40, 41, 42, 45, 46, 50, 63, 63-4, 154, 177, 348, 357; Drygalski Fjord;
MS 14, 68.

Generally distributed in the shallower stations, almost entirely absent from deep water,
a depth of about 300 m. marking its limit with very rare exceptions, e.g. a single small
specimen at WS 63 in 1752 m. It is very common at Sts. 126, 144, 149, WS 28 and MS
68, and only less so at Sts. 27, 148 and WS 25, 32, 42 and 50. At the majority of the
stations, however, it is rare or very rare. The dominant type everywhere is an evolute
flattened form of normal size ; but, wherever the species is abundant, a much smaller

78 DISCOVERY REPORTS

involute form and an abnormally large form, which is also involute, occur in numbers
(e.g. at Sts. 123, 140, 144, 148, WS 25 and 154, MS 68). Abnormal and encysted
specimens were seen at several stations, and at St. 149 a specimen of the large form
occurred with six young individuals attached. These young shells are apparently
megalospheric and consist of three or four chambers. The majority of them lie near the
aperture of the parent shell, which from its size (i-2 mm. diameter) is probably micro-
spheric. They are presumably a young brood, formed outside the parent shell by the
breaking up of its protoplasm into particles surrounding a nucleus, these particular
young having remained in contact with the parent instead of achieving freedom.

A somewhat similar phenomenon of young individuals attached to their presumed
parent shell is of fairly frequent occurrence in the Globigerinae and will be noticed under
that genus (see p. 120).

no. Haplophragmoides crassimargo (Norman) (F 92).
Two stations: WS 51, 373.
Extremely rare. A very good and typical specimen at WS 51.

111. Haplophragmoides sphaeriloculum, Cushman (F 93).

Three stations: WS 334, 348, 353.

Extremely rare, the best specimen at WS 334.

112. Haplophragmoides scitulum (Brady) (F 94) (Plate III, figs. 11, 12).

Eight stations: 151 ; WS 48, 50, 63, 63-4, 334, 336, 351.

Very rare everywhere, the best and most typical specimens at WS 48 and 63-4. At
St. 151, WS 63 and 334 the species is represented by an oval variety, with umbiHcal
region only slightly depressed.

113. Haplophragmoides subglobosus (G. O. Sars) (F 95).

Ten stations: 151 ; WS 63, 63-4, 334, 336, 351, 353, 429, 521, 522.

Very rare everywhere except at WS 334 and 521, where the species was frequent.
Some excellent specimens at WS 522, rather smaller at WS 63, 63-4 and 521, very small
and starved elsewhere.

1 14. Haplophragmoides glomeratus (Brady).

Lituola glomerata, Brady, 1878, RRNP, p. 433, pi. xx, fig. i. Haplophragmium glotneratum,
Brady, 1884, FC, p. 309, pi. xxxiv, figs. 15-18.

Haplophragmium glomeratum, Heron-Allen and Earland, 1913, CI, p. 46, pi. ii, fig. 14.
Haplophragmoides glomeratum, Cushman, 1918, etc., FAO, 1920, p. 47, pi. ix, fig. 6.

Three stations: 136; WS 334, 429.

Very rare everywhere, but quite typical.

LITUOLIDAE 79

115. Haplophragmoides rotulatus (Brady).

Haplophragmium rotulatimi, Brady, 1879, etc., RRC, 1881, p. 50; 1884, FC, p. 306, pi. xxxiv,

figs. 5,6.

Haplophragmoides rotulatiim, Cushman, 1918, etc., FAO, 1920, p. 47, pi. ix, figs. 3, 4.

Two stations: WS 334, 351.

Rare at WS 334, where the best specimens were found, and verj' rare at the other
station. None of the specimens is very typical.

Genus Ammobaculites, Cushman, 19 10

116. Ammobaculites agglutinans (d'Orbigny) (F 96) (Plate II, fig. 22).
Eight stations: 151; WS63, 113, 334, 351, 353, 429, 472.

Generally very rare, often represented by a single specimen. At WS 63 a single large
and typical individual. At all the other stations the specimens are of the minute elongate
form figured by Brady (B. 1884, FC, pi. xxxii, fig. 22). They are most numerous at
St. 151, WS 334 and 472. In length they range up to o-6 mm.

Cushman (C. 1918, etc., FAO, 1920, p. 67) has made this particular figure of
Brady's a synonym for his species Ammobaculites reophaciformis (C. 19 10, NAFP,
p. 440, figs. 12-14; 1921, FP, p. 92, pi. xi, fig. 3 ; pi. xiv, fig. 3) on grounds which are not
very evident to me. His species is stated to live in shallow water on coral reefs, and has
a length up to 3-5 mm. It bears little resemblance to Brady's figure, which was drawn
from a specimen from Challenger St. 5, in 1090 fathoms to the west of Gibraltar
Strait. Brady's specimen is about 0-32 mm. in length and appears to be nothing more
than a pauperate condition of the common type. I have no hesitation in referring the
specimens to A. agglutinans, and attribute their small size to depth and unsuitable
conditions of life. At most they are only worthy of varietal distinction.

117. Ammobaculites americanus, Cushman (F 97).

Twenty-two stations: 14, 16, 20, 27, 30, 45, 123, 131, 136, 144, 148, 149; WS 28, 41, 42, 45, 47,
50, 349. 357; MS 14, 68.

Widely distributed and sometimes frequent to common. It appears to be almost
confined to depths between 100 and 300 m., none of the records being from less than
loom, and only two from over 300 m. The deepest record was at St. 16 in 727 m., where
a single specimen was obtained. The best stations were St. 144 (frequent) and WS 28
(common). At the last station an abnormal specimen was found in which the apertural
end was surrounded with an ovate mass or cyst of loosely aggregated sand and mud.
Whether this is a development from the organism, or a post-mortem attachment, cannot
be definitely stated. The latter case is the more probable as the sand grains embodied
are black as compared with the quartz in the actual test. The line of demarcation be-
tween A. americanus, Cushman, A. temiimargo, Brady, A. rostratus, Heron-Allen and
Earland, and A. bargmanni, sp.n., is very obscure in the early stages, which are practi-
cally inseparable. Only adult specimens show marked specific distinctions.

8o DISCOVERY REPORTS

ii8. Ammobaculites bargmanni, sp.n. (Plate II, figs. 23-6).
Five stations: 27, 45, 123, 144, 148.

Test large, free, planispiral, involute; consisting of about three convolutions of
chambers rapidly increasing in size and thickness. Umbilical area much depressed;
seven to nine chambers in final convolution, the last chamber, or at most the last two
chambers, becoming extended into a straight growth bearing the terminal aperture, a
large elongate slit. Sutures depressed, peripheral edge sub-acute, becoming rounded in
the final chambers. Surface smooth but unpolished, generally composed of very fine
sand grains, neatly and firmly bound together without visible cement. Coarser grains
are sometimes used, giving a rough exterior to the test. Walls thin, rough internally.
Colour grey, the umbilical portion sometimes brown. Size, up to 3-5 mm. in diameter,
sometimes even larger.

Young individuals appear to be inseparable from the young of A. americanus, except
for the slightly depressed sutural lines. There is a tendency to abandon the usual neatness
of construction in the final stages, the last chamber or two often having a few very large
sand grains incorporated in the walls.

A. bargtnamii appears to be a very distinctive species, linking the genera Haplo-
phragmoides and Ammobaailites, and so perhaps demonstrating the zoological continuity
of the group, and the futility of their separation. Up to the time of the formation of the
last two or three chambers, the test is a typical Haplophragmoides, its nearest ally being
apparently H. compressiim (Goes) (= H. ernaciatum, Brady) (G. 1882, RRCS, p. 141,
pi. xii, figs. 421-3 ; 1896, DOA, p. 31). With the formation of these final chambers, it
assumes the external characteristics of Ammobaadites, and bears some resemblance to
A. americanus, from which, however, it is easily separable owing to its involute con-
struction, depressed sutures, inflated chambers, and its size, which is nearly double that
of A. americanus. It appears to be one of the largest recorded species in either genus.

A. bargmanni is not infrequent at a depth of 270 m. at St. 45, which is at the entrance
to Cumberland Bay; it is rarer at the lesser depths of other adjacent stations. The
species appears, therefore, to be confined to a very limited area. At Sts. 27 and 148,
coarser material is employed for construction of the test than at the remaining stations,
with the result that the specimens have a very rough and untidy appearance. At St. 144,
both smooth and rough-walled specimens were noted.

The species is named after H. E. Bargmann, Ph.D., of the Discovery Staff, who has
assisted me in the preparation of index, charts, etc.

119. Ammobaculites rostratus, Heron-Allen and Earland (Plate V, figs. 22-5).

Ammobaculites rostratus, Heron-Allen and Earland, 1929, etc., FSA, 1929, p. 328, pi. ii, figs.
14-17.

Fourteen stations: 27, 41, 45, 123, 126, 143, 149, 660; WS 28, 32, 40, 42, 63-4; MS 68.

Test free, thin-walled and rather fragile, compressed, piano-spiral and evolute.
Consisting of two to three convolutions, with from five to seven chambers in the last con-

LITUOLIDAE 8i

volution. The final chamber is produced at a tangent to the spiral and terminates in a
flattened nipple with slit-shaped aperture. Umbilical region depressed on both sides,
marking the area of the inner convolutions. The chambers of the last convolution are
sHghtly inflated and do not extend to the edge of the test, which is thus furnished with
a solid carina having a rounded peripheral edge. The sutural lines are generally
obscure. Colour light grey, surface rough and unpolished. Walls constructed either of
fine sand and mud with a considerable proportion of grey cement, or of mud without
apparent cement, according to the environment. When sand grains are employed they are
generally minute and of uniform size, but whether sand or mud is used, symmetrical
construction is often spoiled by the inclusion of one or two large grains. The nipple-like
terminal of the final chamber is more neatly constructed than the rest of the shell.

The characteristic terminal chamber presumably marks the completion of growth, as
it is only found in the largest specimens. For this reason, coupled with the fragility of
the test, specimens exhibiting this feature are rare. Broken and immature specimens are
of frequent occurrence, and in this condition the species closely resembles Ammobacii-
lites americmms, Cushman (= Haplophragmium fontinense, Brady hoh Terquem), to
which our species is probably closely allied.

Dimensions. Length of perfect specimens 2-0-2-4 mm.; breadth 1-4-1 -6 mm.

The species is very generally distributed but is never so common in occurrence as
A. amerkamis. The best stations for the mud-building form are Sts. 45, 126 and 149;
for the sandy form St. 27. At St. 660 and MS 68 both mud and sand-building forms
occur in considerable numbers and attain large dimensions.

120. Ammobaculites tenuimargo (Brady).

Haplophragmium tenuimargo, Brady, 1882, FKE, xi, p. 715; 1884, FC, p. 303, pi. xxxiii, figs.

13-16.

Haplopliragmium tenuimargo, Flint, 1899, RFA, p. 275, pi. xix, fig. 3.

Ammobaculites tenuimargo, Cushman, 1918, etc., FAO, 1920, p. 65, pi. xiii, figs. 3-5.

Four stations: 27, 144, 148; WS 429.

Always very rare. The best and most typical specimen was found at WS 429 at a
depth of 2549 m. More numerous at St. 144. All the specimens except the single one
found at WS 429 are very broad in the spiral portion as compared with Brady's figure,
suggesting a close relationship with A. americanus.

121. Ammobaculites foliaceus (Brady).

Haplophragmium foliaceum, Brady, 1879, etc., RRC, 1881, p. 50; 1884, FC, p. 304, pi. xxxiii,

figs. 20-5.

Ammobaculites foliaceus, Cushman, 1918, etc., FAO, 1920, p. 64, pi. xiii, figs, i, 2.

Three stations: 151 ; WS 63, 334.

Rare at WS 334 and very rare at St. 151 and WS 63. All the specimens are small and
less neatly constructed than usual.

82 DISCOVERY REPORTS

Genus Ammomarginulina, Wiesner, 1931

122. Ammomarginulina ensis, Wiesner (Plate III, figs. 1-4).
Ammomarginulina ensis, Wiesner, 193 1, FDSE, p. 97, pi. xii, fig. 147.
Two stations: WS 199, 334.

Test free, much compressed, especially in the early stage which is planospiral, rather
loosely coiled, and contains about two convolutions, the last convolution containing
nine to ten narrow chambers. These are followed by a straight series of chambers,
four to seven in number, narrow and compressed at first, tending to become more
inflated towards the aperture, which is central in the early stages, sometimes becoming
slightly produced and marginal in the adult form. Sutural lines curved butvery indistinct,
the chambers being difficult to distinguish except in balsam-mounted specimens. Con-
structed of rather large mineral grains firmly agglutinated, but without much visible
cement; surface rather rough, colour pale straw to golden yellow.

Length, about 0-4 mm. ; breadth at spiral portion ranging between 0-2 and 0-3 mm. ;
breadth of uniserial portion ranging between o-i and 0-15 mm.

Many specimens occur at WS 334 at a depth of 3705 m. This station is well to the
north of South Georgia, and the absence of the species at stations intermediate between
WS 334 and 199, which is far to the south, near the South Orkney Islands, is rather
remarkable. Only a single specimen was noted at WS 199 at a depth of 3813 m., but
the species, though never common, is of very constant occurrence in the deep water
Antarctic material.

I had already described this species in MS. under the name Ammobacidites ensis
before I became aware of Wiesner's prior publication (^ut supra). Wiesner has created a
new genus Ammomarginulina, of which this is the genotype. The characteristic feature of
the genus is the position of the aperture, which is on the outer marginal edge, as in
Margimilina. I had not attributed so much importance to this feature, though the
general resemblance of the form to MarginuWia ensis, Reuss, had, by a coincidence,
caused me to select the same specific name as Wiesner. His definitions of the new genus
and species are brief, but sufficient in connection with his figure drawn from a trans-
parent specimen for identification with my specimens. He writes of the genus Ammo-
marginulina: "The sandy shell is spirally coiled at first, later chambers arranged in a
straight line, sutures curved, aperture on the back edge ". Of the species it is stated that
"it bears the features of the genus. The shell wall is built up of flat mica-like mineral
flakes with little cement ". No dimensions, locality or frequency are given, but from the
figure his specimen agrees in size with ours. As Wiesner's material was from the opposite
side of the Antarctic area, the species is apparently universally distributed over the
Antarctic seas.

LITUOLIDAE 83

Genus Placopsilina, d'Orbigny, 1850

123. Placopsilina cenomana, d'Orbigny (F 98).
Five stations: 45, 136; WS 51, 61, 154.

The best stations were WS 51 and 61, where many specimens of the small neat form
found in British seas were observed encrusting stones. Otherwise the specimens are few
and rather feeble. At St. 45 two unattached specimens were found, probably dislodged
from their original host.

Sub-family TROCHAMMININAE

Genus Ammolagena, Eimer and Fickert, 1899

124. Ammolagena clavata (Jones and Parker) (F 99).
Three stations: WS 334, 521, 522.

Several good specimens at each station.

Genus Tolypammina, Rhumbler, 1895

125. Tolypammina vagans (Brady) (F 100).

Sixteen stations: 17, 27, 126, 136, 144, 145, 14S; WS 31, 33, 51, 61, 154, 344, 353, 363, 365.

Common at WS 51 and 61, frequent at St. 17 and WS 33, but rare elsewhere. All the
specimens are rather small, the best being at Sts. 27, 144 and WS 33.

Genus Ammodiscus, Reuss, 1861

126. Ammodiscus incertus (d'Orbigny) (F loi).

Eighteen stations: 14, 20, 45, 123, 131, 140, 144, 148; WS 27, 33, 40, 42, 113, 154, 348, 349, 353,
426.

Widely distributed and often fairly common, the best stations being St. 144 and WS 1 54.
The specimens are larger and better developed with more ferruginous cement than in
the Falkland area, but present the same variations as we commented on in that Report.
Flat and evenly coiled specimens are very rare, the best being at Sts. 45, 148, WS 33,
42 and 353. At nearly all the other stations the specimens are more or less irregular in
coiling and pass imperceptibly into Glomospira gordialis. None of the specimens are
very large, seldom exceeding 0-3 mm. in diameter, and they are all microspheric. At
WS 27 a sessile specimen was seen. All the specimens have a matt surface, i.e. there is
never sufficient cement to envelop the minute sand grains. Their colour varies from
white through grey to ferruginous brown.

Genus Glomospira, Rzehak, 1885

127. Glomospira gordialis (Jones and Parker) (F 102).

Twenty-nine stations: 15, 17, 27, 30, 42, 45, 123, 126, 140, 144, 149; WS 25, 27, 31, 33, 40, 41,
42, 43, 51, 61, 63, 63-4, 113, 349, 365, 426, 428; MS 14.

84 DISCOVERY REPORTS

More generally distributed and more abundant than Ammodisciis incertus, from the
irregularly formed local variety of which it is often hardly separable.

The typical gordialis in which the coiling of the tube is entirely irregular is found only
at Sts. 27, 30, 140, 149, WS 25, 27, 31 and 63, MS 14. At WS 63 the specimens were
highly polished owing to excess of cement, all others presenting the matt surface
characteristic of the genus in this area. The typical form also occurs at WS 33, in com-
pany with the local variation akin to A. incertus. The best specimens, both in numbers
and development, were obtained at this station.

The local variety, in which the earlier convolutions are regularly coiled as in A.
incertus and the later convolutions very irregular, is found at all the remaining stations.
Sessile forins often occur, notably at WS 51.

I have followed recent practice in assigning this species to Glomospira, but the
difficulty of separating many of the specimens from A. incertus convinces me that, so far
at least as Glomospira gordialis is concerned, the generic distinction has no real value. In
my opinion this also holds good as regards many old genera revived and new genera
created in recent years by systematists.

The remarks as to the colour of Ammodiscus incertus hold good as regards G. gordialis
also.

128. Glomospira charoides (Jones and Parker) (F 103).
Six stations: 151; 53^ 00' S, 34° 22' W; WS 334, 353, 521, 522.

Singularly rare, one specimen only at each station, all very small except at St. 151 and

WS 521.

Genus Turritellella, Rhumbler, 1903

129. Turritellella shoneana (Siddall) (Plate III, figs. 9, 10).

Trochatnmina shoneana, Siddall, 1878, FRD, p. 46, figs, i, 2.

Ammodiscus shoneanus, Brady, 1884, FC, p. 335, pi. xxxviii, figs. 17-19.

TurritiUella shoneana, Rhumbler, 1903, ZRR, p. 283, fig. 135.

Ammodiscus shoneanus, Heron-Allen and Earland, 1913, CI, p. 49, pi. iii, fig. 6; 1922, TN.

p. no, pi. i, fig. 22.

Turritellella shoneana, Cushman, 1918, etc., FAO, 1918, p. 102, pi. xxxviii, figs. 5-7.

One station: 145.

A number of specimens were found at St. 145, but none at any other station. This
may be due to the nature of the material (trawl-washings) and the shallowness of the
water, 26-35 m.

All the specimens are of the minute form usually found in British gatherings and
identical with such. Both megalospheric and microspheric individuals occur, the
former being characterized by a globular proloculum almost equal in diameter to the
subsequent coil of chambers, which consequently has nearly parallel sides. In the
microspheric form, the proloculum is quite small and the test forms a slender cone. The
texture is finely granular owing to the small proportion of cement employed. Colour
light brown, fading into white at the oral extremity.

LITUOLIDAE 85

130. Turritellella laevigata, sp.n. (Plate III, figs. 5-8).
One station: WS 33.

Test free, straight or irregularly curved, cylindrical, but frequently constricted at
irregular intervals, rounded at the proloculum, which often exceeds the maximum
diameter elsewhere, truncate at the oral extremity, which is slightly narrowed and bears a
rather large central, simple aperture. Surface smooth and highly polished owing to the
large amount of cement employed. Colour pale yellow, generally rather darker at the
aboral end. Convolutions variable in width and number, ranging up to twenty or more
in large individuals.

Viewed as a transparent object under a high magnification, the test is seen to be
constructed of very minute mineral grains embedded in an excess of cement. The walls
are thin, the central columella being comparatively thick. The constrictions visible
externally appear to mark stages of growth ; a fresh coil is formed at the beginning of
each stage, the central columella being often at a slightly difl:'erent angle from the pre-
ceding one. The first convolution of the new coil is wider than the preceding and
following convolution, because it embraces the apertural end of the previous growth.
This gives an appearance of segmentation, but we have been unable to trace any actual
septum at the point of junction. All the specimens are megalospheric.

Size up to 0-7 mm. long, o-i mm. broad.

T. laevigata cannot be confused with T. shoneana, from which it diff'ers in the ex-
cessive quantity of cement employed, the external constrictions and the consequently
varying axis of growth.

A good many specimens were found at WS 33 at a depth of 130 m. The species was
not found elsewhere in the South Georgia area but occurs also at St. 175 in the South
Shetlands (200 m.) and WS 482 in the same area (100 m.). See note A, Appendix,
p. 132.

Genus Trochammina, Parker and Jones, i860

131. Trochammina squamata, Jones and Parker (F 104).
Five stations: 145; WS 27, 33, 154, 334.

Rare everywhere and generally small. Sessile specimens at WS 154 and 334.

132. Trochammina rotaliformis, J. Wright (F 105).

Two stations: 149; WS 33.

Very rare, only one or two specimens at each station. At WS 33 the specimens are
very highly convex on the dorsal side.

133. Trochammina ochracea (Williamson) (F 107).
Three stations: WS 41, 51 ; MS 14.

Extremely rare. A sessile specimen at WS 51.

8-2

86 DISCOVERY REPORTS

134. Trochammina inflata (Montagu) (F 108).
Six stations: 149, 151 ; WS 353, 429, 522, 523.
Very rare and small, the best at WS 429.

135. Trochammina malovensis, Heron-Allen and Earland(F 109) (Plate IV, figs. 38-40).

Fifty-one stations: 13, 20, 23, 27, 30, 31, 42, 45, 123, 126, 131, 136, 140, 143, 144, 145, 148, 149,
i5i;WSi8, 25,27, 28, 31,32, 33,40, 41, 42, 43, 45,46,47,48, 50, 52, 63, 113, 154, 177, 334, 336,
343, 348, 349, 357, 418, 523 ; Drygalski Fjord; MS 14, 68.

This species has proved to be more widely distributed in the South Georgia area than
was anticipated from the records which were available when it was originally described.
It proves to be almost ubiquitous, often common or even very common, the best
stations being 126, 144, 149; WS 25, 31-3 and 40. The species is more subject to
variation than in the Falklands, the dorsal surface varying from the highly convex
trochoid spiral of the original figures to an almost flat dorsal surface. These flat speci-
mens can hardly be separated from T. nana. Every intermediate stage of convexity is
found and the two forms often occur together, though one or the other is generally
dominant. The South Georgia specimens are seldom so ferruginous in colour as the
Falkland types, and at many stations the tests are normally grey. Abnormalities are very
rare but a few were observed, also a case of encystment at WS 113. The range of the
species extends between depths of 26 and 3705 m., but it is most at home in moderate
depths below 200 m. and becomes very rare and small in deep water.

136. Trochammina nana (Brady) (F no).
Two stations: 151 ; WS 28.

A single typical specimen only at each station,

137. Trochammina bradyi, Robertson (F in).
Three stations: 131, 151 ; WS 334.

Very rare, never more than one or two specimens at each station.

138. Trochammina nitida, Brady.

Trochammina niiida, Brady, 1879, etc., RRC, 1881, p. 52; 1884, PC, p. 339, pi. xli, figs. 5, 6.
Trochammina nitida. Goes, 1894, ASF, p. 30, pi. vi, figs. 225-30.
Trochammitia nitida, Heron-Allen and Earland, 1916, FWS, p. 228, pi. xl, figs. 19-21.
Trochammina nitida, Cushman, 1918, etc., FAO, 1920, p. 75, pi. xv, fig. 2.

Ten stations: 16, 23, 31, 42, 136, 144; WS 43, 46, 50, 349.

Very rare, but good specimens at most stations, the best at St. 16 and WS 349.

139. Trochammina turbinata (Brady).

Haplophragmium turbinatim, Brady, 1879, etc., RRC, 1881, p. 50; 1884, FC, p. 312, pi. xxxv,
fig. 9.

Trochammina turbinata, Eimer and Fickert, 1899, AVF, p. 695; Cushman, 1918, etc., FAO,
1920, p. 81, pi. xvii, fig. 2.

LITUOLIDAE 87

Thirteen stations: 23, 31, 131 ; WS 37, 40, 42, 43, 46, 50, 63-4, 334, 349, 353.

Very rare and very small; the best at WS 42, 50 and 349. Seldom more than one or
two specimens at a station.

140. Trochammina globigeriniformis (Parker and Jones) (F no a).
Five stations: 14; WS 40, 334, 351, 429.

Very rare and usually very small. The only good specimens at WS 429.

Genus Globotextularia, Eimer and Fickert, 1899

141. Globotextularia anceps (Brady) (F 112).
Two stations: 131 ; WS 523.

Only a single specimen at each station.

Genus Ammosphaeroidina, Cushman, 1910

142. Ammosphaeroidina sphaeroidiniformis (Brady).

Haplophragmium sphaeroidiniforme, Brady, 1884, FC, p. 313.

Haplophragmimn sphaeroidiniforme, Chapman, 1907, TFV, p. 24, pi. iii, figs. 50-1.
Anmtosphaeroidina sphaeroidiniformis, Cushman, 1910, etc., FNP, 1910, p. 128, fig. 202; 1918,
etc., FAO, 1920, p. 87, pi. xvii, fig. 5.

One station : WS 334.

A single small specimen.

Genus Ammochilostoma, Eimer and Fickert, 1899

143. Ammochilostoma galeata (Brady).

Trochammina galeata, Brady, 1879, etc., RRC, 1881, p. 52; 1884, FC, p. 344, pi. xl, figs. 19-23.
Ammochilostoma galeata, Eimer and Fickert, 1899, AVF, p. 692, fig. 39.

Ammochilostoma galeata, Cushman, 1910, etc., FNP, 1910, p. 127, figs. 198-201 ; 1918, etc.,
FAO, 1920, p. 85.

One station: WS 334.

A single specimen.

144. Ammochilostoma pauciloculata (Brady).

Trochammina pauciloculata, Brady, 1879, etc., RRC, 1879, p. 58, pi. v, figs. 13, 14; 1884, FC,

p. 344, pi. xh, figs. I, 2.

Ammochilostoma pauciloculata, Eimer and Fickert, 1899, AVF, p. 692.

Ammochilostoma pauciloculata, Cushman, 1910, etc., FNP, 1910, p. 126, fig. 197; 1918, etc.,

FAO, 1920, p. 86.

Two stations: 151 ; WS 334.

Several excellent specimens at these stations.

88 DISCOVERY REPORTS

145. Ammochilostoma ringens (Brady).

Trochammina ringens, Brady, 1879, etc., RRC, 1879, p. 57, pi. v, figs. 12 a, b; 1884, FC, p. 343,

pi. xl, figs. 17, 18.

Trochammina ringens, Flint, 1899, RFA, p. 281, pi. xxvii, fig. i.

AmmoiJiilostoma ringens, Eimer and Fickert, 1899, AVF, p. 692.

Haplophragmoides ringens, Cushman, 1910, etc., FNP, p. 107, fig. 166; 1918, etc., FAO, 1920,

p. 49, pi. ix, fig. 2.

One station: WS 351.

Only a single specimen, quite typical except as regards the aperture, which is on the
inner edge of the final chamber, but not above the edge and separated from the previous
convolution. This abnormality gives the shell a pseudo-isomorphism with Pidlenia
subcarinata.

Genus Nouria, Heron-Allen and Earland, 19 14

146. Nouria harrisii, Heron-Allen and Earland (F 113 a) (Plate HI, figs. 14-16).
One station: WS 33.

Quite a number of excellent specimens were found at this station. They vary greatly
in size, ranging between 0-3 and o-8 mm. in length, and are rather less broad than the
type. Otherwise they agree very well, though there is a greater tendency to use broken
spicules and more cement, and to employ mineral grains for filling in odd crannies,
especially at the base of the test. The projecting basal spicules are very pronounced in
some of the specimens.

Sub-family LOFTUSINAE
Genus Cyclammina, Brady, 1876

147. Cyclammina cancellata, Brady (F 114).

Nine stations: 151 ; WS 63, 334, 336, 351, 418, 426, 429, 521.

Very rare, but good specimens of both megalospheric and microspheric forms, the
best at WS 334 and 336. With the exception of WS 418, where a single microspheric
individual was found at a depth of 227 m., all the records are from deep water between
1 170 and 3780 m.

148. Cyclammina orbicularis, Brady.

Cyclammina orbicularis, Brady, 1879, etc., RRC, 1881, p. 53; 1884, FC, p. 353, pi. xxxvii, figs.

17-19.

Cyclammina orbicularis, Cushman, 1910, etc., FNP, 1910, p. 113, fig. 173; 1918, etc., FAO,

1920, p. 57, pi. xi, figs. 7-9.
One station : WS 334.
Two rather small specimens from 3705 m. at this station.

LITUOLIDAE 89

Sub-family SILICININAE

Apart from certain fossils of which we have little personal experience, the existence of
really siliceous Foraminifera has always appeared to be a matter of uncertainty. A
statement by Brady (B. 1884, FC, pp. 100, 131) is probably responsible for the opinion
expressed by Lister (L. 1903, F, p. 53) and Cushman (C. 1928, F, pp. 11, 144,
148) that tests of nearly pure silica may be developed under deep-sea conditions.
Writing of the organisms found at Challenger St. 238, in the North Pacific, at a depth
of 3950 fathoms, Brady says: " Miliolae were the only representatives of the calcareous
forms, and the shells of these were no longer calcareous, but consisted of a thin film of
homogeneous silica, unaffected by acids, and iridescent when first taken out of spirit. . . .
A few Miliolae. . .were found to be unaffected by acids, and, upon further examination,
it became apparent that the normal calcareous shell had given place to a delicate homo-
geneous siliceous investment ". It would appear from this that Brady had actually dealt
with these specimens; but Heron-Allen (H.-A. 1915, RPF, pp. 264 and 272) states, on
the authority of Sir John Murray, that the specimens examined on board the ' Challenger '
were not preserved, and that Brady's statement was made on the strength of Murray's
notes, and not from his personal knowledge of actual specimens.

It is almost certain that the objects seen by Murray were chitinous linings such as are
common to all Foraminifera. At any rate, so far as we are aware, no one has subse-
quently described recent siliceous Foraminifera, although chitinous specimens have
been observed in both deep and shallow waters by many persons, ourselves included.
The discovery of truly siliceous Foraminifera in recent dredgings is therefore a matter
of more than ordinary interest.

In the first place let us define the term "siliceous" as meaning " capable of resisting
the action of strong acids without structural change". There are many Foraminifera
belonging to various groups which, on superficial examination, might be considered
siliceous, but which will not withstand this test. All the Astro rhizidae and Lituolidae
make more or less use of siliceous particles in the construction of their agglutinate tests,
encrusting the chitinous membrane, which forms their basic structure, with sand grains
embedded in a cement secreted by the animal, which contains varying proportions of
silica, ferric oxide, and carbonate of lime. Moreover, many species of Miliolidae are in
the habit of encrusting their normally calcareous tests with sand grains of varying sizes,
often in such abundance as to conceal the calcareous structure. But the apparently
siliceous tests of the Miliolidae are instantly dissolved with effervescence on the addition
of acid, while the ferruginous cement of the agglutinate forms, almost without exception,
breaks down under the prolonged action of the acid. It is true that Brady (B. 1884, FC,
p. 286) states that "in rare instances silica or some siliceous compound is employed,
either by itself or in conjunction with other mineral substances", but the only example
he gives is Reophax nodulosiis, of which he says : " The incorporating medium is more or
less siliceous, sometimes to such a degree that large specimens, half an inch or an inch
in length, preserve their form after all the calcareous and ferruginous constituents have

90 DISCOVERY REPORTS

been removed by means of strong acids, and still retain sufficient firmness to bear
handling without injury". This exception can hardly be said to affect the general rule
that acid destroys agglutinate Foraminifera.

The history of the organisms, which we are now describing, begins in 1913, when
Faure-Fremiet (F. 1913-14, FMAF, p. 4, pi. O, fig. 5) figured and described an or-
ganism from the Antarctic, which he assigned to Miliolina alveoliniformis, Brady, a well-
known coral-reef species. Apart from the fact that Faure-Fremiet 's specimens had
agglutinate tests and a cribrate aperture, it is clear that they had little in common with
Brady's species. Faure-Fremiet seems to have had doubts as to his identification, and
raises the question as to whether his organism is not a true arenaceous form, isomorphous
with Brady's species. Faure-Fremiet informs us that his mounted specimens of
Miliolina alveoliniformis, Brady, have been mislaid, and are probably lost, but, having
refreshed his memory by reference to his paper, he has no doubt whatever that his
figure No. 5 b represents the oral end of his specimens, and that the characteristics there
shown were constant and apparently quite distinct from the normal milioline aperture.

We must, therefore, accept the position that there is a species in the Antarctic answer-
ing to Faure-Fremiet's description and figures, and characterized by a cribrate aperture.
As the attribution of the ' Pourquoi-Pas? ' specimens to Miliolina alveoliniformis, Brady,
cannot be upheld, we suggest the new name Miliammina cribrosa for Faure-Fremiet's type.

In 1914 Chapman (C. 1914, FORS, p. 59, pi. i, fig. 7) described and figured under the
name Miliolina oblonga var.n. arenacea some specimens from the Ross Sea in the
Antarctic. He described his organism as " quite a constant form", differing only from
the porcellanous type of Montagu in the finely arenaceous material of the test. He also
remarked that no porcellanous specimens were found in the same dredgings, and that his
organism is readily distinguishable from Miliolina agglutinans (d'Orbigny), which is an
agglutinate species.

Chapman's variety has since been designated as the genotype of the genus Miliammina
by Professor T. D. A. Cockerell under the name Miliammina arenacea (Chapman). (See
correspondence in Nature, June 28 and September 20, 1930.)

In 1922 we published a report on the Foraminifera of the Terra Nova Antarctic
Expedition (H.-A. and E. 1922, TN, p. 66) in which we recorded that Miliolina oblonga
var. arenacea. Chapman (synonym Miliolina alveoliniformis, Faure-Fremiet non Brady)
was the most typical Miliolid of the Terra Nova Antarctic collections, and that it
presented a considerable range of form. Although a minute examination of the test
was made to ascertain its constituents, no doubts were then entertained as to its milioline
nature, and consequently no chemical tests were employed. Nor did we attempt to
separate the different variations, which were all listed under Chapman's name.

In connection with the examination of the Discovery material, it soon became
apparent that an organism similar to Chapman's was a typical constituent of the muds
in the South Georgia area, where it was found in great variety in nearly every sounding
from moderate depths. In the course of experiments to determine the proportion of
calcareous matter in the test (the South Georgia muds being almost entirely mineral

LITUOLIDAE gi

and diatomaceous), we found to our surprise that there was none. Immersions in strong
nitric acid under the microscope produced no effervescence, nor did 24 hours' immersion
in acid affect the constitution or solidity of the test. It became evident, therefore, that we
were not dealing with a Miliolina at all, but with an entirely new organism, for which we
proposed the generic name Miliammina (= sandy or siliceous " Milioliform " isomorph).

Miliammina is evidently closely related to a fossil organism recently described by
Cushman and Church (CC. 1929, CFC, p. 502, pi. xxxvi, figs. 10-12) from the Upper
Cretaceous of California, under the generic name Silicosigmoilhw . The wall is described
as finely arenaceous, with siliceous cement on which the strongest acid makes no im-
pression. The chambers are on a sigmoiline plan, and it is apparently isomorphous with
Sigmoilina in the porcellanous group, except for its simple aperture which lacks the
tooth found in the calcareous genus. It is referred to the Silicinidae, a family established
by Cushman in 1928 {tit supra) for the reception of the fossil genera Silicina, Borne-
mann, 1874, Rzehakina, Cushman, 1927, Iiwohtina, Terquem, 1862, and Problematina,
Bornemann, 1874. In the definition of the family it is stated that the wall is arenaceous,
" usually siliceous, sometimes partly calcareous ". Of the four genera, Rzehakina seems
to have little in common with the others, but its structural plan is certainly near that of
Silicosigmoilina, and its wall is siliceous. To what extent the other genera would with-
stand testing with acid we cannot say, but Involutina at least has always been regarded as
largely calcareous, and is included by Brady in his sub-family Endothyrinae.

Miliammina is evidently one of the Lituolidae in Brady's system of classification, but
is not easily placed in that family. The sub-family Trochammininae includes many
genera, which have little obvious relationship to each other, though all characterized by
neatly agglutinate tests. Many of them are isomorphous with other genera of porcel-
lanous and hyaline Foraminifera. In this connection it is noteworthy that no true
isomorph of that large and important porcellanous sub-family, the Miliolininae, has
been known until recently. Miliammina, Silicosigmoilina, and Rzehakina will now to
some extent fill this gap, and as they agree in the siliceous constitution of their cement as
contrasted with the highly ferruginous cement characteristic of the Trochammininae,
we propose to establish a new sub-family of the Lituolidae, " Sihcininae", for the re-
ception of these three genera. The Silicininae may be defined as having thin agglutinate
tests consisting of numerous chambers, non-labyrinthic, arranged on a milioline plan,
and furnished with a terminal aperture, with or without a tooth, the wall composed of
minute minerals and Diatoms embedded in an excess of siliceous cement, with smooth
or polished exterior, and smooth interior, surfaces.

Genus Miliammina, ^ Heron-Allen and Earland, 1930
Test free, chambers arranged on a triloculine or quinqueloculine plan ; wall imper-
forate, composed of very minute mineral fragments embedded in an excess of siliceous
cement, smooth or polished, rarely rough. Aperture terminal, furnished with a tooth,
perhaps sometimes cribrate.

1 See note B, Appendix, p. 132.

92 DISCOVERY REPORTS

Wiesner, in his recent monograph on the Foraminifera of the German South Polar
Expedition (W. 193 1, FDSE), does not record or figure any specimens which can be
identified with Miliamnihia. This is all the more noteworthy as Wiesner was quite
familiar with our Terra Nova Report (see p. 54, FDSE, et passim) and his material came
from approximately the same area as much of the Terra Nova material, in which the
genus was plentiful.

149. Miliammina oblonga, Heron-Allen and Earland (Plate III, fig. 17; Plate V, figs.
1-5, 7, 8).

Miliammiria oblonga (Chapman), Heron-Allen and Earland, 1929, etc., FSA, 1930, p. 41, pi. i,
figs. 1-6, 22-3.

Forty-seven stations: 13, 14, 20, 23, 27, 30, 31, 42, 45, 123, 126, 129, 131, 136, 140, 143, 144, 148,
149, 660; WS 18, 25, 27, 28, 32, 33, 37, 40, 41, 42, 43, 45, 46, 47, 48, 50, 52, 63, 63-4, 113, 154,
348. 349' 357 ; Drygalski Fjord; MS 14, 68.

Test regularly quinqueloculine ; chambers tubular, larger at the aboral extremity;
peripheral edge rounded to sub-acute; sutural lines almost flush in the young stage,
becoming more or less depressed with increasing size of shell. Aperture crescentiform
at extremity of final chamber, sometimes on a somewhat produced neck, with or without
a reverted collar, furnished with a small simple tooth. Wall thin, composed of minute
mineral grains embedded in an excess of siliceous cement, smooth, often polished, in
which case the mineral particles of which it is partly composed are more distinct.
Colour variable from very light to dark grey, rarely brown, or afltected by the colour of
the mineral particles employed. Size very variable in different localities. Young
individuals have been seen only 0-125 rrinfi- in length, but average well-developed tests
are about 0-40 0-50 mm. in length, 0-20 mm. in breadth, 0-15 mm. in thickness. The
thickness of the wall in an adult shell is estimated at o-oo5-o-oio mm., and the largest
mineral flakes employed by the South Georgian specimens rarely exceed these dimensions.
Elsewhere larger mineral flakes are used, as we noted in 1922 {ut supra). In the South
Georgian material, diatomaceous debris appears to be used to some extent in the con-
struction of the test, contrary to our experience with the Terra Nova specimens, in
which only mineral matter eould be detected (H.-A. and E. 1922, TN, p. 67). But it is
very difficult to verify the nature of the minute constituents of the test.

Almost universally distributed, M. oblonga is probably the commonest and most
characteristic rhizopod of the South Georgia area, occurring in more or less abundance
in nearly every coastal sounding. It reaches its optimum development, both as to size
and numbers, in moderately shallow water. The best stations are Sts. 20, 126, 144,
WS 28, 33, 42 and 154; but occasional small specimens have been found down at a
depth of 1752 m. at WS 63, beyond which depth it has not been seen in any of the
soundings examined. The species is subject to considerable variation, mainly owing to
differences in the rotundity of the tubular chambers, with corresponding changes in the
sutural depressions and peripheral angles.

The original type of Chapman, M. arenacea (Chapman), does not occur in the

LITUOLIDAE 93

South Georgian material, but appears to be confined to tlie Antarctic area and will be
dealt with in our report on the same. A figure of M. arenacea is given for comparison
(Plate V, fig. 6).

It may be mentioned here that the range of M. oblonga is now known to extend beyond
the ice-area. It occurs in material from St. 6 off Tristan d'Acunha (80-140 m.) and
from WS 4 off South-west Africa (40 m.).

150. Miliammina obliqua, Heron-Allen and Earland (Plate V, figs. 9-14).

Miliammina obliqua, Heron-Allen and Earland, 1929, etc., FSA, 1930, p. 42, pi. i, figs. 7-12.

Forty-six stations: 14, 15, 20, 23, 27, 28, 30, 31, 41, 42, 45, 123, 126, 129, 131, 140, 143, 144, 148,
149, 660; WS 25, 28, 32, 33, 37, 40, 41, 42, 43, 45, 46, 47, 48, 52, 63, 63-4, 113, 154, 348, 349,
357, 418; Drygalski Fjord; MS 14, 68.

Test quinqueloculine, with tubular chambers broader at the aboral ends, the early
chambers lying transversely across the centre of the test. Sutural lines depressed and
peripheral edge rounded. Aperture crescentiform, usually flush with the terminal end
of the test, sometimes on a slightly produced neck with collar, furnished with a minute
simple tooth. Walls thin and smooth, sometimes polished, embodying the fine mineral
grains which appear to be somewhat larger than in M, oblonga. Colour varying from
nearly white to dark grey, occasionally brownish. Size variable, but never attaining the
proportions of M. oblonga. Average length o-30-o-35 mm., breadth o-i5-o-i8 mm.,
thickness o-i2 mm.

This little form, which frequently, but not invariably, occurs in company with M.
oblonga, differs from that species mainly in the transverse disposition of the central
chambers. It may be considered to be isomorphous with Miliolina bosciana (d'Orbigny),
and occupies the same position with regard to Miliammina oblonga, as Miliolina bosciana
does to Miliolina oblonga (Montagu). Its distribution is probably co-extensive with that
of Miliammina oblonga, as it was found in the Terra Nova material and at Kerguelen.

M. obliqua is practically co-extensive in its distribution in the South Georgia area with
M. oblonga, but it is nowhere quite so common, and there are stations at which one or
other species appears exclusively, though as a rule the best specimens of each are found
at the same stations, notably St. 126, WS 28 and 154.

151. Miliammina lata, Heron-Allen and Earland (Plate HI, fig. 17 and Plate V, figs.
15-19)-
Miliammina lata, Heron-Allen and Earland, 1929, etc., FSA, 1930, p. 43, pi. i, figs. 13-17.

Nineteen stations : 13, 14, 16, 31, 126, 131, 136, 145; WS 28, 32, 33, 37, 40, 42, 50, 51, 52, 348;
MS 68.

Test quinqueloculine, but frequently with very little exposure of the earlier chambers,
so little, in fact, that to a casual inspection the test appears triloculine. Chambers
inflated, broadening at the aboral end, broadly rounded at the periphery. Sutural lines
flush or only slightly depressed. Aperture rather small, crescentiform, with slightly
reverted collar, situated on the end of the terminal chamber (never on a produced neck),
furnished with a simple tooth. Wall rather thick, smooth, rarely polished, containing

94 DISCOVERY REPORTS

fine mineral particles in an excess of siliceous cement. Colour light grey. Size varying
between o-30-o-45 mm. in length, o-20-o-32 mm. in breadth, and o-i3-o-22mm. in
thickness.

This is a very distinctive species and may be considered isomorphous with Miliolina
subrotunda (Montagu).

M. lata is much rarer in the South Georgian material than either M. oblonga or M.
obliqua, both as regards the number of stations at which it occurs and its relative
abundance. It nowhere forms a dominant aspect in the material as the forementioned
species frequently do, and generally speaking is rare. But it occurs frequently at WS 32
and 37, and very fine and typical specimens were found at other stations, notably St.
136 and WS 42 and 348. The deepest record is at St. 16 in 727 m., where only a small
specimen was obtained, the species evidently favouring shallower waters than its
relatives. At WS 32 a very extraordinary abnormality was found (Plate III, fig. 17) in
which a specimen of M. lata was fused with one of M. oblonga, the latter projecting
vertically to a height of 0-2 mm. from the centre of the lateral side of the former.

Miliammina lata occurs at seven of the ten Terra Nova stations from which we
recorded Miliolina oblofiga var. arenacea, so it may be presumed that the species is
universally distributed in the Antarctic, like M. oblonga and M. obliqua.

We figure an abnormal specimen from WS 50, at a depth of 230 m., characterized by
the prolongation of the final chamber into an enveloping curve terminated by a very
large aperture (Plate V, fig. 18). Such abnormalities are of frequent occurrence in
gatherings of its porcellanous isomorph, Miliolina subrotunda (Montagu).

Family TEXTULARIIDAE

Sub-family TEXTULARIINAE

Genus Spiroplectammina, Cushman, 1910

152. Spiroplectammina biformis (Parker and Jones) (F 115).

Twelve stations: 28, 144, 149; WS 32, 33, 41, 42, 46, 348; Drygalski Fjord; MS 14, 68.

Rare or very rare everywhere, but very good specimens at Sts. 144, 149, WS 32 and
41. An exceptionally large specimen at WS 32. The specimens everjrwhere are of the
normal type, the very thin form recorded from the Falklands does not occur.

Genus Textularia, Defrance, 1824

153. Textularia agglutinans, d'Orbigny (F 116).
One station: WS 353.

A single very small but otherwise typical example from a depth of 4041 m.

154. Textularia abbreviata, d'Orbigny (F 118).
One station: WS 521.

TEXTULARIIDAE 95

A single small specimen from 3780 m. is attributed with some hesitation to this
species. It is far from typical, the edges being rounded and the median line depressed.

155. Textularia wiesneri, sp.n. (Plate III, figs. 18-20).

Textularia parvula, Wiesner {non Cushman), 1931, FDSE, p. 98, pi. xiii, fig. 152.

Five stations: 144, 145, 151 ; WS 33; MS 14.

Test minute, elongate, often twisted, consisting of between eight and fourteen pairs of
rather inflated chambers. Sutures flush at commencement, becoming depressed. Edge
lobulate, especially at the oral end. Built of fine sand with much cement. Aperture a
normal transverse slit. Colour brown. Length up to 0-4 mm., breadth 0-12 mm.,
thickness 0-07 mm.

Wiesner {lit supra) figures a small Textularia, which does not agree with Cushman 's
description or figures of T. parvula, but represents admirably a species which occurs
both in South Georgia and in the Antarctic material, and is therefore probably widely
distributed over the whole Antarctic area, as Wiesner's records are from the opposite
side of the Antarctic Continent. He records the discovery of both megalospheric and
microspheric specimens.

Cushman figures both stages of T. parvula. His figure purporting to represent the
microspheric form is very similar to Wiesner's figure and our own megalospheric form,
but his figure of the megalospheric form does not resemble either Wiesner's specimens
or our own. Cushman's species is from the Caribbean Sea.

T. wtestieri is rare, or very rare, in the South Georgian material, most numerous at
WS 33. It appears to be more frequent at some Antarctic stations.

It is not very readily distinguishable, at least in the megalospheric form, from T.
tenuissima, as the rounded megalosphere bears some resemblance to the primary
spiroplectine coil or planispire of that species. The brown colour, however, is in strong
contrast to the silvery grey of T. tenuissima, and of course, when viewed by transmitted
light, it is readily distinguishable, owing to its large proloculum and the invariable
absence of a primary planispire.

156. Textularia tenuissima, nom.n. (Plate III, figs. 21-30).

Textularia elegans, Lacroix, 1932, TPCM, p. 8, figs. 4 and 6 (not fig. 5).

Twenty-five stations: 15, 30, 42, 45, 144, 149, 660; WS 28, 32, 33, 37, 40, 41, 42, 47, 63-4, 154,
334. 343. 348, 349. 353. 429; Drygalski Fjord; MS 68.

Test minute, very elongate, straight or slightly curved, oval in section, early chambers
closely coiled in both megalospheric and microspheric forms, the initial end being
rounded in the former and more or less pointed in the latter ; tapering very gradually to
the oral extremity, which is the thickest portion of the test, and rounded ; edge straight
for the first half of the shell, then becoming slightly lobulate, rounded throughout;
chambers very numerous, up to twelve or more pairs following the initial spiral,
distinct, regularly increasing in size and thickness, finally becoming slightly inflated;
sutures distinct, depressed ; aperture distinct, a curved slit on the inner edge of the

96 DISCOVERY REPORTS

terminal chamber; wall thin and smoothly finished, built of very minute mineral grains
and other particles, attached in a single layer to the chitinous membrane and with little
sign of interstitial cement; colour light grey, often silvery, very rarely tinged with
ferruginous cement.

Megalospheric : length up to 0-4 mm., breadth up to o-i mm.

Microspheric : length up to 0-5 mm., breadth up to o-i mm., thickness o-o6 mm.

This very delicate and distinctive little species is widely distributed in the shallower
waters of South Georgia between depths of 130 and 318 m., and also in many Antarctic
gatherings. A few doubtful fragments and a single small specimen were seen in deeper
water at WS 343 in 2856 m.

It is common at St. 144 and WS 32, frequent at St. 45, rare, or very rare, elsewhere,
though owing to its small size and fragility it may perhaps be more abundant than it
seems at some stations. It varies considerably in size and development, the dimensions
given above being the maxima for South Georgia.

The species is apparently trimorphic, two very distinct megalospheric forms are to be
found. One (A i) is much shorter and has fewer chambers than the other, but the
initial spiral is large and neatly rounded. The other megalospheric form (A 2) has the
apex almost as acutely pointed as in the microspheric form, and is equally long.

The microspheric form appears to be acutely pointed. When examined in fluid some
specimens are seen to possess a complete initial spiral coil or planispire, while others
have apparently only a single apical chamber, the outermost chambers of the spiral
being worn away. A similar feature in T. sogittula, Defrance, has been observed by
Lacroix (L. 1929, TS, pp. 1-12, figs, i-io) and confirmed by us (H.-A. and E. 1930,
FPD, p. 72).

On this analogy, I propose to refer my specimens to Textularia rather than to
Spiroplectammina, in which the initial planispire is distinct in both forms.

After the foregoing description of what appeared to be a new species had been written
and was awaiting the press, Dr E. Lacroix of Lyons published a paper on the Textu-
lariidae of the Mediterranean Continental Shelf. My attention was at once arrested by
one of his new species, Textularia elegam (L. 1932, TPCM, p. 8, figs. 4-6), as his
figure of the microspheric form was in general agreement with my specimens, except as
regards size, the Mediterranean specimens attaining little more than half the maximum
dimensions of those from South Georgia.

I sent specimens from the material to Dr Lacroix, who has no doubt as to their
identity with his species. He has been so good as to supply me with a Mediterranean
specimen for comparison.

Apart from the difference in size, there appears to be little doubt as to the identity of
the organisms; the formation and number of chambers are the same, although the
texture of the shell is somewhat different. The Mediterranean specimens have a more
compact arrangement of the mineral particles, ferruginous cement uniting the grains
being visible, when the specimen is examined as a transparent object under a high
power. In the South Georgian specimens, there is no visible cement and the minute

TEXTULARIIDAE

97

mineral grains appear to be attached directly to the chitinous membrane, often with
gaps between the grains.

These, however, appear to be but trifling differences when compared with the
numerous points of agreement, and I should have accepted Lacroix's specific name
elegans for our South Georgian specimens, but for the fact that it has been anticipated by
Hantken for an entirely different but typical Textidaria, Plecanium elegans (H. 1868,
KTF, p. 83, pi. i, fig. 5). This necessitates the abandonment of Lacroix's specific name,
and I propose the new name temdssima, which I had already adopted for the South
Georgian specimens, as a substitute.

Textidaria temdssima

■

Proloculum

Pairs of chambers

Length

Breadth

Observations

St. 149
(South Georgia)

0-008?
0-008

II

10 + I final

0-184

0-212

0-052
0-060

A typical example of my T.
cleaans. Form B
Form B

St. 30
(South Georgia)

0-012
o-oo8
o-oi6

0-012

8

II + I final

8

II

0-248
0-284
0-328

0-392

o-o68
0-060
0-080

0-072

Form A 2 — with planospire
Form B — with planospire
Form A 2 — with planospire.

Proloculum difficult to

measure.
Form A 2 — with planospire

WS 494 A
(Antarctic)

Length "j

0-020
Breadth

0-0185

0-020

0-024

0-020

9+1 final

10 + I final

10 + I final

8 + I final

0-280

0-284
0-292
0-200

0-076

0-080
o-o8o
o-o68.

Form A i — without plano-
spire

St. 144
(South Georgia)

0-024

5+1 final

0-320

—

Five chambers in planospire.
From the size of the prolo-
culum probably Form A i

WS4I
(South Georgia)

0-020?

7+1 final

0-340

Five chambers in planospire.
The proloculum is wrinkled,
almost quadrangular. Its
dimensions vary between
0016 and o-oi8 mm. and if
smoothed out would per-
haps measure 0-020 mm.
Regarded doubtfully as
Form A 2

Lacroix {iit supra, fig. 5) figures what is described as the megalospheric form of his
species. It is extremely rare, only two specimens having been found. They have no
initial spiral and the angle of the sutures is not the same as in his figures of the micro-
spheric form (figs. 4, 6). I regard this figure as a distinctive and genuine Textidaria,
possibly a new species, but, in any case, unconnected with his microspheric specimens.

98 DISCOVERY REPORTS

In Other words, it would appear that the megalospheric form is still undiscovered in the
Mediterranean.

In South Georgia, on the other hand, the megalospheric form, in one or other of its
two conditions, occurs in some abundance, though, contrary to the usual rule, it is less
common than the microspheric form.

Dr Lacroix has been so good as to measure the proloculum of some of my specimens
and has furnished me with the table on p. 97. (In counting the pairs of chambers those
round the proloculum forming the planispire are disregarded.)

Textidaria tenmssima is no doubt a very widely distributed form, and its discovery
may be expected at many localities intermediate between Lacroix's stations off Monaco
and our own South Georgian localities. I am now able to record its existence in the
Falkland Islands area at WS 93 off West Falkland Island from a depth of 133 m.
(TS 503/22), a record omitted from the Falkland Report owing to paucity of material.
It also occurs at many Antarctic stations, at some of which the specimens are much
larger and even more deserving of the name tenmssima than those from South Georgia.

157. Textularia nitens, sp.n. (Plate III, figs. 31-5).

Fifteen stations: 131, 144, 149; WS 33, 41, 42, 47, 63, 334, 343, 348, 373, 428, 429, 523.

Test very minute and fragile, compressed and leaf-shaped, consisting of five to seven
pairs of chambers regularly increasing in size and breadth ; early chambers compressed,
later becoming slightly inflated ; marginal edge rounded, slightly lobulate. Aperture a
terminal slit parallel to face of test. Wall thin, constructed of minute sand grains, pale
gold colour, glistening. Length up to 0-3 mm., breadth 0-13 mm., thickness 0-05 mm.

This delicate little species, though widely distributed in South Georgia, is very rare
everywhere. Its range in depth is considerable (130-3705 m.), but it is most at home
and usually attains its best proportions at the shallower stations. Single specimens of
this species were noted at a depth of 161 m. at WS 210, to north of Falkland Islands,
at WS 433, and at a depth of 1035 m. between the Falkland Islands and South Georgia.
These stations were dealt with in the Falkland Report. The species occurs also at
several Discovery stations in the Antarctic.

Genus Bigenerina, d'Orbigny, 1826

158. Bigenerina minutissima, sp.n. (Plate III, figs. 36-8).
Two stations: WS 199, 472.

Test very minute, rod-shaped, consisting of a large proloculum followed by three to
four pairs of long narrow chambers increasing rapidly in size, but very little in width,
ending with three cylindrical moniliform chambers and a terminal orifice. Sutures
depressed. Constructed of small sand grains, rather large for the size of the organism,
embedded in cement on a chitinous membrane. Colour pale brown. Length 0-35 mm.,
width 0-04 mm. Only a single specimen at WS 199 and a few at 472. The organism is so
small that it might easily have been overlooked at other stations. Its structure is not
easily seen, unless the specimens are mounted in balsam.

TEXTULARIIDAE 99

Genus Verneuilina, d'Orbigny, 1840

159. Verneuilina advena, Cushman (F 121) (Plate III, figs. 43-6).
Ten stations: 144, 149, 151, 660; WS 28, 32, 42, 47, 63-4, 373.

Common at St. 144, very rare elsewhere. The specimens are rather distinctive, being
nearly always constructed of very minute micaceous particles embedded in the chitinous
wall of the test. They are therefore more delicate and thin-walled than the finely
arenaceous specimens so familiar in British gatherings, and the chambers appear to be
more inflated owing to the thinness of the wall. But dimensions and number of
chambers agree, and there appears to be no reason for separating the South Georgian
specimens on account of their construction. Both long and short forms occur, as is
usual in British gatherings also.

160. Verneuilina bradyi, Cushman.

Verneuilina pygmaca, Brady {non Egger), 1884, FC, p. 385, pi. xlvii, figs. 4-7.

Verneuilina pygmaea, Flint, 1899, RFA, p. 285, pi. xxxi, fig. i.

Verneuilina bradyi, Cushman, 1910, etc., FNP, 191 1, p. 54, fig. 87; 1918, etc., FAO, 1922, p. 59.

pi. xi, fig. I.

Eight stations: 53° 00' S, 34° 22' W; WS 334, 336, 353, 429, 521, 522, 523.
Confined to deep water at depths between 1697 and 4041 m. It is frequent at WS 429,
522 and 523, very rare elsewhere.

Genus Gaudryina, d'Orbigny, 1839

161. Gaudryina bradyi, Cushman.

Gaudryina pupoides, Brady {non d'Orbigny), 1884, FC, p. 378, pi. xlvi, figs. 1-4.
Gaudryina pupoides, Flint, 1899, RFA, p. 287, pi. xxxii, fig. 4.

Gaudryina bradyi, Cushman, 1910, etc., FNP, 191 1, p. 67, fig. 107; 1918, etc., FAO, 1922,
p. 74, pi. xii, fig. 8.

One station: WS 523.

Frequent well-grown specimens at a depth of 1697 m.

162. Gaudryina flintii, Cushman.

Gaudryina suhrotundata, Flint {non Schwager), 1899, RFA, p. 287, pi. xxxiii, fig. i.
Gaudryina flintii, Cushman, 1910, etc., FNP, 191 1, p. 63, fig. 102; 1918, etc., FAO, 1922, p. 69,
pi. xii, figs. I, 2.

One station : WS 522.

A single large specimen at a depth of 2550 m.

163. Gaudryina baccata, Schwager.

Gaudryina baccata, Schwager, 1866, FKN, p. 200, pi. iv, figs. 12 a, b.
Gaudryina baccata, Brady, 1884, FC, p. 379, pi. xlvi, figs. 8-1 1.
Gaudryina baccata, Flint, 1899, RFA, p. 287, pi. xxxii, fig. 5.
Gaudryina baccata, Cushman, 1910, etc., FNP, 191 1, p. 68, fig. 108.

One station: WS 521.

A single specimen.

loo DISCOVERY REPORTS

164. Gaudryina apicularis, Cushman.

Gaudryina siphoncUa, Brady {non Reuss), 1884, FC, p. 382, pi. xlvi, figs. 17-19.

Gaudryina siplionella, Flint, 1899, RFA, p. 288, pi. xxxiv, fig. 2.

Gaudryina apicularis, Cushman, 1910, etc., FNP, 1911, p. 69, fig. no; 1918, etc., FAO, 1922,

p. 72, pi. viii, fig. 4.

One station: WS 353.

One perfect specimen and a broken one at this station at a depth of 4041 m.

Genus Clavulina, d'Orbigny, 1826

165. Clavulina communis, d'Orbigny (Plate III, figs. 39-42).

Clavulina communis, d'Orbigny, 1826, TMC, p. 268, no. 4; 1846, FFV, p. 196, pi. xii, figs, i, 2.

Clavulina communis, Brady, 1884, FC, p. 394, pi. xlviii, figs. 1-13.

Clavulina communis, Cushman, 1918, etc., FAO, 1922, p. 84, pi. xvi, figs. 4, 5.

Nine stations: 53° 00' S, 34° 22' W; WS 199, 334, 336, 353, 365, 429, 521, 522.

Confined to the deep-water stations, the shallowest depth from which the species was
recorded being 2549 m. at WS429. Here the specimens were few and small, except for
a fragment which must have belonged to a specimen far larger than any perfect one
found. At WS 353 a large fragmentary specimen was obtained at a depth of 4041 m.
The best specimens were from WS 199 and from a sounding of Diatom ooze taken in
53° 00' S, 34° 22' W (2472 m.). All the specimens except one were uniform in ap-
pearance, very smooth and neatly finished with fine grey mud, no coarse material being
used. The exception occurred at WS 522, where of two specimens found, one was
normal, the other very large and constructed of nearly black sand and cement. Young
individuals, which have not grown beyond the triserial stage, occur at most stations.
At WS 336 there were many of these, but no adult shells. Megalospheric specimens
predominate everywhere.

Sub-family BULIMININAE

Genus Bulimina, d'Orbigny, 1826

166. Bulimina fusiformis, Williamson (F 124).
Four stations: 15, 129, 151 ; WS 18.

Common at St. 15, very rare elsewhere.

167. Bulimina elegans, d'Orbigny (F 127).
Five stations: 16, 136, 157; WS 37, 426.

Rare everywhere but well developed. Best and most numerous at St. 157 and WS 37.

168. Bulimina marginata, d'Orbigny (F 129).
Five stations: 20, 29, 149; WS 41, 47.

Only a single specimen at each station except at St. 149, where several were obtained.
They are uniformly small and pauperate, except the specimen from St. 29, which is
normal.

TEXTULARIIDAE loi

169. Bulimina patagonica, d'Orbigny (F 130).
Two stations: WS 349, 522.

Very rare and not very typical, the best at WS 522.

170. Bulimina aculeata, d'Orbigny (F 131).

Seven stations: 16, 149; WS 37, 63, 351, 429, 523.

Rare and extremely pauperate, except at WS 523, where a good many excellent
specimens were found.

171. Bulimina subteres, Brady (F 134).
Three stations: 123, 144; WS 33.

Very rare and poorly developed at all stations.

172. Bulimina elegantissima, d'Orbigny (F 135) (Plate III, fig. 47).

One station : WS 27.

Represented in the South Georgian material by a single very fine ' ' budding "specimen.
In spite of the fact that WS 27 is off" the west end of South Georgia, it is possible that
this occurrence of only a single specimen of a typical Falkland species may be due
to a foul net or sieve, rather than to an extension of habitat.

173. Bulimina buchiana, d'Orbigny.

Bulimina buchiana, d'Orbigny, 1846, FFV, p. 186, pi. xi, figs. 15-18.
Bulimina buchiana, Brady, 1884, FC, p. 407, pi. li, figs. 18, 19.
Bulimina buchiana, Cushman, 1918, etc., FAO, 1922, p. 95, pi. xx, fig. 4.

Four stations: WS 63, 429, 522; MS 14.

Rare or very rare. All the specimens are rather smaller than the average but other-
wise typical; the largest and best were at WS 522.

Genus Virgulina, d'Orbigny, 1826

174. Virgulina schreibersiana, Czjzek (F 138).

Forty-five stations: 13, 14, 15,20,23,28,30,31,42,45, 123, 129, 131, 136, 144, 145, 149, 151, 157,
660; WS 25, 27, 28, 32, 33, 37, 41, 42, 43, 45, 47, 50, 63-4, 113, 154, 348, 349, 351, 357, 418, 429,

522, 523; MS 14, 68.

Almost universally distributed and one of the most abundant species round South
Georgia. It is common at most stations in the above list and very common at several,
while at St. 144 and WS 63-4 it is a dominant form. By contrast there are stations
(Sts. 15, 28, WS 28, 523) at which the species is extremely rare, and others (as at St. 30)
where, though frequent, the specimens are pauperate and small. The best specimens
were noted at MS 68, WS 25, 32, 33, 349 and 429. The variation referred to in the
Falkland Report, viz. long and short forms usually occurring together, was noticed at
most stations.

102 DISCOVERY REPORTS

175. Virgulina subsquamosa, Egger (F 140).

Twelve stations: 20, 23, 28, 30, 136, 149; WS 25, 33, 40, 48, 349, 418.

Very rare and generally small and pauperate. The best specimens at WS 33 and 48.

176. Virgulina bradyi, Cushman (F 141).

Twenty-two stations: 15, 30, 45, 126, 136, 140, 144, 149, 660; WS 27, 32, 33, 40, 42, 48, 349, 357,
373.429. 522, 523; MS 14.

Common at Sts. 144, 149, 660 and WS 42, elsewhere usually rare. There is consider-
able variety in the length as compared with the breadth of specimens at different
stations. The best specimens were noted at Sts. 136, 144, 149, WS 32, 33, 373 and 523.

Genus Bolivina, d'Orbigny, 1839

177. Bolivina punctata, d'Orbigny (F 143).

Eighteen stations: 15, 16, 30, 43, 136, 143, 144, 149; WS 27, 32, 33, 47, 63-4, 113, 348, 429, 522;
MS 14

Very rare and mostly represented by one or two specimens at each station except at
St. 149, where it is very common. Some of the single specimens are very fine and typical ;
the majority are megalospheric.

178. Bolivina textilarioides, Reuss (F 144).
Two stations: 15; WS 33.

Very rare and pauperate.

179. Bolivina robusta, Brady (F 146).
One station: 149.

A few small and weak specimens only.

180. Bolivina dilatata, Reuss (F 148).
Two stations: 149; WS 47.
Extremely rare but very typical.

181. Bolivina difformis (Williamson) (F 149) (Plate III, figs. 50, 51).

Textiilaria variabilis var. difformis, Williamson, 1858, RFGB, p. 77, pi. vi, figs. 166, 167.

Bolivina dijformis, Heron-Allen and Earland, 1913, CI, p. 65.

Bolivina dijformis, Cushman, 1918, etc., FAO, 1922, p. 32, pi. iv, fig. i.

One station: MS 14.

A single specimen, small, but typical.

182. Bolivina malovensis, Heron- Allen and Earland (F 153).
One station : WS 27.

This species, so abundant in the Falkland area, is represented by a few very weak
individuals only, recorded with some hesitation.

TEXTULARIIDAE 103

183. Bolivina cincta, Heron- Allen and Earland (F 154).
Three stations: WS 351, 521, 522.

Rare, but typical and well developed.

Typical specimens of B. cincta occur frequently in the Challenger material from
Challenger St. 300, to the north of Juan Fernandez Island in the Pacific (depth 1375
fathoms), so it is evident that the species has a wide distribution. It has some super-
ficial resemblance to Bolivina caelata, Cushman (C. 1925, etc., LFR, v, p. 93, pi. xiii,
fig. 28), but may be distinguished by its edge, which is broad and flat, whereas the
edge of B. caelata is described as acute. No edge view of that species has been figured
so far as I know.

184. Bolivina decussata, Brady (Plate III, figs. 48, 49).

Bolivina decussata, Brady, 1879, etc., RRC, 1881, p. 58; 18S4, FC, p. 423, pi. liii, figs. 12, 13.

Bolivina decussata, J. Wright, 1891, SWI, p. 475.

Bolivina decussata, Cushman, 1910, etc., FNP, 191 1, p. 47, fig. 77; 1918, etc., FAO, 1922, p. 32.

Two stations: WS 521, 522.

A single specimen at WS 521 and two at the other station, all quite typical. The
records are of great interest as, with a single exception, the species has hitherto only been
noted in the Pacific, viz. at Challenger Sts. 300 and 302 (1375 and 1450 fathoms re-
spectively) both near Juan Fernandez Island, and at the Albatross St. 4839 ofl^ Japan
(140 fathoms). The depth in the case of the Albatross specimens is remarkable as com-
pared with the Challenger records and our own, which are from 3780 and 2550 m.

The only record outside the Pacific, so far , is of material from the south-west of Ireland,
(50° 52' N, 1 1 " 27' W) in which the species was noted by Joseph Wright to be " common
at 1020 fathoms ". In the absence of a figure it cannot be stated definitely what Wright's
form was, and Cushman (1922, ut supra) states that he has found specimens in Wright's
material which are evidently those referred to, but differing from the Pacific material
he has seen. It is not clear whether he is referring to the Challenger types or to the
material from Japan.

I have not found anything resembling B. decussata in Irish material, though I have
dredgings from approximately the same locality as Wright.

In view of the present extension of the range of the species into the South Atlantic,
there seems no inherent impossibility in Wright's record. It may be noted that the
species shows a wide range of variation in the Challenger material.

Sub-family CASSIDULININAE
Genus Cassidulina, d'Orbigny, 1826

185. Cassidulina laevigata, d'Orbigny (F 157).
Three stations: WS 428, 521, 523.

Singularly rare, except at WS 428, where several specimens were observed.

I04 DISCOVERY REPORTS

1 86. Cassidulina laevigata var. tumida, Heron- Allen and Earland.

Cassidulina laevigata var. tumida, Heron-Allen and Earland, 1922, TN, p. 137, pi. v, figs. 8-10.
Cassidulina laevigata var. tumida, Cushman, 1925, LFR, i, p. 54, pi. viii, figs. 40-2.

One station : WS 66.

Only a single and far from typical specimen. The variety was originally described from
off Three Kings Islands, New Zealand (North Island) at a depth of 90-300 fathoms, but
we have good specimens from a sounding made by the ' William Scoresby ' off Gough
Island (2000-3000 m.).

187. Cassidulina pulchella, d'Orbigny (F 158).
Two stations: WS 27, 521.

Extremely rare and very pauperate. Only a few specimens in all.

188. Cassidulina crassa, d'Orbigny (F 160).

Thirty-four stations: 20, 27, 30, 42, 45, 123, 126, 131, 136, 140, 143, 144, 148, 149, 660; WS 25,
27, 28, 31, 32, 33, 40, 42, 66, 113, 154, 314, 351, 418, 429, 521, 522, 523; MS 68.

The species, though widely distributed, does not occupy the dominant position
which it assumes in the Falkland area and is very rare at many stations. All the forms
met with in the Falkland area were found, but the large typical crossa is rare, and occurs
only at Sts. 144, 148, 149, WS 27, 154 and 521, nearly always accompanied by the small
hyaline form, which is the dominant type round South Georgia and is sometimes very
common, especially at St. 144, WS 25, 27, 66 and 314. The intermediate type referred to
in the Falkland Report also occurs at several stations, notably Sts. 123, 144 and WS 33.
With increase of depth there is a marked tendency to decrease in size. Thus at WS 429
at a depth of 2549 m. the small type becomes quite minute, whereas at WS 522 and
523 at depths of 2550 and 1649 m. respectively the intermediate type is much reduced in
size while acquiring a much thicker shell.

189. Cassidulina subglobosa, Brady (F 162).

Thirty-eight stations: 13, 20, 30, 31, 42, 123, 126, 131, 136, 140, 144, 149, 660; WS 25, 27, 28,
32,33,37,40,42,43,45,47, 50, 51,63-4,66,113,154,314,348,351,418,521, 522, 523; MS 68.

Very generally distributed but usually rare or very rare. It was, however, very common
at St. 144, WS 33 and 66. As a rule the specimens are small and far from typical, the
flattened form referred to in the Falkland Report occurring at many stations, sometimes
exclusively. The best specimens of the typical form were seen at St. 149, WS 33, 522
and 523.

190. Cassidulina parkeriana, Brady (F 163).

Forty-five stations: 13, 14, 16, 20, 23, 27, 30, 31, 42, 45, 123, 126, 131, 136, 140, 144, 148, 149,
157; WS 25, 27, 28, 32, 33, 37, 40, 42, 43, 45, 46, 47, 48, 50, 52, 63, 66, 113, 154, 314, 348, 349, 357,
418, 426; MS 68.

This is the most generally distributed species of Cassidulina, but is seldom abundant
except at a few stations, notably St. 23 where it is very common. It is common at Sts.

TEXTULARIIDAE 105

16, 20, 136, WS 27, 33, 48 and 418, elsewhere generally rare to very rare. Both megalo-
spheric and microspheric specimens are found, wherever it is present in any numbers.
Nowhere does it attain such size as in the Falkland area ; the best specimens were at
WS 33 and 42.

Genus Ehrenbergina, Reuss, 1850

191. Ehrenbergina pupa (d'Orbigny) (F 164).
Five stations: 30, 149; WS 25, 27, 33.

The extreme rarity of this species round South Georgia as compared with its abun-
dance in the Falkland area is most marked, and points to the fact that it has not succeeded
in establishing its footing in the colder water. It is moderately frequent at WS 27
which is off the north-west corner of the island and therefore on its line of approach
from the Falkland area. At the remaining stations the species is represented by one or
two specimens only.

192. Ehrenbergina hystrix var. glabra, Heron-Allen and Earland (F 165).
Five stations: WS 33, 51, 66, 314, 418.

This Antarctic form is common at WS 51, but the specimens are rather small though
typical. At all the other stations it is very rare, but the few specimens found were
generally large, though often weakly developed as regards the marginal spines.

193. Ehrenbergina crassa, Heron-Allen and Earland (Plate VI, figs. 1-9).

Ehrenbergina crassa, Heron-Allen and Earland, 1929, etc., FSA, 1929, p. 329, pi. iii, figs. 18-26.

Fifty stations: 13, 14, 20, 23, 27, 28, 30, 31, 41, 42, 45, 123, 126, 131, 136, 140, 143, 144, 148, 149,
157, 660; WS 20, 25, 27,28,32, 33,37,40,41,42,43,45,46,47,48, 50, 51, 52, 66, 113,154, 314,
348. 349. 357, 418; MS 14, 68.

Test very thick-walled throughout, finely perforate, hyaline in the younger stages, but
frequently becoming white and semi-opaque in the adult shell. Constructed of a
variable number of chambers regularly increasing in size, arranged biserially about an
elongate axis and presenting well-marked dorsal and ventral sides. The main axis of the
shell is normally straight, but occasionally exhibits a spiral tendency, thus giving a
virguline appearance to such tests. The dorsal side of the test is flatter and wider than
the ventral, and the sutural fines, which are depressed, are more prominent on the
dorsal side owing to their greater thickness.

The sutures are most noticeable in the oral half of the shell, those of the initial half
being usually more or less obscured by a secondary layer of shell substance which is
sometimes granular or even feebly striated. The oral half of the shell is quite smooth
and devoid of ornament. The aperture is a loop-like sfit set obliquely on the inner face of
the final chamber.

The initial portion of the shell consists of a more or less prominent knob, curved
over towards the ventral side and excentric to the main axis of the shell. In the megalo-
spheric form, this knob is very prominent, and contains the proloculum situated on its
ventral face, behind which is the first pair of chambers. In the microspheric form, it

io6 DISCOVERY REPORTS

contains the proloculum followed by one piano-spiral convolution of minute biserial
chambers regularly increasing in size, the axis of the spiral being at right angles to the
main axis of the shell.

The megalospheric form predominates everywhere, sometimes to the entire exclusion
of the microspheric. Its appearance is very distinctive, especially in young specimens,
where the proloculum shows like a glassy bubble on the ventral side of the bulbous top.
As the shell develops, it becomes less noticeable owing to the deposition of secondary
shell matter. The pair of chambers immediately following the proloculum are com-
pressed in shape, owing to their position on the dorsal side behind the proloculum.
After them the chambers are regularly arranged biserially. Average specimens exhibit
four to five pairs of chambers, but several specimens having as many as seven to eight
pairs have been seen.

The microspheric form is easily distinguished by its narrower initial end, which is
rather wedge-shaped than bulbous. Owing to the thickness of the wall, the internal
structure is difficult to determine, but, as previously stated, the very small proloculum is
followed by a piano-spiral coil of minute biserial chambers (probably four to five
pairs), regularly increasing in size, and followed by a further series of six to eight
pairs of straight chambers. The microspheric form thus attains a greater average
length than the megalospheric, the other dimensions remaining much the same.

An average megalospheric specimen measures about 070 mm. in length by 0-30 mm.
in greatest breadth and 0-25 mm. in thickness. One very large specimen attained
I -08 mm. in length. The megalosphere as measured in optical section (internal)
averages o-i mm. in diameter.

Microspheric specimens average about 0-90 mm. in length by 0-35 mm. in greatest
width and 0-30 mm. in thickness. A large specimen attained 1-35 mm. in length. The
microsphere could not be measured with any certainty, owing to the thickness of the
shell. It is certainly very minute.

E. crassa is a very abnormal type, and until its relationships were definitely established
by the discovery of the microspheric form with its coiled initial portion, its systematic
position remained uncertain. Its nearest ally is unquestionably E. pupa (d'Orbigny),
from which it differs in many points, notably in the marked development of the produced
series of chambers and their regular Bolivine arrangement. E. pupa was first described
by d'Orbigny from the Falkland Islands, and it is one of the most abundant and
variable species in the Discovery collections from that area. None of the variations,
however, approaches E. crassa, which does not occur at all in the Falkland area, but
is relatively common in South Georgia and adjacent waters, and, so far as our present
knowledge goes, is almost confined to that area, where it constitutes one of the most
characteristic local species. Presumably the two species, E. pupa and E. crassa, are de-
rivatives from a common ancestor which inhabited both areas.

I have little to add to the information already published and quoted above. E.
crassa is certainly the most typical if not the commonest species round South Georgia,
occurring in greater or less abundance at more than half of all the stations examined.

CHILOSTOMELLIDAE 107

The depths range between 18 and 346 m., the latter depth being at WS 28, where the
species is frequent and well developed. This range has only been extended by the
finding of a single fragmentary specimen at St. 157 in 970 m., to which we should not
attach any importance as its presence may be accidental. The species is most at home in
shallower water, the best stations being Sts. 30, 45, 123, 126, 149, 660, WS 25, 42 and
154, with an average depth of about 190 m. Megalospheric and microspheric forms
were found together at twenty-five stations, including the nine just mentioned. At the
other stations megalospheric individuals only were observed. There appears to be great
range in the size of the megalospheric proloculum; exceptionally large primordial
chambers were seen in specimens obtained from St. 27 and WS 66.

The texture of the shell is normally smooth and glossy, becoming smooth but dull
with age. After death the secondary deposit of shell substance appears to disintegrate
rather quickly, leaving a roughened surface. This is quite distinct from a superficial
roughening due to the formation of minute beads or spines or striae, a variation which
appears to be rather rare, having been observed only at MS 68, WS 25, 27, 33, 37, 66
and 349. At MS 68 nearly all the specimens were coated with a thin layer of fine mud,
so that the nature could only be guessed at from the characteristic shape of the mud
lumps. The coating was readily dislodged and its nature is obscure ; it appears to be too
thin to be regarded as encystment.

194. Ehrenbergina bradyi, Cushman (F 166).
Two stations: WS 521, 522.

A few small individuals at WS 521 and 522. Both the stations are in deep water.

Family CHILOSTOMELLIDAE
Genus Seabrookia, Brady, 1890

195. Seabrookia earlandi, J. Wright (F 168).
One station : WS 429.

Rare, but a good many specimens were found at this station at a depth of 2549 m.

Family LAGENIDAE

Sub-family LAGENINAE

Genus Lagena, Walker and Boys, 1784

Owing to the number of species recorded, they have been placed in alphabetical
order for facility of reference.

196. Lagena acuta (Reuss) (F 211).

Eighteen stations: 20, 30, 126, 144, 149; WS 25, 28, 33, 42, 46, 48, 63-4, 66, 113, 154, 349, 521 ;
MS 14.

io8 DISCOVERY REPORTS

Widely distributed, but less frequent than L. biancae, into which it merges imper-
ceptibly, presenting similar variations in the degree of compression and length of test.
The spinous base is very poorly developed every^vhere, except at WS 521, where it is
prominent. At this station and at WS 154 the species is most favourably represented.
At most stations the specimens are small, few in number and very near L. biancae, the
basal process being more or less rudimentary.

197. Lagena acuticosta, Reuss (F 196) (Plate III, fig. 52).

Six stations: 27, 33, 123; WS 27, 33, 418.

Usually only a single typical specimen at each station. At St. 33 a very fine example
of a double shell was found. The individuals, which are large, are as usual fixed apex to
base.

198. Lagena alveolata, var. substriata, Brady.

Lagena auriciilata var. substriata, Brady, 1879, etc., RRC, 1881, p. 61.

Lagena alveolata var. substriata, Brady, 1884, FC, p. 488, pi. Ix, fig. 34.

Lagena alveolata var. substriata, Cushman, 1910, etc., FNP, 1913, p. 34, pi. xviii, fig. 5.

One station: WS 522.

A single good specimen.

199. Lagena annectens. Burrows and Holland (F 215).
Seven stations: 30, 144, 149; WS 27, 33, 50; MS 14.

Never very common, usually very feeble specimens, the best at St. 149 and WS 33.

200. Lagena apiculata (Reuss) (F 174) (Plate IV, figs. 1-3).

Twenty-three stations: 13, 42, 45, 131, 660; WS 32, 33, 37, 40, 42, 43, 46, 47, 48, 50, 52, 63, 113,
154. 314. 349. 357. 522.

This is one of the most widely distributed species in the South Georgian material, and
though never common, some stations yielded many specimens, which often exhibited
variety of form. The average specimen is of a small regular type, but the aperture is
variable. Usually radial and central, the aperture is sometimes radial but excentric,
sometimes fissurine, and at WS 349 fissurine and hooded.

A very large form compared with the others occurs at many stations. It is character-
ized by a long, regularly tapering shell ranging up to 0-7 mm. in length, broadest towards
the base which is furnished either with a small circular pit or a short projecting tube.
The pit or tube may be pierced so as to form a secondary aperture but is usually closed.
The aperture is prominent and radial. This form is the sole representative of the species
at Sts. 42, 131, 660, WS 43, 46, 48, 50, 52, 154 and 357. At some other stations it is
accompanied by other varieties. At St. 131, WS 42, 46 and 50, a few specimens had
basal striae radiating from the tube. I was at first inclined to assign the specimens to
L. stelligera, but on the whole the evidence points to L. apiculata.

LAGENIDAE 109

201. Lagena auriculata, Brady (F 245).
Two stations: WS 63, 521.

Only a single specimen at each station, quite typical.

202. Lagena biancae (Seguenza) (F 210).

Twenty-two stations: 15, 17, 27, 123, 126, 136, 144, 145, 149; WS 25, 27, 33, 37, 42, 43, 47, 48,
418, 429, 521, 522; MS 68.

Probably the most common species of Lagena round South Georgia, though never so
abundant as in some of the Falkland gatherings, or presenting such a range of variation.
Most abundant and varied at WS 27 and 33, where fine and typical specimens were
found, also the strongly punctate form. Neither of these, however, were so abundant
here or elsewhere as a small rather globose form which is the commonest variety at most
stations and is frequently the sole representative of the species. The type was found at
Sts. 27, 126, 136, 144, WS 27, 33 and 418. At WS 429 and 521, single specimens of a
very elongate fissurine variety were found.

203. Lagena bicarinata (Terquem) (F 236).
Three stations: WS 25, 33, 373.

Extremely rare; a single specimen only at each station, the best being at WS 373.

204. Lagena bisulcata, Heron-Allen and Earland (F 239).

Six stations: 136; WS 25, 27, 33, 357, 418.

Good and typical specimens were not infrequent at the few stations at which the
species was observed, except at WS 27 and 33, where it was very rare. They are all of a
coarse, thick-walled form except at WS 27, where the single specimen was small and
weak. The species was most abundant and most strongly developed at WS 357.

205. Lagena catenulata, Reuss (F 201).

Five stations: WS 27, 33, 42, 521, 522.

Rare. With the exception of very strongly marked specimens at WS 521 and 522,
all the specimens are far from typical, the cross-bars being almost absent over much of
the basal half of the test. They are thus intermediate between L. catemilata and L.
sqiiamoso-sulcata, Heron-Allen and Earland (No. 244).

206. Lagena clathrata, Brady (F 243).

One station : WS 429.

A single specimen from WS 429 is probably referable to Brady's species, though very
unlike the usual types. It is elongate, broadest near the base, which is furnished with a
supplementary tubular aperture. Of the three marginal keels, the mid-keel is less
prominent than the outer two, except in the oral region where it extends beyond them.
The faces of the shell are finely striate.

jio DISCOVERY REPORTS

Sidebottom figures a somewhat similar specimen (S. 1912, etc., LSP, 1913, p. 196
pi. xvii, fig. 14) from the South-west Pacific, under the name L. orbignyana var. clath-
rata. His figure, however, has only three costae on the face, as against many in ours, and
is devoid of the basal aperture.

207. Lagena costata (Williamson) (F 195).
Three stations: WS 27, 429, 521.

Only a few specimens, those at WS 27 being characterized by numerous fine costae,
while at the other stations the costae are few but prominent.

208. Lagena danica, Madsen (F 234).
Two stations: WS 33, 48.

Single small specimens only.

209. Lagena distoma, Parker and Jones (F 186).

Two stations: 144; WS 32.

Very rare, two specimens only at St. 144, and one at WS 32, all small and feebly
striate.

210. Lagena fasciata (Egger) (F 212).
Two stations: WS 25, 32.

A few feeble examples at each station.

211. Lagena fasciata var. faba, Balkwill and Millett (F 213).
One station: WS 522.

A single specimen.

212. Lagena felsinea, Fornasini (Plate IV, figs. 4, 5).

Lagena emaciata var. felsinea, Fornasini, 1901, NNI, p. 47, fig. i.

Lagena felsinea, Cushman, 1910, etc., FNP, 1913, p. 10, pi. iv, fig. i ; 1918, etc., FAO, 1923, p. 17.

Three stations: 144; WS 42; MS 68.

Several specimens at St. 144, single ones only at the other stations.

213. Lagena fimbriata, Brady (F 232).

SevpT^ sUcions: 45, 136, 144; WS 33, 40, 42, 351.

A single typical example only at WS 351, resembling one of Brady's figures (B. 1884,
FC, pi. Ix, fig. 26). Otherwise the species is represented round South Georgia only by a
minute variety in which the basal wings are reduced to a minimum, often represented
only by an opaque white line curving round each side of the base. A few specimens of
this form at each station.

LAGENIDAE iii

214. Lagena formosa, Schwager.

Lageiia formosa, Schwager, 1866, FKN, p. 206, pi. iv, figs. 19 a-d; pi. vii, fig. i.
Lagena formosa, Brady, 1884, FC, p. 480, p. Ix, figs. 10, iS-20.
Lagena formosa, Millett, 1898, etc., FM, 1901, p. 624, pi. xiv, figs. 10-12.
Lagena formosa, Cushman, 1910, etc., FNP, 1913, p. 41, pi- xi, fig. 6.

Three stations: WS 33, 521, 522.

A single very fine specimen at WS 33 and 521, and a small one at WS 522.

215. Lagena formosa, Schwager, var.n. costata (Plate IV, figs. 16-18).
One station: WS 522.

At this station a few specimens of a very distinctive variety were found. The test is
rather small and generally presents all the characteristics of the species, but the faces
of the shell are ornamented centrally with a few straight costae, four to six in number.
These occupy only the centre of the face and do not extend to the raised edge as in the
well-known variety comata, Brady. The wing is rather broad and coarsely tubulated and
the basal cleft very distinct.

There is some variation in the relative dimensions of specimens, two of which were
respectively 0-35 and 0-40 mm. long and 0-22 and 0-20 mm. broad.

216. Lagena foveolata, Reuss (F 204).

Two stations: 126; WS 429.

Extremely rare. The single specimen found at St. 1 26 is very weak, but at WS 429 two
specimens were found which, although very small, are most typical in their markings.

217. Lagena globosa (Montagu) (F 169).

Eight stations: 16; WS 25, 32, 33, 66, 314, 521, 523.

Very rare, seldom more than one or two specimens at each station. The best station is
WS 33, where the species occurs typical and large. A compressed variety also occurs
with the type at this station and at WS 521. At St. 16 the compressed variety only
occurs.

218. Lagena gracilis, Williamson (F 185).

Eight stations: 45; WS 37, 45, 47, 113, 429, 521, 522.

Rare except at WS 37, where a good many typical specimens were obtained. There is
the usual wide range of variation in the strength of the markings, which are very feeble at
some stations. Those from WS 521 are extremely coarse. At WS 522 there is a series
with costae ranging from feeble to very coarse.

219. Lagena gracillima (Seguenza) (F 177).

Fourteen stations: 14, 20, 42, 45, 131, 136, 144; WS 33, 37, 42, 50, 154, 349, 357.

Rare and represented by one or two specimens only at most of the stations. More
abundant at St. 131. All the specimens are well developed. Two forms occur, the typical

113 DISCOVERY REPORTS

Straight form of Seguenza and a curved variety which is well ilkistrated by Brady
(B. 1884, FC, pi. Ivi, fig. 24). The two forms occur together at Sts. 42 and 144, the
curved variety only at St. 45, WS 33, 42, 50 and 154, and only the typical straight form,
which sometimes reaches large dimensions, at all the remaining stations.

220. Lagena hartiana, sp.n. (Plate IV, figs. 12, 13).
One station: WS 522.

Test minute, hyaline, compressed, broadest and thickest at a point about three-
quarters of the length of the shell from the aperture, which is fissurine and situated on a
slightly thickened and produced collar. In edge-view it shows five narrow carinae, the
central and outer carinae projecting beyond the intermediates. The face of the test is
covered with longitudinal costae, five in number, converging towards the extremities of
the test. The facial costae are of about the same strength as the intermediate carinae.
Length 0-25 mm., breadth o-i mm.

This is rather a striking little form without any very definite affinities. It may be
compared with a figure assigned by Sidebottom to L. orbignyana var. dathrata, Brady
(S. 1912, etc. LSP, 1913, p. 196, pi. xvii, fig. 14), but his specimen though agreeing in
general construction has only three marginal carinae and three costae running down each
face. Its assignment to L. dathrata seems rather far-fetched, but it is evidently closely
allied to my species.

This species is named after T. J. Hart, B.Sc, of the Discovery staff.

321. Lagena herdmani, sp.n. (Plate IV, figs. 10, 11).
One station: WS 523.

Test minute, hyaline, compressed, consisting of a flask-shaped body with a produced
neck. The flask has two narrow thickened carinae, whitish in colour owing to the presence
of numerous tubuli. The carinae merge at the oral end of the flask into a tapering un-
tubulated wing, which extends up the sides of the neck. This neck is almost as long as
the flask and is continued as an entosolenian tube into the flask. The faces of the flask
are convex and the carinae are separated at the edges of the test by a broad space
showing no trace of a median keel. Length 0-22 mm., breadth 0-09 mm.

This rather distinctive little form combines the tubulations of L. lagenoides with the
untubulated double carinae of L. bicarinata, and appears to be undescribed, although
there are several records closely resembling it, but diff"ering in essential points (compare
S. 1912, LSP, etc., 1912, p. 414, pi. xix, fig. 9, a description of a variety oi L.formosa).
His figure closely resembles my species except for its apiculate base and median
carina. Sidebottom states, however, that "in a few cases the wing or keel dies away
almost as soon as it reaches the body of the test". This variation would be practically
identical with my form but could no longer be ranked with L. formosa.

Another figure deserving of reference is that of L. lagenoides, Sidebottom (S. ut

LAGENIDAE 113

supra, p. 413, pi. xix, fig. 3), differing in the absence of a produced neck and the broader
and more coarsely tubulated carinae.
This species is named after H. F. P. Herdman, M.Sc, of the Discovery staff.

222. Lagena hexagona (Williamson) (F 202).

Eight stations: 27, 31, 131 ; WS 27, 33, 42, 154, 522.

Never very frequent but good and typical specimens, especially at St. 131 and WS 154.
The single specimen at WS 522 has exceptionally large and coarse markings.

223. Lagena hispidula, Cushman (F 180).
Four stations: 123; WS 33, 429, 522.

Single specimens at St. 123, WS 33 and 522. Many smaller but very typical examples
at WS 429.

224. Lagena laevis (Montagu) (F 179).

Seven stations: 129, 149; WS 37, 50, 66, 429, 521.

Rare, most of the specimens being small and narrow. At WS 66 a single curious
specimen occurred, characterized by a thick glassy shell with a milk-white annulus
encircling the test near the base.

225. Lagena lagenoides (Williamson) (F 226).
Two stations: 144; WS 522.

At St. 144 a single specimen only occurred, agreeing with Williamson's type figures,
and one of a large deep-water type at WS 522.

226. Lagena lagenoides var. tenuistriata, Brady (F 228).
Two stations: WS 429, 522.

A single very fine specimen at each station.

227. Lagena lineata (Williamson) (F 183).

Six stations: 123, 144; WS 25, 33, 429, 522.

Rarely more than a single specimen at a station and all very feebly marked. The usual
form is almost globular with a broad oral end, but Williamson's typical oval form also
occurs more sparingly. Both varieties are found at WS 33, where the species occurs
more frequently than at any other station.

228. Lagena lucida (Williamson) (F 214).
Two stations: 144; WS 33.

Only a few specimens at each station, the best developed at WS 33.

229. Lagena lyellii (Seguenza) (F 190).
Two stations: 20, 149.

A single weakly marked specimen at each station.

114 DISCOVERY REPORTS

230. Lagena mackintoshiana, sp.n. (Plate IV, figs. 14, 15).
Two stations: WS 27, 33.

Test pear-shaped with a produced and regularly tapering neck. From a small basal
ring some 10-12 very feeble flattened costae radiate and extend to the base of the neck
where they coalesce. These costae, which are so slightly elevated above the body of the
shell as to be visible only with very oblique illumination, are decorated with a chain of
depressed pits. The surface between the costae is covered with similar pits less regularly
arranged. Length o-6 mm., breadth 0-3 mm.

Only a few specimens, all but one in a poor state of preservation. It is possible that in
the perfect condition the whole of the ornament is enclosed by an outer pellicle or skin,
as in L. scottii, Heron- Allen and Earland (H.-A. and E. 1922, TN, p. 150, pi. vi, figs.

3-4).

Apart from this conjecture the affinities of the species appear to lie near L. torqiiata,
Brady (B. 1884, FC, p. 469, pi. Iviii, fig. 41).

This species is named after Dr N. A. Mackintosh of the Discovery staff.

231. Lagena macroptera (Seguenza) (Plate IV, figs. 6, 7).
Fissurina macroptera, Seguenza, 1862, FMMM, p. 70, pi. ii, iig. 44.

One station : WS 429.

The single specimen which we figure appears to agree sufficiently well with Seguenza's
description and illustration to justify the revival of a name which has apparently never
been used by any subsequent author. In the specimen from South Georgia the keel is
somewhat narrower and the test more compressed than the original figure would suggest.
Length 0-25 mm., breadth 0-13 mm.

The species appears to have affinities with L. marginata. Sidebottom (S. 1912, etc.,
LSP, 1912, p. 406, pi. xvii, fig. 29) figures a very similar form under that name, but his
figure does not suggest the hyaline texture of our specimen. It cannot be confused with
L. margiiiato var. semimargbiota, Reuss, in which the carina is confined to the oral
extremity of the shell round the produced neck.

232. Lagena marginata (Walker and Boys) (F 221).

Fifteen stations: 123, 144, 149; WS 25, 27, 33, 37, 48, 66, 314, 357, 418, 429, 521, 522.

Never very common, often represented by one or two specimens only. Not a single
specimen with a fully developed carina was seen, the usual type being a somewhat
globose fissurine form with only a suggestion of a keel.

At a few stations the carina is better developed, a feature apparently accompanied by
a flattening of the test. The most highly carinate individuals were seen at WS 25, 33,
418 and 429, but the carina is always feebly developed.

233. Lagena marginata var. quadricarinata, Sidebottom (Plate IV, fig. 41).

Lagena staphyllearia var. quadricarinata, Sidebottom, 1912, etc., LSP, 1912, p. 404, pi. xxi,
fig. 16.

LAGENIDAE 115

One station: WS 522.

A single specimen 0-20 mm. in width occurred, differing from Sidebottom's figure
only in the absence of the few basal spines and the lesser development of the principal
carina. It may be observed that no mention of these spines is made in the text, although
they are presumably responsible for the assignment of the form to L. staphyUearia. But
they appear to bear no relationship to the marginal spines of that species, and both
Sidebottom's figure and my own specimen would appear to be more nearly related to
L. margiiiata.

234. Lagena marginata var. striolata, Sidebottom.

Lagena marginata var. striolata, Sidebottom, 1912, etc., LSP, 1912, p. 408, pi. 18, figs. 10, 11.

One station: WS 522.

Two specimens only were found. They resemble the finely striate specimen (fig. 10)
figured by Sidebottom, but the markings are much more delicate. Though very distinct
over the base and edges of the flask, they become indistinct over the central and upper
half, appearing more as hair-streaks than as definite striae. There is no basal cleft as in
Sidebottom's figure, the carina being continuous round the base.

235. Lagena melo (d'Orbigny) (F 200).
One station: WS 33.

A single, rather weak specimen at WS 33.

236. Lagena orbignyana (Seguenza) (F 240).

Five stations: 15; WS 33, 37, 42, 314.

Never more than a few specimens at a station, very good and typical specimens of the
nearly circular form at WS 33 and 314. At WS 37 and 42 a small variety occurs with the
central carina abnormally developed in the basal half and all three carinae partly sup-
pressed in the oral half of the shell.

237. Lagena quadralata, Brady (F 230).

Two stations: 136; WS 33.

A single specimen at St. 136 and four at WS 33. They are all fine examples of the
many-winged form described by us from the Falkland area.

238. Lagena quadrata (Williamson) (F 218).
One station: WS 521.

A single good specimen.

239. Lagena reticulata (Macgillivray) (F 199).
One station: WS 25.

A single good specimen.

ii6 DISCOVERY REPORTS

240. Lagena revertens, Heron-Allen and Earland (F 238).
Two stations: WS 27, 418.

A single typical specimen at WS 27 and a feeble and broken shell at WS 418.

241. Lagena schlichti (A. Silvestri) (F 225).

Four stations: 136; WS 27, 33, 522.

One very fine specimen at St. 136, others not so good at WS 522, and doubtful
specimens at the other stations.

242. Lagena spumosa, Millett (F 205).
Three stations: 16; WS 43, 46.

A single specimen at each station, particularly fine and typical at WS 46.

243. Lagena squamosa (Montagu) (F 197).

Four stations: 131 ; WS 25, 33, 349.

Singularly rare, represented by one or two specimens only at each station, the best at
St. 131. One specimen at WS 33 might have been attributed to L. sqiiamoso-sulcata, but
as the squamous markings covered quite three-quarters of the test it was referred to
L. squamosa.

244. Lagena squamoso-sulcata, Heron- Allen and Earland (F 196 a).
One station: WS 33.

Only a single typical specimen.

245. Lagena staphyllearia (Schwager) (F 224).
One station: WS 33.

A few small but typical specimens.

246. Lagena stewartii, J. Wright (F 171).
Two stations: WS 27, 33.

A single specimen at each station, particularly typical at WS 33.

247. Lagena striata (d'Orbigny) (F 188).

Nine stations: 16, 20, 129, 144; WS 25, 33, 37, 48, 349.

Good and well-marked shells occur, usually as single specimens, except at WS 33 and
37, where the species is more abundant. They are all of the elongate type first figured by
Williamson.

LAGENIDAE 117

248. Lagena sulcata (Walker and Jacob) (F 189).

Five stations: 149; WS 27, 33, 429, 522.

Extremely rare, a single specimen only at each station, and with the exception of
WS 429 and 522 all very feebly sulcate.

249. Lagena ventricosa, A. Silvestri (Plate IV, figs. 8, 9).

Lagena ventricosa, Silvestri, 1903, PMP, p. 10, figs. 6 a-c.

Lagena globosa, Sidebottom, 1912, etc., LSP, 1912, p. 379, pi. xiv, figs. 4, 5.

Ellipsolagena, gen.n., Silvestri, 1923, SE, p. 265.

Ellipsolagena ventricosa, Cushman, 1928, F, p. 265, pi. xxxviii, fig. 10.

One station : WS 522.

Two specimens only with well-developed hoods. One specimen is decidedly com-
pressed, resembling Sidebottom's fig. 5 (tit supra).

250. Lagena williamsoni (Alcock) (F 192).

Three stations: WS 25, 429, 522.

A single feeble specimen of the English type at WS 25 , and strongly costate individuals
at WS 429 and 522.

Sub-family NODOSARIINAE
Genus Nodosaria, Lamarck, 18 12

251. Nodosaria rotundata (Reuss) (F 247).
One station : WS 349.

A single large specimen only.

252. Nodosaria laevigata, d'Orbigny (F 248).
One station: WS 523.

Recognizable fragments of a large specimen at this station constitute the only record
for this species.

253. Nodosaria scalaris (Batsch) (F 250).
Two stations: 149; WS 349.

Very rare, but several good and typical specimens at St. 149.

254. Nodosaria calomorpha, Reuss (F 252) (Plate IV, fig. 19).

Four stations: 144; WS 33, 47, 429.

Very rare but very fine specimens ranging up to the exceptional number of 6-9
chambers at St. 144 and WS 33. At the other stations the specimens are of the normal
minute type with 2-3 chambers.

ii8 DISCOVERY REPORTS

255. Nodosaria consobrina, d'Orbigny.

Dentalina consohriua, d'Orbigny, 1846, FFV, p. 46, pi. ii, figs. 1-3.
Nodosaria consobrina, Brady, 1884, FC, p. 501, pi. Ixii, figs. 23, 24.
Nodosaria consobrina, Halkyard, 1919, BMB, p. 67, pi. iv, fig. 7.

Three stations: 45, 144; WS 33.

Rare, but several excellent specimens were found at St. 45.

256. Nodosaria communis, d'Orbigny (F 254).
One station: WS 522.

A single very small specimen.

257. Nodosaria pauperata (d'Orbigny) (F 255).
Two stations: 136, 144.

The species is poorly represented by a fragment of a large individual at St. 136, and
one three-chambered specimen at St. 144.

Genus Lingulina, d'Orbigny, 1826

258. Lingulina vitrea, Heron-Allen and Earland (F 264).
One station: WS 33.

One specimen only, not very typical, being thicker than the Falkland types.

Genus Vaginulina d'Orbigny 1826

259. Vaginulina legumen (Linne) (F 265).
One station: WS 429.

A small specimen only.

260. Vaginulina linearis (Montagu).

Nautilus linearis, Montagu, 1803-8, TB, Suppl. p. 87, pi. xxx, fig. 9.
Vaginulina linearis, Brady, 1S84, FC, p. 532, pi. Ixvii, figs. 10, 12.
Vaginulina linearis, Goes, 1894, ASF, p. 66, pi. xii, fig. 664.

One station: WS 521.

Represented by a single immature specimen, consisting of a megalospheric pro-
loculum and one subsequent chamber.

Genus Marginulina, d'Orbigny, 1826

261. Marginulina glabra, d'Orbigny.

Marginulina glabra, d'Orbigny, 1826, TMC, p. 259, No. 6, Modele no. 55.

Marginulina glabra, Brady, 1884, FC, p. 527, pi. Ixv, figs. 5, 6.

Marginulina glabra, Cushman, 1918, etc., FAO, 1923, p. 127, pi. xxxvi, figs. 5, 6.

Two stations: WS 429, 523.

A single good specimen at each station at depths of 2549 and 1697 m. respectively.

LAGENIDAE 119

Genus Cristellaria, Lamarck, 18 12

262. Cristellaria crepidula (Fichtel and Moll) (F 268).
One station: WS 418.

Only a single very small specimen. The absence of this species at other stations is
quite extraordinary.

263. Cristellaria gibba, d'Orbigny (F 274).
Two stations: WS 522, 523.

Extremely rare and small, the best specimen at WS 523.

264. Cristellaria cultrata (Montfort) (F 278).
One station: WS 522.

A single small specimen.

265. Cristellaria convergens, Bornemann (F 281).
One station: WS418.

A single small specimen only.

Sub-family POLYMORPHININAE
Polymorphina, d'Orbigny, 1826

266. Polymorphina lactea (Walker and Jacob) (F 283).
One station: WS 521.

The aboral half of a large specimen was found at WS 521 ; it is worthy of record only
on account of the depth (3780 m.) at this station which lies in the deep water between
the Falkland Islands and South Georgia. The species is normally of shallow water
habitat, but Brady records "exceedingly small" specimens at 1990 fathoms in the
South Atlantic and at 2350 fathoms in the South Pacific.

267. Polymorphina williamsoni, Terquem (F 285).
One station: WS 314.

A single specimen only.

268. Polymorphina compressa, d'Orbigny (F 292).
Three stations: 149; MS 14; WS 33.

A few very minute and pauperate specimens are with some hesitation referred to this
species.

Genus Uvigerina, d'Orbigny, 1826

269. Uvigerina canariensis, d'Orbigny (F 294).
One station: WS 429.

A single poor specimen.

I20 DISCOVERY REPORTS

270. Uvigerina pygmaea, d'Orbigny (F 297).
Two stations: WS 33, 349.

Very rare and weakly developed.

271. Uvigerina aculeata, d'Orbigny.

Uvigerina aculeata, d'Orbigny, 1846, FFV, p. 191, pi. xi, figs. 27, 28.

Uvigerina aculeata, Brady, 1884, FC, p. 578, pi. Ixxv, figs. 1-3.

Uvigerina aculeata, Cushman, 1910, etc., FNP, 1913, p. 100, pi. xliii, fig. 4.

Two stations: WS 429, 522.

Rare at WS 522, but very good specimens. Very rare and weaker at the other
station.

272. Uvigerina raricosta, d'Orbigny (F 299).
One station: WS 521.

Only a single specimen was noted, but it is possible that the species having no very
striking characteristics has been overlooked elsewhere among the variations of U.
angitlosa.

273. Uvigerina striata, d'Orbigny (F 300).

Nine stations: 123; WS 27, 33, 50, 66, 314, 349, 418, 521.

Extremely rare. Never more than one or two specimens at a station.

274. Uvigerina angulosa, Williamson (F 301).

Forty-one stations: 13, 16, 23, 27, 28, 30, 45, 123, 126, 131, 136, 140, 144, 149, 157; WS 25, 27, 28,
33. 37. 40, 42, 43. -SO, 51, 52, 66, 113, 154, 314, 348, 349, 351, 357, 373, 418, 426, 428, 521, 522, 523.

Generally distributed, but by contrast with the Falkland area the species is very rare
at the majority of stations. It is, however, common or very common at St. 136, WS 66,
314, 357 and 418, and at these stations practically all the variations referred to in the
Falkland Report were observed, notably at WS 357.

275. Uvigerina angulosa var. spinipes, Brady.

Uvigerina spinipes, Brady, 1879, etc., RRC, 1881, p. 64.

Uvigerina angulosa, var. spinipes, Brady, 1884, FC, p. 577, pi. Ixxiv, figs. 19, 20.

Uvigerina angulosa, var. spinipes, Cushman, 1910, etc., FNP, 1913, p. 99, pi. xliii, fig. 3.

One station: WS 351.

Extremely rare, represented by a few specimens only.

Family GLOBIGERINIDAE
Genus Globigerina, d'Orbigny, 1826

Note. At WS 522 at a depth of 2550 m. a few specimens were observed of G.
inflata, G. pachyderma and G. dutertrei. What are apparently the parent shells have
smaller individuals firmly attached, usually to the last chamber of their tests (see Plate
IV, figs. 20-22). They might be regarded merely as monstrosities, but for the fact that

GLOBIGERINIDAE 121

similar specimens have been observed more frequently in some of the Antarctic material,
sometimes bearing more than one accessory shell. Their presence is not easily explained
unless they are young specimens, discharged or budding from a parent shell, which have
failed to attain separate existence. Compare the note on Haplophragmoides canariensis
(No. 109). They are quite different from the "wild growing monstrous forms" figured
by Brady (B. 1884, FC, p. 593, pi. Ixxxi, figs. 6, 7), which appear to be due to fusion
of fully grown individuals.

276. Globigerina bulloides, d'Orbigny (F 304).

Twenty-seven stations: 27, 30, 123, 131, 133, 138, 144, 149; WS 20, 25, 27, 28, 32, 33, 36, 40,
42, 44. 47. 52, 66, 314, 351, 429, 521, 522, 523.

This species is usually rare or very rare and small, but at a few stations, notably WS
429 and 521 it is very common and well developed. At WS 33 it is common, but all the
specimens are small ; at WS 25 they are frequent but pauperate.

277. Globigerina triloba, Reuss (F 305).

Nine stations: 133, 149; WS 47, 351, 429, 521, 522, 523; MS 14.

Very common at WS 521, common at WS 522, frequent at WS 429 and rare at the
remaining stations. At the stations where it occurs in any numbers some of the speci-
mens attain a large size.

278. Globigerina inflata, d'Orbigny (F 306).

Ten stations: 131, i49;WS28, 314, 351,429, 521, 522, 523; MS 14.

Extremely common at WS 521 and 522, and common at WS 429. Rare or very rare
at the other stations. At most of the stations a very thick-walled type is dominant, but at
St. 149 a thin- walled type only occurs.

279. Globigerina dutertrei, d'Orbigny (F 307).

Thirty-eight stations: 15, 20, 23, 27, 31, 45, 123, 131, 133, 136, 138, 139, 140, 144, 149, 660;
WS 20, 28, 32, 33, 36, 38, 40, 41, 42, 43, 44, 63, 66, 154, 314, 351, 428, 429, 521, 522; MS 14, 68.

Typical specimens are always rare or very rare, which in fact is the general record of
the species in the area. A form intermediate between G. dutertrei and G. pachyderma is,
however, extremely common at WS 429, 521 and 522, and rare at many other stations.
It is more compact than the type, the chambers being less inflated and the aperture a
large hooded arch directed inwards towards the umbilicus. The two forms often occur
together, especially at WS 522. The best typical specimens are found at stations where
the pachyderma variety does not occur, notably at Sts. 27, 45 and 144. Intermediate
varieties are frequent.

280. Globigerina conglomerata, Schwager (F 308).

Sixteen stations: 30, 45, 133, 138, 139, 144, 149; WS 28, 33, 36, 44, 314, 351, 429, 522, 523.

Very common at WS 429 and 522, common at WS 314 and 523, frequent at St. 45
and rare or very rare at the remaining stations. The best and most typical specimens at
Sts. 45, 149, WS 314 and 522. There is great variation in development. At WS 429 a
complete series running into G. pachyderma was obtained.

122 DISCOVERY REPORTS

281. Globigerina pachyderma (Ehrenberg) (F 310).

Fifty-two stations : 14, 16,17,20,23,30,31,42,45,131, 136,138,140, 143, 144,145,149, 157,660;
WS 20, 25, 27, 32, 33, 36, 37, 40, 41, 42, 43, 44, 45, 50, 63, 63-4, 66, 113, 154, 314, 349, 351, 361, 373,
418, 426, 428, 429, 521, 522, 523; MS 14,68.

Almost universally distributed and often extremely abundant, notably at WS 63,
63-4, 314, 418, 429, 521, 522 and 523. At many of these stations it is the dominant
species and forms a high percentage of the organic remains. There are, however, many
stations at which it is very rare. There is a wide range of variety in the size and develop-
ment of the species. At WS 25, where it is frequent, the specimens are all small and
thick walled, and the aperture a mere central puncture. Between this and a large thin-
walled form with large arched aperture every degree of variation is to be found.

282. Globigerina rubra, d'Orbigny (F 311).

Five stations: WS 25, 32, 41, 47; MS 68.

Only a single very small specimen at each of the WS stations, and three small speci-
mens at MS 68.

283. Globigerina conglobata, Brady.

Globigerina conglobata, Brady, 1879, etc., RRC, 1879, p. 286; 1884, FC, p. 603, pi. Ixxx, figs.

1-5-

Globigerina conglobata, Brady, Parker and Jones, 1888, AB, p. 225, pi. xlv, fig. 13.

Globigerina conglobata, Cushman, 1918, etc., FAO, 1924, p. 18, pi. iii, figs. 8-13.

Two stations: 149; WS 52.

One very small but typical specimen occurs at a depth of 184 m. at WS 52 off the
western extremity of South Georgia, and four small specimens at St. 149 in Cumberland
Bay. The southern range of this species is given by Brady as about 35° S, and the
specimens from St. 149 would not be above suspicion (see note on station, p. 37) but
for the occurrence of the specimen from the other end of the island.

284. Globigerina elevata, d'Orbigny (F 312).

Six stations: 15, 149; WS 18, 428, 429, 521.

Very common at WS 429, rare or very rare at the other stations, but good specimens
at all.

Genus Orbulina, d'Orbigny, 1826

285. Orbulina universa, d'Orbigny (F 314).

One station: 149.

The occurrence of several well-grown individuals at St. 149 only inside Cumberland
Bay (200-234 m.) is not easily accounted for, and they should perhaps be disregarded as
"strays" (see note on station, p. 37).

ROTALIIDAE 123

Genus Pullenia, Parker and Jones, 1862

286. Pullenia sphaeroides (d'Orbigny) (F 315).

Six stations: 148; WS 351, 428, 429, 522, 523.

Always rare. The finest specimens were at WS 522, where great size was attained.
More numerous but much smaller at WS 429 and 523.

287. Pullenia subcarinata (d'Orbigny) (F 316).

Thirty stations: 20, 27, 30, 123, 126, 136, 140, 144, 148, 149; WS 27, 28, 33, 40, 42, 43, 47, 48,
52, 63, 113, 154, 314, 334, 348, 357, 418, 429, 522; MS 68.

Very variable in its frequency, abundant at some stations, notably WS 27 and 33.
Here, as might be expected, there was a great range of variation : specimens inseparable
from P. qiiinqueloba were noticed at many stations, but as we have already explained in
our Falkland Report we can see no reason for separating Reuss's species from the earlier
species of d'Orbigny, the two being linked by endless variations.

288. Pullenia obliquiloculata, Parker and Jones (F 317).

Two stations: 30; WS 522.

A thick-walled variety is very common in the deep water Globigerina oozes of WS 522.
This is the only record except a single small specimen from St. 30, which is in Cumber-
land Bay (251 m.). Its presence there and the presence of Orbuliua universa at a neigh-
bouring station are probably due to the influence of currents.

Genus Sphaeroidina, d'Orbigny, 1826

289. Sphaeroidina bulloides, d'Orbigny.

Sphaeroidina bulloides, d'Orbigny, 1826, TMC, p. 267, no. i, Modele no. 65.
Sphaeroidina bulloides, Brady, 1884, FC, p. 620, pi. Ixxxiv, figs. 1-7.
Sphaeroidina bulloides, Brady, Parker and Jones, i888, AB, p. 226, pi. xlv, figs. 9-11.
Sphaeroidina bulloides, Cushman, 1918, etc., FAO, 1924, p. 36, pi. vii, figs. 1-6.

One station: WS 522.

A single small but quite typical specimen at a depth of 2550 m. The position of this
station is several hundred miles farther south than the highest record given by Brady,
which was in the Southern Ocean, at a position of 46° 46' S, 45° 31' E with a depth of
1375 fathoms.

Family ROTALIIDAE

Sub-family SPIRILLININAE

Genus Spirillina, Ehrenberg, 1841

290. Spirillina vivipara, Ehrenberg (F 319).
One station: 145.

Two very minute specimens only.

DVII 13

124 DISCOVERY REPORTS

291. Spirillina obconica, Brady (F 321).
Two stations: 145, 149.

A single specimen at each station.

292. Spirillina obconica var. carinata, Halkyard (F 321 a).

Spirillina vivipara, var. carinata, Halkyard, 1889, RFJ, p. 69, pi. ii, fig. 6.

Spirillina vivipara, var. carinata, Sidebottom, 1904, etc., RFD, 1908, p. 8, pi. ii. fig. 4.

Spirillina obconica var. carinata, Heron-Allen and Earland, 1913, CI, p. 108, pi. ix, figs. 6, 7.

One station: WS 33.
A single specimen.

Sub-family ROTALIINAE
Genus Patellina, Williamson, 1858

293. Patellina corrugata, Williamson (F 326).

Fourteen stations: 30, 123, 140, 144, 145, 149, 157; WS 18, 25, 27, 33, 51, 154, 314.

Never very common except at St. 123, WS 27 and 33. The largest specimens, however,
were noted at 149 and WS 25, where the species was comparatively rare. At most
stations both circular and oval types occur. Most of the specimens are megalospheric,
but microspheric individuals were seen at several stations. The species nowhere attains
the large size of some of the Falkland individuals, but is quite up to average dimensions.

Genus Discorbis, Lamarck, 1804

294. Discorbis globularis (d'Orbigny) (F 331).

Twenty stations: 27, 30, 45, 123, 126, 136, 140, 144, 145, 148, 660; WS 25, 27, 33, 154, 314, 348,
418; MS 14, 68.

Generally distributed and often very abundant. The species occurs both free and
sessile and the specimens cover a very wide range of variation in size, convexity and
shape. At some of the stations, notably St. 123, WS 25 and 33, specimens grow to a
large size and become very irregular with age. Encystment was observed at WS 33 and a
budding specimen at WS 25, at which station also specimens were found entirely covered
with a Diatom, Cocconeis sp.

295. Discorbis globularis var. anglicus, Cushman (Plate IV, figs. 26, 27).

Discorbina irregularis, Heron-Allen and Earland {nan Rhumbler), 1913, CI, p. 120, pi. x, figs.

2-4.

Discorbis globularis, var. anglica, Cushman, 1918, etc., FAO, 1931, p. 23, pi. iv, figs. 10 a-c.

Two stations : 45 ; WS 25.

Particularly fine examples of the irregular formation figured by us from Clare Island
{lit supra) and ascribed to Rhumbler's species, D. irregularis, occur at these two stations.
Many less wild-growing occur at other stations and were not separated from D. globularis.

ROTALIIDAE 125

Cushman has separated our form apparently on account of the absence of the
secondary apertures described by Rhumbler. His varietal name is not very happy, as
we have found these irregular forms not only in British waters, but wherever D.
globidaris is plentiful.

296. Discorbis mediterranensis (d'Orbigny) (F 332).
Five stations: 27, 145 ; WS 25, 33 ; MS 68.

Very common at WS 25, frequent to rare at the other stations.

297. Discorbis vilardeboanus (d'Orbigny) (F 333).
Four stations: 144, 149; WS 33 ; MS 68.

Common at St. 149 and MS 68 and very variable in height at the latter station.
Rarer at St. 144, where the most typical specimens were found, and very rare at WS 33.

298. Discorbis rosaceus (d'Orbigny) (F 334).

Eleven stations: 123, 131 ; WS 25, 31, 33, 40, 42, 51, 66, 351; MS 68.

Small specimens and rare everjrwhere, except at Sts. WS 25, 33 and 66. At the first
of these it is very common.

299. Discorbis chasteri (Heron-Allen and Earland) (F 352).
Two stations: 140; WS 33.

A single large individual at St. 140, many large and small at WS 33.

300. Discorbis margaritaceus, sp.n. (Plate IV, figs. 23-25).
One station: WS 25.

Test free, circular in outline, dorsal side convex, peripheral edge rounded, ventral
side rounded and sinking into a deep umbilicus ; consisting of about two convolutions,
six chambers in the final convolution ; sutures on dorsal side oblique, flush but strongly
marked; sutures on ventral side obscure and flush; dorsal surface granular, ventral
surface covered with very minute beads arranged in radial lines across the chambers
into the umbilical recess. Aperture a minute slit on inner edge of final chamber. The
general aspect of the shell is lustrous or pearly, owing to the difi'raction eff'ects produced
by the granulation of the dorsal surface, and the beads on the ventral surface which is the
brighter of the two.

Diameter up to 0-3 mm., height about 0-05 mm.

A few specimens only of this pretty shell, which is probably allied to D. parisiensis
(d'Orbigny), were found.

301. Discorbis subobtusus, Cushman.

Discorbis subobfusa, Cushman, 1921, FP, p. 304, pi. Ixx, fig. 2.
One station: WS 351.

Three specimens found at this station at a depth of 1170 m. agree fairly well with
Cushman's figure except in size. They are much smaller, the largest being only 0-35 mm.

13-2

126 DISCOVERY REPORTS

in greatest diameter, as compared with 1-5 mm. for his specimens from a depth of
494 fathoms off the Philippines.

Genus Heronallenia, Chapman and Parr, 1930

302. Heronallenia wilsoni (Heron-Allen and Earland).

Discorbina wilsoni, Heron-Allen and Earland, 1922, TN, p. 206, pi. vii, fig. 17-19; 1924, FQM,

p. 172.

Heronallenia wilsoni. Chapman and Parr, 1931, NAF, p. 237, pi. ix, figs. 7, 8.

Two stations: 17; WS 314.

A single specimen at each station. The original records were from the Antarctic, and
showed the same wide variation in depth as the specimens from South Georgia, which
were found at depths of 137 and 1950 m. respectively.

Genus Truncatulina, d'Orbigny, 1826

303. Truncatulina refulgens (Montfort) (F 355).

Twenty stations: 16, 20, 27, 30, 123, 136, 140; WS 25, 27, 33, 40, 43, 45, 46, 51, 63, 357, 373,
428, 521.

Always rare and often very small. The best specimens at Sts. 16 and 123.

304. Truncatulina lobatula (Walker and Jacob) (F 356).

Sixteen stations: 30, 123, 140, 145, 149, 660; WS 18, 25, 27, 33, 40, 66, no, 348, 521, 522.
Frequent at WS 27, elsewhere rare or very rare and often represented by a single
small or immature specimen. The best specimens at St. 149.

305. Truncatulina dispars, d'Orbigny (F 357).
One station: WS 32.

Only a single specimen of this species which is so characteristic of the Falkland area.

306. Truncatulina wuellerstorfi (Schwager) (F 361).
Two stations: WS 521, 522.

Very rare, but good specimens.

307. Truncatulina akneriana (d'Orbigny) (F 362).
Eight stations: 30, 123, 149; WS 25, 27, 33, 113, 522.

Frequent and good specimens at WS 25, rare or very rare elsewhere, but good
specimens at WS 27 and 522.

308. Truncatulina pseudoungeriana (Cushman) (F 363).

Seventeen stations : 30, 42, 123, 131, 140, 149; WS25, 27,33,42,66, 113, 154,314,348,428, 521.
Common but small at WS 66. Elsewhere very rare, often only a single specimen, but
good and typical, notably at St. 123, WS 33 and 154.

ROTALIIDAE 127

309. Truncatulina praecincta (Karrer) (Plate IV, figs. 28-30).

Rotcilia praecincta, Karrer, i868, MFKB, p. 189, pi. v, fig. 7.

Tnaicatiilina praecincta, Brady, 1884, FC, p. 667, pi. xcv, figs. 1-3.

Trmicatulina praecincta, Cushman, 1910, etc., FNP, 1915, p. 39, fig. 42 in text, pi. xxvi, fig. 2.

Two stations: 123, 144.

Many excellent specimens, especially at St. 144. They agree even better with Karrer's
original figure of the fossil form than with recent figures of the species. Its presence
in South Georgia is very anomalous, as recent records appear to be confined to tropical
seas.

310. Truncatulina haidingerii (d'Orbigny) (F 365).

Four stations: WS 28, 33, 521, 522.

Very rare, never more than one or two at each station, the best at WS 522. The speci-
mens agree very well with Brady's figure but are probably not the same organism as the
fossil recorded by d'Orbigny from the Vienna Basin, if his figure of the latter is reliable.

311. Truncatulina robertsoniana, Brady.

Truncatulina robertsoniana, Brady, 1879, etc., RRC, 1881, p. 65 ; 1884, FC, p. 664, pi. xcv, fig. 4.
Cibicides robertsoniana, Cushman, 1918, etc., FAO, 1931, p. 121, pi. xxiii, fig. 6.

One station : WS 27.

A single small specimen referred with some doubt to this species. While presenting
most of the typical features, the sutures on the dorsal side are more oblique than usual.

312. Truncatulina tumidula, Brady (F 366).
Six stations: WS 33, 428, 429, 521, 522, 523.

Very common at WS 429 at a depth of 2549 m. ; rare to very rare at the other stations.

313. Truncatulina bradyana (Cushman) (F 367).
Two stations: WS 33, 521.

All the specimens are very poor except at WS 521. It is very rare at both stations.

Genus Anomalina, d'Orbigny, 1826

314. Anomalina vermiculata (d'Orbigny) (F 369).
Three stations: 30, 149; WS 27.

Extremely rare. A nearly fully developed specimen was found at St. 149. At WS 27
two well-developed specimens in the intermediate stage occurred and a similar one at
St. 30. The rarity of the species as compared with its abundance in the Falkland area is
no doubt due to the difference of temperature.

138 DISCOVERY REPORTS

Genus Globorotalia, Cushman, 1927

315. Globorotalia hirsuta (d'Orbigny) (F 374).

Six stations: WS 28, 351, 429, 521, 522, 523.

Common at WS 522 as a large thick-walled bottom form. Very rare at all the other
stations.

316. Globorotalia scitula (Brady) (F 375).
Three stations: 138; WS 351, 522.

Rare. The specimens exhibit every degree of variation in the thickness of the walls of
the test.

317. Globorotalia crassa (d'Orbigny) (F 376).
Five stations: 149; WS 429, 521, 522, 523.

A large and thick-walled benthic form is very common at WS 523, and a small thick-
walled form is equally abundant at WS 521. Otherwise the species is rare but very good
and typical, especially at St. 149 and WS 522.

318. Globorotalia truncatulinoides (d'Orbigny) (F 377) (Plate IV, figs. 35-7).
Five stations: 149; WS 33, 351, 521, 522.

Common or very common at WS 521 and 522. The majority of the specimens at these
stations belong to a large and very thick-walled form with slightly convex dorsal surface.
It is very distinctive and may prove to be specifically different from the small thin-walled
type which occurs, but very rarely, at the other stations.

Genus Pulvinulina, Parker and Jones, 1862

319. Pulvinulina berthelotiana (d'Orbigny) (F 383).
One station: WS 522.

A few specimens from Globigerina ooze (2550 m.) appear to agree in general with
d'Orbigny's species. But although the sutures are heavily limbate on the dorsal side,
they are depressed on the ventral.

320. Pulvinulina punctulata (d'Orbigny).

Rotalia punctulata, d'Orbigny, 1826, TMC, p. 273, No. 25, Modele No. 12.

Pidvimilina repanda var. punctulata, Parker and Jones, 1865, NAAF, p. 394, pi. xiv, figs. 12, 13.

Pulvinulina punctulata, Brady, 1884, FC, p. 685, pi. civ, iig. 17.

Eponides punctulata, Cushman, 1918, etc., FAO, 193 1, p. 48, pi. x, fig. 6.

One station : WS 27.

A single very fine specimen about 3 mm. in diameter.

321. Pulvinulina elegans (d'Orbigny) (F 385).
One station : WS 522.

Rare but large and typical shells.

ROTALIIDAE 129

322. Pulvinulina umbonata (Reuss) (F 386).
Two stations: WS 429, 522.

Very rare. All the specimens, though typical, are small, except one very large shell at

WS 522.

323. Pulvinulina exigua, Brady (F 387).

Eight stations: 15, 145; WS 33, 42, 429, 521, 522; MS 68.

Frequent at WS 33, elsewhere rare or very rare; but some very good specimens at

WS 522.

324. Pulvinulina karsteni (Reuss) (F 391).

Thirty-eight stations : 13, 14, 16, 20, 23, 27, 42, 45, 123, 126, 131, 136, 140, 144, 148, 149, 157;
WS 25, 27, 28, 33, 37, 40, 41, 42, 43, 45, 48, 50, 66, 113, 154, 348, 349, 357, 418, 521 ; MS 14.

Although generally distributed in the area, the species is usually rare. It is, however,
very common at St. 144 and WS 28, and common or frequent at Sts. 45, 126, 149,
WS 33, 154 and 349. The most generally distributed type is bi-convex, with a somewhat
flattened dorsal side, but practically all the variations noted in the Falkland area were
observed. At WS 27 a small form with the dorsal surface quite flat occurs, all the con-
volutions are visible and the ventral surface is highly conical. This form occurs also at
140 and 144, in company with the normal type and intermediate variations. Very large
specimens were observed at several stations, notably St. 27.

325. Pulvinulina peruviana (d'Orbigny) (F 392).
Seven stations: 27, 42, 144, 148; WS 25, 27, 33.

Very common at St. 144, frequent at St. 27 and WS 27, but elsewhere very rare.
Some very large specimens at WS 27.

326. Pulvinulina pauperata, Parker and Jones.

Pulvimdiiia paiiperata, Parker and Jones, 1865, NAAF, p. 395, pi. xvi, figs. 50, 51 a, b.

Pulvinulina pauperata, Brady, 1884, PC, p. 696, pi. civ, figs. 3-11.

Pulvinulina pauperata, Flint, 1899, RFA, p. 330, pi. Ixxiv, fig. 3.

Laticarinina pauperata, Cushman, 1918, etc., FAO, 1931, p. 114, pi. xx, fig. 4, pi. xxi, fig. i.

One station: WS 522.

Several fairly large specimens from a depth of 2550 m.

Genus Rotalia, Lamarck, 1804

327. Rotalia beccarii (Linne) (F 393).
Three stations: 20, 149; WS 25.

Two specimens at WS 25 and one at each of the other stations. All typical but rather
small.

I30 DISCOVERY REPORTS

328. Rotalia soldanii, d'Orbigny (F 394 a).

Four stations: WS 429, 521, 522, 523.

Many large and typical specimens at WS 521 and 522, pauperate or young only at the
other stations.

Family NUMMULINIDAE

Sub-family NONIONINAE

Genus Nonion, Montfort, 1808

329. Nonion depressulum (Walker and Jacob) (F 399).

Forty-five stations: 13, 16, 20, 23, 27, 30, 31, 42, 45, 123, 126, 131, 140, 144, 148, 149, 157, 660;
WS 20, 25, 27, 28, 33, 37, 40, 42, 43, 45, 46, 47, 48, 50, 52, 63, 63-4, 66, 113, 154, 314, 349, 357,
418, 522; Drygalski Fjord; MS 68.

Generally distributed, very common at WS 50, common at Sts. 20, 45, 144, 149 and
MS 68, varying from frequent to rare elsewhere. The best series at Sts. 45, 149 and
MS 68.

The specimens are not in agreement with the British type but do not present sufficient
difference to warrant specific separation. The type from South Georgia is somewhat
inflated with flush thick sutures and an unbroken peripheral edge. The umbilicus well
marked owing to the coalescence of the ends of the sutural lines, but not so pronounced
as in the figure which Brady gives of A^. asterizans (B. 1884, FC, pi. cix, figs, i, 2).

330. Nonion asterizans (Fichtel and Moll) (F 400).
Three stations: 136; WS 79, 255.

Very rare and pauperate.

331. Nonion orbiculare (Brady).

Nomonina orbicularis, Brady, 1881, HNPE, p. 105, pi. ii, fig. 5 a, h; 1884, FC, p. 727, pi. cix,

figs. 20, 21.

Nonion orbiculare, Cushman, 1918, etc., FAO, 1930, p. 12, pi. v, figs. 1-3.

Two stations: 131 ; WS 37.

A few rather doubtful specimens, which are nearer to Brady's species than any other
with which I am acquainted.

332. Nonion umbilicatulum (Walker and Jacob) (F 401).

Twenty stations: 13, 16, 45, 123, 131, 136, 140, 143, 144, 149; WS 33, 42, 63-4, 113, 349, 351,
418, 429, 521, 522.

Always rare except at WS 429, where it is frequent and typical. Good specimens also
at Sts. 16, 123, WS 113, 521 and 522, but elsewhere always weak and far from typical.

333. Nonion pompilioides (Fichtel and Moll) (F 402).
Three stations : 660 ; WS 63-4, 349.

Very rare, only one or two specimens at each station.

NUMMULINIDAE 131

334. Nonion sloanii (d'Orbigny) (F 403).
Three stations: WS 37, 47, 154.

Rare, but very good specimens at WS 154. Very rare elsewhere.

335. Nonion stelligerum (d'Orbigny) (F 404).
One station : WS 429.

Very rare.

336. Nonion boueanum (d'Orbigny) (F 405).

Nine stations: 123, 131, 136, 144, 149; WS 40, 42, 113, 349.

Very good specimens are frequent at Sts. 136, 144, 149, the best at 149. At the other
stations this species is rare or very rare, but still typical.

337. Nonion grateloupi (d'Orbigny) (F 406).
One station: 144.

Very rare and pauperate, hardly separable from A'^. sloanii.

338. Nonion scapha (Fichtel and Moll) (F 407).

Nineteen stations: 16, 20, 23, 30, 31, 42, 123, 131, 144; WS 27, 28, 33, 40, 43, 50, 348, 349, 418,
429.

Very rare except at St. 123, WS 40 and 349, where it is frequent. Specimens are
rather small but quite typical except for a frequent tendency to inequilateral disposition
of the chambers, especially noticeable in specimens from WS 43. These inequilateral
specimens are inseparable from Nonionella and are evidence that, however convenient
from a taxonomic point of view, the genus Nonionella has no zoological value.

Genus Nonionella, Cushman, 1926

339. Nonionella iridea, Heron-Allen and Earland (F 410).

Forty-three stations: 14, 15,20,23,30,31,42,45, 123, 126, 131, 136, 140, 144, 148, 149, 157,660;
WS 25, 27, 32, 33, 37, 40, 42, 43, 45, 47, 48, 50, 63, 63-4, 113, 154, 314, 349, 351, 418, 429, 521, 522,
523; MS 68.

Universally distributed and perhaps the most typical species of the area round South
Georgia. It is common or very common at many stations, notably St. 144 and WS 33,
429 and 523, but rare at some others.

340. Nonionella turgida (Williamson).

Rotalina turgida, Williamson, 1858, RFGB, p. 50, pi. iv, figs. 95-7.
Nonionina turgida, Brady, 1884, FC, p. 731, pi. cix, figs. 17-19.
Nonionina turgida, Cushman, 1918, etc., FAO, 1930, p. 15, pi. vi, figs. 1-4.

Five stations: 15, 45, 131, 149; WS 33.

Always rare or very rare except at St. 149 where it is frequent.

DVII 14

132 DISCOVERY REPORTS

Genus Elphidium, Montfort, 1808

341. Elphidium incertum (Williamson) (F 412).

Four stations: 140; WS 28, 46, 348.

Very rare and very far from typical, the sutural openings being confined to the neigh-
bourhood of the umbilicus.

342. Elphidium articulatum (d'Orbigny) (F414).
Two stations: 149; WS 314.

Only a single small specimen at each station.

343. Elphidium alvarezianum (d'Orbigny) (F 415).
One station : 30.

A single small specimen only.

344. Elphidium lessonii (d'Orbigny) (F 417).

Two stations: 145 ; WS 25.

Extremely rare. A single small specimen only at WS 25. Several good specimens at
St. 145, but still very small compared with the dimensions attained in the Falkland area.

345. Elphidium owenianum (d'Orbigny) (F 419).
Three stations: WS 25, 27, 33.

Very rare, only a few small specimens at each station, the best at WS 25.

APPENDIX

A No. 130. TuniteUella laevigata. While this report was in the press, I received for identifica-
tion some specimens which had been collected in shore sands of the Torbay area, South Devon,
by Mr E. Milton, F.R.M.S., of Torquay. They are unquestionably referable to this species, differing
only from the South Georgia specimens in their smaller size (the largest is only 0-34 mm. long),
and slightly darker colour, due to the presence of more iron in the cement.

Their occurrence is one more instance of the mysteries of distribution. Perhaps in time we shall
obtain intermediate records, but in the meantime such widely separated records illustrate the
futihty of many theories based only on local lists of faunas.

B The genus Miliammina has now been recorded as a fossil from the Upper Cretaceous of
Manitoba. A new species, Miliammina manitobetisis, Wickenden, is described and figured in Nezv
Species of Foraminifera from the Upper Cretaceous of the Prairie Provinces (Robert T. D. Wickenden,
Trans. Roy. Soc. Canada, ser. 3, xxvi. Sect, iv, 1932, p. 90, pi. i, figs. 11 a-c).

SUPPLEMENTARY BIBLIOGRAPHY

See Part I, pp. 443-51.

The following are the titles of papers other than those referred to in Part I of this Report.

B. 1912, PPSC. R. M. Bagg, jr. Pliocene and Pleistocene Foraminifera from Southern California. U.S.

Geol. Bur., Bull. 513, Washington, D.C., 1912, pp. 1-153, pi. i-xxviii.

C. igio.NAFP. J. A. CuSHMAxN. New Arenaceous Foratninif era from the Philippines. Proc. U.S. Nat. Mus.,

XXXVIII, 1910, pp. 437-42, text-figs. 1-19.
C. 1910, BL. F. Chapman. A Study of the Batesford Limestone. Proc. Roy. Soc. Victoria, xxii (N.S.),

1909-10, Art. 20, pp. 263-314, pi. lii-lv.
C. 1917, NFP. J. A. Cushman. New Species and Varieties of Foraminifera from the Philippines, and adjacent

waters. Proc. U.S. Nat. Mus., Li, pp. 651-62. (No plates.)
C. 1920, CAE. J. A. Cushman. Report of the Canadian Arctic Expedition, 1913-18, ix, pt M, Foraminifera,

Ottawa, 1920, pp. 1-12, pi. i.
C. & C. 1929, CFC. J. A. Cushman and C. C. Church. Some Upper Cretaceous Foraminifera from near

Coalinga, California. Proc. Calif. Acad. Sci., ser. 4, xviii, No. i6, pp. 497-530, pis. xxxvi-xli.
de F. 1887, RR. L. (Marquis) de Folin. Les Rhizopodes Reticulaires. Le Naturaliste (Paris), 1887, pp.

102-3, ii3-i5> 127-8, 139-40, figs. 1-20 in text, etc. (1888).
F. 1887, B. L. (Marquis) DE Folin. Les Bathysiphons, premieres pages d'une Monographic du Genre. Actes

Soc. Linn. Bordeaux, XL, 1887, pt 5, pp. 271-91 ; pt 6, 1888, pis. v-viii.
F. 1889, MPPS. C. Fornasini. Minute forme di Rizopodi reticolari nella mama pliocenica del Ponticello di

Sdvena presso Bologna. Bologna, 1889, pp. 2, i fig.
de F. 1895, SRR. L. (Marquis) de Folin. Aperf us sur le Sarcode des Rhizopodes reticulaires . Bull. Soc. Hist.

Nat. Colmar, 1895-6 (reprint, pp. 25), pi. O.
H. 1868, KTF. M. Hantken. A kis-czelli tdlyag foraminiferdi. Magyar, foldt. tarsulat munkalatai. iv,

Pesth, 1868, pp. 75-96, pis. i, ii.
H. 1883, ALB. R. Haeusler. Die Astrorhizidenund Lituolidender Bimammatus-zone. Neues Jahrb. Min.,

I' PP- 55~6i, pi. iii-iv. (See also Notes on some Upper Jurassic Astrorhizida and Lituolidce.

QJGS, 1883, XXXIX, pp. 25-9, pis. iii, iv.)
H. 1883, JVT. R. Haeusler. On the Jurassic Varieties 0/ Thurammina papillata -Br«(/y. AMNH, ser. 5,

XI, pp. 262-6, pi. viii.
H.-A. 191 5, RPF. E. Heron- Allen. Contributions to the Study of the Bionomics and Reproductive

Processes of the Foraminifera. Phil. Trans. Roy. Soc. (London) B, ccvi. 1915, pp. 227-79,

pis. 13-18.
K. 1899, NNAE. H. Kiaer. Norwegian North-Atlantic Expedition 1876-8, Zoology. Thalamophora,

Christiania, 1899, pp. 15, pi. i and map.
L. 1903, F. J. J. Lister. In E. Ray Lankester, A Treatise on Zoology, pt i, fasc. ii, pp. 47-149, The Fora-
minifera. London, 1903.
L. 1929, TS. E. Lacroix. Textularia sagittula on Spiroplecta Wrightii? Bull. Inst. Ocean. Monaco, No.

532, 1929, 12 pp., 10 text-figs.
L. 1932, TPCM. E. Lacroix. Textularidae du plateau continental mediterraneen entre Saint-Raphael et

Monaco. Bull. Inst. Ocean. Monaco, No. 591, 1932, pp. 28, text-figs. 1-33.
N. 1878, GH. A. M. Norman. On the Genus Haliphysema zvith description of several forms apparently allied

to it. AMNH, ser. 5, I, pp. 265-84, pi. xvi.
P. 1870, GStL. W.K.Parker. (In G. M. Dawson.) On Foraminifera from the Gulf and River St Lawrence.

Canadian Nat. (Montreal), N.S., v, p. 172. Amer. Journ. Sci. (New Haven, Conn.), ser. 3, 1, 1871,

pp. 204-10; AMNH (London), ser. 4, vii, 1871, pp. 83-9.
R. 1905, MF. L. Rhu.mbler. Mitteilung iiber Foraminiferen. Verh. Deutsch. Zool. Ges. 1905, pp. 97-106,

figs. 1-9.
S. 1903, PMP. A. Silvestri. Forme nuove o poco conosciute di Protozoi Miocenici Piemontesi. Ace. R. Sci.

Torino, xxxix, 1903, pp. 4-15.
S. 1923, SE. A. Silvestri. Lo Stipite delle Elissaforme e le sua affinitd. Mem. Pont. Ace. Nuovi Lincei

(Rome), ser. 2, vi, pp. 233-71, pi. i.
W. ?i93i,FDSE. H.WiESNER. Die Foraminiferen der deutschen Sildpolar-Expedition igoi-igo;^. Deutsche

Siidpolar-Expedition, xx, Zoologie, Berlin, pp. 49-165, pis. i-xxiv.

14-2

INDEX

N.B. See Note at commencement of Index, Part I.

abbreviata, Textularia, 154

abyssorum, Crithionina, (S) 49

aculeata, Bidimina, 170

acideata, Uvigerina, 271

acuta, Lagena, 196

acitticosta, Lagena, 197

adunciis, Reophax, 107

advena, Verneuilina, 159

agglutinans, AmniobacuUtes , 116

agghithians, MilioJina, p. 90

agglutinans, Textularia, 153

akneriana, Truncatulina , 307

«/6rt, Hippocrepinella, 84

albicans, Thurammina, 75

algaeformis, Rhizammina, 94

alvarezianum, Elphidium, 343

alveolata var. substriata, Lagena, 198

alveoli niformis, Miliolina, p. 90

americanus, Ammobaculites, 117, (TT) 118-20

AmniobacuUtes, 1 16-21

Animochilostoma, 143-5

Ammodiscus, 126

Ammolagena, 124

Anunomargimdina, 122

Ammosphaeroidina, 142

anceps, Globotextularia, 141

angulosa, Astrorliisa, (T) 36

angulosa, Uvigerina, 274

angulosa var. spinipes, Uvigerina, 275

annectens, Lagena, 199

anomala, Biloculina, 8

Anomalina, 314

apicularis, Gaudryina, 164

apiculata, Lagena, 200

arenacea, Miliammina, (T) 149 p. 90, 92-3

Armorella, 71

articidatum, Elphidium, 342

Articulina, (T) 24

asterisans, Noinon, 330, (T) 329

Astrorhiza, 34-6

auriculata, Lagena, 201

baccata, Gaudryina, 163

bargmanni, AmniobacuUtes, 118, (T) 117

Bathy siphon, 51-3

bfccarii, Rotalia, 327

berthelotiana, Pulvinulina, 319

biancae, Lagena, 202, (T) 196

hicarinata, Lagena, 203, (T) 221

biformis, Spiroplectammina, 152

Bigenerina, 158

Biloculina, i-io

bisulcata, Lagena, 204

Bolivina, 177-84

bosciana, Miliolina, 15, (T) 150

boueanum, Nonion, 336

bradyana, Truncatulina, 313

bradyi, Biloculina, 3

bradyi, Ehrenbergina, 194

bradyi, Gaudryina, 161

bradyi, Trochammina, 137

bradyi, Verneuilina, 160

bradyi, Virgulina, 176

buccidenta, Planispirina, 27

bucculenta var. placcntiformis, MUiolina, (S) 28

bucculenta var. placcntiformis, Planispirina, 28

buchiana, Bulimina, 173

Bulimina, 166-73

6«//a, Tholosina, 67

bulloides, Globigerina, 276

bulloides, Sphaeroidina, 289

caelata, Bolivina, (T) 183

calomorpha, Nodosaria, 254

canaliculata, Thuramminopsis, p. 65

canariensis, Haplophragmoidcs, 109, p. 121

canariensis, Uvigerina, 269

cancellata, Cyclammina, 147

capillare, Bathysiphon, 51

Cassidulina, 185-90

catemdata, Lagena, 205

cenomana, Placopsilina, 123

charoides, Glomospira, 128

chasteri, Discorbis, 299

Cibicides, (S) 311

cincia, Bolivina, 183

circularis, Miliolina, 21

clathrata, Lagena, 206, (T) 220

clavata, Ammolagena, 124

Clavulina, 165

communis, Clavulina, 165

communis, Nodosaria, 256

compressa, Polymorphina, 268

compressum, Haplophragmoidcs, (T) 118

confusa, Sorosphaera, 54

conglobata, Globigerina, 283

conglomerata, Globigerina, 280

consobrina, Nodosaria, 255

convergens, Cristellaria, 265

Cornuspira, 29-32

Cornuspirella = Cornuspira, (S) 32

cornuta, Pelosphaera, 59

corrugata, Patellina, 293

costata, Lagena, 207

crassa, Cassidulina, 188

crassa, Ehrenbergina, 193

crassa, Globorotalia, 317

crassatina, Astrorhiza, 35

crassimargo, Haplophragmoidcs, no

crepidula, Cristellaria, 262

cribrosa, Miliammina, p. 90

Cristellaria, 262-5

INDEX

135

Crithionina, 46-9, (T) 44
cultrata, Cristellaria, 264
curlus, Reophax, (T) 96
Cydammina, 147, 148
cylindrica, Marsipella, 92

danica, Lagena, 208

decorata, Proteoniua, 63

deaissata, Bolivina, 184

Dendroiiina, 50

dentaliniformis , Reophax, loi, (T) 96

depressa, Sorospltaera, 55

depressa, Wehhinella, 65, (T) 66

depressulum, N onion, 329

diaphana, Iridia, 37

diffliigiformis, Proteonuia, 62, (T) 81

difformis, Bolivina, 181

diffusa, Cornuspira, 32

diffusa, Cornuspirella, (S) 32

di/a/ata, Bolivina, 180

Discorbina, (S) 295

Discorbis, 294-301

discreta, Rhabdammina, 93

dispars, Truncatulina, 305

distans, Reophax, 104, (T) 105

distans var. gracilis, Reophax, 105

distoma, Lagena, 209

diitertrei, Globigerina, 279, p. 120

earlandi, Seabrookia, 195

Ehrenbergina, 19 1-4

elegans, Bulimina, 167

elegans, Plecanium, (T) 156

elegans, Pulvinulina, 321

elegans, Textulaiia, (ST) 156

elegantissima, Bulimina, 172

elevata, Globigerina, 284

Ellipsolagena, (S) 249

elongata, Biloculina, 6

elongata, Hyperammina, 85

elongata, Marsipella, 91

elongata, Storthosphaera, 44

elongata var. impudica, Storthosphaera, 45

Elphidium, 341-5

emaciate var.felsinea, Lagena, (S) 212

emaciatum, Haplopliragmoides, (T) 118

e«i/.r, Anmwbaculites, (T) 122

e««.y, Ammomargimdina, 122

e/ww, Marginulina, (T) 122

Eponides (S) 320

exigua, Pulvinulina, 323

fasciata, Lagena, 210
fasciata var. f aba, Lagena, 211
felsinea, Lagena, 212
filiformis, Bathysiphon, (T) 51
fimbriata, Lagena, 213
Fissurina (S) 23 1
flavidum, Bathysiphon, (T) 53
flexibilis, Hippocrepina, 81

flexibilis, Reophax, 102

flexibilis, Technitella, (S) 81, p. 69

Flintia, u

flintii, Gaudryina, 162

foliacea, Cornuspira, 31, (S) 32

foliaceum, Haplopliragmium, (S) 121

foliaceus, Ammobaculites , 121

fontinense, Haplopliragmium, (T) 119

formosa, Lagena, 214, (T) 221

formosa var. comata, Lagena, (T) 215

formosa var. costata, Lagena, 215

foveolata, Lagena, 216

fragilis, Gordiospira, 33

funalis, Tubinella, 24

funalis var. inornata, Articulina, (T) 24

fusca, Psammosphaera, 56, (S) 57, p. 60

fusiformis, Bulimina, 166

fusiformis, Pelosina, 41

fusiformis, Reophax, 99

galeata, Ammochilostoma, 143
galeata, Trochammina, (S) 143
Gaudryina, 161-4
gaussi, Vanhoeffenella, 38
gibba, Cristellaria, 263
glabra, Marginulina, 261
Globigerina, 276-84
globigeriniformis, Trochammina, 140
Globorotalia, 315-18
globosa, Lagena, 217, (S) 249
Globotextularia ,141
globularis, Discorbis, 294, (T) 295
globularis var. anglicus, Discorbis, 295
globulifera, Hormosina, 108
globulus, Biloculina, 10
glomerata, Lituola, (S) 114
glo7neratum, Haplopliragmium, (S) 114
glomeratus, Haplophragmoides, 114
Glomospira, 127, 128, p. 51
gordialis, Glomospira, 127, (T) 126
Gordiospira, 33
gracilis, Lagena, 218
gracillima, Lagena, 219
granum, Crithionina, 46
grateloiipi, Nonion, 337

haeusleri, Thurammina, 73, (T) 76

haidiiigerii, Truncatulina, 310

Haplophragmium, (SS) 114, 115, 120, 121, 142

Haplophragmoides, 109-15

harrisii, Nouria, 146

hartiana, Lagena, 220

hemisphaerica, Thurammina, (T) 76

herdniani, Lagena, 221

Heronallenia, 302

hexagona, Lagena, 222

Hippocrepina, 79-81, p. 69

Hippocrepinella, 82-4

hirsuta, Globorotalia, 315

hirudinea, Hippocrepinella, 82, (TT) 83, 84

136

DISCOVERY REPORTS

hirtidinea var. crassa, Hippocrepinella, 83

hispidiila, Lagena, 223

Honnosina, 108

Hyperammina, 85-8

hystrix var. glabra, Ehrenhergina, 192

incertum, Elphidimn, 341
incertus, Ammodisais, 126, (T) 127
indivisa, Hippocrepina, 79
indivisa, Rhizammina, (T) 94
iiidivisiim, Psammatodendron, 90
iiiflata, Glubigeriua, 278, p. 120
iuflata, Trochammiiia, 134
Invohitina, p. 91
involvens, Cornuspira, 29
iridea, Nonionella, 339
Iridia, 37

irregnlaris, Discorbina, (ST) 295
irregularis, Planispirina , 25
isabelleana, Biloculiiia, (T) 11

Jaculella, 78

liarsteni, Pulvinulina, 324

lactea, Polymorphina, 266

laevigata, Cassidulina, 185

laevigata, Hyperammina, 86

laevigata, Nodosaria, 252

laevigata, Turrilellella, 130, p. 132

laevigata var. tiimida, Cassidulina, 186

laevis, Lagena, 224

laevis, Tholosina, (T) 67

Lagena, 196-250

Lagenammina, (T) 105

lagenoides, Lagena, 225, (T) 221

lagenoides var. tenuistriata, Lagena, 226

laguncula, Lagenammina, (T) 105

lamarckiana, Miliolina, 17

/ate, Miliammina, 151

Laticarinina paupcrata, (S) 326

legutnen, Vaginulina, 259

lessonii, Elphidimn, 344

limicola, Astrorhiza, 34

limosa, Webbinella, 66

linearis, Nautilus, (S) 260

linearis, Vaginulina, 260

lineata, Lagena, 227

Lingulina, 258

Lituola, glomerata, (S) 114

lobatula, Truncatulina, 304

lucida, Lagena, 228

lyellii, Lagena, 229

mackintoshiana, Lagena, 230
macroptera, Lagena, 231
macroptera, Fissurina, (S) 231
malovensis, Bolivina, 182
malovensis, Trochammina, 135
mamilla, Crithionina, 47

manitobensis, Miliammina, p. 132

margaritaceus, Discorbis, 300

?narginata, Bulimina, 168

marginata, Lagena, 232, (TT) 231, 233

tnarginata var. quadricarinata, Lagena, 233

marginata var. semimarginata, Lagena, (T) 231

marginata var. striolata, Lagena, 234

Margimiliiia, 261

Marsipella, 91, 92

jnediterranensis, Discorbis, 296

wefo, Lagena, 235, (S) 244

we/o, Technitella, (S) 49

Miliammina, 149-51, pp. 91, 132

Miliolina, 12-21, (S) 28

}nilne-edwardsi, Biloculina, 5

minuta, Saccammina, 61

minutissima, Bigenerina, 158

murrliyna, Biloculina, i

«a«(7, Troc/iammina, 136, (T) 135

Nautilus, (S) 260

nitens, Textularia, 157

nitida, Trochammina, 138

Nodosaria, 251-7

nodulosus, Reophax, 103, (T) 62, p. 89

Nonion, 329-38

Nonionella, 339, 340, (T) 338

Nonionina, (S) 331, 340

Nouria, 146

novae-zealandiae, Hyperammina, 87

obconica, Spirillina, 291

obconica var. carinata, Spirillina, 292

obesa, Sigmoilina, 22

obliqua, Miliammina, 150, (T) 151

oblicjuiloculata, Pullenia, 288

oblonga, Miliammina, 149, (TT) 150, 151

oblonga, Miliolina, 14, (T) 150

oblonga var. arenacea, Miliolina, (T) 151, pp. 90, 94

obtusa, Jaculella, 78

ochracea, Trochammina, 133

oculus, Vanhoejfenella, 39

orbiculare, Nonion, 331

orbicularis, Cyclammiua, 148

orbicularis, Nonionina, (S) 331

orbignyana, Lagena, 236

orbignyana var. clathrata, Lagena, (TT) 206, 220

Orbulina, 285

oviformis, Llippocrepina, 80, (T) 81

ozvenianum, Elphidium, 345

pachyderma, Globigerina, 281, (TT) 279, 280, p. 120
papillata, Dendronina, 50

papillata, Thurammina, 72, (S) 73, (T) 76, pp. 64, 65
papillata var. albicans, Thurammina, (S) 75
papillata var. haeusleri, Thurammina, (S) 73
papillata var. parallela, Tliurammina, (S) 74
papillata var. tuberosa, Thurammina, (S) 77
parallela, Thura?nmina, 74
parisiensis, Discorbis, (T) 300

INDEX

137

parkeriaim, CassiduUna, 190

parva, Psammosphaera, 57, (T) 48, pp. 57, 65

parvula, Textularia, (ST) 155

patagonica, Biloculbia, 7

patagonica, Bulimina, 169

Patellina, 293

pauciloculata, Ammochilostoma, 144

paiiciloculata, Trochammina, (S) 144

pauperata, Laticarinina, (S) 326

pauper ata, Nodosaria, 257

pauperata, Pulvinulina, 326

Pelosina, 40-3, (T) 59

Pelosphaera, 59

perforata, Tubinella, (T) 24

peruviana, Biloculina, (T) 1 1

peruviana, Pulvinulina, 325

pilulifer, Rcophax, 97

pisutn, Biloculina, 9

pisum, Crithionina, 48

/)W«7« var. hispida, Crithionina, 49, (T) 48

Placopsilina, 123

Planispiri?ia, 25-8

Plecanium, (T) 156

Polymorphina, 266-8

pompilioides, Nonion, 333

praecincta, Rotalia, (S) 309

praecincta, Truncatulina , 309

Problematina, p. 91

protea, Tholosina, 68

protea, Thurarnmina, 76, (T) 74

Proteonina, 62-4

Psammatodendron, 90

Psammosphaera, 56-8, pp. 60-1

pseudoiingeriana, Truncatulina, 308

piilchella, CassiduUna, 187

Pullenia, 286-8

Pulvinulina, 319-26

punctata, Bolivina, 177

punctulata, Eponides, (S) 320

punctulata, Pulvinulina, 320

punctulata, Rotalia, (S) 320

/)«/)«, Ehrenbergina, 191, (T) 193

pupoides, Gaudryina, (S) 161

pygmaea, Miliolina, 18

pygmaea, Uvigerina, 270

pygmaea, Verneuilina, (S) 160

quadralata, Lagena, 237
quadrata, Lagena, 238
quinqueloba, Pullenia, (T) 287

ramosa, Hyperammina, (S) 89

ramosa, Saccorhiza, 89

raricosta, Uvigerina, 272

refulgens, Truncatulina, 303

reophaciformis, Atnmobaculites, (T) 116

Reophax, 95-107

repanda var. punctulata, Pulvinulijia, (S) 320

reticulata, Lagena, 239

revertens, Lagena, 240

Rhabdammifia, 93

Rliizatnmina, 94

ringens, Ammochilostoma, 145

ringens, Biloculina, (SS) 3, 4

ringens, Haplophragtnoides, (S) 145

ringens, Trochammina, (S) 145

robertsoniana, Cibicides, (S) 311

robertsoniana, Truncatulina, 311

robusta, Bolivina, 179

robustus, Reophax, 98

rosaceus, Discorbis, 298

rostratus, Ammobaculites, 119, (T) 117

i?o/«/«j, 327, 328, (S) 309, 320

rotaliformis, Trochammina, 132

Rotalina, (S) 340

rotulatus, Haplophragmoides, 115

rotulatum, Haplophragmium, (S) 115

rotundata, Nodosaria, 251

rotundata, Pelosina, 40, (T) 41

n/6ra, Globigerina, 282

n/i^M, Reophax, (S) 106

rufescens, Bathysiphon, 52

rufum, Bathysiphon, 53, (T) 52

rustica, Psammosphaera, 58, p. 64

Rzehakina, p. 91

sabulosus, Reophax, 106

Saccammina, 60, 61, (T) 64

Saccorhiza, 89

sagittula, Textularia, (T) 156

scalaris, Nodosaria, 253

scapha, Nonion, 338

schlichti, Lagena, 241

schreibersiana, Virgulitia, i-j^

scitula, Globorotalia, 316

scitulum, Haplophragmoides, 112

scorpiurus, Reophax, 95, (ST) 96

scottii, Lagena, (T) 230

Seabrookia, 195

selseyensis, Cornuspira, 30

seminidum, Miliolina, 12

serrata, Biloculina, 2

shoneana, Trochammina, (S) 129

shoneana, Turritellella, 129, (T) 130

shoneanus, Atnmodiscus, (S) 129

Sigmoilina, 22, 23, p. 91

Silicina, p. 91

Silicininae, p. 91

Silicosigmoilina, p. 91

siphonella, Gaudryina, (S) 164

sloanii, Nonion, 334, (T) 337

soldanii, Rotalia, 328

soluta, Flintia, 11

Sorosphaera, 54-5

sphaera, Planispirina, 26

sphaerica, Armorella, 71

sphaerica, Saccammina, 60

sphaeriloculum, Haplophragmoides, 11 1

sphaeroides, Pidlenia, 286

Sphaeroidina, 289

138

DISCOVERY REPORTS

sphaeroidiniformc , Haphphragmium , (S) 142

sphaeroidinifonnis, Ammosphaeroidina, 142

spiculifer, Reophax, 100

spinipes, Uvigerina, (S) 275

Spirillina, 290-2

Spiroplectammina, 152

spumosa, Lagena, 242

squamata, Trochammina, 131

squamosa, Lagena, 243, (T) 244

squamoso-sulcala, Lagena, 244, (TT) 205, 243

staphylkan'a, Lagena, 245, (T) 233

slaphyllcaria var. quadricarinata, Lagena, (S) 233

stelligera, Lagena, (T) 200

stelligerum, Nonion, 335

stewaitii, Lagena, 246

Storthosphaera, 44, 45

striata, Lagena, 247

striata, Uvigerina, 273

subcarinata, Pullenia, 287, (T) 145

siihfiisiformis, Reophax, 96

subglobosa, Cassidulina, 189

siibglobosiis, Haplophragmoides, 113

siibnodosa, Hyperammina, 88, (TT) 47, 76

subobtiisus, Diseorbis, 301

subrotunda, Miliolina, 16, (T) 151

subrotundata , Gaudryina, (S) 162

subsqiiamosa , ]'irgulina, 175

subteres, Btdimina, 171

sulcata, Lagena, 248

Technitella, (S) 49, 81

tenuimargo, Ammobacidites, 120, (T) 117

tenuimargo, Haplopliragmiiim, (S) 120

tenuis, Sigmoilina, 23

tenuissima, Textularia, 156, (T) 155

textilarioides, Bolivina, 178

Textularia, 153-7

Tholosina, 67-70, p. 64

Thurammina, 72-7, pp. 64-5

Thuramtninopsis, p. 65

Tolypammina, 125

torquata, Lagena, (T) 230

triangularis, Astrorhiza, 36

triangularis, Cruciloculina, (T) 20

tricarinata, Miliolina, 20

tricarinata var. crucioralis, (T) 20

triloba, Globigerina, 277

Trochammina, 131-40, (S) 143-5

Truncatulina, 303-13
truncatulinoides, Globorotalia, 318
tuberosa, Thurammina, 77
Tubinclla, 24
Tubinellina, (T) 24
tubulata, Proteonina, 64
tubulata, Saccammina, (S) 64
tumidula, Truncatulina, 312
turbinata, Trochammina, 139
turbinatum, Haplophragmium, (S) 139
turgida, Nonionella, 340
turgida, Nonionina, (S) 340
turgida, Rotalina, (S) 340
Turritellella, 129, 130

umbilicatulum, Nonion, 332
umbonata, Puhinidina, 322
universa, Orbulina, 285, (T) 288
Uvigerina, 269-75

vagans, Tolypammina, 125

Vaginulina, 259, 260

Vanhoeffenella, 38, 39

variabilis, Pelosina, 42

variabilis var. constricta, Pelosina, 43, (T) 42

variabilis var. dijformis, Textularia, (S) 181

ventricosa, Ellipsolagena, (S) 249

ventricosa, Lagena, 249

venusta, Miliolina, 19

vermiculata, Anomalina, 314

Verneuilina, 159, 160

vesicularis, Tholosina, 69, (T) 67

vesicidaris var. erecta, Tholosina, 70

vesper tilio, Biloculina, 4

vilardeboanus, Diseorbis, 297

Virgulina, 174-6

vitrea, Lingulina, 258

vivipara, Spirillina, 290

vivipara var. carinata, Spirillina, (S) 292

vulgaris, Miliolina, 13

Webbinella, 65, 66
zdesneri, Textularia, 155
tcilliamsoni, Lagena, 250
zvilliamsoni, Polymorphina, 267
uilsoni, Discorbina, (S) 302
wilsoni, Heronallenia, 302
wuellerstorfi, Truncatulina, 306

PLATE I

Figs. 1-4. FlitUia soluia, sp.n. (No. 11): x 15. Figs. 1, 2, edge views. Fig. 3, front view.
Fig. 4, end-oral view.

Figs. 5-7. Cornuspira diffusa, Heron-Allen and Earland (No. 32): x 25. Figs. 5, 6,
fragments. Fig. 7, abnormal fragment.

Figs. 8, 9. Astrorhisa triangularis, sp.n. (No. 36): x 16.

Figs. 10-12. Pelosiiia fusifonnis, sp.n. (No. 41): x 19. Fig. 10, specimen with produced
neck. Fig. 11, neck broken away. Fig. 12, in section.

Figs. 13-15. Pelosiiia variabilis, Brady, var.n. constricta (No. 43). Fig. 13, a South
Georgia specimen : x 14. Figs. 14, 1 5 , Antarctic specimens : x 6.

Figs. 16-21. Vanhoeffenella gaussi, Rhumbler (No. 38): x 21. Fig. 16, early stage.
Figs. i8, 19, advanced stages. Figs. 17, 20, 21, stages of subdivision.

Fig. 22. Vanhoeffenella oculus, sp.n. (No. 39): x 21.

Figs. 23, 24. Storthosphaera elongata, Cushman, var.n. impudica (No. 45): x 18.

Fig. 25. Bathysiphon rufescens, Cushman (No. 52): x 14.

Fig. 26. Bathysiphon capillare, de Folin (No. 51): x 15.

Fig. 27. Psammosphaera rustica, Heron-Allen and Earland (No. 58): x 19.

Figs. 28, 29. Proteonina decorata, sp.n. (No. 63): x 26. Fig. 28, side view. Fig. 29,
end-oral view.

Figs. 30, 31. Proteonina tubulata (Rhumbler) (No. 64): x 21.

Fig. 32. Astrorhiza limicola, Sandahl (No. 34): x 9. Specimen incorporating sponge
spicules.

DISCOVERY REPORTS VOL. VII.

PLATE I

V. li

SOUTH GEORGIA FORAMINIFERA

15.

JonnBrOt StrfU 4 Duin^iMtfiL'" Loiulur

PLATE II

Figs. I, 2. Webbinella litnosa, sp.n. (No. 66): x 13. Fig. 2 shows a specimen laid open,
with ingested diatoms.

Figs. 3-10. Thurammina protea, sp.n. (No. 76). Figs. 3, 4, sessile specimens in cavity of
Hyperammina: x 12. Figs. 5, 6, 8-10, free specimens illustrating range of
form: x 17. Fig. 7, a specimen built round sponge spicules: x 18.

Fig. II. Jaculella obtiisa, Brady (No. 78): x 30.

Figs. 12-15. Hippocrepinaflexibilis (Wiesner) {No. Si): X ^o. Fig. 12, collapsed specimen.
Fig. 14, oral view. Fig. 15, specimen in balsam viewed as a transparent
object. The dark masses are protoplasm.

Figs. 16-19. Reophax subfusiformis, sp.n. (No. 96): x 19.

Fig. 20. Reophax spiculifer, Brady (No. 100): x 45.

Fig. 21. Reophax distans, Brady, var.n. gracilis (No. 105): x 40. A reconstructed
specimen.

Fig. 22. Ammobaculites agglutinans (d'Orbigny) (No. 116): x 52.

Figs. 23-6. Ammobaculites bargmanni, sp.n. (No. 118): x 15. Fig. 23, immature stage.
Figs. 24, 25, adult stage. Fig. 26, edge-oral view.

DISCO \'ERV REPORTS. VOL VII.

PLATE I I

SOUTH GEORGIA FORAMINIFERA

PLATE III

Figs. 1-4. Ammomaiginulina ends, Wiesner (No. 122): x 50. Fig. i, viewed as a trans-
parent object. Figs. 2, 3, side views. Fig. 4, edge view.
Figs. 5-8. TurriteUella laevigata, sp.n. (No. 130): x 40. Fig. 8, viewed as a transparent

object.
Figs. 9, 10. TurriteUella shoneana (Siddall) (No. 129): x 70. Fig. 9, megalospheric form.

Fig. 10, microspheric form.
Figs. II, 12. Haplophragnwides scituliim (Brady) (No. 112): Oval variety, x 30. Fig. 11,

side view. Fig. 12, edge-oral view.
Fig. 13. Haplophragmoides caiiariemis (d'Orbigny) (No. 109): x 17. Abnormal specimen

with young shells attached.
Figs. 14-16. Nouria harrisii, Heron-Allen and Earland (No. 146): x 32.
Fig. 17. Miliammina lata (No. 151) and M. oblonga, Heron-Allen and Earland: x 45.

Abnormal individual incorporating both species.
Figs. 18-20. Textularia luiesneri, sp.n. (No. 155): x 75. Figs. 18, 19, front views.

Fig. 20, edge-oral view.
Figs. 21-30. Textularia teimissima, nom.n.iKo. 156). Figs. 21-3 - 75 : Figs. 24-30 x 7c.
Figs. 23-5 are drawn from balsam-mounted specimens, the others from
opaque specimens. Figs. 21, 22, microspheric forms. Figs. 23-30, megalo-
spheric A I and A 2.
Figs. 31-5. Textularia nitens, sp.n. (No. 157). Figs. 31, 32, microspheric, opaque: x 55.
Fig. 33, megalospheric, opaque: x 55. Fig. 34, microspheric, transparent:
X 60. Fig. 35, megalospheric, transparent: x 60.
Figs. 36-8. Bigenerina minutissima, sp.n. (No. 158): x 70.
Figs. 39-42. Clavulina cotnmunis, d'Orbigny (No. 165): x 43. Fig. 39, young, edge view.

Fig. 40, young, oral view.
Figs. 43-6. Verneuiliua advena, Cushman (No. 159). Figs. 43, 44, opaque: x 65.

Figs. 45, 46, transparent: x 60.
Fig. 47. Bulimina elegantissima, d'Orhigny {No. 172): x 42. Plastogamic specimens.
Figs. 48, 49. Bolivina decussata, Brady (No. 184): x 60. Fig. 48, front view. Fig. 49,

edge view.
Figs. 50, 51. Bolivina difformis (Williamson) (No. 181): x 66. Fig. 50, front view.

Fig. 51, edge view.
Fig. 52. Lagena acuticosta, Reuss (No. 197). Abnormal double specimen: x 52.

DISCOVERY REPORTS, VOL. VII.

PLATE III.

40

B Hopkins a=l

42. 46. 47 50.

SOUTH GEORGIA FORAMINIFERA

Jer.n&ilit Sons 4 [iBniBisaan I*'* LnnOan

PLATE IV

Figs. 1-3. Lagena apiculala (Reuss) (No. 200) Variety: x 39. Fig. 2, oral view. Fig. 3,

aboral view.
Figs. 4, 5. Lawfna/e/«Wa, Fornasini(No. 212): X 35. Fig. 4, oral view. Fig. 5, side view.
Figs. 6, 7. Lagena macroptera (Seguenza) (No. 231): x 30. Fig. 7, edge view.
Figs. 8, 9. Lagena ventricosa, A. Silvestri (No. 249): x 30. Fig. 8, side view. Fig. 9,

oral view.
Figs. 10, II. Lagena herdmani, sp.n. (No. 221): x 60. Fig. 10, side view. Fig. 11, edge

view.
Figs. 12, 13. Lagena hartiana, sp.n. (No. 220): x 42. Fig. 12, side view. Fig. 13, edge

view.
Figs. 14, 15. Lagena mackintoshiana, sp.n. (No. 230): x 34. Fig. 14, side view. Fig. 15,

oral view.
Figs. 16-18. Lagena formosa, Schwager, var.n. costata (No. 215): x 55. Figs. 16, 17, side

views. Fig. 18, edge view.
Fig. 19. Nodosaria calomorpha, Reuss (No. 254): x 48.
Figs. 20-2. Abnonnal Globigerinae with young individuals attached: x 40. See p. 120.

Figs. 23-5. Discorbis margaritaceus, sp.n. (No. 300): x 35. Fig. 23, dorsal view. Fig. 24,

ventral view. Fig. 25, edge-oral view.
Figs. 26, 27. Discorbis globularis var. angUcus, Cushman (No. 295). Fig. 26, dorsal view:

x 47. Fig. 27, ventral view: x 40.
Figs. 28-30. Truncatulina praecincta, Karrer (No. 309): x 52. Fig. 28, dorsal view.

Fig. 29, ventral view. Fig. 30, edge-oral view.

Figs. 31-4. Hippocrepina indivisa, Parker (No. 79): x 55. Fig. 33, normal type. Fig. 34,
apertural view of same. Fig. 3 1 , abnormal specimen with irregular aperture.
Fig. 32, apertural view of same.

Figs. 35-7. Globorotalia truncatulinoides (d'Orbigny) (No. 318): x 52. Fig. 35, dorsal
view. Fig. 36, ventral view. Fig. 37, edge-oral view.

Figs. 38-40. Trochammina malovensis, Heron-Allen and Earland (No. 135): x 35. Variety
with flattened dorsal surface. Fig. 38, ventral view. Fig. 39, edge-oral view.
Fig. 40, dorsal view.

Fig. 41. Lagena marginata var. quadricarinata, Sidebottom (No. 233): x 40.

DISCOVERS' REPORTS, VOL. VII.

PLATE IV.

B Hopkins ^1

SOUTH GEORGIA FORAMINIFERA

V DiLMelii.;.'. L*^ Lundno

PLATE V

Figs. 1-5. Miliammhia oblonga, Heron-Allen and Earland (No. 149): x 50. Fig. i, front
view. Fig. 2, back view. Fig. 3, front view. Fig. 4, edge view. Fig. 5,
oral view.

Fig. 6. Miliammina arenacea (Chapman) (sub No. 149). Isomorphous with MilioUna
oblonga (Montagu). Drawn from an Antarctic specimen, the species not
being found in South Georgia. Front view: x 50.

Figs. 7, 8. Miliammina oblonga, Heron-Allen and Earland (No. 149): x 50. Fig. 7,
viewed as a transparent object; front view. Fig. 8, viewed as a transparent
object; edge view.

Figs. 9-14. Miliammina obliqua. Heron- Allen and Earland (No. 150): x 50. Fig. 9, back
view. Fig. 10, front view. Fig. 11, edge view. Fig. 12, front view. Fig. 13,
young specimen; front view. Fig. 14, oral view.

Figs. 15-19. Miliammina lata, Heron-Allen and Earland (No. 151): x 50. Fig. 15, front
view. Fig. 16, edge view. Fig. 17, back view. Fig. 1 8, abnormal specimen;
front view. Fig. 19, oral view of normal specimen.

Figs, 20, 21. Sorosphaera depressa, Heron-Allen and Earland (No. 55). Fig. 20, a young
colony attached to a flexible organic base: x 25. Fig. 21, a colony of
individuals attached to a stone; several specimens show openings due to
accidental fractures : x 8.

Figs. 22-5. Ammobaculites rostratiis, Htron-Ml&n and Earland (No. 119): x 17. Fig. 22,
using mud for construction of test. Fig. 23, immature specimen. Figs. 24, 25,
using sand for construction of test.

DISCOVERY REPORTS, VOL. VII

PLATE V

SOUTH GEORGIA FORAMINIFERA

PLATE VI

Figs. 1-9. Ehrenbergina crassa, Heron-Allen and Earland (No. 193): x 90. Figs, i, 2,
megalospheric ; edge views. Fig. 3, megalospheric ; ventral side. Fig. 4,
megalospheric; dorsal side. Fig. 5, microspheric ; ventral side. Fig. 6,
microspheric ; edge view. Fig. 7, microspheric; edge view of specimen in
balsam. Fig. 8, microspheric; ventral side of specimen in balsam. Fig. 9,
megalospheric ; ventral side of specimen in balsam.

Figs. 10-15. Gordtospira fragilis, Heron-Allen and Earland (No. 33). Fig. 10, young
individual ; side view : x 40. Figs. 11, 12, aduh individuals; side views: x 30.
Fig. 13, adult individual; edge-oral view: x 30. Fig. 14, young individual
viewed as a transparent object to show the irregular coiling of the early
chambers; protoplasmic body dark: x 160. Fig. 15, young individual
viewed as a transparent object to show the irregular coiling of the early
chambers ; protoplasmic body dark : x 80.

DISCOVERY REPORTS, VOL. VII

PLATE VI

#0

SOUTH GEORGIA FORAAIINIFERA

PLATE VII

Figs. 1-9. Hippocrepinella hirudinea, Hemn-AWenandEarhnd (No. 82). Fig. i, abnormal
individual with bifurcate extremity: x 20. Fig. 2, thin section (transparent)
showing protoplasmic body loaded with diatoms: x 20. Fig. 3, opaque
section; the variations in the thickness of the wall of the test at different
places are due to the angle at which the section is cut : x 20. Fig. 4, oral-end
view: x 15. Figs. 5-9, side views illustrating variations in shape due to
compression, shrinkage, etc. Figs. 7-9 show accessory openings in the
walls, possibly due to parasites: x 20.

Figs. 10-12. Hippocrepinella alba, Heron-Allen and Earland (No. 84): x 20. Side views
of specimens of various sizes.

Figs. 13-15. Hippocrepinella hirudinea var. crassa, Heron-Allen and Earland (No. 83):
X 24. Figs. 13, 14, side views. Fig. 15, end-oral view.

Figs. 16-23. Armorella sphaerica, Heron-Allen and Earland (No. 71). Figs. 16-20,
illustrating variations in size, number of tubes, etc.: x 35. Fig. 21, with
incorporated sponge spicule : x 25. Fig. 22, using sponge spicules and coarse
sand for construction : x 25. Fig. 23, using coarse material for construction :
X 25.

Figs. 24-7. Pelosphaera cornuta, Heron-Allen and Earland (No. 59) : x 9. Fig. 24, young
individual. Figs. 25, 26, stages in development. Fig. 27, a specimen laid
open; the white lines between the sand grains indicate the highly finished
surface of the incorporating cement in the interior of the test, as compared
with the rough external layer.

DISCOVERY REPORTS, VOL. VII

PLATE VII

SOUTH GEORGIA FORA.MINIFERA

Discovery Reports. Vol. VII, pp. i7,()-i-]o, November, 1933.]

ON VERTICAL CIRCULATION IN THE
OCEAN DUE TO THE ACTION OF THE
WIND WITH APPLICATION TO CON-
DITIONS WITHIN THE ANTARCTIC
CIRCUMPOLAR CURRENT

By
H. U. SVERDRUP

ON VERTICAL CIRCULATION IN THE
OCEAN DUE TO THE ACTION OF THE
WIND WITH APPLICATION TO CON-
DITIONS WITHIN THE ANTARCTIC
CIRCUMPOLAR CURRENT

T

By H. U. Sverdrup

(Text-figs. 1-23)

H E theoretical studies of the ocean currents have principally dealt with the horizontal
currents which arise because of the effect of the wind and because of the distribution
of density. The question as to the relative importance of the wind and the distribution
of density has often been discussed and seems now to be answered in favour of the wind.
Defant, in his excellent survey of the present status of dynamic oceanography, says
(1929, p. 136) that the greater influence on the development of the horizontal currents
must be ascribed to the wind, while the differences in density are of special importance
to the vertical circulation. In this paper it is intended to show that the wind also main-
tains systems of vertical circulation and that the factors which influence the density of
the sea water, such as heating, cooUng, evaporation and precipitation, are of equal im-
portance to the horizontal currents as is the wind. Before doing so it is necessary to give
a brief review of our present theoretical knowledge of the ocean currents. Compre-
hensive investigations have especially been undertaken by Ekman (1928), who has
examined the currents which arise in homogeneous and non-homogeneous water both
under the action of the wind and as a resuh of the distribution of density. We shall
follow his classification and terminology.

In homogeneous water Ekman discriminates between three different current systems,
which all arise under the action of the wind :

(i) The pure drift current, which is limited to the uppermost layer in which the
current at the surface is directed at 45° cum sole from the direction of the wind. With in-
creasing depth this current turns cum sole and decreases in velocity until it becomes
negligible at the " depth of frictional resistance." Within this current the total transport
of the water is directed at 90° cum sole from the direction of the wind.

(2) The slope current, which arises because the oceans are limited and, therefore, the
water is piled up along the coast towards which the transport by the pure drift current
is directed. Because of this piling up, the surface of the sea becomes inclined and a
current is set up which runs with uniform direction and velocity from the surface to the
bottom, except in the immediate vicinity of the bottom where the influence of the
friction along the bottom has to be considered. Above the "lower depth of frictional
resistance" the slope current is directed at right angles cum sole from the direction of
the slope. Below the " lower depth of frictional resistance " we find the third constituent
of the current system.

142 DISCOVERY REPORTS

(3) The bottom current, which is part of the slope current and which reaches from the
top of the "lower depth of frictional resistance" to the bottom, and turns contra solem
with increasing depth.

This current system is illustrated by considering the currents which must arise in a
channel of uniform depth, running all around the earth between two parallels of latitude,
if a wind of uniform velocity blows in the direction of the channel. With applications to
conditions in the Antarctic in mind, it may be supposed that the channel is situated in
the southern hemisphere, that the wind blows from west to east, and that the depth is
considerably greater than the upper and lower depths of frictional resistance. Within
the pure drift current, which is limited to the upper depth of frictional resistance, the
transport is directed 90° to the left of the wind, which means in this case to the north.
This transport causes a piling up of the water along the northern boundary of the
channel and, therefore, the surface of the water must slope downwards from north to
south. This slope causes, on the other hand, a current which runs in a direction 90° to
the left of the slope, which means in the direction of the wind, from west to east. The
velocity and direction of the slope current are the same at all levels, except within the
bottom layer. When approaching the bottom the current turns to the right and de-
creases. Between the top of the lower depth of frictional resistance and the bottom, the
total transport has a component in the direction of the slope, which means from north
to south. Stationary conditions can exist when the total transport towards the north by
the pure drift current above the depth of frictional resistance equals the transport to-
wards the south below the lower depth of frictional resistance. On the assumption that
the upper and the lower depths of frictional resistance are the same, Ekman has found
that the velocity of the slope current must be equal to the velocity of the pure drift
current at the surface multiplied by -v/a.

Thus, the principal effect of the wind is to uphold a slope current in the direction of
the wind and in addition to maintain a transversal circulation, which in the upper layer
transports water from south to north and near the bottom transports water from north
to south. At the northern boundary descending motion, and at the southern boundary
ascending motion must take place because of the continuity. The problem has not been
treated in three dimensions, but supposing the depth of the channel to be small in com-
parison with the width, the vertical components of the currents must always remain
small and, therefore, it is improbable that a complete analysis would give results which
would deviate considerably from the above.

The simple system of three currents, the pure drift current, the slope current and
the bottom current is modified if the water is non-homogeneous. In this case the iso-
steres are generally not horizontal and, therefore, currents must be present which are
due to the distribution of density or in Ekman's terminology " convection currents." The
velocity and direction of these currents can be computed by means of the Bjerknes
theorem of circulation. Ekman considers the convection current as the fourth con-
stituent of the currents in the sea. He points out that the greatest changes of density in
a horizontal direction are found in the upper layers, from the surface to 1000 m. or less,

VERTICAL CIRCULATION IN THE OCEAN 143

and that in the deep water the isosteres are nearly horizontal. This implies that the
greatest number of solenoids in the sense of Bjerknes are present in the upper layers and
that very few or no solenoids are found in the deep water. Consequently the velocity of
the convection currents decreases rapidly with depth and approaches zero at great
depth, contrary to the velocity of the slope current, which remains constant from the
surface to the top of the lower layer of frictional resistance.

Ekman, furthermore, points out that in a sea in which the density increases with
depth, a pure drift current must give rise to a convection current, running generally in
the direction of the wind. In order to illustrate this effect of the wind we may again
consider the channel around the earth, but now we will assume that the water is not
homogeneous but that the density increases with depth. The immediate effect of a pure
drift current will be to transport the light surface water towards the left-hand side of
the channel and, therefore, near the surface the isosteres cannot remain horizontal but
must soon slope downwards from south to north. In the upper layers the light water is
accumulated on the left-hand side of the wind and we get a solenoid field which must
cause a current in the direction of the wind, but the velocities within this current de-
crease downwards since the inclination of the isosteres decreases. At the same time we
may get a piling up of the water at the northern boundary, and as a result of this a slope
current may arise, which also runs in the direction of the wind but remains constant
down to the top of the lower layer of frictional resistance.

Similarly we may consider a circular wind system within an area which is situated
completely in one hemisphere. If the circulation is contra solem the surface layers are
driven away from the centre and a field of solenoids is built up which results in a current
in the direction of the wind. If the circulation is cum sole the light surface water will be
carried towards the centre and a density distribution is brought about which gives rise
to a convection current circulating cum sole.

These considerations show the character of the currents which may be produced in
homogeneous and non-homogeneous water under the action of the wind, and also that
conditions become much more complicated when the water is non-homogeneous. In
order to undertake a complete analysis of the ocean currents one has, furthermore, to
consider changes in the distribution of density, which are caused by such factors as
heating and cooling, evaporation, etc. The problem becomes so complicated that at
present it cannot be made the subject of any mathematical analysis, but it seems possible
to follow some simple lines of reasoning which lead a step further and which give some
indications of the structure of the circulations in the sea.

In the first place it is of much interest to ascertain whether any evidence is present
for the existence of slope currents in a non-homogeneous sea. It is not a priori given
that such currents are developed. As will be shown below, under the action of the wind
on the surface of a channel of uniform depth running around the world, a stationary
current can exist only in the absence of a slope current, and generally it is possible that
the principal effect of the wind is to build up a solenoid field which gives rise to a con-
vection current and not to bring about a piling up of the water along a boundary.

144 DISCOVERY REPORTS

It has, however, to be borne in mind that the surface of a non-homogeneous sea is,
as a rule, not horizontal. Suppose that in the upper layers the density varies horizontally
but is constant at great depths and that the current is practically zero at great depths.
The latter assumption involves that at great depths the isobaric surfaces are horizontal
or level surfaces, and it follows that the surface of the sea shows elevations where the
average density is small and depressions where the average density is great. Within a
convection current the light water is accumulated on the left-hand side (in the southern
hemisphere) and, therefore, the surface of the sea is higher on the left-hand side of the
current. The surface, therefore, is inclined even if no slope current in the sense of
Ekman is present, and in the presence of such a slope current the actual inclination is
the combined effect of the distribution of density and the piling up of the water.

By means of very precise levelling the variations of the sea-level along coasts can be
determined. By comparing the results with the corresponding variations which can be
derived from the distribution of density, it should be possible to decide if this distribu-
tion is alone responsible for the variation, or if in addition a piling up effect has to be
considered. Some results of highly precise levelling in the United States and Central
America have been discussed by Avers (1927). Between Colon and Panama a difference
in sea-level of 17-8 cm. has been found, the level being highest on the Pacific side. By
means of the observations of the ' Carnegie ' in the Caribbean Sea and the Gulf of
Panama, a corresponding difference of about 35 cm. is found, supposing the pressure at
a depth of 3000 m. to be the same in both seas. If these figures can be brought in re-
lation to each other, they would indicate that the difference of level, which should be
expected because of the difference in density, has been reduced. Such a reduction
would be brought about if the water were piled up in the Caribbean Sea and drawn
away from the coast in the Gulf of Panama.

Avers, furthermore, points out that the sea-level increases from south to north both
along the Atlantic and Pacific coasts. Along the Atlantic coast an increase of 3 1 cm. is
recorded from St Augustine, Florida, to Portland, Maine, while along the Pacific coast
an increase of 34 cm. is found between San Pedro, California, and Seattle, Washington.
The sea-level is from 49 to 59 cm. higher at the Pacific coast than at the Atlantic. It
seems possible that the increase along the Atlantic coast can be associated with varia-
tions in the density of the water, but observations which are suitable for a test are not
available. It may, however, be pointed out that to the north of Portland, Maine, the
sea-level must sink, if the height depends upon the density of the water, because in that
region the average density increases rapidly to the north. Along the Pacific coast the
increase in the height of the sea-level cannot be attributed to a variation in the density,
since this, according to the observations of the 'Carnegie,' should rather tend to cause
a decrease from south to north. It is, on the other hand, not very probable that the
prevailing winds will cause a considerable piling up of the water at the two coasts, be-
cause the general directions are such that the surface water is drawn away. The results
of these levellings, therefore, throw no more light upon the question of the relative im-
portance of the distribution of density and the piling up effect of the wind. Apart from

VERTICAL CIRCULATION IN THE OCEAN 145

the result, that the sea-level is higher along the Pacific coasts than along the Atlantic,
which can be brought into agreement with the distribution of density, the introduction
of the results of the levelling has raised new questions instead of solving old ones.

There exist, on the other hand, observations which indicate that in the oceans the
piling up effect of the wind is negligible. The observations of the International Ice
Patrol around the Grand Banks of Newfoundland have been used for constructing
charts, showing the topography of the surface relative to some level at which the velocity
of the current presumably must be small. From these topographic charts the currents
at the surface can be derived, and since the tracks of icebergs have shown a remarkable
agreement with the computed currents, the latter must closely represent the actual flow
of the water. It is unnecessary to assume the existence of a slope current, and thus no
indication of a piling up of the water exists.

A still more striking example is found in the Strait of Florida where, according to
Wiist (1924), the currents which are computed from the distribution of density agree
perfectly with the measurements by Pillsbury. There the sea-level rises 30 cm. over a
distance of 40 km. in a direction at right angles to the coast. This very big rise over a
short distance clearly shows the importance of the distribution of density to the in-
clination of the sea-level. We must, however, leave open the question as to the possible
piling up effect of the wind and the existence of slope currents, and we shall proceed
to a discussion of stationary wind currents in non-homogeneous water.

Let us again consider the case of a channel running all around the earth in the southern
hemisphere. As a primary effect of the wind the light surface water will, as previously
stated, be transported to one side of the channel, a field of solenoids will be produced
and a convection current in the direction of the wind will commence. This convection
current must gradually become stronger as long as the pure drift current continues to
transport the light surface water across the channel. Stationary conditions can be
reached only when all transversal motion stops, because any transversal motion must
cause changes in the distribution of the density, if no other factors tend to maintain a
certain distribution.

It is easy to find the conditions which must be fulfilled if stationary conditions are to
exist. The equations of motion have in the stationary case the form :

9/> . , d I dvx

dp d I dv^

•(I)

where a represents the specific volume, p the pressure, v the coefficient of eddy viscosity
and V the velocity. The positive s^-axis is directed downwards. Placing the positive
j'-axis in the direction of the channel and the x-axis at right angles to the channel we
have:

«!=/(-). -«|-- w

146 DISCOVERY REPORTS

Assuming, furthermore, that the wind blows in the direction of the channel, the
boundary condition at the surface takes the form

where T represents the tangential stress of the wind.

Conditions can be stationary only when no transport takes place across the channel,
or when ?;^ = o at all levels. From equations (i) and (3) it follows that this condition is
fulfilled when ^^

V -r- = — T = constant.
dz

Since the coefficient v is always positive, it follows that the velocity must decrease at all
depths, and since it is improbable that v increases with depth it also follows that the
velocity decreases more and more rapidly with increasing depth. Furthermore, the
pressure gradient and the velocity must be zero at the bottom. If the velocity at the
bottom differs from zero the influence of the friction at the bottom must give rise to a
transversal current, in which case the original condition that v^ shall be zero at all levels
cannot be fulfilled.

The solution which gives stationary conditions can be written in the following form
if h is the depth to the bottom :

~"E=-^^^^' ^ = ''''/(-) = °' -'^1 = °'

2a>sm<^-'

dVy _ _ rp

dz

The condition that the pressure gradient shall be zero at the bottom leads, as already
stated, to the conclusion that a slope current cannot exist, because within the slope
current the gradient remains unchanged from the surface to the bottom.

It can easily be shown that such currents, as the above solution demonstrates, are not
met with in the sea. In order to do this it is necessary to introduce some function which
shows the relationship between the tangential stress of the wind and the wind velocity.
According to Ekman's and Taylor's investigations we have approximately

T= 3-2 X 10-6 (IF- Vf,

where W is the velocity of the wind and V the surface velocity of the water. The latter
is usually so small compared with the former that it can be disregarded, but in the
present case it has to be considered. Furthermore, it should be noted that with given
values of T, and .'o the coeflicient of eddy viscosity at the surface, the smallest surface
velocity is reached if the velocity is a linear function of depth, which means that the
eddy viscosity is constant. This follows from the assumption that the eddy viscosity

VERTICAL CIRCULATION IN THE OCEAN

'47

does not increase with depth. Consequently we can find a minimum value of the surface
velocity when we introduce t^

where h is the depth. We obtain:

3-2 X 10-8 {W - Fmin.)^ h- v^ Fmin. = 0.

In the first place, it is seen that we always find two values of Fmin. , one which is smaller
than W and one which is greater, but only the former is of interest. It is also seen that
the velocity of the current at the surface approaches the wind velocity asymptotically
when the depth increases towards infinity.

In order to show the surface velocities which may exist under the above conditions,
numerical values have been computed. The values of i',, at different wind velocities are
those which have been given by W. Schmidt. We find the following minimum values of
the surface velocity (in cm./sec.) at given wind velocities and at given values of the
depth of the channel :

Depth m.

Wind velocity, cm./sec.
5 X 10^ 10 X 10^ I

5 X 10^

Coeff
no

icient of eddy vise
430

;osity

950

5°

100

500

1000

5000

32

57

180

223

346

38

66

206

332

599

40
69

259
404

808

Hence, if the wind velocity is lo m./sec. we should obtain a surface velocity of at least
3-32 m./sec. if the depth were looo m. and of 5-99 m./sec. if the depth were 5000 m.
These results are quite unreasonable, and the obvious conclusion is that stationary
currents, which are due to the effect of the wind on the distribution of density, do not
exist in the oceans.

Ekman has already drawn this conclusion, and the above computation has only been
made in order to emphasize a well-known feature. Ekman has, furthermore, pointed
out that stationary currents can exist only if they are directed along the parallels of
latitude and when the depth is constant. The latter conditions are never fulfilled in the
sea and, therefore, no stationary currents can exist.

The circumstance which must be emphasized here, however, is that a pure drift
current must be present unless :

dv' y

dz

where v' represents the velocity of the convection current. The slope current is in-
dependent of depth and need not be considered in this connection. This condition
leads to a decrease of the velocity near the surface of an order of magnitude which is

148 DISCOVERY REPORTS

never observed in the open ocean, where the decrease very often is small in the surface
layers and greater at some intermediate depth. Hence, a pure drift current must always
be present when a wind blows, and within this a transport of water takes place, which
is directed 90° cum sole from the direction of the wind. Fjeldstad (1930) has shown that
the total transport can be derived from the wind resultant and, therefore, all the irregu-
larities in the wind systems need not be considered. Because of this transport vertical
circulations must be developed, partly because of the limitations of the oceans and
partly because of convergences and divergences in the wind systems.

The vertical circulations, which are maintained by the wind systems, would lead to
changes in the distribution of density, which, as previously shown, would further develop
the solenoid field which the wind builds up, and thus further increase the velocity of the
convection currents. Since the distribution of density in the oceans appears to have a
stationary character, it follows that the eff'ect of the vertical circulation, which is caused
by the wind, must be counteracted by other factors which influence the distribution of
density. The great importance of these other factors should thus be to maintain a certain
distribution of density and, therefore, when considering the total development of the
ocean currents, they must be given the same weight as the effect of the wind. These
factors also contribute to the development of vertical circulations, especially because
cold and heavy water sinks in high latitudes.

It has been shown that vertical circulations probably exist because of the action of the
wind, and several systems of this nature are well known. The so-called "upwelling"
along the west coasts of Africa and America result from the action of the wind. The
convection currents along these coasts are directed such that heavy water is accumulated
along the coast but no vertical component would be present if the prevailing wind had
not a component along the coast, which caused a transport of the surface layers away
from the coast.

Little attention has been paid to other systems of vertical circulation, but a very
interesting example is found within the current around the Antarctic Continent. This
current carries two typically different water masses, the Antarctic and the sub-Antarctic
water (Clowes, 1933). These two water masses are separated by a boundary surface
which at sea-level is recognized as the Antarctic Convergence or the Antarctic Ocean
Polar Front in the terminology of the German oceanographers. The boundary surface
between the two water masses takes in this case the place of the coast and makes possible
a vertical circulation within the Antarctic water. Before dealing with this circulation it
is necessary to consider the Antarctic Convergence more closely.

Fig. I gives in the lower part a section of the density of the water along the meridian
30° W according to the Discovery observations. The boundary surface between the
Antarctic and the sub-Antarctic water masses is indicated by a dashed line, and this
boundary surface reaches sea-level in about latitude 50° S, where the Antarctic Con-
vergence is situated. It is seen that the light sub-Antarctic water covers the heavy
Antarctic water in the form of a wedge. In the upper part of the figure the inclinations
of the sea surface and the 1000 decibar surface, relative to the 3000 decibar surface, are

VERTICAL CIRCULATION IN THE OCEAN

149

shown. Both surfaces are inchned from north to south and the current is directed
towards the east both within the sub-Antarctic and the Antarctic water. The incHnation
is much greater to the north of the convergence and, therefore, the easterly current is
much stronger on the northern side. These features are in agreement with the theoretical
conclusions as to the character of the distribution of density and the flow of the water
within two neighbouring water masses of different density (see Defant, 1929, p. 103),
but they do not explain why such a boundary surface exists within the easterly circum-
polar current. It seems obvious that no discontinuity in the variation of the density would

Fig. I. Vertical distribution of density (a,) and inclination of the sea surface and the

1000 decibar surface, relative to the 3000 decibar surface, between latitudes 42° and 58° S

in the Atlantic Ocean along the meridian 30° W.

exist if the direction of the current were due east in all latitudes and the occurrence of the
convergence must, therefore, be associated with the presence of north-south components.

Two factors may contribute to the development of a component towards the north
within the Antarctic water: in the first place the prevailing westerly winds, and in the
second place the supply of water of low salinity owing to the melting of the pack-ice.
The pack-ice is, however, carried towards the north by the wind, and thus the action
of the wind seems to be the principal factor which causes a flow of the surface layers to
the north.

Westerly winds prevail within the regions of both the sub-Antarctic and the Antarctic
waters. The wind is not blowing permanently from the west, but when considering the
effect of the wind it is, as already mentioned, sufficient to take the wind resultant into
account. Between latitudes 40 and 65° S the wind resultant is generally from the west,
and between these latitudes a pure drift current must exist which transports the surface
layers towards the north. The question arises whether the wind resultant from the west

ISO DISCOVERY REPORTS

and, therefore, the transport varies with latitude. The wind observations from the
southern part of the South Atlantic are too scanty to allow a definite answer, but the
available data point in the direction that in summer the strongest westerly winds are
met with between 50 and 60° S. This is, for instance, evident from Koppen's chart
which has been reproduced in Shaw's Maniiol of Meteorology, Vol. 11. This would mean
that we have a great transport of water to the north between 60 and 50° S and a smaller
transport north of 50° S and that, therefore, a convergence in the northward transport
exists at about 50° S. This convergence must take place over a wide belt, since the change
in the wind resultant must be gradual, but it may give rise to the development of a
discontinuity in the ocean. Two factors may contribute to such a development: the
tendency to the conservation of the angular momentum, and the tendency to a minimum
consumption of eddy energy.

Thus, it seems possible that the Antarctic Convergence in the South Atlantic Ocean,
which in summer is found in about 50° S, is due to a convergence in the pure drift
current. Similar conditions may exist in the southern Indian and Pacific Oceans, but
in the Drake Passage, where the convergence is met with in about 60° S, the location
must depend upon other circumstances. This great displacement to the south may be
due to the fact that the easterly current, which carries both the sub-Antarctic and the
Antarctic waters, is forced through the relatively narrow passage.

All these questions deserve a closer examination, but for the present purpose it is
sufficient to state that the convergence exists and that within the Antarctic water the
surface layers must be transported to the north by a pure drift current. Since the
westerly winds are dominant around the Antarctic Continent, the transport must take
place in all longitudes. If, however, this is the case, a compensation current must carry
water back to the south at some deeper level. The level at which this return flow takes
place can be found only by examining the hydrographic conditions in the area in question.

An examination can now be undertaken thanks to the work of the Discovery Ex-
peditions which have accumulated a great number of observations from the Drake
Passage, the Scotia Sea and the South Atlantic Ocean. By means of the published data
(1930, 1932) it is possible to construct vertical sections in these areas, and through the
courtesy of Dr Kemp unpublished observations from a section along meridian 75° W,
in the Pacific Ocean, have also been placed at my disposal. Six temperature and salinity
sections have been constructed, and in addition sections showing the vertical distribution
of oxygen and phosphate have been prepared when observations have been available.
The positions of the sections are shown in Fig. 2 in which they are numbered I-VI.

In considering these sections we shall begin with the most westerly, section I (Figs. 3
and 4) along the meridian 75° W in the Pacific Ocean, to the west of Drake Passage.
The observations were taken by the 'Discovery II' in November, 1931. The Antarctic
Convergence is found at the surface in about 60° S. The downwards bend of the iso-
therms to the north of 60° S indicates sinking motion of the water, and the salinity
section shows that there the origin of the Antarctic intermediate current must be sought.
From the salinity diagram it seems as if the water which has left the surface continues

VERTICAL CIRCULATION IN THE OCEAN

151

to the north, but the temperature section gives a different picture. North of the con-
vergence the isotherms bend towards the south and a tongue of water with temperature
higher than 2° extends past latitude 64° S between the depths of 400 and 1 500 m. This
distribution of temperatures indicates that between these depths the current has a com-
ponent which carries back to the south part of the water which has been transported to
the north by the pure drift current. To the north of the convergence and below the
Antarctic intermediate current, processes of mixing seem to take place, and through this
mixing the salinity of the sinking water increases and the temperature decreases. The
returning current, which is composed of Antarctic water and deep water, therefore
has a greater density than the water of the pure drift current, but before it again
reaches the surface it is diluted by melting water from the ice and by precipitation,
and thus a further development of the solenoid field is prevented.

Fig. 2. Chart showing the location of the vertical sections.

It seems, however, that only part of the water which has been transported to the north
returns again to the south. In order to complete the compensation, the deep water
below 1500 m. must also have a component towards the south, and finally the bottom
water appears to have a component to the north. Below the vertical circulation of the
upper layers, we, therefore, must have a similar circulation, which, when looking in the
direction of the current, rotates clockwise, while the circulation of the upper layers
rotates counter-clockwise. The possible circulations are indicated in the sections by
means of arrows. In this and the following sections the arrows were first plotted by
means of the temperature distribution and afterwards entered on the other diagrams.

The author is aware that great caution must be exercised when drawing conclusions
as to presence of currents from tongue-like distributions of oceanographic elements.
Later, other features will be discussed which confirm our conclusions, but now we shall
proceed with a discussion of the other vertical sections.

Figs. 5-10 show the sections II and III through the Drake Passage, one made by the

Fig. 3. Section I. Distribution of temperature (° C).

Fig. 4. Section I. Distribution of salinity (7oo)-

58" 60° 62°

WS-i06 WS 405 V.'S 4-04 WS 403 h'S 400 t/.S -'lOf ivS 400 IVS 399

Fig. 5. Section II. Distribution of temperature (° C).

WS 406

58° 60° 62°

WS 405 WS 404 WS 403 WS 402 WS 401 WS 400 WS 399

Fig. 6. Section II. Distribution of salinity (7oo)-

Fig. 7- Section II. Distribution of phosphate (P.^Os mg./ni.^).

Fig. 8. Section III. Distribution of temperature (° C).

Fig. 9- Section III. Distribution of salinity (7^ J.

Fig. 10. Section III. Distribution of phosphate (PjOg mg./m. 3).

IS6 DISCOVERY REPORTS

'William Scoresby ' at the end of February, 1929, and one made by the ' Discovery 11'
in the middle of April, 1930. In the two years the conditions were rather different, but
in both years the tongue of water with high temperature was present. In 1929 it reached
almost from the northern to the southern side of the Passage, but in 1930 it extended
only from the middle of the Passage to the southern side. In 1929 the tongue appeared
to be divided, indicating that the southward flow took place at approximately the levels
of 400 and 900 m., except in the most southerly region where the two branches appeared
to unite. In 1930 only one tongue with an axis at about 500 m. was present. In the
southern part of the Passage the 1° isotherm shows in both sections a downwards bend,
which perhaps indicates that part of the deep water which flows south turns back and
joins with the north-flowing bottom water.

In both sections the distribution of salinity indicates that on the northern side of the
convergence the intermediate Antarctic current flows with a northerly component at
depths between 400 and 700 m. The salinity of the north-flowing water is between
34-2 and 34-3 °/oo, while the surface saUnity of the upper layers is lower. The salinity of
the descending water must, therefore, have been increased by mixing with deep water,
and such an increase, owing to mixing, appears to have still more influenced the salinity
of the intermediate current which returns to the south. The water of this current has
a salinity of 34-5-34-6 °l^^. The mixing seems principally to take place at some depth
below the convergence.

Observations of phosphate are available from both years, and by means of these the
vertical sections in Figs. 7 and 10 have been constructed. These show that the maximum
phosphate values are found directly above or within the intermediate return current.
Both the vertical and the horizontal extensions of the phosphate maxima agree re-
markably well with the position of the current, and later on we shall discuss the
significance of this feature.

The next section, section IV (Figs. 11, 12 and 13), lies about 300 km. to the east of
sections II and III. The observations were taken by the 'Discovery II' in the middle
of March, 193 1 . The temperature section shows features which are quite similar to those
of the more western sections. To the north of the convergence we find a region of mixing
within which the descending current is divided into two branches, one which continues
to the north at depths of 400-700 m., carrying water of a temperature of about 4° and
salinity 34-2 °l'^^, and one which returns to the south at a depth of about 400 m., carrying
water of about 2° and 34-6 °l^^ . North of the convergence the region of mixing appears
to reach to a depth of about 1500 m. It is interesting to note that the temperature dis-
tribution in section II shows features which are intermediate between those of sections I
and IV.

Below the intermediate currents a vertical circulation of the deep water again appears
to be present. The downward bend of the 1° isotherm which was seen in sections II
and III is also conspicuous in section IV.

No phosphate observations were made at section IV, but observations of oxygen were
undertaken and they give a very interesting picture (Fig. 13). The Antarctic intermediate

VERTICAL CIRCULATION IN THE OCEAN iS7

water to the north of the convergence has a high oxygen content of nearly 6 c.c./litre.
On the northern side of the convergence the region of mixing is clearly seen, reaching to
about 1 500 m. The water of the intermediate return current contains much less oxygen,
i.e. less than 4 c.c./litre. Within the deep water we find a small oxygen content in the
northern part, but the ascending deep water contains more, while the north-flowing
bottom water shows the highest oxygen content.

Section V (Figs. 14-17) passes through the Scotia Sea. The observations were taken
by the ' Discovery II ' in the beginning of March, 193 1. Several features are similar to
those of the more western sections. Examining the temperature and saUnity sections
we recognize in the extreme northern part the Antarctic intermediate water of about
3° and 34-I-34-2 °/oo and, in the central part of the section, the intermediate return
current. This rises from about 1000 m. in latitude 54° 30' S to about 400 m. in latitude
58° S and carries water of about 2° and 34-5 °/oo- One new feature is seen. In the
southern part we find cold water of a temperature below 0° and of low salinity repre-
senting water from the Weddell Sea which has flowed over the ridge between the South
Shetland and the South Orkney Islands.

The phosphate and oxygen sections do not extend as far north as the temperature and
salinity sections because no observations of phosphate and oxygen were made at St. 653.
The incomplete sections show, nevertheless, features which are quite similar to those of
the sections farther to the west. The water of the intermediate return current shows
high phosphate and a low oxygen content. The cold water of the Weddell Sea is rich in
oxygen but contains less phosphate than the south-flowing water.

The last section, section VI (Figs. 18-21), follows the meridian of 30° W in the
Atlantic and reaches from about 58° S to almost 34° S. The observations were taken by
the 'Discovery 11' at the end of April, 1931. The distances between the stations are
great and, therefore, the horizontal scale has been reduced to one-half. This should be
borne in mind when studying the section.

The north-south extension of the section VI is so great that both the Antarctic and
the sub-tropical convergences are seen. The former is here situated slightly to the north
of 50° S, while the latter is found at about 40° S. To the north of the Antarctic Con-
vergence the sinking of the water and the northerly flow of the Antarctic intermediate
current are clearly seen from the salinity section. After having left the upper layers the
Antarctic intermediate water flows horizontally at a depth of about 400 m., but shortly
before reaching the region of the sub-tropical convergence it sinks further down, to
about 1000 m. in the northern part of the section. The salinity increases somewhat,
from less than 34-2 7„„ to nearly 34-3 7oo> but the temperature remains nearly constant
at 4°. The oxygen content is high, between 5 and 5-5 c.c./litre, but the phosphate con-
tent is relatively low and varies between no and 120 mg. PgOg/m.^ The temperature
section indicates as previously the existence of a warm intermediate return current to-
wards the south. This current begins in the region of mixing to the north of the Ant-
arctic Convergence and on its southwards course it rises from a depth of about 900 m.
in 46° S to about 400 m. in 56° S. The water of this current is rich in phosphate,

3-2

Fig. II. Section IV. Distribution of temperature ^ C).

Fig. 12. Section IV. Distribution of salinity (%o)-

VERTICAL CIRCULATION IN THE OCEAN

159

containing 140-150 mg. PgOs/m.^, but it contains relatively small amounts of oxygen.
The lowest oxygen values, 3-8 c.c. /litre, are found directly below the return current.

The deep water below the return current again appears to have a component towards
the south and shows an ascending motion, while the cold bottom water from the
Weddell Sea sinks and pushes to the north.

In the northern part of the section, features are met with which were not present in
the Pacific and in the Drake Passage. Below the sub-tropical convergence Atlantic deep
water with salinities above 34-8 °j^^ pushes to the south at depths between 2500 and

Fig. 13. Section IV. Distribution of oxygen (0^ c.c. /litre).

3000 m. This southward flow of deep water is also indicated by the bends of the iso-
therms at depths between 1500 and 2500 m. From the temperature and salinity sections
it appears as if the water of high temperature and relatively high salinity at 500 m. in
56° S represents the last traces of the Atlantic deep water, but it must, as pointed out
by Clowes (1933), be borne in mind that the general direction of the current in this
region is from west and that, therefore, the water in 56° S does not come from the Atlantic
but from the Pacific. A section farther to the east in the Atlantic might, on the other
hand, show traces of Atlantic water in high latitudes. Near Bouvet Island salinities of
3475 °/oo were observed below 600 m. (Discovery St. 453 in 54° 06' S and 04° E),
and this water probably represents water from the Pacific which has been mixed with
deep water from the Atlantic.

Fig. 14. Section V. Distribution of temperature (° C).

Fig. 15. Section V. Distribution of salinity (°/^ J.

Fig. i6. Section V. Distribution of phosphate (PjO^ mg./m.').

Fig. 17. Section V. Distribution of oxygen (O2 c.c./litre).

Fig. i8. Section VI. Distribution of temperature (° C).

661

Fig. 19. Section VI. Distribution of salinity (°/^j,).

Fig. 20. Section VI. Distribution of phosphate (P2O5 mg./m.^).

38" 40'

67J

66i

Fig. 21. Section VI. Distribution of oxygen (O^ c.c./litre).

164

DISCOVERY REPORTS

The sections which have been studied all show below the Antarctic water a tongue
of maximum temperature at intermediate depth, which has been interpreted as showing
a flow towards the south. It is, as already mentioned, necessary to exercise great caution
when drawing conclusions as to the existence of a horizontal current from tongue-like
distributions of oceanographic elements, but in this case the conclusion must be correct.
From the Discovery Reports it is evident that at all stations within the body of Antarctic
water an intermediate layer with water of high temperature is found, regardless of the
longitude of the station. Fig. 22 shows temperature-salinity diagrams from four stations.
Latitude and longitude of these stations are entered in the figure. Since all these stations
from widely different areas show a similar feature it must be concluded that the inter-
mediate layer with high temperature extends all around the Antarctic Continent. It

34 2

Fig. 22.

Temperature-salinity diagrams at four stations around the Antarctic Continent. The numbers at
the marks on the curves give the depths in hectometres.

follows that within this layer the current has a component from the north because the
maximum would soon disappear on account of mixing processes if no supply of warm
water from the north took place.

A tongue of water of low temperature is seen in all sections, stretching towards the
north at depths between 80 and 200 m. In considering this tongue it must be borne
in mind that all observations were taken in summer (November to April) and that
great seasonal variations occur in the upper layers. In winter an effective cooling from
above takes place and the water may become intermixed and attain a uniform density
to a depth of 100 m. or more, as evident from the observations at station 471 (54° 57' S,
27° 59I' W) on November ist, 1930, when the temperature in the layer o to 100 m.
varied between —1-62° and —1-77° and the salinity between 34-04°/^^ and 34"077oo-

VERTICAL CIRCULATION IN THE OCEAN 165

In winter the pure drift current may therefore reach to depths greater than 100 m. In
summer the temperature of the surface layers rises and the sahnity decreases because
of melting of ice and excess precipitation. The density therefore decreases and at a
depth of 40 to 80 m. a sudden increase in the density is developed. In summer the
lower limit of the pure drift current is found at this depth. During the summer the
temperature and the salinity of the water below 80 m. increase because of admixture
of the ascending water of the return current and the deep water. The observations
around South Georgia show that such an increase takes place as the season advances
and the feature can be explained only by admixture from below since the surface
salinity decreases. The ascending motion is, however, so slow that the layer of low
temperature is never completely removed in summer, but many stations are found at
which only traces of the cold water are present. The layer of cold water may, therefore,
be interpreted not as indicating a flow towards the north but as representing evidence
of the cooling in the preceding winter.

We can now summarize our conceptions as to the structure of the circumpolar
Antarctic current. Within the Antarctic water we find a vertical circulation which, when
looking towards the east, rotates counter-clockwise. The surface layers are carried to-
wards the north by the pure drift current. At the Antarctic Convergence part of the
water continues towards the north as one constituent of the Antarctic intermediate
current, but part is mixed with deep water and returns to the south as the warm
intermediate current within the Antarctic Zone. The surface layer of the Antarctic water
is very rich in plant and animal life. Within the Antarctic Zone the dead organisms sink
? and are decomposed and, therefore, the water of the intermediate return current shows
a high phosphate content. This phosphate is carried back to the south and can begin a
new cycle. The low oxygen values within or directly below the intermediate return
current indicate that oxidation of organic matter takes place.

Since only part of the water which flows north returns to the south it follows that the
motion of the deep water must also have a component to the south. The active cooling
of the water along the Antarctic Continent leads, on the other hand, to a sinking of the
water in high southerly latitudes, and the cold bottom water which is formed must flow
away from the Continent towards the north. Below the upper circulation we, therefore,
get a lower vertical circulation which rotates clockwise when looking towards the east.
These vertical circulations are superimposed on the general easterly current, and the
resulting motion has the character of complicated spirals. Within these spirals the
physical properties of the water are changed in the upper layers because of heating,
cooling, evaporation, precipitation or melting of ice, while in the lower layers they are
changed by admixture of water from other regions. Few water particles will describe
complete spirals. Some of the water, which at the surface moves north, will continue
towards the north within the Antarctic intermediate current, and only part will be
brought back again to the south by the intermediate return current. The water, which
in high southerly latitudes must be drawn to the surface in order to replace what is
carried north, consists partly of water from the intermediate return current and partly

4-2

t^

z66

DISCOVERY REPORTS

of deep water. One part of this deep water comes from more northerly latitudes, while
another part belongs to the lower vertical circulation of the Antarctic.

The north-south and vertical components of this complicated circulation are shown
schematically in Fig. 23. When studying this figure it must be remembered that the
components of the currents at right angles to the paper (away from the reader) are much
greater than any of the north-south or vertical components.

Our conception of the character of the circulation deviates considerably from that
commonly accepted. Merz and Wiist (1928) consider the relatively warm and saline
water at intermediate depth in high latitudes in the South Atlantic as the last traces of
Atlantic deep water, which in these regions approaches the surface. Clowes (1933) has,

A.C.

Sub antarcTic water,'
)

Antarctic water

Fig. 23. Schematic representation of the north-south circulations within the
Antarctic circumpolar current.

however, shown that this water is of Pacific origin, and according to our view it belongs
to the transversal circulation within the Antarctic Zone. It is true that the deep water
also moves south and ascends in high southerly latitudes, but traces of the Atlantic deep
water are not found in the South Georgia region but much farther to the east, since the
easterly flow of the water is considerable. In the Bouvet region the admixture of water
from lower latitudes leads to an increase of the sahnity within the intermediate return
current.

The conceptions of Merz and Wiist have been accepted by biologists who have studied
the conditions in Antarctic waters. Ruud (1932), for instance, states that the oxygen of
the relatively warm and saline layer is low, and says that " these water masses were last in
contact with the atmosphere somewhere north of the Sargasso Sea, so a very long time
has elapsed since they were aerated ". He says, furthermore, "that the surface layer (of

VERTICAL CIRCULATION IN THE OCEAN 167

the Antarctic water) receives its nutrient substances from the richer intermediate layer
of Atlantic origin ". Our sections show, on the other hand, that the nutrient substances
are carried back to the high southerly latitudes by the transversal circulation within the
Antarctic Zone, and that the low oxygen values of the intermediate layer are local and
must be due to processes within the transversal circulation.

An estimate of the velocity of the north-south circulation gives at present uncertain
results, but a few numerical values may, nevertheless, be communicated. Let us suppose
that the average wind velocity is about 10 m./sec. and that the tangential stress of the
wind is T = V2 x 10-^ W

n

where PFis measured in cm. /sec. We then find T = 3-2, and in latitude 55° S the total
transport of water towards the north through a vertical surface which is i cm. wide will be
S = 2-7 X 10* cm.^ In summer the thickness of the upper homogeneous layerwhich moves
under the action of the wind appears to be 40-80 m. Introducing 60 m. as a probable
mean value we find an average velocity towards the north of the pure drift current of
about 4-5 cm. /sec. If half of the return flow to the south takes place within an interval of
depth of 300 m. the velocity of the southwards flow will be about 0-5 cm. /sec, and the
average speed of the transversal circulation within the upper layer about 2-5 cm. /sec.

We have also other means of estimating the velocity within the intermediate return
current. If the temperature distribution within this current is stationary we must have

where the :x;-axis is placed in north-south and where A is the coefficient of eddy con-
ductivity and t the temperature. From the observations we find

-~2XI0-, g^~2XI0-B, ,~I,

or Vx ~ o-i^.

Our knowledge of the eddy conductivity in the sea is scanty, but at greater depths the
values of the coefficient appear to range between i and 20 (Helland-Hansen, 1930).
We, therefore, obtain o-i < ■y,. < 2-0.

This estimate only shows the order of magnitude of the component, and the agreement
with our preceding value is, therefore, quite satisfactory.

It is also of interest to examine the time which one water particle would need for a
complete transversal circulation in the upper layers. Within the Drake Passage and the
Scotia Sea the north-south extension of the transversal circulation appears to be about
300 km., and thus the total distance which a particle would travel is about 600 km.
Supposing the average speed to be 2-5 cm. /sec. and neglecting the time which is needed
during vertical displacements, we find that the particle returns to its most southern
position after 2-4 x 10' sec. or nearly 275 days. The time needed for vertical displace-
ments can also be estimated. Suppose that the volume of water which is transported to
the north through a vertical surface i cm. wide, 2-7 x 10* cm.^/sec, is replaced by water
which is drawn towards the surface within a belt, which is 100 km. wide. The velocity

i68 DISCOVERY REPORTS

of the vertical component will then be only 0-0027 cm./sec, and a water particle needs
80 days in order to cover a vertical distance of 200 m. The total time used for a complete
circuit in the region to the east of Drake Passage is, therefore, at least one year.

The transversal circulation is no doubt more rapid within the Drake Passage and the
Scotia Sea than in any other region because the Antarctic water is there pressed to-
gether. Comparisons of the sections in the Drake Passage with those in the Pacific and
the Atlantic show this clearly. This feature is perhaps of considerable importance to
several of the biological problems of the South Atlantic. It is possible that the surface
water to the east of South Georgia left the surface in the Drake Passage or the Scotia
Sea in the preceding year, while water masses which are drawn to the surface in the
Pacific Antarctic Ocean have spent years at intermediate depths after they left the sur-
face. This difference in the history of the water masses may in part explain the difference
in the development of plant and animal life in the two regions.

Until now we have disregarded the circulation which, within wide areas, takes place
to the south of the Antarctic current, along the Antarctic Continent. Near the Continent
the wind direction is easterly, and in agreement with this direction we find currents to-
wards the west in high southerly latitudes. Several whirls of stationary character appear
to exist, and these are present partly because of the prevailing winds and partly because
of the bathymetric features.

Within the westerly currents along the Continent we must also find a vertical circula-
tion which is similar to the vertical circulation within the Antarctic easterly current.
The surface layers are transported to the south by the pure drift current which is upheld
bv the prevailing easterly winds, and as compensation the water must flow north at some
lower depth. We know that the bottom water flows towards the north, and within the
Weddell Sea the water at a depth of about 500 m. also appears to have a component to
the north. The latter current represents, however, not a direct return of surface water
because it has a higher temperature than any other water mass in high latitudes. It must
have been drawn south at an intermediate depth within some of the whirls.

In this connection it is of interest to draw attention to the following feature : The light
water which is carried by the wind towards the shelf of the Antarctic Continent is
effectively cooled on the shelf. The tendency of the wind is to accumulate light water
along the coast and thus build up a stronger and stronger solenoid field, but this
tendency is effectively counteracted by the cooling of the water. Here we thus have an
example showing how factors which influence the density of the water prevent the de-
velopment of a strong solenoid field under the action of the wind.

In the preceding part of this paper it has been attempted to give a picture of the
vertical circulations within the Antarctic circumpolar current, starting out from general
considerations as to the eftect of the wind. The observations of the Discovery Expedi-
tions indicate that circulations of the supposed character are present, but many questions
still remain open and in conclusion some of these will be pointed out. The existence of the
Antarctic Convergence and the character of the movement of the water near the con-
vergence must be explained. In connection with the latter question one difficulty will

VERTICAL CIRCULATION IN THE OCEAN 169

be mentioned. Since the general wind direction is westerly on both sides of the con-
vergence a transport to the north must take place on both sides. We have assumed that
the transport is greater on the southern side and that, therefore, part of the water must
sink near the convergence. It seems to follow that another part continues across the
convergence, but the abrupt change in temperature when crossing the convergence is
decidedly against such a view. The convergence seems to represent a narrow zone of
mixing, but no continuous flow across takes place. A possible motion near the con-
vergence is indicated in Fig. 23 by dashed lines. It is supposed that a return flow to-
wards the south exists also on the northern side of the convergence, but this return
flow takes place near the surface and at the convergence part of the water bends up
towards the surface and is carried north, while another part bends down and joins with
the Antarctic water and forms one constituent of the Antarctic intermediate current.

Another difficulty is met with in the southern part of the Antarctic current where re-
turning water is supposed to reach the surface again. The observations show that the
water of the intermediate return current approaches the surface in high southerly
latitudes, but at no station is water of this type met with above 150-200 m. If our con-
ception is correct, it must be assumed that the factors which influence the temperature
and the salinity of the upper layers are so efi'ective that the properties of the inter-
mediate return water are changed when this approaches the surface. This assumption
is not unreasonable, since the ascending motion is very slow.

The questions which have been mentioned here can perhaps be cleared up, partly by
means of theoretical considerations, partly by means of new observations or by a closer
study of the existing data. At present the above representation of the vertical circulation
within the Antarctic circumpolar current must be regarded as hypothetical, but it is
hoped that the consideration of vertical circulations which are maintained by the action
of the prevaiUng wind may be helpful when studying the structure of other currents.

REFERENCES

Avers, Henry G., 1927. A study of the variation of mean sea-level from a level surface. Bulletin of the

National Research Council, No. 61, Washington, D.C.
Clowes, A. J., 1933. Influence of the Pacific on the circulation in the south-west Atlantic Ocean. Nature,

Vol. cxxxi, February 11.
Defant, a., 1929. Dynamische Ozeanographie. Naturwissenschaftliche Monographien und Lehrbiicher,

neunter Band, in, Berlin.
Discovery Reports, 1930. Vol. in, Station List, 1927-9, Cambridge.
Discovery Reports, 1932. Vol. iv, Station List, 1929-31, Cambridge.
Ekman, V. Walfrid, 1928. A survey of some theoretical investigations on ocean-currents. Journal du Conseil

Internationale pour Texploration de la mer, vol. iii, Copenhagen.
FjELDSTAD, J. E., 1930. Ein Problem aus der Winds tromtheorie. Zeitschrift fiir angewandte Mathematik und

Mechanik, B. x.
Helland-Hansen, B., 1930. Physical Oceanography and Meteorology. Report on the scientific results of the

'Michael Sars' North Atlantic Deep Sea Expedition, 1910, vol. i, Bergen.
RuuD, JoHAN T., 1932. On the biology of southern Euphausiidae. Hvalradets Skrifter, Scientific results of

biological research, No. 2, Oslo.
WiJST, G., 1924. Florida und AntiUcnstrom. Veroff. Inst, fur Meereskunde, Heft 12, Berlin.
WiJST, G., 1928. Der Ursprung der Atlantischen Tiefenwdsser . Sonderband der Z. Ges. Erdkunde zur

Hundertjahrfeier, Berlin.

[Discovery Reports. Vol. VII, pp. 171-238, Plates VIII-X, November, 1933.]

A GENERAL ACCOUNT OF THE HYDROLOGY
OF THE SOUTH ATLANTIC OCEAN

By

G. E. R. DEACON, B.Sc.

CONTENTS

Introduction P^g^ '73

The Surface Waters of the South Atlantic Ocean

Antarctic surface water

Nature of the Antarctic surface layer i73

Origin of the Antarctic surface water 1 79

Movements of the Antarctic surface water in the Falkland Sector . . 182

Extent of the Antarctic Zone 19°

Depth of the Antarctic surface layer 193

Temperature and salinity of the Antarctic surface water .... 194

Oxygen content of the Antarctic surface water 202

Phosphate and nitrate content of the Antarctic surface water ... 205

Sub-Antarctic water

Nature and depth of the sub- Antarctic surface layer, and the origin and

movements of sub-Antarctic water 206

Extent of the sub-Antarctic Zone 210

Temperature and salinity of the sub-Antarctic water 212

Oxygen content of the sub-Antarctic water 215

Phosphate and nitrate content of the sub-Antarctic water .... 215

Sub-tropical and tropical waters

Sub-tropical surface and under-layers, and the tropical surface layer . 216

Temperature and salinity of the sub-tropical and tropical waters . . 218

Oxygen content of the sub-tropical and tropical waters 219

Phosphate and nitrate content of the sub-tropical and tropical waters . 219

Extent of the sub-Tropical and Tropical Zones 220

The Deep Waters of the South Atlantic Ocean

Antarctic intermediate water

Structure and depth of the Antarctic intermediate layer, and the origin

and movements of Antarctic intermediate water 221

Temperature, salinity and oxygen content of the Antarctic intermediate

water 222

Warm deep water

Structure and depth of the warm deep layer, and the origin and move-
ments of the warm deep water, the North Atlantic deep water and
the Pacific deep water 226

Temperature, salinity and oxygen content of the warm deep water . 230

Antarctic bottom water

The Antarctic bottom layer, and the origin and movement of Antarctic

bottom water 231

Temperature, salinity and oxygen content of the Antarctic bottom water 233
Distribution of phosphate and nitrate in the deep waters of the South

Atlantic Ocean 233

Appendix. The winds of the Atlantic Ocean south of 40° S, by Lieut.

R. A. B. Ardley, R.N.R 235

List of Literature 237

A GENERAL ACCOUNT OF THE HYDROLOGY
OF THE SOUTH ATLANTIC OCEAN

By G. E. R. Deacon, b.Sc.
(Plates VIII-X; text-figs. 1-24.)

INTRODUCTION

OR some years past the Discovery Committee has been conducting a series of

F

observations on the hydrology and plankton of the South Atlantic and Southern
Oceans. During this period a large quantity of hydrological data has been accumulated,
and as soon as the analyses of the results are completed, detailed papers on difli'erent
aspects of the work will be published in this series of reports. In the meantime, and
because such knowledge is of importance to those who are studying the plankton, the
following general account of the hydrology of the area has been prepared.

THE SURFACE WATERS OF THE SOUTH
ATLANTIC OCEAN

An examination of the distribution of temperature and salinity in the surface layers of
the South Atlantic Ocean shows that there are four kinds of water at the surface. Each
of these waters has a typical range of temperature and salinity, and typical conditions for
the support of animal and plant life. The different bodies of water are closely dependent
on the climatic conditions of the regions in which they lie, and they owe their existence
partly to these conditions. The Antarctic climate in the south gives rise to a cold poorly
saline surface layer of Antarctic surface water, and the tropical conditions near the
Equator to a very warm surface layer of tropical water. Also, in addition to these two
extremes, there are two intermediate types, sub-Antarctic and sub-tropical waters,
which are affected less directly by Antarctic and tropical conditions. The geographical
distribution of the four kinds of water affords a very convenient and significant method
by which the surface of the South Atlantic Ocean can be divided into four zones, and a
description of the hydrological conditions in each zone is almost all that is necessary to
define the conditions over the whole ocean. The four zones have been called : Antarctic,
sub-Antarctic, sub-Tropical, and Tropical.

In this report an account is given of all four types of water, and of the geographical
limits in which they are found ; but the paper is concerned principally with the hydro-
logical conditions in the Antarctic and sub-Antarctic Zones.

ANTARCTIC SURFACE WATER

NATURE OF THE ANTARCTIC SURFACE LAYER
In the Antarctic Zone the surface layer is composed of cold poorly saline water, which
lies in a shallow well-defined layer above warmer deep water. It has a depth of 100-

174

DISCOVERY REPORTS

250 m., and is separated from the warm water below it by a discontinuity layer, within
which the temperature and salinity increase rapidly with depth. Figs, i and 2 show
the vertical distribution of temperature and salinity in a section through the sur-
face 1000 m. of water along longitude 80° W, from 60 to 67° S. The position of the

LATITUDE SI S

STATION 989

250i

500m

750m

Fig. I. Section I, distribution of temperature (° C).

LATITUDE 615

STATION 989

£50

500«

750

Fig. 2. Section I, distribution of salinity.

section is shown as section I in Fig. 1 1 (p. 191). The observations on which the sections
are based were made in October, when the hydrological conditions had hardly changed
from those of winter.

In Fig. I the Antarctic surface water is seen at the southern end of the section as a

HYDROLOGY OF THE SOUTH ATLANTIC

175

cold surface layer, in which the temperature is for the greater part less than - i° C.
The discontinuity layer is shown where the -i°, o°, and i° C. isotherms crowd close
together, and the depth of this layer below the surface increases gradually from about
100-150 m. in 67° S to 150-200 m. in 64'' S. Below the discontinuity layer lies the

LATITUDE
STATION 590

66°S
589

67°S

588

587

586

585

584

250 m

500m

750

I'l

Fig. 3. Section II, distribution of temperature (° C).

warm deep water. At each of the four stations made in Antarctic surface water the
temperature does not alter until at least 100 m. below the surface, and the temperature
of this loo-m. stratum at each station is shown in the section. The water is not coldest
in the extreme south, for there, as will be shown later, it is warmed slightly by up-
welling deep water. Instead, the temperature decreases from — 170° C. at the edge of

176

DISCOVERY REPORTS

the pack-ice to - I-86'' C. at St. 993 and then increases gradually towards the north.
Towards the northern limit of the Antarctic surface layer the increase in temperature
is more rapid, until in 62'' 37' S the convergence of Antarctic with sub-Antarctic
water is reached.

LATITUDE
STATION 590

3373%

66°5

5B9 588

67°S
567 585 585 584

332l%o 33l7%o

250 m

500m

750 m

Fig. 4. Section II, distribution of salinity.

In Fig. 2 the Antarctic surface water is seen as a poorly saline surface layer with a
salinity less than 347oo- Its salinity does not change with depth in the first 100 m.
below the surface. The salinity is greatest at the edge of the pack-ice and decreases
slowly towards the north. Figs, i and 2 together show that the Antarctic surface layer

HYDROLOGY OF THE SOUTH ATLANTIC

177

in winter is composed of almost homogeneous, cold, poorly saline water. There is prac-
tically no change in it with depth above the discontinuity level. The changes from
south to north are also very small, except in the neighbourhood of the convergence
with sub-Antarctic water.

In summer the conditions are quite different. Figs. 3 and 4 show the distribution
of temperature and salinity in a vertical section through the southern half of the
Antarctic surface layer in summer. The section was made along a line north-west of
Adelaide Island and its position is shown as section II in Fig. 11 (p. 191). Although
section II was made in shallower water than section I the positions are not very far

Om

-1"C

o°c

l"C

E°C

250.

500m

750m

I 000 m

AUGUST 1
I
1

..■•■'novembe

^

DECEMBER,^

'FEBRUARY

^'^^^r;:^^^^^

^^^^

i

^:

I;

li
1'

li

1;

ij

0-

It

13

1
ij

M

:i\

Fig. 5. The change in temperature with depth in the Antarctic Zone.

apart, and a comparison of Figs. 3 and 4 with Figs, i and 2 shows how the layer changes
from winter to summer. Fig. 3 shows that in summer the surface of the layer is warmed,
and the temperature no longer remains constant with depth down to 100 m. Instead,
the temperature decreases with depth until a level is reached in which it is not much
different from the value it would have in winter. In summer there are two strata in the
layer, a surface stratum which has been warmed by the greater absorption of heat from
radiation, and by conduction, and a cold stratum where the water has been warmed
hardly at all. The minimum temperature in the cold stratum at each station is shown in
the section. In winter the salinity of the surface layer will be greater in section II than
it is in section I, because section II is closer inshore and affected to a greater extent by

178

DISCOVERY REPORTS

upwelling deep water. In summer, however, as shown in Fig. 4, the surface salinity is
much less than it is in section I. At the surface the minimum salinity is found at St. 587,
which is directly in the path of the surface current from pack-ice which was lying to the
south-west. The surface stratum has been diluted with fresh water from mehing ice, and
by drainage from Antarctic land. The salinity of the water in the cold stratum is still
greater than it is at the same depth in section I : it has not been lessened by dilution
to anything Hke the same extent.

The changes which take place in the layer with the approach of summer are also shown
by diagrams illustrating the way in which the temperature, salinity, and density of the
water change with depth in different
seasons of the year. The four curves in
Fig. 5 show the change of temperature
with depth at a point 50 miles north of
Prince Olaf Harbour, South Georgia, in
August, November, December, and
January. They showthe changes which
take place in the Antarctic surface layer
not very far from its northern boundary .
In August the layer is cold and homo-
geneous, a condition which is typical
of winter. In November the surface
has been warmed, and so to a much
lesser extent has the cold stratum.
Warming goes on throughout the
summer, until in February the surface
is S'Q^C. warmer than it is in winter
and there has been a corresponding
increase of i-3°C. in the cold stratum.

Fig. 6 shows the changes in salinity
with depth at the same stations in
August, December, and February.
In August the salinity does not alter

with depth down to 100 m., and the layer has it greatest salinity. In December the
salinity is less, but it is still uniform down to 80 m. In February the layer has its least
salinity and there is a greater increase in salinity with depth. The changes in density
which take place in the layer are important because of their effect on vertical mixing
both in the layer itself and between the layer and the warm deep layer below it. Fig. 7
shows the vertical distribution of a, at the same stations ^

In August the surface layer has its greatest density. In December it is warmer, less
saline and lighter, and in February, when the effects of warming and dilution are

1 o-( = (Sf - i) 1000, where St is the specific gravity of the sea water at t° C. referred to distilled water
at 4° C.

3400%o 34S0%o

\ DECEd/lBER'!
'\ 1 IAUGU5T

FEBRUARY ^^^--ij,^

250 m

''^^

1

500m

750m

Fig. 6. The change in salinity with depth in
the Antarctic Zone.

HYDROLOGY OF THE SOUTH ATLANTIC

179

Om

2700

2750

greatest, its density is least. Since the density of the layer is greatest in winter and least
in summer, the discontinuity between it and the warm deep water (whose density
remains fairly constant) is in winter
least effective as a hindrance to vertical
mixing between the two layers, and in
summer most effective.

The conclusions drawn from the
changes in temperature and salinity
at these stations near the northern
boundary of Antarctic surface water
are in general the same as those drawn
from sections I and II farther south.
The surface of the layer is warmed
and diluted with fresh water in sum-
mer, and there is still a cold stratum
which the conditions are least

250k

500m

in

750-

lOOOi

FEBRUARY^.
\

iDECEMBERi
\. 1 AUGUST

\ ^ \

\\ \

\\ \
\\
V '

\

\
\

Fig. 7. The change in density (ctj) with depth
in the Antarctic Zone.

changed from those of winter. The

results show, however, that there is

an increase in vertical mixing from

south to north in the layer. Far south

there are sudden changes in salinity

with depth near the surface. Farther

north these changes are more gradual,

and although the surface stratum no

longer has such a low salinity there has been a greater reduction in the salinity of the

cold stratum.

ORIGIN OF THE ANTARCTIC SURFACE WATER

The Antarctic surface layer is bounded in the south by the Antarctic Continent, and
in the north by sub-Antarctic water along a convergence which has recently been shown
to be continuous throughout the Southern Ocean. North of 66° S the direction of move-
ment of the water in the surface of the layer is in general towards the north-east ; south
of 66° S it moves towards the west or south-west. The movement north of 66° S can be
considered as having two components, one to the east and the other to the north. The
easterly component will maintain a continuous current towards the east across the
Southern Ocean but the northerly component will move water northwards towards sub-
Antarctic water.

It will be shown later that along the Antarctic convergence, where the two waters
meet, Antarctic surface water sinks below sub-Antarctic water, and after a good deal of
mixing with it continues to flow northwards as a deep current. This water is lost from
the Antarctic surface layer, and must be replaced.

It has been found that the surface currents in the Antarctic Zone are much more

i8o DISCOVERY REPORTS

rapid in summer than in winter. Tlie principal evidence in support of this conclusion
is obtained from topographical maps, which show the height of the surface, calculated
from dynamic data, above some deeper isobaric surface which can be assumed to be
level. These maps indicate that the Antarctic surface currents in the Falkland Sector
are about twice as fast in summer as in winter. Further evidence, which is, however,
not conclusive, can be obtained from the movements of pack-ice. In the year 1930,
for example, there was no pack-ice north of South Georgia in August, September, or
October, and its appearance in November can be attributed to the speeding up of the
surface currents with the approach of summer. It would also be expected that the
Antarctic convergence would move northwards as the speed of the surface currents
increases; but the evidence we possess does not support this conclusion. Few obser-
vations have, however, been made in winter and those available are not sufficient to
disprove a seasonal movement of the convergence.

Although it is only in the Falkland Sector that there are sufficient data to examine
the seasonal changes in the speed of the surface currents and in the extent of the layer
it can safely be assumed that there is a greater production of Antarctic surface water
in summer than in winter. This increased production can be explained by the great
additions of fresh water which the layer receives in summer from melting ice and as
drainage from the land. In winter these sources are for the greater part stopped, and
the activity in the layer diminishes. The origin of Antarctic surface water does not lie
entirely in the addition of fresh water to the layer in summer, or to a lesser extent in
winter. There must be further additions to the layer from a different source to maintain
its salinity. It will be shown later that besides the transport of Antarctic water towards
the north in the surface there is also a flow of Antarctic water northwards in the bottom
layer. To compensate for the loss of water towards the north by these two currents,
which are known to exist all round the Antarctic Continent, there must be a
return current towards the south in the intermediate depths. Also, between the cold
Antarctic surface layer and the cold Antarctic bottom layer there is a maximum tem-
perature in the warm deep layer. This maximum temperature, in water which lies
between two bodies of colder water, can be maintained only by a movement of water
southwards in the warm deep layer. A second source of Antarctic surface water is
therefore to be expected in the warm deep water.

A certain amount of mixing must always take place between the two layers across the
discontinuity layer which separates them, especially in winter. It is principally to be
looked for in places where the density difference between the two layers is least, and in
places where for hydrodynamical reasons, based on the direction of winds, currents, or
the shape of land masses, the warm deep water is made to rise towards the surface.
Warm deep water has never itself been found at the surface, although it has been found
with its maximum temperature at a depth of only 100 m.: it is always covered with
Antarctic surface water. In section I the salinity of the water at the station made at the
edge of the pack-ice was greater than at those farther north ; its temperature was also
higher, and as will be shown in Fig. 18 (p. 204) its oxygen content was less. Each of these

HYDROLOGY OF THE SOUTH ATLANTIC i8i

facts is explained by a greater mixture of the surface water with the warmer, more
saline, and poorly oxygenated deep water. In section II there is shown a great up-
welling of deep water at the edge of the continental shelf. The effect of the upwelling
on the Antarctic surface layer is shown in the temperature and salinity sections, Figs.
3, 4, and also very plainly in Fig. 19 (p. 204), which shows the percentage of oxygen
saturation of the water in section II.

Fig. 17 (p. 203) shows the percentage of oxygen saturation of a surface stratum of
water 100 m. thick, in the Falkland Sector. It is a fairly accurate guide to the places
where deep water is upwelling. There are objections to its use, but it has been found that
where low percentages in the diagram show that poorly oxygenated deep water is up-
welling, the upwelling is confirmed by the temperature and salinity data. From the
figure it will be seen that the surface layer in the Falkland Sector receives most water
from the warm deep layer along the west coast of Graham Land and the South
Shetland Islands, and along that part of the ridge known as the Scotia Arc (Herdman,
1932, p. 214), which joins Joinville Island to the South Orkney Islands, and the South
Sandwich Islands.

There is also considerable upwelling in other places outside the scope of the diagram,
particularly in the centre of the cyclonic current system in the Weddell Sea. In
winter the effect of addition of water from the warm deep layer to the Antarctic surface
layer is partly the cause of the increasing salinity of the surface layer. In summer its
effect is exceeded by that of additions of fresh water, and the salinity of the surface
layer decreases. The relative importance of the two sources and their efl'ect on the
nutritive conditions in the surface layer has yet to be worked out.

The heavy precipitation in the Antarctic Zone is a third source of Antarctic surface
water. G. Schott (1926) shows that the annual precipitation in the West Atlantic Ocean
between 50 and 60° S is greater than 1000 mm. per annum, and this amount is about
700 mm. per annum greater than the amount of evaporation. A fourth influence, which
increases the salinity of the layer in winter, is the deposition of salt when sea ice is
formed. Neither of these last two factors is negligible and their importance is being
investigated.

The sources of Antarctic surface water can be summarized as follows :

(i) Fresh water from melting ice and snow.

(2) Warm water of high salinity from the warm deep layer.

(3) Fresh water from the excess of the precipitation over evaporation, partly allowed
for in (i).

(4) Salt which is deposited when sea ice is formed.

There are also several influences directly due to the Antarctic climate which are
necessary to maintain the layer. These are (i) conduction and radiation of heat into
the layer in summer, and out of it in winter, and (ii) strong vertical mixing due to
turbulent movement under the influence of strong winds. It is important to remember
that north of 66^ S the water in the layer is part of a continuous movement towards the
east, and that to the south of this latitude its movement has a westerly component.

i82 DISCOVERY REPORTS

The additions to the layer are made therefore to a water mass which is large enough,
and lasting enough, to show some resilience to changes. If Antarctic water flowed only
towards the north much greater seasonal changes would be expected.

MOVEMENTS OF THE ANTARCTIC SURFACE WATER IN
THE FALKLAxND SECTOR

Present knowledge of the surface currents in the Falkland Sector is based principally
on evidence obtained from charts showing the distribution of temperature and salinity,
and from charts showing the topography of the sea surface relative to a deep isobaric
surface, which because of the lesser movement at great depths has been assumed to be
horizontal. The general conclusions from these charts are shown as current arrows in
Fig. 8. The length of the arrows is intended to give an approximate idea of the relative
speed of the surface currents, but more precise information and topographical charts
will be given in a later report.

There are three principal factors which cause movement in the Antarctic surface
layer, and they can be summarized as follows :

(i) A thermohaline influence, which produces a vertical circulation in the ocean as
a result of density diff'erences maintained in different latitudes largely by diff"erent
climatic conditions.

(2) The efl^ect of the prevailing wind systems and pressure difl^erences.

(3) An eff'ect due to the excess of precipitation over evaporation, and to the liberation
of fresh water when ice and snow melt in summer.

The low air temperature and the small heating effect of the sun in the Antarctic
regions give rise to heavy water at the surface : this water, although it has a low salinity,
is still heavier than the more saline but warmer water in the tropical and sub-tropical
regions. The presence of this heavy surface water in the south is shown by the slope
of the isobaric surfaces in the sea downwards from south to north. The slopes of the
o, 300, 600, and 1000 decibar surfaces relative to the 3000 decibar surface are shown in
Fig. 9 (p. 184). Since the 3000 decibar surface can be regarded as almost horizontal
they may be taken to represent the actual slopes. The vertical scale is in "dynamic
centimetres," each of which equals 1-02 cm., and the figure shows the difference be-
tween the depths at which the isobaric surfaces were found and the depths of the hori-
zontal positions they would occupy in a uniform and motionless sea of 0° C. and 35 7 00
salinity. Because the water at the southern end of the section is heavier than that in
the north the same pressure is reached at lesser depths. The presence of heavier surface
water in the south is partly due to the presence of a movement towards the east, south
of 40° S, which is maintained by an external factor, the wind. In this movement the
layers of equal density will attain a more or less constant slope downwards to the north ;
but there is ample evidence to show that the slope is also the result of a continuous
meridional circulation.

u

-a

fc

3
o

T3

CO

M
E

1 84

DISCOVERY REPORTS

It has been shown by Jeffreys (1925) that if a difference in temperature is maintained
over any level surface within, or in contact with a fluid, the fluid will move, and continue
to move, until such difference of temperature ceases to exist. From the application of
this principle to the South Atlantic Ocean it is evident that there must be a movement
of Antarctic water to the north, and of warmer water to the south, in a continuous
circulation.

The circulation is not so simple as that described by Lenz (1847), who imagined a
symmetrical circulation in both hemispheres with water sinking near the poles and rising
towards the surface near the Equator. It is, on the contrary, the result of heavy Antarctic
surface and bottom waters sinking to different levels in the south, and of water of sub-
tropical origin sinking into the intermediate level from the North Atlantic. The circula-
tion will be discussed in detail in the section of the report dealing with the deep waters.

LATITUDE 35S

STATION 675 '

40 5

45°S

673

671

668

50"S
66G

663

55"S

G£l

lOCM

20 CM

1

_JliJ

Si

0- 0

tD

D

m ••

u

z
u

PS

2 Q

■5>

, - — ~'

^===5!—

^ //

^^

1 _^y

300 DECI BARS

^

/^'^

^

/-

SURFACE ■ —

---^

/

Fig. 9. The slope of the o, 300, 600, and looo decibar isobaric surfaces in 30° W.

For the present it can be seen from Plate VIII that the shape of the isotherms and iso-
halines in vertical sections which show the distribution of temperature and saHnity in
the South Atlantic Ocean indicates that both Antarctic surface water and sub-Antarctic
water sink towards the north. This is also seen in Plates IX and X which give the
vertical distribution of phosphate, nitrate, and oxygen content, and it is also indicated
by the presence of water farther north in the deep levels which can only have had its
origin south of the convergences. Such a movement of Antarctic surface water is ex-
plained by Jeffreys' work. Although unable to sink vertically because of the still heavier
water which is found below it, it can sink towards the north and still remain above the
heavier water. The south to north movements of Antarctic surface water can be explained
therefore by the effect of the thermohaline influence.

The greatest movement in the layer, except in some parts of the Falkland Sector, is,
however, probably to the east, north of 66° S, and to the west, south of 66° S, and
these movements can only be explained as due to the influence of the prevailing winds.

HYDROLOGY OF THE SOUTH ATLANTIC 185

North of 66° S the wind moves water northwards and assists the thermohahne influence.
It is considered by some to be the sole cause of the south to north movement and also
of the circulation in the ocean as a whole. Defant (1929, p. 133) says: "The thermo-
hahne circulation is therefore quite a surface phenomenon; the lower layers in the
ocean have no part in it ", and on p. 134, " It is not to be doubted that in contrast with
the thermohaline influences just mentioned, ocean currents produced by winds are
more important, and give rise to the final structure of current systems in the sea".
Defant also says on p. 136: "While acknowledging the winds to be the principal factor
giving rise to a horizontal circulation, we must not overlook the importance of density
differences as the origin of vertical circulation ". There seems to be some contradiction
between the three quotations, unless a circulation caused by density differences is dis-
tinct from a thermohaline circulation.

Our observations show that all the movement in the Antarctic surface layer cannot
be explained entirely as the result of wind currents although the zonal movements are
probably due almost entirely to wind forces. The south to north movements, particu-
larly below the surface stratum, are part of a continuous circulation, which is the result
of density differences. In the opinion of the author the density differences are largely
thermohaline differences which are the result of the different climates in different parts
of the ocean. The distribution of density may however be modified by the effects of
wind and of neighbouring land masses, which give rise to movements of water from
one climatic region to another. In the deep layers the density distribution is influenced
by the topography of the sea bottom : a movement of heavy deep water may be forced
to follow a particular channel, or it may be stopped by a ridge.

Before discussing the effect of wind on the movement of Antarctic surface water, it is
necessary to describe the general conclusions which Ekman (1928) has reached on the
effect of winds on the sea. Ekman distinguishes four types of current :

(i) A PURE DRIFT CURRENT caused by the effect of the resultant tangential force of the
wind on the surface of the sea. The velocity of the water at the surface is deduced on the
assumption that the sea is homogeneous, and that the coefficient of friction is constant,
to be directed at 45° cum sole^ from the direction of the wind. The direction of the surface
current turns uniformly cum sole with depth, and at the same time its velocity decreases
according to a logarithmic spiral until it is only one twenty-third of its surface value
when the direction of movement has turned through two right angles. The depth at
which this takes place has been called by Ekman "the depth of frictional influence",
since, if small velocities less than 5 per cent of the original one are neglected, it shows
the greatest depth to which horizontal motion can be transferred by friction. The depth

of frictional influence is written D, and according to Ekman D = / . , , where iv is

Vsm(/)

the speed of the wind in m./sec, and </> is the latitude. The formula, as pointed out by

Ekman, can only be regarded as approximate; it gives a depth of 80 m. for a wind of

^ Cum sole is the direction of the sun's apparent azimuthal motion — anticlockwise in the southern hemi-
sphere. Contra solein is the opposite direction — clockwise in the southern hemisphere.

i86 DISCOVERY REPORTS

about 20 knots in 60° S. Brennecke (1921) found by measurement a depth of 50 m. in
the Weddell Sea. The total transport of water in a pure drift current is directed at 90°
cum sole from the direction of the wind ; and if the wind is from the west in the southern
hemisphere, the total transport is towards the north. If the depth of the sea is less than
1-25 D, the angle between the surface current and the wind decreases; and when the
depth is o-i D, the surface current is in approximately the same direction as the wind.

(2) A SLOPE CURRENT is the direct result of the transport of surface water in the
pure drift current which will cause the level of the sea to rise at one place and fall at
another. To prevent this slope of the sea surface a deep current flows in a direction at
right angles am sole from the direction of the slope, i.e. in the direction of the wind.
The velocity of this slope current is constant with depth, if the sea is homogeneous and
friction constant, until a depth is reached where the frictional influence of the bottom
is felt.

(3) A BOTTOM CURRENT, which is the slope current modified by the effect of friction
with the sea-bottom. Below a height from the bottom which is analogous to the depth
of frictional influence at the surface— it may be called the depth of lower frictional
influence or the effective height of bottom friction — the slope current is slowed down by
friction and is turned coiitra soleni to flow in the direction of the surface slope. When
stationary conditions have been reached, the bottom flow in the direction of the slope is
all that compensates for the surface flow up the slope.

4. Convection currents. Ekman also recognizes the presence of currents, if the
water is not homogeneous, which are due to instability in the distribution in density.
Their speeds and directions depend on the angles and directions of slope of the surfaces
of equal density.

In the Antarctic surface layer the effect of the prevailing west wind north of 66° S is
to set up pure drift currents with surface velocities towards the north-east, and with a
total transport of water towards the north. As far as can be shown at present the depth
of frictional influence is about 60-80 m. The surface stratum of Antarctic water is very
often uniform to this depth, especially after a storm, and it will be shown in the section
on salinity and temperature of Antarctic surface water, that the changes with depth
below 80 m. are always much greater than those above it. In sub-Antarctic water, also,
there are changes at about 80 m. which are caused by a component of movement
southwards. It therefore seems probable that the effect of pure drift currents is confined
to the surface 80 m. of water, and the movement of the colder water at the bottom of the
layer will be part of the deep current. According to Ekman it will move as a slope
current or a convection current or a combination of the two. The total effect of the
prevailing westerly wind on the Antarctic surface layer will be to produce in the cold
stratum a current towards the east, and in the surface stratum a current which has at
each level the velocity of the pure drift current at that level superimposed on the easterly
movement of the cold stratum.

South of 66° S the prevailing wind is from the east and the direction of movement
will be towards the west at the bottom of the layer, and at the surface, south of west.

HYDROLOGY OF THE SOUTH ATLANTIC 187

These movements are known to exist and the wind is probably almost entirely re-
sponsible for them. At the northern end of Graham Land a little water flows westwards
against the wind, into the Bransfield Strait. This may be due to the eff'ect of the earth's
rotation on the Weddell Sea current flowing towards the north-east out of the Weddell
Sea, or to the presence of a counter-current between the Weddell Sea current and the
Bellingshausen Sea current which also flows towards the north-east outside the Weddell
Sea water. The effect is a very minor one compared with the easterly movement in the
same region.

There must, however, also be a northerly movement in the cold stratum of the
Antarctic surface layer, almost as strong as the northerly movement in the surface
stratum. This is shown by the amount of water sinking at the convergence which is
colder than the water in the surface stratum. Also, if the water found in the cold stratum
near the northern boundary of the zone were not continually renewed from the south,
the minimum temperature in the cold stratum would disappear, as a result of mixing
with the warmer water above and below it. Such a northerly movement can be explained
as part of the thermohaline circulation, and it is to this circulation, in all probability,
that north and south movements are mainly due.

Because Antarctic surface water sinks below the surface at the Antarctic convergence
there must be a greater movement towards the north in the Antarctic surface layer than
there is in the surface layer in the sub-Antarctic Zone. This is not the result of a dif-
ference in the strength of the pure drift currents in the two waters, since, as is shown by
Table XII (p. 236), the wind is strongest north of the convergence. The convergence
is, moreover, sharp and the Antarctic water sinks suddenly, and there is no corresponding
sudden change in wind velocity which could produce a sharp convergence. These are
facts which afford additional reasons for believing that the northerly movement is not
purely a wind current but a convection current caused by density differences.

The greater speed with which the surface water moves in summer cannot be explained
as the result of stronger winds in summer than in winter, since the winds are approxi-
mately of the same strength in both seasons (Schott, 1926, p. 224, and Ardley,
p. 235, infra). The increased speed can be explained by the thermohaline circulation,
in so far that there is a greater production in summer of surface water which is heavier
than that farther north ; but it is perhaps best explained as the result of a third factor
— the liberation of fresh water in summer. According to Schott the precipitation
between 50 and 60° S in the West Atlantic exceeds 1000 mm. per annum, and ac-
cording to Cherubim (1931, pp. 325-35) the amount of evaporation is just less than
300 mm. per annum.

Ekman (1926, p. 261) has studied theoretically the production of ocean currents in
the sea which would be caused directly by the difference in precipitation in different
regions in the ocean. He has shown that the greatest velocity of such currents is
1-2 cm./sec, and that they will probably be only a fraction of this. In the Southern
Ocean, however, the snow which falls far south does not melt until summer approaches,
and then it is released quickly. There are the additional effects of drainage from the

i88 DISCOVERY REPORTS

land, where the snow also melts in summer, and the melting of sea ice which has been
forming during winter and of land ice brought down to the sea by glaciers.

It is this fresh water, released in summer and mixed with Antarctic surface water,
which streams away from the ice and eventually flows north-east, speeding up the sur-
face movements. The eflFect of fresh water on the surface layer will be discussed in greater
detail in the section on the temperature and salinity of Antarctic surface water. The
direction of the current formed by the addition of fresh water to the Antarctic surface
layer will be away from the ice, so that accumulation of water near the ice is prevented.
Its direction will be modified by the effect of winds and the eflFect of the earth's rotation.
It will also be influenced by the shape of the land masses and the configuration of
the sea bottom.

There are two principal currents of Antarctic surface water in the Falkland Sector.
One of these flows towards the north-east out of the Bellingshausen Sea. It has been
shown that the westerly winds extend farther south in the Pacific than they do in either
the Atlantic or Indian Oceans, and south of 66° S there is still a surface current towards
the east. This current is turned towards the north by the west coast of Graham Land
and then flows towards the north-east through the- Drake Passage. The distribution
of temperature in vertical sections across the Drake Passage shows that the current is
strongest at the edge of the continental shelf, for it is there, oflF the South Shetland Islands ,
that the lowest temperatures are found.

In the western half of the Scotia Sea the direction of the current is largely influenced
by the shape of the bottom. Ekman (1928) has shown that in high latitudes the deep
currents in the sea tend to follow the lines of equal depth, and in the Southern Ocean this
is not only true of the deep currents but also to a certain extent of the surface currents,
to which the effect is transferred by friction or by the effect which a change in direction
of the deep current has on the density distribution. North of the Scotia Sea there is a
well-defined ridge, the Scotia Arc, extending from Cape Horn and the Burdwood Bank
to South Georgia. There is a gap in the ridge between 48 and 49° W^, where the water
is deeper than 3000 m., but on the ridge between Cape Horn and 50° W there is nowhere
a depth of water much greater than 1000 m. West of 50° W there is a trough of deeper
water north of the ridge. The Bellingshausen Sea current is prevented from flowing
northwards by the ridge : it follows the easterly direction of the ridge until the gap is
reached. At the gap it turns northwards, and then westwards into the deep trough north
of the ridge, as well as eastwards across the Atlantic Ocean. The changes in direction of
the surface current are reflected in the shape of the convergence in the neighbourhood
of 50° W, which is shown in Fig. 11 (p. 191).

East of 50° W the Bellingshausen Sea current north of the Scotia arc probably has a
small component southwards, towards the western end of South Georgia, since the
isotherms in the surface recede towards the south. Such a component might be the
result of the southward flow of water in the Brazil current, the effect of which on the
temperature distribution can be seen in Fig. 8 to the north-east of the Falkland Islands.

1 Herdman, 1932, pi. xlv.

HYDR0L0GY:0F THE SOUTH ATLANTIC 189

If it is caused by the Brazil current the strength of the southerly component will decrease
towards the south and Antarctic surface water will still sink below the surface at the
Antarctic convergence.

As it leaves the Drake Passage the Bellingshausen Sea current is joined on its right-
hand side by the Weddell Sea current. This current has its origin in water which has
drifted south and west across the Atlantic Ocean south of 66° S under the influence of
the easterly winds. This drift towards the west into the Weddell Sea, shown far south
in Fig. 8, is turned northwards at the east coast of Graham Land, and finally flows out
of the north of the sea past the South Orkney Islands. Part of it flows northwards towards
the Shag Rocks, but most of it turns more eastwards under the influence of the westerly
winds. A httle turns westwards round the northern end of Graham Land into the Bransfield
Strait, as has been mentioned earlier. The principal part of the current flowing east and
north-east spreads out across the eastern half of the Scotia Sea. East of South Georgia
it follows the lines of equal depth, and after crossing the ridge joining South Georgia
to the South Sandwich Islands, part turns towards the west along the edge of the con-
tinental shelf east of South Georgia. It meets the Bellingshausen Sea water, which
flows towards the western end of the island, and is turned back to the east. Then, with
the rest of the current, it flows eastwards across the Atlantic. Even as far east as the
longitude of the Cape of Good Hope Weddell Sea water can be distinguished on the
surface as a cold current by a minimum temperature between 55 and 56° S. North of
the minimum temperature is warmer water, which is the remains of the Bellingshausen
Sea water, and south of it warmer water which is falling away into the drift towards
the south-west.

Between the drift into the Weddell Sea and the Weddell Sea current flowing out of it,
there is a cyclonic water movement with a long zonal axis. The axis lies between about
63 and 65° S, probably nearer to 63° S, and it seems to be several degrees north of the
latitude in which easterly and westerly winds are equally prevalent. The path of the
cyclonic movement in the western part of the Weddell Sea is shown by the drifts of the
'Endurance', and the 'Deutschland', whilst they were beset in the ice, both of which
have been charted by Brennecke (1921, pi. ii). There seems to be only this one movement
in the South Atlantic and not the two cyclonic movements shown by Meyer (1923) in
40° W and 30° E. The centre of the cyclonic movement is also about 5 or 6° farther
north than shown by Meyer.

In Fig. II (p. 191) an attempt is made to show approximately the boundary between
the water from the Bellingshausen Sea and that from the Weddell Sea. It must be re-
membered, however, that along the boundary there is certain to be considerable mixing
between the two waters, since they are so similar, and the position of the boundary will
vary from the mean approximately indicated in Fig. 1 1 . Each season's observations must
be examined separately when a more accurate position is needed, and a position decided
from the small differences between the two waters which will be described later in this
report (p. 198). South Georgia lies in the path of both currents and is influenced to a
certain extent by both of them. The Bellingshausen Sea water is predominant west of

190

DISCOVERY REPORTS

the island, and the Weddell Sea water east of it. North of the island there is Weddell
Sea water with some Bellingshausen Sea water and south of the island a greater propor-
tion of Bellingshausen Sea water.

EXTENT OF THE ANTARCTIC ZONE

The Antarctic convergence, which is the northern limit of Antarctic surface water at
the surface, is usually sharp and well defined. It is distinguished by a sudden change
in the surface temperature of the sea when it is crossed. From south to north in winter
the temperature increases suddenly from about i to 3-5° C. and in summer from about
3 to 5-5° C. The continuous temperature record on the left-hand side of Fig. 10 shows

Fig. 10. Thermograph records showing the sudden change in temperature at the Antarctic (left) and sub-
tropical (right) convergences (vertical scale in ° C).

a drop of surface temperature from 3-2 to — 0-2' C, registered when the Antarctic
convergence was crossed from north to south: it represents a sharp convergence.
Occasionally, and especially if the convergence is crossed obliquely, we have found
that it is not straight, and the ship passes through cold and warm patches of water
alternately. This has been noticed particularly after bad weather, when the speed and
direction of the surface pure drift currents will have been varying considerably. It is,
however, found that there is always either a sharp convergence or these patches, and
never a gradual change from one kind of water to the other. The convergence is usually
not ver}^ well defined in the bend to the west in about 50° W, nor in the northerly bend
to the north of South Georgia, and in both places there is a tendency for sub-Antarctic

193 DISCOVERY REPORTS

surface water to move eastwards across the convergence. According to the evidence
which has been obtained so far, the position of the convergence does not vary very
much, certainly not more than 60 miles between its extreme positions, although the
temperature and salinity of the water sinking into the sub-Antarctic layer vary con-
siderably. The position of the Antarctic convergence shown in Fig. 1 1 has been obtained
from the position of the sudden change in surface temperature or from the position at
which the temperature minimum in the cold stratum of the Antarctic surface layer sinks
below 200 m. The two positions are usually about the same ; but in the two bends of the
convergence which have just been mentioned, a strong flow of sub- Antarctic water is
sometimes found at the surface, whilst the cold stratum of the Antarctic surface layer
has not sunk. The surface conditions in these places are thus sub-Antarctic, and are
probably only temporary, whilst only about 100 m. below the surface there is still
Antarctic water and its presence is permanent. The latitude of the Antarctic convergence
for longitudes between 80° W and 30° E is shown in the following table.

Table I

Latitude 80° W 75° W 70° W 65° W 60° W^ 55° W^ ^

Longitude of Antarctic convergence 62° 40' S 60° 30' S 59° 30' S 58° 40' S 58° 20' S 55° 50' S

Latitude 50° W 48° W 45° W 40° W 35° W 30° W^ ^

Longitude of Antarctic convergence 55° 50' S 55° 30' 50° 30' S 51° 00' S 49° 30' S 50° 10' S

53 10'

51 10

Latitude 25° W 20° W 15° W^ 10° W^ 5° W^ 0°

Longitude of Antarctic convergence 50° 30' S 49° 40' S 48° 40' 48° 00' 47° 40' 47° 40'

Latitude 5°E 10° E 15° E 20° E 25° E 30° E

Longitude of Antarctic convergence 48° 00' 48' 30' 48° 50' 49" 10' 49° 40' S 50° 10' S

The sudden rise in surface temperature was first noticed by Meinardus (1923, p. 531
et seq.), who explained the convergence as the line along which the ice water spreading
northwards sinks below the surface. He gave a table (p. 544) showing the latitude of
the convergence from 105° W to 80° E which is not much different from Table I. The
convergence was next described by Schott (1926, p. 241) in the South Atlantic Ocean,
who called it the " Meinardus Line ". It has since been described by Wiist (1928, p. 518)
and Defant (1928, p. 475). Both authors call it the " Oceanic Polar Front ", and base its
position on a chart of the surface currents in the Atlantic by H. H. F. Meyer (1923). They
give its position correctly between 50° W and 10° E, but between 40 and 60° W they
have confused it with the sub-tropical convergence between the Falkland current and
the Brazil current. We have used the name antarctic convergence, which means the
line of the convergence of Antarctic and sub-Antarctic waters.

The Antarctic Zone extends northwards from the Antarctic Continent to the Antarctic
convergence, and includes islands which have been considered from climatic reasons to
be sub-Antarctic. The only logical subdivision of the zone on hydrological grounds is
into two regions. In the first of these the surface water is part of the westerly drift south
of 66 ' S, or has its origin in this drift and is not far removed from it: in the second it

HYDROLOGY OF THE SOUTH ATLANTIC 193

is part of the drift towards the east in the region of westerly winds. In the Falkland
Sector this subdivision separates regions affected by Weddell Sea water and Bellings-
hausen Sea water. Over the whole zone, however, the layer of Antarctic surface water
is continuous: it is deepest in the north and shallowest in the south.

There is already considerable evidence of the importance of the Antarctic convergence
as a boundary in the distribution of marine plankton. There are some species which are
confined to waters to the south of it and others which are found only north of it. At least
one species has been shown to have different broods north and south of the convergence,
with only such mixing between the two as can be explained by some of the Antarctic
brood being carried northwards in the sinking Antarctic water, and some of the sub-
Antarctic brood being carried southwards in deeper water. The Antarctic convergence
is probably the extreme northerly limit of pack-ice ; but pack is rarely found so far north.
Ice will stop at the convergence because there is a smaller movement northwards in
sub-Antarctic water than there is in Antarctic water, and because of the sudden increase
of temperature.

Judged from a hydrological standpoint the Falkland Island Dependencies are all
Antarctic, although the Falkland Islands themselves are sub-Antarctic. Farther east
Bouvet Island, Heard Island, and MacDonald Island are Antarctic. Kerguelen lies just
on the convergence and there is a mixture of Antarctic surface water and sub-Antarctic
water near it. Marion and Prince Edward Islands, the Crozets and Possession Island,
are just north of the convergence, but so close to it that the Antarctic surface water has
not had time to sink far below the surface. When water upwells, as it will do particularly
on the north side of the islands, their hydrological conditions and marine life will be
influenced by Antarctic water, as well as by sub-Antarctic water.

DEPTH OF THE ANTARCTIC SURFACE LAYER

It has already been mentioned that between the Antarctic surface layer and the warm
deep water there is a well-defined discontinuity in the change of temperature and salinity
with depth. This indicates the presence of well-defined laj'ering which hinders vertical
mixing, but there must be a small amount of mixing taking place, especially in winter.
The level at which the temperature, salinity and density change most rapidly with depth
can be considered the boundary between the two layers. It has been found that the
changes are greatest about 40 m. below the depth of the minimum temperature of the
cold stratum of the Antarctic surface layer, and it is also at about this depth that the
temperature is the mean of the minimum of the cold stratum and the maximum of
the warm deep water. It is in this way that the depth of the layer has been measured.

The depth of the layer varies with the speed and direction of the surface currents . The
influence of the earth's rotation on the layers of equal density in a current is to make them
slope downwards to the left of the current in the southern hemisphere. The depth of the
discontinuity is therefore greater on the left-hand side of a current, and less on the right
and for this reason the depth of the layer changes within the zone. The effect is greatest in
the neighbourhood of land masses because the layers slope more steeply when the inflow

194 DISCOVERY REPORTS

of water laterally into the current is prevented. In the centre of a cyclonic movement,
clockwise in the southern hemisphere, the layer is shallowest and the deep water nearest
to the surface. In the centre of an anticyclonic movement the layer will be deepest.

Besides the variation in depth caused by the difference in the surface currents there
is a general increase from south to north from about lOO to 250 m., and this is probably
explained by the relative densities of the surface and deep layer and their movements.
Although the Antarctic surface water sinks suddenly at the convergence there is a
gradual sinking from south to north in the layer and also a rise in the deep water below
it in the opposite direction. There is also some variation in the depth of the layer as a
result of internal waves in the sea which have been shown to give a small vertical
oscillation to layers of equal density.

TEMPERATURE AND SALINITY OF THE ANTARCTIC SURFACE WATER

An examination of salinity sections, such as Fig. 4, which have been made near the
edge of the ice or near to the land in summer, shows how great is the influence of
additions of fresh water to the layer. At St. 587 (Fig. 4) there is a low salinity of 32-72700
at the surface, and still lower values are occasionally obtained closer to ice or land. At
most stations made far south in summer there is a surface stratum 20 or 30 m. thick
which contains such water with a very low salinity. A sharp discontinuity in the change
of salinity with depth separates this stratum from the rest of the Antarctic surface layer
below it, in which the conditions are not much different from those in winter. Later in
the season, or farther north, the sharp discontinuity becomes destroyed by vertical
mixing and the change in salinity from the surface stratum to the colder stratum is more
gradual.

In the south, at least in summer, the winds are not so strong as they are farther north
and the low salinity stratum is stable. Because of its stability and the absence of vertical
mixing between it and the water below, it does not pass on to the deeper water the heat
which it receives by radiation and conduction at the sea surface. The temperature of the
stratum therefore rises rapidly and a discontinuity also appears in the change of tem-
perature with depth — a factor which still further increases the stability of the stratum.
As water is added to the stratum it will flow away towards the north rather than sink :
a stream will flow away from the ice, or in the direction of the surface current which is
maintained by other additional influences.

On a very fine day, and especially in sheltered waters, a discontinuity may appear in the
surface layer due to the effect of temperature alone. Such a discontinuity is illustrated
by observations made at St. 606. At this station the temperature decreased rapidly
from 3*30° C. at the surface to i"84° C. at a depth of 5 m., and to i-oo° C. at 10 m.
The salinity at 5 m., 33*89 °/„„, was 0-02° j^^ greater than that at 10 m., but o-qi °j^^ less
than the surface salinity, which had probably been increased by evaporation. Owing
to the almost complete absorption of the sun's radiation near the surface there was a
difference of 2-30° C. in temperature, and 0-19 in a, , between the water at the surface and
that at a depth of 10 m., whilst the salinity was almost the same. Shallow discon-

HYDROLOGY OF THE SOUTH ATLANTIC 195

tinuities and very thin surface strata are often found on fine days in the vicinity of icebergs
or pieces of drift ice. Sudden jumps of as much as 4^ C. in surface temperature have
been recorded by the continuous thermograph when such patches have been crossed.
Such thin surface strata do not last, and are only found on calm sunny days. They break
up, and help to reduce the salinity and increase the temperature of the surface layer.

To show the distribution of temperature and salinity in the surface layer in the
Falkland Sector, charts have been prepared giving the average temperature and salinity
of a surface stratum 100 m. deep. They are based on a survey which was completed in
seven weeks, between October 25 and December 13, so that they are not appreci-
ably distorted by the changes which had been taking place whilst the work was being
carried out. The distribution of temperature and salinity in the Falkland Sector for
different seasons of the year and for different years will be published later, when allow-
ance has been made for such changes. The isotherms are shown in Fig. 12.

The cold water west of Graham Land belongs to the Bellingshausen Sea current.
On the northern side of the current the sea is warmer, and the temperature rises
slowly until there is a jump of about 3° C. at the Antarctic convergence. The path of the
current can be followed by the low temperatures to which it gives rise north of Elephant
Island. The cold water east of Graham Land belongs to the Weddell Sea current. The
temperatures observed in Weddell Sea water were not so low as those recorded in
Bellingshausen Sea water : in the Bellingshausen Sea the temperature at the edge of
the pack-ice was —1-63 to — i-89°C., and in the Weddell Sea — i-ig to — i-5i°C.
The northerly movement of Weddell Sea water is shown by the low temperature of the
eastern half of the Scotia Sea. The bend of the isotherms east of South Georgia shows
how Weddell Sea water flows round the north-east coast.

The shape of the i ° and 2° isotherms shows approximately the path of the Bellings-
hausen Sea water north of the Scotia Sea. Near the western end of South Georgia the
isotherms bend southwards, so that the south and west of the island are influenced by
this warmer water which has its origin in the Bellingshausen Sea and has been warmed
on its way northwards.

The Antarctic convergence in November follows approximately the 3^ C. isotherm
in the west ; it lies where the i °, 2°, and 3'' isotherms crowd together in the Drake Passage.
Farther north, where the convergence is not so sharp, it follows the 3-5° isotherm and
lies between the 3° and 4° isotherms. The total range of the average temperature of a
surface stratum 100 m. deep, which in November is almost uniform, is from — i-86° C.
in Bellingshausen Sea water to about 3-5° C. at the Antarctic convergence north of
South Georgia.

The salinity of the 100 m. surface layer is shown in Fig. 13. It is reproduced prin-
cipally to show the difference in salinity of the Bellingshausen and Weddell Sea waters.
The lowest salinities are those on the northern side of the Bellingshausen Sea current,
and the diagram indicates that a belt of water of low salinity exists just south of the
convergence. The salinity of the water in this belt of minimum salinity decreases from
about 34-847oo ^^ 8°° W to 33-797oo farther to the north-east, where the current has

-a
a

C3

V

o
o

3

o
>,

■§

CO

0)

4-2

ic,8 DISCOVERY REPORTS

received more additions of fresh water. In the eastern half of the Scotia Sea the saHnity
of the water is greater because it has its origin in the Weddell Sea. There is still a
minimum salinity in the east of the sector just south of the convergence ; but it is not
so well defined because the salinity of the Bellingshausen Sea water increases after it
reaches its minimum in the west of the Sea. Near South Georgia Bellingshausen Sea
water is warmer than Weddell Sea water, because it has travelled farther from its source
where it is as cold as, or even colder than, Weddell Sea water.

The greater salinity of the Weddell Sea water is probably due to the fact that whilst
it has been flowing round the Weddell Sea in the cyclonic movement described above,
it has received continual additions of water of greater salinity from the deep water,
which upwells in the centre of the movement. In addition, more ice will probably be
formed over Weddell Sea water in winter because the sea extends to such high latitudes,
and this too will increase the salinity of the water in winter.

To decide the origin of a body of water which is found near South Georgia it is best
to rely on charts showing the distribution of temperature and salinity based on data
collected at the time. These diagrams will show the greater influence of Bellingshausen
Sea water in the west and of Weddell Sea water in the east. The shape of the isotherms
and isohalines will show the direction of continuous movements. The lines will approach
close together where the boundary between the two waters is well defined and will be
farther apart when the waters are well mixed. The water of greater salinity, provided it
is continuous with similar water farther east, can be assumed to be Weddell Sea water
and the water of lesser saHnity to be Bellingshausen Sea water. Also, colder water will
be Weddell Sea water and warmer water Bellingshausen Sea water. Further evidence
can be obtained from a chart showing the topography of the sea surface based on cal-
culations of the depth of some deep isobaric surface, which can be considered to be
horizontal, below it. Such a chart also shows the directions of continuous movements,
for these follow approximately the contours on the chart and have velocities which are
inversely proportional to the distance of the contours apart.

Fig. 8, which has been used to show the direction of the surface currents in the
Falkland Sector, also shows the distribution of temperature over a wider area than that
covered by Fig. 12. The shapes of the isotherms are not quite so accurate, since the
observations on which the diagram is based were spread over a slightly longer time,
November lo-January 10; the surface also is more rapidly affected by changes due
to the approach of summer than is a surface layer 100 m. deep. The isotherms are not,
however, seriously distorted, and they give additional information about the Weddell
Sea current. They show the drift south and west into the Weddell Sea, as well as the
direction of the Weddell Sea current flowing out of the sea. The lowest surface tem-
perature found in Antarctic surface water was -i-68° C. in the middle of the Weddell
Sea, and the highest 3-5° C. at the convergence north of South Georgia.

Fig. 14 uses data, already employed in Fig. 5, to show the changes which take place
in the temperature of the surface layer during the year. It shows the changes at the
surface and at depths of 50, 100, and 150 m., and is based on Sts. WS 160, 268, and

HYDROLOGY OF THE SOUTH ATLANTIC

199

342 made in 1928-9: in the diagram there is also a curve which shows the surface
temperature changes at Grytviken, in South Georgia. At Grytviken the surface tempera-
ture is at a maximum in February and at a minimum in July. In the open sea the
maximum and minimum are reached a little later than they are in inshore waters, which
react more quickly to seasonal changes : the maximum temperature is probably at the
end of February or early in March, and the minimum in August. In summer the tem-
perature in the open sea decreases with depth from 2-45° C. at the surface to 0-17° C.
at a depth of 105 m. With the approach of winter the surface is cooled, and in August the

MARCH

JUNE

SEPTEMBER DECEMBER

Fig. 14. The seasonal changes in temperature of Antarctic surface water 50 miles
north of Prince Olaf Harbour, South Georgia.

surface was found to be colder than the rest of the layer. This is no doubt a temporary
phenomenon due to the presence of water of slightly less salinity in the surface, but it
shows that warming of the surface had not started at the end of August. Apart from the
surface 5 m. the layer is homogeneous down to a depth of 80 m., but it begins to lose its
homogeneous nature after November. The surface is warmed first, and in early summer
the difference in temperature between the water at the surface and at 50 m. is greatest.
By midsummer the layer is warmed down to a depth of about 80 m. and then there is a
rapid decrease in temperature from 1-89° C. at 80 m. to 0-31° C. at 90 m. The surface
80 m. is still almost uniform, because of the great amount of vertical mixing which takes
place in it during the turbulent movement of the pure drift currents, caused by the wind.

DISCOVERY REPORTS

The winter and summer temperatures, and the annual range of temperature, at the
different levels off South Georgia are shown in Table II.

Table II.

Depth
m.

Temperature ° C.

Winter

Summer

Annual range

Surface

50
100

ISO

— I -40

— I -06

— I -08
-0-15

2-45

2-l6

0-19
0-59

3-85

3-22
1-27
074

Fig. 15 shows the change of salinity during the year. The surface salinity is greatest
in September and least at the end of February. In summer it increases with depth,

MARCH

JUNE

SEPTEMBER DECEMBER

33-60 %o

3370%,

33-80%

33-90%

34-00%

34-10%

Fig. 15. The seasonal changes in salinity of Antarctic surface water 50 miles
north of Prince Olaf Harbour, South Georgia.

slowly in the first 50 m., and then more rapidly, especially below 80 m. The difference
in salinity between the water at the surface and at a depth of 100 m. is greatest in summer
and least in winter. At a depth of 150 m. the minimum salinity is found in winter. This

HYDROLOGY OF THE SOUTH ATLANTIC

201

is either because the layer is deeper in this particular region in winter, when the surface
currents are slower, or because there is a smaller tendency for deep water to upwell.

Fig. 1 6 shows the seasonal change in density in the surface layer. The Antarctic sur-
face water has its least density at about the end of February, and its greatest density in
August and September. In summer the density increases with depth, slowly in the
surface 50 m., and rapidly below 80 m.; in winter it is almost uniform down to 100 m.

MARCH

JUNE

SEPTEMBER DECEMBER

26' 80

2690

27 00

27' 10

27-20

2730

27-40

Fig. 1 6. The seasonal changes in density (a,) of Antarctic surface water 50 miles
north of Prince Olaf Harbour, South Georgia.

It can easily be seen how such uniformity in winter is brought about. In the two
previous diagrams it is shown that the surface 50 m. of water is practically uniform, even
in summer. Below 50 m. the changes are slightly greater; but the curves in Fig. 16
show how closely the water at 80 m. is related to the surface water in its properties,
and how great is the difference between the 0-80 m. stratum and the rest of the layer.
These facts have led to the assumption that the surface 80 m. is affected by pure drift

202 DISCOVERY REPORTS

currents, whilst the water below 80 m. is not — an assumption which has been confirmed
in sub- Antarctic water by the presence of a component of movement below 80 m. in
the opposite direction to the surface transport. With the approach of winter cooHng
commences in the surface, and its eff'ect is carried into the rest of the surface stratum
by the sinking of the surface water as its density increases, and by vertical mixing caused
by the turbulent movement of the pure drift currents. The density of the whole of the
surface stratum is increased in this way, and also by the inflow of heavier water from
the south, until it can hardly be distinguished from the cold stratum and the whole
layer becomes almost uniform. The density at 150 m. is least in winter, probably
because the layer is deeper at this season.

In Table III the changes in temperature of the Antarctic surface layer from south to
north are summarized.

Table III.

Depth

Temperature ° C.

In winter

In summer

At the surface

In the cold stratum

At the bottom of the layer

— 1-86 to I-o

— 1-86 to I-o

— 1-5 to I-o

- I-o to 3-5

- 175 to i-S

- 1-5 to 2-0

The salinity at the surface can vary enormously, from very low values in the south —
in summer when ice is melting — to as much as 34-5 °l^^ when ice is forming, or where
deep water is upwelling. In the layer as a whole the salinity decreases towards the north,
although at the surface the lowest salinities are found near the ice in summer.

OXYGEN CONTENT OF THE ANTARCTIC SURFACE WATER

The oxygen content of the surface 100 m. of water in the Falkland Sector is shown in
Fig. 17. The oxygen content is expressed as a percentage of the amount that the water
would hold if it were saturated. The diagram shows the conditions in spring before the
diatom growth was considerable. The smallest oxygen contents, 75-2 and 747 per cent,
were found off the west coast of Graham Land and in the Bransfield Strait ; low values
were also found over the ridge joining Graham Land to the South Orkney Islands and
to the South Sandwich Islands. It is probable that pack-ice was present in these
localities shortly before the observations were made, and also that the surface water
was largely mixed with deep water which had upwelled from below.

The oxygen content of the water increases from south to north. There is also a small
increase from west to east in water of the same temperature, about 5 per cent where it
is greatest. This may be due to a change which had taken place whilst the survey was
being made, or it may be due to a richer phytoplankton grov^th in the east than in the
west.

Fig. 18 shows the vertical distribution of oxygen content in Antarctic surface water

c

fa

J3

O
O

e

3

o

o

C

bo

&•
o

j3

o

p

60

ers

STATION 9B9

930

e.z'^

991

G4°5

992

65°S

993

BB"S

994

250«

500u

750*

1682%

90% :-

882%

lag 2%

i 68 5%

i885%

^// !b4o3

— jII^---'^°E^^::^

=^5

""l!!!-— — "

_^ — ^ —

^

^^-""""''^ 63%- ^(^ ^^

^

/

:::>

^y

50%

•

y

so/^

>50% 50%

LATITUDE
STATION 590

Fig. i8. Section I, distribution of oxygen content (percentage of saturation)

66°S
5B9 588 587

67 °S
586 585 584

Fig. 19. Section II, distribution of oxygen content (percentage of saturation).

HYDROLOGY OF THE SOUTH ATLANTIC 205

in winter along section I. It shows that apart from the station farthest south, which
is influenced by upwelhng deep water, the oxygen content of the surface 100 m. is
about 88-5 per cent in the south and 89-2 per cent near the convergence: there is also
very little difi'erence between the content of the surface water and that at 100 m. Fig. 9
shows the vertical distribution along section II in summer. The oxygen content at the
surface is greater than it is in winter ; it varies from 96-5 to 99-9 per cent, and it decreases
with depth. The content of the cold stratum in section II is less than that in section I,
but the difference is due to the greater upwelling of deep water in section II.

Near South Georgia, in January, surface water has been found to be supersaturated
with a content as great as 1 10 per cent. At the same time large catches of diatoms were
obtained, and the water had a high/)H value.

PHOSPHATE AND NITRATE CONTENT OF THE ANTARCTIC

SURFACE WATER

The amount of phosphate in Antarctic surface water is always large compared with
that in the surface of more temperate seas. In winter it has been found that the average
phosphate content of the surface stratum of 100 m., calculated from twenty-one stations
near South Georgia, was 138 mg. of PoOs/m.^ In the following January at a time when
large catches of diatoms were being made, the phosphate content was 1 10 mg. PaOg/m.*
These values are probably both rather high, as the samples on which the phosphate
determination was made had been stored for some time.

In January and February of 1930 the average content calculated from fifty-seven
stations was 82 mg. PoOj/m.^; and in November 1930 from forty-nine stations,
89 mg. PgOs/m.^ The minimum phosphate content of the sea water close inshore at
Grytviken has been found to be about 77 mg. PoOs/m.^' in December or January, and
the maximum to be about 130 mg. PaOg/m.^ in September or October. The inshore
results, however, are likely to be increased by the effect of the effluent from the neigh-
bouring whaling station, and decreased by the large amounts of fresh water, containmg
very little phosphate, which flow into Cumberland Bay. The lowest phosphate content
found in the open sea has been about 70 mg. PaOj/m.', and so many measurements have
been made that it can safely be said that the phosphate content of Antarctic surface
water in the open sea never falls below this figure. There is no evidence of the existence
of a secondary maximum of phosphate content in the autumn, such as is recorded m
the EngUsh Channel, and the 0-80 m. stratum is so well mixed that it is unlikely to
occur. Moreover, in the presence of such an abundance of phosphate a small increase in
autumn will not cause a second outburst of diatom production.

Antarctic surface water also contains large amounts of nitrate; the lowest nitrate
content found has been 290 mg. nitrate + nitrite N2/m.» in summer, and the highest
about 550 mg. in October-November.

The abundance of phosphate in the Antarctic surface layer is maintained by up-
welling from the warm deep layer. This layer in the Antarctic Zone contains at least
140 mg. PaOs/m.s The source of the high phosphate content in this layer will be

5-2

2o6 DISCOVERY REPORTS

described later. The high nitrate content of the Antarctic surface layer is probably
also maintained by upwelling.

SUB-ANTARCTIC WATER

NATURE AND DEPTH OF THE SUB-ANTARCTIC SURFACE LAYER, AND
THE ORIGIN AND MOVEMENTS OF SUB-ANTARCTIC WATER

The surface layer in the sub-Antarctic Zone is a much thicker layer than that in
the Antarctic Zone, and it is composed of much warmer water. Its depth increases from
south to north, and it is about five times as thick as the Antarctic surface layer. Its sur-
face temperature increases from south to north from about 3 to 11-5^ C. in winter, and
from 6 to 14-5° C. in summer. It hes entirely within the region of westerly winds, and
the water in it flows towards the east in a continuous movement around the Southern
Ocean.

Plate VIII shows the vertical distribution of temperature in the layer along 30° W,
where the sub-Antarctic Zone extends from 49° 30' to 40° 30' S. Between these lati-
tudes the temperature generally decreases with depth down to a level to which the
Antarctic surface water has sunk. This level is shown in Plate VIII by the bends in the
isotherms towards the north, which indicate the presence of a cold stratum of water con-
tinuous with the cold stratum of the Antarctic surface layer. An Antarctic layer will not,
however, be distinguished below sub-Antarctic water, since the two are so mixed that it
is not possible to decide where one ends and the other begins. The whole layer will be
called the sub-Antarctic surface layer, but it is useful to remember that the water in the
bottom of the layer is probably more Antarctic in origin than the rest.

Below the cold stratum, which contains the remains of the coldest Antarctic water,
the temperature rises owing to the presence of a warm deep layer. This is continuous
with the warm deep layer which lies below Antarctic surface water, although all the
warm deep water in it is not necessarily of the same origin. In the top of the warm
deep layer there is a secondary temperature maximum which is maintained by warm
water flowing southwards, and at the bottom of the sub- Antarctic layer a secondary
temperature minimum maintained by cold water flowing northwards. Between these
two levels lies the boundary between the two layers, and it is sufficiently accurate to
assume that it lies half way between them. In 30° W the depth of the sub-Antarctic
surface layer is about iioo m. in 45° S and 1450 m. in 40° 30' S.

Plate VIII also shows the vertical distribution of saHnity in 30° W. In the sub-
Antarctic Zone the change of salinity with depth is much less rapid than it is in the
Antarctic Zone, but there is still a discontinuity between the sub-Antarctic water and
the warm deep water. The discontinuity is not so well defined as that south of the
Antarctic convergence, and vertical mixing takes place across it to a greater extent,
particularly north of 45' S. Between the surface and 60-80 m. there is a transport to
the north as a result of the pure wind drift currents, and it is unusual to find any
appreciable change of salinity or temperature with depth. The stratum is uniform

HYDROLOGY OF THE SOUTH ATLANTIC 207

because of the vigorous mixing which takes place in the turbulent movements of the
drift currents. The total movement in the stratum will be the resultant of the pure
wind drift current and the deep current; and at the surface, it is shown by all the
available current charts to be a little north of east.

The winds of the Atlantic Ocean south of 40' S are described by Lieut. R. A.B. Ardley,
R.N.R., in the Appendix to this report. The resultant strength of the westerly winds—
their average strength multiplied by their frequency — in different latitudes is shown
in Table XII, column 4 (p. 236). This resultant is greatest between 45 and 50° S, and
the percentage of easterly winds is also least between these latitudes. The surface
transport towards the north due to wind influence will therefore be greatest between
45 and 50° S, and it will decrease towards the north. Because of this there will be a
tendency for the surface water to sink towards the north, and also, in the south of the
zone, for water to well up towards the surface. Below the surface stratum the move-
ment in the layer is towards the east in a deep current which is the result of the westerly
wind. But at the same time there are north and south components of movement, and
these are responsible for the maintenance of the temperature and salinity distributions
which are typical of the layer.

On the western side of the South Atlantic, and particularly outside a region of intense
mixing, which reaches about 100 miles north of the Antarctic convergence, there is an
increase in salinity with depth below about 80 m. Sometimes the temperature also
increases, or there is a decrease in the temperature-depth gradient. A stratum of water
of greater salinity is found between 80 and 100 m. at St. 668, in 46° 43' S ; and between
100 and 150 m. at St. 671 in 43° 08' S. Such a stratum can have a continued existence
between waters of lesser salinity only if the water in it has a component of movement
towards the south, at least relative to the waters above and below it; but it is not quite
certain how such a component can be explained. It may be a return current to com-
pensate for the flow of surface water towards the north in the drift current. This ex-
planation appears the more likely to be valid because in the South Atlantic Ocean
north of 45"' S the amount of northerly transport decreases towards the north, so that
the surface water must sink towards the north. Within 100 miles of the Antarctic con-
vergence there is not so much evidence of this southward movement below 80 m. ; this
may be due to the greater vertical mixing, or to a falling oft' towards the south of the
northerly transport which might cause the water of the return current to well up towards
the surface. On the western side of the South Atlantic, however, the stratum between
80 and 200 m. is more remarkable for its high salinity than its high temperature, and
it also has a low oxygen content. If therefore it is composed of water which has sunk
below the surface, it must be mixed with water from deeper sources. Just south of
the sub-tropical convergence it contains sub-tropical water, and the last traces of this
may reach to within 100-150 miles of the Antarctic convergence.

The presence of a southerly component can also be explained as the result of a small
thermohaline circulation, confined to the surface 200 m. of sub-Antarctic water and
due to the difference in temperature between the climate in the south and north of the

2o8

DISCOVERY REPORTS

zone. The southerly component has been found in the east of the South Atlantic Ocean,
but in the east of the Pacific Ocean it is only found near the sub-tropical convergence.
It is not confined to sub-Antarctic water, but also appears in sub-tropical water when
there is a prevailing westerly wind.

Below 200 m. the water in the sub- Antarctic layer has a component of movement
towards the north, except perhaps in the eastern corner of the Pacific Ocean. Near
Cape Horn, there is a general component in the layer towards the south, for in this
region part of the easterly current across the Pacific is deflected southwards to flow
through the Drake Passage.

Conditions in the southern and northern parts of the sub-Antarctic Zone differ
rather widely, and the two areas must be considered separately.

STATION 1021

200

W53I8 1 026

1022 I 1023 1024 |ipg5

1027

1028

400m

600m

800m

200
M1LE5 NORTH OF THE CONVERGENCE MILES SOUTH OF THE CONVERGENCE

Fig. 20. The depth of the minimum temperature in Antarctic surface water and sub-Antarctic
water, the depth of the minimum salinity of sub-Antarctic water.

In the Southern Half of the Zone in the region of intense vertical mixing extend-
ing about 100-150 miles north of the Antarctic convergence there are only small changes
from west to east. In the Pacific Ocean in 80° W the salinity of the layer is almost
uniform with depth down to about 300 or 400 m., or it decreases to a weak minimum
at that depth. The temperature is usually uniform down to 150 m. and then decreases
gradually. There is sometimes a minimum temperature at the same depth as the minimum
salinity, and both are caused by the greater percentage of Antarctic surface water in the
mixture of the two waters at that depth. Below the level of minimum salinity the salinity
increases, and there are only small temperature changes. At a depth of 600-800 m. the
temperature decreases to a minimum, and there is a stratum of comparatively uniform
salinity. This is composed of water which has sunk from the cold stratum of the Ant-
arctic surface layer. When the Antarctic surface water sinks below the surface, the
water in the cold stratum will be least mixed with sub-Antarctic water, because it is
much heavier. The warmed and diluted surface stratum of Antarctic surface water has,
however, a density not so different from sub-Antarctic water, and mixes more readily

HYDROLOGY OF THE SOUTH ATLANTIC 209

with it. So complete is the mixing in the region extending 100-150 miles north of the
convergence, that it is not often that the minimum saHnity at a depth of 300-400 m.
can be detected. In summer, or at the end of summer, vertical mixing is not so com-
plete, and a surface stratum is usually formed which is similar to that in the Antarctic
Zone but not nearly so well defined.

In the Drake Passage, where the components of movement perpendicular to the con-
vergence are restricted by the land masses to the north and south, neither the minimum
salinity nor the minimum temperature can be detected. Near Cape Horn the surface
is considerably diluted by coastal water.

At stations east of the Falkland Islands both the minimum salinity caused by the
Antarctic surface water, and also the minimum temperature caused by the cold stratum
of Antarctic surface water, are found. The depth of the minimum temperature is shown
in Fig. 20. South of the Antarctic convergence it shows the depth of the cold stratum of
Antarctic surface water, and north of the convergence the depth to which the water
from the cold stratum sinks. The depth of the minimum salinity north of the convergence
is also shown ; at this depth there is more water from the surface stratum mixed with
sub-Antarctic water.

In the bend of the Antarctic convergence to the east of the Falkland Islands, and
also in the bend north of South Georgia (Fig. 1 1) there appears to be the strongest flow
of Antarctic water below sub-Antarctic water and also of sub-Antarctic water eastwards
over Antarctic water. In both regions the convergence is never very well defined, and
it is the more difficult to fix with accuracy because the rise in surface temperature usually
occurs south or east of the position where the Antarctic surface water starts to sink
rapidly. A stratum of sub-Antarctic water is often found above a stratum of Antarctic
water which is still at its usual depth and conditions in these regions can change
suddenly with depth.

In the eastern half of the South Atlantic, particularly south of the Cape of Good
Hope, the movement at right angles to the convergence seems to be restricted, as a result
of a general southerly component of movement in sub-Antarctic water. Such a com-
ponent is probably caused by the sub-Antarctic water being driven southwards to round
the Cape and the southerly extension of sub-tropical water caused by the Agulhas
current. As a result the minimum salinity is not found except at a greater distance north
of the convergence, and the minimum temperature is not well defined.

The mixture of water which results from the addition of Antarctic surface water to the
sub-Antarctic water moves towards the east, and at the same time sinks towards the
north . Its path is shown by the depth of the levels of minimum temperature and minimum
salinity which can be followed a long way north in the Atlantic — the minimum salinity
as far north as 25° N. This movement will be described in the section on the Antarctic
intermediate water.

In the Northern Half of the Zone there is a surface stratum 60-80 m. thick in
which the saHnity and temperature are almost uniform with depth. Below this surface
stratum, the salinity increases until it reaches a secondary maximum value between 80

2IO DISCOVERY REPORTS

and 200 m., where there is a movement southwards. The temperature starts to fall
rapidly immediately below the surface stratum, but its decrease is arrested some-
what, and occasionally is changed to an increase between 80 and 200 m. Below this
stratum the temperature and salinity both decrease rapidly until more uniform water is
reached near the level of minimum salinity between 400 and 600 m. Their decrease is then
less rapid. Below the level of minimum salinity the salinity increases rapidly, but the
temperature continues to decrease slowly until the level of minimum temperature is
reached, between 800 and 1200 m. In this level, too, the change of saHnity with depth
becomes slightly less. Below the level of minimum temperature both the temperature
and salinity increase, the temperature to a secondary maximum value in the upper stratum
of the warm deep layer, and the salinity to a maximum deeper down in that layer.
The origins of sub-Antarctic water can be summarized as follows :
(i) In Antarctic surface water which sinks below the surface at the Antarctic con-
vergence; and, to a much smaller extent, in Antarctic surface water which mixes
across the Antarctic convergence at the surface.

(2) From the heavy precipitation in the sub-Antarctic Zone which exceeds the
amount of evaporation by about 700 mm. per annum, and also from coastal
waters.

(3) In additions from the sub-tropical surface layer between 80 and 200 m., and
also by mixing across the sub-tropical convergence at the surface.

The effect of mixing across the sub-tropical convergence will be described in the next
section.

EXTENT OF THE SUB-ANTARCTIC ZONE

The sub-tropical convergence, which is the northern boundary of the sub-Antarctic
Zone and the convergence of sub-Antarctic and sub-tropical waters, is not so well known
as the Antarctic convergence. It is, however, usually a much sharper convergence, and
is marked by a sudden change of surface temperature of at least 4° C, and a change of
salinity of at least 0-50 7oo- A continuous temperature record showing the sudden fall
in surface temperature on crossing the subtropical convergence is shown on the right-
hand side of Fig. 10 (p. 190). At the convergence sub- Antarctic water sinks below the
surface. It does not sink downwards to the north because between 80 and 200 m. sub-
tropical water is moving southwards into sub-Antarctic water. Instead, it must mix with
the water in this movement, and then, on its way southwards it either mixes again with
the deeper water moving northwards, or wells up towards the surface.

To explain the sharp convergence at the surface that is almost always encountered,
there must be a much greater movement northwards in the surface of sub-Antarctic
water than there is in sub-tropical water. The sub-tropical convergence is still within
the region of westerly winds, but since it is found in about 40° S the winds are much
weaker north of it. The convergence can therefore be explained as a result of the greater
transport of water towards the north in the surface stratum of sub-Antarctic water ; but
this reason is not sufficient to explain the sharpness of the convergence, since the falling
off of wind strength is gradual.

HYDROLOGY OF THE SOUTH ATLANTIC zii

South of the Brazil current and the Agulhas current there is a southward movement
at the surface which is probably stronger than the transport of water northwards due to
wind currents, and south of these two currents the convergence lies farthest south. In
30° W the salinity and temperature distribution show that there is a southward movement
below 80 m. : if such a movement also occurs at the surface it is reduced by the effect
of the wind. In mid-ocean the convergence is farther north; it is not so far removed
from the region of the south-east trade winds, which give rise to a water transport in the
surface towards the south-west.

When the sub-tropical convergence is crossed, and especially when it is crossed
obliquely, it has been noticed that there are large areas of sub-tropical water cut off by
sub-Antarctic water, and the ship passes through patches of sub-tropical and sub-
Antarctic water alternately. When a station is made in such a patch of sub-tropical
water, it is found to be lying above sub-Antarctic water in a layer about 200 m. deep,
but when the patches of sub-Antarctic water are examined they show the changes of
temperature and salinity with depth which are typical of the sub-Antarctic Zone. It is
therefore the sub-tropical water which pushes southwards over the convergence to cause
these isolated patches of water. These patches are formed so extensively south of the
Brazil current and the Agulhas current that it is impossible to fix the convergence
definitely; but an attempt has been made in Fig. 11 to show the convergence as the
northern extent of sub-Antarctic water. South of the Cape of Good Hope it is most
likely that sub-Antarctic water will not be encountered north of the position shown, but
south of the Brazil current there is not enough information to be certain even of this.
The latitude of the convergence between 50° W and 30° E, based on our own observa-
tions, is given in the following table.

Table IV

Longitude 5°° W 45° W 40° W^ 35° W 30° W^ 25° W^

Latitude of sub-tropical convergence 44° 30' S 43° 30' S 41° 30' S 40° S 40° 30' S 39° 30' S

Longitude 20° W 15° W 10° W 5° W 0° 5° E

Latitude of sub-tropical convergence 38° 30' S 37° 30' S 37° S 37° S 37° S 37° 30' S

Longitude 10° E 15° E 16° E 20° E 25° E ^ 30° E

Latitude of sub-tropical convergence 37° 30' S 37° 30' S 37° 30' S 43° S 43° 30' S 44° S

41° S

42° 30' S

When a station is made near the convergence, the salinity and temperature of the
water must be used to decide whether the water is sub-tropical or sub-Antarctic, or how
much the two waters are mixed.

The sub-Antarctic Zone includes Cape Horn, and the Falkland Islands, and as much
of the Patagonian coast as is influenced by the Falkland current which flows as far north
as the River Plate. Gough Island is sub-Antarctic, and Tristan da Cunha is just sub-
Antarctic. The Cape of Good Hope is north of the sub-tropical convergence, but the
cold water off the south-west coast of Africa is water which has upwelled from the Ant-
arctic intermediate layer into which sub-Antarctic water sinks. The conditions off the

212 DISCOVERY REPORTS

south-west coast are therefore partly sub-Antarctic as well as sub-tropical. Off the south
and east coasts the water is sub-tropical, and in summer it is almost tropical owing to
the influence of the Agulhas current.

TEMPERATURE AND SALINITY OF THE SUB-ANTARCTIC WATER

Both the temperature and salinity of the sub-Antarctic water increase towards the
north. The increase is greatest between 80 and 200 m., as would be expected if the
water in this stratum was flowing southwards as well as eastwards and losing its heat and
salt content by mixing. It is next greatest in the surface stratum, in which the water has
a component of movement to the north. The changes are least at the level of minimum
salinity, where a large volume of water is sinking towards the north ; and they are least
because the properties of a large volume do not change readily. At the level of minimum
temperature, and at the bottom of the layer, there is a considerable increase in salinity
towards the north because of the mixing which takes place with the warm deep layer.

The surface temperature of sub-Antarctic water just north of the Antarctic converg-
ence in the West Atlantic Ocean is about 3 to 3-5° C. in winter and 6° C. in summer. It
increases from south to north, slowly in the southern half of the zone and more rapidly
in the northern half, where vertical mixing is not so complete. Near the sub-tropical
convergence the surface temperature is about 1 1 -5° C. in winter and 14-5° C. in summer.
The salinity in the south of the zone varies between about 33-95 °/oo ^t the end of summer
and about 34* 10 °/oo ^t the end of winter. The salinity also increases towards the north.
Just south of the sub-tropical convergence it has a value of about 34-4700' t)ut it may be
much increased south of the Brazil current and south of the Agulhas current, by mixing
with sub-tropical water which has crossed the convergence at the surface. In the east
of the Atlantic Ocean the surface stratum has a lesser salinity than in the west, probably
because of the additions it has received from the Antarctic surface water as it crosses the
ocean.

Between 80 and 200 m. the changes are slightly greater, starting from about the same
values just north of the Antarctic convergence.

The very small changes in the temperature and salinity at the level of minimum
salinity are shown in Table XI (p. 223). Between 47° S and the sub-tropical convergence
there is only an increase of about 0-04 7oo in salinity and about 0-5° C. in temperature.

At the level of minimum temperature near the bottom of the layer there is a large
increase of salinity towards the north as the water mixes with warm deep water, and
there is also a small increase in temperature. Between 47 and 38° S there is an increase
in salinity from 34-23 to 34-53 7oo ^^d an increase in temperature from 2-36 to 2-76° C.
The increase in temperature is rapid as the Antarctic water sinks below the surface at the
convergence ; it then becomes slower and is approximately regular between 47 and 38° S.

The way in which the temperature and salinity change with depth has already been
described. In the south of the zone the difference in temperature between surface water
and that at the bottom of the layer is about 3° C. and near the sub-tropical convergence
about 12° C.

HYDROLOGY OF THE SOUTH ATLANTIC

213

To measure the seasonal changes in the temperature and sahnity of sub-Antarctic
water five stations have been used, WS 69, 138, 252, 318, and 432, which are fairly close
together in a position about 100 miles from the Antarctic convergence in approximately
52° 30' S, 52° 30' W. The results obtained from so few observations are necessarily only

TEMPERATURE °C.

MARCH JUNE

SEPTEMBER DECEMBER

MARCH

SALINITY %o

MARCH JUNE SEPTEMBER DECEMBER MARCH

6'00

S'OO-

4-00

3-00

200

34-30

34-20

34- 10

34-00

Fig. 21. The seasonal changes of the temperature and saUnity of sub- Antarctic water in

52rs, szrw.

approximate, but they are sufficient to show the nature of the changes which take place.
The four curves in the upper part of Fig. 21 show the change in temperature from month
to month of the water at the surface and at depths of 100, 400, and 600 m. They show
that the annual range of surface temperature, at a station about 100 miles north of the
convergence, is about 34° C: from 3-5° C in September to 6-9° C. at the end of
February. In the deeper layers the annual range decreases, until at 600 m. it is about

214 DISCOVERY REPORTS

1° C. This is, however, a large seasonal variation of temperature for water at so great a
depth, and it is a result of the seasonal variation in the temperature of the Antarctic
surface water which sinks to that depth. The maximum temperature at a depth of
600 m. is only reached about two months after the maximum temperature at the surface.
This is because the Antarctic surface water, though it has its maximum temperature at
about the same time as sub-Antarctic surface water, takes two months to sink to 600 m.

The temperature curves for the water at depths of 100 and 400 m. show that the water
at these depths has a maximum temperature at times intermediate between the times of
the surface and 600-m. maxima. The surface water has its minimum temperature at
the end of August and the water at 600 m. about two months afterwards.

The curves in the lower half of the diagram show the changes in salinity of the water
at the surface and at a depth of 600 m., and the dotted line shows the seasonal change
in the average salinity of the whole water column from the surface down to 600 m. The
annual range at the surface is about o-i^°l ^^ and about 0-04 "/^^ at a depth of 600 m.
At 600 m. the salinity is at a minimum when the temperature is at its maximum, which
is what happens in the Antarctic surface water itself. At the surface, however, the
minimum salinity is only reached about two months after the maximum temperature,
and it appears that the minimum salinity is due to the mixing of Antarctic surface
water from below.

The diagram also shows that the difference in temperature between the water at the
surface and that at a depth of 600 m. is least in September, when the layer is almost com-
pletely mixed, and greatest in February, when, especially in the surface 200 m., there are
fairly stable strata in the layer. The line showing the variation in the average salinity of
the upper 600 m. of water shows that the layer as a whole, or at least that part of it in the
region of intense mixing (which extends to about 100 miles north of the convergence)
has its minimum salinity at the end of April and its maximum salinity in November.
The curve is not quite symmetrical : the time taken for the water to be reduced from its
greatest to its least salinity is about 5^ months, whilst that taken to regain its maximum
salinity is about 6^ months. This would be expected if the speeding up of the currents
of Antarctic surface water in spring is quicker than the slowing down of the same
currents in the autumn.

The nature of these seasonal changes helps to show that sub- Antarctic water has its
origin principally in the Antarctic water. The changes, except in surface temperature,
lag behind those of Antarctic water and the layer is less directly affected by Antarctic
weather conditions.

The distribution of temperature at the surface of the sub-Antarctic Zone is shown in
Figs. 8 and 12 (pp. 183, 196). The surface water north of the 3° C. isotherm in the west,
and north of 3-5 ' C. in the east is sub-Antarctic. The isotherms follow the direction of the
surface currents. They run approximately south-west to north-east through the Drake
Passage and then turn towards the north. Part of the surface water flows between the
Falkland Islands and the Patagonian coast, but since the greatest depth in this channel
is less than 250 m. the movement is almost confined to the surface stratum. This water

HYDROLOGY OF THE SOUTH ATLANTIC 215

together with some which flows northwards to the east of the Falkland Islands forms
the Falkland current. The remainder turns eastwards and flows across the Atlantic Ocean.
The high surface temperatures north-east of the Falkland Islands are the efltect of sub-
tropical water, originating in the Brazil current, which has crossed the sub-tropical
convergence at the surface and mixed with sub-Antarctic water.

Fig. 13 (p. 197) shows the saHnity of the surface 100 m. of water in the Falkland
Sector. It shows how great is the effect of the coastal water which flows southwards
down the west coast of South America and round Cape Horn, on the salinity of the sub-
Antarctic water as far east as the Falkland Islands. When this survey was made the sur-
face salinity south of Cape Horn increased from about 33-30 7oo or less close inshore, to
34-207^0 about 150 miles offshore. Between the Patagonian coast and the Falkland
Islands it increased from 33-28 to 337 '/q,.,, and then to 34-20 about 150 miles east of the
Falkland Islands. The temperature of the sub-Antarctic water affected by the coastal
water is about 0-5 to 1° C. higher than that outside its influence.

OXYGEN CONTENT OF THE SUB-ANTARCTIC WATER
Fig. 18 (p. 204) shows that at the end of winter in the south of the sub-Antarctic Zone
in 80^ W the surface water is about 90 per cent saturated with oxygen ; at the level of
minimum salinity it is 80 per cent saturated. In 30" W the observations were made at
the end of April. The water was about 95 per cent saturated with oxygen at the surface,
90 per cent at a depth of 100 m. and still about 80 per cent in the region of minimum
salinity. There are not enough observations to tell how the oxygen content at one
particular place varies from season to season; but an examination of the Antarctic
intermediate layer, which has its origin in the region of intense mixing north of the
Antarctic convergence, shows that the water in this region has its maximum oxygen
content at about the same time as its minimum salinity. This indicates that the oxygen
content of the layer is renewed by oxygen carried down by the Antarctic surface water.
The distribution of oxygen in the sub-Antarctic surface layer is shown in Plate X,
which gives the distribution along 30° W expressed in cc. of oxygen per litre of sea
water. In the northern half of the zone just south of 40° S the oxygen content decreases
below the surface stratum to a secondary minimum between 80 and 100 m. Below this
stratum there is a small increase in oxygen content because the oxygen is renewed by
water from the region of intense mixing. The oxygen content then decreases with
depth ; but owing to the sinking of the Antarctic surface water, and the water from the
region of intense mixing, the water of the sub-Antarctic layer has a high oxygen content
compared with the rest of the water in the South Atlantic Ocean at the same level.

PHOSPHATE AND NITRATE CONTENT OF THE SUB-ANTARCTIC WATER

Measurements made along 30° W show that the phosphate content of sub-Antarctic

water is about 80 mg. PgOg/m.* just north of the Antarctic convergence and about

60 mg. PgOj/m.s just south of the sub-tropical convergence. Plate IX shows the vertical

distribution of phosphate content along 30° W expressed as mg. PgOg/m.* It will be

2i6 DISCOVERY REPORTS

seen that the water of greatest phosphate content in the South Atlantic Ocean is found in
the bottom of the sub-Antarctic layer and in the top of the warm deep layer. Above this
greatest content the rest of the layer has also a very high phosphate content, and so great
is the ease with which vertical mixing can take place in the upper strata of the layer, that
the surface is never depleted of its phosphate. The lowest content measured has been
50 mg. PoOj/m.^; this was found in sub-Antarctic water south of the Agulhas current,
which contained sub-tropical water mixed with it. Close inshore near Cape Horn at
the end of summer the surface phosphate content was 70 mg. PgOg/m.^

Measurements of nitrate in sub-Antarctic water are rather few, but those made in the
southern half of the zone show that in 30° W the nitrate content of the water at the end of
summer was 210 mg. nitrate Na/m.'* ; on the eastern side of the South Atlantic a little
farther south it was 220 mg. nitrate Ng/m.^ There are no measurements in sub-Antarctic
water farther north, but from the vertical distribution of nitrate on the western side of
the ocean, which is shown in Plate IX, it appears that the surface content in the northern
half of the zone in summer will be about 150 mg. nitrate N,/m.^

SUB-TROPICAL AND TROPICAL WATERS

SUB-TROPICAL SURFACE AND UNDER-LAYERS, AND THE
TROPICAL SURFACE LAYER

In the sub-tropical Zone the surface water is much warmer and more saline than
sub-Antarctic water. Just north of the sub-tropical convergence the surface temperature
varies from about i5"5° C. in winter to 18-5° C. in summer. The water between the
surface and 50 to 60 m. is well mixed; its temperature only decreases 1° C. or less with
depth, and its salinity does not alter appreciably with depth. The complete mixing of the
surface stratum is brought about by the turbulent movement of pure wind drift currents,
and also probably by convection currents resulting from periodical changes in the tem-
perature and salinity of the surface water, which are themselves due to differences in
radiation, conduction, and evaporation.

Below the surface stratum the salinity of the sub-tropical water increases slightly, and it
is greatest in a stratum between 80 and 100 m. There is probably a component of move-
ment southwards in this stratum, and where the whole layer of water moves southwards
the movement will be greatest in this stratum. In 30'' W, where the observations on the
layer were made, the wind south of 30' S blows principally from the west. It is much
weaker than the wind in the sub-Antarctic Zone, but it will give rise to a transport
northwards in the surface and a return current at a depth of 80-100 m.

Below the stratum at 80 to 100 m. both temperature and salinity decrease with depth
until a sharp discontinuity marks the boundary between sub-tropical water and the sub-
Antarctic water which has sunk below the surface south of the sub-tropical convergence.
This water is now known as Antarctic intermediate water, because of its intermediate
position between surface and deep layers of sub-tropical origin.

The temperature of sub-tropical water in 30° W increases gradually towards the

HYDROLOGY OF THE SOUTH ATLANTIC 217

north until, in about 28° S (when the temperature at the surface is 23° C), the increase
becomes rapid, and another surface layer of water can be distinguished.

This second layer lies above sub-tropical water, and is separated from it by another
discontinuity of temperature and salinity, forming so sharp a density gradient that vertical
mixing between the two layers is almost entirely prevented. The surface layer has been
called tropical water. It is almost depleted of its dissolved phosphate and nitrate ; and
they cannot be renewed, except perhaps very slowly, because the sharpness of the dis-
continuity makes vertical mixing across it almost impossible. The mean temperature
in the discontinuity between the tropical water and the sub-tropical water is 23° C,
and the depth of this isotherm has been assumed to be the bottom of the tropical
water.

The sub-tropical water which lies below tropical water has been called sub-tropical
under-water, and the layer the sub-tropical under-layer, to distinguish it from sub-
tropical surface water and the sub-tropical surface layer. The name sub-tropical under-
water has already been used by Wiist (1928, p. 514) for water with approximately the
same limits of temperature and salinity. The mean temperature in the lower discon-
tinuity which separates sub-tropical surface and under-waters from Antarctic inter-
mediate water is about 10-5" C, and the mean sahnity 34-85 to 35-0 °/o„. The depth of
this temperature and salinity has been assumed to be the depth of the sub-tropical
waters. Its depth along 30° W is shown in the following table.

Table V

;8°S 35° S 30° S

25° S

20° s

15° s

10° S

5°S

0°

320 440 510

490

440

360

260

290

350

Latitude

Depth of sub-tropical waters in metres

The slope of the discontinuity in a north and south direction, follows approximately
the slope of the isobaric surfaces and from it the direction of movement of the water in
the sub-tropical layer can be obtained.

Between 38 and 30° S sub-tropical water moves eastwards under the influence of the
westerly winds, and the discontinuity slopes downwards towards the north. The easterly
movement carries water across the Atlantic Ocean from the Brazil current towards
Africa. Some of this water flows south of the Cape of Good Hope, joins water which is
turned back from the Agulhas current and flows eastwards across the Indian Ocean.
The remainder turns northwards, joins the Benguela current and flows north and east
along the west coast of Africa.

Between 30 and 10° S sub-tropical water flows westwards, and the discontinuity
slopes upwards towards the north. This westerly movement is the result of the south-
east trade winds; it carries water from the Benguela current back to the Brazil current.

The two movements of sub-tropical water complete an anticyclonic water movement
which extends over the whole width of the ocean. In the centre of this movement, there
is a tendency for the surface water to accumulate and to sink downwards owing to the
effect of the earth's rotation on the currents. The sub-tropical water is therefore deepest
in the centre of the movement, and in 30° S and 30° W its depth is about 500 m.

2i8 DISCOVERY REPORTS

Between lo" S and the Equator sub-tropical water flows eastwards; probably as a
counter-current below the westward flowing tropical water of the south equatorial
current. North of 28° S in 30° W the sub-tropical water is covered by tropical water.

The following table shows the depth of the 23" C. isotherm which has been assumed
to show the depth of the tropical water.

Table VI

25° S 20° s

15° s

10° S

5°S

0°

1 70 130

145

115

80

70

Latitude

Depth of tropical water in metres

North of 18° S tropical water flows westwards in the south equatorial current and as
it approaches the Brazilian coast this current is partly deflected southwards into the
Brazil current. As it flows southwards, the Brazil current loses water from its left-hand
side in a surface drift towards the east, and in 30° W the tropical water south of 18^ S
is moving eastwards. The tropical water is deepest in 18° S, in the centre of the anti-
cyclonic movement between the south equatorial current and this movement towards
the east. These movements are shown on the chart of surface currents in the Atlantic
Ocean by Meyer (1923).

TEMPERATURE AND SALINITY OF THE SUB-TROPICAL AND

TROPICAL WATERS

The surface temperature of sub-tropical water increases towards the north from
about 15-5° C. in winter or 18-5° C. in summer near the sub-tropical convergence; it
increases to 23° C. before it becomes covered with the tropical surface layer. Its surface
salinity increases from about 35-0 to 36-0 °/„o over the same distance. In the bottom of
the layer the salinity increases from 34-85 to 35-oo7„<, and the temperature remains

about 10-5° C.

From the surface to 50-60 m. sub-tropical water is almost uniform. Below it there
is a stratum of slightly higher salinity, and below this again the temperature and salinity
decrease gradually with depth until the discontinuity which forms the boundary of the
Antarctic intermediate layer is reached.

The surface temperature of tropical water increases from 23° C. in 28° S to a maximum
of 28-29° C. just north of the Equator. Between the surface and the discontinuity which
separates the layer from the sub-tropical under-layer the temperature of tropical water
varies very little with depth.

The salinity of the tropical surface layer increases from about 367„o in 28° S to a
maximum of 37-4 7 00 i" about 18" S. It then decreases towards the north in the south
equatorial current. In 9° 47' S the salinity remained uniform with depth down to 50 m.,
and then increased to a maximum at 100 m. This is probably because the transport of
surface water, which under the influence of the east winds was towards the south, must
be compensated by a movement northwards between 80 and 100 m., similar to that which
has been found to move southwards in the region of westerly winds. The lowest salinity
found in the surface stratum was about 35-9700 just north of the Equator. The salinity

HYDROLOGY OF THE SOUTH ATLANTIC 219

of tropical water in the discontinuity separating it from the sub-tropical under-water
also varies from about 35-8 to 37-0 7oo- It is greatest between 20 and 10° S, and least
just north of the Equator.

OXYGEN CONTENT OF THE SUB-TROPICAL AND TROPICAL WATERS

The oxygen content of the sub-tropical water is approximately 95 per cent of satura-
tion, and the oxygen content between the surface and 50-60 m. (which must be in
constant circulation) is almost uniform with depth. There is, however, always a small
maximum at a depth of 60 m., which must be caused by the maximum growth of
diatoms at that level. The oxygen content in the bottom of the layer decreases from
about 80 per cent in 40° S, to 75 per cent in 28° S.

The surface of tropical water is about 96 per cent saturated with oxygen between 28
and 20° S, but then the oxygen content decreases slightly to a minimum of about
93 per cent between 15 and 10'^ S in the south equatorial current. The oxygen content
of tropical water increases with depth and is at a maximum just above the discontinuity
which separates the layer from the sub-tropical under-layer. This maximum is again
probably due to the greater growth of diatoms at that depth, the diatoms utilizing the
small amounts of phosphate and nitrate liberated in tropical water as a result of the
slight vertical mixing which takes place across the discontinuity.

The oxygen content of the sub-tropical under-layer can only be replenished by a
horizontal inflow of water, since the layer is shut ofl^ from the surface by a discontinuity
which does not allow vertical mixing. The oxygen content of the layer falls from about
80 per cent of saturation in 28° S to about 40 per cent at the Equator.

PHOSPHATE AND NITRATE CONTENT OF THE SUB-TROPICAL

AND TROPICAL WATERS

In 30" W the phosphate content of the surface stratum falls at once from about
60 mg. PgOs/m.^ to 10 mg. on passing from sub-Antarctic to sub-tropical water. The
nitrate content also decreases from about 150 mg. nitrate N.,lm.^ to 50 mg., and a little
farther north to 10 mg. Both decrease towards the north until just south of 28° S, the
phosphate content of the surface stratum is 8 mg. PgOs/m.^, and the nitrate content is
5-6 mg. nitrate NJm.^

In the surface stratum of tropical water there is no phosphate and it must be com-
pletely used up as soon as it is regenerated. There is usually a small nitrate content of
0-5 mg. nitrate NJm.^, and this is replenished from rain water, which has been found
to contain as much as 10 mg. nitrate Nj/m.^ after a thunderstorm. The nitrate is probably
not used up because of the absence of phosphate.

In sub-tropical water the phosphate and nitrate contents increase with depth until
there is about 35 mg. PoOs/m.^ and 150 mg. nitrate Na/m.^ in the discontinuity level.

The sub-tropical Zone is not very favourable for diatom growth in general because
of the small amounts of phosphate and nitrate at the surface. It is not, however, entirely
depleted of them, and where they are replenished by upwelling, such as takes place off
the African coast, the conditions are very good. The tropical water can only support

220 DISCOVERY REPORTS

such growth as can exist on the small amounts of phosphate which are obtained from
> the sub-tropical under-layer.

EXTENT OF THE SUB-TROPICAL AND TROPICAL ZONES
The tropical convergence, between tropical and sub-tropical waters, is not so well
defined as the other convergences. It is found at the surface where the increase in sur-
face temperature from south to north becomes more rapid, and where the surface
stratum becomes shut off from sub-tropical water by a sharp discontinuity. In the
South Atlantic Ocean it corresponds in summer with the 23° C. isotherm.

The tropical Zone extends to about 28° S in the western half of the ocean where the
Brazil current carries tropical water southwards, but only to 10-15° S in ^^^ ^^^ where
the Benguela current carries sub-tropical water northwards. The boundary between
it and the sub-tropical Zone lies between St Helena and Ascension. Hydrologically
St Helena is sub-tropical, and Ascension tropical. The water off the Cape of Good
Hope, and ofT the west coast of Africa as far as 10-15° S, is sub-tropical; but close
inshore, particularly south of 20° S, it is also influenced by upwelling water from the
Antarctic intermediate layer.

THE DEEP WATERS OF THE SOUTH ATLANTIC OCEAN

There are three principal deep layers in the South Atlantic Ocean. They are: the
Antarctic intermediate layer, the warm deep layer, the Antarctic bottom layer.

It has already been shown that the Antarctic surface water which sinks at the Ant-
arctic convergence mixes with sub-Antarctic water in a region of intense vertical mixing
just north of the Antarctic convergence. The mixture of water then sinks downwards
towards the north, and spreads over the whole of the South Atlantic Ocean. It gives
rise to a layer of water which is both Antarctic and sub-Antarctic in origin, and it is
found farther north between layers of sub-tropical water, and warm deep water, both
of which are of sub-tropical origin. The layer has therefore been called the Antarctic
intermediate layer, and the water Antarctic intermediate water.

Below the Antarctic surface water in the Antarctic Zone, below the sub-Antarctic
water in the sub-Antarctic Zone, and below the Antarctic intermediate layer farther
north there is a layer of water which has been called the warm deep layer. It is most
probable that all the water in the warm deep layer is not of the same origin. Brennecke
(1921) and WiJst (1928) have shown that a stream of North Atlantic deep water flows
southwards in the South Atlantic from its origin in the sub-tropical regions of the North
Atlantic. Clowes (1933) has shown that south of 46° S the water in the layer is derived
from the Pacific Ocean. In the Antarctic Zone the maximum temperature in a vertical
column of water is found in the warm deep layer, but north of the Antarctic convergence
there is only a secondary temperature maximum in the layer because there are warmer
layers of water nearer the surface.

Below the warm deep layer there is a heavy type of Antarctic water which has been
formed by the cooling of warm deep water without appreciable dilution. It sinks near

HYDROLOGY OF THE SOUTH ATLANTIC 221

the Antarctic Continent to fill the deep polar basins and then flows northwards along
the sea bottom. It has been called Antarctic bottom water.

ANTARCTIC INTERMEDIATE WATER

STRUCTURE AND DEPTH OF THE ANTARCTIC INTERMEDIATE
LAYER, AND THE ORIGIN AND MOVEMENTS OF ANTARCTIC

INTERMEDIATE WATER

The Antarctic intermediate layer is distinguished in any vertical series of observations
made in the sub-tropical or tropical Zones, by its low salinity. Below the sub-tropical
water the salinity decreases until it reaches a minimum value, and then it increases. The
water on either side of the level of minimum salinity is Antarctic intermediate water.
It has its origin in the region of intense mixing just north of the Antarctic convergence.
The path of the water forming the layer can easily be followed in Plate VIII which
shows the vertical distribution of salinity along the meridian of 30° W.

In the Antarctic intermediate layer three strata may be distinguished, (i) At the
bottom of the layer, where the temperature is lowest, there are the last traces of Antarctic
surface water, which have sunk from the cold stratum of the Antarctic surface layer,
(ii) At the level of minimum salinity the water has its origin in the mixture of lighter
Antarctic surface water and sub- Antarctic water which is formed just north of the
Antarctic convergence, (iii) Above the level of minimum salinity there is water which
has sunk below the surface in the sub-Antarctic Zone, but north of the region of intense
mixing, and is less Antarctic, or more sub-Antarctic, in origin. The greater part of the
layer is, however, composed of the water of low salinity which has its origin in the region
of mixing just north of the Antarctic convergence.

The following table shows the depth of the water of minimum salinity in different

latitudes in 30 ' W.

Table VII

Latitude 45° S 40° S 35° S 30° S 25° S 20° S 15° S

Depth of minimum salinity in metres 380 630 900 940 900 830 700

Latitude 10° S 5° S 0° 5° N 10° N 15° N

Depth of minimum salinity in metres 740 700 700 770 800 800

These depths show the path of the water of minimum salinity which forms the
nucleus of the layer as it flows northwards.

The temperature in the layer decreases until it reaches a minimum value near the
bottom of the layer. This level of minimum temperature can be followed as far back as
the Antarctic Zone and it is continuous with the level of minimum temperature in the
cold stratum of the Antarctic surface layer. The depth of the level of minimum tem-
perature is given in the following table for latitudes from 57° 30' S to 5° S in 30° W.

Table VIII

Latitude

Depth of minimum temperature in metres

57°3o'S
80

55° S
80

50° s
125

45° S
900

40° s

1200

35° S
1660

Latitude

Depth of minimum temperature in metres

30° S
1700

25° s

1400

20° S

1300

15° s

1060

10° s
1050

5°S
1000

7-2

222 DISCOVERY REPORTS

Below the level of minimum temperature, the temperature increases to a maximum
value in the upper stratum of the warm deep layer, and the water in this stratum,
whatever its origin may be, has a component of movement southwards. Table IX
shows the depth of this maximum temperature in 30° W, together with the depth of
the maximum salinity in the warm deep layer.

Table IX

Latitude

57° 30' S

55° S

50° s

45° S

40° s

35° S

Depth of maximum temperature of warm

deep water in metres
Depth of the maximum salinity of the

warm deep layer in metres

600

700

600
640

600
1650

1300
2300

1800

2950

2260

Latitude

30° s

25° s

20° S

15° s

10° S

5°S

Depth of maximum temperature of warm

2260

2080

1800

1480

1450

1300

deep water in metres

Depth of the maximum salinity of the 3400 2800 2650 2000 2000 20DO

warm deep layer in metres

The boundary between sub-Antarctic or Antarctic intermediate water and the
warmer layer beneath — that is to say the boundary between the north- and south-
going movements — may be regarded with sufficient accuracy as lying midway between
the levels of minimum and maximum temperatures.

The following table shows the depth of this boundary in 30° W.

Table X

Latitude 45° S 40° S 35° S 30° S 25° S 20° S 15° S 10° S 5° S

Depth of the boundary between sub-Ant- iioo 1500 i960 1980 1740 1550 1270 1250 1150

arctic or Antarctic intermediate water,

and warm deep water in metres

The origin of the water in the Antarctic intermediate layer has been discussed by
Brennecke (1921, p. 140), who describes the movement in the layer as the sub-Antarctic
deep current ; he gives its origin as the surface drift out of the Weddell Sea, which sinks
in about 50" S and pushes its way, at an average level of about 1000 m., through the
South Atlantic Ocean as far as 25° N. The layer is also described by Drygalski (1927,
p. 498) who describes its origin as Antarctic polar water. Both descriptions are correct
although not complete.

TEMPERATURE, SALINITY AND OXYGEN CONTENT OF
ANTARCTIC INTERMEDIATE WATER

Although it has probable movements from west to east and east to west, the Antarctic
intermediate water flows northwards. This is proved by the fact that the layer which
extends as far north as 25 ° N in the West Atlantic contains water which can only have had
its origin south of 40° S. As the water flows northwards its properties change as a result
of the vertical mixing which takes place between the water in the layer and the waters
above and below it. The changes are greatest at the top and bottom of the layer where

233

HYDROLOGY OF THE SOUTH ATLANTIC

the effect of vertical mixing is greatest, and least at the level of minimum salinity which
forms the nucleus of the layer.

It is difficult to trace the changes in the upper stratum of the layer, since the path of
the water is not indicated by any maximum or minimum in temperature or salinity which
can be followed. The water which has a temperature of more than 10-5° C. has been
considered to be sub-tropical, and the origin of much of the water in the sub-tropical
layer must be in Antarctic intermediate water, which has been warmed, and had its
salinity increased, by mixing across the discontinuity.

In the bottom of the Antarctic intermediate layer, at the level of minimum tempera-
ture, the changes can be followed more easily. The temperature at this level increases
from 278° C. in 34° 08' S to 4-12° C. in 03° 19' S, and the salinity from 34-62 '7oo in
34° 08' S to a maximum of 3473 7 00 in about 18° S. The increase in temperature is the
result of mixing with the warmer waters which lie both above and below it. The increase
in salinity is the result of mixing with the more saline water below and the less saline
water above. North of 18° S the salinity in the level of minimum temperature de-
creases from 3473 to 34-59 7oo in 03° 19' S, because there is an increasing percentage
of Antarctic intermediate water in the mixture at this level.

The changes of temperature and salinity at the level of minimum salinity, where the
layer is least changed by vertical mixing, are most interesting. They are shown, together
with the figures for oxygen content, in the following table.

Table XI

Depth of

Minimum

At depth of minimum salinity

Latitude

minimum

salinity

salinity

°l

I 00

Temperature

Oxygen
content

m.

°C.

c.c./litre

South

46° 43'

330

34-15

3-23

6-02

43° 08'

410

34-16

3-52

5-94

38° 17'

800

34-22*

3-86

5-45

34° 08'

910

34-26

3-96

5-19

31° 16'

940

34-28

4-00

4-86

26° 07'

910

34-33

4-00

4-59

21° 13'

850

34-38

4-1 1

4-24

15° 37'

700

34-41

4-56

4-08

09° 47'

75°

34-49

4-40

3-50

03° 19'

690

34-50

4-68

3-36

North

02° 59'

740

34-51

5-02

3-06

08° 54'

800

34-69

5-86

2-18

14° 27'

800

34-83

6-41

2-37

* This value obtained graphically.

It will be seen from this table that the temperature and salinity both increase towards
the north as a result of vertical mixing. The increase is approximately proportional to
the distance the water travels ; but it is not regular, and there are patches of warmer

224

DISCOVERY REPORTS

and less saline water alternating with patches of cooler and more saline water. The oxygen
content of the water decreases towards the north as the oxygen is used up by animal
life and by oxidizable organic matter in the water. The decrease in oxygen content is also
approximately proportional to the distance the water flows, and it is not regular. There
are alternate patches of water with high oxygen content and low oxygen content.

The irregularity in the increase in temperature and salinity, and in the decrease in
oxygen content, is partly due to the difference between the water which leaves the region
of intense mixing just north of the Antarctic convergence in winter, and that which

LATITUDE
20° 10"

34 10

3420

3470

Fig. 22. The change in the saHnity and oxygen content towards the north in the
Antarctic intermediate current.

leaves in summer. The mixture of waters in this region has its maximum temperature
and minimum salinity at the end of the summer. Both are the result of the higher
temperature and lower salinity of the Antarctic surface water which sinks in summer,
and it is to be expected that the mixture will have its maximum oxygen content at about
the same time. This will be so because the Antarctic surface water has its greatest
oxygen content in spring and early summer, when the phytoplankton is at its maxi-
mum, and also because there is probably a greater percentage of Antarctic surface
water in the mixture in summer than there is in winter when the surface currents are
slower.

Fig. 22 shows the oxygen content and salinity of the water at the level of minimum
salinity in different latitudes in 30° W. To make comparison of the curves easier the

HYDROLOGY OF THE SOUTH ATLANTIC

225

salinity scale has been made to read in a negative direction, the downward slope of the
curves towards the north thus showing a decrease in oxygen content, but an increase in
salinity.

The two curves, which are approximately parallel, show maxima and minima which
correspond to the water which flows northwards from the region of mixing in different
seasons. Water which sinks from the Antarctic surface layer and flows northwards in
summer has a greater oxygen content and a lower salinity than water which sinks and
flows northwards in winter. There are not sufficient points on the curves to allow them
to be drawn exactly, but if they have any meaning at all their shape cannot diff^er very

>

O CALCULATED FROM OXYGEN MAXIMA
A CALCULATED FROM SALINITY MINIMA
\ CALCULATED FROM OXYGEN MINIMA
L CALCULATED FROM SALINITY MAXIMA

Fig. 2j. The change in the speed of the Antarctic intermediate current towards the north.

much from that shown. From the distance apart of the consecutive maxima or minima
the distance travelled by the water in one year can be calculated.

In Fig. 23 the velocities calculated in this way have been plotted against latitude.
South of 20° S the velocities calculated from consecutive oxygen maxima and salinity
minima are greater than those calculated from the oxygen minima and salinity maxima.
This diff'erence would arise if the water which sinks in winter flows, at first, more quickly
than that which sinks in summer, and this might be the result of the greater density of
Antarctic surface water in winter, causing it to sink more quickly and take a more direct
path northwards. The difl^erence could also be explained if the water of minimum salinity

226 DISCOVERY REPORTS

and maximum oxygen content were the first to increase its speed in the general increase
in velocity towards the north.

At the depth of minimum salinity the average velocity of the water in the layer in-
creases from about 1-3 miles a day in 40° S to 2-5 miles a day in 7° S. The increase in
velocity may be caused by the narrowing of the South Atlantic Ocean towards the
north, as well as to the decreasing thickness of the layer, which becomes more confined
between sub-tropical water and North Atlantic deep water.

In addition to this movement towards the north, there are easterly and westerly
movements in the layer; but about these little is known. The layer is deepest in about
30° S between the region of west and east winds. South of 30" S there will probably
be a movement in the layer towards the east, which is partly responsible for the slope
of the isotherms and isohalines downwards towards the north, and north of 30° S a
weaker movement towards the west. These zonal movements will aftect the curves in
Fig. 22, but not enough to destroy them.

The velocity of the Antarctic intermediate current towards the north has recently
been measured on the eastern side of the South Atlantic, and similar results have been
obtained to those already described. A preliminary account of the measurement of
these velocities (Deacon, 1931) has also been confirmed by Castens (i93i),who obtained
a similar result from the changes in temperature with latitude. The movement north-
wards can be considered as a series of waves of water of different properties moving
outwards from the Antarctic regions, each of which, in the Atlantic Ocean, takes about
4^ years to travel from the Antarctic convergence to the Equator.

The salinity of the water at the level of minimum salinity only increases from 34- 1 5
to 34-56°/ CO between 47° S and the Equator. The temperature increases from 3-2 to
4-85° C. over the same distance, and the oxygen content decreases from 6-o to 3-3 cc./litre.
North of the Equator the increase of temperature and salinity is more rapid, and the
most northerly observations which show the minimum salinity typical of the layer are
those of the ' Planet ' (1909) in 24° 20' N, 22° 37' W, and the ' Mowe ' (1914) in 26° 10' N,
16° 42' W. Antarctic intermediate water will reach this point about six or seven years
after it left the surface in the Antarctic.

THE WARM DEEP WATER

STRUCTURE AND DEPTH OF THE WARM DEEP LAYER, AND THE ORIGIN

AND MOVEMENTS OF THE WARM DEEP WATER, THE NORTH ATLANTIC

DEEP WATER AND THE PACIFIC DEEP WATER

The movement of Antarctic surface water away from the Antarctic regions towards
the north is known to take place all round the pole, and it is also known that there is a
similar movement of Antarctic bottom water towards the north near the sea-bottom.
To make up for this transport of water away from the Antarctic regions in the surface
and bottom layers there must be a movement towards the pole in the intermediate layer.
This is supplied by a movement southwards in the warm deep layer.

HYDROLOGY OF THE SOUTH ATLANTIC 227

In any series of vertical observations made inside the Antarctic Zone it will be seen
that the temperature either remains constant or decreases with depth down to the level
of the cold stratum of the Antarctic surface layer. Below this level it increases to a
maximum in the warm deep water, and then it decreases again, down to the sea-bottom.
If there were no current southwards in the warm deep layer, the maximum temperature
of the warm deep water would soon disappear as a result of vertical mixing between the
water in the layer and the colder waters above and below it.

There can be no doubt then that the water at the level of maximum temperature in
the Antarctic Zone has a component of movement southwards. The way in which the
warm deep layer feeds the surface layer has already been described in the first section
of this report, and it will be shown later that it also gives rise to Antarctic bottom water.
The origin of the warm deep water in the Antarctic Zone can best be decided from sec-
tions which show the vertical distribution of temperature and salinity .

In both the sections shown on Plate VIII it appears that the water in the warm deep
layer south of 50° S has its origin in the North Atlantic deep current, which climbs rapidly
towards the surface at the Antarctic convergence. It must, however, be remembered
that the deep currents may have zonal as well as meridional movements : a deep current
from the Pacific Ocean, in which the water is very similar to that in the North Atlantic
deep current, will not be distinguished in the sections. None the less the sections show
that the warm deep layer in the Antarctic Zone is continuous with the warm deep layer
below the sub- Antarctic and Antarctic intermediate waters, even ahhough the warm
deep water in the continuous layer is not all of the same origin.

Clowes (1933) has shown that the warm deep water south of 46° S in 30" W has its
origin in the Pacific Ocean. Further work shows that most of the Pacific deep water is
probably stopped at the Scotia Arc, and also that in 40° S between the observations in
30' W and those of the ' Deutschland' in 50° W, there is a strong movement of North
Atlantic deep water southwards. The conclusion that the water south of 46° S had its
origin in the Pacific Ocean was based partly on the differences of salinity and tempera-
ture at St. 666 and 671. These differences are alternatively explained by the slope of the
warm deep layer and the greater reduction of the layer at St. 666 by vertical mixing.
So great is the similarity between North Atlantic and Pacific deep water that there is
as yet no very reliable evidence to show the origin of the warm deep water just north of
the Scotia Arc. The temperature at its level of maximum salinity is greater than the
temperature of the water of maximum salinity in the Pacific Ocean, and at the same time
the maximum salinity itself is greater. The water north of the Scotia Arc must therefore
contain additions from the Atlantic Ocean. The warm deep water in the Scotia Sea
probably has a Pacific origin, but this in my opinion is as yet uncertam.

Fig. 24 shows the temperature of the warm deep layer in the Falkland Sector. It
has been constructed by plotting the maximum temperature which has been found in
the layer, and not the temperature at any particular depth. The diagram does not
necessarily show direction of movement, but there is usually a component of movement
from warm to cold regions and not from cold to warm. Regions of low temperature

HYDROLOGY OF THE SOUTH ATLANTIC 229

are regions where there is greater mixing of the warm deep layer with Antarctic surface
and bottom waters.

Although the isotherms do not show the direction of movement of the warm deep
water, they show the relative strength of the deep and bottom movements. When the
strength of the bottom current towards the north or north-west increases, or when the
strength of the warm deep current towards the south-east decreases, the isotherms
recede towards the north. When there is a stronger flow of warm deep water, or a lesser
flow of cold bottom water they advance towards the south.

In an investigation of the origin of the deep water in the Falkland Sector the problem
is complicated by the presence of a cold deep current flowing out of the Weddell Sea.
It has been shown by Wiist in 5° E, by Brennecke in 30° W, and by ourselves in
I5°E, 15° W and 20° W, that south of about 66° S the temperature in the warm deep layer
increases towards the south, and it is now almost certain that north of this warmer deep
water there is a continuous belt of colder water stretching from the Weddell Sea to I5°E.
East of is" E, however, warm deep water gets far south without interruption, and the
warm deep current in the Atlantic south of 66° S has its origin east of 15° E. It flows
westwards as a result of the prevailing easterly winds south of 66° S, and Wiist has
pointed out that this is proved by its decrease in temperature towards the west. As it
flows round the Weddell Sea in a cyclonic movement similar to that which takes place
at the surface the deep current is cooled by vertical mixing.

There is, however, always a warm deep layer in the Weddell Sea (except perhaps very
close inshore in the west of the sea) : the temperature always increases below the
Antarctic surface layer to a maximum in this deep water before it decreases to a
minimum value at the bottom. By the time it flows out of the northern side of the
sea the deep water has only a maximum temperature of about 0-40° C. or less, and
as soon as it meets warm deep waters of Pacific and Atlantic origin it both sinks
below them and mixes with them. It is then perhaps best described as a cold deep
current.

The cold deep current from the Weddell Sea follows the bottom current. Part of it
flows eastwards across the Atlantic Ocean; part of it flows north across the Scotia Sea,
or round the outside of the Scotia Arc, and sinks below the warm deep water north of
South Georgia. The cold deep current is partly responsible for the sudden climb of
the warm deep layer north of South Georgia, and also for the sudden difl'erences in
temperature and salinity in deep levels south of 40° S, north of South Georgia. Its
effect on the deep temperature distribution in the east of the Scotia Sea and north-east
of South Georgia is shown in Fig. 24.

As mentioned above, Brennecke and Wiist have shown that the warm deep water of
Atlantic origin is formed in the sub-tropical region of the North Atlantic. It sinks
below the surface in the convergence region between the Canary current, the North
Equatorial current, and the Gulf Stream. It is also mixed with the last traces of Antarctic
intermediate water from the same region, and in the east with water from the Mediter-
ranean Sea (Wust, 1928, p. 522). As it flows southwards it mixes with Antarctic inter-

230 DISCOVERY REPORTS

mediate water and sub-Antarctic water and picks up from them a large concentration
of phosphate and nitrate in its upper stratum.

Not much is known of the origin of Pacific deep water, but when it is in the Atlantic
it differs very little in properties from North Atlantic deep water.

The warm deep water which flows westwards in the south of the Weddell Sea, and
eventually out of the Weddell Sea as the cold deep current, flows from the Indian Ocean.
It cannot yet be definitely stated whether it flows into the Indian Ocean from the
Atlantic or whether it has its origin in the north of the Indian Ocean.

TEMPERATURE, SALINITY AND OXYGEN CONTENT OF THE WARM

DEEP WATER

The upper stratum of warm deep water is characterized by a secondary temperature
maximum and it is warmer than the bottom stratum of Antarctic, sub-Antarctic, and
Antarctic intermediate waters. The temperature in the warm deep layer decreases with
depth, but its salinity increases with depth until it reaches a maximum at a level which
can be considered the nucleus of the layer — just as the minimum salinity level was
considered to be the nucleus of the north-going Antarctic intermediate water.

The depth interval between the depths of maximum temperature and maximum
salinity in the layer is a measure of the thickness of the layer. The thickness can be
obtained from Table IX (p. 222), and it will be seen that it is not uniform. The depth
of the level of the maximum salinity, and the thickness of the layer, are greatest in the
deep basins and least above the zonal ridges and rises which cross the West Atlantic
basin.

The temperature and salinity in the nucleus of the layer decrease towards the south,
and their decrease is due to vertical mixing with the Antarctic intermediate and bottom
waters. The decrease is not regular, partly because the North Atlantic deep water which
supplies the current must vary in just the same way as the water supplying the Antarctic
intermediate current. The rate of decrease also shows some dependence on the thickness
of the layer, owing to the changing speed of the water, or because of lateral inflow of
water in the basins. Attempts to measure the speed of flow of the water are rendered
diflicult by these changes, which are fortunately not communicated to the layer above.

In 50° S and 30° W the depth between the maximum temperature and maximum salinity
of the warm deep water is about 1000 m., but in 55^ S it is only 40 m. This is probably
because the lower strata of warm deep water, whether their origin is Atlantic or Pacific,
have been changed or deflected by the strong flow of Antarctic bottom water, or of cold
deep water, from the Weddell Sea towards the Argentine Basin past the north-east coast
of South Georgia.

The depth between the maximum temperature and the maximum salinity generally
decreases in the direction of movement, or as the temperature itself decreases. The
maximum temperatures of the layer in the Falkland Sector have been shown in Fig. 24
The maximum salinity varies from about 3473 or 3474 °l^^ below the Antarctic con-
vergence to 34-67 7 00 ii^ the cold deep water of the Weddell Sea.

HYDROLOGY OF THE SOUTH ATLANTIC 231

From north to south the temperature at the level of maximum saHnity decreases from
3-93° C. in 14° 27' N to 0-67° C. in 57° 36' S, and the saHnity decreases from 34-97 to
34-68 °/q^ over the same distance.

The stream of North Atlantic deep water is most saturated with oxygen north of the
Rio Grande ridge, which crosses the West Atlantic basin in about 34° S. It contains
most oxygen at a depth of 2000-3000 m., where the water is about 75 per cent saturated
and contains from 5 to 5-5 cc. Oo/litre. There is a second stratum of high saturation in
the layer between 3000 and 4000 m., and the average saturation of the whole layer is
about 70 per cent.

The oxygen content decreases only slowly towards the south until the water has
crossed the Rio Grande ridge. Then the decrease is more rapid, and in the upper
stratum the oxygen content falls to 3-8 cc./litre and the saturation to 50 per cent. The
rapid decrease south of the ridge is probably due to the greater consumption of oxygen
by the greater amount of animal life and oxidizable matter in the water. It may also be
due to a decrease in the southerly component of the deep-water movement.

The deep water in the Antarctic Zone is thus poorly oxygenated compared with the
deep water farther north ; it contains most oxygen near the bottom of the layer, where it
is enriched by mixing with Antarctic bottom water. The oxygen content of the layer in
the Antarctic Zone is shown in Figs. 18 and 19 (p. 204) expressed as a percentage
saturation, and in Plate X as cc. OJ litre.

The seasonal changes in the temperature and salinity of the warm deep water have
not yet been worked out, but it has been shown that there are considerable changes
whose period is not yet known. There are patches of different temperature and salinity
in the North Atlantic deep current as it flows southwards, and the changes in the pro-
perties of the deep water at a particular station may be the result of the arrival of these
patches at different times.

ANTARCTIC BOTTOM WATER

THE ANTARCTIC BOTTOM LAYER, AND THE ORIGIN AND MOVE-
MENT OF ANTARCTIC BOTTOM WATER

Below the warm deep layer there is a colder layer of water, which has its origin in the
Antarctic regions. In the Antarctic Zone this layer is not separated from the warm deep
layer by a discontinuity in the changes of temperature and salinity with depth, and it is
difficult to decide upon the level at which the movement of warm deep water southwards
changes to one of Antarctic bottom water northwards.

The formation of Antarctic bottom water has been explained in various ways. It is
clear from its properties, and the level at which it is found, that it must consist of warm
deep water which has been cooled without being appreciably diluted. It could be formed
if the warm deep water were exposed for some period at the surface, especially in winter ;
but such exposure is not known to take place, for the layer is always found to be covered

by colder and less saline Antarctic surface water.

8-3

232 DISCOVERY REPORTS

Nansen (191 2) explains the formation of similar water in the Arctic Ocean as due to
a convective circulation which carries water downwards from the surface to the bottom
in autumn and early winter. Brennecke (1921, p. 140) says that it is formed along the
continental shelf of the Antarctic Continent where the surface water is cooled right
through and sinks down the continental slope. Drygalski (1927, pp. 495 et seq.) states
that the bottom water is a mixture of the cold water formed on the continental shelf
with the deep water. Wiist (1928, p. 525) at first followed Nansen's explanation. He
thought that when the surface currents became slower in autumn and winter Antarctic
bottom water was formed by the sinking of highly saline surface water in the centre of
two cyclonic movements, one in the Weddell Sea and one farther east.

The difficulty which prevented acceptance of the explanations of Brennecke and
Drygalski was that there was always found to be a continuous warm deep layer, even
at stations very far south, through which the cold water formed on the continental shelf
did not appear to sink. There are, however, insufficient observations to show that this
sinking of water does not take place, particularly in winter along the east coast of
Graham Land.

In a more recent publication (1933, p. 48) WiJst has concluded from the distribution
of potential temperature (the temperature to which the water would be adiabatically
cooled if it were raised to the surface) that Antarctic bottom water in the Atlantic is formed
in the way suggested by Brennecke. He has also distinguished a slightly warmer bottom
water which he has called Antarctic deep water. He suggests that the two cyclonic water
movements distinguished by Meyer (1923) in the Weddell Sea, and in 30° E, do not
exist, and that there is instead a convergence region similar to that shown by Willimzik
(1927) and Moller (1929) in the Indian Ocean. He considers that the convergence
region lies between 60° and 65° S in the Weddell Sea, and between 56° and 60° S
farther east. Its position is closely related to that of the northern edge of the pack-ice.
Antarctic deep water — according to Wiist — would be water which sinks in this region
in autumn and winter.

The existence of a convergence region in the Weddell Sea is however hypothetical ;
our observations and those of Brennecke seem to show that there is simply a diverg-
ence region between the currents flowing west and east. In this divergence region
warm deep water and Antarctic bottom water upwell, and tend to flow outwards (to
the left) in, or just below, the surface layer. At the edges of the sea, and particularly
along the east coast of Graham Land, heavy water will sink downwards. Most water
will probably sink in winter when the surface water is coldest and most saline. The
water which sinks need not have a salinity as high as the warm deep water because it
is colder, and because the sinking is partly due to dynamic forces.

The observations of the ' Deutschland ' (Brennecke, 1921) show that the surface
water in winter has a salinity of almost 34-50700' ^^^ closer inshore it may be much
more. By the sinking of this water and its mixture with warm deep water, which in the
west of the Weddell Sea has a maximum temperature of 0-40° C. or less, both Antarctic
bottom water and Wust's "Antarctic deep water" can be formed. If they are both

HYDROLOGY OF THE SOUTH ATLANTIC 233

formed in this way it is difficult to distinguish between them, and it is perhaps best
to describe all the water which spreads northwards below the warm deep water as
Antarctic bottom water. If not, in a region such as the Drake Passage, where the
bottom water is not so cold as it is in the Weddell Sea, it must be called Antarctic
deep water. It is worth while making an exception of the cold deep current flowing
out of the Weddell Sea, because it can still be clearly recognized that it is the remains
of the warm deep current which flows westwards into the sea south of 66° S.

In all probability the formation of Antarctic bottom water in the Weddell Sea is
largely the result of the warm deep current which flows into the southern edge of the
sea and is turned northwards off the east coast of Graham Land.

TEMPERATURE, SALINITY AND OXYGEN CONTENT OF THE
ANTARCTIC BOTTOM WATER

Because of the very gradual changes in temperature and salinity with depth between
the warm deep and Antarctic bottom layers, it is impossible to give a definite boundary
to the layer. The lowest temperature that we have found in Antarctic bottom water in
the open sea north of 70° S is -0-55° C. and the lowest salinity 34-65700; salinities of
34-63 7oo have, however, been found by the German Atlantic Expedition. Above this
water there are mixtures with varying amounts of warm deep water. The difference in
the properties of the Antarctic bottom layer and the warm deep layer is greater farther
north. Wiist (1933, pp. 71 et seq.) shows that water which has its origin in the Antarctic
bottom and deep layers can be traced as far as 40° N on the western side and as far as
35° N on the eastern side of the South Atlantic. There is, however, only a sharp dis-
tinction between the two layers south of the Equator in the west and south of the Walfisch
ridge in the east. North of the Antarctic Zone the small discontinuity in the density
gradient which shows the boundary between the two layers is found when the tempera-
ture is about 1° C. and the salinity 34-70700-

VERTICAL DISTRIBUTION OF PHOSPHATE AND NITRATE
IN THE DEEP WATERS OF THE SOUTH ATLANTIC OCEAN

The vertical distribution of phosphate along the meridian of 30° W is shown in
Plate IX. North of the Rio Grande ridge the greatest phosphate content is found in
the Antarctic intermediate layer, and the warm deep water contains less phosphate.
The warm deep water is not therefore the source of the high phosphate concentrations
which are found in the Antarctic Zone. Phosphate is added to the warm deep current
south of the Rio Grande ridge, and the source of the phosphate is probably in the rich
plankton of the Antarctic Zone which decomposes as it is carried downwards at the
Antarctic convergence into the bottom of sub-Antarctic water. The decomposition is
probably greatest in the region of greatest phosphate content and least oxygen content,
that is to say, between the north-going sub-Antarctic and Antarctic intermediate waters,
and the south-going warm deep water, between 43 and 38° S. The hydrogen-ion con-
centration is also greatest between these currents.

234 DISCOVERY REPORTS

The greatest vertical mixing between the two currents also takes place between 38
and 43° S, so that the phosphate is not lost from the Antarctic Zone but is returned to
it in the warm deep current.

North of the Rio Grande ridge, where there is a sharper discontinuity between the
Antarctic intermediate water and the warm deep water, the surface of the warm deep
water is not enriched as much as it is farther south.

In the Tropical Zone the surface stratum is depleted of phosphate, but lower down
the phosphate content of the water increases with depth to a maximum of about 120-
140 mg. PjOs/m.^ in Antarctic intermediate water.

There is a minimum phosphate content at the level of greatest oxygen content in
the North Atlantic deep water, and it is probable that the water at this level has
most recently been at the surface. In Antarctic bottom water the content increases to
90-100 mg. PaOs/m.^

The distribution of nitrate which is also shown in Plate IX is similar to that of
phosphate. The greatest concentration is found in 38-43° S between the warm deep
water and the sub-Antarctic and Antarctic intermediate waters. Nitrate is returned to
the Antarctic Zone in a cycle similar to that which returns phosphate.

The vertical distribution of nitrate shows that the regeneration of nitrate from de-
composition products takes place principally south of the Rio Grande ridge : the nitrate
content of the water north of the ridge is much less than that of the water south of
the ridge.

The surface water in the Tropical Zone is almost depleted of nitrate. Most is found
in the Antarctic intermediate layer, and a smaller concentration in the North Atlantic
deep water. The bottom water is richer in nitrate than the deep water.

The examination of the nitrite content of sea water has shown that in the Antarctic
Zone there can be as much as 6-7 mg. nitrite Ng/m.^ at the surface and 8 mg. nitrite
Na/m.^* at a depth of 80-100 m.; but as soon as a discontinuity appears in the water,
which makes vertical mixing with the surface water difficult, no nitrite is found below it.
None was ever found below 150 m.

Nitrite was found in sub-Antarctic water in amounts decreasing from 5-5 to 3-5 mg.
nitrite N./m.^ towards the north. North of the sub-tropical convergence there were
only small amounts of nitrite; but farther north, just below the sharp discontinuity at
the bottom of tropical water, amounts as large as 30 mg. nitrite Na/m.^* have since been
found. Nitrite is a stage in the formation of nitrate from animal decomposition products,
but we have rarely found it at a depth from which vertical mixing with the surface water
was difficult. The absence of nitrite from the deep layers in which there are the greatest
phosphate and nitrate contents is surprising.

HYDROLOGY OF THE SOUTH ATLANTIC 235

APPENDIX

THE WINDS OF THE ATLANTIC OCEAN
SOUTH OF 40° S

By Lieut. R. A. B. ARDLEY, R.N.R.

The accompanying table is compiled from the meteorological logs of the Research
ships 'Discovery', 'William Scoresby', and 'Discovery H', and checked from the
Meteorological Charts of the South Atlantic. In that publication, the data given for
latitudes south of 45° S are very meagre, since no trade route lies across the higher
latitudes of the South Atlantic. The data extracted from the logs of the three ships of
the Discovery Committee, covering a period of seven years, give a good general
analysis of wind conditions. A favourable indication of the accuracy of the final average
is that though each ship's observations were extracted and compiled separately, the
resulting averages for latitude agreed fairly well.

The only other possible source of information concerning oceanic winds in the high
latitudes of the South Atlantic would be the logs of the whaling factories voyaging to
and from the ice-edge ; but in the writer's opinion the averages here given are reliable
and would merely be confirmed by further information. The wind forces are measured
on the Beaufort Scale.

The relatively high average westerly wind force between the parallels of 40 and 50° S
is maintained by the almost constant prevalence of fresh to strong breezes in these
latitudes. Between 50 and 60' S, the wind forces are generally more erratic. The region
with the highest percentage of gales lies between 45 and 55° S.

The data obtainable are insufficient for compiling an average for easterly winds. Within
the latitudes under consideration, easterly winds are usually erratic and of short
duration; they blow either as light winds or as brief, violent gales.

In measuring the average force of the westerly winds all winds with a westerly com-
ponent are taken into account, for it is found that winds from the north-west and south-
west quadrants are fairly evenly balanced. Almost all the region north of latitude 55° S
lies within the northern semicircles of depressions, and the normal wind sequence,
subject to minor variation, is north-west, west, south-west.

South of latitude 55^ S, it will be seen that the percentage of easterly winds grows
steadily greater. Probably at the Antarctic Circle the easterly and westerly winds are
about balanced, and, south of that, easterly winds would predominate.

All the observations collected here were made in spring, summer and autumn, but in
this great stretch of ocean it is doubtful if the winter winds would show any large
deviation from the average. No account is taken of the ice conditions, though probably
when the pack is far north a slightly higher percentage of easterly winds and calms
would be found.

236

DISCOVERY REPORTS

The observations cover the whole of the South Atlantic from the meridian of Cape
Horn to the meridian of the Cape of Good Hope, and there appears to be no marked
difference in the incidence and force of westerly winds in different longitudes. We are
however hampered by the paucity of observations in different longitudes.

On the Patagonian shelf, and locally round the Falkland Islands, though strong
westerly winds are generally prevalent, they are affected by the proximity of the con-
tinent to the westward ; the averages are more erratic, and do not agree very well with
the averages in the table, which are for oceanic winds. Therefore they have not been
included.

South of Cape Horn, within a radius of about 150 miles of the Cape, there are plenty
of observations given in the meteorological charts, which agree very well with the results
given in Table XH.

Table XH
Westerly winds of the South Atlantic Ocean

No. of sets

Average force

Average force

Percentage of
easterlies

Latitude

of

of

X percentage of

°S

observations

westerly winds
(Beaufort)

westerly winds
(Beaufort)

and calnas

40

25

4-65 ■"!

5-28

3-94

16

41

18

4-39

16

42

17

5-13

4-51

II

43

18

5-64

4-35

22

44

18

5-05

470

5

45

19

5-08

4-29

16

46

18

5-25

4-93

5

47

22

5-13

473

9

48

19

5-45

5-14

5

49

21

-^ 5-80 —

4-65

17

50

29

5-58

4-41

21

SI

57

4-85 t

471

2

52

70

4-82J

4-58

7

53

105

4-53

3-90

14

54

140

4-64

378

18

55

82

5-33

4-44

16

56

74

5-33

4-48

15

57

46

4-61

3-68

20

58

50

4-56

3-55

19

59

43

4-52

372

22

60

47

4-21

2-95

30

61

48

4-00

2-37

39

62

37

4-31

2-66

43

63

26

3-97

2-37

43

Sum

mary for 5° intervals

40-45

—

5-14

4-37

15

45-50

—

5-38

473

12

50-55

—

4-95

4-31

13

55-60

—

476

3-8i

20

60-63

—

4-10

2-50

39

LIST OF LITERATURE

Brennecke, W., 1921. Die ozeanographischen Arbeiten der deutschen Antarktischen Expedition, 1911-1912.

Arch, deutsch. Seewarte, xxxix. Hamburg.
Castens, G., 1931. Stromgeschwindigkeiten und Wasserumsdtze in den Tiefen des Ozeans. Annal. Hydrog.

Mar. Meteorol., Hamburg, pp. 335-39.
Cherubim, R., 193 i. tjber Veidunstungsmessung aiif See. Ann. Hydrog. Mar. Meteorol., Hamburg, lix (9),

PP- 325-35-
Clowes, A. J., 1933. Influence of the Pacific on the circulation in the south-west Atlantic Ocean. Nature,

London, cxxxi, pp. 189-91.
Deacon, G. E. R., 1931. Velocity of Deep Currents in the South Atlantic. Nature, London, cxxvm,

p. 267.
Defant, a., 1928. Die systematische Erforschung des Weltmeeres. Zeit. Ges. Erdkunde, Berlin.

1929. Dynamische Ozeanographie. Berlin.

Drygalski, E. v., 1927. Deutsche Sudpolar-Expedition, 1901-1903, vii, Ozeanographie.
Ekman, V. W., 1926. Konnen Verdunstung und Niederschlag im Meere tnerkliche Kompensationsstrome verur-
sachen? Ann. Hydrog. Mar. Meteorol., Liv, p. 261.

1928. A survey of some theoretical investigations in ocean currents. Journ. Conseil, in. No. 3, pp. 295-397.

Herdman, H. F. p., 1932. Report on Soundings taken durijig the Discovery Investigations. Discovery Reports,

VI, pp. 205-36.
Jeffreys, H., 1925. On fluid motions produced by differences in temperature and humidity. Quart. Journ. Roy.

Meteorol. Soc, li, Oct. 1925.
Lenz, E. von, 1847. Bericht iiber die ozeanischen Temperaturen in verschiedenen Tiefen. Bull. Class, hist.-

philos. Acad. Sci. St Petersburg.
Meinardus, W., 1923. Ergebnisse der Seefahrt des Gauss. Deutsche Sudpolar-Expedition, 1901-1903, in,

Meteorologie.
Meyer, H. H. F., 1923. Die Oberfldchenstromungen des Atlantischen Ozeans im Februar. Veroffentl. Inst.

Meereskunde, Berlin, Reihe A, Heft xi, p. 35.
Moller, L., 1929. Die Zirkulation des Indischen Ozeans. Veroffentl. Inst. Meereskunde, Berlin, pp. 1-48.
'Mowe', S.M.S., 1914. Forschungsreise S.M.S. Mowe 1911. Arch, deutsch. Seewarte, xxxvii, pp. 51-2.
Nansen, F., 1912. Das Bodenwasser und die Abkuhhmg des Meeres. Int. Rev. gesamt. Hydrobiol. und

Hydrog., Leipzig, v, pp. 1-42.
'Planet', S.M.S. , 1909. Forschungsreise S.M.S. Planet, iii, Ozeanographie. Berlin.
ScHOTT, G., 1926. Geographie des Atlantischen Ozeans. Hamburg.
WiLLiMZiK, M., 1927. Die Antarktischen Oberfldchenstromungen Zwischen 50° O und 110° O. Veroffentl.

Inst. Meereskunde, Berhn, pp. 25-29.
WusT, G. O., 1928. Der Ursprung der Atlantischen Tiefemvdsser . Zeit. Ges. Erdkunde, Berlin, 1928,

pp. 506-34.
■ 1933. Schichtung und Zirkulation des Atlantischen Ozeans. Erste Lieferung. Wiss. Ergeb. der deutschen

Atlantischen Expedition auf dem Forschungs- und Vermessungsschiff Meteor, 1925-1927.

DISCOVERY REPORTS, VOL. Ml

PLATE VIII

Fig. I. Section III, distribution of temperature (°C.). April-May, 1931.

LATITUDE S 5Q 40° 3D 20

STATION „,: 663 666 868 671 673 675 677 ' 579 661 ' 684

(OOOm-

10°
687

aOOOM-

4000 m

Fig. 2. .Section III, distribution of salinity. April-May, I93'-
For position of Section III, see Fig. 11, p. igi-

N
699

DISCOVERY REPORTS. VOL. VII

PLATE IX

10° N

696 699

LATITUDE 5 5g° 40° 30°

STATION 663 666 668 671 ' 673 675 677 ' 679

Fig. I. Section. Ill, distribution of pliosphate content (mg. PjOs/m.^). .^pril-May, 1931

20°

6SI

684

687

20D0m

3000m-

4000 k,

I

Fig- 2. Section III, distribution of nitrate content (mg. nitrate-nitrite Nj/m.'). April-May, igjj
For position of Secliun III, see Fig. u, P- I9'-

N
599

DISCOVERY REPORTS, VOL. VII

PLATE X

;Fig. I. Section III, distribution of oxygen content (cc. O./litre). April-May, 193 1.
For position of Section III, see Fig. 11, p. 191.

PLATES I-VII

Discovery Reports. Vol. VII, pp. 239-252, Plates XI-XIII, December, 1933.]

WHALING IN THE DOMINION OF
NEW ZEALAND

By
F. D. OMMANNEY, A.R.C.S., B.Sc.

CONTENTS

Introduction P'^Se 241

History 241

BayWiialing 243

Modern whaling in New Zealand 247

List of Literature 252

Plates XI-XIII following page 252

T

WHALING IN THE DOMINION OF
NEW ZEALAND

By F. D. Ommanney, a.r.c.s., b.Sc.
(Plates XI-XIII; text-fig. i.)

INTRODUCTION

HE following account does not pretend to be more than a brief outline of the
history of the whaling industry in New Zealand from the beginning of the nine-
teenth century, and a sketch of the small industry as it exists to-day. No claim is made
that it embodies in any way the result of direct observations made by the author.

The chief sources of information in the first section of the paper were Dr Robert
McNab's The Old Whaling Days (Whitcombe and Toombs, Melbourne and London) and
the writings of Dr Ernest Dieff"enbach {Travels in New Zealand, Murray, London, 1843),
who was naturalist on board the 'Tory', the exploration ship of the New Zealand
Company. In addition to the above very little has been written on the subject of
whaling in the Dominion. A few other references, however, were used and appear in
the list of literature.

The second section of the paper, dealing with the two modern whaling stations, is
compiled almost entirely from conversations with Mr H. F. Cook, manager and part
owner of the station at Whangamumu, who visited the R.R.S. 'Discovery 11' at
Auckland, and with Mr Joseph Perano, owner and manager of the whahng station at
Te-Awaiti, Tory Channel.

The author takes this opportunity of thanking these gentlemen for their courtesy in
supplying information and for their interest in the work of the Discovery Committee.
When the R.R.S. ' Discovery II ' was at Wellington in August 1932 some of the scientific
staff, including the author, paid a visit to Te-Awaiti and were received by Mr Perano
and his family with the greatest hospitality. The author is also indebted to Messrs
A. W. B. Powell and R. A. Falla, of the Auckland Museum and Institute, who reviewed
the manuscript of this paper and made certain corrections.

HISTORY

Attention was first drawn to the whaling grounds of New Zealand by the writings of
Cook, who visited the islands in 1770, 1773 and 1774. In 1791 a fleet of whalers, bound
for the coast of South America, was carrying convicts and stores to Australia. They
reported great numbers of Sperm whales in Australian waters. After the convicts and
stores had been landed they made a trial of the grounds, but reported that, while whales
were abundant, the weather was too bad for profitable fishing. They continued to the
old grounds oflF the coast of Chile and Peru.

In 1798 it became no longer possible for whalers to recruit their ships at Pacific ports

242

DISCOVERY REPORTS

owing to the war between England and Spain. Until that year whalers had been
prevented from operating in Australasian waters by the limits imposed by the British

Fig. I

East India Company, who controlled by permit all whaling operations between longitudes
51° and 180° E. New Zealand Sperm whaling thus really dates from the year 1798, when,
as a result of the reported presence of Spanish warships off Cape Horn, whalers were
instructed to pass into Australasian waters for the duration of the war (McNab, 1914).

NEW ZEALAND WHALING 243

The Sperm whaling trade was carried on mainly by British and American ships,
while a small though increasing proportion was done by Sydney and Hobart firms
(McNab, 1913, p. 260). The Sperm whalers called mainly at the Bay of Islands at
the northern end of the North Island, but also at Doubtless Bay in the extreme north,
and at Cloudy Bay at the northern end of the South Island. This latter locality, with
the neighbouring Queen Charlotte and Marlborough Sounds, later became the main
centre of the New Zealand Right whale industry. The Sperm whalers established bases
in these bays at which the ships could be repaired and careened, and arrangements
were made with the natives for the supply of foodstuffs, the whalers trading iron for
potatoes and flax (McNab, 1914). The natives often went aboard the whaling ships
to assist in the disposal of the carcase and the getting of the oil.

At the end of the third decade of the nineteenth century (1827-30) the number of
Sperm whales in New Zealand waters became seriously diminished (McNab, 1913,
p. 2), and the increasing demand for Right whale oil and whalebone drew greater
attention to the habits and pursuit of the southern Right whale. The year 1830 saw the
beginning of the Right whale industry, which, for some years, continued side by side
with Sperm whaling and finally became the more important of the two. Right whaling
continued as a flourishing industry until 1840, when ruthless overfishing began to tell
seriously upon the stock. The fishing was carried on both in the open sea and among
the bays and sounds around the coast (bay whaling). The vessels engaged in the trade
were at first almost all fitted out and owned at Sydney or Hobart. In the year 1834 the
first British and American vessels took part, and after that year the number of American
ships whaling around New Zealand steadily increased. In 1840 there were between six
and seven hundred American whalers distributed around the coasts of the two islands
(CondHffe, 1930, p. 115). This was about half the total number for all nations. The
Americans carried on most of their operations at Cloudy Bay in Cook Strait, but also
at Kapiti Island and farther south at Banks Peninsula or Bluff Harbour or farther north
at the Bay of Islands and Doubtless Bay. The British whalers never seriously rivalled
the Americans, even when a tariff' of ^26. 125. od. per ton was introduced by the British
Government upon oil carried in foreign ships (McNab, 1913, p. 261). In 1838 a
number of French ships were operating at Banks Peninsula and the corvette ' Heroine '
was sent out "to overlook the interests of France and to maintain order in, and give
help to, their numerous whalers in the South Pacific". Portuguese and Dutch ships
also entered the business at about this time.

BAY WHALING
Much of the Right whale industry was carried on by the method known as "bay
whaling". This branch of the fishery derived its name from the Right whales' habit
of entering shallow bays and inlets along the coast for the purpose of giving birth to
their calves. "These fish", wrote Dieffenbach (1843, p. 47), "approach the shores and
bays with the flood tide and quit them with the ebb . . . they are often seen in places
where the depth of the water does not much exceed their own breadth, rubbing their

244 DISCOVERY REPORTS

huge bodies against the rocks and freeing themselves from the barnacles and other
parasitical animals with which they are covered". The Right whales arrived off the
coast of New Zealand at the beginning of May from the northward, and the cows
entered the bays to calve throughout May, June and July. Nearly all the Right whales
killed were cows with their calves, since the bulls rarely approached the land so closely
and were much more shy and wild. The cows were joined by the bulls later in the season
and both cows and bulls put to sea together. In October and November they returned
north and east, some through Cook Strait and some through Foveaux Strait. According
to Dieffenbach they began to show themselves at the Chatham Islands from June
onwards and their numbers increased in that locality towards the end of the season.
' ' During the six remaining months of the year the ships cruising in the ' whaling ground '
fall in with many whales. This whaling ground extends from the Chatham Islands to
the eastward of the North Island of New Zealand and thence to Norfolk Islands."

Whaling operations were carried on from ships anchored in the bays, particularly
Cloudy Bay or the Bay of Islands, or from stations established on shore. These shore
establishments eventually became very numerous and came to resemble well-equipped
factories with houses and plant erected in suitable bays, from which the bay whaling
could be carried on by means of a number of boats. In the early bay whaling days,
however, they were no more than try pots erected on the beach with huts for the
storage of whaling gear and boats. McNab quotes a Mr Bell, who owned a station
in Cloudy Bay, and who gave a description of bay whaling. " If the fishing is to be
carried on by means of a shore party the try pots and huts are erected on the beach
and the vessel which brought the party down is either employed in collecting flax along
the coast or returns to Sydney and is sent down again at the end of the season to bring
them up with what oil they may have caught. The boats are sent out at daylight every
morning and, when they are so fortunate as to kill a fish, it is towed ashore and ' flinched '
and boiled up on the beach. When the fishing is carried on in a vessel the blubber is
boiled in try pots erected on deck as in a Sperm whaler. . . . The whales are seldom
killed nearer than two miles from the harbour and sometimes seven or eight. . . . The
depth of the water in the bays where the whales are killed is 14-20 fathoms." Look-out
posts were established on an elevated part of the coast near the plant or the anchorage
so that warning could be given when the spout of a whale was seen. The operation of
"flensing" or "cutting-in" was presumably carried out in the manner described by
Scammon (1874, p. 235) at Californian bay whaling stations, since Dieffenbach
mentioned the "shears", a gallows or scaffolding erected near the shore, by means of
which the carcase was suspended at the surface of the water. It could then be turned
and rolled over by tackle while the blubber was being stripped off. When the whaling
was carried on from a ship the shears were lowered over the side. "The blubber was
cut off in square pieces by means of a sharp spade, carried to the shore and put into
the try pots " (Plate XIII, fig. 2). At American whaUng stations the blubber was cut off
in a series of spiral folds.

That this was the method of flensing employed by the bay whalers seems fairly

NEW ZEALAND WHALING 245

certain from Plate XIII, fig. 2, which illustrates the process being carried out in the early
days of the modern station at Whangamumu, Bay of Islands. The shears can be seen
in the background, while the blubber of one of the carcases is being cut into square
sections.

The yield of oil per whale varied from two to thirteen tons and averaged six tons.
Cows were larger than bulls and yielded more oil, but became thin towards the end of
the season from supporting the calves. "It is a pity", wrote Mr Bell, "that it should
often be necessary to fasten to the calf in order to secure the cow." The whalers made
a practice of taking the calf first since it was inexperienced and slow, and the cow then
became an easy prey, refusing to leave the calf. The boat which killed the calf claimed
the cow, even if the latter were captured by a different crew.

With the development of bay whaling, more elaborate shore stations were established
all round the coast, especially at Preservation Inlet, Cloudy Bay, Otago and Marlborough
Sound. Many of them were well-equipped factories employing, perhaps, a hundred
Europeans. The earliest shore stations were established in New Zealand about 1830 at
Te-Awaiti in Tory Channel, and at Preservation Inlet. At the latter place there was a
dwelling-house for the manager and his family and a large storehouse. There were also
six houses for the use of other whaling companies and a shed for sixteen boats (McNab,
1913, p. 85). Thus it seems that at some at least of the stations accommodation was
available for the use of whaHng ships. Diefl^enbach (1843, p. 371) records that at the
Te-Awaiti station some of the houses "were substantial wooden buildings, but the
majority had thatched walls of hands and bulrushes".

Preservation Inlet employed some fifty or sixty Europeans, who were engaged in
sealing or sawing timber when the whaling season was over. At Te-Awaiti, however,
Diefl^enbach wrote that in the summer season the whalers lived "dispersed over the
(Queen Charlotte) Sound, sometimes trading in a small way with passing ships in
potatoes and pigs. . .but more generally passing their lives in idleness". Most of the
stations possessed a ship which brought stores from Australia and took back oil at the
end of the season. The main lines, in fact, upon which the stations were run do not
seem to have been very diflFerent from those of a modern Norwegian shore station. Each
one manned perhaps half a dozen boats.

Each boat's crew at most of the stations, as on the ships, consisted of the usual five
or six oars (either Europeans or natives), a "headsman" or "boat-header", who was
officer-in-charge, and a " boat-steerer ". The headsman and steerer were always
Europeans. Scammon (1874, p. 226) wrote, "The officer-in-charge, or boat-header,
in the stern . . . steers the boat with the steering oar, which is usually 22 ft. long ; the
boat-steerer pulls the oar farthest forward, which is called the harpooner oar. (The
boat-steerer) darts the harpoon, and, after the boat is fast, changes ends with the
boat-header and steers the boat while the latter attends to killing the whale ". Scammon
was writing of the Californian bay whalers, but the methods used in New Zealand seem
to have been the same.

The whalers in New Zealand came into close contact with the natives and often into

246 DISCOVERY REPORTS

conflict with them. The Maoris made frequent raids upon the shore stations and burnt
down buildings ; but they also traded with the whalers for fire-arms, supplying in return
potatoes and vegetables generally. The Maoris, also, in many places, used to engage
in whaling on their own account and sell their captures to the whaling ships or to the
shore stations. At Te-Awaiti they used to obtain j^zo apiece for the carcases of the
whales which they took.

The whalers used to arrange with the local chiefs for the use of their coves and bays
for whaling, and McNab records that these arrangements not infrequently led to
jealousies and feuds between the chiefs. The natives supplied the whalers with wood,
fuel and water, and often came on board the ships to assist in dealing with the carcases.
In this connection McNab states that the American whalers were at a serious disad-
vantage compared with their Australian competitors because they lacked intimate
knowledge of the Maori and of his language. Owing to the great distance from their
home ports the Americans were especially dependent on the natives for food supplies
and the recruiting of labour. They frequently had to make use of interpreters and, in
Cloudy Bay, there were two or three men who acted in this capacity. These interpreters
were usually Europeans, runaways from ships, and were locally known as "tonguers".
Each "tonguer" had a boat and a number of natives attached to him. On the arrival
of a vessel in Cloudy Bay he Vv'ent on board and canvassed for employment, which
consisted of interpreting and furnishing a boat's crew to help to tow in the dead
whales and to cut them up. The remuneration for these services was the carcase
and tongue of the whale. Only the blubber was taken by the whalers and the tongue
was left untouched. This provided the "tonguer" with about six or eight barrels
of oil.

McNab records that, although American and English whalers were carrying on their
trade side by side during a time when their countries were at war, the most friendly
relations existed between them. They seldom quarrelled and frequently formed treaties
of mutual assistance, under which they helped to tow in each other's whales and com-
bined against the assaults of the natives.

The year 1839 saw the maximum development of the Right-whale industry in New
Zealand. After 1840 its history is only that of a steady decline resulting directly from
overfishing. In 1892, the first year for which figures are given in the New Zealand
Year Book, the whaling industry in the Dominion had become insignificant, 3100 gallons
of Sperm-whale oil and 1572 gallons of Right- whale oil being taken.

In the years immediately before the war of 1914-18 the attention of Norwegian
whalers was directed towards the waters around New Zealand. The rapid development
of the South African fishery during the years 1908-11 led to the supposition that a
similar whaling ground would be found to exist off the coast of the Australasian land
masses, to which whales from the Antarctic Ocean would migrate during the southern
winter. The somewhat vague reports brought back from these waters by sealers and
others, and the knowledge that flourishing Sperm and Right-whale industries had once
existed in this region, encouraged several firms to fit out costly expeditions to explore

NEW ZEALAND WHALING 247

the seas around Australia and New Zealand. Risting (1922) gives some account of
these ventures, all of which met with disastrous failure.

The A/S 'Laboremus', in the autumn of 191 1, sent the sealer 'Mimosa', fitted out
as a floating factory, with a whale-catcher, to explore the coast of Tasmania. She failed
to find whales around Tasmania and explored the New Zealand seas without any
success. The "New Zealand Whaling Company" of Larvik, under the same manage-
ment as the "West Australian Whaling Company ", sent the factory ship ' Rakiura ' and
four whale boats to New Zealand in March 1912. The ship was for a time stationed
in the Bay of Islands, and Lillie (1915) records its presence there during the months
July to October, 19 12, when he visited the Whangamumu whaling station. Lillie
mentions also the sailing ship ' Prince George ' with which the company was apparently
also working in the Bay of Islands at that time. The ' Rakiura ' worked only for one
season and thoroughly explored the seas around New Zealand from the Antarctic to
the Kermadec Islands in the north and Campbell Island in the east. No success was
met with, however, and the Company abandoned the area and obtained whaling rights
on the west coast of Australia, becoming the " Fremantle Whaling Company ". Another
expedition also met with failure — that of the A/S 'Australia', which fitted out a ship,
the 'Loch Tay', and two whale boats, to explore the eastern coast of Australia. After
extensive voyages over the whole area, southward to New Zealand and eastward to
Campbell Island, the expedition went to Blufl^, in the South Island of New Zealand,
and for a short time, from January to April 1913, carried on a small Sperm-whale
fishery.

These failures were made only a little less disastrous by the discovery of pieces of
ambergris on one or two occasions. They demonstrated beyond question that the
number of whales to be found in Australasian waters is much too small for large-scale
undertakings, and no further attempts have been made to exploit the region on these lines.

MODERN WHALING IN NEW ZEALAND
There are at present two whaling stations operating on a small scale in New Zealand.
The older of these two stations is situated at Whangamumu, Bay of Islands, and the
younger in Tory Channel, Queen Charlotte Sound, at the northern end of the South
Island. In the bay whaling days there were several shore stations in the Tory Channel
near the site of the modern one.

The Whangamumu station was established by Mr H. F. Cook in 1890, and is still
managed and part-owned by him. Mr Cook shipped to New Bedford and back in one
of the last American Sperm- whaHng ships which visited New Zealand before he started
his own whaling factory. The station was not operating during the winter of 1932 so
that no visit was made to it, but something of the methods employed there was learnt
from a conversation with Mr Cook in Auckland. D. G. Lillie, in the winters of 191 1
and 1912, visited the Whangamumu station and the factory ship 'Rakiura' when she
was lying in the Bay of Islands. He made a study (Lillie, 1915) of the Humpbacks
brought in during the time he was at Whangamumu, and his account is confirmed in

248 DISCOVERY REPORTS

many respects by that which follows. The station was established for the purpose of
taking the Humpback whales which come close in shore along the coast of New Zealand.
Right whales and Blue or Fin whales are very rarely taken. The total catches at both
of the New Zealand stations are very small, seldom exceeding 70 whales, and averaging
about 50 whales, in a season.

At Whangamumu open boats and hand harpoons were used at first for the capture
of Humpbacks. The method used was an unusual one. A rope cable was stretched
between a rock and the shore, across a channel some 50 yards wide, or was buoyed out
from the shore in positions where whales were known to pass regularly and were likely
to be intercepted. Sections of net were suspended from the cable and the whales became
entangled in them (Plate XHI, fig. i). In their struggles to free themselves they used to
carry away sections of the net, so that their progress was impeded and they became an
easy prey for the harpooner. The nets were at first made of rope but later meshes made
of I in. wire, shackled together, were found to be more durable and less bulky. A similar
process has been used in Japan for the capture of Humpbacks in shallow channels
(Mobius, 1893). However, the largest annual catch by means of the nets was only
nineteen whales.

In 1910 a steam whaler was built for the firm and the use of the nets was abandoned.
The station now uses this catcher and two fast motor-boats mounted with light harpoon
guns. These boats are of the same type as those used by the Te-Awaiti station, illustrated
in Plate XI, fig. 2. The average catch per season amounts to 50 to 60 Humpback whales,
while 70 is a record catch. About 5 tons (30 barrels) of oil per whale are obtained.
The whole of the carcase is used, the blubber being reduced in open boilers and the
meat and bone in pressure boilers. Bone meal and guano are made from the residues.
The station is modelled upon Norwegian lines but is on a relatively very small scale.
The workers are all Maoris whose families have been in the firm's employment since
the inception of the station.

The station operates for two separate periods each year. The first period covers the
months of June, July and part of August. The second period begins on October i and
continues until the end of November. During the first of these two periods the whales
are travelling northwards. Lillie wrote (p. no), "The first whales began to pass the
Bay of Islands on their way northward about the middle of April. They continued to
go north until the end of August. The greatest number passed northward of this locality
in May and the early part of June ". During the October-November period the whales
are travelling southwards. "After the middle of September", continued Lillie, "at the
Bay of Islands, the first members of the long procession were seen going southwards.
The majority passed south of the Bay during October and by the middle of December
they were all south of this place."

The direction of the Humpbacks' seasonal migration is thus opposite to that of the
Right whales but occurs at the same time of year. At the beginning of the season the
majority of the cows are carrying large foetuses. In July Lillie saw a foetus 13I ft. in
length ; the length of the Humpback calf at birth being about 15 ft. During the south-

NEW ZEALAND WHALING 249

ward migration the majority of the cows are lactating and the calves not much more
than 15 ft. long — evidently newly born. Mr Cook stated that many of the cows with
sucking calves of about this length are also carrying foetuses about 2 in. in length. In
October Lillie saw a foetus 2I in. in length. The south-bound whales have a greater
or lesser quantity of food in the stomach, while the stomachs of the north-bound
whales are nearly always empty. Matthews (1932), from the various accounts given
to him, identifies the food of New Zealand whales as the Grimothea larva of Munida
gregaria. This animal occurs in shoals around the coast and Mr Cook spoke of it as
colouring the sea red. The present author had no opportunity of verifying any of the
somewhat confused descriptions which he heard of this "whale feed", but Matthews
is in no doubt about its identity.

Occasional Right whales are taken at the Whangamumu station. When the steam
whale-catcher was first built Mr Cook made an expedition with it to Campbell Island,
having heard reports of numerous Right whales in that area: the expedition was,
however, unsuccessful.

During the month of June, around the Friendly Islands, the copulation of Humpback
whales has been observed. According to Mr Cook the animals lie together in the water
obliquely with the axes of their bodies at an angle of about 45° with the surface and
their ventral surfaces apposed, so that only their heads project. The male assumes the
uppermost position and embraces the female with his flippers.

The Tory Channel whaling station at Te-Awaiti was established in 1909 by Mr
Joseph Perano, who was a fisherman without previous experience of whaling. He has
evolved a unique method of chase and capture without any knowledge at all of Norwegian
methods.

Shore whaling stations were situated in Tory Channel in the old bay whaling days.
The region of Queen Charlotte and Marlborough Sounds is remarkable in that it is
an area of very recent post-Pleistocene subsidence, which " drowned " the valleys between
the mountains, leaving in their place long parallel fjords, often less than two miles
across. The Tory Channel is a narrow strip of water at the outer end of this system
of fjords connecting Queen Charlotte Sound with Cook Strait. The whaling ships found
these tortuous channels difficult to negotiate and resorted to the more convenient
Cloudy Bay. Tory Channel was therefore left entirely to the shore establishments
and, during the bay whaling period, quite a large number of stations were operating in
the channel. When they were all fully working and bay whaling was at its height some
twenty boats used to go out from Te-Awaiti alone. The earliest station at Te-Awaiti
was that of Captain John Guard, who was driven into Tory Channel in 1827, when
in command of a sealing expedition. He built a home for himself and, as seals became
scarcer in the neighbourhood, took up the chase of the Right whale, "with great risk
and annoyance from the natives", who frequently burnt down his buildings, "and no
profit to himself". He was unable to keep the oil of the whales he killed, owing to lack
of men and gear, and took only the baleen which he sold to passing ships (McNab, 1913,

250 DISCOVERY REPORTS

p. 297). Guard was succeeded at Te-Awaiti by Messrs Barrett and Thorns, who were
also at first without resources and killed the whales only for their baleen (Dieffenbach,
1843, p. 39). Later, however, speculators in Sydney supported them with gear and
ships for the transport of oil. Besides the two stations of Barrett and Thorns at Te-Awaiti
there were others, and one in the adjoining cove Jackson's Bay. The Maoris also from
the adjoining "pas" or villages used to pursue whales on their own account with their
own boats, selling their captures to the stations for £20 each. All these stations,
however, fell into disuse when bay whaling came to an end about 1840.

The modern station at Te-Awaiti operates only during June, July and August.
There is no second appearance of the Humpback whales during October and November
as at Whangamumu, apparently because they do not pass through Cook Strait during
their southward migration. In June, July and August they are travelling northwards,
as at the Bay of Islands, and appear in Cook Straits usually on the flood tides. Mr Perano
said that most of the whales taken by his station are bulls. Such cows as are caught are
carrying large foetuses about 14 ft. in length. A few Southern Right whales are taken,
especially towards the end of the season.

The method of capturing and killing the Humpbacks which Mr Perano employs is
so diff'erent from the Norwegian method that some description of it may be worth
while. As is well known, the Norwegians use fast steam whale-catchers with a heavy
harpoon gun mounted in the bow. The harpoon is fired by means of a charge of
black powder and has, at its head, three hinged iron barbs which are lashed to the
stem of the harpoon. The barbs become unlashed on entering the whale's body and
project outwards so as to prevent the withdrawal of the harpoon. In front of the
barbs at the head of the harpoon is a heavy conical grenade of cast iron, filled also with
black powder and exploded by means of a time fuse. The fuse is ignited by the discharge
of the harpoon from the breech of the gun. The whale is killed by explosion of the
grenade after the harpoon has entered its body. The harpoon carries out with it a stout
rope line which is coiled in a hold of the ship abaft the gun-mounting and runs from
the hold to the gun round the drums of the steam winch and over a system of
accumulator blocks fitted to the mainmast. The first fifty fathoms of this line are of
smaller diameter than the remainder and are coiled in readiness upon a platform in
front of the gun mounting. This lighter rope is the ' ' forerunner ' ' which is carried
out by the harpoon immediately it leaves the gun. After the whale has been struck
the rest of the harpoon line follows and the whale is, as it were, "played" like a fish
at the end of the harpoon line. He usually "sounds", or dives deeply, after being
struck, or thrashes about in the water. If the harpoon has not hit fairly into the back
it may be necessary to fire a second harpoon. If that is so the whale is hauled within
range by the harpoon fine wound upon the drum of a steam winch. After the death
of the animal the carcase is made fast by the tail alongside the ship, the flukes being
cut off for convenience. The body wall is pierced by a long lance carrying the end
of a pipe leading from an air pump. Air is pumped into the carcase so as to render
it buoyant, since, after expulsion of the air from the lungs, the carcase of the Blue or

NEW ZEALAND WHALING 251

Fin whale sinks. The carcase is then towed back to the whaHng station tail foremost
alongside the ship.

The Norwegian method is not suitable for use in the narrow Tory Channel and
the shallow bays around the Queen Charlotte Sound where the Humpback whales
are found. Mr Perano uses three fast motor launches, 34 ft. long and capable of
maintaining a speed of 30 to 40 knots (Plate XI, fig. 2). These boats can be stopped or
turned almost within their own length. Each has a light harpoon gun (i| in. bore)
mounted in the bows (Plate XII, fig. 3). The harpoon is similar to that used by the
Norwegians but much lighter in build (Plate XII, fig. 4). It has slightly curved barbs
and the cast iron grenade is triangular in section. The harpoon line, which is con-
siderably lighter than that used by the Norwegians in their steam catchers, is coiled
in the stern of the launch and pays out from that position when the shot is fired. The
explosion of the grenade stuns but does not kill the whale. After the shot the launch
is brought up close to the whale and the body is inflated with air in the Norwegian
manner. After inflation the whale is finally despatched by inserting into the thorax
ventrally a long lance with a hollow cast iron head. The head is filled with a pound
and a half of gelignite which is exploded within the whale's thorax by means of an
electric detonator. The carcase is then towed back to the station tail foremost at the
stern of the launch.

Mr Perano and his two sons encamp during the season on a headland overlooking
Cook Strait, where they keep a look-out for whales spouting. As soon as one is sighted
they put out in pursuit of it in their launches which are anchored below. When Mr
Perano 's station was first established there was, for some years, a rival station on the
other side of Tory Channel, and Mr Perano told the writer of races between the boats
of the rival stations for the same whale. These trials of skill were frequently attended
with great risk and damage sometimes occurred, nor did the fact that the other station
was owned by Mr Perano's brother diminish the keenness of their rivalry.

The above method of pursuit and capture is very successful for the chase of the
Humpback or Southern Right whale, especially within a confined space such as the
Tory Channel, but Mr Cook, who works more in the open sea around the Bay of
Islands, has been more successful with the steam chaser. Southern Right whales can
be towed easily by the launches since they float when dead. Humpbacks float just long
enough to enable them to be inflated with the small gear available. Blue and Fin
whales apparently sink almost immediately and have to be towed back along the
bottom of the channel with great difficulty.

The factory plant itself at Te-Awaiti (Plate XII, fig. 2) is very small. There are two
pressure boilers, a small flensing slip and three storage tanks. The capturing and
dismemberment of the whales and the general work of the station is done by Mr Perano
and his two sons, with the assistance of a small staff of Maoris. Only the blubber of
the whale is used and the oil is sold in Australia, where it is used mostly in the manu-
facture of hemp rope.

It is, perhaps, not out of place to conclude this short account by recording two

253 DISCOVERY REPORTS

instances of recovered harpoons which Mr Perano related to the writer. They are
remarkable in that they seem to point to the fact that Humpbacks at any rate, if they
do not use fixed migratory routes, are at least in the habit of returning at intervals to
the same place. Before the introduction of the harpoon with the explosive grenade,
Mr Perano used a type of hand harpoon. This had an S-shaped head swivelled about
its centre at the end of a long shaft. With this harpoon it sometimes happened that
whales were lost and carried the harpoon away inside them. It was not, however, a
very successful pattern and its use was practically discontinued after the introduction
of the grenade. At any rate no whale was lost with one of these hand harpoons inside
it after the introduction of the grenade. Mr Perano stated that a Humpback had been
taken at Te-Awaiti, not very long before the visit of the writer to the station, with a
swivel head embedded in the musculature, and that this harpoon must have been fired
at least eighteen years previous to the capture of the whale. He was able to identify
the harpoon head as one of his own.

When the explosive grenade was first introduced at Te-Awaiti a pattern was used
which had three flanges about its tip. Its use was discontinued when the triangular
grenade replaced it. The flanged pattern was not successful and was only used for one
season. Seven years after that season a Humpback whale was taken in which was found
a portion of a grenade of the flanged type. A harpoon grenade of that pattern was
not used at any other whaling station or factory. These two harpoon heads are now
in the Auckland Museum.

LIST OF LITERATURE

CoNDLiFFE, J. B., 1930. New Zealand in the Making. A survey of economic and social development. Allen
and Unwin, London.

DiEFFENBACH, E., 1843. Travels in New Zealand, Vol. i. Murray, London.

LiLLiE, D. G., 1915. Cetacea. British Antarctic (Terra Nova) Expedition, 1910. Zoology, i, No. 3,
British Museum.

Matthews, L. Harrison, 1932. Lobster-Krill, Anomuran Crustacea that are the food of whales. Dis-
covery Reports, v, pp. 467-484.

McNab, R., 1913. The Old Whaling Days. A history of Southern New Zealand from 1830 to 1840.
Whitcombe and Toombs, Melbourne and London.

1914- T'rom Tasman to Marsden. Wilkie, Dunedin, N.Z.

MoBius, K., 1893. Ueber den Fang und Verwerthung der Walfische in Japan. Sitzungsberichte der Koniglich
Preussischen Akademie der Wissenschaften zu Berlin, Lii, pp. 1-20.

New Zealand Government Year Book, 1893.

RiSTiNG, S., 1922. Av Hvalfangstens Historic. Kristiania.

ScAMMON, C. M., 1874. The Marine Mammals of the North-Western Coast of North America. San Francisco.

: lii ■:■!./

>swim'imi^toO s^'oiW -iiVx "^p \vomu> \'.\^o*\<\'d5\.

PLATE XI

Chasing Humpback whales off the coast of New Zealand.

Fig. I. The whale "sounding" after being struck, throwing the tail

flukes clear of the water.

Fig. 2. The whale has "sounded" and the slack line is being hauled in.

The photograph illustrates clearly the type of motor launch used at the

New Zealand stations.

Fig. 3. The whale has returned to the surface after the "sound", his

nose protruding above the surface in his struggles. The gunner is

preparing to fire a second shot.

Fig. 4. The whale has been stunned, but not killed, by the explosion

of the harpoon grenade inside its body. The body is now being inflated

with air.

Reproduced by permission of the High Commissioner for New Zealand.

DISCOVERY REPORTS, VOL. VII.

PLATE XI.

NEW ZEALAND WHALING

f

PLATE XII

Fig. I. A Humpback whale on the flensing slip at Whangamumu, Bay
of Islands, New Zealand.

Fig. 2. The whaling station at Te-Awaiti, Tory Channel, with Hump-
back whale on the slip.

Fig. 3. Light harpoon gun used by the New Zealand whalers for the
capture of Humpbacks with motor launches.

Fig. 4. Harpoon of the type used at the Te-Awaiti station. Note the
slender shaft, curved barbs and triangular grenade at the head.

Figs. I and 2 reproduced by permission of the High Commissioner of
New Zealand.

DISCOVERY REPORTS, VOL. VII.

PLATE Xn.

S^*'^^

^ 1 1

,^'.4'4ttUkii

Jaliii Bale Suns & DHiiii.-luan.L''' Lanilim,

NEW ZEALAND WHALING

PLATE XIII

Fig. I. Capture of Humpback whales by means of nets near Whan-
gamumu, Bay of Islands. The nets are suspended from a cable attached
to a line of buoys for a distance of 50-100 yards from the shore, as
shown in the photograph, or suspended between rocks across a channel
through which the whales are known to pass. The photograph, taken
about 1900 by the late Josiah Martin, shows the wash of a whale which
has become entangled in the net. The rowing boat is standing off in the
distance with the harpooners. At the time when this photograph was
taken hand harpoons were used.

Fig. 2. Two Humpback whales in the water at Whangamumu, Bay of
Islands. This photograph, taken also by the late Josiah Martin about
1900, shows the old method of flensing the carcases in the water. The
"shears" for manipulating the carcase in the water can be seen in the
background. The blubber of the whale farthest from the camera is being
cut into square sections.

From photographs lent to the author by Mr A. W. B. Powell.

DISCOVERY REPORTS, VOL. Vn.

PLATE Xni.

•

NEW ZEALAND WHALING

[Discovery Reports. Vol. VII, pp. 253-362, Plate XIV, December, 1933]

ISOPOD CRUSTACEA

PART I. THE FAMILY SEROLIDAE

By
EDITH M. SHEPPARD, M.Sc.

CONTENTS

Introduction pftge 255

List of species 256

List of stations 256

Geographical distribution 264

Classification 267

General morphology 271

Key to all known species of Serolis 278

Description of species . . . . • 282

List of references 360

Plate XIV following page 362

ISOPOD CRUSTACEA

PART I. THE FAMILY SEROLIDAE

By Edith M. Sheppard, m.Sc.
(Plate XIV; text-figs. 1-22)

INTRODUCTION

THE present report is based on the collections of Serolids made by the R.R.S.
'Discovery', the R.R.S. 'Discovery 11', the R.R.S. 'William Scoresby' and the
staff of the Marine Biological Station at South Georgia during the years 1925-32. Most
of the material was obtained from the shallow^ waters around South Georgia, the South
Orkneys, the South Shetlands, the South Sandwich Islands, Palmer Archipelago, and
the Falkland Islands, as well as off the coast of the southern part of South America and
in the shallow waters between the latter and the Falkland Islands.

The present collection is undoubtedly the most complete ever made both with regard
to the number of species as well as to the actual number of specimens procured. Of
the fifteen shallow-water species (excluding the two doubtful ones, Serolis serrei, Lucas,
and S. plana, Dana) previously recorded from these waters, all excepting three, S. polaris,
Richardson, and S. laevis, Richardson, from the shores of the South Sandwich Islands,
and S. paradoxa, Fabricius, from Patagonian waters, have again made their appearance,
and to this number may be added a further seven, which are new to science. Perhaps
the most striking feature of the collection is the extraordinary abundance, throughout the
year, of the two species S. schythei, Liitken, and S. exigiia, Nordenstam, both of which
have been recorded from a large number of stations to the north and south of the
Falkland Islands, as well as in the shallow waters between the islands and the mainland.

By kind permission of the Discovery Committee this report also contains a revised
account of the genus Serolis with diagnostic characters of all the known species, to-
gether with notes on their geographical distribution and general morphology.

I wish to take this opportunity of thanking the Discovery Committee for entrusting
me with the examination of this valuable collection. I also wish to thank Dr Caiman
for his helpful advice, and for the facilities he has given me for the examination of the
specimens at the British Museum. My thanks are due also to Professor Ch. Gravier, of
the Museum d'Histoire Naturelle, Paris, Dr P. H. Grimshaw, of the Royal Scottish
Museum, Edinburgh, Dr Waldo Schmitt, of the U.S. National Museum, Washington,
and Professor Dr von Straelen, Directeur du Musee Royal d'Histoire Naturelle de
Belgique, who, by the loan of specimens, have enabled me to examine all except one
{S. bakeri, Chilton) of the existing species of Serolis.

Lastly, I wish to express my deep gratitude to Professor W. M. Tattersall, University
College, Cardiff, for his valuable advice and unfailing interest in the preparation of this
report, as well as for the loan of numerous papers from his valuable carcinological library.

256

DISCOVERY REPORTS

Since sending this paper to press, further zoological results of the Swedish Antarctic
Expedition, 1901-3, have been published in a paper on " Marine Isopoda" by Norden-
stam (1933). This contains a section on the family Serolidae and wherever necessary
references to it have been inserted in the present paper.

LIST OF SPECIES

The genus Serolis contains thirty-seven species, excluding the two doubtful ones,
S. serrei, Lucas, and S. piano, Dana, and nineteen of these, six of which are new to
science, are represented in the present collection.

In the following list, species appearing in this collection are marked with an asterisk
(*), other Antarctic Expeditions are indicated by the following letters: B = Belgica;
C = Challenger ; D = Discovery ; F = Franfais ; G = Gazelle ; Ga = Gauss ; P =
Pourquoi Pas?; S = Scotia; and T = Terra Nova.

*i. S. beddardi, CdXman.

2. S. lafifrons. White. C, G, Ga.

3. S. gracilis, Beddard. C
— 4. S. paradoxa, Fabricius. C, G.

*5. 5. schytlici, Liitken. C, G.

6. S.polaiis, Richardson. F.
*7. S. glaciaIis,TntteTsalL B, T.
*8. S. septemcarinata, Miers. C, G.
*9. S. kempi, n.sp.

S. polita, Vfefler. F, T.

S. elUptica, n.sp.

S. exigiia, Nordenstam.

S. carinata, Lockington.

S. convexa, Cunningham. C, G.

S. gaudichaudii, Aud. et Edw. C.

S. laevis, Richardson. F.

S. geiIachei,Monod. B.

S. meridionalis, Hodgson. Ga, S.

iS. cornuta, Studer.

*io.
*ii.

*I2.
13-

*i4.

*i5-
16.

17-

18.

*i9.

'20.
21.
22.

*23-
24.

25-

26.

*27.
*28.

*2g.

*3o.
*3i.

*32.

33-
34-
35-
36.
37-

S. trilobitoides, Eights. B, C, D, G, P.

S. antarclica, Beddard. C.

S. biomkyana, Suhm. C.

5. neaera, Beddard. C.

5. minuta, Beddard. C.

S. bakeri, Chilton.

S. yongei, Hale.

S. orbiculata, n.sp.

S. nototropis, n.sp.

S. pagenstecheri, Pfeffer. T.

S . platygaster , n.sp.

S. bouvieri, Richardson. F, P.

S. aspera, n.sp.

S. australiemis , Beddard. C.

5. clongata, Beddard. C.

S. longicaudata , Beddard. C.

S. iuberculata, Grube. C.

S. pallida, Beddard. C.

LIST OF STATIONS

In the following list the stations made by the ' Discovery ' and the ' Discovery II ' have
no letters prefixed to the numbers; those of the 'William Scoresby ' have the prefix WS,
and those of the Marine Biological Station MS. A considerable part of the material was
obtained in fine nets of varying mesh attached to the back of the trawl, a method of
great efficiency for the collection of Crustacea and other small invertebrates which live
on the sea-floor.

St. 32. 17. iii. 26. South Georgia, 22-8 miles N 70^° E of Jason Light, i m. tow-net, 0-5 m.
S. septemcarinata, Miers.

St. 39. 25. iii. 26. South Georgia, East Cumberland Bay, from 8 cables S 81° W of Merton Rock
to 1-3 miles N 7 ' E of Macmahon Rock. Trawl and attached nets, grey mud, 179-235 m.

S. pagenstecheri, PfefFer; S. aspera, n.sp.; S. septemcarinata, Miers.

LIST OF STATIONS 257

St. 42. I. iv. 26. South Georgia, off mouth of Cumberland Bay, from 6-3 miles N 89° E to
4 miles N 39° E of Jason Light. Net attached to trawl, mud, 120-204 m.
S. pagetistecheri, PfefFer.

St. 45. 6. iv. 26. South Georgia, 27 miles S 85° E of Jason Light. Net attached to trawl, grey
mud, 238-270 m.

S. septemcaririala, Miers.

St. 51. 4. V. 26. East Falkland Islands, off Eddystone Rock, from 7 miles N 50 E to 7-6 miles
N 63° E of Eddystone Rock. Large dredge, fine sand, 115 m.; nets attached to trawl, fine sand,
105-115 m.

S. schvtliei, Liitken; S. convexa, Cunningham; S. exigua, Nordenstam.

St. 56. 16. V. 26. East Falkland Islands, Sparrow Cove, Port William, li cables N 50' E of
Sparrow Point. Net attached to small trawl, ioJ-16 m.
S. elliptiai, n.sp.

St. 123. 15. xii. 26. South Georgia, off mouth of Cumberland Bay, from 4-1 miles N 54° E of
Larsen Point' to i-2 miles S 62° W of Merton Rock. Net attached to trawl, grey mud, 230-250 m.

S. aspera, n.sp.

St. 140. 23. xii. 26. South Georgia, Stromness Harbour to Larsen Point, from 54° 02' S,36"38'W
to 54= 11' 30" S, 36^ 29' W. Net attached to trawl, green mud and stones, 122-136 m.
S. pagenstecheri, Pfeffer; S. aspera, n.sp.

St. 144. 5. i. 27. South Georgia, oft" mouth of Stromness Harbour, from 54" 04' S, 36° 27' W to
53° 58' S, 36° 26' W. Net attached to trawl, green mud and sand, 155-178 m.
S. pagenstecheri, Pfeffer.

St. 146. 8. i. 27. 53'' 48' S, 35" 37' 30" W. Large dredge, rock, 728 m.
S. pjatygaster, n.sp.

St. 148. 9. i. 27. South Georgia, off Cape Saunders, from 54° 03' S, 36" 39' W to 54' 05' S,
36° 36' 30" W. Net attached to trawl, grey mud and stones, 132-148 m.
S. aspera, n.sp.

St. 149. 10. i. 27. South Georgia, mouth of East Cumberland Bay, from 1-15 miles N 76.^° W
to 2-62 miles S 11° W of Merton Rock. Net attached to trawl, mud, 200-234 m.
S. pagenstecheri , Pfeffer.

St. 157. 20. i. 27. South Georgia, 53" 51' S, 36" 11' 15" W. Large dredge, diatom ooze, stones,
fine sand, 970 m.

S. pagenstecheri, Pfeffer.

St. 160. 7. ii. 27. Near Shag Rocks, 53° 43' 40" S, 40° 57' W. Large dredge, grey mud, stones
and rock, 177 m.

S. aspera, n.sp.

St. 164. 18. ii. 27. South Orkneys, east end of Normanna Strait, near Cape Hansen, Coronation
Island. Small beam trawl, 24-36 m.
5. cornuta, Studer.

St. 170. 23. ii. 27. Clarence Island, off Cape Bowles, 61" 25' 30" S, 53° 46' W. Large dredge,
rock, 342 m.

S. trilobitoides, Eights.

258 DISCOVERY REPORTS

St. 172. 26. ii. 27. South Shetlands, off Deception Island, 62° 59' S, 60° 28' W. Large dredge,
rock, 525 m.

S. Irilobitoides, Eights.

St. 174. 28. ii.-2. iii. 27. South Shetlands, Deception Island, outside entrance, W of Light.
Large fish-trap, 5-10 m.
S. beddardi, Cahnan.

St. 180. II. iii. 27. Palmer Archipelago, 1-7 miles W of N point of Gand Island, Schollaert
Channel. Net attached to trawl, mud and stones, 160-330 m.
S. glacialis, Tattersall.

St. 181. 12. iii. 27. Palmer Archipelago, Schollaert Channel, 64° 20' S, 63" 01' W. Netsattached
to trawl, mud, 160-335 m.

S. glacialis, Tattersall ; S. bouvieri, Richardson.

St. 182. 14. iii. 27. Palmer Archipelago, Schollaert Channel, 64°2i'S, 62°58'W. Nets attached
to trawl, mud, 278-500 m.
5. glacialis, Tattersall.

St. 187. 18. iii. 27. Palmer Archipelago, Neumayr Channel, 64° 48' 30" S, 63° 31' 30" W. Large
dredge, mud, 259 m.

S. glacialis, Tattersall.

St. 190. 24. iii. 27. Palmer Archipelago, Bismarck Strait, 64° 56' S, 65° 35' W.
(i) Large dredge and large rectangular net, rock, stones and mud, 93-130 m.
S. bouvieri, Richardson.

(2) Large dredge, mud and rock, 315 m.
S. bouvieri, Richardson.

St. 195. 30. iii. 27. South Shetlands, Admiralty Bay, King George Island, 62° 07' S, 58° 28' 30" W.
Nets attached to trawl, mud and stones, 391 m.
S. bouvieri, Richardson.

St. 223. 27. iv. 27. Cape Horn, St Francis' Bay, 55° 51' 15" S, 67° 29' 30" W. Large rectangular
net, sand, 63 m.

S. schythei, Liitken.

St. 363. 26. ii. 30. South Sandwich Islands, 2-5 miles S 80 E of SE point of Zavodovski Island.
Large dredge, 329-278 m.
S. cornuta, Studer.

St. 371. 14. iii. 30. South Sandwich Islands, I mile E of Montagu Island. Net attached to trawl
99-161 m.

S. polita, PfefFer.

St. 388. 16. iv. 30. 56° 19J' S, 67° ogf W. Large dredge, 121 m.
S. kempi, n.sp.

St. 456. 18. X. 30. I mile E of Bouvet Island. Large dredge, 40-45 m.
S. septemcarinata, Miers.

St. WS 25. 17. xii. 26. South Georgia, Undine Harbour (North). Small beam trawl, mud, sand,
18-27 Ti-

■S. polita, Pfeffer.

LIST OF STATIONS 259

St. WS 62. 19. i. 27. South Georgia, Wilson Harbour. Small beam trawl, 26-83 m.
S. pagenstecheri, Pfeffer.

St. WS 72. 5. iii. 27. 51" 07' S, 57^ 34' W. Net attached to trawl, sand, shells, 79 m.
S. schythei, Liitken.

St. WS 73. 6. iii. 27. 51" 01' S, 58° 54' W. Nets attached to trawl, fine dark sand, 121-130 m.
S. schythei, Liitken.

St. WS75. ID. iii. 27. 51° 01' 30" S, 60° 31' W. Trawl, 72 m.
S. schythei, Liitken; S. convcxa, Cunningham.

St. WS 76. II. iii. 27. 51° 00' S, 62° 02' 30" W. Trawl, fine dark sand, 207-205 m.
S. schythei, Liitken.

St. WS 78. 13. iii. 27. 5i"oi'S, 68' 04' 30" W, from 51 01' S, 68" 02' W to 51 ' 01' S,
68° 07' W. Net attached to trawl, fine dark sand, 91-95 m.
S. schythei, Liitken.

St. WS79. 13. iii. 27. 51° 01' 30" S, 64° 59' 30" W. Net attached to trawl, fine dark sand, 132 m.
S. schythei, Liitken.

St. WS 80. 14. iii. 27. 50 57' S, 63" 37' 30" W. Trawl and attached nets, fine dark sand,
152-156 m.

S. exigua, Nordenstam ; S, schythei, Liitken.

St. WS83. 24. iii. 27. 14 miles S 64° W of George Island, East Falkland Islands. Net attached
to trawl, fine green sand and shells, 137-129 m.
S. schythei, Liitken.

St. WS 86. 3. iv. 27. 53° 53' 30" S, 60" 34' 30" W. Net attached to trawl, sand, shells and stones,
151-147 m.

S. kern pi, n.sp.

St. WS 90. 7. iv. 27. 13 miles N 83° E of Cape Virgins Light, Argentine Republic. Net attached
to trawl, fine dark sand, 82-81 m.

S. schythei, Liitken; S. convexa, Cunningham.

St. WS 210. 29. V. 28. 50' 17' S, 60' 06' W. Net attached to trawl, green sand, 161 m.

S. schythei, Liitken; S. exigua, Nordenstam.
St. WS 211. 29. V. 28. 50° 17' S, 60° 06' W. Net attached to trawl, green sand, 174 m.

S. schythei, Liitken.

St. WS 212. 30. V. 28. 49° 22' S, 60" 10' W. Nets attached to trawl, green sand, mud and pebbles,
242-249 m.

S. exigua, Nordenstam; S. neaera, Beddard.

St.WS2i3. 30. V. 28. 49° 22' S, 60° 10' W. Net attached to trawl, green sand, mud and pebbles,

249-239 m.

S. neaera, Beddard.

St. WS214. 31.V. 28. 48° 25' S, 60° 40' W. Nets attached to trawl, fine dark sand, 208-219 m.
S. exigua, Nordenstam; S. scliythei, Liitken.

St. WS 215. 31. V. 28. 47° 37' S, 60° 50' W. Net attached to trawl, fine green sand, 219-146 m.
S. schythei, Liitken; S. exigua, Nordenstam.

26o DISCOVERY REPORTS

St. WS216. i.vi. 28. 47" 37' S, 60° 50' W. Net attached to trawl, fine sand, 219-133 m.
S. schythei, Liitken; S. exigua, Nordenstam.

St. WS 219. 3. vi. 28. 47'^ 06' S, 62°i2'W. Net attached to trawl, dark sand, 116-11401.
S. schythei, Liitken; S. exigua, Nordenstam.

St. WS 220. 3. vi. 28. 47' 56' S, 62° 38' W. Net attached to trawl, brown sand, 108-104 m.
S. exigua, Nordenstam; S. couvcxa, Cunningham.

St. WS 221. 4. vi. 28. 48° 23' S, 65^ 10' W. Net attached to trawl, brown sand and mud, pebbles,
large stones and shells, 76-91 m.

S. exigua, Nordenstam; S. gaudichaudii. And. et Edw.

St. WS 222. 8. vi. 28. 48° 23' S, 65" W. Net attached to trawl, coarse brown sand and shells,
100-106 m.

S. exigua, Nordenstam; S. orhiculaia, n.sp.; S. convexa, Cunningham.

St. WS 225. 9. vi.28. 50- 20' S, 62° 30' W. Net attached to trawl, green sand, shells and pebbles,
162-161 m.

5. exigua, Nordenstam; S. schythei, Liitken.

St. WS 226. 10. vi. 28. 49° 20' S, 62° 30' W. Net attached to trawl, green sand, 144-152 m.
5 schythei, Liitken.

St. WS 227. 12. vi. 28. 51 08' S, 56" 50' W. Net attached to trawl, fine green sand, 295 m.
5. exigua, Nordenstam.

St. WS 228. 30. vi. 28. 50° 50' S, 56° 58' W. Nets attached to trawl, shells and coarse white
sand, 229-236 m.

S. exigua, Nordenstam.

St. WS229. i.vii. 28. 50° 35' S, 57° 20' W. Net attached to trawl, fine green sand, 210-271 m.
S. exigua, Nordenstam; S. schythei, Liitken.

St. WS231. 4. vii. 28. 50° 10' S, 58° 42' W. Net attached to trawl, fine green sand, 167-159 m.
S. exigua, Nordenstam.

St. WS 233. 5. vii. 28. 49 ' 25' S, 59 '45' W. Net attached to trawl, fine green sand, 185-175 m.
S. exigua, Nordenstam; S. schythei, Liitken.

St. WS 234. 5. vii. 28. 48° 52' S, 60° 25' W. Net attached to trawl, fine green sand, 195-207 m.
S. exigua, Nordenstam; S. schythei, Liitken.

St. WS 235. 6. vii. 28. 47" 56' S, 61° lo' W. Net attached to trawl, dark green sand, 155-155 m-
S. schythei, Liitken.

St. WS 236. 6. vii. 28. 46- 55' S, 60' 40' W. Net attached to trawl, dark green sand and mud,
273-300 m.

S. exigua, Nordenstam; 5. tieaera, Beddard; S. schythei, Liitken.

St. WS237. 7. vii. 28. 46° 00' S, 60° 05' W. Nets attached to trawl, coarse brown sand and shells,
150-256 m.

S. schythei, Liitken; S. exigua, Nordenstam.

St. WS239. 15. vii. 28. 51° 10' S, 62° 10' W. Netattachedtotrawl, coarse dark sand, 196-192 m.
5. schythei, Liitken.

LIST OF STATIONS 261

St.WS243. 17.vii.28. 5i°o6' S, 64°3o'W. Net attached to trawl, coarse dark sand, 144-141 m.
S. exigua, Nordenstam ; S. cUiptica, n.sp.; S. schythei, Lutken.

St. WS 244. 18. vii. 28. 52"" 00' S, 62" 40' W. Net attached to trawl, fine dark sand and mud,
253-248 m.

S. schythei, Lutken; S. neaera, Beddard; S. kempi, n.sp.

St. WS 245. 18. vii. 28. 52° 36' S, 63° 40' W. Net attached to trawl, dark green sand, madrepore
sand, pebbles and shells, 304-290 m.

S. kempi, n.sp.; S. schythei, Lutken.

St. WS 246. 19. vii. 28. 52' 25' S, 6i'oo'W. Nets attached to trawl, coarse green sand and
pebbles, 267-208 m.

S. kempi, n.sp.; S. exigua, Nordenstam.
St. WS742. 5.ix. 31. 38° 22' S, 73° 41' W. Small beam trawl, 47-35 m.

S. gaudichaudii, Aud. et Edw.
St. WS 752. 19-20. ix. 31. 51° 20' S, 63° 17' W. Rectangular net, 160 m.

S. schythei, Lutken.
St. WS 754. 20. ix. 31. 51" 09' 30" S, 58° 54' W. Rectangular net, 106 m.

S. schythei, Lutken.
St. WS 758. 12. X. 31. 48" 32' S, di' 19' W. Rectangular net, rock, 112 m.

S. schythei, Lutken.
St. WS 763. 16. X. 31. 44' 14' S, 63'' 28' W. Net attached to trawl, mud and sand, 87-82 m.

S. schythei, Lutken.
St. WS765. 17. X. 31. 45°07'S, 60^ 28' 15" W. Trawl, brown and green mud and sand,
113-118 m.

S. schythei, Lutken.
St. WS766. 18-19. X. 31. 45^13'S, 59° 56' 30" W. Net attached to trawl, fine dark grey
sand, 545 m.

S. exigua, Nordenstam.
St. WS 771. 29. X. 31. 42" 41' 45" S, 60" 31' W. Trawl, dark green sand, 90 m.

S. schythei, Lutken.
St. WS772. 31.X. 31. 47°28' S, 6o°5i' W. Nets attached to trawl, grey sand, 309-162 m.

S. schythei, Lutken.
St. WS773. 31.X. 31. 47'28'S, 6o"5i'W. Net attached to trawl, green sand and mud,
291-296 m.

S. neaera, Beddard ; S. exigua, Nordenstam.
St. WS 774. I. xi. 31. 47° 08' S, 62° 02' W. Net attached to trawl, dark green sand and mud,
139-144 m.

S. schythei, Lutken.
St. WS775. 2. xi. 31. 46° 44' 45" S, 63" 33' W. Nets attached to trawl, gravel and fine grey
sand, 115-110 m.

S. schythei, Lutken.

262 DISCOVERY REPORTS

St. WS 776. 3. xi. 31. 46" 18' 15" S, 65° 02' 15" W. Net attached to trawl, green mud and sand,
107-99 m.

S. schythei, Liitken.

St. WS 781. 6. xi. 31. 50° 30' S, 58- 50' W. Net attached to trawl, dark green sand and mud,
148 m.

S. exigiia, Nordenstam; 5. schythei, Lutken.

St. WS 782. 4. xii. 31. 50° 29' 15" S, 58° 23' 45" W. Net attached to trawl, green sand, 141 m.
S. exigua, Nordenstam; S. schythei, Lutken.

St. WS783. 5. xii. 31. 50° 02' 45" S, 6o°i4'W. Net attached to trawl, rock, mud and sand,
159-0 m.

S. schythei, Lutken.

St. WS 786. 7. xii. 31. 49° 07' S, 63° 5s' W. Net attached to trawl, dark sand, i33-ii9m.
S. schythei, Lutken; S. exigua, Nordenstam.

St. WS787. 7. xii. 31. 48°44'S, 65° 24' 30" W. Net attached to trawl, coarse brown sand,
106-110 m.

S. exigua Nordenstam ; S. schythei, Lutken ; S. convexa, Cunningham.

St. WS 791. 14. xii. 31. 45-41' 45" S, 62° 45' W. Net attached to trawl, 96-101 m.
S. schythei, Lutken.

St. WS 796. 19. xii. 31. 47° 49' 37" S, 63° 42' 30" W. Net attached to trawl, coarse brown sand,
106-113 m.

S. nototropis, n.sp.; S. convexa, Cunningham; S. exigua, Nordenstam.

St. WS 797. 19. xii. 31. 47"" 45' 36" S, 64° 20' W. Net attached to trawl, 115-111 m.

5. nototropis, n.sp.; S. convexa, Cunningham; S. schythei, Lutken.
St. WS 801. 22. xii. 31. 48° 26' 15" S, 6i° 28' W. Net attached to trawl, dark sand, 165 m.

S. exigua, Nordenstam.
St. WS 802. 5. i. 32. 50° 45' 45" S, 61° 22' W. Net attached to trawl, 128-132 m.

S. schythei, Lutken; S. exigua, Nordenstam.
St. WS 802. 5. i. 32. 50° 43' 45" S, 61= 26' W. Net attached to trawl, 132-139 m.

S. schythei, Lutken.
St. WS 804. 6. i. 32. 50° 21' 15" S, 62" 53' W. Net attached to trawl, gravel and sand, 143-150 m.

5. exigua, Nordenstam.
St. WS 805. 6. i. 32. 50° 10' 15" S, 63° 29' W. Net attached to trawl, coarse dark sand, 148 m.

S. exigua, Nordenstam.
St. WS 806. 7. i. 32. 50° 03' 30" S, 64° 21' W. Net attached to trawl, 129-122 m.

S. exigua, Nordenstam.
St. WS 808. 8. i. 32. 49° 28' 15" S, 65° 42' W. Net attached to trawl, brown and green sand,
109-107 m.

S. exigua, Nordenstam; S. orbiculata, n.sp.; S. convexa, Cunningham.

St. WS 809. 8. i. 32. 49° 28' 15" S, 66° 29' W. Net attached to trawl, brown sand, 107-104 m.
S.gaudichaudii, Aud. et Edw.; S. orbiculata, n.sp.

LIST OF STATIONS 263

St. WS811. 12.1.32. 51' 24' 30" S, 67° 53' W. Net attached to trawl, 96-98 m.

S. convexa, Cunningham.
St. WS 813. 13. i. 32. 51° 3S' 15" S, 67° 16' 15" W. Net attached to trawl, dark sand, 106 m.

S. nolotropis, n.sp.; S. exigua, Nordenstam ; S. convexa, Cunningham; S. orbiculata, n.sp.
St. WS 814. 13. i. 32. 51° 45' 15" S, 66° 40' W. Net attached to trawl, 111-118 m.

S. exigua, Nordenstam; S. schythei, Lutken; S. convexa, Cunningham.
St. WS 815. 13. i. 32. 51 ' 51' 45" S, 65° 44' W. Net attached to trawl, 132-162 m.

S. nototropis, n.sp.; S. exigua, Nordenstam; S. convexa, Cunningham; S. orbiculata, n.sp.
5. schythei, Lutken.
St. WS 816. 14. i. 32. 52' 09' 45" S, 64° 56' W. Net attached to trawl, sand, 150 m.

S. nototropis, n.sp.; S. exigua, Nordenstam; S. schythei, Lutken.
St. WS818. 17.1.32. 52° 31' 15" S, 63' 25' W. Net attached to trawl, dark sand, 272-278 m.

S. schythei, Lutken; 5. ketnpi, n.sp.; S. exigua, Nordenstam.
St. WS 820. 18. 1. 32. 52' 53' 15" S, 61 ' 51' W. Net attached to trawl, fine dark sand and mud,
351-36701.

S. neaera, Beddard.
St. WS 821. 18. 1. 32. 52° 55' 45" S, 60° 55' W. Net attached to trawl, green and grey fine sand
and mud, 461-468 m.

S. neaera, Beddard; S. exigua, Nordenstam.
St. WS 824. 19.1.32. 52^29' 15" S, 58° 27' 15" W. Net attached to trawl, green sand and shells,
146-137 m.

S. schythei, Lutken.
St. WS 825. 28-29. 1. 32. 50° 50' S, 57° 15' 15" W. Net attached to trawl, green sand, mud and
shells, 135-144 m.

5. schythei, Lutken; S. exigua, Nordenstam.
St. WS 837. 3. 11. 32. 52° 49' 15" S, 66° 28' W. Net attached to trawl, coarse, dark green sand and
pebbles, 98-102 m.

S. nototropis, n.sp.; S. exigua, Nordenstam.
St. WS839. 5.11.32. 53° 30' 15" S, 63°29'W. Net attached to trawl, fine sand and mud,
403-434 m.

S. neaera, Beddard.
St. WS 856. 23. 111. 32. 46° 45' S, 64° 11' W. Small beam trawl, 104 m.

S.gaudichaudii, Aud. et Edw.
St. WS 864. 28. ill. 32. 49° 33' 30" S, 64° 16' W. Net attached to trawl, 128-126 m.

5. exigua, Nordenstam.
St. WS 866. 29. 111. 32. 50° 37' 45" S, 64° 15' W. Net attached to trawl, 137-144 m.

5. exigua, Nordenstam; S. schythei, Lutken.
St. MS 10. 14. ii. 25. East Cumberland Bay, \ mile SE of Hope Point to { mile S of Government
Flagstaff. Small beam trawl, 26 m.
S. pagenstecheri, Pfeffer.

264 DISCOVERY REPORTS

St. MS 65. 28. ii. 25. East Cumberland Bay, i-6 miles SE of Hobart Rock to i cable N of
Dartmouth Point. Net attached to small beam trawl, 18 m.

S. polita, Pfeffer.

St. MS 66. 28. ii. 25. East Cumberland Bay, 2} miles SE of King Edward Point Light to i\
cables W x N of Macmahon Rock. Small beam trawl, 18 m.
5. polita, Pfeffer.

St MS 67. 28. ii. 25. East Cumberland Bay, 3 cables NE of Hobart Rock to I cable W of Hope
Point. Small beam trawl, 38 m.

S. polita, Pfeffer; S. pagenstecheri, Pfeffer.

St. MS 71. 9. iii. 26. East Cumberland Bay, 9J cables E x S to 1-2 miles E x S of Sappho Point.
Small beam trawl, 110-60 m.

S. septemcarinata, Miers.

GEOGRAPHICAL DISTRIBUTION

The geographical range of the genus Serolis, with the exception of S. carinata,
Lockington, which is recorded as far north as San Diego, CaHfornia, is entirely restricted
to the southern hemisphere.

In the following account the distribution of the species has been considered with
reference to the Antarctic Convergence. At this line, which surrounds the Antarctic
Continent, there is an abrupt change in salinity and temperature, and it has been found
that in some groups of animals (e.g. fishes) there are two distinct faunas, one on either
side of it. The line lies roughly midway between Cape Horn and Graham Land m the
west, and passes north of the Shag Rocks and South Georgia, then eastward, crossing
latitude 50° S at 20° E, passing a Httle south of Marion Island and the Crozet Islands
and then through the middle of Kerguelen. The temperature along this line ranges
from 0-50° to 3-0° in the winter to 3-50° to 5-50° in the summer. The majority of species
of Serolis are confined to shallow waters ; the deep-sea forms are comparatively few in
number and have a inuch wider vertical as well as horizontal distribution.

According to Beddard (18846, p. 82) "the shallow-water forms never pass the 300-
fathom limit, nor are any of the deep-sea species known to inhabit shallow water".
The study of accumulated data proves that this statement is no longer accurate, for at
least two shallow-water forms are now known to inhabit depths greater than 300
fathoms. S. pagenstecheri, Pfefll'er, is found in depths ranging from 15 to 970 m.
(approximately 530 fathoms), and the new species, S . platygaster , which froin its general
characters is more closely related to the shallow-water species, was collected at a depth
of 728 m. (approximately 393 fathoms). Further, S. 7ieaera, Beddard, one of the deep-
sea species, previously recorded from depths ranging from 600 to 2040 fathoms, is now
known to exist in comparatively shallow waters: specimens in this collection occur at
depths of 239-300 m. (approximately 130-164 fathoms).

The shallow-water species fall into four groups :

(i) Those which are found outside the Antarctic Convergence, off the coasts of the

GEOGRAPHICAL DISTRIBUTION 265

southern part of South America, and on the shores of the Falkland Islands, as well as
in the comparatively shallow waters between the two.

To this group belong the species S. paradoxa, Fabricius, S. schythei, Liitken,
S. convexa, Cunningham, S.gaudichaudii, Aud. et Edw., S. exigiia, Nordenstam, the new
species S. kempi, S. orbiciilata, S. nototropis and S. elliptica, and the two doubtful
species S. plana, Dana, and S. serrei, Lucas. Of these S. gatidichaudii extends farther
north than the rest, the original specimen having been collected near Valparaiso.

Miers in his list of New Zealand Crustacea includes S. paradoxa, apparently (as
Beddard, 18846, p. 80, points out) "on the authority of a specimen at the British
Museum", and the same collection of Crustacea contains a single example of a species
which Beddard identifies as S. schythei, which is also labelled " New Zealand ". Beddard
continues : " I believe the locaHty is not authenticated beyond a doubt ". With the excep-
tion of these two specimens, the species of this group so far collected are restricted to
the above-mentioned localities.

(2) Those which occur within the Antarctic Convergence, on the shores of the South
Sandwich Islands, South Georgia, the South Shetlands and the Palmer Archipelago, as
well as off Coats Land (long. 20° W) and Oates Land (long. 155° E).

To this group belong S. beddardi, Caiman, S. glacialis, Tattersall, S. gerlachei,
Monod, S. bouvieri, Richardson, S. polita, Pfeffer, S. pagenstecheri, Pfefi^er, S. polaris,
Richardson, S. laevis, Richardson, S. cormita, Studer, S. trilobitoides, Eights, S. septem-
carinata, Miers, and the two new species, S. aspera and S. platygaster.

In Beddard's description of S. cornuta (18846, pp. 52 and 53: iS. trilobitoides of this
paper), an error occurs in the list of stations at which the species was collected. The
longitude in each case is given as west of Greenwich, with the result that the given
stations lie within the Continent of South America. The actual stations are situated
around the shores of Kerguelen in corresponding longitudes east of Greenwich. With
the exception of a record of doubtful value (that of a specimen of S. trilobitoides de-
scribed by Eights from the stomach of a fish collected in Patagonian waters), the species
found within the Antarctic Convergence are all different from those present in the
South American-Falkland Islands area.

(3) Those species which are found around the shores of Kerguelen Island, Crozet
Island, Marion Island, and Prince Edward Island. These localities with the exception
of Kerguelen, lie north of the convergence, considerably farther east than either of the
localities referred to in groups (i) and (2) above. Kerguelen Island actually lies on the
convergence, its southern shores being within it.

With the exception of S. latifrons. White, which is also recorded from Auckland
Island, New Zealand, the species belonging to this group — namely, S. cornuta, Studer,
S. trilobitoides, Eights, and S. septemcarinata, Miers — are also members of group (2) above.

(4) This group contains the species which are found off the shores of South and East
Australia. These are: S. australiensis, Beddard, S. bakeri, Chilton, S. mimita, Beddard,
S. elongata, Beddard, S. tuber culata, Grube, S. lo?igicaudata, Beddard, S. pallida,
Beddard, and S. yongei, Hale, from the Barrier Reef.

266 DISCOVERY REPORTS

The marked difference in the shallow-water faunas of the two areas represented in
groups (i) and (2) is not surprising considering that they are separated from each other
not only by the sudden hydrographical change represented by the Antarctic Con-
vergence but also by Drake Strait. This channel of water, which reaches a depth of over
2000 fathoms, may prove to be a barrier of even greater importance than the con-
vergence to the bottom-living forms. It is interesting to note that this difference in the
shallow-water faunas of these two areas has been noted by other investigators. Barnard,
for example, states that the species of Amphipods in the two areas are entirely different.

An examination of the geological history of the area shows that, during Eocene times,
the South Polar Continent, Antarctica, was united on the one hand with South America
through Graham Land, and on the other with Australia. It is probable that the genus
Serolis was already represented in the shallow- water fauna of the northern coast of this
southern land-mass, and that its centre of distribution lay in the waters off the shores of
Graham Land. This seat of origin would account for the northward spread of the genus
and the almost entire absence of species from the shores of the west coast of South

America.

During the Older Quaternary period, Australia broke away from the main southern
continent, and this resulted in the isolation of certain species of Serolis, the further
modification of which may be seen in the species of Australia to-day (group (4) above).
It is significant that the existing species are found off the coast of South-east Australia
in a ret^ion which was the last to lose its connection with the southern continent.

In the course of time, the formation and gradual deepening of the channel between
South America and Graham Land would be sufficient to account for the separation of,
and variation in, the species found in the shallow waters represented by the areas in
groups (i) and (2) above.

It is possible that the presence of the same species in the two localities represented by
groups (2) and (3), may be accounted for by the fact that conditions of life are much the
same in the two areas. As already mentioned, the southern shores of Kerguelen lie
within the convergence, and Marion Island, and Crozet Island, are respectively 120
and 160 miles north of it. According to information supplied to me by Mr G. E. R.
Deacon of the Discovery scientific staff, the two latter islands are so close to the con-
vergence that upwelling, which must take place somewhere near them, will bring to the
surface Antarctic water which has not been below a depth of 200-300 m. St. 42, in
47° 53' S, 61° 25' E, and St. 43 in 47° 53' S, 66° 26' E, of the Gauss Expedition, are
both north of the convergence. The former, 200 miles north, has Antarctic water at
a depth of about 400-600 m.; the latter, 100 miles north, at a depth of 200-300 m.
From these data it seems quite probable that the shallow water off the shores of these
islands is partly of Antarctic origin, in which case the presence of similar species in
the groups (2) and (3) above is made possible.

The distribution of the various species mentioned by Nordenstam (1933), with the
exceptions of those occurring at stations 33 and 34 b, agrees with the grouping of species
as shown above.

GEOGRAPHICAL DISTRIBUTION 267

St. 33 of Nordenstam's paper is given as South Georgia, and at this station the three
species S. paradoxa, S. schythei and S. convexa were collected : this is the only reference
to the presence of these species within the Antarctic Convergence either in Nordenstam's
or any other paper. Their distribution is outside the convergence, off the coasts of the
southern part of South America and on the shores of the Falkland Islands as well as in
the comparatively shallow waters between the two.

The locality of St. 34 b is given as "Atlantic Ocean, North of Falkland Islands and
East of Patagonia, lat. 44° 19' S., long. 57° 34' W." and at this station S. polita and
S. septemcarinata were collected ; two species which occur otherwise either zvithin the
convergence or, in the case of S. septemcarinata, around the shores of Kerguelen Island
as well.

It is possible that the labels referring to the material collected at these two stations
have been interchanged, since the three species collected at St. 33 are characteristic of
the locality represented by St. 34 b, and the two species collected at the latter station
are characteristic of the locality represented by St. 33. This view is supported by the
fact that at neither station were species characteristic of that locality collected, and,
further, by the facts that both these areas have been extensively explored and that, in
the numerous records, the distribution of the species has been so remarkably consistent.

CLASSIFICATION

ISOPODA, Latreille

FLABELLIFERA, Sars

Family SEROLIDAE

The family Serolidae, together with the families Anthuridae, Cymothoidae and
Sphaeromidae, is included in the sub-order Flabellifera, Sars ; it is perhaps more closely
allied to the Sphaeromidae than to either of the other families. There are, however,
fundamental differences between the members of the two families, for in the Serolidae
the first and second thoracic somites are fused with the head, and the tergum of the
eighth thoracic somite is usually absent and when present is never complete. In the
Sphaeromidae only the first thoracic somite is fused with the head, and the remaining
seven somites are all free and complete.

The family Serolidae contains the single genus Serolis, Leach, consequently the
following definition will serve equally well for the family.

Genus Serolis, Leach

Onisais, Fabricius, 1787, p. 240.
Aselliis, Olivier, p. 252.
Cymothoa, Fabricius, 1793, p. 503.
Serolis, Leach, 1825, p. 340.
Brongiartia, Eights, 1833, p. 53.

268 DISCOVERY REPORTS

An historical account of this genus is inckided in Beddard's report (18846, pp. 2-4),
and for this reason I shall not refer to its earlier history.

The characters of the genus may be defined in the following terms :

Body depressed, flattened and usually broad, with the first two thoracic somites fused
with the head, the sides of which are fused with the forward lateral extensions of the
second somite ; the tergum and coxal plates of the last (eighth) thoracic somite never
complete, usually wanting ; the tergum of the seventh thoracic somite may also be in-
complete. Coxal plates of the first three free thoracic (3rd-5th) somites always separated
from them by sutures ; sometimes, in addition, those of the fourth, or those of all the
remaining somites are also separated from them by sutures. First abdominal segment
without pleural plates, sometimes partially fused with the tergum of the seventh thoracic
somite; second and third abdominal segments free, with either short or long pleural
plates; last two fused with telson to form a large terminal segment. Mouth-parts
normal ; antennule with four and antenna with five peduncular joints ; each with a
multi-articulate flagellum. Second pair of thoracic appendages of both sexes and third
pair of adult male modified into a prehensile organ, the dactylus folding back upon the
greatly dilated propodus ; last thoracic appendage usually smaller than the others, and
sometimes modified in the adult male. First three pairs of pleopods natatory, consisting
of protopodite, exopod and endopod, the two latter fringed with long plumose setae;
endopod of male prolonged into a long penial filament; fourth and fifth pleopods
branchial. Uropods lateral in position, lamellar and usually biramous.

Caiman (1920, p. 299), in his paper on the new species S. beddardi, suggested that it
might form a distinct genus with the allied species S. laiifrom, Miers, and further that
a regrouping of the remaining species might be advisable. In his suggested classification
which is based on the form of the uropods and of the terga of the posterior thoracic
somites, he divides the species into three groups :

(i) A group containing S. latifrons and S. beddardi, in which the endopod of the
uropod is absent, and the tergum of the last (eighth) thoracic somite persists as a pair of
minute lateral sclerites each with a coxal plate separated by a suture. In both the
remaining groups the uropod bears an exopod and an endopod.

(ii) A group containing the six AustraUan species, S. tuberctdata, Grube, S. aiistra-
liensis, Beddard, S. longicaudata, Beddard, S. elongata, Beddard, S. mimita, Beddard,
and S. pallida, Beddard, which Beddard (1884Z), pp. 66 and 81) states form "a well-
marked subdivision of the genus", together with S. bakeri, Chilton.

(iii) A group, represented by S. paradoxa, Fabricius, containing all the remaining
species.

The following points result from a consideration of these groups as they stand :

(i) That the absence of the endopod of the uropod is a character which is no longer
peculiar to the members of group (i). It is absent from the uropod of S. platygaster,
n.sp., a species which is in no other way closely related to members of this group.

(2) That if this character (absence of endopod) is disregarded the only character
separating members of this group from those of group (iii) is the presence of small

CLASSIFICATION 269

lateral portions of the tergum of the last thoracic somite, with its corresponding coxal
plates.

(3) That group (iii), containing the remaining species, includes S. pogenstecheri,
Pfeffer, in which the tergum of the seventh thoracic somite is fused with that of the first
abdominal segment for a short distance on either side of the mid-dorsal line. Conse-
quently members of the "Australian group" do not bear characters which distinguish
them from all the remaining species.

(4) That group (ii), containing the Australian species, is itself by no means a uniform
group and should be divided into three:

[a) A group to contain those species in which the seventh thoracic somite is fused
with the first abdominal segment for a short distance on either side of the mid-dorsal
line.

{b) A group to contain those species in which the seventh thoracic somite is not
only fused with the first abdominal segment but also with the tergum of the sixth
thoracic somite for a short distance on either side of the middle line.

(c) A group to contain the species S. pallida, Beddard, and S. tuber ailata, Grube,
in which the middle portion of the seventh thoracic somite is absent, and the first
abdominal segment comes in contact with the tergum of the sixth thoracic somite.
The coxal plates of the seventh are well developed and fused with the small lateral
portions of the tergum of that somite.

In group {a) would be included S. minuta, Beddard, S. bakeri, Chilton, S. pagen-
stecheri, Pfeffer, S.yongei, Hale, S. orbiciilata, n.sp., and S. nototropis, n.sp.; and in
group {b) S. australiensis, Beddard, S. elongata, Beddard, S. longicaudata, Beddard,
S. bouvieri, Richardson, S. platygaster, n.sp., and S. aspera, n.sp.

(5) That members of group (ii) are not restricted to Australian waters, for the group
now includes, in addition to the Australian forms, S. pogenstecheri, Pfeffer, S. bouvieri,
Richardson, and the new species S. orbicidata, S. nototropis, S. platygaster and S. aspera.

(6) That any new regrouping of the species would have to be based on the form of
the more posterior thoracic somites, and would result in the formation of five genera,
which would include the species represented in group (i), those in group (ii) subdivided
as shown above into {a), (b) and (c), and those in group (iii) excluding S . pagenstecheri
and S. bouvieri.

(7) That S. beddardi and S. latifrons, in which the lateral parts of the tergum and
coxal plates of the last thoracic somite are present, and S. pallida and S. tuberctdata, in
which the central portion of the tergum of the seventh thoracic somite has disappeared,
represent the opposite ends of a series in which the remaining species are graded.

After careful consideration, I have come to the conclusion that since the species
represent such a compact genus, it is undesirable to subdivide it by the formation of
four additional genera.

Since writing the above, Nordenstam's paper (1933) has been published; in this the
author divides the genus Serolis into four sub-genera, namely, Spinoserolis, Serolis,
Homoserolis and Heteroserolis.

270 DISCOVERY REPORTS

The subgenus Spinoserolis, which corresponds to group (i) above, contains the two
species S. latifrons and S. beddardi.

The subgenus Serolis, containing the majority of species, and corresponding to
group (iii) above, is further sub-divided into five sections (called groups in Nordenstam's
paper). If these sections are compared with the key (p. 278) of the present paper, it will
be seen that they correspond very closely to the main subdivisions shown.

Section I contains the single species S. gracilis and corresponds with B. I. A. of the
key.

Section II, containing the species S. paradoxa, S. schythei and S. polaris, agrees with
the division B. I. B.

Section IV comprises the species S. gaudichaiidii, S. co/ivexa and S. plana, and corre-
sponds to the section B. I. G. II. a. 2, to which group also belongs the species S. laevis.

Section V includes the single species S. carinata. The formation of this group depends
on the statement that "the dorsal sutures of all the coxal plates are lacking on all the
pereion segments". This statement, however, is inaccurate, for as in the majority of
Serolids the terga of the first three free thoracic somites are separated from their
respective coxal plates by sutures. Consequently S. carinata should be included in
Section III.

Section III contains the remaining species of the subgenus Serolis: S. irilobitoides,
S. septemcarinata, S. antarctica, S. neaera, S. bronileyana, S. polita, S. meridionalis,
S. glacialis, S. gerlachei and S. exigua. In the key this section is represented by the
subdivision B. I. C, excluding the subdivision B. I. C. II. a. 2; and includes, in addition
to the species mentioned by Nordenstam, S. cortiuta and the new species ^S'. kempi and
S. elliptica, as well as the transferred species S. carinata.

Nordenstam places the remaining species in his two subgenera Homoserolis and
Heteroserolis. The former is characterized by the fact that "the tergum of the sixth
(seventh actual) thoracic somite is coalesced with the first abdominal segment so that the
suture-line between this segment and the abdomen has been efi^aced in the middle ". In
the latter the central portion of the tergum of the sixth (seventh actual) is entirely
missing.

The definitions of these two sub-genera correspond with those of my groups ii a and
ii c respectively, but Nordenstam has no subgenus corresponding to group ii b above,
in which the terga of the sixth and seventh thoracic somites are fused with each other
and with that of the first abdominal segment so that the suture-lines between these
segments have been effaced in the middle.

The omission of this subdivision has resulted in a certain amount of confusion in
Nordenstam's classification, as he has placed one member of this group, S. bouvieri, in
his subgenus Homoserolis, and three others, S. australiensis, S. longicaiidata and
S. elongata, in the subgenus Heteroserolis. Thus, Homoserolis and Heteroserolis contain
certain species which have characters in common, and which do not agree with the
diagnostic characters of either subgenus. Since Nordenstam has based his classification
on the degree of the reduction of, and coalescence between, the last three thoracic

CLASSIFICATION

271

somites, it seems strange that he should have omitted this very definite type of modifica-
tion, more especially since he had in his possession material representing three of the
species concerned, namely S. bouvieri, S. australieusis and S. longicaiidata.

It is obvious from the above, that Nordenstam's classification, as it stands, cannot be
accepted. If the sub-division of the genus Serolis along the lines indicated above is to
be established, then it is essential that a fifth subgenus should be formed to contain
S. bouvieri, S. elongata, S. australieusis, S. longicaiidata and the two new species
S. platygaster and S. aspera. I maintain, however, that the genus is best left intact, and
for this reason I do not suggest a name for this extra subdivision.

GENERAL MORPHOLOGY

The following notes, dealing with certain features in the anatomy of the Serolids, are
included in the introductory part of this paper as they apply to the group as a whole and
have not been described in detail in earlier papers. They include the structure of the
maxillula, the maxilla and maxilliped, the interpretation of parts being based on the
views expressed by Hansen (1925), as well as certain points of more general interest.

According to Hansen (1925) the maxillula in the Isopods consists only of a sympod
made up of three segments : the endopod and exopod are wanting.

-I o. i^v

Fig. I . The maxillula and maxilla of species of Serolis.

a, maxillula of S. cornuta, Studer: x 30. b, maxillula of S. orhiculata, n.sp. : x 140. c, maxilla of S. cornuta,
Studer: : 30. d, maxilla of 5. nototropis, n.sp.: x 104. e, maxilla of S. orbicidata, n.sp.: x 140.

272 DISCOVERY REPORTS

In the majority of forms, as in the members of this family (Figs, i a, b), each maxillula
consists of two endites directed distally and corresponding to the endites of segments
I and 3 of the primitive axis ; the inner endite D is usually about two-thirds the length
of the outer one, and is in the form of a curved rod, which may enlarge towards its
distal end and bears one or two very short setae ; it springs from the centre of a basal
plate (i) which is the first segment of the primitive axis. At the outer side of the basal
plate is a smaller though broader plate (2), corresponding to the second segment of the
primitive axis, and from its outer distal extremity the second lobe of the maxillula arises,
consisting of the fused third segment and its endite (L^). This endite, which is stout and
slightly curved inwards, broadens distally and has its obliquely truncate extremity
armed with a double row of short stout spines.

In all the species of Serolis except S. orbiculata, n.sp., and S. nototropis, n.sp., the
form of the maxilla agrees with that shown in Fig. i c. The three segments of the sympod
are again represented: the first is in the form of a plate (i) from near the outer margin
of which the second segment (2) arises ; this segment is in the form of a stout rod, the
upper third of which is expanded on its inner side to form a plate which appears to act
as a place of attachment for the muscles which hold its corresponding endite (L-) in
position. The latter is a broad lobe which tapers towards its proximal end where it
articulates with the upper half of the inner margin of its segment : its broad somewhat
rounded distal extremity usually bears about twenty pectinate setae. Arising from
a cup-like depression at the outer distal angle of the second segment of the sympod
is the third segment (3) of the primitive axis ; it lies in close proximity to the lower
part of the second endite {L") and at its inner distal oblique extremity articulates with
its corresponding endite. This lobe is cleft into two secondary lobes (L^ and L*), each
of which terminates distally in a truncate extremity which usually bears two strong setae
of the same type as those found on the second endite (L^).

In the maxillae of S. orbiculata (Fig. i e) and S. nototropis (Fig. i d) the three seg-
ments of the primitive axis are again discernible, but the boundary of the second one is
difficult to follow. In some specimens of S. orbiculata the boundary of the plate-like
extension seemed to lie in the position indicated by the dotted line in the figure ; the
third segment (3), except for its proximal end, appears in S. orbiculata to be directly
continuous with the endite (L^) of the second segment as well as with its own endite (L^) ;
in S. nototropis the endite {D) of the third segment is separated by a suture.

In both species the endite {U) is small and undivided (cf. condition in other species),
and that of S. orbiculata bears two and .S. nototropis four long, pectinate setae on its
truncate distal extremity. This type of maxilla may represent either a primitive or a
secondary condition; in support of the former is the fact that according to Caiman
(1909, p. 198) Hansen regards the two endites L^ and L* of segment 3 as having been
derived from the division of a single endite. In this case the condition seen in
S. orbiculata, where the endite (L^) is directly continuous with its segment, would be more
primitive than that in S. nototropis, where the endite (D) is cut oif from its segment,
and both would be more primitive than the type characteristic of the remaining species.

GENERAL MORPHOLOGY

273

where this endite is divided into two endites [U and L*). Since the condition in
S. nototropis represents a stage prior to the division of the endite into two, the presence of
four pectinate setae on its distal end may be significant, as in the majority of species this
number is represented, there being two on each of the endites L^ and L*. A further point
of some significance is that in both species the maxiUiped is of the most primitive type.
On the other hand, the fact that, as far as can be made out, the endite (L-) of the
second segment appears to be directly continuous with the third segment of the axis is

Fig. 2. The maxilliped, etc., of species of Serolis.
a, maxilliped of S. exigua, Nordenstam : x 87. 6, maxilliped of S. schythei, Liitken, S- x 20. c, maxilliped

of S. schythei, Liitken, $: x 25. d, maxilliped of S. glacialis, Tattersall,

65. e, maxilliped of

S. glacialis, Tattersall, $: x 65. /, ventral view of left second thoracic appendage and its associated brood
lamella in a breeding $ of S. discoverii, n.sp.: x 35.

more difficult to explain, and Dr Caiman, who has very kindly examined the prepara-
tions of the maxillae of these two species, is inclined to believe that the condition seen
is a secondary rather than a primitive one.

The sympod of the maxilliped consists of a short coxa (C) and a large basis {B) pro-
duced distally into a large endite {B') ; the endopod or palp consists of three, sometimes
four, more or less lamellar segments.

According to Hansen's description of the Isopod type of maxilliped, there is no exo-
pod ; but a plate-shaped epipod {E) is present, the proximal part {E) of which is marked
off by a transverse suture from the distal major portion (£") ; such a condition is figured
by him for Glyptonotus sibiriciis.

274 DISCOVERY REPORTS

The parts of the maxilliped of certain species of Serolis, as for example that of
S. glacialis, Tattersall (Fig. 2 d, e), might be interpreted in this way, but in some species,
for example S. schythei, Liitken (Fig. 2 b, c), the larger distal portion of the epipod,
instead of being freely articulated, is directly continuous with the proximal part of
the basis, whilst in S. iiototropis, n.sp. (Fig. 17 b), S. orbicidota, n.sp. (Fig. 15 «),
S. coritiota, Lockington, and S. exigua, Nordenstam (Fig. 2 a), this fusion is continued
beyond the articulation of the base of the endopod, so that the latter appears to spring
from the middle of a plate-like structure, which, according to Hansen's interpretation,
would represent the basis with its endite fused with the distal portion of the epipod. In
the two former species, S. nolotropis and S. orbiculata, the fusion of the endite {B') with
the epipod (£") is not as complete as in the two latter species.

The question then arises whether this condition is a secondary one resuhing from the
fusion of the basis (5) with its endite {B') and the distal portion of the epipodite (£"), or
whether it might not better be interpreted as primitive, in which case the "distal portion
of the epipodite" {£') must be regarded as a lamellar outgrowth of the basis. If this
latter view is correct the condition met with in S. carinata, Lockington (Fig. 11 a), and
S. exigua, Nordenstam (Fig. 2«), is the most primitive; the next stage is represented by
S. orbiculata, n.sp., and S. fiotoiropis, n.sp. (Figs. 15 fl, 17 b), where the separation of the
endite {B') from the lamella (£") has begun ; and this is followed by the stage seen in
S. schythei, Liitken (Fig. 2 b), where the separation of the lamella (£") reaches as far as
the articulation of the endopod, but remains fused with the basis (B) behind this point.
In the final stage, seen in S. glacialis, Tattersall (Figs. 2 d, e), the lamella {£') is com-
pletely separated both from the basis (B) and its endite {B').

In connection with this point I have examined the maxillipeds of a number of the
more typical members of this order. In Cirolana borealis, Lilljeborg, the maxilliped
consists of a sympod made up of a small coxa and a larger basis produced distally into a
small endite; the palp consists of five segments, and a small epipod is also present,
attached to the coxopodite. In the female of this species, during the breeding phase,
a large lamellar outgrowth of the basis, exactly comparable in position with that of the
so-called " distal portion of the epipod " is developed ; lamella-like structures also occur
on the coxa and on the epipod, along the inner and lower margin of the former and on
the outer distal angle of the latter. Similar expansions also occur in the breeding females
of Eurydice elegaiitula, Hansen, Corallana antillensis, Hansen and Corallana tricor/iis,
Hansen, whilst in the female of Ceratothoa banksii. Leach, the lamella of the basis is
fused with the basis, producing a condition directly comparable with that of the maxil-
liped of S. schythei, Liitken.

From these examples it seems reasonable to suppose that the "distal portion of the
epipodite" of Serolis is morphologically the same as the "lamella of the basipodite"
which is developed in the breeding females of the above-mentioned species, and that
the four stages represented by S. exigua, S. orbiculata, S. schythei and S. glacialis,
illustrate stages in the evolution of the appendage, in which the cumbersome structure
represented by the basis, its endite and lamellar outgrowth, is gradually replaced by a

GENERAL MORPHOLOGY 275

more flexible structure in which first the distal portion of the lamella becomes free from
the endite {B') and later from the basis (B) itself.

For these reasons, in this paper, I shall refer to the " distal portion of the epipodite"
(Hansen) as the lamella of the basipodite, and the "proximal portion of the epipodite"
as the epipodite proper.

Beddard (18846, p. 56, pi. iv, fig. 8) and Hodgson (1910, p. 28, pi. iv, fig. 5) describe
and figure the maxillipeds of S. bromleyaim, Suhm, and S. trilobitoides, Eights, re-
spectively : in both the figures the coxopodite is shown fused with the epipodite. I have
examined specimens of these species and find that a suture between the coxopodite and
epipodite is clearly visible in each, so that the maxillipeds of all the species of Serolis fall
into one or other of the four groups mentioned above.

The endopodite of the maxilliped of the majority of species of Serolis consists of three
joints; in a few species, however, near the extremity of the third joint, at its outer angle,
is a fourth small one, the distal extremity of which does not extend beyond that of the
third: its truncate extremity bears several long setae. Such a joint is well developed in
S. platygaster, n.sp. (Fig. 19 o). This small joint probably represents the fourth joint
of the primitive five-jointed endopodite.

A further point of interest in connection with this limb is the modification in the form
of the coxopodite of the female during the breeding phase. In the breeding female there
is a very definite increase in the size of the coxopodite as compared with that of the male
(Figs. 2 b-e), and this increase is brought about by a thin lamella-like extension of
the posterior and inner margins of the joint (C), the distal angle of which is also often
produced to form a small lobe. Delicate setae are developed along its inner margins.

This modification in the form of the coxopodite of Serolis has not been previously
noted, and Caiman (1909, p. 199), in dealing with the development of coxal lobes on
these joints in ovigerous females of members of the order Isopoda, states that in Serolis
no such lobes are developed. Nevertheless, the lamella-like extensions (C) of the
coxopodites in the breeding females of Serolids must be regarded as coxal lobes. They
are similar in form to the corresponding structures seen, for example, in the ovigerous
females of Phreatoicus australis, Chihon, and these are homologous with, though less
well developed than the corresponding lobes of the allied species Ph. latipes (Chilton)
(see Sheppard, 1927, p. 87, fig. i (4) and p. 90, fig. 2 (i)).

Beddard (18846, pp. 15-17) gives an account of a number of secondary sexual
characters in which the male differs from the female. To these may be added the
following :

(i) In certain species, e.g. S. glacialis, Tattersall, S. kcinpi, n.sp., and S. aspera,
n.sp., the number of joints in the flagellum of the antennule is considerably greater in
the male than in the female.

(2) In S. convexa, Cunningham, S. gaudichaiidii, Aud. et Edw., and S. laevis,
Richardson, the form of the modified spines on the propodus of the second thoracic
appendage shows a sexual variation (for details see the descriptions of the species).

(3) Long delicate setae are present on the outer margin of the propodus of the second

276 DISCOVERY REPORTS

thoracic appendage of the aduh male of S. exigua, Nordenstam. These are not found
in the corresponding position in the female.

The so-called "frontal sense organ" first described by Grube (1875, p. 225) in his
account of S. paradoxa, Fabricius, appears as an oval semi-transparent area on either
side of the head on the lateral part of the cephalosome. Beddard {I'&^^b, p. 17) states
that such a structure is present in many species of Serolis, and in a footnote further
remarks that it generally has the form of a deep and narrow groove surrounded by a
specially thickened rim, and that in S. schythei, Lutken, and S. cornuta, Studer, there
is a pore on the under-surface of the epimeron exactly beneath it.

There is no doubt that such an area does appear in many species of Serolis, but I have
failed to find the " thickened rim " surrounding a deep and narrow groove, or the " pore "
on the under-surface. After the careful examination of a number of specimens of
several species I have come to the conclusion that this area simply represents a certain
thinness of the ventral chitinous coat brought about by the rubbing movement of the
distal end of the propodus of the second thoracic appendage, which Ues in this position
when the appendage is not extended, and that it has no connection with any sensory

fvmction.

The oostegites appear in one of two forms: in the non-breeding female as chitinous,
parallel-sided plates which reach the middle of the segment and are found on the
second to the fifth thoracic somites ; and in the breeding female as four pairs of large
lamellae overlapping to form a complete marsupium. This overlapping alternates in
successive segments: in the first and third somites the left lamella overlaps the right,
and in the second and fourth the right overlaps the left. In no case do my observations
agree with those of Studer (1879), who states that the lamellae of the right generally
cover those of the left side.

It is interesting to note that in this family the form of the first pair of oostegites (Fig. 2 /)
is similarly modified to that of Asellus aquatiais and of the members of the family
Phreatoicidae. Each consists of a main posterior portion {A) which form the anterior
boundary of the marsupium, and a smaller anterior portion {B) which envelops the
base of the maxilliped of its own side. When viewed from below these two parts are seen
to be separated from each other by a groove, at the bottom of which is a non-chitinous
strip (C), this latter appearing to act as a hinge between the two parts of the lamella.
It seems probable that, as in Aselliis aquaticus and the Phreatoicidae (Sheppard, 1927,
p. 91), the anterior parts of these lamellae, together with the coxal lobes of the maxil-
lipeds, form an additional aerating apparatus for the marsupium.

In giving an account of the sexual characters of the Serolids, Beddard (18446, pp. 14,
15) states that when the brood lamellae of the female are fully developed "the sterna
of the thoracic segments undergo a retrograde development and almost disappear ", and
that "the young appear to be actually contained within the body of the mother, the
alimentary canal is pressed up against the dorsal surface of the body, and its cavity is
reduced to a minimum".

With regard to the first part of this statement, it seems more correct to say that,

GENERAL MORPHOLOGY

277

following the moult at which the brood plates are replaced by the fully formed brood
lamellae and the eggs are laid, the sterna of the thoracic somites remain soft, their usual
function having been taken on by the chitinous middle portions of the brood lamellae.
As a result of this, the cavity of the marsupium can be enlarged, not only by the bending
downwards of the brood lamellae but also by the pushing upwards of the membranous
ventral body wall. In this way accommodation for the increasing size of the developing
embryos is ensured. The second part of Beddard's statement that "the young appear to
be actually contained within the body of the mother " is inaccurate : the embryos undergo
their development within the marsupium and are separated from the body cavity of the
mother by the membranous ventral body wall.

An interesting result of the analysis of the adult female specimens of the species
represented in the present collection is that breeding goes on all the year round, and
that the number of females in the non-breeding condition is comparatively small.

The following table gives the details for S. schythei and S. exigtia throughout the
year. In the latter species, which is of a small size, the examination of the numerous
specimens for the presence or absence of the brood plates was very tedious and was
therefore omitted.

Month

Serolis schythei, $

Serolis exigua, $

In breeding phase

In non-breeding
phase

In breeding phase

Jan.

Feb.

March

April

May

June

July

Aug.

Sept.

Oct.

Nov.

Dec.

9

None collected

10

9

49

33

32

None collected

None collected

18

2

13

2
None collected

2

I

None collected

None collected

10

I

20

17 +

5

3

None collected

16 +
28 +
16 +

77 +
None collected

5

I

16

From the above table it may be seen that in every month, when the collections in-
cluded these species, females in the breeding phase were present, thus proving that they
exhibit no definite breeding season. In this they agree with the Amphipod Gammarus
chevreuxi (Sexton, 1924) and differ from the fresh-water Isopod Aselhis aquaticus which
appears to have a definite breeding season.

In Aselhis aquaticus, and also in the members of the fresh-water family Phreatoicidae,
females with the narrow plate-like type of brood lamellae as well as those with the broad
type are found, and Unwin's observations (1920, p. 335) on Aselhis show that the adult
female, during the breeding season, passes through alternating phases of breeding and
non-breeding, and that in the latter phase the broad lamellae are replaced by the narrow
plate-like type. On the other hand, Gammarus chevreuxi shows no such alternation ;

278 DISCOVERY REPORTS

when maturity is reached the lamellae are fully formed, and a moult occurs after each
brood is hatched. The brood-lamellae are fully formed at each moult and immediately a
new brood is reared.

The fact that female Serolids with the narrow type of brood plate are comparatively
rare, seems to suggest that those present are immature specimens rather than mature
females in the non-breeding phase, and that, as in Gammarus, once maturity is reached,
the brood-lamellae are fully formed after each moult and a new brood is reared. It
would be interesting to know how far these two breeding methods are characteristic of
fresh water and salt water respectively.

Nordenstam (1933) has deah with "the scales and setae in the family SeroHdae"
(pp. 14-38) in great detail: many of the types described by him are considered in the
present paper under the various descriptions of species.

KEY TO ALL KNOWN SPECIES OF SEROLIS.

A. Small dorso-lateral portions of the tergum and of the coxal plates of the eighth thoracic somite
present. Endopod of the uropod absent. All free thoracic somites (3rd-8th) separated from their
coxal plates by sutures. Hindmost suture of seventh thoracic somite complete.

I. Dorsal surface of head and body strongly sculptured; terminal segment studded with

tubercles 1. S. beddardi,Cz\mAn.

II. Dorsal surface of head and body segments nearly smooth; terminal segment smooth except

for a median and a pair of sub-marginal ridges 2. S. !atifrotis,MVhite.

B. Tergum and coxal plates of the eighth thoracic somite absent. Both endopod and exopod of the
uropod usually present.

I. Tergum of seventh thoracic somite articulating freely with that of the first abdominal segment,
that is, the hindmost suture of the seventh somite is complete.

A. All free thoracic somites (3rd-7th) separated from their coxal plates by sutures ; eyes small

and inconspicuous 3. 5. ^ran&, Beddard.

B. Only the first four free thoracic somites (3rd-6th) separated from their coxal plates by
sutures.

I. Coxal plates of seventh thoracic somite extend backwards to about the middle of the
terminal segment, pleural plates of second and third abdominal segments short, not
extending far beyond the anterior margin of the terminal segment 4. S.paradoxa, Fabricius.

II. Coxal plates of seventh thoracic somite extend backwards, in the male, for some
distance beyond the posterior extremity of the terminal segment; pleural plates of
second abdominal segment long, extending beyond the posterior extremity of the
terminal segment, those of the third segment short 5. S. schythei, Liitken."

III. Coxal plates of seventh thoracic somite extend to the postero-lateral angles of the
terminal segment, but not as far as the pleural plates of the second abdominal segment;
those of the third segment extend backwards to about the middle of the terminal
segment 6. 5. /)o/«m, Richardson.

1 5. serrei, Lucas, has not been recorded since first briefly described in 1877 and is possibly identical with
S. schythei.

KEY TO ALL KNOWN SPECIES OF SEROLIS 279

C. Only the first three free thoracic somites (3rd-5th), separated from their respective coxal
plates by sutures.

L Coxal plates of seventh thoracic somite not extending backwards beyond the pleural
plates of the second and third abdominal segments.

a. Pleural plates of second abdominal segment extending beyond those of third.
Terminal segment with well-developed spine in anterior median dorsal line,
followed by a median carina extending to its extremity; on either side of median,
two lateral oblique carinae terminating in small spines some distance from the
lateral margins y. S.gladalts,Tanersal\.

b. Pleural plates of third abdominal segment extending beyond those of second.

1. Extremity of terminal segment trifid ; dorsal surface of terminal segment with a
median carina and three lateral ones on either side of it. 8. S. septemcaiinafa, Miers.

2. Extremity of terminal segment fiot trifid; terminal segment broader than long
or about as broad as long.

a. Terminal segment with a well-developed spine in the antero-dorsal line,
followed by a median carina.

(i) Median carina slight, terminating in a spine at some distance from the
upturned, pointed posterior extremity; two lateral oblique carinae on
either side, terminating in spines ; two inner terminating in a line with the

end of the median carina g. S. kempt, n.sp.

(ii) Median carina extending to extremity of segment, with two lateral oblique
carinae on either side of it, each terminating in a small spine some distance

from the lateral margins 10. 5. /)o/iVa, Pfeffer.

^. Terminal segment without well-marked dorsal spine in an antero-median
position.

(i) Uropoda broad and extending beyond the posterior extremity of the
terminal segment ; body oval in shape and without spines

U.S. elliptica, n.sp.

(ii) Uropoda not broad and extending beyond the posterior extremity of the
terminal segment; cephalosome with dorsal posterior margin produced
backwards to form a median spine reaching to near the middle of the first
free (3rd) thoracic somite; spine less pronounced in the female

12. S. exigiia, Nordenstam.

IL Coxal plates of seventh thoracic somite extending beyond the pleural plates of the
second and third abdominal segments.

a. Pleural plates very short, not extending beyond the anterior margin of the terminal
segment.

1. Terminal segment with three dorsal carinae, the median one of which is not
mterrupted in the middle: lateral carinae following the outline of the segment
at some distance within it and meeting the central one at its posterior
extremity 13. 5. raww/a, Lockington

2. Terminal segment with three dorsal carinae, the median one of which is inter-
rupted in the middle.

a. Lateral carinae well developed and ending in spines; terminal segment

narrowing posteriorly ; body pear-shaped ... 14. 5. co«wxa, Cunningham^.

(8. Lateral carinae are well marked ; terminal segment almost circular in outline ;

body broadly oval 1^. S. gaudkliauJn, And. et Ed-w.

y. Median and lateral carinae almost obsolete ... 16. 5. /aew, Richardson.

1 S. plana, Dana, is probably identical with S. convexa.

4-2

28o DISCOVERY REPORTS

b. Pleural plates of second and third abdominal segments extending beyond the
anterior margin of the terminal segment.

1. Coxal plates of seventh thoracic somite extending backwards for about two-
thirds of the length of the terminal segment (excluding the terminal spine if
present).

a. Median dorsal portion of the cephalosome produced backwards into a spine
which extends to the middle of the fourth thoracic somite

17. S. gerlackei, Monod.

/3. Median dorsal spine of cephalosome absent; posterior extremity of terminal
segment produced to form a long spiniform process, the lateral margins of
which are serrated ; eyes devoid of pigment ... 18. S. meridionalis, Hodgson.

2. Coxal plates of seventh thoracic somite extending as far, or nearly as far, as the
posterior extremity of the terminal segment.

a. Coxal plates of seventh thoracic somite, in the male, extending backwards
almost as far as the posterior extremity of the terminal segment; pleural
plates of second and third abdominal segments to a short distance beyond
the articulation of the uropod; terminal segment longer than broad; body
almost circular in outline ... ... ... ... 19. S. comuta, Studer.

j8. Coxal plates of seventh thoracic somite extending backwards to a point not
far beyond the articulation of the uropod and only just beyond the posterior
extremities of the pleural plates ; terminal segment broader than long ; body
broadly ovate in shape ... ... ... 20. S. tiiIobitoides,Kights.

3. Coxal plates of seventh thoracic somite long and extending backwards beyond
the posterior extremity of the terminal segment.

a. Eyes absent; coxal plates comparatively short and flat, those of the seventh
thoracic produced backwards to a short distance beyond the posterior ex-
tremity of the terminal segment ; pleural plates of third abdominal segment
slightly longer than those of the second, and reaching to the basal joints of
the uropoda ... ... ... ... ... ...21. S. antarctica,'Btddard.

^. Eyes present but not deeply pigmented; coxal plates long and spiniform,
those of the seventh thoracic somite extending for a considerable distance
beyond the extremity of the terminal segment; pleural plates of third ab-
dominal segment in the male produced backwards as far as extremity of the
terminal segment ; those of second to a short distance beyond : eyes whitish
yellow in colour ... ... ... ... ... ... 22. S. brom ley ana, Suhm.

c. Pleural plates of second abdominal segment extending for some distance beyond
the posterior extremity of the terminal segment ; those of the third segment short,
not reaching beyond the antero-lateral margin of the terminal segment ; eyes large,

with bluish black pigment ... ... ... ... ... 23. S. neaera,'Qe.dda.rd.

n. Hindmost suture of seventh thoracic somite obsolete in the mid-dorsal line, so that for a
short distance the tergum is fused with the first abdominal segment; the third to the fifth
thoracic somites separated from their coxal plates by sutures.

A. Hindmost suture of sixth thoracic somite complete.

I. Coxal plates of seventh thoracic somite tiot extending backwards as far as the posterior

extremities of the pleural plates of the second and third abdominal segments.

a. Terminal segment almost triangular in shape, with a dorsal median keel, ending

posteriorly in a blunt prolongation ; thoracic somites each with median tubercle,

and free somites each with a lateral tubercle on either side, close to the junction of

the tergum with its coxal plate... ... ... ... ... 24. 5. W2«tt/a, Beddard.

KEY TO ALL KNOWN SPECIES OF SEROLIS 281

b. Terminal segment with a conspicuous median dorsal keel; posterior extremity
narrow and rounded ; dorsal surface of body without median or lateral tubercles

25. 5. bakeri, Chilton.

c. Terminal segment with slight median dorsal keel ; tubercles on cephalosome and

on thoracic somites absent ... ... ... ... ... 26. .S. j'OKj'e!, Hale.

IL Coxal plates of seventh thoracic somite extending backwards beyond the posterior
extremities of the pleural plates of the second and third abdominal segments.

a. Pleural plates short, not extending beyond the broad anterior margin of the
terminal segment; terminal segment triangular in shape with rounded angles,
slight median dorsal carina present ... ... ... ... 27. S. orbiculata, n.sp.

b. Pleural plates of second and third abdominal segments extending beyond the
anterior margin of the terminal segment; those of the third slightly the longer.

1. Terminal segment with median dorsal and two lateral carinae on either side, the
outer one of which follows the outline of the terminal segment 28. S. noiotropis, n.sp.

2. Terminal segment roughly five-sided, with a well-developed spine in an antero-
median position on the dorsal surface, followed by a low median carina, on
either side of which are two lateral carinae each terminating in a small spine
some distance from the lateral margin ... ... 29. S. pagenstechen',I'ief[er.

. Hindmost suture of sixth thoracic somite obsolete in the mid-dorsal line, so that for a
short distance the terga of the sixth and seventh somites are fused with each other and with
that of the first abdominal segment.

L Position of missing suture between the sixth and seventh somites indicated by a groove ;

uropod very small, endopod wanting ... ... 30. S. platygaster, n.sp,

II. Groove absent.

a. Coxal plates of seventh thoracic somite of male produced backwards for a short
distance beyond the attachment of the uropod, and beyond the level of the pleural
plates of the third abdominal segment; those of female extending backwards as far
as the pleural plates of the third abdominalsegment, which, in both sexes, are longer
than those of the second.

1. Margins of coxal and pleural plates and of terminal segment very much
thickened 3^- S. bouvieri, Richardson.

2. Margins not thickened; terminal segment with a well-developed spine in
antero-median position, followed by a median carina, on either side of which
are two lateral carinae terminating at some distance from margins in small
spines 32. S. aspera, n.sp.

b. Coxal plates of seventh thoracic somite not reaching backwards as far as the attach-
ment of the uropod. Pleural plates short.

1. Coxal plates of seventh thoracic somite of male extending backwards to Just
beyond the pleural plates of the third abdominal segment; surface of body
covered with tubercles 33. 5. fl;«?;a//e«s/.f, Beddard.

2. Coxal plates of seventh thoracic somite extending backward to the level of the
extremities of the pleural plates of the second abdominal segment but not as far
as those of the third.

a. Each of the segments of the body with a dorsal median curved, hook-like
spine, and with a row of short tubercles near its hinder border

34. S. elongata, Beddard.
j8. Segments with median spines and lateral tubercles. Terminal segment long,
with incurved sides, and with posterior truncate extremity concave

35. S. longicaudata, Beddard.

283 DISCOVERY REPORTS

in. Tergum of seventh thoracic somite absent except for small lateral portions which are con-
tinuous with the coxal plates, these are of normal size; hindmost suture of sixth thoracic
somite complete; third to fifth thoracic somites separated from their coxal plates by sutures.

A. Head triangular in shape, with apex directed backwards and produced into a large median
spine ... ... ... ... ... ... ... ... ... 36. S. pallida, Beddard.

B. Head four-sided ; pointed tubercles along the posterior margins of the anterior thoracic
somites, and median dorsal tubercles on all the segments ... ... t,-]. S. tuherculaia, Gmht.

DESCRIPTION OF SPECIES

I . Serolis beddardi, Caiman (Fig. 3).

S. beddardi. Caiman, 1920, p. 299, figs. 1-3.
Occurrence. St. 174: South Shetlands, 5-10 m.; 4 J(J, i ? (b.) and 9 immature specimens.

Diagnostic characters. The largest male in the collection measures 22 mm. in
length and 17 mm. in greatest breadth, a size approximately the same as that of
Caiman's specimen; the adult female is incomplete, but has a breadth of 15-5 mm.,
which is I mm. less than that of Caiman's largest specimen.

Fig. 3. Serolis beddardi, Caiman.
a, cutting edges of mandible: x 20. b, second thoracic appendage: x 7.
c, third thoracic appendage of ,^: x 16. d, type of setae found on limbs.

The dorsal surface of the head and body segments is strongly sculptured and the
terminal segment is studded with rounded tubercles which lie between the median and

DESCRIPTION OF SPECIES 283

sub-marginal ridges ; a pair of short sub-median ridges is also present ; the eyes are
rather small, reniform in shape and containing black pigment.

The coxal plates, which are all fringed with long setae, are comparatively small and
are all separated from their respective thoracic somites by sutures ; in this species, as
Caiman points out, a vestige of the tergum and of the coxal plate of the last thoracic
somite may be seen on either side of the first abdominal segment. The coxal plates of
the seventh thoracic somite (sixth free) do not extend posteriorly beyond the base of
the terminal segment, which is broad, and extends laterally beyond the pleural plates of
the second and third abdominal segments.

Remarks. The original account of this species does not include descriptions of the
appendages: a few notes on them are given here.

The surfaces of both the antennules and the antennae are covered with minute rounded
scales. The first peduncular joint of the antennule is short, the second and third are
each about twice the length of the first ; the fourth is very small, being about half the
length of the first flagellar joint; the posterior margins of the second, third and fourth
joints are densely fringed with long setae ; the flagellum consists of nineteen joints. The
first three joints of the antenna are very short, the fourth is twice as long as the third,
and the fifth twice as long as the fourth ; the posterior distal angle of the fifth joint is
produced into a rounded process and its posterior border is fringed with long setae ;
similar setae are present in tufts on the anterior margin, and a group is also present on
the posterior distal angle of the fourth joint; the flagellum consists of nineteen joints,
the distal one of which is very small and bears a number of delicate setae.

The mandibles are of the usual form ; the terminal joint of the palp bears the usual
type of setae, but these are absent from the second joint, their place being taken by a
few long simple setae which are scattered along its margin, as well as along that of the
basal joint. The form of the cutting edge of both the right and the left mandible may be
seen from Fig. 3 a. The outer lobe of the maxillula bears eleven strong spines on its
truncate distal extremity; the inner lobe is about two-thirds the length of the outer and
bears a single short seta on its much broadened distal end. There are two pectinate setae
on the outer and four on the inner lobe of the maxilla, and about twenty on the fixed lobe.
The maxilliped has its surface covered with minute imbricating scales ; the basipodite
is separated by a suture from the lamella, which is longer than broad with a concave
distal extremity ; the lower half of its outer margin is fringed with long hairs, whilst the
upper half is toothed : the inner margin of the basipodite is fringed with delicate setae,
scattered amongst which are stronger ones. The second joint of the palp is broad, whilst
the terminal joint is long and almost parallel-sided; setae are present on its rounded
distal extremity and on the upper inner margin of the second joint.

The second thoracic appendage (Fig. 3 h) is short and robust ; the propodus is broad
and its inner margin is armed with a row of broadly rounded processes alternating with
a row of modified spines ; the distal edge of the carpus is crenulate and bears two short
stout spines. The third thoracic appendage of the male (Fig. 3 c) is considerably longer
than the second ; the propodus has its inner margin armed with a double row of spines

284 DISCOVERY REPORTS

of the type usually found on this limb — there are eight spines in each row ; the dactylus
ends distally in a strong spine, arising just behind which are two other spines and three
short setae; its inner margin appears to be raised into a number of minute papillae.
With the exception of the merus and dactylus, the other joints of the 4-7 appendages
are fringed with strong setae, many of which are of the type shown in Fig. 3 d, whilst
some, at the distal ends of the joints, bear much stronger pectinations.

The first three pairs of pleopods are of the usual type ; the protopodite of each is
somewhat triangular, with the prolonged angle bearing three plumose setae on the first
and two on each of the second and third pleopods : long delicate hairs are present around
the margins of the protopodite and also fringe the anterior margins of both the exo-
podite and endopodite. The suture of the exopod of the fourth pleopod is oblique and
towards the distal end ; both margins of the exopod are fringed with hairs, those of the
inner margin are replaced towards the distal end by plumose setae, and a group of these
setae surround the distal extremity ; the ventral surface of the exopod is covered with
long scattered setae which increase in size and number distally; the endopod has a
rounded distal extremity and bears no setae.

The uropoda have already been described by Caiman (1920, pp. 300, 304, fig. 3).
Each consists of a protopodite which is prolonged into an acute point ; the exopod is
small and spiniform; the endopod is wanting.

Distribution. The species has not been recorded since Caiman described it; the
present specimens come from the type locality. Deception Island, South Shetlands.

2. Serolis latifrons, White.

S. latifrons. White, 1847, p. 106; Miers, 1875, p. 74; 1876, p. 116; Smith, 1876, p. 63; 1879, p. 204;
Studer, 1879, p. 26 ; Beddard, 1 8846, p. 44, pi. ii, figs. 1-4 ; Vanhoffen, 1914, p. 5 18 ; Caiman, 1920, p. 299.

Diagnostic characters. The largest specimen in the Challenger collection is a female
32mm. in length and 24 mm. in greatest breadth; the males are proportionately
broader than the females, the largest measuring 28 mm. in length and 24 mm. in
breadth : the body is oval in shape with an almost smooth dorsal surface.

In this species, as in S. beddardi, Caiman, minute lateral portions of the tergum of
the eighth thoracic somite are present, and separated from them by sutures are small
coxal plates . The coxal plates of the remaining free thoracic somites are separated from
their respective somites by sutures ; they are closely applied together and those of the
seventh somite do not extend backwards as far as the bases of the uropoda.

The terminal segment is almost triangular in shape, its anterior margin extends
laterally beyond the pleural plates of the second and third abdominal segments, which
are short ; near the antero-lateral margin, on either side, is a notch into which the base
of the protopodite of the uropod fits. A median and a pair of lateral sub-marginal ridges
are present on the dorsal surface.

The protopodite of the uropod is produced distally as an acute prolongation ; the
exopod is reduced and spiniform and the endopod is absent.

Distribution. Off Kerguelen.

DESCRIPTION OF SPECIES 285

3. Serolis gracilis, Beddard.

S. gracilis, Beddard, 18840, p. 332; 18846, p. 6i, pi. iii, figs. 7-13.

Diagnostic characters. The largest male specimen in the Challenger collection
measures 21 mm. in length and 22 mm. in greatest breadth; the females are much
smaller, being 9 mm. in length and 8 mm. in greatest breadth.

The outline of the body is almost circular and its surface is covered with scattered pits ;
its colour in alcohol is a dark slate-blue, varying to reddish yellow upon the terga of
the posterior thoracic somite and the abdominal segments : the cephalosome is strongly
convex between the eyes, whilst the antero-lateral areas are flat and depressed; the
rostrum is very short ; the eyes are small and appear to contain little or no pigment.

The coxal plates are flat and sickle-shaped and gradually increase in length from
before backwards : in the male those of the seventh somite extend for some distance
beyond the end of the terminal segment, whilst those of the female are much shorter
and do not reach to its extremity. All the free thoracic somites (3-7) are separated from
their respective coxal plates by sutures.

The pleural plates of the second abdominal segment are longer than those of the third ;
in the male they extend for a short distance beyond the end of the terminal segment,
whilst those of the female do not extend farther than the middle of the segment ; the
plates of the third segment in the male are about the same length as those of the second
in the female, whilst the third in the female are considerably shorter.

The terminal segment is squarish in outline, it possesses "a slight longitudinal
median keel which is crossed at right angles by a sinuous ridge with three convexities,
one median and two lateral, which correspond to the spines on the caudal shield of
S. neaera, Beddard and S. schythei, Lutken ; about the end of the anterior third of the
caudal shield is a short flat spine in the middle line and two oblique ridges, one on either
side of this spine; the lateral portions of the caudal shield are bent down".

In the male, plumose hairs are present on the inner margin of the merus and carpus
of the third appendage and on the merus, carpus and propodus of the last.

Remarks. My observations agree with those of Beddard except that in his description
of the maxilliped he says that "the stipes and lamina are not separated by a complete
suture " and in his figure (pi. iii, fig. 10) this suture is also omitted. In the specimens at
the British Museum a suture between the basipodite and the lamella of this appendage
is clearly visible.

Distribution. Off Pernambuco, South America, in 675 fathoms.

4. Serolis paradoxa, Fabricius.

Oniscus paradoxus, Fabricius, 1775, p. 296; 1787, p. 240.

Cymothoa paradoxa, Fabricius, 1792-8, Suppl. pp. 304, 503.

Serolis fabricii, Leach, 18 18, p. 340.

Serolis orbignyi, Audouin and Milne-Edwards, 1841, p. 25.

Serolis orbigniana, Milne-Edwards, 1840, p. 232; Grube, 1875, p. 225.

Serolis paradoxa, Miers, 1881, p. 61 ; Beddard, 1884*, p. 33, pi. v, figs. 12-14; Nordenstam, 1933,

PP- 51-5. text-figs.

■nviT 5

286 DISCOVERY REPORTS

Diagnostic characters. The male of this species is a Httle broader in proportion to
its length than is the female, the largest male in the Challenger collection measuring
24-5 mm. in length and 25 mm. in breadth, and the largest female 27-5 mm. in length
and 26 mm. in breath.

The head is broadest anteriorly and in this point differs from that of S. schythei,
Liitken, and S. poIaris, Richardson, two closely allied species, in both of which it is
broadest at about the level of the eyes. The flagellar joints (7-16, in one specimen) of the
antenna of both sexes are furnished with a series of short recurved hooks which corre-
spond with the "lamellar-like" structures of S. schythei.

The body is less flattened than in the two allied species ; the coxal plates are smaller,
and those of the first four free somites are separated from them by sutures. I noticed
that in this species the positions of the sutures on the tergum of the fifth free somite,
which have disappeared, are indicated by shallow grooves. The coxal plates of the seventh
somite are produced backwards to a level about the middle of the terminal segment ; the
pleural plates of both the second and third abdominal segments are short, and do not
extend far beyond the anterior margin of the terminal segment ; those of the third are
slightly longer than those of the second segment.

The ischium, merus and carpus of the third appendage in the male are furnished with
abundant plumose hairs arranged in two rows : these are absent from the corresponding
joints of this appendage in the two allied species. The terminal segment is more tri-
angular than in the other two species, and it bears on its dorsal surface a median
longitudinal keel on either side of which is a lateral one. A strong median spine is
situated in an anterior position, and running parallel to the anterior margin is a trans-
verse carina of which only traces exist in S. schythei; the transverse ridge of the latter
is absent in S. paradoxa.

Distribution. Off the Falkland Islands, and adjacent shores of Patagonia.

5. Serolis schythei, Liitken (Figs. 2b,c, ^a,b; Plate XIV, fig. i).

S. schytliei, Lutken, 1858, p. 98, pi. i, figs. 12, 13; Grube, 1875, p. 220, pi. vi, fig. i ; Beddard, 18846,
p. 40, pi. ii, figs. 5-13; Tattersall, i92i,p.227; Giambiagi, 1925, pp. 1 1-12, pi. ii, fig. 3 ; Nordenstam,
1933. PP- 55-7. text-figs.
Occurrence. St. 51 : East Falkland Island, 115 m.; 19 (JcJ, 45 ?? (b.), and a number of immature specimens.
St. 223 : Cape Horn, 63 m. ; 6 ?? (b.), 4 immature. St. WS 72 : 51° 07' S, 57° 34' W, 79 m. ; 2 .^cJ. St. WS 73 :
51° 01' S, 58° 53' W, 121-130 m.; 5 <?<?, 4?? (b.), 2 immature. St. WS 75: 51° 01' 30" S, 60° 31' W, 72 m.;
2cJ<?, i$(b.), I immature. St WS76: 5i°oo'S, 62° 02' 30" W, 207-205 m.; a number of immature.
St. WS78: 5i°oi'S, 68° 04' 30" W, 91-95 m.; 3 immature. St. WS79: 5i°oi'3o"S, 64° 59' 30" W,
132 m.; I ?(b.), a number of immature. St.WSSo: 50° 57' S, 63°37'3o" W, 152-157 m.; 2cJo, 4?? (b-).
a number of immature. St. WS 83 : 14 miles S 64° W of George Island, East Falkland Island, 137-129 m.;
I immature. St. WS 90: 13 miles N 83° E of Cape Virgins Light, Argentine Republic, 82-81 m.; 3 ,5,^,
3?? (b.), and a number of immature. St. WS 210: 50" 17' S, 60° 06' W, 161 m.; i o, i $ (b.), i immature.
St. WS211: 50° 17' S, 6o°o6'W, 174 m.; I immature. St. WS214: 48° 25' S, 60° 40' W, 208-
219 m.; 1$ (non-b.), I? (b.), a number of immature. St. WS215: 47° 37' S, 60° 50' W, 219-146 m.;
5 immature. St. WS216: 47° 37' S, 60° 50' W, 219-133 m.; 2,5c?, 5??(b.), a number of immature.
St. WS 219: 47° 06' S, 62° 12' W, ii6-ii4m.; I4(J<J, 22$$ (b.), a number of immature. St. WS225:
50° 20' S, 62° 30' W, 162-161 m.; 2 immature. St. WS 226: 49°2o' S, 62° 30' W, i44-i52m.; i immature.

DESCRIPTION OF SPECIES 287

St. WS 229: 50° 35' S, 57° 20' W, 210-271 m.; i 5 (b.), 5 immature. St. WS 233 : 49° 25' S, 59° 45' VV,
185-175 m.; 3 $? (b.), 2oc?, a number of immature. St. WS 234: 48° 52' S, 60° 25' W, 195-207 m.;
2 $$(b.), 6 immature. St. WS 235 : 47° 56' S, 61° 10' W, 155 m.; i <?, 3 $$ (b.), i immature. St.WS236:
46° 55' S, 6o°4o'W, 273-300 m.; i immature. St.WS237: 46° 00' S, 6o°05'W, 150-256 m.; ioV?(b.)
I $ (non-b.), i2<:JcJ, a number of immature. St. WS 239: 51° 10' S, 62° 10' W, 196-192 m.; 2$? (b.), i <S,
I immature. St. WS 243 : 51° 06' S, 64° 30' W, 144-141 m.; 11 immature. St. WS 244: 52° 00' S,
62° 40' W, 253-248 m.; 2 33, 3 immature. St. WS 245 : 52° 36' S, 63" 40' W, 304-290 m.; 3 immature.
St. WS752: 5i°2o'S, 63°i7'W, 160 m.; i young t?. St. WS754: 5i°09'3o"S, 58° 54' W, 106 m.;

I immature. St. WS 758: 48° 32' S, 61" 19' W, 112 m.; a number of immature. St. WS 763 : 44° 14' S,
63° 28' W, 87-82 m.; I immature. St.WS765: 45°07'S, 60° 28' 15" W, ii3-ii8m.; 3$? (b.), 55 immature.
St. WS771: 42°4i'45"S, 6o°3i'W, 90 m.; y^S, 8 ?$ (b.), 10 ?? (non-b.), a number of immature.
St. WS 772: 47° 28' S, 6o°5i' W, 309-162 m.; 13 J, 7$ (b.), a number of immature. St.WS 774: 47° 08' S,
62°02'W, i39-i44m.; i ? (non-b.). St. WS 775: 46° 44' 45" S, 63° 33' W, 115-110 m.; 6 (?<?, 2 ?? (b.),
a few immature. St. WS776: 46° 18' 15" S, 65° 02' 15" W, 107-99 m.; i <J. St. WS 781 : 50° 30' S,
58°5o'W, 148 m.; 2(J(J, a number of immature. St. WS782: 50° 29' 15" S, 58° 23' 45" W, 141m.;

II (J,^, 7?? (b.), 8?? (non-b.), a number of immature. St. WS 783: 50° 02' 45" S, 60° 14' W, 159-0 m.;
533, I? (b.), 2?? (non-b.), 7 immature. St. WS786: 49° 07' S, 63° 55' W, i33-ii9m.; 7 immature.
St. WS 787 : 48° 44' S, 65° 24' 30" W, 106-1 10 m. ; 5 <J(J, 2 ?? (non-b.), a number of immature. St. WS 791 :
45° 41' 45" S, 62°45'W, 96-101 m.; i? (b.), i? (non-b.), a number of immature. St. WS797:
47°45'36"S, 64° 20' W, 115-111 m.; 2?? (non-b.), a number of immature. St. WS802: Haul A,
50° 45' 45" S, 61° 22' W, 128-13201.; I, ^, 2$$ (b.), I ? (non-b.), several immature. St.WS 802: Haul B,
50° 43' 45" S, 6i°26'W, 132-139 m.; g<S3, 6S?(b.), 6?? (non-b.), several immature. St. WS 814:
5i°45'i5"S, 66°4o'W, iii-ii8m.; i immature. St. WS815: 51° 51' 45" S, 65° 44' W, i32-i62m.;
I immature. St. WS 816: 52° 09' 45" S, 64° 56' W, 150 m.; i cj, 7 immature. St. WS 818:
52° 31' I5"S, 63° 25' W, 272-278 m.; 7 immature. St.WS 824: 52° 29' 15" S, 58° 27' 15" W, 146-137 m.;
2immature. St WS 825: 50° 50' S, 57° 15' 15" W, 135-144 m.; id", 6?? (b.), several immature. St. WS866:
50° 37' 45" S, 64° 15' W, 137-144 m.; I immature specimen.

Diagnostic characters. The largest male in the collection measures 28 mm. in
length and 33 mm. in greatest breadth, whilst a female in the breeding phase measures
23 mm. in length and 23 mm. in greatest breadth. The male (Plate XIV, fig. i) is thus
considerably broader in proportion to its length than the female, the latter being almost
circular in outline.

The head is broadest at the eyes, which are large and prominent. Between them are
three somewhat triangular areas, the two lateral of which are raised to form low
prominences, while the central one forms the posterior boundary of the head. The body
is flattened; the coxal plates are sickle-shaped, with their outer halves projecting freely,
those of the first four free somites are marked off by sutures. In the male the coxal plates
of the seventh thoracic somite are long, extending, in the largest specimen, for the space
of about 5 mm. beyond the extremity of the terminal segment. The posterior margin
of the tergum of each thoracic somite, except the first, is curved backwards in the middle
line into a short spine; these diminish in size from before backwards.

The pleural plates of the second abdominal segment are long and reach beyond the
end of the terminal segment in the male, but not quite to the end in the female ; those
of the third abdominal segment are short and do not extend far down the terminal
segment. The terminal segment is hexagonal in outline, its posterior end projecting as
a short spine. It bears a median dorsal and two lateral keels, the latter starting from the
base of a strong, stout spine, which lies in a median anterior position, and terminating

5-2

288 DISCOVERY REPORTS

in two short spines placed some distance in front of the bases of the uropoda. A trans-
verse curved ridge, which is prolonged backwards into three short spines, the median
one the larger, extends across the segment in a line with the terminal spines of the lateral
keels.

Remarks. Beddard ( 1 884 Z), p. 43) states that the antennules and antennae are of equal
length and that " the antennae of the males possess a series of delicate lamellar processes
(pi. ii, fig. 6) arranged in a single line along the inner side of most joints, being absent
only from a few of the extreme distal and proximal joints". He also says that "the
anterior margin of the middle joints has a series of short curved spines like those already
-described for S. porodoxa". I find on examination that the antenna of this species is
considerably longer than the antennule, the latter only extending to the end of the
peduncle of the former. I also find that there is only one set of spines on the flagellar
joints of the antennae, and that these correspond with the " deHcate lamellar processes "
of Beddard's description. The spines are also present in the female, so that they can no
longer be regarded as a secondary sexual character. These lamellar processes, which are
present on the ventral surface of the joints, presumably correspond with the "short
curved spines" of S. paradoxa; similar though somewhat smaller processes occur in
S. polaris.

The maxillae are of the usual type, except that the inner articulating lobe is slightly
broader than that of most species and bears seven pectinate setae instead of the more
usual numbers of either two or four : the fixed lobe bears fourteen or fifteen of these
setae.

The maxiUipeds (Figs. 2 b, c) agree with Beddard's description except that the "small
protuberance carrying three stoutish hairs", which he states (1884 />, p. 43) to occur
on the summit of the terminal joint of the palp, appears to be an extra, very minute
joint which articulates with the so-called terminal joint at its outer distal angle. This
small joint which occurs in other species of the genus probably represents the vestigial
fourth joint of the five-jointed palp of the ancestral type.

The thoracic appendages have been described already, but the form of the dactylus
of the third appendage of the female and of the immature male, and of the fourth of the
adult male appears to have been overlooked. The dactylus of these appendages (Fig. 40)
is somewhat flattened, and bears, on the anterior edge of the distal third of its length, a
series of curved though flattened spines which increase in size towards the extremity.
Arising between the two most distal of the spines is a strong seta which is densely
plumose, and three setae, the middle one of which is sparsely plumose, spring from a
slight depression on the posterior margin. The dactylus of the other appendages is less
flattened and never bears this row of spines, but a plumose seta (often broken), a couple
of strong spines and a few setae may be present at its distal extremity. The dactylus of
the seventh appendage in the male is shorter than in the female, recurved, bearing a
number of hairs on its outer margin, whilst the propodus and carpus bear groups of
strong setae on their inner margins.

The endopod of the fourth pleopod is bifid, with the inner branch prolonged for some

DESCRIPTION OF SPECIES

289

distance beyond the outer, and with its rounded extremity fringed with delicate hairs.
The exopod of the uropod is shorter than the endopod; the latter is broad, with its
distal end obliquely truncated, and with the inner angle rounded and the outer pointed
and slightly produced (Fig. 4 b).

Colour. Pale brown, becoming darker in the middle of the body, with dark brown
or black spots which vary considerably in number and size in different specimens.

Distribution. From the Gulf of Peiias, in the north to the Falkland Islands in the
south ('Challenger').

i

C.

Fig. 4. Serolis schythei, Liitken, 5. polaris, Richardson, and 5. glacialis, Tattersall.

S. schythei, Lutken. a, dactylus of third thoracic appendage of ?: x 20. b, uropod: ;< 20.

5. polaris, Richardson, c, type of setae on propodus of second thoracic appendage, d, third thoracic

appendage of immature o- e, fifth thoracic appendage of ^J. /, uropod.
S. glacialis, Tattersall. g, cutting edges of mandible: ;■' 42.

[Serolis serrei, Lucas.

S. serrei, Lucas, 1877, pp. cxlv, cxlvi; Beddard, 18846, p. 32.

This species has not been recorded since it was first briefly described by Lucas in
1877 from a collection made by the French vessel ' Magicienne ' in the Strait of Magellan.
The description given by Lucas is repeated in the Challenger Report (Beddard, p. 32).

According to Lucas the species is closely allied to S. trilobitoides. It measures 27 mm.
in length and 36 mm. in breadth ; its dorsal surface is covered with small brown spots,
whilst the ventral surface and the surface of the limbs is of a yellow colour. The coxal
plates are transparent and not serrated at their margins. The terminal segment is broader

290 DISCOVERY REPORTS

than long, rounded, ending posteriorly in a fairly large median acute spine; three
carinae are present on its dorsal surface, and its lateral margins are not toothed.

Beddard (1884 b, p. 32) remarks : "it seems to me, from the above quoted description,
that it is probably more nearly related to Serolis scJiythei". I agree with Beddard that
this may be so. This view is supported by the fact that the distribution of S. serrei agrees
with that of S. schythei and differs from that of S. trilobitoides. The description of the
species is so brief that it is impossible to make any definite statement with regard to its
exact systematic position.]

6. Serolis polaris, Richardson (Figs. 4 c-f; Plate XIV, fig. 2).

Serolis polan's, Richardson, 191 1, pp 396-8, fig. i ; Nordenstam, 1933, p. 58.

Diagnostic characters. This species was instituted to contain six specimens, all of
which were said to be females. Through the kindness of Prof. Ch. Gravier, of the
Museum d'Histoire Naturelle, I have been able to examine the type specimen and find
that it is undoubtedly a male ; for although it is immature the appendix masculina on the
second pleopod is quite well developed. This specimen measures 19 mm. in length and
21 mm. in breadth. The measurements given by Richardson are presumably those of a
female and are length 22 mm. and breadth 22 mm., so that as in so many species of this
genus the male is broader in proportion to its length than is the female.

The head is broader than long, with its broadest portion on a level with the eyes, which
lie in a posterior position just within the lateral margins. The eyes are reniform in shape
and are less prominent and less deeply pigmented than those of S. schythei, Liitken.

The body (Plate XIV, fig. 2) bears a close resemblance to that of S. schythei, although
it is slightly less flattened. As in that species the posterior margin of the free thoracic
somites is produced backwards to form a small median spine, that of the fifth somite
being the largest. The coxal plates of the free thoracic somites have their postero-lateral
angles acutely produced backwards. The first four pairs are separated from their re-
spective somites by sutures, and just within the suture, on either side, the tergum is
raised into an angular prominence. The coxal plates of the seventh thoracic somite
extend backwards to the postero-lateral angles of the terminal segment, the pleural
plates of the second abdominal segment to the end of the terminal segment and those
of the third segment to about the middle of the segment. The posterior margin of each
of the abdominal segments is produced to form a median tooth similar to those of the
thoracic somites.

The terminal segment is roughly hexagonal in outline, with both the anterior and the
posterior angles very obtuse, and with the sides shghtly converging and terminating in
an acute tooth on either side, into the angle of which the protopodite of the uropod is
inserted.

The dorsal surface is furnished on either side with a curving ridge, situated close to
the anterior and lateral margins of the segment ; behind this, arising from a small flat
angular process, one on either side of the base of a large median spine, is a second ridge
which runs obliquely outwards to terminate in an angular process near the postero-

DESCRIPTION OF SPECIES 291

lateral angle of the segment. The median spine is long, prominent and acute, and is
situated in the anterior part of the segment ; behind it the median portion of the segment
is slightly keeled, and a transverse curved ridge, produced into three acute, flat spines,
one in the median line and one on either side, passes across the segment at the level of
the angular process of the oblique ridge.

Remarks. The appendages of this species have not been described; the following
notes are based on observations made on the type specimen:

The antennule and the antenna have been described by Richardson (pp. 396-7), who,
however, has not observed the structure of the flagellum of the antenna. As in S. schythei,
Lutken, a series of delicate lamellar processes is arranged in a single row on the ventral
surface, near the posterior margin of most of the joints— these processes are sHghtly
smaller than those present in 5. schythei.

The basipodite of the maxilliped is fused with the lamella as in S. schythei (Fig. 2 b) :
it was impossible to see whether or not the extra minute joint of the palp was present.

The second thoracic appendage is modified in the usual way; the propodus bears
modified spines of the types shown in Fig. 4 c. The third appendage (Fig. 4 d) of this
specimen has not developed the characters of the adult male. It is, however, stouter and
shorter than the remaining thoracic appendages, and bears a row of toothed spines along
the propodus ; the dactylus is modified in a similar way to the corresponding appendage
of S. schythei (Fig. 4 a). Long plumose setae, similar to those found on the pleopods,
are present on the posterior border of the ischium and at the posterior distal angles of
the merus and carpus of this and the remaining thoracic appendages (Fig. 5 e).

Each of the sternal plates of the first three abdominal segments has the median pos-
terior border produced into a spine. The protopodite of each of the first three pairs of
pleopods is characterized by the absence of plumose setae; in this they agree with
S. schythei and S. paradoxa.

The suture of the exopod of the fourth pair is oblique and the distal extremity of the
endopod is bifid, with the inner branch extended some distance beyond the outer; the
endopod of the fifth pleopod has a rounded distal extremity.

The uropoda (Fig. 5,/) do not extend to the tip of the terminal segment; the exopod
is shorter than the endopod, both are broadly rounded, toothed, and fringed with plumose
setae.

Distribution. South Sandwich Islands.

This species is undoubtedly very closely related to S. schythei, Lutken, and also to
S. paradoxa, Fabricius. The following points result from a comparison of the three
species :

(i) The males of 5. schythei are proportionately broader than those of the other two
species, whilst the body of the former is more flattened than that of the other two.

(2) The head of S. polaris and S. schythei is broadest at the level of the eyes, whilst
that of S. paradoxa is broadest anteriorly. The eyes of S. schythei are larger than those
of S. polaris.

(3) In all three species the first four free somites are separated by sutures from their

292 DISCOVERY REPORTS

respective coxal plates ; in S. paradoxa the position of the suture on the fifth free somite
is indicated by the presence of a slight groove.

(4) In S. polaris the coxal plates of the seventh thoracic somite extend tv .he level of
the bases of the uropoda, whilst the pleural plates of the second abdom" lal segment
extend beyond them to the end of the terminal segment : the pleural plates of the third
segment are a little shorter than the coxal plates of the seventh thoracic somite. In
S. schythei the coxal plates of the seventh thoracic somite extend backwards for some
distance beyond the end of the terminal segment, whilst the pleural plates of the second
abdominal segment extend almost to the postero-lateral angle, and those of the third
to just beyond the anterior margin of the terminal segment. In S. paradoxa the coxal
plates of the seventh thoracic somite extend to about the middle of the terminal segment
and the pleural plates of both the second and third abdominal segments are short, only
just reaching beyond the anterior margin of the terminal segment.

(5) The terminal segment of both S. schythei and S. polaris is more or less hexagonal
in outline, that of S. paradoxa is more triangular ; the spines on the transverse ridge,
which are present in the two former species, are much larger in S. polaris than in
S. schythei. The transverse ridge which is situated near the anterior margin in both
S. polaris and S. paradoxa is hardly seen in S. schythei. The large anterior median spine
is present in all three species.

(6) The lamellar processes which occur on some of the joints of the flagellum of the
antenna in S. schythei are present, only less well developed, in S. polaris. Their place
is taken by strong spines in i?. paradoxa.

(7) The modification of the form of the dactylus of the third appendage in the female
and immature male and of the fourth appendage in the mature male is seen in both
S. schythei and S. polaris.

(8) The setae on the thoracic appendages of S. schythei are either simple, or serrated
or pectinate ; those of S. polaris are either simple or plumose, like those on the pleopods.
The setae on the thoracic appendages of S. paradoxa are like those of S. schythei, but
in this species the inner margin of the ischium, merus and carpus of the third appendage
of the adult male is fringed with delicate plumose setae.

From the above comparison it is obvious that the three species are quite distinct.

7. Serolis glacialis, Tattersall (Figs. 2 d, e, 4^).

S.glacialis, Tattersall, 1921, p. 228, pi. vii, figs. 1-5; Monod, 1926, p. 35, figs. 33, 34.

S. glacialis, Tattersall, var. austrogeorgiensis, Nordenstam, 1933, p. 65, pi. i, fig. i, text-fig. 16.

Occurrence. St. 180: Palmer Archipelago, 160-330 m.; i $ (b.).

St. 181 : Palmer Archipelago, 160-335 m.; 4,^^, 5 ?? (b.)

St. 182: Palmer Archipelago, 278-500 m.; 2 ?? (b.), 6 immature specimens.

St. 187: Palmer Archipelago, 259 m.; i ? (non-b.).

Diagnostic characters. The body is broadly oval, slightly longer than broad, semi-
translucent, especially laterally. The head is nearly twice as wide as long with a small
rostral process between the bases of the antennules. Behind this process is a well-
marked transverse ridge which extends laterally to the sides of the cephalosome, and

DESCRIPTION OF SPECIES 293

between the anterior ends of the eyes is a second transverse ridge behind which is a deep
groove. The portion of the head between the eyes is raised into three oval prominences,
the posterior margins of which are more sharply defined than the anterior. The eyes are
large, reniform, with black pigment.

Each of the third to the seventh thoracic somites is produced in the median dorsal line
into a short but distinct spiniform process. The coxal plates are well developed and those
of the first three free somites are separated by sutures ; those of the seventh thoracic
somite are produced backwards and extend about to the level of the basal joint of the
uropoda. The pleural plates of the second abdominal segment extend backwards beyond,
and those of the third abdominal segment as far as, the coxal plates of the seventh
thoracic somite. The terminal segment is furnished with a well-developed spine in the
anterior median line, followed by a median dorsal keel which extends to the extremity.
On each side of the median keel there are two lateral oblique ones, terminating in small
spines some distance from the lateral margins, which are slightly turned downwards.

Remarks. The only specimens of this species previously recorded are two males, the
type specimen described by Tattersall (1921) and a second one identified by Monod
(1926). The present collection contains nineteen specimens, of which nine are adult
females and four adult males; all these are smaller than the type specimen, which
measures 17 mm. in length and 14-5 mm. in breadth. The largest male in this collection
is only 12 mm. long and 10 mm. broad, and the largest female is 1 1 mm. long and 8 mm.
broad.

The specimens agree with the description given for the type, but there appears to be
a sexual diff'erence in the number of joints present in the flagellum of the antennule; in
the males the number of joints is twenty-eight, compared with thirty-five in the type,
whilst in the females the number of joints is only sixteen. A similar sexual diff'erence
also occurs in S. kempi, n.sp., and S. aspera, n.sp.

The form of the cutting edges of the mandibles is shown in Fig. 4 g. The maxillula
is of the usual type: the outer lobe bears ten stout curved spines, nine of which are
toothed, whilst the tenth innermost one is more dehcate and pectinate; the inner lobe
is about two-thirds the length of the outer, and is somewhat expanded for a short dis-
tance behind its distal extremity ; the latter bears a single short seta.

The maxilla bears two spines on its outer, three on its inner, and nine on its fixed
lobe; these spines each possess a double row of short pectinations. The maxilliped
(Fig. 2 d, e) belongs to the type in which the basipodite is separated from its lamella by
a suture ; the lamella is almost square with the angles of its distal extremity rounded.
The sexual difference in the form of the coxal joint may be seen from the figures.

The thoracic appendages have been described and figured by Tattersall (1921, p. 229,
pi. vii, figs. 2-5). The protopodite of each of the first three pleopods is triangular in
shape, and the produced angle bears three long plumose setae on the first and two on
each of the second and third pleopods. The suture of the exopod of the fourth pleopod
is almost transverse and the endopod is rounded and undivided. The exopod of the
fifth pleopod is also partially divided by a transverse suture which extends from its

294 DISCOVERY REPORTS

outer margin for about a third of its width : two plumose setae are present at its extremity.
The endopod of the fifth pleopod is about the same size as the exopod and is undivided.

Distribution. The specimens in this collection come from the Palmer Archipelago,
a locality on the opposite side of the Antarctic Continent and not so far south as that
given by Tattersall for his type specimens. The latter were taken ofl^ Oates Land,
69° 43' S, 163° 24' E.

Since the above was written, Nordenstam (1933, p. 65) has described a new form of
S. glacialis, a variety austrogeorgiensis. The description is based on a single young female
which appears to differ from the main species. Of the ten points of difference noted
between this variety and the typical form, only three seem to be of real diagnostic
value, namely:

(i) The posterior angles of the coxal plates do not protrude freely.

(2) The distinct proximal spine in the middle line on the pleotelson is missing, and
the lateral keels are not so marked as in the main species and do not terminate in spines.

(3) The pleura of the third abdominal segment reach as far back as those of the second
abdominal segment.

Of these three, number (2) is the only one to which any real importance can be attached.

In the female of the typical form the spines are perhaps not so pronounced as in the
male, and the pleura of the second and third abdominal segments are much more nearly
the same length ; those of the third segment do not reach quite as far back as those of
the second. Since the differences between the species and its variety are so slight, I have
no hesitation in including the varietal name mistrogeorgiensis'mthe synonymy of S. glacial is.

Nordenstam gives a figure of the maxiUiped of the variety (p. 68, text-fig. 16) and
states that he failed to find any suture between "the coxopodite and the proximal epipo-
dite ". I very much doubt the correctness of this statement since in typical S. glacialis,
and in fact in all the known species of Serolis, a suture is present in this position.

8. Serolis septemcarinata, Miers.

S. quadricarinata. White, 1847, p. 106.

5. septemcarinata, Miers, 1875, p. 116; Miers, 1879, p. 206, pi. xi, fig. 3; Studer, 1884, p. 8; Beddard,
1884 b, p. 47, pi. ii, fig. 14, pi. viii, figs. 3-5 ; Pfeffer, 1887, p. 63, pi. ii, figs. 5, 6, pi. iii, figs. 1-26,
pi. iv, fig. 6; Collinge, 1918, p. 74, pis. iii, iv, figs. 1-13; TaUersall, 1921, pp. 227-8; Monod,
1931, p. 26; Nordenstam, 1933, pp. 61-3, text-figs.

S. ovalis, Studer, 1879, p. 24, pi. iii, figs. 8-10.

Occurrence. St. 32: South Georgia, 91-225 m.; 2 ?? (b.).

St. 39: South Georgia, 179-235 m.; 2 ?§ (b.), 5 immature specimens.

St. 45 : South Georgia, 238-270 m.; 5 ?? (b.), i <?•

St. 456: Bouvet Island, 40-45 m.; 16 ?? (b.), 14 ,S<S, a number of immature. Cumberland

Bay, from kelp root washed up on the beach. 2 ?$ (b.).

St. MS 71: East Cumberland Bay, 110-60 m.; i ? (b.).

Diagnostic characters. Beddard (p. 47) points out that in this species the males
differ from the females only in their greater size. In this collection the largest female is
13-5 mm. in length and 10-5 mm. in greatest breadth; it has a greater length than
Beddard's largest male specimen which measured 13 mm. in length and 11 mm. in

DESCRIPTION OF SPECIES 295

greatest breadth. From these facts it is clear that there is Uttle difference in the size of
the sexes, but the male is slightly broader in proportion to its length than the female.

The body is broadly ovate with a sculptured surface of minute anastomosing ridges ;
the surface of the head and the tergal portion of the segments is covered with minute
scattered pits. The head is convex with a crescentic ridge in front of and between the
eyes and a second ridge extending across the head close to its anterior margin: a small
median rostrum is present. The eyes are placed in a postero-lateral position and are
rather small, reniform in shape, and contain black pigment.

The thoracic somites are slightly convex and their coxal plates are compact ; those of
the third, fourth, and fifth are separated by sutures from their respective somites. The
coxal plates of the seventh somite do not extend as far as the distal extremities of the
pleural plates of the second abdominal segment ; each pleural plate of the third segment
extends slightly beyond that of the second to a point just beyond the distal end of the
protopodite of the uropod.

The terminal segment ends in a trifid extremity; its dorsal surface is furnished
with a median longitudinal carina, extending to its tip, and on either side of this are two
lateral carinae which do not extend throughout the length of the segment but end in
freely projecting spines. Close to the lateral margin of the segment, on each side, is
another small carina which extends to a point just anterior to the base of the uropod.

Remarks. There is little to add to the already existing descriptions of this species
except to point out that the last thoracic appendage of the aduh male, as figured by
Pfeffer {loc. cit., pi. iii, fig. 21), does not show the presence of numerous fine hairs along
the inner margins of the propodus and carpus and on the distal half of the merus. The
presence of these hairs was noted by Beddard (1884 h, p. 49) and they are undoubtedly
present.

ColHnge (1918, p. 75, pi. iv, figs. 7, 9), in his description of this species, states that
there is only one variety of sensory spine present on the inner margin of the propodus
of the second thoracic appendage: this fact was noted as incorrect by Tattersall (1921,
p. 228), who confirmed Pfeffer's observations that two types were present. My own
observations agree with those of the two latter workers.

Distribution. The original material of Miers was collected off the Crozet Islands,
Studer's and Collinge's specimens came from Kerguelen, and Tattersall's from Cumber-
land Bay, South Georgia. It is from this last locality that the specimens in the present
collection were obtained.

9. Serolis kempi, n.sp. (Figs. 5, 6).

Occurrence. St. 388: 56° 19^' S, 67° 09!' W, 121 m.; 3 <?(?, 6$? (b.), 7 immature.

St. WS 86: 53° 53' 30" S, 60° 34' 30" W, 151-14? m-; 2 ^c?. 2 ?? (b-). 2 ?? (non-b.) and 3
immature specimens.

St. WS 244: 52° 00' S, 62° 40' W, 253-247 m.; 6 3<S, 13 9? (b-). 6 ?? (non-b.), and n im-
mature specimens.

St. WS 245 : 52° 36' S, 63° 40' W, 304-290 m.; i S, 5 ?? (b-). 33 immature specimens.
St. WS 795: 157-161 m.; I ? (b.).
St. WS 818: 52° 31' 15" S, 63° 25' W, 272-278 m.; I V (b.), 7 immature specimens.

6-2

296

DISCOVERY REPORTS

Fig. 5. Serolis kempi, n.sp., ^: x 6.

Description. This collection contains adult males and females, the latter in both the
breeding and non-breeding phase. An adult male measures 15 mm. in length and
14 mm. in breadth, and an adult female 15 mm. in length and 13 mm. in breadth.

The body (Fig. 5) is broadly oval,
slightly longer than broad and slightly
keeled. The head is about one and a half
times as wide as long, shield-shaped in
outline, with a very small rostral process
between the bases of the antennules:
behind this process a well-marked trans-
verse ridge extends laterally to the sides
of the cephalosome. The portion of the
head between the eyes is convex and
divided into two areas by a slight groove ;
an anterior one bearing two rounded
prominences and a posterior larger one,
divided into three oval prominences, the
central of which is smaller than the two
lateral : this area extends for some distance
behind the eyes.

The eyes are large, about half the length of the head, reniform in shape and containing
black pigment.

The second thoracic somite is fused with the head as in all the species of Serolis. The
lateral portion bears two transverse ridges on each side, a small anterior one which is
continuous with the anterior one already described on the head, and a second which
extends laterally from near the posterior angle of the eye, a point about half-way between
the anterior and the posterior margins of the cephalosome. Each of the third to the
seventh thoracic somites is produced into a short but distinct median dorsal spiniform
process ; that of the fifth somite is considerably larger and longer than those of the other
somites and projects backwards to about the middle of the sixth somite. Low eminences
are present, one on either side of each of the thoracic somites, just within the junction of
the tergum with its coxal plate. The coxal plates are well developed and are sparsely
fringed with short hairs, those of the first three free somites are marked off by distinct
sutures. The coxal plates of the seventh thoracic somite are produced backwards to a
point a short distance behind the level of the bases of the uropoda.

The abdomen (Fig. 5) consists of three free somites and a terminal one, and each of
the three somites has a slight median dorsal process corresponding with those of the
thoracic somites. The pleural plates of the second segment are well produced backwards
beyond the coxal plates of the seventh thoracic somite to about the level of the basal
joint of the uropod; those of the third segment are longer and reach beyond the basal
joint of the uropod. The terminal segment, which is broader than long, has a well-
developed spine in the anterior median dorsal line ; this is followed by a slight median

DESCRIPTION OF SPECIES

297

keel which terminates in a second spine at some distance from the pointed and some-
what upturned posterior extremity of the segment. On each side of this median keel
there are two lateral oblique ridges, each terminating in a spine some distance from the
lateral margins of the segment ; these latter are slightly turned downwards.

The antennule and the antenna are rather long, the latter being longer than the former
and extending back to about the posterior margin of the second free thoracic somite.
The first and second joints of the peduncle of the antennule are sub-equal, but the first
is somewhat bow-shaped ; the third joint is twice as long as the second and six times the
length of the fourth. The flagellum consists of twenty-three joints in the male, but only
seventeen in the female ; the terminal joint is very small and bears a group of four stout
hairs at its tip : an olfactory seta together with several simple setae are present at the
distal end of each joint. The first joint of the peduncle of the antenna is short, the second
and third are sub-equal, the fourth nearly twice as long as the second, and the fifth
slightly longer than the fourth. Tufts of long setae are present on the outer border of
the three distal joints, whilst a few shorter ones are scattered along their inner margins.
The flagellum consists of seventeen joints, the terminal one of which is small and bears
four delicate setae.

The mandibles are of the usual type : the form of the cutting edge of both the left and
right mandible is shown in the figure (Fig. 6 a). The outer lobe of the maxillula bears,

Fig. 6. Serolis kempt, n.sp.
a, cutting edges of mandible: x 24. b, second thoracic appendage of (J: x 24. c, third thoracic appendage
of adult (J: x 24. d, fourth thoracic appendage: x 24. e, eighth thoracic appendage: x 24. /, fourth
pleopod: x 15. g, uropod: x 30.

298 DISCOVERY REPORTS

on its truncated distal extremity, ten simple spines and one pectinate seta, the latter
being innermost in position. The inner lobe is slender and curved with a slightly
broadened distal extremity on which may be found a single, very short seta. The fixed
lobe of the maxilla has its rounded distal end fringed with about twenty pectinate setae;
the distal end of the outer articulated lobe bears two, and that of the inner, four similar
setae.

The basipodite of the maxilliped is not separated by a suture from its lamella;
scattered setae are present on its outer surface. The inner margin of the basipodite is
fringed with fine hairs and groups of these are also found on the outer margins of the
lamella and the epipodite. A very minute extra joint is present near the outer distal end
of the third joint of the palp and bears a number of setae on its truncate extremity
(cf. S. polita, Pfefi"er, S. pageiislecheri, Pfefi^er, and S. platygaster, n.sp.).

The second, third and fourth thoracic appendages are shown in Figs. 6 b-d. The
second appendage, as usual, is of stout build, with the propodus greatly expanded,
having its outer distal angle somewhat pointed and its inner margin armed with a row
of about thirty-eight broadly oval processes ahernating with peculiar stout spines
(Fig. 6 b). The distal edge of the carpus is crenulate and bears a number of simple setae
as well as two stout spines, the upper halves of which are closely pectinate. The third
thoracic appendage of the male (Fig. 6 c) is much more slender than the second. The
propodus is broad at the base but tapers towards its distal end; it bears, on its inner
margin, a double row of six spines of the usual type. The dactylus is longer than the
propodus and ends distally in two strong spines ; a few scattered setae occur along its
outer margin, whilst a number of very short spines are also present along its inner
margin. The inner margins of the carpus and merus are fringed with a dense mass of
fine hairs scattered amongst which are a few long setae ; these hairs extend for a short
distance along the inner margin of the ischium. A further sexual diflFerence in this
species occurs in the last thoracic appendage of the adult male (Fig. 6 e), the inner
margins of the merus, carpus and propodus are fringed with long plumose setae, whilst
a group of three or four is also present at the outer distal extremity of the propodus.
A similar sexual character has been described in S. convexa, Cunningham, S. gmidi-
chaudii, Aud. et Edw., and S. bouvieri, Richardson.

Each of the three sternal plates of the abdominal segments in both sexes has the
median posterior border slightly produced into a spine, increasing in length from the
first to the third. The protopodite of each of the first three pairs of pleopods is not
triangular in shape and does not bear any plumose setae. The exopod of the fourth
pleopod (Fig. 6/) is divided by a transverse suture at right angles to its long axis; the
endopod is bifid at its distal end, and the inner branch is continued for some distance
beyond the outer. The uropoda (Fig. 6^) extend almost to the tip of the terminal seg-
ment ; the basal joint of each is triangular with its inner distal angle prolonged into an
acute spiniform process ; its outer margin bears five long plumose setae. The exopod is
shorter than the endopod and is fringed with plumose setae ; the rounded distal extremity
of the endopod bears a few short hairs.

DESCRIPTION OF SPECIES 299

Distribution. From 52° to 53° 53' S, and 60° to 63° W, between the Falkland
Islands and South America.

This species appears to be most closely related to S. septemcarinata, Miers, and
S. glacialis, Tattersall. A comparison of these two species is included in the description
of the latter by Tattersall (1921, pp. 229-30) and certain points of difference are
noted.

I have compared S. kempt with specimens of these two species in the present collec-
tion and have noted the following points of difference :

(i) The body of S. kempi is broader in proportion to its length than in either of the
other two species: it is also less flattened. S. glacialis is more flattened, less compact
than S. septemcarinata and more transparent than either of the other two species.

(2) The eyes of S. kempi and S. glacialis are larger than in 5. septemcarinata.

(3) In S. kempt and S. glacialis each of the thoracic and first three somites of the
abdomen has the median dorsal posterior border produced into a spine-like process,
which is absent in S. septemcarinata. The spine of the third free thoracic somite in
S. kempi is much larger than any of the others and is also larger than the corresponding
one in S. glacialis, whilst those of the abdomen are proportionately larger in S. glacialis
than those of S. kempi.

(4) In S. kempi the coxal plates of the seventh thoracic segment do not extend back-
wards as far as the pleural plates of the second abdominal segment, and these in turn
are shorter than those of the third abdominal segment. In S. glacialis the coxal plates
of the seventh thoracic somite extend backwards as far as the pleural plates of the third
abdominal segment and not quite as far as those of the second. In S. septemcarinata
the coxal plates of the seventh thoracic somite are shorter than the pleural plates of the
third abdominal segment, which in turn are slightly longer than or are equal to those of
the second.

(5) In S. kempi the last segment of the abdomen bears anteriorly a prominent median
dorsal spine ; this is also present in 5. glacialis, but is absent from S. septemcarinata.
The median dorsal carina posterior to this spine extends to the tip of the segment in
S. glacialis, but ends in a short spine at some distance from the tip in S. kempi. There
are five dorsal carinae on this segment in both S. kempi and S. glacialis and seven on
that of S. septemcarinata. The posterior extremity of the segment is keeled and upturned
in S. kempi; it is trifid in S. septemcarinata, and is pointed in S. glacialis.

(6) There are certain sexual differences which are characteristic of these species. In
S. kempi and S. glacialis the number of joints in the flagellum of the antennule is greater
in the male than in the female. Hairs are present on the merus, carpus and propodus
of the last thoracic appendages of the adult male of S. kempi and S. septemcarinata ;
these hairs are plumose in the former and simple in the latter species and are wanting in
S. glacialis. A further sexual character peculiar to the aduh male of S. septemcarinata
is the development, on the ventral surface of the second thoracic somite, of a row of
ridge-like tubercles, arranged in a semicircle with its convexity directed forwards and
outwards, situated one on either side just external to the articulation of the appendage.

300 DISCOVERY REPORTS

lo. Serolis polita, Pfeffer.

S.poIiia,P{efleT, 1887, p. 8i, pi. ii.figs. 3, 4, pi. iv,fig.4; Richardson, 1906 a, p. 7; Richardson, 1911,
pp. 396; Tattersall, 1921, p. 231, pi. vii, fig. 6; Monod, 1931, pp. 26-7; Nordenstam, 1933,
pp. 63-5, text-figs.

Occurrence. St. 371: South Sandwich Islands, 99-161 m.; 2 immature.

St. WS 25: South Georgia, 18-27 "^■' ' o and 2 immature specimens.

St. MS 65: East Cumberland Bay, 78 m.; i immature specimen.

St. MS 66: East Cumberland Bay, 18 m.; 2 ?? (b.), 2 $? (non-b.), 2 3o, and 2 immature.

St. MS 67: East Cumberland Bay, 38 m.; i immature specimen.

Diagnostic characters. The largest specimens, both male and female, measure
1^-5 mm. in length; the male, as is usually the case, is broader than the female and
measures 15 mm. in breadth; the breadth of the female is 13-5 mm. The body is broadly
ovate and slightly keeled.

The head is broadest anteriorly ; its margin is excavated for the reception of the an-
tennules, between which it is produced forward as a very small rostrum. On either side
of the head a transverse ridge extends from the outer side of the excavation for the
antennule to the lateral margin of the cephalosome. The area of the head behind the
rostrum is raised, and a short transverse ridge extends across between the anterior
margins of the eyes ; immediately behind this is a shallow depression. The area between
the eyes is convex and divided into three areas, the two lateral of which appear as some-
what obliquely placed oval prominences. The central area is not raised as much as the
lateral ones, neither does its straight posterior margin extend back as far : the posterior
margins of these prominences are better defined than the anterior, which merge into the
general surface of the head. The lateral portion of the second thoracic segment is not
divided by a transverse ridge.

Eyes large, reniform in shape, containing black pigment.

The third and fourth thoracic somites are sub-equal, the fifth is slightly longer ; the
sixth and seventh are together equal in length to the fifth. In the mid-dorsal line each
somite is produced into a small spiniform process. The coxal plates are well developed
and those of the third, fourth and fifth somites are marked off by distinct sutures. The
coxal plates of the seventh somite are only slightly more produced than those of the
other somites and do not extend nearly as far backward as do the pleural plates of the
second and third abdominal segments.

Each of the three free abdominal segments bears a short median dorsal spiniform
process. The pleural plates of the second and third segments are broad ; those of the
third segment are produced backward to the level of the articulation of the uropoda,
those of the second to a level a little over half-way between the tip of the coxal plate of
the seventh thoracic somite and that of the pleural plate of the third abdominal segment.

The terminal segment is broader than long, with a well-developed spine in the anterior
median dorsal line, followed by a median keel which extends to the extremity. On
either side of this keel are two lateral oblique keels, the outer one of which is the longer ;
each terminates posteriorly in a small spine at some distance from the lateral margins,
which are somewhat flattened. The posterior extremity of the segment is rounded. The

DESCRIPTION OF SPECIES 301

uropods extend almost to the tip of the segment ; both the endopod and exopod are
elongate oval in shape and fringed with long plumose setae, the latter a little shorter
than the former.

Remarks. The mouth-parts of this species are quite typical. The maxilliped has the
basipodite separated by a suture from its lamella. Previous workers have failed to notice
the presence of a very minute extra joint to the palp of this appendage ; it lies in a slight
groove on the outer side of the third joint, and does not extend as far as its tip, and it
carries on its truncate extremity a group of three or four long setae. A similar joint is
found in a corresponding position on the maxillipeds of S. kempt, n.sp., as well as in
some less closely allied species ; in S. septemcarinata, Miers, its position is indicated by
a notch at the distal extremity of the third joint which separates off, on the outer side of
the joint, a group of three or four long setae.

The types of spines present on the propodus of the second thoracic appendage are
very similar to those figured by S. beddardi, Caiman (see Fig. 3 b). Those on the pro-
podus of the third thoracic appendage of the male are unlike the typical variety. Each
consists of a broad proximal portion, the central axis of which extends as an elongated
slender projection surrounded at its base by a dense mass of short hairs ; the central area
between the rows of spines is toothed, and as Tattersall (1921, p. 228) pointed out, the
inner margins of the merus and carpus of this limb are densely clothed with long
plumose setae. The propodus of the last thoracic appendage in the male is much broader
than is usual and bears a series of stout spines each of which carries a double row of short
pectinations towards its distal extremity; the merus and carpus are also broader in
proportion to their lengths than are the corresponding joints of the other limbs.

Distribution. Pfeffer's and Tattersall's specimens came from South Georgia, as did
the present ones ; Richardson has recorded it from Booth- Wandal Island and from the
South Sandwich group.

II. Serolis elliptica, n.sp. (Figs. 7, 8).

Occurrence. St. 56: East Falkland Island, io|-i6 m.; i immature specimen.
St. WS 243: 52° 00' S, 62° 40' W, 253-248 m.; 3 9? (b.).

Description. The largest specimen (Fig. 7) measures 6 mm. in length and 4 mm.
in greatest breadth. The body, together with the broad almost terminal uropods, is oval
in shape, with a convex rounded dorsal surface covered with microscopic rounded scales,
arranged in alternating rows and overlapping like tiles on a roof. The head is consider-
ably broader than long, equally broad at the anterior margin and at the level of the eyes,
but narrower between these two regions. There is no rostrum, but a slight concavity
exists on either side of the middle line for the reception of the base of the antennules,
behind which is a low transverse ridge which extends across the head and on to the
lateral portions of the cephalosome. The eyes, which are small, oval in shape, and contain
black pigment, are situated far apart, on slightly raised areas in the postero-lateral
angles of the head.

The second thoracic somite is fused with the head, but this fusion may not be quite

302 DISCOVERY REPORTS

complete anteriorly (Fig. 7). The remaining thoracic somites are very broad, and the

coxal plates correspondingly short: those of the third, fourth and fifth somites are

separated from them by sutures. The third,

sixth, and seventh somites are sub-equal, the

fourth and fifth slightly longer. The coxal

plates of the seventh somite are not much

more produced than those of the sixth, and

extend backwards almost to the level of the

base of the uropods.

The first three abdominal segments are
sub-equal, not much narrower than the last
thoracic somite. The pleural plates of the
second and third segments are broad, and
produced backwards beyond the coxal plates
of the seventh thoracic somite ; those of the
second extend almost to the posterior end
of the terminal segment, those of the third
beyond it, reaching nearly to the level of the
outer distal angle of the protopodite of the
uropod. The terminal segment is about as
broad as long. It is almost circular in out-
line, but with the postero-lateral regions
slightly excavated for the reception of the Fig- 7- Serolis elliptka, n.sp.,$: x 22.

uropods, so that the posterior extremity appears to be broadly rounded. A median dorsal
carina is present and the outer margins of the coxal and pleural plates are fringed with
delicate plumose setae.

The antennule is short, about one-half the length of the antenna. The first peduncular
joint is small ; the second is twice as long as the first, broad, with rounded distal end ;
the third joint is the longest, but is narrower than the second; the fourth is a little
shorter than the third. The flagellum consists of seven joints of which the terminal and
penultimate are short, together equal to the fifth. A sensory seta is present at the tip of
the terminal joint and at the anterior distal angle of the fourth, fifth and sixth joints.

The antenna consists of a peduncle of five and a flagellum of nine joints. The first
peduncular joint is short, the second and third are a little longer: the inner margin of the
second is fringed with short hairs and at its inner distal angle is a stout, densely plumose
seta. The fourth and fifth joints are longer and sub-equal and their anterior margins
bear groups of long setae between which is a fringe of shorter ones.

The mouth-parts are of the type usual in members of this genus. The primary cutting
edge of the left mandible is rounded distally and sHghtly produced a Uttle to one side
of the middle ; its secondary edge is in the form of two spines, one of which is slender,
while the other is large and flattened distally to form a finely toothed blade. The primary
edge of the right mandible is also rounded, with a single notch towards one side; the

DESCRIPTION OF SPECIES

303

secondary edge again consists of two spines, a simple one, and a broad one which is
flattened and notched at the distal extremity, but not as broad as the corresponding one
on the left mandible.

The outer lobe of the maxillula bears eight strong spines and a single more deUcate
inner one on its truncate distal extremity. The maxilla is of the ordinary type, only its
fixed lobe is not much broader than the articulating lobes : two long pectinate setae are
present on each of the articulating lobes, and six on the fixed lobe.

The basipodite of the maxilliped is fused with the lamella as far as the articulation of
the endopod : the lamella is broader than long, and fringed with short setae. The first
joint of the endopod is short; the second is almost parallel-sided, expanding slightly at
its distal end ; the third is also parallel-sided, with a rounded tip bearing six setae. Setae
are also present along the distal half of the inner margin of the second joint and two
strong spines are present on the broad distal extremity of the endite.

Fig. 8. Serolis elliplka, n.sp.
a, second thoracic appendage: x 70. b, fourth thoracic appendage of ?: x 85. c, uropod: x 85.

The form of the second and fourth thoracic appendages is shown in Figs. 8 a, b.
The second is stoutly built, with the propodus greatly expanded and with its inner
margin armed with a row of about eighteen stout, modified spines, alternating with a
peculiar toothed and broader variety (Fig. 8 a) : the distal edge of the carpus is crenulate
and bears two stout spines similar to, though smaller than, the longer variety of the pro-
podus. The remaining thoracic appendages (Fig. 8 b) are all of approximately the same
size : the last pair in this species is not smaller than the rest as it is in all other species of
Serolis. Long plumose setae are present on the outer margins of the ischium, merus,

7-2

304 DISCOVERY REPORTS

carpus and propodus, whilst shorter pectinate setae are present at the distal ends of the
carpus and propodus.

The protopodite of the first three pairs of pleopods is triangular with the produced
angle bearing plumose setae — three on that of the first and two on each of the second and
third. The suture of the exopod of the fourth pleopod is oblique, and its outer margin
is sparsely fringed with setae ; the endopod is rather small and broadly rounded distally.

The uropoda (Fig. 8 c) are short and broad, and meet beyond the distal end of the
terminal segment. The protopodite has its inner angle produced into a broad triangular
lobe, at the apex of which is a single plumose seta ; the outer angle is also sHghtly pro-
duced and bears four delicate plumose setae on its outer margin. The endopod is not
much longer than broad, about twice as wide as the exopod, with a short inner margin
and with the distal portion broadly triangular in shape with rounded angles. The
exopod, which is considerably smaller, is oval in shape, with its distal extremity rather
pointed. The outer margin of the exopod, and the rounded distal end as well as the inner
margin of the endopod, are fringed with delicate plumose setae.

12. Serolis exigua, Nordenstam (Figs. 2 a, 9, 10).

S. exigua, Nordenstam, 1933, pp. 70-5, pi. i, fig. 2, text-figs.

Occurrence. St. 51: East Falkland Island, 105-115111.; 9 t?^, 6 ?? (b.). St. WS80: 50° 57' S,
63° 37' 30" W, 152-156 m.; 2??(b.). St. WS210: 50° 17' S, 60° o6'W, 161 m.; 2$$ (b.), a number of
immature. St. WS 212: 49° 22' S, 60° 10' W, 242-249 m.; i ? (b.). St. WS 214: 48° 25' S, 60° 40' W,
208-219 m.; a large number of both ^ and ?. St. WS 215: 47° 37' S, 60° 50' W, 219-146 m.; 7 $$ (b.),

1 immature. St. WS 216: 47° 37' S, 60° 50' W, 219-133 m.; 53 specimens, ^ and ?. St. WS 219: 47° 06' S,
62° 12' W, ii6-ii4m.; 4(J<J, i2??(b.). St. WS 220: 47° 56' S, 62° 38' W, 108-104 m.; 15 <JtJ, 11 ?? (b.),
a number of immature. St. WS221: 48° 23' S, 65° 10' W, 76-91 m.; 2 (JcJ. St. WS222: 48° 23' S,
65°oo'W, ioo-io6m.; i ? (b.). St. WS 227: 5i°o8'S, 56° 50' W, 295m.; 3??(b.); St. WS228:
50° 51' S, 56° 58' W, 229-236 m.; I $ (b.), 3 immature. St. WS 229: 50° 35' S, 57° 20' W, 210-271 m.;

2 ?? (b.). St. WS231: 50^10'S, 58°42'W, 167-159 m.; i ? (b.). St. WS 233: 49° 25' S, 59° 45' W,
185-175 m.; i?(b.). St. WS 234: 48" 52' S, 60° 25' W, 195-207 m.; 3?? (b.), i <?. St. WS 236: 46° 55' S,
60° 40' W, 273-300 m. ; a number of both <J and ? specimens. St. WS 237 : 46° 00' S, 60° 05' W, 150-256 m. ;
a number of both S and $ specimens. St. WS 243: 51° 06' S, 64° 30' W, 144-141 m.; 6 (JJ, 15 $$ (b.).
St. WS 246: 52° 25' S, 61° 00' W, 267-208 m.; 10 ?? (b.), 4 3<S. St. WS 766: 45° 13' S, 59° 56' 30" W,
545 m.; a number of both (J and $ specimens. St. WS 773: 47° 28' S, 60° 51' W, 291-296 m.;
5 ?? (b.), I (J, several immature. St. WS 781 : 50° 30' S, 58° 50' W, 148 m.; i ? (b.). St. WS 782:
50° 29' 15" S, 58° 23' 45" W, 141 m.; II (JcJ, II ?? (b.), 5 immature. St. WS 786: 49° 07' S, 63° 55' W,
i33-ii9m.; 1$ (b.), 7 immature. St. WS787: 48°44'S, 65° 24' 30" W, 106-iiom.; i $ (b.). St.
WS795: 157-161 m.; 2<S<S,i'^ (b.), 10 immature. St. WS 796: 47° 49' 37" S, 63° 42' 30" W, 106-113 m.;
icJ, i$(b.). St. WS 801: 48^26' 15" S, 61° 28' W, 165 m.; I cJ, i?(b.). St. WS 802: A, 50° 45' 45" S,
6i°22'W, 128-132 m.; icJ. St. WS 804: 50° 21' 15" S, 62° 53' W, 143-150 m.; 1$ (b.), i cj. St. WS805:
50° 10' 15" S, 63° 29' W, 148 m.; 4 33, I ?, 3 immature. St. WS 806: 50° 03' 30" S, 64° 21' W, 129-
122 m.; a number of both (J and ? specimens. St. WS 808: 49° 28' 15" S, 65° 42' W, 109-107 m.; i ? (b.).
St. WS813: 51° 35' 15" 8,67° 16' 15" W, io6m.;24o^cJ,4?$(b.), 9immature. St. WS 814: 51° 45' 15" S,
66°4o' W, 111-118 m.; 5 <J<J, 4 ^? (b.), 7 immature. St. WS815: 51° 51' 45" S, 65° 44' W, 132-162 m.;
4cJ(J, 2 immature. St. WS 816: 52° 09' 45" S, 64° 56' W, 150 m.; 13, x ? (b.). St. WS 818: 52°3i' 15" S,
63° 25' W, 272-278 m. ; 6 cJ<J, I $ (b.). St. WS 821 : 52° 55' 45" S, 60° 55' W, 461-468 m. ; a number of both
(J and ? specimens. St. WS 825 : 50° 50' S, 57° 15' 15" W, 135-144 m.; 4?? (b.). St. WS 837 : 52° 49' 15" S,
66°28'W, 98-102 m.; 5 9? (b.), 2c?c?, a number of immature. St. WS864: 49° 33' 30" S, 64° 16' W,
128-126 m.; 2 63. St. WS 866: 50° 37' 45" S, 64° 5' W, 137-144 m.; 3 cJ<J, i ? (b.).

DESCRIPTION OF SPECIES

305

Description. The body (Fig. 9) is broadly oval with a median keel ; the male, the
largest specimen of which measures
7-5 mm. in length and 6-25 mm. in
greatest breadth, is slightly broader in
proportion to its length than is the
female: the latter measures 7 mm. in
length and 5-5 mm. in breadth. The
colour and markings of the body appear
to be very variable.

The head is about twice as broad as
long, shield-shaped in outline, with a
narrow transverse ridge close to its
anterior margin which extends laterally
to the sides of the cephalosome. The
portion of the head between the eyes is
convex, with a strong median dorsal

Fig. 9.

Serolis exigua, Nordenstam, (J : x 12.

keel which is continued backwards as a spiniform process to about the middle of the
third (first free) thoracic somite. The eyes are large, about half the length of the head,
reniform in shape and containing black pigment.

Each of the third to the seventh thoracic somites has its posterior margin produced
into a median dorsal spiniform process ; those of the third, fourth and fifth somites are
larger than those of the following somites, but are smaller than the corresponding pro-
cess of the head: these processes are not so well developed in the female. The coxal
plates of the third to the seventh thoracic somites are well developed, those of the third,
fourth and fifth are separated from their respective somites by sutures. In the male the
coxal plates of the seventh thoracic somite do not extend backwards quite as far as the
pleural plates of the second abdominal segment; in the female they are considerably
shorter.

The abdomen (Fig. 9) consists of three free segments and a large terminal one ; each
of the three free segments is slightly keeled and has a short median dorsal spine corre-
sponding with those of the thoracic somites. The pleural plates of the third abdominal
segment extend backwards beyond those of the second to a point level with the base of
the uropod ; the outer margins of both the coxal and pleural plates are sparsely fringed
with hairs and short hairs are also present on their posterior margins.

The terminal segment is roughly five-sided, with the angles rounded. A median
dorsal keel extends to the posterior extremity of the segment, but is more pronounced in
the anterior half. On either side of this keel, lying midway between it and the lateral
margin of the segment, is a lateral carina which extends nearly to the posterior margin
of the segment ; a secondary carina arises half-way along each of the lateral carinae and
passes forwards and outwards to the lateral margin of the segment.

The antennule (Fig. 9) is very deUcate, and consists of a peduncle of four joints and a
flagellumof eight ; the first peduncular joint is broad and geniculate, the second sub-equal

3o6

DISCOVERY REPORTS

to it, the third about twice as long as the second, but narrower, the fourth about a third
the length of the third. The flagellum is slightly shorter than the peduncle ; the first six
joints are sub-equal, shorter than the fourth peduncular joint; the seventh flagellar
joint is very short and about half the width of the sixth. The terminal joint is very narrow,
but about equal in length to the sixth, with four delicate setae at its distal extremity.
An olfactory seta is present at the distal end of each of the joints of the flagellum except
the first two and the terminal one. The antenna is nearly twice as long as the antennule
and consists of a peduncle of five and a flagellum of ten joints. The first peduncular joint
is very short. The second and third longer and sub-equal, the fourth almost twice, and
the fifth two and a half times as long as the third. The anterior borders of the third,
fourth and fifth joints bear a series of tufts of long hairs. The flagellum is about equal in
length to the fifth joint of the peduncle.

Fig. 10. Serolis exigua, Nordenstam.
a, cutting edges of mandible: x 55. b, second thoracic appendage of adult cJ: x 55. c, third thoracic
appendage of adult ,^: x 55. d, fourth thoracic appendage: x 55.

The mandibles are of the ordinary type ; the cutting edges are comparatively simple
and of the form shown in Fig. 10 a.

The outer lobe of the maxillula bears eight spines and one delicate seta on its distal
truncate extremity. The inner lobe is about half the length of the outer, with its distal
end somewhat expanded and bearing a single, short curved seta with a bifid extremity.
There are two pectinate setae on each of the two outer lobes of the maxilla, and eight
pectinate and four simple setae on the rounded end of the fixed lobe.

DESCRIPTION OF SPECIES 3°?

The maxilliped (Fig. 2 a) has the basipodite completely fused with the lamella, the
fusion being continued beyond the attachment of the basal joint of the endopod, so that
the latter appears to spring from the centre of a plate-Uke structure. A similar arrange-
ment occurs in S. carinata, Lockington.

The second, third and fourth thoracic appendages of the male are shown in Figs.
ID b-d. The propodus of the second appendage is broadly oval with its inner margin
armed with a row of about thirty short stout spines : these spines (Fig. lob) are inclined
slightly backwards, and each consists of two branches the posterior one of which is the
longer; a series of rather long setae flank this row on either side. This and S. nototropis,
n.sp., are the only species in which the propodus bears only one variety of modified
spine on its inner margin ; for although Collinge (p. 75) failed to find more than one type
in S. septemcarinata, Miers, the observations of Beddard, Hodgson and Tattersall on this
species show that in reality there are two. Another character peculiar to this species is
the presence, in the adult male, of a dense covering of long delicate hairs on the outer
distal part of the propodus of this appendage. The distal truncate end of the carpus
bears two stout spines, each with a bifid extremity, and a number of simple setae. The
third thoracic appendage of the male (Fig. \o c) is considerably longer though more
delicate than the second. The propodus is only slightly broader than the carpus; it has
eleven spines of the usual type on its inner margin, arranged in three rows^two lateral
ones of four spines each, three of which are placed near the proximal and one near the
distal end of the propodus, and a median row of three, placed in a position intermediate
between the distal and the proximal spines of the lateral rows. The dactylus terminates
in a stout spine, the tip of which is slightly curved outwards. The last pair of thoracic
appendages is shorter and more delicate than the others.

In each of the three sternal plates of the abdominal segments in both sexes the median
posterior border is produced into a spine which increases in size from before backwards.
The protopodite of the first three pairs of pleopods is somewhat triangular, with the
produced angle bearing three plumose setae on the first, and two on each of the second
and third pleopods. The appendix masculina on the second pleopod of the male reaches
about two-thirds of the way towards the apex of the terminal segment. The exopod of
the fourth pleopod has a transverse suture and a fringe of plumose setae along its outer
margin ; the endopod is entire and with a transverse suture.

The uropoda extend nearly to the tip of the terminal segment. The protopodite, as
usual, has its inner distal extremity drawn out into a spine and the endopod is sHghtly
longer than the exopod; both have rounded distal ends, crenate and fringed with
plumose setae.

Distribution. Off East Falkland Island, and between 66° and 58° W and 52° 55' S.

This species appears to be most nearly related to S. carinata, Lockington, to which it
bears a fairly close resemblance. The following points are noted after comparison of the
two species:

(i) Both species are of small size, the adult males measuring not more than 8 mm.
in length ; the shape of the body is broadly oval with a median dorsal keel.

3o8 DISCOVERY REPORTS

(2) Each of the thoracic and first three abdominal segments has the median dorsal
posterior border produced into a spiniform process, which is much larger in 5. exigua
than in S. carinata.

(3) The median posterior border of the head is also produced into a spiniform process
in both species ; but in S. carinata this process is Y&Yy small and does not extend beyond
the head segment, whereas in S. exigiia it is produced backwards to about the middle
of the third (first free) thoracic somite.

(4) In both species the first three free thoracic somites are separated from their coxal
plates by sutures. In 5. carinata the coxal plates of the seventh thoracic somite extend
beyond the ends of the pleural plates, which are very short and hardly extend beyond the
anterior margin of the terminal segment. In S. exigua they are short and do not ex-
tend beyond the pleural plates of the second and third abdominal segments. The pleural
plates of the third abdominal segment in S. exigua reach to the attachment of the
uropod ; those of the second are a little shorter.

(5) The terminal segment of S. exigua is broader than long and roughly five-sided,
with a median dorsal and two lateral carinae, the latter lying midway between the
median carina and the lateral margin of the terminal segment, and not extending to the
posterior end of the segment. The terminal segment of S. carinata is roughly triangular,
longer than broad, with its posterior extremity truncate, and with a median and two less
pronounced lateral carinae which run parallel with the lateral margins of the segment
and meet the central one at the posterior extremity.

(6) The antennules of both species are small, with a flagellum of ten or eight joints;
the antennae are nearly twice as long, with short flagella.

(7) The maxillipeds of both species agree in having the basipodite fused with the
lamella to a point beyond the attachment of the endopod, so that the latter appears to
spring from the centre of a plate-like structure.

(8) The characters of the spines on the propodus of the second thoracic appendage
of the two species differ. In S. exigua only one variety of spine is found, whilst in
S. carinata two types are present, one of which is peculiar to the species. The propodus
in the aduh male of S. exigua is further characterized by the presence of long delicate
hairs on its outer border ; these are not found in S. carinata.

The above description was written before receiving Nordenstam's paper (1933). It
agrees with that given by Nordenstam for S. exigua and his specific name has therefore
been used. Nordenstam's description is based on the female only: in the above de-
scription the male characteristics are also included.

13. Serolis carinata, Lockington (Fig. 11).

S. carinata, Lockington, 1877, pt. i, p. 36; Richardson, 1899, p. 842; Richardson, 1900, p. 224;
Richardson, 1905, p. 321, figs. 353, 354.

I have been able to examine specimens of both sexes of this species through the
generosity of Dr Waldo Schmitt of the United States National Museum, Washington,
to whom I wish to express my gratitude. Neither the description nor the figures given

DESCRIPTION OF SPECIES 309

in Richardson's Isopods of North America (1905) agree with my own observations, and
for this reason I include here a revised account of the species.

Description. The body of the female is broadly ovate with a pronounced median
dorsal carina: the coxal plates are much flattened. The male is slightly broader in
proportion to its length than the female, and measures 8 mm. in length and 7 mm.
in greatest width; the female is only 5-5 mm. long and 4-25 mm. wide.

The head is about as broad as long, with the anterior margin excavated on either side
of a small median rostrum for the reception of the antennules. Just within the anterior
margin, and running parallel with it, is a shallow transverse groove, behind which the
area of the head between the eyes becomes raised. The central portion of this area,
which is separated from the postero-lateral parts by slight grooves, is somewhat tri-
angular in shape, with the apex directed backwards and terminating in a small spiniform
process at the posterior margin of the head. At about the level of the middle of the eyes
this area is raised to form a second small median process. The eyes, which are situated in
postero-lateral positions on the head, are comparatively large, reniform in shape (not
oval as stated by Richardson) and contain black pigment.

The lateral portion of the second thoracic somite, which is fused with the head, is
broad and flattened, without any transverse ridges. The third, fourth and fifth thoracic
somites are sub-equal, with the median keel produced backwards as a short spine; the
sixth and seventh somites are short, their length together equal to less than that of those
preceding them. Coxal plates well developed ; those of the third to fifth somites marked
oif by distinct sutures, and not "perfectly coalesced, with no indication of the place of
fusion" (Richardson, 1905, p. 322). Those of the seventh somite are produced back-
wards, but do not reach as far as the level of the base of the uropoda.

The abdomen is composed of three free somites and a large terminal one. Each of the
free somites is keeled and bears a short median spiniform process. The process on the
third segment is not on the posterior margin, but on a well-marked transverse ridge
which extends across the segment, disappearing laterally on the pleural plates: this
ridge lies a short distance within the posterior margin. The pleural plates of the second
and third abdominal segments do not extend beyond the anterior margin of the terminal
segment ; those of the third segment are slightly the longer. The terminal segment is
large, somewhat triangular in shape. In the female the apex is obtusely rounded, but not
nearly as truncate as shown in Richardson's figure (1905, p. 321, fig. 353). In the male
specimen in my possession the distal extremity is broader than that of the female and is
slightly concave. A median and two lateral carinae are present ; each of the latter runs
parallel, one on either side, with the lateral margin of the segment, but at some distance
within it, meeting the median carina at its posterior extremity.

The sterna of the three free abdominal segments are produced backwards to form
median spines ; in the female that of the first segment is much longer than those of the
other two, extending almost to the tip of the second ; in the male they are all of ap-
proximately equal size.

In the antennule the first peduncular joint is broad, rounded and geniculate; the

3IO DISCOVERY REPORTS

second is about the same length only narrower ; the third and fourth sub-equal, longer
than the second. The flagellum consists of ten joints in the male and five in the female;
the first joint is short, the second more than twice as long as the first, the remaining
segments gradually decreasing in length and breadth to the penultimate one ; the terminal
segment is very small, bearing a sensory and two long simple setae. A single sensory
seta is present at the distal end of each flagellar joint. In the antenna the peduncular
joints are covered with microscopic imbricating scales, those of the fourth and fifth
joints larger than those of the other joints. The first peduncular joint is short, the second
twice as long as the first, both with fine hairs on their anterior margins ; the third is about
equal in length to the second, but not as broad ; the fourth joint is a little more than
twice the length of the third, its anterior margin bearing four groups of delicate setae on
slightly raised areas at points about equidistant one from the other ; the fifth joint is
about one and a half times the length of the fourth, with a similar arrangement of setae.
The flagellum is composed of seven joints, of which the first is the longest and broadest;
the terminal and penultimate joints are very small, together equal to about a third of the
length of the preceding one. Each joint is furnished with a short toothed spine at its
distal posterior angle and a few delicate setae at its anterior angle.

Richardson's figures (1905, p. 321, fig. 354 b, c) of the maxillula and maxilla are quite
inaccurate : she shows only one distal lobe in the former and two in the latter appendage,
with none of the basal joints. Actually these appendages have the normal Isopod struc-
ture (see Figs, i a, c).

The maxillula has eight curved setae on the truncate distal extremity of the outer
lobe ; the inner lobe is delicate, about two-thirds the length of the outer, curved out-
wards towards its distal end and with a single short seta. The maxilla has the usual two
setae on the distal ends of each of the two outer lobes, and seven or eight on the fixed
lobe.

The lamella of the maxilliped (Fig. 11 rt) is not separated from the basipodite by a
suture, the union of the two extending beyond the articulation of the endopod, so that the
latter appears to spring from the centre of a plate-like structure. Such an arrangement
occurs also in S. exigiia, Nordenstam (Fig. 2 a), S. orbiculata, n.sp. and S. nototropis,
n.sp. The endopod is composed of three joints ; the second is long and narrow with its
distal angle somewhat produced, rounded, and bearing six setae; the terminal segment
has six setae on its rounded distal extremity.

The cutting edges of the mandibles are simple, each with its distal extremity notched
to form three large rounded teeth. In the three-jointed palp the terminal joint is rather
short, with setae of the usual type on its posterior margin, and on that of the second
joint.

The second, third and fourth thoracic appendages of the male are shown in Figs.
1 1 h-d. The propodus of the second is greatly expanded, with its inner margin armed
with a row of twenty-three broad, leaf-like, toothed processes (Fig. 1 1 b), alternating
with peculiar spines. The leaf-like processes are not unlike those found in a similar
position in S. elliptica, n.sp. The distal end of the carpus is crenulate, and bears two

DESCRIPTION OF SPECIES

311

Strong modified spines (Fig. 12 b) as well as two or three more delicate ones. The third
appendage of the male (Fig. 1 1 c) is considerably longer and more delicate than the second,
with the propodus broad and bearing on its inner margin a double row of eleven spines
of the type usually found on this limb. The last thoracic appendage is smaller than the
others and is not modified in the male.

The protopodite of the first three pairs of pleopods is long, without a produced angle,
and with the outer margin fringed with delicate hairs. Both exopod and endopod are
fringed with strong plumose setae. The exopod of the fourth pleopod has an obUque

Fig. II. Serolis carinata, Lockington.
a, maxilliped of (J: x 52. b, second thoracic appendage of <?: x 52. c, propodus and dactylus of third
thoracic appendage of adult J: x 52. d, fourth thoracic appendage: x 35.

suture, and the outer margin fringed with plumose setae ; the endopod is rounded with
a transverse suture. The uropoda extend to the tip of the terminal segment, with the
inner distal angle of the protopodite produced and bearing three plumose setae. The
exopod and endopod are sub-equal, with rounded distal extremities bearing plumose
setae which also extend along their outer margins.

Locality. San Diego, California.
14. Serolis convexa, Cunningham (Fig. 12; Plate XIV, fig. 3).

S. convexa, Cunningham, 1871, p. 498, pi. lix, fig. 3; Studer, 1884, p. 9, pi. i, figs, i a, d; Beddard,
1884 b, p. 37, pi. vi, figs. 9-15; Nordenstam, 1933, pp. 77-82, text-figs.
Occurrence. St. 51: East Falkland Island, 115 m.; i ? (b.), 4 immature. St. WS 75 : 51° 01' 30" S,
60° 31' W, 72 m.; 2 cJ<?, 2 immature. St. WS 90: 13 miles N 83° E of Virgins Light, Argentine RepubUc,

8-a

312 DISCOVERY REPORTS

82-81 m.; I 3. St. WS 220: 47° 56' S, 62° 38' W, 108-104 m.; 2 immature. St. WS 222: 48° 23' S,
65° 00' W, 100-106 m.; I ? (b.), 3 immature. St. WS 787: 48° 44' S, 65° 24' 30" W, 106-iiom.; 2 ?? (b.),
a number immature. St. WS 796: 47° 49' 37" S, 63° 42' 30" W, 106-113 m.; i (J, 2 ?? (b.), 3 immature.
St. WS 797: 47° 45' 36" S, 64° 20' W, 115-111 m.; 2 cJc? (young). St. WS 808: 48° 28' 15" S, 65° 42' W,
109-107 m.; I cJ, II ?? (b.), a number immature. St. WS 811 : 51° 24' 30" S, 67° 53' W, 96-98 m.;
I $ (b.). St. WS813: 51° 35' 15" S, 67° 16' 15" W, 106 m.; icJ. St.WS8i4: 51° 45' 15" S, 66° 40' W,
Iii-ii8m.; 1 ? (non-b.). St. WS 815: 51° 51' 45" S, 65° 44' W, 132-162 m.; i immature specimen.

Diagnostic characters. The body (Plate XIV, fig. 3) is pear-shaped owing to the
comparatively long terminal segment, and to the fact that the sixth and seventh thoracic
somites with their coxal plates are narrower than the preceding ones : the colour of the
body is a uniform pale brown. The largest male in this collection measures 31 mm.
in length and 24 mm. in greatest breadth, and the largest female 22 mm. in length
and 17 mm. in breadth.

The head is sHghtly broader than long, broadest anteriorly, with a median rostrum
behind which a transverse ridge extends laterally to the sides of the cephalosome. The
eyes are reniform in shape, small but prominent, situated on raised areas placed rather
near together. The median portion of the head, behind but between the eyes, is slightly
raised and produced posteriorly into a small median spiniform process.

The dorsal surface of the body is convex with a median carina which in the fourth
and fifth thoracic somites and in the three abdominal segments is produced backwards
as a small spiniform process. The coxal plates of each of the first three thoracic somites
are separated from their respective somites by sutures. The sixth and seventh somites
are narrow and together equal to the fifth ; the coxal plates of the seventh do not extend
backwards as far as the middle of the terminal segment. The pleural plates of the second
and third abdominal segments are short, not extending beyond the anterior margin of
the terminal segment ; those of the third are shghtly the longer.

The terminal segment is roughly five-sided with a rather pointed posterior extremity.
On its dorsal surface are a median and two lateral carinae ; the latter are curved and
follow the lateral margin of the segment, but at some distance from it, and each ends in
a short spiniform process just anterior to the point of articulation of the uropod. The
median carina is found in the anterior part of the segment and also towards its posterior
extremity, the middle part being obsolete.

Remarks. Beddard (1884 b, p. 37) describes a patch of plumose setae on the sternum
of each of the last three thoracic somites as a sexual character pecuhar to the males, and
mentions that a small tuft may also be found on some of the others. This patch is
present, however, on each of the sterna of the third to the seventh somites, that of the
fourth measuring in a large specimen as much as 4 mm. in width. The patch is absent
from the sterna of youiig males, apparently appearing along with certain other characters
when the final moult to the adult stage is reached.

The appendages of this species have already been described, but some further ob-
servations are noted here. The mouth-parts are of the typical form, that of the cutting
edges of the mandibles is shown in Fig. 12 a. The outer lobe of the maxillula bears on
its truncate distal extremity nine simple spines and one pectinate seta having a bifid

DESCRIPTION OF SPECIES

313

b.

Fig. 12. Serolis convexa, Cunningham,
a, cutting edges of mandible: x 40. b,
form of spines on propodus of second
thoracic appendage of ?: > 40. c, form of

extremity ; the inner lobe is delicate, about two-thirds the length of the outer, with its
distal half bent outwards, and with its distal end only slightly expanded and bearing a
single short seta. ~~--J)-_ 0-

The maxilla is figured by Beddard (1884 Z), pi. vi,
fig. 14), but the sutures which correspond with those
already described and figured in the introductory part
of this paper are not shown. The fixed lobe is not much
larger than the other two, each of which bears eight or
nine pectinate setae instead of the more usual number
of two.

The form of the spines on the propodus of the second
thoracic appendage of the adult male is shown in
Fig.i2c; according to Beddard (18846, p. 40) only a
single row of short hairs is present on the distal portion
of the longer variety of spines : as he points out, a tuft
of long plumose hairs is present on the carpus of this
limb. In examining the material of this collection, I
have observed for the first time that in this species
there is a sexual difi^erence in the form of the modified
spines on the propodus of the second thoracic appen-
dage. In the female, and also in the immature male, spines on propodus of second thoracic
as in the male, two kinds of spines are present in an appendage of cj: x 40.
alternating row, and those of the shorter variety are very similar in both cases. But
in the immature male and in the female the longer variety (Fig. iz b) consists of a
spine expanding distally into two processes, one of which is a little longer and broader
than the other ; between these the central axis is continued as a third arm to about the
level of the shorter of the two processes, where it ends in a bifid extremity. A similar con-
dition also occurs in S.gaudichaudii, Aud. et Edw., and probably in S. laevis, Richardson.

The distribution of plumose setae in Beddard's figure of the last thoracic appendage
of the male is incorrect ; he figures them on the posterior margins of the ischium, merus
and carpus, but actually they occur on the merus, carpus and propodus. Incidentally,
he refers to this appendage in the text as the "sixth" instead of the last or eighth.

Distribution. The type locality for this species is the north coast of Tierra del Fuego ;
Beddard's material was collected at Port William, Falkland Islands. The material in this
collection comes from off East Falkland Island and the Argentine RepubHc.

[Serolis plana, Dana.

S. plana, Dana, 1853, pt. ii, p. 794; Beddard, 18846, p. 39.

This species has not been recorded since it was described by Dana in 1853. Beddard
(18846, p. 39) points out that it appears to be identical with S. convexa, Cunningham,
except that the eyes are stated by Dana to be conical in shape, whereas in S. convexa,
as in all the species of the genus, they are reniform.

314 DISCOVERY REPORTS

The eyes of S. convexa are raised on prominences which might be described as conical,
and it is possible that Dana was referring to the shape of these rather than to that of the
eye itself — in which case the two species are identical.

Locality. Patagonia, in a few fathoms.]

15. Serolis gaudichaudii, Aud. et Edw. (Plate XIV, fig. 4).

S.gaudichaudii, Audouin and Milne-Edwards, 1841 , p. 5, pis. i, ii ; Beddard, 1884 b, p. 38 ; Nordenstam,
1933. PP- 76-77. text-figs. 2e-g.

Occurrence. St. WS 221 : 48° 23' S, 65° 10' W, 76-91 m.; i ? (b.).

St. WS 742: 38° 22' S, 73° 41' W, 47-35 m.; i <? (young), 3 ??.
St. WS 809: 49° 28' 15" S, 66° 29' W, 107-104 m.; 3 immature.
St. WS 856: 46° 45' S, 64° 11' W, 104 m.; a number of immature specimens.

Diagnostic characters. The largest specimen in the collection (Plate XIV, fig. 4)
measures 23 mm. in length and 17-5 mm. in breadth; it is approximately the same size
as an adult female of S. convexa, Cunningham, to which species it bears a very close
resemblance. The body, however, is more oval in shape, the decrease in width of the
thoracic somites being gradual: that is, the sixth and seventh somites are not con-
siderably narrower than the preceding ones as they are in S. convexa. The dorsal surface
is less convex than in S. convexa and a median keel is hardly discernible.

The head, as in S. convexa, is slightly broader than long with its broadest part anterior ;
a median rostrum is present, behind which is a transverse ridge extending laterally to
the margins of the cephalosome. The eyes are raised on two postero-lateral rounded
prominences, between which the area of the head is sunken ; behind the eyes the median
area is slightly raised. The body colour is brown, darker in the mid-dorsal region, and
dotted all over with black spots. The eyes are small, but slightly larger in proportion to
the head than in S. convexa ; they are reniform in shape, prominent, and placed near
together.

The first three free thoracic somites are sub-equal ; the fourth and fifth are narrow
and together equal to the third. The coxal plates of the first three free somites are
separated from them by distinct sutures ; those of the fifth free somite are broader than
the corresponding ones in S. convexa and extend backwards as far as the middle of the
terminal segment. The pleural plates of the second and third abdominal segments do
not extend backwards as far as the extremities of the coxal plates of the seventh thoracic
soinite ; those of the second are slightly longer than those of the third segment.

The terminal segment is considerably broader and more rounded than that of
S. convexa, and is furnished with a central and two lateral dorsal carinae. The latter are
curved and follow the outline of the segment as far as the bases of the uropoda, where
they each terminate in a very small spiniform process ; these carinae are very much more
prominent in S. convexa. The middle part of the median carina is obsolete in both
species.

Remarks. The head appendages are of the usual type. Most of the joints of the
flagellum of the antennule bear two sensory hairs, one situated near the distal extremity

DESCRIPTION OF SPECIES 315

of the joint, the other near its posterior end; Beddard noted a similar arrangement in
S. convexa. The two outer lobes of the maxilla, like those of S. convexa, are broad and
bear a number of serrated spines (seventeen on the outer and nine on the inner lobe)
instead of the more usual number of two. The basipodite of the maxilliped is separated
by a suture from the lamella ; the distal joint of the palp is very small.

The second thoracic appendage of the male is figured and described by Audouin and
Milne-Edwards, and the types of sensory hairs present on the propodus are very similar
to those of the males of S. convexa and S. laevis ; as in these two species, I have observed
that there is here also a sexual difference in the form of the spines on the propodus, for
in the female the longer type of spine is replaced by one which has its distal end pro-
duced into two processes, one of which is slightly longer than the other. Between these
processes the central axis is continued as a prolongation extending to about the level of
the shorter process (cf. Fig. izb). A group of plumose setae is present on the carpus of
the second thoracic appendage of the male, as in S. convexa and 5. laevis ; long plumose
setae also fringe the inner margin of the ischium, merus, carpus and propodus of the
last thoracic appendage of the male.

The protopodite of each of the first three pairs of pleopods is triangular in shape, and
the projecting angle is furnished with three plumose setae in the first of these appendages
and two in the second and third. The suture of the exopod of the fourth pleopod is
nearly transverse, and the endopod has a rounded distal extremity. The uropoda extend
to near the tip of the terminal segment ; the exopod is about two-thirds the length of the
endopodite, and has a broadly rounded distal extremity bearing a number of long
plumose setae. The distal extremity of the endopod is more pointed and is also fringed
with plumose setae.

Distribution. Shores of South America; the original specimen was found near
Valparaiso.

16. Serolis laevis, Richardson (Plate XIV, fig. 5).

Serolis laevis, Richardson, 1911, pp. 399, 400, text-fig. 2; Nordenstam, 1933, p. 81.

Through the kindness of Prof. Ch. Gravier I have been able to examine the type
specimen of this species from the Museum d'Histoire Naturelle, Paris, and since
Richardson's account consists merely of a brief comparison with the allied species
S. plana, Dana, S. convexa, Cunningham, and S . gaudichaudii, Aud. etEdw., I include a
description. This, however, is necessarily incomplete, as my observations were restricted
to such details as could be seen without the removal of any appendages.

Description. The type specimen (Plate XIV, fig. 5) is an adult male measuring 1 6 mm.
in length and 11-5 mm. in greatest width. The body is ovate in shape, white in colour,
with a smooth surface, and a slight median dorsal keel. The head is broader than
long, broadest anteriorly, with a well-developed rostrum, and with antero-lateral angles
narrow. A slight transverse ridge extends across the head just behind the rostrum ; the
area between the eyes is convex. The eyes are sHghtly raised, reniform in shape, not

3i6 DISCOVERY REPORTS

deeply pigmented, and larger in proportion to the size of the head than in either
S. convexa or S. gaudichaudii. The lateral portions of the cephalosome are broad, the
posterior margin of the segment measuring 1 1 mm.

The coxal plates of the first three free thoracic somites (3rd-5th) are separated by
distinct sutures ; the plates of the seventh somite extend backwards to about the middle
of the terminal segment, but not as far as the bases of the uropoda : they are longer than
the corresponding plates in S. convexa. In the male the sternum of each of the third
to the seventh thoracic somites bears a small patch of plumose hairs exactly similar to
that described for S. convexa.

In each of the abdominal segments the posterior margin is slightly produced back-
wards to form a median dorsal process ; this is largest on the third segment. The pleural
plates of the second and third abdominal segments are short, not extending beyond the
anterior margin of the terminal segment; those of the second segment are slightly the
longer. The terminal segment is ovate in shape, with a very small longitudinal median
dorsal keel which disappears posteriorly : the extremity of the segment is rounded and
not truncate as shown in Richardson's figure and in this specimen the transverse ridge
shown in her figure is absent.

The antennule is about half the length of the antenna, the latter extending to about
the middle of the third free thoracic somite. The flagellum of the antennule consists of
twenty-two joints ; the penultimate joint is very short — only about half the length of the
terminal one although twice its width : the latter bears four long setae at its distal ex-
tremity. The flagellum of the antenna consists of twenty-one joints, of which the two
most distal are short (cf. S. convexa) ; the terminal joint bears a group of five setae at its
distal extremity.

The form of the maxillula, maxilla and mandible could not be made out in this
specimen. The basipodite of the maxilliped is separated by a suture from the lamella,
and the terminal joint of the palp is small, as it is also in S. convexa and S. gaudichaudii.
It is probable that in this species, as in S. convexa and S. gaudichaudii, there is a sexual
diff"erence in the form of the spines on the propodus of the second thoracic appendage,
for in the male the two types of modified spines are almost exactly similar, though
smaller, to those found on the propodus of the males of the above species (see Fig. 14c).
As in the allied species, a tuft of plumose setae occurs on the carpus.

The third thoracic appendage of the male is more delicate than the second, and the
propodus bears on its inner margin two rows of spines, six in each row. Each spine has
a truncate extremity on which are situated a number of short hairs. The last thoracic
appendage has the posterior margin of the propodus, carpus, merus and anterior half
of the ischium fringed with long plumose setae : the remaining appendages are furnished
with setae which are either simple or serrated.

The protopodite of each of the first three pairs of pleopods is triangular in shape, the
produced angle bearing three plumose setae in the first of these appendages and two in
the others. The suture of the fourth pleopod is almost transverse, and the endopod has a
rounded extremity. The uropoda extend to near the tip of the terminal segment ; the

DESCRIPTION OF SPECIES 317

exopod is about two-thirds the length of the endopod, and both are rather narrow and
fringed with plumose setae.

Distribution. South Sandwich Islands.

Since the three species S. convexa, Cunningham, S.gaudichaiidii, Aud. et Edw. and
S. laevis, Richardson, resemble one another rather closely, a short comparison of the
three may prove to be of value.

(i) The shape of the body differs in the three species; that of S. convexa is pear-
shaped, owing largely to the greater length of the terminal segment, and to the fact that
the sixth and seventh thoracic somites, with their coxal plates, are distinctly narrower
than the preceding ones. This decrease in width is also characteristic of S. laevis; but
in this species, which is not as large as the other two, the terminal segment is ovate. In
S. gaudichaudii the terminal segment is oval, with the sixth and seventh thoracic somites
not distinctly narrower than the preceding ones.

(2) Small median dorsal spines are present on the fourth and fifth thoracic somites
and on the first, second and third abdominal segments of S. convexa, whereas in the
other two species they are only present on the abdominal segments.

(3) The terminal segment of S. convexa is roughly five-sided, with three well-
developed dorsal carinae: a median one, interrupted in the middle, and two lateral ones,
one on either side of the segment, each terminating in a sharp spine. The terminal seg-
ment of S. gaudichaudii is rounded, nearly as broad as long ; it bears three dorsal carinae,
a median one interrupted in the middle, and two lateral ones, which are curved and
follow the outline of the segment as far as the bases of the uropoda where each terminates
in a very small spiniform process : these carinae are less pronounced in this species. In
S. laevis the median and lateral carinae of the terminal segment are almost obsolete.

(4) The colour of the three species differs : that of S. convexa is a uniform pale brown,
that of S. gaudichaudii is brown, darker towards the middle line and dotted all over with
black spots, whilst that of S. laevis is white.

(5) The three species resemble one another and differ from the majority of species in
possessing maxillae in which the two articulating lobes are broad ; and each lobe bears
eight or nine pectinate setae instead of the more usual number of two. The distal joint
of the palp of the maxilliped is also smaller than usual.

(6) The two species S. convexa and S. gaudichaudii, and probably also S. laevis, are
characterized by a sexual difference not found in any other species ; the form of the
modified setae which occur on the propodus of the second thoracic appendage differs in
the two sexes (see descriptions of species).

(7) Long plumose setae are present on the carpus of the second thoracic appendage,
and on the posterior margins of the propodus, carpus, merus and ischium of the last
thoracic appendage of the adult male in all three species.

(8) A further sexual character of the adult male of S. convexa and S. laevis is the
presence of a patch of plumose setae on the sternum of each of the third to the seventh
thoracic somites. I have not been able to examine adult male specimens of S. gaudi-
chaudii, and I am unable to say whether or not it also possesses this character.

3i8 DISCOVERY REPORTS

(9) The distribution of S. convexa and S. goiidichaiidii is somewhat similar. The
former occurs off the coast of the Argentine Republic and around the Falkland Islands ;
the latter has been recorded from the shores of South America and off Valparaiso. The
only record of S. laevis is the South Sandwich Islands, a locality farther east than any
of those recorded for the two former species.

Nordenstam, 1933, includes S. laevis as a synonym of S. convexa.

I'j. Serolis gerlachei, Monod.

S.gerlachei, Monod, 1925, p. 299; 1926, pp. 36, 37, text-figs. 35-7.

Only one specimen of this species, an immature male, has so far been recorded, and
I wish to express my gratitude to Dr von Straelen of the Musee Royal d'Histoire
Naturelle de Belgique for allowing me to examine it.

Diagnostic characters. The body is longer than broad, strongly arched and with a
median longitudinal keel; it measures 10 mm. in length and 8 mm. in greatest breadth.
The breadth of the head is about equal to its length, excluding the end of the median
spine. The head is broadest anteriorly ; the antero-lateral angles are produced beyond
the anterior margin, which is excavated on either side of a minute median rostrum for the
reception of the antennules. The antero-lateral angles are marked off by a transverse
ridge which extends across the head immediately behind the base of the rostrum ; the
area behind this ridge and between the anterior ends of the eyes is raised, and is separ-
ated by a slight groove from the raised area between the eyes. The latter area is divided
into two lateral oval prominences and a median area which is produced backwards as
a long acute spine extending to about the middle of the second free somite. The eyes
are reniform and pigmented.

The first three free somites (3rd-5th) are sub-equal in length. The third has its
posterior margin produced into a small median spine ; the fourth and fifth free somites
are together equal in length to the first free somite, and are without spines. The coxal
plates of all the thoracic somites are compact and those of the first three free somites are
separated from them by sutures. The coxal plates of the seventh thoracic somite extend
for a short distance beyond the bases of the uropoda and farther than the pleural plates
of the second and third abdominal segments, the latter of which are a little longer than
the former and reach to the bases of the uropoda. The terga of the three abdominal
segments are each produced posteriorly into a small median dorsal spine.

The terminal segment is roughly pentagonal in outline, slightly longer than broad,
with a sub-acute posterior extremity. On its dorsal surface, in an anterior median
position, is a small raised triangular area with its apex directed backwards ; this is con-
tinuous with a median carina which extends to the posterior extremity of the segment.
On either side, arising from the angle of the base of the triangle, is another carina which
runs obliquely outwards and terminates in a spine some distance within the lateral
margin, just behind the level of the base of the uropod. Midway between this and the
median carina, and running parallel with the latter, is a third carina which starts an-
teriorly a little behind the apex of the triangular area and terminates posteriorly a short

DESCRIPTION OF SPECIES 3i9

distance within the margin of the segment. Just within the anterior and lateral margins
of the segment is a narrow ridge which follows the outline of the segment as far as the
base of the uropod.

Remarks. Monod (1926, p. 37) does not describe any of the appendages of this
species. He figures most of them ; but some of the figures are inaccurate in so far as they
do not show the complete structure of the appendage concerned.

The flagellum of the antenna consists of fifteen joints ; on the third to the ninth joints,
just within the anterior margin, is a row of small spines, which have been omitted in
Monod's fig. 36 A. Only the distal portion of the outer lobe of the maxillula (fig. 37 D)
is figured, although both lobes can be seen in Monod's preparation; the inner lobe is
broken near its base and is lying partly covered over by the outer lobe. The appendage
is of the usual type ; the inner lobe is about half the length of the outer, with a slightly
broadened distal extremity bearing one short seta. The figure of the maxilla (fig. 37 E)
is also incomplete, as it does not show the sutures or the basal segments of the appendage.
The basipodite of the maxilliped (fig. 37 F) is separated from its lamella by a suture.

In this species a small extra joint bearing a number of setae is found at the outer distal
angle of the third joint of the palp: a similar joint has already been described in
S. schythei, Liitken, and S. polaris, Richardson. The suture of the exopod of the fourth
pleopod is transverse, and its outer distal border is fringed with long plumose setae ;
the endopod is rounded. The uropod is figured by Monod (fig. 37 G), but the scattered
plumose setae on the margins of both endopod and exopod are not shown.

Locality. 71° 19' S, 87° 37' W, at a depth of 400 m.

18. Serolis meridionalis, Hodgson.

S. meridionalis, Vanhoffen, 1914, p. 518, fig. 51.

In the Report of the Scientific Results of the Voyage of S.Y. 'Scotia' (vol. iv (i),
pi. xi) there is a photograph of this species taken by W. S. Bruce. The species was
named by Hodgson and it has since been recorded and briefly described by Vanhoffen
(191 4, p. 518) from material collected during the Deutsche Siidpolar-Expedition.
Through the kindness of the Keeper of the Royal Scottish Museum I have been able to
examine the type specimen.

Diagnostic characters. The type is an adult female in the breeding phase. It is not
possible to give its exact size, as unfortunately the greater portion of the much-produced
terminal process is broken off; nevertheless it reaches a length of 80 mm. and a breadth
of 55 mm., and it is probable that the complete length would be from 10 to 15 mm.
greater. The specimen is considerably larger than that of Vanhoffen ; this is also a female
in the breeding phase, and it is 59 mm. in length and 35 mm. in greatest breadth. The
species is by far the largest in the genus.

The head is broader than long, with its lateral margins parallel and with its anterior
margin excavated for the reception of the antennules. In its antero-lateral angle, on
either side, is a somewhat triangular-shaped prominence, whilst behind this is a second
one, oval in shape, extending obliquely inwards from the postero-lateral angle. On the

9-2

320 DISCOVERY REPORTS

outer side of this prominence is the long narrow eye, which is devoid of pigment.
Vanhoffen (p. 518) does not mention the eyes in his description; but he illustrates them
(fig. 51 a), using cross-hatching which, however, may be intended to represent optical
units and not pigment.

Either the absence or the reduction in the amount of pigment, or the absence or the
reduction in the size of the eyes, seems to be a characteristic of the deep-sea forms, for
the colour of the eyes of S. bromleyana, Suhm, is whitish yellow, and that of S. neaera,
Beddard, is bluish black owing to the comparatively small amount of pigment, whilst in
S. gracilis, Beddard, the eyes are small and inconspicuous and in S. antarctica, Beddard,
they are absent.

On the inner side of each eye prominence is a small backwardly projecting spine ; the
rest of the area between the eyes is raised and divided posteriorly into a median and two
lateral parts, the latter of which have their posterior margins toothed ; the posterior
margin of the median part is straight. The lateral portion of the cephalosome, on either
side, is divided by a transverse groove which extends from the posterior angle of the eye
to the lateral margin.

The coxal plates are curved and sickle-shaped; the articular processes which unite
together the succeeding plates are placed at some distance from the junction between
the terga and the plates ; those of the first three free somites are separated from them by
sutures. The coxal plates of the seventh thoracic somite extend backwards to about the
base of the spine of the terminal segment. The pleural plates of the second and third
abdominal segments are of approximately equal length and do not reach quite as far as
the base of the uropoda.

The terminal segment is considerably longer than broad ; it is broadest at the level of
the bases of the uropoda, from which point it gradually narrows to form a long spiniform
process, the lateral margins of which are deeply serrated. A longitudinal median toothed
keel runs from near the anterior margin to the end of the terminal process, and between
the anterior end of the keel and the anterior margin are two small, backwardly directed,
rounded spines, situated one on either side of the median line. Between these spines and
the median keel, on either side, is an oblique transverse ridge which extends to the base
of the uropod and terminates in a small spine. The uropoda are comparatively small.
Unfortunately in the type specimen one is missing and the other incomplete ; the proto-
podite is only slightly produced on its inner side.

Remarks. Vanhoffen gives no description of the mouth-parts and appendages of this
species. I append a few notes, which are necessarily brief as my observations have had
to be restricted to those made on the entire specimen.

The antennule is not much shorter than the antenna. The first joint of the peduncle
is almost triangular in shape ; the second is 7 mm. long ; the third is shorter than the
second, measuring 5-5 mm., and the fourth is 3-5 mm. in length. Dorsal grooves are
present on the second and third joints. The flagellum measures 18 mm. in length and
consists of thirty-three joints.

The antenna consists of a peduncle of five joints, the first two of which are short;

DESCRIPTION OF SPECIES 3ai

the third measures 7 mm., the fourth 11 mm., and the fifth 14 mm. in length. There
is a deep longitudinal groove on the dorsal surface of each of the last three peduncular
joints. The flagellum is considerably shorter than that of the antennule ; it measures
1 1 mm. in length and consists of only seventeen joints.

The basipodite of the maxilliped is separated from the lamella by a suture ; the suture
between the coxopodite and the epipodite is not so easy to trace, but it appears to
be present. As in the other deep-sea forms a small tubercle, covered with hairs, is
present near the base of the second joint of the palp.

The propodus of the second thoracic appendage is broad, and its inner margin is
armed with a row of thirty-four somewhat triangular processes alternating with a row
of modified spines, the detailed structure of which it was not possible to make out. The
inner distal extremity of the carpus is crenulate and bears two short spines. The re-
maining thoracic appendages are slender ; as in S. bromleyana, Suhm, the fourth joint
is longer than either the third or the fifth. The setae, which are restricted to the distal
ends of the propodus, carpus and merus, and to two groups on the outer margins of the
merus and the carpus, are all of the simple variety.

The sterna of the abdominal segments are produced backwards in the median line
and the resuhing spines increase in size from before backwards. The protopodite of
each of the first three pairs of pleopods is large, and roughly triangular in shape ; I have
been unable to see any spines on the produced angles. The suture of the exopod of both
the fourth and fifth pleopods is oblique, and in both cases the endopod is entire.

Distribution. Off Coats Land, in 2759 m.

A considerable amount of confusion surrounds the identification of the two following
species, S. trilobitoides, Eights (PI. XIV, fig. 7) and S. cornuta, Studer (PI. XIV, fig. 6).
The former species was described by Eights in 1833 and the latter by Studer in 1879,
although he mentioned it in an earlier paper ( 1 876) under the name of Brongniarta cornuta.
Neither the descriptions nor the figures are detailed or accurate, and since the two species
are very much alike and have never both been recorded, until now, in any single collection,
and since Studer (1879) does not bring out the main points of difference between the
two, it is easy to see how such confusion has arisen.

Beddard (18846), in his account of the Challenger Serolids, describes and figures
certain specimens (pp. 49-53) which he assigns to S. cormita. I have examined these
specimens at the British Museum and compared them with those of the present collec-
tion and I have come to the conclusion that his specimens are S. trilobitoides and not
S. cornuta, and that actually S. cornuta is not represented in the Museum collection.

Hodgson (1910), who redescribed S. trilobitoides, considers that the two species are
identical, for he says, "I have, I think, satisfactorily proved that S. cormitus, Studer, is
merely the immature form of S. trilobitoides, Eights", and he includes S. cornuta as a
synonym of S. trilobitoides. In so far as Hodgson's results are drawn from existing
literature and since Beddard 's figures prove to be those of S. trilobitoides and not

333 DISCOVERY REPORTS

S. cornuta, his conclusions are justified ; but I may point out that even the smallest speci-
mens of S. cornuta exhibit the characteristic shape of the adult. It is interesting to note
that Studer does not state the sex of the specimen he figures ; nevertheless, Beddard says
it is a male, whilst Hodgson, referring to the same figure, says, "this figure, from its
great breadth is probably a female ". By this statement Hodgson recognizes one of the
chief difl^erences between the two species, but, as is usual amongst the SeroUds, the
male has the greater proportional width. The only other record of either species is that
given by Monod (1926) for S. trilobitoides, Eights.

The most noticeable point of difi'erence between the two species lies in the relative
proportions of the various parts of the body. In both species, the male is broader in pro-
portion to its length than is the female . The following comparison is drawn between two
adult males.

(i) The shape of the body of S. cornuta, excluding the terminal spine of the last
segment, is almost circular, whilst that of S. trilobitoides is broadly ovate ; the actual
measurements are:

S. cornuta : length minus terminal spine = 49 mm. ;

breadth = 49 mm. ;

terminal spine = 4 mm.
S. trilobitoides: length minus terminal spine =52 mm. ;

breadth = 48 mm. ;

terminal spine = 2 min.
The proportions of the body in the specimens in the British Museum and also of
Hodgson's specimen agree with those of S. trilobitoides, whilst the measurements given
by Eights (70 x 57 mm.) for that species seem to point to the same general shape,
though the ratio of length to breadth is slightly greater, 1-22: i as compared with
I-I2 : I. Studer (1879) gives the length of his largest specimen of S. cornuta as 30 mm.
and makes no mention of its breadth ; but if the specimen is figured correctly his specimen
agrees in shape with those of the present collection, for the length without the terminal
spine is 32 mm., the breadth also 32 mm., and the length of the spine 3-5 mm.

(2) The terminal segments of both species (Fig. 13 a, 6) are pentagonal in outline;
but that of S. cornuta is longer than broad with a much more acute posterior extremity
and with a longer terminal spine, whilst that of S. trilobitoides is broader than long ; the
measurements taken from two adult male specimens are :
S. cornuta: length of segment + spine = 19 mm. ;

length of spine = 4 mm. ;

breadth =17 mm.
S. trilobitoides: length of segment + spine = 18 mm. ;

length of spine = 2 mm. ;

breadth =19 mm.
This difference in shape can be clearly seen if the figures of Beddard (1884^, pi. i,
figs. 1-3) and of Hodgson (1910, pi. iv, figs, i, 2) are compared with those of Studer
(1879, taf. iii, figs. I, 3) for S. cornuta.

DESCRIPTION OF SPECIES

323

(3) The spines on the median dorsal keel of the terminal segment are much larger
and fewer in number in S. cornuta than in S. trilobitoides.

(4) The length of the coxal plates of the seventh thoracic somite also differs in the two
species. In S. cornuta (Fig. 13 Z)) they extend back-
wards almost to the level reached by the tip of the
terminal segment, and a considerable distance
beyond that reached by the pleural plates of the
second and third abdominal segments ; the pleural
plates of the second segment are slightly longer
than those of the third, and both pairs extend back-
wards beyond the distal end of the protopodite
of the uropod. In S. trilobitoides (Fig. it, 0) the
coxal plates reach backwards to a level half-way
between the attachment of the uropod and the tip
of the terminal segment and a little way beyond
the pleural plates of the abdominal segments : the
pleural plates of the second abdominal segment
are very slightly longer than those of the third.

The British Museum specimens of " Serolis
cornuta" agree with S. trilobitoides on this point,
as do also the figures of Hodgson (1910, pi. iv,
figs. I, 2), though those of Beddard (1884 b) show
the coxal plates of the seventh thoracic somite in
the male (pi. i, figs, i, 3) rather more than their
normal length though not as long as those of the
male in the true S. cornuta. Studer (1879), i^^ his
original description of S. cornuta, states that the
coxal plates of the seventh thoracic somite extend
beyond the middle of the caudal shield, and in his
figure (taf. iii, fig. i) shows them extending to the
base of the terminal spine, the level reached by the
corresponding plates in the females of this collec-
tion. In his figure of the abdomen (taf. iii, fig. 3),
which also includes these plates, they are shown
very much shorter, whilst the pleural plates of the
abdominal segments do not reach as far as the level
of attachment of the uropods. Since in all existing
figures, and in all examined specimens of both species, the pleural plates extend back-
wards beyond the level of the articulation of the uropod, it may be presumed that on
these points Studer's figure (fig. 3) is inaccurate, though the actual form of the terminal
segment appears to be correct.

(5) Observations made on living specimens of the two species show that a further

Fig.

It.

13. Seroh's trilobitoides. Eights, and
S. cornuta, Studer.
S. trilobitoides. Eights, a, abdominal and
terminal segments : x 2.
S. cornuta, Studer. b, abdominal and ter-
minal segments: x 2. c, cutting edges of
mandible: x 40.

324 DISCOVERY REPORTS

difference exists between them ; the colour of S. cormita is pale grey with dark brown
spots of pigment (for distribution see Plate XIV, fig. 6), whilst that of S. trilobitoides is
pale horn colour, blotched and with a darker colour which varies from maroon to pale
terra-cotta. The colour given by Studer for S. cormda is pale with a darker posterior
edge to each segment.

(6) With regard to the distribution of the two species, both have been recorded from
Kerguelen, S. cormita from the west of the island, and S. trilobitoides from Betsy Cove.
Studer's material of S. cormita came from this locality and also from the Crozet Islands,
whilst that of the present collection comes from the South Orkneys and South Sandwich
Islands. The specimens of S. trilobitoides in the present collection come from Clarence
Island and the South Shetlands, and the species has previously been recorded from the
latter locaHty. Hodgson's material came from a station in 67° 21' 46" S, 155° 21' 10" E,
which is much farther east than any other recorded for either species.

19. Serolis cornuta, Studer (Figs, i a, c, 13 ^, f, Plate XIV, fig. 6).

S. cormita, Studer, 1879, pp. 21-24, P'- '"> ^8^. 1-7; 1884, p. 7.
Brongniartia cornuta, Studer, 1876, p. 75.
Occurrence. St. 164: South Orkneys, 24-36 m.; 2 ?? (b.), 6 $$ (non-b.), 2 adult cjc?, 4 immature (JcJ, and
a number of small specimens.
St. 363: South Sandwich Islands, 329-278 m.; 2$? (non-b.), i young cj.

Diagnostic characters. The body of the adult male (Plate XIV, fig. 6), excluding
the spine on the terminal segment, is circular in outline, that of the female is not quite
as broad; the largest male in the collection measures 49 mm. in length, excluding the
terminal spine of 4 mm., and 49 mm. in breadth, whilst the female is 51-5 mm. long,
without the terminal spine, and 49 mm. in greatest breadth. The colour of the living
animal is pale grey with dark brown spots of pigment.

The head is shield-shaped, broadest anteriorly, with the anterior margin excavated
on either side of the small median rostrum for the reception of the antennules ; im-
mediately behind this is a transverse ridge extending to the sides of the head. In front
of and between the eyes, is a squarish raised area with its posterior margin slightly
concave, the two ends being somewhat produced backwards; the posterior margin is
separated from the remaining portion of the head by a deep groove. The area between
the eyes is raised, and separated by grooves from the elongated protuberances on the
outer sides of which the eyes are situated. The area between the eyes is divided into three
parts, of which the central is less convex and has its posterior margin produced into a
small, median, somewhat rhombic-shaped area which bears a small rounded pigment
spot ; each of the lateral parts is convex, oval in shape and produced posteriorly into a
small spine.

The eyes are long, reniform in shape, rather narrow, and containing black pigment.
The lateral portion of the cephalosome is divided into an anterior and a posterior part
by a well-marked transverse groove, which extends outwards to the lateral margin from
a point near the postero-lateral angle of the eye.

The thoracic somites are all sub-equal, with well-developed coxal plates, those of the

DESCRIPTION OF SPECIES 325

third to the fifth somites being separated by sutures. The coxal plates are sickle-shaped,
with the terminal portions from the articular processes outwards, curved first forwards
and then outwards, so that a space is left between two consecutive plates. Those of the
seventh somite are produced backwards to near the end of the terminal segment in the
male, and to about the level of the base of the terminal spine in the female.

The first three abdominal segments are sub-equal (Fig. 13 b), each a little more than
half the length of the thoracic somites. The pleural plates of the second and third seg-
ments are long, narrow and produced backwards beyond the point of attachment of the
uropod, those of the second being slightly the longer : the outer margins of the coxal and
pleural plates are serrated. In the female the sterna of the first three abdominal seg-
ments each have their posterior margin produced into a median spine which is largest
on the third segment; these are hardly developed in the male.

The terminal segment (Fig. 13 6) is pentagonal in outline, slightly longer than broad,
terminating in a long spiniform process which is slightly upturned ; on its dorsal surface
is a median dorsal keel bearing five large spines directed upwards and slightly back-
wards, and decreasing in size from before backwards. The area in front of the first spine
is raised and bears two small spines, from the base of each of which a transverse oblique
carina extends outwards to near the lateral margin where it ends in a sharp backwardly
directed spine. The margin between the attachment of the uropod and the terminal
spine is sharply serrated.

Remarks. As the descriptions of the mouth-parts and appendages given by Studer
(1879, PP- 21-24) are very brief, some further details are given here. The appendages
are all of the normal type and bear a very close resemblance to those of S. trilobitoides.
Eights. The antennule and the antenna are described by Studer, but he failed to observe
that a dorsal groove is present on each of the first three peduncular joints of the former,
and of the second to fifth joint of the latter. The surface of each groove, like that of
S. trilobitoides, is covered by short setae which are all directed towards the centre of the
groove in a distal direction. The number of joints in the flagella of both antennule and
antenna appears to be greater than that given by Studer ; his number for the former is
twenty-two and for the latter fourteen, whereas in specimens in the present collection
these numbers are increased to thirty and eighteen respectively. A further detail
omitted from Studer's description is the mention of the row of teeth occurring on the
ventral surface of the middle joints of the flagellum of the antenna : these are also seen
in a corresponding position in S. trilobitoides and were figured by Beddard (1884&, pi. i,
fig- 6).

The cutting edges of the mandibles are strongly chitinous : their form is shown in
Fig. 13 c. The outer lobe of the maxillula (Fig. i a) bears nine strong spines and an inner
pectinate seta on its distal extremity. The inner lobe is broad distally, with a series of
extremely small toothed spines on the outer margin of its rounded end, and a single
longer spine near its inner margin. There are two setae, on both the inner and outer
lobes of the left maxilla (Fig. i c), whilst on the right there are two on the outer and
three on the inner lobe : the fixed lobe of both bears about thirty setae.

326 DISCOVERY REPORTS

The basipodite of the maxilliped is separated from its lamella by a suture. The second
joint of the palp is broad, with its margin fringed with setae, some of which have their
distal ends flattened and toothed ; the third joint is almost parallel-sided with its truncate
distal extremity fringed with setae. The ventral surface of the palp and of the basipodite
is sparsely covered with delicate setae. The modified spines on the inner margin of the
propodus of the second thoracic appendage are similar to those on the corresponding
joint of S. trilobitoides, and are very like those figured for S. platygaster, n.sp. (Fig. 196).
The propodus of the third thoracic appendage of the adult male bears a double row of
seven modified spines at the proximal end of its inner margin and a further five extend
in a median position towards the distal end of joint. These spines are short and stout,
with a delicate distal extremity and with the distal half, facing towards the distal end of
the propodus, covered with short dense hairs. Similar setae are found on the inner
margin of the carpus, and a single one at the inner distal angle of the merus. The re-
maining appendages bear setae which are either simple or pectinate ; the latter, which
are more abundant on the joints of the last appendage, are found towards their distal
ends. The last pair of thoracic appendages is much smaller than the rest : in that of the
mature male the propodus is broader than the corresponding joint in the female, and
its outer margin is armed with a number of strong spines ; the dactylus in the male is also
slightly recurved.

The protopodite of the first three pairs of pleopods has a produced angle which bears
long plumose setae, three on that of the first and two on each of the second and third.
The sutures of the exopod of both fourth and fifth pleopods are oblique and placed
rather near their distal ends: the endopods of both are large and undivided. The proto-
podite of the uropod has its inner angle produced into a sharp spine. The endopod is
a little longer than the exopod; both are elongate-oval in shape with pointed distal
extremities. The apices of both exopod and endopod and the outer margin of the former
are deeply serrated.

Distribution. Studer's specimens were collected from the Crozet Islands and a
locaHty west of Kerguelen Island ; those in the present collection come from the South
Orkneys and South Sandwich Islands.

20. Serolis trilobitoides, Eights (Fig. 13 rt; Plate XIV, fig. 7).

S. trilobitoides, Eights, 1833, pp. 53-7, 2 plates.

5'. cornuta, Beddard, 18846, pp. 49-53, pi. i, figs. 1-6.

S. trilobitoides, Eights, Hodgson, 1910, pp. 23-30, pi. iv, figs. 1-8; Monod, 1926, p. 38; Nordenstam,

1933. PP- 59-60, text-fig. 5 a.
S. zoiphila, Stechow, 1921, pp. 221-3; Nierstrasz, 1931, pp. 222-4.

Occurrence. St. 170: Clarence Island, 324 m.; i ? (non-b), i immature specimen.
St. 172: South Shetlands, 525 m.; i <? and 3 immature specimens.

Diagnostic characters. Body (Plate XIV, fig. 7) broadly ovate. The largest male in
the collection is 52 mm. in length without the terminal spine of 2 mm., and 48 mm. in
greatest breadth; the female is 51-5 mm. in length (including the terminal spine) and

DESCRIPTION OF SPECIES 327

44 mm. in breadth. The colour of living specimens was pale horn, blotched and mottled
with a darker colour which varied from maroon to pale terra-cotta.

The head is shield-shaped and almost identical with that of S. cornuta, Studer: it is
broadest anteriorly, with the anterior margin excavated on either side of the small
median rostrum for the reception of the antennules ; immediately behind this is a trans-
verse ridge extending laterally to the sides of the head. In front of and between the eyes
is a square raised area with its posterior margin slightly concave, the two angles being
somewhat produced backwards. The area between the eyes, which, except anteriorly, is
separated from them by deep grooves, is convex and is divided posteriorly into a median
and two lateral areas, each of the latter being in the form of a flattened, somewhat oval
enlargement, produced backwards into a very small spine. The median area is produced
into a narrow four-sided plate with a small dark tubercle in its centre ; this lies between
and rather behind the two lateral enlargements. The eyes are reniform in shape, rather
narrow, and containing black pigment.

The lateral portion of the cephalosome is divided on either side into an anterior and
a posterior portion by a well-marked transverse groove, which extends outwards from
a point near the postero-lateral margin of the eye to the lateral margin. A further
ridge on the anterior portion arises near the end of the anterior transverse ridge of the
head.

The thoracic somites increase in length from the third (first free) to the fifth ; the fifth,
sixth, and seventh are sub-equal. Each somite has well-developed coxal plates, those
of the third, fourth and fifth being separated by sutures. The plates are somewhat
sickle-shaped, with the terminal portion from the articular process outwards curved
backwards, and outwards, following closely the outline of the preceding one (cf.
S. cornuta, Studer). The coxal plates of the seventh thoracic somite are produced back-
wards to a point half-way between the attachment of the uropod and the end of the
terminal segment, reaching slightly beyond the pleural plates of the abdominal seg-
ments ; the lengths of these plates are almost the same in both sexes.

The first three abdominal segments are sub-equal (Fig. 13 a), with the sterna each
bearing a median spine, increasing in size from before backwards; these are well de-
veloped in the female, but only slightly in the male. The pleural plates of the second and
third abdominal segments extend backwards almost as far as the coxal plates of the
seventh thoracic somite ; those of the second segment are slightly the longer. The outer
margins of the coxal and pleural plates are minutely though not so deeply serrated as
those of S. cornuta.

The terminal segment (Fig. 13 a) is pentagonal in outline, ending posteriorly in a
spiniform process which is not as long or acute as the corresponding one in S. cornuta.
The segment is broader than long, that of the largest specimen measuring 18 mm. in
length (including the terminal spiniform process of 2 mm.) and 19 mm. in greatest
width. On the dorsal surface is a longitudinal median keel bearing five or six recurved
teeth ; the first of these is the largest, but none are as large as those in a similar position
in S. cornuta. The area in front of the first tooth is raised and bears two small spines,

328 DISCOVERY REPORTS

and from the base of each an obUque ridge extends to near the lateral margin at the
level of the attachment of the uropod, where it ends in a small, backwardly directed
spine. The margins of the segment behind the base of the uropoda are serrated.

Remarks. The appendages are of the usual type, and have been described by
Beddard (1884 Z), pp. 51-3, pi. i, figs. 1-16, as S. cornuta), and by Hodgson (1910,
p. 26, pi. iv).

The first three joints of the peduncle of the antennule and the second to fifth of the
antenna, are marked by ridges which extend along the dorsal surface of the joints not
far from their posterior margins, and the surface of the groove towards the anterior side
of each ridge is covered with short setae, all of which are directed towards the centre
of the groove and in a distal direction. This point has not been described previously.
The number of joints in the antennular and antennal flagella appears to be greater in
the present specimens than previously noted ; the antennule has forty-one joints as com-
pared with twenty-five in earlier descriptions, and the antenna twenty as compared with
sixteen. The row of teeth occurring on the middle joints of the antennal flagellum,
figured by Beddard (18846, pi. i, fig. 6), occurs on the ventral surfaces of the joints.

Hodgson (1910, p. 28) states that he is unable to detect any division in the basal plate
of the maxilliped even with a I objective ; my observations agree with those of Beddard
(18846, pi. i, fig. 11), who shows a suture between the coxal plate and the epipod.
Hodgson's figure (pi. iv, fig. 6) is thus inaccurate on this point.

The forms of the modified spines which arm the inner margin of the propodus of
the second thoracic appendage are figured by Hodgson (1910, pi. iv, figs. 7, 8). The
propodus of the third thoracic appendage of the aduh male is modified in the usual way ;
the inner margin of the enlarged propodus bears spines of the type usual to that joint-
seven pairs are arranged around the proximal portion of margin, and a further five
spines form a median row towards its distal end.

Both Beddard (18846, p. 52) and Hodgson (1910, p. 30) state that the exopod of the
uropod is almost twice as long as the endopod. This, however, is not correct: theendo-
pod is the longer, but only slightly so, the actual measurements in a large specimen being :
endopod 8 mm., exopod 7 mm. Hodgson's statement (p. 30) that the exopod is two-
jointed, "the terminal one being scarcely half as long as the other", is also incorrect:
in this species, as in all other Serolids, the exopod consists of a single joint.

Distribution. This species has been recorded from Betsy Cove, Kerguelen, and
from Clarence Island and the South Shetlands, as well as from a locality much farther
east: 67° 21' 46" S, 155° 21' 10" E (Hodgson).

[Serolis zoiphila, Stechow.

S. zoiphila, Stechow, 1921, pp. 221-3; Nierstrasz, 1931, pp. 222-4.

Stechow (1921, pp. 221-3) describes a new species of Obelia (O. longa) which he
found attached to the caudal segment of a species of Serolis. The only statement he
makes concerning this species is that it occurred at Kerguelen and that he wishes to call
it Serolis zoiphila, n.sp. A photograph is given which clearly shows it to be a specimen

DESCRIPTION OF SPECIES 329

of S. trilobitoides, Eights, and I have included the name in the synonymy of that species.
The name is included in the list of Serolids given by Nierstrasz (1931).]

21. Serolis antarctica, Beddard.

iS. antarctica, Beddard, 1884 a, p. 333; 18846, pp. 63-6, pi. iii, figs. 1-6.

Diagnostic characters. This species, which lives in deep water, has not been
recorded since it was originally described by Beddard. The largest male measures
33 mm. in length and 31 mm. in greatest breadth, and the largest female 31 mm. in
length and 26 mm. in breadth, so that the male is broader in proportion to its length
than the female.

The eyes are absent, their place being taken by rounded tubercles covered with
chitinous integument in which there is no trace of any optical structures ; between these
tubercles is a median one which is short and blunt.

The dorsal surface of the body is strongly sculptured and the posterior margin of each
of the thoracic and the abdominal segments is produced backwards as a short, blunt,
median spine. The coxal plates are comparatively short and flat; those of the first three
free thoracic somites are almost rectangular in shape and are separated from their
respective somites by sutures. The coxal plates of the seventh somite of the male are
produced backwards to a short distance beyond the end of the terminal segment ; those
of the female do not extend beyond the terminal segment. The pleural plates of the
second and third abdominal segments reach to the base of the uropoda, those of the
latter being slightly the longer.

The terminal segment is more or less hexagonal in shape ; the lateral edges are turned
downwards so that the uropoda cannot be seen from the dorsal surface; "there is a
median longitudinal keel which bifurcates at about the end of the anterior fifth, on
either side is a Y-shaped keel inclined at an oblique angle ; the portion of the caudal
shield which lies between the median and lateral keels is flat, the part which lies outside
the inner fork of the lateral keel is bent downwards ; the posterior end of the caudal
shield is slightly bent up" (Beddard, 1884 Z», p. 65).

Occurrence. Off Pernambuco (i 375-1 600 fathoms), and between Prince Edward
Island and the Crozets.

22. Serolis bromleyana, Suhm.

S. bromleyana, Suhm, 1874, p. xix; 1876, p. 591; Beddard, 18846, pp. 53-7, pi. iv.

Diagnostic characters. This species can no longer be considered the largest of the
genus, for the length of the largest known male is only 54 mm., which is less than two-
thirds the length of Hodgson's female of S. meridionalis. The breadths of the specimens
are approximately the same, that of the former being 56 mm., that of the latter 55 mm. ;
the length of the female is 45 mm., and its greatest breadth 39 mm.

The body is more or less oval in shape, and its surface is covered with shallow pits
and scattered hairs: the colour (in alcohol) is "violet-grey with whitish yellow patches
upon the caudal shield and posterior portion of the thorax" (Beddard, 18846, p. 53).

The head is longer than broad, owing to the projection of its antero-lateral portions

33° DISCOVERY REPORTS

for some distance beyond the rostrum : these antero-lateral portions are separated from
the rest of the head by a transverse ridge which runs from just behind the rostrum to
the lateral margin on either side. This character distinguishes the species from all other
deep-sea forms. The central portion of the head between the eyes is strongly convex
and is divided into three areas — "two round convexities which lie to the inner side of
and behind each eye, and a median T-shaped elevation, at the upper end of which, on
a level with the anterior portion of the eyes, are four tubercles arranged in a semicircle
with the concavity directed forwards ; at the hinder extremity is another short tubercle "
(Beddard, 1884 /!>, p. 54). The eyes are whitish yellow in colour.

The posterior border of each of the third to seventh thoracic somites is produced into
a minute median dorsal spiniform process; the third, fourth and fifth somites are
separated by sutures from their coxal plates. The coxal plates are long and sickle-shaped
and increase in length from before backwards; those of the seventh thoracic somite
extend for a considerable distance behind the termination of the terminal segment and
are longer in the male than in the female. In the male the pleural plates of the second
abdominal segments extend beyond, whilst those of the third extend as far as the tip of
the terminal segment ; in the female those of the third segment hardly reach as far as
the extremity of the terminal segment. The terminal segment is broader than long,
somewhat pentagonal in outline, with the posterior extremity notched and slightly
turned up ; on its dorsal surface is a median longitudinal keel, and also, on either side,
a short flat spine, placed near the lateral margin on a level with the attachment of the
uropod.

Remarks. There are two points in which my observations on this species difl^er from
those of Beddard. In the first place, he states in his description of the maxilliped, that
" the stipes and lamina are not separated by a complete suture ", or, in other words, that
the basipodite and lamella are not completely separated. In his figure (pi. iv, fig. 8),
however, he shows these two lobes as separated by a suture, but the coxal joint fused
with the epipodite. Actually the basipodite is separated by a suture from the lamella and
the coxa by a suture from the epipod. The second point is that Beddard describes the
third thoracic appendage of the male (p. 56) but calls it the second appendage, whilst
actually he omits any description of the second. The second appendage is of the usual
type and bears on its propodus alternating rows of two varieties of modified spines very
similar in appearance to those found in S. beddardi, Caiman (Fig. 3 b).

Occurrence. Off" the east coast of New Zealand, in 1 100 and 700 fathoms, and close
to the Antarctic Ice-Barrier in 1975 fathoms.

23. Serolis neaera, Beddard.

5. neaera, Beddard, 1884 «, pt. iii, p. 331; Beddard, 18846, p. 57, pi. v, figs. i-ii.

Occurrence. St. WS 212: 49° 22' S, 60° 10' W, 242-249 m.; 8 immature.

St. WS 213: 49° 22' S, 60° 10' W, 249-239 m.; I $ (b.), i immature.
St. WS 236: 46° 55' S, 60° 40' W, 272-300 m.; 2 ?? (b.), 2 SS, i immature.
St. WS 244: 52° 00' S, 62° 40' W, 253-247 m.; 2 (?(?, 2 ?? (b.), 6 immature.
St. WS 773: 47° 28' S, 60° 51' W, 291-296 m.; 1 ? (b.), 3 immature.

DESCRIPTION OF SPECIES 33i

St. WS 820: 52° 53' 15" S, 61° 51' W, 351-367 m.

St. WS 821 : 52° 55' 45" S, 60° 55' W, 461-468 m.; i <J, several damaged specimens.

St. WS 839: 53° 30' 15" S, 63° 29' W, 403-434 m.; I ? (non-b.), 5 immature specimens.

Diagnostic characters. The specimens in the collection are smaller than those
taken by the ' Challenger '. The actual body length of both the largest male and female
is 26 mm. ; but if the length of the seventh coxal plates is included this is increased to
37*5 mm. in the former, and 33 mm. in the latter, the difference in length being due to
the greater length of the coxal plates of the male. As in most species the male is pro-
portionately broader than the female and measures 33 mm. as compared with 28 mm.,
the width of the female.

The head is broader than long, broadest anteriorly, with a small rostrum behind
which a transverse ridge extends laterally to the margin of the cephalosome ; im-
mediately behind this is a fairly deep groove. The area between the eyes is convex, and
extending across this area, not far behind the anterior margins of the eyes, is a transverse
row of four spines, the inner ones of which are the smaller. The posterior margin of the
head is indented and forms three projections, the lateral of which are somewhat tri-
angular in shape ; the central one is elongated transversely and is oval in shape. Unlike
other deep-sea forms the eyes are large, being both long and broad ; they are reniform
in shape but not deeply pigmented, the colour being bluish-black. According to Beddard
(18846, p. 26) the minute structure resembles that of the eyes of S. bromleyana,
Suhm, but is a little less degenerate.

On the lateral portion of the second thoracic somite is a transverse ridge extending
from about the middle of its inner margin to the lateral margin. The posterior margin of
this portion is curved forwards, then backwards, so that it does not extend back as far
as the middle portion of the segment. Each of the third to seventh thoracic somites has
its posterior margin produced in the middle dorsal line to form a spiniform process;
those of the third and fourth somites are larger than those of the other somites. The
third and fourth somites are sub-equal, the fifth a little longer, and the sixth and seventh
each a little shorter than the third. The coxal plates are long, narrow and spiniform,
increasing in length from before backwards ; those of the seventh thoracic somite extend
backwards for some distance beyond the extremity of the terminal segment, and in the
largest male in the collection reach a length of 20 mm. and extend backwards in a
direction almost parallel with the long axis of the body. The coxal plates of the third,
fourth and fifth somites are separated from them by sutures.

The three free abdominal segments each possess a small median dorsal spiniform
process on the posterior border; the pleural plates of the second segment are long,
extending backwards for some distance beyond the end of the terminal segment ; those
of the third segment are short, hardly extending beyond the anterior margin of the
terminal segment. The terminal segment is slightly broader than long, roughly hexa-
gonal in outline. It has a slight median dorsal keel the length of which is divided into
three approximately equal lengths by two spines, the anterior one of which is the larger.
On either side of the posterior spine, connected to it by a curved ridge, is a small lateral

332 DISCOVERY REPORTS

spine, lying midway between the posterior spine and the lateral margin of the segment.
Two small oblique ridges extend from either side of the base of the anterior spine and
terminate in small spines on a level with the lateral ones and midway between them
and the lateral margins of the segment. The uropoda are attached rather far back on the
terminal segment, a little behind the level of the posterior spine; both exopod and endo-
pod are well developed and the latter is the larger.

Remarks. There is little to add to Beddard's account of this species, but a minor
point may be mentioned in connection with the third thoracic appendage of the adult
male. Beddard (18846, p. 59) describes the modified spines which are found on the
propodus of this limb, but does not mention that they only occur on the proximal half
of the joint, their place being taken distally by a double row of long branched setae
similar to those found on the inner margins of the merus and carpus. Similar setae,
though not so densely arranged, are found on the inner margins of the carpus, merus
and propodus of the last thoracic appendage of the adult male.

Distribution. Beddard's specimens were collected at two stations, 42° 32' S,
56° 29' W, at a depth of 2040 fathoms, and 37° 17' S, 53° 52' W, in 600 fathoms. Those
in the present collection come from a locality a little farther to the south-west, between
46°-53° S and 6o°-63° W, and from depths of 242-434 m.

24. Serolis minuta, Beddard.

S. minuta, Beddard, 1884 i, p. 77, pi. vii, figs. 2-7; Whitelegge, 1901, p. 237; Chilton, 1917, p. 397.
S. minuta, Beddard var. eugeniae, Nordenstam, 1933, pp. 82-5, pi. i, fig. 3, text-figs. 11 h, 20.

Diagnostic characters. The Challenger collection contained a single male mea-
suring 5 mm. in length, and 5 mm. in greatest breadth ; the two specimens mentioned
by Chilton (p. 397) are both females in the breeding condition, the larger of which
measures 8 mm. in length and 7 mm. in breadth. The male is therefore slightly broader
in proportion to its length than is the female, and the outline of the body is almost
circular.

The head terminates in front in a short rostrum; the posterior margin is provided
with three blunt tubercles, the middle of which is the larger. Each of the following
thoracic somites has its posterior border produced into a median blunt spine. The third
to the seventh somites have well-developed coxal plates and those of the third to the
fifth somites are marked off by distinct sutures ; lateral tubercles are present on the
somites close to their junctions with the coxal plates. The tergum of the seventh somite
is partially fused with that of the first abdominal segment, the posterior suture of the
former being incomplete for a short space on either side of the middle line.

The pleural plates of the second and third abdominal segments extend beyond the
coxal plates of the thoracic somites ; those of the third are sHghtly longer than those of
the second and extend to a point just beyond the base of the articulation of the
uropod.

The terminal segment is "almost triangular in shape, and ends in a blunt prolonga-
tion ; the upper surface slopes gently downward on either side from the central keel ; the

DESCRIPTION OF SPECIES 333

lateral margins, commencing from the attachment of the terminal appendages, curve
inwards and then slightly outwards, becoming almost parallel just before their termina-
tion". Colour in spirit: brown.

Nordenstam (1933, pp. 82-5) describes a new variety of the above species which he
names eugetnae. It differs from the main species in the following respects :

(i) "in its more oblong shape of body;

(2) in that the free triangular tips of the pereion segments laterally, from their
spimera are lacking;

(3) in having two lateral tuberculae on the pleotelson;

(4) in having the triangular tip in the middle of the posterior margins of the pereion
segments missing on the first and second segments".

These differences do not seem sufficient to warrant the formation of a new variety ;
consequently I have placed the name in the synonymy of S. mimita.

Occurrence. Port Philip, Australia, 38 fathoms; and off St Francis Island, from
6 to 13 fathoms.

25. Serolis bakeri, Chilton.

S. bakeri, Chilton, 1917, pp. 398, 399, figs. 12-14.

Diagnostic characters. This species, according to Chilton, bears a close resemblance
to S. mimita, Beddard, but specimens are slightly smaller than those of the latter species,
the largest breeding female measuring 6 mm. in length and 5 mm. in breadth : the males
are approximately the same size as the females.

As in S. mimita three rounded tubercles are present on the posterior margin of the
head, but neither the median tubercles nor the lateral tubercles which are seen on the
thoracic somites of the former species are found here.

No definite information is given as to whether the coxal plates of the first three free
thoracic somites are separated from them by sutures ; but since sutures are present in
S. mimita, and Chilton, after mentioning the points in which the two species differ,
states that in all other respects they are alike, it is probably correct to assume that they
are also present in this species.

The tergum of the seventh thoracic somite is partially fused with that of the first
abdominal segment, the posterior suture of the former being obsolete for a short
distance on either side of the middle line. The coxal plates of this somite extend back-
wards through the anterior third of the terminal segment and the pleural plates of the
third abdominal segment reach to the middle of the segment: the terminal segment
bears a conspicuous median keel, and its posterior extremity is narrow and rounded.
The uropoda are described as having " both branches narrow, outer with external margin
obscurely serrate, end dentate or crenate, and with one or two hairs, inner branch a little
longer and with end less distinctly serrate".

Nordenstam (1933, p. 85) regards this species as a variety of S. minuta, Beddard.

Occurrence. Encounter Bay, South Australian coast.

334 DISCOVERY REPORTS

26. Serolis yongei, Hale.

S.yongei, Hall, 1933, pp. 560-1, i text-fig.

Hale's description of this species is extremely short ; it is based on two small specimens
collected outside Trinity Opening, Low Isles, during the Great Barrier Reef Expedition.
Through the kindness of Dr Caiman I have been able to examine these specimens, and
as a result of my observations am able to amplify the original description in a number
of points.

Diagnostic characters. The larger specimen, which is a female in the non-breeding
condition with the four pairs of brood plates well developed, measures 2-9 mm. in
length and 2-5 mm. in greatest breadth. The shape of the body is described as "wide,
pyriform, dorsal surface almost smooth with shallow pitting".

The head is broader than long, about the same width anteriorly as at the level of the
eyes, slightly narrower between these two levels; the anterior margin is excavated on
either side of a fairly long rostrum for the reception of the antennules. The antero-
lateral margins of the head are not completely united with the forward lateral extensions
of the second thoracic somite, which is here, as in all species of Serolis, fused with the
head. The head is roundly convex between the eyes, which are large and are situated
on either side in a postero-lateral position. The eyes are reniform and contain black
pigment.

The first three free thoracic somites (3rd-5th) are almost sub-equal, decreasing
slightly in length from before backwards ; the sixth is about half the length of the fifth.
The seventh thoracic somite is extremely narrow, with its tergum fused with that of the
first abdominal segment for some distance on either side of the median line. Hale makes
no mention of this important fact in the text of his paper, but he illustrates it in his
figure.

The coxal plates are well developed, and those of the first three free somites are
separated from them by sutures (they are not fused as in Hale's figure). Those of the
seventh somite are not much more produced than those of the preceding somites ; they
extend backwards to a point slightly beyond that reached by the pleural plates of the
second abdominal segment, at about the level of the bases of the protopodites of the
uropods.

The abdominal segments are sub-equal, with the pleural plates of the second and
third segments extending for a short distance beyond the broad anterior end of the
terminal segment ; those of the third segment are slightly the longer, and extend back as
far as about the level of the middle of the protopodite of the uropod.

The terminal segment is broader than long, roughly triangular in shape, with the
apex directed backwards, and with the angles rounded ; it is broadest at the level of the
base of the uropods, and has a low median dorsal carina. The sternum of each of the
first three abdominal segments is produced into a median backwardly directed spine
which is largest on the first segment.

The antennules and antennae have already been described. The form of the antennule
is given as "second joint of peduncle nearly half as long again as the fourth, and four

DESCRIPTION OF SPECIES 335

times as long as the third, which is very short ; flagellum of four joints ". This description
of the peduncle fits the condition seen in the larger specimen, but in the smaller one the
third joint is not much shorter than the fourth; the second is the longest in both
specimens. The flagellum consists of six, not four joints. The two distal joints are about
half the breadth of those which precede it ; the penuhimate is short and bears a sensory
seta at its anterior distal angle, the terminal is long and bears three long deUcate setae
at its extremity (cf. a similar arrangement in S. mimita, Beddard). A sensory seta is
present at the anterior distal angle of each of the third and fourth joints.

According to Hale the two terminal joints of the peduncle of the antenna are sub-
equal ; actually, however, the terminal one is considerably longer than the fourth, and is,
in fact, slightly longer than the third and fourth together. Groups of delicate setae
spring from shallow depressions on the anterior margins of the fourth and fifth joints
and one or two delicate plumose setae are also present on the dorsal surface of the fifth
joint. The flagellum consists of eight, not six joints. The terminal one is narrow and bears
a group of long delicate setae at its extremity ; similar setae are present at the anterior
distal end of each of the other flagellar joints except the first.

It is impossible to make out any of the structure of the mouth-parts from the speci-
mens. The second pair of thoracic appendages is modified in the usual way: the inner
margin of the dilated propodus is armed with a row of three-pronged spines alternating
with a row of broad leaf-like structures very similar to, though slightly broader than, the
corresponding structures in S. minuta, Beddard, which are figured in the Challenger
Report (1884, pi. vii, fig. 7). These are flanked by another row of spines, the distal ends
of which appear to be covered with very short closely-set hairs. The dactylus bears a
strong short spine near the distal extremity of its inner border. The carpus is crenulate
and bears two short stout spines, the distal ends of which are obliquely truncate and
covered with short hairs ; a few more delicate spines similar to those which flank the
inner margin of the propodus are also present.

The remaining appendages are all more or less alike, the last pair being somewhat
smaller than the rest ; the setae on these appendages are all simple, long and delicate
(cf. S. mimita, Beddard). The protopodite of the first three pairs of pleopods is com-
paratively long and slender, and, as in the other Australian species, the inner margin is
not produced to form a triangular process bearing plumose setae. The suture of the
exopod of the fourth pleopod is transverse ; short plumose setae are present on its outer
margin, at any rate at the distal end. The endopod is considerably smaller than the
exopod and has a rounded extremity.

The protopodite of the uropod has its inner angle not much produced ; the endopod
is longer than the exopod and extends nearly to the tip of the terminal segment. Both
lobes are long, with their distal rounded ends sparsely fringed with delicate plumose
setae. The colour of the spirit specimen is rather dark brown.

Distribution. Outside Trinity Opening, Low Isles, Great Barrier Reef, 16° 17' S,
146° 2' E, in 200 m.

The species appears to resemble very closely the two Australian species -S. mimita,

336 DISCOVERY REPORTS

Beddard, and S. bakeri, Chilton ; the three are together characterized by the facts that
(i) the tergum of the seventh thoracic somite is fused with that of the first abdominal
segment ; (ii) the coxal plates of the three free somites are separated from them by
sutures, and (iii) the coxal plates of the seventh somite are comparatively short. In
each species the pleural plates of the third abdominal segment extend back farther
than do the coxal plates of the seventh thoracic somite.

The form of the antennule is the same in the three species ; the second joint of the
peduncle is the longest, the penultimate joint of the flagellum very short and the
terminal one long and delicate. The form of the modified spines on the inner margin
of the propodus of the second thoracic appendage is very much the same in all three
species, and in each the setae on the remaining appendages are simple, long and delicate.

If the size of the specimen of .S. yongei (2-9 mm. long, 2-5 mm. broad) is any indica-
tion of that of the adult, this species is considerably smaller than either 5. bakeri or-
S. minuta. The main difference appears to lie in the ornamentation ; the dorsal surface
of the body of S. yongei is practically smooth and there is only a sHght median dorsal
carina on the terminal segment, whereas in S. mimita there are three tubercles on the
posterior part of the cephalosome and one in a median position on each of the thoracic
and abdominal segments. In S. bakeri the head bears three rounded tubercles on its
posterior margin, but these are absent from the thoracic and abdominal segments ; the
terminal segment bears a conspicuous median dorsal carina and its posterior extremity
is narrow and rounded.

The three species are all recorded from the Australian coast ; but S. yongei, from the
Great Barrier Reef, comes from a locality considerably farther north than the other two.
S. bakeri is recorded from Encounter Bay, South Australia, and S. jniniUa from Port
Philip, South Australia, and off St Francis Island.

27. Serolis orbiculata, n.sp. (Figs, i b, e, 14, 15).

Occurrence. St. WS 222: 48° 23' S, 65° 00' W, 100-106 m.; i ? (non-b.).
St. WS 808: 49° 40' 15" 8,65° 42' W, 1 10-107 m.; 1$.
St. WS 809: 49° 28' 15" S, 66° 29' W, 107-104 m.; i ? (b.).
St. WS 813: 51° 35' 15" 3,67° 16' 15" W, 106-92 m.; i ? (b.).
St. WS815: 51° 51' 45" S, 65°44' W, 132-162 m.; 3 ?? (b.), 3 (non-b.), and i <?.

Description. The body (Fig. 14) is broadly ovate, not much longer than broad;
the male is 8-5 mm. in length and 8 mm. in greatest breadth, and the largest female
10-5 mm. in length and 9 mm. in greatest breadth. The male, as in so many species of
Serolis, is slightly broader in proportion to its length than is the female.

The surface of the body is smooth, sparsely covered with small microscopic conical-
shaped projections. The colour of spirit specimens is brownish towards the centre of
the body, paler towards the margins, the whole covered, except for the anterior margins
of the somites and their coxal plates, with small spots of black pigment.

The head is convex dorsally, shield-shaped, broader than long, with its greatest
breadth anteriorly. The anterior margin is excavated on either side of a short median

DESCRIPTION OF SPECIES

337

Fig. 14. Serolis orbiculata, n.sp., $: x lo.

rostrum for the reception of the antennules, and behind the rostrum is a low transverse
ridge extending to the lateral margins of the head. The eyes, which are oval or slightly
reniform in shape and contain black pigment, are situated on raised areas towards the
postero-lateral angles of the
head; the area between the
eyes is hollowed out and the
lateral portions of the cepha-
losome are smooth.

The third and fourth (first
two free) thoracic somites are
sub-equal, the fifth not quite
as long, the sixth about half
the length of the third; the
seventh is fused dorsally with
the first abdominal segment
for a considerable distance on
either side of the middle line.

The coxal plates are well
developed, those of the first
three free somites are separ-
ated from them by sutures ; those of the seventh somite extend backwards to about the
middle of the protopodite of the uropod and just beyond the middle of the terminal
segment: those of the male are very slightly longer than those of the female.

The first three abdominal segments are broad, with a length equal to that of the fifth
thoracic somite ; the pleural plates of the second and third segments are short and do
not extend beyond the anterior margin of the broad terminal segment. Each of the
sternal plates of these segments has the median posterior border produced into a small
spine, increasing in length from the first to the third. The terminal segment is triangular
in shape, with the angles approximately equal and well rounded, and with the apex
directed backwards ; on its dorsal surface is an ill-defined median dorsal carina, on either
side of which, placed at some distance from it, is a short lateral carina. This is made
more conspicuous by the presence of a line of deeper pigmentation.

The antennule is less than half the length of the antenna ; it consists of a peduncle of
four and a flagellum of eight joints. The first peduncular joint is short ; the second and
third are both about twice the length of the first, whilst the fourth is shorter, being only
about one-third the length of the third : the second and third joints are sparsely fringed
with delicate setae. The flagellum consists of seven joints of about equal length, but
narrowing distally, and a very minute terminal joint which bears a single sensory and
several simple setae ; a sensory seta is present at the anterior distal angle of all except the
first two joints of the flagellum.

The antenna consists of a peduncle of five and a flagellum of six joints. The first
peduncular joint is short, broader than long, with its inner border fringed with short

338

DISCOVERY REPORTS

hairs ; the fourth joint is about three times the length of the second, and the fifth over
twice as long as the fourth ; bundles of fine setae arising from round knob-like bases are
scattered over the ventral surfaces of the third, fourth and fifth joints. The flagellum is
not much longer than the fifth peduncular joint ; its first two joints are sub-equal and
twice as long as each of the remaining four ; the terminal joint bears a bunch of fine setae
at its extremity.

The mouth-parts are built up in the normal way. The cutting edge of the left mandible
has its primary edge divided into three teeth ; the secondary one is in the form of two
spines, one of which is truncate with a broad rounded extremity: the secondary cutting
edge of the right mandible is in the form of two spines, the inner one of which is larger
than the outer.

Fig. 15. Serolis orbiculata, n.sp.

a, maxillipedofcJ: x 75. 6, second thoracic appendage of t?: x 35. c, third thoracic appendage of
adult cj: X 50. </, fourth thoracic appendage of cj : x 50. e, eighth thoracic appendage of cJ : x 60.
/, uropod: x 50.

The maxillula (Fig. i b) has nine spines on the truncate extremity of its outer lobe ;
its inner lobe is small, curved, not expanded distally, and with a small seta at its rounded
end. The maxilla (Fig. i e) resembles that of S. nototropis, n.sp., and differs from that
of all other species in having only one articulating lobe. This lobe (L^) is less well de-
veloped than in S. nototropis and bears only two pectinate setae. The lobe is continuous
with the segment from which it arises, and the latter appears also to be continuous with

DESCRIPTION OF SPECIES 339

the lamella of the second segment (L-) : in some preparations a faint line, shown by
the dotted line in the figure seems to indicate the boundary of the plate-like portion
of the second segment (for discussion on the morphology of this appendage see
p. 272).

The maxilliped (Fig. 15^) closely resembles the type seen in S. nofotropis, in which
the union of the basipodite with its endite (B') and lamella (£") extends for some distance
beyond the base of the endopod, the latter springing from the centre of a plate-hke struc-
ture. The groove separating the endite {B') from the lamella (£") is slightly deeper than
that of S. nototropis. The endopod consists of the usual three joints ; the second is almost
parallel-sided, narrowing slightly towards its distal end, the third joint is small and both
are fringed with setae along their inner margins. The forms of the second, third, fourth,
and last thoracic appendages of the male are shown in Figs. 15 b-e. The second is stout,
with the propodus expanded and bearing a row of about fourteen modified spines
alternating with pecuHar processes, the form of which is shown in Fig. 156. This row
is flanked by another row of modified spines of the type shown in the figure, and beyond
these, and alternating with them, is a row of simple setae. The distal edge of the carpus is
crenulate and bears three setae in form similar to, but smaller than those flanking the row
on the propodus. The third thoracic appendage (Fig. 15 0 is considerably longer than
the second ; the propodus is broad and has its inner margin armed with a double row
of nine modified spines ; the other joints of the appendage, as well as those of the
remaining ones, bear a series of short transverse rows of long setae, some of which,
towards the distal ends of the merus, carpus and propodus, are pectinate. The last
appendage, in both sexes, only diff'ers from the remaining ones in being smaller.

The protopodite of the first three pairs of pleopods is not produced as is usually the
case; both margins are fringed with long delicate setae. The appendix mascuHna on the
second pleopod of the male reaches three-quarters of the way towards the apex of the
terminal segment. The exopod of the fourth pleopod is divided by a transverse suture ;
it is fringed with long plumose setae on its outer margin.

The protopodite of the uropod (Fig. 15/) is exceptionally long, being considerably
longer than the endopod ; its inner distal angle is not produced as is usually the case.
The exopod and endopod are narrowly oval in shape, the latter being longer than the
former ; both are fringed along their inner margins with long plumose setae, which are
continued for a short distance on to the margin of the protopodite. The outer margins
of both the protopodite and the exopod are fringed with delicate setae.

This species resembles most closely S. nototropis, n.sp., but can be readily distinguished
from it by the shape of the body and terminal segment, by the form of the modified
spines on the propodus of the second thoracic appendage, as well as by the difference in
length of the protopodite of the uropod. The two species resemble each other and diff'er
from all other species in the form of the maxilliped, and the absence of the second
articulating lobe in the maxilla.

Distribution. The species occurs off the coast of South America to the north-west
of the Falkland Islands, in depths of from 92 to 162 m.

34°

DISCOVERY REPORTS

28. Serolis nototropis, n.sp. (Figs, id, 16, 17).

Occurrence. St. WS 796: 47° 49' 37" S, 63° 42' 30" W, 106 m.; 3 ?? (b.).
St. WS 797: 47° 45' i8" S, 64° 10' 30" W, 117 m.; I 9 (b.).
St. WS813: 51° 35' I5"S, 67° 16' 15" W, 106 m.; i ? (b.), 2 immature (J^.
St. WS815: 51° 51' 45" S, 65° 44' W, 132-162 m.; I (J.
St. WS 816: 52° 09' 45" S, 64° 56' W, 150 m.: I ?.
St. WS837: 52° 49' i5"S, 66°28'W, 98 m.; 13.

Description. The body (Fig. 16) is longer than broad, almost oval in shape, but
narrowing posteriorly; the dorsal surface is arched, with a pronounced median keel
which is better developed in the male than
the female. The surface of the body is
slightly pitted and covered with rows of
conical papillar projections (Fig. ly a).
The only adult male measures 1 1 mm. in
length and 8 mm. in greatest breadth,
the largest female 12 mm. in length and
9 mm. in greatest breadth, so that in this
species the ratio of length to breadth is
approximately the same for the two sexes.

The head is broader than long, shield-
shaped, with the greatest width anteriorly.
A short rostrum is present, on either side
of which the anterior margin is excavated
for the reception of the antennules, and
immediately behind this margin a trans-
verse ridge extends across the width of
the head. The eyes, which are situated
on raised areas, are placed rather close
together near the posterior border of the head ; the area of the latter between the eyes
and the anterior transverse ridge is raised. The eyes are small, oval in shape, and contain
black pigment.

As in all the Serolidae the second thoracic somite is fused with the head ; the lateral
portion is smooth, but in the middle line, behind the head, the surface is raised into a
small dorsal spiniform process. The third and fourth thoracic somites are sub-equal,
the fifth and sixth are together equal to the third, whilst the seventh is equal to the sixth
but is fused with the first abdominal segment for a short distance on either side of the
middle line. Small median spiniform processes, decreasing in size from before back-
wards, are present near the posterior borders of the third, fourth and fifth somites. The
coxal plates of the third to the seventh somites are well developed and those of the third,
fourth and fifth are separated from their respective somites by sutures ; in both sexes
those of the seventh somite are more produced than those of the other somites and
extend backwards to about the level of the base of the uropod.

Fig. 16. Serolis nototropis, n.sp.,?: x 10.

DESCRIPTION OF SPECIES 34i

The abdomen consists of three free segments and a large terminal one. Each of the
first three is slightly keeled and bears a small spiniform process corresponding in position
with those of the thoracic somites ; the pleural plates of the second and third abdominal
segments are well developed, but do not extend backwards quite so far as the coxal
plates of the seventh thoracic somite ; those of the third segment are slightly the longer.
Each of the sternal plates of these segments has the median posterior border produced
into a spine, increasing in length from the first to the third.

The terminal segment (Fig. i6) is roughly triangular in shape, with the apex directed
backwards and with the basal angles broadly rounded. On its dorsal surface is a strong
median carina which extends to the end of the upturned posterior extremity ; this carina
is broadest anteriorly where there is a small oval pit in a median position on its dorsal
surface. Two pairs of lateral carinae are also present ; one on either side runs parallel
with the lateral margin of the segment but a short distance within it; the other lies
almost midway between the latter and the median carina, with which it runs almost
parallel, meeting the outer lateral carina at a point level with the posterior end of the
coxal plate of the seventh thoracic somite.

The antennule is short and consists of a peduncle of four joints, and a flagellum,
which in the male comprises ten joints, in the female seven. The first peduncular joint
is not much longer than broad ; the second is about the same length with its postero-
distal angle slightly produced ; the third joint is more than twice as long as the second,
broad proximally, but narrowing distally ; the fourth is about half the length of the third.
The flagellum is not quite as long as the third peduncular joint ; its joints, except for the
terminal one, which is minute, are all of approximately the same size : the third to the
tenth each bear a sensory seta of the usual type, placed near the anterior distal angle.

The antenna is nearly twice as long as the antennule: both the peduncle and the
flagellum consist of five joints. The first joint of the peduncle is short, broader than long;
the second is about three times the length of the first, with its anterior distal angle pro-
duced, and with the proximal part of the anterior margin covered with short hairs; in
the third peduncular joint the anterior margin is equal in length to that of the second
joint, but with its posterior margin curved, and considerably shorter; this joint is
broadest distally, has its anterior distal angle somewhat produced, and short hairs are
present on its posterior margin. The fourth peduncular joint is twice the length of the
third, with a dorsal longitudinal ridge extending along its upper surface near the
posterior margin; the fifth joint is nearly twice as long as the fourth but considerably
narrower. On the ventral surfaces of the third, fourth, and fifth joints is a series of
short transverse rows of setae which are arranged in bundles ; each bundle consists of
about five long dehcate setae arising from a common basal "knob", and the transverse
rows, each of which contains three or four of these bundles, extend from about the
middle of the ventral surface as far as the anterior margin. There are two transverse
rows on the third, five on the fourth and seven on the fifth joint.

The mandibles are built on the usual plan and their cutting edges are comparatively
simple ; that of the right mandible is divided into four by three shallow notches, with

342

DISCOVERY REPORTS

the secondary blade in the form of a single toothed spine. The cutting edge of the left
mandible is narrower than that of the right and is undivided ; its inner secondary blade
is in the form of two spines, the larger one of which has a bifid extremity.

The outer lobe of the maxillula bears ten stout spines, and a single more slender
pectinate one, on its truncate distal extremity. The maxilla (Fig. i d) differs from that of
all the other species of Serolis except S. orbicidata, n.sp., in possessing only one articu-
lating lobe instead of two. This lobe bears four pectinate setae at its distal extremity,
and fourteen similar setae are present on the distal end of the fixed lobe. There appears
to be no suture between the third joint of the primary axis and the lamella of the second
joint (for discussion on the morphology see p. 272).

b.

Fig. 17. Serolis nototropis, n.sp.

a, two joints of the antennal flagellum, to show the conical scales, b. maxilliped of ^: x 50. c, second
thoracic appendage of (J: x 64. d, third thoracic appendage of (J: x 90. e, fourth thoracic appendage of
(J: X 67. /, uropod: x 83.

The type of maxilliped met with in this species (Fig. 17 6) is intermediate between
that of S. exigua, Nordenstam, and that of S. schythei, LiJtken. The endopod springs from
the middle of a plate-like structure consisting of the basipodite [B) plus its endite {B')
and its lamella (£") : distally the endite and the lamella are separate for a short distance,
but the division does not extend as far as the articulation of the endopod. The endopod
is comparatively slender, and its distal joint is small and fringed on its inner margin
with long setae ; similar setae are also found on the distal half of the inner margin of the
second joint.

DESCRIPTION OF SPECIES 343

The form of the second, third and fourth thoracic appendages is shown in the figures
(Fig. 17 c-e). The second appendage is stout; the propodus is oval in shape, and bears
on its inner margin a row of about thirty-six rather delicate spines which appear to be
simple but with a high magnification show fine, short pectinations near their distal ends.
The spines are of two lengths, the longer being slightly stouter and alternating with the
shorter type. The species agrees with S. exigiia and differs from all other species in
having only one variety of modified spine on this joint. At the proximal end, the inner
surface is considerably raised into a median prominence, and its steep face (towards the
distal end of the joint) is covered with very short hairs ; the dactylus extends for some
distance beyond the flat face of this raised area.

The distal truncate extremity of the carpus bears two stout spines, each with a bifid
extremity, and three or four delicate setae. The third appendage of the male (Fig. 17 d)
is longer but less robust than the second. The propodus is broad and its inner margin
is armed with a double row of five modified setae of the usual type ; near its proximal
end is a small group of short hairs and at the outer angle of its truncate distal extremity
is a group of five long setae : the dactylus terminates in a stout spine at the base of which
are a few short hairs. The setae present on the remaining joints as well as those of the
other appendages are long, usually simple, and arranged in short transverse rows in
notches on the margins of the joints (Fig. 17 e). The last thoracic appendage is similar
in the two sexes ; it is shorter than the others, and its joints are relatively broader in
proportion to their length.

The protopodite of the first three pairs of pleopods is somewhat triangular in shape
with the produced angle bearing long plumose setae ; there are three of these on the first
and two on each of the second and third pleopods. The appendix masculina on the second
pleopod of the male reaches about two-thirds of the way towards the apex of the terminal
segment. Both the exopod and endopod of the fourth pleopod are divided by a trans-
verse suture ; the outer margin of the exopod is fringed with plumose setae.

The protopodite of the uropod (Fig. 17/) is rather long, with its inner angle produced
and bearing a short plumose seta; the exopod is shorter than the endopod, and both
are fringed along their inner margins with long plumose setae which, however, do not
extend quite to their distal ends.

Distribution. This species was collected from localities north and west of the
Falkland Islands, in water from 98 to 150 m. in depth.

29. Serolis pagenstecheri, Pfeffer.

S. pagenstecheri, Pfeffer, 1887, p. 73, pi. ii, figs, i, 2, pi. iv, figs. 1-3; Tattersall, 1921, p. 231 ; Monod,

1931, p. 26; Nordenstam, 1933, pp. 85-8, text-figs.
S. pagenstecheri var. albida, Nordenstam, 1933, pp. 88-9, pi. i, figs. 4, 5, text-fig. 21 e.

Occurrence. St. 39: South Georgia, 179-235 m.; i ? (b.), 2 3S (immature).

St. 42: South Georgia, 120-204 m.; i ? (b.); 2 ?? (non-b.), 2 immature <J<J, and a number of

small specimens.

St. 140: South Georgia, 122-136 m.; a number of immature specimens.

St. 144: South Georgia, 155-178 m.; 4 immature specimens.

344 DISCOVERY REPORTS

St. 149: South Georgia, 200-234 ^i-; i immature specimen.

St. 157: South Georgia, 970 m.; i $ (b.).

St. WS 62: South Georgia, 26-83 m.; 2 ?$ (b.), 2 33, 4 immature specimens.

St. MS 10: East Cumberland Bay, 26 m.; 2 ?? (non-b.), 3 immature specimens.

St. MS 67: East Cumberland Bay, 38 m.; i immature specimen.

Beach, Grytviken, South Georgia; i $ (non-b.).

Diagnostic characters. The largest male in the collection measures 37 mm. in
length and 34 mm. in greatest breadth ; it is slightly broader in proportion to its length
than the female, the largest of which measures 38 mm. in length and 32 mm. in breadth.
The colour of the living animal is given as dark grey above and white beneath, with the
gonads often showing through as orange-red masses.

The body is broadly ovate, with its dorsal surface covered with scattered hairs which
are most numerous along the margins of the segments. The head is broadest anteriorly,
with a short median rostrum, behind which is a transverse ridge extending laterally to
the sides of the cephalosome. Behind this again is another transverse ridge extending
between and in a line with the anterior end of the eyes ; it is lower in the middle than at
the sides and immediately posterior to it is a deep groove. The portion of the head
between the eyes is strongly convex, and divided into three prominences : in these the
posterior margins are much more sharply defined than the anterior which gradually
merge into the surface of the head. The lateral prominences are oval and larger than the
central one, the latter being in the form of a triangle with the apex directed forwards.
The eyes are about half the length of the head, rather narrow, reniform in shape, and
containing black pigment.

The lateral portion of the second thoracic somite, which is fused with the head, has
a transverse ridge on each side, commencing at a point in line with the middle of the
length of the eye, and running transversely and then slightly backwards, fading away
into the lateral margins ; a second short ridge, commencing on either side on the anterior
transverse ridge, half-way along the course of the latter on the cephalosome, extends
backwards and slightly outwards for a short distance.

The first three free thoracic somites are sub-equal, with their posterior borders
slightly raised to form small median spiniform processes which increase in size from
before backwards. The sixth and seventh thoracic somites are narrow, together equal to
the preceding one ; the posterior border of the seventh is obsolete in the middle line, so
that for a short distance the tergum of the seventh somite is fused with that of the first
abdominal segment. The third to the seventh thoracic somites have well developed coxal
plates, each of which bears a well marked groove which follows the outline of its anterior
margin a short distance within it. The coxal plates of the first three free somites are
separated from their respective somites by sutures: those of the seventh somite are
produced backwards beyond the pleural plates of the abdominal segments to a point
just beyond the level of the articulation of the uropods ; they are longer in the male than
in the female.

Each of the three free abdominal segments has a short median dorsal spiniform
process ; the pleural plates of the third segment extend backwards to the level of the

DESCRIPTION OF SPECIES 345

articulation of the uropods, and those of the second somite are shorter than those of the
third. The terminal segment is roughly five-sided, with rounded angles and with a
truncate posterior extremity. It has a well developed spine in the anterior mid-dorsal
line, followed by a low median keel which is more pronounced towards the posterior
extremity; on each side of this keel are two lateral longitudinal keels, terminating in
small spines some distance from the lateral margins which are slightly bent downward.
The uropods do not extend to the tip of the terminal segment : the exopod is about two-
thirds the length of the endopod ; both are elongated oval in shape, with their margins
fringed with long plumose setae.

Remarks. The mouth-parts are of the normal type. The outer lobe of the maxillula
bears eleven spines on its distal extremity : the inner lobe differs from the usual type in
having the distal third of its length much expanded so as to be almost circular in outline ;
it bears a single small curved spine. The lamella of the maxilliped is separated from the
basipodite by a suture. A point not previously noted in this appendage is that at the
distal outer extremity of the third joint of the palp there is a groove in which a very
small fourth joint is articulated. This joint does not extend quite as far as the distal end
of the third, and it carries five long setae on its truncate extremity. A similar joint has
been observed on the maxillipeds of S. kempi, n.sp., and S. platygaster, n.sp. (Fig. 19 a).
The modified spines on the propodus of the second thoracic appendage are very similar
to those figured for S. beddardi, Caiman (Fig. 3 b).

According to Pfeffer the third thoracic appendage of the male ' ' resembles those of the
following segments and does not show the cheliform formation present in other species ;
the only outstanding character is the slightly vigorous formation of the penuhimate
joint". It is obvious that the males in Pfeffer 's collection were all immature, for in the
adult male this appendage shows a typically cheliform structure. The propodus is
broad and bears on its inner border a double row of eleven strong curved spines, each
of which is covered with dense short hairs ; the axis of the spine extends beyond the
hairs as a delicate curved process and between these spines are a few simple setae. The
inner border of the carpus, merus, ischium and distal half of the basipodite is densely
clothed with short fine hairs. Similar hairs occur on the propodus, carpus, merus and
ischium of the eighth thoracic appendage of the adult male ; they are found on both inner
and outer margins, but are more abundant on the former ; in the latter position, on the
merus and the ischium, they are restricted to the distal half. The joints of this appendage
are broader in proportion to their length than those of the other appendages.

Nordenstam (1933, p. 88) describes a new form of S. pagemtecheri, a variety albida.
The variety is smaller in size than the typical form, and of "slightly yellowish, almost
whitish" colour. The only other difl'erence between this and the typical species is that
"the male has a slightly more oblong shape of body than the main species ". The shape
of the female is the same in both forms. The differences in size and colour are not
sufficiently dependable characters on which to base the formation of even a variety, and
since the shape of the female is the same in both I am unable to admit the validity of the
varietal name albida.

346

DISCOVERY REPORTS

Distribution. Pfeffer's and Tattersall's specimens came from South Georgia, as do
those in the present collection.

30. Serolis platygaster, n.sp. (Figs. 18, 19).

Occurrence. St. 146: 53° 48' 00" S, 35° 37' 30" W, 728 m.; i adult $.

Description. The specimen (Fig. 18) measures 33 mm. in length and 27 mm. in
greatest breadth ; it is broadly ovate in outline and compact in form. The head is broadest
anteriorly and widens in front of the eyes,
the distance between the eyes being three-
fifths of the length of the anterior margin.
A small but distinct rostrum is present ; a
short distance behind it is a transverse ridge
which meets the boundary of the head
midway between the eyes and its anterior
margin. On either side of the rostrum,
the anterior margin is excavated for the re-
ception of the base of the antennule and a
groove running within the margin extends
laterally to the sides of the head. Behind
the transverse ridge, and between it and
the eyes, the middle portion of the head is
convex, whilst the area between the eyes
is also raised and divided into two large
rounded prominences, the posterior mar-
gins of which are better defined than the
anterior. In an anterior position between
these prominences is a small triangular
area, with its apex directed backwards.

Fig. 18. Serolis platygaster, n.sp., cJ: x 4-

which is more deeply pigmented than is the rest of the body. The eyes are very small
in proportion to the size of the body, postero-lateral in position, about a third of the
length of the head, and very narrow ; they contain black pigment.

As in all Serolidae, the second thoracic segment is united with the head. The lateral
portion bears a low ridge extending from near the posterior angle of the eye ; it runs
first in a transverse direction outwards, and then branches, with one branch curving
forwards and disappearing on the antero-lateral margin of the head, the other curving
backwards, running almost parallel with but some distance from the margin of the
cephalosome, and disappearing near the posterior border. The third, fourth, and fifth
thoracic somites are sub-equal, slightly keeled, with the fifth somite raised near its
posterior margin into a small median dorsal protuberance. The sixth and seventh
somites are sub-equal, together equal in length to the preceding somite: the posterior
sutures of the terga of both somites are absent for a short distance on either side of the
mid-dorsal line, so that for a short distance they are fused with each other and with the

DESCRIPTION OF SPECIES 347

first abdominal segment ; the position of the former suture between the two thoracic
somites is indicated by a slight groove. In this respect the species agrees with the con-
dition seen in the female of S. aspera, n.sp. (in the male the groove is absent), and is
intermediate between S. pagenstecheri, Pfeffer, where only the suture between the seventh
thoracic and first abdominal segment is obsolete, and 5. boiwieri, Richardson, and the
male of S. aspera, where fusion is complete and no groove indicates the position of the
former suture. The coxal plates, which are rather short and closely applied together for
their whole length, are developed on each of the third to the seventh thoracic somites ;
those of the third to the fifth are separated from their respective somites by sutures.
A low prominence is present on either side of each somite near its posterior border and
just within the point of union with it and its coxal plate : a groove is present on the dorsal
surface of each coxal plate, running parallel with, though a little distance within its
anterior margin. Long delicate setae fringe the inner lateral margins of both the coxal
and pleural plates. The coxal plates of the seventh thoracic somite extend back to a level
corresponding with that of the bases of the uropoda, and slightly beyond that reached
by the pleural plates of the third abdominal segment. The pleural plates of the second
abdominal segment are a little shorter than those of the third.

The terminal segment is somewhat pentagonal in outline, terminating posteriorly in
a blunt point which is slightly uptilted. The base of the pentagon also forms the base of
a somewhat raised, triangular area, which bears a low median keel and extends to about
the middle of the segment; the posterior quarter is also slightly keeled. Two longi-
tudinal ridges, one on either side, lie near the lateral margins of the segment.

The peduncle of the antennule is four-jointed ; the first two joints are small, the third
one and a half times the length of the second, the fourth about a third of the length of the
third. Delicate setae are present on the anterior margin of the first joint and along the
anterior and posterior margins of the second and third joints. Neither flagellum is com-
plete, but one consists of twenty-five joints, the first eight of which do not bear sensory
setae.

The first peduncular joint of the antenna is short ; the second is three times the length
of the first and is fringed with delicate setae ; the third is about two-thirds the length of
the second with tufts of setae on its posterior margin ; the fourth and fifth are each three
times the length of the third, the fifth being narrower than the fourth. The incomplete
flagellum consists of fourteen joints.

The mandibles are of the usual type ; the primary cutting edges of both are strongly
chitinized. The maxillula is of the usual type, with a stout outer lobe bearing eight stout
spines and an inner more delicate one on its obliquely truncate distal extremity. The
inner lobe is two-thirds the length of the outer and is flattened distally to form an oval
plate, the outer margin of which is more curved than the inner. The distal ends of both
outer and inner articulated lobes of the maxilla are provided with two strong setae which
are minutely pectinate on their inner margins ; the fixed lobe has eight large and about
twelve smaller setae at its distal end.

The basipodite of the maxilliped (Fig. 190) is separated from its lamella, which is

348

DISCOVERY REPORTS

slightly broader than long, by a suture. The second joint of the palp is broadest distally,
with its inner margin fringed with setae; the third joint is more or less parallel-sided,
with its inner distal end rounded and fringed with setae. Near the extremity of this
joint, at its outer angle, is a fourth small joint ; the distal extremity of this joint does not
extend beyond that of the third, and it is truncate, bearing six long setae. Similar joints
are present in a corresponding position on the maxillipeds of several other species.

Fig. 19. Serolis platygaster, n.sp.
a, maxilliped: x 14. b, second thoracic appendage: x 14, c, third thoracic appendage of (J: x 14.

£f,uropod: x 24.

The second thoracic appendage has its propodus greatly expanded, with its inner
margin armed with a row of about thirty-five broadly oval processes alternating with
modified spines (Fig. 19 b). As is usual in the adult male the third thoracic appendage
is modified (Fig. 19 c). The proximal angle of the expanded propodus is produced
so that the inner margin appears to be considerably hollowed out ; this margin is armed
with a double row of modified spines (see figure), three pairs of which are situated
on the inner surface of the produced angle, whilst a further six pairs are arranged along
the remaining part. The dactylus is shorter than the propodus and terminates distally
in two short spines, one of which is larger and stronger than the other ; the inner margins
of the carpus and merus are densely fringed with short hairs. Most of the thoracic
appendages are incomplete, the last does not differ except in size from those that pre-
cede it.

The protopodite of the first three pairs of pleopods is triangular in shape with the
produced angle bearing plumose setae, three on the first, and two on the second and

DESCRIPTION OF SPECIES 349

third. The appendix mascuHna on the second pleopod reaches about two-thirds of the
way towards the apex of the terminal segment. The suture of the exopod of the fourth
pleopod is slightly oblique and the outer border is fringed with plumose setae; the
endopod is not divided by a suture, and does not bear setae ; a suture extends for a short
distance across the exopod of the fifth pleopod.

Unfortunately the right uropod is damaged, but if the left is complete, as it appears
to be (Fig. 19 d), it resembles that of S. beddardi, Caiman, and S. latifrons, White, in
the absence of the endopod. It differs from them, however, in having the protopodite
and its produced spine short, not produced into a long spiniform process. The exopod
is very small, oval in shape and 1-5 mm. in length, with its outer margin and rounded
end notched and bearing a few delicate setae.

This species most closely resembles S. boiivieri, Richardson, but can be readily dis-
tinguished from it by the fact that in the latter the outer edges of the coxal and pleural
plates and also the margins of the terminal segment are greatly thickened, and that the
posterior region of the head is raised into two prominent tubercles which meet in the
middle line. In the latter species, also, the eyes are well developed and contain black
pigment, whereas in the former the eyes are small and very narrow.

Distribution. 53° 48' 00" S, 35° 37' 30" W, north-east of South Georgia.

31. Serolis bouvieri, Richardson (Fig. 20).

S.bouvieri, Richardson, 1906 «, p. 7, text-figs. 12, 13, pi. i, fig. i; Nordenstam, 1933, pp. 89-90,
text-figs. 5 a, 11 a.
Occurrence. St. 181 : Palmer Archipelago, 160-335 "^-S 2 immature specimens.

St. 190: Palmer Archipelago, 93-130 m.; i adult cJ and i ? (non-b.).

St. 190: Palmer Archipelago, 315 m.; i ? (non-b.).

St. 195: South Shetlands, 391 m.; i adult cJ, 16 immature specimens.

Diagnostic characters. The body is oval, longer than broad, with the whole sur-
face covered with scattered shallow pits. On either side of the long rostrum the head is
excavated for the reception of the antennules, and the posterior region is raised into two
prominent tubercles which meet in the middle line. The eyes are placed one on the
outer side of each tubercle ; they are reniform in shape and contain black pigment.

The postero-lateral angles of the coxal plates of the thoracic somites are produced
backwards and have their distal ends rounded and thickened, a thickening which gives
the species a very characteristic appearance. The terga of the first three free thoracic
somites are separated by sutures from their respective coxal plates ; the terga of the
sixth and seventh thoracic somites are fused with each other and with the tergum of the
first abdominal segment in the mid-dorsal line.

The terminal segment is a little broader than long, with its greatest width towards
the middle of the segment ; it narrows posteriorly to about a third of its width. The
rounded postero-lateral angles are thickened and are produced for about 1-5 mm.
beyond the central tip of the segment, the median dorsal carina of which extends as a
short projection. In some specimens a small lateral carina is present, one on either side
of the median; this is less noticeable in large specimens. The uropoda are very small,

DVII 13

3SO DISCOVERY REPORTS

but both exopod and endopod are developed. In the male the coxal plates of the seventh
thoracic somite extend posteriorly to a short distance behind the attachment of the
uropod ; in the female they are not as long: in both sexes they extend beyond the pleural
plates of the second abdominal segment, but in the female they only reach as far as the
extremities of the third.

Remarks. The specimens were collected to the south of the South Shetland Islands
at depths of from 93 to 335 m. ; the largest adult male measures 41 mm. in length and
3 1 mm. in breadth and is considerably larger than any of Richardson's specimens. The
largest specimen, which was collected at a depth of 20-40 m., measures 35 mm. in
length and 25 mm. in breadth.

The appendages, with the exception of the antennule and the antenna, are not
described in the original paper: a few additional observations are therefore included
here.

The antennule (Fig. 20 a) consists of a peduncle of four and a flagellum of thirty-six
joints. The first joint of the peduncle is geniculate, and bears on its dorsal surface a well-
marked ridge, which has its anterior distal angle produced into a strong spiniform pro-
cess. The second joint is two-thirds the length of the first and has a similar dorsal ridge,
which, however, is produced to form two more processes, one midway along its posterior
margin, and the other, on its anterior margin, slightly nearer the distal extremity. The
third joint is about equal in length to the first, but is slightly curved and only half its
width ; the fourth is a quarter the length of the third and is broader at its distal than at
its proximal end. Long delicate hairs are present on the anterior margins of the first
two, and on the posterior of the second peduncle joints. The first joint of the flagellum
is about half the length of the last peduncular joint; the second and succeeding joints are
half the length of the first, the length increasing towards the distal end of the flagellum ;
the terminal joint is equal in length to the penultimate, but is only half its breadth.
A sensory hair is present at the distal end of each of the last twenty joints.

The antenna is made up of the usual five peduncular joints and a flagellum of nine-
teen. The first joint of the peduncle is short ; the second is three times the length of the
first with short hairs on its anterior margin, whilst the third is the same length as the
first. A rounded tubercle is present on the dorsal surface at the distal end of both second
and third joints and is larger on the former than the latter. The fourth and fifth joints
are each twice as long as the second, but the fourth is considerably broader than the
fifth ; the flagellum is one and a half times as long as the fifth joint of the peduncle.

The mouth-parts are of the usual type ; the form of the cutting edges of the mandibles
may be seen from Fig. 20 b. The outer lobe of the maxillula bears ten stout spines and
an inner, more delicate, pectinate one on its truncate distal extremity. There are two
pectinate setae on the outer, three on the inner and about fifteen on the fixed lobe of the
maxilla.

The basipodite of the maxilliped is separated by a suture from the lamella. The second
joint of the palp is broad ; the inner and outer margins of the third joint are almost
parallel, but at about one-quarter of the distance from the distal end, the outer margin

DESCRIPTION OF SPECIES

3SI

is notched for the reception of a fourth joint which is short and narrow, not extending
to the end of the third joint and bearing six long setae at its distal end. A corresponding
joint is found on the palp of S. platygaster, n.sp. (Fig. 19 a), S.pagenstecheri, Pfeffer,
and S. gerlachei, Monod.

a 5^ "^Z

Fig. 20. Serolis bouvieri, Richardson.
a, peduncle and ten flagellar joints of antennule: x 20. b, cutting edges of mandible: x 15. c, second
thoracic appendage: x 7. d, third thoracic appendage of adult cJ: x 20. e, fourth thoracic appendage of tj:
X 6. /, last thoracic appendage of cJ: x 20.

The second, third, fourth and eighth thoracic appendages of the male are shown in
Figs. 20 c-f. The propodus of the second appendage has its inner margin armed with a
row of thirty-six broadly oval processes, alternating with peculiar modified spines. The
distal edge of the carpus is crenulate and it bears two stout spines, each with a bifid
extremity. In the third thoracic appendage of the male (Fig. 20 d) the propodus is slightly
concave, broader at its proximal than at its distal end ; the broad end bears a group of six
pairs of stout spines ; three or four pairs are arranged along the proximal two-thirds of
its surface, and a few simple setae are also present. The spines are of the type shown in
the figure. The dactylus is rather short, curved, terminating in two spines, one of which
is larger than the other ; the inner margins of the carpus and the merus are fringed
with numerous fine hairs, which also extend along part of the margin of the ischium.
The last thoracic appendage is much smaller than the preceding. In the male (Fig.
20/) the dactylus is short and curved, and the inner margins of the propodus, carpus
and merus bear a dense fringe of long delicate hairs ; shorter hairs also occur on the
outer margins of these joints as well as on the inner margin of the ischium. The setae

13-2

352 DISCOVERY REPORTS

on the remaining appendages are stout and simple. The entire surfaces of the appendages
are covered with minute rounded imbricating scales.

The first three pairs of pleopods are of the type in which the protopodite is triangular,
with its produced angle bearing plumose setae— three in the first pair, and two in each
of the other two. The suture of the exopod and of the endopod of the fourth pleopod is
almost transverse; the outer margin of the former is fringed with plumose setae. The
inner distal angle of the protopodite of the uropod is produced and bears strong plumose
setae at its tip. The exopod is a little more than half the length of the endopod ; both are
small and fringed with delicate hairs amongst which, at the distal and outer margins, are
strong plumose setae.

Distribution. Richardson's specimens were collected at Booth- Wandel Island and
the Bay of Flanders ; those in the present collection come from the Palmer Archipelago
and the South Shetland Islands.

32. Serolis aspera, n.sp. (Figs. 21, 22).

Occurrence. St. 39: South Georgia, 179-235 m.; i immature specimen.
St. 123: South Georgia, 230-250 m.; i immature specimen.
St. 148: South Georgia, 132-148 m.; 3 immature specimens.
St. 160: Shag Rocks, 177 m.; i adult S, i ? (b.).

Description. The length of the aduh male and female is the same (26 mm.), but the
former is slightly broader than the latter and measures 21 mm. in greatest width as com-
pared with 20 mm. The body (Fig. 21) is
broadly ovate in shape, with its dorsal sur-
face covered with shallow pits. The colour of
spirit specimens is variegated, being partly
grey and partly cream: in the adult male,
for example, the lateral parts of the cephalo-
some, the dorsal median spines, the tips and
edges of the coxal plates and the posterior
part of the terminal segment are all of the
lighter colour.

The head is nearly twice as broad as long,
broadest anteriorly. The anterior margin
shows two shallow excavations, one on
either side of the well-developed rostrum ;
behind the latter is a well-marked transverse
ridge which extends across the width of the
head. Just in front of and between the eyes
is another transverse ridge, the lateral parts
of which are higher than the middle. The
portionofthehead between theeyesisraised F'g- ^i- Serolis aspera, n.sp., <?: x 4.

into two rounded prominences, the surfaces of which are covered with rounded pits ;
between these two, posteriorly, is a third small raised area, the centre of which is marked

DESCRIPTION OF SPECIES 353

by a small blunt spine : the posterior margin of these prominences is better defined than
the anterior. The eyes (Fig. 21) are about half the length of the head, reniform in shape,
rather narrow in proportion to the size of the head, and containing black pigment.

The lateral or coxal portion of the second thoracic somite, which is fused with the
head, is divided into an anterior and a posterior portion by a transverse ridge. This
ridge commences on either side at a point approximately midway along the length of
the eye, and runs at first transversely and then obUquely backwards and outwards, finally
disappearing near the lateral margin. The posterior margin of the second thoracic
somite is produced backwards as a small median dorsal spiniform process. The third,
fourth and fifth thoracic somites are sub-equal ; each is produced backwards in the mid-
dorsal line into a small spiniform process which increases in length from before back-
wards. The sixth and seventh thoracic somites are together equal in length to one of the
preceding somites ; the terga of these two somites are fused with each other and with the
tergum of the first abdominal segment for a short distance on either side of the mid-
dorsal hne. In this character the species agrees with S. bouvieri, Richardson, and with the
Australian species S. australiensis, Beddard, S. elongata, Beddard, and S. longicaiidata,
Beddard. In the female the position of the suture between the sixth and seventh somites
is indicated by a slight groove.

Well-developed coxal plates are present on each of the third to seventh thoracic
somites ; these are closely applied together for their whole length in the female, but in
the male the tip of each projects slightly beyond that of the preceding one. The coxal
plates of the third, fourth and fifth are marked off by distinct sutures from their re-
spective somites ; just within these sutures the lateral portion of the somite on each side
is raised into a low eminence and similar raised areas are present in corresponding
positions on the remaining somites. The coxal plates of the seventh thoracic somite are
longer than those of the other segments and, in the male, extend backwards beyond the
pleural plates of the second and third abdominal segments to a point level with the distal
end of the protopodite of the uropod ; in the female they are somewhat shorter and
extend to a level equal to that reached by the pleural plates of the third abdominal
segment.

The posterior margin of each of the three free abdominal segments is produced into
a median dorsal spiniform process similar to those of the thoracic somites ; the posterior
border of the sternum of each of these segments is produced backwards into a median
spine ; these increase in size from before backwards, and are considerably larger in the
female than in the male. The pleural plates of the third abdominal segment are longer
than those of the second and extend backwards to the level of the articulation of the
uropod.

The terminal segment (Fig. 21) is broader than long, terminating posteriorly in a
slightly up-turned trifid extremity, the middle projection of which is formed by the
extension backwards of a low median dorsal keel. Anterior to this keel is a strong median
dorsal spine, whilst on either side of it, in a lateral position, is a short ridge ending in
a spiniform process at about the level of the articulation of the uropod. Slightly posterior

354 DISCOVERY REPORTS

to these ridges, and nearer the median keel, are two small round prominences, one on
either side.

The antennule has a four-jointed peduncle: the first joint is broad, geniculated,
covered with fine hairs ; the second is about the same length as the first, broader distally
and covered with dense hairs ; the third is nearly one and a half times as long as the second,
but narrower, and three times as long as the fourth. The flagellum comprises twenty-
five joints in the female, and thirty-seven in the male. The first joint is about equal in
length to the fourth peduncular joint; the second to ninth are each about half the length
of the first, then gradually increasing in length ; the terminal and penultimate joints are
narrow, the former bearing two delicate terminal setae. A single sensory seta is found
at the distal end of all except the first four joints. The antenna has a five-jointed peduncle.
The first joint is short and fringed with delicate setae ; the second is a little more than twice
as long as the first with longer setae on the inner margin ; the third is about equal to the
second, broader at the distal end, with a row of five groups of setae on the dorsal surface
a little within the margin and a second row of four groups arranged somewhat obhquely.
The fourth joint is about twice as long as the third, with six transverse rows each of four
groups of setae equidistantly placed, but with the distal row near the distal end of the
joint. The surface is covered with short hairs, amongst which longer ones are scattered.
The fifth joint is more than twice as long as the third, with five short transverse rows
each of four groups of setae. These are near the anterior margin, and arranged more or
less equidistantly, the distal row being near the distal end of the joint. The flagellum of
the adults of both sexes is incomplete, but in an immature specimen is composed of
sixteen joints, of which the terminal and penultimate ones are small.

The mandibles are of the usual type, the form of the right is shown in Fig. 22 a. The
outer lobe of the maxillula bears ten stout spines and one delicate seta on the inner angle
of its distal truncate extremity ; the inner lobe is about two-thirds the length of the outer,
fairly stout, with its oval distal end bearing two minute setae. The maxilla has its two
outer lobes each bearing two long setae with a few delicate pectinations along their
inner edges ; the distal end of the fixed lobe is truncate and bears about ten stout and
eight more delicate setae.

The maxilliped is broad, with the basipodite separated from its lamella by a suture;
the inner margin of the basipodite is covered with fine dense hairs amongst which are
scattered a number of long setae ; the second joint of the palp is not much broader at its
distal than at its proximal end, and its under surface is covered with scattered setae.
The terminal joint is almost parallel-sided, with the rounded distal end fringed with
setae.

The second, third, fourth and last thoracic appendages of the adult male are shown in
Figs. 22 b-e. The propodus of the second is greatly expanded, with its inner margin
armed with a row of about thirty-nine broad processes alternating with peculiar
modified spines (Fig. 22 b). The distal end of the carpus is crenulate, and bears two stout
spines surrounded by numerous setae. The third thoracic appendage of the male is
much smaller than the second, with the propodus expanded and its inner margin armed

DESCRIPTION OF SPECIES

355

with a double row of modified spines (Fig. 22 c) ; the inner margins of the carpus,
merus and ischium are fringed with a dense mass of short hairs. The last thoracic
appendage (Fig. 22 e) is smaller than the rest and is similar in the two sexes.

-— 6. —

Fig. 22. Serolis aspera, n.sp.
a, right mandible : x 14. b, second thoracic appendage of (J : x 14. c, third thoracic appendage of adult ^ :
X 14. <^, fourth thoracic appendage : x 14. e, last thoracic appendage of cj, propodus and dactylus missing:
X 14.

The protopodite of the first three pairs of pleopods is somewhat triangular in shape,
with the produced angle bearing three plumose setae on the first, and two on each of the
second and third pleopods. The appendix masculina on the second pleopod of the male
reaches about two-thirds of the way towards the apex of the terminal segment. The
sutures of the exopod of the fourth and fifth pleopods are almost transverse.

The uropoda do not extend to the tip of the terminal segment ; the protopodite has
its inner angle produced into a long spinous process and the exopod is about two-
thirds the length of the endopod. Plumose setae are present on the outer margin and
rounded distal extremity of both the exopod and endopod, and their inner margins are
fringed with delicate setae.

Distribution. South Georgia.

This species very closely resembles .S. pagenstecheri, Pfeffer, but differs from it in the
following points :

(i) The most important difference is that in S. aspera the terga of the sixth and seventh
thoracic somites are fused together and with the first abdominal segment for a short
distance on either side of the median dorsal line. In S. pagenstecheri the suture between

356 DISCOVERY REPORTS

the sixth and seventh thoracic somites is complete, only the posterior suture of the
seventh segment being obsolete for a short distance in the mid-dorsal region ; in this
species therefore the sixth thoracic somite is free, and the seventh is fused for a short
space with the first abdominal segment.

(2) The body of S. pagenstecheri is slightly broader in proportion to its length than
that of S. aspera.

(3) The body of S. pagenstecheri is covered with scattered hairs, which are most
numerous along the posterior margins of the somites ; in S. aspera they are very sparse
and are not concentrated along the margins of the somites, but are moderately abundant
along the posterior free margins of the coxal and pleural plates.

(4) The terminal segment of both species is broader than long, but that of S. aspera
narrows more rapidly behind the articulation of the uropods ; the posterior extremity is
definitely trifid in S. aspera, in most specimens of S. pagenstecheri it is truncate or only
slightly trifid.

(5) In both species the terminal segment bears on its dorsal surface a strong median
spine in an anterior position, which is followed by a low median carina extending to the
posterior extremity of the segment ; on either side of this are two lateral carinae ending
in spiniform processes at some distance from the lateral margins. In S. aspera the
carinae next to the median are smaller than the outer, whereas in S. pagenstecheri they
are all of approximately equal size.

(6) The third joint of the palp of the maxilliped of S. pagenstecheri bears at its outer
distal angle a minute extra joint which does not extend beyond the end of the third,
and bears a group of long setae. This joint is absent from the maxilliped of S. aspera.

(7) The thoracic appendages are very much alike in both species ; the modified spines
on the propodus of both the second and third appendages of the male are similar, and
the inner borders of the carpus, merus and ischium are fringed with a dense mass of
fine hairs in both species. Similar hairs are present on the propodus, carpus, merus and
ischium of the last appendage of the mature male of S. pagenstecheri, but these are
absent from the corresponding appendage in S. aspera.

The distribution of the two species is the same : all recorded specimens have been
collected near South Georgia.

33. Serolis australiensis, Beddard.

S. australiensis, Beddard, 1884 ^>, p.6g, pl.vi, figs. 3-6; Whitelegge, i90i,p.237; Chilton, 1917, pp.396,
400, fig. 10; Nordenstam, 1933, pp. 90-2.

Diagnostic characters. The largest specimen of this species so far collected is a
female 14 mm. in length by 11 mm. in breadth, whilst the largest male measures lomm.
in length and 9 mm. in breadth: as in many Serolids, the males appear to be pro-
portionately broader than the females.

The form of the body is oval; its most characteristic feature is the presence of a large
number of tubercles, which are scattered over its surface and are especially large on the
posterior borders of the somites and on the dorsal surface of the terminal segment.

DESCRIPTION OF SPECIES 357

The head is broadest at the level of the eyes and ends anteriorly in a very long rostrum ;
a median spine is present on its posterior border and also on those of the thoracic somites.
The central portions of the terga of the sixth and seventh thoracic somites are fused
with the tergum of the first abdominal segment, that is, the hindmost sutures of the
sixth and seventh somites are obsolete in the middle line. The coxal plates are closely
applied together; according to Beddard those of the free somites are all separated by
sutures, but I have failed to find more than three pairs of sutures present on the third,
fourth and fifth somites respectively.

In the male the coxal plates of the seventh thoracic somite extend for a short distance
beyond the extremities of the pleural plates of both second and third abdominal seg-
ments, but not as far as the points of attachment of the uropoda ; in the female they do
not extend beyond the level reached by the pleural plates of the second abdominal
segment.

The terminal segment is more or less triangular in shape, with a median keel and a
truncate, slightly emarginate extremity. Its dorsal surface is covered with tubercles
which are arranged irregularly except for a transverse row which extends across the
segment slightly posterior to the articulation of the uropoda : one of these tubercles,
close to the lateral margin on either side, is considerably larger than the rest. The
exopod of the uropod is shorter than the endopod and its distal end is truncate ; the
endopod is rounded distally and reaches to the end of the terminal segment.

Occurrence. South Australian coast, 38 fathoms.

34. Serolis elongata, Beddard.

S.elongata, Beddard, 1884a, p. 335; 18846, p. 71; Whitelegge, 1901, p. 236; Chilton, 1917, p. 393.

The original description of this species, which is based on an examination of a single
adult female, is extremely brief, and no account of the form of the mouth parts and
other appendages is given. More recently the species has been collected on two occa-
sions: it was recorded by Whitelegge in 1901 in his report on the Isopoda of the Thetis
Expedition and in 1917 by Chilton, who examined a specimen in the collections made
by the F.I.S. 'Endeavour'. Neither of these authors have added anything further to
Beddard 's description.

Diagnostic characters. The adult female measures 10 mm. in length and 6-5 mm.
in greatest breadth. The surface of the body is almost smooth, except that each thoracic
somite is furnished with a median dorsal spine on either side of which a row of short
tubercles extends along its hinder border and is prolonged on to the coxal plate. The
hindmost sutures of the sixth and seventh thoracic somites are obsolete for a short
distance on either side of the middle line : I have examined the British Museum specimen
and find that the first three free somites are separated from their respective coxal plates
by sutures, and also that the coxal plates of the seventh somite extend beyond the pleural
plates of the second abdominal but not as far as those of the third. The pleural plates
of these segments are short and hardly extend beyond the anterior margin of the
terminal segment. The margins of the terminal segment are serrated. A median

D VII ^4

358 DISCOVERY REPORTS

longitudinal carina is present on the dorsal surface, and on either side of this is a
lateral serrated carina; between these and the central carina is "a short ridge running
obliquely towards the margin of the caudal shield from a point a little below and to
one side of the commencement of the central carina".
Occurrence. Port Jackson, Sydney, 30 fathoms.

35. Serolis longicaudata, Beddard.

S. longicaudata, Beddard, 1884a, p. 336; 18846, p. 72, pi. vii, figs. 8-10, pi. viii, figs. 1,2; Whitelegge,
1901, p. 237; Chilton, 1917, pp, 397, 400, fig. II ; Nordenstam, 1933, pp. 92-3.

Diagnostic characters. The only specimen in the Challenger collection is an im-
mature female 7 mm. in length and 5 mm. in breadth ; the two specimens described by
Chilton are both males, 8 mm. in length and 5 mm. in breadth. The head is broad and
terminates in a comparatively long rostrum. The body, excluding the terminal segment,
is circular in outline, and the terminal segment projects back for some considerable
distance and is proportionately longer than in any other species. The first three free
thoracic somites are separated from their respective coxal plates by sutures; the coxal
plates are all short and truncate at their distal extremities. Beddard states (18846, p. 73)
that "the tergum of the sixth segment is entirely absent " ; this, however, is not the case,
but the tergum of both the sixth and seventh somites is very narrow, and the hindmost
sutures of both are obsolete for some distance on either side of the middle Hne, so that
the middle portions of these somites and that of the first abdominal segment are fused
together. A similar condition is seen in S. australiensis, Beddard, S. elongata, Beddard,
S. bouvieri, Richardson, and S. aspera, n.sp.

The pleural plates of the second and third abdominal segments are also truncate and
do not extend beyond the anterior margin of the terminal segment ; the latter is roughly
pentagonal in outline, terminates posteriorly in a truncated extremity which is slightly
concave, and bears a median and two lateral carinae on its dorsal surface.

Occurrence. Off St Francis Island, South Australia, 6-13 fathoms.

36. Serolis tuberculata, Grube.

S.tuberculata, Grube, 1875, p. 227; Beddard, 18846, p. 67, pi. vi, figs, i, 2; Whitelegge, 190 1, p. 236;
Chilton, 1917, pp. 392, 394, text-figs. 1-9.

Diagnostic characters. The largest recorded specimen of this species is a female
measuring 19 mm. in length and 17 mm. in breadth; the measurements given for the
male are 12 mm. in length and about the same in breadth, so that, as is usually the case,
the male is proportionately broader than the female.

As Chilton (1917, p. 394) points out this species is readily distinguished from other
Australian species by " the series of pointed tubercles along the posterior margins of the
anterior segments of the peraeon, and by the median tubercles on all the segments ".
The tubercles are most numerous on the fifth segment, where there may be as many as
nine on either side of the median one. The tergal portion of the sixth thoracic somite is
very narrow, whilst the middle portion of the seventh somite has disappeared; the
first three free thoracic somites are separated from their respective coxal plates by

DESCRIPTION OF SPECIES 3S9

sutures, a fact which is not mentioned by either Beddard or Chilton, neither does the latter
show them in his figure of the entire animal : nevertheless they are quite distinct in the
specimens at the British Museum.

The coxal plates of the seventh thoracic somite extend backwards to just beyond the
base of the uropoda. The pleural plates of the second and third abdominal segments
project only slightly beyond the anterior margin of the terminal segment. The latter is
broader than long, its broadest part being at the base of the uropoda, from which point
it narrows rapidly posteriorly and terminates in a truncate extremity: it bears, on its
dorsal surface, a longitudinal median carina on either side of which is a lateral sub-
marginal ridge which extends to just beyond the base of the uropoda, at which point the
two are united by a transverse curved ridge.

Occurrence. Bass Strait, Gulf St Vincent, and St Francis Island, South AustraHa.

37. Serolis pallida, Beddard.

S. pallida, Beddard, 1884 a, p. 335 ; 1884 b, p. 74, pi. vii, fig. i, pi. viii, figs. 6-16; Whitelegge, 1901,
p. 236.

Diagnostic characters. This species is described by Beddard from two specimens,
a male and a female ; the latter is the larger and measures 16 mm. in length and 13 mm.
in breadth.

The body is oval, smooth, and of a pale brown colour. The head is almost triangular
with the apex directed backwards; a long, slender rostrum is present and the eyes
are small. The antennules and antennae are of approximately equal length; the first
peduncular joint of the former bears three strong tubercles on its upper surface and
a tubercle with its apex directed backwards is present on the posterior border of the
second joint ; the filament consists of twenty-four joints each of which bears two sensory
hairs. The fifth peduncular joint of the antenna is much enlarged; the filament is short,
consisting of only nine joints. The tergum of each of the third, fourth and fifth thoracic
somites is produced in the mid-dorsal line to form a backwardly directed spine ; similar,
though smaller spines also occur on the abdominal segments. The tergum of the sixth
thoracic somite is very narrow, whilst the middle portion of the seventh has disappeared.
The coxal plates are short and closely applied together. The terminal segment of the
body is roughly hexagonal in outHne with a median dorsal keel ; its posterior extremity
is notched.

After examining Beddard 's specimens at the British Museum, I can only add to his
description that the first three free thoracic somites have their coxal plates marked off
from the terga by distinct sutures, and that the pleural plates of the abdominal segments
are short and only extend for a short distance beyond the anterior margin of the terminal
segment ; the coxal plates of the seventh thoracic somite extend posteriorly for a short
distance beyond the pleural plates.

Occurrence. Off Port Jackson, Sydney, 30-35 fathoms.

14-2

LIST OF REFERENCES

AuDOUiN and Milne-Edwards, H., 1841. Description des Crustach nouveaux ou peu connus, etc. Arch. d.

Mus. d'Hist. Nat., 11, p. 5, pis. i, ii.
Beddard, F. E., 1884 a. Preliminary Notice of Isopoda collected during the Voyage of H.M.S. ' Challenger'.

Part I. The Genus Serolis. Proc. Zool. Soc. Lond., 1884, p. 330.
1884 ^i. Report on the Isopoda collected by H. M.S.' Challenger' during the years 1873-76. Parti. The

Genus Serolis. 'Challenger' Reports, Zool., xi, 85 pp., 10 pis.
Calman, W.T., 1909. Crustacea, in Ray Lankester's Treatise on Zoology. VII. Appendiculata, 3rd Fascicle,

p. 199.
1920. A new species of the Isopod genus Serolis. Ann. Mag. Nat. Hist. (9), VI, pp. 299-304,

3 text-figs.
Chilton, C, 1917. Notes on Australian Isopoda. Trans. Roy. Soc. S. Australia, xli, pp. 391-400, text-figs.

1-14.
CoLLiNGE, W. E., 1918. Some observations upon two rare marine Isopods. Journ. Zool. Research, ill, pp.

73-8, pis. iii, iv.
Cunningham, R. O., 1871. Notes on the Reptiles, Amphibia, Fishes, Mollusca and Crustacea obtained during

the voyage of H.M.S. ' Nassau ' in the years 1866-69. Trans. Linn. Soc. Lond., xxvii, pp. 465-502,

pis. iviii, lix.
Dana, J. D., 1853. Crustacea. U.S. E.xploring Expedition, xiv, pt. ii, pp. 794, 795, pi. liii, %• i-
Eights, J., 1833. Description of a new Crustaceous aninud found on the shores of the South Shetland Islands.

Trans. Albany Inst., 11, pp. 53-7, 2 pis.
Fabricius, J. C, 1775. Systema Entomologiae, Flensburgi et Lipsiae, p. 296.

1787. Mantissa Insectorum, Hafniae, i, p. 240.

GiAMBiAGi, D., 1925. Crustdceos Isopodos. Resultados de la Primera E.xpedicion a Tierra del Fuego

(1921). Buenos .\ires, 1925.
Grube, E. a., 1875. Beitrag ziir Kenntniss der Gattung Serolis. Arch. f. Naturg., LXI, i, p. 208, pis. v, vi.
Hale, H. M., 1933. Tanaidacea and Isopoda collected by the Great Barrier Reef Expedition, 1928-9. Ann.

Mag. Nat. Hist. (10), xi, pp. 560-1, i text-fig.
Hansen, H. J., 1925. Studies on Arthropoda. II. Order Isopoda, pp. IS17-22 and pi. vii, figs. 2-9.
Hodgson, T. v., 1910. Crustacea. IX, Isopoda. National Antarctic Expedition, 1901-1904. Nat. Hist., v,

PP- 23-30, pi. iv.
Leach, W. E., 1818. Dictionnaire des Sciences Naturelles, xii, p. 339.
Lockington, W. N., 1877. Remarks on the Crustacea of the Pacific, with Descriptions of some New Species.

Proc. Cal. Acad. Sci., vii, i, p. 36.
Lucas, H., 1877. Note and description of Serolis serrei. Bull. Soc. Franf. (5), vii, pp. cxlv, cxlvi.
Lutke'n, C. F., 1858. Beskrivelse af en ny Serolis Art. Vidensk. Medd. nat. Foren. Kjobenhavn, p. 98, pi. i.
MiERS, E. J., 1875. Descriptions of new species of Crustacea collected at Kerguelen Island by the Rev. A. E.

Eaton. Ann. Mag. Nat. Hist. (4), XVI, p. 73.
1879. Crustacea, in an Account of the Petrological , Botanical and Zoological Collectiom made in

Kerguelen's Land and Rodriguez during the Transit of Venus Expedition. Phil. Trans., CLXVIII, extra

vol., p. 200, pi. xi.

1881. Crustacea collected during Survey of H.M.S. 'Alert'. Proc. Zool. Soc. Lond., p. 61.

Milne-Edwards, H., 1840. Histoire Naturelle des Crustaces, iii, p. 228.

Monod, Th., 1925. Isopodes et Amphipodes de I'Exped. Antarct. Beige, 2' 7iote prdlim. Bull. Mus.

Paris, no. 4, pp. 296-9.
1926. Tanaidaces, Isopodes et Amphipodes. Resultats du voyage de la Belgica en 1897-9, PP- 35-8,

text-figs. 33-7.
1931. Tanaidaces et Isopodes sub-antarctiques de la collection Kohl-Larsen du Senckenberg Museum.

Senckenbergiana, xiii, No. i, pp. 26, 27.
Nierstrasz, H. F., 1931. Die Isopoden der Siboga-Expedition. Ill, Isopoda Genuina: ii, Flabellifera.
Siboga-Expeditie, Monographic xxxiic, pp. 222-4.

LIST OF REFERENCES 361

NORDENSTAM, A., 1933. Marine Isopoda of the Families Serolidae, Idotheidae, Pseudidotheidae, Arcturidae

Parasellidae and Stenetriidae mainly from the South Atlantic. Further Zoological Results of the

Swedish Antarctic Expedition 1901-3, Vol. iii, No. i. Stockholm.
Pfeffer, G., 1887. Die Krebse von Siid-Georgien nach der Ausbeute der Deutschen Station, 1882-3. Jahrb.

Hamburg, wiss. Anst., iv, pp. 43-150, pis. i-vii.
Richardson, H., 1899. Key to the Isopods of the Pacific coast of North America, with descriptions of

twenty-two 7iew species. Proc. U.S. Nat. Mus., xxi, p. 842.
1900. Synopses of North American Invertebrates. VIII. The Isopoda. American Naturalist,

XXXIV, p. 224.

1905. A Monograph of the Isopods of North America. Bull. U.S. Nat. Mus., No. 54, pp. 320-2,

text-figs. 353, 354.

1906 a. Isopodes. Expedition Antarctique Fran^aise (1903-1905) commandee par le Dr Jean Charcot,

pp. 7-10, pi. i, fig. I, text-figs. 12, 13.

19066. Sur les Isopodes de V expedition fraji^aise antarctique. Bull. Mus. Paris, pp. 187-8.

1906 f. Compt. Rendus, CLXII, pp. 849-51.

1911. Isopodes du Sandioich du Sud. Ann. Mus. Nac. Buenos Aires, ser. 3 a, xiv, pp. 395-400,

2 text-figs.
Sexton, E. W., 1924. The moulting and growth-stages of Gammarus, w-ith descriptions of normals and

intersexes ofG. chevreuxi. Journ. Marine Biol. Assoc, (n.s.), xiii, pp. 340-96.
Sheppard, E. M., 1927. Revision of the Family Phreatoicida (Crustacea), with a description of two

new species. Proc. Zool. Soc. Lond., pp. 86-90; text-figs, i and 2.
Smith, S. I., 1876. Contributions to the Natural History of Kerguekn Island, etc. Part ii. Bull.

U.S. Nat. Mus., No. 3, p. 63.
Stechow, E., 1921. Symbiosen zwischen Isopoden und Hydroiden. Zool. Anz., liii, pp. 221-3.
Studer, Th., 1876. Ueber neue Seethiere aus dem Antarktischen Meere. Mitth. Naturf. Gesellsch. Bern,

P- 75-
1879. Beitrdge sur Kenntniss niederer Thiere von Kerguelensland . Die Arten der Gattung Scrolls

von Kerguelensland. Arch. f. Naturg., XLV, I, pp. 19-34. taf. iii.
1884. Isopoden gesammelt wdhrend der Reise S.M.S. 'Gazelle' urn die Erde 1874-76. Abhand.

K. Preuss. Akad. Wiss. Berlin, pp. 1-28, pis. i, ii.
SuHM, R. von W., 1874. Von der Challenger-Expedition. Briefe an C. Th. v. Siebold: no, ii. Zeitschr. f.

wiss. Zool. XXIV, p. xix.
1876. Preliminary report. . .on Crustacea observed during the cruise of H.M.S. 'Challenger' in the

Southern Sea. Proc. Roy. Soc. Lond., xxiv, p. 591.
Tattersall, W. M., 1921. Tanaidacea and Isopoda. Brit. Antarct. (' Terra Nova') Exped., 1910, Zool., in,

No. 8, pp. 227-32, pi. vii
Unwin, E. E., 1920. Notes on the Reproduction 0/ Asellus aquaticus. Journ. Linn. Soc, Zool., xxxiv, p. 335.
Vanhoffen, E., 1914. Die Isopoden dei- Deutschen Siidpolar-Expedttion, 1901-1903. Deutsche Siidpolar-

Expedition, xv (Zool. vii), pp. 518-19, abd. 51.
White, A., 1843. Descriptions of New species of Insects and other Annulosa. Ann. Mag. Nat. Hist., xii, p. 346.

1847. List of specimens of Crustacea in the collection of the British Museum, pp. viii, 143.

Whitelegge, T., 1901. ' Thetis' Scientific results. Australian Museum, Mem. iv.

PLATE XIV.

Fig. I. Serolis schythei, Liitken, (?: x 2.

Fig. 2. Serolis polaris, Richardson, (J : x 3.

Fig. 3. Serolis convexa, Cunningham, $: x 3.

Fig. 4. Serolis gaudichaudii. And. et Edw., ?: x 4.

Fig. 5. Serolis laevis, Richardson,,?: x 2.

Fig. 6. Serolis cornuta, Studer, x i.

Fig. 7. Serolis trilobitoides. Eights, x i|.

DISCOVERY REPORTS. VOL. VII

PLATE XIV

John Ba>: Smis & Drtrueluon, L'^ l-aruUa.

SE ROLLS

[Discovery Reports. Vol. VII, pp. 363-406, Plate XV, December, 1933]

SOME ASPECTS OF RESPIRATION IN BLUE

AND FIN WHALES

By

ALEC H. LAURIE, M.A.

CONTENTS

Introduction page 365

Southern whales and their environment 365

Diving 368

Hydrostatic pressure 370

Gas Analyses 374

Urine 376

Allantoic fluid 382

Liquor folliculi 383

Blubber 383

Blood 383

Disappearance of nitrogen from blood 389

Mechanism of nitrogen removal 392

Nature of X organisms 395

Oxygen capacity of whale blood 396

Carbon dioxide in whale blood 398

Summary 399

Literature cited 400

Appendix 400

A note on the composition of whale blood 400

A note on the composition of the allantoic fluid of Blue whales ... 401

A note on the weights of some Blue whales 402

Plate XV following page 406

SOME ASPECTS OF RESPIRATION IN BLUE

AND FIN WHALES

By Alec H. Laurie, m.a.
(Plate XV; text-figs. 1-4)

INTRODUCTION

THE investigations which form the subject of this report were performed in two parts.
Preliminary examination of the general physiological background of cetacean life
was made at the Marine Biological Station in South Georgia in 1 930-1. Part of the
ensuing year was spent in studying problems of respiration and gaining experience in
the technique of blood gas work at Cambridge under the guidance of Professor Barcroft,
to whom I am much indebted for advice and laboratory facilities. Field work was
resumed in the season 1932-3, when The Southern Whaling and Sealing Company
very kindly placed a laboratory at my disposal on board the pelagic whaler
' Southern Princess '. The freshness of whales brought in to a pelagic whaler outweighs
the inconveniences of a floating laboratory, since it is important to examine post-mortem
material with a minimum of delay.

Some qualification is necessary for the title of this paper. The physiological properties
of living southern Blue and Fin whales will in all probability never be known by direct
observation, though the smaller whales and porpoises might conceivably be confined
and observed. An attempt has been made in the present work to examine the properties
of fresh carcasses of Blue and Fin whales from which tentative deductions as to their
mode of life have been made. It is felt that such an investigation, which would be
superfluous with smaller land mammals, has its justification in that it may help to throw
light on the life of one of the world's most inaccessible creatures.

At the suggestion of Professor Barcroft the manuscript of this paper was submitted
to Professor August Krogh of Copenhagen. Professor Krogh is renowned for his re-
searches into problems of respiration and blood gases, and I am very grateful to him for
a number of criticisms which I have incorporated in the text.

I wish also to thank Professor Barcroft, Dr F. J. W. Roughton, Mr G. S. Adair,
Professor D. Keilin, and Mr D. Ward Cutler for the interest they have taken in this
work and for the useful suggestions and criticisms they have offered.

SOUTHERN WHALES AND THEIR ENVIRONMENT

It will be convenient to review briefly the external conditions of a Blue or Fin whale's
life, and later to see how much of its mode of living can be deduced from the applica-
tion of physiological principles. Blue and Fin whales are very large, entirely aquatic
mammals, preferring for the most part to live on the high seas and seldom venturing

366 DISCOVERY REPORTS

into shoal water. More than half the year is spent in Antarctic waters, where the
temperature of the surface water is rarely more than 4° C. and is frequently below zero.
The rest of the year is spent in tropical waters, whose temperature may be as high as
29° C. The whale's only external protection against cold is the blubber, which is
generally beheved to be thickest at the time of migration from the Antarctic to the
tropics and thinnest on the return migration. Since whales cannot regulate their tem-
perature by the use of sweat glands, it follows that they must be able to modify body
temperature greatly by variation of blood supply to the skin, if not by control of basal
metabolism.

Temperature. Some effort was made in South Georgia to determine the normal
temperature maintained in Blue and Fin whales. The procedure was to insert a thermo-
meter into the longissimus dorsi muscle as near as possible to the vertebral column.
This was done while whales were being dismembered at the whaling station. It was
found that fresh carcasses cooled little before dismemberment and that a slight but
irregular rise in temperature was observed when decay commenced. The average tem-
perature of thirty fresh whales was 35-1° C. High temperatures are found near the
abdomen, due doubtless to the presence of decaying matter in the stomach and in-
testinal tract. Morimoto, Takata, and Sudzuki (1921) quote a body temperature of
36-6-36-9° C. from the early work on northern Balaenopterids, but no mention is made
of the part of the body in which the temperature was taken. Both these and the South
Georgia temperatures suggest strongly that the whale's temperature is lower than that
of most mammals. Smith (1895) quotes observations by Seidangrotsky on domestic

mammals :

Horse 38-oo-38-2o° C.

Cattle 3876-38-96° C.

Swine 39-44° C. average

Sheep 39-72-40-22° C.

Basal metabolism. The problem of temperature maintenance and control leads to
a consideration of the minimum energy requirements of a whale which would be
sufficient to keep the animal alive in a state of rest (basal metabolism). In this respect,
as in others, it will be necessary to make constant comparisons with man's physiological
properties, partly because human physiology is better known than that of other land
mammals, and partly because man in his submarine ventures has attempted to adapt
himself by mechanical means to the conditions under which the whale lives.

One general feature of basal metabolism according to Starling may be applied to
whales: "We may say that a warm-blooded animal requires a daily expenditure of
about 1000 calories per square metre of body surface to carry out such motor processes
as are essential to life" (p. 510). This figure of basal metabolism can be taken as a
mammalian constant irrespective of the natural or acquired integument of the animal ;
presumably each animal has taken steps according to its environment to provide itself
with adequate insulation, such as blubber, fur, feathers, and so forth. Basal metabolism
is greater in small animals than in large ones, since in small animals the surface area is

RESPIRATION IN SOUTHERN WHALES

367

greater in proportion to body weight ; conversely the basal metabolism of a whale may
be expected to be small. There is no convenient method of measuring the surface of
a whale, though various formulae have been suggested for calculating the area. For the
present purpose it will be sufficient to take the Blue whale as representing two cones
joined together by their bases, as Guldberg (1907) suggested. The base of each cone
may be taken to be a transverse section of the body at the occipital condyles.

A female Blue whale was measured and weighed by Capt. Th. S0rlle at Stromness,
South Georgia, in 1926: the length was 27-18 m. (see Appendix, p. 404). The approxi-
mate area of the whole whale, calculated as above, was 275 sq. m. ; the basal metabolism
therefore would be 275,000 calories per day. The weight of this whale was approximately
122,000 kg. The calories per kg. per day necessary to support hfe would be 2-25. For
comparison the following figures are shown (Starling) :

Weight (kg.)

Calories per kg.
per day

Man

Rabbit

Guinea-pig

70
2-os
0-67

32-9

S8-S
223-1

It would seem therefore that the Blue whale, of which this specimen is fairly typical,
has a very small basal metaboHsm in comparison with smaller mammals. A man of
70 kg. at rest absorbs about 300 cc. oxygen per minute on behalf of his basal metabolism,
that is to say 4-28 cc. oxygen per kg. of body weight per minute. The Blue whale's basal
metabolism is 14-6 times less than man's, so that the whale's oxygen requirements will
be only 0-293 cc per minute per kg. This whale, then, required 35-75 1. per minute of
oxygen, which is contained in 178-75 1. of air.

Vital capacity. Direct measurement of a Blue whale's vital capacity, or the greatest
amount of air which can be taken in after the most forcible expiration, is impracticable.
The weight of the lungs of the whale under consideration was 1226 kg., 1-24 per cent
of the weight of the soft parts, whereas human lungs average 2-37 per cent. Another
Blue whale measured and weighed at the same whahng station in 1924 was 20-3 m.
long and weighed 48,903 kg. (see Appendix, p. 403). The lungs weighed 1-53 per cent
of the soft parts.

If for the moment it is assumed that the lung's capacity is proportional to its weight
in whales and human beings, two facts emerge: (i) the whale's vital capacity is ap-
proximately half that of man in proportion to its weight ; (2) if the human vital capacity
is 3-50 1. for a man weighing 70 kg. (Starling), then the vital capacity of the whale first
mentioned would be 3050 1. (3-05 cm.). The minimum air requirements of the whale
were estimated above to be 178-75 1. of air per minute. At this rate the whale can stay
submerged for 17 min., assuming that all the oxygen in the lungs is used and that no
muscular exertion is taking place. The calculation, for lack of first-hand data, has been
based on the assumption that the capacity of the lung is a function of its weight. No

368 DISCOVERY REPORTS

account so far has been taken of the possibility of a much greater elasticity in whales'
lungs than in those of man. Unfortunately no figures are available to show the maximum
internal volume of the thorax in the particular whale under consideration, but in a Fin
whale measured by myself in South Georgia, length 22 m., the volume of the thorax
was 8 cm. after death, when probably it was not fully expanded. Of this the heart and
blood vessels occupied approximately 0-5 cm., so that there were at least 7-5 cm. of
space available for the accommodation of the lungs. If the vital capacity is taken to be
7 cm., the whale could remain submerged for 39 min. It would be unprofitable to try
to assess more accurately the probable vital capacity of the Blue whale originally con-
sidered; but these figures indicate that a submersion period of 17-30 min. for a whale
at rest is a very conservative estimate.

DIVING

Depth and duration. What the energy requirements of an active whale would be
while submerged it is hard to say. But there is reason to suppose that during long sub-
mergence whales do not move rapidly since they frequently reappear after a quarter of
an hour close to the spot at which they dived. Little reliable information is available
on the normal period of diving of Blue whales. Most observations have been taken
when the whale either is being chased or has already been harpooned. A whale tends
to hide beneath the surface when chased and the time of submergence is 10 min. or
more. A harpooned whale will stay down as long as 25 min., returning to the surface
with a rush, and will not again remain submerged for a long period until it has had an
opportunity to ventilate the lungs thoroughly. I have noticed that estimates of the
time of submergence of whales during the chase are liable to be faulty and greatly
exaggerated owing to the suspense which prevails. I have gathered from conversations
with Norwegian gunners that the average time of submergence of a Blue whale which
is not alarmed is about 10 min. On being chased or harpooned the dive may last as
long as half an hour.

The depth to which whales dive is a matter of the greatest interest physiologically,
since the pressure on the animal becomes of significance in all considerations of the
whale's respiration. The external hydrostatic pressure becomes communicated to the
lungs and the air in them, so that pressure afl'ects the most vital processes. There have
been many wild estimates of the depth to which whales can dive ; but W, Scoresby jun.
(1820) has given a conservative account of the behaviour of the Greenland Right whale.
"When fish have been struck by myself, I have on different occasions estimated their
rate of descent. For the first 300 fathoms the average velocity was usually after the rate
of 8 to 10 miles per hour. In one instance the third line of 120 fathoms was run out in
61 sec, that is at the rate of 8^ English miles or 7^ nautical miles per hour.... The
average stay under water of a wounded whale, which descends steadily when struck,
according to the most usual conduct of the animal is about 30 min. But in shallow water,
I have been informed, it has sometimes been known to remain an hour and a half at the
bottom after being struck and yet has returned to the surface aUve. . . .The remarkable

RESPIRATION IN SOUTHERN WHALES 369

exhaustion observed on the first appearance of a wounded whale at the surface after a
descent of 700 or 800 fathoms perpendicular does not depend on the nature of the
wound it has received, . . . but is the effect of the almost incredible pressure to which the
animal must have been subjected."

Roy Chapman Andrews (19 16) records being told that a Blue whale dived straight
down, taking one-quarter of a mile of rope (220 fathoms), and remained below 32 min.
A Southern Blue whale sometimes takes out 300 fathoms of rope in a steep or nearly
vertical dive.

So much for diving under stimulus. Very little is known about the depth to which
whales normally dive. Racovitza (1903) estimated 100 m. as the normal maximum.
It has been found in the course of plankton investigations in Antarctic waters that
whales' food, Eiiphausia superba, occurs anywhere between the surface and 200 m. or so,
and that large swarms, such as would concern the whale more than stray individuals,
are found either at the surface or anywhere between the surface and 100 m. This sug-
gests that the lower limit of a Blue whale's normal activities in the vertical plane may
be 100 m.

A case was related to me in 1931 of a dead Sperm whale which was found off the
Peruvian coast entangled in a submarine cable which had broken at a depth of 500
fathoms. The mate of the cable ship 'AH America' informed me that when the cable
was hauled to the surface it was caught in the angle of the whale's jaws and a loop was
twisted round the tail. From these observations it is probable that the whale became
entangled in the cable while actively engaged, possibly in pursuit of a cephalopod. The
evidence available thus goes to show that whales of various species are able to sustain
hydrostatic pressure up to 10 atmospheres in normal life and even greater pressures
under provocation. It should, however, be borne in mind that Sperm and Right whales
may differ widely in their diving capabilities from Balaenopterids.

Breath retention. It will be appropriate at this point to consider what may be
expected to be the effects of the whale's habit of life on the respiratory system. In the
first place a whale must hold its breath. Instead of enjoying a continuous ventilation
of the lungs, as do land mammals, the whale has to depend on an intermittent filling and
emptying of the lungs which occurs once in 10 or 15 min. (A land mammal, of which in
this paper man has been taken as an example, is quite unable to hold the breath for
more than i min., though longer periods may be endured after training [Starling].) The
natural consequences of this compulsory holding of the breath are a tendency to
shortage of oxygen and accumulation of carbon dioxide.

It may reasonably be expected that a whale will return to the surface to breathe when
actuated to do so by some mechanism analogous to the respiratory centre in man. The
system of ventilation in human beings does not provide for a complete change of air at
every breath. Only a part of the air in the alveoli, or ultimate subdivisions of the bron-
chioles, is changed at each breath, so that continuous breathing is necessary. The whale,
on the other hand, takes only one breath and disappears below the surface again. It
would thus be to the whale's advantage to effect as complete a change of air in the lungs

370 DISCOVERY REPORTS

as possible at each opportunity, to get rid of all the carbon dioxide excreted from the
blood and take in air containing its full complement of oxygen.

The histological structure of the whale's lungs has been examined in search of any
adaptations to the whale's peculiar mode of life. Photomicrographs of distal portions
of lungs stained to show elastic tissue and cartilage are shown in Plate XV. Lungs of
two species of southern whale are shown : Humpback and Fin whale, each magnified 540
times. The main features of these lung sections are the great thickness of the walls of
the infundibula and alveoli compared to those of a man or pig, and the presence of thick
bands of elastic tissue surrounding each infundibulum. The epithelial cells of the in-
fundibula seem to be embedded in a mass of spongy material. The Fin whale lung
(PI. XV, fig. 2) is quite collapsed, so that many infundibula are almost invisible. The
Humpback lung (PI. XV, fig. i) was full of water some time before the dissection was
begun, and this may account for the smaller degree of collapse which seems to have
occurred. The presence of such large amounts of elastic tissue in the infundibula probably
enables the lung to collapse readily and completely when the whale expires, so that the
only air left in the lungs is that in the "dead space", or non-flexible portions of the
lungs, such as the trachea and bronchi.

Inspiration and expiration. There is further reason to suppose that the whale
makes a more complete change of air in the lungs on each occasion than any other
mammal. The blast of expiration, which in the Antarctic is clearly visible as a column of
condensing vapour, is usually seen to rise to a height of 20 ft. in half a second ; the noise
of the expiration is audible on a still day for a distance of at least half a mile. It is
obvious that expiration is extremely forcible and that a vast volume of air passes through
the blowhole with great velocity.

I have been able to time the acts of expiration and inspiration on occasions when the
whale was so close that the movements of the blowhole could be seen. Sometimes it
was possible to distinguish the sound of expiration and the more subdued sound of air
rushing into the lungs. Otherwise the time of inspiration was taken to be the interval
between cessation of the blast and the closing of the blowhole, allowance being made
for sound lag due to distance. The total time taken from the commencement of the
expiratory blast to the closing of the blowhole preparatory to submersion averages
I -5 sec. Expiration lasts o-6 sec. ; inspiration takes longer, 0-9 sec. Bennett (1931), from
personal observation, estimates 2 sec. for the complete act. No attempt seems to be
made by the whale to start expiration before the surface of the sea has been broken.
A Blue whale may breathe twice in succession with an interval of about i min. after a
long dive, but normally only one breath is taken at a time.

HYDROSTATIC PRESSURE

As has been indicated above whales are accustomed to undergo considerable pressure
in the course of submersion. For every 10 m. depth the pressure is increased by i atmo-
sphere, or 14 lb. to the sq. in., so that the absolute pressure bearing on the surface of a
whale is 14 lb. per sq. in. plus another 14 lb. per sq. in. for every 10 m. submersion.

RESPIRATION IN SOUTHERN WHALES 37i

A whale at loo m. is subjected to an absolute pressure of ii atmospheres (1541b. per
sq. in.). This pressure bears equally on the whole surface of the whale and compresses
the internal organs uniformly throughout. It might perhaps be thought that the lungs
and heart were immune from compression through the protection of the thorax, but it
will readily be seen that if the viscera and musculature with their attendant blood vessels
were compressed and the heart, lungs, and great veins not compressed there would
immediately be such a surge of blood to the latter organs as would derange all blood-
pressure control and engorge them so that life would be endangered. Nor is it to be
supposed that the blubber surrounding a Blue whale acts as a rigid or semi-rigid sheath.
The blubber, though hard, is flexible and is quite pliable and slack ventrally from the
chin to the umbilicus in the region of the ventral grooves. Thus there is less protection
afforded by the blubber in the thoracic region than in any other part of the body.

Uptake of oxygen. In order to understand fully the implications of hydrostatic
pressure on the whale's respiration it is necessary to imagine a whale submerged and
staying at a depth of 100 m. for some minutes. Compression of the whole body, in-
cluding the thorax, results, and with it the air in the lungs also becomes compressed
ten times. The effect of this compression on the oxygen uptake will be to raise the partial
pressure of the oxygen in the lungs, as pointed out by Ommanney (1932). The partial
pressure of oxygen in normal air is about 1 50 mm. of mercury. A partial pressure of
60 mm. is sufficient to keep the blood of most mammals above 80 per cent saturation
with oxygen. The whale's oxygen will be at 1500 mm. of mercury when a depth of
100 m. is first reached and may become ten times as depleted at that depth before in-
adequate oxygenation of the blood takes place. The oxygen in the lungs would be at a
partial pressure of only 6 mm. of mercury when the whale returned to the surface. In
other words the whale is able to make the fullest use of the oxygen in the inspired air.
There is evidence to show that in animals there is great power of endurance under
conditions of progressive anoxaemia. Experiments on dogs (C. W. and C. H. Greene,
1922) show extraordinary endurance, provided the blood pressure and minute volume
are maintained.

Ommanney (1932), in discussing this matter, has reached the conclusion that whales
do not descend much deeper than 130 ft., but this estimate would appear to be too low
even for normal dives, and, though satisfactory evidence is lacking, it appears most
probable that when harpooned they can reach still greater depths. Ommanney thinks
that whales could not reach a depth of 600 m. because at this depth the hydrostatic
pressure, 60 atmospheres, would raise the partial pressure of the oxygen in the lungs
to 4 atmospheres and remarks that this has been found by Paul Bert to be toxic in certain
animals. Oxygen in the lungs would, however, be toxic at that pressure only if there
were an unlimited supply, whereas it will be seen from the figures of blood volume and
vital capacity above that one lungful of air (say 7000 1.), containing approximately
1400 1. of oxygen, would serve to supply the blood (8000 1.) with 17 vol. per cent
oxygen, which is somewhat below the normal oxygen capacity of human blood but
considerably above that of whales, as will be shown later. And, even if it were supposed

372 DISCOVERY REPORTS

that a whale dived with its blood fully oxygenated and the lungs full of fresh air, the
fluid portion of the whale alone, 60 per cent of the total weight, or 73,200 1., could
dissolve 1793 1. of oxygen, or about 25 per cent more than the available volume, if the
partial pressure of oxygen were equal to i atmosphere. In other words, the volume of
oxygen taken down by the whale could be dissolved more than once over in the body
fluids before the partial pressure of the oxygen exceeded i atmosphere. In any case, as
Professor Krogh has pointed out, a large proportion of the oxygen in the lungs would
be used up before the whale reached any considerable depth. There is therefore no
theoretical obstacle to prevent the Sperm whale mentioned above from reaching the
depth of 900 m. at which it became entangled with the submarine cable.

Excretion of carbon dioxide. The accumulation of carbon dioxide in the blood and
hence, by diff^usion, in the lungs will be considerable. In contrast to the conditions
underlying oxygen intake, the heightened pressure is in this case a disadvantage to the
whale. The eff'ect of carbon dioxide on the respiratory centre depends on the partial
pressure of the gas in the lungs and not on the percentage. A small percentage of
excreted carbon dioxide under a high pressure will have the same eff'ect as a large
percentage under a low pressure. Accumulation of carbon dioxide is bound to occur,
particularly in the blood, and, as will be shown later, considerable volumes of this gas
are found dissolved in body fluids as well as in the blood. The conditions which must
result from deep diving postulate a less delicate regulatory mechanism for the control
of carbon dioxide tension in the blood than exists in man and other land animals. The
respiratory centre, if it is similar in action to that of man, must be either adapted
to respond to a diff'erent range of carbon dioxide tensions or dependent entirely on
stimulation by oxygen shortage. One interesting consequence of stimulation by oxygen
shortage would be that if the stimulation began at 100 m. the whale's ascent to the sur-
face would cause greater and greater stimulation as the hydrostatic pressure decreased
and with it the partial pressure of oxygen in the lungs.

Dissolved nitrogen and caisson sickness. The third consequence of hydrostatic
pressure on the whale is the passage of gases from the lungs into solution in the blood
and hence into the body generally. The physical solution of oxygen and carbon dioxide
in the blood is a minor phenomenon compared with the mechanism for taking these gases
into chemical combination. It remains therefore to consider the solution of nitrogen
and the inert gases (these latter being of minor importance). The transference of gases
from the lungs to the blood follows Dalton's Law of the solubility of gases. Blood dis-
solves about 1-2 vol. per cent of nitrogen from air at atmospheric pressure. For every
atmosphere increase in pressure the blood takes up another 1-2 vol. per cent. This pro-
cess in human blood is the basis of the trouble known to divers as caisson sickness, in
which nitrogen is dissolved in the blood under pressure and on decompression fails
to return to the air through the lungs if the diver's return to the surface is too rapid.
A summary of a typical case of caisson sickness is here abstracted from Sir Leonard
Hill's book on the subject (1912). A petty officer diving at Lamlash in 24I fathoms of
water (= 6 atmospheres) took 40 min. to reach the bottom, remained there 40 min., and

RESPIRATION IN SOUTHERN WHALES 373

took 20 min. to return to the surface. Shortly after conning up he was taken ill and died
in 7 minutes. Post-mortem examination showed black or very dark blood in the blood
vessels and large bubbles of air in the veins of Galen and the choroid plexus, in the right
ventricle of the heart, and in the veins covering the brain. Small vessels of the mesentery
attached to the gut were found full of nitrogen. It has been shown that the illness is
due to the accumulation of nitrogen in the vessels of the heart, certain blood vessels,
and the central nervous system.

In order to see how far caisson sickness may be expected to be a danger to the normal
life of a Blue whale it will be convenient to summarize the main features of the illness
as it appears in man and other mammals, according to Hill. " The rate at which various
animals, and different organs of the same animal, become saturated and desaturated in
compressed air is proportional to the volume of blood relative to that of the tissues, and
to the velocity of the circulation." The volume of blood in the large Blue whale which
was weighed at Stromness in 1924 was 6-6 per cent of the total weight. The blood
volume in man is 4-9 per cent of the body weight (Haldane), in the horse 6-6, ox 77,
sheep 8-OI, and pig 4-6 (Ellenburger). No information is available as to the rate of
circulation in whales, though it is to be expected that as among land mammals, other
things being equal, the rate would be lower in a large animal such as a whale. The low
basal metabolism indicated above also suggests sluggish circulation. The heart of the
Blue whale according to two observations is about 0-59 per cent of the body weight,
the human heart is 0-41 per cent. These figures suggest, but by no means prove, that
the whale's heart is not notably larger or more powerful in proportion to the size of the
animal than the human heart. " If the activity of the circulation is great enough, the
excess of gas may escape by diffusion through the lungs without the formation of
bubbles" (Hill). For this reason small animals with rapid circulation are relatively
immune from the effects of decompression, while large animals are more liable to the
sickness. "When a liquid saturated with gas under pressure is suddenly decompressed
the excess of gas does not immediately come out of solution either by bubbles or
diffusion. This delay is specially marked in the colloidal body fluids." Hill goes on to
say that bubbles are not found in gland cells or muscular fibres (places of rapid circula-
tion of blood) but may, on the other hand, occur abundantly in the collections of body
fluids — bile, urine, and synovial and amniotic fluids. Bubbles are found most frequently
in the fat, because fat absorbs five times as much nitrogen as water and the circulation
through fat is relatively poor. "Oxygen bubbles are not a factor in caisson sickness — "
The fat in all cases of decompression is honeycombed like whisked white of egg.

So far the probability that a whale will be liable to caisson sickness seems high. It is
hard to imagine that an animal which stayed submerged until the need for fresh air
became urgent should come to the surface so gradually as to allow the dissolved gases
to diffuse back into the lungs. The time required for desaturation is more than 20 min.
per atmosphere in man, so that, on the basis of human performance, if a whale were to
dive 100 m. and stay there for 15 min., by which time the supersaturation of the body
tissues might be half complete, it would, in order to avoid caisson sickness, have to

374 DISCOVERY REPORTS

return slowly to the surface during a period of at least loo min., which is contrary to
observation.

The influence of fatness on the liability to caisson sickness is marked in land mammals
because of the high solubility of nitrogen in fat mentioned above. Boycott, Damant and
Haldane (1908) have tested guinea-pigs and concluded that fatness increases thesuscepti-
bihty to death from caisson sickness. The coefficient of solubility of the watery part of
the body may be taken to be 0-9 per cent and of the fatty part about 5-0 per cent. Man
contains on the average 66 per cent water and 1 5-20 per cent fat. The fat content of
a Blue whale including internal fats varies seasonally but is seldom more than 24 per
cent of the weight and frequently as low as 20, so that whales are in proportion little
fatter than man.

The Blue whale whose weight has been considered above contained 6o-o per cent
water and 2275 per cent fat. The weight was 122,000 kg. The watery part, 73,200 kg.,
would dissolve 658-8 1. of nitrogen from air at atmospheric pressure, while the fat,
27,755 ^§-> would take up 1387-7 1. The total uptake of nitrogen is therefore 2046-5 1.
at atmospheric pressure, and in addition the same amount can be absorbed from the
air in the lungs for every atmosphere of compression which the whale undergoes. The
vital capacity of this whale was estimated conservatively at 3050 1. of air, of which
2440 1. would be nitrogen, so that if the whale stays submerged, even at a depth of
10 m., at which i extra atmosphere pressure is appHed over and above atmospheric
pressure, all the nitrogen in the lungs will eventually disappear into solution in the body
fat and fluids. The nitrogen from one breath would obviously be insufficient to over-
charge the blood ; but the cumulative effect of successive inspirations followed by sub-
mersions would be to overcharge the blood, so that in the event of the whale's electing
to stay at the surface longer than the instant which is usual caisson sickness might
supervene. After a certain number of breaths at the beginning of the whale's hfe, there
would be an equilibrium established between the nitrogen in the blood and the nitrogen
in the lungs at the pressure which represented the average depth of submersion. After
a time only a little nitrogen would go into solution if the whale dived a little deeper than
usual, or only a little gas would come out if the dive were shallower than usual, but
there would still be the same danger from decompression bubbles if the whale hngered
at or near the surface.

GAS ANALYSES

The theoretical conditions of a whale's respiration construct a scheme of life which
condemns the whale never to delay at or near the surface on penalty of caisson sickness.
Actually a whale can linger at the surface, as for instance when suckling a calf, and an
explanation of this is to be found in the condition of gases in solution in freshly killed
whales. It will never be possible to know everything on this subject of whales' respira-
tion for obvious reasons, but, just as the life history of Blue and Fin whales has been
reconstructed in considerable detail from examination of their carcasses, so the study
of the gas contents of whales' body fluids has thrown light on respiratory activities. It

RESPIRATION IN SOUTHERN WHALES 375

must be clearly realized that this kind of work has limitations and that post-mortem
gas analyses may give misleading results. Every precaution was taken to ensure that
the specimens used in analysis were as nearly as possible in the condition which
prevailed at the moment of the whale's decease. For instance, a sample of blood was
never taken from an abdominal blood vessel, where the fermentation of stomach and
intestinal contents might have a local effect on the blood gases. Freshness was of
course essential ; though some specimens were from whales which had been dead 24
hours or more in order to see the effect of staleness and general decomposition on gas
values.

Technique. A few words should be said about the technique employed. The method
of collecting the samples will be described under the appropriate headings. The samples
were taken with as little delay as possible from the deck down to the laboratory in the
'tweendecks. Here was installed a gas burette similar to the Van Slyke constant-pressure
type (Van Slyke, 1917). This burette was specially constructed by Messrs W. G. Flaig
and Sons for use on board ship where the accommodation is limited. In addition to the
usual features of a gas burette, there was an extra branch tube at the top for the ad-
mission of gases, while below the main tap there were three limbs of large capacity
instead of the two tubes in the ordinary model. There was a hole near the top of one of
the limbs, which was covered by a strong rubber band. The purpose of this hole was to
permit the withdrawal of samples of evacuated fluids from a lower limb through a
hypodermic syringe and the injection of small quantities of reagents into the lower half
of the burette. Strong spiral springs were fitted to the stopcocks so that the burette
could be used for mixing gases with fluids under pressure, and the whole apparatus was
constructed of specially hard glass and on generous Hues. These modifications were
introduced so that the burette might be used as a tonometer and as a Haldane gas
analyser as well as in its normal capacity.

The technique of gas analysis of fresh samples was exactly the normal one. Samples
were introduced into the burette, which had previously been evacuated and tested,
without exposure to air and were evacuated with vigorous shaking. After allowing
10 min. for evacuation, the sample was isolated below the main stopcock while the gases
were measured. Carbon dioxide was absorbed by addition of semi-normal caustic soda
and oxygen by anthraquinone-beta-sulphonate with sodium hydrosulphite in semi-
normal caustic soda. The residual gas in the burette was taken to be nitrogen and inert
gases. Some difficulty was at first experienced in bad weather in reading the gas volumes,
since the roUing of the ship caused a pulsation of the mercury column. This was over-
come by keeping the burette with its rubber tube and levelling bulb all in a fore-and-aft
plane and by restricting the surge of the mercury in the tube by a screw clip. Tempera-
tures were taken from a thermometer suspended beside the graduated part of the burette ;
barometric readings from the ship's barometer. All gas volumes mentioned are corrected
to normal temperature and pressure unless otherwise stated. Tests were mads to ensure
that gases were being fully extracted from liquids by evacuating water which had been
allowed to stand in contact with air at various temperatures. The results given by the

376 DISCOVERY REPORTS

burette agreed, after correction, with the theoretical values to within 5 per cent. Each
sample was re-evacuated after the gases had been measured to ensure that extraction
was complete.

The various body fluids will now be dealt with in detail in the following order : urine,
allantoic fluid, liquor folliculi, maternal and foetal blood. The gas contents of blubber
and connective tissue will also be considered.

URINE
Hill records the accuracy with which the pressure conditions in the lungs of an animal
were reflected by the nitrogen content of the urine ; advantage was taken of this to find
out the degree of supersaturation of urine caused by various pressures in the lungs
by drawing the fluid under cover and analysing the dissolved gases. Since the bladder
of a Blue whale is nearly always full after death with the walls in a state of tight contrac-
tion, against which the sphincters seem able to retain the urine, it was thought that
samples of urine from freshly killed whales would be valuable in giving some indication
of the gaseous conditions obtaining during the last half hour or so of the whale's life.
The bladder is extremely tough and thick, and since there was never gas in the bladder
except in one case it is fairly certain that if supersaturation of the urine was found it
indicated at least the same amount of supersaturation before death. Supersaturation
could not have occurred after death.

The procedure was to puncture the bladder with a sharp knife and quickly immerse
a large test-tube in the urine. The tube was immediately closed with a rubber bung
while still submerged and promptly placed in a pot of ice-cold water. The last move
prevented any tendency to effervescence, in the event of supersaturation, by increasing
the solubility of gases in general through lowering the temperature. The tube was then
conveyed to the laboratory, and some of the contents drawn from the bottom under
paraffin were pipetted into the receiving cup of the gas burette. 0-05 A^ lactic acid,
previously evacuated, was used in the liberation of carbon dioxide.

Several estimations were performed on the same sample of urine. The specific gravity
of some specimens was taken, the sodium chloride content by Volhard's method, total
carbon dioxide content, and carbon dioxide combined as carbonate and bicarbonate.
The dissolved and combined carbon dioxides were differentiated in this way. The total
carbon dioxide in the urine was liberated in the burette with the aid of lactic acid.
Another portion of the same sample was thoroughly aerated and treated in the burette
similarly. Aeration was repeated until a constant minimum carbon dioxide content was
found. The difference between the first and second estimations of urine was taken to
be the volume of dissolved gas. Professor Krogh has stated that this method of dis-
criminating between free and combined carbon dioxide is approximately valid only
when the fluid is fairly acid (pH < 5). It was found, however, that the volume of
carbon dioxide which was removed by aeration was exactly equal to the volume which
could be liberated by evacuation without reagents. In other words the method was
approximately accurate for the measurement of dissolved carbon dioxide.

Table I. Gas and salt contents of whale urine

I

2

3

4

5

6

No.

N2 in
solution
vol. %

N2 capacity
vol. %

Total CO2

vol. %

CO3 as CO2
vol. %

NaCl

mg./cc.

Remarks

I

0-88

—

9-60

—

—

2

0-53

—

—

—

—

3

o-8o

—

—

—

—

4

070

—

—

—

—

5

0-89

—

—

—

—

6

1-04

—

7-20

—

—

7

0-94

o-8o (18°)

10-50

—

—

8

0-53

0-89 (15°)

33-40

3-95

20-60

9

0-41

0-88 (15°)

40-00

3-90

16-10

10

0-65

0-83 (15°)

22-00

—

18-70

II

0-64

—

24-00

—

12

0-88

073 (35°)

io-6o

3-58

23-36

13

o-8i

13-60

—

—

14

1-14

0-S2 (36°)

8-30

1-20

—

IS

0-87

—

29-00

—

21-72

16

I -06

—

13-20

—

—

17

1-05

1-09 (12°)

13-80

1-40

23-35

18

171

—

6-30

—

22-78

19

o-8i (18')

32-80

—

13-57

20

072

0-63 (36°)

35-20

11-50

13-33

21

1-04

io-8o

—

26-60

22

1-26

0-65 (36°)

17-60

—

25-99

23

078

o-6o (36°)

10-80

2-50

23-68

24

0-92

o-6o (36°)

15-50

0-75

22-8o

25

o-6o

—

40-10

—

13-50

26

0-94

—

10-80

—

22-54

27

1-39

o-6i (36°)

io-6o

2-60

23-35

28

0-35

—

43-00

—

15-25

29

0-87

—

14-40

—

26-25

30

0-85

—

7-40

—

24-30

31

0-36

—

38-00

—

14-44

32

0-68

—

26-00

—

2i-8o

33

0-91

i-oo (15°)

I2-20

—

22-49

34

0-69

—

13-30

—

i8-8o

35

0-36

—

14-00

—

15-78

Some gas in

36

079

—

48-00

—

19-70

[bladder

37

I -20

—

14-20

—

23-00

38

0-98

—

21-20

—

22-40

39

1-53

—

14-80

—

25-80

40

0-62

0-95 (14°)

35-50

0-73

15-02

41

072

—

13-60

0-17

23-30

S.G. 1-033

42

0-90

—

13-60

0-37

22-95

1-034

43

—

—

22-6o

—

9-99

1-029

44

—

—

21-00

0-72

23-00

1-038

45

—

—

15-60

1-20

20-90

1-033

46

—

—

11-00

0-33

22-00

1-034

47

—

—

15-40

—

23-20

1-036

48

—

—

io-8o

—

23-20

1-035

49

—

—

11-50

—

24-00

—

50

—

—

22-10

2-10

19-72

1-034

51

—

—

18-60

0-53

21-18

1-034

52

—

—

17-40

20-02

I -03 1

53

—

—

20-I0

25-65

I -03 1

54

—

—

1 1 -20

25-60

—

55

—

—

7-80

24-30

1-037

56

—

—

83-20

42-00

7-10

i-i8o*

* Protein 30 grm./iooo cc.

378 DISCOVERY REPORTS

In a few cases the solubility of air nitrogen in the urine at atmospheric pressure was
estimated at various temperatures. The procedure consisted in shaking the urine in a
flask open to the air with a thermometer dipping in the Hquid, the whole operation being
performed in a water bath. The urine was pipetted ofl^ and introduced into the gas
burette in the usual way. In order to form a standard of comparison, the technique was
as far as possible the same as in treating the sample originally. As will be seen from the
figures in column 3, Table I, the solubility of nitrogen in whale urine is low both at
36° C. and at room temperature. Hill quotes two figures for normal nitrogen content
in human urine at atmospheric pressure, 1-14 and 0-90 vol. per cent, of which he prefers
the latter. The corresponding figure for whale urine is o-6o vol. per cent. Whale urine
contains more salt than human urine, which should not have much eff'ect on the solu-
bility of a gas ; but whale urine also contains considerable quantities of protein. ^ Every
sample produced a thick precipitate when boiled ; a few samples were estimated gravi-
metrically and an average of 7 g./iooo cc. was obtained. It was apparent from the
appearance of many samples which were boiled that the protein content varied con-
siderably, reaching in one case (no. 56) 30 g./iooo cc. The eff'ect of protein would
be to reduce the solubility of gases in the liquid, but there may be other reasons for the
low figures obtained. The technique of estimating the solubility was the same as that of
estimating the original gas content, and the results can therefore be taken to be relative.
The conclusion is then that the urine is supersaturated with nitrogen in some cases and
not in others, though the balance is in favour of supersaturation of a low order. The
average content of forty-one samples was 073 vol. per cent, while the solubility was
0-62 at the temperature of the living whale. The issue may have been clouded by the
presence of numerous small organisms in the urine, further reference to which will be
made in dealing with blood.

The carbon dioxide content of the urine is of interest. In column 4, Table I, some
large volumes of total carbon dioxide are recorded in fresh urines ; considerable variation
is shown, and there is also a variation in the amount of combined carbon dioxide present
in those samples which were estimated. Large amounts of combined carbon dioxide are
rare and probably have no connection with the balance of respiratory gases. Specimens
20 and 56 had large combined carbon dioxide, especially no. 56, which is an exceptional
specimen in several other ways — high specific gravity, very low salinity, and very high
protein.

Control experiments were made with human urine, which was treated in exactly the
same way. Variable results were obtained from the same subject, and the dissolved
carbon dioxide ranged from 1-24 to 5-40 vol. per cent. The combined carbon dioxide
was below i-o vol. per cent.

In Fig. I the volumes of carbon dioxide found in whales' urine have been arranged
along a line which represents the solubihty graph of water at 36° C. Each volume found

1 A number of specimens were examined under the microscope and found to contain millions of organisms
identical in appearance with those referred to in the section on blood. It is possible that these organisms
are excreted through the kidneys.

RESPIRATION IN SOUTHERN WHALES

379

is placed on the line opposite the appropriate ordinate so that it is possible to read off on
the abscissa the partial pressure of carbon dioxide which is necessary to cause that
volume of gas to go into solution. The results of the control experiments are also re-
corded. By analogy with the behaviour of nitrogen it appears that the gas content of
human urine tends to be slightly above that which would be expected from an equili-
brium of the carbon dioxide in the alveolar air, standing constantly at 40 mm. of
mercury, with the blood and hence with the other body fluids. The average human

40

30

in

>

S20-

X

a

z

o

en

Qi

<

© HUMAN CONTROLS
• WHALE RESULTS

200 300

PARTIAL PRESSURE OF CO^ [mm, HgJ

400

500

Fig. I. Carbon dioxide contents of whole urine superimposed on the solubihty
curve of carbon dioxide in water at 36° C.

urine gas content corresponds to a partial pressure of about 60 mm. of mercury m
contrast to the 40 mm. which was expected, a state of affairs which might be accounted
for by the diffusion direct into the urine of the carbon dioxide generated in the
capillaries surrounding the kidney tubules. The main point is that in comparing the
carbon dioxide pressures in human and whale urine a variation of 20 mm. either way
is of no great significance. It should be mentioned that in those cases where the com-
bined carbon dioxide has not been directly estimated the dissolved carbon dioxide has
been computed by subtracting the average combined gas, 2-20 vol. per cent, from the

3

D VII

38o DISCOVERY REPORTS

total carbon dioxide. This average does not include the combined carbon dioxide in
sample 56, which is taken to be exceptional.

The graph shows that twenty-six out of the fifty-six gas contents are grouped round
a partial pressure of 140 mm. of mercury, or roughly twice that of the human controls,
three agree approximately with human tensions, and the remainder are spread fairly
evenly over tensions ranging from 240 to 540 mm. of mercury. It must be remembered
that the tensions of carbon dioxide in the lungs which gave rise to these volumes of gas
in the urine through the medium of the blood can have originated in one of three ways :
by accumulation of carbon dioxide in the lungs during a long dive, or by a compression
of the carbon dioxide, or by a combination of the two. It is not therefore possible to
judge the depth of diving which caused a given tension of carbon dioxide unless the
duration of the dive is also known. But it is perhaps significant that such a large
accumulation of gas per se would derange the hydrogen-ion concentration of the blood
and other fluids, while increased pressure would have the effect of transmitting carbon
dioxide through the blood to the urine at high tension, but without seriously altering
the reaction of the blood. It is therefore likely that the carbon dioxide tensions implied
by the condition of the urine are an indication of the depth to which the whale dived
during the last half hour or so of its life. The group mentioned above which centres on
140 mm. of mercury in all probability represents the general state of carbon dioxide
conditions in the body, and it suggests a habitual depth of submersion of between 20
and 30 m. The remaining whales showing tensions higher than 140 mm. of mercury are
those which dived deep and long in their efforts to escape from the harpoon, while the
group of three low tensions is evidence of a small minority which had been basking or
feeding at the surface for some hours previous to capture. (A small proportion of whales
are killed by the first harpoon without a struggle.)

The interpretation of these figures must be open to suspicion because urine is con-
stantly being generated and excreted and hence reflects only the whale's recent activities.
A stationary body of fluid, such as allantoic fluid, which will be considered later, shows
with more precision the whale's normal or habitual condition.

The salt contents of these samples of urine, some of which are recorded in column 6,
Table I, show considerable variation. That there appears to be some relation between
the dissolved carbon dioxide and the salinity of the solvent is shown in Fig. 2, from
which it may be concluded that to some extent the salinity is inversely proportional to
the volume of dissolved gas. These results may indicate that the presence of carbon
dioxide at high pressures in the body of the whale is accompanied by an increased ex-
cretion of urine and that this diuresis serves to carry away considerable quantities of
carbon dioxide which would otherwise discommode the whale. It is impossible to say
whether this diuresis, which can be said to coincide with times of high external pressure,
is a direct result of increased difference between the blood pressure in the glomeruli and
the pressure in the ureter, giving rise to increased flltration, or whether it is caused by
more deliberate control of kidney activity, e.g. by dilatation of kidney blood vessels.
It is plain that an aquatic animal has ample opportunity for passing water through the

RESPIRATION IN SOUTHERN WHALES

381

body if need be, though the absorption of water isotonic with the blood from sea water
would entail the expenditure of energy. Under the circumstances it is probable that the
extra water thus excreted, whatever the reason, is withdrawn from the blood plasma
temporarily when the whale is under high pressure, compensation being made from the
sea at leisure when severe demands are not being made upon the whale's metabolic
resources.

DISSOLVED CARBON DIOXIDE VOLS/.

Fig. 2. The relation between the dissolved carbon dioxide and the
sah content of whale urine.

Professor Krogh objects to this theory of carbon dioxide excretion on the ground that
the quantities thus disposed of would be negligible, since they would correspond at
most only to the volume of urine compared with the volume of the whale. It is not
known by what means whales are able to avoid taking up salts from their food or from
sea water swallowed. It is evident that a mechanism of some sort exists, so that there is
no reason why whales should not extract from the sea unlimited quantities of water
isotonic with the blood and pass it immediately through the kidneys.

3-2

382

DISCOVERY REPORTS

ALLANTOIC FLUID
Table II. Gas and salt content of allantoic fluid {Blue whales).

The dissolved COj is obtained by subtracting " CO3 as COg" from "Total CO2".

No.

No in
solution
vol. %

Nj capacity
vol. %

Total CO2
vol. %

CO3 as CO2
vol. "0

NaCl
mg./cc.

I

0-86

o-6o (35°)

19-1

3-5

176

2

1-17

—

29-0

I i-o

5-70

3

078

—

27-5

—

2- 1 6

4

1-67

0-87 (36°)

34-0

II-O

4-15

5

078

—

30-1

—

3-15

6

0-69

—

40-0

—

o-8i

7

0-55

—

28-6

—

0-94

8

0-87

—

30-5

—

i-i6

9

079

—

31-0

—

6-34

ID

—

—

—

—

1-42

11

—

—

34-1

13-2

Trace

12

1-17

—

31-0

—

3-50

13

I-I3

—

367

—

2-88

H

0-62

—

75-°

51-0

7-08

IS

1-62

—

55-2

33-0

4-05

16

1-07

—

37-8

15-0

1-35

17

—

—

41-0

19-3

1-35

18

—

—

24-6

II-2

1-35

19

—

—

70-6

36-9

271

20*

—

—

24-2

17-8

I -61

21

—

—

74-4

44-1

6-93

22

—

—

41-1

7-15

* Gas found over fluid in allantois.

The results of analyses of allantoic fluid, performed in exactly the same way as has
been described for urine, are shown in Table II. The allantois in Blue whales is a large
bag occupying the fork of the division of the umbilical chord, which bifurcates at some
distance from the umbilicus — about 2 m. when the foetus is half developed. The
urachus runs in the middle of the chord and opens direct into the allantois. The capacity
of the allantois increases with the growth of the foetus and contains about 50 1. of fluid
when the foetus is half developed.

Slight supersaturation with nitrogen is again apparent. More attention has been paid
to carbon dioxide content, which is seen to be high, as in urine. Large amounts of com-
bined carbon dioxide were found in the twelve samples in which this was estimated.
The dissolved carbon dioxide, obtained as before by subtracting the combined from the
total carbon dioxide, shows less variation than in the urine; the range is from 13-4 to
337 vol. per cent, excluding no. 20, in which a large bubble of gas was seen above the
liquid before the allantois was punctured. The average content is 22-4 vol. per cent;
four samples have contents agreeing with the average to within i vol. per cent. These
figures for allantoic fluid are considered to represent more nearly the normal conditions

RESPIRATION IN SOUTHERN WHALES 383

of carbon dioxide tension in the whale in contrast to the temporary states reflected in
the urine. The average carbon dioxide tension responsible for these contents is about
320 mm. of mercury.

LIQUOR FOLLICULI
One sample only was taken from a ripe ovarian follicle and submitted to the same
tests. The total carbon dioxide was 69-6 vol. per cent. The combined carbon dioxide
was 45 vol. per cent, so that the dissolved carbon dioxide was 24-6 vol. per cent, a
figure which approximates to the average content of allantoic fluid.

BLUBBER
In view of the findings mentioned above (Hill) on the high solubility of nitrogen in
fat, special attention was paid to the possibilities of extracting the gases from blubber.
The spongy connective tissue between the blubber and the muscles, which is particularly
plentiful on the throat and belly of the Blue whale in the region of the ventral grooves,
was found to be inflated and in a condition which resembled whipped white of egg.
Portions of this material were taken and the gases extracted. It was found that the
tissue was inflated with almost pure air. The average percentage of oxygen in five ex-
periments (five different whales) was 21-33 with a range of 2 1-09-21 -60; the remaining
78-67 per cent consisted of nitrogen and inert gases. This is doubtless due to the practice
of inflating whales by compressed air after death in order to render them buoyant for
towing. The syringe is inserted through the blubber and the air is blown into the space
between the blubber and the musculature. There is no evidence that the air has any
effect on the gas contents of blood and body fluids except perhaps on the blood in the
superficial musculature, since the volumes of oxygen in body fluids have always been
found negligible. Occasional deep penetration of the syringe has resulted in an inflation
of the abdominal cavity. The percentage of oxygen in the injected air is too high since
atmospheric air contains not more than 20-9 per cent. The explanation of this is pro-
bably to be found in the great solubility of nitrogen in fat as a result of which more
nitrogen than oxygen has dissolved into and diffused away in the blubber before the
sample is taken. The gas contents of blubber thus have no significance. The only other
convenient source of fat was the peritoneal fat, which is plentiful in whales in good con-
dition. But it was considered dangerous to rely on results from this material, as the
proximity of the stomach and its fermenting contents might introduce large errors.

BLOOD
Method of collection. For reasons which have been mentioned in other sections,
it was desirable to draw samples of blood from parts of the body in which one might
expect the blood to be as nearly as possible as it was when the whale died. It was there-
fore decided to take samples from the small arteries which are found in profusion
running in the blubber on the top of the head. The head was chosen because it is a large

384 DISCOVERY REPORTS

inflexible mass wherein the blood would not be disturbed by the undulations of the
body during the towing of the whale to the factory ship, and because it is far distant
from both the air blown subcutaneously into the belly region and the fermenting ali-
mentary tract. When the blubber is stripped off the head the blood spurts from these
small arteries. In order to collect the blood without exposure to the air, a small glass
cannula was inserted into an artery and connected to a large test-tube which contained
paraffin oil. The cannula and connecting tubing were filled with oil. The blood then
ran through the cannula into the test-tube under the oil until the oil reached the top of
the tube. A cork was then inserted, displacing some of the oil and sealing the tube,
which was then taken below to the laboratory. Foetal blood was collected by cutting
open the thorax and heart in one motion and rapidly submerging a tube in the blood
which welled up freely. The cork was inserted under the blood as in collecting urine.
On occasions when the blood could not be taken below immediately the tube was im-
mersed in ice cold water.

The technique of gas analysis. In this series of estimations of the blood gases
special attention was paid to the conditions of oxygen and nitrogen. Less attention was
given to the carbon dioxide values of fresh blood, for it was found that plain evacuation
of a sample of blood did not extract carbon dioxide completely although all the nitrogen
and oxygen were extracted. Since the main object was to assess the dissolved nitrogen,
no lactic acid was used to expel the carbon dioxide, as the addition of acid might have
caused slight flocculation of the protein constituents of the blood to the occlusion of
some of the nitrogen. The extraction of oxygen and nitrogen was carried out without
reagents. At first gas-free water was used as a diluent to hasten evacuation, but, after
some control experiments had shown that extraction of nitrogen was as rapid without
the water, the diluent was abandoned to simplify the manipulation. The procedure was
to evacuate the blood with frequent shaking for 10 min., isolate the blood in the lower
half of the burette, and absorb carbon dioxide and oxygen with the usual reagents.

Nitrogen and oxygen in fresh whale blood. The nitrogen and oxygen content of
samples of blood are shown in Table III. It was impossible to obtain accurate data as
to the time that had elapsed from the death of the whale to the drawing of the sample,
but in no case had the whale been dead more than 12 hours and the majority less than 6.
The only exception is sample no. 99, which was deliberately taken from a decomposing
whale. The table shows the resuks of all the analyses which were performed. The sig-
nificance of the nitrogen and oxygen contents will be discussed later ; for the moment
it will be enough to point out that only six samples (nos. 27, 49, 68, 69, 74, 99) show
supersaturation to correspond to a pressure of between 2 and 3 atmospheres, while only
one of the six (no. 69) indicates a pressure of more than 4 atmospheres in the air
nitrogen in the lungs. Thus supersaturation in blood, a viscous colloidal fluid, is more
rare than in urine and allantoic fluid.

It will also be seen in Table III that the majority of the blood nitrogen volumes are
less than the normal solubility of atmospheric nitrogen in blood at barometric pressure.
The average content for the whole series, including the supersaturated samples men-

RESPIRATION IN SOUTHERN WHALES

38s

Table III. Nitrogen and oxygen contents of arterial whale blood.

F after the serial number denotes foetal blood. The sample numbers are not consecutive because not all
the samples were examined for these gases.

Sample no.

N2 vol. %

O2 vol. %

Sample no.

N2 vol. %

O2 vol. %

6

1-20

0-96

64

I -00

0-55

10

1-30

0-89

66 F

0-22

0-00

12

0-53

3-52

67

1-55

0-66

14

1-44

1-40

68

2-95

0-00

15^

071

0-35

69

5-60

0-00

16

i-i8

0-35

70

1-50

0-00

17

1-03

0-45

71

170

0-15

18

I-IO

07s

72 F*

176

0-54

22

178

0-65

73

1-05

0-35

23

o-6i

—

74

270

0-00

25^

0-53

—

75

1-03

0-50

26 F

0-26

0-82

76F

0-40

0-20

27

2-75

0-20

77

0-89

0-89

28F

0-33

0-22

78

0-88

0-23

29

0-66

0-33

79

I-OI

0-33

30F

1-56

I -60

81

0-33

0-33

31

i-io

2-00

82 F

0-46

0-00

32

1-03

0-99

83 F

0-30

0-45

33

°-55

0-55

84

1-30

0-23

34

077

0-62

85 F*

0-65

o-io

36

1-09

1-31

86

077

0-99

37

0-85

0-20

87

I-I3

I-I3

38

0-63

0-30

88

1-69

o-io

39

0-90

o-8o

89

0-89

0-23

40

1-30

0-22

90

0-89

I-IO

42

1-04

0-26

91

I-IO

0-89

43

174

0-17

93

I -08

0-04

44

1-07

0-17

94

0-68

0-22

45

i-ii

0-44

95^

1-23

0-33

48

1-54

0-66

96

1-34

o-o8

49

2-95

o-oo

97

1-68

o-oo

50

I -06

o-6o

98 F

0-44

0-00

51

0-56

1-46

99t

270

0-34

52

0-48

0-56

100

I-I2

0-22

54

1-34

0-34

loiF

0-37

0-00

55

0-82

0-51

102

o-i8

0-00

56

0-55

0-22

103

0-66

0-44

57

i-ii

104

1-56

O-IO

58

i-ii

0-22

105

i-iS

0-25

59

073

0-49

106

075

0-35

60

1-23

0-45

107

I -06

o-i8

61

1-23

0-45

108

0-98

0-34

62

072

0-57

no

1-22

0-15

63 F

0-66

0-66

III

0-92

0-22

* Foetus of preceding whale.

-f Stale blood. 5 hours later the inert gas content had increased to 4-60 vol. per cent.

tioned, is i-ii6 vol. per cent, or slightly below the human blood figure, which is
I -20 vol. per cent.

386

DISCOVERY REPORTS

Solubility of nitrogen in whale blood. So far comparison has been, made be-
tween the nitrogen volumes existing in whale's blood and the human solubility coefficient
for dissolved atmospheric nitrogen. The peculiar circumstance has been revealed of a
mammal having less nitrogen in solution than should have been dissolved normally in
accordance with established physical laws, even without high pressures which should
cause the solution of even greater volumes of gas. When, however, control experiments
were performed to ascertain the solubility of air nitrogen in whale's blood, it was found
that more than the expected i-20 vol. per cent were taken up. These experiments con-
sisted simply of exposing blood samples to air by shaking them in an open flask, both
at room temperature and at 36° C. Gas analysis was performed as for the original
sample. Variable results were obtained by this method, the full significance of which
was not realized until it was found that inevitable delay in manipulation influenced the
volume of residual nitrogen. A number of nitrogen "capacities", some of which
represent residual nitrogen in estimations of the oxygen capacity of the blood, are
shown in Table IV. Reagents used in oxygen estimations were, of course, carefully
evacuated and tested before introducing the blood into the burette.

Table IV. Nitrogen capacity of whale blood from aeration followed by gas analysis.

Sample no.

Nj capacity
vol. %

Sample no.

N2 capacity
vol. %

3

2-8l

27

3-33

5

2-58

29

177

6

2-02

30 F

2-i8

7

2-95

33

1-58

8

2-30

ZSF

1-77

9

2-62

37

1-42

10

i-8i

38

1-62

II

2-53

40 a

1-43*

13

1-63

406

3-20

14

1-28

45

1-93

16

1-42

65

2-36

17

2-02

78

1-58

21F

1-55

79

1-68

22

2-20

81

1-99

23

2-o6

94

1-59

2SF

2-83

* 10 minutes' interval between aeration and gas analysis.

It is evident that whale blood is able to take up more than the normal amount of
nitrogen from the air. The highest figure found by this method was 3-20 vol. per cent
in no. 40. The plasma of no. 7, which sedimented rapidly, was treated in the same way
in two separate experiments; the nitrogen capacity was in each case 3-14 vol. per cent.
Unfortunately not much attention could be paid to plasma for lack of an effective
centrifuge, but the results obtained with this sample indicate that the ability to hold
extra nitrogen resides in the plasma and not in the corpuscles, a conclusion which will
be substantiated in other ways later.

RESPIRATION IN SOUTHERN WHALES

387

Another method, in which the progress of nitrogen into solution was directly ob-
served, was used for determining the nitrogen capacity of the blood. The procedure is
as follows. A sample of blood whose initial nitrogen content is known is put in the gas
burette and a known volume of air is introduced over the blood. The exact volume of
air is measured wet as it stands over the blood after the upper stopcock has been sealed
with mercury. The burette is now inverted so that the air in the narrow graduated
portion is displaced by mercury and forced into contact with the blood in the bulb of
the burette. Vigorous shaking of the burette helps the mixing of the blood and air. At
intervals, the burette, in which the blood and air have been carefully kept at atmo-
spheric pressure, is restored to the vertical position and the volume of gas standing over
the blood is read. Inversion and shaking are repeated between readings, and at each
reading a trace of sodium hydroxide solution is run into the burette to absorb the carbon
dioxide evolved by the blood. All the oxygen in the air is absorbed in the first two
shakings. Before reading the volumes of nitrogen a trace of oxygen absorption mixture
is added to ensure absence of oxygen. Protocols of an experiment on these lines are as
follows :

Sample no. 100: nitrogen content 1-12 vol. per cent; uncorrected volume 1-25 vol. per cent.
10 cc. of this blood enclosed in burette with 0-945 ^^- of ^ii" measured wet at 15° C.

Initial nitrogen

in the air

0-747 cc

After 5 min.

oxygen +

nitrogen

0-760 „

» 10 „

nitrogen

0-710 „

„ 20 „

0-650 „

,. 25 „

0-640 ,,

,. 30 "

0-620 ,,

.. 35 >.

0-610 „

„ 40 ..

o-6io ,,

At the end of 35 min. the blood had absorbed 0-747 — 0-610 cc. = 0-137 cc.^

cent at room temperature and pressure.
So that the nitrogen in the blood was then 1-25 + 1-37 = 2-62 vol. per cent.

1-37 vol. per

The results of this experiment and others are shown graphically in Fig. 3, where the
volume of nitrogen absorbed in addition to the volume already in solution have been
plotted against time. It is of course to be realized that the means used for mixing the
blood and air were inefficient, and that under ideal conditions of mixing the absorption
of nitrogen would in all probability proceed much faster.

While different samples of blood give different nitrogen capacities, the nitrogen
capacity of any one sample is fairly constant. For example, the nitrogen capacity of
no. 108 was estimated four times: (i) starting from the initial nitrogen content of
I -08 vol. per cent (uncorrected), (2) after evacuation, (3) after a second evacuation and
treatment with pure nitrogen instead of air, (4) after removal of nitrogen after (3) by
a third evacuation. The nitrogen capacities were respectively 2-32, 2-34, 2-66, and 2-36.
The higher capacity in (3) is the result of using pure nitrogen, and the difference be-
tween it and the other capacities is approximately the volume of extra nitrogen which
would go into solution in water when the partial pressure of the gas was 760 instead of

388

DISCOVERY REPORTS

608 mm. of mercury as in air. Octylic alcohol, which is commonly used in gas analysis
to prevent frothing, tends to diminish the nitrogen capacity. Fortunately whale blood

SAMPLE N5I00
SAMPLE N2I0IF
SAMPLE N2 103
SAMPLE N5 104
SAMPLE N2 lOG
SAMPLE N2 lOB

ID

30

40

20
TIME- MINUTES

Fig. 3, Progressive solution of nitrogen in whale blood.
* Octylic alcohol was added at this point and nitrogen came out of solution again.

seldom froths or these experiments would have been impossible. During one of these
experiments, on sample loi, it became difficult to make a direct reading of the volume
of gas overlaying the blood, and a trace of octylic alcohol was added with the sodium

RESPIRATION IN SOUTHERN WHALES 389

hydroxide. Immediately the absorption of nitrogen was reversed as is shown m the
graph. Octyhc alcohol was added at 23 min. after the commencement of the experiment.
The nitrogen content of this sample dropped from 2-29 to 1-40 at the end of i hr. and
20 min., a figure little above the solubility of nitrogen in other types of mammalian
blood. Subsequently it was found that whale blood would never take up nitrogen to its
full capacity when there was contamination with octylic alcohol. This effect was not
observed in control experiments on the solubility of nitrogen in fresh pig's blood with
and without octyHc alcohol, wherein the solubility was i-ii vol. per cent at 36° C.

It now appears in the face of these nitrogen "capacities" that the samples of whale
blood, whose average content was i-ii vol. per cent, instead of being merely not
supersaturated by comparison with the normal nitrogen solubility of blood are in fact
all "sub-saturated " by comparison with their own capacity for nitrogen. Such a state
of affairs is plainly at variance with the known application of Dalton's Law on the
solubility of gases. It can only be explained by postulating either a state of negative gas
pressure in the lungs of living whales, which is unthinkable, or a mechanism for dis-
posing of dissolved nitrogen from the blood.

DISAPPEARANCE OF NITROGEN FROM BLOOD
In the figures relating to the estimation of nitrogen capacity by aeration of the blood
followed by gas analysis (Table IV), it was shown that the volume of nitrogen extracted
from the blood was variable. The explanation of this is to be found in the peculiar fact
that nitrogen which goes into solution in the blood is not completely recoverable by the
application of a negative pressure to the blood as in gas analysis. That is to say that at
any rate part of the nitrogen which goes into solution in the blood is retained and does
not obey normal solubility laws ; the passage of nitrogen into the blood is to some extent
irreversible.

Evidence of this phenomenon was found in one of the early experiments on nitrogen
capacity when the accuracy of the method was being tested. Sample 1 1 was aerated
and 20 cc. of the blood were immediately enclosed in the lower Umb of the burette,
safely bottled between the main stopcock above and mercury below, so that portions of
the blood could be raised to the upper half for gas analysis at intervals with the minimum
of manipulation. The blood was aerated at mo hours and immediately transferred to
the burette. After evacuation of the upper half of the burette and testing for leaks the
first sample was ready to be analysed at

1 1 17 hours when nitrogen content of blood was 2-53 vol. per cent.

1140 ,, )) >) )> >) ^'35 "

1210 ,, „ „ „ ,, I'l? "

The nitrogen content of this sample, whose capacity was not less than 2-53 vol. per cent
(and possibly more since there was inevitable delay in manipulation), decreased by
1-36 vol. per cent in 53 min. The method employed precludes any escape of nitrogen
from the stored blood ; and in the event of nitrogen being evolved the gas would only

4-2

390 DISCOVERY REPORTS

rise to the top of the stored fluid and accompany the next sample of blood into the upper
part for analysis.

Experiments of the kind performed with sample no. loo may be extended by checking
the final nitrogen content of the blood as soon as the absorption of the nitrogen is com-
plete. This is done by driving out the remaining air nitrogen, sealing the upper stopcock
with mercury, and immediately evacuating the blood. Nitrogen in the reagents is of
course deducted from the volume obtained. If this part of the experiment is done the
instant the absorption is complete, the same volume of nitrogen can be extracted from
the blood as is theoretically contained in it, namely the original content plus nitrogen
gone into solution. But if this extraction is delayed nitrogen disappears and the final
content is less than the original plus added gas. The nitrogen capacity experiment on
sample loo (p. 387) was completed by extraction of the gases from the blood at 50 min.
after the beginning of the experiment. The nitrogen content was 2-47 vol. per cent,
whereas the volume of nitrogen dissolved in the blood plus its original content was
2-62 vol. per cent. The volume of nitrogen which had disappeared was therefore
0-15 vol. per cent uncorrected, or 0-13 vol. per cent at n.t.p.

A number of experiments on these lines were performed both with adult and foetal
whale blood, difl^erent times being allowed after complete solution of nitrogen before the
final blood gases were estimated. The volumes of blood and air employed were varied
at random, so that there was no possibility of the disappearance of nitrogen being the
result of a constant error. The experiments were all performed at room temperature
since it was not practicable to reproduce the temperature conditions of the living whale
in these experiments on board ship. Similar results have been obtained by aerating
blood and immediately covering it with a thick layer of paraffin. Successive samples are
pipetted from beneath the oil and analysed as before.

This disappearance of nitrogen goes far towards explaining the low nitrogen contents
of the fresh samples of blood (Table III). It is impossible that the blood could have been
exposed to nitrogen at less than atmospheric pressure during the last minutes of a
whale's life; the probabilities are, on the evidence of urine and allantoic fluid, that the
air pressure in the lungs was considerably above normal. The slight supersaturation
which has been observed in the urine indicates, in conjunction with the resuhs of these
experiments, that the blood was momentarily supersaturated and had not become de-
nitrogenated before it reached the kidneys through which some of the excess nitrogen
diffused into the urine. At atmospheric pressure, as we have seen, the blood can take
more than 2 vol. per cent into solution, and yet such a nitrogen content is seldom found.
It seems very likely that the sequence of events which formed the basis of the experi-
ments described above is in fact a recapitulation of what happens to the blood when it
passes through the lungs, becomes aerated, and subsequently in its passage through the
arteries and veins becomes denitrogenated by some internal mechanism. The samples,
which were drawn from arteries in the heads of dead whales, had stood long enough for
more or less denitrogenation to have occurred according to the time which had elapsed
since death and other factors.

RESPIRATION IN SOUTHERN WHALES

391

Opportunity was taken to determine the consumption of oxygen which took place in
the blood during the experiments. It had been noticed that blood became appreciably
deoxygenated when it was aerated and left to stand for the determination of the pro-
gressive decrease in nitrogen content. This was perceptible by the change in colour of
the blood apart from the evidence of gas analysis. If pure nitrogen is supplied to blood
which has been deprived of oxygen there is no disappearance of nitrogen.

The results of these experiments together with the volume of blood taken, oxygen
consumed, time, temperature, and ratio of nitrogen disappearance to oxygen consump-
tion are set out in Table V. The object of the experiments was to establish the fact of
nitrogen disappearance and not the rate, which could not be determined by such an
inefficient method of mixing blood and gases.

Table V. Disappearance of nitrogen in adult and foetal whale blood.

O2 con-

Sample
no.

Vol.

of blood

cc.

Vol. of air
supplied

cc.

N2 dis-
appearance
vol. %

Time
min.

O2

consumed

vol. %

sumed/N,

disappeared

ratio

Temp.
°C.

78

22

1-03

0-58

80

1-22

2-10

20

79

ID

I -02

0-85

135

0-65

0-76

16

81

10

0-92

1-32

45

3-53

2-68

i8-s

83/'

10

0-91

1-29

240

3-30

2-56

17-5

86 a

8

0-96

0-46

75

2-50

5-44

19

86/)

18

0-97

0-43

180

1-17

272

12

87

20

0-95

0-41

60

1-32

3-22

13

89

10

0-94

0-26

120

1-58

6-10

i8-5

92

10

I -00

0-44

90

I -80

4-10

18

94

10

0-97

0-58

140

279

4-81

13

99*

10

0-89

2-70

40

2-22

0-82

20

100

10

0-95

013

5°

1-77

13-60

15

101 F

10

0-96

0-43

21

1-43

3-33

15-5

104

10

0-97

0-45

180

1-84

4-10

17

105

10

0-98

060

60

1-92

3-20

18

106

s

0-97

0-85

25

3-60

4-22

17

loSa

10

0-93

0-65

20

1-67

2-57

15

1086

10

0-98

0-78

17

2-30

2-95

15

io8c

9

o-93t

000

18

0-15

—

15

io8(/

8-5

o-97t

0 00

23

0-00

—

15

io8e

8

0-79

0-72

20

1-90

2-64

15

109^

10

o-9ot

000

60

o-o6

—

15

1096

8

0-93

1-31

30

2-30

177

16

no

6

0-97

1-38

45

3-13

2-27

12

* Stale blood. f Pure nitrogen supplied instead of air.

The considerable variation in volumes of nitrogen disappeaiance which is shown
bears no relation to the duration of the experiment. Consumption of oxygen is a notable
accompaniment and indicates catabolic activity in the blood. While it is impossible to
draw any conclusions from the nitrogen-oxygen ratio, it is thought that the ratios shown
against samples 78, 81, 83 F, 866, and 1080, b, and e are more significant than the rest,
since in these samples the blood, before being subjected to the experiment, was refreshed

392 DISCOVERY REPORTS

by being aerated. The volume of oxygen available in the other samples (one-fifth of the
volume of air supplied) would not serve materially to refresh the blood in its deoxygen-
ated state. The ratios recorded against these samples range from 2-io to 2-95 and their
average is 2-60. No nitrogen disappears when no oxygen is supplied (nos. io8c, d, logo),
the amount already in the blood being very small. This suggests that the main oxygen
consumption, apart from some small metabolic activity which the blood may possess,
is a factor in nitrogen disappearance. Control experiments showed no disappearance of
nitrogen in fresh pig's blood.

If nitrogen is constantly being withdrawn from the blood by some mechanism such
as has been indicated above, it follows that on each successive dive there will be more
accommodation in the blood for nitrogen than would normally be the case, with the
result that but little nitrogen would be left in the lungs after a dive of some minutes'
duration ; but in actual fact the terrific blast of expiration shows without doubt that there
is plenty of nitrogen left in the lungs at the end of a longish dive. It is difficult to re-
concile these two facts with each other unless it be assumed that the circulation of a
whale is much slower than that of other mammals. In this connection it is perhaps
significant that the wall of the aorta of a Blue whale is remarkably thin, not more than
I cm. thick while the internal diameter is about 20 cm. If blood were being pumped at
high pressure the walls of the aorta might well be much thicker than this. Such blood
as does pass through the lungs will become supersaturated with nitrogen and will then
be cleared, but it seems unlikely that the whole of the whale's blood could course
through the lungs in the space of one dive.

The application of the phenomenon of nitrogen removal to the prevention of caisson
sickness is obvious. The effect of slow circulation would be of course to diminish the
risk of caisson sickness, although not to abolish it altogether. Even if, as has been
suggested in the preceding paragraph, only a small portion of the blood becomes super-
saturated with nitrogen, that blood, in passing to certain organs, for instance the central
nervous system, will cause these in their turn to become supersaturated with nitrogen
which would become dangerous in the event of sudden decompression.

MECHANISM OF NITROGEN REMOVAL

Blood smears. Special samples of blood were taken from fresh whales and foetuses
for microscopic examination. These were drawn from arteries in the head, as mentioned
above, but with special precautions to ensure that all the apparatus used was sterile.
The collecting cannula and tube into which the blood flowed were wrapped in cloths
and steamed for 5 hours. The wrapping was removed at the instant of using, and the
tube was plugged with sterile cotton wool after being filled.

Plain wet smears of this blood were examined microscopically with a 1/12 objective.
Apart from the usual formed elements of blood which were seen, such as erythrocytes,
leucocytes, etc., vast numbers of very small particles (hereinafter referred to as X
organisms) were seen in the plasma. Their shape was indefinite but approximately

RESPIRATION IN SOUTHERN WHALES 393

spherical, and their diameter varied from 0-5 to 2-0/^. They were motile, insomuch that
they moved with an irregular motion distinct from Brownian movement. They were
white or sometimes pale blue and highly refractive. Unsuccessful attempts were made
to count these particles by haemocytometer. The best way of assessing the number
present was by comparison with the number of erythrocytes. A great variety of numbers
was found, ranging from 10 to 30 million per c.mm. These particles retained their
activity at all room temperatures below 40° C.

Cultures. In the absence of more elaborate facilities than are afforded on board a
floating factory, it was impossible to make as comprehensive an investigation of these
organisms as was desired. They were cultured in a crude manner by placing the whole
blood in a sterile Petri dish, diluting it with freshly boiled 07 per cent solution of
sodium chloride, and keeping it at 30° C. No figures are available to show the numbers
of organisms present from day to day, because of the difficulty of counting mentioned
above, but superficial examination of smears from the culture was ample to show that
the organisms were increasing at a rapid rate. A sample of fresh blood treated in this way
became so congested with X organisms in 4 days that the surface of the culture was
covered with a grey scum which was composed exclusively of the organisms. After some
days, of course, the other formed elements of the blood became disrupted, the haemo-
globin turned to methaemoglobin, and the blood became unrecognizable as such. As
far as could be seen the cultures consisted entirely of the X organisms.

Cultures showed no signs of infection until some days had elapsed, when, owing to
frequent lifting of the lid in an insanitary laboratory, foreign bodies found their way in
and multiplied. An infected culture usually cleared itself if kept covered and undisturbed
for a week. For comparison, a sample of blood was taken from a rotten whale and was
found to be teeming with many kinds of bacteria, including streptococci, in addition to
the X organisms. This sample was treated in the same way as the fresh samples, and in
one week there were no bacteria visible in the blood ; the only sign of life was in the X
organisms, which had increased substantially. The aggregation of organisms at the
surface of the culture suggested that they were aerobic. One portion of a culture was
aerated continually for a week, another was covered with paraffin oil, and the remainder
was left standing open to the air. The X organisms in the aerated portion diminished
steadily, in the covered portion they remained almost stationary, while in the third they
showed a steady increase. It appeared that too much air was harmful.

Experiments were performed on samples of blood which were kept in Petri dishes
until all the corpuscles had disappeared to determine the gaseous condition of the
culture. The methods were exactly the same as for fresh blood. The blood of a Blue
whale (taken 2. i. 33), v/hich had been kept in a Petri dish for 23 days, was subjected to
one of these experiments. The sample had been open to the air under the lid of the dish.

Gas content of the culture: Nitrogen o-66 vol. per cent n.t.p.

Oxygen 0-22 ,, ,,

Carbon dioxide 18-50 ,, ,,

394 DISCOVERY REPORTS

A portion of this culture was treated in the way which has been described for the
experiments to show nitrogen disappearance.

Nitrogen which disappeared ... ... ... 0-36 vol. per cent

Time 23 minutes

Oxygen consumption ... ... ... ... 1-58 vol. per cent

Ratio of oxygen consumed to nitrogen removed 4-40

It will be noticed that the initial nitrogen content of the culture was below the expected
figure of I vol. per cent.

A sample of foetal blood, taken 13. i. 33 and kept for 14 days, was examined. The
nitrogen content of the sample, which had been standing open to the air, was 0-89 vol.
per cent, and the oxygen content o-o8 vol. per cent. On being shaken with air in the
burette this culture immediately absorbed 1-53 vol. per cent nitrogen, making the
nitrogen capacity up to 2-42 vol. per cent. At the same time 1-57 vol. per cent oxygen
were absorbed from the air. Spectroscopic examination of this sample showed that
there was no trace of haemoglobin present, which suggests that the oxygen was ap-
propriated by the X organisms. After z\ hours the sample had disposed of 0-64 vol. per
cent nitrogen, and all the oxygen, 1-57 vol. per cent, which had been taken up was
consumed. The ratio of oxygen consumption to nitrogen removal was 2-45.

Professor Krogh has rightly pointed out that the oxygen consumption recorded in
this and other experiments can hardly represent the oxygen requirements of the
mechanism of nitrogen removal, since the total oxygen in the blood at full saturation
would be insufficient to account for the disappearance of nitrogen taken up at 10 atmo-
spheres pressure and there would be nothing left for the whale's metabolism proper.
He has also pointed out that asphvxic blood may absorb much oxygen which is not
applied to nitrogen fixation and that therefore the observed ratio does not reflect normal
conditions.

Infected pig's blood. Since it was apparent that the ^organisms reproduced rapidly,
opportunity was taken to infect fresh pig's blood from a culture. The fresh pig's blood
was divided into two portions, and all the usual manipulations were performed with the
first. No absorption of nitrogen either temporary or permanent was observed. The
second portion, 100 cc, was infected (20. i. 33) with 5 cc. of a three weeks old culture
and left overnight at 30° C. The following day the blood was seen to have become de-
oxygenated, and on aerating the blood it was apparent that something in it was consuming
oxygen. Experiments with nitrogen showed a decrease in nit.''ogen content of 0-02 vol.
per cent in 50 min., a volume so small as to be within the limits of error proper to the
experiment.

Another sample (20 cc.) of fresh pig's blood was infected (21. i. 33) with i cc. of a
sixteen days old culture of foetal blood. Distinct de-oxygenation of the infected blood
was apparent after i hour. This blood disposed of 0-62 vol. per cent nitrogen in 2 hr.
10 min. A smear of this blood was seen to contain about 10 million organisms per c.mm.
No disappearance of nitrogen occurred in the uninfected control.

RESPIRATION IN SOUTHERN WHALES

395

A third sample of blood was taken as cleanly as is possible on board a factory ship
from a pig at the moment of being killed (25. i. 33) and frozen to —10° C. for 5 days.
The bottle was then allowed to stand for 3 days unopened at room temperature. At the
end of this time, the blood was seen to contain great numbers of the X organisms. No
other infection was apparent. The small organisms were seen to be attacking the
erythrocytes, 10 or 12 organisms to each corpuscle, a feature of their activity which
had been noted in fresh adult and foetal blood when there was little oxygen present.
A number of experiments were performed in the burette. The nitrogen content of the
blood was i • 1 1 vol. per cent. The following amounts of nitrogen were removed from the
blood in six separate experiments :

Time
min.

Nitrogen removed
vol. %

Oxygen consumed
vol. %

Nitrogen-oxygen
ratio

25

25*

°-53

0-00

176
1-76

3-3

30

0-35

1-65

47

30
30

0-29
0-63

0-87
3-26

3-0

5-2

5°

0-47

1-52

3-2

* With octylic alcohol.

So far as the crude technique can show, this blood alone of the samples of pig's blood
became infected with X organisms and behaved exactly like whale blood. Whether
the infection occurred m vivo or in vitro it is impossible to say ; the former is possible
since the pig-sty was close to a deck running with whale's blood.

NATURE OF X ORGANISMS

The behaviour of the organisms which were found in all samples of Blue and Fin
whale blood, adult and foetal, suggests that they may be a kind of bacterium. Their
presence in foetal blood militates against their being any known species of bacterium
since, so far as is known, the very few bacteria which are able to penetrate the foetal
membrane are pathogenic.

The experiments with cultures and pig's blood suggest that the nitrogen removal or
fixation, as perhaps it may be called, is performed by the X organisms. The appropriate-
ness of this fixation in dealing with the problem of caisson sickness has been suggested,
but at the same time the possibility that the bacteria are present in the blood as a result
of post-mortem infection has not been overlooked. The presence of the bacteria in such
huge numbers in the freshest of whales and particularly in their foetuses is strongly in
favour of their being present in the blood of living whales. However, these conclusions
are offered with some degree of diffidence, since it is unheard of that even a benignant
organism should be found in such numbers in the blood of a mammal.

The X organisms' resistance to freezing was indicated in the last experiment with
pig's blood. Subsequently a number of samples of adult and foetal whale blood, which

396 DISCOVERY REPORTS

had been taken in a sterile manner and brought home from the Antarctic in a refrigerator
at — lo'^ C, were thawed and examined. All were found to have A' organisms alive and
active. One bottle was taken to the Rothamsted Experimental Station, where Mr
D. Ward Cutler and Miss L. M. Crump very kindly undertook to examine it. Mr Cutler
has made the following statement :

Whale's blood brought from the Antarctic frozen has been found to contain bodies which are
almost identical in appearance and movement with small particles which may be seen in newly shed
human blood^ (diam. o-^-z-Ofi). They arc definitely motile, though their action is irregular, and not
amoeboid. Small quantities of this frozen blood were introduced into a nutrient solution containing
I per cent glucose, 07 per cent sodium chloride, and a trace of sodium phosphate. In this solution
the bodies were found to grow and reproduce. Various stages in the simple division of the bodies
were observed and at the same time their total number increased :

Thawed blood: number of bodies 30 :, lo^/c.mm.
Blood diluted in nutrient solution 10 times: number of bodies
After 3 days 53 x lo^/c.mm.
„ 4 ,, 28 X lo^/c.mm.
5 „ 82 X io6/c.mm.

Blood diluted in nutrient solution 100 times: number of bodies
After 3 days 36 :- lo^/c.mm.
„ 4 ,, 64 : lo^cmm.
5 .- 52 -■■ io8/c.mm.
Considerable powers of endurance are evidenced by the fact that the whales' blood was kept at
— 10° C. for 2 months with occasional thawing.

OXYGEN CAPACITY OF WHALE BLOOD

Attempts were made in South Georgia in 1 930-1 to estimate the oxygen capacity of

the blood of Blue and Fin whales and their foetuses. A Barcroft differential manometer

was used. The results obtained in the estimation of the capacities of forty samples of

fresh blood were compared with their haemoglobin content. The haemoglobin in the

blood varied greatly from one sample to another, but the oxygen capacity, also variable,

was always far in excess of the theoretical capacity calculated from the percentage of

haemoglobin.

Average haemoglobin percentage 9-62

Oxygen equivalent — vol. per cent 12-97

Average oxygen capacity — vol. per cent 21-30

In the absence of other methods of blood gas analysis, it was thought that the extra
capacity was caused by another respiratory pigment supplementary to haemoglobin.
In the light of the results of the season 1932-3 it is plain that the Barcroft manometer
gave false readings with whale blood. While oxygen was being liberated from the blood
on one side of the manometer by treatment with potassium ferricyanide, which poisoned
any living organisms in this blood, the blood in the dummy bottle was absorbing nitrogen

1 These appear to originate from the disintegration of leucocytes and since they are unable to reproduce
are quite different from the X organisms (A. H. L.).

RESPIRATION IN SOUTHERN WHALES 397

from the air and consuming oxygen, so that the manometer was further deflected by a
negative pressure on the dummy side of the apparatus. While a more or less constant
manometer reading was obtained after 10 min. shaking, it was found that if the system
was allowed to stand for some hours a further deflection of the manometer was recorded,
indicating that some form of metabolism was proceeding at a slow rate.

Oxygen capacities were again estimated in 1932-3 with the Van Slyke burette.
A Haldane haemoglobinometer was used in conjunction with the gas analyses, but the
results are subject to an error of 10 per cent, since, as afterwards appeared, the haemo-
globinometer was faulty. Variable capacities and haemoglobin percentages were again
found, but the actual and theoretical capacities agreed to within 3 vol. per cent. The
average oxygen capacity of seventeen samples was 14-1 vol. per cent. The haemoglobin
in the same samples averaged 9-00 per cent, which is equivalent to an oxygen capacity
of 1 2- 1 vol. per cent.

The haemoglobin percentages from South Georgia and the oxygen capacities found
in 1932-3 indicate that the blood of Blue and Fin whales has small haemoglobin content
and hence small oxygen capacity. It might perhaps be thought that it would be to the
whale's advantage to have blood with a large oxygen capacity for purposes of storage.
This would be so if the whale stayed long enough at the surface for all the blood to come
into contact with fresh air. But, since a whale seldom takes more than two or three
breaths in quick succession, its need for haemoglobin is regulated by the volume of
oxygen which can be contained in the lungs^ for transport throughout the body during
submersion.

The oxygen capacity of porpoise blood was estimated by Morimoto, Takata, and
Sudzuki (1921). Two samples gave capacities of 42 and 45 vol. per cent. The red
corpuscles amounted to 8-4 and 11-2 million per c.mm. respectively, and it was inferred
by the writers that the high oxygen capacity was due to the increased number of cor-
puscles rather than to a high percentage of haemoglobin in each corpuscle. The haemo-
globin percentage is not recorded, nor is the method by which the oxygen capacity was
estimated. It may be that such a high oxygen capacity was obtained by the use of a
Barcroft manometer, while the blood may have had some features in common with the
blood of Blue and Fin whales.

Oxygen dissociation curve. It was not found possible to work out a dissociation
curve of whale blood because the blood when fresh from the dead whale is always acid
in reaction. It was necessary always to make the blood alkaline before a successful
estimate could be made of oxygen capacity. In the circumstances it would not have been
profitable to attempt a dissociation curve since it would have been necessary to derange
the blood reaction and salt content to some extent before starting.

1 For this reason the function of oxygen storage, suggested by Ommanney (1932) for the fat of the retia
mirabiha, seems improbable.

5-2

398

DISCOVERY REPORTS

CARBON DIOXIDE IN WHALE BLOOD
Carbon dioxide content. A number of estimations of the carbon dioxide content
of whale blood were made both at South Georgia and in the factory ship. Since samples
of blood were always taken from arterioles in the head, the contents recorded are, of
course, those of arterial blood. In view of the high concentrations of carbon dioxide
found in various body fluids, it is unlikely that at the end of a long dive, such as usually
preceded the death of the whale, there would be much difference in this respect between
arterial and venous blood. The following figures were obtained:

Carbon dioxide

Carbon dioxide

Sample no.

vol. %

Sample no.

vol. %

31

76-6

Ill

76-0

36

41-0

112

63-0

37

47-0

113

33-o(?)

38

44-0

114

58-4

40

38-0

115F

95-4

43

50-4

116

51-9

48

40-0

117

57-5

49

51-0

119

79-0

110

71-4

120

73-0

In human blood the carbon dioxide content of whole blood oxygenated is approxi-
mately 68 vol. per cent when the partial pressure of carbon dioxide is equal to 100 mm.
of mercury (Bock, Field, and Adair, 1924) and 43 vol. per cent when the partial pressure
equals 30 mm. of mercury. It therefore appears by comparison with human blood that
the partial pressure of carbon dioxide responsible for the contents of carbon dioxide
found in whale blood varies for the most part between 30 and 100 mm. of mercury.
If partial pressures above 100 mm. of mercury occurred during diving, it is probable
that the blood would become saturated in respect of its power of taking carbon dioxide
into combination and any further carbon dioxide would be in physical solution and
easily transmitted to the other body fluids by diffusion. Some attempt was made to
assess the alkali reserve by equilibrating blood with air containing carbon dioxide at
40 mm. of mercury. Variable results were obtained, ranging from 30-0 to 50-9 vol. per
cent, and it is considered that the acid condition of the blood mentioned above was
sufficient to render observations of this kind useless.

Carbonic anhydrase. Two specimens of frozen blood were submitted to Dr F. J. W.
Roughton, for the estimation of the catalyst which accelerates both reactions of the
reversible process HaCOg^HgO + CO2, and which has been found in blood corpuscles
(Brinkman, Margaria, Meldrum, and Roughton, 1932). Dr Roughton kindly estimated
the potency of this enzyme in adult and foetal blood. The following communication has
been received from him :

I define provisionally the unit of enzyme as that amount which, when added to 4 cc. of a mixture
of M/5 phosphate buffer (pH 6-8) with M/5 NaHCOj in equal parts, doubles the rate of COj

RESPIRATION IN SOUTHERN WHALES 399

evolution at 15° C. Ox blood on the average contains per c.mm. about i-i units of enzyme as so
defined.

The adult Blue w^hale blood contained 1-4 units per c.mm.

The foetal Blue whale blood contained 0-35 unit per c.mm. i.e. about 25 per cent of the
mother blood.
The goat foetal blood had usually less than 10 per cent of the activity of the mother blood, so
that the whale foetus blood was relatively potent.

SUMMARY

1. The conditions underlying the respiratory activities of Southern Blue and Fin
whales were investigated in South Georgia and on board the pelagic whaler ' Southern
Princess '.

2. The environment of these whales and some aspects of their aquatic existence have
been considered in relation to their respiratory demands and limitations. Instances of
deep and prolonged diving have been quoted. The inference has been drawn that unless
whales are different in some respects from other mammals they would be liable to
caisson sickness. Histological evidence suggests that whales expel as much air as pos-
sible at each expiration in order to compensate for the hardships of prolonged holding
of the breath.

3. Gas analyses were undertaken to show the condition of urine, allantoic fluid, and
blood, and the possible influence on these of the whale's submarine activity. Large
volumes of carbon dioxide have been found dissolved in urine and allantoic fluid, which
indicate that high partial pressures of this gas are common within the whale. Slight
supersaturation of urine and allantoic fluid with nitrogen has been observed.

Adult and foetal blood hardly ever contain as much dissolved nitrogen as is soluble
in the blood of other mammals at atmospheric pressure. In addition, the nitrogen
capacity of whale blood has been shown to be more than twice that of human blood.
Whale blood therefore is found with very much less nitrogen than could be dissolved
in it from the air in the lungs.

Nitrogen disappears in the blood and cannot be extracted from it by evacuation. It
has not been possible to measure the maximum rate of disappearance of nitrogen ; the
greatest volume removed was 27 vol. per cent in 40 min. The disappearance of nitrogen
is contingent on the presence of oxygen.

4. Small organisms, referred to provisionally as X organisms, have been observed in
all samples of adult and foetal blood. Their diameter varies from 0-5 to 2-o/t. They re-
produce rapidly in vitro and are resistant to freezing. Crude cultures of these organisms
were found to have the power to take up more nitrogen than should be soluble in physical
solution and of disposing of the nitrogen in some way so that it was not recoverable by
evacuation. Pig's blood infected with these organisms behaves like whale blood in dis-
posing of nitrogen. In the face of the evidence which has been collected it is difficult to
avoid the conclusion that the organisms are responsible for a kind of nitrogen " fixation "
and that their presence in whale blood serves to protect the whale from caisson sickness.

400 DISCOVERY REPORTS

Samples of whale blood brought to England from the Antarctic at — io° C. were shown
to contain organisms which reproduced in nutrient solution and were apparently
bacteria. Investigations are still in progress.

5. The haemoglobin content of whale blood is low, approximately 9 per cent, com-
pared with 13-8 per cent in human blood. The oxygen capacity of the blood is roughly
proportional to the haemoglobin content and averages 14 vol. per cent. Whale blood
therefore has a smaller capacity for oxygen than human blood.

6. The carbon dioxide contents of the blood have been found to be slightly greater
than the human equivalent. Carbonic anhydrase has been noted in adult and foetal
blood ; there is more of this enzyme in the foetal blood than has been found in the
blood of foetal goats.

LITERATURE CITED

Andrews, R. Chapman, 1916. Whale Hunting with Gun and Camera. New York and London.

Barcroft, J., 1925. The Respiratory Function of the Blood . Cambridge.

Bennett, A. G., 1931. Whaling in the Antarctic. Edinburgh.

Bock, A. V., Field, H., and Adair, G. S., 1924. Oxygen and Carbon Dioxide Dissociation Curves in Blood.

Journ. biol. Chein., Lix, p. 353.
Boycott, A. E., Damant, G. C. C, and Haldane, J. S., 1908. The Prevention of Compressed Air Illness.

Journ. Hyg., viii, p. 356.
Brinkman, R., Margaria, R., Meldrum, N. U., and Roughton, F. J. W., 1932. The CO^ Catalyst Present

in Blood. Journ. Physiol., lxxiv P.
Ellenburger, W. Die Physiologie der Hausgethiere. Quoted by Smith, q.v.
Evans, C. Lovatt, 1930. Starling's Principles of Human Physiology, 5th ed., London.
Greene, C. W. and C. H., 1922. Oxygen in Blood in Induced Anoxaemia. Journ. biol. Chem., Lii, p. 137.
Goldberg, G., 1907. Ueber das Verhahren bei Berechnung des Rauminhaltes und Gewichtes der grossen

Waltiere. Christiania Videnskabs-Selskabs Forhandlinger for 1907, No. 3.
Hill, L., 1912. Caisson Sickness. London.
MoRiMOTO, Y., Takata, M., and Sudzuki, M., 1921 . Untersuchungen tiber Cetacea. Tohoku Journ. of Exp.

Med., II, \, passim.
Ommanney, F. D., 1932. The Vascular Networks {Retia Mirabilia) of the Fin Whale (Balaenoptera Physalus).

Discovery Reports, v, pp. 327-62, text-figs. i-io.
ScoRESBY, W., 1820. An Account of the Arctic Regions. Edinburgh.
Smith, A., 1895. Manual of Veterinary Physiology . London.
Van Slyke, D. D., 1917. Studies in Acidosis. II. A Method for the Determination of Carbon Dioxide and

Carbonates in Solution. Journ. biol. Chem., xxx, p. 347.

APPENDIX

A NOTE ON THE COMPOSITION OF WHALE BLOOD

Appearance. Whale blood drawn from an artery after death is dark red, almost
maroon. After aeration it is darker than aerated human blood. If aerated whale blood is
diluted with weak ammonia, as in the estimation of oxygen capacity by Barcroft's
method, the red colour is tinged with blue which is particularly noticeable by trans-
mitted light. (It is possible that the blue colour has some connection with the occasional
blueness observed in the X organisms.)

RESPIRATION IN SOUTHERN WHALES

401

Smell. The blood of all Rorquals has a peculiar pungent odour which can only be
described as resembling the smell of decaying Crustacea. This smell has been noticed
in the blood of pigs which have been fed largely on whale meat.

Erythrocytes. Red corpuscles examined under the microscope appear as discs of
about the same size as human corpuscles. Occasionally they appear to be slightly
concave on one side and convex on the other. " Rouleaux " of corpuscles form in fresh
smears as in human blood. An attempt was made to study the rate of sedimentation by
Fahraeus' method, but accurate observations were impossible because all the samples
of blood were not of the same age. It was, however, noticed that the blood of pregnant
whales sedimented most rapidly, as does also the blood of pregnant human beings.

Number of erythrocytes. Enumeration of corpuscles was made by the standard
method using a haemocytometer. The average of eleven samples was 3-84 million
corpuscles per c.mm. (in Blue whales). In drawing samples for corpuscle counts care
was taken to secure large volumes of blood from whale arteries in case local sedimenta-
tion had occurred in the vessels. The blood was well stirred and mixed before the
haemocytometer pipette was filled from it.

Plasma chlorides. Twenty samples of blood (ten Fin and ten Blue whales) were
centrifuged to concentrate the corpuscles . A known volume of standard silver nitrate
was added to the prepared plasma, and the excess titrated against standard potassium
thiocyanate (Whitehorn's method). The average chloride content was equivalent to 6-98
mg. of sodium chloride per cc. of plasma. The silver nitrate was frequently standardized
against " Standard Sea Water " supplied by the Hydrographic Laboratory, Copenhagen.
The plasma chlorides were constant to within 5 per cent. Samples of Blue and Fin
whale blood were found to have identical plasma chloride contents.

A NOTE ON THE COMPOSITION OF THE
ALLANTOIC FLUID OF BLUE WHALES

Reducing sugar. All the samples of allantoic fluid which were taken for gas analysis
(Table II) contained a reducing sugar which was appreciable to taste and gave a strong
reaction with Benedict's reagent. Quantitative estimation of the sugar was made in three
samples by titration against standard Benedict's solution. The following results were
obtained :

Sample no.

Sugar g./iooo cc.

19
20

21

49-0

59-0
6o-8

Takata (1922) found 22-5 g. of fructose per 1. in the allantoic fluid of a Sei whale and
9 "75 g- P^f '• 'ri that of a Sperm whale. No facilities were available for discriminating
between glucose and fructose in the results given above.

Uric acid. Some samples of allantoic fluid were submitted to Mr A. Smith, chemist

403 DISCOVERY REPORTS

on board the ' Southern Princess ', who kindly tested them for the colour reaction with
phosphotungstic acid. Negative results were obtained. Takata records 0-003 P^^ ^^^^
uric acid in the allantoic fluid of a Sei whale.

A NOTE ON THE WEIGHTS OF SOME BLUE WHALES

The weighing of a Blue whale is laborious and has seldom been performed, but some
data have been collected at various times. In the tables below are shown measurements,
weights of various organs (expressed also as percentages of the total), and total weights
of two Blue whales which were dismembered and weighed piecemeal by Capt. S0rlle
at Stromness, South Georgia.

Sir Sidney Harmer has kindly supplied me with another record in Norsk Hvalfangst-
Tidende (1924, No. 9, p. 108, quoted from Andrews, 1916) of a Blue whale of 23-72 m.
Dr F. A. Lucas weighed this whale piecemeal in Newfoundland and obtained the
following results :

Flesh 40 tons Blood j

7 tons

Flesh 40 tons

Blood

Blubber 8 „

Viscera
Baleen

Bones 8 „

Total

63 tons (61-7 tonnes)

Harmer in his account of "Cervical Vertebrae of a Gigantic Blue Whale from
Panama" {Proc. Zool. Soc. London, 1923, p. 1085) relates that the whale was 29-8 m.
long (98 ft.) and that attempts to lift it out of the water with 75-ton cranes failed. The
weight was estimated at 100 tons.

Messrs the Southern Whaling and Sealing Company have kindly allowed me to quote
the approximate weight of a large Blue whale estimated by their chemist Mr R. Squire
at Prince Olaf, South Georgia. The whale was 29-5 m. (97 ft.) long. From the number
of blubber, meat, and bone cookers which were filled by this whale the weight was
deduced to be 163-7 tons (160-4 tonnes) exclusive of blood. The total weight of the
whale was probably 174 tons (170-5 tonnes).

D'Arcy Thompson, in Science of the Sea (second ed., p. 492), calculates that the
weight of a Rorqual 85 ft. long would be about 370 tons. In arriving at this result he
has made use of the principle that "in bodies of similar shape the bulk or weight will
vary as the cube of the linear dimensions ". His calculation is based on the weight of a
foetal Rorqual i ft. long. Small foetuses of Rorquals tend to differ considerably in shape
from the adults, and if a whale of the shape of, say, a foetal Humpback were to attain
to a length of 85 ft. it would doubtless have the weight calculated by D'Arcy Thompson ;
but on reference to Table VII it will be seen that this estimate is rather more than 300
per cent too high for Blue whales, the largest known species, and it is apparent that
weights may not be calculated from foetal material.

On the basis of the three accurate weights known, a number of calculations have been
made to ascertain the weights of Blue whales of various lengths. These calculations are

RESPIRATION IN SOUTHERN WHALES

403

based on the same formula as that used by D'Arcy Thompson and they are vaUd only
on the assumption that a Blue whale during its growth does not differ in form from

Table VI. Measurements, proportional weights, and analyses of
the various parts of a Blue whale.

From Norsk Hvalfangst-Tidende, May 5, 1924, No. 7, p. 74.

The following measurements were taken :

m.

Length

20-30 (66 ft.)

Height, whale lying on its side .

270

Breadth back to front

3-IO

Greatest circumference ...

11-05

Tail fins from tip to tip ...

...

4-25

Jaw bones, length ...

4-35

Flippers, length

2-50

From flipper to flipper across the breast

5-15

The individual parts* had the following weights:

kg.

/o

Meat

.

25.940

53-00

Blubber

9,116

18-65

Bone

9.433

19-25

Tongue

1,102

2-26

Lungs

588

1-20

Liver

409

0-84

Heart

329

0-67

Kidneys

220

0-45

Stomach

200

0-41

Intestines

1,164

2-39

Whale bone

402

0-82

Total...

4«.903

The individual parts of the skeleton had the followi

ng weights :

kg-

Jaw bones ...

928

Head bone (front)

... ...

1971

Backbone

4410

Ribs

917

Breast bone

56

Tail fins

575

Flippers

576

Total...

9433

Analyses of the parts :

Meat 0/
/o

Blubber

%

Water 69-3

Water

7-0

Oil 6-5

Oil

87-8

Fibre 23-6

Fibre

4-2

Other constituents o-6

Other

constituents

i-o

Bone 0/

/o

Oil

Barrels

Water ... ... 13-2

Blubber oil

48-25

Oil 45-5

Meat oil

9-95

Bone material ... 41-1

Bone oil

25-40

Other constituents 0-2

Total...

83-60 (13,985 kg.)

* Blood not mentioned.

404

DISCOVERY REPORTS

Table VII. Particulars of a Blue whale kindly supplied to the Discovery
Committee by Capt. S0rlle, Stromness, South Georgia, November, 1926.

Measurements m

Length

1

27-18

(89 ft.)

Height

(whale lying on

1 side)

3-10

Greatest girth

13-90

Length

1 of jaw bone

6-95

Length

1 of baleen

5-90

Length

. of swimmers

3-00

Weights

kg.

0/
/o

Weights

of bones w^

Blubber

25,651

21-01

Jaw bone

2,117

Meat

56,444

46-27

Head

4,508

Bone

22,280

18-27

Spine

10,230

Tongue

3,158

2-58

Bones

; of swimmers 3,863

Lungs

1,226

1-05

Swimmers

960

Heart

631

0-51

Blades of swimmers 602

Kidneys

547

0-44

Total 22,280

Stomach

416

0-34

Intestines

1,563

1-30

Liver

935

0-76

Biggest dorsal vertebra 240

Baleen

1,153

0-94

Blood (ca.)

8,000

6-56

Total ']

122,004

Analysis of the parts:

Oil

Fibres

Water

Blubber

0/

/o

/o

%

Tongue

33-30

19-40

46-80

Ventral grooves

32-15

20-00

48-06

Tail flank

81-80

6-20

12-50

Back

75-70

8-25

16-30

Top of head

38-90

22-30

38-20

Meat

%

/o

Tail:

Water

73-90 Average of 1 2 samples :

69-75

Oil

6-65

8-00

Fibres

19-00

22-90

Bone

%

/o

Jaw:

Water

13-20

Ribs:

Water

ii-i6

Oil

63-60

Oil

64-90

Solids

24-10
/o

Solids

23-50

0/
/o

Head

1: Water

8-16

Vertebra :

Water

9-10

Oil

68-60

Oil

56-30

Solids

23-60

Solids

34-40

Oil production

kg.

Blubber

13,604

Meat

6,880

Bone

7,224

Total

27,708 (166 barre

Is)

METRES 10

APPR0X.FEET33

14 16 IB 20 22 24

41 49 57 65 73

LENGTH OF WHALE

26 28

I 89

30
97

Fig. 4. Calculated relation between the length and weight of Blue whales.

4o6 DISCOVERY REPORTS

those for which actual weights are recorded. Mackintosh and Wheeler {Discovery
Reports, l, pp. 296-7) have shown that in actual fact some alteration takes place,
but it is so small that it makes very little difference to the validity of the weights shown
in Fig. 4. In this figure three curves are shown, each based on one whale: curve A on
a whale of 20-3 m., 48-9 tonnes (without blood); curve B, 27-18 m., 122 tonnes; curve
C, 2378 m., 617 tonnes. The theoretical weight of the whale of 20-3 m., as found on
curve B, approximates very closely to its actual weight, as does also that of the whale
of 27-18 m. There is less agreement between these curves and curve C (Lucas's whale).
The blubber content of this whale was very low, only 12-8 per cent of the total weight
against an average for the first two whales mentioned of 19-83 per cent. It is to be sup-
posed that this whale, in any case different from the others in being a Northern Blue,
was in poor condition and that its weight was low for its length. On the whole, the
mean of curves A and B may be taken to furnish an approximate guide to the weights
of Southern Blue whales.

PLATE XV

Sections of whale lungs stained to show elastic tissue ( v 540).
Fig. I . Lung of Humpback whale.
Fig. 2. Lung of Fin whale.

DISCOVERY REPORTS, VOL. VII

PLATE XV

A. Saunders phot.

John Bole Sww i\ DHnjelsson-L^ Ltoidnn

SECTIONS OF LUNGS OF HUMPBACK AND FIN WHALES

DISCOVERY
REPORTS

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the Falkhnd Islands

Vol. VII, pp. i-vi

TITLE-PAGE AND LIST OF CONTENTS

CAMBRIDGE

AT THE UNIVERSITY PRESS

1933

Price ninepence net

/"■h^ ff

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

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VII, pp. 1-16

Issued by the Discovery Committee, Colonial Office, London
on behalf of the Qovernment of the Dependencies of the FalMand Islands

FOSSIL FORAMINIFERA FROM THE

BURDWOOD BANK AND THEIR

GEOLOGICAL SIGNIFICANCE

W. A. Macfadyen, M.C., M.A., Ph.D., F.G.S.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1933
Price Three shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo
Manizen Company, Ltd

All rights reserved

PRINTED

I N GREAT BRITAIN

BV

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. Vn, pp. 17-26

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

FAECAL PELLETS FROM MARINE

DEPOSITS

Hilary B. Moore, B.Sc.

CAMBRIDGE

AT THE UNIVERSITY PRESS

1933

Price two shillings 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

DISCOVERY
REPORTS

Vol. VII, pp. 27-138, plates I-VII

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

FORAMINIFERA

PART II. SOUTH GEORGIA

by
Arthur Earland, F.R.M.S.

CAMBRIDGE

AT THE UNIVERSITY PRESS

1933

Price sixteen shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

PRI NTED

IN GREAT BRITAIN

BY

WALTER fetfl

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VII, pp. 139-170

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

ON VERTICAL CIRCULATION IN THE
OCEAN DUE TO THE ACTION OF THE
WIND WITH APPLICATION TO CON^
DITIONS WITHIN THE ANTARCTIC
CIRCUMPOLAR CURRENT

by
H. U. Sverdrup

CAMBRIDGE
AT THE UNIVERSITY PRESS

1933
Price five shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcuua, 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. VII, pp. 171-238, plates VIII-X

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

A GENERAL ACCOUNT OF THE HYDROLOGY
OF THE SOUTH ATLANTIC OCEAN

by
G. E. R. Deacon, B.Sc.

CAMBRIDGE

AT THE UNIVERSITY PRESS

1933
Price eleven shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macraillan

Tokyo

Manizen Company, Ltd

All rights reserved

PRINTED

I N GREAT BRITAIN

BY

™/ LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

PRINTED

I N GREAT BRITAIN

BY

WALTER

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS

DISCOVERY
REPORTS

Vol. VII, pp. 239-252, plates XI-XIII

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

WHALING IN THE DOMINION OF
NEW ZEALAND

by
R D. Ommanney, A.R.C.S., B.Sc.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1933

Price four shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights reserved

DISCOVERY
REPORTS

Vol. VII, pp. 253-362, plate XIV

hstted by the Discovery Committee, Colonial Office, London
on behalf of the Qovemment of the Dependencies of the Falkland Islands

ISOPOD CRUSTACEA

PART I. THE FAMILY SEROLIDAE

by
Edith M. Sheppard, M.Sc.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1933
Price sixteen 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. VII, pp. 363-406, plate XV

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

SOME ASPECTS OF RESPIRATION IN BLUE

AND FIN WHALES

Alec H. Laurie, M.A.

CAMBRIDGE
AT THE UNIVERSITY PRESS

1933

Price six shillings net

Cambridge University Press
Fetter Lane, London

New York

Bombay, Calcutta, Madras

Toronto

Macmillan

Tokyo

Maruzen Company, Ltd

All rights teserved

PRINTED

I N oREAT BRITAIN

BY

LEWIS MA

AT

THE CAMBRIDGE

UNIVERSITY

PRESS