Skip to main content

Full text of "Quarterly journal of microscopical science"

See other formats 


14 iy . . 4 an . . 
A j 7 Vi eo = | ; ot “s 
. re : . . hs evan ty 
. 3H » 3 ‘ a os ea TOES: 
é ih) Ses 
; Ra ) : +E AN hh ae Te LO ELON, 
' ' , , pues wy at Aner * 
ae ' , 


ake b isl. jerry aee 


$8 per 0K “ss 


eer 


: 
ie pret egies.) 
ae Fear pen 


apie 


Hin eee 
Cao 


cae 
1 lweetted tp de 


ry 
DEY, 


bie, 
ae 
a i ees se 


et ae ee a 


shoe & 
ae? he ir) iui 


toate 
ee 
Asie one 
i ee 


i 


aN AE eae ik 
bo 4 
My diale “6 Roe 


FOR THE PEOPLE 
FOR EDVCATION 
FOR SCIENCE 


LIBRARY 


OF 
THE AMERICAN MUSEUM 
OF 


NATURAL HISTORY 


QUARTERLY JOURNAL 


OF 


MICROSCOPICAL SCIENCE. 


EDITED BY 


Size RAY LANKESYER, K.C.B., M.A., D.Sc., LL.D., F.RS., 


HONORARY FELLOW OF EXETER COLLEGE, OXFORD 5 
MEMBER OF THE INSTITUTE OF FRANCE (associf ETRANGER DE L’ACADEMIE DES SCIENCES) 3 
CORRESPONDENT OF THK IMPKRIAT ACADEMY OF SCIENCES OF ST. PETEKSBUKG, AND OF THE 
ACADKMY OF SCIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIENCES 
OF TURIN; FORKIGN MEMBER OF THE ROYAL SOCIETY OF SCIENCES OF 
GOTTINGEN, AND OF THE ROYAL BOHEMIAN SOCIETY OF SCIENCES, AND 
OF THE ACADEMY OF THE WLINCEI OF ROME, AND OF THR AMERICAN 
ACADEMY OF ARTS AND SCIENCES OF BOSTON } ASSOCIATE OF THE 
ROYAL ACADEMY OF BELGIUM: HONORARY MEMERR OF THE 
NEW YORK ACADEMY OF SCIKNCES, AND OF THER 
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OP 
THE ROYAL PHYSICAL SOCIETY OF EDIN- 
BURGH, AND OF THE 
KIOLOGICAL SOCIETY OF PAKIS, AND OF THE CALIFORNIA ACADEMY OF SCIENCES OF SAN 
OF THK ROYAL ZOOLOGICAL AND MALACOLOGICAL SOCIETY OF BELGIUM; 
CORRESPONDING MEMBER OF THE SENKENBERG ACADEMY OF FRANKFURT-A-M} 
FOREIGN ASSOCIATE OF THK NATIONAL ACADEMY OF SCIENCES, U-S., AND MEMBER OF THE 
AMERICAN PHILOSOPHICAL SOCIETY 3 
HONORARY FELLOW OF THE ROYAT SOCIETY OF EDINBURGH; 
LATE DIRECTOR OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSKUM, LATE PRESIDENT OF THE 
BRITISH ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE; LATE FULLERIAN PROFESSOR OP 
PHYSIOLOGY IN THE ROYAL INSTITUTION OF GREAT BRITAIN; 
LATR LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFORD) 
EMERITUS PROFESSOR OF ZOOLOGY AND COMPARATIVE ANATOMY IN UNIVERSITY COLLEGE, UNIVERSITY OF LONDON. 


WITH THE CO-OPERATION OF 


ADAM SEDGWICK, M.A., F.R.S., 


FELLOW OF TRINITY COLLEGE, CAMBRIDGE. AND PROFESSOR OF ZOOLOGY IN THE IMPERIAL COLLEGE OF 
SCFENCE AND TECHNOLOGY, LONDON ; 


SYDNEY J. HICKSON, M.A., F.R. 


BEYER PROFESSOR OF ZOOLOGY IN THE UNIVERSITY OF MANCHESTER, 


E. A. MINCHIN, M.A., 


PROFESSOR OF PROTOZOOLOGY IN THE UNIVERSITY OF LONDON, 


AND 


GILBERT C. BOURNE, M.A., D.Sc., F.R.S., 


OXFORD. 


FRANCISCO, AND 


LINACRE PROFESSOR OF COMPARATIVE ANATOMY, AND FKLLOW OF MERTON COLLEGE, 


VOLUME 55.—New SERIEs. 
With Vithographic Plates und Text-Figures 


AL 


AE 


in 
AS 


LONDON: 
J. & A. CHURCHILL, 7, GREAT MARLBOROUGH STREET. 


1910: 


if eee AAU ut 


Men ts 
TVA, GU 


Aen ie Tcthinc A 
i 7 KUHEW UGE, BAAR | 
; | YAOT ETE SANTA io 


CON Lewy tS 


f 
CONTENTS OF No. 217, N.S., APRIL, 1910. 


MEMOIRS : 

On the Anatomy and Systematic Position of Incisura (Scissu- 
rella) Lytteltonensis. By GitBertr C. Bourne, Fellow’ of 
Merton College, Oxford, and Linacre Professor of Comparative 
Anatomy. (With Plates 1-5) 

The Eye of Pecten. By W. J. Daxin, M.Sc., ipamoristeatae aad 
Assistant Lecturer in Zoology, University of Belfast. (With 
Plates 6 and 7, and 2 Text-figures) : 

Observations on Certain Blood-Parasites of Biahiee occurring at 
Rovigno. By Prof. E. A. Mincurn, M.A., and H. M. Woopcock, 
D.Sc. (With Plates 8-10) : : 

On Ganymedes anaspidis (nov. gen., nov. ere a Gregarine 
from the digestive tract of Anaspides tasmaniz (Thomp- 
son). By Junian S. Huxusy. (With Plate 11, and 5 Text- 
figures) : 2 : 

The Fetal Membianes of tiie Vetteitatas By A. A. W. Husrecur 


CONTENTS OF No. 218, N.S., JUNE, 1910. 
MEMOIRS : 

The Structure and Life-History of Crithidia melophagia 
(Flu), an Endo-parasite of the Sheep-Ked, Melophagus 
ovinus. By ANNIE Porter, B.Sc.Lond., Zoological Research 
Laboratory, University College, London. (With Plates 12 and 
13, and 15 Text-figures) : : 

Studies in the Experimental epee of Bee. By GEOFFREY 
Smiru, Fellow of New College, Oxford. (With Plate 14) 

Some Observations on a Flagellate of the Genus Cercomonas. By 
C. M. Wenyon, M.B., B.S., B.Sc., Protozoologist to the London 
School of Tropical Medicine. (With 19 Text-figures) ‘ 

_ Some Observations on a New Gregarine (Metamera schubergi 
nov. gen., nov. spec.). By H. Lynpuurst Dvuxs, B.A., B.C. 
Cantab. (With Plates 15 and 16) . : ‘ 

On the Anatomy of Histriobdella Homans By CRESSWELL 
Surarer, M.A., Trinity College, Cambridge. (With Plates 17— 
20, and 5 Text-figures) : 

On the Artificial Culture of Marine Plankton @npunisrad: By 
E. J. Auten, D.Se., Director of Laboratories and Secretary of 
the Marine Biological Association, and E. W. Netson, Assistant 
Naturalist 


PAGE 


49 


113 


241 


1V CONTEN''S. 


CONTENTS OF No. 219, N.S., SEPTEMBER, 1910. 


MEMOIRS: 

Notes on the Free-living Nematodes. I. The Hermaphrodite 
Species. By F. A. Ports, M.A., Fellow of Trinity Hall, Cam- 
bridge, and Demonstrator of Comparative Anatomy in the 
University. (With 11 Text-figures) : : : 

J Observations on Trypanoplasma Congeri. Part I. The Division 
of the Active Form. By C. H. Martin, B.A., Demonstrator of 
Zoology, University of Glasgow. (With Plate 21, and 1 Text- 
figure) . ; : ; 

The Development ‘of Agila Se tert By A. M. Carr 
SaunpERs and Maraaret Pootr. (With Plate 22,and 20 Text- 
figures) 

y The Relation pate: een Light and eee for acetate in Creni- 
labrus and Hippolyte. By F. W. Gamsuz, F.R.S., Mason 
Professor of Zoology, University of Birmingham. (With Plate 23) 

Is the Trophoblast of Hypoblastic Origin as Assheton will have 
it? By A. A. W. Huprscutr. (With 7 Text-figures) . 

The Origin and Formation of Fibrous Tissue produced as a 
Reaction to Injury in Pecten Maximus, as a Type of the 
Lamellibranchiata. By G. H. Drew, B.A., Beit Memorial 
Research Fellow, and W. Dr Moraan, F.Z.S. (With Plate 24) 


CONTENTS OF No. 220, N.S., NOVEMBER, 1910. 


MEMOIRS : 

The Division of the Collar-Cells of Clathrina coriacea 
(Montagu); a Contribution to the Theory of the Centrosome 
and Blepharoplast. By Murizt Roperrson, M.A., and E. A. 
Mincuin, M.A. (With Plates 25 and 26) : 

Studies on Avian Hemoprotozoa. I.—On Certain Parasites of 
the Chaffinch (Fringilla celebs) and the Redpoll (Linota 
rufescens). By H. M. Woovcock, D.Se.(Lond.), Assistant to 
the University Professor of Protozoology. (With Plates 27-31) 

Studies on Ceylon Hematozoa. No. II.—Notes on the Life-Cycle 
of Hemogregarina nicoriz, Cast. and Willey. By Murren 
Rosertson, M.A. (With Plates 32-41 and 1 Text-figure) 

On the Origin and Migration of the Stinging-Cells in Craspedote 
meduse. By CuHarueEs L. BoutencErR, M.A.Camb., Lecturer 
on Zoology in the University of Birmingham. (With Plates 42 
and 43 and 5 Text-figures) 

The Researches of Bouvier and Bowiage on Maikationss in Crustal 
of the Family Atyide. By W. T. Cauman, D.Sc. (With.4 
Text-figures) : : : 

Tire, INDEX, AND CONTENTS. 


PAGE 


433 


485 


497 


611 


641 


785 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 1 


On the Anatomy and Systematic Position of 
Incisura (Scissurella) lytteltonensis. 


By 
Gilbert C. Bourne, 


Fellow of Merton College, Oxford, and Linacre Professor 
of Comparative Anatomy. 


With Plates 1—5. 


Wuen Mr. Geoffrey W. Smith was in Tasmania in 1907-08 
T asked him to collect for me any rare or remarkable speci- 
mens of gastropod molluscs and preserve them in a form 
suitable for anatomical and histological examination. Among 
other forms Mr. Smith obtained for me, through the kind 
offices of Mr. C. Hedley, of the Australian Museum, Sidney, 
a number of specimens of the little gastropod which is the 
subject of the present memoir. They were preserved in 
Perenyi’s fluid, which of course dissolved the shells, but 
except for the difficulty of staining always resulting from a 
prolonged immersion in this reagent, the histological condition 
of the specimens leaves little to be desired. 

Scissurella lytteltonensis was described in 1893 by 
H. A. Smith (16), who noted certain differences between the 
shell of this and other species of the genus Scissurella, but 
evidently did not consider them of generic importance. In 
1904 C. Hedley (8) recalled attention to these differences, 
and founded the new genus Incisura for the reception of 
the species which, he maintained, is marked off from all 
other Scissurellide as also from all Pleurotomariide by the 
brevity of the slit in the shell, by the absence of raised rims 
or keels on either side of the sht, by the subterminal apex, 

voL. 55, PART 1.—NEW SERIES. 1 


2 . GILBERT C. BOURNE. 


by the absence of spiral sculpture, and by the remarkable 
solidity of the shell. He further asserted that his new genus 
cannot, because of the above-mentioned differences, be in- 
cluded among the Scissurellide, and suggested that it is a 
member of the Fissurellide in which development has been 
arrested, so that the larval characters of the shell have 
persisted in adult life. Hedley was evidently unacquainted 
with Pelseneer’s (12) memoir, containing an account of the 
anatomy of this very species and of Scissurella costata, 
which, brief as it is, leaves no doubt that the New Zealand 
and the Mediterranean species are members of the same 
family, but at the same time discloses so many anatomical as 
well as conchological differences that they may well be placed 
in different genera. After some consideration I am of the 
opinion that Hedley’s genus should stand, because the New 
Zealand species, in addition to the conchological characters 
enumerated above, differs from the Mediterranean species in 
the following particulars: (1) In the shape of the radular 
teeth. (2) In the shape of the foot, which is long and 
narrow in 8. costata and S. crispata, but short and 
broad in Incisura lytteltonensis. (3) In the absence of 
cirrhi below the epipodial tentacles in Incisura. (4) The 
greater development of the right columellar muscle, and the 
more symmetrical disposition of the mantle in Incisura. In 
its general anatomical features Incisura bears much the 
same relation to Scissurella as Septaria bears to Paranerita 
among the Neritide. The systematic position of the Scissu- 
rellidee will more conveniently be discussed at the end of 
this paper. 

Scissurella is placed by most authors among the Pleuro- 
tomariide, though a few recognise the Scissurellide as a 
separate but closely allied family. A full description of its 
anatomy is therefore much to be desired, but the accounts 
that have hitherto been published are insufficient. Vayssiére 
(18) has given a short and, as far as it goes, a good account 
of the external features of S. costata var. levigata, and 
has figured and described the radula and jaws of this species. 


INCISURA (SCISSURELLA) LYTTHLTONENSIS. 3 


Pelseneer (12), in his well-known memoir on the morphology 
of primitive mollusca, gives seven figures of sections of 
S. costata and two of Incisura lytteltonensis in addi- 
tion to three figures of the external features of the latter 
species. ‘The description he gives in the text is concise, and 
furnishes a good general idea of the anatomy of the family ; 
but he does not give sufficient detail to enable one to make 
a critical examination of its systematic position. Hence, 
having sufficient material at my disposal, I have thought it 
worth while to make a thorough study of the anatomy of 
Incisura lytteltonensis. 

Incisura, as Mr. Hedley states in a letter accompanying the 
specimens, is found on the seaweed Cystophora in rock-pools 
in Lyttleton Harbour, where it is associated with Rissoina, 
Cantharides, and Gibbula. It may be inferred from its shape 
and structure that it is semi-sessile in habit, but it is not 
attached to one spot lke a limpet. On the contrary, it is 
fairly active, and one of the specimens was observed to crawl 
for a distance of nearly half an inch in the space of a quarter 
of an hour. When alive it is of a pink colour, and this tmge 
is sometimes preserved in the shell. The length of the 
animal, when contracted in spirit, is about 1 mm. 

External features.—These have been correctly if somewhat 
diagrammatically figured by Pelseneer. A three-quarter 
ventral view of the animal is given in fig. 1. Attention may 
be called to the following points: The visceral spire is 
attenuated and much reduced, its coiled apex containing only 
some lobes of the liver and, in some specimens, a portion of 
the gonad. ‘The last whorl is greatly expanded laterally, 
compressed dorso-ventrally, and contains all the important 
organs of the body. The snout is moderately long, termina- 
ting in a trumpet-shaped expansion, on the ventral side of 
which is the mouth. The mantle is large, and in contracted 
specimens completely covers the head and the greater part of 
the snout. The mantle slit, corresponding to the labral 
incision in the shell, is short, and situated nearly opposite the 
right eye; its margins are furnished with short digitiform 


4, GILBERT C. BOURNE. 


processes bearing projecting sense-papille, such as have 
been described by Vayssiére in Scissurella costata. The 
cephalic tentacles are moderately long, reaching in their 
contracted state as far forward as the end of the snout. They 
are fringed with a large number of small, conical sense- 
papille, which, in Incisura, are not scattered all over the 
surface of the tentacles as figured by Vayssiére for 
S. costata, but are arranged in two multiple rows on the 
inner and outer margins of each tentacle (fig. 27), somewhat 
like the pinnules on the tentacles of an Alcyonarian polyp. 
The structure of these sense-papillee will be described further 
on. The eyes, which are closed and provided with a cornea 
and lens, are situated on prominences at the outer sides of 
the bases of the tentacles. Just below and behind the 
tentacle of each side is a short sub-ocular tentacle which 
does not bear sense-papillz like the cephalic tentacles, but is 
richly ciliated and glandular in structure. In the single male 
specimen of which I have cut sections, the sub-ocular tentacle 
of the right side is somewhat enlarged, spatulate in form, 
and more abundantly provided with gland-cells than in the 
females. In all the females I have examined the sub-ocular 
tentacle of both sides is digitiform. 

The foot, as is shown in fig. 1, is rather short and triangular 
in shape, the apex of the triangle being posterior. In shape 
and in the size of the broad, creeping sole it differs consider- 
ably from the narrower elongated foot of S. costata and 
S. crispata. The epipodium begins as a low ridge in about 
the middle third of the foot, and increases in size posteriorly. 
As described by previous authors it bears three moderately 
long epipodial tentacles on each side of the body towards the 
posterior end of its course. These tentacles bear lateral rows 
of sense-papillee exactly like those of the cephalic tentacles, 
but there are no ventral cirrhi in connection with them as 
in 8. costata. The epipodial folds meet posteriorly above 
the posterior end of the foot, and just dorsal to and in front 
of their union is a muscular opercular lobe bearing the 
operculum. The last-named structure is small, horny, and 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 3) 


multispiral, as in other Scissurellide. It must be regarded 
as vestigial since, as is the case in Pleurotomaria, it cannot 
be of any use in closing the aperture of the shell. There are 
two columellar or shell-muscles (fig. 2) symmetrically disposed 
right and left of the middle of the body, the right muscle 
being slightly larger and extending rather further back than 
the left. 

As it is almost impossible to make dissections of an animal 
scarcely exceeding 1 mm. in length, the following account of 
the anatomy of Incisura is mainly founded on reconstructions 
from sections, but I succeeded in making some satisfactory 
whole preparations of the ctenidia, and have checked the 
results of my reconstructions as far as possible by the study 
of whole specimens cleared in various ways.. Fig. 2 is a 
camera drawing of a specimen stained in picro-carmine and 
mounted in oil of cloves; it shows as much of the general 
anatomy as can be made out by this method. Figs. 3, 4, and 
5 are reconstructions from sections showing respectively the 
anatomical relations of the alimentary tract, the kidneys and 
pericardium, and the nervous system. Figs. 6 to 12 are 
camera drawings of some of the sections from which the re- 
constructions were made. 

Organs of the pallial complex.—lIncisura is typically zygo- 
branchiate, and the position and general characters of the 
ctenidia, hypobranchial glands, left kidney, and pericardium 
have been correctly described by Pelseneer. 

The ctenidia.—Both right and left ctenidia take their 
origin from the roof of the mantle-cavity, close to the anterior 
end of the columellar muscle of their respective sides of the 
body. The left ctenidium lies almost transversely across the 
neck of the animal, its anterior extremity reaching nearly as far 
as the base of the right tentacle (fig. 2), and it is closely com- 
pressed between the body-wall and the roof of the mantle. 
The right ctenidium, on the other hand, lies for the most 
part in front of the right columellar muscle, and the bulk of 
it hangs vertically downwards in the space enclosed between 
the mantle and the outer side of the foot (fig. 7). Pelseneer 


6 GILBERT C. BOURNE. 


has described the right ctenidium as mono-pectinate, but, as 
may be seen in fig. 7, it is really bi-pectinate; the external 
lamelle, however, are few in number, and in some specimens 
are so feebly developed that they might easily be overlooked. 
It is at first rather difficult to make out the details of the 
structure of the ctenidia and to institute an exact comparison 
between them and those of closely allied Aspidobranchs, but 
a careful study of sections and whole preparations shows that 
they are constructed on the familiar pattern. Hach ctenidium 
consists of an axis, the posterior part of which is fused to the 
roof of the mantle-cavity and extends back in the angle of 
that side of the mantle-cavity to which it belongs, lying just 
above the columellar muscle. The anterior end of the axis 
is free, and the large osphradial ganglion, as is always the 
case in Aspidobranchia, is situated at the point where the 
axis becomes free from the mantle. ‘This point, in Incisura, 
corresponds with the anterior end of the columellar muscle. 
In the case of the left ctenidium that part of the axis which 
is fused to the mantle bears no filaments, but, as will be des- 
cribed further on, this statement does not hold good for the 
right ctenidium. Taking the left ctenidium for the purpose 
of description: its free apex projects into the mantle- 
cavity in front of the columellar muscle as a thin, tri- 
angular lamina, which, as already explained, is bent over 
to the right, and also is twisted about its own axis from 
right to left in such wise that the morphologically outer row 
of filaments become posterior in position, the morphologically 
inner row anterior. ‘The efferent branchial vessel runs, as is 
always the case, along the dorsal, here the posterior margin, 
and the afferent vessel along the ventral, here the anterior 
margin of the axis. ‘lhe inner and now anterior filaments 
borne on the free portion of the axis are short and not more 
than four or five in number, and are folded backwards over 
the upper (morphologically ventral) side of the axis, appa- 
rently as a result of the latter being twisted from right to 
left in a narrow space. ‘lhe morphologically dorsal edges of 
the anterior filaments are consequently maintained in a dorsal 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 7 


position. But in the case of the posterior filaments, which 
are eight in number and much longer than the anterior fila- 
ments, the twisting of the axis has brought the ventral 
surfaces into a dorsal position. Fig. 16 represents a section 
through the anterior and fig. 17 a section through the posterior 
row of filaments. Hach is more or less quadrangular in out- 
line, its lateral walls formed of long columnar cells bearing 
long and fine cilia, which in contracted specimens appear to 
interlock like the cilia of the ciliated discs of filibranch 
Lamellibranchia. I do not think, however, that their function 
is to hold the filaments together, but simply to create 
currents over the surfaces of the filaments. Their inter- 
locking is simply due to their becoming matted together 
in consequence of the contraction of the gillin spirit. On the 
ventral surface of each filament is a band of very short cilia. 
The dorsal edge of the filament bears no cilia externally, but, 
as shown in the figures, is produced to form a peculiar bolster- 
shaped swelling, which, as far as I am aware, has no analogue 
in the gills of any other mollusc. ‘his dorsal glandular 
ridge, as I will call it, takes its origin from near the free 
distal end of the filament, and extending along the dorsal 
face of the latter is closely fused to it for the greater part of 
its length, but on approaching the proximal end of the fila- 
ment the glandular ridge becomes free and ends in a small 
rounded projection. ‘lhe ridge is traversed throughout its 
length by a small ciliated canal, which makes no communica- 
tion with the blood channel of the filament, but opens into 
the mantle-cavity in the angle between the free proximal 
extremity of the ridge and the filament. ‘This communication 
with the mantle-cavity, as seen in section, is shown in the 
central filament in fig. 17. In the filament on the right hand 
in the same figure the section passes through the middle of 
the glandular ridge, and the ciliated canal is seen to be closed 
in on all sides and to be situated near the ventral, i.e. the 
filamentary side of the ridge. ‘The same features are shown 
in the ridge attached to the right-hand filament in fig. 18, 
but in the case of the left-hand filament in this figure the 


8 GILBERT C. BOURNE. 


section passes through the more distal part of the ridge, and 
the ciliated canal is seen to be smaller and situated near the 
dorsal side of the ridge. A little further on it ends blindly. 
As the figures show, the ridge is made up of a sheath or 
cortex of elongated, fusiform cells, which pass nearly trans- 
versely round the periphery of the ridge, and a medulla of 
large, closely packed ovoid or fusiform cells having large nuclei 
and granular cell contents. The cells abutting on the lumen 
of the ciliated canal are usually larger and more granular 
than those more peripherally situated, and their histological 
characters leave little doubt that they are secretory. It is 
noticeable that there are very few if any glandular cells 
interspersed among the columnar ciliated cells of the filament, 
and the glandular ridge appears to have taken over the 
secretory functions, and to replace the secretory cells scattered 
over the surface of gill-filaments of other Mollusca. The 
extreme specialisation exhibited by the formation of a closed 
canal into which the secretory cells discharge their products 
is certainly a remarkable feature in Incisura. 

The central blood-channels of the filaments, as may clearly 
be seen in the figures, are elongate-oval in shape, and their 
walls are strengthened, for about half their extent, by 
flattened, chitinous, skeletal bars, which, as in other molluses, 
may be traced to the proximal end of each filament, where 
they diverge from one another, and curve round to run up 
in the walls of the blood-spaces of the adjacent filaments 
(fig. 19). As M. F. Woodward (19) has shown that in 
Pleurotomaria these skeletal bars run along the dorsal edges 
of the gill-filaments, whereas in Nucula they run along the 
ventral edges, it is of some interest to determine the position 
of these bars in Incisura, which is usually reckoned as 
belonging with the genera Scissurella and Schismope to the 
Pleurotomariide. It is clear from an inspection of fig. 17, 
representing a transverse section through the posterior gill- 
filaments of the left ctenidium, which, as explained above, 
are turned upside down, that the skeletal bars lie on the 
dorsal sides of the filaments, and the same thing can be 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 9 


seen still more clearly by inspection of fig. 7, in which the 
relations of the gill-filaments to the axisare obvious. In the 
anterior gill-filaments of the left ctenidium the skeletal bars 
appear to be ventral in position, but this is because these 
filaments are reflected backwards and their natural surfaces 
are reversed. Incisura, then, agrees with Pleurotomaria, and 
also with Trochus (fide Fleure and Gettings) and Fissurella, 
and differs from Nucula. But it must be observed that 
Woodward went further than the facts warranted when he 
asserted that the position of the gill-bars indicated a more 
remote affinity between Pleurotomaria and the primitive 
Lamellibranchia than is generally supposed. As a matter of 
fact the skeletal bars differ considerably in position in some 
not remotely related mollusca. In Solenomya, for instance, 
they are shifted to a more dorsal position than in Nucula, and 
in the Filibranchia they are actually dorsal. The fact is, as 
Woodward himself pointed out, these skeletal bars have a 
physiological rather than a morphological significance, and 
are always developed in close relation to the tracts of cells 
bearing specially long or functionally important cilia. Hence, 
in Filibranchia we find them related to the ciliated discs, 
which are near the dorsal edges of the filaments. 

In so small an object as Incisura it is very difficult to make 
sure of the presence or absence of a septum dividing the 
blood-channel into an afferent and an efferent moiety, but I 
am tolerably certain that such a septum exists, as shown in 
fig. 18. But it is not always placed transversely, but may 
be oblique or even nearly longitudinal. 

The attached portion of the axis of the right ctenidium 
extends far back in the extreme right-hand corner of the 
mantle cavity, lying close above the columellar muscle of 
that side, and gives off some three or four short filaments 
before reaching the level of the osphradial ganglion. At 
this spot there is a break in the continuity of the filaments, 
none being formed in the immediate proximity of the ganglion, 
but in front of it the ctenidial axis becomes free, and drops 
vertically down in front of the columellar muscle to hang in 


10 GILBER! C. BOURNE. 


the space between the foot and mantle, as shown in fig. 7. 
The basal portion of the axis is also enlarged at this point, 
and gill-filaments are given off from both sides, both of the 
free apex and of the broad basal portion. These filaments 
are not simply digitiform like those of the left ctenidium, 
but are plate-like, with the glandular ridge running along 
their dorsal margins, as shown in fig. 7. As the skeletal 
bars and glandular ridges are on the inner side of the 
filaments of the inner row, the free axis must have been 
rotated through 45° to bring the dorsal surface inwards. 
The plate-like filaments springing from the expanded base of 
the free part of the axis spread out on, and are attached to, 
the adjacent parts of the mantle; the filaments, or as they 
more appropriately might be called, the ‘‘ gill-lamelle” of the 
inner row extending dorsally along the inner surface of the 
mantle, while those of the outer row, two or three in number, 
pass round the front edge of the columellar muscle and run 
back for some distance below it as ridges projecting inwards 
from the dependent margin of the mantle (fig. 8, m. br.) The 
blood supply to the ctenidia will be described in connection 
with the heart. 

The rectum runs diagonally from left to right in the roof of 
the mantle-cavity, and the anus opens opposite the slit in the 
mantle edge. In much contracted specimens, such as that 
from which fig. 2 was drawn, the anus is situated some dis- 
tance from the slit, but in other less contracted specimens 
it is close to it. 

The hypobranchial glands lie in the roof of the mantle on 
either side of the rectum, between it and the ctenidia. Both 
consist of a more or less extensive modified glandular patch 
of the internal epithelium of the mantle. The glaud-cells are 
very large relatively to the size of the animal, and’are of two 
kinds: large ovoid cells filled with large granules which 
stain deeply in hematoxylin and green in picro-indigo-car- 
mine; these are therefore mucigenous cells. ‘he other 
gland-cells are- of nearly the same size and shape, but have 
clear or minutely granular contents. The left hypobranchial 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 11 


gland is much the smaller of the two (figs. 7 and 8) ; posteriorly 
it is a narrow strip of glandular epithelium lying between the 
terminal part of the rectum; anteriorly in front of the anus 
it becomes broader and extends about as far forward as the 
level of the mantle slit, but stops far short of the anterior 
border of the mantle. In this pre-anal region the right and 
left hypobranchial glands are very closely approximated in 
the middle line. The right hypobranchial gland has approxi- 
mately the same anterior extension as the left, but runs back- 
wards on the right side of the rectum nearly to the posterior 
end of the mantle-cavity. Comparing the arrangement with 
that described by Woodward for Pleurotomaria, it is obvious 
that the pre-anal portions of the two glands of Incisura 
correspond to the large anterior hypobranchial gland, “ par- 
tially divided by a median longitudinal furrow into two 
halves,” of Pleurotomaria, and the posterior portions of the 
two glands of Incisura correspond to the two “ additional 
mucous glands” lying on either side of the rectum of Pleuro- 
tomaria. But whereas in the latter genus the left additional 
gland is conspicuously the larger, in Incisura it is the right 
posterior portion of the gland which preponderates in size, 
the left gland being small, no doubt because of the relatively 
large size of the left kidney, for the hypobranchial gland 
does not extend beneath this organ. 

The pericardium, as in all Rhipidoglossa except the Helici- 
nide, is traversed by the rectum. It is relatively of large 
size, and can always be distinguished in whole specimens as 
a clear space surrounding the first bend of the rectum on the 
left side of the body behind the columellar muscle. At this 
point it lies close to the surface of the body, and its outer 
wall is very thin and transparent (fig. 11). The exact limits 
of its extension to the right are very hard to make out, 
because the left kidney projects into it from above, and its 
cavity is largely blocked by the auricles. Its extent, as far 
as I am able to determine it by reconstruction from sections, 
is represented by the thick black line in fig. 4. The large 
transverse extension of the pericardial space, as compared 


Le GILBERT C. BOURNE. 


with its narrow limits in Pleurotomaria, Haliotis, or Trochus, 
is correlated with the tendency towards a secondary bilateral 
symmetry, the development of two columellar muscles, and 
the position of the ctenidia wide apart from one another on 
the right and left sides of the body. The necessary result is 
an increased breadth of the body, and the blood returning to 
the heart by the efferent branchial vessels has to traverse a 
considerable distance before reaching the ventricle. In other 
words, the auricles are considerably elongated, and the peri- 
cardium has to be extended to receive them. Very similar 
relations are seen in Fissurella. 

The heart and circulatory system.—The ventricle 
is placed rather far forward on the rectum; no further for- 
ward than in Fissurella, but much further forward than in 
either Pleurotomaria or Haliotis. The walls of the ventricle 
are so thin and feebly muscular that they are difficult to 
recognise, even with the highest powers of the microscope. 
The auricles also have very thin walls but are more easily 
recognisable. ‘The left auricle is relatively very large (fig. 10), 
and its anterior border gives off a number of short and wide 
sinuses, which penetrate the folds of the wall of the left 
kidney and vascularise this organ. The right auricle is of 
smaller size. The course of the blood-vessels, as far as I was 
able to determine it, is of the usual diotocardiate type, and is 
diagrammatically represented in fig. 4, which is fully lettered 
and needs no further description. I was unable to trace the 
course of the aorta, but the blood, after passing to the foot 
and the various viscera, is evidently collected in a large sinus 
lying below the pedal ganglia, and is returned to the afferent 
branchial vessels by sinuses running over the dorsal side of 
the great mass of muscle-fibres which diverge on each side of 
the foot to form the columellar muscles. 

The kidneys.—The left kidney (figs. 8, 9, and 13) is of 
comparatively large size, but its structure and histological 
characters leave no doubt that it corresponds physiologically 
to the papillary sac of the Pleurotomariidz, Haliotide, and 
Turbonidz, for it is unquestionably phagocytic and not depu- 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 13 


ratory. It is a triangular sac lying close alongside of the 
rectum and projecting largely into the pericardium. It opens 
into the mantle-cavity by a simple slit-shaped aperture with 
somewhat tumid lips (fig. 9). The majority of the specimens 
of which I cut sections were females, and in all of them the 
cavity of the sac was large and but slightly broken up by 
ridges or papillae projecting into it. In all the specimens the 
epithelium lining the cavity of the sac and covering the 
papillae had the characters shown in fig. 14. The cells are 
large and pale, with pale nuclei, and most of them are stuffed 
with rod-shaped masses which stain very deeply with iron 
hematoxylin. Whatever may be the nature of these rods, 
which, as shown in the figure, have rhomboid outlines and 
are apparently crystalline, they have clearly been taken up by 
the amceboid cells of the left kidney from the adjoining blood- 
spaces, for these latter are also filled with similar rods, which, 
however, are smaller, more transparent, and stain less deeply 
in hematoxylin. The left kidney differs considerably in appear- 
ance according to its functional activity. In some specimens 
no rod-shaped bodies can be detected in the cells, and the 
walls of the kidney sac then appear pale and thin. In other 
specimens, again, no rod-shaped bodies can be seen in the 
blood-sinuses, but the kidney-cells are stuffed so full of them 
that their outlines are no longer distinguishable. In other 
specimens, again, the rod-shaped bodies are abundant in the 
blood-sinuses and more or fewer are present in the kidney- 
cells. A portion of the epithelium of a specimen in the last 
condition is represented in fig. 14. The fact that the histo- 
logical character of the left kidney or papillary sac in Haliotis 
and Trochus is different from that of the right kidney was 
established by Rémy Perrier in his careful studies on the 
kidneys of prosobranch Gastropoda, and Pelseneer (11) after- 
wards showed that the amcebocytes of the papillary sac take 
up solid particles, such as carmine or Indian ink, injected 
into the blood-sinuses, whereas the secretory cells of the right 
kidney eliminate sulphindigotate of soda injected in solution 
into the blood. Both kidneys of Patella are depuratory, that 


14, GILBERT C. BOURNE. 


is to say, they take up sulphindigotate of carmine from the 
blood, but there is still some doubt as to the very rudimentary 
left kidney of Fissurella. Rémy Perrier (14) describes its 
histological structure as identical with that of the right 
kidney, and consequently it has been generally assumed that, 
like the left kidney of Patella, it is depuratory in function, 
but this is not certain and the subject requires renewed 
investigation. All observers agree in describing the left 
kidney of the Fissurellide as being in a rudimentary con- 
dition, and it is possibly nearly if not quite functionless. It 
may even be absent in some species of Fissurella, for I have 
been unable to find a trace of it in transverse and longitudinal 
sections of F. greca. 

In the single male specimen of Incisura of which I have 
sections the left kidney is larger than in any of the females; 
the papilla projecting into its cavity are more numerous, are 
covered with a much more definite layer of epithelial cells, 
and I could not find any trace of phagocytosis in the latter. 
Whether this is a constant sexual difference I cannot say, as 
I was unable to find another male. A section through this 
kidney is represented in fig. 15, which also shows the left reno- 
pericardial canal. The last-named structure is found in the 
same position in both male and female. It opens into the 
kidney close to the external aperture of the latter, and runs 
towards the left as a very fine canal which traverses the floor 
of the kidney and opens into the left-hand corner of the peri- 
cardium, as indicated in the figure. ‘lhe cells lining the 
nephric end of the canal appear to bear very fine cilia, but I 
am unable to speak with certainty on this point. The right 
kidney of Scissurella and Incisura has been very briefly 
described by Pelseneer (12), who figures it as a very small 
tube lying below the rectum in 8. costata and to the right 
of the rectum in Incisura. He describes it as being rather 
narrow in its anterior portion and says further: “Il s’etend 
partiellement sous le rectum, comme chez Trochus, et pénétre 
dans la masse viscerale, au cOté droit de ce corps, sur et entre 
les convolutions de l’intestin.” 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 15 


I may amplify this account by saying that the right kidney 
of Incisura is a structure of considerable size and importance 
which may be described as consisting of three lobes. The 
most anterior lobe varies considerably in size: it lies in the 
roof of the mantle-cavity to the right of the rectum (figs. 4 
and 10) and somewhat posterior to the left kidney. It opens 
byasimple slit-shaped aperture (fig. 10, /.7.0.) into the mantle- 
cavity, and a few sections further back than the one figured 
it extends over to the right, forming a considerable projection 
into the posterior part of the mantle-cavity. Posteriorly it 
gives off two lobes. That on the right runs nearly vertically 
downwards close to the right side of the vertical loop of the 
intestine and passes inward among the viscera, curving round 
the floor of the middle part of the stomach and eventually 
coming in contact with the gonad, but it does not effect any 
communication with this organ. The left posterior lobe 
passes below the rectum and overlies the anterior cecal end 
of the stomach. 

The excretory cells of the depuratory kidney of Gastropods 
are notoriously difficult to preserve, and in my specimens 
were too much macerated to admit of a satisfactory study of 
their structure. For the same reason I have been unable to 
satisfy myself completely as to the relations of the right reno- 
pericardial canal. For some time I was uncertain whether 
any communication existed between the right kidney and the 
pericardium, but the series of sections represented in figs. 22 
to 26 demonstrate that this connection does exist, and that, as 
in Trochus, there is an intimate connection between the right 
reno-pericardial canal and the gonaduct. In fig. 22 the ovi- 
duct (od.) is seen lying close to the right side of the anterior 
lobe of the kidney, and from it a narrow canal leads upwards 
and inwards. The histological features of this canal are not 
well preserved in any of my specimens, but its walls appear 
to be formed by cubical epithelial cells containing small, deeply 
staining nuclei, whose characters as shown in figs. 22 and 23, 
suggest that they bear cilia and form a ciliated funnel open- 
ing into the pericardium. The connection between the canal 


16 GILBERT C. BOURNE. 


and the pericardium is clearly shown in fig. 24, and figs. 28 
to 25 show that the lower end of the canal is, in fact, con- 
tinuous with the gonaduct, and opens along with it into the 
kidney, close to the external orifice of the latter. It should 
be noted as a peculiar feature in Incisura that there is no 
distinct duct to the right kidney ; its simple slit-like opening 
into the mantle-cavity is a Pectinibranch character. 

The gonad, in both sexes, is a simple tubular structure lying 
to the left side of the stomach, and in the case of the ovary 
partly embracing this organ. The anterior end of the gonad 
extends as far forward as the posterior limit of the mantle- 
cavity and ends blindly below the first bend of the rectum. 
The cavity of the ovary, in all my specimens, is filled with ova 
in all stages of development, the ripe ova being very large 
relatively to the size of the animal, and abundantly supplied 
with yolk-granules. The testis, in the single male I have 
been able to examine, is very small, and I think the individual 
must have been a spent one, as the cavity of the testis only 
contained a few free spermatozoa and I could find no trace of 
spermatogenesis. 

The course of the gonaduct and its connection with the 
right kidney has been correctly but all too briefly described 
and insufficiently figured by Pelseneer. He only says of it: 
‘Ta glande genitale est unique et occupe le sommet de la 
masse viscerale. Elle n’a pas d’orifice exterieure ; son con- 
duit arrive au rein droit.”” But it would be difficult for any- 
body to guess the course of the gonaduct before its arrival 
at the right kidney by an inspection of his fig. 115, perfectly 
correct as it is. As shown in fig. 4 the ovary, which in 
the more anterior and broader part of the visceral mass 
is on the left side of the stomach, extends into the narrow 
commencement of the terminal whorl of the spire, and here 
its posterior end is produced from left to right imto a 
fairly spacious thin-walled sac which hes between the upper 
and lower of the two posterior lobes of the liver extend- 
ing into the spire. The walls of this sac are not lined by a 
germinal epithelium but its cavity often contains a ripe ovum. 


INCISURA (SCISSURELLA) LYTTELTONENSIS. NG 


It is the commencement of the oviduct. Rapidly narrowing 
in diameter it passes forward to the right of the posterior end 
of the stomach and the liver lobes originating from it, and 
maintaining a position close below the external body-wall, it 
passes as a very much flattened and very thin-walled duct 
along the right side of the visceral mass, gradually mounting 
from a more ventral to a more dorsal position till it arrives 
above the right-hand loop of the intestine. All this while it 
has laid close to the outer body-wall, and it is extremely 
difficult to follow its course, owing to its being flattened 
between the liver and the external integuments. It turns in- 
ward just above and in front of the right visceral ganglion 
and runs in the roof of the posterior end of the mantle-cavity 
towards the right kidney (fig. 11). Here its walls become 
thicker and are lined by a distinct cubical epithelium. The 
duct does not at once enter the kidney but runs along its 
outer wall and opens into it in close proximity to the renal 
orifice. As stated above the gonaduct opens into the kidney 
coincidently with a reno-pericardial canal, the relations 
being very similar to those in Trochus. The vas deferens 
takes the same course as the oviduct. 

The alimentary tract.—The buccal bulb is relatively of 
enormous size. There are two large odontophoral cartilages 
on either side whose shape, as seen in section, is very similar 
to that of the cartilages of Fissurella as figured by Boutan. 
As shown in figs. 6, 7, and 8, the anterior and dorsal cartilages 
are the larger, and support the radula ; the posterior cartilages 
lie ventrad of the hinder ends of the anterior cartilages and 
have concave upper surfaces, with which the hinder ends of 
the latter articulate. A similar arrangement obtains in 
Trochus, and has been well described by Randles (15). The 
musculature of the buccal bulb is powerful, but I have not 
attempted to follow it out in detail. It is noticeable, how- 
ever, that the cross-striation, both of the intrinsic and ex- 
trinsic muscles of the odontophore, is very well marked. 
Though it is well known that these muscles are cross-striped 
in Gastropods, I am not aware that the character of the stria- 

VOL. 55, PART 1.—NEW SERIES. 2 


18 GILBERT ©. BOURNE. 


tions has been carefully studied, and I take this opportunity 
of giving a drawing (fig. 15) of three fibres of the extrinsic 
muscles attaching the anterior end of the odontophore to the 
integuments of the snout. ‘These were specially well-stained, 
and it is obvious that the ends of the fibres nearest the snout 
are in a state of contraction, while their odontophoral ends— 
the lower ends in the figure—are relaxed. The fibres are not 
round but elongate oval in cross-section. That on the right 
has been cut through its long axis; in the two fibres on the 
left the section passes through the shorter axis, near the edge 
of the fibres. It can be seen that each fibre is a single 
metamorphosed cell, with a single nucleus situated near its 
broader end. The central portion of the cell, in which lies 
the nucleus, is composed of but little-altered cytoplasm, 
exhibiting an alveolar or reticular structure, differing from 
the normal only in the fact that the meshes of the reticulum 
are very regularly disposed in rectangular fashion. This 
cytoplasmic core of the fibre is invested by a sheath of con- 
tractile substance, which is thickest at the two ends of the 
long axis of the oval, and therefore appears as two bands in 
the right-hand fibre in the figure, while in the two left-hand 
fibres only the contractile substance is cut through. The 
whole is invested by a delicate sarcolemma. ‘The most 
interesting thing about these fibres is that the reticular 
arrangement of the cytoplasmic core corresponds exactly with 
the striations of the contractile substance in the upper part of 
the fibre on the right side of the figure, and in the left-hand 
fibre the cross-striations are very obvious and close together 
in the uppermost contracted part of the fibre, but lower down 
as the fibre becomes more relaxed, the dark transverse lines 
become progressively broader and fainter, and each may be 
seen to be made up of a number of dark longitudinal strie, 
which may well be interpreted as nodal thickenings of a 
reticulum. It is, of course, possible that the difference 
between the two ends of the fibres is due, not to a difference 
in the state of contraction, but to a greater specialisation of 
the broader end. Whichever interpretation is correct, the 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 19 


appearances lend support to the reticular theory of the con- 
stitution of striped muscle-fibre, and are inconsistent with 
the opposing theory of sarcomeres. 

The mandibles occupy the usual position at the sides of the 
mouth, and are composed of a number of plates or “tesserae” 
as described by Vayssiére for Scissurella costata. Randles 
has shown that each tessera is the product of a single epithe- 
lial cell in Trochus, and the same is evidently the case in 
Incisura. The radular sac occupies the usnal position. 
Lying at first between the upper horns of the odontophoral 
cartilages it maintains a median position to the posterior end 
of the buccal bulb, and then curves to the right between the 
right cesophageal pouch and the pedal ganglia and soon ter- 
minates in a swollen bilobed extremity lying on the right side 
of the hemocele. The radular teeth are represented in fig. 
20. The centrals are squarish, with an expanded basal plate ; 
their anterior margins decurved, and furnished with five very 
distinct and sharp-pointed denticulations. The next three 
teeth (medio-laterals) are oblong, with decurved denticulate 
margins ; they decrease somewhat in size from within out- 
wards. The next tooth is much smaller, has a somewhat 
sigmoid curvature, a thickened base, a narrow neck, and a 
single recurved marginal denticulation. The next tooth is 
very large, shaped somewhat like a rake with a crooked 
handle, its expanded margin decurved and bearing about a 
dozen denticulations. Then follow the marginals or uncini, 
which are numerous, curved, slender, with expanded and re- 
curved denticulate margins. The radular formula may be 
written : 

Oi1(4+1+44) 10 

Vayssiére has given a good figure of the radula of Scis- 
surella costata, which is similar to but differs in small 
details from that of Incisura. The radula of the Scissurellidee 
is usually described as resembling that of 'Trochus, but it is 
much more nearly like that of the Fissurellide. A reference 
to Thiele’s figures in the concluding chapter of Tréschel’s 
‘Gebiss der Schnecken’ shows that the radula of Incisura 


20 GILBERT C. BOURNE. 


very closely resembles that of Subemarginula picta, the 
shape and relative size of the outer medio-lateral tooth bemg 
almost identical, as also the characters of the centrals and 
uncini. ‘The large and specialised outer lateral tooth, though 
it differs widely in detail in different species, is characteristic 
of the Fissurellide. A close resemblance also exists between 
the radula of Incisura and that of Emarginula pileolus, 
and a less clearly marked resemblance can be seen in the 
radulee of various species of Fissurella. On the other hand, 
no comparison with the radula of Pleurotomaria is possible. 
A general view of the alimentary tract, as determined by 
reconstruction from sections, is given in fig. 3, which so far 
explains itself that little description is necessary. The ceso- 
phagus is enormously dilated in the anterior part of its course, 
forming in addition to the wide lateral diverticula or cesopha- 
geal pouches (figs. 9 and 10, w. p.) a spacious ventral pocket 
or “‘jabot.” These are all lined by a soft-looking glandular 
epithelium. Behind the level of the pedal ganglia the 
posterior section of the cesophagus leaves the jabot as a 
narrow tube with thick, longitudinally ridged walls formed by 
a long ciliated columnar epithelium. It runs back below the 
stomach and opens into the latter near its posterior end. 
Near the cesophageal opening numerous liver czeca open into 
the posterior end of the stomach. There is no spiral cecum 
connected with the entry of the liver-ducts as in Pleuroto- 
maria, Haliotis, and Trochus, but there is a deep ciliated 
ventral groove, the lips of which are bordered by specially 
long ciliated columnar cells, extending along the floor of the 
stomach from the cesophageal opening to the pylorus. A 
precisely similar groove occurs in the stomach of Fissurella, 
and has been well described and figured by Boutan (2). 
Randles has shown that in Trochus a cxcal groove, bounded 
by two conspicuous folds, extends into the spiral caecum from 
the cesophageal opening, and that the larger of the two bile- 
ducts opens into this groove. Though the spiral cecum is 
absent there can be little doubt that the ventral groove of the 
Fissurellidz and Incisura corresponds in function to the cecal 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 21 


groove of the Trochide, and it has the same relation to the 
liver-ducts. It should be noted in this place that Incisura, in 
the possession of numerous biliary apertures, resembles 
Fissurella and differs from Trochus, which has two, and 
Pleurotomaria, which has only one bile-duct. The intestine 
leaves the stomach on the ventral side of the anterior third of 
the stomach in Incisura. Beyond it the stomach narrows 
rather abruptly, and is continued forward as a small cecal 
diverticulum, the front end of which is inserted in the loop 
formed by the left-hand bend of the rectum. The walls of 
the blind end of this diverticulum are covered internally by a 
thick chitinoid layer, and thrown into complicated folds and 
ridges, but the cecum is not spirally coiled, and situated as it 
is at the end of the stomach furthest from the bile-ducts, it 
cannot be homologised with the spiral caecum of Pleuroto- 
maria, Haliotis, or Trochus. It must, however, be the cecum 
referred to by Pelseneer (12). The walls of the intestinal end 
of the stomach of Incisura have the columnar cells with 
striated borders and thick cuticle so fully described by 
Randles for Trochus. 

The intestine is provided throughout its length with a 
single longitudinal ridge or typhlosole. On leaving the 
stomach it makes a sharp bend from left to right, passes 
vertically upwards to above the level of the stomach, thence 
turns sharply to the right, describes a wide loop on the right 
hand, as shown in fig. 3, and bending sharply again to the 
left, passes nearly straight across the body till it reaches the 
left-hand corner of the pericardium, when it turns upward 
and to the right in the mantle roof, and becoming rectum, 
traverses the pericardium in its diagonal passage across the 
roof of the mantle-cavity to end in the anus opposite the 
mantle-slit. 

The liver ceca, as may be seen in figs. 3 and 11, are few 
in number, of relatively large size, with large lumina bordered 
by large secretory cells. As far as I could determine they 
do not branch, but have somewhat convoluted courses, and 
open independently into the cesophageal end of the stomach. 


22 / GILBERT C. BOURNE. 


A few details may be added relative to the structures 
connected with the buccal cavity and cesophagus. 

In the mid-dorsal line the roof of the buccal cavity is 
deeply folded to form a median ridge containing a narrow 
lumen T-shaped in transverse section. This lumen of course 
communicates freely below with the buccal cavity. This 
median fold or ridge is deepest anteriorly over the mouth, 
and extends backwards for about two thirds of the length 
of the buccal bulb, gradually shallowing posteriorly till it 
dies out altogether. The walls of this ridge are composed of 
simple columnar cells, the internal ends of which have a 
striated border, and bear short cilia. On either side of the 
anterior part of this mid-dorsal ciliated groove is a somewhat 
shallower but still conspicuous groove appearing on the 
dorsal surface as a pair of folds lying close and parallel to 
the median ridge. These may be called the salivary grooves, 
for the small, simple, tubular buccal or anterior salivary 
glands open into them near their anterior extremities (fig. 21, 
sg. and s. d.). These anterior salivary glands are simple 
short ceca lined by an epithelium, consisting mainly of large 
finely granular secretory cells with a few columnar supporting 
cells between them. ‘lhe salivary grooves die out posteriorly 
at the point where the cesophagus leaves the buccal cavity, 
and at this level a second or posterior pair of salivary glands 
opens into the roof of the buccal cavity, just to the outside 
of the salivary grooves. These posterior salivary glands are 
very small tubalar structures with minute lateral diverticula. 
They correspond in position to, but are much smaller than, 
and not so much branched as the second pair of salivary 
glands in Fissurella. Otherwise the structures just described 
are identical in the two genera. As soon as the cesophagus 
is separated from the buccal cavity its right and left walls 
are produced into the broad and flattened cesophageal pouches, 
but from the first the right-hand pouch is considerably larger 
than the left. The T-shaped lumen of the dorsal ciliated 
groove may be traced for some way along the roof of the 
esophagus, but presently it dies out, and is replaced by a 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 23 


band of ciliated cells which diverges towards the left, and 
eventually passes completely over to the left side and passes 
into the narrow posterior part of the cesophagus. Ventrally, 
to the right side of the narrow cesophageal tube, the floor of 
the spacious anterior cesophageal cavity is produced into a 
capacious pouch or “jabot,” which runs back for some 
distance alongside of the narrow cesophageal tube (fig. 11, 7), 
and eventually ends blindly. The deviation of the cesophagus 
to the left and the preponderant size of the right cesophageal 
pouch have been noted by Boutan in Fissurella, and it is 
indeed a common feature in the Rhipidoglossa, indicative, as 
Amandrut has pointed out, of the larval torsion which brings 
about the asymmetry of the adult Gastropod. 

The nervous system.—Fig.5 is adiagram of the prin- 
cipal ganglia and nerve-trunks, as reconstructed from sections. 
Pelseneer’s description of this system in Scissurella costata 
and Incisura lytteltonesis is as follows: “ Dans les deux 
espéces, les cordons pédieux sont dans la masse musculaire 
du pied, et s’étendent jusqu’a la partie postérieure. A leur 
extrémité tout a fait antérieure se trouvent des ganglions 
pleuraux bien distincts. La commissure viscerale nait de ces 
derniers ; elle est croiseé et porte un ganglion supra-intes- 
tinal presque accollé au ganglion branchial ou osphradial 
gauche, comme dans ‘Trochus. ‘'‘l'out ce systéme nerveux 
ressemble done beaucoup plus a celui de Trochus qu’aux 
parties correspondantes connues de Pleurotomaria, telles 
que les ont décrites Bouvier et Fischer.” Since this was written 
we have had the more complete account of the anatomy of 
Pleurotomaria by M. F. Woodward, and the difference between 
the nervous system of this genus and that of the Scissurellidee 
is even more apparent than before. 

As may be seen from the diagram, the nervous system of 
Incisura is at once typically Rhipidoglossate and specialised. 
As the nervous systems of various Rhipidoglossa have been 
described in great detail by sundry authors, it will only be 
necessary here to mention the more important and pecuhar 


features. 


24, GILBERT C. BOURNE. 


The cerebral commissure is long and situated far forward 
in front of the anterior pair of salivary glands. It is a true 
nerve, not ensheathed by a layer of ganglion cells, differing 
in this from Pleurotomaria. The cerebral ganglia are of large 
size, sub-triangular in transverse section, and produced into 
prominent lobes at the origins of the more important nerves. 
The tentacular and optic nerves have separate origins from 
the cerebral ganglia, Incisura agreeing in this point with 
Trochus and Fissurella but differing from Pleurotomaria. 
The labial lobe is very large, and forms a long, conical, taper- 
ing, antero-ventral process of the cerebral ganglion, which 
curves inward below the odontophore on either side, maintain- 
ing its thickness for about two thirds of its course towards 
the middle line. Then it tapers abruptly to form a thin labial 
nerve, which passes between the muscles of the lower lip, and 
as far as I can determine is connected by an extremely fine 
prolongation with its fellow of the opposite side, thus com- 
pleting the labial commissure. The buccal commissure is 
given off from the labial lobe about half way between the 
cerebral ganglion and the mid-ventral line. It passes inwards 
among the muscles of the odontophore and at once turns 
abruptly upwards to run between the extrinsic and intrinsic 
muscles to the top of the buccal bulb. Here it enlarges to 
form a ganglion of considerable size, lying close to the inside 
of the cerebral ganglion, and from this a stout nerve—a true 
nerve without a sheath of ganglion cells—passes inwards and 
backwards over the top of the odontophore and enlarges below 
the origin of the cesophagus into a small ganglion, which is 
connected by a very short commissure with its closely adjacent 
fellow of the opposite side. Bouvier (8) has figured and 
described two swellings at the ends of each of the elongated 
buccal ganglia of Turbo setosus, but I infer from his descrip- 
tion that they are not separate ganglia, but merely swellings at 
the ends of a long and ill-defined ganglion. I find precisely 
the same arrangement in Fissurella greca, but Boutan 
figures four clearly defined ganglia in F. reticulata. The 
sub-division of this elongated ganglion into two distinct 


——————— ee 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 25 


ganglia isan indication of specialisation and a peculiar feature in 
Incisura. For the rest the characters of the cerebral ganglia, 
the size of their labial lobes, and the relations of the buccal 
ganglia are very similar in Turbo, Fissurella, and Incisura. 

The cerebro-pleural connective, as is commonly the case, is 
larger than the cerebro-pedal; both are true nerves, devoid of 
any sheath or local accumulations of ganglion cells. The 
pleural ganglia are distinct and that of the right side is 
relatively large, but both are fused to the dorsal surfaces of 
the pedal ganglia. The visceral commissure is typically 
streptoneurous, and for the same reason that the osphradial 
ganglia are situated far forward, the whole commissure is con- 
tracted antero-posteriorly as in Patella; on the other hand, it 
is considerably extended right and left. The sub-intestinal 
ganglion is distinct, but elongated and rather ill-defined; as 
Pelseneer remarks it is connected by a very short nerve with 
the large left osphradial ganglion. The left symmetrical 
pallial nerve passes straight out from the left pleural ganglion 
almost immediately below the supra-intestinal ganglion, 
and traverses the posterior fibres of the left columellar 
muscle, turning nearly verticaily downwards to enter the 
thickened border of the mantle. Before turning downwards 
it gives off a very fine branch, which makes connection with 
the short nerve uniting the supra-intestinal with the osphradial 
ganglion, thus establishing a left-hand dialyneury very 
similar to that of Trochus. 

The subintestinal nerve is very stout, and crosses over the 
dorsal surface of the hinder part of the pedal ganglia almost 
at right angles to the long axis of the body. ‘The sub-intes- 
tinal ganglion is fairly large and distinctly indicated by an 
accumulation of nerve-ganglion cells. It is triangular in 
shape, and from its right-hand lower corner the visceral nerve, 
and from its right-hand upper corner the osphradial nerve is 
given off. The latter is a very slender nerve, which passes 
into the substance of the columellar muscle, and turns verti- 
cally downward and then forward along the dependent edge 
of the mantle, running in this part of its course at the base 


26 GILBERT ©. BOURNE. 


of the gill-filaments, which, as has been explained above, run 
back along this region of the mantle. At the anterior edge 
of the columellar muscle the nerve expands to form the large 
right osphradial ganglion. ‘The right symmetrical pallial 
nerve takes its origin from the ventral side of the right 
pleural ganglion, just where the latter becomes fused to the 
pedal ganglion. It runs outward, traverses the columellar 
muscle some way in front of the osphradial nerve, and takes 
a direct course to the right osphradial ganglion, which it 
crosses dorsally, and in so doing enlarges and makes an 
intimate connection with it. Just in front of the osphradial 
ganglion the pallial nerve divides into two branches. The 
posterior branch, which is slender, runs back along the 
thickened border of the posterior part of the mantle. The 
anterior branch runs forward to the mantle-slit, where it 
expands to form a small ganglion, indicated by a distinct 
accumulation of nerve-gauglion cells, and is here joined by a 
slender nerve from the anterior end of the osphradial ganglion. 
This little ganglion at the hinder border of the mantle- 
slit gives off an external branch supplying the posterior sense- 
papillz of the mantle-slit, and a stout anterior branch which 
passes round the mantle-slit and is continued forward as the 
peripheral pallial nerve, meeting and uniting with its fellow 
of the opposite side on the anterior border of the mantle. 
There is thus a very intimate dialyneury on the right side. 
These relations are very hard to make out, and require careful 
study with high powers of the microscope, but I can vouch for 
the correctness of the account here given of them. ‘The rela- 
tions in Fissurella are somewhat similar, but the proportions of 
the lengths of the nerves differ greatly, and apparently differ 
in different species, for in my sections of F. greca the sub- 
intestinal is close to the right osphradial ganglion, whereas 
in F. reticulata Boutan figures them as far apart and con- 
nected by a long slender nerve, as in Incisura. The origin 
of the right symmetrical pallial nerve from the upper surface 
of the pedal ganglion rather than from the right pleural 
ganglion is identical in Incisura and Fissurella. 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 27 


The visceral loop bears three distinct accumulations of 
ganglion cells, forming as many ganglia. The right ganglion 
lies close below the gonaduct and gives off a slender nerve to 
that organ. ‘The pedal ganglia, as may be seen in fig. 5, are 
very much concentrated. Anteriorly they are rather flat, but 
in about the middle of their length they increase considerably 
in thickness, this increase being due to the addition of a 
considerabie ventral thickening to each ganglion. In this 
region, in fact, each pedal ganglion consists of a dorsal and a 
ventral moiety, as is the case in all Rhipidoglossa (fig. 9). 
Here also the whole of the pedal ganglia lies in the hemoccele, 
as is the case with the more elongated pedal cords of Fissurella. 
But in Incisura the dorsal moieties of the pedal ganglia have 
very little posterior extension. The ventral moieties, on the 
other hand, extend back behind the dorsal moieties, and, 
narrowing in diameter, plunge into the muscular substance of 
the foot (fig. 10). ‘There they are continued backwards for a 
short distance, giving off nerves from their outer edges, and 
diminishing rapidly in diameter, partly because of fibrils 
given off to the different nerves, but also largely because of 
the thinning out and eventual disappearance of their coating 
of nerve ganglion cells. Posteriorly the cords become simple 
nerves, and end some distance in front of the posterior end of 
the foot. Pelseneer states of Scissurella costata and 
Incisura lytteltonensis: “Dans les deux espéces, les 
cordons pedieux sont dans la masse musculaire du pied, et 
s’étendent jusqu’a la partie posterieure.’ This is certainly 
not the case in Incisura; the left pedal cord, or rather nerve, 
dies out at a distance of 125 w from the posterior end of 
the foot in two specimens in which I calculated its extent, and 
remembering that the animal is only 1 mm. long this is a 
considerable distance. In short, one can hardly speak of 
pedal cords. The pedal centres, particularly the dorsal 
portions of them, have become concentrated into two clearly 
defined pedal ganglia, and it is only the ventral portions that 
are continued backwards to represent in some measure the 
elongated pedal centres of other Rhipidoglossa. In addition 


28 GILBERT C. BOURNE. 


to the thick anterior commissure connecting the dorsal 
portions of the ganglia, there is a single anterior thin com- 
missure connecting the ventral portions, but this is the only 
trace of the usually numerous cross commissures of other 
lowly organised Gastropoda. Such a concentration of the 
pedal centres is very unusual if not unique among Aspido- 
branchia, and indicates that Incisura, and, if one may judge 
from the similar relations indicated in Pelseneer’s figures of 
S. costata, the Scissurellide in general are highly specialised. 
Much has been written about the significance of the dorsal 
and ventral moieties of the pedal cords of archaic Gastropods. 
The French authors hold that the upper moiety is pleural, or, 
as they say, pallial, the lower moiety pedal in character. 
Pelseneer and most English and German authors hold that 
both moieties represent pedal centres. The facts in Incisura 
seem to uphold the latter view. I have no wish to re-enter 
upon a controversy which has become almost wearisome by 
repetition, but may state that in Incisura the cerebro-pedal 
connectives certainly join the dorsal moieties of the ganglia; 
that the epipodial nerves are certainly given off from the 
dorsal moieties, and that whereas the left symmetrical pallial 
nerve is undoubtedly given off from the left pleural ganglion, 
the right symmetrical pallial nerve certainly appears to be 
given off from the dorsal moiety of the right pedal ganglion 
and not from the right pleural, both in Incisura and Fissurella. 
Advocates of the French view will take this last fact as 
evidence in support of their theory. The nervous system of 
Incisura certainly bears no resemblance to that of Pleuro- 
tomaria. On the whole it most nearly resembles that of the 
Fissurellide, in which family the pedal cords, though still 
elongate and ganglionic, and provided with several cross- 
commissures, have undergone a considerable reduction in 
length as with those of other Rhipidoglossa. 

The sense organs.—The eyes, as already stated, are closed 
and provided with a distinct lens. Their structure resembles 
that of the eyes of the Fissurellidz, and differs from the eyes 
of the Pleurotomariide and Trochidz, which are open. 


ee ee, Eo 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 29 


The otocysts occupy the usual position on the dorsal surfaces 
of the pedal ganglia and present no unusual features (fig. 9). 

The osphradia are strips of modified epithelium running for 
some little distance along the lower side of the gill-axes in 
front of the osphradial gangha and just ventral to the osphra- 
dial or branchial nerve (fig. 16). They are very similar in 
structure and position to the osphradia of Fissurella greca. 

Sense-papillee occur not only on the cephalic tentacles but 
also on the epipodial tentacles, all round the margins of the 
mantle and on the cirrhi bordering the manile-slit. Those on 
the cephalic tentacles are by far the largest, those on the 
margins of the mantle are very minute, but all have essenti- 
ally the same structure. Fig. 28 represents a longitudinal 
section through three of the papillee of the cephalic tentacles. 
Each papilla is a conical projection of the integument of the 
tentacle and is composed of a number of elongated cells of 
two kinds, closely packed together like the cells in a taste- 
bud from the human tongue. ‘I'he larger cells with larger, 
pale nuclei are evidently supporting cells, their characters 
being similar to the adjoining epithelial cells. The more 
slender, finely granular cells with smaller, deeply staining 
nuclei are the sense-cells, and each ends in a short stiff 
cilium projecting from a small cup-shaped depression at the 
end of the cone. According to Vayssiére these cilia are in 
constant movement in the living animal. The tentacles of 
Fissurella are clothed with a vast number of minute papille 
giving a velvety texture to the surface. These papille, 
though not so highly specialised, have each a single apical 
sense-bulb, the structure of which is similar to that of the 
sense-papille of Incisura. 

Finally, mention may be made of the pedal glands. The 
anterior pedal gland consists of a mass of unicellular glands 
lying in the hemoccele below the buccal bulb (fig. 7, p. gl.). 
It extends back nearly as far as the pedal ganglia. Ante- 
riorly these glands become more deeply seated and pass into 
the muscular mass of the foot, where they debouch into a 
median ciliated duct (fig. 6) which runs forward and opens 


30 GILBERT C. BOURNE. 


on the anterior face of the foot in the groove between it and 
the lower surface of the snout. The posterior pedal glands 
are a mass of unicellular glands lying above the epithelial 
cells of the sole of the whole posterior surface of the foot. 
Kach unicellular gland has its own duct, which runs between 
the epithelial cells to open on the surface. 

The genera Scissurella, Schizotrochus, Incisura and Schis- 
mope, which have been grouped as a separate family Scissu- 
rellidee by some few authors, are generally placed in the family 
Pleurotomariide because they are zygobranchiate Rhipido- 
gloss, with a labral incision of variable length and position 
in the shell. ‘There is no frontal veil between the cephalic 
tentacles, an epipodial ridge is present, and there is a 
corneous multispiral operculum. Fischer (5) writes: “ Quel- 
ques auteurs distinguent deux familles, Scissurellide et 
Pleurotomariide, mais les differences qui existent entre ces 
deux types n’ont pas plus d’importance que celles qu’on con- 
state entre les divers groupes de Trochidee. Je les considere 
comme des sous-familles.”” Pelseneer (18), who had studied 
their anatomy, retains these forms in the family Pleuroto- 
mariide. Yet it is obvious, from what precedes, that the 
Scissurellidee cannot possibly be retained in this position. 
The differences in the radula alone are sufficient to distinguish 
the two types. But in addition to this the Scissurellide 
differ from the Pleurotomariidz in a number of characters, 
which may be summarised as follows: 

(1) The Scissurellidz have two columellar muscles ; Pleuro- 
tomaria has only one. 

(2) The eyes of Scissurellidee are closed ; thoso of Pleuro- 
tomaria are open. 

(3) The subocular tentacles of the Scissurellide are absent 
in Pleurotomaria. 

(4) The epipodium of Pleurotomaria is destitute of tentacles, 
cirrhi, or lappets. 

(5) he wide distance apart of the ctenidia, the large size 
of the pericardial cavity, the forward position of the ventricle 
of the heart, and the more distinct shifting of the organs of 


———— 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 31 


the pallial complex into a median position in the roof of the 
mantle-cavity are all points in which the Scissurellide differ 
from Pleurotomaria. 

(6) In Pleurotomaria the right kidney has a distinct duct, 
with thickened glandular walls in the female; in the Scis- 
surellidze there is no such duct. 

(7) There is no spiral caecum to the stomach in the Scis- 
surellide, and the form of the stomach differs largely from 
that of Pleurotomaria. 

(8) The hepatic orifices are numerous in Scissurellide, 
whereas there is only a single orifice in Pleurotomaria. 

(9) The nervous system of the Scissurellide differs in detail 
in almost every point from that of Pleurotomaria, particularly 
in the concentration of the cerebral ganglia; the extreme fine- 
ness of the labial commissure; the presence of distinct pleural 
ganglia; the well-developed symmetrical pallial nerves estab- 
lishing a right and left dialyneury; the presence of distinct 
supra- and sub-intestinal ganglia; the shortness of the visceral 
loop; the concentration and abbreviation of the pedal centres. 

Not only are the Scissurellide distinct from the Pleuro- 
tomaride, but they are clearly less closely related to them 
than the Halitide or even than the Trochide and Turbonide, 
for the last-named families, though they have lost the labral 
incision in the shell, as also the right ctenidium and the 
structures correlated to it, have retained many anatomical 
features which find their counterpart in Pleurotomaria. 

Where, then, shall we find the nearest relatives of the 
Scissurellide ? Though Mr. Hedley was clearly in error in 
removing Incisura from the Scissurellide, I think he came 
very near the truth in suggesting the affinity of this genus 
with the Fissurellidee. His comparison of the adult Incisura 
with the post-larval stage of Fissurella is a just one. Almost 
all the differential external features which serve to distinguish 
the adults disappear on comparison of the adult of the one 
type with the post-larval stage of the other. In the young 
Fissurella we see a coiled shell with spiral sculpture, a labral 
incision of considerable length to the right of the middle line. 


32 GILBERT C. BOURNE. 


There is a pair of ciliated post-ocular tentacles on either side 
of the head (I find vestiges of these structures in the adult of 
F. greca), a well-developed pair of ciliated epipodial ten- 
tacles in the vicinity of the opercular lobe, and a corneous 
multispiral operculum. Even the gills, if one may judge from 
Boutan’s figure (PI. xlii, fig. 8), have a close resemblance to 
those of a Scissurellid. Ifthe animal were sexually mature one 
would not hesitate to place it among the Scissurellide. In the 
next or Rimuliform stage the epipodial tentacles are multi- 
plied ; Boutan figures six in addition to the sub-ocular tentacles 
in F. reticulata and two in F. gibba. The labral incision 
has been converted into a foramen by the approximation of its 
edges at the labrum, but a suture still connects the foramen 
with the margin of the shell. his condition is exactly paral- 
leled by the Scissurellid genus Schismope. Subsequent de- 
velopment leads to the assumption of Fissurellid characters. 
The visceral spire, and with it the spiral coils of the shell, become 
obsolete. The foramen in shell and mantle become situated 
at the summit of the Patelliform shell, the post-ocular and 
epipodial tentacles (which obviously belong to the same 
series) degenerate, the operculum is cast off, and the oper- 
cular lobe disappears. In short, the Fissurellid develops 
along lines which remove it further and further from the 
Scissurellid condition of the larva. 

But, as must be apparent from the preceding pages, there 
is a considerable number of anatomical features in which the 
adult Scissurellid more nearly resembles the adult Fissurellid 
than any other family of the Rhipidoglossa. These features 
may be shortly recapitulated, Incisura being taken as a type 
of Scissureilid structure. 

The jaws of Incisura in position and structure very closely 
resemble those of a Fissurella. The radula of Incisura 
lytteltonensis finds its nearest counterpart in the radula 
of Subemarginula picta, and in general is distinctly 
Fissurellid in character. In the alimentary tract the characters 
of the salivary glands and cesophageal pouches, the absence 
of a spiral caecum in the stomach, the presence of an ceso- 


EEE 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 33 


phageo-intestinal groove in the capacious stomach, the 
existence of numerous hepatic ducts, are all points in which 
Incisura agrees with Fissurella, and differs, to a greater or 
less degree, from the Pleurotomaride, Haliotide, Trochide, 
and Turbonide. The presence of a right and left columellar 
muscle in the Scissurellidz is evidently an antecedent stage 
of the horse-shoe shaped columellar muscle of the Fissurellide. 

The eyes, which are open in Pleurotomariide, Haliotide, 
and 'I'rochide, are closed in both the Scissurellide and the 
Fissurellide. 

The subocular and posterior epipodial tentacles of the 
Scissurellidz are paralleled by the similar larval organs in 
the Fissurellide, 

In both the Scissurellidz and Fissurellide the increased 
size of the last whorl of the shell and the diminution of the 
visceral spire has led to a broadening of the dorsal part of 
the body, in consequence of which the bases of the ctenidia 
are widely separated on the right and left sides of the body, 
the pericardium is transversely elongated, and the heart and 
kidneys are shifted towards the mid-dorsal line in the roof 
of the mantle-cavity. In these respects Incisura is inter- 
mediate between Fissurella and the other families of Rhipido- 
glossa enumerated above. 

The nervous system of Incisura, though much specialised, 
shows more resemblance to that of the Fissurellidze than to 
that of any other Rhipidoglossa, as has been explained in 
detail in the descriptive part of this paper. The corre- 
spondence in the labial commissure, the buccal ganglia, and 
the visceral commissure is very exact. The pedal centres of 
the Scissurellidee have undergone great concentration, but 
this is foreshadowed in the pedal cords of the Fissurellide, 
which are much shortened in comparison with the elongated 
scalariform pedal centres of such families as the Pleuroto- 
maruidee, Haliotide, and Trochide. 

There can be little doubt, then, as to the affinity of the 
Scissurellidz with the Fissurellide, but the exact relationship 
of the two families remains to be considered. In my opinion 

VOL. 55, PART 1,—NEW SERIES. 3 


34 GILBERT C. BOURNE. 


it is not exact tosay, as Hedley has, that Incisura represents 
an arrested stage of development of a Fissurellid. It is a 
more reasonable inference from the facts that the two 
families have descended from a common stock, and have 
diverged in different directions. There are several arguments 
in favour of this inference. One which in my opinion has 
great weight is derived from the condition of the left kidney 
in the two families. In the Scissurellidz, as I have shown, 
the left kidney is relatively of large size, and is a true 
‘papillary sac,”’ phagocytic in function like the left kidney 
of the Pleurotomariidz, Haliotide, and Trochide. In the 
Fissurellidz this organ is reduced to a mere rudiment, and 
may, I believe, disappear altogether in some species, for I 
have failed to find a trace of it in transverse and horizontal 
sections of I. greca. , 

Remy Perrier (14) has stated that the epithelium of the 
left kidney of Fissurella is identical with that of the right 
kidney, but there is some doubt about this, and a renewed 
investigation of the left kidney of several species of the 
Fissurellide is much to be desired. But there is no doubt 
that it is a vestigial organ, and that in this respect the 
Fissurellide have been specialised along a different line to 
the Scissurellide, which have retained the left kidney in a 
fully functional state. Per contra, while the Fissurellide 
retain to a large extent the primitive scalariform character of 
the pedal centres, the Scissurellide have in this respect sur- 
passed them in specialisation, for their pedal centres are 
concentrated to a degree elsewhere unknown among the 
Rhipidoglossa. The divergence of the two types is obvious, 
and one may conclude that both have been derived from a 
stock very nearly represented by the so-called Emarginuliform 
larva of Fissurella, which had a spirally coiled shell with a 
large umbilicus, spiral sculpture and a considerable labral 
incision. A corneous multi-spiral operculum and a well- 
developed epipodial ridge bearing sub-ocular as well as 
posterior epipodial tentacles were present. The left kidney 
was a well-developed papillary sac, and the pedal centres were 


INCISURA (SCISSURELLA) LYTTHLTONENSIS. 35 


elongate and scalariform. Such an ancestral form would not 
be far removed from a Pleurotomaria, but would differ from 
it in the development of a double columellar muscle and in 
the tendency to acquire a secondary symmetry always 
correlated with the doubling of this muscle. The Scissurellidee 
have retained most of the features of this parent form, but 
have undergone considerable specialisation in the nervous 
system. The Fissurellid branch must early have acquired a 
“ sessile ” habit, and have been much modified in connection 
with it, but its members have largely retained the primitive 
condition of the pedal centres. The Scissurellide, though for 
the most part constant to the primitive type, are also under- 
going modification in the same direction as the Fissurellide. 
In Incisura the visceral spire is reduced, the shell is becom- 
ing thick and solid, the spiral sculpture is absent, the margins 
of the aperture are in one plane, the foot is becoming short 
and broad, and its whole organisation is indicative of a semi- 
sessile habit. Further specialisation along these lines would 
give it Fissurelliform or rather Emarginuliform characters. 
It is interesting to note that another member of the family, 
Schismope, while retaining its spiral coil and widely open 
umbilicus, has undergone specialisation in another direction, 
for the labral shit has been converted into a foramen by the 
approximation of its edges, so that although distant from the 
margin it is connected with it by a suture. In this respect 
it closely resembles Semperia, a sub-genus of Emarginula. 
Semperia leads on to Rimula, and as we have seen there are 
Emarginuliform and Rimuliform stages in the development of 
Fissurella. This is an undoubted example of the develop- 
mental stages of one form resembling the adult stages of 
other forms, a phenomenon the occurrence of which some 
persons are inclined to deny nowadays, though the evidence 
in favour of it is very large. 

The parallel stages of evolution among the Scissurellidz 
and Fissurellide afford interesting examples of the pheno- 
menon of convergence, and illustrate a principle which, I think, 
has not been sufficiently attended to in drawing inferences as to 


36 GILBERT C. BOURNE, 


the affinities of animals from morphological evidence, namely, 
that a similar environment and similar habits of life reacting 
on a similar organisation may often produce very similar struc- 
tural results. Not, however, identical, for however similar the 
results may appear at first sight in all cases of convergence 
a close analysis will always disclose differences which exclude 
the idea of direct descent of the animals in question, This 
instance is particularly instructive; the Haliotide, Scissu- 
rellidee and Fissurellidz have all inherited the same structure 
from a presumably Pleurotomariid ancestor, viz. the slit in 
the mantle and the corresponding labral incision in the shell. 
It has been variously modified, and similar modifications are 
displayed independently by different groups, the similarity of 
the evolutionary series being, as far as one can judge, 
correlated with the adoption of similar habits, 


ADDENDUM. 


It is long since I first read the short but profound essay 
of Sir Ray Lankester (9) “On the Use of the term Homology 
in Modern Zoology,and the Distinction between Homogenetic 
and Homoplastic Agreements.” On referring again to this 
essay, I find that the conclusions arrived at in the foregoing 
paragraph, as also similar conclusions arrived at after a 
detailed study of various members of the Neritide (1), are 
unconsciously expressed in nearly the same words that he 
used forty years ago. I have to beg Sir Ray Lankester’s 
pardon for not making specific reference to his essay in my 
former paper. But I find a certain satisfaction in not having 
had the form of his argument clearly in my mind while I was 
working to the same conclusion from evidence gathered from 
the study of the probable lines of descent of animals belong- 
ing to a different class to that which he used to illustrate his 
original thesis. Had I consciously set out to prove, or even 
to disprove, his contention, I could not have avoided a certain 
amount of bias. To have arrived unconsciously—or sub- 
consciously, for the idea of homoplasy inculcated by him was 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 37 


always present to my mind—at an identical conclusion is to 
give unequivocal support to the validity of the arguments by 
which it was sustained. In the essay in question Lankester 
showed that the term homology, which really belonged to 
the platonic school of the natural philosophers of the end of 
the eighteenth and the beginning of the nineteenth century, 
acquired a new connotation after the publication of the ‘Origin 
of Species.’ But this new connotation was indefinite. On 
the one hand structures were said to be homologous which 
“are genetically related, in so far as they have a single 
representative in a common ancestor.” For this kind of 
homology Lankester proposed to substitute the term ‘ homo- 
geny.” On the other hand, various organs were described as 
homologous which could not possibly be included under the 
idea of homogeny, because, over and above general resem- 
blances such as might be referred to inheritance from a 
common ancestor, they exhibited a number of detailed 
resemblances such as could not possibly be supposed to have 
been represented, in like detail, in a generalised ancestral 
form. Therefore, Lankester pointed out, there must be a 
second quantity covered by the term homology, and he 
described it in the following words: ‘‘ When identical or 
nearly similar forces or environments. act on two or more 
parts of an organism which are exactly or nearly alike, the 
resulting modifications of the various parts will be exactly or 
nearly alike. Further, if, instead of similar parts in the same 
organism, we suppose the same forces to act on parts in two 
organisms, which parts are exactly or nearly alike and some- 
times homogenetic, the resulting correspondences called forth 
in the several parts of the two organisms will be nearly or 
exactly alike. . . . I propose to call this kind of agree- 
ment homoplasis or homoplasy. . . . What exactly 
is to be ascribed to homogeny and what to homoplasy in the 
relations of a series of structures is a matter for careful con- 
sideration.” Somewhat further on in the essay homoplasy is 
defined as “‘ depending on a common action of evoking causes 
or moulding environment on homogenous (= homogenetic) 


38 GILBERT C. BOURNE. 


parts, or on parts which for other reasons offer a likeness of 
material to begin with.” 

The term “ homoplasy ” has passed into current use, and 
the principle expressed by it has been freely used to explain 
numerous large and general resemblances which have obviously 
been evolved independently, such as the general resemblances 
between different kinds of patelliform gastropod shells, e.g. 
between Patella, Fissurella, Septaria, Capulus, and 
Siphonaria, or the general resemblances of external mor- 
phology of fishes and cetacea. But the term homogeny 
has not been so generally accepted, and many, if not most, 
zoologists have preferred to retain the old word homology, 
and in so doing it is clear that many of them have failed to 
distinguish between the two quantities contained within the 
single term, of which the differences were so clearly pointed 
out in Lankester’s essay. For it must be evident to anybody 
who is well acquainted with the morphological literature of 
the last thirty years that, so far from attempting to distinguish 
between homogenetic and homoplastic resemblances, a large 
number of authors have shown a vast amount of ingenuity in 
referring the most minute resemblances in the organs of 
animals, which are certainly not very closely related to one 
another, to homology. ‘The most extreme instances of this 
tendency to ascribe every resemblance, however detailed, to 
inheritance, ignoring the possibility that similar structural 
changes may be induced by the incidence of similar forces, 
are to be found in the works of those authors who attempt to 
derive the lower members of one phylum of the animal 
kingdom from highly differentiated members of another 
phylum. 

It is, of course, true that several of the most thoughtful 
and best informed among contemporary zoologists have been 
fully aware of the error lurking in the indiscriminate use of 
the term “ homology,” notably Gegenbauer and Fiirbringer in 
Germany ; Cope, W. B. Scott, E. B. Wilson, and Osborn in 
America. It is not my present intention to enter upon along 
discussion of this subject, which I hope to return to on a future 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 39 


occasion. But I take the opportunity of dealing with an 
interesting and suggestive essay by Osborn (10), in which 
Gegenbauer’s admirable analysis of the different forms of 
resemblances obtaining among animal structures is largely 
quoted. 

In the first place Osborn makes it evident that I, in common 
with others, have fallen into an error in using the term “ con- 
vergence” to denote the parallel stages of evolution among 
the Fissurellid and Scissurellidee. In the common meaning 
of the word, convergence might appropriately be used to 
signify that apparent approximation of structural character- 
istics which not infrequently leads to two forms being classified 
together in the absence of sufficiently complete information as 
to their internal anatomy. But it has acquired a special 
meaning, defined by Osborn as the “independent similar 
development of unrelated animals, bringing them apparently 
closer together.” As it has been the purpose of my paper to 
show that the families of Molluscs treated of are related, and 
closely related, the term convergence is not applicable to 
resemblances recurring in those families. But when I come 
to consider whether other resemblances between various 
mollusca should be described as due to “ parallelism” or 
“homoplasy”’ I find myself in a difficulty. Parallelism is 
defined as the ‘independent similar development of related 
animals, plants, or organs’’; homoplasy as the “‘ independent 
similar development of homologous organs or regions giving 
rise to new parts.’ It is added that homoplasy always 
involves homology, while parallelism and convergence may 
or may not involve homology. 

In Incisura the reduction of the visceral spine, the oblitera- 
tion of spiral sculpture, the levelling of the margins of 
the aperture, the alteration in the shape of the foot are 
changes parallel to those observed in the ontogeny of a 
Fissurellid, and they involve homogenetic organs; the 
parallelism in this case involves homology and should be 
called homoplasy. In Schismope the conversion of the labral 
slit into a foramen is a change parallel to that observed in the 


40 GILBERT C. BOURNE. 


ontogeny of a Fissurellid and it involves a homogenetic 
character, therefore it also is due to homoplasy. On the 
same reasoning the resemblances in the shell, foot, and mantle 
of more distantly related forms, the Patellidee, Septaria, the 
Capulidz, and Siphonariidz are homoplastic. But should the 
pallial branchize of a Patella and the gill of a Siphonaria, be 
attributed to parallelism or homoplasy? They are certainly 
not genetically derived from the typical molluscan ctenidium, 
and to this extent are deficient in the element of homology 
which Osborn says should always be associated with homology. 
On the other hand they are vascular outgrowths of the mantle, 
which is assuredly a homogenetic structure in all the forms in 
question, and therefore there is an element, though a more 
remote element, of homology. Im this case it is simply a 
question of the importance attached to the degree of homo- 
logy whether these structures should be ascribed to parallel 
or homoplastic development. But Lankester’s term, homo- 
plasy, as originally defined, covers all the cases. It appears 
to me that, while there is a contrast between homoplasy and 
convergence, there is no such contrast between homoplasy 
and parallelism, and that for the sake of clarity the last term 
should be abandoned, homoplasy being retained in the sense 
originally defined by Lankester. It has the priority over 
Firbringer’s term homomorphy, which, as Osborn points 
out, has the same connotation ; and it has the advantage of 
indicating a resemblance due to the moulding influence of 
environment, whereas homomorphy only calls attention to 
similarity of form. 

In the latter half of his essay Osborn raises a most interest- 
ing question, which has presented itself with various degrees 
of insistence to workers in various groups of the animal 
kingdom. Drawing his evidence from paleontological as 
well as recent types, he points out that the accessory cusps in 
the molar teeth of Mammalia arise in the same order and with 
the same relations to the primary cusps in groups which can 
be proved to have diverged widely from one another before 
any complication of the tooth pattern arose. Here, then, are 


INCISURA (SCISSURELLA) LYTTELTONENSIS. Al 


examples of detailed resemblances which cannot be due to 
inheritance nor yet can they be due to external forces acting 
upon homogenetic parts, for the teeth are formed below the 
gum and the cusps are in place before any mechanical forces 
are brought to bear on them. ‘The characters of the teeth 
are clearly congenital, and the resemblances between the 
patterns which have arisen independently in different groups 
cannot be accounted for by the preservation of fortuitous 
variations by natural selection, for paleontological evidence 
shows that variation has in each case proceeded along one 
line and not along several lines, one of which has been 
selected. 

Calling to mind Lankester’s suggestion of the “common 
action of evoking causes . . . on parts which for other 
reasons (than homogeny) offer a likeness of material to begin 
with,” Osborn pleads for the recognition of a latent or 
potential homology, by which term I understand him to mean 
a tendency or capacity to produce a definite structure, which 
capacity must have been present in the ancestors of the 
existing orders of Mammalia, but has only manifested itself in 
such groups as possessed or were subject to the co-operating 
factors necessary for evoking the latent capacity, and thus 
producing the structure in question. 

The objections to a principle of this kind are that, in the 
first place, as Osborn himself admits, it leads us on the 
dangerous ground of teleological speculation; and, in the 
second place, that it might, if loosely applied, be used to 
explain anything or everything by a phrase. 

Nevertheless, I think that some such principle may be 
admitted, with due caution, in explanation of a large number 
of difficulties which present themselves, with increasing 
insistence, to every class of zoological workers. In a recent 
paper on the Neritide I alluded to the great difficulty of 
finding a satisfactory theory to account for the distribution of 
the fresh-water Neritids, described as species of the sub- 
genera Paranerita and Septaria, in remote oceanic islands. 
As their general anatomical and conchological characters 


42 GILBERT (C. BOURNE. 


differ in a very small degree from those of the marine species 
of the genus Nerita, abounding in the seas in which the 
oceanic islands inhabited by the fresh-water Neritids are 
placed, it did not seem an unwarrantable assumption that in 
each locality the marine species had ascended from estuaries 
into rivers (just as prawns do in so many parts of the tropics), 
and had been similarly modified as a result of the fresh-water 
environment. But when I found that the accessory genera- 
tive organs of the fresh-water species from different localities 
were always alike, and differed in the same direction and to 
the same degree from the accessory generative organs of the 
marine species from the same localities, particularly in the 
fact that the female gonaducts of the freshwater species are 
always triaulic, whereas those of the marine species are diaulic, 
I was no longer able to sustain the opinion that I had first 
formed as to the possibility of the independent but similar 
modification of the marine species in different parts of the 
world. It seemed to me impossible that the triaulic condition 
should have been evolved several times over. The problem, 
however, is of the same kind as, though of less magnitude 
than, that presented by the cusps of mammalian molar teeth. 
If we can conceive the presence in the germ-plasm of Neritidz 
of factors competent to produce the triaulic condition of the 
genital ducts, but that the activity of these factors is only 
excited by the co-operating action of other factors—in this 
case by reduction of the salinity of the water—the detailed 
resemblances between structures existing in animals living 
so far apart but under similar conditions are susceptible of 
explanation. 

A few years ago such an explanation would have been 
inadmissible. But since Mendelian experiments have shown 
that definite changes affecting parts of the organism in a 
similar manner may require the co-operation of two or more 
factors, and cannot be produced unless those factors are 
brought together; and since such experiments as those of 
Stockard on Fundulus have shown that a relatively slight 
change in the salts dissolved in water may induce profound 


INCISURA (SCISSURELLA) LYTTHLTONENSIS. 43 


changes in certain organs of developing embryos, it is no 
longer possible to reject such suppositions as fanciful and 
incredible. 

Those who have given unprejudiced consideration to the 
objections raised against the all-sufficiency of natural selec- 
tion, must have felt that a term is wanting somewhere in the 
current forms of argument used to explain resemblances 
between structures which are only doubtfully homogenetic. 
The missing term may possibly be found when we have a 
more exact knowledge of the kinds of factors whose co-opera- 
tion is necessary to produce specific structure. Some of these 
factors must be germinal, but evidence is accumulating that 
germinal factors are not simple but compound, and may be 
split into subordinate factors which, taken alone, do not pro- 
duce the specific result. There is further evidence that 
germinal factors react differently to different external factors, 
and if this be so many kinds of resemblances and differences 
may be accounted for by laws of interaction of which we are 
as yet only dimly aware. 

The evidence on these matters is insufficient to enable us to 
arrive at definite conclusions, but it is at any rate sufficient 
to earn respect for a suggestion supported by such a large 
number of positive facts as that of Osborn. 

I believe that in the future morphologists, in conjunction 
with systematists, will be largely occupied in attempting to 
discriminate between the different kinds of resemblances 
among animal structures, between similarities due to the 
“ common action of evoking action or moulding environment,” 
and similarities due to direct descent, and I venture to think 
that such morphological studies, carried out with scrupulous 
attention to detail, are not useless, but will give precision to, 
and perhaps modify our views on, the causation of modifica- 
tion of animal structure. 


Ae 


GILBERT C., BOURNE. 


List oF tHE PrincipAL PAPERS REFERRED TO IN THE TEXT. 


i 


2. 


3. 


13. 


14. 


15. 


16. 


iv 
18. 


Bourne, G. C.—* Contributions to the Morphology of the Neri- 
tana,” ‘ Proc. Zool. Soc. Lond.,’ 1908. 

Boutan, L.—* Recherches sur la Fissurelle,” ‘ Arch. de Zool. Expér. 
et gén.,’ (2), iii bis, 1885. 

Bouvier, E. L.—Systéme nerveux, morphologie générale et 
classification des Gastéropodes Prosobranches,” ‘Ann. des Sci. 
Nat. (Zool.),’ (7), iii, 1887. 

and Fischer, H.—* Etude monographique des Pleurotomaires 

actuels,” ‘ Arch. de Zool. Expér. et gén.,’ (3), vi, 1898. 


. Fischer, H.— Manuel de Conchyliologie,’ Paris, 1887. 
- Fleure, H. J., and Gettings, M. M.—** Notes on Common Species of 


Trochus,” ‘ Quart. Journ. Mier. Sci.,’ 51, 1907. 


- Haller, B.—“Beitriige zur Kenntniss der Niere der Prosobranchier,” 


‘Morph. Jahrb.,’ xi, 1886. 


. Hedley, C.—* Additions to the Marine Molluscan Fauna of New 


Zealand,” ‘ Records of the Australian Museum,’ v, 1904. 


. Lankester, E. R.—‘ On the Use of the term Homology in Modern 


Zoology, and the distinction between Homogenetic and Homo- 
plastic Agreements,” ‘ Ann. Mag. Nat. Hist.’ (4), vi, 1870, p. 34. 


. Osborn, H. F.—* Homoplasy as a Law of Latent or Potential 


Homology,” ‘ American Naturalist,’ xxxvi, 1902, p. 259. 


. Pelseneer, P.—* Les reins, les glandes génitales et leurs conduits 


dans les Mollusques,” ‘ Zool. Anzeiger,’ xix, 1896. 


“Recherches morphologiques et phylogénétiques sur les 
Mollusques» Archaiques,” ‘Mém. couronnés de Acad. Roy. de 
Belgique,’ lvii, 1898-99. 

“The Mollusca,” Lankester’s ‘Treatise on Zoology,’ part v, 
London, 1906. 

Perrier, R.—‘* Recherches sur l’anatomie et l’histologie du rein des 
Gastéropodes Prosobranches,” ‘Ann. des Sci. Nat. (Zool.),’ (7), 
viii, 1889. 

Randles, W. B.—* Observations on the Anatomy and Affinities of 
the Trochide,” ‘Quart. Journ. Mier. Sci.,’ 48, 1904. 

Smith, E. A—** A Description of some Shells from New Zealand,” 
‘Proc. Malacol. Soe.,’ i, 1894. 

Thiele, J., in Tréschel’s ‘ Gebiss der Schnecken,’ vol. ii. Berlin, 1891. 

Vayssiére, A—‘ Etude zoologique de la Scissurella costata, 
var. levigata,” ‘ Journ. de Conchyliologie,’ (5), xxxiv, 1894, 


INCISURA (SCISSURELLA) LYTTELTONENSIS. 45 


19. Woodward, M. F.—“The Anatomy of Pleurotomaria Beyrichii,” 
‘Quart. Journ. Mier, Sci.,’ 44, 1901. 


EXPLANATION OF PLATES 1—5, 


Illustrating Mr. Gilbert C. Bourne’s paper “On the Anatomy 
and Systematic Position of Incisura (Scissurella) 
lytteltonensis.” 


LETTERING FOR ALL THE FIGURES. 


An, Anus. a.g.f. Anterior gill-filaments. aw. 1. Left auricle. au.7. 
Right auricle. B, Buccal ganglia. 0b. b. buccal bulb. 6. ¢. Buccal 
cavity, br. x. Branchial nerve. b. sk. Branchial skeleton. b.v. Blood- 
vessel. C. Cerebral ganglia. car. buccal cartilage. c. ce. cerebral com- 
missure. cl. Cilia. cdl. c. Ciliated canal of dorsal ridge of gill-filament. 
eil.l. Lateral ciliated cells. c.pd. Cerebro-pedal connective. c. pl. Cerebro- 
pleural connective. di.1. Left dialyneurous connection. d.g.7. Dorsal 
glandular ridge of gill-filament. e. Eye. ep. Epipodium. ep. n. Epipodial 
nerve. ep.t. Epipodial tentacle. F. Foot. f.c. Frontal cilia. g. Ganglion 
behind mantle-slit. g.f. Gill-filaments. g.n. Genital nerve. hy.g.l. Left 
hypobranchial gland. hy.g.r. Right hypobranchial gland. 7. Intestine. 
z.v. Vertical loop of intestine. j. Jabot. jw. Jaws. k.l. Left kidney. 
k.r. Right kidney. J. b. Left branchial ganglion, lb. 1. Labial lobe. 
l.c. Labial commissure. J. c¢.m. Left columellar muscle. J. ct. Left 
etenidium. Jl. Liver. Ui. d. Liver-ducts. 1. @. p. Left cesophageal 
pouch. /é. Lateral tooth of radula. m. Mouth. m. c. Mantle-cavity. 
md. t. Medio-lateral teeth of radula. m. f. Median dorsal fold of buccal 
cavity. m.s. Mantle-slit. mt. Mantle. mn. Nucleus. od. Oviduct. 
od.o. Opening of oviduct into right kidney. o@. Csophagus. os. 
Osphradium. of. Otocysts. ov. Ovary. P.Pedalganglia. pa. Pallial 
nerve. pc. Pericardium. p.g.f. Posterior gill-filaments. p. gl. Pedal 
gland. phg. Phagocytic cells of left kidney. pl. 1. Left pleural ganglion. 
pl. r. Right pleural ganglion. p.n. Pedal nerves. p.v. Pallial blood- 
vessels. Rk. Rectum. r.b. Right branchial ganglion. +. c. m. Right 
columellar muscle. +. ct. Right ctenidium. rd, Radula sac. 1. @. p. 
Right cesophageal pouch. rp. d. Reno-pericardial duct. sb. 7. Sub- 
intestinal ganglion. sg.! Anterior salivary glands.  sg.2 Posterior 
salivary glands. sn. Snout. s. 0. ¢. Sub-ocular tentacle. sp.z. Supra- 
intestinal ganglion. sp. 1. Left symmetrical pallial nerve. sp. . Supra- 
intestinal nerve. sp. 7. Right symmetrical pallial nerve. st. Stomach. 


46 GILBERT C. BOURNE. 


‘sy. p. Sensory papille. TT. Cephalic tentacle. tn. Tentacular nerve. 
un. Uncini. V. Ventricle of heart. v. aff. Afferent branchial vessel. 
v. eff. Efferent branchial vessel. v. g. 1. Left visceral ganglion. v. g. 7. 
Right visceral ganglion. v.. Visceral nerve. 


[All the figures are of Incisura lytteltonensis.} 


Fig. 1.—A specimen viewed from the left side and below to show the 
size and shape of the foot, the operculum, the epipodial tentacles, ete. 
x about 40. 

Fig. 2.—Dorsal view of a female specimen which has been stained and 
mounted as a transparent object. x 80. 

Fig. 3.—A reconstruction of the alimentary tract ; semi-diagrammatie. 
x 80. 

Fig. 4.—A diagram showing the relations of the right and left kidneys, 
the heart, pericardium, ovary and oviduct. The extent of the pericardial 
cavity is indicated by a thick black line. x 80. 

Fig. 5.—The nervous system as determined by reconstructions from 
sections. x 80. 

Fig. 6.—¢. A transverse section through the posterior part of the 
head, including both eyes. x 135. 

Fig. 7.—¢. A transverse section taken just behind the mantle-slit, 
showing the bi-pectinate character of the right ctenidium. x 135. 

Fig.8.—¢. A transverse section through the anterior ends of the 
pedal ganglia. x 135. 

Fig. 9—Q. A transverse section through the hinder ends of the 
pedal ganglia. Note the position of the left kidney, k./., and its open- 
ing into the mantle-cavity ; the size and extent of the right and left 
cesophageal pouches, 7.@.p. and l.@.p.; the size and position of the 
right and left pleural ganglia, pl.r. and pl.l., and the supra-intestinal 
ganglion, sp.7.; the pedal ganglia, P., are clearly seen to be composed 
of a dorsal and a ventral moiety. x 135. 

Fig. 10.—?. A transverse section passing through the posterior end 
of the mantle-cavity, showing the large size of the left auricle, au.L.; 
the orifice of the left kidney, k.r.o.; the pedal nerves, p.n., which are 
the posterior continuations of the ventral moieties of the pedal ganglia 
shown in fig. 9. x 135. 

Fig. 11— 9. A transverse section passing through the posterior 
end of the foot. Note that the large pedal nerves shown in the previous 
figure do not extend into the hind part of the foot. x 135. 

Fig 12— 9. A transverse section taken near the terminal part of 


INCISURA (SCISSURELLA) LYTTELTONENSIS, A7 


the visceral spire, showing the opening of the oviduct, od., into the 
hinder end of the ovary, ov. x 155. 

Fig. 13.—. A section through the left kidney showing the reno- 
pericardial duct, *p.d. Note the band of ciliated cells, ezl., on the floor 
of the mantle-cavity opposite the opening of the left kidney. x 225. 


Fig. 14.— 9. A portion of a section through the left kidney showing 
the rounded phagocytic cells, phy., which have taken up solid rod- 
shaped bodies from the subjacent blood-vessel, b.v. x 1000. 


Fig. 15.—Striped muscle-fibres attaching the anterior end of the 
buccal bulb to the integument. x 1000. 


Fig. 16.—A transverse section through the anterior filaments of the 
left ctenidium. Note the osphradium, os., lymg under the branchial 
nerve, b.n. xX 535. 

Fig. 17.—A transverse section somewhat posterior to that drawn in 
fig. 16, passing through the posterior filaments of the left ctenidium. 
In this and the previous figure note, d.g.7., the dorsal glandular ridges 
of the gill-filaments. x 535. 

Fig. 18.—A transverse section through two gill-filaments of the right 
ctenidium ; cél.c., the ciliated canal traversing the dorsal glandular 
ridges of the filaments. x 1000. 

Fig. 19.—The left ctenidium stained and viewed from above as a 
transparent object. x 225. 


Fig. 20.—A portion of the radula. x 800. 


Fig. 21—Part of a transverse section passing through the anterior 
end of the buccal bulb to show, m.f., the median dorsal fold of the 
buccal cavity and, s.g.', the anterior salivary glands and their ducts. 
x 535. 

Figs. 22-26.—A series of transverse sections through the right-hand 
posterior corner of the mantle-cavity showing the relations of the ovi- 
duct and the right reno-pericardial duct to the right kidney and the 
pericardium. x 225. (These figures are drawn as seen reversed under 
the microscope.) 

Fig. 27.—A cephalic tentacle showing the two multiple rows of 
sensory papille. x 225. 

Fig. 28.—A longitudinal section through three sensory papille of a 
cephalic tentacle. n', pale nucleus of a supporting cell; n*, deeply 
stained nucleus of a sense-cell ; c7l., cilia borne at the ends of the sense- 
cells. x 1000. 


Quart. fourm Meer Sev. Vat, 55, NS FEL 


eS Vy : 4 tae 
2 a Oar 
OSE os: 4A 
EO Yo) bee yf 


TSO INOS / 
ROO aoe LY / 
ef} eed S 
st/ Cle FEBs 5. 
G C. Bourne del. 


ANATOMY OF INCISURA. 


7 
‘ 


——" ai = 


“VYNSIONI 40 AWOLVNVYV 


‘TPP euinog 9 #9) 


~ Ye A y 
Ut ») 
A (Acres cil iG 1 Z 
2 = aa in J 
Nee \Y Oem oN Ne ¥ 
Me eS ) 
eer TL ony ns a b, 
: S ® TAZIND LEME BNL, "4 J 
we ate XN Ok HONS BE f 7 
i \ ° Nae SS SS /, 34 \t xy, 
Ne st ~~ 2 
St SEN ) SEF LE 
<i g +B 5 ea 


CIBSN SS 70 MS CU? mine pang 


a a a IR St NAR aa ea iM Ais ait “Cet li 


G.C Bourne del 


A 
ae 


GR 


Qa 


ANATOMY OF 


INCISURA, 


TET! 
STFVALEIST i 


‘ 


Quart. Sourr.Micn Sci. bt. Ef NEA 4: 


am? 


ge Me 
> SES 


er No 
g SEs 
2027 < 
NS, 
WEE RRSe Sx. 


G.C.Bourne del 


ANATOMY OF INCISURA. 


G.C.Bourne del 


ANATOMY OF INCISURA. 


ey 


THE EYE OF PECTEN. 49 


The Eye of Pecten. 


By 


Ww. J. Dakin, M.Sce., 


Demonstrator and Assistant Lecturer in Zoology, University 
of Belfast. 


With Plates 6 and 7, and 2 Text-figs. 


Tue first reference to the eyes of Pecten that I have been 
able to find is that of Poli in 1795. Since that date more 
than a score of investigators have studied these small organs 
and treated in more or less greater detail the histology. 
Hach has made new discoveries, which have in very many 
cases been refuted by their immediate successors, to such an 
extent, in fact, that it was almost impossible to determine 
from the literature on the subject the truth in regard to cer- 
tain parts. One of the last and most reliable papers was that 
of Hesse, published in 1900 (84). He pointed out that some 
points were still unsolved (though adding one or two dis- 
coveries himself), and that the success of the methylene-blue 
method, if attained, would possibly elucidate all. 

In 1904 a paper appeared by Miss Hyde (89), embodying 
the results of a successful employment (according to the 
author) of the methylene-blue methods for nerve-endings in 
the retina, but these results were certainly not those expected 
by Hesse nor probably by other authors, for they stand in 
striking opposition to the views previously held. Whilst 
working at a memoir on Pecten in 1907, I came to the con- 
clusion that this, the latest investigation of the Pecten eye, 
differed greatly from the preceding ones, and that only one 

VOL. 55, PART 1.—NEW SERIES. A 


50 W. J. DAKIN. 


more confusing series of results had been added to the already 
existing multiplicity. 

I determined therefore to make a complete study of the 
histology of the eye. The privilege of occupying the British 
Association Table at Naples enabled me to carry out this 
investigation on a species previously examined by most 
writers on the eye—Pecten jacobeus—and this was com- 
pleted by a considerable stay at the Port Erin Biological 
Station. The results have been the discovery of several new 
points, the confirmation and refutation of many discoveries of 
different workers, and I hope the complete elucidation of the 
structure of the retina. It has been due to the frequency of 
occurrence of artefacts and the difficult histological work 
required for such complicated organs that the structure of 
these eyes has remained so long a puzzle. 

By the use, however, of numerous methods it has been 
possible to eliminate to the greatest extent the artefacts, and 
incidentally the trial of so many fixatives, etc., has enabled 
me to obtain practically all the appearances seen and figured 
by the various investigators. 

The account of the structure will be given at some length, 
since a comparison of the various views is necessary, and, 
with the exception of Hickson’s and Patten’s papers very 
little has appeared in English. I am indebted to the British 
Association for permission to use their table at the Zoological 
Station of Naples, and also to the staff of that well-known 
institution. My thanks are also due to Professor Herdman 
and to the Curator of the Port Erin Biological Station for the 
trouble taken in supplying me with material and apparatus 
for carrying out detailed work at the latter place, and to Pro- 
fessor Drew, of Maine, for specimens of P. tenuicostatus. 


History. 


Only the history of the references to the Pecten eye before 
and including the fundamental paper of Hensen will be given 
in this section, since the other works will be discussed more 


THE EYE OF .PECTEN. at 


fully when describing the structures involved, and it will 
avoid repetition if they be omitted here. In 1795 Poli, in his 
large work on the Mollusca (1), gave figures illustrating the 
general anatomy of Pecten, in which the eyes are depicted, 
and also a view of the mantle-edge showing more clearly the 
tentacles and eyes, but no details of structure are given what- 
ever except the external pigmented ring bounding the cornea 
and the pigment stripe on the tentacles. 

He recognised a likeness to the human eye, and as usual. 
applied some of the names given to parts of the latter, a 
feature followed by his successors, who naturally recognised 
at once the resemblance to the vertebrate eye, which is such 
a striking character of the eyes of Pecten. These organs 
were mentioned, though left practically undescribed by suc- 
ceeding naturalists. Cuvier refers to them as “globules 
verdatres,”’ and Lamarck as “tubercules oculiformes.” 

The next description is to be found in Robert Grant’s 
“Comparative Anatomy’ (2), where reference is made to the 
“smooth cornea,” the “iridescent choroidea,” and a “small 
crystalline lens.” Another English writer, Robert Garner (8), 
1837, continued the work. He states that Pecten, Spondylus, 
and Ostrea (probably Pecten Jacobus, Ostrea jacobeus 
of Poli) possess “small, brilliant, emerald-like ocelli, which, 
from their structure, having each a minute nerve, a pupil, a 
pigmentum, a striated body, and a lens, and from their situa- 
tion at the edge of the mantle, where alone such organs could 
be useful, and also placed, as in Gasteropoda, with the 
tentacles, must be organs of vision.” There are no figures 
illustrating his shortaccount. Almost simultaneously Krohn (5) 
and Grube (4) published descriptions of the eye. Grube des- 
cribed the position and number of the eyes in P. jacobeus, 
P. varius, and P. opercularis. Krohn gave a much more 
detailed account. He stated that the eye was a closed 
spherical vesicle containing two transparent bodies separated 
by a septum (he was therefore the first observer to see this 
structure). The hinder of these bodies he described as being 
of fibrous texture. Krohn was the first investigator to notice 


52 W. J. DAKIN. 


that the nerve in the eye-stalk divided into two branches, 
one of which ran up to the optic vesicle, where he lost it, 
whilst the other passed up the side and entered the vesicle, 
lying on the septum. 

Will (6) noticed the cellular structure of the lens, and 
Keferstein (12) recognised the retina in the hinder trans- 
parent body of Krohn. ‘This brings us to Hensen’s paper (18) 
published in 1865, which is the first account of the histology 
of the retina. Hensen divided this part of the eye into five 
layers : 

1. First cell layer. 

. Second cell layer. 
. Rods. 

. Tapetum. 

. Pigment layer. 

The cells of the first layer, which may be arranged in a 
single or double row, are spindle-shaped. The second layer 
is made up of cylindrical cells (the rod-cells), the third layer 
is that of the rods, and then follow two others—the tapetum 
(first demonstrated by Krohn), and the pigment layer. 

The innervation is described as follows: The proximal 
branch of the optic nerve does not bore through the optic 
vesicle below, as Keferstein had assumed, but splits into a 
number of small branches which enclose the lower part of the 
optic vesicle, and these branches of the nerve form a plexus 
in the peripheral region of the retina. Apparently Hensen 
assumed that they were connected with his second cell-layer 
(the rod-cells)—“ Der Zellenauslaufer geht so continuirlich in 
den Nerven iiber, dass man nicht sagen kann, wo der eine 
anfingt und der andere aufhort.” 

The other nerve-branch penetrates the septum, and the 
fibres become connected to the cells of the first layer. 
Hensen, it will be seen, discovered the different groups of 
cells in the retina, described the nerve innervation correctly 
(though since he did not recognise two types of cells in the 
outer layer and in the rod-cell layer, this was probably more 
accidental than otherwise), and saw the axial fibre in the 


Or & © pO 


THE. EYE OF PECTEN, 53 


rods—truly a marked advance in the knowledge of the eye- 
structure. 


TECHNIQUE. 


This investigation of the eye has been carried out by the 
study of sections (paraffin and paraffin-celloidin), by macera- 
tion preparations and by the teasing of fixed material. 

It is impossible to over-estimate the value of macerations in 
conjunction with section work, and the true shape of many 
cells could not have been determined without this method. 
For both fixation and maceration it was found that different 
reagents were necessary according to the cells to be studied. 
In the retina alone the various elements reacted very differently 
to fixatives and macerating fluids, and it was surprising to 
notice how different the preservation of the different cells 
might be after treatment with the same fixative. 

The fixatives giving the best general fixation of all parts 
were Zenker’s fluid and Carnoy’s mixture. Zenker was used 
as follows: Fixation lasted for about twelve to twenty-four 
hours, and was followed by washing first with water and then 
in alcohol of gradually rising strength. Sections were made 
after paraffin embedding, the usual thickness being that of 
the rod-cells, namely 6 u, but others were only 2 uw, and some 
were 10 » thick. ‘The stains used after Zenker were Mallory 
(connective-tissue stain), iron hematoxylin (Heidenhain), a 
modified Weigert, and picric acid—siurefuchsin. 

Mallory’s connective-tissue stain.—The sections, 
on slides, were stained in an aqueous solution of saurefuchsin, 
0:05 per cent., for ten minutes, then rinsed quickly in water 
and placed in | per cent. solution of phosphormolybdic acid 
for three to five minutes. After washing in several changes 
of water for five to ten minutes the sections were stained in 
the following solution for eight to fifteen minutes : 


Aqueous aniline blue (Griibler) ; ‘ 0°5 gr. 
OrangeG.  . : : : : : 2:0 gr. 
Oxalic acid. : ; ; ; : 2°0 gr. 


Water -. " : : : : = L00-Olei: 


54 , W. J. DAKIN. 


This was followed by a rapid washing out in water, dehydra- 
tion in 90 per cent. alcohol to absolute, and mounting after 
xylol or origanum oil in balsam. 

Iron hematoxylin.—The sections were mordanted for 
twenty-four hours in a 4 per cent. solution of iron alum 
washed in water, and stained in a 0°5 per cent. to 1 per cent. 
solution (aqueous) of hematoxylin for twelve to twenty-four 
hours. This was followed by differentiation under microscopic 
observation with 2 per cent. iron alum solution. ‘Tap-water, 
alcohol dehydration, etc., as usual. 

The modified Weigert method was only used after 
Zenker fixation. It was partly like that used by Schreiner 
(80), but modified in combination with Zenker. 

Schreiner used a 10 per cent. alcoholic solution of haema- 
toxylin (P. Mayer says that the “10 per cent.” must be a 
misprint). 

I used a 5 per cent. solution, but did not investigate the 
effects of a stronger nor of al per cent. solution, which Mayer 
believes to be the one intended by Schreiner. 

The sections (on slides) were placed ina 3 per cent. solution 
of potassium bichromate for twenty-four hours, then rinsed 
in water and alcohol, and placed in a 5 per cent. solution of 
hematoxylin (alcoholic) for a time varying from ten minutes 
toan hour. The sections must be black, and this takes place 
much quicker after the hematoxylin solution has been used 
once or twice and is oxidised by contamination with bichro- 
mate. After staining, the sections were rinsed in water and 
placed in a saturated aqueous solution of copper acetate, 
which turns them a steel-blue colour. Differentiation was 


carried out (under microscopic observation) in the following 
solution : 


Borax . ‘ : : : 2°0 gr. 
Pot. fernioyanite : : ; : 2°5 gr. 
Distilled water ‘ : i . 100°0 c.c. 


The sections were then aaahod in tap water and mounted 
in the usual way, after alcohols and xylol, in Canada-balsam. 
Picric acid—saurefuchsin (van Gieson).—The sections 


THE -EYE OF. PEOTEN. 5) 
were stained in Delafield’s hematoxylin and washed well in 
tap water. This was followed by staining for five minutes in 
a mixture of— 

1 per cent. solution (aqueous) siurefuchsin . 50 
Saturated solution of picric acid in water . 1000 
The stained sections were washed in tap-water and taken 
up to balsam as usual. 
Carnoy’s fixative was used in the following strength : 


Chloroform ; ; E ; , ; 10°0 
Acetic acid f : : : t ; 30°0 
Absolute alcohol : F , t 60:0 


This is the best fixative for the retina. Iron hematoxylin 
and Bethe’s toluidin blue were the stains used on material 
so fixed. 

Bouin’s fluid (‘ Lee,’ edit. vi., p. 76) gave excellent results 
for rod-cells and rods, especially when followed by Mallory’s 
stain. The axial fibre of the rods was stained better by the 
siurefuchsin in this method than by any other except the 
modified Weigert. Zenker’s fluid, Mann’s fluid,' and a mix- 
ture of equal parts of corrosive sublimate saturated aqueous 
solution, and Hermann’s platinum-osmic fluid were useful for 
the lens, especially the latter. 

Other fixatives used were 4 per cent. formol, corrosive 
sublimate (aqueous solution and solution in salt water), 
Mayer’s picronitric mixture (‘ Lee and Mayer,’ ed. vi, p. 68), 
Flemming, Von Rath’s picro-platinum-osmic mixture, and 
treatment with pyroligneous acid. The latter did not give 
particularly good results. There were also special fixing and 
other processes connected with the following methods—Golgi’s 
silver process (Cajal’s modification), Bielschowsky-Paton silver 
method for neurofibrille (41), Apathy’s nachvergoldung and 
hematein IA methods, Nabias’ gold method, Lists’ eosin 
method, and methylene blue processes. The latter were 
failures, though injection methods, staining in aqueous solu- 
tions, solutions in Pecten serum, and dusting powder over the 
eye were all tried. The results given by the other and more 

1 Mann, ‘ Physiological Histology,’ p. 96 (solution d). 


56 W. J. DAKIN. 


ordinary methods were more complete than by the complicated 
ones, and there was usually a far greater freedom from 
artefacts. There remains finally the maceration methods to 
be referred to, The lens-cells, with all their peculiar 
processes, were easily isolated after immersion of the eyes 
directly in a 3 per cent. solution of chloral hydrate in sea- 
water for about four hours. The same solution was used for 
the retinal cells, and the eyes were placed, as above, directly 
into this medium. After two hours the retina was dissected 
out from the eye, placed in a drop of water on a slide, and a 
cover-glass supported by wax feet placed above it. Gentle 
tapping on the cover-glass separated the elements. Chromic 
acid solutions in sea-water of =, per cent. strength gave very 
good results for macerations of the rod-cells and rods. 

This was also used as advised by Patten after fixation of 
the eyes in } per cent. chromic acid for five minutes. 

The maceration preparations were examined unstained, and 
stained with picro-carmine. 

The chief species examined have been Pecten maximus 
and P. jacobeus, with the following others: Pecten oper- 
cularis, P. varius, P. tigrinus, and P. tenuicostatus. 


Position AND NumBer or Kyzs. 

The eyes of Pecten occur on the mantle-edges of both 
valves. ‘The mantle-edge can be said to be divisible into 
three folds, the periostracal fold, the ophthalmic fold, and 
the velum (Pl. 7, fig. 2, V.). All three possess tentacles, 
those situated on the first two being long and mobile sensory 
structures, well provided with sense-cells for the perception 
of tactile and olfactory stimuli, whilst those on the velum 
are short and rather immobile. 

The eyes are situated. on the median fold, between the 
periostracal groove and the base of -the velum (PI. 7, fig. 2, 
Hye), and amongst the long tentacles. Poli in 1795 noticed 
a certain resemblance of the eye-stalks to the tentacles, and 
considered them as modifications of the latter. 

The number of eyes present varies considerably for the 


THE. EYE OF PECTEN. 57 


different species, and there is, further, considerable variation 
among’ the individual members of any species. 

Carriére (21) stated that those species with large eyes 
possessed fewer than those with small eyes; that there were 
always more on the upper mantle-lobe than on the lower; and 
that in general, large specimens had more eyes than smaller 
ones of the same species. 

This latter sentence was an important assertion, since it 
implied growth and development of new eyes during life, and 
certainly it appeared supported by the fact that large and 
small eyes exist side by side. 

Patten (22) also pointed out that there were more eyes 
present on the left valve than on the right, and that they 
were larger, but he disagreed with Carriére, stating that no 
new eyes develop after a size of 2 centimetres has been 
attained. Rawitz (25) found similarly more eyes on the left 
mantle-lobe than on the right, and agreed with Patten on 
the development. Schreiner (30) agrees also with reference 
to the number of eyes on the two mantle-lobes, but states 
that those of the right are not smaller than those of the 
left (Patten). Had Schreiner examined P. jacobexus, the 
chief species investigated by Patten, he would not have 
made this assertion. The eyes are always more numerous 
on the left mantle-lobe than on the right, as all observers 
have found. ‘The exact relations, however, vary in different 
species. The eyes are situated in three groups, on each 
mantle-fold, one group on the anterior auricular area (two 
to seven eyes close together), another on the posterior 
auricular area, close up against the hinge-line, and the third 
and largest group along the ventral margin of the mantle. 
Spaces without eyes separate these three regions. In each 
series the eyes vary considerably in size. Patten (22) asserted, 
in fact, that a regular arrangement of small and large eyes 
existed, and Rawitz (25), though denying the existence of 
Patten’s arrangement, stated that a large eye was always 
followed by a small one. I have examined all the species 
referred to by Patten and Rawitz and find no such arrange- 


58 W. J. DAKIN. 


ment. There is a quite irregular series, and a small eye may 
be followed by another small one or by two large ones, or a 
group of large eyes may exist together. The eyes on the 
left mantle-lobe exceed in number those on the right, in 
particular in species with the most inequivalve shells (as far 
as the species I have examined are concerned), that is, in 
P. maximus and P. jacobeus, and this difference in 
numbers is greatest in P. jacobeus. 

The eyes in this species are far less numerous on the right 
lobe, and are alsovery much smaller (PI. 7, fig. 2) (contra 
Schreiner). 

I believe, however, that the greater number of eyes on 
the left mantle-lobe is due primarily to the fact that this 
valve is always uppermost, and not to its shape; and if a 
Pecten is turned over on to the left valve, it very soon rights 
itself by a peculiar turning movement. Patten (22) connected 
this numerical superiority of eyes on the left valve with its 
position but was puzzled to see how this could be an advan- 
tage to the animal, since the eyes on the lower mantle-fold 
received the light direct from above, and the eyes on the 
upper one were apparently directed downwards, 

Schreiner (30) also figures them as lying pointed to the 
ground and at an angle of 45° to the valve. Ifa Pecten be 
watched as it opens the valves, it will be seen that the 
eyes of the left mantle-lobe project just outside the shell, 
and their field of view is practically as much above the 
animal as that of the eyes in the right valve. The upper 
valve is alsoa little shorter than the lower one, and lies inside 
it when the shell is closed; the mantle lining the lower valve 
is retracted accordingly to a greater extent when the shell 
is closing. ‘The valves of the almost equivalve species meet, 
however, ventrally, and the conditions appear either more 
favourable to the eyes of the right mantle-lobe than in P. 
jacobeus, or else, as will be referred to again, this form is 
an older and more specialised one, and the eyes have begun 
to degenerate in the lower valve. Some figures are appended 
which will give an idea of the number of eyes in the three 


THE EYE OF PECTEN. 59 


groups on the mantle-edge of P. opercularis and also of 
the individual variation in this species (the specimens were 
from the Irish Sea). 


Left mantle-lobe. Right mantle-lobe. 

Length of ant.- Li faved 

post. diameter. 

of shellin cm. | Total | No.on | No. on | Total No. on | No. on 

No. ant. ear. |post. ear.| No. ant. ear. |post. ear. 
| 

75 59 6 4 42, 0 3 
6°4 50 4, 6 41 0 4 
5'8 48 4 4 39 0 3 
38 37 4 5 Bil 2 5 
64 54 2 5 40 ) 4 
58 50 4 4 39 1 4, 
4-25 44 4 5 41 2 4 
51 52 3 6 42 2 4 
51 Dee || 4 5 50 2 6 
5°25 45 3 4. 39 2 3 
5:15 54 5 5 35 il 3 
515 53 5 6 48 3 3 
5°85 55 5 5 49 3 4, 
7:25 58 a 5 47 2 5 
4:10 59 4 Se ane 1 5 
50 55 4 3 50 2 4 
7A 51 2 4 38 i 5 
0:25 61 i] 7 47 4 HY 
475 55 5 6 43 il 4 
4-70 54 5 Sys oli I 4 
O38 62 6 a aig) i 4 
44, 51 6 Dl Ail 1 5 
49 57 5 5 | 43 1 4 


It will be seen from these figures that there is no relation 
between the size of the animal and the number of eyes, though 
if the first five only had been taken the reverse would have 
appeared to be the case. Possibly Carriére only examined a 
few and chanced to get an accidental series. No one appears 
to have examined the very small eyes occurring with the 
large ones. I sectioned some of those taken from the right 
mantle-lobe of Pecten jacobeus and found that they 
agreed in every respect with the large eyes of the left lobe, 
all parts being represented and in the normal positions. The 


60 W. J. DAKIN. 


only difference was in the number of cells present; they were 
apparently as large as usual but fewer in number. ‘These 
eyes, in fact, appeared to be young ones, or rather, they had 
been arrested in development and had remained with the 
small number of component cells characteristic of young eyes, 
though they were just as old as the large ones. 

In examining hundreds of eyes one meets some strange 
abnormalities, though the latter are of rather rare occurrence. 
In a specimen of P. opercularis two eyes were fused 
together so that the pupil was oval with a slight constriction 
indicating the boundary of the separate organs. Often the 
eyes appeared with very little black pigment—that is, all the 
eyes of a specimen, even the “iris” cells being almost un- 
pigmented. 

I never found any of the eyes completely covered with 
pigment as stated by Patten, nor has this feature been met 
with by any of his successors. 


GENERAL STRUCTURE OF EYE-STALK. 


The eyes are situated at the ends of short stalks (Pl. 6, 
fig. 1), which, as already pointed out, were considered by 
Poli as modified tentacles. This eye-stalk is made up of 
connective tissue, which is a direct continuation of that of the 
mantle-edge and is clothed by an epithelial layer, also a direct 
continuation of the pallial epithelium. 

The connective tissue is more homogeneous or hyaline in 
appearance than that of the tentacles, and is not broken up 
so much by crossing muscle-fibres, which, as might be ex- 
pected, are a prominent feature of the retractile tentacles. 
This homogeneous tissue extends also below the eye-stalk for 
some distance, and the transverse muscle-fibres which raise 
the velum are absent under the eyes, being arranged in 
bundles situated between these sense-organs. Large blood- 
spaces occur irregularly scattered in the stalk, communicating 
with one another and usually containing blood-corpuscles 
(Pl. 6, fig. 1, Lac.). There is, however, scarcely such a 


THE EYE OF PECTEN. 61 


defined space as a “ Hauptader des Augenstieles” to which 
these lacunz belong (Rawitz [25], p. 105). Neither do they 
always surround the nerve (Schreiner, p. 11). 

Whilst the long sensory tentacles are, in the living animal, 
continually in motion, being retracted and again extended, 
and moved from side to side, the eyes are practically motion- 
less and point fixedly in one direction only. They contract 
and may move away from a point of stimulation, this bemg 
rendered possible by means of muscle-tibres, which lie longi- 
tudinally arranged, near the epithelium (PI. 6, fig. 1, Mus.). 
The latter are narrow fibres, and are not striated, as figured 
by Patten. Striated muscles do occur, though elsewhere, in 
the mantle-edge of Pecten (45). The muscles occur on all 
sides of the eye-stalk. They terminate, according to Rawitz, 
always at the proximal end of the optic vesicle and are never 
to be found higher ([25], p. 105). Rawitz has presumably 
taken the finer muscle-fibres, which do extend up to the 
cornea, for connective-tissue fibres. Schreiner found prac- 
tically no muscles in small eyes ([30], p. 11),and states that in 
P. islandicus, where they were exceptionally well developed 
on the shell side, they could be traced to the entrance of the 
distal branch of the optic nerve. I have traced them to this 
point in P. maximus, but more delicate fibrils (Pl. 6, fig. 1, 
M.f.), staining quite differently from connective-tissue fibres, 
extend under the epithelium as far as the edge of the cornea, 
and are, moreover, present between the cornea and the lens 
(oie; fic. 1, .N.Lf.; Pl. 7, fis. 7, Lf): These are, evidently 
the “fine smooth fibres” mentioned by Patten in contra- 
distinction to his “long striated muscle-cells” of the lower 
part of the eye-stalk. These fibrils do not, however, enter 
into any connection with the epithelial cells bounding the 
cornea, and Patten’s “ciliaris ” does not exist. They will be 
referred to again wher discussing the fibres situated between 
the lens and cornea. 

Ganglion cells do not occur scattered in the connective 
tissue of the eye-stalk, a fact already noted by Patten’s 
successors, who criticised his observations on this, as on other 


62 W. J. DAKIN. 


details, somewhat severely. The epithelium covering the 
eye-stalk is a direct continuation of the pallial epithelium, 
but is modified in various regions of the eye-stalk and 
becomes a transparent cornea over the free pole. Below the 
optic vesicle the cells are small and cubical, or rather deeper 
than wide (Pl. 6, fig. 1). They contain no pigment here, 
and the nucleus is situated near the base. A distinct cuticle’ 
is present. Some little distance below the optic vesicle these 
cells increase in depth and at the same time begin to contain 
pigment. This pigment extends further down that side of 
the eye which is uppermost (see fig. 1, Pl. 6; the right-hand 
side is the shell side of the eye and also the uppermost, since 
it is an eye from the left valve). At the level of the middle 
of the optic vesicle, that is, about the plane of the septum, 
the epithelial cells have attained their greatest depth and are 

almost filled with dark pigment, occurring in the form of fine 
- granules. The external portion of the cells is usually less 
thickly crowded, and if the sections are stained to bring out 
the nuclei it will be seen that these have moved, with the 
acquisition of pigment, so that they reside near the surface 
instead of at the basal end. The statements of Rawitz and 
Schreiner in regard to the colour of this pigment in the 
different species appear to me to be of little importance, and 
in any case I can hardly confirm them. ‘The colour of the 
granules in Pecten jacobeus, P. maximus, and P. 
opercularis is dark brown, and the exact shade varies in 
any one species and according to fixation and preservation; 
moreover, the cells are completely filled in P. jacobeus, or 
at least those of the upper side of the eye-stalk. 

Another point that may be noted here is that the increase 
in height of the epithelial cells opposite the optic vesicle is 
common to all the species I have examined, though Rawitz 
states that in P. jacobeus the epithelium is everywhere the 
same in height and figures it as such ((25], p. 106). Pecten 
abyssorum possesses (Schreiner) no pigment in the cells of 
the mantle-edge or of the eye-stalk. Patten appears to be 
the only one who has noticed that there is more pigment 


THE EYE OF PECTEN. 63 


present on the upper side of the eye-stalk, and there is really 
a longitudinal band present, exactly similar (though not so 
definite) to the one on the corresponding side of the tentacles. 
The pigmented area bounding the cornea was termed the 
“iris” by Patten. Since, however, as described above, these 
pigmented cells extend far down the eye-stalk on both sides, 
it is difficult to make any division into regions or to define a 
boundary. If, moreover, the physiological action of the iris 
were considered solely to be that of a diaphragm, keeping 
out oblique rays, the name might perhaps be applied, but, as 
Rawitz pointed out, there is no proof whatever of this area 
being capable of contraction with diminution of the “ pupil,” 
and since this region is not to be homologised with the 
vertebrate structure of the same name it is better to use the 
term pigment-mantle (Pl. 6, fig. 1, P. man.) if a special 
one is necessary. Patten considered that the “ pupil” could 
be diminished to almost half its previous diameter (p. 571), 
but I have been unable to find any trace of this under natural 
conditions, nor do any other authors appear to have been 
more fortunate. The same writer states that on the shell 
side even in fully formed eyes the pigment may sometimes be 
absent so that a colourless fissure is left—termed by him the 
“choroid fissure ” (p. 578).. I have not seen this in any eye 
examined, and fail to find any references confirming the 
statement of its existence. 

The pigmented epithelial cells pass suddenly into the 
transparent cells of the cornea (PI. 6, fig. 1, Co.), through 
which is seen in the living specimen the silvery glance of the 
subretinal structures. In P. maximus the depth of the tall 
epithelial cells may decrease slightly in one or two cells, and 
then the next is much lower and completely free from 
pigment. Sometimes, however, the decrease in height takes 
place after the pigment becomes absent. 

The nuclei take up again a central position or a position 
nearer the base in the corneal cells, but there are certain 
exceptions which will be considered later. The cells are 
hexagonal in surface view and are much flatter than those of 


64 W. J. DAKIN. 


the pigment-mantle. ‘They are usually constricted in the 
middle, so that they appear hour-glass-shaped in section, an 
intercellular space being left between them (PI. 7, figs. 4 and 
10). Externally there is a very distinct striated cuticle 
(Pl. 7, fig. 10, Cut.) which forms a hexagonal plate over the 
cell, and if the cornea is carefully focussed down upon from 
above these hexagonal plates are seen with their edges in 
close contact forming a definite mosaic (Pl. 7, fig. 3). If the 
corneal cells are now brought into focus at about the level 
of the nucleus, they appear still hexagonal in section though 
rather irregular, and the cell-walls do not touch. ‘he spaces 
left between the cells on each side are crossed by numerous 
intercellular bridges (Pl. 7, fig. 4). I have no doubt 
that these are what Patten took to be interlocking processes 
of the cells. Carriére (26) was the first to discover their true 
nature, but asserted that Patten could not have seen them at 
all, since they were finer than his interlocking processes. 
Schreiner (30) stated that the intercellular spaces were filled 
with a prominent cement substance which, through shrinkage 
during fixation, caused the appearance seen by Patten, and 
does not mention any intercellular bridges whatever. Rawitz 
was also of the same opinion and does not refer to Carriére’s 
statement (Rawitz [25], p. 109). I have seen them quite 
distinctly in the pigmented cells of the pigment-mantle as 
well as in the cornea, and they have the same structure in 
both places. There is another detail to be mentioned here 
which illustrates the difficulties caused by artefacts. Patten 
stated that the corneal cells had basal processes like the 
lateral ones, but which were longer and penetrated the 
underlying connective tissue, reaching the lens. This has 
been denied by all investigators since, and J had seen no 
traces of any such structures in hundreds of sections examined. 
After using the Bielschowsky-Paton silver method, however, 
the result figured (Pl. 7, fig. 10) was obtained. The tissues 
were fixed in 4 per cent. formol and lay in | per cent. silver 
nitrate solution for three weeks, which one might say was a 
likely method for artefacts. On the other hand, the structures 


THE BYE OF PECTEN. 65 


appeared well preserved and very little contraction had taken 
place. The processes were very definite, and had I found 
them by other confirming methods [I should not have 
hesitated to describe them as actual cell processes. I have 
figured, however, the preparation, and prefer to leave the 
question of their true nature open. The type of cornea just 
described is that of Pecten jacobeus, P. maximus, and 
P. opercularis. 

Rawitz ([25],p. 108) divides the types of cornea into three 
classes: (1) Cells of cornea considerably smaller than those 
of the pigment-mantle, ex. P. flexuosus, P. glaber, and 
P. opercularis; (2) cells of cornea, smaller at periphery 
against the pigment-mantle, but rapidly increase towards the 
centre, where they equal the pigment-cells in height, ex. 
P. jacobeus and P. varius; (3) corneal cells are as high 
as cells of the pigment-mantle at periphery, but increase 
rapidly in height towards the centre, the nucleus lying near 
the base, ex. P. pusio. I hardly think it advisable to make 
such a division, since, in the first place, the appearance often 
varies with the size of the eye, and it is difficult to fix a 
boundary between the two first groups. The corneal cells of 
P. jacobzus are, moreover, not equal in height to those of 
the pigment-mantle, though they are much higher in com- 
parison with the same cells in P. maximus. There is, 
however, a well-marked division in which Pecten pusio 
and also P. tigrinus can be placed. ‘The latter is figured 
(Pl. 7, fig. 12). In this species the corneal cells are very 
different from those of P. maximus. Those next to the 
pigment-mantle are of similar size, or smaller than the 
adjoining pigment-holding cells, but towards the centre 
the cells increase in height very considerably until they are 
deeper than the pigment-cells, the height of the corneal cells 
being double that of the latter. The cell-boundaries are not 
very distinct, and intercellular bridges are not to be seen. 
I have been unable to make out any reason for the peculiar 
difference in these two forms. 

The connective tissue of the eye-stalk has already been 

VOL. 55, PART 1.—NEW SERIES. 5 


66 W. J. DAKIN. 


referred to; it is continued around the optic vesicle (PI. 6, 
fig. 1, Con.) forming the inner wall of this (the outer being 
formed by the epithelium), and finally persists much diminished 
in thickness as a thin, transparent, and practically structureless 
layer underlying the cornea and separating this from the lens 
(PI. 6, fig. 1, Co. S.). Thisis the ‘‘ pseudo-cornea ” of Patten, 
and the “ innere Pellucidaschicht” of Rawitz. Nuclei are on 
rare occasions to be seen in it, but generally it is free from 
the connective-tissue fibrils and muscle-fibrillz, which appear 
in that part just outside the corneal area, under the pigment- 
mantle (Pl. 6, fig. 1, M./.).. This more hyaline character is 
in all probability due to the fact that light rays have to pass 
through this layer before entering the optic vesicle. 


Tue Lens.—The lens (Pl. 6, fig. 1, Z.) is one of the. 


structures that gave much trouble to the early investigators, 
but has lately been considered, entirely understood, and 
passed over somewhat lightly. Hesse made out some new 
and highly interesting structures, which I have been able to 
confirm. I find, however, that the shape of the lens-cells 
has been quite misunderstood, and the cells are certainly of 
a very peculiar nature. 

The early authors could not determine the correct shape 
of the lens itself. Kefersteim believed it to be spherical ; 
Hensen was uncertain, but believed it to be bi-convex ({13], 
p- 222) ; Hickson considered it, however, as elliptical ([18], 
p. 447). 

_ The confusion was again due to artefacts. It may be taken 
as definitely proved that the lens is bi-convex. ‘The distal 
surface is, however, almost flat, whilst the proximal is very 
convex, and may appear dome-shaped. The actual degree 
of convexity depends largely on the contraction which has 
taken place in the eye during fixation, and the lens, dissected 
free from its limiting elements.in a living specimen, probably 
alters in shape considerably, since it is not of very firm con- 
sistency. The lens is suspended from the subcorneal connective 
tissue (Pl. 6, fig. 1, Co. S.), against which its lesser convex 
surface is fastened. In surface view this face is circular and 


THE -EYE OF. PECTEN. 67 


not elliptical. Its diameter is a little greater than the cornea, 
since its periphery extends under the pigment-mantle for 
a short distance (PI. 6, fig. 1). 

The space in which the lens is suspended is bounded by 
the connective-tissue wall of the optic vesicle, the subcorneal 
extension of the same, and by the septum (PI. 6, fig. 1, Sep.), 
a membrane separating the dioptric part of the eye from the 
retina. This space was regarded by Patten as a blood-space. 
Carriére (21) first saw the blood-corpuscles in this part of the 
eye, and Patten, though also finding them, was at a loss to 
account for their presence, since the retina seemed to shut off 
all communication with the blood-lacunz of the eye-stalk. 
Rawitz appears to have found a definite vessel running on the 
outer surface of the optic vesicle and entering the distal part 
of the eye ([25], p. 113). Schreiner considers these cor- 
puscles due to pathological conditions, and remarks that 
the three other authors named above considered them as 
normal ([80], p.17). This is not strictly correct, since Patten 
stated that they might be forced into the cavity artificially 
by the contraction of the connective tissue through the 
action of reagents. 

I have only found blood-corpuscles present in this space 
on extremely few occasions, and on one of these, when there 
were many, I could trace quite easily a series of spaces in 
the connective tissue, connecting up the lacune of the eye- 
stall with the lens-cavity. This may of course have been an 
abnormal condition, and the lacune may have been produced 
artificially. These corpuscles had been forced in on the 
inner side of the eye, and I find no traces of Rawitz’s blood- 
vessel on the outer side. 

The blood plays an important part in the extension of the 
tentacles, and if a small living Pecten is watched under the 
microscope, the corpuscles can be traced running rapidly 
along the cavities of the tentacles as they are extended and 
back in the reverse direction as they contract. I believe 
their presence in the eye is due to contraction, and that they 
are forced there from the lacune of the eye-stalk. 


68 W. J. DAKIN. 


There is no membrane covering the lens and helping it to 
retain its shape. Hensen and Hickson could not find such, 
but Patten described a “suspensory ligament,” and also 
stated that the lens was attached to the septum by a con- 
nective-tissue ligament (P. varius). None of Patten’s 
successors could find any suspending capsule, neither does 
the connective-tissue ligament exist. The lens may touch the 
septum (it very often appears so in sections), but this depends 
on the contraction during fixation, and usually the retina 
leaves the posterior wall of the optic vesicle and lies across 
the middle, coming naturally against the proximal end of the 
lens. Patten’s connective tissue was in all probability the 
sheath of the distal nerve-branch (PI. 6, fig. 1, Op. Ds.), 
which would be touching the lens and lying between this and 
the septum if the retina had been forced up. Patten’s 
theories of accommodation as expressed at some length on 
p- 571 I cannot confirm, and they are somewhat irrational. 
They have not been referred to at all by his successors. He 
believed that the contraction of certain muscles supposed to 
be attached to the suspensory ligament would cause a move- 
ment of the lens towards the retina. This meant an inward 
movement of the septal membrane to which the lens (accord- 
ing to Patten) was attached. ‘The elevation of the lens was 
to be brought about “by the tendency of the elastic septal 
membrane to return to its natural position, after the contrac- 
tion of its peripheral circular fibres has relaxed the tension 
upon the central portion.” 

There is, however, no suspensory ligament nor attached 
muscles, and the lens is not attached to the septum. The 
septum, moreover, cannot move forward without taking the 
whole of the retina with it, and if this was the case (rather an 
absurdity) the recipient elements would always be the same 
distance behind the lens, whether it had been elevated or 
otherwise. Accommodation will be referred to later when 
discussing Hesse’s theory. 

The lens cells had received little attention until Hesse 
described them (34). Hensen stated that the lens consisted 


THE EYE OF PECTEN. 69 


of polygonal cells with thick walls. Patten described them 
as irregular with excentric nuclei, which appear in many 
cases to have disappeared from the cells near the inner sur- 
face. Rawitz described them as polygonal and membraneless 
with small nuclei, and Schreiner terms them “pretty large” 
vesicular cells, the peripheral ones flattened, with a large 
nucleus and no cell-membrane. ‘lhe latter writer noticed 
that in sections of the lens some cells appeared to be without 
a nucleus (see Pl. 6, fig. 1), but went no further into the 
question. 

Hesse says (84) the lens ‘‘besteht wie schon lange bekannt, 
aus zahlreichen, dicht neben einander gepackten Zellen, 
deren K6rper sich an einander abplatten und bisweilen 
eigenthiimliche Formen auf den Durchschnitten zeigen.” 

Later he adds (p. 395) “. . . da man ferner aus einem 
Durchschnitt auf die Gesammtgestalt der Zellen nicht 
schliessen kann, so ist es nicht méglich hier einen Zusam- 
menhang zwischen Lage des Centralkérperchens und Gestalt 
der Zelle festzustellen.”” Hesse, however, did not adopt any 
maceration methods to solve the difficulty presented by 
sections. In sections through the lens, which is well pre- 
served in formol-fixed specimens or Hermann-sublimate, the 
cells only rarely possess a polyhedral shape, in fact it is 
only here and there that they appear sharply angular. ‘The 
cell contours are very distinct and appear rounded, so that 
there are irregular oval, pear-shaped and long band-shaped 
cells (Pl. 7, figs. 5 and 6). The size, too, varies considerably, 
and a very small, apparently non-nucleated cell may adjoin 
a large one. If, however, this small cell be followed through 
several sections, it will be found to be merely the continua- 
tion of a cell which is elongated to an extraordinary degree. 
The true shape of the cells was found after macerating the 
eye in 23-3 per cent. chloral hydrate solution in sea-water for 
four to six hours. This medium preserves admirably the 
delicate processes of the cells, and the preparation gives the 
lens-cells, separated, uncontracted, and with all details of 
structure undamaged. 


70 W. Ji DAKIN. 


The cells vary considerably in shape. ‘Those near the 
surface of the lens, particularly the proximal surface, are 
flattened and are strap-shaped (PI. 7, fig. 6, ¢.), or are con- 
stricted in the middle and have two bulging ends. The 
length may be very considerable. The common appearance 
is that depicted in fig.5 (PI. 7). The cells are pyriform, with 
the cell-body drawn out into extraordinary long tapering pro- 
cesses many times the length of the swollen part. In addition 
to this, processes are often given off very abruptly from the 
broad end. Other cells are more rectangular, yet also with 
rounded contours and the same abrupt fine processes. 
‘hese extensions are wedged between adjacent cells (PI. 7, 
fig. 5), which fit close together, and the result is a mass of 
great compactness, whose components, though having the 
most varied shape, fit together without intercellular spaces 
being left between them. 

It is often quite difficult to separate some of the cells in 
macerations. It is now quite obvious why there appears to 
be no nucleus in many cells in sections, for it may be at one 
end and the cell be so long that many sections may cut 
through the latter without touching the nucleus. 

The cells have a very distinct membrane, and it is difficult 
to imagine how this could have been missed by Rawitz and 
Schreiner, especially after Carriére. had asserted its presence. 
It is easier now to understand why there is no need of a lens- 
capsule or supporting ligament, for the soft protoplasmic cells 
are tied together by their processes and the superficial cells are 
practically converted into fibres or straps. The contour of 
the lens is, in fact, as even as if formed by a connective-tissue 
sheath or a layer of pavement epithelial cells. The cell 
contents are finely granular, with a slight trace of pigment, 
and stain intensely with eosin. ‘The nuclei are similar in 
size to those of the epithelial cells, and since the lens-cells 
are usually somewhat larger than the latter the nuclei can 
hardly be termed pretty large (Schreiner), though such 
terms are purely arbitrary. Hesse (84) discovered in the 
lens-cells of P. jacobeus, which had been fixed in sublimate 


=e too Oa 


pa 


THE. EYE OF PECTEN. 71 


and stained in Heidenhain’s iron hematoxylin, a remarkable 
structure. In addition to the nucleus there was present a 
dark staining body from which delicate but very distinct 
fibrils radiated out to the periphery and became attached to 
the cell-wall. Most of them were straight, some were bent, 
but all went out from the one point and all could be followed 
to the cell membrane if their whole length lay in the section. 
I have found the same structures (PI. 7, fig. 6, 5.), not only in 
material fixed and stained as above but also after the 
following treatment : 

After fixation in Hermann-sublimate mixture and staining 
in iron hematoxylin, the shape of the cells is well preserved, 
the contents are homogeneous or very finely granular and 
stain grey, the nucleus is black, and radiating fibrillee appear 
distinctly in many cells though not in all. After Zenker 
fixation and Mallory’s stain the cell contents are very granular 
in appearance and stained deep red, the nuclei being yellow- 
red, and there is just a slight trace of the fibrille. They are 
also to be made out, though not distinctly, after Bouin fixa- 
tion. Von Rath’s treatment caused the cell contents to 
appear very granular and vesicular (PI. 7, fig. 6, d.) the radi- 
ating fibrille were often very distinct, but the central dark 
staining body did not look exactly like the normal centrosome 
of dividing cells. 

This permanent centrosome (PI. 7, fig. 6, Cent.), if it be such, 
does not appear to have any definite position, but since it 
cannot be made out in macerations it is almost impossible to 
determine its true position, for sections cut the cells in all 
directions. In addition to the species enumerated by Hesse 
I have found these structures in P. tenuicostatus, and 
probably they are present in all species. Hesse naturally 
compared these with the centrosome and astral rays which 
appear in cells undergoing mitotic division. Such structures 
have been demonstrated as persisting in the resting stages 
of certain cells, in pigment-cells of fishes, and more particularly 
in leucocytes. It has not been possible for Hesse or myself 
to determine any connection with cell-division. ‘he astral 


72 W. J. DAKIN. 


rays are very fine and remarkably definite, ‘There are three 
explanations of these structuresthat may be given. ‘The first 
and most unlikely is that they are artificial productions ; the 
second, that they are modified astral rays and centrosome 
kept permanently for another function; the third, that they 
are entirely different from those functioning in the cell divi- 
sion, but have arisen ina similar way and are purely supporting 
fibrilla. The appearance of the structures and their presence 
after such varied treatment is against the first view. It 
would only be possible to demonstrate which of the latter 
were correct if the origin of the aster had been observed. 
I believe they are supporting fibrillee whatever be their mode 
of origin, and this is Hesse’s view, he considering they are 
for the purpose of increasing the elasticity of the cells, This 
is put forward in an interesting theory of accommodation, 
and the fibrillee are considered to form the antagonistic appa- 
ratus to another, to be referred to presently, which alters the 
shape of the lens, Between the sub-corneal connective tissue 
and the lens is a layer of peculiar fibres, first seen, though 
incorrectly described, by Patten. He made out two layers, 
a series of radiating fibres extending from the centre of the 
distal surface of the lens to the periphery, superimposed on 
a layer of strong circular fibres concentrically arranged (p. 
581). As such do no fibres exist. Rawitz saw none here 
whatever, and regarded Patten’s structures as artefacts ( [25], 
p- 113). Hesse discovered the true conditions, which I can 
confirm with some slight additional features. There is one 
layer of fibres only (Pl. 7, fig. 8), and these have a kind of 
spiral arrangement, so that towards the centre of the lens 
surface they are running at almost right angles to their 
previous course. Near the periphery they run more or less 
concentrically (Pl. 7, fig. 8). They do not terminate at the 
centre of this surface, but continue across for some distance, 
and there results a series of fibres crossing one another in all 
directions. 

In thin sections cut parallel with the plane of the cornea 
it is possible to see a number of nuclei here, with very deli- 


THE EYE OF PECTEN. 73 


cate celi-outlines enclosing them (PI. 7, figs. 7 and 9). These 
cells have their ends drawn out into the long fibres seen in 
macerations so easily, and which are many times the length 
of the cell-body (Pl. 7, fig. 2). In some cases, as Hesse 
pointed out, a number of fine parallel fibrils appear to pass 
out of and through the cells (Pl. 7, fig. 7). He regards the 
fibres as muscle-fibres, and the cell-body as containing the 
remaining myosare and nucleus. This view is based on the 
reaction to picric acid—saurefuchsin, which stains muscle 
yellow and connective-tissue red. I was not sure that they 
were not connective tissue cells, and in fact believed them to 
be such. For this reason Mallory’s connective-tissue stain 
was used as recorded on p. 53. The fibres and cells were 
stained by this process an intense red, against the blue sub- 
corneal tissue above (Pl. 7, fig. 7). They stain therefore as 
muscle-fibres. Hesse says ([84], p. 397) that these fibres 
extend to the edge of the lens but not further. 

The same fibres, however, are to be found in the connective 
tissue extending down the sides of the optic vesicle (PI. 6, 
fig. 1, W.f.) and often quite near or even on the inner surface 
of the same. I believe they have a far wider distribution 
than Hesse supposed. ‘This is the apparatus that, aided by 
the lens-cells, is (according to Hesse) concerned with accom- 
modation. Through the contraction of these fibres the outer 
surface of the lens becomes reduced in extent, the lens-cells 
are compressed together here, and, being plastic, change 
their shape, the contents swelling towards the inner surface 
where there is less tension. ‘The result is an alteration in the 
shape of the lens and hence of the focus. If the muscles are 
relaxed the elastic cells (aided by the fibrille) return to 
their previous shape and the focus is adapted for more distant 
objects. No physiological proof has yet been brought to 
support this theory, and, as far as experiments go, I could 
find no evidence of accommodation (see p. 102). 

Hesse has built up his theory simply to account for the 
fibres on the lens and the persistent astral rays in the cells. 
The function of the latter may be simply to give greater 


74, W. J: DAKIN. 


rigidity to the lens, and if the former were accommodation 
muscles one would expect a more definite and efficient 
arrangement. ‘I'he same red-staining fibres can be traced, 
however, down the sides of the optic vesicle in the connec- 
tive tissue, and those present between the lens and cornea 
may be simply for the purpose of tying the lens to the sub- 
corneal layer. Before leaving the lens it will be advisable to 
refer to another condition seen in some of the lens-cells. 
This is a peculiar condition of the nucleus (perhaps patho- 
logical) observed in one or two cells in preparations fixed in 
von Rath’s fluid and also in Hermann-sublimate mixture 
(preparations stained with Heidenhain’s iron hematoxylin). 
The latter specimen was an eye from a small P. opercularis 
or P. varius. The nucleus (PI. 7, fig. 6, a, wuc.) is perfectly 
spherical and much larger than the normal ones. The size of 
the normal nuclei was 5°3 wby 4 (they are oval in shape), 
whereas the spherical ones attained a diameter of 10°6 p. 
These nuclei were homogeneous, not staining deep black as 
the normal ones, but rather grey, slightly darker than the 
cytoplasm. <A very delicate nuclear membrane appeared to 
be present with the remains of deeply stained chromatin 
substance attached to it. The cells containing these nuclei 
do not look distorted nor vacuolated by fixatives and the 
nucleus appears perfectly natural; no other stages could be 
found connecting these with the normal nuclei. 


THe REetTINA. 


The retina, being the recipient region of the eye, is of great 
interest, and this is increased by the wonderful complexity 
for an invertebrate and by the numerous conflicting views 
that have been published as to its histological structure. 

IT agree with Rawitz when he said that to Patten must be 
given the credit of solving much of this structure. He was 
the first to reduce chaos to order, and though he was unfor- 
tunately. carried a little too far by his imagination, he 
published a very creditable work, especially since very little 


= a 


THE .KYE OF, PECTEN. 75 


was known previously about this part of the eye. I believe, 
also, that most of Patten’s good work was due to the great 
use of maceration preparations, though perhaps owing to the 
more primitive methods of section work he did not check his 
results as much as he possibly could by this means. It is a 
great pity, therefore, that lis description should have been 
couched in terms which, accentuated by his theories, did 
much to bring the whole paper into some disrepute. 

The retina covers almost exactly half of the interior of the 
optic vesicle, and since it is of considerable thickness com- 
pared with the size of the eye there is not much space left 
in the proximal hemisphere. ‘The retina and underlying 
layers will be considered together. ‘hey are separated from 
that part of the eye previously considered by a membrane, 
the septum, first discovered by Krohn (5). 

This septum is a homogeneous sheet of connective tissue 
which is slightly thicker in the middle than at the sides, and 
at the periphery it appears to become continuous with the 
inner wall of the proximal half of the optic vesicle, that part 
termed the “sclerotica” by Patten (PI. 6, fig. 1, Sc.). This 
author described it as cellular, but no traces of cells or nuclei 
are to be seen, though the corresponding structure in the 
eye of Spondylus is formed of distinet cells. Patten also 
stated that it was double. This has not been alluded to by 
other observers, but I thought I had detected this double 
nature (44). I have since found out my error and I believe 
also the cause of Patten’s mistake. He writes that the distal 
branch of the optic nerve, which lies across the septum, has 
no sheath, since the latter terminates where the nerve enters 
the optic vesicle. The nerve, however, has a distinct sheath, 
and this accompanies it to the middle of the retinal surface, 
where just as the nerve branches (PI. 7, fig. 18) and spreads 
out over the centre, the nerve-sheath spreads out too, covering 
all the diverging nerve-fibres which lie therefore between 
two sheets of connective tissue, the nerve-sheath above and 
the septum below (Pl. 6, fig. 1). This nerve-sheath fuses 
with the septum, and I think the two sheets of tissue were 


76 W. J. DAKIN. 


regarded by Patten as both belonging to the septum. In 
preparations stained by Mallory’s method the blue connective 
tissue is brought out very distinctly against the retina, whose 
elements are stained red, and hence both septum and nerve- 
sheath can be easily followed. In some sections there appears 
to be a delicate concentric striation in the septum, but this is 
all the structure to be made out. The distal branch of the 
optic nerve penetrates the septum, the fibres boring through 
separately, 

The retina has been divided into several layers by previous 
writers, but anatomically as well as for purposes of description 
it will be better to consider it as made up of two layers only: 

(1) he outer layer of distal sense-cells with their inter- 
stitial supporting cells (Pl. 6, fig. 1, D. S.; Pl. 7, fig. 13, O. I.¢.). 

(2) The inner layer of rod-cells and their continuations 
the rods, together with interstitial supporting cells (Pl. 6, 
Deel Pl aie, 138 Rh, CO. and Tete 

A table is appended (p. 77) giving the synonyms that have 
been used, which shows also the gradual changes that have 
taken place in our knowledge of these structures. 

Hensen (18), as will be seen from the table, placed all the 
cells present in the retina distally to the rod-cells and rods 
in one category, called this stratum the “ first cell layer,” and 
said it was composed of one or two layers of spindle-formed 
cells, whose contours were not very distinct. The layer of rod- 
cells was called the “second cell layer” and the nuclei of 
the inner interstitial cells considered to be their nuclei. 

Patten found that the outer cells of Hensen were not all of 
the same shape. He supposes, however, that physiologically 
they are alike and calls them all outer ganglionic cells. Of 
these he described three types, one of which had broad ends 
bearing many fibrous processes which penetrated the septal 
membrane and became continuous with the nerve-fibres of 
the distal branch of the optic nerve. 

One of his most important discoveries was the finding of 
the interstitial cells of the rod-cell layer, which he termed 
“inner ganglionic cells” (Pl. 7, fig. 13, I. [.c.). Only the 


ri 


OF PECTEN. 


HYE 


THE 


‘(4x0} 90s) ,, UBICUTOUTQaTY ,, ot[} 07 yueTeaAmnbe ATJOVXe You st amnyonays posoddns sty, 


‘O[OISOA 
odo fo [Va 
PeYIpour 19ynO 


“ToAVT YWOULST 
‘uangede y, 
‘ouTRAG, 
“MoU JUITIIOSe 


“XLIQVU-poy 
"Spory 


‘ST[90-poy 
‘s]T]90 OUT 
-yaoddns ye1q14s 
-19JUL TOUUT 
‘s]T]09 Sut 
-yaoddns [ery144s 
~19YUT [VULI4 XG 
"sT]O0 
-9stes [VISICL 


PPT OSPUOULS Ig 
tingode yz, 


| URIG WOU YOO 
ZURYS 
~({SUOTOSTM 7, 
pee gets 


UBC ULIUIG alg 


Us[[PEZUIYOqvyg 


UWo[[OZWOPSTM 7, 


WoT[OZ 
“MOTPOSTMZ, 


UsT[eZ Vfeysiq. 


QOUpOSPUIULST 
unjode J, 
UBIG THOU 
-(aIQ otouuy 


suaTpOqryig 


Sep [eyURyL 
waToqRyAS 


WRAL ULOTI aT 
o.lossny 


UdT[OZTETIGLIG 
FIP IPOSTMOT OZ 


-“MOT[OURS IOUUT 


OO 
“WOT [OZUITPOUL | 
e.10ssn Vy 


quVyUSULST 
wngjodey, 


STTTOGRIG 
sep JoyUL] 
WARIS 
Uo[[eZTayoq VAG 
pun weyoqryg 
WOA OZOL4) 


UeT[ezZMaTOqVyG 


Ud] [eZMET[Suvs 
alrpundoeg 


AIST [OZ 
-UWoT[ ours) 


BOTJO.AOPOS 


tangede J, 
RoPUISLY 
YOM 
-JoU SNOBT4I A 
spot 
Jo qavd 109nO 


BRO 


, OUR 
“WOU [VUTULLO,T, 


‘e1oydourjeyy 
LOAVT 


OMMOTLSUVS LOUUT 


Tak] 
ONLOTLSUBS 1oyNO 


QIPTIPOSPUOUAST | 
twingodey, 


weToquyG weyo 
-SIMZ ZURISQNG 
TWaYOqRyG 


UBL WLITIQelg 


Ud[[EZMayoqVyg 


Ue][eZ 
twesturosjepurds 


OP PPPS 


TIN} B48 
qUOULST 
wngjoede yz, 


weToqRyIG 


qypryos 
“U9T[OZ O41OM7, 


gyros 
-“UdT[OZ, 94SAGL 


‘redud yuosaig 


“006T ‘ossoH 


“9681 “LOTLAIYOS 


“S887 ‘Z9IM BY 


“OSS ‘U9}4Bq 


“GgsL ‘oLQTIIByD) 


*cggT ‘mesuay, 


‘BULJOY 9} fo suo1ydraosoq Ur posn sudoy, 


78 W. J. DAKIN. 


nuclei of these cells had been seen before and they were 
thought to lie inside the rod-cells. 

Rawitz agreed with Patten in almost all respects, but made 
a retrograde step in asserting that a division of the outer 
cells into three types was unnecessary because “die gesamten 
Zellen dieser Schicht vollstiindig einander gleichen, abgesehen 
natiirlich von den nebensichlichen Differenzen im iusseren 
Habitus, und weil sie, vielfach miteinander in direkter Kom- 
munikation stehend, eine physiologische EHinheit repriasen- 
tieren.” Schreiner also refers to the two layers of ganglionic 
cells (the outer being a mixed layer, see table, and the inner 
one the non-nervous inner interstitial cells), and states that 
the outer layer is four or five cells deep in the middle of the 
retina. He noticed, however, that the cells of the outermost 
row (Patten’s first type) differed from the others, though 
considers that all are of the same physiological nature. 
Hesse in 1901 (84) was the first to upset the prevalent ideas 
of these cells. He stated that there was only a single layer 
of cells, and that the fibres of the distal nerve were not con- 
nected with them. Hesse had forgotten, however, that the 
previous observers would also have considered the outer 
ganglionic layer to be of but one layer of cells if they had 
only meant it to include the cells of Patten’s first type. ‘lhe 
other cells Hesse alludes to as being pushed in between those 
of the outer row, which he states are of epithelial-like nature. 
In any case, to Hesse belongs the credit of having separated 
off the outer interstitial cells from those of the most distal 
layer, and breaking up the idea that all were ganglion-cells 
and alike in function. 

In addition to the difference in shape and the fact that the 
outer cells bear cilia-like processes (PI. 7, fig. 15, D. S.), he 
also noticed that the nuclei of the outer cells were somewhat 
different from those of the others, now termed “ Zwisclien- 
zellen.”’? This difference has often been very apparent to me, 
and it is strange that the earlier writers missed this point 
unless fixation and staining of these cells had been rather 
indifferent. 


THE EYE OF. PECYEN. 79 


Hesse finally noticed the resemblance of these outer inter- 
stitial cells to the inner ganglionic cells of Patten, Rawitz and 
Schreiner, and called all of them “ Zwischenzellen,” stating 
at the same time that they did not bear exactly the appear- 
ance of nerve-cells, but his preparations showed that the 
fibres of the distal nerve arose from them. He did not regard 
them as ganglion-cells but considered them to be optic sense- 
cells. The function of the outer cells is not stated, but they 
are not supposed to be connected with the distal branch of 
the optic nerve. 

The next mention of these cells occurs in Schneider’s 
‘'Text-book of Histology’ (88). Schneider finds no connec- 
tion existing between the ‘ Zwischenzellen” and the distal 
branch of the optic nerve, nor any junction of the latter with 
the outer layer of cells, but finds that the nerve-fibres pene- 
trate between them and cannot be traced further. He also 
describes how at the edges of the retina the cells of the outer 
layer at various places surround, collar-like, branches of the 
nerve. I believe this (see his illustrations, p. 560) must have 
been caused by artefacts. The interstitial cells are not con- 
sidered to be sense-cells. 

In 1904 appeared Hyde’s remarkable account of the nerve- 
endings in the retina, which really caused my attention to be 
drawn to the Pecten eye. Hesse had previously stated that 
methylene-blue methods had failed him, but that the problems 
of the retina would in all probability be solved by the attain- 
ment of success with this stain. According to Hyde, methy- 
lene-blue methods were perfectly successful and solved all, 
the result being a description of the retina which stands in 
striking opposition to all previous work. Hyde finds that the 
inner interstitial cells are the nerve-cells connected with 
the axial fibre of the rods, and only mentions one row 
of outer cells which are supposed to be connected to the 
fibres of the optic nerve. 

So much for the outer cells ; I shall have occasion to make 
further reference to Hyde’s work later. In 1908 Hesse refers 
again to the Pecten eye (48), and now finds a connection 


80 W. J. DAKIN. 


existing between the distal cells and the distal branch of the 
optic nerve, so that these are also included as sense-cells, but 
his views of the interstitial cells remain unaltered. He had 
apparently neither seen nor heard of Hyde’s paper, which has 
remained, therefore, uncriticised. Such is the mass of con- 
flicting evidence at present existing. ‘There is no doubt that 
the relation of the distal nerve to the outer distal cells and 
interstitial cells is the most difficult histological problem of 
the retina. It is extremely difficult to trace the endings of 
the nerve-fibres in sections, and impossible to make out the 
shape of the interstitial cells. I have been able to make out, 
however, the shape of the latter from macerations, and to 
trace the extent of their branches, which can be confirmed by 
sections. A schematic figure has been built up from macera- 
tions and sections which shows the relation of the cells to one 
another (Pl. 7, fig. 13). 

The structures are as follows: The distal surface of the 
retina is bounded by a single layer of cells (Pl. 7, fig. 15, 
D.S.), the distal cells of Hesse, and the first type of Patten’s 
outer ganglionic cells. They are somewhat regularly placed 
so that an epithelial-like layer is formed. The outer ends of 
these cells, which are directed towards the septum, are broad 
and bear cilia-like processes, so that a space exists between 
septum and cell-layer, which is crossed by the nerve-fibres 
from the distal nerve and filled by the processes of the distal 
cells, which for the most part do not reach the septum (this 
may be caused, however, by breakage of the fine processes 
during fixation). The cells are cylindrical, transverse sections 
cut in the plane of the retina, being perfectly circular (Pl. 7, 
fig. 16, D.S.). Their lower ends are rounded, and in some 
cases appear to terminate in a short pointed process. This, 
however, could not be followed far, and I have only seen it in 
some maceration preparations. 

The cell contents are finely granular. Dark-staining 
granules (basal granules) are present at the bases of the cilia- 
like processes (Pl. 7, fig. 13), and these sometimes produce 
the appearance of a dark-staining edge. There are also 


THE EYE OF PECTEN. 81 


delicate longitudinal fibrillee in the protoplasm of the distal 
ends of the cells, running to the bases of the processes 
(Pl. 7, fig. 13, D.S.). Like Hesse I have found no motion 
of the processes in living cells. Between the cells pass 
branches of the distal nerve, which can be traced quite easily 
through the septum, but with great difficulty in the retina, 
where it has been uncertain whether they entered into con- 
nection with the outer cells, interstitial cells, or ended free. 

I think it is certain that they terminate, however, in the 
distal cell-layer and become connected with the cells, not by 
the cilia-like processes, but to their sides (Pl. 7, fig. 13). It 
is easy to see in Sections the nerve-fibre passing to the side or 
apparently one corner of the distal cell, and in macerations 
each distal cell can be seen to possess a long, thicker process 
which appears to arise at the edge of the distal end, but can 
often be traced some distance down the side wall. This is 
unfortunately very difficult to make out, but is confirmed, I 
think, by the character of the distal cells, which are those of 
sense-cells, and by sections of young eyes, where the inter- 
stitial cells are only slightly or not at all developed (as 
noticed by Hesse). 

The nucleus requires special consideration since it differs 
from that of the interstitial cells. Fig. 16 (Pl. 7) illustrates 
a transverse section through distal and interstitial cells 
stained with Mallory. The nucleus of the first-named 
(fig. 16, D.S.n.) is large, perfectly round, and contains a 
number of small chromatin granules, which stain orange red 
(orange G. and siéurefuchsin) in addition to the distinct nucle- 
olus which is always present and stains more distinctly 
orange (there may be two nucleoli present). The cytoplasm 
is stained red. hese nuclei are very similar in appearance 
to those of the rod-cells to be considered below and to the 
nuclei of nerve-cells from the various ganglia. The character 
of the outer interstitial cells (fig. 13, O.J.c.) is very 
different, and I have termed them “supporting cells.” They 
bear no resemblance to sense- or nerve-cells, and no connec- 
tion between them and the inner interstitial cells or the fibres 

VOL. 55, PART 1.—NEW SERIES. 6 


82 W. J. DAKIN. 


of the distal nerve could be found. The isolated cells 
obtained by macerating the retina in chloral hydrate solution 
are illustrated in fig. 15 (Pl. 7), but these were only obtained 
on a few occasions and after a long search, for it is most 
difficult to separate them from the distal cells. 

The cell-body is very small and there is but little cyto- 
plasm left surrounding the nucleus, but from this extend 
long branched processes. The nucleus retains the blue stain 
after Mallory when it has been taken from the nuclei of the 
distal sense-cells, and generally it may be said that the inter- 
stitial cell-nuclei stain darker and are more homogeneous, it 
being more difficult to resolve the granules, They are further- 
more flattened and are only about half the size of the sense- 
cell nuclei. The processes lie in close contact with the distal 
sense-cells, there being often two clasping them and extending 
between them towards the septum (PI. 7, fig. 15, a.). 

From the proximal end of the cell may arise one or more 
irregular processes which branch and penetrate some distance 
between the rod-cells. It is quite easy to understand how 
these long processes, which in my opinion tie and support 
the sense-cells, have been for a long time considered as 
nerve-endings, either of nerve-cells or of the fibres from the 
distal nerve. In many ways the interstitial cells resemble 
in shape and staining the neuroglia cells found clasping the 
nerve-cells in the various ganglia of Pecten and other lamelli- 
branchs. The outer ganglionic layer of Patten is composed, 
therefore, of two types of cells—sensory cells forming an 
outer layer and connected with the distal nerve, and support- 
ing non-sensory cells interpolated between them. Miss Hyde 
did not recognise the latter at all. I have had no success 
with methylene-blue methods, but I do not think they would 
be of much advantage unless fixation was very good (a thing 
not by any means easy to attain with many special methods), 
for it would be almost impossible to check the results and to 
determine whether, in the confusing mass of fibres, nervous 
or both these and non-nervous processes had taken the stain, 

We have now to consider the second sensory part of this 


———— 


THE EYE OF PECTEN. 83 


remarkable retina, innervated by the proximal branch of the 
optic nerve. ‘This region is most obvious in sections and is 
composed of a row of pillar-like rod-cells, bearing rods, with 
a series of interstitial cells lying between the former and 
once supposed to be their nuclei. 

The rod-cells (retinophore of Patten) (Pl. 7, fig, 18, 
RR. &.) occupy a very large part of the retina. In very young 
eyes, however, the distal cells are more prominent and occupy 
a proportionately much larger part. They are extremely long 
cells, especially those situated in the centre of the retina. 
The outer ends, to be found at the periphery of the retina, 
are attenuated and pass gradually into the nerve-fibres of the 
proximal branch of the optic nerve (PI. 6, fig. 1, Op. P.”), so 
that it is impossible to say where one ends and the other 
begins. From this point they increase in thickness, the first 
third of their length or more lying almost horizontally under 
the outer layers of cells, embraced by the processes of the 
supporting cells. Some little distance from the periphery, 
not very different for cells from different parts of the retina, 
each swells rather suddenly round its nucleus (Pl. 7, fig. 13, 
R. C..), and from this point the thickness remains practically 
the same to the basal end, though there is a slightly more 
constricted part below the nucleus. All the rod-cell nuclei 
are situated in a scattered cluster not far from the edge of 
the retina, so that the nucleus is nearer the proximal end in 
rod-cells belonging to the centre of the retina, whilst in the 
middle or shghtly nearer the base of rod-cells from the 
peripheral regions. 

The distal cylindrical portions of the rod-cells lie parallel 
with one another, perpendicular to the plane of the retina, 
and terminate at the same level, forming a well-defined line 
between them and the layer of rods. This line (PI. 7, fig. 13, 
S.m.) has been described as the section of a membrane (see 
table), which extended across the retina and was pierced by 
the rods (Pl. 7, fig. 15, Rod). These are direct continuations 
of the rod-cells, and rod and rod-cell form together one 
entity—the product of one cell. Patten described a delicate 


84, W. J. DAKIN. 


membrane supposed to separate the protoplasm of these two 
parts, but there is no trace of one, and the cell contents of 
both are continuous. 

The appearance of two definite structures separated by 
a membrane is due to an external flange or projection existing 
on the wall of the rod-cells at their junction with the rods, by 
means of which adjoining rod-cells are connected. This pro- 
duces in sections the effect of a “sieve-membrane ” with 
circular holes through which the rod-cells and rods pro- 
trude. 

It is a rather difficult point to decide. Hensen, so far back 
as 1865, said that by reason of the rod-cells ending at the 
same level a sharp bounding line was formed, which could 
easily be mistaken for a membrane, but this was not present. 
Patten did not see it either, but, as stated above, believed 
there was a delicate membrane, the ‘‘ terminal membrane,” 
in each rod-cell. Rawitz found no membrane either inside 
or external to the cells, but Schreiner and Carriére both 
affirmed its presence. Hesse (34) refers to a sieve-membrane, 
and on p. 409 he remarks that in some specimens of P. 
jacobeus and P. maximus the inner interstitial cells 
can be followed up to the sieve-membrane, which is possibly 
a product of these cells. 

In my opinion the sieve-membrane is, as above stated, due 
to the extended walls of the rod-cells, and has no part from 
the interstitial cells. This line is usually well marked in the 
marginal regions of the retina, where there are no rods borne 
by the rod-cells (P]. 6, fig. 1, M. ret.). Where necessary, the 
well-marked line above referred to will be called a “‘ pseudo- 
sieve-membrane” for convenience in description. In macera- 
tions of the retina, in =4, per cent. chromic acid (the pre- 
parations being stained with picro-carmine and examined 
with the oil-immersion) a series of very delicate parallel 
fibres could be seen running longitudinally on the surface of 
the rod-cells (Pl. 7, fig. 13, Cells A). It was not possible to 
follow them proximally to the nucleus. At the junction of 
rod-cell and rod they bear thickenings (PI. 7, fig. 13, S.m.), 


THE BYE OF PEOCIEN. 85 


which stain more distinctly, and it is probably these only 
which form the “flange” and the attachment of rod-cells to 
each other. ; 

The fibres are supporting fibrilla, and in preparations 
where the rods had broken off (Pl. 7, fig. 13a) the tube of 
fibrils could be distinctly seen. Where the rods remained 
attached to the rod-cells the fibres were continued below the 
thickenings, but had left the surface of the rod, enclosing 
the latter in a kind of sheath (PI. 7, fig. 13). 

Whether they lie on the rod-wall in the normal condition 
or in the interstitial substance to be presently considered 
I cannot say. Another point concerning the shape of the 
rod-cells remains to be referred to. Above the nucleus the 
rod-cell does not become gradually less in diameter, but after 
a constriction there often occurs one or more irregular swell- 
ings, which give the attenuated end of the rod-cell a more 
or less varicose appearance. 

Patten saw one of these and described it as a delicate 
oblong vesicle containing a second faintly staining and 
often invisible nucleus. Rawitz would not consider the 
presence of a nucleus, but saw the enlargement and said it 
might be artificial, Schreiner also figures it. It is most 
easily seen in isolated rod-cells, in a maceration. I find that 
there may be one or more, and that they are simply due to 
the rod-cell being flattened in places by the pressure of 
adjacent cells; the flattened part appears as an enlargement 
if not seen in edge view. 

The rods are cone-shaped with the apices rounded. The 
base has the same diameter as the rod-cell, that is, where 
they are continuous, and from here the diameter gradually 
decreases towards the lower end, though at first very gradu- 
ally. ‘They are separated and surrounded by a homogeneous 
substance (PI. 7, fig. 15, R. mat.), which fills up all the cavities 
that would otherwise have remained between them, and also 
forms a layer below them. ‘This substance is stained black 
by iron hematoxylin, it is blackened by osmic acid, and is 
stained blue by Mallory’s connective-tissue stain. I believe 


86 W. J. DAKIN. 


it is a semi-fluid substance of connective-tissue-like nature, 
which contains some oil or fatty body, and I have called it 
the rod-matrix (PI. 7, fig. 17, R. mat.). 

Patten described the rods, which are very difficult to 
preserve, as consisting of a “hyaline refractive sheath 
surrounding a pyramidal axial. core filled with a watery non- 
refractive fluid, and a short distance from the inner ends of 
the rods, terminating in a rounded apex” ({22], p. 585). 
This axial core is, in my opinion, the true rod, and what he 
described as the sheath is the surrounding rod-matrix. 
Carriére (21), had noticed this before Patten, and described 
the rods as being immersed in a fatty substance. Patten, 
however, adds that this was due to the fusion of the sheaths 
of the poorly preserved rods. Rawitz agreed with Patten 
about this sheath, though he differed slightly in regard to 
its optical properties, and Schreiner also does not accept 
Carriére’s view. Hesse’s view is, however, the same as mine, 
and he has emphasised the error of Patten, Rawitz, and 
Schreiner, whose peculiar idea of the rod was due to the fact 
that they believed an outer sheath to be necessary. ‘lhe rod 
structure differs from that of the rod-cell in the fact that 
there is much less stainable protoplasm, and this is usually 
ageregated round an axial fibre (Pl. 7, fig. 13, da.f.). It 
will be unnecessary here to go into further comparisons of 
the previous views on these structures. The rod-cells have 
been described almost correctly, though with deficiencies by 
most observers, with the great exception of Hyde, whose 
account I am leaving until later. 

In sections of well-preserved rod-cells and rods, such as 
those fixed in Bouin or Zenker and stained in Mallory’s stain, 
an axial fibril will be easily seen running through the rod. 
It is with reference to this structure that most of the con- 
fusion has arisen. Patten stated that each rod-cell contained 
an axial nerve-fibre which entered the attenuated end, passed 
through the first vesicle-like swelling, passed the large 
nucleus, and went on down to the lower end of the rod, 
whence it issued, and divided into two main branches which 


THE EYE OF PECTEN. 87 


became connected with the axial fibres of neighbouring cells 
(see Patten’s fig. 140, Taf. 32). Furthermore, he describes 
how towards the lower ends of the rod-cells the axial nerve- 
fibre begins to give off radiating fibrille, which are so 
numerous in the rods as to constitute the greater part of 
their substance. Hensen was the first to see the axial fibre 
in the rod. Patten figured it as being equally distinct and 
of the same diameter in rod-cell and rod. Rawitz found, 
however, that there was a fine canal running through the 
former in which lay the fibre, which, he adds, is the continua- 
tion of a nerve-fibre from the proximal branch of the optic 
nerve. ‘This central canal and fibre was supposed to be 
present in the rod but terminated without the complicated 
connections of Patten. Carriére, in his second paper (an 
answer to Patten’s criticisms of his first) (26), could not bring 
the existence of a nerve-fibre inside a cell into line with 
histological teaching, and hence said that what was present 
was simply a differentiation of the cell-substance. Schreiner 
came to the conclusion that a detailed examination was 
necessary owing to the diverging opinions of previous authors, 
and found after making sections and teased preparations 
that there was no axial fibre at all in the rod-cells, and what 
had been seen there was only one of the contours of a rod- 
cell produced by pressure causing these normally cylindrical 
cells to be angular. He found it very distinctly stained, 
however, in the rods, and it ran straight to the end where it 
terminated in a point. He adds that it differs somewhat 
in staining qualities from nervous tissue and is too thick for 
a nerve fibre (p. 72). 

Hesse found after all this research that it was necessary to 
go back to the earlier views, for he made out the axial fibril 
running through both rod-cell and rod. 

He states, however, that it is far more easily seen in the 
rods, and even there it varies in the same preparation. 

It is less distinct in the rod-cells because thinner (except 
in P. aratus), and in some cases Hesse saw more than one 
present. This bring us to Hyde’s views (39) regarding rod- 


88 W. J. DAKIN. 


cell, rod, and axial fibre, which are based on methylene-blue 
methods. It appears somewhat difficult for me to understand 
‘how the material presumably stained could remain in good 
condition for four years until taken up for completion. 

We are told that a rod consists of a nerve-cell whose small 
anterior end (upper end 7) projects slightly beyond the median 
limiting membrane, aud whose much elongated posterior 
portion is tubular and bluntly terminated. ‘This portion is 
encased in a hyaline sheath, with the end capped by a homo- 
geneous cuticular substance, which in methylene-blue pre- 
parations appears like the matrix separating the rods. A 
small nucleus lies in the anterior end of the rods and from 
this an axial fibre extends to the posterior (lower end). 

There is another series of important elements in the retina 
—pipolar cells.” These extend from the median limiting 
membrane (presumably the same as the line dividing rod- 
cells from rods) outwards towards the margin of the retina. 
“Their large granular elliptical nuclei may be seen in longi- 
tudinal sections extending in a row, a short distance from 
the median limiting membrane. ‘lhe whole cell with its 
afferent and efferent axon is encased in a hyaline sheath, 
under which are scattered blue granules of various sizes.” 
The rest of Hyde’s conclusions are difficult to understand, 
but putting figures and descriptions together, one gathers 
that the rod-cells of all previous writers are the same as 
certain “supporting cells of the median layer”’ of Hyde. 
‘The bipolar nerve-ce!ls above referred to are the inner inter- 
stitial cells (Pl. 7, fig. 18, I. .c.) or inner ganglionic cells of 
other authors, and from them arise two fibres, one of which 
runs to the edge of the retina and the other to the pseudo 
“‘sieve-membrane,” following the course of the median 
supporting cells of Hyde and lying between them. These 
are the afferent and efferent axons. Distally the afferent 
axon has a dendritic termination, which comes into relation 
with the upper end of the axial fibre of the rod. Proximally 
the efferent axon terminates with other efferent axons in a 
common large ganglionic cell. These marginal ganglionic 


re 


THE EYE OF PHECYIEN. 89 


cells, besides connecting up various axons of bipolar cells, 
give off fibres which make up the proximal branch of the 
optic nerve. ‘This means in short that the sensory structures 
(the rod-cells) of all other writers are merely median sup- 
porting cells, the inner ganglionic cells of Patten and Rawitz 
(the interstitial supporting cells) are bipolar nerve-cells, and 
the marginal ganglionic cells of Hyde have not been seen by 
any other investigators. Patten and others must have mis- 
taken, adds Hyde, the axons of the bipolar cells for axial 
fibres in the rod-cells! 

I took some little trouble to see if it were possible for any 
of these results to be correct, though from a priori reasons, 
assuming @ little of the previous work to be satisfactory, it 
appeared very doubtful. 

In the first place Patten and his successors could not have 
seen the bipolar cell axon inside a rod-cell, since they all 
described it as being outside and possible of separation in 
teased preparations. 

In the second place, the bipolar cell of Hyde has always 
been described as multipolar, and hence though two long 
afferent and efferent axons might have been missed, her 
predecessors had a better idea of its true shape. Finally, 
since rod-cell and rod are in direct continuation it is impos- 
sible for the axial fibre of the latter to become connected with 
the process of a cell lying between the former. The results 
are, in fact, impossible. ‘he rod-cell in its general features I 
have found to be exactly as described by most other writers. 
The ‘ bipolar cell” is the interstitial supporting cell to be 
subsequently described, and the rod contains no nucleus at all. 
The marginal ganglionic cells as described by Hyde do not 
exist. J must now refer to the axial fibre and the internal 
structure of the rods. ‘he first idea striking an observer is 
that the true condition of things is like that described by 
Schreiner, viz. an axial fibre is present in the rods, but not 
in the rod-cells. After staining with iron hematoxylin, but 
especially after using Mallory’s stain, with Bouin’s fluid as 
fixative, traces of a much thinner fibre or fibres are to be seen 


90 W. J. DAKIN. 


in the rod-cells (Pl. 7, fig. 18, cells B.). In a memoir on 
Pecten (44) I made the statement that this was probably the 
true condition, and J find that Schreiner in his text-book on 
histology (88) has done the same. ‘he latter author refers to 
the axial fibre as a neurofibril, a structure which has risen in 
importance since Apathy’s work in 1897 and about which 
very much has been written, chiefly on the continent, in the 
last few years. I believed that the thick neurofibril easily 
seen in the rods divided into numerous delicate, more 
elementary fibrille in the rod-cells, a view rendered more 
probable by the fact that whilst the contents of the latter are 
uniformly distributed, filling the cell, the protoplasm of the 
rods is usually aggregated in the middle. I could not at that 
time, however, find proof of this in Pecten, although Hesse 
stated that sometimes he had seen more than one fibre 
present. 

Usually the axial fibre is thickest and stains most darkly in 
the upper half of the rod, though sometimes the whole length 
in the rod is much the same in appearance. 

It begins to disappear a little below the line of junction 
with the rod-cells, but again sometimes extends quite as 
distinctly a little above this. ‘This disappearance, or partial 
disappearance, is due to the separation into delicate branches 
which extend right through the rod-cell (Pl. 7, fig. 18, 
BoC.) , 

The separation is irregular, and sometimes one fibril is left 
much thicker and may be followed easily through the rod- 
cell: presumably this feature gave rise to Patten’s view. 

‘The point of separation of the axial fibril of the rods into 
finer fibrille varies even in the same section, and in rod-cells 
situated near the margin of the retina (young rod-cells) the 
axial fibre may often be seen as thick and distinct as in the 
rods. In macerations in chloral hydrate solution or chromic 
acid and also in teased fresh material the axial fibril is seen 
as distinctly as in stained sections. 

It is rather thick and quite stiff like a bristle in these 
preparations, never having normally the snaky course 


THE EYE OF PECTEN. 9] 


ascribed to it by Hesse. Often the more delicate rod is 
broken up in maceration and the axial fibre is then left 
sticking out from the protoplasmic remains of the cell (PI. 7, 
fig. 14, Aw. f.). 

After seeing these preparations one is rather inclined to 
believe that this is also a supporting structure. 

In sections, however, the appearances are more favourable 
to the nervous view. ‘I'he separation of the components of 
the axial fibre is similar to that often taking place in neuro- 
fibrillee, and the fibre occurs in a sense-cell and stains always 
like the nerve-fibres in the same preparation. In the rods 
the axial fibre differs somewhat in appearance from a typical 
neurofibril in thickness and distinctness. These structures 
considered as the conducting elements of the nervous system 
were unknown to the earlier writers on the Pecten eye. 

There are two views, then, that may be taken of the function 
of these fibrille. We may regard the axial fibril in the rod 
as a true neurofibril, a ‘primitive fibril” formed by the 
apposition of several elementary fibrillae which pass through 
the rod-cell, the apposition occurring normally or through 
fixation. These neurofibrille have, then, the function assigned 
to them by Apathy and Bethe—the conduction of nerve 
inpulses. On the other hand we may consider the whole to 
have only the function of a system of supporting fibres. The 
latter view would resemble that put forward by Nansen and 
accepted by several investigators, who consider the neuro- 
fibrillee to be the supporting, and not the conducting elements 
of the nerve-cells. It is also conceivable, of course, that the 
structures are not homologous with the neurofibrille of nerve- 
cells at all. There is at present, to my mind, much confusion 
existing in reference to fibrous structures in nerve-cells, 
especially since Holmgren has shown (87) that processes of 
the neuroglia actually penetrate into ganglion cells and act 
as supporting fibres. 

An axial fibril of the same type as that occurring in the 
Pecten eye is a feature of the rod-cells of many other inver- 
tebrate eyes. For example, in the Lamellibranchiata it is 


92 . OW. I. DARIN 


present in Arca, Lima, Spondylus (84), and Cardium (42) ; 
in the Cephalopoda it is probably of general occurrence. It 
is very definite in the rods of the Alciopiden, and has been 
found in the Polychates Nereis and Lysidice by Hesse (88). 
In Gastropods a definite bundle of neurofibrille has been 
found in the visual cells of Limax (Smith [40]). In other 
forms there occur, instead of one thick axial fibril, a number 
of fibrillee which terminate in a comb-like margin (‘Stiftchen- 
saum”’ of Hesse). ‘This is a feature of the distal cells of the 
Pecten eye, andaccording to Hesse is practically universal, the 
fibrille occurring also in the rods and cones of vertebrates. 

‘The rods or analogous structures are also of widespread 
occurrence in optic sense-orgaus, though it would be difficult 
to homologise many of the rod-like structures with one another. 
Hensen, and later Grenacher, looked upon all the rods as 
cuticular structures, but I doubt now if any rod can be shown 
to be cuticular, not even the rhabdome of the Arthropods, 
a differentiated part of the reticular cells. Hesse regards 
the neurofibrillae then as the universal actual recipient ele- 
ments of the visual cell and the plasmatic part of the rod as 
a support for the fibrils. Experimentally it is impossible to 
determine whether the neurofibrille are the recipient elements 
or not, but from the constancy of their presence I believe 
they play a great part in this process. I have shown how in 
macerations the rod-cell may break up, leaving the axial fibre 
(Pl. 7, fig. 14). It does not appear from this as if the rod 
could give much support to the latter, but the true state of 
things in the living eye may possibly be different. I am 
rather inclined to believe, however, that the plasmatic portion 
of the rod acts conjointly as a recipient organ, and that the 
stimulus is passed on to the neurofibrille which conduct the 
nerve impulse wider. 

I consider Hesse’s estimation of the number of rods in a 
retina to be rather low for the large eyes of P. jacobeus 
or P. maximus. In the latter species there were about 
ten thousand in the retina of one specimen examined, and the 
number of rod-cells therefore exceeded this number, since the 
marginal ones do not bear rods, 


THE EYE OF PECTEN. 93 


Below the rod-matrix which underlies the rods is a limiting 
membrane, the basement membrane (PI. 7, fig. 13, B. m.), 
which extends completely across the eye. It corresponds to 
Schreiner’s “Innere Siebmembran,” but is a perfectly con- 
tinuous thin sheet. It is stained by hematoxylin similarly 
to the matrix but darker, and since the rods terminate a little 
distance above it it is obvious that they cannot pass through 
it. It occupies a similar position to Patten’s “ vitreous net- 
work,” but his description also refers toa thin layer of hyaline 
substance perforated by large holes into which the inner ends 
of the rods fit, and Schreiner states that the points of the 
rods come to lie against the tapetum. No traces of any cell- 
structure have been made out in this bounding membrane, 
which, as noted above, is not perforated by the rods. 

Reference has already been made to the marginal area 
of the retina (Pl. 6, fig. 1, W. Ret.). Thisis best studied from 
specimens fixed in Carnoy’s fluid. The rods remain practically 
similar in size until about the tenth from the margin of the 
rod-bearing region, and then follows a rapid decrease in size, 
leading to the apparently fibrous lateral parts where no rods 
are present. Careful examination will reveal the fact that 
the so-called outer sieve-membrane can be traced to the very 
edge of the retina, but the space between it and the basement- 
membrane is exceedingly small. ‘This corresponds, however, 
to the space occupied by the rods in the middle part of the 
retina, The axial fibre or neurofibril can be seen more dis- 
tinctly in these marginal rod-cells, which for a little distance 
are similar in diameter to the much longer ones in the centre 
of the retina. They next become much less in diameter until 
finally the boundaries become difficult to detect, and the axial 
fibril is the most distinct part of the cell. It can also be seen 
extending below the line of the pseudo sieve-membrane, 
though without any rod. Between these modified rod-cells 
are more supporting cells. 

The marginal region differs, therefore, from the central 
part of the retina in being composed of rod-cells which are 
far shorter than those of the latter region, whose diameter is 


94, W. J. DAKIN. 


reduced, and which bear practically no rods, though the 
axial fibril, which is very distinct, appears to extend a little 
way below the pseudo sieve-membrane. I believe that this 
region is occupied by young rod-cells and rods, and it can be 
seen how the rod isa gradual product of the rod-cells, as the 
appearance of the former in other parts of the retina naturally 
suggests. The gradual increase in size of the rods at the 
junction of the marginal and the central rod-bearing region 
is well marked. Probably the former region does not play 
any active part in vision at all. 

Hensen called this area the “ Retinawiilste,” because of the 
folded appearance in sections, and Hickson’s figures also 
show the retina in thisform, Ihave found the same condition 
after several fixatives, including Von Rath’s fluid and Bethe’s 
fixative for methylene blue. It is due to contraction, and is 
not normal. 

The inner interstitial supporting cells (PI. 7, fig. 
18, I. I. c.) have already been referred to several times. They 
lie in close contact with the rod-cells, between which they 
send their processes, and they are situated not far from the 
pseudo-membrane (PI. 7, fig. 18, S.m.). Patten was the first 
to recognise that the nuclei of these cells really belonged to 
cells lying between the rod-cells; they had been considered 
the nuclei of the latter by his predecessors. He figured them 
correctly as multipolar cells, but fell into error in regard to 
the nucleus, just as he and most of his successors considered 
that all the cells between the rod-cell layer and the septum 
had the same type of nucleus and were physiologically alike, 
It is quite easy to see in preparations stained with Mallory or 
iron hematoxylin that these nuclei resemble exactly those of 
the outer interstitial cells. There is a considerable difference 
between them and the large nuclei with distinct nucleolus 
and chromatin granules, which are present in both the distal 
sense-cells and the rod-cells (Pl. 7, figs. 13 and 16, R. C. n. 
and D. S.1.). 

The shape of the cells can be best seen in isolated retinas 
after macerating in =, per cent. chromic acid for several days 


a0 


—- 


THE EYE OF PECTEN, 95 


and staining in picro-carmine. There is very little protoplasm 
round the nucleus, and the processes are so irregular that 
beyond the fact that the cells are multipolar no definite shape 
ean be ascribed to them. ‘They are, on the whole, slightly 
larger than most of the outer interstitial cells. The processes 
wrap round the rod-cells, and may even extend through the 
basal pseudo-membrane between the rods. It has been said 
by Hesse that the inner interstitial cells are so rare in the 
centre of the retina that there is only one to four or five rod- 
cells. They are just as numerous here as elsewhere, except, 
perhaps, the peripheral modified region. 

Patten and Rawitz considered these cells to be ganglion- 
cells. Schreiner figured their shape incorrectly (as did J 
myself in a previous memoir) and found them to be connected 
with the distal nerve. Hesse also believed these cells to be 
nervous, for he states that the connection with the distal 
nerve is sometimes very distinct. In his last paper, however, 
he has altered his views of the relations between the distal 
cells and the nerve, and the question of the interstitial cells 
is therefore left open. Hyde, as already noted, regarded 
them as bi-polar nerve-cells connected with the axial fibre of 
the rods. Hverything, however, points to the conclusion that 
the inner interstitial cells, like the outer, are simply support- 
ing cells, their structure being quite unlike that of nerve- or 
sense-cells, and no connection with nerves having been found. 


SUB-RETINAL LAYERS. 


Below the retina there is generally a space, a split between 
it and the next layer, which may be of considerable size. All 
writers have figured this, but 1t 1s impossible, in most cases, 
to discover whether they regarded it as normal or not, since 
only Hesse refers to it, and he regarded it as due to shrinkage, 
I have figured it as it usually occurs in sections (PI. 6, fig. 1), 
but it must be remembered that this space is simply due to 
fixation, etc. In some cases, for example, the next layer, 
the tapetum (PI. 7, fig. 1, Ta.), will be found for some distance 


96 W. J. DAKIN. 


attached to the retina, and then will occur a stretch where it 
has evidently been torn off, and remains attached to the 
underlying pigment-layer (PI. 6, fig. 1, Pg.). This layer is 
also very often pulled away from the wall of the optic vesicle, 
and, whilst remaining attached to the tapetum, leaves 
fragments adhering to the wall, indicating where it once has 
been. Inthe normal eye, retina, tapetum, and pigment-layer 
are all in contact with one another, and no space occurs 
between the latter and the wall of the eye. 

The tapetum.—tThis layer is very conspicuous both in 
the living eye and in sections, and was very early discovered 
by Krohn (5). Hensen stated that it consisted of polyhedral 
cells. Patten called it “the argentea” (a name which I 
previously employed, but since “tapetum” is more correct 
by order of priority I have gone back to it). It is unfortunate 
that the term “ tapetum” has been used to designate two 
different layers. 

Hickson and Carriére believed the structure was formed of 
a number of fine fibres crossing at right angles. Patten 
considered it to be a modification of two layers of cells into 
refractive laminated membranes composed of minute square 
plates. Hesse found the tapetum to contain always a single 
nucleus surrounded by some residual protoplasm and there- 
fore derives this layer from a single large cell. 

The tapetum is made up of several layers of minute square 
plates (Pl. 7, fig. 19), which are yellow by transmitted light 
and reflect the light like silvery plates. 

This gives the diamond-like lustre to the living eye, and I 
have even a series of transverse sections, mounted in canada- 
balsam, which retain the same property. The layer is 
thickest in the centre and shades off gradually to a very thin 
peripheral region, which can. be traced between the retina 
and the pigment-layer to the wall of the optic vesicle. 

I have been unable to trace Hesse’s nucleus, and in adult 
eyes it is impossible to detect any remains of cells. I believe 
rather that this layer is formed by the underlying pigment 
containing cells or by other cells which disappear, but more 


ee 


—__, 


en 


THE EYE OF PECTEN. 97 


probably by the former, since some of the granules con- 
tained in these cells may resemble the substance of the 
tapetum. 

The pigment layer was also an early discovery because 
of its conspicuous appearance, and it is often possible to see 
the red pigment through the substance of the eye-stalk if 
there is little pigment in the epithelium of the latter. This 
layer was Patten’s tapetum. Hickson had regarded it as a 
fluid with no cellular elements at all. Carriére thought it 
was a continuation of the septum, and Rawitz describes it as 
being differently coloured in the various species. Schreiner 
explains Hickson’s view on the grounds that in P. maximus, 
which he examined, the pigment was really a fluid mass 
containing large and small granules, but adds that in other 
species this layer is a single or double row of rather large 
polygonal cells. 

I have investigated several species and find that this layer 
is cellular in all, though the boundaries of the cells may be 
difficult to see in the adult. In young specimens of Pecten, 
only a few millimetres in diameter, the pigment-layer appears 
to be composed of a single layer of epithelial-like cells with 
little or no pigment present. 

As the eye grows the pigment increases, the cells become 
filled and usually very irregular in shape, so that in large 
eyes of P. maximus the epithelial arrangement persists 
often only in the marginal part, and in the middle the layer 
may be irregularly two cells thick. 

The actual colour of the pigment is of little importance, 
since it varies in specimens of the same species and often in 
cells of the same eye. It is some shade of red-brown, and 
generally the cells are filled with a finely granular dark 
brown pigment, but with here and there frequently large, 
more darkly coloured bodies, like round concretions (Pl. 6, 
fig. 1, Ta.c.). There are large and small bodies of this 
nature, and sometimes also iridescent granules resembling in 
appearance the substance of the tapetum. The nuclei are 
best seen in iron hematoxylin preparations. In P. maximus 

VOL. 55, PART 1.—NEW SERIES. a 


98 W. J. DAKIN. 


they are round and contain a conspicuous nucleolus together 
with scattered chromatin granules. 

The cells of the pigment-layer appear to be continuous 
with the retinal cells at the periphery of the retina, Patten 
considered this layer, in fact, to be homologous with his outer 
ganglionic layer. I am unable to say whether it should be 
considered as a modified continuation of the distal sense-cells 
or of the outer interstitial cells. The nuclei are much more 
like those of the former, but the development of the Pecten 
eye still requires elucidation. This completes the account of 
the structures enclosed in the optic vesicle. A reference 
must be made here to the inner wall of the proximal hemi- 
sphere of the latter. It is formed of connective tissue, and 
Patten called the surface layer the “sclerotica” (PI. 6, 
fig. 1, Sc.). He described it as a two-layered, tough, hyaline, 
connective-tissue membrane continuous with the septum. 

Rawitz disagreed entirely with this and objected to the 
term “sclerotica,” because of its inappropriateness, considering 
the use of this term in the nomenclature of the vertebrate 
eye. This membrane of Patten is, however, well marked in 
longitudinal sections of the eye, though it is simply the 
limiting or surface layer of the connective tissue of the eye- 
stalk and directly continuous with it. In sections stained 
with Mallory’s fluid it is very conspicuous (Pl. 6, fig. 1, Se.), 
and stains a deep blue against the light blue of the ground 
tissue of the eye-stalk. It also differs from the latter in 
being hyaline and containing neither fibrous elements nor 
nuclei. The connective tissue forming the wall of the distal 
part of the optic vesicle lacks this differentiated surface layer 
entirely. In reactions to several stains it resembles the 
septum, and it also appears to be continuous with this mem- 
brane. The layer is thus obvious, but is not to be considered 
as a separate structure in Patten’s sense, and the term 
“ sclerotica”’ is certainly inapplicable. 

T have called it simply “‘ the modified connective tissue-wall 
of the optic vesicle.” It must be remembered that the terms 
“cornea,” “sclerotica,” “iris,” etc., used by Patten and others 


THE EYE OF. PECTEN. ; 99 


cannot be compared directly with those designations in ‘the 
vertebrate eye, for the structures bearing these names are not 
homologous, and in fact the whole structure of the eye is not 
to be homologised with that of the vertebrate optic organ. 
The resemblances are pure cases of homoplasy, and there is 
absolutely no proof of a genetic community of origin. 


INNERVATION AND GENERAL CONCLUSIONS. 


It has already been pointed out that the retina is innervated 
by two branches of an optic nerve which passes down the 
centre of the eye-stalk (Pl. 6, fig. 1; Pl. 7, fig. 2, Op. N.). 

This nerve has been considered as an offshoot from the 
circumpallial nerve. In sections which cut the optic nerve 
obliquely, so that only a small part appears in a section, this 
may very easily appear to be the case, but if a section cuts 
the mantle exactly in the plane of the optic nerve, so that a 
long stretch appears in one section, it will be seen that the 
real state of things is somewhat different. At irregular 
intervals nerves pass radially through the mantle- lobes 
(between the radial pallial muscles) from the visceral ganglion 
to the circumpallial nerve (PI. 7, fig. 2, Circ. N.). Some of the 
fibres of these nerves pass into the latter, but at certain places 
(below the eye-stalks) the bulk of the fibres pass round the 
eircumpallial nerve (on the shell side of it, Pl. 7, fig. 2), 
touching it, but not entering it, and these innervate the eye. 
Some fibres appear also to leave the circumpallial nerve and 
to enter this optic nerve, but it will be evident that most of 
the nerve-fibres come directly from the visceral ganglion. 

Now the visceral ganglion of Pecten is extremely compli- 
cated in build and I think unique among the Lamellibranchiata. 
No details will be given here, since a paper is being prepared 
on this subject, but it will be seen from the figure (text- 
fig. 1) that there are several lobes, of which two lateral 
ones are very conspicuous. From these radiate out on either 
side the pallial nerves (Pall. N.). The ganglion is asym- 
metrical, the left lateral lobe being larger than the right, and 


100 W. J. DAKIN. 


it is from these lobes that the nerves arise which innervate 
the eyes. It is interesting, therefore, to observe how the 
development of the eyes has affected the ganglion, for in P. 
jacobeus and P, maximus, where the number of eyes on 
the left, mantle-lobe exceeds that on the right, the left lateral 
lobe of the visceral ganglion is considerably larger than the 
right, especially in the former species, whereas in P. oper- 
cularis, where the number of eyes is more equal on both 
sides, the left lobe is but slightly larger than the right. 
Probably the presence of both lateral lobes is due in the first 
instance to the great development of pallial structures. 


TrExtT-Fic. 1. 


The retina of Pecten is of the inverted type, that is (like 
the vertebrate eye), the recipient bodies, the rods, are directed 
towards the tapetum, and away from the source of light 
(text-fig. 2). In addition to this feature we have a com- 
plexity only paralleled in a few cases in the invertebrata 
(and even then without the inversion), for there are two 
series of recipient cells. Inversion occurs in the Platyhelmia, 
though the eyes are much simpler than the Pecten eye. In 
the Lamellibranchiata the eyes are either absent or much 
more simple as a rule than the eye of Pecten, but we have. as 
a matter of fact the two eyes most like the one we are con- 
sidering in this group, namely, the pallial eyes of Spondylus, 
which are practically the same as Pecten, and the eyes 


THE, EYE OF PECTEN. 101 


(siphonal) of Cardium. In both cases there are two series of 
recipient cells and the retinas are inverted. 
There are some interesting analogies; thus, for example, 
the ocelli of Agrion (a dragon fly) possess a retina which has 
also two series of recipient cells very like the rod-cells with 
rods and the distal cells of Pecten, but there is no inversion. 


TEXT-FIG. 2. 


“ORR pees pret iact MoS lane 
D.C. Dista) Sense Celis. Re. Red cells ith Rods. 

We are also familiar in the vertebrate eye with two kinds of 
recipient structures—the rods and cones—though these bodies 
are situated in practically the same layer (Bernard [36] has, 
however, stated that in Amphibia the cones are earlier stages 
in the development of new rods). 

When all things are taken into consideration the eye [of 
Pecten. and also of Spondylus appears a very remarkable 


102 W. J. DAKIN. 


development, especially for a Lamellibranch, and the com- 
plexity of structure, together with the large number of eyes, 
has been a difficulty felt by most writers who have sought for 
an explanation of these organs. Patten put forward an 
extraordinary theory, calling the eyes “ heliophags.” It is 
hardly necessary to go into this here, since a criticism 
appeared in the ‘ Quarterly Journal of Microscopical Science,’ 
vol. 27, which may be referred to. 

The eyes have shown no evidence of being phosphorescent 
organs, though I have observed and stimulated them at night 
and in the dark. A shadow thrown on to the eyes of an 
open Pecten causes a closure of the valves, and this reaction 
usually takes place very rapidly, though very often the per- 
ception of light stimuli does not appear to be any better than 
by Arca with very simple eyes or others with pigment spots. 
If, however, the shadow thrown on to a Pecten does not 
extend over a number of eyes there appears to be no reac- 
tion, and, just as Rawitz observed some time ago, a small 
object quite near produces no effect unless its shadow falls 
on a large number of eyes in quick succession. No evidence 
of accommodation could be obtained experimentally. Further- 
more, it is hardly possible to correlate the presence of these 
structures with the active habits of the animal, e.g. swim- 
ming, for Lima swims just as well as Pecten, but has ex- 
tremely simple eyes. Again, Spondylus has eyes practically 
identical with those of Pecten, but does not swim, and the 
same thing applies to the only other Lamellibranch with an 
eye approaching that of Pecten in structure, namely Cardium. 
In the latter case the eyes are confined to the tentacles of the 
siphons. It would be interesting to determine by biometric 
‘methods whether these organs were still being kept up, or 
were degenerating, especially in forms like P. jacobeus 
and P. maximus, where there exist very small eyes side by 
side with the large ones. 

These may be growing, or they may be eyes which haye 
retained their young form, have not grown, and will not 
grow. They. agree with young eyes in structure. The 


aA Ss 


6 


THE EYE OF. PECTEN. 103 


variation, however, in specimens of the same size renders the 
examination of a large number a necessity, and I have been 
unable to obtain a fraction of that number. It is possible 
that P. jacobeus and P. maximus are more highly 
developed forms than P. opercularis and P. varius, for 
they possess no byssus, though the gland is present and they 
have passed through a byssus stage, and the retractor muscles 
of the foot, of which one is left in P. opercularis, are even 
more vestigial in P. maximus. If these two forms are con- 
sidered older we find that there has been a reduction in the 
number of the eyes, for they are more numerous in P. opercu- 
laris, P. tigrinus, and other smaller forms, and this reduc- 
tion has then taken place to a greater extent on the under 
convex valve than on the upper flat one. The increase in 
convexity and difference between the two valves, reaching a 
maximum in P. jacobeus, has been accompanied by a 
reduction of the eyes on the convex mantle-lobe both in 
number and size. ‘These are, however, only hypotheses. 
The large number of eyes present is probably to be accounted 
for by the reason put forward by Rawitz, namely, that the 
actual recipient area in each eye is small, that oblique rays 
are cut off, and that in life the eye-stalks remain still; a 
large field of view is therefore only possible with numerous 
eyes. 

The presence of two series of recipient elements has not 
been explained by previous writers and has in fact been 
usually passed over. No experiments have enabled me to 
state anything definitely about this, except that, as already 
mentioned, there appears little evidence of accommodation. 
It might be advisable to point out here that the removal of 
an animal like Pecten from the dim regions at the bottom to 
the daylight and shallow water of the aquarium has possibly 
an injurious effect, and probably it would be a delicate com- 
plicated structure like the eye that would suffer most. Hence 
it may be that our aquarium experiments are almost useless 
in this respect. 

The presence of the distal layer of sense-cells as well as 


104 3. DARIN. 


that of the rod-cells and rods may be a device for increasing 
the area of the recipient elements without increasing to any 
extent the size of the retina, but more probable is perhaps 
the following view. There has not yet been definitely proved 
to exist any special apparatus for accommodation in the eye 
(though Hesse’s theory has not been disproved). Now 
it may be that the two layers of recipient cells are for 
the reception of images of objects situated at different 
distances from the eye, which are focussed at different 
distances from the lens. ‘hus the image of near objects 
would be focussed on the rods and that of distant objects on 
the outer distal cells. A similar condition would apply to 
the ocelli of Agrion, and, in fact, Hesse describes such (85), 
but adds, “Ich kenne nirgends eine ihnliche Hinrichtung.” 
In the Heteropod eye there also appears to be a device for 
the reception of rays from objects at different distances from 
the eye. There is, however, only one series of cells, but the 
free ends bearing the comb-like margins are turned so that 
they are at right angles to the plane of the retina, and some 
are nearer the lens than others. 

The development of the Pecten eye still remains incom- 
pletely known, and Patten’s observations need confirmation. 
The derivation of the various layers will certainly throw 
much light on the structure of the adult eye and the inversion 
of the retina. Unfortunately the material for such a research 
is somewhat difficult to acquire as all the elements are formed 
in extremely young specimens, and I have been unable there- 
fore, so far, to follow out this line of inquiry. 

It will be perhaps useful if the most interesting features in 
the general structure of the Pecten eye are summarised here 
and a few comparisons made with other eyes, which may bear 
some resemblance to the former. The eye is a closed vesicle ; 
there is a cellular cornea continuous with the surface epithe- 
hum, and below this a cellular lens. The retina is made up 
of two series of recipient cells innervated by two branches of 
an optic nerve. The cells of the distal layer have each a 
comb-like margin, and the proximal visual cells bear rods 


THE. EYE OF PECTEN. 105 


with an axial neurofibril.. The retina is of the inverted type. 
The eyes are not cephalic, but occur on the mantle-lobes. 

There is no ground whatever for placing the Pecten eye in 
the same class as the vertebrate eye, for the resemblance is 
very superficial, and though the retina is inverted in both 
cases this has been produced in very different ways. If we 
consider Biitschli’s observations as correct the retina of the 
Pecten eye has been formed from an invagination of the ecto- 
derm, which forms a closed vesicle cut off from the surface. 
The distal wall of this gives rise to the retina, and the 
proximal to the pigment layer. 

Amongst invertebrate eyes that of Spondylus is the only 
one that can be safely homologised with the Pecten eye. 
The structure of these organs is identical but for one point, a 
layer of cells in Spondylus takes the place of the non-cellular 
septum of the Pecten eye. The eye of Cardium can also be 
homologised, though with less certainty. ‘There is a cellular 
lens, an inverted retina with two series of recipient cells, and 
also layers corresponding in position to the tapetum and pig- 
ment layer of Pecten. There is, however, another layer (the 
choroid) interpolated between the retina and tapetum, which 
may be taken as equivalent to the interstitial cells of the 
Pecten eye. 

These are, so far as I am aware, the only vesicular eyes 
occurring in the Lamellibranchiata. 

In the highly organised cephalopod eye we do not meet any 
resemblance to the Pecten eye, except that the visual cells 
bear rods with an axial neurofibril like these recipient struc- 
tures in the latter. There is a single layer of recipient cells 
directed towards the light, and the lens is not cellular and 
arises quite differently from the lens of the arthropod eyes. 

Amongst the Polycheta there are some highly organised 
visual organs, in particular those of the Alciopina, ex. Alciopa 
and Vanadis, and the large and complex organs of these 
forms have been studied in detail by Greeff and Hesse. The 
eye takes the form of a closed vesicle as in Pecten, the free 
pole being formed by a cellular cornea, a continuation of the 


106 W. J. DAKIN. 


general epithelium of the body-wall.. The inner wall of this 
optic vesicle is, however, also made up of a layer of cells, 
which though forming a complete hollow sphere, are differen- 
tiated in three regions, in structure and function. Those 
cells immediately under the cornea just spoken of are low and 
form a second and inner cornea. The cells lining the proximal 
half of the optic vesicle are the retinal cells, and between this 
area and the inner cornea the cells are again different and 
contain pigment. 

There is only one series of recipient cells in the retina, and 
they bear rods which resemble those of the Pecten eye and 
contain a very distinct axial neurofibril. They are, however, 
directed towards the lens, that is, not inverted. ‘The lens is 
spherical and non-cellular, and another difference from the 
eye of Pecten is produced by the presence of a vitreous body 
between lens and retina. 

There are several interesting arthropod eyes that may be 
briefly referred to. The ocelli of Cloéon (one of the May- 
flies) are distinctly peculiar and are superficially rather like 
the Pecten eye, but this resemblance is due to the dioptric 
part of the eye, and not to the retina. We have again a 
closed vesicle. ‘The cuticle extends over the cornea, but 
remains thin and does not form a corneal lens. The hypo- 
dermis forms a cornea similar to that of Pecten. Under this 
cornea and lying in the optic vesicle is a cellular lens 
strikingly like that of Pecten and altogether unlike other 
arthropod eyes. ‘The retina is made up of two layers of cells, 
but the distal ones are not visual and the proximal ones 
forming the retina proper are not inverted. 

Another interesting arthropod eye is the ocellus of Agrion. 
This bears some resemblance to the Pecten eye in the fact 
that there are two series of recipient cells in the retina. 
They are, however, not inyerted. The distal part of the 
optic vesicle is quite different, and the chitinous exoskeleton 
or cuticle is thickened over the free surface, forming a 
corneal lens. This is a monomeniscous arthropod eye there- 
fore, and the arrangement of the retinal cells is interesting. 


THE EYE OF PECTEN. 107 


The distal layer of sense-cells lie touching the lens, almost 
like the outer cells of Pecten touch the septum. 

A striking difference from the Pecten retina is, however, 
present which lends at the same time support to the view of 
Leydig, upheld by Lankester in 1883, namely that the com- 
pound eye is formed by the segregation of the elements of a 
simple eye, and this is the segregation of the retinal cells. 
The visual cells do not remain, as in the Pecten eye, alto- 
gether independent with their recipient ends directed towards 
or away from the lens, but bear a comb-like margin of 
neurofibril endings laterally and are collected in groups 
of threes, each group being a retinula. Thus we have a 
monomeniscous eye with a retinulate retina, the whole being 
very different from the Pecten retina except in the one point 
—the presence of visual cells arranged in two layers. 

The central eyes of the Scorpions may finally be mentioned 
here. These are also monomeniscous and present a far 
greater resemblance to the Pecten eye than appears at first 
sight. They are vesicular, though the cavity of the vesicle 
has disappeared and the retina is inverted, though, owing to 
a secondary reversion during development, this is not 
obvious. 

The eyes are developed from an involution of the hypo- 
dermis or ectoderm, which, however, does not le vertical to 
the surface. The outer wall becomes thickened and forms 
the retina; the inner wall remains thin and represents the 
post-retinal layer of ectoderm cells in the adult. This is 
strikingly like the process in the Pecten eye where the 
inner wall becomes the pigment layer. The retinal cells 
are of course inverted. The nerve-fibres are attached to 
the outer ends of these cells in the embryo, but, owing to 
reversion in the course of development, become connected 
to the inner ends in the adult eye. In the course of these 
changes the optic nerve must penetrate the post-retinal 
layer, and this has been shown by Ray Lankester and 
Bourne (46) to be the condition actually prevailing in the adult. 
Beyond this remarkable similarity in development the eyes 


108 W..J. DAKIN. 


are very different: there is a retinulate retina of one layer 


of 


recipient cells which are segregated in groups of fives, 


and the dioptric part is again represented by a corneal lens. 


It will be seen, therefore, that no eye outside the Lamelli- 


branch group presents anything but isolated features of 
resemblance, and the only common structures appear to be 
the general occurrence of rods with axial neurofibrille or 
visual cells with a margin of cilia-like processes arranged 
like the teeth of a very fine comb, and these margins may 
form rhabdomes. 


TUE 


12. 


LIvrERATURE. 


. Poli, Josepho.— Testacea utriusque Siciliz eorumque historia et 


anatome eneis illustrata,’ Parmee, 1795, vol. ii, p. 153. 


. Grant, Robert.—‘ Umrisse der vergl. Anat.,’ Leipzig, 1835, p. 311. 
. Garner, Robert.—‘ On the Nervous System of Molluscous Animals,” 


‘Trans. Linn. Soc. Lond.,’ vol. xvii, 1837, p. 488. 


. Grube, E.—* Uber Augen bei Muscheln.,” ‘Arch. f. Anat. u. Physiol.,’ 


Jahrg. 1840, Berlin, p. 27. 


. Krohn, A.—‘ Ueber augenihnliche Organe bei Pecten und 


Spondylus,” ‘ Arch. f. Anat. u. Physiol.,’ Jahrg. 1840, p. 381. 


. Will, Fr.—* Uber die Augen der Bivalven und der Ascidien,” 


‘Froriep’s Neue Notizen aus dem Gebiete d. Nat. u. Heilkunde,’ 
Bd. xxix, Weimar, 1844, p. 81. 


. Chiaje, St. delle.—‘ Miscellanea Anatomico Pathologica,’ tome ii, 


Napoli, 1847. 


. Siebold, C. Th. v— Lehrbuch der vergl. Anat. der wirbellosen 


Thiere,’ Berlin, 1848, p. 261. 


. Duvernoy, M.—* Mémoires sur le system nerveux des mollusque 


acéphales lamellibranches ou bivalves,” ‘Mémoires de l’academie 
des science de l’inst. de France,’ tome xxiv, Paris, 1854, p. 75. 


. Leydig, Franz.— Lehrbuch der Histologie des Menschen und der 


Thiere,’ Frankfort-a-M., 1857, p. 261. 

Bronn, H. G.— Classen und ord. des Tierreiches,’ Bd. 3, Abth. 1, 
1862. 

Keferstein, Wilhelm.—* Untersuchungen iiber niedere Seethiere, 
Ueber den Bau der Augen von Pecten,” ‘ Zeit. f. wiss. Zool.,’ 
Bd. xii, Leipzig, 1863, p. 133. 


13. 


14. 


15. 


16. 


iia 
18. 


19. 


20. 


21. 


22. 


23. 


24. 


25. 


26. 


27. 


28. 


29. 


30. 


31. 


32. 


33. 


THE HYE OF PECTEN. 109 


Hensen, v.—‘ Ueber das Auge einiger Cephalopoden,” ‘ Zeit. 
f. wiss. Zool.,’ xv, Leipzig, 1865, p. 220. 

“Ueber Sehpurpur bei Mollusken,” ‘Zool. Anz.,’ Bd. i, 
No. 2, Leipzig, 1878, p. 30. 

Chatin, J.—‘ Bull. de la société philomathique,’ Paris, 1877, pp. 8-14, 
44, 45. 

Gegenbauer, C.—‘ Grundriss d. vergl. Anatomie,’ Zweite Auflage, 
Leipzig, 1878, p. 373. 

Grenacher, H.— Unters. iiber das Sehorg. der Arth. Géttingen,’ 1879. 

Hickson, Sidney J.—‘‘ The Eye of Pecten,” ‘ Quart. Journ. Mier. 
Sci.,’ vol. 20, London, 1880. 

Schmidt, E. O.—‘ Handbuch d. vergl. Anatomie,’ 8 Auf., Jena, 1882. 

Sharp, Benj.—‘‘ On the Visual Organ in Lamellib.,” ‘ Mitth. a. d. 
Zool. Stat. zu Neapel,’ Bd. v, Leipzig, 1884, p. 455. 

Carriére, J—‘* Die Sehorgane der Thiere,” ‘Minch. u. Leipzig,’ 
1885, pp. 99-107. 

Patten, W.—* Eyes of Molluscs and Arthropods,” ‘Mitth. a. d. 
Zool. Stat. zu Neapel,’ Bd. vi, Berlin, 1886, p. 568. 

Butschli, O.— Notiz zur Morfologie des Auges der Muscheln,’ 
Heidelburg, 1886. 

Nansen, F.—‘* Die Nervenelemente, ihre Struktur und Verbindung,” 
‘Anat. Anz.,’ Bd. 3, 1888. 

Rawitz, B.—‘ Der Mantelrand der Acephalen,” ‘ Jenaisch. Zeit. f. 
Naturw., Bd. xxii (N. F. Bd. xv), Jena, 1888, p. 508. 

Carriére, J.—* Uber Molluskenaugen,” ‘ Arch. f. Mikr, Anat.,’ Bd. 
Xxxili, p. 390. 


Fraisse, P—*‘ Uber Molluskenaugen mit embryonalem Typus,” 
‘ Zeit. f. wiss. Zool.,’ Bd. xxxv, p. 461, 

Kishinouye.—* Note onthe Eye of Cardium muticum,” ‘ Journ. 
Coll. of Science, Tokyo,’ vol. vi, 1894. 

Parker, G. H.— Retina and Optic Ganglia in Decapods,” ‘ Mitth. 
Zool. Stat. Neapel, Bd. xii, 1895. 

Schreiner, K. H.—* Die Pectenaugen,” ‘ Bergens Museums Aarbog,’ 
1896, pp. 5-51. 

Apathy, S.—‘‘ Das leitende Element des Nervensystems,” ‘ Mitth. 
Zool. Stat. Neapel,’ Bd. xii, 1897. 

Hesse, R.— Die Augen der Plathelminthen,” ‘ Zeit. f. wiss. Zool.,’ 
Bad. Ixii, 1897. 


“ Die Augen der polychiten Anneliden,” ‘ Zeit. f. wiss. Zool.,’ 
Bd. lxv 1898. 


110 W. J. DAKIN. 
34. Hesse, R.—‘‘ Die Augen einiger Mollusken,” ‘ Zeit. f. wiss. Zool., 
Bd. Ixviii, 1900. 


“Von den Arthropoden-Augen,” ‘ Zeit. f. wiss. Zool.,’ Bd. 
Ixx, 1901. 


86. Bernard. H. M.—“ Studies in the Retina,” ‘ Quart. Journ. Mier. Sei.,’ 
vol. 43, 1901. 


87. Holmgren.—* Nervenzellen,” ‘ Anat. Hefte,’ H. 59, 1901. 
88. Schneider, K. C.—‘ Histologie der Tiere,’ Jena, 1902. 


35. 


89. Hyde, I—‘“ The Nerve Distribution in the Eye of Pecten 
irradians,” ‘Mark Anniv. Vol., New York,’ 1903. 


40. Smith, Gr.—* Eyes of certain Pulmonate Gasteropods,” ‘ Bull. Mus. 
Comp. Zool. Harvard,’ vol. xlviii, 1906. 

41. Paton, St.—‘* Reaction of the Vertebrate Embryo to Stimulation, 
etc.,” ‘Mitth. Neapel,’ Heft 3, Bd. xviii, 1907. 

42. Weber, F. L.—* Uber Sinnesorgane des genus Cardium,” ‘Arb. 
Zool. Inst. Wien.,’ 1908. 

43. Hesse, R.—‘‘ Das sehen d. nieder. Tiere, ete.,”’ Jena, 1908. 

44. Dakin, W. J.—‘ Pecten,” ‘L. M. B. C. Memoirs,’ xvii, 1909. 

45. 


“Striped Muscle in the Mantle of Lamellibranchia,” ‘ Anat. 
Anz.,’ Bd. xxxiv, 1909. 

46. Lankester and Bourne.—‘ The Eyes of Scorpio and Limulus,” 
‘Quart. Journ. Micr. Sci.,’ vol. 23, 1883. 


EXPLANATION OF PLATES 6 ann 7, 
Illustrating Mr. W. J. Dakin’s paper on “The Eye of Pecten.” 


List OF REFERENCE LETTERS. 


Aw. f. Axial fibril of rods. B.m. Basement-membrane. Cire. n. 
Circumpallial nerve. Cent. Centrosome of lens-cells. Co. Cornea. 
Co. S. Sub-corneal connective tissue. Con. Connective tissue of eye- 
stalk. Cut. Cuticle. D.S. Distal sense-cells. D. Sn. Nuclei of distal 
sense-cells. Hye. Kye. J. I. C. Inner interstitial cells. Z.Lens. LZ. C. 
Lens-cells. Lac. Blood lacune of eye-stalk. J. 7. Muscle-fibres on 
distal surface of lens. M. Mantle. M. 7. Muscle-fibres of connective 
tissue of optic vesicle. M. lf. Muscle-fibres of lens surface. Mus. 


THE EYE OF PECTEN. did als 


Muscles of eye-stalk. M. Ret. Marginal area of retina. N. Lf. Nuclei 
of muscle-fibres of lens surface. nuc. Nucleus. O. TJ. ¢c. Outer inter- 
stitial cells. Op. D. Distal branch optic nerve. Op. Ds. Sheath of 
distal nerve. Op. N. Optic nerve. Op. P. Proximal branch optic 
nerve. Op. P". Fibres (separated) of proximal branch of optic nerve. 
P. man. Pigment-mantle. Pg. Pigment-layer. R. C. Rod-cells. 
R. C.n. Nuclei of rod-cells. R. mat. Rod-matrix. Rod. Rod. S. m. 
Pseudo sieve-membrane (see text). Sc. Modified connective-tissue wall 
of optic vesicle. Sep. Septum. Ta. Tapetum. fa. c. Pigment layer 
concretion. V. Velum. 


PEATE 6: 


Fig. 1.—Section through eye-stalk and eye, P. maximus, in a plane 
at right angles to that of the mantle surface; the right side of the 
5 t=) 5 
fizure represents the shell side of the eye. The various parts, lens, 
tan) I Pp ’ 
retina, etc., have been drawn with the camera lucida, but from different 
preparations, each showing best the part drawn. x 270. 


PILATES JZ: 


Fig. 2.—Diagrammatic section through both mantle-lobes of P. 
jacobeus, illustrating the course of the nerves and difference in size 
of the eyes. The left mantle-lobe is to the left in the figure. 


Fig. 5—Upper surface of cornea, P. maximus. x 1000. 


Fig. 4.—Transverse section of corneal cells at about the middle of 
their height. P.maximus. xX 1000. 


Fig. 5.—Isolated cells from the lens. P. maximus, maceration in 
chloral hydrate solution. x 570. 


Fig. 6.—lLens-cells as seen in sections. 6. Normal cells from 
Hermann-sublimate fixed specimen, P. varius. «a. Cell from same 
specimen with large nucleus. Stain iron hematoxylin. d. Cell from 
lens fixed in von Rath’s fluid. x 660. 


Fig. 7.—Transverse section cutting layer of fibres between lens and 
subcorneal tissue. The fibres and cells are stained red with Mallory’s 
connective-tissue stain, the subcorneal tissue blue. P. tenuicostatus. 
x 310. 


Fig. 8.—Fibres between lens and subcorneal tissue; attached to the 


latter in a maceration preparation (chromic acid). P. jacobeus. 
x 300. 


Fig. 9.—Cells and nuclei between lens and subcorneal tissue, as seen 
through the cornea, which has been teased from an eye fixed in Zenker’s 
fluid. Iron hematoxylin. P.maximus. xX 330. 


112 W. J. DAKIN. 


Fig. 10.—Transverse section of cornea and subcorneal tissue of 
P. jacobzus (Bielschowsky-Paton method). x 650. 


Fig. 11.—Isolated cells from distal surface of lens. P. maximus, 
chromic acid maceration, x 330. 


Fig. 12.—Transverse section of cornea and pigment-mantle of 
P.tigrinus. Fixed Zenker, stained Mallory. x 300. 


Fig. 18.—Schematic view of retinal elements, reconstructed from 
sections and macerations. The two left-hand rod-cells are shown in 
external view, from macerations, and the two right-hand ones in section. 
x about 920. 


Fig. 18a.—Distal ends of two rod-cells (chromic acid maceration). 

Fig. 14.—Rod-cells with partly broken-up rods, showing the bristle- 
like appearance of axial fibre. P. maximus (chromic acid maceration). 
x 900. 

Fig, 15.—Isolated interstitial supporting cells from retina; a and b 
are two outer interstitial cells. P. maximus. Chromic acid and 
chloral hydrate macerations. x 900. 


Fig. 16.—Transverse section of distal sense-cells and outer inter- 
stitial cells. Mallory’s connective-tissue stain. P.maximus. x 940. 

Fig. 17.Transverse section of rods and rod-matrix. P.jacobeus. 
Fixed Zenker, stained by modified Weigert method. x 800. 

Fig. 18.—Distal branch of optic nerve, breaking up into branches on 
surface of septum. P.jacobeus. From teased preparations. x 250. 


Fig. 19.—Tapetum in surface view. From sections. The large circle 
shows relative size of a rod-cell in section. x 1600. 


Miia 


Au bY 


i 


| 
| 


IU SN GS 2B) PS AOY? Vou nOge PDI 


Fath Tath” London, 


ing 


FS 


iy 


AM 


se 


~ 


ps 
z 


= Nv 


J* 


~ 


1.55 NS LUT, 


Jo 


Huth Lath? Londen 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 113 


Observations on certain Blood-parasites of Fishes 
occurring at Rovigno. 


By 
Prof. E. A. Minchin, M.A., and H. MW. Woodcock, D.Sc. 


With Plates 8-10. 


THe objects of this paper are twofold. The first is the 
description of a hemogregarine and a trypanosome occur- 
ring in a species of Trigla (T’. lineata), which is a common 
gurnard in the Northern Adriatic. We are unaware of any 
published observations of blood-parasites in this fish, and we 
regard the hremogregarine as a new species, for which we pro- 
pose the name He mogregarina rovignensis. Neumann, 
in his recent account of heematozoa of marine fishes at Naples 
(15), mentions the occurrence of a trypanosome which he 
names T’. triglz in another species of Trigla, namely, T. 
corax. It is not at all improbable that the trypanosome 
we have found is the same species as that described by 
Neumann; we therefore do not propose to give it a new 
specific name. 

The second object that we have in view is the comparison 
of the minute structure of the nucleus of the above-named 
hemogregarine with that of the trophonucleus of a trypano- 
some. for this purpose, as the trypanosome of the Trigla 
is very rare in our material, a trypanosome which occurs in 
much greater abundance in skates (Raia, sp.) in the 
neighbourhood of Rovigno has been selected. ‘This form 
is most probably T. raiz Lav. and Mesn. 

With regard to the technique, permanent preparations were 
made according to both the principal methods in use. Smears 

VOL. 99, PART 1.—NEW SERIES. 8 


114 E. A. MINCHIN AND H. M. WOODCOCK. 


on slides were fixed in osmic vapour! for half a minute and 
then placed in absolute alcohol for ten minutes or so. ‘They 
were stained with Giemsa’s solution in the ordinary manner 
for two and a half to three hours and afterwards differentiated 
with orange-tannin and acetone. Smears on cover-slips, on 
the other hand, were at once dropped face downwards into 
sublimate-acetic mixture or Schaudinn’s fluid for fixation, 
after which they were brought up to 90 per cent. alcohol. 
These smears were stained either with iron-hematoxylin or 
with Twort’s stain and finally mounted in balsam. Fuller 
details regarding this mode of procedure will be found in 
Minchin (11), the method there described being that followed 
in the present case. The figures of the hamogregarine are 
all drawn to a magnification of 3000, those of the trypano- 
somes to a magnification of 2000. We are indebted to Miss 
Rhodes for drawing the greater number of the figures and for 
colouring nearly all, and take this opportunity of expressing 
our thanks to her. 


J. HmMoGREGARINA ROVIGNENSIS, N. SP. 
(Figs. 1-28, 39-50, and 58-62.) 


This parasite was found on three occasions in the early 
spring. In the three gurnards in which it was seen it was 
not at all infrequent in the blood, as we have since learnt by 
the study of stained preparations. The parasite, chiefly in its 
small intra-cellular form, was observed in the living condition 
in cover-slip preparations of fresh drops of blood, but only 
with difficulty, since it was not readily distinguishable from 
the cytoplasm of the blood-corpuscle. When made out with 
certainty it was seen as a small oval area, somewhat clearer 
and paler than the surrounding cytoplasm. No prominent 
refractile granules were noticed, nor could the nucleus be 
made out satisfactorily. We did not observe any alteration 


1 A 4 per cent. solution of osmic acid was used, to which a drop or 
two of acetic acid had been added. 


BLOOD-PARASITES OF FISHES OCCURRING Al ROVIGNO. 115 


or movement in these intra-cellular forms so long as they 
were kept under notice, but in a few of the living preparations 
one or two small forms were seen free. These in all proba- 
bility corresponded to the small intra-cellular parasites. They 
appeared as sausage-shaped bodies very slightly curved. One 
third or so of the body was granular in character ; this prob- 
ably was the region of the nucleus (cf. description of fixed 
and stained parasites below). ‘They underwent no active 
displacement, and it was difficult to feel sure whether they 
really moved or altered in shape at all or not. Now and 
again the concavity would appear to lie first on one side of 
the body and then on the other, and now and then the body 
would stand on one end, appearing in transverse optical sec- 
tion as a round globule; but these changes were extremely 
slow, and may have been due to passive rotation of the vermi- 
cule, caused by very delicate currents in the blood-plasma at 
that spot, not sufficient to disturb the corpuscles. As we did 
not see any intra-cellular forms actively liberating themselves 
from the corpuscle, it may possibly be that these free indi- 
viduals were accidentally liberated by the rupture of the 
corpuscle during the process of manipulation, which would 
explain their apparent immobility. In the stained smears 
only one such free form has been found (fig. 3). 

In most of the permanent preparations made only a little 
searching with the oil-immersion lens is required to find a 
hemogregarine. ‘I'he parasite does not appear to be in quite 
the same period of its development in all three gumnards. In 
gurnards 1 and 3 the hemogregarine is present chiefly or 
almost entirely in the one phase—the small type. Large 
forms (see below) do occur, but they are evidently very rare. 
Two or three individuals of the latter type have been found 
both in the general circulation (fishes 1 and 3) and in a 
smear from the kidney (fish 1), but they have not been 
observed in smears from the liver or spleen, and no other 
phases have been seen in any of the organs. In fish 2, on 
the other hand, in which the parasites are rather more 
numerous, besides the small forms many large forms occur 


116 E. A. MINCHIN AND H. M. WOODCOCK. 


in the general circulation, and these belong to two distinct 
types. Most unfortunately, as it has turned out, no smears 
from the organs were made in the case of this fish. 

Considering first the relation of the parasites to their host- 
cells, we find that nearly always a single individual is present 
in a blood-corpuscle. Instances of a double infection occur, 
but they are very rare; we have only noticed two. In these 
cases two small forms lie side by side in one half of the corpuscle, 
the nucleus being displaced to the opposite end (fig. 27). 
The only effect which the parasites appear to have on the 
corpuscles is a mechanical one. In no case does the haemo- 
eregarine attack the nucleus or cause its hypertrophy or other 
degenerative effect.1 In corpuscles infected with the small 
forms the nucleus is generally displaced to one side (ef. figs. 
1, 2, 16,17). In the case of the large forms, the nucleus of 
the host-cell is often displaced quite to one side, lying at the 
periphery of the corpuscle (figs. 18-26). It may be slightly 
flattened or compressed (figs. 18-20), but shows no other sign 
of alteration. 

In a few cases indications of the presence of a delicate 
membrane or envelope around the intra-cellular parasite 
appear to be furnished by the stained preparations in a rather 
interesting manner. Usually no signs of such a structure 
can be detected; even where, as occasionally happens, the 
parasite has shrunk slightly from the enclosing corpuscle, 
leaving a clear space around itself, it cannot be said, as a 
rule, that a definite sheath is apparent. But now and again 
the stain is deposited in a marked manner in the protoplasm 
of the corpuscle immediately around the parasite, which 
points to the existence of a layer of somewhat altered 
character, acting as a hindrance to the further penetration of 
the stain. ‘This is especially well seen in those cases where 
the cytoplasm of the parasite happens to have partly or 
entirely shrunk (as in figs. 6, 39), leaving a deeply stained 
line bordering the outer edge of the clear space. We have 
noticed this appearance occasionally around both the small 

1 Hence this parasite does not belong to the genus Karyolysus. 


BLOOD-PARASITES OF FISHES OCCURRING A’ ROVIGNO. 117 


forms and the large ones of the wide type (fig. 11), but it is 
not shown clearly in the case of the large forms of the narrow 
kind, perhaps because the deeply stained cytoplasm of these 
individuals renders it less apparent. We regard this sheath 
as probably in the nature of a cytocyst, i. e. an altered layer 
of the blood-corpuscle.! 

In describing the structure of the parasites we may begin 
with the small forms. These show, on the whole, great 
uniformity in size and appearance. The body is nearly always 
oval in shape, either a fairly regular oval (figs. 39-42) or slightly 
pyriform, the half in which the nucleus lies being in this case 
somewhat narrower than the other half, and the end less 
broadly rounded (figs. 2, 8, 16). The average size of these 
small individuals—the mean being taken of several measure- 
ments—is 4°8 w in length by 2°34 in greatest width on “ wet”’ 
smears (sublimate-acetic mixtures, iron-hematoxylin), and 
4-9 in length by 2-4 in width on “dry” smears (osmic, 
absolute alcohol, Giemsa). We regard the above figures as 
representing, as nearly as possible, the true or correct average 
size of this phase. The largest “normal” dimensions noticed 
are 0°2 u by 2°6u, and the smallest 4°44 by 2°24. We add 
“normal”? because it is interesting and instructive to note 
one or two instances which well illustrate the considerable 
variation in this respect, which may be caused solely by the 
technique. One of them” is afforded by an extremely thin, 
“dry ” smear in which both corpuscles and parasites are 
greatly flattened out and consequently enlarged. On this 
smear the apparent average size of the small form is 564 
by 3u, and some individuals measure as much as 6:2 by 
3:3. On the other hand, in a particular ‘“‘ wet”? smear 
(Schaudinn, T'wort) both corpuscles and parasites are uni- 
formly smaller than the normal average on other films made 


1 We have observed nothing so well marked as the capsule, with lines 
of rupture, described by Sambon and Seligmann, for instance, around 
certain hemogregarines of snakes (21). 

2 It is, perhaps, scarcely necessary to mention that these cases are 
not included in the above “ standard” average. 


118 E. A. MINCHIN AND H. M. WOODCOCK. 


at the same time. Here the average is only 42 by 2°2 
(cf. figs. 58, 60, 61). We are rather at a loss to account for 
this case, but it seems evident that shrinkage must have 
taken place sometime during the manipulation. 

In the small type of the parasite the nucleus is relatively 
large. As already mentioned, it is generally situated entirely 
in one half of the body and near one end, but now and then 
it occupies a more median position (fig. 5). In Giemsa-stained 
preparations the nucleus appears to take up quite half the 
body, or even more, and to occupy its entire width or even to 
bulge out slightly at the sides (figs. 1, 2, 7). From prepara- 
tions stained with iron-hematoxylin, however, it is apparent 
(cf. figs. 39-45) that in the former smears there is a certain 
amount of artificial enlargement, due to the characteristic 
overloading with the stain which so detracts from the value 
of the Romanowsky method. We reserve a detailed descrip- 
tion of the structure of the nucleus until later, when we 
compare it with that of a trypanosome. ‘lhe general cyto- 
plasm appears fairly homogeneous in character, and with the 
exceptions to be mentioned, is usually devoid of granules. A 
conspicuous vacuole is frequently present; this lies about the 
middle of the body, generally close to the nucleus. In 
Giemsa smears it is sharply defined, becanse the cytoplasm, 
as arule, is distinctly stained, either purplish or bluish in tint 
(figs. 1, 8, 16, 17); but in iron-hematoxylin preparations it 
is often difficult to make out, probably owing to the fact that 
the cytoplasm of the parasites in these smears is itself very - 
pale and scarcely stained at all; sometimes, however, it is 
well seen (fig. 40). Rarely the cytoplasm contains two or 
three vacuole-like areas, with less sharply defined limits (ef. 
figs. 2, 41). 

The most striking feature in the morphology of the para- 
sites, as brought out by iron-hematoxylin, is afforded by one 
or two large granules, which take up this stain with intense 
avidity (cf. figs. 39, 40, 43, 46). They are often present, but. 
not always (figs. 41,42). When they occur these granules 
are by far the most prominent objects in the body, appearing 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 119 


always larger and much blacker than the chromatic grains of 
the nucleus. The granules are of about the same size whether 
two are present or one. They are always distinctly outside 
the nucleus, lying usually close to it, however, about the 
middle of the body. They were first noticed in the iron- 
hematoxylin preparations, and we surmised at first that they 
might correspond to part or all of the conspicuous vacuolar 
region seen in Giemsa smears, but have found since that this 
is not the case. In some individuals both grains and vacuole 
are seen to be present (figs. 1, 13, 16, 40). Sometimes one of 
the grains appears to lie in the vacuole, but we think that, 
in such a case, the grain is really outside the vacuole, lying 
above or below it instead of at the side (cf. especially fig. 43, 
where one grain is at the side, the other apparently in the 
vacuole). Although the grains (or single grain) usually lie 
close to the vacuole, this is not always so; for instance, in 
the parasite drawn in fig. 46 the grain is well removed from 
the nuclear zone, while in fig. 39 it is on the opposite side of 
the nucleus, near the other end of the body. 

Turning next to Giemsa smears for indications of these 
grains, we have found that there is often considerable difficulty 
in recognising them with certainty. This is chiefly because 
none of the individuals show any signs of bodies which have 
taken up the Giemsa stain in the same intense manner in 
which the above-described granules stain with iron-hema- 
toxylin. In individuals which do show granules that can be 
reasonably identified with those, the granules are markedly 
smaller in size. Hence care has to be taken not to be misled 
by stray, reddish-staining grains of the ordinary chromatoid 
character, of which occasionally one or two are present in 
the cytoplasm. Making all possible allowance for such, we 
do nevertheless find in some parasites one or two definite 
granules, situated close to the cytoplasmic vacuole, in a posi- 
tion similar to that often occupied by the granules in the iron- 
hematoxylin smears, which there can be little doubt actually 
correspond to those bodies. They stain dark-reddish in most 
smears, about the same colour as the chromatin-masses of 


120 BE. A. MINCHIN AND H. M. WOODCOCK. 


the nucleus, whereas isolated chromatoid grains are a fainter 
red. Figs. 16, 17 show instances of which we feel fairly 
sure, as also figs. 4,25 of a large form. In the latter the 
two granules contrast distinctly with the mass of chromatoid 
substance, of which we have more to say below. ‘Their posi- 
tion in this case, some distance from the nucleus, agrees 
closely, it will be noticed, with that of the one grain in the 
iron-heematoxylin individual drawn in fig. 49. On the specially 
thin smear, already alluded to, which was only lightly stained, 
most of the parasites show no signs of these granules; only 
in one or two individuals is a round, faintly pink-staining 
body present, which probably represents one (figs. 8, 9). 

There is another point to mention in this connection. In 
Giemsa-stained smears, parasites which show these granules 
clearly are much scarcer than in iron-hematoxylin smears. The 
explanation is probably as follows: ‘lo judge from the iron- 
hematoxylin smears, in certain of the parasites the grains (or 
grain) are more or less separated from the uucleus (cf. figs. 44, 
46, 49), but in others they are close to it and may be in 
contact with it. Hence in the Giemsa films, where the nucleus 
is so obviously enlarged by being overloaded with stain, the 
granules may be swamped, as it were, and not distinguishable 
from the nuclear mass. We may expect, therefore, to be 
able to recognise them only where they are well out of the 
nuclear zone. 

Since the granules appear distinctly smaller in Giemsa- 
stained preparations than they do in iron-hematoxylin ones, it 
is most likely that the former depict more nearly the true size, 
and that in the latter there is some overloading, due to the 
strong affinity the granules have for the iron-hematoxylin 
stain. This is rather important in considering the significance 
to be attached to these bodies. For if they were of chromatin 
in the ordinary sense—like the grains and masses in the 
nucleus—we should expect to find them apparently quite as 
large in the Giemsa smears as in the iron-hematoxylin ones ; 
since, as is well known, the Romanowsky stain is deposited 
most heavily in and around chromatic structures, with result- 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 121 


ing enlargement of size (cf., for imstance, the kinetonucleus 
of atrypanosome). The fact that this is obviously not the 
case here suggests that the granules are not chromatic 
structures. 

In this connection certain of our Twort preparations are very 
instructive. As has been shown by Minchin (l.c.) the Twort 
stain is in one respect superior even to iron-hematoxylin, 
in that chromatic elements can be distinguished from others 
by the fact that they alone stain red, everything else being 
green. These particular Twort smears were examined soon 
after being made to see if they showed granules corres- 
ponding to the ones we had already found in iron-hematoxyln 
films. In several individuals one or two granules were seen, 
which we regarded as those for which we sought. ‘They 
were small, however, and not particularly conspicuous. They 
were very faintly stained red, nothing lhke so deeply or 
sharply as the grains in the nucleus. Some of the individuals 
showing these granules were noted and sketched at the time, 
and then the smears were put aside to work out on our return 
to England. On finding the same parasites again recently, 
in order to draw them, we conld no longer see the red 
granules in any of the individuals marked. All signs of 
them have vanished, although the red of the nucleus has not 
faded at all. In one or two cases, however, in about the 
position which was occupied by the red granules (according 
to our sketches), small rounded areas, somewhat diffuse in 
outline, can be made out, staining a rather deeper green 
than the surrounding cytoplasm (cf. fig. 59). It may be 
that these greener areas mark the position of the structures 
which stain so intensely withiron-hematoxylin. Inthe great 
majority of the individuals stained with 'wort, however, 
the body is uniformly pale green in colour, and cannot be 
said to show any indications of the granules. 

To sum up, we regard the above-described characteristic 
bodies as composed, at any rate, chiefly of achromatic material. 
Our opinion is based on the one hand upon a comparison of 
their staining reactions to iron-hematoxylin and to Giemsa, and 


122 E. A. MINCHIN AND H. M. WOODCOCK. 


on the other hand upon the fact that they show very little, if 
any, real affinity for the neutral red of 'wort’s stain. It may 
be pointed out that, besides the fact of their being frequently 
double, these grains differ from kinetonuclei in their staining 
reactions, After Giemsa, they appear only reddish and have 
not the characteristic dark purple or almost black colour 
associated with true kinetonuclear elements; compare, for 
instance, the intensely staining chromatic grain described by 
Woodcock (28) in the case of a Halteridium of the 
chaffinch, which is quite distinct from the ordinary nucleus. 
Again, their appearance after T'wort’s stain shows no resem- 
blance at all to that of kinetonuclei. With regard to the 
staining of these grains by wort and iron-hematoxylin, one 
of us (K. A. M.) has found an interesting parallel in the case 
of the blepharoplast (basal granule) of the flagellum of Try- 
panosoma lewisi. In the multiplying (not the adult) 
forms of this parasite the blepharoplast appears after T'wort’s 
stain as a diffuse green patch, but after iron-hematoxylin it is 
seen as a definite black granule. ‘This comparison suggests 
that the grains seen in H. rovignensis are centrosomic in 
nature, but apart from the above facts we have no further 
evidence to bring forward in support of this conclusion.! 

We pass now to the consideration of the large forms of the 
parasite. These are, as already mentioned, of two distinct 
types: (1) Long, comparatively slender individuals, often 
slightly curved or bow-shaped, which possess a small nucleus 
(figs. 18-21); and (2) broad forms, oval or bean-shaped, 
which have a much larger nucleus (figs. 22-26). In the wide 
forms both ends of the hemogregarine are generally similar 


1 We have noticed one or two references to the occurrence of bodies 
in other hemogregarines, which may perhaps relate to a similar 
organella. Thus Miss Robertson (18) describes and figures two large 
oval “red bodies,” staining red with Giemsa, in H. vittatz; and again 
(19) she mentions the occurrence of an eosinophile body or vacuole, 
beside which is a sharply staining grain, in a hemogregarine from 
Pleuronectids. Whether, on the other hand, the “ Plastinkerne” 
described and figured by Prowazek (16) in H. platydactyli also 
represent a corresponding body appears more doubtful. 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 123 


and more or less bluntly rounded. In the slender forms, on 
the other hand, one half of the body is usually narrower than 
the other and tapers towards its extremity, the opposite end 
being bluntly rounded. ‘The more pointed end may be curved 
or partially bent up on itself (figs. 18, 21), this being most 
likely due to limitations of space; the drawn-out corpuscle is 
apparently incapable of being stretched to the full extent of 
the length of these forms. (We have not seen any phases in 
this parasite, it should be said, indicating the development of 
a definite U-form such as occurs in many hemogregarines.) 
The average size of the slender type is 12m by 2-1; and 
individuals up to 144 long have been observed, the breadth in 
this case being 2°24. ‘he stout forms average 10°6 u by 3°4 pu, 
and their extreme size is about ll « by 3:5. The size of the 
nucleus differs greatly in the two cases, and this is a constant 
feature. In Giemsa smears the nucleus of the slender type 
averages 2°6 « by 1:7 w and that of the wide type 48 mu by 
2°9 w; in iron-hematoxylin preparations the former is 2°5 wu 
by 1:4 and the latter 4 by 2°24. In the slender forms the 
nucleus is always in the narrower, tapering half. 

The structure of the nucleus appears practically the same 
in both the stout and slender individuals ; it agrees closely 
also with the structure of the nucleus in the small forms. 
‘The cytoplasm of these large forms, however, differs greatly 
in character from that of the small parasites to judge from the 
effect produced by the Giemsa stain. The difference is 
especially noticeable in the case of the slender forms. Here 
the broader half of the body (not that in which the nucleus is 
situated) is nearly always more or less completely filled with 
some substance which stains red, and which may indeed 
appear at times almost as deeply and intensely stained as the 
nucleus (cf. figs. 19,21). Often the colour increases in depth 
regularly towards the broad end, as if the substance which 
attracts the stain were most concentrated in that region of the 
body (fig. 18). From the effect produced by the Giemsa 
stain it would seem as if the mass were of a finely granular 
character and consisted of chromatoid material. Some of the 


124 E. A. MINCHIN AND H. M. WOODCOCK. 


wide individuals show little or no indications of this substance 
(figs. 22, 23), but in others the condition is present to a 
greater or less degree (figs. 24, 25). It is never so prominent 
as it frequently is in the slender type of form. It is some- 
what remarkable that neither iron-hematoxylin nor Twort 
show anything at all corresponding to this appearance so far 
as can be seen. After T'wort’s stain the body of the large 
form either appears finely granular and faintly tinted green 
(tig. 62) or else it is very pale, scarcely, if at all, stained (fig. 
61).' Yet it is quite evident that there must be something 
more than merely the ordinary cytoplasm present to account 
for the appearance seen after Giemsa. 

We are inclined to doubt whether the characteristic granules 
above described occur so frequently in these large forms as 
they do in the small ones. They are not visible, for instance, 
in either fig. 50, of a slender individual, or in fig. 48 of a 
broad one. One granule is seen in the stout form of fig. 49, 
however, and a couple in that drawn in fig. 25. Also in 
the slender parasite in fig. 4 there are a couple, but this is 
the only full-sized individual of this type in which we have 
made them out with certainty. Fig. 45 represents a slender 
form of intermediate size, and this shows a single prominent 
granule. 

Nearly all the individuals of the slender and broad types 
appear to be full grown. We have been able to find, however, 
two or three examples of what are undoubtedly young, 
growing individuals of these forms. They occur in slides of 
the series containing many of the large forms. ‘T'wo young 
parasites of the slender type are seen in figs. 12,13; anda 
young individual which would probably have developed into 
a stout form is drawn in fig. 14. One of the slender forms 
shows very conspicuously the two characteristic granules; 
and in the small broad individual we think it not at all 


1 The shading, often irregular, of the iron-hematoxylin parasites in 
figs. 45 and 49, for instance, is meant to indicate the slightly varying 
thickness of the densely stained cytoplasm of the corpuscle lying over 
the parasite. 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 125 


unlikely that one or both of the grains at the edge of the 
nucleus, on its inner side, may represent the same _ bodies. 
Individuals which are somewhat larger, but still distinctly 
intermediate in size, are seen in figs. 45 and 15, the former 
being of the slender type, the latter of the broad type. 

We have not found any other phases of the hamogregarine, 
besides those above described, in our preparations. In the 
eurnards from which smears of the internal organs were made 
the parasites happened to be nearly all small forms, and 
large forms are very scarce. In these cases, although pre- 
parations from the liver, kidney and spleen have been searched, 
no signs of schizogony have been noticed.! Perhaps if smears 
had been taken from these organs in the case of the gurnard 
in which the large types of form are frequent, multiplication 
phases of a particular kind might have been found. 

It is important to note that we have never found any indivi- 
duals of either of the large types free from the blood- 
corpuscle. Further, in only a solitary instance has one of the 
small forms been noticed free (fig. 5). This occurs on a smear 
from the kidney. The parasite is not very vermicule-like, 
and resembles the small intra-cellular type of form. This 
scarcity of free forms quite agrees with our observations of 
the parasites in the living condition (cf. above, p. 115). 


GENERAL CONSIDERATIONS. 


We have now to consider what is the significance of the 
different phases observed, in what relation do they stand to 
one another, and how do they compare with the known forms 
of other heemogregarines ? 

The ordinary small forms doubtless represent an early 


1 The only possible indication of commencing schizogony which we 
have noticed is found in two or three of the small forms in a prepara- 
tion from the infected gurnard examined. The nucleus of the parasite 
shows a constriction about the middle, which causes it to appear some- 
what dumbbell-shaped (fig. 28). This may, perhaps, signify commencing 
nuclear division, prior to fission of the parasite, but we do not feel at 
all certain upon the point. 


126 E. A. MINGHIN AND H. M. WOODCOCK. 


stage of the infection, and may be regarded as derived from 
sporozoites which have penetrated the blood-corpuscle. They 
are for the most part very uniform in size and appearance 
and probably destined to become schizonts. 

The large forms are, we consider, of two distinct types, and 
not directly connected with one another—that is to say, an 
individual of one kind, e.g. a broad form of parasite, does 
not pass into one of the other kind, the slender type, by a 
process of elongation and simultaneous bending-up, such as 
is described by Borner (4) in H. stepanow1; nor, on the 
other hand, is the broad form to be derived from the slender 
type by a process of change comparable to that frequently 
described among Reptilian haemogregarines, where a U-shaped 
form gives rise, by the fusion of its two arms, to an oval or 
bean-shaped form (cf., for example, H. tunisensis of Bufo 
mauritanicus, Billet [2], H. bagensis of Emys leprosa, 
Billet [3], etc.). Our reasons for regarding the two large 
forms of H. rovignensis as independent are three: In the 
first place, we have noticed no transitional forms indicatmg 
such a connection as just mentioned; secondly, there is 
always a well-marked difference between the nucleus of forms 
belonging to these two types; and lastly, young individuals 
of each type are clearly distinct. We suggest that these 
large forms show sexual differentiation, the slender type with 
the small nucleus being of male character, and the stout form 
with the large nucleus being of female sex. 

With regard to their origin, it is very probable that they 
have been developed from two forms of merozoites, smaller 
and larger, for the young individuals of the slender or male 
type in figs. 12, 13 are manifestly different from the other 
small forms, and not likely to have arisen from them. There 
is one point, however, in this connection which at first was 
not at all clear. In the second gurnard examined (figs. 12— 
14, 16-26, 44-50) the number of the young parasites found 
belonging to the type which we regard as male is very small 
indeed as compared with the number of the ordinary small 
forms present, whereas about equal numbers of the full-grown 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 127 


parasites of both types occur, and these are fairly plentiful. 
We think the most probable explanation is that we have to 
deal here with more than one infection. The great majority 
of the small forms must be regarded as the early stages of a 
new or recent infection; the forms showing sexual differen- 
tiation, on the other hand, belong, we think, to an older 
infection. Most of them occur as adult large forms, and only 
a few are to be met with as young or intermediate-sized indi- 
viduals. The young parasites of the female type (belonging 
to the older infection) are probably not very different in 
appearance from the numerous small forms of ordinary type 
(of the recent infection), and thus are only distinguishable 
where they are beginning to grow into the stout type (fig. 14). 
The two forms of merozoites, which give rise in time to the 
large individuals, have doubtless been developed by a schizo- 
gonous process, probably occurring in the internal organs. 
The large forms themselves must represent one of two 
phases, schizonts or gametocytes respectively. If, according 
to the first of these interpretations, we have here micro- and 
macro-schizonts which will give rise again to a fresh series of 
micro- and macro-merozoites, it is evident that sexual dimor- 
phism is apparent throughout the schizogonous series of 
generations (as is known to occur among certain Coccidia, 
e.g. Adelea, Cyclospora). If, on the other hand, our 
two types are micro- and macro-gametocytes respectively, we 
have no indications with regard to the nature of the ordinary 
schizogony (or fission) in this parasite, which may be very 
likely all of one kind, that is to say, “ indifferent” in 
character, with no sexual dimorphism manifest (as in many 
Coccidia and all known Heemosporidia). In this case it will 
only be in the last generation produced by schizogony that 
sexual dimorphism appears, in the formation of what are 
really the young micro- and macro-gametocytes (cf. above). 
Unfortunately, from our own preparations alone, we cannot 
1 The process is, perhaps, comparable to the formation of merozoites 


ot two sizes in cysts of Karyolysus (cf. Labbé [5}), both as regards 
the development and the significance of the small elements formed. 


128 Ek. A. MINCHIN AND H. M. WOODCOCK. 


pronounce definitely between these two views. When, how- 
ever, we compare the facts which we have learnt concerning 
H. rovignensis with what is known in the case of other 
piscine hemogregarines, we are strongly inclined to consider 
the latter view the correct one, and that the two distinct 
types of form are micro- and macro-gametocytes. Fission or 
schizogony is now known to occur in many piscine heemogre- 
garines, and in most of the instances described it appears to 
be of one kind, no indications of sexual dimorphism being 
mentioned (cf. H. bigemina, Laveran & Mesnil [6], H. 
quadrigemina, Brumpt & Leb. [9, fig. 3, p. 382], H. 
simondi, Lav. & Mesn. [6], etc.). Neumann, in his account 
of piscine hemogregarines (15), regards the schizogony in 
these instances as resulting in the formation of “ gametes,” 
but does not attempt to explain why only one kind is described 
and figured. Further, in his account of H. polypartita 
from Gobius pagenellus, he regards similar crescentic 
forms, four of which are developed in a_blood-corpuscle, 
also as “ gametes.’ He endeavours to show that these forms 
exhibit sexual differentiation, but such distinctions as are 
apparent in his figures seem to us to be due merély to slight 
differences in size (or age) and in tint of colour (otherwise 
degree of staining). We do not find anything approaching 
the pronounced and constant differences, both in form and 
in the size of the nucleus, which are shown by the large 
types of H. rovignensis. We certainly consider the curved 
forms of H. polypartita—equally with those of Neumann’s 
other new species, H. clavata—as 
really “indifferent ”’ in character,’ and quite comparable with 
those described in the above-mentioned parasites. 


‘ 


‘ merozoites,” probably 


1 It is quite possible, of course, that in some hemogregarines the 
schizogonic forms exhibit sexual differentiation. Up to the present, 
however, we do not consider this has been shown to be the case. 
Wenyon (22) has figured “ barillets” both of micro- and mega-mero- 
zoites in connection with H. gracilis from the liver of Mabuia. 
Certain of Wenyon’s figures suggest strongly the schizogony of a 
Coccidian, and we think this explanation is not at all unlikely, both in 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO, 129 


There is another point which is of considerable importance. 
In all the above instances these long, slightly curved forms 
(adult merozoites) were readily observed free, as ‘‘ vermicules,” 
in smears as well as in the living condition, This is quite 
natural if, as we consider, these vermicules are ready to 
penetrate a fresh host-cell, probably in an internal organ, and 
there give rise by schizogony to the sexually differentiated 
forms. We think it will be useful to distinguish these free 
individuals as schizokinetes, meaning thereby a_ tem- 
porarily motile schizont. 

In striking contrast to the above cases we have the entire 
absence, so far as we are aware, of free individuals of either 
of the large types of H. rovignensis, although the great 
majority of them seem to be full-grown and mature. We have 
no grounds whatever for thinking that these individuals 
become free from the corpuscle while still in the fish. 

This difference in behaviour also points to these large forms 
of the Trigla-parasite being gametocytes and not schizo- 
kinetes. On the supposition, which is most probable, that 
the sexual process of these hemogregarines takes place in a 
leech, these is no reason why we should expect to find gameto- 
cytes liberated in the blood of the fish, because they are await- 
ing the stimulus of the invertebrate host before being set free 
from the corpuscles. We may compare in this respect other 
intra-cellular blood-parasites, for instance, Halteridium, 
Proteosoma, or Leucocytozoon of birds, with which we 
have had much to do. If a drop of blood containing these 
parasites is taken and smeared quickly, scarcely any of the 
ripe gametocytes present have ruptured the host-cell and 
become rounded off. It is only when the blood is allowed to 
cool for a short time in the living condition that the sexual 
forms become free, as indeed is well known. It may be said 
that if the large forms of H. rovignensis are gametocytes, 
we ought to have seen some of them at any rate become free 
in the living preparations examined. ‘This does not follow at 
view of the situation and also on account of the nuclear structure (cf. 
also below, p. 149, footnote). 

VOL. 55, PART 1.—NEW SERIES. 9 


130 E. A. MINCHIN AND H. M. WOODCOCK. 


all, for the particular stimulus which effects the liberation of 
these elements in the above cases is here lacking, namely, the 
fall in temperature. Lastly, it is not out of place, perhaps, 
to refer in this connection to Miller’s account of “ Hepato- 
zoon” in rats (10). In this case the sexual forms (which 
apparently show little or no differentiation) are encysted in 
lymphocytes. Miller found that when blood containing the 
parasites was mixed with the expressed juices of the mite 
(Lelaps), it was ten to thirty minutes before the host-cells 
were dissolved and thirty minutes or more in addition before 
the gametocytes were liberated from their capsules and 
became motile vermicules. 

From all these facts we conclude that where large free 
vermicules of a hemogregarine are found in the circulating 
blood, at least in fishes, they are schizokinetes, which 
have yet to give rise to the true sexual forms, and not them- 
selves the gametocytes (‘‘ gametes” according to Neumann). 
It is very doubtful whether the full-grown gametocytes of 
hemogregarines ever become free until the blood is drawn 
from the body.? 

Characteristics of H. rovignensis nobis.—The prin- 
cipal characters of this species from Trigla lineata, so far 
as we have been able to ascertain them, are as follows: A very 
small parasite, one of the smallest piscine hemogregarines yet 
described. Schizonts usually oval in form. Average size 
(full grown ?), before fission has commenced, 4°84 by 2°5 pw. 
Gametocytes large and well differentiated. Female forms 
wide, ovoid or bean-shaped; average size 10‘6m by 3°4y. 
They possess a large nucleus. Male forms fairly slender, with 
one end somewhat club-shaped, the other end usually slightly 
recurved ; average size 12 4 by 2°lj. These possess a small 
nucleus. Individuals of all types may show one or two 
characteristic granules, extra-nuclear in position, and most 
probably achromatic in nature; they are particularly promi- 

1 From the published descriptions and figures of reptilian hemo- 


gregarines which we have seen we consider it most probable that a 
similar state of affairs obtains in their case also. 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 151 


nent in parasites stained with iron-hematoxylin, which stain 
they take up with great avidity. 


IJ. Trypanosoma sp. (CF. TRIGLH) IN TRIGLA LINEATA. 
(Figs. 29-31, 51, and 65.) 


In the three gurnards which were infected with Hamo- 
gregarina rovignensis a trypanosome was also found. 
This parasite was always very rare, and never more than two 
individuals could be seen in a cover-slip preparation of living © 
blood. Neumann (15) gives only a very brief notice of 
T. trigle, from a single Trigla corax, and does not 
remark upon the strength of the infection in this particular 
case, but he states at the beginning of his paper that as a 
rule trypanosomes appear to be very rare in the blood of 
infected fish. 

Observed in the living condition, the trypanosomes per- 
formed very active movements of contortion, but did not 
actually displace themselves much in the blood-fluid. The 
movements were of the wriggling and twisting kind so 
characteristic of fish trypanosomes. The body would be 
coiled into a spiral or S-shape and then unbent again only 
to become twisted in the reverse sense with the greatest 
rapidity. Occasionally the parasite would burrow into a mass 
of corpuscles and pass through them: Particular parasites, 
whose position was noted, were found to remain alive from 
five to eight hours in the drawn blood or cover-slip prepara- 
tions. Their movements had become extremely sluggish by the 
end of this time, but no alteration inform was noticed. Only 
in a single case was a parasite seen alive after a longer interval 
(twenty-four hours) ; the body of the trypanosome was then 
bent up and motionless but the flagellum waved feebly, show- 
ing that the parasite was still alive. he parasites did not live 
any better in preparations in which the blood had been mixed 
with a drop of salt-citrate solution or of sea-water. In remain- 
ing alive such a short time in drawn blood this trypanosome 


132 E. A. MINCHIN AND H. M. WOODCOCK. 


differs markedly from T. raizw, as will be seen from our 
statements below. 

Owing to the scarcity of this parasite very few individuals 
are present on our stained preparations. At the most two 
occur on a film; more usually only one has been found, and 
on some smears there appear to be none. All the parasites seen 
belong to one type and show no pronounced variation in size, 
Individuals on “wet” smears are generally a trifle smaller 
than those on “dry” ones; we think the former are slightly 
contracted. ‘The average size of the trypanosome, as seen on 
Giemsa-stained smears, is 59 u in total length (i.e. inclusive 
of the flagellum) by 4°5 » in total width (inclusive of the mem- 
brane). The free flagellum averages 8 » in length. Fig. 29 
shows a typical example of the parasite, with dimensions 
almost as given. The longest parasite observed (fig. 31) has 
a length of 62 4 and a breadth of 4°7 ». The length may be 
really 2 or 3 « longer, as the flagellum is very faintly stained, 
and its free portion probably continues a little farther than 
can be made out. On the other hand the parasite from 
an iron-hematoxylin film drawn in fig. 51 has a total length 
of only 54 4 and a width of only 4 , inclusive of the mem- 
brane; but here also the free part of the flagellum is so 
faintly stained that we cannot be certain its entire length is 
represented. 

All the trypanosomes on our slides belong to this long, 
slender type. The flagellar extremity is narrow and finely 
tapering (figs. 29-31), a feature which is more usual in these 
parasites of marine fishes than in those of fresh-water ones. 
The distance of the kinetonucleus from this end of the body is 
generally about 6 u, and may be as much as7 yu. The free 
part of the flagellum at the opposite end is comparatively 
short, and varies from 6 to 10 uw. The trophonucleus is 
generally in the flagellar half of the body (figs. 29-31), but 
may be occasionally more centrally placed (fig. 30). The 
undulating membrane is well developed. It appears under 
two aspects. In the first, which we think represents the more 
natural condition, it shows well-marked folds and pleats, of 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 133 


which there are six to eight, of varying prominence and 
depth (figs. 29,31). In other cases the membrane appears 
as an extensive flap or fin, with a slightly wavy border, and 
widest about the middle of the body (fig. 30). We think the 
difference between these two aspects 1s more apparent than 
real, due to a large extent to the flattening out of the parasite 
on the slide, in the second case. On the other hand, in the few 


“ wet ” films the membrane 


individuals we have observed on 
appears narrow and inconspicuous (cf. figs. 51, 65), and we 
should certainly be inclined to say it had undergone some 
shrinkage here. 

One or two remarks which we have to make upon the 
nuclear structure may be deferred until we consider the 
trypanosome-nucleus generally in the second part of this 
paper. ‘lhe only detail with regard to the general cytoplasm 
that requires mention is the occurrence in certain cases of 
numerous granules. These granules are only noticeable in 
parasites on Giemsa-stained smears. In some individuals 
(fig. 31) the cytoplasm is quite free from them; in others 
there are a certain number of small dark granules, chiefly in 
that part ot the body lying between the two nuclei. The 
flattened-out individuals, however, are rendered conspicuous 
by the presence of numbers of large granules, apparently 
occupying most of the body (fig. 50). Moreover, many are 
seen lying apparently in the membrane. ‘These granules 
stain a purple colour of a somewhat deeper tint than the 
lilac of the cytoplasm. 

The trypanosome from Trigla corax, to which Neumann 
has given (l.c.) the specific name of trigle, is described 
very briefly and without any figures. According to Neumann’s 
account the parasite is of a different type of form from that 
which we have above described. Its total length is about the 
same, namely, 60 uw, but the free flagellum is rather longer, 
being 15 u. The chief difference, however, is in the breadth, 
which appears to be much greater. Neumann gives the 
width of the body as 8 mw and that of the membrane as 4 p, 
the entire width being thus 12 4 as compared with an average 


134 E. A. MINCHIN AND H. M. WOODCOCK. 


of 4°5 w in the case of our form.) Further, the aflagellar 
end of the body is short and somewhat blunt, as indeed is 
often the case in these ‘‘stumpy ” types of trypanosome, and 
the membrane does not show well-developed folds. 

At first sight these two trypanosomes, from different 
species of Trigla, might be considered to belong without 
doubt to different species ; and probably many authors would 
not hesitate to give both parasites a distinct name. In our 
opinion this would. be decidedly premature, for we think it 
quite likely that both are merely different forms of one and 
the same species. Polymorphism is now known to be of 
common occurrence among trypanosomes.  'l'’o give only one 
or two instances, Minchin (11) has recently shown clearly, by 
photographs, the marked (sexual) polymorphism in ‘I’. gam- 
biense, while one of us (H. M. W.) has been struck by the 
polymorphism, of a character quite similar to that implied in 
the case before us, which occurs in an Avian trypanosome. 
We do not intend here to ascribe any special sexual signifi- 
cance to the “stout” (wide) form of the Trigla trypanosome ; 
we suggest, however, that Neumann’s form isa particular type 
of the parasite we have described above, the latter being the 
one which we are inclined to regard as the more “ ordinary ” 
type. Hence our reasons for including the trypanosome 
trom T’rigla in the species T’. trigle. 

In connection with this point, it may be remarked, it is 
especially among piscine trypanosomes, where the parasites 
attain to such large dimensions that marked variations in 
form and appearance may be expected to occur, due either to 
young forms or to different types of the parasite. It seems 
to us that there has been too much tendency to ascribe hard 
and fast limits to the size of a specific trypanosome. Many 
authors, in describing new species, appear to have overlooked 


1 It is quite possible that the real width is not so much as these 
figures indicate ; for we have found that it is especially in such stout or 
stumpy types that flattening-out in dry smears may be most appreciable 
and most liable to give an incorrect idea of the true width of the para- 
site. 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 155 


the possibility of considerable variation in size and appear- 
ance, and as a result have given fresh names to parasites 
which are in all probability only phases of trypanosomes 
already kuown from the same or closely allied hosts. We 
have little doubt that Neumann has made such a mistake in 
distinguishing his “Trypanosoma variabile” from '’. 
rai. of skates. 


III]. T. nar# rrom Rata sp. 
(Figs. 832-38, 52-57, and 66-68.) 


As this parasite has been described already by several 
workers, we need not give a general account of it here. One 
or two points of interest have been observed, however, which 
may be mentioned. In the fish examined the trypanosomes 
were exceedingly abundant, but no dividing forms were seen, 
nor were any hemogregarines found. 

With regard to the behaviour of the trypanosomes in 
freshly taken blood under a cover-slip, we found that they 
remained alive and active for a much longer period than did 
T’. trigle kept in a quite similar condition, Many were 
seen quite unaltered, but undergoing less active moments, 
after twenty-four hours. Also after fifty-four hours several 
were seen, their movements being now sluggish. At the end 
of seventy-two hours only four were found, two of them 
being individuals which had been noted and marked after 
twenty-four hours. They showed no alteration, but their 
movements were very slow and feeble, being contined to little 
jerks of the flagellum and a very slight twisting of the body. 
Three of these individuals were seen again on the fourth day, 
after ninety-six hours, when they appeared in much the same 
condition. Lastly, on the fifth day one trypanosome only 
was still seen living, extremely feeble and moving very slightly 
at intervals. The remarkable point is that none of these para- 
sites showed any alteration in form; nor during the earlier 
periods, when several individuals were still alive and fairly 


136 K. A. MINCHIN AND H. M. WOODCOCK. 


active, did we notice anything comparable to the rounded-oft 
phases described by Miss Robertson (17). It is probable that 
the percentage of individuals which undergo this alteration 
on the slide is very small compared with the number that do 
so when the parasites pass into the leech. 

The trypanosomes in our permanent preparations show 
considerable differences in size. On Giemsa-stained smears 
the largest individual observed has a total length of 72 wand 
a breadth of 5°6 u, including the membrane. The correspond- 
ing dimensions of the smallest form seen are only 55 a by 4 
(fig. 33). These two extremes are connected by intermediate 
forms of varying size (cf. figs. 34, 35). The average size 
works out at about 67 « by 5:2 yu. The free flagellum varies 
from about 10 » to 15 w, with an average length of 13 p. The 
length of the flagellum does not seem to stand in any very 
close relation to the size of the parasite, and now and again 
is shorter in a large individual than in one of intermediate 
size. The largest trypanosome noticed on “wet” films 
stained with iron-hematoxylin is drawn in fig. 53; it isa com- 
paratively wide, plump individual, which would probably 
have seemed even wider on a Giemsa-stained smear. It 
measures 65 « by 6°6 u. It is probably somewhat longer in 
reality, for the free flagellum, which is unusually short in the 
drawing, comes into contact, at the point where it apparently 
ends, with a corpuscle which is stained deep black ; although 
it probably runs across this for some distance, its course 
cannot be followed. Other parasites on iron-hematoxylin- 
stained films are seen in figs. 52-56. They are mostly a little 
shorter than the parasites on ‘“ dry ”’ films; this difference is 
most noticeable in comparing the body-protoplasm, for the free 
flagellum itself is in most cases actually longer and averages 
14°8 y against 13 4 on the dry smears. We are inclined to 
think this is due to the contraction of the general cytoplasm 
to a greater extent than the entire flagellum (i. e. the flagellar 
border + free flagellum) in iron-hematoxylin tilms. 

There is another rather interesting point brought out by a 
comparison of the trypanosomes fixed and stained by the 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 137 


two methods, which, incidentally, may also help to explain 
this difference in apparent length. ‘lhe great majority of the 
individuals on the “ wet,” iron-hzmatoxylin-stained films are 
in a different position from those on the “ dry,” Giemsa-stained 
smears. In the former they are usually found in a twisted S 
or corkscrew-like position (figs. 52-56), while in the latter 
the parasites are nearly always simply rolled or coiled up to 
a greater or less extent (figs. 32-37). Now, in life the 
trypanosomes are generally observed in a twisted or S-shaped 
condition, and only rarely, and as it were transiently, in the 
form of a simple ©. We may conclude, also, that death and 
fixation are at least as instantaneous in the case where the 
parasites actually come into contact with sublimate and acetic 
as in the cause where the slide is placed in a tube containing 
osmic vapour. Hence we consider that the position of the 
parasites on the wet films approximates most nearly, as a 
rule, to that in which they were the instant before death. 
What is the cause, then, of the parasites assuming the very 
different rolled or coiled-up form on “dry” films ? 

The manner of attachment of the undulating membrane to 
the body has an important bearing upon this question. Figs. 
52, 638, and 56, from iron-hematoxylin slides, show very 
clearly that in these individuals the undulating membrane 
was wound spirally round the body at the instant of death. 
The flagellar border runs now under, now over the general 
cytoplasm, and in fig. 56 it is seen to run twice under. Are 
we to regard the membrane as actually attached spirally to 
the body (when the latter is in a “ passive ” condition), or as 
being merely twisted round it at the time by the voluntary 
contortion of the protoplasm? We think the latter view 
affords the true explanation. ‘The appearance of the parasites 
on Giemsa-stained films gives probably a fairly correct repre- 
sentation, froma morphological point of view, of the manner 
of attachment of the membrane—that is to say, it hes along 
one side of the body, more or less in one plane. The 
membrane itself, especially on its outer side, is longer than 
that part of the body to which it is attached. In life the 


138 E. A. MINCHIN AND H. M. WOODCOCK. 


membrane is usually twisted in a spiral fashion round the 
body by the voluntary contortion of the protoplasm, this 
being in all probability effected by the contraction of 
myonemes. Minchin (18) has recently published figures 
clearly showing myonemes inl’. perce and'!’. granulosum, 
and we have no doubt they are present in other fish-trypano- 
somes, though we have not had the good fortune to see 
them in ‘I’. raizw. In wet films the parasites have retained 
their twisted position. In “dry” smears, on the other 
hand, the body becomes untwisted, and, at the same time, 
passively or mechanically coiled up in one plane, by the 
mere fact of the attached membrane being longer than the 
body is. 

We do not think this different behaviour on “dry ” smears 
is to be explained by a flattening-out process due to actual 
drying. In the first place, in our procedure, the slides are 
removed from the osmic-acid tube and placed in absolute 
alcohol before the moisture dries off from the greater part, 
at all events, of the slide; it is only along the edges that 
drying sometimes occurs. And after the smear has been 
hardened in alcohol little or no alteration, we consider, takes 
place in the form of the parasites, even though the smear is 
allowed to dry off ultimately. In fact, as Minchin has 
already shown in his account of the technique in connection 
with ‘I’. lewisi (14), this method is probably the best for 
the general form and size of the parasites. Secondly, now 
and again where the body of the parasite really appears to 
be somewhat flattened out due to an actual drying at first, 
this C-form is not shown (cf. fig. 30 of T. trigie). Indeed, 
this process of untwisting and coiling would seem to require 
the presence of a film of moisture for its accomplishment. 
The following explanation appears to us the most probable. 
In fixation by the “wet” method, both death and fixation 
are practically instantaneous. In fixation by osmic vapour, 
on the other hand, death probably occurs appreciably before 
fixation. In the twisted condition, during life, the flagellar 
border of the membrane is probably to a certain extent in 


BLOOD=-PARASITES OF FISHES OCCURRING AT ROVIGNO. 189 


a state of tension, from which it relaxes, in virtue of its 
elasticity, on the death of the body; in so doing it auto- 
matically unwinds the body, at the same time causing it to 
become more or less C-like, before actual fixation occurs. 
In this connection it should be pointed out that Danilewsky 
(4a), who studied trypanosomes carefully in the living 
condition, frequently figured them with the undulating 
membrane spirally wound round the body, but in some cases 
he shows it attached along one side of the body. 

Lastly, if, as we have been led to consider, the parasites on 
wet films are generally in a spirally twisted condition, we 
might expect to find a slight shortening in length; this, 
together with a certain amount of contraction due to shrinkage, 
would be sufficient to explain the differénce in average length 
between the parasites on wet films and those on dry ones. 

In many of the parasites on Giemsa-stained smears numbers 
of small bodies occur, which appear to be prominent granules 
(fig. 37). They are deep black at the middle focus, but are 
bright and glistening at the upper focus. They are not com- 
parable to ordinary chromatoid granules, which stain more or 
less redin colour. Moreover they are most abundant in the 
aflagellar part of the body, especially between that extremity 
and the kinetonucleus, a region which is generally free from 
chromatoid grains. ‘lhey are also scattered throughout the 
body, and some, which cannot be distinguished by appearance 
from the others, lie occasionally in or on the undulating mem- 
brane. In wet films, stained either with iron-hematoxylin or 
with wort, the same bodies, if present, are not nearly so con- 
spicuous. In the body generally no sign of them is to be 
seen; but near the aflagellar end, which is often slightly 
vacuolated in character, a certain number of granules can be 
seen, not stained very differently from the cytoplasm (figs. 
54-56). We are not sure, however, if these granules are the 
same. 

Returning to the parasites on dry smears, we have recently 
noticed the peculiar fact that, simce the individual of fig. 37 
was drawn, all the black granules have vanished, leaving only 


140 iE. A. MINCHIN AND H. M. WOODCOCK. 


small, clear areas, like spaces, in the position in which they 
were. In fig. 34 is another parasite which showed originally 
a very similar condition as regards the granules ; this has been 
drawn since they disappeared. ‘Two or three of the granules 
are still seen, and the small vacuoles indicate the position 
originally occupied by many others. It seems most probable 
that these black “ granules” are really minute globules of 
oily or fatty substance, which are blackened by the osmic 
acid used in fixation, and are liable to be dissolved away by 
the frequent washings with xylol given to the slide, of course 
after immersion-oil has been upon it. 

We entertain no doubt that this trypanosome belongs to 
the species T’. raiz, Lav. and Mesn. ‘These authors, in des- 
cribing this species, gave its size as from 75 4 to 80 uw in total 
length, by about 6 « in width (inclusive of the membrane). 
Apparently, as has been so often the case, the species was 
characterised solely from the full-grown individuals of the ordi- 
nary type which were encountered, and no reference is made 
to young forms or to variations in type. Further, Laveran and 
Mesnil found trypanosomes which they regarded as belonging 
to the same species in four species of Raia, namely R. 
punctata, R. mosaica, R. clavata, and R. macro- 
rhynchus. We do not know the name of the species of 
Raia in which we found the parasite. ‘The dimensions of the 
largest individuals we have observed, however, are only very 
slightly less than those above mentioned, and the general 
appearance of the parasites, as shown in our figures, agrees 
so closely with that of the individual figured in Laveran and 
Mesnil’s original account (7) that there is every probability 
that the trypansome is the same in the two cases. Neumann 
(15) has given the name 'T’. variabile to a trypanosome 
fron R. punctata, principally or solely because he has 
found variations in size and form in the parasite which are 
not referred to by Laveran and Mesnil ; though he states more 
than once that his parasite resembles T’. raiz closely and in 
its largest form agrees with that trypanosome. As we have 
described above, we have found forms of T. raize very much 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO, 141 


smaller than the full-sized ones; and many of Neumann’s 
figures of T. variabile resemble strongly these smaller 
forms. In fact, neither from Neumann’s description nor from 
his figures is there any reason to suppose that ‘I’, variabile 
is not a synonym of T. raie. 


CoMPARISON OF tHE NUCLEAR STRUCTURE OF A H&@MOGREGARINE 
WITH THAT OF A T'RYPANOSOME. 


A most interesting and important result of our study on 
the above-mentioned blood-parasites of fishes is afforded by 
the evidence it has given us of the essential difference between 
the nucleus of a hemogregarine and the trophonucleus of a 
trypanosome. ‘This difference is brought out forcibly by all 
the three methods of technique employed, though, of course, 
one method may show a certain feature or detail better than 
another, 

Considering first the case of Hemogregarina rovig- 
nensis in Trigla sp., the conclusion arrived at by comparing 
and combining the impressions given by different stains is 
that the nucleus in this parasite consists of a regular or 
irregular meshwork or reticulum, itself chromatic or impreg- 
nated with chromatin, on which are suspended chromatin 
grains and masses of varying size and form. The reticular 
ground-work is best seen in iron-hematoxylin or T'wort pre- 
parations (figs. 39-50, 59-64); in the latter it is always 
distinctly red (chromatic) in colour. The limit or border of 
the nucleus appears to be itself part and parcel of the 
reticulum, the peripheral segments of the latter being usually 
arranged so as to give a fairly regular oval contour or 
“membrane.” ‘his structure is well shown in figs. 48 and 
49 of large forms. Hence one cannot speak here of a true 
nuclear membrane as something distinct and separate from 
the general nuclear substance. This ‘‘membrane”’ also has 
numerous chromatic granules strung upon it; these are 
generally smaller than those occurring in the more central 
parts of the reticulum. The chief chromatic aggregations 


142 E. A. MINCHIN AND H. M. WOODCOCK. 


sometimes tend to run together, or to lie in short streaks. In 
no case have we observed any signs whatever of a definite 
central body or karyosome in the nucleus. 

In individuals stained very faintly and sharply on the 
particular Giemsa-stained smear to which allusion has been 
made several times, the above-described characters of the 
nucleus can be made out quite well (cf. figs. 5-10). In such 
cases the picture represents fairly accurately the true con- 
dition. In other individuals, however, slightly more deeply 
stained, the nucleus appears more granular and already some- 
what “blotchy”; this is due to the enlargement of the 
chromatic grains and to the deposition of the red stain in the 
nuclear sap, more or less occluding and obliterating the 
reticulum, This leads on naturally to the appearance generally 
seen in deeply stained Giemsa smears of a dense mass, staining 
red or purple, in which bodies and streaks still more darkly 
coloured can be made out, representing the chromatic grains. 

We may add that we have been struck by the considerable 
resemblance between the nucleus of the parasite and that of 
its host-cell; this will, indeed, be apparent from many of the 
figures (fig. 14). 

Turning now to the trophonucleus of a trypanosome, we 
find a remarkable contrast. T. raiz being a very large 
trypanosome and possessing a correspondingly large nucleus 
is a most advantageous form to study for this purpose, since 
the various nuclear details—particularly of the karyosome— 
can be made out more readily than in the case of a compara- 
tively small parasite, such as T. lewisi, for example. 

Our description is based upon the appearances yielded 
after iron-hematoxylin and Twort, for in this case—far more 
so than when considering the hemogregarine nucleus—it 
would be difficult, if not impossible, to arrive at what we 
regard as the correct interpretation of the nuclear structure 
by studying the Giemsa appearance alone. Having obtained 
a fairly accurate idea of the nuclear constitution from iron- 
hematoxylin and Twort preparations, we can then interpret 
the widely different picture seen after Giemsa. Miss Robertson 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 145 


(17) has recently described the nuclear structure in certain 
developmental forms of a trypanosome occurring in Pontob- 
della, which trypanosome she regards (and we think correctly) 
as T. raiz. Our own observations quite agree with her 
account; we are able, perhaps, to add a few more details 
with regard to the karyosomatic mass. (Our main purpose, 
as we have already said, is to emphasise the contrast between 
this type of nucleus and that of a hemogregarine.) 

The nucleus is very generally oval in shape and always 
possesses a well-defined, regular contour. Its size varies not 
inconsiderably (cf. figs. 52-56), and, as might be expected, is 
im accordance with the size of the parasite, small- or inter- 
mediate-sized individuals having a smaller nucleus than the 
large ones. ‘he size may be as small as 2°44 by 1°71, or as 
large as 3'2 yn by 26. The greater part of the nucleus is 
the karyosomatic 


occupied by aprominent deep-staining body 
mass. Around this appears a practically clear space, which 
is bordered or limited by a sharply marked line, the nuclear 
membrane. Any space or halo surrounding this on the outside 
again, as sometimes occurs (cf. figs. 52 and 54), is most likely 
a shrinkage-space, Delicate rays, sometimes four or five in 
number, sometimes more, proceed from the central mass to 
the membrane; these are usually very faintly stained, but 
can be made out with a good illumination, especially in iron- 
hematoxylin preparations. Both membrane and rays are 
always green after Twort (figs. 66, 67, 68a). Hence we may 
regard them as achromatic in structure. The rays are pro- 
bably comparable to a linin framework for the support of 
the karyosome. The membrane is a much more definite 
structure than in the case of the hemogregarine-nucleus. 
In this parasite both membrane and rays appear to be, as a 
whole, remarkably free from chromatin, very different in this 
respect from the chromatic reticulum of the hemogregarine. 
The only possible indications of chromatin are furnished by 
small dots or condensations at the junctions of the rays with 
the membrane; they are best seen in iron-hematoxylin pre- 
parations (fig. 54). We can get no evidence of a red colour 


144 BK. A. MINCHIN AND H. M. WOODCOCK. 


at these points after Twort, however, and so do not feel at all 
certain that they are chromatin. 

A correct interpretation of the characteristic central body 
in the nucleus is best gained from Twort films, Iron-hzema- 
toxylin films must be very well extracted, and then the same 
or a similar condition is revealed. But in those iron-hzma- 
toxylin smears where the whole karyosomatic mass is stained 
almost uniformly black (as in figs. 53 and 56 for example), it 
is safe to say that an excess of stain still prevents the details 
from being apparent. The true structure appears to be as 
follows: In the centre is a fairly large, clear region, oval or 
rounded, which is stained grey in iron-hematoxylin films and 
a pale green (distinctly paler than the rays) in ‘l'wort prepara- 
tions. This is the basis or ground-work of the karyosome and 
is probably of a plastin-like nature. The chromatin is located 
at the surface, or at any rate in the peripheral region of this 
plastinoid basis. In the smaller nuclei the chromatin is mostly 
in the form of granules or small masses of varying number 
(usually three to five) and size, and more or less separate from 
each other (figs. 55, 57a, 68a and b) ; but in the large nuclei 
we frequently find the chromatin forming a complete zone or 
ring around the paler area, with thickenings or bulgings here 
and there (figs. 57, f and g, 68, c, d) corresponding to the 
small masses in the other case. 

One important detail remains to be mentioned, namely, the 
presence of a small, distinct granule in the centre of the 
plastinoid area, which is probably of constant occurrence. It 
is readily made out in T'wort preparations (cf. fig. 67, 68, a—f); 
sometimes it is green in colour, but in other cases the colour 
appears to be a mixture of both the red and the green ; it is, 
however, never of the same sharp red colour that the chromatic 
masses are stained. This granule can be distinguished also in 
the individuals on iron-hematoxylin smears, but not so 
easily. 

Comparing, now, the appearance of the nucleus in Giemsa 
smears, a condition is generally found which at first sight 
seems to be diametrically opposite to that shown by iron- 


BLOOD-PARASITES OF FISHES OCCURRING AT: ROVIGNO. 145 


hematoxylin films—that is to say, centrally or excentrically 
is a comparatively clear, faintly stained area, while all the rest 
of the nucleus is stained red more or less deeply (figs. 34, 35, 
38, a—-d), the periphery, in the neighbourhood of the mem- 
brane, being perhaps darkest. The clear area corresponds 
without doubt to the central part of the karyosome, i.e. to 
the plastinoid basis free from chromatin. Rather curiously, 
the central granule, referred to above as occurring in the 
plastinoid part of the karyosome, is often very conspicuous, 
probably because it is to a certain extent artificially enlarged 
by the stam. The remarkable feature about these Giemsa- 
stained nuclei, and the one which creates such a false impres- 
sion, is that the nuclear sap is often so loaded with stain that 
not only the rays but also the chromatic zone or ring imme- 
diately surrounding the central area is indistinguishable as 
such. Occasionally, in more favourable pictures, the chro- 
matic zone is more deeply stained than the nuclear sap and 
can be distinguished somewhat better (fig, 37) ; and now and 
then coarse indications of the rays proceeding to the periphery 
can also be made out (fig. 38, b,c). Hence we have little 
doubt that here also the structure of the nucleus agrees really 
with that above described. 

Owing to the scarcity of Trypanosoma trigle in our 
preparations the few individuais present on wet films do not 
show the nuclear structure very satisfactorily. Extraction 
had to be carried on quite in the dark, asit were, and neither 
in the individual drawn in fig. 51 from a film stained with 
iron-hematoxylin, nor in that of fig. 65, from a preparation 
stained with Twort, has the extraction been carried far 
enough. From these two examples, however, it is quite 
obvious that the nucleus is of the same karyosomatic 
type, and fig. 51 affords indications that the structure of the 
karyosome itself is similar to that above described. 

We regard the above instances as indicative of the typical 
character, speaking broadly, of the nucleus of a hemogre- 
garine and the trophonucleus of a trypanosome. 

So far as the case of the trypanosome is concerned, it is 

VOL. 55, PART 1.—NEW SERIES. 10 


146 E. A. MINCHIN AND H. M. WOODCOCK. 


already quite evident, from figures published during the last 
year or two, since the use of the iron-hematoxylin stain 
became more general, that the trophonucleus is in the main 
constituted on the same plan, having most of its chromatin 
associated with a definite karyosomatic body. Besides Miss 
Robertson (17), Minchin has shown this to be the case both 
in various fish-trypanosomes (18) and also in T. lewisi (14). 
We may mention that during the progress of our work at 
Rovigno we have obtained a similar result in the case of the 
trypanosomes in the little owl; and we see that recently 
Rosenbusch has published figures (20) of cultural forms of 
these trypanosomes (which he calls “ Hwemoproteus 
noctue” and ‘ Leucocytozoon ziemanni”) showing 
this same nuclear structure. 

Further, we are inclined to think that in many cases the 
minute details of the karyosome will be found to be similar, 
that is to say, as regards the tendency of the chromatin to 
be located at the periphery of the plastinoid basis, and the 
presence in the central, clearer zone of a definite granule. 
Of course, nuclei with a large karyosome may be expected to 
show this more distinctly than those with a very small karyo- 
some (such as, for example, T’. lewisi.) One of us (H. M. W.) 
has several times observed, in Giemsa preparations of a 
trypanosome of the chaffinch,! an appearance of the nucleus 
quite similar to that in fig. 37, namely, a conspicuous granule 
occurring in the centre of a clear zone in the middle of the 
nucleus; and the interpretation of the whole appearance is 
doubtless also the same. It is interesting to note that some 
years ago Laveran and Mesnil, in their account of certain 
trypanosomes of fishes (7), published a figure of T. remaki, 
of the pike, which showed the same nuclear appearance. 
We remember thinking this unusual at the time, as it was 
quite different from the uniform granular character which the 
nucleus was generally described as possessing. 

These finer details of the karyosome are best revealed, 


' It is hoped to publish an account of this parasite, and of others in 
the chaffinch, very shortly. 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 147 


perhaps, by wort; in the case of films stained with iron- 
hematoxylin, the stain must be very well extracted, or else 
the whole karyosomatic mass is too heavily stained for them 
to be made out. ‘This is evident by comparing our various 
figures. From many of the figures in the above-mentioned 
papers we should say that extraction in those cases had not 
been carried far enough for this purpose. For instance, in 
most of the resting nuclei of the various trypanosomes 
drawn by Resenbusch (l.c.), the karyosome is too dark to 
show the central granule. ? 

There can be no doubt that this granule or centriole is the 
intra-nuclear centrosome first described by Schaudinn in the 
trophonucleus of his trypanosome in the little owl. It is also 
evident that it acts as a division-centre, and forms an intra- 
nuclear spindle at the commencement of nuclear division. 
This phase is well shown by several of Rosenbusch’s figures. 
Again, to compare a dividing stage described from a Giemsa- 
stained preparation, Minchin, in his account of T’. grayi (11), 
figures an intra-nuclear granule at each end of the spindle 
still connecting two daughter-nuciei, immediately after division 
has taken place. Hence this intra-nuclear centrosome! is 
doubtless a regular constituent of the trophonucleus of a 
Trypanosome. 

It will be noticed from our figures that, in the Giemsa- 
stained preparations of T’. rai, the red-stained part of the 
nucleus is fairly uniform or homogeneous in appearance. It 
is more usual, however, for the nucleus of trypanosomes 
stained by the Romanowsky method to appear granular in 
structure, apparently consisting of small or medium-sized 
eranules in close contact, and forming a compact mass (cf. 
the selected figures, either in the article on “Trypanosomes” in 
Lankester’s ‘Treatise on Zoology,’ or in Liihe’s article in 
Mense’s ‘ Handbuch der Tropenkrankheiten’). This appear- 


1 Moore and Breinl (144) use the term “intra-nuclear centrosome” 
in a different sense from ourselves, namely, for the entire central body 
which we regard as the karyosome. They do not seem to have dis- 
tinguished at all the centriole contained in the karyosome. 


148 E. A. MINCHIN AND H. M. WOODCOCK. 


ance 1s quite easily capable of explanation when the known 
tendency of the Romanowsky stain to be deposited in excess 
around anything of a chromatic nature is borne inmind. We 
may suppose that in such cases there is a certain amount of 
chromatin distributed in the nuclear sap or karyolymph (in 
addition to that associated with the karyosome); this is most 
probably in the form of very fine granules, which are of 
course magnified by the stain to many times their real size, 
Hence the effect is produced of a granular mass, such as has 
been so often described. By this means the clear central 
area, indicating the position of the karyosomatic body, is 
generally occluded completely and indistinguishable. 

We have now to consider, briefly, the hamogregarine- 
nucleus. Here, too, there can be no question but that the 
true nuclear structure is better revealed by stains like iron- 
hematoxylin and Twort than by the Romanowsky method of 
staining. Nearly all the illustrations of hemogregarines 
which we have seen are from parasites obviously stained by 
the latter method. Prowazek, it may be mentioned, in his 
paper on H. platydactyli (16), has three figures which 
were drawn from preparations stained by Grenacher’s hema- 
toxylin, and these also give indications of the same type of 
structure—an irregular reticulum carrying chromatin-grains 
and masses of various sizes—which we have found in H. 
rovignensis. ‘lhe nuclei in these figures of Prowazek’s 
differ very greatly from those he hasdrawn from Romanowsky 
preparations ; many of the latter, we are convinced, do not 
give at all an accurate idea of the nuclear constitution. 

Of all the other figures of hemogregarine-nuclei at which 
we have looked, those of Bérner, in his account some years 
ago (4) of reptilian hemogregarines, seem to convey most 
approximately the true idea of the nucleus. From his 
“Tafelerkliarung” we gather they were drawn from 
Romanowsky preparations ; but for this intimation we should 
have regarded them as from preparations stained by some 
hematoxylin method, both from the appearance of the nuclei 
and from the manner in which the figures are coloured. We 


BLOOD-PARASITES OF FISHES OCCURRING AT. ROVIGNO. 149 


are inclined to think the nuclei in some of his figures may be 
possibly a trifleschematic both as regards the uniform size of 
the granules and the rather suspicious regularity of the 
reticulum; but in many of the other figures there is, in our 
opinion, an indication of the nuclear structure which is pro- 
bably as correct as it is possible to obtain it by the 
Romanowsky method. In none of Borner’s figures is there 
anything remotely resembling a karyosome, and, in fact, the 
author expressly mentions that he never observed such a 
structure in the nucleus.’ 

Numerous figures accompanying the description of new 
hemogregarines have been published during recent years, 
all of them naturally from Romanowsky preparations. — It 
would take too long to cite them; nor is it necessary. It is 
sufficient to say that in no case can the structural details of 
the nucleus be deciphered. Inall cases it is obvious that the 
nucleus drawn was still hopelessly overloaded with stain. At 
the best the nucleus is figured as a dense granular mass, 
bearing often a strong resemblance to those in our figures 
from Giemsa-stained preparations, from which it may be 

1 The only instance of which we are aware, where anything resembling 
a karyosomatic nucleus appears to be present, is in certain figures of 
Wenyon’s (22) on Pl. 12, purporting to represent H. gracilis in the 
liver of Mabuia. The figures are from preparations stained by hema- 
doxylin. We think it most likely that Wenyon has figured besides 
phases of a hemogregarine, also phases of a coccidian, the latter being 
the ones in which the nuclei show a karyosome. His fig. 29 shows 
undoubtedly the development of typical merozoites of a hemogregarine 
(ef. H. simondi); and it is only these merozoites which he figures also 
in the red corpuscles. His figs. 27, 22, and 31, on the other hand, we 
consider, represent a Coccidian ; the two latter especially appear very 
like young coccidian schizonts. Since our MS. was sent to the 
printers the memoir by Hahn (48) has appeared. We can only point 
out here that Hahn uses the term ‘“ karyosome” in a sense quite 
different from that in which we understand the word, namely, to mean 
the entire nucleus when in a condition ‘“ devoid of visible chromatin 
bodies ” (p. 331). He terms such bodies * achromatic nuclei” (which 
seems to us rather a contradiction in terms), and calls them “* karyo-. 
somes, in the sense that they are the bodies from which the chromatin 
bodies subsequently arise.” 


150 E. A. MINCHIN AND H. M. WOODCOCK. 


inferred that its structure conforms to that of H. rovig- 
nensis and to what we consider is the general plan. At 
other times nothing but a “ splotch” of colour, from which it 
is impossible to ascertain anything, is depicted. 

In conclusion we have only to point out that it seems clear 
that the nucleus of a hemogregarine is of a very different 
type from that of a trypanosome. The former is characterised 
by its chromatic reticulum, with chromatin grains or masses 
more or less generally distributed upon it. In the latter the 
greater part or nearly all of the chromatin is, as it were, con- 
densed around a plastinoid basis, the whole forming the con- 
spicuous karyosome ; and in the centre of this plastinoid area 
is a definite granule, the intra-nuclear centrosome. 

So far, therefore, as the hemogregarines at least are 
concerned, we are totally unable to agree with Hartmann (4c), 
who proposes to remove the Hemosporidia entirely from the 
Sporozoa, and place them with the trypanosomes and their 
allies among the Flagellata as a group named Binucleata. 


PostTscRIPT. 


We had not intended to refer in this paper to the nucleus 
of Halteridium. Quite by chance, however, we have 
noticed a couple of sentences at the end of Berliner’s account 
of the cytology of certain Flagellates (1), which relate to 
the structure of Halteridium noctuze and Leucocyto- 
z0on ziemanni, as shown by iron-hematoxylin. There is 
no reference to this point in the title or list of contents, and 
we have only had our attention drawn to Berliner’s figures 
since our present paper was completed. We refer to 
Berliner’s note because we have ourselves obtained similar 
indications of the nuclear structure of these parasites during 
our work at Rovigno. We will only point out here that from 
Berliner’s published figures, and equally from our own pre- 
parations, there can be no doubt that the nuclear structure of 
Halteridium is quite different from that of a hemo- 
gregarine, and, on the other hand, remarkably like that of the 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO, 151 


trophonucleus of a trypanosome, in being of the karyosomatic 
type. Further, Berliner mentions and figures the occurrence 
at times of a distinct extra-nuclear granule, connected by a 
fibril with the main nucleus, which he regards as representing 
the kinetonuclear element of a trypanosome perhaps in a 
somewhat reduced (*‘ riickgebildet”) condition, consequent on 
the intra-cellular, “‘ resting” condition of Halteridium. We 
are very pleased to have this independent confirmation, and 
from iron-hematoxylin preparations, of the occurrence of 
nuclear dimorphism in Halteridium, a feature which one 
of us (H. M. W.) has already described (23) in the case of 
another species parasitic in the chaffinch, though unfortu- 
nately in this instance only Giemsa-stained smears were 
available. ‘There can be little doubt, therefore, that Halteri- 
dium, in regard to its nuclear structure at ali events, shows 
very much closer affinity to the trypanosomes than do the 
hemogregarines. 


LIstER INSTITUTE ; 
November 27th, 1909. 


BIBLIOGRAPHY. 
1. Berliner, E.—* Flagellaten-Studien,” ‘ Arch. Protistenk.,’ xv, 1909, 
p. 297, with plates. 


2. Billet, A—‘ A propos de l’Hémogrégarine du crapaud de I’ Afrique 
du Nord,” ‘ C.R. Soe. Biol.,’ lvi, 1904, p. 482, with figs. 


3. “A propos de l’Hémogrégarine de l’Emyde lépreuse (Emys 
leprosa) de Afrique du Nord,” t.c., p. 601, with figs. 
4, Borner, C.—‘ Untersuchungen iiber Himosporidien, ete.,”’ ‘ Zeit. 


Wiss. Zool.,’ lxix, 1901, p. 398, one plate. 

4a. Danilewsky, B.—‘ Recherches sur la Parasitologie comparée du 
sang des Oiseaux,’ Kharkoff, 1888-89. 

4s. Hahn, C. W.—‘* The Stages of Hemogregarina stepanowi 
Danilewsky found in the Blood of Turtles, with Special Reference 
to Changes in the Nucleus,’ ‘Arch. Protistenk.,’ xvii, 1909, 
pp. 307-376, pls. xvi-xviii. : 

4c. Hartmann, M.—‘‘Das System der Protozoen, ete.,”” ‘Arch. Pro- 
tistenk.,’ x, 1907, pp. 139-158, 3 text-figs. 


152. E. A. MINCHIN AND H. M. WOODCOCK. 


5. Labbé, A.—“ Recherches. . . sur les Parasites endoglobulaires 
du sang des Vertébrés,’ ‘Arch. Zool. Exp.’ (3), ii, 1894, p. 
55, with plates. 

6. Laveran, A., and Mesnil, F.—* Deux Hémogrégarines nouvelles des 
Poissons,” ‘C.R. Acad. Sci.,’ exxxiii, 1901, p. 572, with figs. 


M: “Des Trypanosomes des Poissons,” ‘Arch. Pro- 
tistenk.,’ i, 1902, p. 475, with figs. 
8. ‘Trypanosomes et Trypanosomiases,’ Paris (Masson 


et Cie), 1904. 
9. Lebailly, C.—‘ Recherches sur lés Hématozoaires parasites de 

Téléostéens marins,” ‘ Arch Parasitol.,’ x, 1906, p. 348, with figs. 

10, Miller, W.—‘ Hepatozion perniciosum, n.g., n. sp., a Hemo- 
gregarine,” * Washington Treas. Dept., Public Health, Hyg. Lab. 
Bull.,’ No. 46, 1908, with plates. 

11. Minchin, E. A.—“ Investigations on the Development of Trypano- 
somes in 'T'setse-flies, etc.,” ‘Quart. Journ. Mier. Sci.,’ 52, 1908, 
p. 159, with plates. 


12. “Note on the Polymorphism of Trypanosoma gam- 
biense,” ‘ Parasitol.,’ i, 1908, p. 236, with plate. 

13. * Observations on the Flagellates parasitic in the Blood of 
Fresh-water Fishes,” ‘Proc. Zool. Soc.,’ i, 1909, p. 2, with plates. 

14. “The Structure of Trypanosoma lewisi in Relation to 


Microscopic Technique,” ‘Quart. Journ. Micr. Sci.,’ 53, 1909, p. 
755, with plates. 

144. Moore, J. EK. 8., and Breinl, A.—* The Cytology of the Trypano- 
somes,” ‘Ann. Trop. Med. Parasitol. i, 1907, pp. 441-480, 
pls. xxxviii—xlii, with one text-fig. 

15. Neumann, R. O.—“ Studien iiber protozoische Blutparasiten im 
Blut von Meeresfischen,” ‘Zeit. Hyg.,’ lxiv, 1909, p. 1, with 
plates. 

16. Prowazek, 8. von.—‘* Untersuchungen iiber Himogregarinen,” ‘Arb. 
kais. Gesundhts.,’ xxvi, 1907, p. 32, with plate. 

17. Robertson, M.—‘* Further Notes on a Trypanosome Found in Pon- 
tobdella muricata,” ‘ Quart. Journ. Mier. Sci.,’ 54, 1909, p. 119, 

- with plate. 


18. “A Preliminary Note on Hematozoa from some Ceylon 
Reptiles,” ‘Spolia Zeylanica,’ v, 1908, p. 178, with plate. 
19. * Notes on Certain Blood-inhabiting Protozoa,” ‘ Proc. Phys. 


Soc., Edin.,’ xvi, 1906, p. 232, with plates. 
20. Rosenbusch, F.—‘* Trypanosomen-Studien,” ‘Arch. Protistenk.,’ xv, 
1909, p. 265, with plates. 


BLOOD-PARASITES OF FISHES OCCURRING AT ROVIGNO. 153 


21. Sambon, L. W., and Seligmann, C. G.—* The Hemogregarines of 
Snakes,” ‘ Trans. Path. Soc.,’ lviii, 1907, p. 310, with plates. 

22. Wenyon, C. M.— Report of Travelling Protozoologist,’”’ ‘Rep. 
Wellcome Res. Lab.,’ iii, 1908, p. 121, with plates. 

23. Woodcock, H. M.—“ On the Occurrence of Nuclear Dimorphism in 
a Halteridium Parasitic in the Chaffinch, etc.,” ‘Quart. Journ. 
Mier. Sci.,’ 53, 1909, p. 339, with figs. 


EXPLANATION OF PLATES 8-10, 


Illustrating Professor E. A. Minchin’s and Dr. H. M. Wood- 
cock’s paper on “ Observations on certain Blood-parasites 
of Fishes occurring at Rovigno.” 


[The drawings on Pl. 8 are all from Giemsa-stained preparations, 
those on Pl. 9 are from films stained with iron-hematoxylin, and those 
on Pl. 10 from films stained by Twort’s method. All the figures relat- 
ing to Hemogregarina rovienensis are magnified 3000 times linear ; 
those relating to Trypanosoma trigle and T. raizw 2000 times. ] 


PERATE 8. 


Figs. 1-28—Hemogregarina rovignensis. Figs. 1-11 are from 
the first infected gurnard, figs. 12-27 from the second, and fig. 28 from 
the third infected fish. Figs. 5-11 are from a very thin smear, which 
was faintly stained; both corpuscles and parasites are uniformly flattened 
out, but the nuclei of the parasites come out better than in any other 
Giemsa-stained smears. 

Fig. 1, 2, 5-9, 16, 17.—Ordinary small forms, schizonts. 

Fig. 3—A small form, free from the blood-corpuscle, occurring 
in a smear from the kidney. 

Figs. 4, 10, 18-21.—Large forms of the slender or male type. 

Figs. 11, 22-26.—Large forms of the broad or female type. 

Figs. 12, 13.—Young individuals of the slender, male type. 

Figs. 14, 15.—Young and intermediate-sized individuals respec- 
tively of the broad, female type. 

Fig. 27.—A double infection of the corpuscle, with two small 
forms of the parasite. 

Fig. 28.—Small form showing a constriction of the nucleus in the 
middle. 


154 K. A. MINCHIN AND H. M. WOODCOCK. 


Figs. 29-31.—Trypanosoma trigle. 
Figs, 32-37.—T.. raiw. Fig. 33 is of a small parasite, fig. 35 of an 
intermediate-sized one; the rest are of large individuals. 
In fig. 34 the small vacuolar spaces were originally oceupied by 
black grains similar to those seen in the parasite of fig. 37. 
Fig. 38, a-d.—T. raiw, trophonuclei of different parasites (x 3000). 


PLATE 9. 


Figs. 39-50.—H. rovignensis. Figs. 39-43 are from the first in- 
fected fish, figs. 44-50 from the second one. 

Figs. 39-43, 46, 47.—Ordinary small forms (schizonts). 

Figs. 44, 45.—Young and intermediate-sized individuals respec- 
tively of the slender or male type. 

Figs. 48, 49.—Large forms of the broad, female type. 

Fig. 50.—Large individual of the slender, male type. 

Fig. 51.—Trypanosomatrigle. 
Figs. 52-56.—T. raie. 

Fig. 55.—The trypanosome of this figure is on a different film 
from the others, one from which the stain has been consider- 
able more extracted. 

Fig. 57, a-i.—T. raiw, trophonuclei from various trypanosomes ; 
e and h are from large trypanosomes, the rest from smaller or inter- 
mediate-sized parasites (x 3000). 


PLATE 10. 


Figs. 58-64.—H.rovignensis. Fig. 58 is on a film from which the 
red part of the stain (neutral red) has been much less extracted than in 
other cases. 

Figs. 58-61.—Ordinary small forms. 
Figs. 62, 64.—Large broad forms. 
Fig. 63.—Large slender form. 

Fig. 65.—T. trigle. 

Figs. 66, 67.—T. raie. 

Fig. 68, a-f—T. raiw, trophonuclei from various individuals; d-f 
from large parasites (x 3000). 


rd 


—— 


a. Celie, 


= 
y 


e 


ee oe 


a ee eee 


6. 


Quart, Fourn Micrdec. Vob.55.ME I. 


Quand, fou rN Mac. Sch, Dot, wie N. i SY, g. 


Sie Wnn Abas Condes 


” 


al ee —_ 


Quart. Foaurn Mier bei. Uo. 5S, NS , 10 


68. 


Huth, Lith? London 


s 


a 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. SP.), 155 


On Ganymedes anaspidis (nov. gen., nov. sp.), 
a Gregarine from the digestive tract of Anas- 
pides tasmanize (Thompson). 


By 


Julian S. Huxley. 


With Plate 11, and 5 text-figures. 


INTRODUCTION. 


In 1907 Mr. Geoffrey Smith was in Tasmania on a zoolo- 
gical errand, his object being especially to investigate the 
structure and development of Anaspides, the Mountain 
Shrimp of that country. 

After his return to England, he noticed, while examining 
his sections in detail, some curious structures in the liver, 
which on investigation proved to be large binucleate cells, 
obviously of parasitic origin. ‘Turning his attention to the 
gut, he found that it was in some cases inhabited by large 
numbers of Gregarines of an unusual type, and surmised 
that there was a connection between these and the non- 
motile parasites in the liver. 

This was enough to show that Anaspides, so interesting 
in every detail of its structure, is no less so in regard to its 
parasites ; and, as he had much work of his own on hand, he 
kindly offered me the congenial task of describing this new 
Sporozoan, at the same time providing me with all his surplus 
specimens of Anaspides. For this, and for much help and 
advice, I must here tender my best thanks; nor must I 


156 “JULIAN S. HUXLEY. 


forget to express my gratitude to Prof. Bourne for much 
kind assistance. 


Mernops, Eve. 


Preservation.—Some of the Anaspides had been pickled 
in formalin, some in corrosive sublimate; these latter were 
much better preserved, and were exclusively used in the 
work. 

Preparation of the Gregarines.—Mr. Smith’s speci- 
mens of Anaspides had been kept in captivity for some time 
before they were preserved; and, either they had had very 
little to eat, or else all the fare provided for them was 
digestible—at all events their guts were almost empty, save 
of parasites. Thus it was easy to make preparations of large 
numbers of the Gregarines by staining the gut and liver- 
tubes whole in paracarmine for a couple of hours, and then, 
after taking up to xylol, teasing in Canada balsam on the 
slide, and removing as much of the débris of the gut as 
possible, leaving the parasites behind. 

This was quite good for general features, but, as I found 
to my cost later, did not bring out certain important cyto- 
plasmic structures. 

Subsequently some more Anaspides were sent over from 
Tasmania ; these had been preserved at the moment of 
capture, and their guts were filled with a mass of sand, 
swallowed for the sake of the contained organic fragments. 
This made matters more difficult. The Ganymedes had to 
be picked one by one out of the débris by means of a eapil- 
lary pipette under the binocular microscope. They were 
then mounted from 90 per cent. alcohol on to a film of egg- 
albumen smeared over a slide, so that they could be stained 
with Heidenhain’s iron hematoxylin, which proved much the 
best reagent for picking out the details of the complicated 
structures in the cytoplasm. 

Besides making these whole preparations, I had sections 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. SP.). 157 


cut of individual parasites, and of the gut and liver of the 
host. Most of these were stained with iron hematoxylin, 
some with Khrlich’s hematoxylin and eosin, and some with 
methyl-blue eosin by Mann’s method. Iron hematoxylin 
was the best for most purposes, but Mann’s method was very 
interesting in revealing some of the complexity of the purely 
vegetative processes that take place in the nucleus and 
nucleolus. 

It was of course impossible to get any Anaspides over to 
England alive, and thus several questions of structure and 
hfe-history which could probably have been easily elucidated 
by observations and cultures of the living Gregarine, have 
had to be left to await the verdict of some investigator who 
has not got the Tropics between himself and the source of 
his material. 


GeneraL Account (Lirz-nisrory, Hasrrar, Erc.). 


As above mentioned, Ganymedes is a parasite of the 
Synearidan Crustacean Anaspides tasmania, inhabiting 
various portions of its digestive tract. Before proceeding to 
a detailed account of its structure, it will be best here to 
give a brief general survey of its life-history, as far as such a 
continuous record can be pieced together from the mere snap- 
shots which are all that preserved material can give. 

I came across no sporozoite stage. The smallest Gregarines 
found were only about one-eighth the length of the full- 
grown motile trophozoite, but otherwise similar in every 
way. These elongated motile forms, obviously belonging 
to the class Gregarinida, are in what I shall call the first 
trophic period, which is spent within the very long mid- 
gut of the host. Here some are attached to the epithelium 
(fig. 9), but the majority are found free in the lumen. If 
the host has recently been feeding, the gut is crammed with 
sand-grains and organic particles; and when this is the case, 
the parasites collect between this food-mass and the gut-wall, 


158 JULIAN S. HUXLEY. 


where there is plenty of food that they can absorb, and least 

chance of their being carried away to the exterior. 
Sometimes the parasites, instead of having their typical 

straight or gently-curved form (fig. 1) lie coiled and con- 


Trxt-Fia. 1. 


An associated couple of Gany medes, showing the cup-individual 
(B) grasping in its cup the ball (6) of the other associate (A). 
The ball end of Bis abnormal. The cup end of A has a large 
vacuole within it. The bodies are slightly dilated round the 


nucleus. 


torted against the intestinal wall; and when this is so, 
many are usually congregated in patches, and are stuck 
together, presumably by the coagulated secretion of the 
endodermic cells. What are the reasons for this condition I 


could not discover. 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. SP.). 159 


Finally, a certain number of the Gregarines are found 
associated in pairs, the attachment being by dissimilar ends 
(text-fig. 1). Not very many are in this state, but I suspect 
that the shock of killing, and the subsequent manipulation, 
manage to sever the connection between a large number of 
couples, and therefore cannot say if association always super- 
venes when the parasites reach a certain size, nor what are 
the proportion of couples to free Gregarines. 

Association marks the close of the first trophic period. 
In the second trophic period the Gregarines are non- 
motile, have lost all the complex structure they had before, 
and are characterised by their (probably rapid) growth to a 
very large size. In this state they are found in the liver- 
tubes, of which there are twenty or thirty, lying free in the 
hemococeele, and not intertwined. It follows that the asso- 
ciated couples must migrate forwards to the junction of mid- 
and fore-gut, where the liver tubes open, and thence back 
into one of these. On penetrating a safe distance along the 
tube, a transformation must take place, the two Gregarines 
undergoing complete cytoplasmic fusion, a state of affairs 
known hitherto only in those neogamous forms from Holo- 
thurians, Cystobia and allied genera (Woodcock, 6). 

These fused couples, looking just like one cell with two 
nuclei, are found wedged in between the cells of the wall, 
with a considerable free surface for absorption towards the 
lumen of the tube. There is often another free surface on 
the exterior, due, I should say, simply to the growth of the 
creature, and the consequent forcing apart of the liver-cells 
(text-fie. 3). For this growth, Ganymedes is here in a very 
favourable place, since the so-called liver, in addition to 
producing digestive ferments, is the organ where a great 
part of the food is absorbed; and so, while the parasites may 
enter on this period when measuring no more than 70 x 60 p, 
they often attain to the considerable size of 200 x 130 w, and 
I have seen one that measured 300 uw in its greatest length, 
though its breadth was only 1204. The shape is variable, 
from a nearly perfect sphere toa long ellipsoid or ovoid. 


160 JULIAN 8S. HUXLEY. 


The two nuclei meanwhile become round and very large, and 
possess on one side a large lenticular nucleolus. 

The next step in the cycle is for the associated couple, 
while still in the liver-tube, to form a thick resistant coat 
round itself: in so doing it becomes perfectly spherical, and 
a process of concentration of cytoplasmic materials must take 
place, as I have found none of these cysts with a diameter of 
more than 115 u, and one only 85 » across, the average being 
about 100 p. 

The formation of the cyst wall of necessity closes the 
trophic periods, and sporogony now presumably begins. I 
say presumably, for I have seen no spores, nor even any of 
the preparatory nuclear divisions. ‘Two cysts in the liver of 
a particular host showed nuclei with central nucleoli emitting 
chromatin (fig. 17)—a phenomenon very common in Protozoa 
at the close of vegetative life: and I have found a number of 
the usual type of cysts free in the gut. 

From these facts, and from analogy with other intestinal 
Gregarines, we must suppose that after the formation of the 
smooth cyst wall, the couples can be expelled from the liver 
tubes (while those in the second trophic period remain in 
place by virtue of their soft surface adhering to the similar 
surfaces of the liver-cells), that they are then passed out by 
the anus, and that it is only here, under the stimulus of the 
changed conditions, that the processes leading to the pro- 
duction of spores can take place. 

This being so, it is probable that infection is casual, the 
spores or sporocysts being taken in with the food—as, 
indeed, might have been deduced from the feeding habits of 
Anaspides. The infection is usually heavy (text-fig. 3), 
and frequently seems to be multiple, cysts, motile Gregarines, 
and associated immobile forms being often found all in one 
host. The proportion of infected hosts was over 50 per cent. 
in the case of those that were captured by Mr. Smith ina 
small pool on one of the mountain becks of Mt. Wellington ; 
but in those he obtained from a larger piece of water, the 
infection was nil—or at least no parasites were forthcoming 


GANYMEDES ANASPIDIS (NOV. GEN., Nov. spP.). 161 


in the dozen or so of hosts that [ examined. The time of 
year seems to have no effect on any of the processes of the 
parasites’ life. 

As regards the effects produced by Gany medes, no incon- 
venience seems to be suffered by the organism of the host as 
a whole, and only trifling damage is done to individual 
tissues. ‘Those few cells of the gut epithelium to which the 
Gregarines attach themselves look generally unhealthy, and 
their nucleus becomes hyperchromatic (fig. 10) ; and the walls 
of the liver tubes get more or less distorted by the growth of 
the large couples in the second trophic phase: but in neither 
of these ways can any serious harm be done. 

After these preliminary remarks, we may now proceed to 
consider in detail the structure of Ganymedes in its various 
stages. 


DETAILED AccouNT. 
(i) The First Trophic Period. 

Although the size of the smallest free Gregarines seen was 
only 80—100 nu, yet I could find no points of difference 
between them and the adults, save that in the young forms 
the body has not attained to its full size relative to the 
structures (soon to be described) situated at the extremities. 
From these small forms all stages may be seen to Gregarines 
400—425 uw long, and 23—30 w broad, though the average 
size is 250—300u x 17—20 pu. 

The shape of the body is cylindrical, tapering slightly 
towards one end, and considerably towards the other. The 
thinner end is almost certainly anterior in progression, and 
when attachment takes place, it is by means of a structure at 
this extremity. This structure in favourable specimens is seen 
to consist of a sphere connected by a thinner neck to the 
main body: I propose to call it the ball, and the thin 
extremity on which it is placed, the ball end. The other 
end may be called the cup end, for here many individuals 
possess a perfectly regular hemispherical depression, whose 

VOL. 55, PARY 1.—NEW SERIES. i 


162 JULIAN S. HUXLEY. 


outside walls continue the lines of the body: the whole is 
marked off by a circular groove, thus rather resembling the 
sucker of an Octopus. 

Leaving the details of these organelle for the present, I 
will now describe the main body of Ganymedes. This is of 
the usual type seen in motile Gregarines. It is covered with 
a firm cuticle, the longitudinal striations on which can be 
easily seen (figs. 6, 10, 11). Just beneath this appears in 
many cases a pale ectoplasmic layer, lacking the granules 
of the central endoplasm: and though I have never been able 
to demonstrate actual myonemes, yet from what we know of 
other Gregarines it is probable that this layer is the seat of 
the contractile structures which this free-swimming creature 
must possess. The endoplasm proper is denser, and con- 
tains granules. The whole cytoplasm is of reticular or 
alveolar structure. 

‘ne nucleus lies more or less in the centre of the body: 
it is ellipsoidal: the folds and processes sometimes seen at 
one end of it (fig. 15) being probably artefacts. Its breadth 
is often very nearly that of the Gregarine, and it would some- 
times touch the cuticle except that when it is large the body 
bulges out slightly round it. It possesses a thin but distinct 
nuclear membrane, within which is a reticulum with granules 
on the threads—sometimes loose with largish grains (fig. 14), 
sometimes finer (fig. 15). In addition there is present a 
deeply staining spherical nucleolus, usually towards the cup 
end of the nucleus. In it, a thin outer rind usually stains 
deeper than the central medulla, which is filled with clear 
vacuoles of various sizes (figs. 14, 15). With Mann’s 
methyl-blue-eosin it stains usually bright crimson to claret- 
colour, often with a violet crescentic area on one side. 

Returning now to the anterior extremity, we find that in 
some cases there is, as above stated, a distinct stalked sphere 
(figs. 7—10). This is covered with a cuticle thinner and less 
firm than that of the body, the two passing into each other 
round the narrowest part of the neck (fig. 7). The sphere is 
filled with a quite homogeneous fluid, except at the extreme 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. SP.). 163 


front end, where there is usually a sort of pad of fine-grained 
cytoplasm projecting back into the cavity (fig. 8). In the 
main body, behind the neck, is another spherical cavity, 
apparently separated from that of the ball proper, and con- 
taining a fluid that is not quite clear, but of a loose reticulate 
structure (figs. 6, 7). Hnclosing the hinder part of these 
may sometimes be seena dark crescent of nearly homogeneous 
material (fig. 9). 

So far, so good. In other cases, however, we find quite a 
different appearance, there being only one cavity present, and 
all traces of a neck having vanished (figs. 3, 5). Closer 
inspection shows that the cavity corresponds with that of the 
true ball, as its contents are perfectly clear, and it has a 
pad of cytoplasm anteriorly. The dark crescent may come 
directly behind it (fig. 5), while the thick body cuticle 
extends completely over it. The question then is, what is 
the relation between these two conditions ? 

It seems obvious that the ball can be extruded at will 
—but in what way? Is it evaginated (pleurecbolic) or 
is it acrecbolic, and, if the latter, is it pulled out by 
muscular or elastic action or pushed out by some other 
means; and how is it retracted? An examination of many 
Gregarines (a task necessary owing to the absence of living 
material, but laborious from the small size of the ball— 
8—10 pw across), has made it seem probable that it is acrec- 
bolic, and pushed out by the accumulation of a watery fluid 
behind it. As far as I can make out, the structures and 
processes concerned are as follows :—The dark crescent (s.t. 
in figs.) is a tissue which has the power of secreting a fluid 
(w.) into a space anterior to it, thus driving the ball out 
through an opening in the body cuticle. When the ball is 
retracted, the elastic cuticle would be closed over the anterior 
end; and when extrusion has taken place, it would press in 
and form the thin neck. One animal (fig. 2) shows what I 
suppose to be an early stage of extrusion: the hole is just 
being enlarged, so that the cuticle at its edge stands out as a 
well-marked rim (cut. rim). In later stages (figs. 4 and 6) 


164 JULIAN 8S. HUXLEY. 


this rim will press against the convexity of the ball and thus 
be difficult to see; it is only in the early stages of extrusion 
that its inner surface will form an angle with the surface of 
the ball, and thus stand out. The pad of cytoplasm (p.) is 
always seen at the anterior end of the ball vesicle, showing 
that there can be no question of invagination. 

Retraction would then take place by the resorption of the 
secretion ; while the ball seems to be kept in place by strands 
from the ectoplasm (probable muscular layer), for this, and 
this only, usually extends up the sides of the secreted fluid to 
the ball vesicle (figs. 3, 5, 6). 

When fixation takes place, the condition of things looks 
somewhat different (fig. 10),and there is an open communica- 
tion from the ball to the space behind it. Very possibly the 
cytoplasm at the neck is temporarily dissolved so as to leave 
this passage-way for the food absorbed by the ball to pass 
further into the substance of the animal. 

Finally, in association, the ball of one is extruded into 
the cup of the other, and the cup then seemingly contracts so 
as to hold the ball firm (fig. 9; text-fig. 1). It may be here 
remarked that the free ball end in the couple in text-fig. 1 is 
quite abnormal: it was pointed, and contained a pointed 
cavity within it, but otherwise had none of the typical 
structure. 

The cup-end also presents various difficulties. When 
well formed its structure is simple enough, and has already 
been described. But at other times the hollow cup may be 
quite wanting, the body ending simply in a rounded end with 
rather thick ectoplasm (fig. 12); or, more often, there are 
numerous vacuoles beneath the cuticle (fig. 13), with some- 
times an irregular aperture in addition (text-fig. 2). What 
the meaning of these variations is, and whether the cup-end 
can pass from one state to another, I fear I cannot say. 

It was from the presence of the cup that I ventured to call 
this new genus Ganymedes, though the pedant will perhaps 
maintain that this name should have been reserved for some 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. SP.). 165 


parasite of Aquila. With the specific title anaspidis, 
however, I think no one will quarrel. 


(ii) Second T'rophic Phase. 


Between the two phases of trophic life no intermediate 
stages were found, all the couples in the liver having lost 
every trace of the cytoplasmic structures of the Gregarinoid 
form. All they possess is a thin cuticle (fig. 18), investing 
a delicately-meshed cytoplasm. 

The nucleus, on the other hand, has increased in com- 
plexity (fig. 18). It is large and more or less spherical, 


TEXT-FIG. 2. 


Diagrammatic view of the cup end of a Gregarine, to show the 
opening on one side, and the numerous vacuolar spaces in the 
cytoplasm. 


with a thin nuclear membrane, and an achromatic network 
in which there is very little chromatin present. The chief 
interest lies in the nucleolus, which is peculiar in two 
ways. First, it occupies an unusual position, right on one 
side of the nucleus, somewhat like the lens of an eye, with a 
considerable surface in contact with the cytoplasm—a state 
of things not, I believe, known in any other Gregarine, 
though Awerinzew (1) has described something similar for 
a Myxosporidian ; and secondly, it possesses itself another 
lens-like structure, projecting more or less into the cell- 
body, and composed of a very pale-staining meshwork, 
with its outer border not a smooth curve, but formed of the 
slightly projecting parts of the component alveoli (fig. 18). 


166 JULIAN S. HUXLEY. 


This is perhaps the absorptive part of the nucleolus, taking 
up from the cytoplasm the soluble food which this in its 
turn has abstracted from the liver-tubes. 

In the centre of the nucleolus, abutting on the absorptive 
part, is often an area, with a reticular structure, staining 
blue-violet with Mann’s method. The remainder is com- 
posed of a dense material staining deep red, in which are 
embedded definite clear pink vacuoles. Towards the cyto- 


TEXT-FIG. 3. 


ak 
us f 
3. re} 


Portion of a liver-tube of Anaspides with four couples of 
Ganymedes in it. The nuclei of the liver cells are repre- 
sented only in one corner. J = lumen of liver tube. The 
lighter parts of the parasites (e) are exposed on the exterior 
of the liver-tube. 


plasm these vacuoles project slightly ; when one sticks right 
out, as at a, fig. 18 b, it is colourless, showing that the others 
look pink only because there is red substance above and 
below them. Towards the nucleus, on the other hand, the 
vacuoles rarely project, the edge of the nucleolus being 
usually clean cut. Text-fig. 4 represents diagrammatically 
another nucleolus in which the absorptive area was ex- 
tremely large. 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. SP.). 167 


The nucleolus thus seems obviously to be the chief agent 
concerned in the manufacture of food-stuffs (for theories 
regarding the action of Mann’s methyl blue eosin see Léger 
and Duboscq (2) ). 

What is the function of the rest of the nucleus in this 
period remains uncertain, though its large size shows that 
it must play some important part in metabolism. The chief 
interest here lies in the behaviour of the nucleolus, which 
migrates out to enter into direct relations with the cytoplasm 
at the beginning of the second trophic period, when assimila- 
tion begins to be greatest, and at its close, when all 


TExT-FIG. 4. 


Ex > 

VA. oe: &3 
y ay 

ee ay 
~~“ 


Section of one of the nuclei of a couple in the second trophic 
phase. The nucleolus does not project very far, and the 
surface of the absorptive area is flush with that of the 
nucleolus, although the area itself is very large. 


assimilation ceases, returns, as will be seen later, to the 
interior of the nucleus. 


(iii) Encysted Phase. 


The cyst-wall, though always fairly strong, varies a good 
deal in thickness. It stains bright blue by Mann’s method, 
bright red with carmine, but not strongly with hematoxylin. 
From it often project radially inwards curious irregular, 
branching filaments, never reaching much more than a third 
of the way to the centre, as to whose nature and function I 
am quite in the dark (fig. 16). 

The cytoplasm is reticular, with minute granules on the 


168 JULIAN 8S. HUXLEY. 


threads, and larger, chromatic granules here and there. It 
always looks denser than in the unencysted forms. 

The nuclei in what I take to be the earlier cysts are much 
like those described for the second trophic phase, except that 
they stain a little deeper, and that the nucleoli do not pro- 
ject so far out from the surface (text-fig. 5). In the next 


TrxT-Fia. 5. 


A cyst found in the gut. The nuclei are not actually touching, 
but very near to each other. The cyst-wall is very thick in 
this specimen. 


stage (fig. 16) the nuclei, bounded only by a very thin mem- 
brane, stain quite deeply, as they are almost filled with 
chromatic granules of various sizes. The nucleolus is still 
in contact with the cytoplasm, but its outer surface is now 
flush with that of the nucleus. This outer border of the 
nucleolus is made up of rows of minute vacuoles, while the 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. spP.). 169 


rest is dense, with a clean-drawn boundary towards the 
interior, and homogeneous except for a few large vacuoles. 

To this stage probably belongs the cyst in fig. 19, stained 
by Mann’s method. The nucleolus is blue, having given up 
most of its chromatin to the nucleus, which is violet with 
dark purple grains. 

In fig. 17 we have another state of affairs: The nu- 
cleolus, now retreated from the surface, seems to be giving 
off chromatin to the nucleus in the shape of hollow spherules. 
It is itself formed of a single central vacuole, surrounded by 
a layer of small ones embedded in a dense chromatic cortex 
(the lower nucleolus is cut tangentially, and so does not show 
this condition). The nucleus, apart from the chromatic 
spherules, appears perfectly homogeneous, with no achro- 
matic network, and differs also from the nuclei of other 
stages in being ameebiform, with ‘‘ pseudopodia” that can 
be very clearly seen on focussing up and down. 

From what we know of other Gregarines, it is clear that 
these stages are preliminary to the breakdown of the large 
trophic nuclear apparatus, and the reconstitution of the 
idiochromatin to form the gametocyte nucleus. But, as 
above mentioned, the cysts soon after this pass into the gut 
and out by the anus, so that their further development 
must remain for the present unknown. 


Conciusions: Systematic Postrion. 


Though here more than ever must we lament the absence 
of spores, it is still possible to draw some fairly definite 
conclusions. To start with,Ganymedes is not a Polycystid, 
nor does it belong to any existing family among the Mono- 
cystids. ‘hus a new family, the Ganymedida, must be 
created, whose characters will provisionally be those of the 
genus: these may be here conveniently summarised as 
follows : 

(1) The possession by the motile form of a special exten- 
sible organ at the front end, which may serve for fixation 
to the cells of the host. 


170 JULIAN 8. HUXLEY. 


(2) The presence of a special cup-like structure at the 
posterior end, which co-operates with the epimeritic organ 
at the anterior end to effect a close union of two individuals 
during association. Association is thus by dissimilar ends, 
and lasts for some time. 

(3) The eventual complete cytoplasmic fusion of the 
associated couples, and the existence of a second trophic 
phase, when the animals grow very fast, but are morpho- 
logically quite degenerate. 

(4) The position of the nucleolus in this phase, on one 
side of the nucleus, partly in contact with the cytoplasm. 

(5) The habitat, in the gut and liver of Syncaridan 
Crustacea. 

Considering these characters in relation with other members 
of the class, we find that no known Gregarines inhabit the 
liver of any Crustacean; none have the nucleolus in the same 
position; none go through two trophic phases; none have 
any special structure for association at the posterior end; 
and none have a protrusible organ of the same sort at the 
front end. It is thus at least obvions that Ganymedes is the 
representative of a very divergent line. The suggestion I 
would make is that, while nearer to the Monocystid type, 
Ganymedes is partly intermediate between the two great 
groups of Hugregarines, as represented diagrammatically in 
the following tree: 


Polycystidea. Ganymedes. Monocystidea. 


Ancestral Eugregarines. 


GANYMEDES ANASPIDIS (NOV. GEN., Nov. SP.). 171 


In the first place, the ball and the cavity containing the 
secreted fiuid represent with great probability an epimerite 
and protomerite. ‘'rue, there is no cuticular septum; but 
the secreting tissue forms a fairly definite barrier between 
these on one side, and on the other the deutomeritic posterior 
part. Here alone, it is to be remarked, do we find the true 
granular endoplasm. Occasionally, too, this latter can be 
seen ending off with a definite contour within the secreting 
tissue (fig. 3). The ball itself, when extruded, would pass 
for a typical epimerite save for the absence of a septum 
behind it; but in so far asit is protrusible, it is only paralleled 
by the anterior extremity of Lankesteria ascidiz (Sied- 
lecki, 4). his, however, seems to be merely a pseudo- 
podium, or a drop of the hyaline inter-reticular substance of 
the cytoplasm pressed out through a hole by contraction of 
the animal, and its extrusibility has obviously been inde- 
pendently evolved. 

The fact of its being a parasite of the digestive tract is the 
second link with the Polycystidea. The only Monocystid 
eut-parasite whose life-history has been thoroughly worked 
out is Lankesteria, and this possesses an “ epimeritic” 
organ. The three or four other genera of this sub-class that 
live in the gut, such as Callyntrochlamys and Ancora, 
are very insufficiently known; it is even possible that they 
may be Polycystid in early stages. 

Regarding the matter phylogenetically, we find that the 
early Eugregarine stock must have been motile, Polycystid 
gut-parasites ; their association was by dissimilar ends, and 
took place only at the very end of the trophic period ; and 
they showed well-marked anisogamy. 

One of the first steps towards the typical Monocystid 
condition was the change of habitat, due very likely in the 
first instance to the evagination of the full-grown tropho- 
zoites from the gut into the coelom—as takes place to-day 
in certain insect-parasites at the time of the host’s meta- 
morphosis. For a full discussion of the further stages, 
leading eventually to complete isogamy, coupled with entirely 


172 JULIAN S. HUXLEY. 


coelomic habitat, precocious ‘association, and degenerate 
structure, the reader is referred to Woodcock (6). Suffice 
it here to say that the course of affairs in Ganymedes must 
have been somewhat different. It is probable that Gany- 
medes at first associated only at the close of the trophozoite 
stage. Some of the couples having migrated into the liver, 
found it (like the ccelom for other Monocystidea) a safe 
retreat and abounding in soluble food. Here too the 
Gregarine could afford to dispense with all the structures 
necessary for a life in the open gut, and devote all its 
energies to growing. One might have thought then that 
Ganymedes would have associated in the sporozoite stage, 
like Cystobia, and migrated at once into the liver; but, 
whether non-motile couples below a certain size could be 
expelled from the tubes or be engulfed and digested by the 
activity of the liver-cells (see Smith, 5, p. 536), or from some 
other cause, Ganymedes has found it necessary to remain 
in the gut till it has attained a definite bulk, thus presenting 
to us the phenomenon of two sharply-distinct trophic phases 
after the sporozoite stage. As the parasites are non-motile 
when they are about to sporulate, conjugation must needs be 
precocious, so that no Gregarine shall migrate alone into the 
liver, and thus be, from the point of view of the species, 
wasted. For this fairly lasting association some special 
mechanism was imperative, hence the cup and ball; while 
the necessity of remaining some time in the gut has led to 
Ganymedes retaining more of the original Polycystid 
structures than is usual in the morphologically degenerate 
Monocystidea. Finally, although the sporogony remains 
unknown, it may be confidently prophesied that this Grega- 
rine will be found to be completely isogamous. 

Thus it will be seen that the Ganymedide diverged 
very early from the Monocystid stock, and possess now 
many new and peculiar characters intermixed with those 
they have inherited from the common ancestor. For the 
complete disentangling of these from each other, further 
work must be done on Ganymedes, and in addition all 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. SP.). 178 


Syncaridan Crustacea should be searched for allied parasites, 
whose structure would at once give us new standpoints from 
whence to view the problem. 


LITERATURE REFERRED TO. 


_ 


. Awerinzew.—‘ Studien tiber parasitische Protozoen : I. Ceratomyxa 
Drepanopsettae,” ‘ Arch. f. Protistenkunde,’ vol. 14, 1909, p. 74. 


i) 


. Léger and Duboseq.—‘ L’Evolution schizogonique de Aggregata 
eberthi,” ‘ Arch. f. Protistenkunde,’ vol. 12, 1908, p. 44. 


3. Minchin.—Article ‘‘ Sporozoa,” in Lankester’s ‘ Treatise on Zoology,’ 
Part I, 2nd fascicle, 1903. 


4. Siedlecki.—‘‘ Ueber die geschlechtliche Vermehrung der Mono- 
cystis ascidiae, R. Lank.,” ‘ Bull. Internat. Acad. Sci. Cracovie,’ 
Dec., 1899. 


5. Smith.—‘“On the Anaspidacea, Living and Fossil,” ‘Quart. Journ. 
Mier. Sci.,’ vol. 53, 1909, p. 489. 


6. Woodcock.—* The Life-cycle of ‘Cystobia’ irregularis, Minch., 
together with observations on other ‘Neogamous’ Gregarines,” 
‘Quart. Journ. Micr. Sci.,’ vol. 50, 1906, p. 1. 


EXPLANATION OF PLATE 11, 


Illustrating Mr. Julian Huxley’s paper “On Ganymedes 
anaspidis (nov. gen., nov. sp.).” 


REFERENCE LETTERS FOR THE FIGURES. 


b. Ball-cavity. c¢.s. Cuticular strie. ect. Ectoplasm (probable myo- 
cyte layer). p. Cytoplasmic pad at anterior end of ball. s.¢. Secreting 
tissue. v. Vacuoles. w. Secreted fluid that accumulates to drive the 
ball out. 


Bor.-car. Borax carmine. Paracarm. Paracarmine. Hem. Hema- 
toxylin. M. B. EH. Methyl-blue eosin (Mann’s method). 


174 JULIAN S. HUXLEY. 


PoATH, 11. 


Fig. 1—Large individual at the close of the first trophic stage, with 
well-formed cup. (Paracarm. x 640.) 


Fig. 2—Ball end of the same, to show the ball being pushed out 
through the hole in the cuticle; the edges of this hole stand out 
markedly as a rim (cut. rim). (x 1300.) 


Figs. 3-8.—Ball ends of various Gregarines in different condi- 
tions. 

Fig. 5.—(Semi-diagrammatic). Very slightly extended. 
Secreting tissue very large, with the granular endoplasm 
(e,) ending off within it. Outside is a non-granular layer 
(e,), and just beneath the cuticle the still paler ectoplasm, 
extending on the left to touch the ball-vesicle. (Bor. 
Carm. X 1875.) 

Fig. 4.—Semi-extended. The secretion of the secreting 
tissue is fairly dense. The double contour of the hinder 
part of the ball is well seen. There seems to be no 
ectoplasm. (Iron Hem. x 1875.) 

Fig. 5.—(Semi-diagrammatic.) Completely retracted. Very 
large cytoplasmic pad (p) with dark grains in it. A 
large dark granule in the secreting tissue. The ectoplasm 
extends to touch the ball. (Iron Hem. x 1875.) 

Fig. 6.—Almost extended. The secreted fluid has here a wide- 
meshed structure. The thick body-cuticle ends abruptly 
where it touches the ball, which possesses only a thin 
cuticle. Cuticular striae are seen on the under surface. 
No well-differentiated ectoplasm. (Iron Hem. x 1875.) 

Figs. 7 and 8.—(Semi-diagrammatic.) Completely extruded. 

In fig. 7 the neck of the ball is well seen, also the more delicate 
nature of the ball’s cuticle. No cytoplasmic pad is visible. 

In fig. 8 the ball is directed slightly upwards. The cuticle is 
distended round the secreted fluid, showing that this is 
under pressure. (Paracarm., fig. 7 x 1300; fig. 8 x 1875.) 


Fig. 9.—Section (5 ») through the point of junction of an associated 
couple in the first trophic phase. The cytoplasm of the ball individual 
(A) is denser than that of the other (B). (M.B. E. x 1500.) 


Fig. 10.—Section (5) through the point of attachment of a mobile 
Ganymedes toa cell of the host’s gut. The cuticular striz are well 
seen. The ball is thrust into the host-cell, and contains a fluid that is 
not clear, the reticular structure being probably due to the coagulation 
of absorbed food. There isan open passage through the neck into a 


GANYMEDES ANASPIDIS (NOV. GEN., NOV. sp.). 175 


cavity in the body of the parasite. The cytoplasm contains numerous 
deeply-staining granules. The nucleus of the host-cell (n) is large, 
darkly-stained, and homogeneous, except for some dark grains. (Iron 
hem. x 1340). 


Figs, 11-15.—(Semi-diagrammatic). Cup-ends. 

Fig. 11.—Cup-end of the Gregarine whose ball-end is shown in 
fig. 5; (a) is focussed near the upper surface, and shows 
how the cup is separated from the body by a circular 
groove; (b) shows the greatest diameter ot the cup. (Iron 
Hem. x 1875.) 

Fig. 12.—Cup-end of another Gregarine, to show absence of 
all differentiation. The ectoplasm is thicker at the end 
than elsewhere. (Paracarm. x 13500.) 

». 13.—-Section of the cup-end of Gregarine a in fig. 9, to 
show the numerous vacuolar spaces beneath the cuticle. 
(M. B. E. x 1300.) 


Figs. 14 and 15.—Sections (5 ») to show the structure of the nucleus 
in the first trophic phase. (M.B. E. x 1300.) 
Figs. 16 and 17.—Sections of cysts. 

In fig. 16 the filamentous inward projections from the cyst-wall 
can be seen. Small chromatic granules fill up the nucleus ; 
there is no sign of an achromatic network. The nucleoli 
are retreating to the interior of the nucleus. (Iron Hem. 
5 p X 970.) 

In fig. 17 the nuclei are ameeboid, filled with a homogeneous 
sap in which are hollow chromatic spherules, apparently 
emanating from the nucleoli. The cyst-wall is crumpled, 
and in one place a flap of it has got detached so that 
its surface-structure is seen. (Ehrlich’s hem. + eosin 
10 » x 800.) 


Fig. 18a.—Section (5 1) through an associated couple in the second 
trophic phase. The reticular nature of the cytoplasm is not indicated. 
(M. B. E. x 610.) 


Fig. 18b.—The next section in the series. The nucleolus and the 
outline of the nucleus are given, more highly magnified. The three 
areas of the nucleolus and their structures are shown (see text). At x 
a vacuole projects beyond the general surface, and is seen to be colour- 


less. (M.B.E. x 870.) 


Fig. 19.—Section of a cyst, to show the alteration in staining reactions 
of nucleus and nucleolus in this stage (see text). (M.B.E. x 400.) 


Ki 


JQ 


* : = Saal 
- © 
4 
- 
5 
, 
- 
a 
« 
a = 


Iuanrt. PurrMicrici. Vol. 55 NS MY, I/ 


ee a 


oie Steck 
Senses 


Huth Tith? Londow 


oy 


ing, 


THE F@TAL MEMBRANES OF THE VERTEBRATES. 177 


The Foetal Membranes of the Vertebrates. 


AN ADDRESS REPRINTED FROM THE “ PROCEEDINGS OF THE 
SEVENTH INTERNATIONAL ZOOLOGICAL CONGRESS ” 
HELD AT Boston, 1907.! 


By 
A. A. W. Hubrecht. 


I was honoured by the request of the Executive Committee 
to give an address at the first meeting of the Embryological 
Section of the Seventh International Zoological Congress. 

I hope that in choosing for my subject the present state of 
our knowledge concerning the foetal membranes of vertebrates 
I can avoid the disadvantages of too much special detail, and 
can at the same time call your attention to the fact that these 
foetal membranes offer a very wide field for theoretical specu- 
lation, that may in its turn influence our views concerning 
certain important phylogenetic problems. 

The foetal membranes of vertebrates are known to occur in 
reptiles, birds, and mammals. The embryological hand-books 
tell us that they are absent in amphibians and fishes. 

In consequence, a primary subdivision of the vertebrates 
has been instituted, those with foetal membranes being classed 
as Amniota allantoidea, those without them as Anamnia anal- 
lantoidea. From this nomenclature any close observer, even 
when he is not a zoologist, may safely conclude that one of 
the foetal membranes carries the name of amnion, the other 

1 At Professor Hubrecht’s request this address is here reprinted. It 
will assist readers in apprehending the conclusions which Professor 


Hubrecht holds to be rendered probable by the large memoir published 
in this Journal in November, 1908. 


VOL. 55, PART 1.—NEW SERIES. 12 


178 A. A. W. HUBRECHT. 


the name of allantois. An older, now more obsolete, sub- 
division into Achoria and Choriata reveals the presence of a 
third membrane, the chorion, about which we will have more 
to say hereafter, and which will explain how this third mem- 
brane came to fall—so to say—between two stools, when the 
division into Amniota and Anamnia was established. 

If we now take into account that neither chorion nor 
amnion nor allantois was ever detected in fishes or in am- 
phibians, then we must recognise that the problem, how these 
foetal membranes of the vertebrates did arise, is one well 
worthy of full consideration. 

Up to now attempts to explain their gradual evolution have 
utterly failed. So, for example, the suggestion of van Beneden 
and others that the amnion, as a protective membrane, arose 
in consequence of the early embryo sinking into the yolk-sac, 
which closed up above it, has long since been abandoned. 
Also Haeckel’s idea that the allantois arose by a precocious 
segregation of the urinary bladder of an early amphibian 
which took the habit of carrying blood-vessels, at a very 
early stage, to the outer wall of the blastocyst, must be 
dropped by all who object to predestination in evolutionary 
processes. Whenever an explanation offers itself which does 
afford a clue to a more logical sequence of events, it should 
be preferred. 

And turning finally to the outer layer, the chorion, who can 
be satisfied with the lame explanation that the appearance of 
this membrane is a necessary sequel to the formation of the 
amnion, which we find inside of it, and which later, in so 
many orders of mammals, never even arises by folds, which, 
however, in their turn are necessary to explain the chorion’s 
appearance ? 

The subsidiary explanation of all the three embryonic 
envelopes, which I am going to offer you on this occasion, 
seems to me to have the great advantage of simplifying 
matters ; especially in this sense, that henceforth we can link 
them all three to one simpler and earlier stage (which must 
have preceded in the Carboniferous and in earlier geological 


THE F@TAL MEMBRANES OF THE VERTEBRATES. 179 


epochs) without having to look for incipient stages of any of 
them among our present ichthyopsids. Nay, we may even 
say that of this earlier, archaic starting-point evident traces 
have been preserved in the teleostomes, the dipnoi, and the 
amphibians, so that we have to reconsider most seriously 
whether it will be wise to go on subdividing the vertebrates 
into the two subdivisions of those that have and those that 
have not the fcetal envelopes above mentioned. 

Now let us consider the facts as they present themselves to 
us, when we want to test the question whether one single 
original foetal envelope could not after all be at the bottom of 
the three complicated involucra we have just mentioned. As 
far as I can see, we are only in need of this one assumption, 
that an invertebrate ancestor was possessed of what we call 
an exterior larval layer (such as are not uncommon among 
different worms, and as we find them, with certain further 
complications, in some arthropods), to be able to explain how, 
in their vertebrate descendants, chorion, amnion, and allantois 
gradually came into being. 

Part of this hypothetical assumption we see actually realised 
under our eyes wherever one of the mammals goes through 
its normal stages of development. 

We find that the cell-material out of which the embryo is 
going to be built up is surrounded by an expanded cell-layer, 
which takes no part whatever in the composition of the future 
embryo. Here we actually have our single larval layer that 
will be stripped off later, and that surrounds what are going 
to be the formative cells. 

In all mammals it is this very larval layer which will become 
the outer wall of the blastocyst, what we have above called 
the chorion. 

But before following it in its further transformations, we 
have to ask ourselves, what can be the reason that this outer 
larval layer, this trophoblast, is so far away from the formative 
cells of the embryo which adhere to it only at one point? 

We have only to recall the fact of the pilidium larva, in 
which, similarly, the distance between the outer layer and the 

12$ 


180 A. A. W. HUBRECHT. 


cell-material, which is going to be the new worm, is also very 
considerable, to remove the objection that in this respect 
mammals would stand isolated. And we may go one step 
further and say that it is easy to understand why this con- 
siderable extension of the outer larval layer has come into 
existence. When we look back along the line of phylogenetic 
descent we can imagine that at the period when, for the first 
time, aquatic animals became inhabitants of the land, four- 
footed instead of four-finned, and adapted for aérial breathing 
in addition to their respiration by the aid of gills, it may have 
been a great advantage to them to become viviparous at the 
same time, i.e. to keep their developing eggs inside of them, 
where they are better protected and can be better nourished 
than outside of the mother. The atmosphere and the dry 
land offer less favourable conditions for the development of 
that small amount of protoplasm that forms the primordium of 
each new being than does the water, and so viviparity is likely 
to have been a parallel phenomenon to the exchange of the 
aquatic for the terrestrial existence. 

We can see clearly that once an embryonic envelope, one 
cell-layer thick, being present (on our original assumption, as 
far back as the invertebrate ancestor), that this one-layered 
larval envelope could obtain high efficiency for the incipient 
viviparity if only it bulged out as much as possible, thereby— 

(1) Preventing the egg from passing through the genital 
ducts rapidly and being deposited, so to say, accidentally. 

(2) Enabling the egg to adhere in various ways to the 
maternal tissues, either as a simple mechanical improvement 
of what was attained (1), or at the same time inducing phago- 
cytotic attacks on that maternal tissue. 

(3) Creating the occasion for individual trophoblast cells 
of this outer layer to absorb fluids either from the uterine 
cavity or accessory to the phagocytic processes alluded to 
under (2), and thus accumulating nutritive material inside the 
blastocyst. 

Furthermore, it is equally clear that, once the viviparity 
having been establised, and the surface extension of the 


THE FQ@TAL MEMBRANES OF THE VERTEBRATES. 181 


trophoblast going parallel with it,a yet more efficient mode of 
nutrition than the one alluded to above under (3) might be 
obtained if the embryonic vascular system, which was slowly 
coming into existence on the hereditary plan of development, 
succeeded in spreading out, in one way or another, on this 
outer trophoblastic layer, and would enter into osmotic inter- 
change with maternal blood. 

Finally, the protection of the embryonic shield during its 
further development by some sort of appliance resembling a 
water cushion would, in these incipient viviparous animals, 
undoubtedly have been a most efficient variation, for the 
earliest origin of which we have simply to go back to the 
early stage in which we noticed the formative cells of the 
embryo adhering to the larval layer, the trophoblast, in one 
spot only. Suppose that in further development this sessile 
attachment to have become converted into a circular adhesion 
—by fluid accumulating between the trophoblast cells and the 
formative cells, as we see it happen under our eyes in Hrinaceus 
and Gymnura—we then find that the water-cushion, in casu 
the amnion, took its origin in a most simple fashion, whereas 
the chorion is in no way dependent on it, but has preceded it 
as as earlier formation. 

The rapid summary here given shows us that the assumption 
of a single monodermic larval layer is quite far-reaching 
enough to allow us to understand how, out of it, chorion, 
amnion, and allantois (the latter as representing one form of 
early vascularisation of the trophoblast) have gradually come 
about. 

The only change we have to make, in what I might 
designate the present “ fashion” in comparative embryology, 
is that we look upon the earliest ancestors of mammals not as 
oviparous, yolk-laden vertebrates, but that we acknowledge 
them to have been viviparous animals with blastocysts that 
obtained vesicular shape from quite other motives than an 
eventual “loss of yolk,” such as Rabl has attempted to prove, 
Here, then, is the place for an appeal to paleontologists. They 
haye no shadow of direct interest in foetal envelopes which are 


182 A. A. W. HUBRECHT. 


never met with in the fossil condition! But they may, never- 
theless, be all the more impartial jadges when we have to 
choose between two different assumptions: the one given in the 
hand-books, according to which mammals must, through the 
Ornithodelphia, be derived from some oviparous sauropsidian 
ancestor, or the one here advocated, according to which a 
viviparous Prototrapod, provided with an adhesive and dis- 
tending larval layer diverged into various directions, some of 
the descendants utilising the conditions of growth and develop- 
ment (such as they find them) with the highest degree of 
intensity and becoming primates, others applying their tropho- 
blast to nutritive purposes in more diverse and less direct 
ways, becoming the ancestors of most of our other Mono- 
delphia and Didelphia. Others, again, going a certain distance 
with the preceding, but then acquiring yolk-laden eggs 
(Ornithodelphia), whilst yet other very effective branchings 
off in various directions gave rise to the primitive sauropsidian 
ancestors. 

The difference between the sauropsidian and the amphibian 
descendants of the protetrapods need no longer be so incisive 
—as those zoologists that divide the Vertebrates into Amniota 
and Anamnia would make it. The hypothesis here brought 
forward proposes to look upon what we know as the Decks- 
chicht of the early larval Amphibia and Dipnoi, and even of 
the teleostomes, as a last remnant of the very larval layer from 
which we started in trying to explain the foetal membranes of 
vertebrates according to what seems to me a simple plan. 

We have now to look a little closer into certain details, by 
which we may be enabled to judge of the greater or smaller 
degree of tenability of some of the views here brought forward. 

We notice that all the Mammalia-monodelphia, that have 
up to now been observed in very early stages, fully confirm 
the strong antithesis which in those early stages prevails 
between the trophoblast and the embryonic cells strictiori 
sensu. We also notice this in the Didelphia, as far at least 
as Selenka’s figures for the opossum go, although he himself 
has not interpreted the facts he brought to light in the same 


THE FETAL MEMBRANES OF THE VERTEBRATES. 183 


way asI do. Similarly, Wilson and Hill, in their latest paper 
on the development of the duck-bill, give us figures of sections 
which make it probable that the distinction between tropho- 
blast and formative cells holds good here, even though the 
development of yolk has obliterated the sharp outlines of the 
process. 

Again, in reptiles and birds traces of the larval layer have, 
in later years, been unmistakably noticed. Schauinsland, 
Mitsukuri, and Mehnert were among the foremost to contri- 
bute facts in th}. direction, although at the same time they 
failed to see the essential points of comparison with the 
mammals. ‘This failing on their partis all the more explicable 
as the bird’s egg, which has always served as the prototype 
even of mammalian development, does not clearly bring out 
the fundamental distinction that exists between trophoblast 
and formative matter of the embryo. 

The gradual obliteration of this distinction may, perhaps, 
be ascribed to the fact that in these sauropsids, as in the 
ornithodelphia, a shell has developed, which naturally tends 
to relegate any outer larval layer to the pension list. 

Concerning the yolk accumulation in the sauropsidian ege, 
there is no trouble at all to suppose that the vesicular blasto- 
cyst of an early viviparous ancestor has gradually become 
yolk-laden. The contrary assumption, found in the hand- 
books, that the mammalian egg, while totally losing its yolk, 
has yet preserved the identical developmental features as the 
sauropsid, is, in reality, much more difficult to reconcile with 
sound evolutionary principles. 

We have seen that a simple clue to our understanding of 
the complicated foetal envelopes of the sauropsids and the 
mammialia is the assumption of a simple larval layer, one cell 
thick, among the invertebrate ancestors. 

We must be ready to admit that this one factor has un- 
doubtedly given rise to an endless number of variations and 
modifications in those innumerable families, genera, and 
species which have come and have gone, ever since the time 
when viviparity and terrestrial life became an established fact 


184. A, A. W. HUBRECHT. 


in the vertebrate kingdom. What is preserved to us in the 
recent fauna inhabiting this planet is only the faintest echo of 
the multitudinous and protean changes that have, during the 
course of time, succeeded one another. And it has been our 
mistake to attempt to co-ordinate the present stages of de- 
velopment with each other in such a sense that they were 
expected to represent, in lineary arrangement, the successive 
evolutionary stages of those foetal envelopes. 

How false the conclusions may be to which this method may 
lead us is best exemplified by what is at present often taught 
concerning, e.g., placentation, a phenomenon in which the 
outer larval layer, the trophoblast, plays such a prominent 
part. You will find in the text-books that this was started 
by what is called the diffuse placentation as it is at present 
met with in many ungulates, in the lemurs, and in certain 
Edentates. It is my conviction that this doctrine is utterly 
false. The diffuse placentation is no placentation at all! The 
horse and the lemur are, by birthright, aplacental animals, 
much more so than marsupials, such as Perameles aud 
Dasyurus, which have hitherto ranked among the Mammalia 
aplacentalia. And still, by careful comparison of various 
data, we can soon discover that the diffuse placentation, and 
that variety of it which is styled the polycotyledonary, far 
from being archaic or primitive, is, on the contrary, very largely 
a secondary modification. Among the living Carnivora we 
find several intermediate stages, not in the sense that these 
have been phylogenetic transitions, but in that wider sense 
that these Carnivora demonstrate the possibility how more 
intricate placentary structures may finally have led up to a 
diffuse placentation, as that of the horse and the pig, conse- 
quent upon an increase in the area of surface contact between 
mother and foetus. What was originally a small surface of 
intense interchange (Procavia) has then gradually become an 
extended surface, along which two epithelial layers, one 
maternal and one foetal, between the blood of the mother and 
the blood of the embryo, offered no impediment for a sufficient 
interchange of nutritive matter and of oxygen. 


THE F@TAL MEMBRANES OF THE VERTEBRATES. 185 


If we do not accept the starting-point in the placentation- 
process to be represented in the ungulate arrangement, a 
proposal which the systematic position of the Ungulata would 
in itself render doubtful, we must then look for another phylo- 
genetic sequence which will help us to rightly interpret that 
momentous process of placentation. And here the important 
results of Hill’s investigation of very intense placental pheno- 
mena in some marsupials, such as Perameles, have great 
weight. 

We may fairly conclude that kangaroos, phalangers, oppos- 
sums and other marsupials have only gradually become 
aplacentary, parallel to those other formidable changes which 
must have accompanied the elaboration of that peculiar type 
which we call our recent Didelphia, in which the dentition, 
the lactation, and those adaptations of the new-born animals 
for nutrition during their life inside the marsupium form such 
distinctive characters. 

And so if the Didelphia are in reality erratic Monodelphia 
secondarily modified and with an allantois that has been thrown 
out of the line of its normal development, with the exception 
of Perameles, Dasyurus, and in part Phascolarctos, then we 
have again to look, not amongst them, but amongst the 
Monodelphia, for such forms that can give us an indication as 
to what may have been the primitive stage of placentation. 

And I may here state that my own researches on the 
placentation of both primates and of insectivores have led me 
to the conclusion that we should look in quite another direction 
than the one alluded to above, which starts from diffuse 
placentation. In the earlier part of this address I have con- 
sidered those early phylogenetic stages when, in viviparous, 
air-breathing tetrapods, the larval layer, the trophoblast, 
found the most diverse possibilities open to it. 

I believe that those forms of which the embryonic tropho- 
blast actually attacked the maternal uterine mucosa phago- 
cytically were the pioneers towards the formation of what has 
later become the discoid placenta. In some forms, even 
among our recent mammals, that phagocytic attack is com- 


186 A. A. W. HUBRECHT. 


bined with a penetration of the whole blastocyst inside the 
maternal tissue, e.g. man, anthropomorphe, hedgehog, Gym- 
nura, and many rodents. This was naturally a far higher 
position of vantage than any peculiar fixation inside the 
lumen of the uterus, for now, when once the blastocyst was 
encapsuled inside its mother’s tissues, it could be most 
thoroughly bathed in maternal blood without any extravasa~ 
tion into the uterine lumen. ‘To take three examples of this 
we may allude to the guinea-pig, the hedgehog, and man. 
Still, all these utilise the favourable conditions offered to them, 
thanks to their situation inside a capsula or decidua capsularis, 
in a very different manner. 

here is a most remarkable amount of similarity between 
the hedgehog and man, as far as the conditions are concerned, 
which the mother offers to the young. But then the embryo 
itself of man has seen its way to much more intense utilisation 
of these favourable conditions than the hedgehog embryo has. 
Principally because the vascular system of the hedgehog 
develops in a sequence of stages, which serve to bring its 
area vasculosa on the umbilical vesicle in primary contact 
with the profusion of maternal blood by which the blastocyst 
is surrounded. 

On the contrary, in man this area vasculosa on the umbilical 
vesicle is not in contact at all with the maternal circulation. 
In man it is more devoted to hematopoietic functions, i. e. 
to the formation of new blood-corpuscles for the embryonic 
circulation. But in another respect the human blastocyst has 
got far ahead of that of the hedgehog, in so far as the de- 
veloping embryo has succeeded in vascularising its outer 
larval layer, its trophoblast, at a quite exceptionally early 
moment, without the aid of any allantoic outgrowth, and 
simply in consequence of a very early segregation of certain 
portions of the mesoblast, into which the entoderm sends both 
blood-vessels and blood-corpuscles. This very early vascula- 
risation of the trophoblast leads to a most intense osmotic 
interchange between the blood of mother aud child—far more 
intense that what obtains in the hedgehog, where an ompha- 


THE FETAL MEMBRANES OF THE VERTEBRATES. 187 


loidean placentation precedes an allantoidean one, the allan- 
tois being a vesicular outgrowth, as it is in so many mammals 
and in all sauropsids. 

I cannot refrain from looking upon the vascularisation of 
the outer larval layer or trophoblast, such as it occurs in man, 
in the monkeys, and in 'arsius, as the more primitive arrange- 
ment of the two. And inthat case the presence of a connect- 
ing stalk (Haftstiel) and the absence of a free allantois in 
man, monkeys, and Tarsius is not a secondary simplication, 
but a primary fact of high importance. What is known as 
the allantois tube inside the so-called Haftstiel or Bauch- 
stiel of man, monkeys, and T'arsius, is not the remnant of 
what was once a vesicular allantois, but a remnant of that 
part of the entoderm which has served towards the vasculari- 
sation of the trohpoblast. It is this portion of the entodermal 
surface which will become the free allantois in those other 
descendants of the primitive tetrapods, which have not 
adhered to the very direct line of utilising most fully and as 
early as possible all favourable circumstances. This most 
direct line leads up straight to the primates. Less direct lines, 
in which conditions of different or of slower vascularisation 
have.come to the foreground, are, however, represented in 
various orders of monodelphian mammals, and further in the 
Didelphia, the Ornithodelphia, and in the different subclasses of 
sauropsids. In the latter the allantois has grown to the 
dignity of a separate foetal membrane, which co-operates to 
the further ensheathing of the developing embryo, and which 
carries the blood-vessels for respiratory purposes to the inner 
surface of the egg-shell, whereas, in the ancestral viviparous 
forms, the same vessels were more directly distributed over 
the inner surface of the outer embryonic larval layer, in order 
to improve the nutritory conditions which had been inaugu- 
rated by phagocytic action of the trophoblast cells on the 
maternal tissues. 

This, then, is a short sketch and a rapid review of how the 
foetal membranes of the vertebrates may be looked upon if 
we make certain changes in the interpretations that have 


188 A. A. W. HUBRECHT. 


been hitherto adhered to, but by which latter nobody has as 
yet succeeded in clearing up the actual phylogenesis of these 
foetal membranes. 

Full and extensive investigations of all those numerous 
genera of mammals that have not yet been examined will, I 
hope, in due time give us occasion to complete or to modify the 
views here advocated. 

It was a great pleasure to me to offer them, tentatively, in 
an address which I was invited to give in the section of 
embryology of this Seventh International Congress—a section 
which, with good right, has been called into life for the first 
time at this meeting in Boston. Kmbryological problems 
have been attacked by American investigators with wonder- 
ful results, and the lucidity of exposition that is characteristic 
of so many of your embryological workers is only equalled 
by the beautiful transparency of the eggs of those marine 
animals on which so many important researches on cell-lineage 
have been conducted. 

‘hat I have been less clear is not only a congenital defect, 
but is parallel with the utter hopelessness of our expecting 
that we shall ever be able to follow the cell-lineage in the 
deeply hidden and exceedingly small mammalianeggs. Still, 
a full knowledge of that very cell-lineage would be eminently 
decisive for many of the questions that have occupied us in 
the course of this address, to which you have listened with so 
much patience. 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 189 


The Structure and Life-History of Crithidia 
melophagia (Flu), an Endo-parasite of the 
Sheep-Ked, Melophagus ovinus. 


By 
Annie Porter, B.Sc.Lond., 
Zoological Research Laboratory, University College, London. 


With Plates 12 and 15 and 15 Text-figures. 


ConrENTS. 
PAGE 
Introduction ‘ 3 : 5 : 190 
Material and Methods J : , 190 
Distribution of Parasites in the Tose : : 192 
Movements ‘ E , : 194 
Morphology: Pre- feselbes Stage . : E 197 
Flagellate Stage ‘ 197 
Post-flagellate Stage in the Beceun of 
the Host 2 
Longitudinal Division ; 202 
Hereditary Infection of Melo Attias us ovinus by C. 
melophagia : ; : 204 
Casual Infection . : : ' ; 207 
Environmental Effects 3 ; : 208 
General Remarks : : ; : 210 
Summary : 211 
Appendix 208 the Geemmcaes of a Spicoeiete (S. 
melophagi, n.sp.)in Melophagus ovinus . 213 


Appendix I]—Note on a Fungus found in the Mal- 
pighian Tubules and Intestine of Melophagus 
ovinus P 214 

Appendix IJI.—On the Oeeneencs of an fake coneultn 
in the Alimentary Canal of Melophagus ovinus, 
and its Significance in Relation to Crithidia 


melophagia : : , 216 
References to Literature . : 218 
Explanation of Plates. : ‘ ; 220 


VOL. 55, PART 2.—NEW SERIES. 13 


190 ANNIE PORTER. 


INTRODUCTION. 

Tue part played by insects as agents in the transmission of 
the pathogenic organisms of sleeping sickness and other pro- 
tozoal diseases gives great importance to the investigation of 
the parasites found within them. It is necessary for any- 
one seeking developmental stages of pathogenic flagellate 
Protozoa to have also a first-hand working knowledge of the 
possible flagellates that may be purely parasites of the insect 
involved, for certain stages of insect flagellates may resemble 
possible developmental phases of such organisms as 'l'rypano- 
somes. Much useful information regarding stages of flagel- 
lates can be gained from the study of such a parasite as 
Crithidia melophagia (Flu), occurring in the alimentary 
tract, ovaries, and ova of the sheep-‘‘ ked,’ Melophagus 
ovinus. This insect, which is blood-sucking, is also known 
as the sheep-‘‘ tick” or sheep-‘‘louse.” It belongs really to 
the Diptera (Hippoboscide), possessing extremely reduced 
wings. 

Crithidia melophagia (Flu) was recorded by E. Pfeiffer 
in 1905, but not named by him. The parasite is of peculiar 
interest, for I am able to bring forward evidence of a double 
mode of infection, both hereditary and casual. Swingle (1909) 
studied the flagellate stages and briefly described infection in 
the egg of Melophagus. Flu (1908) found parasites in the 
gut, ovaries, and larva, but was not clear as to the mode of 
infection (see p. 211). 

Owing to conditions of environment it was impossible to 
conduct the whole of this investigation in a large city. Con- 
sequently the work has entailed travelling, and I have to 
thank many friends in agricultural centres for their kindly 


help. 


MatertaL AND Meruops. 
Many specimens of Melophagus ovinus were examined 
during a long period of investigation, but owing to the 
effective operation of the dip laws in England there was 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 191 


much difficulty in obtaining the “keds.’! Indeed, it seems 
probable that the sheep-ked may soon become almost extinct 
in England. Those obtained came in very small numbers from 
many localities in the south of Hngland, namely, Sussex, 
Hampshire, Kent, Middlesex, and Gloucestershire. I also 
received a number of keds from different parts of Scotland, 
but these never contained the Crithidia. 

Many of the Melophagus, however, were infected by a 
fungus (see Appendix II). Where fungus was present 
Crithidia very rarely occurred. I shall show later, from 
experimental evidence, that this fungus was fatal to the 
Crithidia (p. 210). 

Unlike Swingle (1909), who found that practically every 
Melophagus he examined in Nebraska was infected with 
Crithidia, I found that this was very far from being the 
case. Much depended on the locality from which the Melo- 
phagus was obtained. The more heavily infected individuals 
came from the southern districts of England. Often entire 
stocks of keds from one locality proved to be uninfected. 
Again, it was impossible to keep keds alive more than three 
days after their removal from the sheep. 

Both young and adult Melophagus and many puparia in 
all stages of development were carefully examined. Raising 
puparia naturally upon a sheep was tried, but was not an easy 
matter, and as one could not be sure of having infected keds, 
there was always a percentage of uninfected puparia. 

For observations of the living organism two methods of pro- 
cedure were followed. ‘The alimentary canal was isolated and 
divided into cesophageal, crop, stomach, intestinal and rectal 
portions, which were separated one from another. These 
were either teased with needles, mounted in 0°75 per cent. 
salt solution, and covered, the cover-slip being carefully 
vaselined, or the contents of the isolated portions of the gut 
were expelled by gentle pressure, and these only were 
examined, being mounted as before. Alkaline methylene 


‘In this paper I shall frequently use the term “ked” to denote 
Melophagus ovinus. 


192 ANNIE PORTER. 


blue and neutral red were occasionally used as intra-vitam 
stains and were sometimes useful. 

For fresh preparations used in work on hereditary infection, 
the ovaries and gut were dissected out very carefully, kept 
as far as possible relatively in situ, and mounted in 0°75 per 
cent. NaCl solution. The behaviour of the Crithidia visible 
through the walls of the gut and their action when they 
passed out from it were then most carefully watched. 

I have attached very great importance to the study of the 
living organism in all its phases. 

For making permanent preparations the alimentary tract 
of the Dipteran host was carefully removed and divided into 
portions as before. ‘These isolated portions were usually 
teased very finely and fixed wet. Formalin vapour and osmic 
acid vapour were chiefly used for instantaneous fixation of the 
hanging-drop preparations, which were then spread. The 
preparations were subsequently treated with methyl or ethyl 
alcohol. Corrosive-acetic-alcohol (Schaudinn’s fluid) and 
Bouin’s fluid (sightly modified and containing a little alcohol) 
were also used for fixation. 

Various stains were employed. Giemsa’s stain gave some 
pretty results; thionin acted rapidly and well; iron-hema- 
toxylin, carefully differentiated with iron-alum, was very 
serviceable ; while gentian violet and Delafield’s hematoxylin 
were of great use, particularly in obtaining details of the 
membrane and flagellum. 

In the investigation of Crithidia melophagia, as in all 
other flagellates on which I have worked, I found that pre- 
parations mounted in neutral Canada balsam were superior 
to dry films or to films mounted in any other media. 

Preparations of ovaries, eggs, and puparia were treated 
similarly. Special methods adopted are detailed in the section 
dealing with hereditary infection (p. 204). 


DISTRIBUTION OF THE PARASITE IN THE Host. 


The Crithidia parasitic in the alimentary canal of Melo- 
phagus are often mixed with the blood obtained by the ked 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 193 


from the sheep. This blood from the sheep in the cesophagus, 
crop, and anterior part of the stomach of Melophagus is 
always fluid, and of an extremely bright red colour. That in 
the remaining part of the stomach is duller red but fluid, and 
in the intestine the blood, now semi-digested, is always 
darker in hue, sometimes brownish or greenish, while in the 
extreme rectum it is black. ‘The enhanced red colour in the 
anterior portions of the alimentary canal has been shown 
experimentally to be associated apparently with the presence 
of an anti-coagulin in the digestive tract of the sheep-ked 
(see Appendix III). 

Crithidia can be found throughout the length of the 
alimentary canal of Melophagus ovinus. In the anterior 
parts of the canal they are small, rounded, non-flagellated 
forms, which, when they come in contact with the blood, 
rapidly develop and divide, the products of division becoming 
the typical flagellates found throughout the rest of the canal. 
The parasites, after this rapid development, pass backwards 
towards the partly digested blood, which would appear to be 
a medium more suited to their requirements. In the posterior 
third of the stomach there are large numbers of young 
flagellates which form great aggregation rosettes (Pl. 12, 
fig. 45) and clumps, while true division rosettes are also 
present (Pl. 12, fig. 56). 

In the intestine the: same holds good. When many 
Crithidia are present in a ked, they usually swarm in the 
fore-part of the intestine. Repeated division occurs in the 
intestine, so that small flagellates are found in the rectum. 
Most of these attach themselves to the gut-wall or to débris 
and encyst, the resting (post-flagellate) stage of the parasite 
then being found on the walls of the rectum and in the 
feeces. 

The ovaries and ova serve as places in which a kind of 
post-flagellate development occurs, the ova being penetrated 
by flagellate forms of Crithidia, which rapidly lose their 
flagella and ultimately round themselves off, and pass through 
a resting stage (Pl. 15, figs. 57-94). 


194 ANNIE PORTER. 


The Malpighian tubules of Melophagus ovinus are 
sometimes invaded by Crithidia melophagia, but this is 
not common. 

Parasites were more numerous in female than in male keds. 

Repeated investigation of sheep’s blood failed to show the 
presence of any flagellate therein. Flu and Swingle obtained 
similar results. C. melophagia is, then, purely a parasite 
of Melophagus ovinus, 


MoveMENTs. 


The movements of C. melophagia are very vigorous. 
The parasites are even more active than C. gerridis (see 
Porter [1909], p. 352). As in C. gerridis, the membrane 
takes an important share in locomotion, but the movements 
of the body of C. melophagia are not so noticeable as in 
the parasite of the water-bug. 

When C. melophagia was examined under the water 
immersion (2°5 mm.) objective, the movements of the less 
active organisms could be analysed. In progression the 
organism moves with its flagellum foremost, and the latter 
executes vigorous, slightly spiral, boring movements. The 
body also aids in progression, for waves pass from the 
posterior end towards the flagellum, causing a series of 
peristaltic-like swellings. The body of the parasite seems to 
become shorter during this period, and then by relaxing to 
move forwards. The bead-like swellings due to undulatory 
movements are more noticeable in certain areas, and in the 
hving organism myonemes could be sometimes seen both on 
the body and in the membrane in these regions. Flu has 
also figured myonemes on some of the parasites he drew, and 
observation of them in life confirms his work, but it was with 
the greatest difficulty that I could find myonemes in stained 
specimens (Pl. 12, figs. 17, 18, 40, 42, 45). 

The body of C. melophagia, compared with that of C. 
gerridis, is relatively rigid, but slight twisting movements do 
occur. The previous workers on C. melophagia are agreed 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA, 195 


as to this rigidity. The anterior end, to which the flagellum 
and undulating membrane is attached, is naturally more 
flexible than the posterior end, and its movements are more 
marked. 

Movements of contraction of the posterior end of the body 
of C. melophagia result ina temporary concentration of the 
protoplasm around the nucleus of the organism. The body 
then resembles a short, ‘hick pear, drawn out at its anterior 
end into a long, narrow stalk. Sometimes the body remains 
in this condition, which is fairly common in forms about to 
encyst, and in such forms withdrawal or degeneration of the 
flagellum, followed by the secretion of a thin gelatinous wall, 
completes the encystment. In other parasites from the 
stomach, where no encystment occurs, this concentration of 
the protoplasm in the nuclear region is not so marked, and 
when relaxation occurs the organism is propelled forward with 
a very slight jerk, and repetition of the contraction follows, 
as has been before described. ‘I'he jerking is never so 
marked as in Herpetomonas, for the membrane has the 
effect of producing smoothness of motion. 

Reversal of the direction of motion occurs and is very 
rapid. ‘he flagellum swings out, describing a semi-circle, of 
which the body acts as the diameter for an instant, but the 
force of the movement of the flagellum is so great that the 
body also swings outwards in a line with the flagellum, and 
the organism moves away, not exactly in the same course as 
before, but in one at a very small angle to it. The path of 
the organism is frequently parabolic in nature. 

Many peculiar movements can be observed when C. melo- 
phagia is endeavouring to free itself from débris in the 
lumen of the gut. Much writhing, both of the flagellum and 
body of such a parasite, is then seen, and the organism often 
swings round and round, the point of attachment serving as 
the centre of rotation. If the posterior end should be 
attached, the flagellum executes violent lashings and spiral 
movements, these latter not being, as a rule, very noticeable 
in the normal organism. 


196 ANNIE PORTER. 


Occasionally I have seen the flagellum and membrane of 
specimens of C. melophagia torn away from the body, and 
for a few seconds after, the flagellum executed intermittent 
flickers or lashing movements before it finally became still. 

Ageregation-rosettes (Pl. 12, figs. 41,45; Pl. 15, figs. 95, 
96) are common in C. melophagia. Rosettes seem to move 
fairly as a whole, and I have watched them rotate rather 
quickly. Each individual of such a rosette is attached by 
its flagellum to débris, usually epithelial in nature, and 
moves up and down in a slightly inclined plane. 

In division the movements of the daughter organisms are 
very noticeable. I will defer the description of their motion 
until division is discussed. 

During encystiment in the rectum of the host, which occurs 
with some of the parasites, movement of the nucleus towards 
the flagellar end of the organism occurred. I have also seen 
the migration of the nucleus from the mid-region of the body 
to near the flagellum during periods of violent movement of 
the latter organella. I have never seen migration of the 
blepharoplast in living organisms under similar conditions, 
though it may occur at times, since blepharoplasts can occa- 
sionally be found in the post-nuclear region (PI. 12, figs. 40, 
42), as well as by the side of the nucleus (PI. 12, fig. 33) in 
different stained specimens. By far the commonest position 
for the blepharoplast is the pre-nuclear one. ‘The other 
movements occurring during encystment will be described in 
the section of the paper dealing with that subject (see p. 200 
and text-figures 1-10). 


MorpuHo.oey. 


The life-cycle of Crithidia melophagia may be con- 
veniently divided into three stages, which gradually merge 
into one another. They are—the pre-flagellate, flagellate, and 
post-flagellate stages. The morphology of these forms may 
now be described. 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA, 197 


The Pre-flagellate Stage. 

The early pre-flagellate stages of C. melophagia are 
more or less oval or rounded bodies (PI. 12, figs. 1-6), varying 
from 4°54 to 6 long, and from ly to 45 broad. They 
are most abundant in the fore-gut of young Melophagus, 
but the pre-flagellate stage is passed through with great 
rapidity and is easily missed, ‘This probably accounts for the 
very brief references to these small forms by Flu and Swingle. 

The protoplasm of the pre-flagellate forms is very finely 
granular (Pl. 12, figs. 1-5). ‘he nucleus is usually round 
and not quite central in position (Pl. 12, figs. 1, 9-12). The 
bar-like blepharoplast (kinetonucleus) is very deeply staining, 
and lies either below (Pl. 12, figs. 2, 10) or to one side of the 
nucleus (Pl. 12, figs. 1, 6). A chromatophile area with its 
chromatin in a very diffuse condition is sometimes fairly 
prominent, and from this a fine thread arises, which grows 
outwards, forming the flagellum (PI. 12, figs. 9, 10), and 
appearing to draw out the end of the body with it (PI. 12, 
figs. 11-13), while the periplast of the body forms the mem- 
brane (Pl. 12, figs. 14-20). The posterior end elongates at 
the same time (figs. 16-18) and the flagellate form (PI. 12, 
figs. 19, 20) is assumed. This.development is in accord with 
that of C. gerridis and C. tabani, and I have watched these 
processes in living specimens of both C. gerridis and C. 
melophagia. 

Division of pre-flagellate forms can occur before the develop- 
ment of the flagella (Pl. 12, figs. 5, 4). 'his will be described 
in the section dealing with division. 


The Flagellate Stage. 

The mature flagellates vary very much in size, the variation 
being due to division and growth. Very large forms (PI. 12, 
figs. 44, 45) may beas much as 50 uw to 754 long, this measure- 
ment including the flagellum,’ while short forms just flagel- 


1 Tt is almost impossible to differentiate between the limiting areas 
of the body, the membrane and the free flagellum of C. melophagia, 
as so much variation occurs in different specimens. 


198 ANNIE PORTER. 


lated (Pl. 12, figs. 18, 19) in the crop, or the small forms 
produced by division prior to encystment (PI. 12, figs. 20, 
21; 99) are very much smaller (12 to 20, long). The 
breadth of the flagellates varies from 1°5 x to 2°8 yu. 

The protoplasm of C. melophagia is very slightly alveolar 
or almost hyaline, differing therein from the more alveolar 
protoplasm of C. gerridis. ‘There is no suggestion of large 
permanent vacuoles or of a cyto-pharynx. Occasionally the 
protoplasm is more granular at the posterior end (PI. 12, figs. 
30, 34) and slight alveolation occurs there. At the anterior 
end, near the origin of the flagellum, the remains of the 
chromatic area, from which the flagellum arose, sometimes 
persist. 

The nucleus (trophonucleus) of C. melophagia is oval 
(Pl. 12, figs. 21-24) or rounded (figs. 26, 50, 32) and some- 
what vesicular. There is a fair amount of chromatin present, 
which may consist of a number of very fine granules, evenly 
distributed (fig. 32), or the chromatin may be concentrated 
into about eight masses (fig. 44), or, as is often the case, the 
chromatin is present in the form of bars (figs. 25-29), which 
sometimes extend across the whole breadth of the nucleus 
(figs. 34-37), less frequently across part of its breadth 
(figs. 24, 42), or in an even more rare condition dots and 
bars occur in the nucleus of the same organism (figs. 30, 59). 
In certain cases the chromatin ot the nucleus may be con- 
centrated into a central mass (fig. 25). 

The nuclear membrane is fairly distinct in most of the 
specimens I have examined. | think that such a membrane 
must be present to keep together the nuclear material during 
the migrations of the nucleus seen during life. 

The blepharoplast (kinetonucleus) of C. melophagia 
is very evident in a stained preparation, for it colours deeply 
whatever stain be employed. Like the nucleus, it can also 
be seen in life as a small bright refractile bar. In some 
cases it is slightly bowed or curved (Pl. 12, fig. 32), or oval 
(Pl. 12, fig. 34). It is dumb-bell-shaped in forms about to 
divide (Pl. 12, fig. 44). The blepharoplast, which is typically 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 199 


rod-like, usually lies transversely across the organism (PI. 12, 
figs. 21-28). It is exceptional to find it in any position other 
than anterior to the nucleus, though on a few occasions the 
blepharoplast was at the posterior end of the body (PI. 12, 
figs. 46, 42), but in these cases the flagellum originated in a 
pre-nuclear position. 

As a rule the blepharoplast shows no differentiation of 
structure (PI. 12, figs. 21-39), but sometimes in dividing 
forms, in which the blepharoplast is dumb-bell shaped, there 
seems to be a concentration of chromatin in the ends of the 
dumb-bell (Pl. 12, figs. 40, 44, 45). A clear area (PI. 12, 
fig. 51) is often present around the blepharoplast. 

Chromidia are present, scattered in the general proto- 
plasm (Pl. 12, figs. 25, 37, 39). They stain in the same way 
as the nucleus, aud less densely than the blepharoplast. 
The occurrence of such chromatoid granules at division 
(Pl. 12, fig. 45) suggests that they have been given off from 
the nucleus into the general protoplasm, and exercise some 
controlling influence over the same. 

The undulating membrane and the flagellum.— 
The flagellum originates from a chromatic area in the pre- 
flagellate form, and is attached to the body by a narrow mem- 
brane (PI. 12, figs. 21-46), which is a periplastic outgrowth 
of the anterior end of the body. There is but one flagellum 
in any single, undividing individual (Pl. 12, figs. 21-39). 
The flagellum is thick, but gets thinner towards its free end 
(Pl. 12, figs. 40, 45). At times it appears to show very fine 
transverse striations. 

In stained specimens the membrane sometimes shows myo- 
nemes (PI. 12, figs. 39, 42, 45), though, curiously enough, the 
myonemes were much more obvious in some of the living 
specimens that I examined. Flu described myonemes in 
C. melophagia, but figured the myonemes as accompanying 
a central spindle. ‘This latter feature I have never seen. 

A basal granule (blepharoplast of Minchin) is often 
present (Pl. 12, figs. 17, 27, 33, 42, 45) between the point of 
origin of the flagellum and the blepharoplast (kinetonucleus). 


200 ANNIE PORTER. 


The Post-flagellate Stage of C. melophagia in th 
Rectum of Melophagus ovinus. 


The preparation of Crithidia melophagia for life outside 
the body of the host occurs in the rectum of the sheep-ked. 
Large numbers of small flagellates (Pl. 12, figs. 27-29) are 
present in the hind gut, also some forms in process of division 
(Pl. 12, figs. 97, 98). The small forms attach themselves to 
the wall of the rectum and encyst there, but encystment can 
be watched when the rectal contents are expressed on to a 
slide and examined under the microscope. ‘The flagellate 
(text-fig. 1) at first executes violent lashing movements with 
its fagellum, and during this motion migration of the nucleus 
nearer the tlagellar end of the organism frequently occurs 
(text-fio. 2). At the same time the body of the Crithidia 
shortens and thickens (text-figs. 3, 4; Pl. 13, fig. 100), waves 
of contraction passing rhythmically down the body, which 
gradually may become somewhat triangular (text-fig. 5; Pl. 13, 
fig. 101). The flagellum meanwhile shortens (text-figs. 5, 6), 
and the organism may bend on itself (text-figs. 6,7) during 
this period. Concentration of the protoplasm occurs, the 
flagellum becomes less wavy (text-fig. 7), and, little by little, 
it contracts nearer the body (text-figs. 8,9; Pl. 13, figs. 102- 
106) and is withdrawn, the parasite becoming oval (text-fig. 
10; Pl. 18, figs. 109-112). The organism at this time 
becomes surrounded by a thin layer of refractile, gelatinous 
substance, which rapidly hardens to form a closely adherent 
resistant cyst-wall. The oval bodies (Pl. 15, figs. 109-114) 
so produced are post-flagellate forms, which become detached 
from the walls of the rectum, and pass out with the feeces of 
the ked, from which feces they can be recovered. These 
cysts, which measure from 2°5 «7 to 5°5 uw by 15 to 3, serve 
for the infection of other Melophagus ovinus. 

All Crithidia melophagia do not go through a post- 
flagellate stage in the gut of their host. Some, after passing 
a portion of their existence as flagellates in the gut of the ked, 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA, 201 


pierce the walls of the alimentary tract and make their way 
to the ovaries of the ked, where their development is con- 


tinued. 


TEXT-FIGURES 1-10. 


9 10. 


Encystment of Crithidia melophagia in the rectum. 


Text- figs. 1-5.—Parasite rounding off and flagellum disappearing. 
Text- figs. 6-7.—Show bending of | parasite on “itself. 
Text- figs. 8-10.—Final stages in loss of flagellum and assumption 


of typical eye form. 


Swingle (1909, p. 104) has described thick-walled cysts. 
I have but rarely seen the thick-walled forms (PI. 13, fig. 114), 
most of the cysts found being thin-walled. 


202 ANNIE PORTER. 


LONGITUDINAL DIvIsIon. 


The longitudinal division of the living organism has been 
frequently watched. While the movements of the dividing 
flagellates are noticeable, those of the smaller dividing pre- 
flagellates are far less marked. 

When a flagellate is about to divide, the protoplasm of the 
posterior end concentrates somewhat in the nuclear region, 
and the organism appears to shorten. The protoplasm 
migrates from the centre of the parasite towards the sides, so 
that a comparatively clear area is left at the centre (PI. 12, 
fiz. 46). The greatest change at this stage is seen in the 
blepharoplast and flagellum. The blepharoplast becomes 
slightly dumb-bell-shaped (Pl. 12, figs. 44, 45) and gradually 
constricts into two (Pl. 12, fig. 46). The flagellum splits 
rapidly at the body end (PI. 12, fig. 46), and then, more slowly, 
the halves become free. The nucleus meanwhile becomes 
slightly indented in the median line (PI, 12, fig. 46) and then 
gradually constricts into two, the halves migrating to the 
periphery (PI. 12, fig. 47). During this nuclear division the 
daughter-flagella execute very vigorous lashing movements, 
and a constriction appears at the flagellar end of the parent 
organism. A split appears at this end (Pl. 12, figs. 47-49), 
and, at the same time, vacuoles in the clear median area fuse, 
and thus the extension of the split is facilitated. The daughter- 
organisms rapidly separate from one another, their appear- 
ance at times being suggestive of diverging curved calipers 
(Pl. 12, figs. 51, 52). At length the two are practically in a 
straight line (Pl. 12, figs. 53-55), in which condition they 
remain for a short time and then finally separate. 

The division of the pre-flagellate forms is initiated by the 
division of the blepharoplast, and is followed by the division 
of the nucleus and the appearance of vacuoles. A slight split 
appears at one end (PI. 12, fig. 3), and the organism remains 
in this condition until the flagellum of each half has partly 
grown, when final separation 1s effected by their movements. 

Sometimes repeated division of a pre-flagellate form occurs 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA, 203 


and a rosette (Pl. 12, fig. 4) is produced, but the rapidity of 
the process of formation of flagella causes short duration of 
the rosette stage. On the other hand, repeated longitudinal 
division of flagellated individuals occurs, and as the individuals 
so produced do not separate immediately, rosettes (PI. 12, 
fig. 56) are formed. In division, the posterior ends of the 
daughter-organisms are the last parts to separate. As the 
daughter-forms remain in proximity and themselves proceed 
to divide with rapidity, true division-rosettes are formed, in 
which the posterior ends of the organisms are central, while 
the flagella radiate out from the common centre. Such 
division-rosettes (PI. 12, fig. 56) differ from the aggregation- 
rosettes (Pl. 12, figs. 41, 43; Pl. 15, figs. 95, 96) where the 
organisms become attached by their flagella. The distinction 
between the two forms of rosettes has not been shown by 
previous workers on C. melophagia. 

Longitudinal division results in the formation of both 
equal and sub-equal daughter forms. 

While the occurrence of equal longitudinal fission is the 
commoner (Pl. 12, figs. 50, 54, 55), I have seen cases of 
marked inequality in the size of the daughter-parasites, the 
one being very thin and narrow, the other considerably 
broader and thicker (Pl. 12, figs. 51, 53). As the entire 
process of sub-equal division has been watched in living 
organisms, there is no possibility of it beimg mistaken for 
anything else. The polymorphism resultant on division is 
strongly against the idea that there are sexual forms of 
Crithidia, and I have never seen the slightest indication 
that there is sexual dimorphism, in C. melophagia, C. 
gerridis, Herpetomonas jaculum, H. musce domes- 
tice, H. culicis, and a new Herpetomonas from Vespa 
crabro, all of which [I have examined in the living con- 
dition (see Porter [1909] on C. gerridis and H. jaculum). 

Division, usually twice repeated, is found to occur in 
parasites destined to encyst, and the resultant forms are very 
small. The first division is of the usual flagellate type 
(Pl. 13, fig. 97). The process of the second division rather 


204 ANNIE PORTER. 


resembles that of the pre-flagellate stages, for before it is 
accomplished the flagella have almost disappeared. Some- 
times no flagellum is visible at all, and the parasites look 
hike dividing cysts. 

On rare occasions the posterior end of a flagellate has 
divided before the anterior end (PI. 15, fig. 98). 


Tue Herepirary InFectioN OF MELOPHAGUS OVINUS BY 
CRITHIDIA MELOPHAGIA. 


Casual infection of Melophagus ovinus by the ingestion 
of post-flagellate cysts of Crithidia melophagia is fairly 
easily observed. The development of the parasite in the egg 
can only be studied with difficulty. I now wish to give a 
fuller account than exists up to the present of the processes 
leading up to the birth of Melophagus infected with 
Crithidia melophagia. 

The first point to be determined was the way in which the 
Crithidia reached the egg. Infected Melophagus were 
carefully dissected so that no rupture of the gut was made. 
The ovaries also were dissected out and kept as far as 
possible in the position beside the gut that they occupied in 
life. Crithidia could be seen through the gut-wall moving 
actively about. Suddenly they concentrated in one place 
and soon began to pass through the wall, their posterior 
(blunt) end first. They rapidly swam direct to the ovaries 
and penetrated them in the same way, that is, with the non- 
flagellar end first. The flagellum was very rarely used as a 
boring organ to allow of the passage of the organism. 

Penetration of the ovaries of their host by the parasites 
occurs in other cases, e.g. C. gerridis, H. jaculum, but 
the ova are apparently unattacked and the flagellates simply 
degenerate. But in the case of C. melophagia the organisms 
(Pl. 13, figs. 57, 59) make their way rapidly to the ova, to 
which they cling, whether the ova are mature or immature. 
In some cases one Crithidia only enters the egg (Pl. 13, 
fig. 58); at other times several penetrate it at once. In 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 205 


penetration the blunt end of the flagellate enters the egg 
first. Occasionally the flagella are cast off as the Crithidia 
pass into the egg and remain on the outside. | 

In the case of older ova, the parasites seem to penetrate 
the egg at a definite spot (Pl. 13, fig. 58), which probably 
becomes the mouth of the embryo. Parasites invading older 
embryos enter by the embryonic mouth. Like Swingle I did 
not find parasites in the milk-glands or milk of Melo- 
phagus. 

In investigations of the stages of C. melophagia in 
the egg and puparia I found that smear preparations 
were preferable to sections. Greater rapidity of manipula- 
tion and thinner preparations could be obtained by this 
means. 

The method adopted was to prick the egg or open the 
young puparium and express the contents on to a slide. The 
contents were at once fixed and then were allowed to flow 
over the slide, so that no artificial spreading was required, 
and therefore no mechanical distortion or tearing of the 
parasites could occur. The preparations so made contained 
much fatty matter. The slides were treated with ether to 
remove the fat, and then after washing with absolute alcohol 
were stained and mounted in the usual manner. 

Once within the egg the parasite gradually loses its flagel- 
lum (PI. 13, figs. 61-63). This may be cast off entire, for 
flagella are found floating freely in the vitellus of eggs that 
had been treated with the utmost care in the manner pre- 
viously detailed. In many cases the flagellum appears to 
be gradually absorbed (PI. 13, figs. 64, 66). Longitudinal 
division of the flagellates in the egg may occur, though 
rarely. 

The protoplasm of the Crithidia then concentrates round 
the nucleus and blepharoplast (PI. 13, figs. 64-69) and the para- 
site gradually becomes more or less rounded (PI. 13, figs. 70- 
73). Multiple division of both nucleus and blepharoplast 
then occurs (PI. 13, figs. 74-77), and the daughter-blepharo- 
plasts appear to pass outwards towards the periphery (PI. 13, 

VOL. 55, PART 2,—NEW SERIES. 14 


206 ANNIE PORTER. 


figs. 76, 77). A “plasmodial”! form (PI. 13, figs. 75, 77) is 
thus assumed. The protoplasm collects around the nuclei, 
and gradually fragmentation of the ‘ plasmodium ” occurs, 
the result being the formation of a number of small bodies, 
which rapidly round off, forming definite resting bodies (PI. 15, 
figs. 78-81). Sometimes these resting bodies remain in 
proximity to one another, so forming groups (Pl. 13, figs. 80, 
81). The parasites now measure only l'5 to 4m long and 
ly to 2°54 broad. Sometimes one chromatic mass (Pl. 13, 
fig. 82) only can be distinguished. Often both nucleus and 
blepharoplast (PI. 13, figs. 80, 81, 83, 84) are present. 

As the embryo grows the rounded forms of the parasite in 
the stomach (which is the chief cavity within the young Melo- 
phagus) also grow (PI. 13, figs. 82-84). The Crithidia 
then undergo multiple division, small rosettes (Pl. 13, figs. 
85-88), analogous to pre-flagellate rosettes, being produced. 
The division clusters may separate, giving rise to small, pear- 
shaped or ovoid individuals (PI. 13, figs. 89-94), or they may 
remain as a rosette (Pl. 13, fig. 88) for some time. Whether 
the Crithidia remain as groups or become isolated as oval 
non-flagellated bodies, they undergo no further development 
for a considerable period. In fact, when the young Melo- 
phagus is hatched, a month after extrusion of the puparium, 
there is still no further differentiation in the parasite. 

Freshly hatched Melophagus do not contain the fully 
developed flagellates, but the rounded or pear-shaped pre- 
flagellate forms (Pl. 13, figs. 92-94) and rosettes (Pl. 13, fig. 
88) may be present. The parasites appear to lie dormant for 
a day or two, during which time the young insect does not 
appear to suck blood. Soon after the first meal of blood is 
taken, rapid development of the pre-flagellate forms occurs, 
and the adult flagellate form of the Crithidia is quickly 
assumed. 


‘A plasmodium is really a multinucleate mass of protoplasm 
formed by fusion of small amcebe. However, the term is sometimes 
used, as in describing certain Haplosporidia, for a multinucleate 
mass of protoplasm formed by division. 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 207 


Casual INFECTION. 


The method of cross-infection in many species of Crithidia 
has not been demonstrated, but in the cases known the casual 
or contaminative method seems to prevail. The post-flagellate 
stages of Crithidia gerridis and C. tabani are known, 
and the cysts of these parasites are shed in the feces of the 
insectan hosts. ‘The crithidian cysts are swallowed by new 
hosts when they feed on material accidentally contaminated 
by the faces of their neighbours. The cysts then develop 
in the alimentary tracts of the new hosts. Melophagus 
ovinus also becomes infected with its Crithidia by the 
casual method. 

When studying C. melophagia I have noticed that the 
feeces of Melophagus ovinus are voided near spots on the 
sheep from which blood has recently been sucked (particularly 
is this the case at times of extrusion of puparia); that the 
feeces contain crithidian, post-flagellate cysts, and sometimes 
active flagellates; and that other Melophagus, feeding at 
the same spot, have thrust their proboscides into the semi- 
fluid faeces to reach the blood of the sheep. Ingestion of 
cysts under such circumstances is easy. The ingestion of 
feces has been seen particularly well when a number of keds 
have been kept confined to a small area of the sheep’s body. 

At shearing a slight injury was caused to one sheep, and 
the keds seemed to collect round the small bleeding patch. 
Their habits were carefully observed then, and were similar 
to those described above. I do not agree with Swingle that 
casual infection is only a remote possibility ; to my mind itis 
a certainty. 

A modified contaminative cross-infection is rendered pos- 
sible by the cannibalistic habit of Melophagus ovinus. 
The keds have been seen to attack one another, the point of 
seizure invariably being at the end of the abdomen near the 
anus. When a ked so attacked has been freed from its 
aggressor and then dissected, I have found that the abdominal 
cavity was almost empty, the viscera having been devoured 


208 ANNIE PORTER. 


by the attacking ked. By this cannibalistic habit it is 
possible for Melophagus ovinus to acquire practically 
every stage of Crithidia melophagia direct, and this is 
probably a subsidiary method of spreading the parasite. 


ENVIRONMENTAL EFFECTS. 


Crithidia melophagia shows less response to slight 
changes of environment than does C. gerridis or Her- 
petomonas jaculum, both of which I have studied. Never- 
theless, under certain conditions remarkable effects have 
been produced by relatively simple means, and these may now 
be recorded. 

(1) Response to light.—lIncreased intensity of white 
light produces increased velocity of movement of Crithidia 
melophagia. 

Green light somewhat retards the movements of the 
organism. This is also the case with Herpetomonas 
jaculum. 

Intense light causes aggregation-rosettes of C. melo- 
phagia to separate. 

C.melophagia lives very much longer in diffuse light 
than in bright light. 

(2) Responseto changes of temperature.—C. melo- 
phagia can live at a temperature just below that of the 
blood of the sheep, but the flagellates are killed at a tem- 
perature above 40° C. 

At room temperature (15° C.) the parasites will live for 
several hours. 

(3) Response to change of medium.—Though the 
flagellates normally live surrounded by fluid blood (a diseus- 
sion of which will be given in Appendix III), yet they can 
live in other media and can resist the effects of such media 
to varying degrees. 

(a) Tap-water when added to the parasites in the gut- 
liquid seemed to have little effect. Though the movements 
of the flagellate became slightly more active, this was possibly 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 209 


due to the greater space in which the parasites could move, 
the débris being distributed over « greater area than 
before. 

(6) 0°75 per cent. NaCl solution increased the 
activity of the parasites. 

Five parts of tap-water added to one part of 0°75 per cent. 
NaCl solution containing Crithidia caused the flagellates 
to move more rapidly, the spiral boring movements of the 
flagellum becoming more exaggerated. 

(c) Caustic potash.—T'wo per cent. solution killed all 
the Crithidia within a minute; 1 per cent. potash solution 
killed them in from seven to twelve minutes, but their bodies 
were not dissolved, this pointing to the chitinoid nature of 
the thin periplast or ectoplasm. 

(d) Acetic acid.—One third per cent. aqueous solution 
had the effect of swelling the parasites, which then died. 

(e) Grape-sugar.—A most remarkable effect was that 
produced on C. melophagia by a solution containing a very 
small amount of grape-sugar. When this was added to the 
parasites they commenced to divide very rapidly, and many 
divisions occurred. ‘To ascertain if there were a connection 
between this division and the occurrence of sugar in the 
natural medium of the parasites, some experiments were 
made. ‘The results were as follows: 

(i) Sheep-serum contains a very small amount of grape- 
sugar. 

(ii) The liquid obtained when wool cut from the sheep was 
‘boiled with water and then concentrated also showed traces 
of sugar. There were, then, these two sources from which the 
ked probably could obtain minute quantities of sugar. It is 
possible that the traces of sugar may take a small share in 
stimulating division of C. melophagia, which goes on more 
rapidly in the stomach of the ked than elsewhere. 

(f) Fresh blood (human or sheep’s) added to a pre- 
paration of living Crithidia caused the parasites to move 
away to areas where the blood was somewhat less concentrated, 
where they proceeded to divide rapidly. 


210 ANNIE PORTER. 


(gy) Dilute glycerine killed C. melophagia almost at 
once. Vaseline had the same effect after a very short time. 

(4) Effect of a parasitic fungus of Melophagus 
ovinus on C. melophagia.—The presence of a fungus in 
Melophagus ovinus has already been mentioned. As I 
very rarely found the fungus and Crithidia co-existing in a 
ked, it was deemed advisable to find out any possible inter- 
relation of the two parasites. The Malpighian tubules of the 
ked—often blocked with fungus—were the most heavily 
infected organs. Fungus taken from the Malpighian tubes 
was crushed with a little water. The emulsion, which 
probably contained an enzyme, was added to a preparation of 
actively moving C. melophagia. ‘he movements of the 
flagellates slowed at once, their protoplasm became much 
more vacuolated, and the parasites appeared to burst. After 
seven to nine minutes no living Crithidia were to be seen. 

The fungus-infected Melophagus ovinus seems widely 
distributed. Specimens from Scotland were practically 
always heavily infected with it, and some keds from each 
locality tried in England also were infected. These keds 
very rarely contained Crithidia. The fungus seems to 
have a pathogenic action upon the flagellate, and I believe 
that the co-existence of the fungus and Crithidia for long 
together is almost impossible. 


GENERAL REMARKS. 

Regarding the previous work done on the genus Crithidia, 
I have already noted most of the memoirs dealing with the 
subject in my paper on Crithidia gerridis (1909). Conse- 
quently the remarks now appended relate especially to the 
flagellate of Melophagus ovinus. 

E. Pfeiffer (1905) first briefly described a flagellate as 
occurring in the gut of Melophagus ovinus. He mentions 
that L. Pfeiffer had seen and recorded the parasite in 1895. 
The flagellate stage only was described, and no definite name 
was given to the organism, which was stated to be “like a 
trypanosome.” 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 211 


P. C. Flu (1908) published an account of the flagellate 
under the name of Crithidia melophagia. Flu stated that 
he saw parasites in the ovary of Melophagus, and small 
forms in the larva, but was unable to determine the mode of 
infection of the host. 

L. D. Swingle (1909), working in Nebraska, wrote a 
description of the parasite, calling it C. melophagi. From 
a private communication I learn that Swingle’s work was 
completed, but not published, before Flu’s paper appeared, 
thus accounting for the specific name melophagi (described 
as new), which cannot stand. ‘he chief value of Swingle’s 
work lies in the fact that he described rounded and 
*‘»nlasmodial”’ stages of the parasite as occurring in the egg 
of the host. 

While Swingle was working in Nebraska, I was investiga- 
ting the parasite independently in England. It gives me 
great pleasure to be able to confirm Swingle’s work, and to 
add many more details concerning the modes of infection of 
the parasite and its general life-history. 


SUMMARY. 


(1) Crithidia melophagia is a flagellate occurring in 
the alimentary tract, ovaries, ova, and puparia of Melo- 
phagus ovinus. 

(2) The parasite has three stages in its existence, a pre- 
flagellate stage (PI. 12, figs. 1-20), passed chiefly in the crop 
and fore-gut of the insect host, a flagellate stage (PI. 12, figs. 
21-44), occurring chiefly in the posterior two thirds of the 
gut, and a post-flagellate stage, occurring either in the 
rectum and feces (Pl. 13, figs. 97-114) or in the ova and 
pupe (Pl. 13, figs. 57-94). 

(3) ‘The pre-flagellate stage is passed through very rapidly. 
These parasites are small, usually oval bodies, lu to 4°5 uw by 
4°5 uw to 6 w, with round nuclei and bar-lke blepharoplasts. 
The flagellum arises near the blepharoplast from a chromato- 
phile area. Division of pre-flagellates may occur (Pl. 12, fig. 4). 


212 ANNIE PORTER. 


(4) The flagellate forms are from 12 » to 75 long, and 
15 w to 2°8 « broad (including the flagellum). The general 
protoplasm is slightly alveolar. ‘The nucleus is vesicular. 
he blepharoplast is well marked, rod-like, usually anterior 
to the nucleus, and generally homogeneous. 

Chromidia may occur as isolated granules. 

(0) The undulating membrane and flagellum are well 
marked. ‘here are indications of myonemes (PI. 12, figs. 40, 
45) in some stained specimens, but the myonemes are more 
evident in some living specimens. The membrane is of great 
use in securing smoothness of motion. ‘lhe flagellum is long 
and forms a chromatic edge to the membrane. A _ basal 
granule may occur near the root of the flagellum. 

(6) The post-flagellate stage in the host’s rectum (Pl. 13, 
figs. 97-114) gives rise to resistant (resting) bodies that are 
passed out as cysts with the feces and serve to infect new 
hosts. ‘The cysts measure, on the average, 4 u by 2°5 uw. The 
flagellates divide, usually twice, and the four small forms 
thus produced lose their flagella, become round, and then 
invested with a thin gelatinous wall. 

(7) The post-flagellate stages in the ova and puparia of 
Melophagus (Pl. 15, figs. 57-94) serve for the hereditary 
transmission of C. melophagia. ‘The flagellates pass 
through the wall of the gut near the anterior ends of the 
ovaries, swarm towards and enter the ovaries and penetrate 
the ova—the posterior (aflagellar) end of the parasite being 
used in penetration. Within the ova each parasite loses its 
flagellum and becomes ovoid or rounded (PI. 13, figs. 64-73). 
Nuclear multiplication follows and “plasmodial” forms are 
produced (Pl. 13, figs. 74-77). These give rise to small, 
rounded bodies (Pl. 13, figs. 83, 84) about 3 u by 2 « which 
undergo multiple fission to form rosettes (PI. 13, fig. 88), 
which give rise to the typical pre- flagellates. 

(8) The young Melophagus do not show flagellates until 
after their first feed of blood, the blood stimulating the pre- 
flagellates to form flagella. 


(9) Multiplication of C. melophagia by longitudinal 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGTA, 2138 


division occurs in both the pre-flagellate and the flagellate 
stages of the parasite. 

(10) Infection of Melophagus ovinus with C. melo- 
phagia is either hereditary or casual. In the case of casual 
infection the insects ingest the post-flagellates voided with 
the feces of other Melophagus. 

(11) A very dilute solution of grape-sugar causes the 
parasites to divide. There are only traces of sugar in both 
sheep-serum and wool extract. 

(12) Sheep’s blood or human blood added to the Crithidia 
also increased the rapidity of their growth and division. 

(13) A fungus present in the Malpighian tubules and gut 
of the ked (see Appendix II) has a rapid, fatal effect on the 
Crithidia. 

(14) An anti-coagulin is present in the salivary glands, 
stomach and intestine of Melophagus ovinus (see Appen- 
dice TUT). 

(15) A new spirochete, S. melophagi, was found in 
the gut, ovaries and puparia of a few of the Melophagus 
examined (see Appendix I). 


APprENDIX I. 


On the Occurrence of a Spirochete, 8S. melophagi, 
n.sp.,in Melophagus ovtnus. 


I wish to record the occurrence of a rare spirochete in the 
gut, ovaries and puparia of Melophagus ovinus. The 
spirochete was observed in life in the above-mentioned 
organs of a very few of the Melophagus examined, and at 
very different periods of the year (February, September, 
October). Very few spirochetes occurred, and consequently it 
is impossible to give full details regarding size and structure. 
The organisms seen were from 10 to 30 long and were 
narrow. ‘They vary in length, some being practically half 
the length of others, indicating the probability of transverse 
division. As some parasites were thicker than others there 


214 ANNIE PORTER. 


is the inference that longitudinal division takes place. 
This would be in accordance with the behaviour of other 
spirocheetes, for Fantham (1907-8-9) has shown that both 
forms of division occur in §. balbianii and S. anodonte. 
I (1909) also have observed the same, while the joint 
researches of Fantham and myself (1909) have demonstrated 
that both directions of division occur in 8. recurrentis and 
S. duttoni, and that there is a periodicity in the direction 
of division. 

The movements of 8. melophagi are fairly active, and 
are of the typical spirochete nature, namely, of forward 
progression accompanied by spiral or corkscrew rotation on 
its course. 

The occurrence of S. melophagi in the ovaries, ova and 
puparia of the ked is of much interest, for it indicates that 
the spirochete is transmitted hereditarily. Hence Melo- 
phagus ovinus can transmit both Crithidia melophagia 
and Spirocheta melophagi to its offspring. 


Aprenpix IT. 


Noteona Fungus found in the Malpighian Tubules 
and Intestine of Melophagus ovinus. 


A fungus was present in many specimens of Melophagus 
ovinus examined, especially those obtained from Scotland. 
Crithidia were not seen in the “keds” received from 
Scotland, and I have shown experimentally that the action 
of the fungus is fatal to the flagellate. 

The fungus occurred chiefly in the Malpighian tubules of 
the insect, and to a lesser extent in the intestine. The 
Malpighian tubules were frequently blocked by the fungus. 
A brief description of the fungus may now be given. 

The hyphe were long and filamentous with few septa. 
Many spores were produced. At the extremity of some 
hyphe globular heads were formed, possibly due to sexual 
processes. ‘The globular bodies contained many nuclei (text- 
fig. 11) fairly evenly distributed through the protoplasm. 


STRUCTURE AND LIFE-HISTORY OF OCRITHIDIA MELOPHAGIA. 215 


Nuclei and protoplasm then shrank away from the wall of 
the rounded body—provisionally called a sporangium (text- 
fig. 12)—so that a space intervened. Segregation of the 
protoplasm round the nuclei followed (text-fig. 15), and a 
morula-like body resulted. The morula differentiated into a 
mass of rounded spores (text-fig. 14), each of which formed a 
spore coat for itself. The sporangium ultimately ruptured 


TEXT-FIGURES 11-15. 


Fungus parasitic in Melophagus ovinus. 


Text-fig. 11.—Hypha with globular head. 

Text-fig. 12.—Differentiation of nuclei within the head (sporan- 
gium). 

Text-fig. 13.—Spores forming in sporangium. 

Text-fig. 14.—Mature sporangium. 

Text-fig. 15.—Dehiscing sporangium. 


(text-fig. 15), and the numerous small spores were set free. 
Some spores remained in the Malpighian tubes, others passed 
out into the intestine and were voided with the feces. 
Parasitic fungi have been previously recorded in insects, 
for example, in the house-fly, caterpillar, mosquito. ‘The 
fungus mentioned by Schaudinn in Culex was probably a 
member of the Entomophthoree, or related thereto. The 


216 ANNIE PORTER. 


fungus infesting Melophagus ovinus seems to be more 
nearly allied to the Peronosporee. 

I learn from a private communication that a similar fungus 
was found last year by Dr. H. B. Fantham, of Cambridge, in 
the alimentary tract and Malpighian tubes of the grouse-fly, 
Ornithomyia lagopodis. From examination of a pre- 
paration of the fungus of Ornithomyia, kindly lent to me, 
I believe that the fungi of the grouse-fly and the sheep-ked 
are very closely related. 


AppENDIX III. 


On the Occurrence of an Anti-coagulin in the Ali- 
mentary Canal of Melophagus ovinus, and its 
Significance in Relation to Crithidia melo- 
phagia. 

‘The pronounced and peculiar brightness of the blood in the 
crop and fore-part of the stomach of the keds examined was 
noticed very early in the investigation. ‘The blood of the 
sheep in the stomach of keds that had not fed for as long as 
three days was still practically fluid and had not coagulated 
much, while twelve to twenty-four hours after feeding the 
blood had not coagulated at all. This led me to suspect that 
an anti-coagulin, such as had been described ina tick (Argas 
persicus) by Nuttall and Strickland (1908), was present here 
also, and a series of tests were performed at different times 
which verified this inference. Every test that I performed 
had the same result—coagulation was delayed. 

The method of testing was simple. Separate emulsions of 
the salivary glands, stomach, and intestine of Melophagus 
ovinus were made with 0°75 per cent. NaCl solution. A 
known quantity—about 0°5 c.c. of human blood from a pricked 
finger—was then mixed with the same quantity of organ- 
emulsion, while for control purposes the same quantity of 
blood mixed with 0°75 per cent. NaCl solution was used. 
The test fluid and the control fluid were taken up in small 
glass capillaries, and the test was applied by blowing out the 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 217 


liquid at stated times and noting when coagulation occurred 
ineach. ‘Typical results of these experiments are tabulated 
below: 

A. Adult Melophagus. 


Coagulation period Coagulation period 


Experiment. of bloodandorgan- — of blood and °75 
emulsion. NaCl solution. 
(1) Salivary gland . 20min. : 7-8 min. 
er. y ot 220) Saal. 
(3) Intestine ; Rome Cre 5 Se a 
(4) * : SP TAs oOSecs. + Smee 


Obviously an anti-coagulin was present, for considerable 
delay of clotting occurred. 

B. Young Melophagus.—Here the interval between the 
clotting of the test and control preparations was noted. A 
few typical results are given: 

(1) Blood mixed with emulsion of the salivary glands 
clotted nine minutes after the control. 

(2) Emulsions of intestine added to blood caused the latter 
to take three times as long to clot as the control preparations 
took. 

Comparing the behaviour of the emulsions of the salivary 
glands of young and of older keds, the anti-coagulin seems to 
be more strongly developed in the salivary glands of the 
older keds, while a similar comparison between the intestinal 
emulsions would tend to show that the anti-coagulin was 
more abundant in the intestines of young keds. 

The temperature at which the anti-coagulin was destroyed 
was also investigated. It was found that below 50° C. the 
anti-coagulin would act. At about 55° C. its action was 
checked. When 60° C. was reached it was destroyed. 

Human blood mixed with emulsions of any part of the 
alimentary canal at once assumed the vivid red hue so notice- 
able in the blood removed from the gut of the keds. 

The red blood-corpuscles of the sheep, seen en masse, 
appear far brighter on adding emulsions of the gut of the 
ked containing the anti-coagulin. When much water was 
added to normal blood, hemolysis occurred, and the colour 


218 ANNIE PORTER. 


of the solution so obtained was made much brighter when an 
emulsion of crushed salivary glands of the ked was added to 
it. The leucocytes of the sheep’s blood occurring in the gut 
of the ked do not appear to be affected in any way by the 
anti-coagulin. 

Anti-coagulin appears to be found in all parts of the alimen- 
tary canal of the ked and to decrease in amount from before 
backwards. As before mentioned, I determined experi- 
mentally that freshly shed, and therefore fluid, blood acted 
as a stimulant to division of the Crithidia. This artificial 
condition is the counterpart of the natural condition of the 
blood within the fore-gut of the ked. here, owing to the 
action of the anti-coagulin, the freshly ingested sheep’s 
blood does not clot, but remains fluid. It is probable that 
Crithidia within the gut are stimulated by this fluid blood, 
and divide rapidly. I obtained similar results in the case of 
Herpetomonas jaculum, where “division of the flagellate 
Herpetomonad takes place rapidly under natural conditions 
after ingestion of blood by the host” (Porter [1909], p. 382). 
If the Critsidia are in the pre-flagellate condition the rapid 
multiplication is followed by the outgrowth of flagella, after 
which the organisms separate and pass further down the 
alimentary canal. The presence of anti-coagulin, from the 
salivary glands, in the contents of the fore-gut of the ked 
may be the cause of the rapidity with which the pre-flagellate 
stage of Crithidia melophagia is passed through, the 
blood, kept fluid by the anti-coagulin, acting as a stimulus 
to further development. 


. 


REFERENCES TO LITERATURE. 


Further references will be found at the ends of some of the papers 
quoted. 


Bruce, Sir David. Hamerton, A. E., Bateman, H. R., and Mackie, F. P. 
(x, 1909).—** The Development of Trypanosoma gambiense 
in Glossina palpalis,” ‘Proc. Roy. Soc., ser. B, lxxxi, 
pp. 405-414, pls. 10, 11. 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 219 


Fantham, H. B. (i, 1908)—**Spirocheta (Trypanosoma) balbianii 
(Certes) and Spirocheta anodonte (Keysselitz) : their Move- 
ments, Structure and Affinities,” ‘Quart. Journ. Micr. Sci.,’ 52, 
pp. 1-73, 5 pls. 

and Porter, Annie (1909).—** The Modes of Division of Spiro- 
cheta recurrentis and S. duttoni as observed in the Living 
Organisms,” ‘ Proc. Roy. Soe.,’ ser. B, Ixxxi, pp. 500-505. 
Flu, P. C. (1908).—* Ueber die Flagellaten im Darm von Melophagus 

ovinus,’ ‘ Archiv f. Protistenkunde,’ xii, pp. 147-153, 1 pl. 


Léger, L. (1902).—‘* Sur un flagellé parasite de /Anopheles maculi- 
pennis,” ‘C.R. Soc. Biol.,’ liv, pp. 354-6, 10 figs. 

Mackinnon, D. L. (1909).—‘* Note on two New Flagellate Parasites in 
Fleas—Herpetomonas ctenophthalmi, n. sp., and Cri- 
thidia hystrichopsylle, n. sp.,” ‘ Parasitology,’ ii, pp. 288- 
296, 1 pl. 

Minchin, E. A. (1908).—* Investigations on the Development of Try- 
panosomes in Tsetse Flies and other Diptera,” ‘Quart. Journ. 
Mier. Sci.,’ 52, pp. 159-260, 6 pls. 

Novy, F. G., MacNeal, W. J., and Torrey, H. N. (1907).—“ The Try- 
panosomes of Mosquitoes and other Insects,” ‘Journ. Infect. 
Diseases,’ iv, pp. 225-276, 7 pls. 

Nuttall, G. H. F., and Strickland, C. (1908).—‘* On the Presence of an 
Anti-coagulin in the Salivary Glands and Intestines of Argus 
persicus,” ‘ Parasitology,’ i, pp. 302-310. 

Patton, W. 8. (1908).—‘* The Life-Cycle of a Species of Cr thidia 
Parasitic in the Intestinal Tract of Gerris fossarum Fabr.,” 
‘Archiv f. Protistenkunde,’ xii, pp. 151-146, 1 pl. 

—— (1909).—* The Life-Cycle of a Species of Crithidia Parasitic 
in the Intestinal Tracts of Tabanus hilarius and Tabanus 
sp.,’ ‘Archiv f. Protistenkunde,’ xv, pp. 333-362, 1 pl. 

—— (1909).—* A Critical Review of our Present Knowledge of the 

Hemoflagellates and Allied Forms,” ‘ Parasitology,’ ii, pp. 91-143. 

and Strickland, C. (1908).—‘* A Critical Review of the Relation 

of Blood-sucking Invertebrates to the Life-Cycles of the Try- 

panosomes of Vertebrates, etc.,” ‘ Parasitology,’ i, pp. 322-346. 


Pfeiffer, E. (1905).—‘* Ueber Trypanosomenihnliche Flagellaten im 
Darm von Melophagus ovinus,” ‘Zeitschr. f. Hyg., 1, 
pp. 324-29, 1 pl. 

Porter, Annie (1909).—“*The Morphology and Life-History of Cri- 
thidia gerridis, as found in the British Water-Bug, Gerris 
paludum,” ‘ Parasitology,’ ii, pp. 348-366, 1 pl. 


220 ANNIE PORTER. 


Porter, Annie (1909).—“The Life-Cycle of Herpetomonas jaculum 
(Léger), Parasitic in the Alimentary Tract of Nepa cinerea,” 
‘Parasitology,’ ii, pp. 867-391, 1 pl. 

Pratt, H. 8. (1895).—* Beitriige zur Kenntnis der Pupiparen (Die Larve 

von Melophagus ovinus),” ‘Archiv f. Naturgesch.,’ liii, pp. 

151-200, 1 pl. 

(1899).—“*The Anatomy of the Female Genital Tract of the 
Pupipara as observed in Melophagus ovinus,” ‘ Zeitschr. f. 
wiss. Zool.,’ xvi, pp. 16-42, 2 pls. 

Robertson, Muriel (1909).—‘ Studies on Ceylon Hematozoa: I, The 
Life-Cycle of Trypanosoma vittate,” ‘Quart. Journ. Mier. 
Sci.,’ 53, pp. 665-695, 2 pls. 

Schaudinn, F. (1904).—‘* Generations- und Wirtswechsel bei Try pano- 
soma und Spirochete (Vorl. Mitt.),” ‘Arbeit. a. d. Kaiser. 
Gesundheitsamte,” xx, pp. 387-430, 20 figs. 

Swingle, L. D. (1909).—* A Study on the Life-History of a Flagellate 
(Crithidia melophagi, n. sp.) in the Alimentary Tract of the 
Sheep-Tick (Melophagus ovinus),” ‘Journ. Infect. Diseases,’ 
vi, pp. 98-121, 3 pls. 

Woodeock, H. M. (i, 1909).—‘* The Hiemoflagellates and Allied Forms,” 
article in ‘ Treatise on Zoology, edited by Sir Ray Lankester, 
pt. i, fase. i, sect. G., pp. 193-273. 


EXPLANATION OF PLATES 12 ann 13, 


Illustrating Miss Annie Porter’s paper on “ Crithidia 
melophagia.” 


[All figures were outlined with an Abbé-Zeiss camera-lucida, using a 


2 mm. apochromatic (Zeiss), or ~ inch achromatic (Zeiss) objective, 


and compensating oculars 8 and 12 of Zeiss. The magnification is in 
all cases approximately 1500 diameters, except where otherwise stated. ] 


PLATE 12. 


Figs. 1-20.—Pre-flagellate Stages. 
Fig. 1.—Pre-flagellate with round nucleus, bar-like blepharoplast. 
No flagellum. Crop. Giemsa. 
Fig. 2—Oval pre-flagellate. Blepharoplast slightly constricted. 
Crop. Delafield’s hematoxylin. 
Fig. 3.—Dividing pre-flagellate. Crop. Delafield’s hematoxylin. 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 221 


Fig. 4.—Division rosette of pre-flagellates. Two individuals again 
dividing. Crop. Delafield’s hematoxylin. 

Figs. 5-8.—Elongating pre-flagellates. Crop. Thionin. 

Fig. 9.—Large preflagellate, with round nucleus, rod-like blepharo- 
plast, flagellum just differentiating. Crop. Giemsa. 

Fig. 10.—Rounded form. Flagellum longer than in fig..9. Crop. 
Giemsa. 

Fig. 11.—Smaller parasite with large nucleus and long flagellum. 
Crop. Delafield’s hematoxylin. 

Fig. 12.—Parasite, showing elongation of flagellar (anterior) end of 
the body. Crop. Giemsa. 

Figs. 15 and 14.—Crithidia with elongated posterior ends. Anterior 
part of stomach. Giemsa. 

Fig. 15.—Pre-flagellate with posterior blepharoplast. Crop. Giemsa. 

Fig. 16.—Parasite with anterior end more developed. Crop. Giemsa. 

Figs. 17 and 18.—Almost mature flagellates, membranes showing 
myonemes. Crop. Giesma. 

Figs. 19 and 20.—Practically adult flagellates. Fore-part of stomach. 
‘Thionin. 

Figs. 21-43.—Flagellate Stage. 

Fig. 21.—Small flagellate. Nucleus with chromatin in granules 
extending part way across the body. Rod-like blepharoplast. Intestine. 
Giemsa. 

Fig. 22.—Flagellate, with well-marked myonemes on the body. 
Stomach. Gentian violet. x 2250 approximately. 

Fig. 23.—Parasite, with flagellum almost continuous with the 
blepharoplast. Nucleus with centralchromatin. Stomach. Delafield’s 
hematoxylin. 

Fig. 24.—Crithidia showing blepharoplast posterior to the nucleus 
—an uncommon condition. Stomach. Giemsa. 

Figs. 25, 26.—Flagellates showing chromidia in their posterior ends. 
Chromatin of nucleus in bars. Stomach. Giemsa. 

Figs. 27-29.—Parasites with somewhat pointed posterior ends. 
Chromidia present in fig. 29. Intestine. Thionin. 

Figs. 30, 31.—Crithidia showing somewhat alveolar protoplasm. 
Stomach. Thionin. Xx 2250 approximately. 

Fig. 32.—Flagellate with blunt posterior end, round nucleus contain- 
ing large chromatin granules, and extending across complete breadth 
of body; blepharoplast curved. Stomach. Thionin. 2250 approxi- 
mately. 

VOL. 99, PART 2.—NEW SERIES. 15 


222 ANNIE PORTER. 


Fig. 33.—Parasite with scattered chromidia. Blepharoplast slightly 
posterior to and to the side of the nucleus. End of crop. Giemsa. 

Fig. 34.—Crithidia with large oval blepharoplast. Stomach. 
Giemsa. 

Fig. 35.—Narrow parasite. Intestine. Giemsa. 

Figs. 36, 37.—Longer parasites with many chromidia. Stomach. 
Tron-hvematoxylin. 

Fig. 38.—Flagellate showing alveolar protoplasm, nucleus and 
blepharoplast almost in contact. Intestine. Thionin. x 2250 approxi- 
mately. 


Fig. 39.—Long form. Nucleus with chromatin arranged in bars. 
Oval blepharoplast. Membrane distinct. Intestine. Giemsa. 

Fig. 40.—Long parasite with thick flagellum. Myonemes present 
on body. Blepharoplast showing constriction, so about to divide. 
Chromatin of nucleus in large masses. Stomach. Delafield’s hwema- 
toxylin. 

Fig. 41.—Small aggregation-rosette, showing entanglement of large 
and small flagellates. Stomach. Giemsa. 

Fig. 42.— Flagellate with rounded nucleus and posterior blepharoplast. 
Basal granule near root of flagellum. Myonemes in membrane. 
Intestine. Iron-hxematoxylin. 

Fig. 43.—Large rosette. Many parasites shown aggregated around 
a piece of débris. The flagella serve as points of attachment, therein 
differing from a division-rosette. Common in stomach and intestine. 
Delafield’s hematoxylin. 


Figs. 44-56—Stages in Division. 

Fig. 44.—Parasite showing constricted blepharoplast with clear area 
around it. Chromatin in nucleus arranged in masses at periphery. 
Intestine. Thionin. xX 2250 approximately. 

Fig. 45.—Stage similar to fig. 44. Well-marked myonemes on body 
and membrane. Giemsa. X 2250 approximately. 

Fig. 46.—Parasite with both nucleus and blepharoplast constricted. 
Flagellum beginning to split at base. Stomach. Delafield’s hematoxylin. 

Fig. 47.—Flagellate with anterior end of body, nucleus and blepharo- 
plast all divided. Stomach. Delafield’s hematoxylin. 

Figs. 48, 49.—Somewhat rounded parasites ; bodies of daughter-forms 
not yet diverging from one another. Stomach. Thionin. 


Fig. 50.—Daughter-organisms forming a V. Stomach. Giemsa. 


Figs. 51, 52.—Further stages in the divergence of the bodies of the 


STRUCTURE AND LIFE-HISTORY OF CRITHIDIA MELOPHAGIA. 223 


daughter-forms. The flagella have interlocked. Intestine. Delafield’s 
hematoxylin. The parasites represented in fig. 51 divided sub-equally. 

Fig. 53.—Sub-equal division. Daughter-organisms are almost 
separated. Intestine. Delafield’s hematoxylin. 

Figs. 54, 55.—Parasites about to separate. Stomach. Giemsa. 

Fig. 56.—True division-rosette. The separation of the daughter- 
individuals takes place from the flagellar end backwards, so that in 


a rosette the posterior ends of the organisms are centrally directed. 
Stomach. Thionin. 


PLATE 13. 
Figs. 57-94.—Stages of the Parasite in the Ovary, Eggs, 
and Puparia. 
(The eggs in figs. 58, 64, 65 are represented diagrammatically.) 

Fig. 57.—The flagellate as it penetrates the ovary. Delafield’s 
hematoxylin. 

Fig. 58.—Flagellate in the act of penetrating a young egg, the blunt 
end of the parasite being used. Thionin. The egg of Melophagus 
ovinus is represented diagrammatically. 

Figs. 59, 60.—Flagellates from ovary. Flagella somewhat reduced. 
Giemsa. 

Figs. 61-63.—F lagellates from within the egg. Giemsa. 

Figs. 64, 65—Rounding-up forms of C. meloplagia within eggs. 
Delafield’s hematoxylin and fresh preparations. Eggs of Melophagus 
represented diagrammatically. 

Figs. 66-72.—Series of parasites showing successive stages in shorten- 
ing and rounding-up of flagellates when within the eggs. Delafield’s 
hematoxylin. 

Figs. 73, 74.—Parasites showing nuclear division. Very young 
puparium. Giemsa. 

Figs. 75-77.—* Plasmodial ” stages of C. melophagia in developing 
puparia. Peripheral blepharoplasts seen. Giemsa and fresh prepara- 
tions. 

Figs. 78-81.—Rounded parasites resulting from plasmodial forms. 
Delafield’s hematoxylin. 

Figs. 82-84.—Parasites produced by growth of forms similar to those 
shown in fig. 81. Giemsa. 

Figs. 85-87.—Rosettes of somewhat oval parasites from young 
puparium. Delafield’s hematoxylin. 

Fig. 88. — Well-defined division-rosette from mature puparium. 
Giemsa. 


224: ANNIE PORTER. 


Figs. 89-91.—Dividing forms. Mature puparium. Giemsa. 

Figs. 92-94.—Parasites resembling pre-flagellates produced from cyst. 
Mature puparium. Delafield’s hematoxylin. 

Figs. 95, 96.—Small aggregation-rosettes. Intestine. Thionin. 


Figs. 97-114.—Post-flagellate Stages in Rectum. 

Fig. 97.—Parasite dividing prior toencystment. Intestine. Thionin, 

Fig. 98.—Uncommon form of division, occasionally seen in living 
specimens. Rectum. Giemsa. 

Fig. 99.—Small form. Flagellum in process of absorption. Rectum. 
Giemsa. 

Fig. 100.—Parasite showing concentration of protoplasm in the region 
of the nucleus. Rectum. Giemsa. 


Fig. 101.—Form common in rectum. Body much flattened. Flagellum 
disappearing. Delafield’s hematoxylin. 

Figs. 102-108.— Parasites showing progressive disappearance of 
flagellum. Rectum. Thionin. 

Figs. 109-112.—Post-flagellate cysts from rectum. Giemsa. 

Fig, 113.—Post-flagellate cyst from feces of Melophagus ovinus. 
Giemsa. 

Fig. 114.—Thick-walled cyst. Rectum. Giemsa. 


eee) 


23. 


A.Porter, del. 


CRITHIDIA 


Muth, Lith? London, 


Iuant.bunn Mécnr8i.Ubl 65 NSA, 12 


ELOPHAGIA. 


Cc 


* 


ee 


A Portes, dal. 


Quant. Sourn Mier Sci. Ut.58. NS A12 


9 6 ,., © ® 
& 1s, & uw. ® ae 
107. 109. Tes 


Huth Sith’ Londan. 


GCRITHIDIA MELOPHAGIA. 


STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX. 225 


Studies in the Experimental Analysis of Sex. 


By 


Geofirey Smith, 
Fellow of New College, Oxford. 


With Plate 14. 


3. FurrHer OBsERVATIONS ON Parasitic CastTRATION. 

Durine my occupation of the British Association Table at 
Naples this winter I took the opportunity of re-examining 
certain points connected with the effect of Sacculina 
neglecta on the spider-crab Inachus mauretanicus, 
with the especial purpose of trying to settle the exact way 
in which the gonad of infected individuals degenerates and 
is absorbed. The mid-winter months being the most favour- 
able season at Naples for finding numerous individuals of 
Inachus very profoundly modified by the presence of the 
parasite, | was able to re-investigate many crucial stages in 
the modification of the external and internal sexual organs, 
with the result that, while certain new facts of interest have 
come to light, I see no reason whatever for departing in any 
respect from the statement of facts made in my earlier work, 
or from the deductions drawn from them (‘Naples Mono- 
graph,’ No. 29, Chap. V). In this paper, besides giving the 
results arrived at in respect to the degeneration of the gonad, 
I propose to describe certain new instances of infected 
Inachus which afford incontestable proof that male crabs 
with differentiated though reduced male internal organs can 
assume all the adult female secondary sexual characters. 
It will also be shown both for the male and female sex that 
the effect of parasitic castration can on no account be 


226 GEOFFREY SMITH. 


interpreted as a return to a juvenile undifferen- 
tiated condition. 

In Part 2 of these studies it has already been pointed out 
that this interpretation is ruled out by the facts, and this was 
also pointed out in my earlier work, but not in so detailed 
and categorical a form, with the unfortunate result that 
Professor T. H. Morgan, in a recent paper on ‘‘ Sex Determina- 
tion” (‘Journal Exper. Zoology,’ vol. vii, 1909, pp. 343, 344), 
has adopted this very explanation of my observations. Thus 
he writes: “The broad abdomen of the castrated male might 
be considered to correspond to the juvenile state. ‘The only 
external structure cited by Smith that might seem to indicate 
that the characters of the castrated males are female rather 
that juvenile ones is the presence of hairs on the abdominal 
appendages of Inachus, absent in the young crab, but present 
in the adult female. Such evidence would not in itself be 
conclusive, since the presence of hairs may be due to increase 
in size or to a later moult rather than to latent female 
characters. Smith concludes that the male sex, and pro- 
bably the male sex alone, can be so radically modified in its 
sexual nature as to assume a perfect external hermaphroditism. 
If, on the contrary, we assume that we have here, not herma- 
phroditism, but an imperfect development of male characters 
combined with the juvenile condition, we might offer a 
plausible explanation of the facts.” 

I am sorry that any want of explicitness on my part should 
have misled Professor Morgan, but I cannot accept the state- 
ment that the only characteristically adult female character, 
cited by me as being assumed by the infected males, is the 
presence of hairs on the abdominal appendages. I pointed 
out in my earlier work (‘Naples Monograph,’ xxix, pp. 67, 
70 and 71) that in the young stages of the female, before the 
adult breeding form is assumed, the abdoinen is a com- 
paratively small flat plate, whereas in the adult it becomes 
suddenly widened and also takes on a hollowed trough-like 
shape, so that the two forms of abdomen are absolutely 
distinct morphological structures, distinguishable from one 


STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX. 227 


another at a glance (see figs. 1, 2, 3 and 4, PI. VII, ‘ Naples 
Monograph’ and again ‘ Quart. Journ. Micr. Sci.,’ vol. 54, Pl. 
30, figs. 10, 11, 18,14). Now, when the infected males take on 
the female external characters they have never been found to 
assume the juvenile flattened form of abdomen which charac- 
terixes the young stages of both males and females, but they 
invariably take on the hollow trough-like form characteristic 
of the adult breeding female and of her alone (see the 
numerous figures on the plates referred to above). That is 
the first objection to the view that the alteration of the male 
is merely towards a juvenile condition, and anyone who will 
examine the series of specimens exhibited in the South 
Kensington Museum or in the Oxford Museum, or those 
deposited by me at the Zoological Station at Naples, will at 
once perceive the entire morphological difference of the 
abdomen in the young and adult female, and the identity of 
the modified male abdomen with that of the adult female. 

Secondly, with regard to the abdominal appendages. It is 
not a question of the mere presence or absence of a few hairs, 
as Professor Morgan has unfortunately been led to suppose. 
The abdominal appendages of the juvenile and adult indi- 
viduals differ as radically, if not more radically from one 
another, than the form of the abdomen. In the young form 
of the female these appendages are short, stout and rod-lke, 
and provided with a very few short bristles, as shown in PI. 
14, fig. 7 of this paper. In very young males similar 
appendages are present, but they are lost at a very early 
stage indeed, only the two anterior appendages being kept as 
the copulatory styles. ‘he form of these two appendages in 
the young male is shown in PI. 14, figs. 1 and 2. 

The adult female, at the same moult at which it acquires 
the characteristic adult form of abdomen, assumes a totally 
different kind of appendage of the form shown in PI. 14, fig. 
4. Here it is seen that instead of being stout and rod-like 
with a few stiff hairs, as in the young females, the appendage 
has become transformed iuto two wisp-like branches, the 
exopodite being densely clothed with long plumose hairs, the 


228 GHOFFREY SMITH. 


endopodite, now a slender-jointed structure, being furnished 
with exceedingly long pointed hairs for the attachment of 
the eggs. The structure of these abdominal appendages in 
the adult female, adapted as they evidently are for repro- 
ductive purposes, is as morphologically distinct from that of 
the young individuals of either sex as anything very well 
could be. 

Now let us inquire in what form the infected modified 
males assume the abdominal appendages. ‘The answer is 
plainly given by reference to Pl. 14, fig. 4. This figure is an 
actual camera drawing of the second abdominal appendage of 
an infected individual, which was proved to be a male by the 
presence of a copulatory style of a somewhat modified form 
(Pl. 14, fig. 38), and internally by the presence of testes and 
vesicule seminales of a typical character on either side. The 
testis and vesicula seminalis of one side of this individual are 
shown in Pl. 14, fig. 10. The form of the abdominal ap- 
pendages (PI. 14, fig. 4), of which there were four on each 
side in addition to the copulatory styles, is identical with that 
of a normal adult female; in fact, since this figure serves 
equally well to depict the abdominal appendage of a normal 
adult female, [ have not considered it necessary to give 
another figure, which would simply mean repeating the same 
structures. 

‘he infected male individual to which figs. 3, 4, and 10 on 
Pl. 14 refer is a particularly favourable type for showing con- 
clusively thatthe abdominal appendages, when assumed by 
the infected males, are of the characteristically adult female 
type. Asamatter of fact acommoner condition is that shown 
in Pl. 14, fig. 5. : In this infected male the copulatory 
style was greatly reduced (fig. 6) and the abdominal appen- 
dages were also developed in an imperfect condition, with 
almost complete suppression of the endopodites. Neverthe- 
less, the characteristic plumose hairs are present on the exo- 
podite, which is of.a slender shape, thus conforming to the 
adult type of female appendage and not really approaching to 
the juvenile condition. This figure might equally well refer 


STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX. 229 


to the abdominal appendage of an infected female, in which 
the endopodites are very frequently thus reduced. 

We have now examined in some detail two of the mest im- 
portant characters in which the infected male Inachus is 
modified by the presence of the parasite Sacculina, viz. the 
shape of the abdomen aud the form of the abdominal appen- 
dages, and we have seen that Professor Morgan’s attempt tu 
explain the modification of the male as a return to a 
juvenile condition is quite at variance with the facts. But 
we have still two more points to consider, which render that 
explanation still more impossible. 

The most important of these two points is the fact that in 
a certain small percentage of cases the infected males, on 
recovery from the parasitic disease, have been observed to 
have regenerated the gonad, and to have developed large ova 
measuring about 1 mm. in diameter and full of the reddish- 
coloured yolk characteristic of the mature ova of the female 
Inachus. Professor Morgan himself admits the cogency of 
this fact, so that I need not labour it here, its significance, 
indeed, being obvious. 

The second point is one which I have only been able to 
settle finally during my recent visit to Naples. In my earlier 
work (loc. cit., p.68) I inclined to the view that the presence 
of Sacculina caused the young females under 15 mm. in 
carapace length to assume prematurely the adult type of 
abdomen and abdominal appendage, and I emphasised this 
point as being of importance in precluding the view that the 
effect of the parasite was merely to arrest development or 
“ause a return to a juvenile state. By a careful examination 
of the large amount of material put at my disposal by Dr. 
Lo Bianco this winter, I have found that this premature 
assumption of adult characters by infected females undoubtedly 
occurs. During December and January all the uninfected 
females of carapace length up to 14 mm. had the immature 
juvenile form ot abdomen and appendage, but all the intected 
females measuring froin 6-14 mm. had the fully adult type of 
both those structures. The real theoretical significance of 


230 GEOFFREY SMITH. 


this fact, which has an important bearing on the whole 
meaning of parasitic castration, will be discussed later, but it 
has been introduced here as a final nail in the coffin of the 
theory which attempts to explain the effects of parasitic 
castration as due to arrested development or the assumption 
of juvenile characters. Possibly the use of the term “ parasitic 
castration” has had something to do with perpetuating this 
unfortunate error, the analogy between ordinary operative 
castration or mechanical removal of the gonads and their 
degeneration owing to the presence of a parasite being, as 
Professor Sedgwick has pointed out, extremely small. In 
parasitic ‘ castration” the degeneration of the gonad is not 
brought about by the parasite mechanically removing or 
attacking the gonad, but by its setting up a deep-seated 
Alteration of the metabolism of the host which secondarily 
reacts on the gonad. We may now enter into the question 
of the method of degeneration of the gonad. In the above 
paragraphs I trust that the following conclusion has been 
thoroughly vindicated. The modification of the male Inachus 
by the parasite Sacculina consists in the assumption by the 
male of adult female sexual characters to a greater or less 
degree of perfection; in neither sex can the modifica- 
tion be ascribed to arrest of development or the 
assumption of a juvenile immature condition. 

As I have shown in my earlier work (loc. cit., pp. 72-74) 
the degenerate condition of the ovaries and testes with their 
ducts in infected Inachus is due to two causes: firstly, an 
arrest of growth, so that the gonad tends to remain in the 
same condition as it was when infection took hold, and 
secondly, to an actual absorption of the tissues of the gonad 
and their final disappearance, a process which was often 
accompanied by an actual irruption of the roots of the 
purasite into the germinal tissues. ‘I'he arrest of growth of 
the gonad and the first stages of degeneration, at any rate in 
the male, were shown to be independent of the irruption of 
the Sacculina roots. 

The method of absorption and disappearance of the gonad 


STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX. 231 


was not clearly made out, and it was my chief object this 
year to obtain some idea of how this process takes place, to 
observe, for example, whether phagocytosis takes any active 
part in it. 

The condition of arrested growth without any signs of 
active degeneration is well exemplified by the testis and 
vesicula seminalis figured on PI. 14, fig. 10, which was dis- 
sected out of the perfectly modified male whose appendages 
are represented in figs. 3 aud 4. In Pl. 14, figs. 8 and 9, are 
drawn, on the same scale, the vesicula seminalis and a small 
portion of the coiled testis of an uninfected male of the same 
size, Showing that the gonad of the infected individual has 
remained very small and undeveloped. Spermatozoa were 
entirely absent from the infected individual, whereas the 
vesicula of the normal individual was crowded with them. 
There is, however, another point to be observed. Investing 
the gonad of the normal individual is a thin sheath of con- 
nective tissue with flattened, darkly staming nuclei (c.s., P]. 14, 
fig. 9). In the infected individual tlis sheath is seen to be of 
proportionately greater thickuess. 

In Pl. 14, fig. 11, is shown a portion of the testis of an 
infected male in which the process of absorption of the gonad 
has proceeded to a considerable extent. In three places small 
disconnected masses of testicular cells (¢.) are seen lying 
ensheathed by connective tissue; between the disconnected 
pieces of germinal tissue nothing remains but the connective- 
tissue sheath. By staining such preparations with a triacid 
stain, e.g. Ehrlich-Biondi, small globules are seen lying 
between the germinal nuclei and the sheath, which take up 
the orange stain. These globules may be looked upon as 
degeneration products of the germinal tissue in process of 
absorption. In none of the preparations which I have made 
of degenerating gonads is there any sign of phagocytosis, the 
degeneration appearing to take place by some process of 
auto-digestion. 

Turning to the degeneration of the ovary, PI. 14, figs. 12 
and 13, we find exactly the same process. Fig. 12 represents 


ot 


232 GKOFEREY SMITH. 


a portion of degenerate ovary of an infected female, in which 
islets of ovarian tissue containing disintegrating ova are seen 
encapsuled in the connective-tissue sheath. Fig. 13 is a high 
power drawing of a small portion of the ovary showing the 
clear distinction between the germinal nuclei (VV), the nuclei 
of the connective-tissue sheath (cs) and the degenerating ova. 

In a very great number of infected crabs dissected no trace 
could be found of the remains of a gonad; and in these, allowing 
for .a certain number in which I overlooked the degenerating 
remains, one must suppose that the process of encapsula- 
tion by connective tissue and auto-digestion had led to com- 
plete disappearance. Iam unable to state for certain whether 
the connective-tissue sheath plays an active part in the 
absorption of the germinal tissue; the chief part is clearly 
due to a simple disintegration of the same nature as is now 
known to occur in the destruction and absorption of the 
larval organs of insects during metamorphosis. In this latter 
process it was formerly held that phagocytosis played thie 
principal part, but it is now kuown that a process of auto- 
digestion by fluids is at least as active an agent. 

T'o conelude this part, I will attempt to outline, in a more 
satisfactory manner than was possible before, an explanation 
of why it is that the presence of a parasite should bring 
about such profound physiological and morphological changes 
in its host. 

We must clearly define, in the first place, what these 
changes essentially consist in. It has been shown in my 
earlier papers, and I trust still more fully brought out in this 
paper, that the effect of Sacculinaon Inachus is to cause thie 
infected individuals of botl. sexes to assume adult female 
characteristics. ‘his results not only in transforming the 
males into hermaphrodites with preponderating female char- 
acters, but also in hastening on the assumption of adult 
female characters by immature females. The problem, there- 
fore, resolves itself into this, Why should the presence of 
Sacculina cause the host of either sex to become adult female 


in nature ? 


STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX. 233 


Let us examine what the process of becoming adult involves 
in an ordinary female crab. Plainly the most important 
change is the rapid elaboration of yolk material which accu- 
mulates in the ovary; causing the latter to grow to a very great 
size. This elaboration of food material: in the ovary is the 
fundamental point in which the adolescence of the female 
gonad differs from that of the male. In the male gonad at 
maturity we have an immense multiplication of nuclei and of 
chromatin but a small development of cytoplasmic material 
and no deposit of yolk; in the female we have the exact 
opposite of this process. The most important part, then, in 
the process of becoming adult female, is the active elaboration 
of yolk material. 

We have arrived, therefore, at this point of the argument: 
that the presence of Sacculina causes the crab of either sex 
to become adult female in nature, and the most important 
activity of this state is the elaboration of yolk material. Can 
we prove that the presence of Sacculina actually causes its 
host of either sex to produce yolk material? [believe we can. 
If the roots of Sacculina which fill the body of an infected 
Inachus be examined, they will be found to be packed with 
small globules of an oily material, and if the roots are stained 
with such a mixture as Ehrlich-Biondi’s tri-acid stain it may 
be observed that the Sacculina roots take up the same consti- 
tuent in the stain, namely the acid fuchsin, as the yolk of 
an adult female crab’s ovaries. From the observed contents 
of the Sacculina roots and from their reaction to stains it is 
clear that they are elaborating from the blood of the Inachus 
of both sexes a closely similar yelk material to that which is 
normally accumulated in the ovary of a healthy adult female 
Inachus. 

The effect of Saceculina on Inachus is therefore to force 
the latter to elaborate yolk material of a similar kind to that 
which is normally developed in the ovary of the female at 
maturity. As the Inachus elaborates it the Sacculina 
abstracts it, so that it does not come to be deposited in the 
gonad until after recovery from the disease, when, as we have 


234 GEOFFREY SMITH. 


seen, the yolk-containing ova may be formed in the gonad 
of either sex. 

Meantime the continued production and circulation in the 
blood of the infected Inachus, whether male or female, of 
this yolk material, or rather of the substances from which the 
yolk is built up, is accompanied by the production of the 
secondary sexual characters proper to the adult female. 
These yolk-forming substances, or substance, are therefore 
identical with the “sexual formative substance,’ whose 
existence we deduced in Part 2 of these studies. We may 
summarise the above argument as follows: The Sacculina 
roots require for their nourishment a substance in the blood 
of the crab which they can work up into yolk material. 
This substance is provided for them in the female sexual 
formative substance, which is circulating in small quantities 
in normal male crabs as well as, in greater quantities, in 
female crabs. But the Sacculina roots must have the power, 
not only of abstracting this material from the crab’s blood, 
but also of forcing the crab to go on forming this substance 
in excess. This may seem to be a great assumption; but it 
is exactly here that a very close parallel can be drawn 
between the phenomenon we are dealing with and the 
general processes of immunity to parasites and organic 
poisons. Immunity has been interpreted, especially by 
Ehrlich, to mean that when a poison acts upon an organism it 
combines with and anchors certain organic molecules, which 
are then regenerated in excess and poured out into the blood- 
stream as antibody. If we suppose, therefore, that the 
Sacculina roots anchor the molecules of the female sexual 
formative substance, and this, from the fact of their forming 
yolk material, they appear to do, it is in accordance with the 
facts of immunity to suppose that the molecules of the sexual 
formative substance, wherever they are formed, will be 
regenerated in excess. 

The continued operation of this process, namely, the pro- 
duction of female sexual formative substance in the blood- 
stream, and its abstraction by the Sacculina roots, would 


STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX. 2380 


account for all the observed phenomena, viz. the development 
of adult female characters, which are dependent on the presence 
of this substance in quantity in the blood, and the abortion 
of the gonad owing to the Sacculina roots seizing on its 
proper nutriment and not permitting it to grow or develop. 
Nevertheless in the case of the hermit-crab infected by Pelto- 
gaster, Potts has shown that small eggs may be formed in 
the gonad, even while the parasite is still vigorous, showing 
that the excess of sexual formative substance has to some 
extent been seized on by the gonad. 

In the above manner it appears to me that we not only 


“‘ parasitic castra- 


gain a clear idea of the process involved in 
tion,”’ but the phenomenon, instead of appearing an isolated 
curiosity of a wholly inexplicable nature, falls into line with 
the well-known reactions to parasitic infections which are 
classed under the category of immunity. The clue to the 
whole theory rests in the truth of the statement that ‘ para- 
sitic castration ” consists in the assumption by the infected 
individuals of adult female characteristics, owing to the 
development within them of the female sexual formative sub- 
stance. If this statement of the case is rejected by the reader 
on the evidence which I have adduced, he will naturally reject 
the theory proposed to account forit, and if he can succeed in 
framing a different and more satisfactory theory which will 
include all the facts I shall be very well pleased. 

But any attempt to explain “parasitic castration” by 
vague analogies with the effects of operative castration, or by 
referring the whole phenomenon to arrested development or 
appearance of juvenile characters, is certainly foreordained to 
failure. 

‘The explanation here offered of parasitic castration differs 
from that which I proposed in my first work (‘ Naples Mono- 
graph,’ xxix, p. 82, et seq.) only in its greater precision, not 
in its general outline. In my original statement of the theory 
T ascribed the alteration of the male to an adaptive response 
of the metabolism in order to make good the drain on the 
system caused by the presence of a parasite. The metabolism 


250° . GEOFFREY SMITH. 


was represented as changing from the katabolic male con- 
dition to the more anabolic female, and this change was 
supposed to be effected by the development in the body of 
the female sexual formative substance. It is clear that this 
theory is fundamentally the same as that now proposed, but 
being couched in rather vague and general language, it seems 
to have made very little impression even on those who un- 
reservedly accepted my statement of facts. By showing, 
firstly, that the assumption of the adult female condition 
involves an active elaboration of yolk material, and secondly, 
that the Sacculina roots actually withdraw some substance 
from the ecrab’s blood from which they manufacture a yolk 
substance closely similar to that normally deposited in the 
eggs of the crab, and also by emphasising the fact that in reality 
both sexes of the host react in exactly the same way to the 
parasite, it has been possible to express the theory in a far 
more objective manner. 


a) 


Summary oF Parr 3. 

(1) 'The effect of Sacculina on male Inachus consists in 
the assumption by the male of adult female characteristics, and 
can in nowise be ascribed to arrest of development or acqui- 
sition of juvenile or immature characters, as suggested by 
Professor ‘I’. H. Morgan. 

(2) The effect of Sacculina on young immature females of 
Inachus is to force them to assume prematurely adult female 
characteristics. 

(3) The absorption of the gonad of infected Inachus of 
both sexes is brought about by a process of ensheathment 
with connective tissue and auto-digestion, phagocytosis ap- 
parently playing no part. 

(4) The reason why Sacculina causes theassumption of the 
adult female state in Inachus is found in the facts (1) that 
the roots of Sacculina elaborate a yolk-substance from the 
blood of Inachus of a similar nature to that which is elabo- 
rated in the ovaries of an adult female Inachus; (2) that in 
order to elaborate this yolk substance they take up from the 


STUDIES IN THE. EXPERIMENTAL ANALYSIS OF SEX. 237 


blood of Inachus the female sexual formative substance, 
which is the necessary material for forming the yolk; (8) that 
the female sexual formative substance, being anchored by the 
Sacculina roots, is regenerated in excess; (4) that the presence 
of the female sexual formative substance contin ually circulat- 
ing in large quantities in the body-fluids of the infected crabs 
causes the production of adult female secondary sexual 
characters, and, when the parasite dies, of yolk-containing eges. 


4. On a Case oF Parasitic CAsTravION IN A VERTEBRATE. 


Although numerous cases are now known of .the presence 
of a parasite causing arrest of development or degeneration 
of the reproductive organs in various invertebrates, no clear 
instance of this process has been reported, so far as I am 
aware, among vertebrate animals as the result of bacterial 
disease of organs other than the reproductive organs them- 
selves. Of course, where the reproductive organs themselves 
are the seat of infection, a certain amount of atrophy or 
degeneration may naturally result, but we have here to deal 
with a case of parasitic castration, analogous to the case of 
Sacculina on Inachus, or of Entoniscus on various 
-crabs, where the reproductive organs are not themselves 
necessarily attacked by the parasite, but are secondarily 
affected by the general disturbance of the metabolism, set up 
by the presence of a parasite in other parts of the body. 

During December, 1909, I received a pure-bred Gallus 
bankiva cockerel for breeding purposes. It belonged to 
the breed known as the Indian Jungle Fowl, a breed which 
has departed very little from the wild Gallus bankiva. 
The bird when it arrived appeared in good health; the 
‘plumage was in good condition, the comb and wattles well 
developed and red, the spurs fully developed, the tail carried 
-erect, and the bird crowed in the normal manner. Its age 
was one year and a half. About two weeks after it arrived 
it showed signs of sickness and a tendency to mope in the 
straw at the back of its run. These symptoms became 
gradually worse, and at the beginning of February the whole 

VOL. 5D, PART 2.—NEW SERIES. 16 


238 GEOFFREY SMITH. 


appearance of the bird was changed: the comb and wattles 
were greatly shrunken, and instead of being bright red were 
unhealthy pink patched with grey; the skin round the eyes 
was bloodless; the tail was carried drooping, and the bird 
never crowed. The bird was isolated and treated with 
purgatives, but the illness continued, the comb and wattles 
having withered by the middle of April to about half their 
original size. The spurs and plumage were unchanged, save 
for the fact that the tail was always drooped. The bird was 
killed and dissected on April 8th. 

The post-mortem examination showed that it was suffering 
from very acute avian tuberculosis. The liver was inter- 
penetrated with whitish calcareous nodules swarming with 
the characteristic tubercle bacillus, while the whole course of 
the alimentary canal, pancreas and spleen was covered with 
similar swellings, some of them of the size of a pea, also full 
of living bacteria. Only the alimentary and lymphatic organs 
were infected, the lungs, kidneys, and testes being entirely 
free of infection. 

Although the testes were uninfected, it was at once 
apparent that they were very remarkably reduced in size, 
measuring only 10 mm. in length by 5 mm. in breadth, 
whereas in a normal cockerel of the same breed and age, at 
the same time of year, they measured 40 mm. in length by 
25 mm. in breadth. The vasa deferentia were also reduced 
in size, and this was especially noticeable in the coiled lower 
part of the tubes where they pass into the vesicule seminales: 
no spermatozoa were present. 

Sections of the testes showed the testicular tubes intact, 
with a regular lining of germinal epithelium cells with nuclei 
in a resting condition. There was no sign of any mitosis or 
of any other stages in the process of spermogenesis. The 
testicular tubes, in fact, presented the appearance charac- 
teristic of immature birds of a few weeks old. 

In a certain number of the tubes degenerating germinal 
cells with abnormal nuclei could be seen. 

In contrast to this extreme reduction and arrest of develop- 


STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX. 239 


ment in the germinal part of the glands, the interstitial cells, 
forming islets everywhere between the testicular tubes, were 
well marked. 

There was no trace of infection by the tubercle bacilli in 
either testis. 

It is clear from the course of the disease and from the post- 
mortem examination that the reduction of the comb and 
wattles and the atrophy of the testes went hand in hand with 
the acute development of the tuberculosis. We know from 
numerous experiments that the effect of the removal of the 
testes in Gallus is to arrest the development of the comb and 
wattles ; otherwise, except for the loss of the crowing and 
the drooping of the tail, the other secondary sexual characters 
are not affected. We have seen that as the bird in question 
became ill, the principal symptom was the reduction in the 
comb and wattles, and the post-mortem showed that the 
testis must have been accompanying these organs in a process 
of atrophy. 

We have, therefore, in this case, an instance of parasitic 
castration caused by a bacterial infection of a vertebrate host, 
exactly parallel to the cases of parasitic castration in various 
Invertebrata caused by such various parasites as Crustacea, 
Sporozoa, and worms of various kinds. Ina great number 
of these cases the effect of the parasitic castration is to 
arrest the development or cause the atrophy of the primary 
and secondary sexual characters without actively calling 
forth the production of the female sexual characters in 
the parasitised male. In other cases (as far as we know 
only in the Crustacea) besides the suppression of the sexual 
characters both primary and secondary proper to the 
infected individual, we find the active assumption of female 
characters by the parasitised male, as described in Parts 2 
and 3 of these studies. ‘he particular case just described 
belongs, as far as the evidence goes, to the former of these 
two categories, v. e. that in which certain of the male sexual 
characters atrophy without the active assumption of female 
characters. The principal interest attaching to this case 


24.0 GEOFFREY SMITH. 


consists, firstly, in establishing a bacterial disease of a verte- 
brate asa cause of parasitic castration and thus extending 
the operation of this principle to two new classes of organisms, 
and secondly, in bringing out the correlation between the 
activity of the testes and the development of the comb and 
wattles of Gallus bankiva. Inthe next part this correlation 
will be dealt with more fully on an experimental basis. 


LETTERING. 
C. 8S. Connective tissue sheath. Hn. Endopodite. Ex. Exopodite. 
N. Germinal nuclei. O. Ovary. TT. Testis. V. S. Vesicula seminalis. 


EXPLANATION OF PLATE 14, 
Illustrating Mr. Geoffrey Smith’s paper on “ Studies in the 
Experimental Analysis of Sex.” 


All the figures refer to Inachus mauretanicus (Lucas). 

Fig. 1.—First abdominal appendage (copulatory style) of normal 
uninfected male. x 5. 

Fig. 2.—Second abdominal appendage of normal uninfeeted male. x 5. 

Fig. 3.—First abdominal appendage of infected male “A.” x 5. 

Fig. 4.—Second abdominal appendage of infected male “A.” x 5. 
(This figure might serve equally well for the abdominal appendage of 
an adult female.) 

Fig. 5.—Second abdominal appendage of infected male“ B.” x 5. 


= 


Fig. 6.—First abdominal appendage of infected male * BB.” ee 

Fig. 7—Second abdominal appendage of normal uninfected female, 
before adult condition is assumed. x 5. (The adult form of this 
appendage is practically identical with that given in fig. 4.) 

Fig. 8—Vesicula seminalis of a small normal male, measuring 14 mm. 
carapace length. xX 20. 

Fig. 9.— Coils of testis of the same male. X 20. 

Fig. 10—Vesicula seminalis, duct, and coils of testis of infected 
male“ A.” xX 20. 

Fig. 11.—Portion of testis of an infected male, showing absorption 
of germinal cells in connective-tissue sheath. x 30. 

Fig. 12.—Portion of ovary of an infected female, showing absorption 
of ova and germinal cells in connective-tissue sheath. x 930. 

Fig. 13.—Another portion, higher magnification, of ovary of infected 
female. X 60. 


PEER ELOOR AAO 
RESIS SA 
pi kt te 


Se ea i Se Lies 
SOOO SR d < ee 7 08 
a Siece SS if PES 
SAE 
Mcafee 
~ ’ Z 


& = 


ae sq 


eet 


aes wee = ‘ 
: meng eR Se SO > 
Rea eee eS cca = 
——— Sey =_ 
Se a 
Sh 


Huth Tath? London 


s 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 241 


Some Observations on a Flagellate of the Genus 
Cercomonas. 


By 
Cc. M. Wenyon, M.B., B.S., B.Sc., 
Protozoologist to the London School of Tropical Medicine. 


With 19 Text-figures. 


In the present paper I shall describe a flagellate of the 
genus Cercomonas, a genus first created by Dujardin, in his 
‘Historie Naturelle des Zoophytes Infusoires,’ published in 
1841. Since Dujardin’s original description numerous flagel- 
lates have incorrectly been attributed to this genus, so much so 
that Klebs, in his ‘ Flagellatenstudien’ (1893), says that this 
genus has not been defined with sufficient accuracy, that it 
has been confused with Heteromila and Bodo by the over- 
looking of the tail flagellum, and that the genus Cercomonas 
must be rejected. It is undoubtedly true that the genus 
Cercomonas is very confused, and this confusion has been 
considerably heightened by the description of Cercomonas 
from the intestine of man and other animals. Davaine (1854) 
was the first to record the presence of Cercomonas in the 
evacuations of a man suffering from cholera. Without going 
into the question of the correctness or otherwise of Davaine’s 
conclusions, it is undoubtedly a fact that many observers, 
noting the presence of active flagellates in the intestinal 
contents, have attributed them at once to the genus Cerco- 
monas, and as a result of this various species of 'l'richomonas, 
Lamblia, and possibly other flagellates have been included in 
this genus. In the present instance the flagellate to be 


242 Cc. M. WENYON. 


described was found in the feces of a patient in the Albert 
Dock Hospital at the London School of Tropical Medicine. 
This patient was infected with Entamceba coli, and in order 
to observe changes inthe encysted forms of this amceba some 
of the feces were placed in a clean glass-stoppered bottle. 
In the course of a few days it was noticed that large numbers 
of flagellates were present. It is probable they had developed 
from cysts which must have been present in the feces. On 
first examination it was seen that these flagellates corre- 
sponded very closely with the original description of Dujardin 
for the genus Cercomonas, and for this I took them to be. 
On more careful examination I found that the tapering 
posterior end was in reality a second flagellum, and that this 
could be traced along the surface of the body to which it was 
attached as far as the insertion of the long anterior flagellum. 
The presence of this posterior flagellum and its attachment 
to the body required very careful observation to make out, 
for it can only be clearly seen in certain portions of the 
animal, and it ig quite conceivable, as Klebs maintains, that 
Dujardin overlooked this posterior flagellum. Dujardin’s 
original description of the genus is as follows: 

“Genre Cercomonas. 

“An, arrondi ou discoide, tuberculeux, avec un prolonge- 
ment postérieur variable, en forme de queue, plus ou moins 
long, plus ou moins filifornie. 

Les Cercomonas ne different absolument des Monads que 
par un prolongemert postérieur, formé par la substance 
méme du corps qui s’agglutine au porte-objet, et s’étire plus 
ou moins, de maniére a n’étre tantét qu’un tubercule aminci, 
tantot une queue allongée transparente, tantot enfin un tila- 
ment presque aussi fin que le filament antérieur, et suscep- 
tible @un mouvement oudulatoire; mais bien scuvent j’ai cru 
voir les Monades passer par degrés |’état de Cercomonas.” 

A comparison of this description with that now to be given 
will show how closely the two agree. 

The occurrence of this flagellate has been described above. 
By transplanting into other media I have been able to keep 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 243 


cultures of this flagellate free from other Protozoa for about 
a year, and it is only that cireumstances preventing ine from 
continuing these observations I now describe what results I 
have already obtained. 


MerHop or OBSERVATION. 


I have found the best liquid culture medium to be hay 
infusion to which a small quantity of faeces has been added. 
The flagellates will live and multiply in hay infusion alone, 
but, as in other thin media, the numbers of flagellates are 
always very small, sothat any observation is difficult to make. 
In the thicker medium the numbers are not only larger but the 
movements of the Hagellates are slower and accordingly more 
easily followed. For keeping stock cultures small test-tubes 
were used as in bacteriological methods, but for making 
observations hanging-drops in the moist chambers of Max 
Schultze were most useful. In these hanging-drop prepara- 
tions the flagellates would live for weeks, till finally, all nutri- 
ment being used up, encystment followed. | By the addition 
of fresh nutriment to the hanging-drop the culture would 
commence again. 

In addition to the liquid medium I have found the solid 
agar medium used for the culture of amcebe most useful. It 
was first employed for the culture of flagellates by Berliner. 
This observer, working with Copromonas major, found 
that on the solid medium the flagellates multiplied rapidly 
till enormous numbers were present. I can fully confirm 
this, and for the study of the details of nuclear division 
the presence of such large numbers of dividing forms is 
very useful. ‘The medium I employed differed slightly from 
that used by Berliner. For the culture of amcebe I have 
used with success the medium first invented by Musgrave 
and Clegg, and I have found it equally good for the flagel- 
lates at present under discussion. I have employed it in 
the ordinary Petri dishes. By unveiling the dishes the 
progress of the culture may be watched under the low 
powers of the microscope. A very useful method for the 


244 Cc, M. WENYON. 


use of this medium, and one which will allow observations to 
be made with high, powers, is the following: A long cover- 
glass (1} inches) is taken and carefully cleaned, On a clean 
slide ridges of Czokor’s wax, first recommended to me by 
Professor Minchin, are so arranged, about an eighth of an inch 
high, that the cover-glass will form the lid of a box. Some 
of the medium is melted by placing the test-tube in boiling 
water, and a small drop of this is allowed to fall on to the 
cover glass, which is lying on the top of the hot-water oven. 
By careful tilting of the cover-glass the melted medium will 
form a very thin layer over the cover-glass, which is then 
removed so that the medium may solidify. he surface of the 
medium is then inoculated with a small quantity of material 
from a previous culture and the cover-glass inverted on the 
wax ridges. By means of a hot wire and more wax the whole 
may be completely sealed up. It is most essential that not 
the smallest opening be left, or it will be found that the 
medium will quickly dry and the culture end. 

In this way it is easy to follow the multiplication of the 
flagellates with the in. objective, and if the film of medium 
has been made sufficiently thin the oil-immersion may be 
employed. 

In every case where the flagellates grow in the solid 
medium their chief nourishment seems to be the numerous 
bacteria that grow at the same time. 

For studying the flagellates in the fixed and stained con- 
dition the cover-glass method has been mostly used. . Some 
of the liquid medium or some of the culture scraped from 
the surface of the agar is spread on a clean cover-glass, and 
without allowing it to dry it is dropped, film side down, on to 
the surface of some fixing fluid. Another method of obtain- 
ing a film from the agar cultures is tlis: A cover-glass is 
dropped on to the surface of the agar culture in a Petri dish. 
It is gently pressed down till its surface is seen to have 
touched the culture. On raising it with a needle it will be 
found that a layer of the culture is adherent to the cover- 
glass, and it may be fixed as before. 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 2495 


For fixing the flagellates the most useful fixative has been 
Schaudinn’s mixture of two thirds saturated aqueous solution 
of sublimate and one third alcohol, slightly acidified with 
acetic acid. ‘his has been used in the manner just described 
by Schaudinn or ina slightly modified form. The films are 
best stained with iron-hematoxylin. 


DESCRIPTION OF THE Livinac FLAGELLATES. 


When examined in a drop of liquid medium on a slide the 


Text-figs. 1-8.—Drawings from life. 
Text-Fi@. 1. TEXT-FIG. 2. 


Ameeboid form in early division Ameeboid form. 
stage. 


flagellates appear as pear-shaped organisms, with a long 
flagellum, about twice the length of the body, arising from 
the blunt end. he posterior end of the body is, as a rule, 
drawn out to a fine and tapering point. By the constant 
lashing of this long anterior flagellum the animal is drawn 
along. Sometimes the flagellum is, as it were, hooked around 
some distant object, and by its flexion pulls the body towards 
this point. ‘The posterior end of the body, which, as stated 
above, is also a flagellum, moves much less vigorously than the 
anterior. Its movements may be quite passive, being only 
the accidental changes in position produced by the changes 


246 C. M. WENYON. 


in shape of the body. At other times there is a distinct 
to-and-fro or lashing movement, but at its maximum it is 
much less violent than that of the long anterior flagellum. 
The protoplasm of the body may be continued along this 
posterior flagellum for a considerable distance. On very 
careful focussing it can be seen that the posterior flagellum 


TEXT-FIG. 3. 


Two ameboid forms with entangled flagella. 


is attached to one side of the body, and really arises from the 
insertion of the anterior flagellum.% This is very well shown 
in some of the figures,e.g.3,5,9. When the body is viewed 
in certain positions it is seen that it is distinctly flattened 
aloug the line of attachment of the posterior flagellum (fig. 10), 
and when the posterior flagellum is moving at its maximum 
rate this flattened edge of the body shows slight but distinct 
undulatory movement, reminding one most strikingly of the 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 247 


movements of the blood inhabiting Trypanoplasma. 
Indeed, this flagellate in many respects occupies a position 
intermediate between the genus Bodo and Trypano- 
plasma. 

The nucleus is clearly visible in the living animal. There 
is a distinct membrane, and at the centre of the nucleus is a 
large karyosome. The nuclear membrane is drawn out at 
one pole towards the insertion of the two flagella, and occa- 
sionally a clear line may be detected connecting the apex of 


TEXT-FIG. 4. 


Division-stage of free-swimming form. 


the nucleus with the base of the two flagella. The details of 
these structures are much more evident in the fixed and 
stained films. The protoplasm of the body contains food 
and other vacuoles, but contractile vacuole is not present. 
Sometimes the nucleus is surrounded with refractile granules, 
having the same greenish line and refraction as the karyo- 
some within the nucleus. ‘hese may be present in sufficient 
numbers as to completely obscure the nucleus. Similar 
granules occur in the protoplasm of eucysted forms (fig. 6). 
These granules stain deeply, and are possibly of a chromatin 
nature. 

In the hanging-drop preparations especially this organism 


248 C. M. WENYON. 


exhibited a peculiar polymorphism. In the central part of the 
hanging drop, where the fluid was deep, the flagellates had 
the typical pear-shaped appearance, with the long, tapering, 
posterior extremity. At the sides of the hanging drop, where 
there was only a thin layer of moisture on the cover-glass, the 
typical pear shape was lost and the flagellates had the appear- 
ance of amcebe. When first I observed this I thought my 
culture had become contaminated with an amoeba, but the 


TEXT-FIG. 5. 


Ordinary free-swimming type. 


presence of the long anterior flagellum and the short posterior 
one disproved this idea. It was possible to watch a single 
individual swimming in the deep part towards the edge. On 
reaching the shallow part the character ot the organism 
changes at once to the amceboid form. Pseudopodia are pro- 
truded and withdrawn, and the animal creeps about in a 
typical amceboid manner. All this while the long anterior 
flagellum is lashing to and fro, but appears powerless to draw 
the animal across the surface of the cover-glass. It is only in 
the deeper part of the hanging drop that the flagellum is useful. 
The posterior flagellum is often not visible, and its prolonga- 
tion across the surface of the body is more difficult to detect. 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS, 249 


When seen it is inert and only moves in a passive manner. 
It seems to take little share in movements of progression. 

On the surface of the agar medium the organism is generally 
of the amceboid form. 

At the edge of the hanging-drop preparations or on the 
surface of the agar it is easy to watch these amceboid forms 
ingesting food by surrounding objects with pseudopodia. As 
a rule the amceboid forms contain many more food-vacuoles 
than those swimming in the deeper layers. 


= 


PExXT-ETG. 7. 


Less regular encysted form. 


Reproduction is by longitudinal division. ‘There is first 
multiplication of the flagella, whether by new formation or 
division of those already existing has not been determined. 
The nucleus next divides. The karyosome is divided into two 
parts, and finally the elongated nuclear membrane becomes 
constricted and two nuclei are formed. After a short time 
the protoplasm becomes drawn out and finally a constriction 
appears, which ultimately. ends in complete division. ‘The 
process of this division is very readily watched on the cover- 


250 OC. M. WENYON. 


elass cultures described above. Both the amceboid and the 
tree-living forms divide in this manner, but on account of the 
more sluggish movements of the former they are more readily 
kept under observation. 

In the cultures encysted forms commence to appear after a 
few days. In the liquid cultures they are to be found in the 
scum on the surface or in the deposit at the bottom. On the 
agar cultures the cysts appear in the older parts of the culture. 
On this medium the margin of bacterial growths spreads over 
the surface, and in this margin the actively reproducing 
flagellates are to be sought. In the oldest part of the culture 
no free flagellates can be found, but only the cyst. 


TEXT-FIG. 6. Text-FIG. 8. 


Encysted forms showing refrac- Free form with refractile granules. 
tile granules — surrounding Probable preparation for en- 
nucleus. cystment. 


In the fresh condition these cysts appear as slightly brownish 
spherical bodies, with a wall of double contour. 

At the centre of the cyst is the spherical nucleus, which 
has similar characters to that of the free form, except for the 
prolongation towards the flagella. ‘The nucleus is surrounded 
by the bright refractile granules, which were described as 
occurring in some of the free forms. It is probable these 
granules are of a chromatin nature, and that they arise from 
chromatin passed out from the nucleus, though this process 
has not been followed. 

‘hough these organisms have been kept under observation 
for a year or more conjugation has not been seen, nor has 
any sexual process been detected. it is possible that some 
sexual process is bound up with the encystment, but as the 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 251 


entrance into and emergence from the cyst has not been 
directly observed and no multiplication within the cyst could 
be seen nothing definite on this point can be stated. 


Text-figs. 9-19—Drawings from stained preparations. 
TEXT-FIG. 9. 


Free-swimming form with granules round nucleus. 


FIXED AND STAINED SPECIMENS. 


In the fixed and stained specimens, in addition to the 
details which were so clearly visible in the living organism, 
others could be made out. 

The: protoplasm of the body has a marked alveolar structure. 
The anteriorly placed nucleus shows a large, deeply staining 


Poe CG. M. WENYON. 


karyosome, while connecting this latter body to the nuclear 
membrane is a coarse linin network. All the chromatin of 
the nucleus appears to be concentrated in the karyosome. 
The prolongation of the nuclear membrane towards the 


Trxt-Fic. 10. 


Side view of free form showing the flattened side along which 
the flagellum runs. 


flagella is clearly shown, while the base of these organs is 
connected to the apex of the nucleus by a rod-like rhizoplast. 
In some cases the drawn-out apex of the nuclear membrane 
shows longitudinal markings, which converged toward the 
rhizoplast, while in others there is a connection in the form 
of a more deeply staining pyramin between this body and the 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 253 


karyosome (fig. 11). _Prowazek describes for Cerco- 
monas longicaudia a “ein Art undeutlichen Zwischen- 
fibrille,’ which connects the karyosome to the insertion of 
the flagella. Prowazek figures this:connection as a dark line 
running from the karyosome to the apex of the nucleus, but 
I have not been able to detect any structure as definite as the 
one he figures. 

This flagellate is a very excellent illustration of the fallacy 
of relying for detail on the old dry Romanowsky methods of 


Taxt-rie, 11, 


Shows connection of karyosome and rhizoplast. 


staining. The nucleus of this organism is clearly visible in 
the living condition. There is a definite nuclear membrane. 
At the centre of the nucleus is a large refractile karyosome, 
while the space between this body and the nuclear membrane 
is free from granules. The nuclear membrane is drawn out 
at one point towards the insertion of the flagella. Now if a 
film of the material containing this flagellate is allowed to 
dry as in the usual method for the preparation of blood for 
staining trypanosomes, and stained by one of the modifications 
of the Romanowsky method, the result may be very beautiful 
from the colour point of view, but totally misleading in the 
structure of the nucleus. This latter organ appears in these 


VOL. 55, PAR’ 2.—NEW SERIES. il 


254 0. M. WENYON. 


dried films as an irregular clump of red staining granules. 
In other words, its appearances are like those of the nuclei of 
trypanosomes in similarly prepared films. In films fixed and 
stained by the wet method described above the structure of 
the nucleus is comparable with the appearances to be made 
out in the living organisms. 

The details of longitudinal division can be followed in the 


Taxt-ries, 12, 13. 


Dividing forms. 


stained preparations. The large karyosome becomes elongated 
and constricted, and finally divided into two parts (fig. 15). 
I was never able to detect within the karyosome a centriole, 
spindle, and zquitorial plate, as described by Berliner in the 
division of Copromonas major, but the division takes place 
in an amitotic manner, resembling that of Copromonas 
subtilis (Dobell). Most usually the karyosome becomes 
distinctly dumb-bell shaped as in fig. 12, but at other times 
the division is along the longitudinal axis of the elongated 
karyosome, the resulting daughter-karyosomes each being 
elongated (figs. 15, 16). Following the division of the karyo- 
some the nuclear membrane elongates while the daughter- 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 255 


karyosomes separate. The flagella are duplicated at this stage, 
but they still have a common rhizoplast, which is inserted 
into one point of the elongated nuclear membrane, which is 
drawn out slightly at this point towards the anterior end of 


Trxt-Fies. 14-16. 


: re, ots 2, ; 
ae * {eps a, ~ ge 
py = Pas a : St : 
+ os er 4 
pt < , boa, aren ae ee me 
LEON en ee ‘ . 
ving fe : LZ ere F a 
“re epee wos 
*G. eae a “a= 
*, : e.:3 Lee ae 
Q ¥ wee LS 4 é 
; A Psd 
“n ay rae “5 a 
i4. len 16 
. 


Dividing forms. 


the body of the flagellate. Division of the nuclear membrane 
commences by a constriction at the point opposite the in- 
sertion of the rhizoplast. ‘The division is completed, and the 
two nuclei, each with an apex, are connected to the base of 
the rhizoplast. The rhizoplast finally divides longitudinally, 
so that there result two nuclei, each with arhizoplast and two 
flagella. The exact method of origin of the flageila I was 


256 Cc. M. WENYON. 


unable to trace, though some of the appearances seem to 
indicate the formation of two new ones by outgrowth from the 
rhizoplast. In fig. 18 is the nuclear apparatus of a flagellate 
partially broken up on the film. It shows very clearly the 
single rhizoplast with the duplicated flagella. The last stage 
in the division process is thus the splitting of the rhizoplast, 
while the first stage is the multiplication of the flagella and 
the commencing division of the karyosome. After complete 
division the nuclei pass to opposite poles of the body (fig. 17), 


TeExt-FIG. 18. 


Is. 


Part of nucleus, rhizoplast, and flagella of partly broken-down 
individual, to show the multiplication of the flagella before 
division of the nucleus and rhizoplast. 


and after a varying interval of time the body is divided into 
two equal parts. 

The bright refractile granules which were described above 
as occurring in the protoplasm around the nucleus in the 
encysted forms and in some of the free forms appear in the 
stained specimens as dark-staining granules. Whether these 
are chromatin granules of the nature of a chromidium or 
whether they are capable of some other interpretation cannot 
be definitely stated, since their fate has not been followed. 
They certainly stain as chromatin, and their presence within 
the cyst (fig. 19) would seem to suggest the possibility of 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 257 


their being nuclei of spores destined to escape from the cyst 
and ultimately to develop, with or without conjugation, into 
the adult flagellate form. Though cysts have been constantly 
kept under observation and every inducement possible to 
encourage the emergence from the cyst has been tried, I have 
never been fortunate enough to witness this process. That it 
does occur is borne out by the experiment of adding dried 
cysts to fresh medium, resulting in a culture of flagellates. 

In the stained preparation certain appearances are capable 
of interpretation as a conjugation of the flagellates, and some 
of the nuclear appearances as processes of maturation, but as 
no undoubted conjugation was observed in the living flagel- 


Trxt-Fic. 19. 


Cyst showing dark-staining granules surrounding the large 
central nucleus. 

lates [refrain from describing these. Without the control of 
observation on the living forms descriptions of conjugation 
and the accompanying nuclear changes are of little value, 
since the possibility of error in interpretation is very great. 
For Copromonas major Berliner has described from stained 
preparations such a process of conjugation, but without the 
necessary controls it is always possible that abnormal or in- 
volution forms have been mistaken for such stages. 

In rich cultures of the flagellates there is a very great 
variation in size. Some individuals are comparatively large, 
reaching a length of 15 or more, excluding the flagella. 
Others are very minute, being not more than 2-3 u in longest 
diameter. All intermediate sizes are to be met with in the 
cultures. ‘he encysted forms have a diameter of about 6 or 


258 C. M. WENYON. 


fod 


7. These cysts will withstand drying at ordinary laboratory 
temperatures, and are capable of giving rise to fresh cultures 
when brought into suitable media. 


NOMENCLATURE. 


It is certain that Dujardin’s original description of the 
genus Cercomonas is incomplete, but it seems to me quite 
clear from his account that he was dealing with flagellates 
similar to the one described in this paper. ‘Though he did 
not definitely state that the fine drawn-out posterior extremity 
of the body was a flagellum, still, he says that it was at times 
so fine as to resemble the anterior flagellum, and that it was 
capable of independent movements. Further, in his table of 
classification he divides the Monads into two groups. In the 
first he includes forms with “un seul filament flagelliform,” 
while in the second those with ‘pleusieurs filaments ou 
appendices.” The genus Cercomonas appears in the second 
of these groups as a form with ‘‘un second filament ou 
appendice postérieur.” It is therefore quite evident that 
Dujardin regarded this posterior termination of the body as 
of the nature of a flagellum. Stein and Blochmaun describe 
the genus Cercomonas as having a drawn-out posterior end, 
though they do not describe a definite flagellum. ‘The genus 
Cercomonas was not accurately defined by Kent or Biitschli, 
and to Klebs the confusion seemed so great that he proposed 
the rejection of this generic name and the substitution of 
Gruber’s name Dimorpha, which was created for a_bi- 
flagellate showing at certain stages definite heliozooid 
characters. In this genus Dimorpha Klebs included forms 
which he identified. with those described originally by 
Dujardin as Cercomonas, and he suggests that this observer 
has overlooked the second flagellum. We have seen how 
near Dujardin was to definitely describing this second 
flagellum, so that the action of Klebs in rejecting this 
genus is hardly sound. It seems to me clear that the forms 
described by Dujardin really possessed two flagella, though 


OBSERVATIONS ON A FLAGELLATE OF CERCOMONAS. 259 


he failed to see this clearly. On this account I think it safer 
to retain the genus Cercomonas for flagellates of the 
character described in this paper, viz. flagellates with an 
anterior blunt end from which arises a single long flagellum 
and a posterior tapering end also with a flagellum, trace- 
able over the surface of the body towards the insertion of the 
anterior flagellum. This conclusion is come to by Prowazek 
also, who figures Cercomonas longicauda with two 
flagella arising from the nucleus. 

The specific name of this flagellate is difficult to determine. 
Dujardin named several species of Cercomonas, though he 
was careful to state that he was far from regarding these as 
true species, but as a convenient means of distinguishing the 
forms met with in different infusions. From the figures of 
Dujardin and Stein it is possible that the flagellate belongs 
to the species longicauda, so that the flagellate described 
here may be assumed to be Cercomonas longicauda 


Dujardin. 


REFERENCES. 

Dujardin (1841).—* Histoire naturelle des Zoophytes Infusoires,’ Paris. 

Dallinger, W. H., and Drysdale, J.—‘* Researches in the Life-history of 
a Cercomonad, a Lesson in Biogenesis,” ‘ Monthly Micr. Journ.,’ 
VOlaxa Oe: 

Davaine, C. (1875).—** Monadiens,” in ‘ Dict. Eneycloped. des Sciences 
Médie.,’ t. ix. 

Kent, W. S— A Manual of Infusoria,’ London, 1880-82. 

Klebs, G. (1893).—* Flagellatenstudien,” ‘ Zeit. wiss. Zool.,’ vol. lv, 
p- 265. 

Blochmann, F. (1895).—** Die mikroskopische Tierwelt des Susswassers,” 
Abt. I, ‘ Protozoa,’ Hamburg. 

Stein, F.—‘ Der Organismus der Infusionsthiere,’ Abt. III, 2 Halfte. 

Prowazek, 8S. (1903).—* Flagellatenstudien,’ ‘Arch. fiir Protisten- 
kunde,’ Bd. ii. 

Biutschli, O—In Bronn’s ‘ Klassen u. Ordn. d. Tierreichs,’ 1885-87. 

Dobell, C.—** The Structure and Life-History of Copromonas sub- 


tilis, nov. gen., nov. sp.,” ‘Quart. Journ. Mier. Sci., vol. 52, 
1908. 


260 Cc. M. WENYON. 


Gruber, A. (1881).—‘“Dimorpha mutans,” ‘ Zeit. wiss. Zool.,’ Bd. 
XXXVI. 
Berliner, E. (1909).—* Flagellatenstudien,” ‘ Arch. fiir Protistenkunde,’ 
Bd. xv, H. 3, p. 297. 


SOME OBSERVATIONS ON A NEW GREGARINE. 261 


Some Observations on a New Gregarine (Meta- 
mera schubergi nov. gen., nov. spec.). 


By 
H. Lyndhurst Duke, B.A., B.C.Cantab., 


With Plates 15 and 16. 


ConreEn'ts. 
PAGE 
Introduction : , : : : » e206 
Material and Methods : ‘ : ; . 263 
Structure of the Trophozoite ; : . 266 
Cyst-formation and Development of the ene es ‘ = 210 
Discussion of Some Special Points in the Life-cycle. 28 
Diagnosis of Metamera schubergi ; 282 
Literature References : : : : . 282 
Explanation of Plates : : : ‘ . 284 
INTRODUCTION. 


Wate working at Heidelberg in 1906, under Professors 
Biitschli and Schuberg, the latter kindly called my attention 
to a new species of gregarine in the gut of Glossosiphonia 
complanata L. (Clepsine sexoculata), and suggested its 
further investigation. The preceding summer, while busied 
with a recently discovered coccidium occurring in the leech 
Herpobdella atomaria Car. (= Nephelis vulgaris), 
Professor Schuberg turned his attention to Glossosiphonia 
complanata, which occurs in company with Herpobdella 
in the Neckar and occasional ponds in the Heidelberg district. 
Deeming it probable that two forms so alike in habit and 
environment might harbour the same parasites, he dissected 


262 H. LYNDHURST DUKE. 


several specimens of this leech, and, though the results were 
in the main negative, he found several animals infected with 
a species of gregarine. Reference to the literature proved 
the parasite to be identical with a species briefly mentioned 
by Bolsius in 1895 (2), and the subject of a more detailed 
but still fragmentary paper in 1896 (3). Beyond a super- 
ficial study carried on incidentally during his work ou the 
Glossosiphonia Bolsius seems to have paid no further 
attention to the parasite, which remained unnoticed until 
1900, when Castle (5), in an exhaustive treatise on the 
N. American Rhynchobdellide and their parasites, mentions 
having observed the gregarine seen by Bolsius in about half 
the specimens of Clepsine elongata which he examined. 
He adds, however, that he only finds the animals in the 
stomach diverticula, and never in the intestine or crop, as 
indicated by Bolsius in his diagrams. Castle also mentions 
encysted protozoa which he found in C, fusca, and suggests 
the possibility of their relationship to the form in G. com- 
planata. ‘he cysts he found in the muscle-layers of the 
body-wall, so that they probably have nothing to do with 
the gregarine in question. 

Liihe (14) quotes the parasite as having been mentioned 
by Bolsius, and suggests that it probably belongs to the 
tricystid gregarines. 

The gregarine is thus a new and previously undescribed 
form, for which I propose the name Metamera schubergi.! 

In the preparation of the sections and the study of the 
living animal, during the last few weeks of my stay in Heidel- 
berg, Professor Schuberg assisted me most kindly in every 
way in his power; and it is due solely to him that I was able 
to obtain Bolsius’ principal pamphlet. My thanks are also 
due to Geheimrat Prof. Biitschli, whose practical suggestions 
I found of the greatest value. 


1 The form which appears most closely allied as regards structure of 
the trophozoite is Echinomera. A study of the life-history, however, 
has revealed points of difference which seem to warrant the creating of 
a new genus for the form under consideration. 


SOME OBSERVATIONS ON A NEW GREGARINE. 263 


By the kindness of Professor Sedgwick, who allowed me a 
free hand in the laboratory of the Imperial College of Science, 
8. Kensington, I was able eventually to complete my study of 
the sections. And in this connection I must express my 
indebtedness to Mr. C. C. Dobell, who is at present lecturing 
at the College. His unrivalled knowledge of protozoan life- 
history and technique has always been most generously 
placed at my disposal, and has proved of the greatest value in 
the preparation of this paper. 


Mareriat AND MeruHops. 


The leech which serves as host to Metamera schubergi 
is Glossosiphonia complanata Linn. A few specimens 
of Hemiclepsis marginata werealso found infected. The 
leeches live under stones in shallow water—running by pre- 
ference—though I have found them in smaller numbers in still 
pools. The material was collected at Heidelberg from the 
shallows left by the summer fall of the Neckar in the neigh- 
bourhood of the electric power station, below the new bridge, 
and also from the opposite bank, along the wall separating 
the skating rink from the river itself. The leeches are fairly 
common, and may be found clinging firmly to the under-side 
of stones at the water’s edge, especially in the numerous lumps 
of red sandstone which hitter the shore everywhere. 

Recently I examined some specimens of Glossosiphonia 
complanata sent me from the neighbourhood of Cambridge, 
and found them well infected. These latter were obtained 
in January, when the leeches are hard to find owing to the 
scanty vegetation in the ponds in winter. In all the speci- 
mens I examined from this source I only obtained one cyst, 
and that a very small and early one. 

The leeches can be kept for an indefinite period in a good- 
sized glass jar, provided the water be aérated by passing 
bubbles of air through it. Food is not necessary, though a 


' For this I have to thank Mr. Harding, and also for his kindness in 
assisting me to determine the species. 


264, H. LYNDHURST DUKE. 


few small water-snails are much appreciated. Owing to the 
transparent nature of the integument in Glossosiphonia, 
the parasites are visible in the living leech ; and if the latter be 
forcibly pressed between two slides provided with wax corners, 
and examined under a low magnification, the gregarines may 
sometimes be detected in the stomach diverticula and intes- 
tine. Unfortunately, however, this method of diagnosis is 
by no means infallible, as the numerous pigment-cells with 
their clear nuclei look very like gregarines, and render 
accurate observation impossible. ‘lhe gregarines occur in 
the hindermost stomach diverticula and the intestine, just as 
indicated by Bolsius in his diagram. The cysts are found in 
the same regions of the alimentary canal, but are especially 
numerous in the intestine. 

Kxamination of sections shows that cysts can develop as 
far as the sporoblast stage in the intestinal canal of the host, 
though they are often expelled with the feces at a much 
earlier stage in development. 

In sections just above the anus no cysts were to be seen. 
This part of the gut was almost occluded by a mass of 
cephalonts and some sporonts of a peculiarly blunt outline. 
The leech from which these sections were cut had previously 
evacuated feces containing a few very early cysts among 
a greater number in which sporoblasts could be distinguished. 
As many as ten cysts have been counted in one section. 

‘To obtain the gregarine, the infected leeches were partially 
dried on blotting-paper and the under-surface opened by 
three incisions—two parallel and close to the margins, and 
one at right angles to the long axis of the animal, at about 
the junction of the middle and anterior thirds. The flap of 
tissue was then carefully turned backwards towards the anal 
sucker, the animal being placed in a watch-glass containing 
normal saline solution. ‘The gut-contents were thus emptied 
into the saline, together with connective tissue, which is of 
no account. By the aid of a hand-lens the gregarines could 
now be seen sticking to the bottom of the glass, or still fixed 
to fragments of the host-tissue. These latter are useful in 


SOME OBSERVATIONS ON A NEW GREGARINE. 265 


studying the structure of the epimerite, as this organ, in the 
course of the teasing out, 1s very easily torn away, leaving 
decapitated individuals which may be confused with true 
sporonts. By gentle coaxing with a pipette the gregarines 
can be freed from the bottom of the watch-glass and trans- 
ferred to a slide for further handling. 

Preparations in toto were made originally under a cover- 
slip provided with wax feet, and the various reagents drawn 
through with blotting-paper. In this way, by fixing the 
gregarines with alcohol and glacial acetic acid (9 : 1), a large 
number of animals may be treated under one cover-slip, which 
is an obvious advantage. More recently I made some pre- 
parations by fixing the selected gregarines in a watch-glass 
with picro-acetic acid (3:1) and adding the various fluids by 
means of a pipette and eventually pickmg out and mounting 
the stained gregarines under a low magnification. I consider 
the former method of treatment the more satisfactory and 
certainly less laborious. As stains for these preparations I 
used Grenacher’s alcoholic carmine solution and Schuberg’s 
modification of Mayer’s acid carmine. This latter solution, 
being acid in reaction and not neutral, has the power of 
penetrating the cuticle, and in employing it the preparations 
must be very rapidly washed through with } per cent. 
solution of HCl to prevent precipitation of the carmine during 
the further treatment with the alcohols. Leeches destined 
for sections were fixed either in Gilson’s fluid or in the above- 
mentioned alcoholand acetic mixture. Gilson’s fluid should act 
for two or three hours, and the sublimate constituent be most 
carefully washed out with iodine-alcohol or a solution of KI 
in 75 per cent. alcohol. As staining reagents hematoxylin 
(Delafield’s) and eosin, safranin, and Heidenhain’s iron-hema- 
toxylin were employed. Owing to the paucity of material, 
the laborious expedient of applying both methods in succession 
on the same preparation had to be employed. It was found 
that hematoxylin and eosin were satisfactory for the cepha- 
lonts and sporonts, but gave very incomplete and misleading 
results with the nuclear changes of the encysted forms, which 


266 H. LYNDHURST DUKE. 


were defined much more distinctly with the iron-heematoxylin 
method. All tissues were embedded in paraffin, with chloro- 
form as the intermediary fluid. 

Culture of the cysts.—To obtain the ripe spores the 
cysts were simply placed in the moist chamber, where, in the 
course of seven or eight days, the spores were developed. 
The cysts were either placed simply on a slide in a drop of 
Neckar water or under a cover-slip provided with wax feet. 
The cysts dehisced by simple rupture after about seven or 
eight days. Cysts placed in normal NaCl solution in the 
moist chamber did not develop successfully. 


STRUCTURE OF THE ‘l'ROPHOZOITE. 


The body is divided by septa into epi-, proto-, and 
deutomerite, and is elongated in form (figs. 1-6). Some 
individuals have a more thick-set appearance than others, 
especially in the extreme hinder end of the gut, where the 
eregarines are often crowded together. The animal measures 
about 150u by 45u. At the posterior end of the deutomerite 
there are often present indications of further subdivision of 
the body, and occasionally as many as three complete segments 
are seen (fig. 4). This segmentation is not confined to 
eregarines of any pecuhar build, being present in both long 
and short forms, and it varies in the degree of development 
of the segments. It was present in about a third of the 
gregarines examined alive in Heidelberg, and is also very 
distinct in the preparations of these animals made at the time. 
The Cambridge gregarines also showed segmentation, though 
it was distinctly less in evidence, both in the living animal 
and in carmine preparations of it. It appears to vary greatly 
—from the very faintest indication to quite definite septa. 
It must be stated in this connection that no segmented 
eregarines were seen in the sections of the infected leeches, 
though constantly found in preparations made by teasing out 
the host-tissues. This compels one to consider the possibility 
of injury during extraction being the cause of this segmenta- 


SOME OBSERVATIONS ON A NEW GREGARINE. 267 


tion, although the stained preparations do not in the least 
degree support this suggestion. 

The epimerite is a dome-shaped structure. It is provided 
with short club-like processes, recalling those of Kchino- 
mera, but often branched, arranged in a dense ring around 
the line of junction with the protomerite, and also on the roof 
of the dome (figs. 4 and 5). These latter processes are 
markedly shorter than those of the ring, and decrease in size 
as the apex of the epimerite is approached. The processes 
are perforated at their somewhat clubbed ends by small 
pores, clearly to be seen in the freshly mounted living 
eregarine by the aid of a ;4; in. oil-immersion lens. Judging 
from analogy with such forms as Echinomera and Ptero- 
cephalus (Nina), and also from the appearance seen in 
sections across the point of fixation to the host, there is no 
doubt that fine pseudopodia are protruded through these 
pores, which fix the gregarine to the intestinal mucous mem- 
brane of the host. The fixing apparatus is by no means easy 
to identify, as, owing to the unavoidable roughness of the 
dissection, the gregarines are rudely torn from their moorings, 
and almost invariably carry away with them a crown-like 
fringe—derived from the host-cells—which surrounds the 
epimerite in the zone of the processes, and obscures all 
details of its structure (fig. 3). 

When kept under observation for some time—say an hour 
or so—in NaCl solution, a curious phenomenon ensues. Just 
at the line of junction between the protomerite and epimerite 
a bubble-like vacuole appears, which gradually increases in 
size, and carries with it the fringe of host tissue with the 
embedded processes till they sit lke a crown on its upper 
pole, sometimes symmetrically, sometimes displaced to one 
side. Having reached a diameter about equal to that of the 
protomerite the vacuole bursts, and the gregarine is suddenly 
deprived of its epimerite (fig. 2). This vacuole formation has 
been seen by Léger and Duboseq to occur in Pyxinia (14), 
and in my opinion has a probable bearing on the mooted 
question regarding the fate of the gregarine epimerite, in the 


268 H. LYNDHURST DUKE. 


transition from cephalont to sporont. Frenzel (14) believed 
the epimerite to be absorbed in a manner similar to the 
assimilation of a tadpole’s tail. He found among numerous 
cephalonts with large epimerites individuals with but a minute 
projection from the protomerite, and he regarded this as a 
scene in the gradual absorption of the epimerite. The sudden 
disappearance he regarded as pathological, and due to 
changes in the surrounding medium. My own observations 
point to the same conclusion. ‘The vacuole formation quoted 
above is plainly due to plasmoptysis, which can be followed 
under the microscope from its earliest onset to the bursting 
of the bubble. Further, when the gregarines were examined 
in a special solution of egg-albumen, NaCl and camphor, as 
prepared by Professor Biitschli, the vacuole formation was 
considerably delayed; a fact explicable on the ground that 
the solution more nearly resembles the natural environment 
of the gregarine. 

The behaviour of the finger-shaped processes also points to 
the epimerite being absorbed rather than directly thrown off 
when the cephalont becomes free. In gregarines which are 
normally lying free in the gut the processes are never to be 
seen (figs. 1 and 6). The epimerite is still present, but the 
processes have been withdrawn during the process of separa- 
tion from the mucous membrane ; just as they are absorbed 
in Echinomera when the cephalont becomes free in the 
gut (17). This applies to all the free-lying specimens seen 
in sections, and to a solitary living form which, together 
with several cysts and some feces, was pressed out through 
the anus during examination of a leech between two slides 
(fies Di 

In the living sporont (fig. 1) the extreme anterior end of 
the animal is quite transparent and devoid of granules, a few 
of which, separate from the main endoplasmic mass of the 
epimerite, may be seen showing Brownian movement along 
its anterior border, After some time the whole granular 
body of the gregarine appears to shrink back somewhat into 
the cuticular sheath which envelopes it, and this clear area 


SOME OBSERVATIONS ON A NEW GREGARINE. 269 


enlarges proportionally until almost the whole of the conical 
knob which forms the epimerite is clear of granules. During 
this process all three divisions of the endoplasm are still quite 
distinct. By the time this stage has been reached osmosis 
asserts itself, and the vacuole formation mentioned above 
commences (fig. 2). In sections, however, the free-lying 
sporonts all showa curious thickening of the extreme anterior 
end of the epimerite, which behaves towards stains in the 
same way as the rest of the cuticle, being, in fact, a thickening 
of the latter anteriorly (fig. 6). It seems a feasible explana- 
tion of this structure to say that it represents the cuticular 
constituents of the numerous processes of the epimerite, which 
have been retracted on the animal becoming free. It may 
here be mentioned that Liithe (14), in his review of the 
gregarines generally, pronounces in favour of the casting off 
of the epimerite as the typical way in which the cephalonts 
become free. 

The nucleus lies in the deutomerite. It consists of a nuclear 
membrane enclosing a clear ground substance, in which lie a 
large vacuolated karyosome and a number of masses of 
chromatic substance (fig. 7). The specimens from which 
figs. 3 and 4 were drawn were very faintly stained owing to 
excessive washing out, but some other preparations stained 
with Grenacher’s carmine confirm the appearancesseen in 
sections, especially as regards the vacuolated nature of the 
karyosome. ‘The nuclear area is about 18 in diameter; the 
karyosome measures about 8 u, and as a rule contains one 
very large vacuole and several small ones. The large 
chromatin masses are scattered irregularly throughout the 
nucleus, and are of varying shape. The nuclear membrane 
is well marked, and in common with the karyosome and the 
chromatin masses stains deeply with both Delafield’s hama- 
toxylin and Heidenhain’s iron-hematoxylin. The ground 
substance takes on a very faint blue tinge with iron-hema- 
toxylin. In some of the sections the karyosome has yielded 
almost completely to the differentiating iron alum, and appears 
grey by contrast with the black chromatin masses. In 

VOL. 55, PART 2.—NEW SERIES. 18 


270 H. LYNDHURST DUKE. 


these cases its vacuolated structure is very plain (fig. 7). As 
a rule, however, the karyosome shows very deeply stained in 
the adult nucleus. Besides the nucleus there are usually to 
be seen scattered throughout the body patches of a substance 
which stains deeply with chromatin stains. These patches 
have been described by Berndt (1) and others, and are espe- 
cially numerous in the protomerite. Comes (7) has recently 
shown that these appearances in Stenophora are probably 
due to metabolic products, and are not nuclear. There are 
also deeply stained granules in connection with the epimerite 
processes in sections stained with iron-hzematoxylin, as 
described by Schellack in Echinomera hispida (17). 


Cyst-FORMATION AND DEVELOPMENT OF THE SPORES. 


The act of association of two animals to form a cyst has 
not been observed inthe livinganimals. As indicated above, 
in the sporont the epimerite tends to become less prominent, 
while a pad of cuticle forms anteriorly. Simultaneously with 
this shortening of the long axis of the body the protomerite 
increases in breadth and bulges, particularly around the 
edges of the apical cuticular pad. From sections it would 
seem that the two animals come together with their epi- 
merites in contact. A ring of cuticle now arises around the 
base of the terminal pad in one animal. Into the cup formed 
by this ring the cuticular pad of the other gregarine is 
inserted, while external to, and dovetailing with the ring of 
the cup, a similar ring of cuticle arises in the second animal 
(fig. 37). In very young cysts in which the nuclei of the 
two animals are still unaltered the above arrangement of 
the parts is very clear; but as development proceeds the 
septum of cuticle dividing the encysted sporonts becomes 
increasingly irregular. In this region in the earlier cysts 
there are patches of deeply stained material suggestive of 
membrane, which are probably the remains of the cuticle of 
the contiguous epimerites (fig. 15). 

Behaviour of the nucleus preparatory to the 


SOME OBSERVATIONS ON A NEW GREGARINE. paral 


formation of the first two daughter-nuclei.— 
Although the material which I was able to collect was 
very limited, I was fortunate in obtaining one leech very 
heavily infected. In the intestine of this animal I found 
numerous cysts, and also an enormous number of adult 
gregarines mostly fixed to the gut-wall. A study of these 
sections has revealed several phases of the first division of 
the nucleus, though to elaborate all the stages is impossible 
without further examples, which I hope shortly to procure. 
In order, therefore, to make the most of this limited material, 
Iemployed first hematoxylin (Delafield’s) and eosin, and 
then after decolorisation with acid alcohol, re-stained by 
Heidenhain’s method. This latter method revealed numerous 
important facts quite indiscernible with the original staining. 
My thanks are due to Dr. Pembrey, of Guy’s Hospital, who 
very kindly provided me with all the necessary apparatus for 
staining. 

For some time at any rate after a definite cyst-wall has 
formed, the nuclei of the encysted gregarines remain appa- 
rently unaltered. Then the chromatin masses begin to frag- 
ment, with the result that chromidia are formed within the 
limits of the nuclear membrane. Simultaneously, this mem- 
brane becomes increasingly thin, and the karyosome throws 
out masses of substance from its interior, becoming in con- 
sequence markedly reduced in size. ‘These masses are more 
or less spherical and of distinct outline; they stain very 
deeply, showing black with iron-hematoxylin. Their number 
and size vary greatly (figs. 9-14). At times one large mass 
is present, almost equal in size to the original karyosome; at 
others, numbers of small massesare seen. The actual process 
of extrusion of one of these masses is shown in fig. 36. After 
their extrusion, the main karyosome-relic shows a blue colour 
with hematoxylin and eosin, as contrasted with the more 
purple hue shown by the intact karyosome and the chromatin 
masses of the trophozoite nucleus. The extruded masses on 
the other hand behave throughout, as regards stains, like 
the chromatin masses. After the fragmentation of the 


272 H. LYNDHURST DUKE. 


chromatin masses and the breaking up of the karyosome 
have proceeded for some time, a new structure appears in 
the nucleus. In close proximity to the main karyosome 
residue, which is seen lying near the periphery of the nucleus, 
an ill-defined mass appears which takes up nuclear stains 
very definitely. The earliest appearance of this mass is shown 
in fig. 9 before the chromidia formation has progressed very 
far. <A. slightly later stage is shown in figs. 10 and 11, where 
the nuclear area presents a homogeneous appearance, without 
any signs of the chromidial elements being discernible, while 
the neighbourhood of the main karyosome residue is occupied 
by a somewhat elongated mass, showing faint longitudinal 
striation (fig. 11). The relative size of this mass, which I will 
call the “achromatic mass,”! is shown in figs. 9, 10,11. It will 
be noticed that the various products of the karyosome are in 
close connection with it. 

At this stage, the absence in my preparations of any 
structures distinguishable as definite chromosomes or cen- 
trosomes is to be emphasised. ‘The achromatic mass stains 
deeply with iron-hematoxylin, but yields to the differentiating 
iron-alum before the karyosome and its products become 
decolorised. 

The next stage in the division represented is shown in figs. 
12 and 13. The achromatic mass has increased in bulk and 
definition, and has become more drawn out. The striation is 
very marked, and for the first time in the course of the division 
the true chromosome element appears. At each pole of the 
achromatic mass, which is now distinguishable as a true 
spindle, there is a small black mass of chromatin; while 
converging towards this mass, like the ribs of a basket, are 
seen deeply stained streaks of granules of chromatin, arranged 
upon the spindle-fibres and obviously en route for the re- 
spective poles of the figure. It may here, again, be seen 
that the spindle stains very deeply with chromatin stains, and 

1 T call this structure the “achromatic mass” because of its function 


—as seen in its later development—and not on account of its staining 
properties. 


SOME OBSERVATIONS ON A NEW GREGARINE. Dil 


it is only on very thorough differentiation that the chromo- 
somes are rendered visible. The spindle. fibres appear to 
merge with the terminal chromatin mass. Distal to this there 
is no true astral arrangement visible. 

Each terminal chromatic aggregation now gives place to 
a definite vesicular structure, situated at the poles of the 
spindle and forming the centre of a definite astral radiation 
(figs. 14 and 15). Simultaneously with the appearance of the 
vesicle, the chromatin streaks and granules disappear from 
the spindle, so that the more definite the terminal vesicle, the 
fewer the chromosomes on the spindle. Fig. 12 shows a ring- 
hke arrangement of the terminal chromatin aggregation at 
one pole of the spindle (a), while fig. 15 shows a true polar 
vesicle containing definite granules of chromatin, in one 
instance arranged indiscriminately around the circumference, 
in the other accumulated at one point upon it. ‘These vesicles 
are the points upon which the very definite spindle-fibres 
converge, and measure from 1$—-24 1 across. In figs. 14 and 
15 it will be noticed, firstly, that—apart. from the granules 
within the vesicles and the karyosome products—there are 
practically no other discrete chromatin elements to be seen ; 
secondly, that some of the spindle-fibres plainly run down 
into the midst of the nuclear area and the karyosome remnants, 
where these latter are not already lying on the spindle. In fig. 
15 will be seen, lying close to the large irregular karyosome 
residue, a collection of deeply stained granules, which are 
connected with the karyosome and with each other by deeply 
stained strands. They have probably been recently thrown 
out from the karyosome, which is much distorted from its 
original spherical shape. 

The latest stage of the first division represented among my 
slides was unfortunately injured before anything more than a 
rough drawing had been made of its structure (fig. 16). It 
represented the spindle very much drawn out, just before the 
final separation of the two daughter-nuclei. There was at 
each pole a well-marked vesicle, containing numerous granules 
of chromatin, and distal to this vesicle was a mass of achro- 


274 H. LYNDHURST DUKE. 


matic substance, showing within it a granule of deeply 
stained substance. ‘he figure was very suggestive of the 
state of affairs seen in fig. 18 a and b, with, however, a single 
polar granule. The sparsity of material unfortunately renders 
a complete account of the first division _ phenomena out of the 
question. From a careful study of the slides at my disposal 
I suggest the following as the more striking points, the 
significance of which I shall revert to later on (see p. 278). 
Firstly, the depth to which the spindle proper stains with 
both Delafield’s and Heidenhain’s hematoxylin: secondly, 
the proximity of the karyosome to the origin of the achro- 
matic mass, and, later on, the very definite spindle-fibres 
running down in among the karyosome remnants and the site 
of the old nucleus: thirdly, the absence of regular chromo- 
somes such as can at any stage be outlined or counted with 
anything approaching certainty: fourthly, the vesicles at 
the poles of the later spindles, which form the centres of 
definite astral figures. The nature of these vesicles it is 
difficult to decide. Are they centrosomes or incipient 
daughter-nuclei? As will be seen later, the daughter-nuclei 
are strikingly vesicular; and the fact that, if these vesicles 
are considered as centrosomes pure and simple, there are no 
other defined chromatic elements in the spindle figure, seems 
to indicate their being early stages of the daughter-nuclei. 
This being the case, the centrosome must be sought either in 
one of the granules on the circumference of the vesicle, or 
distal to the latter. On this point, though tempted to an 
explanation, I dare not base a theory upon a drawing so 
diagrammatic as fig. 16. 

Proceeding to the further division of the daughter-nuclei, 
all uncertainty about the centrosome vanishes. In the 
earliest stages, where eight or nine nuclei are present in each 
cyst (fig. 17 a,b, and c), the astral radiations are very marked, 
and the centrosome consists of a deeply stained mass at the 
periphery of the nuclear vesicle, from which emanate the 
striez. ‘These, where they spring from the centrosome, are 
extremely obvious. In fig. 19 ¢ and d, stained with hema- 


SOME OBSERVATIONS ON A NEW GREGARINE. 275 


toxylin and eosin, the centrosome is differentiated into a 
faintly stained peripheral portion—the centrosphere—in the 
centre of which is a black centriole; this also shows in fig. 20 
stained in the same manner. 

In studying the various generations of daughter-nuclei 
several interesting points demand attention. ‘hey present 
an infinite variety as regards the arrangement of their 
chromatin, Except when actually drawn out into a spindle 
they are invariably vesicular in structure; and, in the great 
majority of cases, in the earlier stages at any rate, they 
contain a distinct karyosome. This is of interest in that in 
Kchinomera hispida, described by Schellack (17), where 
the karyosome invariably appears in the daughter-nuclei, 
its origin is referred to the unpaired chromosome of this 
form, which chromosome thus has a function allotted to it. 
In Stylorhynchus, which also shows this phenomenon, there 
is, however, no such unpaired chromosome (11). The fate of 
these daughter-karyosomes in Metamera schubergi is not 
certain. ‘lhe corresponding spindle figures do not show any 
traces of karyosome fragments in their neighbourhood. On 
the other hand, in such stages as shown in figs. 19 ¢ and d, 
where the nucleus is on the point of elongating into a spindle, 
the karyosome seems to be extruding part of its substance. 
If this is so, the process is one of immediate and complete 
solution, and not exactly parallel with the behaviour of the 
adult karyosome. It must be clearly understood that, as the 
figures show, a karyosome cannot be always with certainty 
identified in these daughter-nuclei. There are always present 
masses of chromatic substance of varying sizes, and their 
arrangement is at times such as to make the distinction 
impossible. In the daughter spindle-figures, as with the first 
division, there is again no definite chromosome formation. 
‘The chromatic elements are sometimes discernible as streaks 
und granules near the poles of the spindle; sometimes the 
deep black appearance of the spindle-fibres, alone present, 
suggests that these latter may be conveying chromatin in 
very minute particles. A constant feature of these young 


276 H. LYNDHURSY DUKE. 


spindles is a black mass of deeply staining matter at the 
extreme poles. In some early spindles shown in fig. 18 a, the 
earhest actual daughter spindle-stage to hand, this polar 
mass 18 seen as two adjacent granules or centrioles lying in a 
definite centrosphere showing radiations. In fig. 18 b these 
two granules are connected by a deeply staining link. This 
I interpret as the early division of the centrosome, occurring 
almost before the daughter-nuclei, which in the figs. 18 a and b 
are distinguishable as faint vesicles, are free from their parent 
spindle. In this connection it is of interest to note that the 
daughter-nuclei always appear provided with two centrosomes. 
I have not been able to discover any with a solitary centro- 
some. ‘This is in keeping with the above suggestion as to 
the early division of the centrosome in the history of each 
daughter-nucleus. As the daughter-nuclei become smaller 
their division-figures become less complicated, while the 
chromatin becomes arranged as a single mass rather than as 
separate particles. Some of the smallest spindles still show 
occasionally distinct chromatin elements near their poles, but 
the majority do not. ‘here appear to be no definite astral 
rays distal to the terminal mass of chromatic substance 
(figs. 21 d, 28, and 24 b). Finally all traces of spindle- 
formation disappear, and the nuclei are reduced to mere 
masses of chromatin about 1 to 1°5 mw in size.. These are 
arranged on the periphery of masses of protoplasm, after 
the fashion of a typical so-called Perlenstadium, and the 
protoplasm soon becomes mammillated round each nucleus 
with the formation of gametes (fig. 25). 

That part of the protoplasm which does not take part in 
the formation of the gametes—the Restk6Orper—contains 
a few nuclei which have not kept pace with the general 
division (fig. 25 6). These laggard nuclei are present here and 
there in all sections of the later daughter-divisions, and are 
noticeable in that: they are larger than their more numerous 
companions. Similar nuclei have been noticed by Léger and 
Duboseq in Hoplorhynchus (18). Scattered throughout 
the later cysts are also seen a number of round clear bodies 


SOME OBSERVATIONS ON A NEW GREGARINE. 277 


(fig. 25a) stained very faintly with iron-hematoxylin. ‘They are 
most obvious in cysts containing gametes or sporoblasts, and 
have not been seen in the earlier cysts, at any rate in the same 
form. ‘Their size varies considerably, and they appear to be 
products of the original karyosome which have lost most of 
their staining properties,and which have become more obvious 
owing to the splitting up of the protoplasm entailed in gamete 
formation. The majority are rather too large to be referred to 
the daughter-karyosomes. The main residue of the original 
karyosome is often to be found, deeply stained, in these later 
cysts. 

The gametes are very like those described for Lankes- 
teria ascidiz by Siedlecki (18), and show no signs of sexual 
differentiation (fig. 26). Considering the fact that there is 
at no time in the history of the encysted animals any difference 
in structure, and that the nuclear changes are practically co- 
incident, this isogamous type of gamete is what one would 
expect. Conjugation has not been observed in the living 
animal, owing to my studies being interrupted by my departure 
from Heidelberg. Fig. 27 shows, however, what is practically 
certain to be a zygote. The gametes measure about 3 u, and 
are roughly circular in outline. Their nuclei consist of small 
masses of chromatin with no definite vesicular structure. 
The zygote measured over 4°5 u, and contained two distinct 
nuclei. Several cysts were found containing sporoblasts, 
(figs. 28 to 33). These are ovoid bodies measuring 6 uw by 4 pu, 
and containing large vesicular nuclei. ‘These sporoblasts gradu- 
ally acquire a spore coat, and grow in size somewhat during 
the process (fig. 33), so that in a cyst of sporoblasts one or 
two may be detected with the outline of a formed spore (fig. 
34). The fully formed sporeis shown in fig. 35. The nuclear 
changes resulting in the formation of the sporozoites have not 
been made out, nor did I obtain a view of a free sporozoite. 
It was easily seen, however, in optical sections of the living 
spores that eight sporozoites were arranged peripherally 
around a granular mass of residual protoplasm. ‘The spores 
measure 9 « by 7 uw, and are navicelliform, provided at each 
end with a little peg-like projection (fig. 55). 


278 H. LYNDHURST DUKE. 


Discussion or Some Specran Pornrs ry tHe Liee-Cyce. 


In the description of the trophozoite mention has been 
made of the traces of further segmentation shown occasionally 
at the posterior end of the deutometrite in Metamera 
schubergi. ‘The presence of segmentation in some gre- 
garines, apart from the three fundamental divisions of the 
body, is a well-established fact, Léger (12) having described 
a form, Tewniocystis, where this phenomenon is so well 
marked as to make the animal resemble a small cestode. 
Porospora (18) also shows a segmentation, which, however, 
appears to be somewhat different in nature, as the animal is 
said to be capable of obliterating its segments merely by 
stretching itself out during movement. 

In Metamera schubergi the segmentation is always 
confined to the posterior end of the deutomerite, and is not 
constantly present. In their full development these posterior 
septa appear in every way as definite as those of the anterior 
part of the gregarine ; but in some animals, on the contrary, 
it requires the most careful focussing to demonstrate their 
existence. I am unable to explain the significance of these 
septa; whether they mark a certain period in the life-cycle or 
whether they are due to some form of plasmolysis I cannot 
say. ‘They are, however, sufficiently often present to form a 
striking feature of this gregarine. 

As regards the explanation of the phenomena shown in the 
division of the nucleus, it is difficult to discover anything of 
the nature of a precedent in the current description of this 
stage. The vacuoles described by Cuénot (6), Prowazek (16), 
and others, in close proximity to the sporont nucleus, or by 
Siedlecki (18) within the latter, have not been seen in 
Metamera schubergi. From the proximity of the com- 
mencing achromatic mass to the actively disintegrating 
karyosome, I suggest that this latter body supplies material 
—more or less, it is impossible to say—which will assist in 
the formation of the two daughter-nuclei. Another point, to 


SOME OBSERVATIONS ON A NEW GREGARINE. 279 


which attention has been frequently called, is the intense 
staining capacity shown by the achromatic mass, both at its 
first appearance and later in the fully formed spindles. This 
applies equally to Delafield’s hematoxylin and to Heidenhain’s 
method, which latter is known to stain plastin-substance 
darkly. Now the chromosome material, when first detected, 
is seen as streaks lying on the spindle-fibres near the poles ; 
or, when the fibres are seen in optical section, as a line of 
contiguous granules (figs. 12 and 13). No_ preparation 
showing an equatorial arrangement of the chromosomes was 
obtained, although, of course, this does not prove the non- 
existence of such a stage. Fig. 12 shows some of the 
chromatin streaks directly continuous with the well-marked 
terminal mass; and it is thus possible that this mass repre- 
sents a collection of chromatin which has been delivered by 
the spindle-fibres. I suggest, therefore, that throughout the 
division the spindle-fibres are carrying chromatin in a form 
unrecognisable as discrete particles, until it undergoes con- 
densation towards the poles of the figure. With the 
appearance of the vesicles the chromatin elements disappear 
from the spindle, leaving only the few scattered granules of 
figs. 12 and 15. These vesicles would thus appear to have 
been formed from the chromosomes of the earlier stages, 
and supposing them to be indeed daughter-nuclei, it is 
conceivable that they go on growing at the expense of 
chromatic substance still uncondensed in the spindle-fibres, 
until finally they become free as the first pair of daughter- 
nuclei. This theory would account for the staining properties 
of the spindle; and the absence, at the earliest stage of the 
division, of definite chromosomes. 

As regards the origin of the chromatin of the daughter- 
nuclei, there is nothing upon which to dogmatise. We have 
the fragmentation of the original chromatin masses, which 
proceeds until the resultant particles are indistinguishable, 
and we have the breaking-up of the karyosome, both of 
which might supply a source for the chromatin. ‘hat this 
chromatin is being in some way drawn up on to the spindle 


4 


280 H. LYNDHURST DUKE. 


from the débris of the old nucleus is obvious from figs. 14 
and 15. 

Siedlecki, in his work on the karyosome of Caryotropha 
(19), reviewing the réle played by this body in Coceidia, 
points out that while in some types the karyosome plays a 
purely vegetative part, in others it has definite responsibilities 
regarding the reproductive functions. ‘I'he latter appears to 
be the case in Metamera schubergi. If, as I believe 
to be the case, the daughter-nucle1 reform their karyosomes, 
may not these daughter-nuclei—which are, after the upheaval 
of the trophozoite nucleus during its first division, presumably 
sexual in nature—throw some light on the functions of the 
karyosome? If the latter be purely vegetative in function, 
why should it recur in the daughter-nuclei, which, with their 
two centrosomes, are plainly not in a vegetative condition ? 

In the face of the facts it is certainly a reasonable sugges- 
tion that the original karyosome consists of two elements at 
least. The one of these is thrown out at the first division of 
the nucleus, and is of no further use in the formation of the 
daughter-nuclei; the other is of vital importance in the 
propagation of the species, as realised in the sexual gametes. 
In the daughter-karyosomes only one of these components 
persists—i. e. that part essential to nuclear division ; the 
other part—for which, in the active reproductive processes 
now proceeding no need remains—is not represented. ‘Thus, 
in the daughter-spindles no karyosome remnants are seen. 
This is hardly the place for a discussion on the binuclearity 
hypotheses, so ably dealt with by Dobell (8), but the above- 
wnentioned differentiation of the karyosome constituents 1s 
sufficiently suggestive. On the one hand, the vegetative and 
reproductive elements of Goldschmidt’s theory may be seen 
in the original karyosome residue and the so-to-speak more 
intense daughter-karyosome respectively. On the other hand, 
one is equally justified in assuming that the karyosome 
residue merely represents elemeuts whose life is over and 
whose functions are exhausted, while the perpetuated 
remainder persists in the daughter-karyosomes, which are 


SOME OBSERVATIONS ON A NEW GREGARINE. 281 


thus thoroughly equipped for their part in the ceremony of 
division. 

It will be noticed that, except in fig. 15, where the vesicles 
attain their maximum development, there is no true striation 
shown distal to the polar aggregation; in other words, 
although the spindle-fibres are throughout very distinct, the 
centrosome element is not. This, again, suggests a bearing 
on the origin of the centrosome. On the one hand, as Dobell 
(8) points out, we have a binucleate condition held as the 
starting-point in the development of the centrosome ; on the 
other there are observers, such as R. Hertwig, who believe 
the centrosome to be a specialisation of the central spindle, so 
that the spindle in the Protozoa is equivalent to centrosome + 
spindle of the Metazoa. Without wishing to claim originality 
for the suggestion, I may say that the first division figures of 
Metamera schubergi have all along pointed forcibly to a 
most interesting lack of differentiation and specialisation 
between the various constituents. The chromatin is not 
marked off in the form of distinct chromosomes, nor are the 
centrosomes—assuming my interpretation of the figures to be 
correct—distinguishable as such. The three elements, chro- 
matin, spindle, and centrosome, act in concert in the formation 
of the first two daughter-nuclei, and it is difficult to say where 
one begins and the other ends. I suggest, therefore, that 
the evidence afforded by Metamera schubergi tends to 
support Siedlecki’s view, expressed in connection with his 
work on Caryotropha (8), that ‘“‘ we have in a protozoan cell 
. . . but a single and simple nuclear apparatus before 
us,’ and not a binuclear arrangement. 

In conclusion, with reference to the apparent isogamy 
shown by this gregarine, it will be noticed that we have 
another apparent exception to what Léger (18) deems the 
general rule in gregarines, i.e. anisogamy. In this connec- 
tion the recent work of Brasil (4) and Hoffmann (10) on 
Monocystis, which had previously been considered isoga- 
mous, is interesting. The work of the latter emphasises the 
futility of drawing conclusions from stained preparations. 


282 H. LYNDHURST DUKE. 


He showed that a very definite anisogamy was visible in the 
living cysts, which, however, became much less marked in 
the process of fixing and staining. This may be so in 
Metamera schubergi, but, considering isogamy as the 
more primitive condition, it is possible that this gregarine, 
whose first spindle suggests a phase in the evolution of 
karyokinesis, may also exhibit true isogamy. 

I hope in the spring to renew my acquaintance with this 
species, and to be able to complete its life-history. 


DiaGNosis OF MErAMERA SCHUBERGI N.G., N.SP. 


A cephaline gregarine belonging to the family Dactylo- 
phoride (Léger).! Trophozoite ca. 150 u by 454. Epimerite 
subconical, with apex excentrically placed, and surrounded 
by numerous branched, digitiform appendages. The deuto- 
merite sometimes (not always) shows a secondary septation 
into one to three segments in the region posterior to the 
nucleus. Conjugation isogamous, no sexual differentiation 
being observable at any stage in the life-cycle. Cyst dehise- 
ing by simple rupture. Spores navicelliform, containing 
eight sporozoites, and measuring 9 u by 7 mu. 

Hosts: Glossosiphonia complanata (Heidelberg and 
Cambridge) and Hemiclepsis marginata (Heidelberg). 

Guy's HospiratL, 


Lonpon, S.E., 
February, 1910. 


LITERATURE. 


1. Berndt, A.—* Beitrag zur Kenntnis der im Darme der Larve von 
Tenebrio molitor lebenden Gregarinen,” *‘ Arch. Protistenk.,’ 
Bd. i, 1902. 

2. Botsius, H.—‘ Ann. Soe. Bruxelles,’ vol. xix, 1895. 


3. ——— “Un parasite de la Glossiphonia sexoculata,” ‘Mem. 
Acad. Lineei,’ vol. xi, 1896. 


1 See Minchin (15). 


SOME OBSERVATIONS ON A NEW GREGARINE. 283 


4. Brasil, L.— Nouvelles recherches sur la réproduction des Gréga- 
rines monocystidées,” ‘ Arch. Zool. Expér.,’ iv, No. 4, 1905. 

5. Castle, W. E—*‘‘Some North American Freshwater Rhynchob- 
dellidz and their Parasites—VI Parasites,” ‘Bull. Mus. Comp. 
Zool.,’ Harvard, vol. xxxvi, 1900. 


& 


Cuénot, L.—* Recherches sur l’évolution des Grégarines,” ‘ Arch. 
Bioles Maxine LO OIe 
Comes, L.—‘* Untersuchungen iiber den Chromidialapparat der 
Gregarinen,” ‘ Arch. Protistenk.,’ Bd. x, 1907. 
Dobell, C. C.—‘* Chromidia and the Binuclearity Hypotheses,” 
‘Quart. Journ. Micr. Sci.,’ vol. 53, 1909. 
9. Doflein, F—‘ Lehrbuch der Protozoenkunde,’ Jena, 1909. 
10. Hoffmann, R.—* Uber Fortpflanzungserscheinungen von Mono- 
vystideen des Lumbricus agricola,” ‘ Arch. Protistenk.,’ Bd. 
xiii, 1908. 
11. Léger, L.—‘‘ La réproduction sexuée chez les Stylorhynchus,” 
‘Arch. Protistenk.,’ Bd. viii, 1904. 


= 


ed 


12. “Etude sur Teniocystis mira Léger, Grégarine meta- 
mérique,” ‘ Arch. Protistenk.,’ Bd. vii, 1906. 
13. and Duboseq, O.—* Etudes sur la sexualité chez les Gréga- 


rines,” ‘ Arch. Protistenk.,’ Bd. xvii, 1909. 


14. Lihe, M—‘“Bau und Entwicklung der Gregarinen,” ‘Arch. 
Protistenk.,’ Bd. iv, 1904. 


15. Minchin, E. A.—‘‘Sporozoa,” ‘Lankester’s Treatise on Zoology, 
pt. 1, fase. 2, 1903. 

16. Prowazek, S.—* Zur Entwicklung der Gregarinen,” ‘ Arch. Protis- 
tenk.,’ Bd. i, 1902. 

17. Schellack, C.—* Uber die Entwicklung und Fortpflanzung von 
Echinomera hispida (A. Schn.),” ‘ Arch. Protistenk.,’ Bd. ix 
1907. 

18. Siedlecki, M.—‘* Uber die geschlechtliche Vermehrung der Mono- 
cystis ascidiz (R. Lank.),” ‘ Bull. Internat. Acad. Sci. Cracow,’ 
1899. 

“Uber die Bedeutung des Karyosoms,” ‘ Bull. Internat. 

Acad. Sci., Cracow,’ 1905. 


’ 


? 


19. 


284, H. LYNDHURST DUKE. 


EXPLANATION OF PLATES 15 ann 16, 


Illustrating Mr. H. Lyndhurst Duke’s paper on “Some 
Observations on a New Gregarine (Metamera schu- 
bergi nov. gen., nov. spec.).” 


PLATE 15. 

[Figs. 1 and 2 were drawn from living animal, figs. 3 and 4 from 
preparations fixed with alcohol and acetic acid and stained with 
Schubereg’s modification of Mayer’s acid carmine. Figs. 5, 6, and 16 
are diagrammatic. Figs. 7-18 were fixed with Gilson’s fluid and stained 
with Heidenhain’s iron-hematoxylin. All these figures were drawn 
with Zeiss oc. 6, obj. 2mm. apochromatic. Figs. 9, 10, 11, and 17, 18 
are to scale at magnification of 2000. Figs. 12, 15, 14 are drawn on 
a slightly smaller scale. | 


Fie. 1.—Living sporont ex sressed through anus of leech. 
5 5 > 

Fie. 2.—Same sporont as Fie. ih showing bub] vle-formation. 
5 5 D5 


Fig. 3.—Cephalont with epimerite embedded in fragment of host- 


Fig. 4.—Showing optical section of epimerite. 

Fig. 5.—Diagram of structure of epimerite, ete. 

Fig. 6.—Diagram of sporont with cuticular pad on epimerite. 

Fig. 7.—Nucleus of trophozoite. 

Fig. 8—Nucleus showing fragmentation of chromatin masses and 
extrusion process of karyosome. 

Fig. 9.—Sporont nucleus showing earliest appearance of the “ achro- 
matic mass,” with fragmentation of the karyosome. 

Figs. 10 and 11.—Successive sections of another nucleus showing 
slightly later stage than fig. 9. 

These three figures (9, 10 and 11) are drawn from same cyst. 

Fig. 12.—First division of sporont nucleus showing at (a) the ring 
arrangement beginning at the pole; also the streaks of chromatin and 
the spindle-fibres in optical section. The two poles are respectively at 
the extreme upper and lower focus. One of the chromatin streaks is 
seen running into the polar aggregation. 

Fig. 13.—An early cyst, containing two associated individuals, with 
remains of epimerites seen at the centre. Nuclei at stage of first 
division. In upper animal the polar aggregation and the chromatin 
streaks are very marked. (Combined from two successive sections.) 


SOME OBSERVATIONS ON A NEW GREGARINE. 285 


Fig. 14.—First division of the sporont nucleus at a somewhat later 
stage than figs. 12 and 15. Shows polar vesicles more distinct. Also 
the distinct fibres running down into neighbourhood of original nucleus 
and karyosome. 


Fig. 15.—First division of sporont nucleus at a later stage than fig. 14. 
Vesicles fully formed and fibres running down towards karyosome. 
The vesicles here shown were 6, apart, lying respectively at top and 
bottom focus. 

Fig. 16.—Diagram of first spindle just before final separation of first 
two daughter-nuclei. 

Fig. 17, a, b and c.—Earliest stage of daughter-nuclei, eight or nine 
in cyst. 

a. Shows centrosomes connected by a thick band. 
b. Shows chromatin bunched as an early spindle figure. 
c. Shows karyosome. 

All from same cyst. 


Fig. 18.—Somewhat later daughter-nuclei at end of division. 

a. Shows two centrioles at each pole; also one daughter-vesicle. 
(The section has not passed through the left vesicle.) 

b. Shows division of the centriole with poorly developed daughter- 
vesicle. (The vesicle at the right end of the figure lies outside 
the plane of this section, and is therefore not seen.) 

c. Shows a separated daughter-vesicle. 

All from same cyst. 


PLATE 16. 


[Figs. 21-34 and 36 were fixed with Gilson’s fluid and stained with 
Heidenhain’s iron-hematoxylin. Figs. 19 and 20 were stained with 
Delafield’s hzematoxylin and eosin. Figs. 19-34 were drawn at 
magnification of 2000. Fig. 35 is not to scale, being relatively too 
large. | 


Fig. 19 (a-e)—From same cyst. Somewhat later daughter-nuclet. 
All show karyosomes. cand d show early stage of spindles, and the 
karyosomes in a state of activity. 

Fig. 20.—Showing differentiation of centrosome into centriole, and 
centrosphere in a daughter-nucleus of same stage as fig. 19. 

Fig. 21.—Similar daughter-nuclei showing karyosomes: also corre- 
sponding spindle. 

Figs. 22 and 23.—Later stages of daughter-nuclei, mostly showing 
karyosomes; also corresponding spindles. 

VOL. 00, PART 2.—NEW SERIES. 19 


286 H. LYNDHURST DUKE. 


Fig, 24.—Smaller daughter-nuclei and spindles. 

Fig. 25.—Shows the Perlenstadium, with «a single free gamete. 
Notice the clear karyosome remnants (a), and the residual nuclei (5). 

Fig. 26.—Gametes. 

Fig. 27.—A zygote with two unfused nuclei. 

Figs. 28-32.—Sporoblasts. Figs. 29 and 31 show these in transverse 
section. 

Fig. 33.—Shows a sporoblast assuming shape of spore. 

Fig. 34.—Shows a spore coat in process of developing. 

Fig. 35.—Fully formed spore, with sporozoites in optical section, 

Fig. 36.—Shows the karyosome in the act of extruding some of its 
substance. 

Fig. 37.—Diagram to show method of apposition of associating 
sporonts in a cyst. 


u ' ™ 
e co , 
vs WG «ot 
- . . 
y - ee Zl 
¥ bd * 
i 
° ‘ 
S = 


Quart. Iounn.MiorSci VAS NETS. 


17a. 
17e. Hath, Lith? Tendon 


Quant. Lurn Maier: ek uot. oo NV, ELL IG 


Huth Lith! Londen 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 


287 


On the Anatomy of Histriobdella Homari.’ 


| CO oes 


joe) 


co On D 


10. 


By 
Cresswell Shearer, M.A., 
Trinity College, Cambridge. 


With Plates 17-20 and 5 Text-figures. 


ConrTENTS. 


. Introduction, Material, and Methods 
. Review of Literature and Remarks on Habits 
. Description of the Nephridia 


A. First Nephridium 
B. Second Nephridium 
c. Third Nephridium. 
p. Fourth Nephridium 


. Body-cavity and Nephridia 
. Muscular System 


A. Longitudinal Muscles 
B. Special Muscles of the Generative Serine 
c. Oblique Muscles 


. Digestive System 

. Nervous System 

. Sense-Organs 

. Reproductive sean. 


A. In the Male 
B. In the Female 
Conclusion and Summary 


PAGE 


288 
291 
300 
303 
303 
304. 
304 
305 
308 
308 
309 
311 
314. 
321 
327 
328 
328 
334. 
346 


1 T have tothank the Director and members of the staff of the Marine 
Biological Association of Plymouth for their kind attention and interest 
in my work while at Plymouth. 


288 ORESSWELL SHEARER. 


1. Inrropuction, Mareriat, AND Merruops. 


Our knowledge of the anatomy of Histriobdella is based 
on the papers of Van Beneden (1858), Foettinger (1884), and 
Haswell (1900). Of these, Foettinger’s account is the most 
extensive, while Haswell’s paper is perhaps the most valuable. 
Both accounts contain a more or less detailed description of 
the internal structure and organisation of the adult. Several 
years ago I described the presence of solenocytes in con- 
nection with the nephridia of Dinophilus. This dis- 
covery rendered it probable that these peculiar structures 
would also be found in Histriobdella, with which 
Dinophilus shows many relationships. Moreover, the 
different description of the nephridial system given by 
Haswell in Stratiodrilus from that of Foettinger for 
Histriobdella called for a re-investigation of these organs. 
For these reasons the present work was begun. I was soon 
led to undertake a detailed examination of the animal. It is 
some twenty-five years since the publication of Foettinger’s 
paper, and during this interval the European species of 
Histriobdella has received no further attention. In the 
following account I have endeavoured to clear up Foettinger’s 
description of several of the organs. I have had the advan- 
tage of having made use of the methyl-blue method of intra 
vitam impregnation, which has proved most valuable. With 
its use I have experienced no difficulty in determining the 
number of the nephridia and their relationship to the segments, 
and to make out new details in their structure quite impos- 
sible from ordinary sections of fixed material. 

Good methyl-blue! preparations of the nephridia can be 
obtained by placing the lobsters bearing the parasites in small 
tanks of sea-water, to which sufficient blue has been added to 
colour the water a light shade. It is necessary for the 
animals to remain in the blue two or three days before it 
appears in the nephridia. As the blue is rapidly absorbed 
by the living tissues of the lobster, an additional quantity has 


1 This is * soluble blue,” and not methylene blue. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 289 


to be added to the water from time to time. With good air- 
circulation anda little attention, a medium-sized lobster can 
be kept alive for several weeks in a tank of four or five litres 
capacity without change of water. 

At the end of the second day the blue will have collected 
in dark granules on the walls of the nephridial canals, so as 
to outline these clearly. By this time it has been discharged 
from the nervous system and the sensory cells of the 
epidermis. About the bases of the legs of the head it shows 
a tendency to remain some time after it has disappeared from 
the brain. It is retained alone by the nephridia on the third 
day. Here it collects in dense masses on the courses and 
openings of the canals. 

In the study of these methyl-blue breeaeioms I fave 
made use of long, thin cover-slips, such as:are used in pre- 
paring large serial sections instead of ordinary slides, on 
which to mount my preparations. The use of a thin cover- 
slip used as a slide allows of the preparation being examined 
from each surface, as desired, under an oil-immersion lens. 
It is thus possible to trace a nephridium first. on one side of 
the preparation, and then turn the slide over and trace it 
further on the other surface. 

Histriobdella is a somewhat difficult animal to fix. The 
only reagent that has given uniform results is a saturated 
sublimate solution, with 5 per cent. acetic, used boiling hot. 
Hermann’s solution and Flemming and the osmic: acid 
mixtures give very irregular results, and are. not to be 
depended on for their action. One lot of material will be 
excellent, while the next, fixed with the same solutions and 
under the same conditions, are useless. Picro-acetic and 
Bouin’s solutions, used hot, give good results, but not as good 
as material fixed with sublimate-acetic. Picro-sulphuric was 
used for preparations to be studied whole, on account of the 
excellent preservation it gives of the external form. As 
stains, the following have given satisfaction: Heemacalcium 
and Benda’s iron-hematoxylin, paracarmine, lithium-carmine 
follewed by Lyon’s blue for eggs. 


290 CRESSWELL SHEARER. 


The nephridial canals are remarkably difficult to recognise 
in sections on account of the retraction they invariably 
undergo during fixing. It is impossible to trace them with 
any degree of certainty through consecutive sections. For 
this reason I have relied mainly in my investigation of the 
nephridia on methy]l-blue impregnation preparations of living 
material. The figures accompanying the present paper there- 
fore represent the appearance of the nephridia in living 
material. It is necessary to use the highest powers of the 
microscope to determine the structure of the nephridia, and 
even then the eye requires considerable practice and training 
to distinguish the motion of their cilia. It is difficult to 
convey any idea of the extreme delicacy and minuteness of 
these structures. The necessity of being compelled to use 
immersion-lenses for their study excludes the use of any of 
the ordinary dark ground systems of illumination. Doubtless 
these would offer an excellent means of investigating struc- 
tures of this nature in an animal so transparent as His- 
triobdella, if they could be used successfully with the 
immersion-lens. 

Of great service in the study of the methyl-blue pre- 
parations is, I have found, the use of a number of sodium 
glass screens of different shades, such as are used in ortho- 
chromatic photography to vary the exposure from five to 
fifteen times. 

To obtain a uniformly constant light I have used an 
ordinary Welsbach gas lamp, with standard screens. ‘This 
gives a light much superior.to that of ordinary daylight 
in bringing out the finer structure of the nephridial canals. 

For sections I have used the ordinary paraffin and the 
paraffin celloidin method. The sections were cut of the 
uniform thickness of 7.4. Inthe reconstructions of the nervous 
system shown in figures I have used a method which is in 
part a modification! of that described by Woodworth (‘ Zeit. 
f. wiss. Mik.,’ xiv, 1897, p. 15). Each section, of which there 


1 This I owe to my friend, Mr. E. W. Nelson, of the Marine Biological 
Association, Plymouth. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 291 


were about eighty, was first outlined on paper with the aid of 
a camera lucida, and the nervous system carefully marked 
in. Each of these drawings was then measured transversely 
from side to side, and the measurements plotted out on milli- 
metre paper, allowance being made for the magnification 
between sections. ‘The nervous system was also measured, 
and likewise put in, all the distances being doubled to give 
an axial line. ‘lhe ends of the plotted points were then joined 
up, and an outline of the external form and the nervous 
system obtained. ‘The figures were then reduced to their 
present size, and at the same time transferred to ordinary 
drawing-paper by means of an eidograph. In the sagittal 
section shown in fig. 15 the dorso-ventral diameter was taken 
instead of the transverse. By this means the relationship of 
the ganglia to the segments can be accurately determined in 
a way that would be impossible with the ordinary recon- 
struction methods (figs. 15, 21, 28). 


_ 2. HisroricaL Review and GENERAL Remarks on Hapsirs, 
ETC. 

Histriobdella was discovered and briefly described by 
J. P. van Beneden (1) in 1853. He found it as a parasite on 
the eggs of some lobsters obtained from Ostend. He con- 
sidered it a larval Serpulid, and placed it among the Poly- 
chete. Subsequently, in 1858, he (2) pointed out that it was 
an adult form. From its peculiar structure he remarked that 
it could not be easily classed with any known group of animals, 
although some of its features he thought were such as to 
place it among the leeches. He gave a more or less detailed 
description of both sexes, and’figured the eggs and immature 
young. 

To Foettinger (8) we owe the most extensive account of this 
animal. He describes the nervous system, nephridia, repro- 
ductive organs, and, in fact, was the first to give a detailed 
account of its anatomy based on sections. He supported the 
conclusions of Edouard van Beneden that it was an Archi- 
annelid, placing it near Polygordius, but separate from it, 


292 CRESSWELL SHEARER. 


in the family Histriodrilides. In his opinion many of its 
characters show its inferiority in organisation to Proto- 
drilus. Among these the absence of any trace of the 
circulatory system, the feeble internal segmentation, marked 
by the complete absence of dissepiments and the small number 
of segments. On the other hand, the presence of well- 
developed ganglia points towards a higher organisation than 
that possessed by any known Archiannelid. Again, the com- 
plicated sexual apparatus of the male is different from any- 
thing at present found in this class. ‘lhe presence also of 
chitinous jaws with striated muscles and the anterior and 
posterior feet he considered as distinguishing it as a type 
superior to Polygordius. 

More recently Haswell (13) has obtained, as already men- 
tioned, a freshwater species from the branchial chamber of a 
Tasmanian crawfish. In the possession of cirri it differs 
externally slightly from Histriobdella. Haswell pointed 
out, among other new features, that the lateral organs which 
Foettinger considered penes are in reality organs that func- 
tion as claspers, while the penis, as in Dinophilus, is a 
median unpaired structure. While Foettinger described the 
seminal vesicles he was unable to trace their ducts to acommon 
receptaculum as Haswell has done in Stratiodrilus. The 
nervous system of Stratiodrilus seems to be different how- 
ever from that of Histriobdella, in being more highly 
differentiated. In Histriobdella the ventral nerve-cord is 
still in complete continuity with the epidermal layer, while 
in Stratiodrilus it is situated much deeper. ‘This difference 
may be in great part due to the close union of the epidermal 
and sub-epidermal tissues in contrast to those of Stratio- 
drilus. A more important difference is the separation shown 
by the two component halves of the ventral nerve-cord in 
Histriobdella, and the somewhat different position and 
number of the ganglia. 

The greatest difference, however, is shown in the excretory 
system. It is impossible to reduce this to a common type. 
In Stratiodrilus the crossing and branching of the canals 


ON THE ANATOMY OF HISTRIOBDELILA HOMARI. 293 


in the anterior region, and their course in some instances 
through more than one segment, seems to preclude any com- 
parison with Histriobdella.' Again, in Stratiodrilus 
the interior feet are retractile, and can be completely drawn 
into the head. ‘his is not the case in Histriobdella, 
where the distal joint alone is retractile. The main mass of 
the foot is incapable of retraction, even under the action of 
strong reagents. 
Histriobdella was found by van Beneden and Foettinger 
on the eggs of the European lobster, and was considered by 
them a parasite on these alone. It is, however, like Stratio- 
drilus, normally an inhabitant of the branchial chamber and 
gills. It passes to the eggs of the female from the gill- 
chamber when these happen to be present,” returning to the 
same situation when the eggs are hatched and the egg-mem- 
branes shed. In the branchial chamber it is quite difficult 
to detect at first,on account of its almost colourless condition 
and the fact that in this situation it does not show the excitable 
movements exhibited while on the eggs, but crawls slowly, 
keeping close to the mucous membrane. Examination of the 
branchial surtace of the carapace, however, once the eye has 
become accustomed to distinguishing them, seldom fails to 
show their presence in this situation in either of the sexes. 
They prefer the carapace to the gill surface, as it affords 
a better footing, and the long hairs under which they 
move prevent their being readily brushed off. ‘To the bases 
of these they attach their eggs in great numbers, especially 
towards the margin of the carapace, where the hairs are long 
and numerous. Comparison of the parasites from the “ berry ” 
with those from the chamber shows no difference between 
them, except that the jaws of the parasites from the chamber 


1 Professor Haswell informs me that since the publication of his 
account of Stratiodrilus he has re-examined the nephridia and has 
re-confirmed his statements regarding them. 

* According to Herrick this takes place once in two years. “The 
Reproductive Period in the Lobster,” ‘ Bull. of U.S. Fish Commission,’ 
vol. xxi, 1901, p. 161. 


294, CRESSWELL SHEARER. 


seem a little better developed than those of the “berry.” 
The parasites are evidently able to migrate rapidly from one 
situation to the other. On female lobsters whose eggs are 
about to hatch, many of them have already migrated to the 
gill-chamber. A certain number, however, are always to be 
found on the old egg-membranes, although the eggs have 
been hatched and the membranes are much discoloured with 
age, showing that the breeding period had passed some time. 
I have taken females in this condition, and placed thein in 
tanks with air circulation and kept them under observation. 
In the course of several. weeks the membranes drop off, but 
no parasites are found about the tank, showing that they have 
all taken refuge in the gill-chamber. In the gill-chamber and 
on the eggs both sexes are present in equal numbers. When 
the lobster ova are well advanced and about to: hatch, the 
male Histriobdellid would seem to preponderate over 
the female. On the ova the immature young are found in 
greater numbers than in the gill-chamber. 

Frequently a large female can be seen carrying a male 
attached to its back by means of its claspers. ‘These would 
seem to throw out some sticky secretion, for once the male 
has taken hold of the female it is unable readily to let go, and 
gets carried about by the female although it makes violent 
efforts to free itself. 

In the gill-chamber, as on the eggs, the parasites show the 
same tendency: to collect in small groups, huddling close 
together and crawling over and over one another. When 
disturbed they separate, to re-collect shortly in another group. 
Why they do this is not obvious, as the individuals are some- 
times all males or immature young, in which the sexual organs 
have not yet developed. ‘This habit of collecting in groups 
therefore can hardly be for the purpose of the impregnation 
of the females. 

I have examined a considerable number of ‘“ berried” 
crabs and rock lobsters, both at Plymouth and Naples, 
without finding Histriobdella. They would seem to be 
exclusively confined to the lobster. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 295 


Nothing is known regarding geographical distribution 
beyond the fact that Histriobdella is common on the 
lobster of the Channel region. J. P. van Beneden (2), as 
already mentioned, obtained it at Ostend. He also states in 
his second paper that he had observed it on the lobster of the 
Norwegian coast. I have been unable to find it on the 
lobster at Naples. ‘My observations were, however, limited 
by the rarity of this animal in the Bay of Naples. I 
only had the opportunity of examining a few adults. So 
far it has not been reported as occurring on the American 
lobster.!. It is remarkable that an animal of such peculiar 
structure should be represented in Europe bya single species, 
while its nearest ally should be found in fresh-water streams 
of Tasmania. 

Little is also known of its life-history and habits. If a 
small mass of lobster ova with the parasites is placed in a 
watch-glass of sea-water, it will be noticed that they never 
crawl on any foreign body brought in contact with them. 
When left to themselves they collect in groups, twisting their 
bodies together, and remaining quiet for long periods. On 
being disturbed they show singular excitement, twisting 
themselves violently and throwing their heads rapidly from 
side to side, all the time remaining firmly attached by their 
powerful hind legs. From time to time they can be seen to 
bite one another with their strong jaws. 

While the parasite can be obtained from the branchial 
chamber or “‘ berry” of almost any lobster on the coast of 
England, the manner in which it gains access and passes 
from one host to another has not been determined. Like 
most parasites, it has limited powers of locomotion, being 
unable to swim, and crawling very slowly. It has no larval 
stage that might assist in its distribution. The eggs are 
attached in capsules to the lobster ova, and the young 
undergo their entire development within this capsule, emerg- 
ing in almost the adult condition. ‘There can be no larval 


1-Professor Herrick informs me that he has never found it on the 
American lobster. 


296 CRESSWELL SHEARER. 


stage during which it can live, either internally or externally, 
on some other host. 

The parasites are able, however, to pass from one host to 
another without apparent difficulty. ‘his can be readily 
demonstrated by placing a lobster in a solution of neutral 
rose in sea-water until the parasites it bears are stained, and 
then placing it in company with a number of normal unstained 
lobsters. In the course of a day many of the stained parasites 
will be found to have gained access to the normal lobsters, 
while many unstained parasites will be found on the stained 
lobster. This takes place readily in large tanks where the 
animals have room to keep well apart. How this passage is 
accomplished under these conditions I have been unable to 
observe, as the female lobster is very shy when ‘in berry,” 
and unsociable, strictly avoiding its mates and companions. 

Both Foettinger and Haswell have drawn attention to the 
remarkable chitinous jaws with which Histriobdella is 
furnished. Haswell has made a careful study of these in 
Stratiodrilus, and has shown how the movements of the 
component parts of the mechanism are brought about. In 
Histriobdella the jaws are almost identical, as far as I can 
determine, with those of Stratiodriius. Foettinger repre- 
sents them as furnished with many more teeth than I can find 
to be the case. Their use is not known, as neither Foettinger 
nor Haswell have made any observations on this head. 
Unfortunately the intestinal contents are reduced to such a 
fine amorphous condition as to afford no evidence as to the 
animal’s food. It is probable that the parasites feed on 
small alge to a certain extent, as the intestinal contents 
are usually of a greenish tint. Diatoms occasionally are 
present, and in some instances would seem to compose the 
greater portion of the food. This is so in the case of the 
parasites living on the “berry.” In the parasites of the 
gill-chamber they seem absent, and the intestinal contents 
consist of a fine brownish mass, among which reddish granules 
are seen. It is certain that the jaws are not used for tearing 
the membranes of the lobster’s ova as has been supposed, 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 297 


When the animals are excited they have a rapid way of © 
opening and closing the jaw teeth, but they are never seen 
to use them to tear open the lobster ova. When suddenly 
disturbed they sometimes secure themselves by means of 
their jaws to the egg membranes. They possess the power 
of protruding the jaw apparatus considerably beyond the 
mouth orifice, and in crawling they are sometimes seen to 
seize some object in front of them by protruding their jaws 
in this manner, after the fashion of many Polychets. 

Histriobdella is remarkably sensitive to any changes in 
the sea-water. The circulation of water through the branchial 
chamber of the lobster insures their receiving a continual 
change of water under normal conditions. Likewise, on the 
“berry ” the water is kept in constant circulation round them 
by the ceaseless motion of the lobster’s swimmerets. With 
any slight impurity of the water they fall off their hosts, and 
are found on the bottom.of the tank in a half paralysed 
condition. They are quickly killed by the addition of small 
quantities of fresh water, and die very readily when exposed 
to bright light. This is of interest when it is recalled that 
Stratiodrilus is found in fresh water. 

Fertilisation takes place internally. ‘The male drives its 
penis through any portion of the body-wall of the female. In 
one instance I saw a male drive its penis into the head and 
discharge a considerable quantity of spermatozoa. These could 
be seen under the microscope working their way down into the 
generative segment. In many cases the males fertilise young 
females without eggs, and the spermatozoa apparently remain 
in the body till the ova develop. Many females can be 
observed carrying spermatozoa but no eggs. 

The female exercises apparently little choice in the selec- 
tion of a site in which to deposit her eggs. On the “ berry ” 
these are usually attached to the membranes of the lobster 
ova, while in the branchial chamber the carapace side is the 
one selected. They are usually deposited in groups of four or 
five, and this would indicate that these are all deposited at one 
time by the female. The eggs are all of one size, and it is 


298 , CRESSWELL SHEARER. 


impossible to distinguish the sex of the immature young. 
They develop at once on being laid, showing that they have 
already been fertilised within the body. Even when the 
female is isolated in pasteurised sea-water the eggs develop 
immediately on being deposited; no sperm can therefore 
fertilise them in the sea-water. 

The egg-laying is done at night, as every morning fresh 
capsules are to be seen adherent to the coatings of the lobster 
“berry.” ‘The eggs within these are always in segmentation 
or gastrulation stages. ‘hey are laid in great numbers, so 
that it is easily possible to obtain all the stages of develop- 
ment up to the time the young worm leaves the capsule. 
Development is direct and would seem to be rapid, for by the 
end of the fortieth hour the young are fully formed and 
appear.ready to quit the capsule. On leaving this they move 
about the gill-chamber or pass immediately to the lobster ova, 
where they soon attain maturity. They are readily distin- 
guishable at this time by their small size and the undeveloped 
condition of the generative segment. The young of both 
sexes resemble the female in shape. Van Beneden (2) has 
figured a number of the young stages, and Haswell (18) 
mentions that he has obtained a number of the stages in the 
development of Stratiodrilus. 

Regarding the nephridia, Foettinger (8) stated that in the 
male there were five pairs of these organs, while in the female 
there were four. Each nephridium consisted of an intra- 
cellular tube running backwards on the border of the longi- 
tudinal muscle-strands. They turn in sharply towards the 
median line, to terminate ventrally, on the surface of the 
succeeding segment to that in which they arise, in a small 
pore. He could observe no internal openings or funnels, 
Their heads at their point of origin are on the dorsal surface; 
since they terminate on the ventral surface they run back- 
wards in an oblique plane between the dorsal and ventral 
muscle-bands. The first pair arise in the neck segment close 
to the head, and run backwards to terminate on the ventral 
surface of the second segment. The second pair arise in the 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 299 


posterior portion of this segment, and terminate in a similar 
manner in the third segment, ‘he third pair arise in the 
third segment to terminate in the fourth. In the female the 
third and fourth pairs overlap, while in the male the fourth 
pair arises much farther back between the posterior portion 
of the fourth and the anterior border of the fifth segment. 

In Stratiodrilus, onthe contrary, according to Haswell, 
the nephridial system would seem to extend into the head 
region. Hach nephridium at its anterior end divides into an 
external and an internal branch. The external branch runs 
forward into the head, while the internal crosses over to join 
the internal branch of the opposite side. From the fact that 
the motion of the cilia of this pair of organs is always from 
behind forward, their openings are probably in the head. 
The other nephridia are not branched. ‘In the female an 
apparently continuous line of cilia is traceable backwards on 
each side from the Lead canals to a point some little distance 
behind the second cirrus, where a canal is clearly traceable, 
which, after bending round in a loop, opens on-the exterior 
on the ventral side. But as the direction of the movement of 
the cilia is from before backwards in the posterior part of this 
line, it would appear probable that there are two pairs of 
canals in this anterior region in the female. In the male, on 
the other hand, there is no such evidence of division, the pair 
of nephridia which branch in the head being traceable back- 
wards, without change in the direction of the cilia, nearly as 
far as the bases of the second cirri, at which point they bend 
in and terminate in the ccelom in the middle line.” In the 
fourth segment, according to Haswell, it is probable that the 
oviducis represent the nephridia, while in the male they are 
represented by the vasa deferentia. In both sexes, in the 
fifth segment there is a pair of organs (beginning in a loop in 
the male) which run back in the caudal region to terminate 
near the anus. ‘I'he direction of the movement of the cilia in 
these organs is from behind forwards. ‘Thus, in the male 
there are three pairs of organs, while in the female there are 
four ; so that the nephridia do not partake of the metamerism 


300 GRESSWELL SHEARER. 


of the body, Stratiodrilus having the same number of 
segments as Histriobdella. In no part of the canals were 
ciliary flames observed. 


3. GENERAL DESCRIPTION OF THE NEPHRIDIA. 


From the inspection of figs. 1, 7, and 9, it will be seen that 
the nephridia have much the same positions as those assigned 
them by Foettinger (8). Apparently in the male the fourth 
pair, figured by him in the genital region, have no existence. 
Like the female, the male has only four pairs of organs. It 
will be seen that they are the narrow, delicate, §-shaped 
structures he has described (figs. 4, 5, 6, 10, 14), running in 
the mesodermic tissue of the body-wall. Their position in 
sections can be seen in figs. 37 and 43. Each organ takes its 
origin in a small space—a prolongation or part of the 
general blastoccelic cavity that surrounds the gut—in the 
anterior portion of the segment to which it properly belongs, 
and runs back to terminate on the ventral surface of the 
following segment near the median line. It arises in a knob- 
like process that projects slightly into the space. This 
process is thick-walled, and sometimes contains refractive 
granules. It is shown in fig. 14. Its structure is difficult to 
determine, and especially the relationship it bears to the 
space. What I take to be the real head of the organ is 
shown in section in fig. 42. Here the space into which it 
projects is surrounded by darkly staining nuclei. These are 
not seen in the living condition. It bears no cells that have 
any resemblance to solenocytes, and these structures would 
seem to be entirely absent in Histriobdella. In a number 
of preparations it was obvious that the internal ends of the 
canals were closed, and that they did not open into the space 
into which they project. 

The main portion of the nephridial canal is a thin-walled 
intra-cellular tube, the anterior end of which contains a few 
refractive granules and nuclei. It runs directly backwards 
in an oblique plane, and is much longer than the terminal 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 301 


portion. It reaches its greatest length in the case of the 
second nephridium (fig. 14). Frequently the lumen can be 
seen to be enlarged into small spaces or lacune, These would 
seem similar to the spaces I have described on the nephridial 
canals of Dinophilus. A number of these are usually seen 
on the course of the second organ (fig. 6), One large one is 
often found on the posterior part of the third. From the 
terminal portion of the canal they seem to be absent. The 
nephridial flagella pass through their centre while their walls 
themselves are unciliated. It is possible they are due to the 
somewhat abnormal conditions under which the parasites are 
kept in the process of their impregnation with methyl-blue, 
as I have never been able to observe their presence in the 
unstained living Histriobdella, although something like 
their appearance can be detected in sections. 

The terminal portion of the nephridial canal turns towards 
the median line, close to which it ends in a darkly staining 
pore (fig. 14). About this the blue usually collects in thick 
granules, which can sometimes be seen vibrating to and fro 
in the fluid escaping from the canal. The lumen of this 
terminal portion is greatly restricted in size. 

Throughout the length of the nephridial canal the ciliary 
action of the flagella in their interior can be plainly observed 
during life. ‘The movement of this is always in the one 
direction—from before backwards—and I have never observed 
any reversal of this motion as Haswell has described in 
Stratiodrilus. Despite numerous observations, I have 
been unable to determine whether the flagella are derived 
from the walls of the canal or from the knob-like head of the 
organ. In some preparations they seemed derived from the 
wall, in others they seemed derived from the nephridial 
heads. In sections they would seem to be derived from the 
walls. 

In no case can the canals be seen dividing, as Haswell has 
described in the case of the first pair in the male of 
Stratiodrilus. I am quite positive in saying no such 
division takes place in Histriobdella, They run through 

veL. 55, PART 2.—NEW SERIES. 20 


302 CRESSWELL SHEARER. 


only one segment, in every case terminating in the next 
segment to that in which they arise in the manner similar to 
the nephridia in Annelids. 

In both sexes the first three pairs of nephridia hold the 
same position, but the fourth varies according to sex. In 
the male it is situated much farther back—at the junction of 
the generative with that of the following segment, while in 
the female it is much more forward—in the anterior portion 
of this segment. In the female the third and fourth pairs 
overlap and cross one another in different planes. 

The most posterior nephridium in Stratiodrilus is in 
that segment that would correspond to the fifth in Histrio- 
bdella. In neither of the sexes are organs found in this 
region in Histriobdella. 

Foettinger, in figuring a pair of nephridia in the region of 
the penis, evidently mistook the motion of the cilia in the vas 
deferens, or the slit in the vesicule, for the ciliary motion 
of excretory organs. The slit in the vesicule was first 
described by Haswell in Stratiodrilus, and as such was 
evidently overlooked by Foettinger. ‘They are even better 
developed in Histriobdella than in Stratiodrilus. They 
are edged with very stout cilia, that could readily be mistaken 
for nephridial flagella. With methyl-blue it is easy to 
determine, however, that no excretory organs exist at this 
point in Histriobdella. 

In Histriobdella, unlike Stratiodrilus, the nephridia 
partake to some extent of the metamerism of the body. In 
the third and fourth segments this is masked in the male by 
the great. development of the reproductive organs. In the 
case of the nephridia there has resulted a pushing forward in 
the female of the fourth organ, while in the male this has 
been reduced in size and moved backwards. 

Unlike Dinophilus, we do not find the sharp specialisation 
of the different parts of the nephridial canal into a thick- 
walled anterior excretory portion and a thin-walled posterior 
part. It is more uniform throughout in structure. 


9 


ON THE ANATOMY OF. HISTRIOBDELLA HOMARI. 303 


A. The First Nephridium (figs. 1, 7, 9, 37, and 42). 


The first nephridium arises in the segment immediately 
behind the head, and opens on the exterior in the second 
segment. The head of the organ I have never been able to 
see plainly in the living condition, as it is hidden by the 
muscle-bands. These are always undergoing contraction 
during life ; the lumen of the canal in its anterior portion is 
continually compressed, and thus the action of its cilia 
rendered very intermittent. The head of the organ at its 
point of origin is very close under the epidermis; in one 
case seeming to be almost under the limiting membrane of 
the epidermis. In section the head of the organ appears as 
shown in fig. 42, which is taken in a horizontal plane in the 
dorsal region of the first segment. The space into which the 
nephridial head projects is shown surrounded by a number 
of darkly staining nuclei. ‘The canal with its flagella is 
shown cut in section in the body-cavity. The actual projec- 
tion of the head into the space is not seen in this section. 

The neck segment is very clear, and were the canal pro- 
longed into the head, asin Stratiodrilus, it could easily be 
seen at this point passing into the head. As this can never 
be done, it is apparent that the organ takes its origin in the 
neck segment and is not prolonged into the head. It is also 
certain that it does not divide and send a branch to join one 
from the opposite side, as in Stratiodrilus. Throughout 
its course it is a simple, unbranched, intra-cellular tube, being 
in the same position in both sexes. 


B. The Second Nephridium (figs. 1, 6, 7,9, 14, and 37). 


The second nephridium arises in the anterior portion of the 
second seement, and runs back to terminate in the anterior 
part of the third. It is much the longest, being twice the 
length of the first. Its course is straight backwards along the 
border of the muscle-bands. ‘The action of its cia is much 
more constant than that of the others, and for this reason it 


304 CRESSWELL SHEARER, 


is the one most readily observed. The main portion of its 
canal is slender and thin-walled. The general course of the 
organ is shown in figs. 1, 6, 7, 9, 14, and 37. It will be seen 
from these figures that Foettinger observed the organ only 
at the pomt where it passes from the second to the third 
segment, and that he was unaware of its considerable exten- 
sion into the anterior region of the second segment. In the 
female the segments through which it stretches are somewhat 
more compressed, and for this reason it appears in the female 
shorter than in the male. 


c. The Third Nephridium (figs. 1, 3, 4, 10, and 13). 


The third nephridium has much the same position in both 
sexes. It arises in the anterior part of the third segment and 
runs back to bend outwards in the male and slightly in- 
wards in the female, and terminates in the anterior part of the 
generative region. When the body is retracted it overlaps 
the posterior third of the second. The head of the organ, as 
already mentioned, at its point of origin is in the normal con- 
dition ona level with the opening of the second. It is situated 
close under the epidermis, as in the case of the first nephridium, 
and on the dorsal surface. It runs backwards, and about the 
middle of its course makes a sharp turn ventralwards (fig. 5), 
In the female it overlaps considerably the fourth, its opening 
on the exterior being internal to the course of this organ, 
While in the male it makes only one turn outwards, in the 
female it is S-shaped, the terminal portion running inwards 
(fig. 3). 


p. The Fourth Nephridium (figs. 1, 3, 10, and 13). 


In the male the fourth nephridium arises in the posterior 
part of the generative segment in the region immediately 
behind the clasper, It runs backwards and terminates in the 
anterior part of the caudal segment. Its course is short and 
somewhat difficult to observe. It is much the smallest of all 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 305 


the nephridia, and its action more feeble than that of the 
others, as the motion of its flagella is the first to stop when 
the animal is compressed under a cover-slip. It, however, 
assumes a much darker colour on impregnation with methyl- 
blue than do the other nephridia, and for this reason seems to 
play a considerable part in the excretion of waste products, 
Its position in the male has been correctly indicated by 
Foettinger, who remarks that no organ is to be found in. 
this position in the female. 

Unlike Stratiodrilus, the organ does not begin in a loop 
or run back so far in the tail region, but opens on the exterior 
just over the line of separation between the generative and 
the caudal segments (fig. 12). 


4. Bopy-Caviry AND NEPHRIDIA. 


As in Dinophilus, there is an extensive blastoccelic cavity 
surrounding the gut, which sends prolongations into the head 
region, and also into the anterior and posterior feet. It has 
been described by Foettinger as lined by a more or less 
definite coelomic epithelium, I cannot find that this is strictly 
the case. The gut surface of the cavity is covered by a 
delicate cuticle, in which at rare intervals are seen small 
flat nuclei. It is difficult to say if this membrane is a definite 
structure or a mere secretion from the blastoccelic ends of 
the cells of the gut-wall. ‘he somatopleuric side of the cavity 
is not lined by any such membrane. The longitudinal muscles, 
as in Stratiodrilus, are surrounded by a similar delicate 
cuticle, but no nuclei are to be seen in it as in the gut 
membrane. I believe in neither of these cases can this 
membrane be considered a true peritoneal or coelomic epithe- 
lium. No mesenteries are present, nor can I observe the 
fusion of the gut to the dorsal ectoderm as mentioned by 
Foettinger. The gut is more or less closely applied to the 
dorsal wall, but I cannot find that any true fusion takes 
place. 

In the head the blastoccelic space sends prolongations into 


306 CRESSWELL SHEARER. 


the feet, and forward on the under side of the brain. It is 
more or less separated from the cavity of the trunk by the 
neck muscles and the narrow constricted condition of this 
region. Its extension into the posterior feet is in free 
communication with the trunk, so that in living preparations 
the eggs in the female can sometimes be forced into the leg 
portion of the cavity by slight compression of the cover- 
glass. They slip back, however, to the main blastoccelic 
space surrounding the gut when this pressure is removed. 
The whole of the blastoccelic cavity is very irregular in out- 
line, and is divided, as has been described, in the trunk 
region by the oblique muscle strands into two lateral 
chambers. 

In every respect it corresponds with the same cavity sur- 
rounding the gut in Dinophilus. There is this difference, 
however, that the numerous brown granules seen in it in 
Dinophilus are wanting in Histriobdella, although 
Histriobdella, like Dinophilus, has no specialised vascular 
system. It is sharply divided from the sae of the ovary, 
there being no communication between the two. When the 
ova are forced into the blastoccelic space of the hind limbs 
the sac of the ovary is either pushed with them, or is definitely 
ruptured, and the ova pass directly into the blastoccele. Both 
at the anterior and posterior regions the wall of the ovary is 
considerably thickened where it crosses the blastoccelic space 
between the body-wall and the gut. In the male the sac of 
the testis is likewise sharply cut off from the blastoccelic 
space in the anterior and posterior part of the generative 
region. Histriobdella, like Dinophilus, shows the primary 
and secondary body-cavity existing together, but sharply 
divided from one another. The nephridia, as in Dinoaphilus, 
are in relation with the blastoccelic cavity alone. 

From the fact that we get two nephridia in the generative 
region in the female, there is considerable reason for concluding 
that the oviduct and its funnel can hardly represent a trans- 
formed nephridium as Haswell has suggested. The arrange- 
ment of the ganglia and the external appearance of the 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI, 307 


segmentation bear out the conclusion that in the male and 
female this region is composed of two segments. Haswell, 
in Stratiodrilus, states that “in the fourth segment the 
nephridia are probably represented in the female by the ovi- 
ducts, in the male by the vasa deferentia.” I have shown 
in the male and female that two nephridia are present in the 
generative region, although holding slightly different positions 
in the two sexes. It is therefore impossible that the ovi- 
duct and vasa deferentia represent transformed nephridia, 
unless we consider the generative region to be composed of 
three segments, for which there is no evidence. 

In my paper on the nephridia of Dinophilus teniatus 
I have given some reasons for opposing the view brought 
forward by Schimkewitsch (28) and Harmer (12), that the 
oviducts and vasa deferentia in the male of this animal 
represent modified nephridia. Here there are four pairs of 
close solenocyte-bearing nephridia in the male and five in the 
female. ‘They Show the same primitive relationship with 
the blastoccelic cavity as do those of Histriobdella. 
Harmer’s suggestion is that in the male the fifth nephridium 
has been modified into the vesicule seminales and vasa 
deferentia, while it remains unmodified in the female as the 
fifth nephridium. In the male he holds that one of the 
pairs of nephridia has lost its primitive relationship with the 
blastoccelic cavity, and here becomes highly modified into 
the large ciliated apparatus of the vesicula seminalis and 
the vasa deferentia. The principal evidence relied on by 
Harmer in making this comparison is the resemblance of the 
funnel-like opening of the vasa deferentia into the cavity 
of the testis, to the funnels with which he thought the 
nephridia were furnished. | have shown that these do not 
exist, and that the nephridia of D. teniatus are definitely 
closed. ‘Therefore the funnels of the vasa deferentia 
cannot be derived through modification from those of the 
nephridia, 

In Histriobdella and Dinophilus, I believe the ovi- 
ducts, funnels, and vasa deferentia represent structures 


308 ORESSWELL SHEARER. 


belonging to an entirely different set of organs from those of 
the nephridia, viz. the ccelomoducts of Lankester’s nomen- 
clature. 


5. Muscunar System. 


The muscular system has been described by Foettinger, 
whose account is correct in its main particulars. ‘he muscles 
of the trunk region, as described by him, consist of two 
groups, the dorsal and ventral longitudinal, and the irregular 
oblique or transverse muscles. It is to these last that I wish 
to call particular attention in the present account, as they 
are only mentioned briefly by Foettinger. 

In addition to this I have been able to add new details in 
the division and arrangement of the fibres of the longitudinal 
muscles that escaped Foettinger’s observation. 


A. Longitudinal Muscles. 


The chief muscles of the body are these powerful longi- 
tudinal bands. They have already been described by 
Foettinger in considerable detail. They consist of two 
dorsal and two ventral sets. Each band is composed of from 
twenty to thirty fibres, flattened dorso-ventrally. They are 
attached by their outer margins to the cuticle, while their 
free edges project into the body-cavity. In the generative 
segment their number seems reduced, but this is due to their 
confinement within a limited space—against the gut dorsally 
and the nerve-cord ventrally. In the caudal region they 
spread out, forming a more or Jess complete wall round the 
segment, only interrupted dorsally by the gut and ventrally 
by the nerve+cord. They split up in the head and tail regions, 
sending fibres to the jaws and the anterior and posterior feet. 
In the head dorsally they converge on one another, uniting 
in the median plane, and are inserted in the anterior surface 
of the jaw apparatus. The ventral bands, on the other hand, 
divide into two sets of fibres, the outer of which split again 
to supply fibres to the anterior and posterior surfaces of the 


ON THE ANATOMY: OF HISTRIOBDELLA HOMARI. 309 


anterior feet, while the other set run forward and are inserted 
ventrally into the anterior part of the jaw mechanism. In 
the posterior region each band splits likewise, the ventral 
sending fibres to the foot of the same side, other fibres cross- 
ing to be inserted in the small appendage of the posterior 
limb. ‘The dorsal send part of their fibres into the leg on the 
same side, while the internal ones cross over to be inserted in 
the leg of the opposite side, these fibres thus forming a cross 
dorsal to the anal part of the gut. The dorsal longitudinal 
bands give off a few fibres to the two segments of the caudal 
region, which run towards the median line and are inserted 
into the cuticle. It is due to the action of these fibres that 
the contraction of the caudal segment is brought about. 

As already mentioned, the ventral bands split in the head 
region into two sets of fibres. One of these runs forward to 
be inserted in the anterior part of the jaw apparatus, while 
the other supplies the extensor and flexor surfaces of the 
anterior feet. The manner of insertion of these last is some- 
what peculiar. The fibres of the external side of the longi- 
tudinal band are not inserted immediately into that side of 
the foot nearest them, but run to the anterior surface of the 
organ, while those of the inner side of the band cross these 
to run to the posterior surface. In addition to these there 
are also other fibres, derived from the bands of the opposite 
side of the body, that also run to the anterior face of the foot. 
These fibres form a cross ventral to the anterior end of the 
stomach. Besides these there are some strands that run 
from the same side of the foot directly towards the median 
line, and appear to be inserted into the anterior end of the 
jaws. All these are inserted into the distal joint of the foot. 
‘The course of these different fibres can be readily understood 
on reference to text-figs. 1 and 2. 


B. Special Muscles of the Generative Segment. 

In the anterior and posterior part of the generative segment, 
inthe intersegmental region, a few transverse fibres are 
present, running beneath the epidermis. 


310 CRESSWELL SHEARER. 


In the male special muscles are developed in relation with 
the claspersand the penis. These are similar in their arrange- 
ment to the same muscles of Stratiodrilus. The penis 
possesses a pair of protractors and retractors. The retractor 
muscles also function as the retractors of theclaspers. They 


TrxtT-FIGS. 1 AND 2. 


I. Male. rr Female. 


The muscles seen from the dorsal side in the male and female. 
The division of the longitudinal bands in the head and 
caudal regions is also represented. 


run from the base of the penis to the base of the claspers, 
and by their contraction at the same time retract the penis 
and claspers. ‘lhe claspers have also, as in Stratiodrilus, 
a set of protractor muscles, which run obliquely forwards and 
inwards in the generative segment, and also a few fibres that 
run from the bottom of the clasper sheath to the anterior lip 
of the same. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 31] 


- It will be seen that there is some difference between the 
arrangement of the main muscles in Histriobdella as com- 
pared with Stratiodrilus. In the neck region I cannot 
find the complicated crossing of fibres shown by Haswell in 
his fig. 1. Nor in the posterior legs can I distinguish some 
of the fibres he represents. .The muscular system of Stratio- 
drilus is much better developed, and the presence of cirri 
and the retractile condition of the anterior feet give it a more 
elaborate muscular system than that of Histriobdella. 


c. Oblique Muscles. 


If we examine a number of transverse sections we will see 
the body-cavity traversed occasionally by oblique strands 
(figs. 39, 40, 41, 43, and 44). Foettinger mentions their 
resemblance to the oblique muscles of Protodrilus, but he 
was somewhat uncertain as to their nature. He says, “Je 
n’ai pu m/assurer si elles étaient de nature musculaire” 
(p. 457). They divide the body-cavity, as in the Archi- 
annelids, into a circular portion surrounding the gut and a right 
and left lateral chamber. In some of my sections they form 
almost a continual sheet of fibres, and they are much better 
developed than one might suppose from Foettinger’s remark. 
They are found as irregular bundles crossing the body- 
cavity from the head to the tail region. They are well marked 
in the posterior part of the head; commencing at a point on 
a line with the chitinous jaws, they are continued back into 
the neck region in an unbroken succession. Inthe middle of 
the segment they almost disappear, while they are more 
prominent in the intersegmental regions. - In the anterior 
and posterior parts of the generative segment they are also 
present, but are entirely missing from the middle in the male, 
being interrupted by the muscles and accessory glands of the 
penis. Anteriorly they divide the testis in two portions, 
forming a right and left chamber (fig. 39). In the anterior 
part of the first segment they are shown in fig. 43. Here, 
during part of their course, they touch the wall of the gut. 


312 CRESSWELL SHEARER. 


In Stratiodrilus their presence has been observed by 
Haswell (13), who states: ‘Throughout the body slender 
oblique bundles occur at fairly regular intervals, running 
from the cuticle of the lateral surface to that of the ventral 
near the nerve-cord” (p. 306). Here, however, they would 
seem to be much less developed. I think there is no doubt 
that they correspond to the oblique muscles of Polygordius. 
It is interesting to note that the nephridial canals, as in 
Polygordius, are always within the limits of the lateral 
cavities formed by them. Another point of similarity consists 
in the manner of their insertion into the dorsal body-wall. 
‘They spread out in a fan-like manner, as Hempelmann (15) 
has shown takes place in Polygordius (see his text-fig. 14). 
This same arrangement of the fibres, it will be seen, is found in 
Histriobdella (fig. 41). Theanterior and posterior feet, in 
addition to the fibres they receive from the longitudinal 
bands, also possess a special musculature of their own. In 
the anterior foot this consists of a series of parallel fibres that 
run from its base to the distal, flat, saucer-like pad of the 
foot. The foot itself is composed of two parts, a distal 
retractile portion and a larger non-retractile, cone-shaped 
basal portion. Some of the fibres are applied closely to the 
cuticle of the outer part of the basal portion, while those of the 
bands, as already mentioned, are inserted into the retractile 
distal portion. They surround and run into the basal gland 
of the foot, There is a collection of granular mucus cells at 
the base of the foot, abutting internally on the jaw muscula- 
ture. ‘They stain deeply with carmine, each cell having a 
darkly granular periphery, with a clear centre hollowed out 
in a small cavity. ‘The gland gives off a number of straight 
tubes, that open on the pad surface of the foot. They run up 
amongst the muscle-fibres, and can be readily distinguished 
from these by the manner in which they take the stain. The 
gland pours out on the surface of the pad some sticky secre- 
tion, by means of which the animal is enabled to obtain a firm 
hold. In the case of the posterior limbs a similar, but larger, 
gland is present.. It extends from the wall of the gut out of 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 313 


the centre of each leg to the commencement of the outer 
third, where it gives off a mass of fine, darkly staining 
tubules, which open on the pyramidal pad of the foot. This 
gland is able to pour out a copious viscid secretion. Fre- 
quently, when the animals are irritated, this secretion can be 


TEXT-FIG. 3. 


m.d.t------tM 


| i we --¢p. f. 


3. 


Head showing the muscles in connection with the feet and 
the jaws. bl.m., Bulb-like muscular organ of the jaws; 
cl. p., ciliated pit of the head; cv. f., crossed strands of the 
dorsal longitudinal muscles ; m.d.f.a., dorsal longitudinal 
muscles running into anterior surface of the foot; m.d.f.p., 
dorsal longitudinal muscles running into posterior surface 
of the foot; m.gl., salivary gland of the mouth; m.d.j., 
dorsal longitudinal muscle running to jaw apparatus ; 
tr,m., transverse muscle-strands running into the feet. 


seen pouring out from the ends of the tubules, forming 
minute round drops on the end of thefoot. Like the anterior 
limb, the posterior has some muscular fibres apart from those 
it receives from the longitudinal muscle-bands. These are a 
delicate set of fibres just under the cuticle on the posterior 
surface, that run from the extremity to be inserted on either 
side of the anus. In addition to these there are some oblique 


314 CRESSWELL SHEARER. 


fibres, asin Stratiodrilus, but they are but feebly developed. 
A considerable prolongation of the blastoccelic cavity takes 
place into the posterior limbs, running out along each leg 
between the muscle-fibres and the glands. Into this space 
the ova in the female are sometimes forced when the animal is 
compressed under a cover-slip, showing that it is in free com- 
munication with the cavity surrounding the gut. 

The movement of the limbs takes place alternately, the 
head being swung from side to side with the movement of 
the feet. It is a most remarkable sight to see the animals 
rear up, as they sometimes do, on their hind feet, and stand 
executing movements with their head while they remain 
firmly attached with their powerful hind feet. ‘They also 
crawl quite readily, by means of the feet, on the underside of 
the surface-film of the water. In the ordinary movements of 
crawling the glands do not appear to throw out any secretion 
on the pads of the feet; only when they are disturbed do 
they pour out a thick secretion, which firmly attaches the 
feet to the surface on which they happen to be. While the 
animal violently twists its head and body, it never moves its 
feet. This hold is remarkably firm. On the lobster ova the 
parasites can be seized by the middle of the body by means 
of a pair of fine forceps, under a dissecting microscope, and 
the body pulled off, leaving the feet still attached, the limbs 
having been torn from the body without loosening their 
hold. 

As already mentioned, the front limbs in Histriobdella 
differ from those of Stratiodrilus in that they are non- 
retractile. I have never been able to observe any retraction 
of the feet in the living condition, or in preserved specimens 
treated with different reagents. 


6. DicEstIvE SysTEmM. 


The digestive system is sharply specialised into a number 
of divisions. These are readily seen in the figure of an 
immature parasite (fig. 30). - Here they are more marked 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 315 


than in the adult. A more or less slender cesophagus leads 
dorsalwards and backwards from a quadrilateral-shaped 
mouth (text-fig. 4). ‘This, although small, is capable of 
considerable expansion. It is completely everted in allowing 
the jaws and teeth to be protruded in the act of biting. It 
commences in a slight ciliated depression, which rapidly 
deepens into a groove in the anterior part of the head. The 
cesophagus terminates, on a line with the posterior boundary 
of the jaw musculature, in a narrow constriction leading into 
the stomach. It is difficult to say where the mouth ends and 
the cesophagus commences. ‘lhe mouth and cesophagus are 
lined throughout with fine cilia, those of the cesophagus 
being much stouter than those of the mouth. 


TEXT-FIG. 4. 


si 


Showing the outline of the mouth when partially closed. 


The stomach may be defined as that portion of the intestinal 
tract lying between the first and the third segment. Its wall 
is composed of a single layer of cubical cells. It is for the 
most part uniform in thickness. The rounded internal ends 
of cells project irregularly into the lumen and are ciliated. 
At the anterior end, near the cesophagus, the cells are very 
columnar and contain many granules. ‘They have probably 
to do with the elaboration of the digestive secretions, as they 
are seen to be very opaque after the animals have taken food. 
Those of the ventral wall in this part are somewhat larger 
than the dorsal. The nuclei are always placed at the bottom 
of the cell, that.is, farthest from the internal ciliated surface. 
In the anterior region they are long and oval in shape, while 
in the middle and posterior regions they are spherical, and 
the cells themselves cubical in outline. In the posterior 


316 CRESSWELL SHEARER. 


region of the stomach the wall is relatively thin in comparison 
with that of the anterior part, and its cells on the ventral 
side are furnished with very long, dense cilia. 

About the middle of the end of the third segment the 
stomach contracts into a narrow mid-gut, which runs 
through the generative region to widen somewhat in the 
caudal region into a more or less straight hind-gut. The 
lumen of the intestinal tract, from the stomach backwards, is 
greatly reduced in size, and, in the contracted condition of 
the animal, somewhat folded on itself. The character of its 
ciliation is also different from that of the stomach. At the 
point where the stomach passes into the mid-gut there is a 
sort of valve formed by the thickening of the stomach-wall. 
A similar valve is found at the point of union with the hind- 
gut. The wall of the mid-gut is relatively the thinnest part 
of the tract, and its cells are not of the marked yellow colour 
of those of the stomach. ‘The course of the mid-gut is 
irregular, from its being slightly folded on itself. That of 
the hind-gut is comparatively straight, bnt its lumen is 
irregular and wavy in outline, due to the irregular thicken- 
ing of the wall at different points on its course. Throughout 
the generative segment the gut is- very closely confined 
against the dorsal body-wall. The anus is dorsal. The cells 
of the hind-gut are of a character quite different from those 
of the other parts of the tract. They are quite irregular in 
size, and extend into the lumen so as to make its outline very 
broken, as if thrown into a number of convolutions. In no 
part of the wall of the stomach or gut are any contractile 
muscular fibres to be-seen. In the body-cavity, ventral to 
the anus, and close to the point where the gut joins the body- 
wall to form the anus, there is usually present a conspicuous 
cell on either side. The anus itself is an oblong, vertically 
placed, T-shaped slit placed more towards the dorsal than the 
ventral side of the animal. It is apparently kept closed by 
some contractile fibres of the cuticle which function as a sort 
of sphincter muscle. ; 

The digestive traet of Stratiodrilus agrees in all essen- 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. Sz. 


tial details with that of Histriobdella as far as can be 
judged from Haswell’s somewhat brief description. There is 
the same reduction of the tract in the generative region, this 
being much greater in the female than in the male, and its 
expansion into a more or less large hind-gut in the caudal 
region. 

Ascompared with Dinophilus there is a greater difference. 
Yet with the exception of the pecuhar mid-gut portion of the 
tract, which is a development due to the peculiar condition 
produced ‘by the presence of a special generative segment, 
there is considerable resemblance between Histriobdella 
and Dinophilus, and in many of the finer histological 
details there is a very close resemblance. In the first place, 
the appearance of the cells of the stomach, each composed of 
a single layer of ciliated cells, the yellow vacuolated appear- 
ance of their protoplasm, and the basal arrangement of the 
nuclei, are the same in the two. The terminal dorsal position 
of the anus and the configuration of the cesophagus and 
pharynx are remarkably the same in both. 

According to Nelson (25) there is a feeble strand of muscle- 
fibres that act in Dinophilus as sphincter ani, as in 
Histriobdella. Throughout the stomach region there is a 
lack of muscular strands, and the stomach is not supported 
by mesenteries, but is closely appled to the dorsal wall, as 
in Histriobdella. The blastoccelic surface of the stomach, 
as in Histriobdella, is covered with a fine cuticle. 

The jaw apparatus of Histriobdella is very similar to 
that of Stratiodrilus. Haswell has given an extensive 
description of this, so that I need only briefly consider it. 
As in Stratiodrilus, it consists of two portions—the upper 
and the lower. The upper consists of a median rod (fig. 36), 
which Haswell has called the fulcrum. This is slender, 
round, and slightly curved; it articulates by means of a 
number of basal pieces with a series of jointed arms, each 
terminating in a curved tooth (text-fig. 5). It lies in the 
median plane dorsal to the two blades of the lower jaws, being 
set at a different angle to these. Its length is somewhat less 

VOL. 55, PART 2.—NEW SERIES. Pal 


318 GRESSWELL SHEARER. 


than these last. The cubical basal pieces with which it 
articulates support four arms on each side, each being com- 
posed of three or four pieces, the last of which is fashioned 
into a sharply-curved claw-like tooth. This is strongly 
serrated on its inner edge. One difference between the jaw 
parts of Histriobdella and Stratiodrilus consists in the 
length of the middle joint of these arms. ‘They are much 


TEXT-FIG. 5. 


Jaw apparatus. Enlarged figure showing the structure of 
the teeth and the arrangement of the jJaw-muscles. ba.p.j., 
basal piece of Jaws; bl. m., bulb-like muscular organ of the 
jaws; f., fulerum; m.d.j., strands of the dorsal longi- 
tudinal muscle-bands running to the jaws; st.m.j.., 
striated muscles of the Jaws; th., teeth. 


longer in Histriobdella, and allow of the teeth being 
folded back in the mouth or cesophagus to a greater extent 
than in Stratiodrilus. When at rest in the ordinary 
position the teeth are not folded back to their full extent. 
The middle piece of the arm projects at right angles to the 
jaws, and in this position the most anterior part of the arm is 
the distal joint, the tooth being strongly flexed. When the 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 319 


arms are folded to their full extent the fulcrum is drawn back 
on a level with the extremity of the blades of the lower jaws. 
These are paired throughout. They consist of two long wide 
blades, thickened at their outer margins, and articulating at 
their basal ends with two curved wedge-like pieces, the 
pointed end of the wedge being directed forward in the 
ventral lip of the mouth. Ventrally they articulate with one 
another in the median line, and turn up dorsally to form a 
support for the upper jaws. The upper anterior angle of 
each plate is turned outwards and backwards, some of the 
fibres of the dorsal longitudinal muscle-bands being inserted 
into it. The internal interior edges of these plates are finely 
serrated, and evidently assist the teeth in their action. As 
far as can be judged from Haswell’s figures, the shape cf 
these plates differs slightly in Histriobdella from that of 
Stratiodrilus. They fold up dorsally to a greater degree. 
Tbe main portion of the lower jaws are the wide blade-like 
portions which project backwards parallel with one another. 
They are widest behind, and taper slightly in front, where 
they articulate with the wedge-like portions. Connecting 
the upper with the lower jaws are the pieces that Haswell 
distinguishes as “ bridles.’ Into the posterior extremities of 
these are inserted the powerful striated muscles. Apart 
from their action in binding together the jaw-sets I have not 
satisfactorily determined their function. They would seem 
to be composed of a single curved piece in Histriobdella, 
and its chitinous substance is broken up into a number of 
dark hairs where the muscle is inserted, giving it a furred 
appearance. The powerful nature of these fibres shows that 
their action in pulling on the bridles has to do with some 
essential movement of the jaws. It is likely that the actual 
process of biting is brought about by their contraction, as 
Haswell has suggested, while the fulcrum has merely to do 
with their protrusion and opening. In addition to these 
there is the peculiar bulbular muscular organ, not unlike the 
sub-cesophageal muscle pad of Dinophilus. This is attached 
to the posterior ends of the ventral surfaces of the lower 


320 CRESSWELL SHEARER. 


jaws. Its fibres form an oval mass attached directly to the 
jaw blades. Into this mass some of the striated muscle-fibres 
are inserted. Its action is hard to understand. It is well 
shown in Foettinger’s figures. In the movements of the 
jaws the lower blades are sometimes seen to separate con- 
siderably from one another posteriorly, and it is possible this 
motion is brought about by them. What this movement has 
to do with the teeth I have been unable to observe. This 
muscular organ appears to be wanting in Stratiodrilus, as 
it is not shown in Haswell’s figures. 

On either side of the jaws about their middle there pro- 
trudes laterally a small pear-shaped gland composed of from 
three to four large granular cells with conspicuous nuclei 
(text-figs. 1 and 2). This gland opens into the mouth or the 
anterior part of the cesophagus, and is evidently of a mucous 
nature, as it absorbs the methyl-blue colour very strongly 
when the parasites are placed in it for a short time. The 
protoplasm of the gland-cells is finely granular, each having 
a very large, darkly staining nucleus with a prominent 
nucleolus. The duct of the gland converges and opens on 
the ventral side of the mouth. The posterior portion of the 
organ lies against the muscular pad of the ends of the lower 
jaws, while its dorsal surface touches the cuticle of the dorsal 
surface of the head. 

In position and structure it is in all respects similar to the 
glands occupying the same position in Dinophilus, and 
undoubtedly answers the same purpose. In Protodrilus, 
also, similar glands are present. It appears to have been 
overlooked by Foettinger. In fig. 1 of his paper he shows a 
mass of tissue on either side of the jaws, which in great part 
belongs to these salivary glands, and not to the jaw muscles, 
as he evidently thought. Haswell makes no mention of its 
presence in Stratiodrilus, although it is probably present 
here also, for he shows a number of round cells in the position 
that it occupies in Histriobdella, 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. S2t 


7. THe Nervous System (figs. 15, 21, and 28). 


The nervous system extends throughout the body, and is 
composed of a brain, cesophageal commissures, and ventral 
nerve-cord, with ganglia at intervals corresponding to the 
external segmentation. The brain is situated well forward 
in the head, its main mass being anterior to the oral opening, 
and close to the dorsal surface. It is composed externally 
of a mass of nerve-cells surrounding a clear fibrous core. 
The nerve-cells are distributed over its dorsal surface. 
Behind, the brain is deeply cleft ventrally, descending in 
lateral lobes on either side of the anterior part of the mouth. 
This cleft runs forward, forming a small closed sinus in the 
anterior end of the organ. 

The brain terminates rather abruptly at a point about on 
a line with the anterior third of the jaws; here it gives off 
two fine commissures that run directly ventralwards and 
backwards, connecting it with the first ganglion of the ventral 
nerve-cord. At the point where these come off some fibres 
go to the anterior legs, and others run directly backwards in 
the dorsal region. ‘They probably correspond with the 
“nerfs sympathiques” of Foettinger. In addition to these, 
the brain supplies nerves to the anterior tentacles. 

The commissures are closely applied to the cesophagus, and 
are difficult to follow in sections on account of their small 
size. 

At about on a line with the posterior boundary of the 
brain, and slightly in front of the anterior feet, there is a 
small ciliated pit on either side of the head. The anterior 
lip of this protrudes slightly, forming a sort of papilla. This 
pit is undoubtedly sensory in nature, and appears to have 
some fine nerve-fibres running to it from the brain. ‘The 
nerve-cells of the dorsal surface of the brain are distinctly 
differentiated from the cells of the ectoderm. ‘hey are 
recognisable by the elliptical outline of their nuclei, and the 
marked way in which they take the stain when treated with 
the hematoxylin mixtures. As compared with the ectoderm 


322) CRESSWELL SHEARER. « 


cells, their nuclei are rich in chromatin. ‘This peculiarity 
renders them distinguishable from the supporting cells of the 
surrounding tissues. Some of the ganglion cells are clearly 
multipolar, but axons and dendrites are not recognisable. At 
the base of the tentacles the cells are bipolar, one process 
going into the tentacle while the other enters the neuropile. 
They form a dense mass of cells on the anterior dorsal surface 
of the brain-core. They are, however, quite distinct from it, 
only sending a few fine threads into its substance. In the 
median plane a small space, a prolongation of the general 
blastoceelic space, extends up under the brain, and separates 
them from the core, dividing them into two lateral masses. 
The central core of the brain is composed of a dense mass 
of interwoven nerve-fibres. It is distinguishable by its 
yellow colour and its non-nucleated character. It is remark- 
able that both in relation with the brain and the ventral cord 
the nerve-cells seem quite apart, and outside the fibrillar part 
of the nervous system. ‘Their relationship seems closer with 
the ectodermic tissues of the head and the mesodermic 
and ectodermic tissues in the trunk than with the fibrillar 
material of the nervous system in these regions. 

‘The fibres of the ventral portion of the neuropile seem to 
run from side to side, while those of the superficial layers ran 
more longitudinally. In sagittal sections it is lenticular in 
outline, and in the median plane is divided by a transverse 
fissure into an anterior and posterior part. Haswell also 
shows these divisions in the brain of Stratiodrilus (fig. 8). 
This division is only limited to the median plane; laterally 
the neuropile swells out into two large lobes on either side. 
Thus it consists, as in Dinophilus, in a median and two 
lateral lobes, the median being in turn divided into an 
anterior and posterior portion. In the figures of the brain 
accompanying this paper these divisions do not show, as the 
brain surface is taken from the ganglion cells and not from 
the central core. Behind the brain, and dorsal to the 
muscular apparatus of the jaws, there is a second accumula- 
tion of nerve-cells. These may possibly have to do with the 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 323 


innervation of the jaw muscles; they are dorsal and median 
to the cesophageal commissures. I have been unable to make 
out their connection with the muscles. ‘They take up methyl- 
blue much more readily than do the other cells of the brain, 
and retain it considerably longer. 

The ventral nerve-cord, like the brain, consists of a similar 
central fibrous core, surrounded with nerve-cells. The two 
halves of the cord are separated in the intersegmental 
regions, joining up in the middle of the segments to form a 
ganglion. From what can be judged from Haswell’s draw- 
ings, in Stratiodrilus this separation is much less than in 
Histriobdella. Unfortunately, most of the sections drawn 
_ by Foettinger are taken through the middle of the segments, 
and do not properly illustrate the extent to which the two 
portions of the cord separate in the intersegmental regions. 
The two halves of the cord are crescentic in transverse 
section, the nerve-cells being imbedded on the ventral surface. 
Where the cords unite these cells are drawn out laterally to 
form considerable masses on either side. 

The main ganglia, as already mentioned, correspond closely 
with the five main segments into which the trunk is divided. 
The first is situated in the anterior region of the first segment, 
and is of considerable size. ‘he second is somewhat smaller, 
and is situated about the middle of the second segment. It 
has fewer nerve-cells, and, like Stratiodrilus, it is placed 
nearer the first than the third ganghon. The third is the 
largest, taking up the greater part of the length of the cord 
in the third segment, and having a great number of nerve- 
cells. Between all the ganglia in the intersegmental regions 
of the anterior segments the component parts of the cord 
separate as already mentioned ; between the third and fourth 
ganglia this is hardly perceptible, and from this point back- 
wards to the tail region the two portions of the cord are in 
close union, with the exception of a small area near the end. 

The fourth ganglion is the largest of all, and occupies the 
middie of the generative segment. ‘The fifth is in the middle 
of the caudal segment. The position of these ganglia can be 


324 CRESSWELL SHEARER. 


seen from the reconstructions shown in figs. 15, 21, and 28. 
In these figures the nervous system is seen from the ventral 
side. The outlines of the cord and ganglia have been 
measured from the nerve-cel!s, as far as these could be 
roughly differentiated from the surrounding tissues.’ From 
fig. 15 it will be seen that the main mass of the fourth gan- 
glion lies just in front of the penis, but many of its cells extend 
backwards in the region dorsal to the penis. Here they 
would almost seem to form a second division of the ganglion. 
I have not attempted to determine its structure, which differs 
considerably from that of the other ganglia, on account of the 
great size of its lateral parts. A few of its cells are distri- 
buted on the penis sheath. Past the fourth ganglion the cord 
diminishes rapidly, but enlarges again rather suddenly in the 
interior part of the caudal region. It is the second in this 
segment that is the largest. In the posterior region the cord 
divides to run into the posterior feet. At this point a number 
of nerve-cells are arranged, forming quite a mass. Itis diffi- 
cult to decide whether each of these ganglia is to be considered 
as representing a segment. If so, then there are three main 
ganglia in the segment itself, and counting the mass of cells 
at the termination of the cord, it would be composed of four 
segments. Foettinger came to the conclusion that it was one 
segment, formed by the partial fusion of three metameres. 
In the female there are a number of differences in the con- 
figuration of the nervous system, due to the somewhat different 
size of the segments as compared with the male. This is 
most pronounced in the generative region. ‘lhe absence in 
the female of the penis and accessory glands results in the 
almost complete disappearance of the cord and ganglia in the 
posterior part of the generative region, and throughout this 
portion of the body the cord and its ganglia are much less 
prominent than in the male. In the absence of the penis the 
cord retains its ventral position. In the caudal region, on 


1 In the reconstruction of the male nervous system shown in fig. 15 
no allowance has been made for the dorsal curvature of the cord in 
the generative region. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. O20 


the contrary, the cord and ganglia are much the same as in 
the male (fig. 15). 

In Stratiodrilus the cord and ganglia are much the same 
asin Histriobdella. In the male the fourth ganglion is 
opposite the claspers. After this the cord is very much 
reduced where it passes dorsalwards over the penis. In 
Histriobdella this reduction is not so marked. In the 
caudal region also the ganglia are smaller. Haswell remarks, 
regarding the nervous system of the caudal region of 
Stratiodrilus, that “the ventral chain may be described 
either as represented by a single elongated ganglion imper- 
fectly divided into five or six portions, or as consisting of five 
or six imperfectly separated ganglia” (p. 315). The nerve- 
cells are arranged on the ventral surface of the cord, and 
the lateral ganglia are much better developed in Histrio- 
bdella, especially those of the generative region. They send 
fibres into the cirri. The second ganglion would seem to be 
double. 

Haswell has drawn attention to the fact that the nervous 
system in Stratiodrilus does not show the complete union 
with the body-wall tissues as does that of Histriobdella. 
1 think, however, no great importance can be attached to this 
point. The separation shown by the nervous system in 
Stratiodrilus is due in great part to the thinness of the 
body-wall as compared with Histriobdella, and not to a 
more highly differentiated condition of the system itself. 

It is of considerable interest to compare the nervous system 
of Histriobdella with that of the Archiannelid it resembles 
most, that is, Dinophilus. From the study of a species 
closely allied to D. gyrociliatus, Nelson (25) has deter- 
mined the main structure of the central nervous system in 
considerable detail. In the first place there is a marked 
separation of the two parts of the ventral nerve-cord in the 
intersegmental regions, much more so than in Histrio- 
bdella. Unhke Histriobdella they do not unite to form 
the ganglia, but are joined by commissures, the two portions 
of cord remaining separated throughout their course. There 


326 CRESSWELL SHEARER. 


are four well-marked ganglia corresponding to the four main 
segments of the trunk. In addition to this, there are a few 
cells that probably form a fifth, corresponding with the 
somewhat reduced caudal segment. If we compare the 
reconstruction figure he gives of the nervous system with 
that of either the male or female Histriobdella given in the 
present paper, it will be seen that, with the exception of this 
greater separation of the cords, there is a remarkable resem- 
blance in the general configuration of the nervous system of 
the two forms. ‘lhe brain and the cesophageal commissures 
are much the same. In transverse sections the cords hold 
similar positions in the ectoderm. The ventral sinus found in 
the brain of Histriobdella, it would seem, is also present 
in Dinophilus asa small closed cavity in the brain substance 
itself, 

In minor histological details they bear a striking resem- 
blance to one another. The brain is clothed dorsally and 
laterally with a mass of nerve-cells, having the peculiar 
eranular nuclei so characteristic of these cells in Histrio- 
bdella. ‘They are similarly differentiated from the supporting 
tissue cells. ‘lhe circum-cesophageal commissures are better 
developed, however, in Dinophilus, and pass backwards 
round the cesophagus just below the dorsal longitudinal 
muscle-strands. ‘he centre of the brain is composed of a 
mass of clear fibrillar material that stains with difficulty. 
As in Histriobdella, tibres are given off by the cesopha- 
geal commissures at the point where these leave the brain. 
They are much bigger in the case of Dinophilus, and are 
more easily traced through consecutive sections. In His- 
triobdella there are no pre-oral commissures, and the 
ganglia are more circumscribed and definite than in Dino- 
philus. 

As compared with the nervous system of Protodrilus 
there is a greater difference than in the case of Dinophilus. 
This is due to the lack of ganglia on the ventral cord. In 
Protodrilus the ventral cord shows no ganglionic divisions 
corresponding to the external segmentation. This is very 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 327 


slight, being shown ouly by the ciliated rings. Internally it 
is better marked by the dissepiments and the nephridia. he 
two halves of the cord remain separate throughout their 
course, uniting at their ends in a small ganglion. Such a 
nervous system can hardly be compared with that of His- 
triobdella. 

According to Pierantoni (26), the nerve-cells in Proto- 
drilus are equally as difficult to distinguish from the sur- 
rounding tissues as in Histriobdella. While retaining 
their primitive position in the ectoderm, they send fibrils to 
the tentacles and the digestive system. In the ventral cord 
there would seem to be no localisation whatever of the nerve- 
cells corresponding to the segmentation. 


8. SEnsr-OrGANS. 


Among the sense-organs of Huistriobdella are to be 
classed the five tentacles of the head and the palps of the 
posterior legs. All these receive nerve-fibres from the central 
nervous system, and are armed with short, stiff, sensory hairs. 
The most essential of the tentacles appears to be the median 
one of the head. In the larva this is the first to appear, and 
its nerve supply in the adult would seem to be greater than 
that of the others. In addition to the tentacles, scattered 
over the cuticle of the body are a number of cells of a sensory 
nature that stain readily with methylene blue. 

On the dorsal lateral parts of the head are the sensory pits 
described by Foettinger. These, as already mentioned, are 
very small, and placed a short distance in front of the anterior 
feet. Foettinger has sought to compare them with the ciliated 
grooves of Archiannelids. They measure about 14 in their 
longest diameter, and are oval in outline. They are therefore 
much smaller than the long grooves of Protodrilus and 
Polygordius. In the bottom of the pit are placed a few 
fine sensory hairs. As described by Foettinger, the anterior 
edge of the pit is developed into a slight lip or ridge that is 
capable of being folded completely over the pit and of 


328 CRESSWELL SHEARER. 


obliterating it. From the way in which this lip is protruded 
and the pit opened when the animal is feeling its way or 
examining any small object it may come across in crawling 
on the bottom of a watch-glass, it is evident that the pit 
functions in some way as an organ of taste or smell. It 
appears to receive a set of nerve-fibres from the brain. ‘There 
is no doubt that these pits correspond to the ciliated pits of 
the Archiannelids, despite their small size. They are present 
in both sexes. According to Haswell they are not present in 
Stratiodrilus. 


9. THe Rupropucrive System. 

The reproductive organs in the male consist of a testis, 
paired in its anterior part, two vesicule seminales, two 
vasa deferentia, and a median penis. Dorsal to each vesicle 
is the so-called granule gland (fig. 11). In relation with the 
penis there is a gland of unknown function, as in Stratio- 
drilus. 

In the female the organs consist of a large sac or ovary 
filling the whole of the generative region. On its ventral 
surface this is furnished with a paired oviduct, armed with a 
large funnel, the dorsal lip of which only is ciliated. On the 
course of the oviducts and close to their external openings 
are the ampulle or shell-glands. 


A. In the Male. 


The testis in the male when fully developed fills the anterior 
and middle third of the generative region. Its extreme 
anterior end is separated into a right and left portion, its 
middle portion is fused in the median line. Behind it ends 
somewhat abruptly in front of the penis. ‘lhe remaining pos- 
terior third of the generative region is taken up with the 
penis and its accessory glands. This portion is sharply 
divided from the anterior two thirds by the limiting membrane 
of the testis. This fact has not been clearly shown by 
Foettinger. He seems to have overlooked the well-defined 


ON THE ANATOMY OF HISTRIOBDELLA HOMARL. 329 


nature of the limiting membrane, and fails to show the sharp 
manner in which the testis is shut off from the general blasto- 
coelic cavity surrounding the gut. He states that the testis 
takes up the whole of the generative region, which is not the 
case, for the penis and its glands take up the posterior third as 
I have mentioned. ‘The anterior paired portion of the testis 
is shown in section in fig. 39, while the main unpaired portion 
is Shown in section in fig. 35. Internally the testis is filled with 
a number of oval bodies, the spermatidia (figs. 27 and 35). 
These consist of a number of nuclei with granular chromatin, 
arranged round the circumference of a small mass of cyto- 
plasm. In the region close to the anterior end of the testis 
they form a solid mass, while in the middle they crowd its 
cavity as a number of oval bodies. The mature spermatozoa 
are found in the spaces of the testis cavity between them. 

If we regard the generative region as due to the fusion of 
two segments, then this conclusion is supported by the arrange- 
ment of the nephridia and the ganglia. ‘he testis itself takes 
up the first and largest of these, while the penis and accessory 
olands take up the second. The division between the testis 
and penis portion comes at just that point we should naturally 
conclude that it should from the position of the ganglia. 

In the female the double nature of the generative region is 
not so clear asin the male, and the metamerism is masked 
by the extensive prolongation backwards of the ovarian sac. 
In the young female, however, the ovary is confined to the 
anterior two-thirds. The double nature of the generative 
region then is almost as distinct in the female as in the male. 

The vesicule seminales are found in the posterior part 
of the testis, and are pear-shaped bodies with their pointed 
ends directed forwards. They are readily recognised on 
account of the large quantities of sperm with which they are 
always crowded. Leading into the lateral surface of each 
vesicle is a fine duct from the granule gland. 

These are a mass of large mucus-like cells that lie against 
the inner surface of the cuticle of the body-wall of the genera- 
tive region. They secrete a granular mucous substance which 


330 CRESSWELL SHEARER. 


they discharge into the vesicule. Hach gland is composed 
of about twenty cells, arranged in a single layer, laterally, 
against the wall of the segment. They fill up the greater 
part of the middle third of the region. They commence 
anteriorly, just behind the orifice of the retracted claspers, 
and stretch back to a point, on a line behind the vesicule 
on either side. Dorso-ventrally they extend from the border 
of the dorsal longitudinal muscles round the sides of the 
segment to the border of the ventral bands. ‘Their cells 
have a waxy appearance, and their cytoplasm, which is rela- 
tively large in amount, is very finely granular. Each cell 
possesses a round nucleus and a dark karyosome. On a line 
with the vesicule the dorsal cell of each group gives off a 
fine duct, that crosses the space of the testis cavity and runs 
into the ventral external surface of the vesicle of the same 
side. The wall of this tube is also, hke the protoplasm of the 
cells of the gland, finely granular. About its middle there 
are usually two large nuclei embedded in the wall. Where 
the tube runs round the outer surface of the vesicule it is 
much thickened, and this appears to be due to the accumulation 
of drops of the gland secretion in its Jumen (fig. 31). 

The vesiculx are roundish bodies with thin walls. The 
lateral and ventral third of their cavities is taken up with 
the mucous secretion derived from the granule glands. This, 
in sections of fixed material, projects upwards into the cavity 
ina mass of finger-like digitations. On the outer ventral 
surface of each vesicula there is a small slit. Its edges are 
armed with short stout cilia. It was the motion of these 
that Foettinger evidently mistook for the presence of a pair 
of excretory organs in this region. Through this slit the 
spermatozoa gain an entrance into the vesicule. 

The vas deferens leads out from the posterior ventral 
portion of each vesicle and turns in towards the median line, 
and is continued as a small tube to the base of the penis. It 
is of considerable diameter, and forms a‘sac-like canal on 
either side. At the base of the penis the vasa deferentia 
of both sides meet, forming a smallreceptaculum seminis, 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. oon 


which lies between the two lateral halves of the organ. 
During life this is always full of very actively moving 
spermatozoa. 

The penis is a firm, semi-solid, pear-shaped body, the 
pointed end being directed backwards. It is always carried 
retracted within the sheath. Unlike Stratiodrilus, it is 
not composed of black chitinous material similar to that of 
the jaws, but of some transparent substance, sufficiently 
rigid, however, to enable its being driven through the firm 
cuticle of the female in the act of copulation. It is pro- 
truded through the quadrilateral-shaped mouth of the penis- 
sheath by the action of the strong protractor muscles. The 
organ itself is composed of two lateral blades, the spermatozoa 
being ejected through the median canal between them during 
copulation. 

In relation with the dorsal surface of the penis on either 
side, and taking up the lateral posterior corners of the 
generative region, are the so-called accessory glands of the 
penis (figs. 11 and 40). ‘These are large vacuolated groups 
of cells forming oval masses running up to the dorsal surface 
on either side of the gut. From each gland a small duct 
leads down to the penis, and is inserted laterally about its 
middle. This opens into the canal on the penis on its ventral 
side. The gland-cells are divided into an anterior and pos- 
terior group. In horizontal sections the gland appears as a 
four-lobed structure, posterior and dorsal to the base of the 
penis. The anterior and smaller of these groups is composed 
of numerous cells, while the posterior, although larger, consists 
of fewer cells. The cytoplasm is granular and very vacuolar. 
This is shown in fig. 40, where their anterior ends come in 
the section on either side of the gut. In transverse section 
the gland will be seen to be composed of two groups of 
cells, one of which is much smaller and more dorsal than 
the other. This is wedged in against the gut on either side. 
Towards the posterior region of the gland the cells are 
somewhat larger. ‘The largest of these contains a vacuole 
of considerable size. This probably acts as a receptacle for 


332 CRESSWELL SHEARER. 


the gland secretion. It is connected with the penis by a 
strand of cytoplasm that runs to its ventral side, and is con- 
tinuous across the median line with a strand from a similar 
cell from the opposite side. Posterior and ventral to this 
are a number of small, darkly staining cells. They are 
lenticular in shape, with prominent nuclei. They fill up the 
corners between the large cells. The largest cell of the 
gland is placed about the middle or slightly towards its 
posterior end. ‘The section shown in fig. 24 passes just 
behind its posterior border. The nuclei of the smaller cells 
are rod-shaped, and frequently bent in a semi-circular form. 
On the inner wall of the gland, close to where it abuts against 
the penis-sheath, are a number of darkly staining masses of 
nuclear material. The ends of all the gland-cells converge 
on the penis. When the cells are charged with secretion 
their nuclei are seen to be large and round, with a well- 
marked karyotheca. The karyoplasm is collected into a 
darkly staining karyosome. In the cells that have dis- 
charged their secretion, on the other hand, the nuclei are 
invariably long and rod-shaped, with a uniformly staining 
karyoplasm, and no karyosome. 

In fig. 23 are represented some of the cells of the posterior 
group under high magnification. The cytoplasm forms a 
superficial layer which throws threads across the vacuolar 
interior of the cell. The nucleus is always situated about the 
middle of the cell and is of considerable size, and contains a 
darkly staining karyosome. 

The compartment of the generative region holding the 
glands is sharply separated from the anterior part of the 
segment, which contains the testis, as already explained. 
This is clearly separated from the granule cells and the 
vesicule, which are within the limits of the testis proper, 
and enclosed by its membrane. 

The region of the accessory gland is often seen distended 
with the accumulation of secretion within the gland. With 
dark ground illumination this appears opaque and whitish in 
colour. In the surface view of a living preparation the two 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 3338 


portions of the gland appear somewhat as shown in fig. 11. 
The anterior lobe seems distinctly separated from the pos- 
terior. The function of these glands is problematical. They 
doubtless pour some secretion into the canal of the penis 
during copulation, which assists in this act in some way. 

They were first described by Haswell in Stratiodrilus, 
where they are much larger and somewhat different in appear- 
ance from those in Histriobdella. They seem to have been 
overlooked by Foettinger, although he plainly figures them in 
his sections. He evidently mistook them for a portion of the 
testis. That they are separate structures from this can be 
easily seen in horizontal sections. They correspond to the 
similar glands found in connection with the male organs in so 
many Turbellaria, as in Proxenetes, Provortex, and 
Plagiostoma. 

Under the heading of the male reproductive organs come 
the claspers. These are usually carried retracted, only being 
protruded when the males are impregnating the females. 
Under the action of strong reagents during fixation they are 
sometimes extended, in which case they are always seen pro- 
jecting ventralwards and never laterally. Each clasper is 
furnished with a protractor and a retractor muscle that runs 
to the base of the penis, as already explained. At the base 
of each organ there is a large mucous cell with a large nucleus. 
This, in the retracted condition, occupies the anterior wall of 
the clasper-sheath, and is a conspicuous feature in a trans- 
verse section through the anterior region of the generative 
segment. In a full-grown male the cell is very large. A fine 
duct leads from it to the tip of the organ and pours some 
adhesive secretion on the surface of the clasper, similar to 
that poured on the surface of the feet. This cell is shown in 
fig. 2. ‘The anterior lip of the orifice formed by the retrac- 
tion of the organ forms a marked projection which overlaps 
the orifice (fig. 13). When the organ is extended this lip is 
obliterated, as shown in fig. 9. ‘The gland cell then occupies 
the middle of the clasper. At the top of the organ there are 
a few short, stiff hairs. I have already mentioned that once 

VOL. 95D, PART 2,—NEW SERIES. 22 


334 CRESSWELL SHEARER. 


the male has seized the female by means of the claspers its 
grip is immediately rendered secure by the gland secretion, 
and then the male is only able to free itself from the female 
with difficulty. Sometimes the male can be seen being carried 
about by the female, making violent efforts to free itself. 
The claspers never seem to be used for any other purpose than 
that of seizing the female, and are never extended to enable 
the animal to hold more securely when an attempt is made 
to brush them off the lobster ova. 


B. In the Female. 


The ovary in the female holds the same position in the 
generative region as the testis in the male. It has a more 
sac-like appearance, however, and its lining membrane is 
thicker than in the case of the testis. In the anterior and 
posterior regions of the segment there is not the great 
thickening of the wall seen in the male. It is more uniform 
in thickness, and the contour of the limiting membrane 
throughout more distinct. In sagittal sections in the median 
line it appears as a long chamber lying ventral to the gut 
(fig. 22). 

Foettinger’s account of the oviduct and funnel is correct, 
the funnel being large and collapsible, ciliated on its dorsal 
side only. It projects downwards into the ventral region of 
the middle third of the generative segment. Its ventral lip 
is a short distance from the nerve-cord on either side; its 
dorsal lip is the longest, and almost meets that of the 
opposite side in the median line. The funnel is composed of 
a large number of flattened cells, a conspicuous one being 
usually seen in the edge of the dorsal lip. The cilia are 
remarkably stiff and short. It leads into a small, round 
ampulla which is usually crowded with spermatozoa. This 
leads into a still larger one, the walls of which are drawn out 
in a number of digitations. This functions as a sort of shell- 
gland. Its lumen is filled with a granular secretion that 
forms the egg-capsule. A short canal Jeads from the second 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. oo 


ampulla to the exterior. When the ovary is full of ova it is 
sometimes difficult to see the funnel and oviduct, as its 
lumen is obliterated by compression against the body-wall. 
The inner surface of the ovary is closely invested with a 
thin layer, of nucleated cells—the true coelomic epithelium. 
It is from this in the anterior region that the primitive ova 
arise. ‘his takes place close to where the sac abuts against 
the end of the third segment. Here certain of the nuclei are 
much larger than the rest. They are the odgonial cells. 
They have relatively little cytoplasm and large, transparent 
nuclei. A considerable number of them are seen at this 
point in different stages of development. The fact that the 
o6gonia arise from a small, circumscribed portion of the 
anterior end of the ovary, and not from its epithelial surface 
in general, recalls the condition described by Nelson (25) in 
Dinophilus conklini, which differs from the other species 
of this group, D. vorticoides, D. teniatus,and D. gigas, 
in that only a small portion of the ovary likewise gives rise 
to the odgonial cells. It is evident that the epithelium of the 
middle and posterior portions of the ovarian cavity play no 
part in their formation. As they pass backwards and become 
the primary odcytes, the epithelium of this part of the cavity 
throws out processes that attach themselves to the growing 
oécytes, folding up round them and forming a supporting 
matrix crowded with small nuclei. ‘They furnish them with 
the material for their growth, but beyond this take no part in 
their formation. As the odcyte grows these follicle cells 
diminish rapidly in size, and their nuclei undergo degenera- 
tion, becoming long and granular. ‘They appear to have 
something to do with the formation of the yolk-granules, but 
how this is accomplished is not plain. These arise in situ, 
as nothing similar to them can be distinguished in the follicle 
cells, which are always clear and transparent. At the time 
of their formation the granules are also clear and transparent, 
and only acquire their dark appearance after they have been 
formed some time. For this reason the {small odcytes, 
although highly granular, are almost as transparent as the 


336 CRESSWELL SHEARER. 


odgonial cells. By the time the odcytes reach the middle of 
the generative region they turn dark brown in colour. In a 
few days they increase greatly in size. Their outline becomes 
regular, and the superficial layer of their cytoplasm seems to 
stain much more intensely than the deeper portion. Their 
nuclei become large, round, and transparent, and are readily 
distinguishable in the living animal. ‘There appears to be no 
yolk-nucleus present, but the germinal nucleus goes through 
a number of changes during the formation of the deutoplasm, 
that probably has to do with the great elaboration of this 
material. 

The mature eggs are found in the posterior region, where 
they take up the greater part of the ovarian chamber. They 
measure from 80-200, in their longest diameter, according to 
the size of the female. They are oval in shape and somewhat 
flattened. They are highly granular, the granules being very 
uniform in size. 

Unlike Stratiodrilus, there may be a number of ripe 
eges within the chamber at one time, although one usually 
predominates in size over the others. In the violent move- 
meuts of the animal small fragments of the egg are some- 
times broken off by compression against the gut, or from 
friction against the other eggs of the cavity. These are seen 
to move about the cavity quite freely, and, by some peculiar 
cohesive process, are capable of joining up with the egg 
again. This can be seen taking place under the microscope. 
The fragments have a membrane of their own, and may be 
seen lying against the egg from which they have separated. 
The membrane between them breaks down, and they flow 
together rapidly. 

Normally the ripe ovum is almost divided in two portions 
by its compression against the gut. When a ripe ovum is 
discharged its place is immediately taken by the next in size. 
I have never actually observed the female in the act of 
depositing her eggs; as I have mentioned, this takes place 
usually at night. From the fact that the funnel in the female 
is well forward at the generative region and the ripe ova are. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 337 


sometimes far back at the caudal end of the ovary, they have 
to move some considerable distance forward before they can 
find exit through the oviducts. In passing through the 
second ampulla the egg is surrounded by its capsule, which 
binds it firmly to the membranes of the lobster’s ova, or the 
hairs of the carapace surface of the branchial chamber. 

I have already drawn attention to the fact that the odcyte 
commences to prepare for the first maturation division and 
the extrusion of the polar bodies when it has acquired only a 
portion of its yolk material. It is noteworthy that only one 
of the odcytes undergoes this change at a time. It is the 
most advanced and the largest. While the amphiaster is 
seen in this egg, I have never observed it in any of the 
younger ones, although some of these to all external appear- 
ances are as large and as mature as the one in which it has 
appeared. 

As the odcyte prepares for maturation its staining reaction 
changes. Up till this time the superficial layer of its cyto- 
plasm stains darkly, while the deeper portions surrounding 
the nucleus do not take the stain. With the appearance of 
the maturation spindle the staining reaction of the cytoplasm 
becomes uniform throughout the cell. 

The first sign of approaching maturation is announced by 
the changes undergone by the nucleus. It is distinguishable 
in the living egg as a clear spot in the middle of the dark 
granular cytoplasm. By a number of changes, which I have 
not followed in detail, the chromosomes form, the germinal 
vesicle breaks down, and the amphiaster of the first polar 
body forms. This at the moment it appears is very small, but 
erows rapidly with the growth of the egg. From the time it 
appears to the time it reaches its full dimensions it at least 
trebles its length, while the egg grows considerably in size. 
From measurements made of the length of the central spindle, 
from centrosome to centrosome, and the diameter of the egg 
in its longest axis, it was found that from the time the central 
spindle was clearly visible to the time it ceased to grow it 
trebled its length, while the egg a little more than doubled 


338 CRESSWELL SHEARER. 


its longest diameter. The spindle seems to grow with the 
egg. ‘lhe size of the amphiaster is always proportional to 
that of the ovum. In the large female, where the eggs are 
almost double the size of those of the small ones, the spindle 
is correspondingly larger. ‘he size of the spindle is appa- 
rently determined by that of the cell. 

In Limulus, according to Munson (23), the growing centre 
of the egg is the vitaline body. ‘This, in the early stages, 
presents all the appearances and features of the centrosome 
and sphere, and, in fact, is the centrosome of the dividing 
oogonia. In later stages it remains as the definite centrosome 
in the cytoplasm. ‘Thus it appears as the primitive basis or 
centre of growth of the cytoplasm, building this in part from 
the granules supplied by the follicle-cells. In Histrio- 
bdella growth does not seem confined to the region near the 
umphiasters, but seems to take place generally throughout 
the cytoplasm of the egg. No yolk-nucleus or vitaline body 
is present. In sections of fixed eggs the cytoplasmic material 
in the immediate vicinity of the spindle is markedly less dense 
than in the peripheral region of the ovum. In some sections 
the middle of the ovum appears as a space, in the middle of 
which is the spindle with its chromosomes. 

The ovum goes through a portion of maturation during the 
time it is still adding material to its cytoplasm. While the 
achromatic threads of the amphiaster can be readily seen in 
the living egg, the chromosomes cannot be detected without 
staining. At the end of the prophase eight chromosomes are 
found in the equatorial plate of the spindle. 

The astral rays are much less definite than the strands of 
the central spindle. While the former seem in the living egg 
as if due to the arrangement of the yolk-granules in definite 
lines, the latter appear as actual threads running between 
the granules themselves. In speaking of the astral rays 
Wilson (82) says: ‘A careful study of their relation to the 
meshwork in the Echinoderm, and in many other forms 
(especially in Nereis, Thalassema, Lamellidoris, and 
Asterias), leaves no doubt in my opinion that they are actual 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 339 


fibrille, that thread their way among the crowded alveolar 
spheres. In my best preparations the astral rays appear like 
wires bending to and fro among the alveoli” (p.13). “From 
a study of 'oxopneustes one would be led to the conclusion 
that they ,arise in rows of granules or microsomes, held 
together by the continuous substance” (p. 15). These words 
exactly describe the appearance of the astral rays in the 
living ege of Histriobdella. 

Towards the centre of the astral figures the rays appear as 
continuous fibres, while peripherally they break up into rows 
of granules. I believe in both the asters and the central 
spindle the granules do not build up the achromatic figure, 
but are merely incidental] to it. This is borne out by the fact 
that they are less numerous within it than in the surrounding 
cytoplasm. For this reason the area of the amphiaster in the 
living egg is always the most transparent. The archoplasm 
ean be distinctly seen as a clear substance running between 
the microsomes. 

The less dense nature of the astral rays, as compared with the 
fibres of the spindle, has been clearly demonstrated recently by 
Lillie (18) on centrifugalised eggs, where the egg-granules are 
readily driven through the substance of the astral rays, while 
they are stopped and forced to go round that of the spindle. 

The chromosomes in Histriobdella are arranged round 
the periphery of the equatorial plate. Hach chromosome lies 
directly against one of the spindle-fibres. These run from 
one centrosome to the other without any break in their con- 
tinuity. It is obvious that the chromosomes have no proper 
mantle-fibres, and that the number of fibres composing the 
spindle is in excess of that of the chromosomes. In sections 
the number of fibres can be counted. There are twenty, 
while there are only eight chromosomes. 

The centrosome itself is not distinguishable as a distinct 
point or granule in the living egg, but its position is indicated 
by a small area where the fibres of the astral rays and those 
of the spindle all converge on one another. No sphere can 
be distinguished. 


340 CRESSWELL SHEARER. 


In the early stages, during the formation of the central 
spindle, its fibres in part appear to arise outside the area of 
the nucleus. In one instance I was able to distinguish the 
spindle-fibres beyond the still evident remains of the nuclear 
wall. ‘Ihe centrosome clearly arises beyond the limits of the 
nucleus, and from the reticulum of the cytoplasm, and its 
presence can be clearly detected before the dissolution of the 
nuclear wall. 

Much has been written on the origin of the spindle and the 
centrosomes as to whether they are of nuclear or cytoplasmic 
origin. It has been established that the spindle-fibres may 
arise from either. In the case of the mantle-fibres they arise 
almost invariably from the nucleus, while the spindle sub- 
stance proper arises from the cytoplasm, as has been shown 
by Meves (22) in Salamandra, Calkins (8) and Ishikawa (17) 
in Noctiluca, Flemming and Heidenhain (14) in leuco- 
eytes. In cases where no central spindle is present the 
astral rays seem to arise from the cytoplasm, as in a number 
of plants, some worms, as ‘l'halassema, according to Griffin 
(11), and in a number of Annelids as described by Mead (21). 
In other cases from the nucleus, according to Flemming (7), 
Rickert (27), Wilson (83), and Korschelt (18). 

According to Watase (81) the centre of the aster is merely 
the point where the greatest number of cytoplasmic filaments 
meet, the centrosome thus produced giving rise in turn to the 
spindle filaments. Thus the spindle-tibres originate from the 
centre of the aster, and not from the nucleus. This is clearly 
shown in the case he instances of the blastomeres of Loligo, 
where the nucleus remains a clear area in the middle of the 
central spindle. ‘There is a short period in the formation of 
the spindle in Histriobdella when almost the same con- 
ditions are shown. Again, the observations on eggs that 
have been artificially fertilised by salt solutions clearly point 
to the origin of the spindle quite independent of the nucleus. 
According to Wilson (84) all degrees exist between the asters 
that lie remote from the nucleus and of undoubted cyto- 
plasmic origin, and those close beside it. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 341 


When the amphiaster attains the prophase, it remains in 
this stage until the egg is fertilised and deposited in the sea- 
water. If this does not take place, or if the conditions for 
egg-laying are unfavourable, it apparently remains in this 
state indefinitely, not making any further progress. 

In one instance I was able to keep a large female under 
observation for the greater part of a week with the amphi- 
aster of its largest egg in the prophase. At the end of this 
time the fibres of the central spindle and the astral rays were 
as distinct as at first, and showed no evidence of dissolution. 
It is evidently contact with the sea-water that 1s necessary 
to cause the completion of maturation and the extrusion of 
the polar body. 

The spindle is of considerable size, measuring from 50- 
60 x from centrosome to centrosome. It can be readily seen 
in the living egg with the aid of a good hand-lens. As the 
animal moves and the egg outline is changed by compression 
against the body-wall, it does not change the position of its 
main axis with regard to that of the egg. According to 
Hertwig’s well-known law, as the result of the interaction of 
the nucleus and protoplasm the spindle comes to lie in such 
a position that its longitudinal axis corresponds with the axis 
that passes through the greatest protoplasmic mass. In figs. 
18-20 are shown the position of the amphiasters in the egg 
as it has undergone change. The axis of the spindle, it will 
be seen, does not always correspond with that of the main 
axis of the egg, but on the whole it lies very close to this, 
and the cytoplasm always shows a tendency to group itself 
symmetrically about the spindle. I have made a number of 
observations that seemed to show that the form of the egg 
does not greatly affect the direction of the spindle-axis. 

In fig. 32 is shown the egg when it has undergone con- 
siderable pressure in its long axis through contraction of the 
animal. The spindle shows no appreciable shortening as the 
result of this pressure. In fig. 18 the egg shows the com- 
mencement of two furrows running into the cytoplasm, due 
to compression against the gut. In fig. 32 a small portion 


342 CRESSWELL SHEARER, 


has been broken off the posterior end. This subsequently 
joined up with the egg again. 

No polar body is given off by the egg while it remains 
within the cavity of the ovary. Ihave had a female under 
observation for several days, and have been able to follow the 
growth and maturation of a particular egg from the first 
without seeing the formation of any polar body taking place. 

I have mentioned that the male is often seen to fertilise the 
female while she is without eges and still immature and in the 
larval state. In these females the sperm can be seen working 
their way through the tissues and finally collecting im the 
oviduct. I believe this invariably takes place. Whether 
the sperm, once in the oviducts, retain their vitality till the 
female reaches maturity aud bears eggs I have been unable to 
determine. It would seem that it is immaterial whether this 
does or does not take place. ‘he female is usually fertilised 
over and over again before she reaches maturity and bears 
eggs, so that fertilisation is probably effected by the last supply 
of sperm she may happen to receive. It is clear that the 
presence or absence of ova in the female play no part as a 
factor in fertilisation. 

No matter where the sperm are injected into the body of 
the female 


and the male exercises no choice in this respect— 
they seem to collect ultimately in the ampulle of the eviducts. 
It would seem as if some substance in this situation exerted a 
chemotactic influence over their movements, causing them to 
collect here from all parts of the body. 

The sperm are frequently seen in the blastoccelic cavity in 
small masses beneath the gut. In this situation they are still 
shut off from the cavity of the ovary and the eggs. 

In the anterior end of the ovary, crowded among the small 
odgonial cells, are frequently seen small masses of sperm. 
‘hese appear to have undergone considerable change and to 
have partially lost their tails. I[t is probable that these sperm 
have gained access to the ovary by way of the oviducts. It 
is remarkable, however, that in the posterior region of the 
ovarian cavity no sperm are seen free among the ova, but they 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 343 


would seem to be confined to its anterior region. Fertilisa- 
tion takes place within the ovarian chamber, as a large 
oblong sperm-nucleus is always found in the ovarian egg, in 
which the amphiaster has appeared. This always lies at some 
distance from the spindle and close to the egg-membrane, 
while the spindle is centrally placed. I have been unable to 
determine at just what stage in the growth of the ovarian 
egg fertilisation takes place. As the egg is seen to increase 
considerably in size after the amphiaster has appeared, and as 
the sperm nucleus is always found in the ovum when this 1s 
present, it is possible that the egg is fertilised at a stage in 
which the yolk-granules are first beginning to appear, hie 
fusion of the pro-nuclei takes place only after the polar body 
is extruded, and this takes place when the egg has been 
deposited in the sea-water. 

In Stratiodrilus Haswell has observed the fertilisation of 
the egg taking place within the ovarian cavity. 

In Dinophilus teniatus, according to Harmer (12), the 
same conditions hold regarding impregnation and fertilisa- 
tion asin Histriobdella. The penis is inserted anywhere 
under the skin, the act of copulation taking place repeatedly 
with the same female. He says, “ the act of copulation has 
no relation to the maturity of the ova of the female, nor is it 
prevented by the fact that the female has already received an 
ample supply of spermatozoa by a preceding operation” 
(p. 13). The spermatozoa can be seen collected in small 
masses beneath the gut. Fertilisation is therefore internal. 
The polar bodies are given off apparently when the eggs reach 
the exterior, or shortly after they are deposited in the sea- 
water. 

In the ripe egg, after the amphiaster of the first polar 
body has been formed in this manner, a remarkable occur- 
rence can be brought about, which demonstrates most clearly 
the semi-solid nature of the spindle itself. In compress- 
ing the cover-glass on a preparation of a living parasite 
I happened in several instances to rupture the body-wall in 
the vicinity of the ovum. The egg-envelope was also broken 


B44 CRESSWELL SHEARER. 


at the samé point. The yolk-granules then rapidly poured 
through the opening into the sea-water, and carried the 
amphiaster with them. It held together asa semi-solid body, 
and could be seen turning over and over as it was pushed 
along by the granules. Once in the water outside the body 
the granules tended to disperse, while the amphiaster remained 
with its immediate surrounding granules, apparently a solid 
body. It remained like this for several minutes until it finally 
dissolved and disintegrated. I have tried to represent this 
taking place in fig. 8. The asters go first, while the central 
spindle still remains intact. This seems to show that the 
substance of the spindle is of firmer texture than that of the 
asters and centrosphere ; and this is borne out, as I have 
mentioned, by the actual appearance of the archoplasmic 
substance of the asters as compared with the sharp, definite 
structure of the spindle. The yolk-granules adhere and 
seem almost a part of the archoplasmic substance of both 
asters and spindle, the amphiaster really appearing as a mass 
of brown yolk-granules held together by the thread-like 
archoplasmic substance. As the spindle begins to dissolve 
the yolk-granules can be seen being liberated from the trans- 
parent substance of the archoplasm and moving away in the 
sea-water. In fact the whole process of the dissolution of 
the amphiaster, as seen under an oil-immersion lens, is similar 
to that of some gelatinous substance slowly dissolved by the 
action of sea-water. That the spindle has some considerable 
rigidity is borne out by the fact that it keeps its shape, and 
can be seen rolling over and over as it is drawn along in the 
sea-water. It shows no tendency at first to flatten under the 
pressure of the cover-glass. ‘This is always considerable, 
although its corners are supported as much as possible by wax 
feet, as the capillary attraction invariably draws the cover- 
glass down somewhat in the middle. It is not till the spindle 
has begun to dissolve that this rigidity is lost, when it under- 
goes flattening. It at the same time becomes more trans- 
parent, the archoplasmic threads appearing as if actually 
undergoing dissolution by the sea-water, leaving the dark 


ON THE ANATOMY. OF HISTRIOBDELLA HOMARI. 345 


yolk-granules behind them arranged in positions that had 
previously been held by the archoplasm. There is a short 
period during which the spindle almost remains alone, the 
asters having completely disappeared from either end. It is 
at this time that the spindle can be seen to roll over as it is 
pushed farther and farther away from the point of rupture 
in the body-wall by the escape from the egg of fresh cyto- 
plasm. 

In a uumber of experiments I subsequently ascertained 
that this rupture of the egg and extrusion of the spindle will 
not occur if the egg is far back in the body-cavity. The 
body-wall ruptures at its thinnest part, which is well forward 
in the generative region. If the egg has to move forward 
some considerable distance under pressure, before it can 
begin to flow through the rupture the amphiaster is usually 
broken and destroyed. It takes place most satisfactorily 
when the egg is only a short distance from the point of 
rupture of the body-wall. It can only be observed to occur 
when the amphiaster itself is fully mature. When not fully 
formed it dissolves immediately any movement of the yolk- 
granules takes place. Unless, moreover, the rupture in the 
body-wall is fairly large, the amphiaster is usually broken 
in the act of being forced through, being destroyed by the 
granules pushing it through from behind. 

In one instance the central spindle had the appearance of 
being composed of a mass of distinct threads, some of which 
on one side of the spindle had been injured and broken, 
the yolk-granules appearing as small grains entangled in 
these fibres. 

I think this observation clearly demonstrates the truth of 
a suggestion that has been put forward, that the achromatic 
threads and amphiasters are firm structures, or at least more 
rigid than the reticulum of the cytoplasm. Gardiner (10), 
in his paper on the egg of Polychcrus caudatus, states 
(p. 89), “That the amphiaster is much more rigid than the 
surrounding cytoplasm is shown by two instructive prepara- 
tions which were the result of accident. Ova containing 


346 GRESSWELL SHEARER. 


amphiasters in the stage now under discussion were ruptured 
just before the worm containing them was placed in Hermann’s 
fluid. The cytoplasm had flowed or been pressed out of the 
ovum, carrying with it the amphiaster. In both cases the 
cytoplasmic network had been completely bent and twisted 
into a confused snarl. The achromatic rays were somewhat, 
but not nearly so much distorted, but the centrospheres were 
almost unchanged. From this I infer that the amphiaster 
and the rays are, on the whole, much more rigid than the 
cytoplasmic network or the cytoplasm from which they are 
formed.” Evidently the same thing took place in this 
instance as I have observed in Histriobdella, where the 
large size of the spindle and the granular nature of the egg 
renders the various steps in the process clearly visible under 
the microscope. 

By pricking the egg-membrane of Allolobophora, Foot 
and Strobell (9) have been able to get the egg contents on 
the slide, and there photograph it after fixation. “ By this 
method the germinal vesicle, and sometimes even the spindle, 
flow out of the egg-membrane intact” (p. 201). Some 
excellent photographs are shown of these in figs. 125-130 
of this paper. In Allolobophora, as in Histriobdella, 
the early stages of the first maturation division are gone 
through by the egg while it is still within the receptacula 
ovorum. 


10. ConcLUSION AND SUMMARY. 


Harmer (12) was the first to point out that Histriobdella 
was more closely related to Dinophilus than to any other 
Archiannelid, although Pierantoni (26), in his recent mono- 
graph, has placed Histriobdella and Dinophilus as an 
appendix to the Polygordide (including Protodrilus). 
Schimkewitsch (28) has contended that Dinophilus is closely 
related to the Rotifers, and Haswell (18) has put forward a 
similar claim for Histriobdella. In Histriobdella it is 
certain that the parasitic mode of lite has resulted in a 
peculiar specialisation, which, combined with its direct mode 


ON THE ANATOMY OF HISTRIOBDELLA HOMART. 347 


of development, renders its relationship hard to determine, 
and hides the primitive characteristics of its organisation. 
That the Rotifers themselves are likewise a highly specialised 
class of somewhat uncertain affinities is an opinion that is 
gaining ground, since so much doubt has been thrown on their 
supposed relation to the Annelid trochophore. The work of 
Wesenberg-Lund (20) has shown that the most simple and 
trochophore-like of the Rotifers are probably the most highly 
specialised and the farthest removed from the Annelids. Yet 
the clearly segmented plan of both Dinophilus and His- 
triobdella, it must be admitted, is essentially similar to that 
of a Chetopod. This, combined with the clearly Polychet 
nature of egg-segmentation in Dinophilus, is sufficient to 
place these forms in direct connection with the Annelids, 
quite apart from either Protodrilus or Polygordius. 
Under the heading of the various organs I have already 
gone into a more or less detailed comparison of Histrio- 
bdella with Dinophilus, so that it is only necessary to 
review the subject here from a more general standpoint. 
In both forms the animal consists of a distinct head and 
trunk, the latter composed of relatively few segments. In 
both the nervous system consists of a well-defined brain or 
neuropile, and a double ventral nerve-cord, with metameri- 
cally arranged ganglia. In Dinophilus these are formed 
by transverse commissures, while in Histriobdella the two 
parts of the cord unite directly to form the ganglia. The 
external segmentation corresponds with that of the nervous 
system. Dinophilus does not possess the feet, cirri, or 
tentacles that so clearly mark segmentation in Histrio- 
bdella. But the metamerism is less definitely shown by the 
ciated bands, mucus glands, and the ring-like constriction 
of the body into a series of segments. On the other hand the 
nephridia show a more matamerically placed arrangement than 
they do in Histriobdella. In both (with the exception of 
Stratiodrilus) the nephridia open to the exterior in the seg- 
ment followimg that in which they arise, asin Annelids. In 
Histriobdella the muscular system shows a very high 


348 GRESSWELL SHEARER. 


degree of development, and for this reason can hardly be 
compared with that of Dinophilus; in both, however, the 
main musculature consists in a series of longitudinal ventro- 
lateral and dorso-lateral muscles. The alimentary canal shows 
the same divisions, although differing considerably in the 
relative proportion of its parts. The strong chitinous jaws 
are wanting in Dinophilus. In each the cavity surround- 
ing the gutis a primitive blastocele with no definite epithelial 
lining. This cavity sends prolongations into the head. The 
equivalent of the ccelom in both is represented by the cavity 
of the reproductive glands. In the male these consist of a 
more or less paired testis, vesicule, vas deferens, and 
median penis, and in the female a large ovarian cavity, 
paired or unpaired, with oviducts. 

With Annelids Dinophilus shows a closer relationship 
than Histriobdella, mainly due to its less direct develop- 
ment. In fact the development of Dinophilus brings it 
into line with that large group of animals such as the Poly- 
chete, Echiuride, Gephyrea, Lamellibranchs, and 
the Gasteropoda, in having the ectoderm arising from the 
first three quartettes, mesoderm from the left posterior cell of 
the fourth quartette (4 p.), and the endoderm from the remain- 
ing cells. In the derivation of a large part of the ectoderm 
of the trunk from the posterior cell of the second quartette 
the resemblance to the Polychzt Annelids is most pronounced. 
This is further enforced in the origin of the bilateral cleavages 
in the cross cells and in the products of 2p. ‘The transition 
from the spiral type of cleavage to the more specialised 
bilateral type occurs in precisely the same directions as in 
the Polychets. Moreover the second bilateral divisions of 
the cells of the posterior arms of the cross continue this 
resemblance. All these characters, if such they may be 
called, when viewed as a whole point in no uncertain way - 
to the descent from the Annelid stem, and at a point not far 
from that at which the Polycheta arose ” (Nelson, p. 728). 

The weight of our evidence, furnished by recent work on 
the morphology and embryology of Dinophilus, is strongly 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 349 


in favour, therefore, of a close relationship with Annelids. 
The general ciliation, the caudal appendage, ciliated rings, 
nervous system, general configuration of the head, trunk, and 
alimentary canal are what are found in a number of Annelids, 
and most clearly in such a form as Ophyotrocha. Nelson 
(25) has even suggested that the pre-oral nerve commissures 
can be satisfactorily explained by deriving them from the 
nerve-ring of the Trochopore. He comes to the conclusion : 
“ On the whole, Dinophilus can best be considered as a very 
young Polychet worm, retaining some of its larval features, 
with sete and parapodia undeveloped, and whose peritoneum 
and ccelom have been transformed into a generative organ” 
(p. 135). 

The relationships of Histriobdella to Polygordius and 
Protodrilus have been gone into fully by Foettinger (8), 
Harmer (12), and Haswell (13), so I need not repeat their 
arguments for this relationship here. It seems to me, from 
the Archiannelid point of view, it is important to determine 
what features of Histriobdella are primitive, and what 
have been derived from its peculiar mode of life. Hisig (5) 
has gone so far as to suggest that in Histriobdella we 
have to do witha highly modified, possibly degenerate animal, 
and not an Archiannelid at all. If Histriobdella is a 
degenerate form then it must be a degenerate Chetopod as 
Haswell (18) has pointed out. “If we are to take this view, 
we must at the same time acknowledge that side by side with 
the supposed degeneration, there must have gone on a special 
development in certain directions; that, while the definite 
characters of the segmentation became lost, a special set of 
locomotor organs with an elaborate musculature became 
evolved.” “'This view appears to me to involve difficulties 
so great that they render the degeneration theory extremely 
improbable, and it seems to me more in accordance with the 
facts of the case to conclude that the Histriobdellide are 
really primitive Annulates, and that the rudiments of their 
specialised features have been inherited from forms lower in 
the scale” (p. 327). 

VOL. 55, PART 2.—NEW SERIES. 25 


3590 CRESSWELL SHEARER. 


Apart from any degeneration I agree with Haswell (18) 
that the relationship of Histriobdella with Polygordius 
“is extremely remote, and not such as to justify their inclu- 
sion in the same class.” The absence in Histriobdella of 
a blood-vascular system, a distinct prostomium and _ peri- 
stomium, the presence of mouth opening well forward in the 
head, chitinous jaws, and complicated generative apparatus 
in the male, paired limbs, and mucous glands, clearly 
separate it from Polygordius and Protodrilus, placing 
it quite apart from these forms, With the Rotifers, on the 
other hand the relationship is undoubtedly more pronounced, 
Haswell has pointed out that all the main features of His- 
triobdella can be traced to this class, although in general 
features the resemblance is greater perhaps with the 
Gastrorichia than with the Rotifers proper. The 
chitinous jaws of Histriobdella can be readily homo- 
logised with the mastix of Rotifers. In the absence of 
solenocytes and the general similarity of the nephridia of 
Histriobdella to the flame-cell type nephridia of Rotifers, 
we have a further resemblance, In both the cuticle is firm 
and shows a tendency to contract into ring-like folds. In 
both, also, the generative organs, especially in the male, can 
be reduced to the same plan. 

In Paraseison we have a Rotifer not unlike Histri- 
obdella in many of its features. The body is elongated and 
worm-like, with a distinct head bearing the mouth at its 
anterior extremity. In the middle of a very rudimentary 
coronal disc which bears no ciliated apparatus are four small 
bundles of hairs, placed in two pairs. Behind the mouth are 
found the orifices of two glands, similar to those found on the 
anterior feet of Histriobdella, On the top of the head is 
a small tubercle representing the dorsal median tentacle of 
Histriobdella, There is a narrow cesophagus, which leads 
into a large cylindrical stomach. There is no gut, and the 
stomach, which is not ciliated, is definitely closed. But this 
condition has plainly been evolved within the limits of the 
genus, as it is not characteristic of other Rotifers, It is 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 351 


noteworthy that Paraseison, like Histriobdella, is para- 
sitic, beimg found on the gills of the Crustacean Nebalia. 
It is undoubtedly with such forms as Paraseison among the 
Rotifers that Histriobdella must be compared, The 
oreatest objection to the comparison of Histriobdella with 
the Rotifer is encountered in regard, to the nervous system. 
This in Histriobdella is already so elaborated, and of that 
type found among the higher Annelids, as to be hardly com- 
parable to the diffuse, and less differentiated, and centralised 
system of Rotifers. 

I cannot agree with Haswell that Zelinka’s (86) discovery 
of a sub-cesophageal ganglion in Callidina and Discopus 
renders this comparison more easy. A further difficulty is 
found in the absence of any true metamerism in the Rotifers. 
This difficulty is possibly not so great when we consider the 
arrangement cf the transverse muscle-cells in such a rotifer 
as Discopus synaptex. Leaving aside any comparison, 
therefore, of the nervous system, it nevertheless remains a 
fact that Histriobdella undoubtedly resembles the Rotifers 
more closely than any other group of animals, 

If Histriobdella is related to the Rotifers it becomes 
necessary to determine the relationship of Dinophilus to 
the same class. Schimkewitsch (28) was the first to point out 
the similarity of the caudal appendage in Dinophilus to 
the foot of the Rotifer. In Dinophilus, as in the Rotifer, 
this is used in attaching the animal. In both forms there is 
a marked sexual dimorphism. But as Nelson (25) has pointed 
out, the caudal appendage in Dinophilus resembles more 
that of some of the polytrochal annelid larve than the foot of 
the Rotatoria, and the sexual dimorphism can have arisen 
within the genus, as it is found in other groups of the 
Annelida besides the Rotifers. One striking difference 
between the Rotifers and Dinophilus is the apparent total 
absence of a definite mesoblast in the Rotifers, while it is 
clearly present in Dinophilus, where it has the same cell- 
origin as in Polychets. In Rotifers the mesoblast would 
seem to be represented by the germ-cells alone, and it is 


852 CRESSWELL SHEARER. 


necessary to suppose that the Rotifers separated from the 
main stem of the Annelida at a stage earlier than that of the 
formation of a definite mesoderm, while Dinophilus arose 
only after the ccelo-mesoblast had definitely appeared. On 
the whole, Dinophilus is not so closely allied to the 
Rotifers as Histriobdella. Unfortunately our lack of 
information with regard to the development of the ccelo-meso- 
blast in Histriobdella prevents our forming any opinion as 
to how much it resembles the Rotifers in this respect. 

It is remarkable with regard to the Rotifers that, despite 
their wide distribution and their great number of species, so 
comparatively few marine forms should be known. What 
has become of these if they have ever existed? Are forms 
like Belatro and Hemidasys (Claparéde, 4), Turbanella 
(Schultz, 29), or the Echinoderes (Zelinka, 37) to be looked 
upon as the modified descendants of a marine branch of 
these animals? Here we have a marked metamerism coupled 
with the main features that characterise both Histriobdella 
and the Rotifers. It is possible that it is with some of 
these somewhat obscure groups that the relationship of 
Histriobdella really lies. 

In conclusion, it may be stated that our present knowledge 
does not warrant us farther than to conclude that Histrio- 
bdella is a highly specialised form, retaining many Rotiferan 
features, and that it is to be grouped with Dinophilus asa 
primitive Annulate, but not directly related to Polygordins 
and Protodrilus. 


SUMMARY. 


(1) Histriobdella homari is a normal inhabitant of the 
branchial chamber of the European lobster. It is found in 
equal numbers throughout the year, on both the male and 
female. 

(2) he anterior feet of the head, unlike those of Stratio- 
drilus, are non-retractile. 

(3) There are four pairs of nephridia in both sexes. They 
are closed, and are of the primitive flame-cell type similar to 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 353 


those of Rotiters. Unlike those of Dinophilus, they bear 
no solenocytes. 

(4) There is a pair of salivary glands in connection with 
the mouth. — 

(5) There are fewer teeth in the jaw-apparatus than, 
Foettinger has represented. 

(6) The ventral nerve-cord is composed of two portions, 
which separate in the intersegmental to unite in the 
segmental regions, in prominent ganglia. The metamerism 
of the nervous system corresponds with that of the external 
form. 

(7) In the male there isa complicated generative apparatus. 
It is similar in all respects to that of the male generative 
apparatus in Stratiodrilus, 

(8) Fertilisation takes place internally. The largest egg is 
usually seen in the prophase stage of the first maturation 
division. ‘he amphiaster and the spindle can be seen to 
pass out through the body-wall with the cytoplasm, when the 
egg is ruptured by pressure. It remains for some seconds 
intact in the sea-water surrounded with yolk-granules. 

(9) In the equatorial plate there are eight chromosomes in 
the first maturation division. 

(10) Histriobdella is to be placed close to Dinophilus. 
It retains many Rotiferan features, and is more closely con- 
nected with this group than Dinophilus. Histriobdella 
and Dinophilus show distantrelationship with Polygordius 
and Protodrilus, but cannot be classed with them as true 
Archiannelids. 


LITERATURE. 


1. Van Beneden, J. P._—‘ Note sur une larve d’annélide d'une forme 
tout particuliere, rapportée avec doute aux Serpules,’ * Bull. 
Acad. Roy. Belgique,’ tome xx, 2nd pte., 1853, p. 69. 

“Histoire naturelle d’un animal nouveau, désigné sous le 

nom d@Histriobdella,” ‘Bull. Acad. Roy. de Belgique,’ 2nd 

serie, tome v, 1858, p. 270. 


354 CRESSWELL SHEARER. 


3. 


10. 


iH 


12. 


13. 


14. 


15. 


16. 


is 


18. 


Calkins, G. N.—* Mitosis in Notiluca miliaris and its Bearing 
on the Nuclear Relations of the Protozoa and Metazoa,” ‘Journ. 
of Morph.,’ vol. xv, 1899, p. 711. 


. Claparede, E.—* Observations sur les Rotateurs,” ‘Annals de Sci. 


Nat. Zool.,’ V ser., t. viii, 1867, p. 5. 


. Hisig, H.—* Die Entwicklungsgeschichte der Capitelliden,”’ ‘ Mitt. 


a. d. Zool. Stat. Neapel,’ Bd. xiii, 1898, p. 1. 


. Flemming, W.— Zellsubstance Kern, und Zellteilung,’ Leipzig, 


1882. 
‘Zur Mechanik der Zelltheilung,” ‘ Arch. f. Mik. Anat.,’ vol. 
xlvi, 1895, p. 696. 


. Foettinger, A.—‘* Recherches sur l’organisation de Histriobdella 


homari,” ‘ Arch. de Biol.,’ vol. v, 1884, p. 435. 


. Foot, K., and Strobell, E, C.—*‘ Prophases and Metaphase of the 


First Maturation Spindle of Allolobophora fwtida,” ‘Amer. 
Journ. Anat.,’ vol. iv, 1905, p. 199. 

Gardiner, E. G.—* The Growth of the Ovum, Formation of the 
Polar Bodies, and the Fertilisation in Polycherus caudatus,” 
* Journ. of Morph.,’ vol. xv, 1898, p. 73. 

Griffin, B. B.—‘ Studies on the Maturation, Fertilisation, and 
Cleavage of Thalassema and Zirphea,” ‘Journ. of Morph.’ 
vol, xv, 1899, p. 583. 

Harmer, 8. F.—* Notes on the Anatomy of Dinophilus,” ‘ Journ. 
of Marine Biol. Assoc.,’ N.S., vol. i, 1889, p. 1. 

Haswell, William A.—* On a New Histriobdellid,” ‘Quart. Journ. 
Mier. Sci.,’ vol. 43, 1900, p. 299. 

Heidenhain, M.—** Neue Untersuchungen iiber die Centralkorper 
und ihre Beziehungen zum Kern und Zellenprotoplasma,” 
‘Arch. f. Mik. Anat.,’ Bd. xliii, 1894, p. 423. 

Hempelmann, F.—* Zur Morphologie von Polygordius lacteus 
Schn. und P. triestinus,” ‘ Zeit. f. wiss. Zool., vol. Ixxxiv, 
1906, p. 527. 

Hermann, F.—* Beitrag zur Lehre von der Entstehung der karyo- 
kinetischen Spindel,” ‘ Arch. f. Mik. Anat., Bd. xxxvii, 1891, p. 
569. 

Ishikawa, C.—“ Studies on Reproductive Elements: II, Notiluca 
miliaris,” ‘Journ. Coll. Sci. Imp. Univ. Japan,’ vol. vi, 1894, 
poze: 

Korschelt, E.—* Ueber Kernteilung, Eireifung und Befruchtung bei 
Ophryotrocha puerilis,” ‘Zeit. Wiss. Zool., Bd. Ix, 1895, 
p. 543. 


19. 


20. 


21. 


22. 


23. 


24. 


25. 


26. 


27. 


28. 


29. 


30. 


31. 


32. 


33. 
34. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 3590 


Lillie, Frank B.— Karyokinetic Figures of Centrifuged Eggs,” 
‘Biological Bull.,’ vol. xvii, 1909, p. 101. 

Lund, C. Wesenberg.—* Danmarks Rotifera: I, Grundtraekkene i 
Rotiferernes Okologi,” ‘Morfologi og Systemstik.” Kobenhavn, 
1899. 


Mead, A. D.—*‘ The Origin of the Egg Centrosomes,” ‘Journ. of 
Morph.,’ vol. xii, 1897, p. 391. 


Meves, F.—‘* Uber eine Metamorphose der Attractionsphire in 
den Spermatogonien von Salamandra maculosa,” ‘ Arch. 


f. Mik. Anat.,’ Bd. xliv, 1894, p. 119. 


Munson, J. P.—‘ The Ovarian Egg of Limulus: A Contribution 
to the Problem of the Centrosome and Yolk Nucleus,” ‘ Journ. 
of Morph.,’ vol. xv, 1898, p. 111. 


Nelson, J. A.—‘* The Early Development of Dinophilus,” ‘ Proc. 
Acad. Nat. Sci. Phil.,’ 1904, p. 687. 


“The Morphology of Dinophilus Conklini n. sp.,” ‘ Proce. 
Acad. Nat. Sci. Phil.,’.1907, p. 82. 


Pierantoni, U.—‘‘ Protodrilus,” ‘Fauna und Flora des Golfes 
von Naple,’ 31 Monographie, 1908. 


Rickert, J—‘ Zur Hireifung bei Copopoden,” ‘ Anat. Heft,’ Bd. iv, 
1894, p. 261. 


Schimkewitsch, W.—‘‘ Zur Kenntniss das Baues und der Entwick- 
lung des Dinophilus vom Weissen Meere,” ‘Zeit. f. Wiss. 
Zool.,’ Bd. lix, 1895, p. 46. 


Schultze, M.—‘* Uber Chetonotus und Ichthydium Ehrb. und eine 
neue verwandte Gattung Turbanella,’ ‘Arch. f. Anat. u. 
Phys.,’ Jabrg. 1853, p. 241. 


Shearer, C.—‘‘ On the Structure of the Nephridia of Dinophilus,” 
‘Quart. Journ. Micr. Sci.,’ vol. 50, 1906, p. 517. 

Watasé, S.—* Homology of the Centrosome,” ‘Journ. of Morph.,’ 
vol. viii, 1893, p. 433. 

Wilson, E. B.—‘On Protoplasmic Structure in the Eggs of 
Echinoderms and some other Animals,” ‘Journ. of Morph..,’ 
vol. xv (Suppl.), 1899, p. I. 


—— ‘The Cell,’ New York, 1900. 


“ Experimental Studies in Cytology : I,” ‘Arch. f. Entwick.,” 
vol. xii, 1901, p. 529. 


356 CRESSWELL SHEARER. 


85. Zelinka, C.— Die Gastroctrichen,” ‘ Zeit. f. Wiss, Zool., Bd. xlix, 
1890, p. 209. 


36. “Studien iiber Raderthiere,”’ ‘ Zeit. f. Wiss. Zool.,’ xliv, 
p. 396, Bd, xlvii, p. 353, Bd. liii, p. 1, 1885-1892. 
37. “Uber Echinoderes,” ‘Verh. d. deutschen Zool. Gesell.,’ 


4th Jahrssam., 1894, p. 46, 


EXPLANATION OF PLATES 17—20. 


Illustrating Mr. Cresswell Shearer’s paper ‘‘ On the Anatomy 
of Histriobdella Homari,” 


LETTERING. 


ac. Accessory glands of the male reproductive apparatus. an. Anal 
aperture. ap. p. Appendage of the posterior leg. ble. Blastoccelie 
cavity. bl. m. Muscular organs of jaws. br. Bridle piece of jaws. 
brn. brain. ed. g. l-ed. g. 3. Ganglia of the caudal region. el. Clasper. 
cl. p. Ciliated pit of the head. ca. Celom. ce. ep. Celomic epithelium. 
com. Nerve commissures. f. Fulcrum of jaws. fol. Follicle cells. jgl. 
Flagella of the nephridial canals. g. 1-g. 5. Ganglia of the ventral nerve- 
cord. gl.cl. Gland-cell of clasper. gr.g. Granule gland. nt. Intestine. 
int.2. Intestine, posterior part. j. Jaws. j.1. Upper ramus of jaws, 
j. 2. Lower ramus of jaws. Jl. a. Anterior legs or feet. 1. p. Posterior 
legs. m. Mouth. m.d. Dorsal longitudinal muscles. m.d.p. Median 
duct of the penis. m. gl. Salivary glands of the mouth. m. ob. Oblique 
muscles. m.v. Ventrallongitudinal muscles. n.c. Ventral nerve-cord. 
neph. 1-neph.4. Nephridia. neph.c. Nephridial canals. neph.h. Head 
of the nephridium. neph. 0. Opening of the nephridial canal on the 
external surface. neph.s. Spaces on the course of the nephridial canals. 
es. G@sophagus. o.im. Immature ova. ov.p. Orifice of the penis 
sheath. ov. Ovary. ovd. Oviduct. p. Penis. 7. Ramus of upper jaw. 
sprm. Spermatidia. spe. Spermatocyte. st. Stomach. ¢. 1. Median 
tentacle. ¢.2 and ¢.3. Lateral tentacles. te. Testis. th. Teeth. v. def. 
Vas deferens. ves. Vesicula seminalis. vn.c. Ventral nerve-cord. vit. 
Vitellarium or shell-gland. 


PLATE 17. 


Fig. 1—Female Histriobdella with eggs. The largest egg shows 
the presence of a maturation amphiaster x 300. 


Fig. 2.—Clasper extended. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARI. 357 


Fig. 3.—Third and fourth nephridium in the female. This and all 
the subsequent figures of the nephridia have been drawn from living 
preparations impregnated with methyl-blue; 2mm. oil-immer., comp. 
oes. 4and 6, x 500 and x 1000. 

Fig. 4.—Third nephridium in the male. 

Fig. 5.—Third nephridium in the male. 

Fig. 6.—Second nephridium in the male. 

Fig. 7.—Male Histriobdella with claspers retracted, x 300. 


PLATE 18. 


Fig. 8.—Rupture of an egg through the body-wall in a living prepara- 
tion by compression of the cover-glass. The first maturation amphi- 
aster is seen outside the body-wall in the sea-water. x 500. 


Fig. 9.—Male with claspers extended. x 300. 

Fig. 10.—Third and fourth nephridium in the male. x 800, 

Fig. 11.—Generative segment in the male. Taken from a living 
preparation, showing the reproductive organs. 

Fig. 12.—Fourth nephridium in the male. x 800. 


Fig. 13.—Fourth nephridium in the male. Segment contracted. 
x 800. 
Fig. 14.—Second nephridium in the female. x 800. 


PLATE 19. 


Fig. 15.—Reconstruction of the nervous system in the male, showing 
the dorsal curve taken by the ventral nerve-cord in the region of the 
penis. Lateral view. x 300. 

Fig. 16.—Young in egg-capsule. 

Fig. 17.—Young in egg-capsule. Harlier stage than that shown in 
fig. 16. 

Fig. 18.—Odcyte with first maturation amphiaster. This, with the 
subsequent figures, 19, 20, 26, and 32, are all drawn from the same egg- 
cell. They show the changes of shape assumed by the egg in the move- 
ments of the animal. They were drawn at intervals of from ten to 
twenty minutes. 


Fig. 19.—Odcyte, same as that shown in fig. 18, drawn twenty minutes 
later. 


Fig. 20.—Odcyte, same as that of fig. 19, fifteen minutes later. 


358 CRESSWELL SHEARER. 


Fig 21.—Reconstruction of thejnervous system in the female. The 
brain surface is measured from the ganglion cells and not from the 
fibrous core. Ventral view. x 300, 


Fig. 22.—Sagittal section in the female showing the sac-like nature 
of the ovarian cavity. 


Fig. 25.—A cell of the accessory gland of the male. 


Fig. 24.—Transverse section in the male in the region of the penis. 
x 400. 


Fig. 25.— Eggs attached to the membranes of the lobster “ berry.” 
Fig. 26.—Odcyte twenty minutes after that shown in fig. 20. 


Fig. 27.—Spermatidia. 


Fig. 28. — Reconstruction of the nervous system of the male. 
Ventral view. x 300. 
Fig. 29.—Longitudinal section of the wall of the intestine in the 


posterior region. 
Fig. 30.—Young, a short time after hatching. x 300. 
Fig. 31.—Section through the generative region in the male showing 
the granule glands. 
5 5 


Fig, 32.—Odcyte twenty minutes later than fig, 26. 


PLATE 20. 


Fig. 33.—Transverse section in the male through the region of the 
vesicule seminales. 

Fig. 34.—The same. Ina region a little posterior to the last. 

Fig. 35.—Transverse section through the middle of the generative 
region in the male. 

Fig. 36.—Chitinous jaws. 

Fig. 57.—Horizontal section in the male. 

Fig. 388.—Transverse section through the middle of the second 
segment. 

Fig. 39.—Transverse section through the anterior region of the 
generative segment in the male, showing the divided nature of the 
anterior portion of the testis. 

Fig. 40.—Transverse section through the posterior region of the 
generative segment in the male. 

Fig. 41.—Transverse section through the dorsal region of the body- 
wall in the caudal segment, showing the insertion of the oblique muscle- 
fibres. 


ON THE ANATOMY OF HISTRIOBDELLA HOMARL. 359 


Fig. 42.—Horizontal section through the region of the first segment 
in the male, showing the head of the first nephridium. 


Fig. 43.—Transverse section through the neck region. 


Fig. 44.—Transverse section through the caudal region, showing the 
oblique muscles. 


: —_ ° 3 1 
- * 
S = | 
os . - 
7) . 
4 & i 
_ 
» - 
a: : 
4 
- : & 
7 
_ y . 
= 
f mf , 
“a Jf i ' - = 
at 
= 4 =" 
cr * 5 ; ; 
caer a i 8 , : 
a P ‘ : 
o . 


Qua rb. umn clio Dek. Vol. I NS HM. £4 


neph Bf ; 


nepl £ 


Hath, Lith? London 


“A 
reph..¢, 


Cae 


burn Mier Sc. Vol.5INS ALIS. 


Duart. 


aheanegens 


mda. 


Huth, Lith™ _ 


- Sourn Mucr Sci. lo b5, NS. Te 


i 
é 
? 
! 
id 
a 
ww 
° 
a 
# 
. 
’ 


: Tyo 


- So ee eee ee elie 


eR sr Ny 
Pe Conenge Pee 
Ss, 


C.S. & G.B.R, del. 


es 


Quart. Fournr Mice. Vb SSNS, 20 


3 ? 
oe Bre aN 


Huth, Lith? London. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 361 


On the Artificial Culture of Marine Plankton 


Organisms. 
By 


E. J. Allen, D.Sc., 


Director of Laboratories and Secretary of the Marine Biological 


Association ; 


and 


E. W. Nelson! 
Assistant Naturalist. 


Con'rEN'TS. 


Introduction . ; : 

I. Culture of Plankton Diatoms 

A. Practical Culture Methods 
. Miquel’s Method : : 
. Houghton Gill’s Method , F : 
. (A) Modification of Miquel’s Methods: “ Miquel 
Sea-water ” : ; , 

(8) English Channel Water . 
(c) Tank-Water 
( 
( 


— 


bo 


ey) 


p) Animal Charcoal Water . 
E) Peroxide of Hydrogen Water 
(r) Cultures in these Media . : ‘ 
p. Experiments with a view to Determining the Con- 
ditions which underlie the Successful Culture of 
Diatoms ; 
Methods. : 
The Sea-water employed : 
The Constituents of Miquel’s Solutions . 
Animal Charcoal and Peroxide of Hydrogen 
Reviving Exhausted Cultures . 
Silica 
Organic Infusions 


Co co 
~I 
bo 


2 O89 GW OD 
~I sI NI <1 
Oe Ww 


“Ss 
eo) 


1 Owing to pressure on our space, this memoir could not be pubiished 
when first in type. It has in the meantime been issued in the ‘ Journal 


of the Marine Biological Association,’ vol. vili, No. 5.—E. R. L. 


862 KE. J. ALLEN AND E. W. NELSON, 


PAGE 
Artificial Sea-water . : , : 395 
Alkalinity . P . , . 395 
Salinity : : 3 : . 402 
Light ; : : f . 403 
Temperature ' ‘ ; . 404 
General Conclusions . : : . 405 
II. Mixed Cultures, : : : . 407 
III. Notes on Particular Species of Diatoms, on their Methods 
of Reproduction, and on other Alge occurring in 
Cultures ‘ ; j : . 412 
IV. Rearing of Marine Larve . é 5 aie 
Methods A ‘ ; : | AZ 
Echinus acutus : : : . 419 
E.esculentus . : : : . 420 
E. miliaris . ; ; : _ wer 
Cucumaria saxicola, 5 . 422 
Pomatoceros triqueter : ; . 422 
Chetopterus variopedatus ; ; . 423 
Sabellaria alveolata . ; ‘ . 423 
Archidoris tuberculata : . . 423 
Calanus finmarchicus. : . . 424 
Echinus hybrid . - ; ’ . 425 
Sacculina carecini : ; ; ; 25 
Summary of Method for Rearing Larve : . 426 
Bibliography . 5 ‘ . ; . 427 


INTRODUCTION, 


THe observations to be recorded in this paper were com- 
menced in March, 1905, They originated in an attempt to 
find a general method for rearing marine larval forms. 
Several investigators had previously succeeded in rearing 
echinoderms, molluscs, and polychztes from artificially 
fertilised eggs, under laboratory conditions, but the process 
was generally difficult and the results more or less uncertain. 
The most promising method seemed to be that adopted by 
Caswell Grave (26), who was able to rear his larve by feed- 
ing them on diatoms. Grave obtained his diatoms by placing 
sand, collected from the sea bottom, in aquaria, and using 
such diatoms as developed from this material. All the 
methods, however, suffered from the uncertainty of not 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS, 363 


knowing what organisms were introduced into the aquaria in 
which the larve were to be reared, either in the original sea- 
water or along with the food supply. 

It appeared, therefore, at an early stage of the work, worth 
while to make an attempt to carry out rearing experiments 
on amore definite and precise plan, to endeavour, in fact, to 
introduce the larve to be reared into sterile sea-water, and 
to feed them with pure cultures of a suitable food. his was 
the-ideal to be aimed at. As a matter of fact it has seldom, 
if ever, been attained in practice; nevertheless, a consider- 
able measure of success has been achieved by working upon 
these lines, and during the course of the work innumerable 
problems relating to the physical conditions under which 
plankton organisms can best flourish have presented them- 
selves. Some account of the experiments made may be of 
interest to other workers, although imany of the problems 
raised are not yet solved, notwithstanding the fact that some 
1500 cultural expernnents have been under observation, It 
is rather with a view of stimulating other work upon similar 
lines than of bringing forward conclusive results, that this 
paper is being published. 

In the summer of 1907 Mr. E. W. Nelson became associated 
with the investigation, and since that date the experimental 
work has been carried out by him. ‘The discussions in this 
paper of a more chemical character, particularly the section 
on alkalinity, are almost entirely the work of Mr, Nelson, and 
we have both had throughout the advantage of the constant 
advice and help of Mr. D. J. Matthews on all such matters. 


I. Curture or Pranxton Dria'roms\ 
(aA) Practical Culture Methods. 


1, Miquel’s Method.—Attention was first directed to 
the culture of plankton diatoms; and the methods, which 
had been elaborated by Miquel (11) for fresh-water diatoms 
and had been found by him to succeed with marine bottom 
diatoms, were tried. 


364 kK. J. ALLEN AND E. W. NELSON. 


The essential features of Miquel’s method, as applied to 
marine diatoms, are as follows : 
Two solutions are prepared : 


Solution A. 


Magnesium sulphate . ‘ . 10 grm. 
Sodium chloride , : > LOOP Se 
Sodium sulphate 3 ‘ : (eS 
Ammonium nitrate L. 
Potassium nitrate, ; «.\ Gea 
Sodium nitrate . BA ie 
Potassium bromide . , >, (eae 
Potassium iodide i : . ts 


Water ‘ , : ; + 0G te 


Solution B.! 


Sodium phosphate ee 
Calcium chloride (dry). ‘ee 
Hydrochloric acid. : . eer 
Ferric chloride . : : . > Shea 
Water : ‘ ; , . ‘SO0'ce 
Forty drops of solution A and ten to twenty drops of solu- 
tion B are added to each 1000 c.c. of sea-water, and the sea- 
water is sterilised by keeping it at 70° C. for about twenty 
minutes. 
According to Miquel it is also necessary to add “ organic 
nutritive material in the form of bran, straw, or filaments of 
1 «The preparation of solution A presents no difficulty. Solution B 
should be made up as follows: To the sodium phosphate dissolved in 
40 ¢.c. of water are added first the 2 ¢.c. of hydrochloric acid, then the 
2 cc. of hydrous ferric chloride, and then the 4 grm. of calcium 
chloride dissolved in 40 c.c. of water, taking care to shake the mixture, 
which I call phospho-ferro-calcic solution. The addition of this last 
solution to the maceration throws down a slight brownish flocculent 
precipitate, formed for the most part of ferric oxide, which should be 
carefully separated from the liquid used for cultivations.” 
> * Acid chlorhydrique pur a 22°.” Presumably meaning degrees 
Baumé = sp. gr. 1169. 
3“ Perchlorure de fer liquide 4 45°." As above = sp. gr. 1421. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS, 865 


weeds, such as Zostera. Macerations of these should be made 
up separately some time before they are required for use, 
and should be carefully filtered and sterilised. Organic 
matter must, however, be used very sparingly, or else putre- 
faction will set in arid the cultures will be irrevocably lost.” 
As a matter of fact we have found that such organic infusions 
are unnecessary when dealing with plankton diatoms, and it 
has not been our practice to employ them (cf., however, p. 392). 
Miquel obtained cultures of single species of diatoms either 
by picking out individual diatoms under the microscope and 
introducing them into the prepared water, or by adding a 
small quantity of water containing a mixture of diatoms and 
other organisms to some prepared water, and subdividing 
this into'a number of tubes. If the subdivision has been 
carried out sufficiently some of the tubes may contain one 
kind of diatom only, from which fresh cultures can be made. 
In this way, by repeated subdivision, cultures can be obtained 
which, by inoculating fresh quantities of prepared water from 
time to time, may, with care, be maintained indefinitely. Such 
cultures, however, must practically always contain bacteria, 
and Miquel distinguishes them from bacteria-free cultures, 
which he terms “ cultures des diatomées a l’état de pureté 
absolue.” The latter he found very difficult to obtain, but 
through repeated washing in sterile water, followed by frac- 
tional subdivision, he succeeded in getting some in which he 
could -find no trace of bacteria by ordinary bacteriological 
methods (Miquel [11], p. 155; ef. also Richter [16-181). 
We propose to call any diatom culture which can be 
carried on practically mae by inoculating fresh 
supplies of prepared water a “per sistent” culture, the 
term “pure” culture being reserved for cultures which 
can be proved to contain not more than one organism. We 
are not satisfied that we have yet succeeded in obtaining’ 
cultures of the latter kind. For the most part our persistent 
cultures contain one species of diatom only, and are free eu 
all organisms larger than small flagellates. 
In our Pacer experiments with plankton diatoms we 
VOL. 55, PART 2.—NEW SERIES. 24 


366 BE. J. ALLEN AND E. W. NELSON. 


obtained persistent cultures, containing a single species of 
diatom, by both of the methods recommended by Miquel. We, 
however, have rarely succeeded by picking out single diatoms 
or chains of diatoms, for although we have passed the selected 
diatom through several changes of sterilised sea-water, the 
resulting cultures, even when the diatoms have multiplied to 
some extent, have generally shown evidence of contamination 
by harmful organisms, and have soon died down, Only in 
one of the earliest experiments, and in one more recent, has 
complete success resulted, In the first case a small chain of 
six or eight frustules of Skeletonema costatum, picked 
out in April, 1905, gave rise to a culture which still persists 
(November, 1909). Subcultures can still be obtained even 
from the original flask inoculated in April, 1905. In the 
second case a chain of eight or nine cells of Chetoceras 
densum, picked out from a Petri dish culture, has given a 
particularly good growth, 

The method of dilution and subdivision has been more suc- 
cessful, and persistent cultures of a number of species have 
been obtained in this way. 

A more ready method of obtaining the cultures is, we have 
found, to add one or two drops of plankton to, say, 250 c.c. 
of a suitable sterile culture medium, and to pour this into 
shallow glass dishes (Petri dishes). The dishes should be 
placed in a position as free as possible from vibration, and 
where they can be easily examined with a lensin situ. The 
temperature should be kept as constant as possible and the 
dishes exposed to light of moderate intensity, direct sunlight 
being avoided. In the course of a few days, colonies of 
diatoms of different species will be seen at different spots on 
the bottom of the Petri dishes. These can be picked out 
with a fine pipette and transferred to flasks containing fresh 
culture medium. The colonies should be picked out from the 
Petri dishes at as early a stage as possible, because if left too 
long some one organism, a diatom or a flagellate, may have 
multiplied so rapidly that the whole of the water in the dish 
becomes infected with it. In this case persistent cultures of 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 867 


a single species would not be obtained. The above method is 
similar to one described by Miquel, excepting that he placed 
gelatinous silica at the bottom of the vessel. Some very 
successful persistent cultures were obtained from the follow- 
ing experiment, which will serve to illustrate the method: A 
sample of plankton, from a very fine-mesh bolting-silk tow-net, 
was diluted down with sterile sea-water, until a single drop 
examined under a 2 in. objective contained on an average 
ten organisms, chiefly diatoms, of various species. Petri 
dishes (4 in.), containing 60 c.c, each of Miquel sea-water, 
were then inoculated with various numbers of drops of the 
diluted plankton. The two dishes, to which two and three 
drops respectively were added, gave the best results, and 
from these persistent cultures of several species of diatoms 
were obtained. Hence we may conclude that the most advan- 
tageous number of single cells or short chains of cells to be 
added to a 4 in. Petri dish, containing 60 c.c. culture medium, 
is about twenty to thirty. 

We have succeeded in obtaining the following species of 
plankton diatoms in persistent cultures: . 

Asterionella japonica Cleve, 

Biddulphia mobiliensis (Bail.) Grun. 

Biddulphia regia (M. Schultze).! 

Chetoceras densum Cleve. 

Chetoceras decipiens Cleve. 

Chetoceras constrictum Gran. 

Cocconeis scutellum EKhr. var. minutissima Grun, | 

Coscinodiscus excentricus Ehr.’ 

Coscinodiscus Granii Gough. 

Ditylium Brightwellii (West) Grun. 

Lanuderia borealis Gran. 

Nitzschia closterium W. Sm. 

Nitzschia closterium W. Sm. forma minutissima. 

Nitzschia seriata Cleve. 

Rhizosolenia stolterfothii H. Perag. 


1 See p. 413. 
7 See p, 412, 


368 E, J. ALLEN AND E. W. NELSON. 


Skeletonema costatum (Grev.). 

Streptotheca thamensis Shrubs. 

Thalassiosira decipiens Grun.! 

It is hardly necessary to add that in dealing with these 
cultures similar precautions to those used in bacteriological 
work must be taken, all vessels and instruments being care- 
fully sterilised before they are brought into contact with the 
prepared sea-water. ‘lhe cultures are best made in small, 
wide-mouthed flasks, which may be plugged with cotton-wool, 
or simply covered with watch-glasses. ‘The flasks should be 
kept at as uniform a temperature as possible (from 12°-17° C.) 
and should be exposed to strong daylight, direct sunlight 
being avoided. A flask should not be more than half filled 
with culture fluid, so that the surface exposed to the air may 
be large in proportion to the volume of fluid. 

Other Methods.—The addition of the solutions devised 
by Miquel to sea-water has in all cases given us good cultures 
of diatoms, and the method is certain in its action. We have, 
however, made numerous experiments by treating sea-water 
in other ways, with a view to finding ont what are the best 
conditions under which plankton diatoms will grow, and of 
arriving at some explanation of the action of the different 
salts contained in Miquel’s solutions. 

2. Houghton Gill’s Method.—H. Houghton Gill (5), 
a contemporary of Miquel, made use of a culture medium not 
essentially different from that employed by the latter. Unfor- 
tunately he died before publishing his work, but an account 
of his principal results is given by Van Heurck. In his final 
method Houghton Gill made use of four distinct solutions, as 
follows : 

Solution 1. 
Crystallised sodium phosphate 2 
Calcium chloride’. . . ee 
Syrup of iron chloride © . 0 
Strong hydrochloric acid ‘ sea 
Water . ; : - 100,48 
: ve p- 412, 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS, 9869 


Solution 2. 


Crystallised magnesium sulphate 4 germ 

Crystallised sodium sulphate . ey: eee 

Crystallised potassium nitrate 4 ,, 

Common salt (sodium chloride) 4 ole 

Potassium bromide . : i ae Ue 

Potassium iodide. e : Or ew: 

Water. ‘ : ‘ : NOW! -& 
Solution 3. 

Crystallised sodium carbonate . Agrm, 

Weater . . . : : F 25LOOE 
Solution 4. 

Well-washed, precipitated calcium 

silicate . ; : : + 2 ST 
Water Loge 


All the salts employed must be chemically pure. Three 
c.c. of each of these liquids are added to 1000 c.c. of fresh 
water or sea-water (according to circumstances), and the 
whole sterilised. In his earlier work Houghton Gill added a 
sterilised infusion of grass or of diatoms, but it is not-clear 
from the accounts whether this was still employed with the 
above solutions. We have obtained very good cultures with 
the above solutions, to which we did not add any organic 
infusion. 

3. (A) Modification of Miquel’s Method: “ Miquel 
Sea-water.”’—Since several of the components in Miquel’s 
formula for solution A (p. 363) are obviously unnecessary 
when sea-water is being used as the basis of the culture 
medium, we adopted for our own work the following modifica- 
tions: After some preliminary experiments it was found, 
as would be expected from the composition of sea-water, 
that the only salts of value to the medium are the three 
nitrates KNO;, NaNO;, NH,NO,, and possibly KBr and KI. 


The omission of the two latter was soon found to make no 


370 E. J. ALLEN AND E. W. NELSON, 


difference. Experiments also showed that the formula for 
solution A could, without any appreciable detriment to 
results, be further simplified to the one salt KNO, or NaNOs, 
but not NH,NO,. At first the amount of KNO, dissolved in 
100 c¢.c. distilled water, used to make the modified solution 
A, was the same as the sum of the weights of the nitrates in 
Miquel’s own formula, viz. 5 grm. But later experiments 
showed that a considerably greater concentration of KNOs 
than this gave more lasting cultures; the strength of solu- 
tion and amount to be added to a litre of sea-water in 
order to obtain the best results being 2 c.c. 2 M KNOs3. 

In the case of solution B no modification has been adopted, 
but it has been found that small variations in the amounts of 
the ingredients used do not affect the results. A convenient 
method for measuring the right amount of FeCl, is to warm 
the salt until it just melts in its own water of crystallisation, 
and to pipette out 2 c.c. with a previously warmed pipette. 
No temperature corrections need be considered. Also 2 ¢.c. 
of the ordinary pure concentrated hydrochloric acid at room- 
temperature will suffice. 

Our own formula for preparing Miquel sea-water is now : 


Solution A.! 


Potassium nitrate, 20 2 grm. t= 2 M KNO,. 
Distilled water, 100 eS 


Solution B.? 


Sodium phosphate (Na,HPO,12H,0) + : 
Calcium chloride (CaCl,6H,O) . :- 
Ferric chloride (melted) ee 
Hydrochloric acid (pure concentrated) 2 ,, 
Distilled water . : : : : SOnee 


‘ This strength has only been used in the most recent experiments ; 
and solution A in this paper, unless otherwise stated, means the 5 % 
solution of KNO3. 

* For preparing this solution see p. 564. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 371 


To each 1000 c¢.c. of sea-water! add 2 c.c. solution A and 
1 c.c. solution Band sterilise by heating to 70°C. When cool, 
decant off the clear liquid from the precipitate, which will 
have formed when solution B is added to the sea-water. 

As a rule our cultures were made in 60 c.c. of this medium 
contained in short-necked, wide-mouthed flasks of 125 c.c. 
capacity, so that the proportion of air-surface to volume of 
liquid was large. 

The medium was found to give constantly satisfactory 
results. On inoculation from a persistent culture of such 
diatoms as Thalassiosira, Skeletonema, Chetoceras, 
etc., a growth visible to the eye is obtained in about ten days, 
and then multiplication takes place very rapidly. In from 
three weeks’ to a month’s timea very considerable growth will 
be seen making a brown, flocculent mass at the bottom and 
back of the vessel containing the culture. 

In from two to four months the culture begins to show 
signs of exhaustion and the frustules lose colour, but they do 
not, as in the case of sterilised outside and tank-water, com- 
pletely die off. A great number certainly do die, but some 
remain in a resting condition, and often, after a period of six 
months or so, these begin to multiply again and the culture 
regains its former vigour. his is probably due to the food- 
stuffs contained in the dead frustules going into solution again, 
possibly by means of bacterial action. This periodicity in 
cultures is interesting in that it resembles what takes place in 
the ocean. Cultures in this medium will persist indefinitely, 
so far as our experience goes. The oldest culture in our 
possession is one of Skeletonema costatum made at the 
very commencement of this work, dated April, 1905. Although 
the frustules in this culture are quite unrecognisable as any 
diatom now, on making a subculture in fresh Miquel a normal 
and healthy growth can always be obtained. 

In old cultures the diatoms are nearly always found to be 
very much deformed, and often appear to be only a mass of 


1 * Miquel water” seems to succeed equally well, whether it is made 
by adding Miquel’s solutions to “outside water” or to “tank-water.” 


372 HE. J. ALLEN AND EK. W. NELSON. 


broken-down chromatophores. Whether regeneration can be 
successfully obtained from a single chromatophore, which 
must presumably be contained within a cell-wall of some 
kind, has not been definitely decided, but results seem to 
point in this direction, 

At the start of a culture a tendency to teratological forms 
is often exhibited, but when the growth is well advanced, the 
shape of the frustules is usually quite normal, 

(8) English Channel Water (“Outside Water”).—In 
a large number of our experiments sea-water brought in from 
outside the Plymouth breakwater, and therefore taken at some 
distance from the shore, has been used. This is referred to as 
“outside water.” 


and the temperature range for the year is from 8° to 16° C. 
2 


It has an average salinity of about 35°0 °/,, 
If a sample of ‘‘ outside water ” is inoculated from a persis- 
tent culture of a plankton diatom, a small growth is obtained 
in from five to fifteen days. . But soon minute bottom forms of 
diatoms, other algze, flagellates, infusoria, etc., appear, and the 
inoculated species is lost. ‘lhe total growth of any form is 
never large. If the growth of these foreign forms is pre- 
vented by sterilising the water before inoculation, a consider- 
ably better growth of the plankton form is obtained. The 
water was, as a rule, sterilised by simply heating to 70° C., 
which temperature was found to be quite adequate. Boiling 
gave equally good results, but the former was preferred, as 
less concentration due to evaporation took place. Even 
under these conditions no permanent culture can be obtained, 
the diatoms soon beginning to lose colour and getting into an 
exhausted condition, Death takes place in from two to three 
months after the culture has been started, and in many cases 
considerably sooner. Long before inability to start new 
cultures, the test of death, has been established, the valves 
appear on examination quite colourless and practically empty. 

Samples of eutside water, taken at times when the quantity 
of plankton was widely different, gave no appreciable varia- 
tion in the results obtained by culture methods. It is, how- 
ever, doubtful whether differences in the amounts of growth 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 3873 


in cultures, proportional to the seasonal variation in the 
quantity of phytoplankton, would be sufficiently marked to 
be appreciable. 

The total growth under cultural conditions, although small 
for a culture, is very much greater than any natural plankton 
that has come within our experience. 

(c) Tank-Water.— Tank-water,” or water taken from 
the supply of sea-water circulating through the tanks of the 
Aquarium at Plymouth, shows some striking and interesting 
differences from “ outside water.”’? This water is pumped up 
from the sea, just below the Laboratory, into two large, 
covered-in, settling reservoirs, with a capacity of 60,000 
gallons each. Pumping is only done at high-water spring 
tides, so as to get the least contaminated water, and no water 
is pumped that does not show a specific gravity, measured 
with a hydrometer, of 0!” = 26:00 (S = 34:00) or over. The 
water is allowed to settle for about a fortnight before being 
used for the general circulation. 

The tanks themselves are made of slate and glass, and the 
pipes which convey the sea-water to them are of vulcanite, 
so that the water does not come in contact with metal, 
excepting in the pumps, which are of cast iron. The two 
settling reservoirs are used alternately for about a week each. 
From time to time, tide and water allowing, waste is re- 
plenished, and about twice a year each reservoir is emptied, 
cleaned out, and refilled. The aquarium takes about 20,000 
gallons, and this is in circulation with one of the two 50,000 
gallon reservoirs. An estimate of the amount of life in the 
tanks of the aquarium must be exceedingly rough, but the 
intensity of the larger forms of life is far greater than any- 
thing met with in natural waters. About 500 fish and 2000 
invertebrates, including all forms as large as an Actinia 
equina, might be somewhere near the mark. So it will be 
seen that the accumulation of excretory products must be a 
by no means negligible factor. The flora of the tanks is very 
restricted, and is chiefly composed of minute forms of alge. 
Minute naviculoid diatoms, Eetocarpus, Cladophora, 


374 E. J. ALLEN AND E. W. NELSON, 


Knteromorpha, Vaucheria, and unicellular alge are the 
commonest forms. ‘The large seaweeds, such as Fucus and 
Laminaria, do not live long if introduced. Plankton 
diatoms, although a great number must be pumped up when 
the reservoirs are being filled, are not represented. 

As in the case of outside water, a sample of ‘‘ tank-water,” 
inoculated from a persistent culture, will only give a very 
small growth, minute forms, etc., soon multiplying and 
choking out the plankton form. The ultimate growth of 
minute unicellular alge other than diatoms is often con- 
siderable, and many quite unknown and unidentified forms 
have been obtained. ‘The total growth of vegetable forms is 
always found to be greater than in the case of outside water. 

In cultures of plankton diatoms made with sterilised tank- 
water, a very great improvement on outside sterilised water 
was always noted. ‘lhe culture of the diatom used to inocu- 
late this medium persists for a considerable period, and the 
colom of the frustules remains normal for two to three months, 

(Dp) Animal-Charcoal Water.—The use of animal 
charcoal, as a means of purifying the water in small aquaria, 
has for a long time been known and practised by those who 
have kept such aquaria in inland places. At an early stage 
in Our experiments, water from a tank, which was not in 
a satisfactory condition, was treated with some powdered 
animal charcoal and filtered. It was noticed that a good 
growth of diatoms took place in this water. Systematic experi- 
ments with the use of animal charcoal were then commenced, 
and these have resulted in a method of great value, both for 
the culture of diatoms and for the rearing of pelagic larve. 

Animal charcoal is made by the carbonisation of bones,! 


1 Analysis of animal charcoal, from Thorpe’s * Dictionary of Applied 


Chemistry “— 
Carbon, , ; ; ; . Ist 
Ca., Mg. phosphates, Ca. fluoride, ete. . . 8021 
Calcium carbonate : : . "SSB 
Other mineral matter. ; : SS 


100°00 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 375 


and is sold in two grades knownas “pure” and “‘ commercial.” 
Our earlier experiments were all made with “ pure” animal 
charcoal, but subsequently the ‘‘ commercial” animal char- 
coal was largely used, and appears to give equally good, if not 
better results. In both cases the animal charcoal is used in the 
powdered form. Animal-charcoal water is prepared as follows: 

(1) A quantity of sea-water is sterilised by heating it in 
a flask to 70°C., at which temperature it should be kept for 
about twenty minutes. At the same time some animal 
charcoal is heated sufficiently to sterilise but not to burn it, 
covered over, and allowed to cool. When both are quite 
cold the charcoal is added to the water (ca. 15 grm. to 
1000 ¢.c.), and well shaken up in it several times. After 
an interval of half an hour or more the water is filtered 
through fine filter-cloth,! the whole filter having been first 
sterilised with boiling sea-water, and is received in a sterile 
flask. It is then ready for use. 

(2) For many experiments, where larger quantities of water 
were required, the sea-water was not sterilised before being 
treated with animal charcoal. In this case, if the first part 
of the filtrate be rejected, the subsequent water will generally 
be practically sterile, and few, if any, extraneous organisms 
will develop in it. 

(3) At a later date an automatic apparatus was set up in 
the Plymouth Laboratory, by which very considerable quanti- 
ties of sea-water could be treated with animal charcoal, and 
subsequently filtered through a “ Berkefeld” filter; water 
treated in this manner we call “ Berkefeld water.” Tank- 
water was always used in this apparatus, and was mixed with 
animal charcoal,” in a clean sulphuric acid carboy, by blowing 
air through with a pair of bellows. ‘lhe mixture was allowed 
to settle for at least twenty-four hours, and then syphoned 


1 The filter-cloth used for this purpose is the same as is made for use 
in filter presses, and is known as extra-super swansdown. To prevent 
this becoming clogged another cloth, known as hydraulic twill, was, as 
a rule, used over it. 

2 Ca. 300 grm. to 20 litres of water. 


376 E. J. ALLEN AND £. W. NELSON. 


over into an inverted bell-jar, with a tubulure at the bottom, 
into which the Berkefeld candle was fitted. Filtration under 
these conditions was found to be rather slow, so in order to 
increase its rate an apparatus was devised by which the 
pressure on the filter was considerably augmented. 

This apparatus (see Fig. 1) consists of a glazed earthenware 
“tobacco jar,” with two tubulures, one at the side, the other 
at the bottom, and a lid which can be screwed down tightly 
on to a rubber washer, by means of a triangular metal 
arrangement fitting into grooves above the lid.! The internal 


Fia. 1.—Diagram of apparatus for preparing sterile sea-water 
by filtration, without contact with metal. 


dimensions of our jars are 11 in. by 6 in., and the diameter of 
the opening at the top is 3Lin. ‘he tubulures are coned, 
with the smaller diameter external, and make a good fit for a 
No. 8 rubber bung. When setting up this apparatus a bung, 
through which a short glass tube bent at right angles is 
passed, is fitted into the side tubulure. This tube is con- 
nected, by means of rubber pressure-tubing, to another glass 

1 These jars were made to our specification by Messrs. Price, Powell, 
and Company, Bristol. The clamps usually supplied with such jars are 
not strong enough to obtain a tight joint, but these are easily repiaced 
by stronger ones. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS, 377 


tube leading down from the bottom of a small inverted bell- 
jar, placed some height above (in our case 14 ft., which gives 
a pressure of ca. 6 lb. to the square inch inside the jar). A 
serew pinch-cock on this connection serves as a tap. ‘The 
carboy containing the treated water stands just above the 
bell-jar, and is fitted with a tightly fitting rubber bung, 
through which two tubes pass. One is an ordinary syphon, 
the other the only air-inlet into the carboy. This latter auto- 
matically keeps the level of the water in the bell-jar constant, 
by closing the air-inlet as soon as the water covers the end 
of the tube. When filtering water the modus operandi is 
as follows: The carboy is filled with tank water, treated, and 
allowed to settle as before. The Berkefeld candle,’ bung, 
delivery tube, and connections (see fig. 1) are sterilised by 
boiling for half an hour, and fitted into place from within. 
(The delivery tube is shaped so that any drops of water, 
accidentally running down outside it, do not enter the vessel 
receiving the filtrate ; and the jar should be large enough to 
allow the hand to fit the filter into place without much 
trouble.) The pinch-cock is closed, and the syphon from the 
carboy started, which will automatically stop if the bung has 
been properly fitted. This should be watched to avoid acci- 
dents. The pinch-cock is then opened until the water rises 
in the jar well above the top of the candle, but still leaving 
some air-space. ‘he lid can now be fitted into place and 
screwed down. The tightness of this joint can be tested by 
pouring a little water into the crack round the lid, and 
observing if any bubbles are formed when the pinch-cock is 
opened. If all is right, no bubbles will be seen, and a good 
stream of water will flow out from the delivery tube. Our 
apparatus will filter about 20 litres an hour, and the filtrate 
is exceptionally bright and clear. ‘The candle should be 
sterilised every three or four days that the apparatus is in 
use to avoid indirect contamination by growths of organisms 


9 


through the substance of the filter.2, The water while passing 


1 No. 5 porcelain mount, length 8 in., diameter 2 in. 
2 See Bulloch and Craw, ‘ Journ. of Hygiene,’ vi, No. 3 (1906), p. 409. 


878 E. J. ALLEN AND E. W. NELSON. 


through this apparatus only comes into contact with glass, 
earthenware, and rubber, the use of metal having been pur- 
posely avoided. 

(r) Peroxide of Hydrogen Water.—As it seemed 
probable that the action of animal charcoal was due to contact 
oxidation with the oxygen occluded in the charcoal, experi- 
ments were made to determine whether a similar effect could 
be produced by the use of hydrogen peroxide (H,O;), This 
was used in two ways. In the first method a sufficient quantity 
of H,O, was added to the sea-water to ensure complete 
sterilisation (1 c.c. of H,O, of twenty vols. strength per 1000 
c.c, of tank-water was found to be satisfactory), and the 
excess of H,0, was decomposed by adding manganese dioxide, 
The water was then filtered through filter-cloth, and the 
filtrate appeared to remain quite sterile. Good cultures of 
Chetoceras constrictum, Biddulphia mobiliensis, 
and Skeletonema costatum were made in this water, 
which seemed to be as good as water treated by the animal 
charcoal method. 

The second way of using the peroxide of hydrogen was to 
start with water sterilised by heating to 70° C. and to add to 
this H,O,, in small quantities at a time, until its presence 
could just be detected on testing the sea-water with perman- 
ganate of potash. In these circumstances, the first amounts 
of H,O, are decomposed in the oxidation of organic substances 
in the water, and a very slight excess of H,O, persists. For 
tank-water 1 c.c. of one vol. H,O, per 1000 c.c. was found to 
give the best general effect. Cultures grown in water 
prepared in this way developed satisfactorily, being practi- 
cally equal to those made in animal-charcoal water, but they 
became exhausted rather quickly. 

The treatment of aquarium water with ozone was also tried, 
as this seems to offer a possibility of treating large quantities 
of water,! such as the whole bulk of water in an aquarium 


1 The use of ozonised air for the purification of fresh water for town 
water supplies has been adopted in some localities. (See Bridge, J. H., 
paper read before Franklin Institute, reprinted in ‘ English Mechanic,’ 
1907, pp. 369 and 392.) 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 379 


circulation, without very considerable expense. Experiments 
on a small scale, which we were able to make, unfortunately 
only with imperfect apparatus, showed that water treated with 
ozonised oxygen gave distinctly better cultures than untreated 
water. Although the sea-water was not absolutely sterilised 
by the treatment to which we actually subjected it, a sample 
of water which was visibly clonded with bacteria became 
quite clear and bright. 

(F) Cultures in these Media.—In order to make clear 
the different results which are obtained by using these 
different waters, we will describe the probable result which 
would be got from a series of flasks set up with the following 
media, and each inoculated with a persistent culture of a 
true plankton diatom, such as Thalassiosira, Skeleto- 
nema, or Chetoceras. 

A. “ Outside water” untreated. 

Small growth in from five to fifteen days, almost 
immediately swamped by growths of foreign forms ; 
the latter, however, will never be large. 

B. Ditto, sterilised. 

Shghtly larger growth, very soon becoming ex- 
hausted. 

c. * Tank-water ” untreated. 

Same result as in a, but growths will be much larger 
and healthier, and will last longer. 

p. ‘'Tank-water”’ sterilised. 

A fair growth of the inoculated species, but the total 
growth will not be as great as in c; the diatoms will 
retain their normal appearance for some time. 

E. ‘Outside water” + Miquel’s solutions A and B, sterilised. 

Best culture in series, both in quantity and quality. 
The diatoms will remain normal and healthy for a 
very long period. 

F. Outside water” sterilised and treated with animal 

charcoal. 

Fair growth, especially at first; diatoms will soon grow 
pale and become exhausted ; better than p. 


380 E. J. ALLEN AND E. W. NELSON. 


a. “Tank water” sterilised and treated with animal 
charcoal. 

As ¥, only growth will be slightly greater and will 
last considerably longer. Third best in series. 

u. ‘Tank-water ” treated with animal charcoal and filtered 
through Berkefeld filter. 

This will usually be the second best culture in the 
series, but the difference between this and G will 
only be slight. 

k. “ Outside water ” treated with H,O,. 

This will most resemble F, but will not be quite so 
good. 

L. “ Tank-water” treated with H,0O,. 

A distinct improvement over k. This medium is 
rather variable, and in some cases the growth 
obtained has been quite equal to r, if not better. 


zB. Experiments with a View to Determining the 
Conditions which underlie. the Successful 
Culture of Diatoms. 


The attempt to make cultures of diatoms for use as food 
when rearing pelagic larve, led naturally to an effort to 
determine the best culture medium and the most favourable 
conditions for the: rapid and continuous growth of diatoms. 
Before success can be attained in this direction exact know- 
ledge as to the nature of the essential food-stuffs, and, -in 
fact, as to the general physiology of the Diatomacez, is 
necessary.! Numerous experiments. extending over the last 
three years, have been carried out, with a view to obtaimng 
such knowledge, and the results, though still by no means 
complete or conclusive, are perhaps worth recording. 

A great difficulty which has to be metin carrying out such 
investigations on marine diatoms, is caused by the fact that 
when sea-water is used as a basis for the culture media, we 


1 For general references to literature see ‘* Bibliography,” especially 
Miquel (12), Richter (18). 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 381 


are dealing with a solution of a very complex and very 
variable character, the exact nature of which it is extremely 
difficult to determine. The most direct method of research, 
namely, chemical analysis, has not proved of much service, 
owing to the uncertainty, and in many cases impossibility, of 
accurate determinations, in sea-water, of such minute quan- 
tities of substances as those upon which the growth of 
plankton diatoms has been found to depend. 

We have had, therefore, to rely, for the most part, on the 
lengthy and tedious process of analysis by “trial and error,” 
the experiments being largely conducted on lines suggested 
by Liebig’s well-known “law of minimums” (Pfeffer, vol. 1, 
p. 413). The ideal at which we aim is to find a culture 
medium with artificially prepared sea-water as its basis, such 
that the absence, or diminution in quantity, of any one of its 
constituents would have a profound effect upon the growth of 
diatoms in it. Whether the conditions regulating growth in 
such a medium would be at all comparable to the natural 
conditions of life in the sea is a question that would have to 
be decided by experiment, but in any case this could be made 
a starting point for much more definite research than has yet 
been attempted. Up to the present time we have not, unfor- 
tunately, succeeded in finding such a culture medium. 
Throughout the work we have had very great difficulty, in 
spite of much care and many precautions, in obtaining 
consistent results. It may even happen that in two flasks 
containing the same culture medium, inoculated with the same 
culture of diatom and standing side by side, under exactly 
identical conditions, as far as can be recognised, quite 
different degrees of growth will be observed. All experiments 
must therefore be frequently repeated before entire confidence 
can be felt in any conclusions which they seem to indicate. 

It must be remembered, also, that in all the persistent 
cultures of diatoms that we have used, bacteria have pro- 
bably been present, and this fact has probably had some 
influence on the result. Unfortunately our attempts to 
obtain absolutely pure cultures have not met with success. 

VOL. 05, PART 2.—NEW SERIES. 25 


382 E. J. ALLEN AND E. W. NELSON. 


Methods.—In carrying out the experiments to be described 
in this section the procedure has been as follows: All media 
have been prepared from sterile sea-water, and sterile vessels 
and instruments have always been used. ‘The cultures have 
usually been made in 60 c.c. of liquid, in short-necked, wide- 
mouthed flasks of 125 c.c. capacity. When a number of 
cultures were to be compared, the flasks were kept standing 
in a row together in such a way as to keep the physical 
conditions as similar as possible. Control cultures in standard 
media were included in each series, so that results from 
different series could be compared by reference to the 
controls. The various media were inoculated from a persisteut 
culture of a species of plankton diatom, which in the great 
majority of cases was Thalassiosira decipiens (p. 412). 
When preparing the different media the methods used were, 
as far as possible, identical, and although only about 60 c.c. 
was needed for a culture, a litre was made up, so that errors 
due to-measuring very minute quantities might be avoided. 
The media were all freshly prepared for each comparative series 
of cultures, the same sample of sea-water being used, when 
the basis of any two or more was the same. Comparative 
estimates of the amount of growth in the different cultures 
were made by eye alone. Any difference between amounts of 
growth that has been described here as appreciable has always 
been accompanied by a marked difference in appearance to 
the eye on holding the cultures up to the light. A few drops 
from each culture were also, from time to time, examined micro- 
scopically, as a test of the quality and purity of the growth. 

The Sea-water Employed.—tThe sea-water employed as 
a basis for the culture media has been either (1) “ outside 
water” or (2) “tank-water.” A general description of these 
will be found on pp. 372-374. An accurate chemical analysis of 
both types of water would probably make clear many difficult 
points, but, as already pointed out, no chemical methods of 
sufficient delicacy have yet been devised. 

We have seen that if we compare “‘ tank-water,” i. e. water 
from the closed circulation of the Plymouth Aquarium, with 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 383 


off-shore sea-water in situ, a most obvious difference is the 
much increased density of the larger forms of animal life in the 
former, combined with the almost complete absence of plant 
life. Hence the concentration of excretory products in the 
tank-water must be very much higher than in outside water. 
Other factors, such as increased bacterial action, artificial 
aération, etc., in tank-water, must also be taken into account 
(cf. Vernon [58], Smith [561). There seems to be direct. 
evidence to show that the concentration of nitrates, possibly 
due to the action of nitrifying bacteria on the products of 
excretion, such as urea, ammonia, etc., is considerably higher 
in the tank-water, and the presence of soluble organic 
matter in concentrations never met with in the sea, can 
almost certainly be assumed. It is probably due to the 
presence of these nitrates and soluble organic substances that 
sterilised tank-water is a much better medium in which to 
grow diatoms than sterilised outside water (see p. 379). 

The Constituents of Miquel’s Solutions.—It has 
been already stated that no better medium for the culture of 
plankton diatoms has been found by us than the solutions 
recommended by Miquel, although these solutions may be 
modified and simplified in various ways with equally good 
results. The formulee recommended by Houghton Gill give 
very similar cultures. The essential features of Miquel’s and 
Houghton Gill’s methods, when adapted to sea-water, are the 
same. Miquel’s solution A and Gill’s solution 2, can both be 
replaced by a solution of potassium nitrate (p. 369). Again, 
Miquel’s solution B and Gill’s solution 1 only differ in the 
proportionate amounts in which the various constituents are 
prescribed. The formule are: 


Miquel’s sol. B. Hi. Gall’sisol.- 1. 
Na,HPO,,12H,0 : 4 germ. . : 2 grm. 
CaCl, 3 : : 4 ,, 2 5 Aye 
FeCl, (syrupus) : PAO ; : 0°. 5; 
HCl (concentrated) . raat : : oe 
Water ; 80, 5 LOOM 2 


Use 1 c.c. per 1000. Use 3 c.c. rer 1000. 


384, HK. J. ALLKN AND E. W. NELSON. 


The proportionate amounts added to equal volumes of sea 
water are : 


Miquel’s sol. B. H. Gill’ssol. 1, 
Na,HPO, . ; se lO > : _. ae 
CaCl, . ' ; AO) ag . 24 
FeCl, . ; ; ‘ 5 : : : 3 
HE 0G) ee : : 5) ; ; ; 6 


Since cultures can be obtained with no appreciable difference 
by using media prepared by adding either of these solutions, 
together with Miquel’s solution A, to sea-water, a con- 
siderable latitude in the proportions of the salts present is 
tolerated. 

We must now consider what is the rédle of the various 
constituents in Miquel sea-water. The part played by any 
salt of a culture medium may be considered as being either, 
firstly, “nutritive,” or secondly, “ protective.”! Under the 
first heading, any direct addition of food material must be 
included; under the second, any removal or neutralisation of 
harmful substances, such as toxins and possibly bacteria, and 
any more remote effects, which, although influencing growth, 
do not directly enter into the metabolism of the plant. 

Our experiments have proved that solution A can be 
reduced to a simple solution of potassium nitrate” without 
detriment (cf. p. 369), and that the amount of growth is, 
within limits, roughly proportional to the amount of KNO, 
added, as the following experiment shows :— 

Inoculated from persistent culture of Thalassiosira 
decipiens: 

A. Normal Miquel sea-water. 
Growth as usual. 
B. Ditto, but only half amount of solution A. 

1 Loeb, ‘ The Dynamics of Living Matter,’ New York, 1906, p. 77. 

2 For the sake of convenience the expression solution A will be used 
throughout the rest of this paper to indicate a simple solution of potas- 
sium nitrate (5 per cent.), and solution B to indicate Miquel’s phospho- 
ferri-calcic solution. Unless otherwise stated the amounts of each 
added to 1000 ¢.c. sea-water will be normal, i,e. 2 ¢.c. solution A and 
1 cc. solution B, 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 385 


Good growth at first, but exhausted sooner than A. 
c. Ditto, but two and a half times amount of solution A. 
Was slower than either A or B at start, but after- 
wards was better than a or B, and lasted longer. 
p. Ditto, but five times amount of solution A. 
As c, but in greater degree. 

Considering the nature of the substance added, and its 
already well-known action in piant metabolism, these results, 
coupled with the fact that exhausted cultures can often be 
regenerated by the simple addition of nitrates (see below, p. 
390), are quite consistent with the assumption that solution 
A is simply nutritive in action. The concentration of nitrates 
in natural sea-water is so low (Brandt [47]) that the amount 
available in a culture of untreated water very soon becomes 
completely exhausted, and it is this deficiency that solution A 
probably corrects. 

Considering now the action of solution B, it must first be 
observed that increased concentration of nitrates alone will 
not explain the whole action of Miquel’s solutions, for no 
increase in growth is obtained when nitrates or solution A 
only are added to sea-water. ‘To illustrate this point an 
account of an actual experiment may be given :— 

Inoculated with Thalassiosira decipiens: 

A. Normal Miquel sea-water. 
Good strong culture, in every way normal. 
B. Outside water sterilised. 
Small growth at first, very soon exhausted. 
c. Ditto + solution A. 
No improvement over B. 
p. Ditto + solution B. 
Fair growth. Great improvement on B and ¢, but 
exhausted considerably before a. 
E. ‘l'ank-water sterilised. 
Appreciably better than 8, but growth not large. 
F. Ditto + solution A. 
Not even as good as £. 
G. Ditto + solution B. 


386 E. J. ALLEN AND E. W. NELSON. 


Next best in series to a; lasted longer than p, and 
had better colour. 
_ To generalise, no improved culture is obtained with solution 
A alone, but a fair, though not very lasting, growth can result 
from using solution B only. 

The action of solution B is to some extent obscured by the 
fact that, when this solution is added to the alkaline sea- 
water, a precipitate is formed. ‘This precipitate is at first 
white, but, on heating or standing for some time, it becomes 
greenish-yellow. We are indebted to Mr, D. J. Matthews for 
the following analyses. 

Ten litres of normal Miquel sea-water were prepared, and 
the precipitate was collected on a filter-paper, washed, and 
dried at 100° C, 

Weight of dry precipitate from 10 litres = 0°2949 grm. 


Analysis of Dry Precipitate. 


P.O. ' : ; ‘ . 26°36 per cent. 
Wes. oe ‘ " ; . 41°31 © 
CaO : : : . ,; ee ES 
H,O ; : : 4 . 24°86 


100°16. . = 
Or the precipitate from 1 litre of normal Miquel sea-water 
contains— 
P.O; oes ‘ ; : . . °00777 grm. 
FeO, « ‘ ; ; . °01218 
CaO . ‘ ; : ~ *OO22R ae 
An analysis of 1 ¢c.c. Miquel solution B, the amount added 
to 1 litre Miquel sea-water, gave— 
12710 ae : : ; ; - *00825 
FeO... - - : : . °0105 
CaOi, : ‘ : : .! 2014S 
Comparing these figures it seems probable that, when added 
to sea-water, all the iron in solution B is precipitated, and 
a certain amount also of the phosphate and calcium. The 


3) 


2) 


33 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS; 387 


additive effect on the sea-water is, therefore, a slightly in- 
creased concentration of phosphate and calcium. 

“An analysis of a sample of tank-water for phosphorus, 
before and after treatment with solution B (1 c.c. per 1000), 
gave the following figures : 

Tank-water ‘5 mgrm. P per litre = ‘00163 grm. P.O. 

Tank-water + solution B (without precipitate) 1:5 mgrm. P 
per litre = ‘00488 grm. P,O;. 

It will be noticed that the figures from the different 
analyses do not agree very well. This is probably due to 
the fact that different samples were used for analysis in each 
case, and also to the fact that the solutions, were made up in 
the ordinary way, without any special precautions, volumes, for 
instance, being measured in cylindrical glasses, pipettes, etc. 

Cultures’ were tried in sea-water containing the normal 
amount of solution A, plus the normal constituents of solution 
B, less all the iron and less the amount of phosphate that would 
combine with the iron to form basic ferric phosphate (P.O, 
2Fe,03;12H,O). This solution should have very nearly the same 
chemical composition as normal Miquel sea-water from which 
the precipitate has been removed. Successful cultures could 
not, however, be obtained init. Neither could cultures be 
grown in sea-water to which had been added the normal 
amount of solution A and 1 mgrm. P (as sodium phosphate) 
per litre. 

To ascertain the effects of the different constituents of 
solution B, experiments were carried out with separate solu- 
tions of these constituents, each of the same strength, as in 
Miquel’s formula. Different combinations of these solutions 
were added, together with solution A, to sterilised sea-water, 
and the resulting media were inoculated in the usual way. It 
was found necessary to repeat these experiments a great 
number of times, as the results obtained were rather contra- 
dictory. ‘To illustrate the methods used a list of the different 
media, and notes of the cultures obtained in them, are given 
below. These media were inoculated from cultures of 
Thalassiosira decipiens, and the cultures were kept 


3888 ir. J. ALLEN AND E. W. NELSON. 


under observation for at least four months. Series were 
made as uniformly as possible, and controls in standard 
media were included in each. The strength of the various 
solutions used in these experiments was the same as in 
Miquel’s formula. 
A. Outside water + solution A + solution B (normal Miquel 
sea-water. 

First control. 

s. Outside water + solution A + Na,HPO, solution + 
FeCl, solution + CaCl, solution. 

Second control. 

Good normal cultures were always obtained in these two 
controls. 
c. Outside water + solution A + Na,HPO, solution. 

A very uncertain medium. Sometimes no growth has 
been recorded, and at other times a fair growth 
results, but these cultures are never equal to normal 
Miquel. 

p. Outside water + solution A + FeCl. solution. 

Occasionally a very small growth has been obtained, 
but at the best it is very poor. 

gr. Outside water + solution A + CaCl, solution. 

About equal to D. 

r. Outside water + solution A + NasHPO, solution + 
FeCl; solution. 

Uncertain as c. No cultures have been obtained equal 
to the best in c, 

g. Outside water + solution A+ Na,HPQ, solution + CaCl, 
solution. 

Some cultures very nearly equal to the controls have 
been obtained in this medium. 

H. Outside water + solution A + FeCl; solution + CaCl, 
solution. 

Poor, about equal to D. 

Analysing the above results we see that— 
(1) None of these modifications of solution B give results 
equal to solution B itself. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 3889 


(2) The best result is obtained from the combination of 
the phosphate and calcium chloride solutions. 

(3) Of the solutions used singly the phosphate is the 
best, the iron and calcium chloride being about equal. 

(4) The addition of FeCl, to Na,HPO,, or the addition of 
CaCl, to FeCl,, does not improve the medium to any extent. 

Experiments were also made to determine whether the 
precipitate thrown down in sea-water by Miquel’s solution B, 
itself had any influence on culture media. A quantity of this 
precipitate was prepared, filtered off, and then added to 
outside sea-water + solution A (nitrates). A small growth 
was obtained, which was a distinct improvement on the 
control without the precipitate, but exhaustion soon set in. 

Further discussion of the mode of action of solution B, 
and as to whether that action is purely nutritive, or partly 
nutritive and partly protective, is better postponed until a 
later section, after the action of animal charcoal and other 
substances has been considered (see p. 405). 

Animal Charcoal and Peroxide of Hydrogen.—The 
most successful culture medium for plankton diatoms, next 
to Miquel sea-water, is that prepared from animal charcoal 
(cf. p. 379). Animal charcoal water gives at first almost 
as good cultures of plankton diatoms as Miquel sea-water, 
but the tendency to paleness and exhaustion appears much 
sooner. ‘he best cultures were obtained in “ Berkefeld 
water,” that is, tank-water from the Plymouth Aquarium 
treated with powdered commercial animal charcoal and filtered 
through a Berkefeld filter. ‘Tank-water as a basis for animal 
charcoal water is very much better than outside water, 
probably on account of the higher concentration of nitrates, 
etc. 

There is a very striking resemblance between the effect of 
animal charcoal and of Miquel’s solution B upon sea-water 
used for diatom cultures, and the growths obtained by using 
tank-water + solution B and tank animal-charcoal water are 
very similar in character. If Miquel’s solution A is added to 
animal-charcoal water there is a great improvement, both in 


390 E. J. ALLEN AND E. W. NELSON. 


the colour and quantity of diatom growth, and in the case of 
Thalassiosira decipiens the chains are long and well 
formed. With animal-charcoal water + solution B, on the 
other hand, practically no growth was obtained. 

It is possible that a certain amount of phosphate, and 
perhaps of calcium, from the animal charcoal, goes into 
solution and serves as a ‘‘ nutritive ” material for the diatoms. 
But we are inclined to think that its chief action is “ protec- 
tive,” and due to its power of occluding gases, such gases 
being in a state of higher chemical activity than under 
normal conditions.! 

As was explained in a previous section (p. 378), the 
possibility that the action of animal charcoal might - have 
some sort of effect comparable to oxidation, led us to experi- 
ment with hydrogen peroxide. Fair growths of diatom could 
be obtained in sea-water prepared in the manner described, 
but they always showed a tendency to rather rapid exhaustion. 
As in the case of animal-charcoal water, tank-water proved 
a much better basis for treatment with H.O, than outside 
water. 

Reviving Exhausted Cultures.—Several experiments. 
were carried out with water from old, exhausted cultures. 
The sediment was filtered off, the filtrate was sterilised bs 
heat, and then treated by various methods. 

In one typical experiment the following was the result :— 

Water from an exhausted -culture of Skeletonema 
costatum in Miquel sea-water, reinoculated with the same 
diatom : , 

A. Filtered and sterilised. 

No growth obtained. 

B. Ditto + solution A (nitrates only). 

Good culture, but did not last very long; further 
addition of nitrates made no improvement. 

c. Ditto + solution B. 

1 Against this view would seem to be the fact that when powdered 


cocoa-nut charcoal, which has a still higher power of occluding gases, 
was used in place of animal charcoal, very poor cultures were obtained. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 391 


No growth. 
p. Ditto + solution A + solution B. 
Very good growth, lasting considerably longer 
than B. 
g. Ditto + animal charcoal. 
No growth. 

Exhausted cultures in animal charcoal water gave. the same 
general results on treatment and reinoculation. In an old 
culture of Biddulphia mobiliensis in outside water + 
solution B only, which was in a very exhausted condition 
(nine months old), the addition of KNO; gave a very rapid 
regeneration, and the diatoms became of normal colour and 
form. This renewed growth, however, did not last very long, 
and a further addition of KNOs did not give any result. The 
addition of sodium phosphate also failed to stimulate growth, 
The same rapid regeneration, on the addition of potassium 
nitrate, has been obtained with almost every medium, but a 
second attempt has always failed. 

Silica.—A very noticeable character of the true plankton 
species of marine diatoms is that their skeletons are very 
markedly less siliceous than the great majority of other forms. 
Their valves are only feebly marked, if at all, and they will 
not stand the vigorous treatment of cleaning with acids and 
heat that is commonly used in the case of fresh-water diatoms. 
In cultural forms this absence of silica is still more obvious, 
and no marking can usually be seen on even those forms, 
which, under natural conditions, are the most siliceous, e. g. 
Coseinodiscus excentricus. Deformed and distorted 
frustules are the rule in certain stages of growth in our 
cultures, and it is often very hard to make out more than the 
thinnest coating of silica. It is quite probable that this 
deformity can be accounted for simply by the absence of a 
strong siliceous skeleton. As a rule, the more rapid the 
growth the more -teratological forms will be found. In 
untreated outside water little deformity will take place, but 
in normal Miquel, where very rapid growth takes place, the 
diatoms may assume almost any conceivable shape. The 


392 E. J. ALLEN AND E. W. NELSON. 


form of the frustules tends to come back to the normal again, 
when the culture is well started, and in old stages the 
majority will be perfectly formed, although small and pale. 
It was found that the addition of silica (in early experiments 
as fragments of potassium silicate) was, as far as could be 
judged, immaterial, which fact led to the conclusion that a 
sufficiency dissolved out from the glass flasks in which the 
cultures were kept. During rapid growth, it is possible that 
the silica does not dissolve out fast enough to supply the 
demand, although it is also possible that diatoms, during rapid 
division, cannot absorb silica and form a perfect skeleton, 
even when the supply is abundant. Richter (18) has proved 
the necessity of either CaSi,O, or K,81,0; for the growth of 
Nitzschia palea, grown in pure cultures. We tried the 
addition of silica in various forms, and in one instance, in a 
culture of Coscinodiscus excentricus, to which a little 
precipitated calcium silicate had been added, the uniformity 
and markings of the valves were much more regular than in 
the control. ‘The presence of a trace of pure, dialysed silica 
also, in one experiment, gave an improved regularity of form, 
but the quantity or rapidity of growth did not seem to be 
affected. No sign of regeneration could be obtained in 
exhausted cultures by the addition of silica. 

Organic Infusions.—Miqnel recommends the use in 
culture media of infusions of organic substances, such as 
bran, straw, diatom broth, etc., in addition to the saline solu- 
tion. He does not make it quite clear if he ever dispensed 
with them at all. In his general directions he certainly 
states that the addition of both saline and organic nutrient 
material is necessary. As would be expected from the general 
metabolism of plants, the saline constituents are sufficient for 
growth. At the same time, excellent cultures have been 
obtained from dilute organic infusions, both with and without 
the addition of Miquel’s solutions A and B. About a square 
inch of Ulva was boiled in 600 c.c. of sea-water for half-an-hour, 
cooled, and filtered. In this medium an excellent growth of 
Coscinodiscus excentricus in one case, and Biddulphia 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 393 


mobiliensis in another, was obtained, the growth lasting 
for some considerable time. 

Infusions, made in the same way from a small piece of fresh 
fish, gave the same results, and although growth was rather 
slower at first, the final result was, if anything, slightly 
better. As Miquel points out, these infusions must be made 
very dilute, otherwise growths of bacteria, moulds, etc., will 
completely swamp the diatoms. Karsten (7), in some interest- 
ing experiments, showed that Nitzschia palea (Kutz) 
W. Sm. could be made to alter completely its mode of nutri- 
tion. On placing this diatom in organic nutrient solutions, 
it lost all chlorophyll and became colourless, but in saline 
media the chlorophyll would not regenerate, and the nutrition 
change back from heterotrophic to autotrophic.! 

Of course, with our infusions, it cannot be said that the 
diatoms were necessarily feeding on dissolved organic 
material, as some necessary, saline, nutritive materials conld 
have dissolved out from the weed or fish. If the former is 
the case, it might explain the superiority of tank-water over 
outside water, since the tank-water must contain a much 
higher percentage of organic substances in solution. If an 
alternative mode of nutrition, autotrophic or mixotrophic, 
could be proved, especially in the case of the ‘‘ bottom” 
forms of diatoms, a great many phenomena could be ex- 
plained, but the evidence is as yet far too slight to warrant 
any such assumption. 

Artificial Sea-water.—As we have explained in a 
previous section, the ideal aimed at in this part of our work 
has been to obtain strong growths of Diatomacez in purely 
artificially prepared solutions of simple salts. If this end could 
be satisfactorily attained the difficulties due to the unknown 
and variable composition of natural sea-water at once dis- 
appear. According to van ’t Hoff (85) sea-water is a solution 
containing salts in the following molecular concentrations: 
NaCl 100-0, KCl 2:2, MgCl, 7°8, MgSO, 3°8, CaCl, 1:0 (varies). 

1 Cf. Zumstein, ‘Zur Morphologie u. Physiologie d. Huglena 
gracilis,’ Leipzig, 1899. 


394 E. J. ALLEN AND E. W. NELSON. 


Using these molecular concentrations, a sea-water of any 
desired salinity can be prepared. ‘The chlorine content of 
average Atlantic water is about Cl = 19-4, and samples of 
artificial sea-water were prepared with the same chlorine 
value, thus : 


NaCl . , ; : ‘ : .  267a 
i S| ‘715 
MgCl, . 3°42 
CaCl, : ; : ; : 5] 
MesO, ; : : , : : Pg | 
Double distilled water ; ; . 966°47 
1000-00 


To make this solution comparable to natural sea-water, the 
“alkalinity”? must be raised by the addition of an alkali such 
as Na,CO;. After the importance of “ alkalinity ” asa factor 
had come before our notice, 2°4 c.c. M/, Na,CO, was always 
added to the above solution in order to make the amount of 
base in equilibrium with CO, equivalent to the usual 40 mgrm. 
OH vree 

The only success we attained with artificial sea-water as a 
basis for culture media was with four isolated cultures in one 
of our earlier experiments. ‘I'wo of these were cultures of 
Coscinodiscus excentricus in artificial sea-water + 
Miquel’s solutions A and B. The two cultures were identical 
except that one was in an ordinary bohemian glass flask and 
the other ina “ resistance glass” flask. No difference between 
these two could be seen. The growth obtained in both was in 
every way equal to normal Miquel sea-water, and is still fair, 
although over two years old. The other two successful 
cultures were growths of the same diatom in the same media, 
plus a small quantity of weed infusion, made by boiling up a 
small piece of Ulva in artificial sea-water. These gave just 
as good results, but the addition of unknown factors from the 
weed detracts from their general interest. In spite of 
frequent attempts, over fifty in number, we have not been 
able to repeat this experiment, which may possibly be due to 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 395 


some accidental impurity in the salts or distilled water from 
which the successful media were prepared. 

Alkalinity.—Tornée (48) and Dittmar (33) were the first 
to investigate the fact that sea-water showed on analysis an 
apparent excess of base over acid, which excess they termed 
“the alkalinity of sea-water.” Dittmar defines the alkalinity 
of sea-water as “‘a measure of its potential carbonate of 
lime,” but this definition and his supposition that this excess 
of base combines directly with dissolved CO, to form car- 
bonates and, further, but only in very small proportion, bicar- 
bonates, is hable to give a quite erroneous idea of the state of 
equilibrium actually occurring in the ocean. For, as Fox (84) 
has shown, “ sea-water reacts in situ very nearly neutral, and 
actually just slightly more acid than distilled water.” This 
is due to the fact that sea-water always contains a consider- 
able quantity of dissolved COQ. 

If a salt solution with neutral reaction, that is, containing 
H: and OH! ions in concentrations equal to one another and 
the same as for pure water, be exposed to an atmosphere con- 
taining CO,, a definite amount, depending on pressure, tem- 
perature, and salinity, would go into solution. This CO, 
would combine with water and form the very weak acid 
H,CO;, which would ionise with the formation of free H- ions, 
thus : 

HCO. 2 Be ECOG 


pbs 
(HCO! 2 - CO”,) 

The second stage of dissociation is so small as to be 
negligible. The concentration of H: being now increased 
and OH! decreased, the solution would have an acid reaction. 
The actual amount of CO, thus dissolved would always be 
small; for instance, a salt solution of strength Cl = 20-00 
(average Atlantic water Cl = 19:4) will at 10° C. dissolve 
about 3 c.c. CO, per litre from an atmosphere containing 
3 loo CO, (about normal). But the ocean is found to contain 
very much greater quantities than this, 60 c.c. or 200 times 
this amount being a not unusual figure for the total CO. 
The difference between this amount and the ‘3 c.c. or so dis- 


396 KE. J. ALLEN AND E. W. NELSON. 


solved by the neutral salt solution, as above, is kept in 
equilibrium with the 3 °/,,, CO, of the atmosphere by the 
amount of “excess” base equivalent to the amount of acid 
neutralised when an acid such as HCl is added to sea-water 
in excess, If a solution identical with sea-water but abso- 
lutely free from CO, (a practical chemical impossibility) could 
be obtained, then there would be present an excess of base 
over acid, and consequently an excess of OH! ions over H, 
ions, and an alkaline reaction. On exposing such a solution 
to the atmosphere, CO, would go into solution, ionise, and the 
H: ions thus set free would react with the OH! ions, due to 
the excess base, to form water. And this reaction would 
continue to take place, on more CO, dissolving, until all the 
excess OH! ions were neutralised, at which point the solution 
would react neutral. Now, as before with the neutral salt 
solution, a further small amount of CO, would go into solu- 
tion, bringing the solution into equilibriam with the atmos- 
phere, and the excess H- ious thus formed would give an 
acid reaction. ‘The final result would be a ‘solution exactly 
identical with natural sea-water. ‘he total CO, found in sea- 
water can be considered as existing in two parts: the larger 
part in equilibrium with free base, its amount depending on 
temperature, pressure, and alkalivity; the smaller in equili- 
briuin withthe partial pressure of CO, in the atmosphere, its 
amount depending on temperature, pressure, and salinity. 
Although sea-water in situ has an acid reaction, it still main- 
tains the property of being able to neutralise a certain amount 
of any acid stronger than H,CO,, that is, any acid which, on 
dissociation, forms a higher concentration of H° ions ; for the 
stronger acid will turn out the H,CO, in equilibrium with the 
“excess base” and CO, will be evolved. 

In consideration of these points, a less confusing definition 
> would perhaps be a 
measure of its potential capability of neutralising a 
strong acid! with the evolution of CO,. This can be 
conveniently expressed, as is usual, in mgrm. OH °/... 


of the “alkalinity of sea-water’ 


1 Such as HCl, with a high degree of ionization. 


ARTIFICIAL CULTURE OF MARINE PLANK'TON ORGANISMS, 397 


Some of our earlier experiments seemed to show that 
“alkalinity ”’ was a factor of considerable importance for the 
successful growth of cultures of plankton diatoms ; so an 
attempt was made to analyse the various samples of water 
both before and after treatment as culture media. The 
method adopted was a modification of that used by Tornde 
and Dittmar. Solutions of NaOH and H,SO, of strength 
N/;), by intention, were made up and stored in five-litre 
“aspirator” bottles. Two accurately graduated burettes 
standing side by side were connected to these by tubes, so 
that they could be readily filled by gravity. All air inlets 
to burettes and stock bottles were fitted with tubes of soda 
lime. A standard solution of Na,CO, of exactly known 
alkalinity, approximately that of average sea-water (40°00 
mgrm. OH °/..), was prepared by diluting down from a N/,,) 
solution, all operations being performed by weighing. These 
standards were stored in stoppered bottles of the fairly 
insoluble dark green glass, but those that had been kept for 
any length of time were not trusted, fresh standards being 
prepared, On analysis these standards agreed with one 
another to well within ‘1 mgrm. OH °/,,. The water used for 
diluting the standards was distilled water from the laboratory 
still, re-distilled in all-glass apparatus with potasstum bichro- 
mate and sulphuric acid. 

When ‘carrying out an analysis, equal volumes (about 100 
c¢.c.) of sample and standard were measured out into Jena 
glass Erlenmeyer flasks with a Knudsen automatic pipette. 
The specific gravity of each was determined by weighing in a 
25 c.c. pyknometer. Sampleand standard were then titrated 
by running in acid from the burette and back titrating with 
alkali, using a 1 per cent. alcoholic solution of aurine as an 
indicator and keeping the liquid boiling. The acid to alkali 
equivalent was determined by titrating a pipetteful of double 
distilled water in the same manner, The mean of at least four 
readings was always used. Let N and n be number of burette 
divisions of alkali equivalent to standard and sample respec- 
tively, and D and d their density at the time of pipetting out. 

VOL. 55, PART 2.—NEW SERIES. 26 


398 EK. J. ALLEN AND E. W. NELSON. 


Then if A is the alkalinity of the standard and X the required 
alkalinity of sample: 
Dn 
x= Nd 

Since all operations were carried out at the same room 
temperature, no corrections for temperature are necessary. 

In spite of the greatest care consistent results could not be 
obtained by this method of analysis. A sample analysed 
against the same standard would sometimes give results 
varying as much as 0°5 mgrm. and occasionally 1°0 mgrm. 
OH °/,,... The work on indicators by Salm (42) and its 
application to this question has only recently come to our 
notice, and it is our intention to experiment on this in future 
research. ‘lhe figures quoted below as the results of analyses 
have been rounded off as whole numbers, since their interest 
lies in their comparative rather than their absolute value. 
For convenience they are quoted as “ alkalinities,” although 
we are fully conscious that the methods used do not warrant 
this assumption, and that their actual chemical significance 
is still obscure. 

The mean value for ‘outside water’? was found to be 
fairly constant at 40°0 mgrm. OH °/,., which figure agrees 
with results obtained by others for average ocean water. 
Samples from the aquarium tanks never gave as high figures 
as this, the average being approximately 37°5 mgrm. OH °/... 
From this it seems that the amount of base in equilibrium 
with CO, in tank-water is appreciably less than in outside 
water. A series of thirteen samples taken from seven miles 
beyond the Eddystone to well inside the Cattewater (an 
inner tidal harbour near Plymouth) showed a gradual lowering 
of the alkalinity from the normal 40, to 388 mgms. OH °/,, as 
the water became more estuarine and polluted. 

The addition of Miquel’s solution B to sea-water was found, 
on analysis, to reduce the “alkalinity ” by an amount equiva- 
lent to 10 mgrm. OH °/,, or more. The 1 c.c. solution 
B added to a litre of sea-water in itself contains a certain 
amount of free acid, equivalent to less than 4 mgrm. OH °/,.. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 399 


But this reduction of alkalinity cannot be accounted for by 
the addition of free acid alone, because if only one quarter the 
amount of solution B is added, the alkalinity of the sample 
will be found to be, if anything, only very slightly higher. 
Also, if the various constituents of solution B are added as 
separate solutions, thus obviating any addition of free acid, a 
reduction equivalent to about 6 mgrm. OH °/,, is still obtained. 
The presence of ferric chloride in solution B gives a possible 
explanation of this phenomenon. Ifa solution of ferric chloride 
is added to a solution of a soluble carbonate, a reaction, 
which can be expressed by the following equation, takes place: 

3R,CO;Aq. + Fe,Cl,Aq. = 6 RC].Aqg. + Fe,0,Aq..+ 3 CO,. 

When the ferric chloride is added to sea-water, the final 
result will be that a certain amount of the “excess base,” 
which was in equilibrium with CO,, will then be in equilibrium 
with the chlorine, available on the precipitation of hydrated 
ferric oxide, with a consequent liberation of CO, and a 
reduction in ‘alkalinity ” will, therefore, take place. 

An analogy between the actions of Miquel’s solution B and 
animal charcoal can be seen in the fact that water treated 
with animal charcoal also shows a reduced “ alkalinity,” the 
amount being very variable in different samples. 

Sea-water treated with H,O, also showed a lowering of the 
alkalinity, but in a much less degree when, as usual, minimal 
quantities were used. 

Control experiments on double distilled water, which had 
been treated with these substances, were tried, but great 
difficulty was found in obtaining an end point with the 
indicator. As far as could be judged, distilled water treated 
with solution B (quantities as with sea-water) showed a 
negative “alkalinity,” equivalent to about 8 mgrm. OH°/,,, 
and in the case of animal charcoal a positive alkalinity 
equivalent to 6 mgrm. OH°/,,, but the colour change was so 
slow that these results are only the roughest estimates. The 
possibility that the above results are due to some effect on 
the indicator, which entirely cloaks the true alkalinity, must 
always be taken into consideration. 


4.00 E. J. ALLEN AND FE. W. NELSON. 


Before any attempts at analysis had been made, the proba- 
bility that considerable differences might be found in the 
alkalimity of the various media had presented itself, Im- 
provement in the growth of diatom cultures was found to 
result from the purely empirical addition of NaHCO,, this 
result being most marked in normal Miquel sea-water, outside 
water + solution B only, and Berkefeld water. No growth 
could be obtained in either ‘‘tank-water” or Miquel sea- 
water to which had been added 1 c.c. HCl (pure, concentrated) 
per litre, but on again raising the alkalinity of the latter 
by the addition of NaHCO, or KOH, good normal growths 
resulted. Richter (18) and H. Gill (5), also, both state 
that a weak alkaline reaction is necessary for the growth of 
diatoms, 

In our most recent experiments, all the media have been 
analysed for alkalinity, and those given in detail below 
illustrate the importance of determining this factor. Cultures 
of Thalassiosira decipiens were made in the following 
media ; 

A, ‘l'ank-water. Control. 

Poor growth, hardly normal. Later, good growth of 
minute forms, etc. 

B. Tank-water, treated with cold commercial animal char- 

coal, and filtered. 
Very good growth indeed. 

c. Tank-water treated with cold, pure animal charcoal, and 

filtered. 
Very poor growth, comparable to a without minute 
forms. 

p. ‘T'ank-water treated with pure animal charcoal as in ¢, 

but the animal charcoal was added red-hot. 
Fair growth, much superior to c, but not up to B. 

The sample of pure animal charcoal used here had been 
previously found to give very poor results, and it was also 
quite contrary to our experience that any improvement in 
growth should be obtained by adding it hot. But if we 
examine the results of analysis of these media for alkalinity 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 401 


a probable explanation presents itself. The following figures 
are only comparative : 
A, 38 mgrm. OH °/,, (used as standard), 


Bora dF 3, » (higher than-usual). 
cr ON as » (very low indeed). 
dD. 34 90s ” 


Tt will be seen that the amount of Beatie in each treated 
sample follows the alkalinity very closely. 

Solutions of Na,CO,, NaHCO; and HCl were made up, so that 
4.c.c. of any one contained an amount of acid or alkali equiva- 
lent to 10 mgrm. OH. From these a series of normal Miquel 
sea-waters of different alkalinities were prepared. Cultures 
of Thalassiosira decipiens were grown in these media. 

A. Normal Miquel sea-water. Control, A = 32°7 mgrm. 

OF */co- 
Perfectly normal growth. 
B. Ditto + 4 c.c. Na,CO; per litre. A = 41°7 megrm. 
OH °/,, (= + 90). | 
No difference between this culture and a. 
c, Ditto +.8 c.c. Na,CO3. per. litre, A = ;50'2 mgrm, 
OH = /5.. (==: 15), 
Best culture in series in quality and quantity. 
p. Ditto + 4 cc. NaHCO; per litre. A = 42:4 merm, 
OH °/.. (= + 9°7). 
Shghtly better than control. 
E. Ditto +. 8 c.c, NaHCOs, per litre. A = 51°5 mgrm, 
OE (S17 18i8). 


PAS! D. 
F, Ditto. +. 4 c.c. HCl per, litre, :- A = 22°2 mgrm. OH 
too (= = 10°5). 


Fair growth, but never up to control; exhausted 
much sooner. 
¢. Ditto + 8 c.c. HCl per litres A = 11:1 mgrm. OH 
fog (= — 21:6). 
Poorest in series. 


1 Figures in parentheses are difference in alkalinity from control, 
in mgrm, OH °/,,. 


402. , kK. J.. ALLEN "AND &£.’ W. NELSON. 


Except in the cases where the alkalinity was lowered by 
the addition of HCl, the results obtained from this series 
were not up to expectation. Nevertheless the majority showed 
a distinct improvement from increased “alkalinity,” and in 
c, where the alkalinity had been raised 17°5 mgrm. OH 
°/ooo this improvement was very marked. 

Another point illustrated by cultural experiment is that in 
two samples of animal-charcoal water, one with “ outside” and 
the other with “tank-water ” as a basis, the amount of growth 
in the latter considerably exceeded that in the former, and at 
the same time it was found that, with the tank-water, the 
alkalinity had not been reduced to the same extent as in the 
case of the outside water. 

How far apparently anomalous results, which have so 
frequently occurred in our experimental work, could be 
explained by unforeseen changes in “ alkalinity,” can only be 
answered by future research. 

Salinity.—The salinity (or amount of salts dissolved in 
1000 grm. sea-water) of the outside water used in these experi- 


ments only varied between small limits, S = 34°5 to 35°5 
“loo: ~=6' The salinity of “tank-water” is also fairly constant, 
the average being about S = 34°9°/,,; water is only pumped 


up into the reservoirs at high water, spring tides, and unless 
the salinity on analysis is weli above S = 345 °/., no water 
is taken. Experiments to show what effect salinity pure and 
simple had on the growth of diatoms were undertaken. 
Samples of sea-water of various salinities were prepared by 
diluting down “outside water” with double distilled water, 
and by concentrating “outside water” by slow evaporation. 
Two litres of “outside water,’ S = 349, were evaporated 
down to the bulk of one litre, giving a 50°/,! concentra- 
tion. Miquel solutions 4 c.c. A, 2 c.c. B, were now added, 
and the solution was divided into ten culture vessels, 20 
c.c. in each. Double distilled water was added, 2 c.c. to 
the first, 4 c.c. to the second, 20 c.c: to the last, so that a 
series of media were obtained, varying in salinity from 


1 i.e. from every 100 c.c. sea-water 50 c.c. H,O had been subtracted. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 403 


normal to nearly 50°/, concentration, each containing the 
same amount of Miquel’s nutrient solutions. These were 
inoculated from a mixed culture of Skeletonema costa- 
tum, Biddulphia mobiliensis, and Coscinodiscus 
excentricus. A good growth took place in all except the 
two with highest concentration. Of these two, the last 
remained praetically sterile and the growth in the other 
was very poor. ‘lhe limit of concentration, therefore, seems 
to lie between 35 and 40°/,. In the same way series of 
lowered salinities were prepared, and cultures of the same 
diatoms were grown in these. Dilution up to 100°/, did 
not seem to make any difference at all in the quantity 
or quality of growth. In a series extending the dilution 
to 200°/,, even in the cultures of lowest salinity a fair 
quantity of growth took place. ‘The range of salinities 
covered by the various series was S = 12 °/,, toS = 60 °,,, 
and within these limits no effect on growth could be observed, 
except in the very highest, where a distinct deterioration 
was noted. 

An attempt to grow Coscinodiscus excentricus in 
tap-water + Miquel’s solutions was tried, and it was thought 
that some slight multiplication took place, although it was 
certainly not at all considerable. Inoculating a culture of 
normal Miquel sea-water from this after six weeks gave no 
growth. 

Light.—Of all the factors controlling the rate of growth 
ofa culture, light seems to be by far the most important. With- 
out light a culture soon dies off completely, showing marked 
signs of malnutrition very soon after having been placed in 
the dark, the brown pigment being the first to go and later 
the chlorophyll. A culture (Thalassiosira) placed in the 
dark for five months was found to be completely killed, the 
diatoms being quite colourless. In cultures kept in bulbous 
flasks or in any spherical vessel, the strongest and earliest 
growth always takes place at the side of the vessel away from 
the source of light, where the light will be found to be con- 
centrated owing to the lens effect of a sphere of water. By 


404 E. J. ALLEN AND E. W. NELSON, 


painting a flask black onthe outside up to the water-line of the 
medium, a very marked diminution in the rate of growth 
was obtained. The total growth was not affected, but depends 
on the available quantity of food-stuffs present. a 

Experiments on the reaction of cultures to different rays of 
the spectrum, obtained by coloured glass, were tried, but no 
results obtained. Miquel obtained marked results with yellow 
light, but in our experiments, with plankton diatoms, satisfac- 
tory cultures could not be obtained under these conditions. 

Temperature.—The highest temperature which diatoms 
and allied forms can stand was about uniform for all the 
species tested, and lay between 35°-40° C. Cultures of the 
following species, viz. Asterionella japonica, Nitzschia 
closterium, minute naviculoid diatom, Pleurococecus 
mucosus, Chilomonas sp., were slowly heated in a water 
bath, and at every rise of 5° C. from 15° C. to 45° C. a few 
drops of the culture were pipetted out and a fresh flask 
inoculated. In all the flasks cultures were obtained where 
the heating process had not been carried above 35° C., but 
none in those where the temperature had exceeded this. 

In the earlier stages of experimentation the cultures of 
diatoms were kept in various places about the laboratory, and 
so were under quite different temperature conditions. Those 
placed in the warmer situations, i. e. near hot-water pipes, as 
arule gave the most satisfactory results. In all the later work 
the cultures have been kept in one room, and an attempt has 
been made to keep the temperature of this room as nearly as pos- 
sible constant at 15° C. A continuous record of its temperature 
has been kept by means of a recording thermograph, and no 
very great change of temperature has been noted. In a few 
isolated cases the temperature has dropped as low as 9° C., 
and in hot weather has risen just above 20° C., but these have 
only been for very short periods, the average temperature 
having kept remarkably constant. An apparatus in which 
flasks could be kept at different uniform temperatures from 
10° to 25° C., by means of hot air, was used, but no really 
satisfactory result could be obtained. About 17° C. seemed 


ARVLIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 405 


to give the maximum growth, and the cultures below this 
temperature were usually superior to those above. 

General Conclusions.—The general conclusions to be 
drawn from the experiments described in this section, which 
were made with a view. to determining the conditions that 
underlie the successful culture of diatoms, may now be dis- 
cussed. Although the experiments have involved the making 
of some 750 different cultures, our conclusions on many of the 
questions raised are still indefinite, and much further work 
will be necessary before a satisfactory answer can be given 
to them. 

If we wish to obtain the maximum quantity of healthy 
growth of a plankton diatom, the diatom must first be 
obtained as free as possible from all other organisms, if not 
in a “pure” culture, at least in a “persistent” culture. All 
culture media should be sterilised either by heat or filtration, 
and the experiments should be conducted under sterile condi- 
tions. Starting with normal sea-water as the basis for the 
culture medium, it seems to be first necessary to raise the 
concentration of the nitrates, and possibly also of the phos- 
phates, in solution. But this simple addition of nutrient 
materials will not in itself suffice. Some other action, such 
as that exerted by Miquel’s solution B, by animal charcoal, 
or by peroxide of hydrogen, seems to be imperative in nearly 
every case. ‘I'he exact nature of this action we have not 
been able conclusively to determine. If the substances con- 
tained in solution B were purely nutritive in character, we 
should expect that, when alterations in the amounts of the 
different ingredients were made, or when any one of the 
ingredients was omitted altogether, the differences in the 
quantity of growth would show a direct relation to the kind 
of modification introduced. But our usual experience has 
been that solution B can be modified within certain limits, 
without producing any appreciable effect upon the resulting 
cultures, whilst, if these limits are exceeded, there is an 
almost complete inhibition of growth. In supplying a neces- 
sary increase of phosphates, both Miquel’s solution B and 


406. HE. J.. ALLEN AND E. W. NELSON. 


animal charcoal may, and probably do, act as ‘‘ nutritive ” sub- 
stances ; but, since the addition of phosphates alone does not 
yield cultures comparable with those produced by either of 
them, and since, excepting phosphates, there is no possible 
common nutritive substance in their composition, we are led 
to conclude that, in addition to any nutritive effect, they 
must exert some other action. This view is supported by the 
results obtained by using H,O,. This substance cannot be 
directly ‘‘nutritive,” although it may be so indirectly, by 
oxidising into useful food-material substances which the 
diatoms are incapable of using in their metabolism, e.g. 
nitrites into nitrates. The absence of any increase in phos- 
phates, when using H,0,, may possibly be the reason why 
better results were not obtained with this medium. The 
action, which, in addition to any nutritive value, we must 
assume that solution B, animal charcoal, and H,O, can all 
effect, would appear to fall into the class of “ protective ”’ 
actions (p. 884). It is quite conceivable that, with different 
samples of sea-water, this “ protective” action is not neces- 
sary in every case, and this wouid account for the anomalous 
results met with when using sea-water + nitrates + phos- 
phates only, in which medium sometimes good cultures, but 
more often the reverse, are obtained. The effect of Miquel’s 
solution B, animal charcoal, and H,O, on the “ alkalinity ” of 
the sea-water, also points to some chemical change, which 
does not directly enter into the metabolism of the plants. 

It may be pointed out that the action of such substances as 
finely powdered carbon, and ferric oxide precipitates, have 
been shown to produce a favourable effect on nutrient solu- 
tions used for the culture of certain higher plants, and it has 
been suggested that the beneficial action of these substances 
is the removal of toxic elements from the media (Breazeale [3]). 
Such removal of toxins would fall under our definition of 
** protective” action. 

Of nutritive substances, other than those already mentioned, 
we have still to consider, (1) silica, and (2) dissolved oxygen 
and carbonic acid. Having regard to the conditions under 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 407 


which our cultures have been grown, i.e. in glass flasks, 
the question of silica does not seem to enter into the problems 
which we have discussed.. A few words must, however, be 
said as to the dissolved gases. Whipple (62) and Baldwin (44) 
have drawn attention to the observed relations, which are 
found in natural waters, between algal growths and the 
amounts of dissolved oxygen and carbonic acid. That these 
factors are of great importance cannot be doubted, but in our 
cultures it seems reasonable to suppose that the conditions of 
saturation of these gases are the same in all, since series of 
cultures in standard media, such as Miquel sea-water or 
Berkefeld water, can be set up with the certainty that, if not 
every one, at least a very high percentage, will give normal 
results. 

Of the purely physical factors, ight is by far the most 
important. Within limits, the rate of growth in a suitable 
medium seems to depend directly on the intensity of the 
light (Whipple [60]). Absence of light, as would be 
expected, soon completely kills the diatoms. 

Temperature also seems to affect the rate of growth to a 
certain extent, but for those temperatures at which we have 
experimented it does not appear to alter the quantity of 
growth. 

Salinity, apart from the quantities of available nutrient 
materials, can be varied within large limits without appreci- 
able effect on the diatoms. 


Il, Mixep CuLtuREs; 


In what has been said up to the present, we have been 
dealing with persistent cultures containing a single species of 
diatom, which are comparatively, if not entirely, free from 
admixture of other organisms. The study of cultures which 
contain a considerable mixture of organisms is not without 
interest. 

-A number of experiments have been made on the following 
lines: About 10,000 ¢c.c. of water, taken at some distance 


408 EK. J. ALLEN AND E. W. NELSON. 


from shore, was placed in a tall bell-jar fitted with a 
“plunger,” which keeps the water in constant movement 
(‘Journ. Mar. Biol. Assoce.,’ vol. v, p. 176). The water was 
treated with Miquel’s solutions in normal proportions, and a 
considerable quantity of plankton taken with a fine-meshed 
net (150 meshes to the inch) was added, say 10 or 20 c.c. of 
a moderately rich sample of tow-netting. ‘The experiments 
were made during the spring and summer months, and the 
general course of events has been the same, with a certain 
amount of difference in detail according to the nature of the 
plankton present at the time. 

During the first two days the water often became cloudy, 
owing to the rapid multiplication of small flagellate infusoria, 
though this was not always the case. Plankton copepods and 
other animals gradually died off, though some survived for as 
long as a week or ten days. ‘The plankton diatoms, on the 
other hand, generally multiplied rapidly during the early 
days of the experiments, the first to become abundant in the 
body of the water being usually Skeletonema costatum, 
which at the end of a week might be so thick that a number 
of chains could be seen in every drop of water examined with 
the microscope. Along with the Skeletonema were found, 
other plankton diatoms, such as Lauderia borealis, 
Chetoceras (two or three species), Biddulphia mobi- 
liensis, Ditylium Brightwellii, and in nearly every case 
Thalassiosira decipiens. These latter diatoms were pre- 
sentin moderate numbers only, when the Skeletonema was at 
its height, butas the Skeletonema died down they increased 
in quantity. At the same time Nitzschia closterium com- 
menced to appear, both amongst the precipitate on the bottom 
of the jar and in the general body of the water. Small green 
flagellates often began to get numerous also at this stage. 
The true plankton diatoms were at their height about a fort- 
night after the experiments were started. At this time a 
great many diatoms of all kinds were to be found amongst 
the precipitate at the bottom of the jar, Asterionella 
japonica and Coscinodiscus excentricus being often 


ARTIFICIAL GULTURE OF MARINE PLANKTON ORGANISMS, 409 


numerous here. During the course of the next week, how- 
ever, Nitzschia closterium rapidly increased in quantity, 
until not only the sides of the jar were coated with it, but the 
whole mass of the water became thick and opaque. By this 
time the plankton diatoms had all disappeared, with the 
exception of those which may survive for a considerable 
period amongst the precipitate at the bottom of the jar. 
Bottom diatoms (Navicula, etc.) had begun to grow on the 
sides of the jar, and small green and brown alge (Pleuro- 
coccus mucosus, Hctocarpus, etc.) also appeared. 
Infusoria (Euplotes and other smaller forms) then became 
numerous, and as the Nitzschia and bottom diatoms in- 
creased on the glass, large numbers of Amcebe made their 
appearance among them. ‘The jars continued in this con- 
dition for many months, the algze becoming more and more 
predominant. 

From these experiments, as well as from instances of mixed 
cultures obtained in the course of our attempts to secure 
persistent cultures of single species of diatoms, it seems usnal 
that, in a culture obtained by inoculating Miquel sea-water 
with plankton taken freshly from the sea, the true plankton 
diatoms are the first to develop in considerable numbers. 
Subsequently bottom diatoms and alge of various kinds 
become abundant, and the true plankton forms die out. 

A complete explanation of this sequence of events would 
probably be of a very complicated character, and we have 
practically no evidence from our experiments which bears 
very directly on the question. It would seem, however, that 
the early predominance of the plankton forms in the cultures 
would naturally follow from the fact ‘that, in the plankton 
material used for inoculation, these plankton forms are 
numerous, whilst bottom diatoms and spores of alge are rare. 
‘he subsequent very great predominance of such a species as 
Nitzschia closterium may be due simply to a very much 
more rapid growth rate, though it is difficult to avoid the 
impression that the organisms, which finally take possession 
of the cultures, are in some way directly inimical to those 


410 E. J. ALLEN AND E. W. NELSON. 


which they supersede, not merely by robbing them of their 
food supply, but perhaps, also, by the production of toxic 
substances. This suggestion does not, however, give an 
adequate explanation of the essential facts concerning these 
organisms. We have to consider two sets of species—(1) the 
true plankton forms, which flourish in the open sea and can 
be grown quite easily in the laboratory, provided the cultures 
yemain pure, and (2) what we may call “aquarium” or 
“‘hottom forms,” which under experimental conditions invari- 
ably take possession, when present in mixed cultures, whilst 
the plankton forms are killed off. Why is it that, although 
species of the second class are always present in small 
numbers in plankton taken from the sea, they are there alto- 
gether outnumbered by the true plankton forms, whereas 
under conditions such as those of our experiments they 
invariably succeed in gaining the upper hand? What are 
the factors which determine the difference in behaviour of 
these two sets of organisms in the sea and in the culture 
vessels? The whole question offers a very fruitful field for 
further experiment. ‘The evidence at present available is so 
slight that further discussion of it here is not likely to be of 
much service. 

The details of two experiments which we have made 
bearing on the subject of mixed cultures may, however, be 
recorded. 

A flask, containing about 1000 ¢.c. of sea-water treated 
with Miquel’s solutions, was inoculated with approximately 
equal amounts of certain persistent cultures of diatoms, which 
we possessed at the time. The following diatoms were in 
this way, introduced: Chetoceras constrictum, Bid- 
dulphia mobiliensis, Skeletonema costatum, Cos- 
cinodiscus excentricus, Streptotheca thamensis. 
'The flagellate (Chilomonas sp.) was also introduced, since 
it was present in the culture of Coscinodiscus. The ex- 
periment was started on August 26th, 1907. On September 
6th (11 days) Biddulphia, Coscinodiscus and Cheto- 
ceras were increasing rapidly and were very healthy, 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 411] 


Skeletonema was not so good, and no Streptotheca was 
found, 

On October 2nd (37 days) Biddulphia was numerous and 
healthy, Coscinodiscus was healthy but not so numerous, 
Skeletonema was poor, and Chetoceras was not seen. 
Flagellates (Chilomonas) had become very numerous. 

On October 31st (66 days) all the diatoms were in very 
poor condition, Coscinodiscus being slightly better than 
the others. The flagellates (Chilomonas) were extremely 
thick, giving the water a deep red colour. 

Subsequently a small green alga (Pleurococcus 
mucosus) appeared, having probably been derived from the 
Coscinodiscus culture. ‘This increased very greatly in 
quantity, whilst the flagellates become inconspicuous. 

On July 28th, 1909 (1 year 11 months) some Coscino- 
discus, which were still in a healthy condition, were seen in a 
sample examined from the flask. A great quantity of 
Pleurococcus, in a healthy condition, was also present, but 
no other oganisms were noted. On tlis date a subculture 
was made from the flask in normal outside Miquel. The 
subculture gave a considerable growth of Skeletonema, 
the cells being, however, of a very abnormal character, and a 
good many normal and healthy Coscinodiscus were found 
in each sample examined. ‘lhe whole culture was crowded with 
Chilomonas in a very active state, which gave the whole 
contents of the flask a deep red-brown colour. Upto August 
24th the green alga (Pleurococcus) had not become suffi- 
ciently abundant to be detected by the naked-eye appearance 
of the flask, though it could be seen in samples examined with 
the microscope. 

In another experiment a flask of Miquel sea-water was 
inoculated (May 4th, 1908) from two cultures, one containing 
the green alga (Pleurococcus mucosus) and the other 
Thalassiosira decipiens. At first both did well, and on 
May 20th (16 days) there was a very good crop both of the 
diatom and the alga. Gradually, however, the alga became 
predominant, and on October 14th (163 days) only quite empty 


412 E. J. ALLEN AND FE. W. NELSON. 


frustules of Thalassiosira could be found, whilst the 
growth of Pleurococcus was abundant and healthy, The 
only case where a diatom was observed to flourish in the 
presence of this green alga was in a culture of Nitzschia, 
a bottom form, In this case a very abundant growth of the 
diatom was obtained, but the Pleurococcus did not 
multiply to any extent although it could always be found on 
microscopic examination. 


III. Nores on Parricutar Species or Diaroms, ON THEIR 
Meruops or Repropucrion, AND ON OTHER ALGm OccuR- 
RING IN CULTURES. 


A list has been already given (p. 367) of those species of 
diatoms which we have obtained in “ persistent”? cultures. 
Of these a species belonging to the genus Thalassiosira 
has been used for experimental work in the great majority of 
cases. We are not quite certain as to the identity of the 
species, but since it most resembles ‘I’. decipiens Grun. we 
have called it by that name, although it does not exactly con- 
form to the published descriptions of that form. The most 
characteristic feature of this particular species is the eccentric 
markings on the valves, which are also seen on the valves of 
the diatom Coscinodiscus excentricus Ehr., and, as is 
typical of the genus, the frustules are united into chains by 
a delicate filament. Jérgensen (50, p. 96) describes the valves 
as “decidedly convex,” Gran (49) as “‘ flat,” and both agree 
that there are marginal spines and a single asymmetrical 
spine. Our cultural forms are united together by a filament 
into chains, some of which are made up of 500 cells and more, 
but the distance between each is considerably smaller than 
that figured by Gran. The valves are quite flat and the 
marginal spines are often present, although this is not always 
the case. The odd, asymmetrical apiculus can nearly always 
be seen. The eccentric markings have only been observed in 
a few isolated cases, and are then usually very indistinct. In 
one culture these markings on the vatves were yery distinct, 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 413 


and were also easily seen on the megafrustules (cf. below), 
which developed in it, but in none of the several generations 
of cultures started from this one have we been able to find 
any traces of marking at all. ‘he genus seems to be in con- 
siderable confusion, and it is probable that the conflicting 
descriptions given by different observers are due to variations 
in what is really one species. 

Persistent cultures of Coscinodiscus excentricus Khr. 
have also been obtained, and it is interesting to note that this 
diatom sometimes forms chains, but they are rather excep- 
tional. ‘hese chains are never as long as those commonly 
found with Thalassiosira, two or four cells only being the 
rule. ‘lhe filament joining the valves is also finer and more 
easily broken. The two species are quite distinct, and cultures 
of them can be discriminated by a practised eye. _ 

Two species of the genus Biddulphia are commonly 
met with in our cultures, namely Biddulphia mobiliensis 
(Bail.) Grun. and Biddulphia regia M. Schultze. These 
two forms are generally regarded as one species, but Osten- 
feld (54) has recently shown that they are really distinct. 
We have obtained persistent cultures of both forms from 
several different samples of plankton, and the two species are 
easily recognisable, never merging into one another. When 
Petri dishes, moculated from plankton (see p. 367), contain 
both species, the colonies can be easily distinguished with a 
small hand lens. 

The most generally accepted theory of the reproduction of 
the Diatomacee is briefly that the cells divide by simple fission, 
but on account of the rigid character of the cell-walls each 
division necessitates a decrease in size of the new valve, since 
this must always be formed inside the old valve. So the 
frustules gradually get smaller and smaller as multiplication 
proceeds, thus necessitating some process by which the 
original size can be re-established. This takes place by the 
formation of what are known as auxospores, which ultimately 
form megafrustules, and these in turn multiply by division 
until the minimum limit of size has again been reached. 

VOL. 55, PART 2.—NEW SERIES. 27 


414 E. J. ALLEN AND E. W. NELSON. 


‘There are also several special processes of reproduction, but 
no occurrence of any of these has been noted in our work 
(cf. Miquel [14]). 

The diatoms in our cultures multiply by simple. fission, and 
although there is, in nearly every case, a considerable diminu- 
tion in size when compared with specimens from the plankton, 
this diminution soon seems to reach a limit, where further 
decrease does not take place. In chains of Thalassiosira, 
several hundred cells in length, no difference in size between 
individuals could be made out. Auxospores are commonly 
formed with every species, but only in cultures of Coscino- 
discus and Thalassiosira have megafrustules been found, 
and in these they are very exceptional. These megafrustules 
seem to divide once or twice and then die or form new auxo- 
spores. What exactly is the fate of these auxospores, which 
are often exceedingly numerous, we have not been able to 
make out. It seems that cultural conditions are not favour- 
able to this mode of reproduction, and that the auxospores do 
not further the multiplication of the diatom at all. If this 
were not the case, stages of the formation of auxospores into 
frustules must have been seen in some at least of the very 
numerous samples examined. As it is, what has been seem to 
take place is, that the cell contents expand and force apart the 
valves of the diatom and emerge as a spherical body about 
three or four times the diameter of the parent cell. The 
chromatophores and diatomin then collect to one side, form- 
ing a compact cap against the cell-wall. Beyond this point 
no stages have been found, except in the case of the few 
cultures where megafrustules were formed. In these the 
chromatophores, etc., gradually formed into the shape of the 
diatom (Coscinodiscus); the siliceous coat with plain 
eccentric markings was easily seen inside the spore; and 
lastly, the cell-wall of the spore burst, leaving the mega- 
frustule free. The megafrustule was measured and found to 
have a diameter three times that of the parent cell. 

In the case of the diatom we have very largely used for 
feeding larve, etc., namely Nitzschia closterium, forma 


ARTIFICIAL CULTURE OF MARINE PLANK'TTON ORGANISMS. 415 


minutissima, a great number of cultures have been made, 
all originating. from the single drop from which -the first 
persistent culture was obtained. The total amount of growth 
in all the various cultures has been enormous, and the number 
of generations must be quite inconceivable. No diminution in 
size has, however, been appreciable, and no sign of any method 
of re-establishment of size has been seen, although these cul- 
tures have been under constant observation for over two years. 
This seems to prove that the theory of gradual decrease in 
size with successive generations cannot be generally applied. 

The following experiment on the rate of multiplication of 
Thalassiosira in normal Miquel sea-water was carried out. 
A single drop from a fresh and vigorous culture was kept 
under a microscope as a hanging-drop preparation in a moist 
chamber. ‘The number of diatoms in this drop was counted 
from time to time and the results are given in the following 
table : 


Number of Geometric 
Day. frustules. progression. 
11th : : ; 59 : : 63 
14th ‘ 7 ; 62 : 68 
19th < : : 85 : : 85 
27th : : 140 : ‘ 120 
34th : : : 170 ; : 160 
41st : : : 190 4 ; 220 


The curve obtained by plotting the number of diatoms 
against the number of days approximates the curve of an 
ordinary geometric progression, where the ratio is 2 and 
the periods are equal to sixteen days. ‘To show this the 
figures read off from the curve at the same intervals as 
the diatoms are appended in the table. From this it will be 
seen that, after a start had been made and before exhaustion 
set in, the numbers obtained agree fairly closely with the 
assumption that every diatom divided once in a period of 
sixteen days. Probably in normal cultural conditions the 
rate of multiplication greatly exceeds this figure on account 
of better lighting, etc. (ef. Miquel 12). 


416 EK. J. ALLEN AND E. W. NELSON. 


Besides diatoms, many other organisms appear in these 
cultures. We are. indebted to. Mr. G. 8. West for the 
identification of a form of unicellular alga, which is very 
common and difficult to avoid when attempting to obtain 
persistent cultures of the Diatomacex, namely, Pleuro- 
coccus muvcosus (Kutz.) Rabenh. This small green alga, if 
once introduced into a culture of a plankton diatom, will soon 
multiply at the expense of the latter with its ultimate extinc- 
tion. It is very hardy, and cultures. of it in almost every 
medium seem to last indefinitely. Multiplication beyond a 
certain point probably does not occur, but the cells retain 
their colour and normal shape, and will start active repro- 
duction if suitable nutrient material is provided. 

In cultures inoculated from plankton, many other forms of 
unicellular and filamentous alge thrive. Several species 
belonging to the classes Rhodophycex and Myxophycew 
commonly occur, but we have not been able to identify them. 
The most usual filamentous forms of Chlorophycezx are 
Enteromorpha, Vauchera, Rhizoclonium, ete. It is 
interesting to note that it was the unintentional appearance 
of young plants of Laminaria digitata in some of our 
Petri dishes that led Mr. Drew (4) to cultivate this alga in 
Miquel sea-water and so discover its early life-history. 
Cultivations of marine alge by these methods would without 
doubt yield many new species, and would also provide rich 
material for the study of their modes of reproduction. 

Many forms of flagellates live either together with diatoms 
or alone. Among: these is an unidentified species of Chilo- 
monas, which we have obtained in persistent culture. It 
multiplies very rapidly, colouring the whole medium a deep 
red-brown. It flourishes in Miquel sea-water and its nutrition 
is evidently autotrophic. In one culture, in Miquel sea-water 
inoculated with plankton, a number of coccospheres developed, 
probably Coccospheraatlantica Ostenf. Other flagellates 
and ciliated infusoria are very commonly met with, such as 
Bodo, Euplotes, Euglena, etc., which all seem to depend 
on the diatoms or other vegetable organisms for their food 
material. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 417 


IV. THe Rearina or Marine Larva. 


In the rearing of pelagic larval forms of marine animals,! 
the principle which we have followed has been to introduce 
into pure, sterile sea-water the larve to be reared, together 
with a-pure culture of a suitable food. As far as practicable 
all other organisms have been excluded from the rearing 
vessels. It should beadded that the food used in all successful 
experiments has been of a vegetable nature, and has continued 
to grow actively in the vessels. This is important from the 
point of view of oxygen supply. Under the above conditions, 
or rather under the nearest approach to them at which we 
have-been able to arrive, no change of water has been found 
necessary. 

Methods.—It will, perhaps, best make the matter plain 
if we first of all describe the actual procedure, which we now 
follow in the case of such ananimalas Hchinus esculentus 
or K.acutus. ‘The water to be used is first of all prepared 
by treating water from the ‘aquarium tanks with powdered 
animal charcoal, filtering it through a Berkefeld filter (p. 375), 
and collecting it in sterilised glass vessels. All instruments 
and pipettes are sterilised by baking in an oven, and a fresh 
sterile pipette is used for each operation during the progress 
of the work. Specimens of Echinus are then opened until 
a perfectly ripe female has been found, that is to say, one in 
which the eggs separate quite freely when a portion of the 
ovary is shaken in sea-water. 

Pieces of ovary, taken from a little below the exposed 
surface, are then placed in sterile sea-water in a shallow glass 
dish, and shaken with forceps in order to get the eggs well 
separated, or a number of eggs from the centre of the ovary 
are drawn up with a pipette and placed in the water. A 
very small quantity of active sperm from a ripe male is then 
added, very little being sufficient to fertilise a large number 
ot eggs. Excess of sperm should be avoided owing to its 

1 See ** Bibliography,” especially Grave (26), MacBride (28-80), Don- 
caster (25), ete. 


418 rE. J. ALLEN AND KE. 'W. NELSON. 


lability to putrefy. After an interval of ten or fifteen 
minutes the water, containing the eggs, is filtered through 
bolting silk of 100 meshes per inch, which just allows single 
eggs to pass through, whilst keeping back clusters of eggs 
or other large material. The filtrate is divided amongst a 
number of tall narrow beakers containing sterile sea-water, 
and the beakers, after being covered with a glass plate, are 
placed where the temperature will be uniform and not rise 
much above 15°C. In the course of twenty-four hours the 
healthy larvee will swim up to the surface and can be easily 
seen and removed from vessels of this shape. They are 
transferred by means of sterile pipettes to jars! of sterile 
sea-water, about fifty to seventy larvae being put in each jar 
of 2000 c.c. sea-water. At the same time, a good pipetteful 
of a pure culture of diatom is added to each jar. The small 
diatom Nitzschia closterium, forma minutissima we 
have found most useful, as its size is suitable, and it grows 
well in animal-charcoal tank-water, floating throughout the 
body of the water, and so being in intimate admixture with 
the larves. The jars are placed in a moderate light and at as 
even a temperature as possible.” No further attention is 
necessary until the larvee have metamorphosed. ‘The meta- 
morphosis takes place in from six to nine weeks after 
fertilisation. Larvee may be taken out from time to time and 
examined to see if they are feeding well. If the diatoms do 
not grow sufficiently rapidly in the jar more should be added 
from the culture flasks. We are more often troubled, however, 
towards the end of an experiment, by an excessive abundance 
of diatoms. In this case the jar may either be put ina 
darker place, or some of the water may be drawn off and 
replaced by a fresh supply of sterile sea-water. Care should 


1 The vessels we use are ordinary green-glass sweet-Jars, having a 
capacity of about 2000 ¢.c., which are kept covered with the glass stoppers 
provided with such jars, from which the cork band has been removed. 

2 In hot weather we often stand the jars in one of the tanks of circu- 
lating aquarium water, which maintains them at a very uniform tempe- 
rature. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 419 


be taken to have a sufficient supply of food at the beginning 
of the experiment, so that the larva may be able to feed as 
soon as they are ready for food. 

The method just described can be modified in various 
ways without detriment to the result. Sufficient sterilisation 
of the water may be effected by heating to 70° C. for 
fifteen minutes, affer which it should be aérated by violent 
shaking. ‘Outside water” may be used instead of ‘ tank- 
water,” and may be treated with Miquel’s solutions in the 
ordinary way, to ensure a satisfactory growth of the food- 
diatom. 

With regard to the food organisms, we have tried to obtain 
as large a variety of these in pure culture as possible, and 
then to make trial of a number of them with each batch of 
larve on which we have experimented. If no suitable pure 
cultures are available, success can sometimes be obtained by 
adding a few drops of tow-netting, collected with a fine- 
meshed net (180 meshes per inch), directly to the treated 
sterile water containing the larve. In this case one depends 
on the chance of a suitable food-organism growing in the 
vessel, unaccompanied by any destructive organism. On 
several occasions a satisfactory result has been reached by 
proceeding in this way, and the method is generally worth a 
trial, seeing that the number of larve obtainable from an 
ordinary fertilisation is very large and many different 
experiments are easily made with them. 

We will now give details of some of the results obtained by 
making use of the methods described, or of their modifications. 

Echinus acutus.—The first successful experiment was 
made with this species. Eggs fertilised on June 13th, 1905, 
produced healthy larve, fifty to seventy-five of which were 
placed, three days later, in a glass jar containing 2000 c.c. of 
ouside sea-water, filtered through animal charcoal, to which 
modified Miquel solutions were added. ‘hey were fed on a 
diatom culture, containing a small species of Chetoceras, 
which did not form chains, a small diatom probably belonging 
to the genus Melosira, a small naviculoid diatom, two 


420. Kk. J.. ALLEN AND KE. W. NELSON. 


minute flagellates, and a small green organism, probably one 
of the Pleurococcacew. ‘The vessel stood in a shallow 
tank, through which a stream of aquarium water was flowing 
and the temperature was fairly constant at 15° or 16° C., 
though there is one record of 19° C. at the end of July. The 
first two young: Echinus were seen on July 25th, forty-two 
days after fertilisation, and on August Ist twenty were 
counted. On August 5th (the fifty-third day) a. careful 
search through the jar gave twenty-one young Echinus of 
normal size attached to the glass, six minute but fully formed 
Hchinus, about twenty-three still in the Pluteus. stage, 
roughly half of which were well advanced. On August 16th 
some of the water, which had not been changed since the 
beginning. of the experiment, was replaced by “outside” 
water. On October 5th (sixteen weeks after fertilisation) 
twelve Echinus were still alive. Some pieces of red seaweed 
were placed in the jar, upon which the Echinus fixed them- 
selves and fed. Several of these specimens lived for over a 
year, but sufficient attention was not given to finding suitable 
food for them after the metamorphosis, so that they did not 
grow very large. 

Kchinus esculentus.—Three successful experiments 
have been made with EK. esculentus. In the first (eggs 
fertilised April 5th, 1907), “outside”? water treated with 
animal charcoal and filtered through filter-cloth, but not 
otherwise sterilised, was used. A number of jars of 2000 e.c. 
capacity containing Jarve were set up, and, to the most of 
these, various diatom cultures then in our possession were 
added, none of which, however, gave a satisfactory result. 
In two jars, on the other hand, to which no culture was 
added, there was considerable growth of diatoms and of a 
flagellate, upon which the Plutei fed. The first young 
Echinus were recorded in both jars on June 8th (sixty-four 
days),, but may have been present a few days earlier. 
Eventually from thirty to forty metamorphosed in one jar 


and about twelve in the other. The temperature varied from 
10°5° Cy. to: 12°5° C. 


ARTIFICIAL CULTURE OF MARINE PLANK'TON ORGANISMS. 421 


Inthe second experiment (eggs fertilised June 8th, 1908), 
made with similar water, the larve were feda on pure culture 
of Nitzschia closterium var., and six had completely 
metamorphosed on July 26th (forty-eight days after fertilisa- 
tion), two more subsequently coming through. ‘The tempera- 
ture was generally 15° to 16° or 17° C. 

In the third experiment (eggs fertilised March 29th, 1909) 
aquarium tank-water treated with animal charcoal and then 
filtered through a Berkefeld filter was used.. Plutei fed witha 
pure culture of a small flagellate (probably Chilomonas sp.) 
grew satisfactorily, and eight young Echinus were found on 
June 5th (sixty-eight days after fertilisation), which. had 
probably metamorphosed some days earlier. ‘I'wo other jars, 
in which Nitzschia closterium var. was used as food, 
were not successful, probably because the growth of diatoms 
became too thick towards the end:of the experiment. 

Echinus miliaris.—In the first experiment with this 
species animal-charcoal Berkefeld water was used, each jar 
containing, as usual, 2000 ¢c.c. In one jar the Plutei, from eggs 
fertilised on August 27th, 1907, were fed on a pure culture of 
Nitzschia closterium, var. On October 4th, 1. e. thirty- 
eight days after fertilisation, one Hchinus has just metamor- 
phosed. On October 29th about a dozen healthy-lookmg 
Echini were climbing about the jar, and many were still in 
a healthy condition on January 8th, 1908. ‘Temperatures : 
September, 15° to 19° C.; October, 16° dropping to 13° C. 
towards end; November, 12° to 11° C.; December, 15° 
to: 10°C. 

To another jar containing larve from the same batch a few 
drops of fresh Plankton were added as food. The Plutei in 
this case fed on flagellates and Nitzschia which grew in the 
jar, and several metamorphosed. 

In a second experiment with eggs fertilised on September 
13th, 1907, the larve were fed with Nitzschia closterium, 
but although there were a few well-advanced plutei still 
living on January 8th, 1908, none completed the meta- 
morphosis. 


422 E. J. ALLEN AND E. W. NELSON. 


Cucumaria saxicola.—A female Cucumaria, one of a 
number in a dish containing “ outside” water, laid eggs, 
which were fertilised, and sezmented on May 12th, 1906. A 
number of these were placed in a flask in 800 c.c. of “ outside”’ 
water, which had been sterilised by heating and then treated 
with animal charcoal and filtered. About 1 c.c. of fine 
plankton, containing diatoms, was added to the flask on May 
12th. On May 25th some of the water was poured off and a 
new supply added. As the amount of food seemed small, some 
culture of a green alga (Pleurococcus mucosus [Kutz.] 
Rabenh.) was added, and this continued to grow well in the 
flask. The larve continued healthy and formed young 
Cucumaria, of which many were still alive on July 25th, 
1907, i. e. fourteen months after fertilisation. Some of the 
water was changed in this flask on May 30th, 1906, June 
18th, 1906, and September 15th, 1906, and July 25th, 1907. 
Although many of these Cucumaria remained quite healthy 
they did not grow to any great size. Probably the food 
which was suitable to the larve and early stages, ought to 
have been changed as the animals grew older. 

Pomatoceras triqueter.—The larve of Pomatoceras 
are perhaps the easiest to rear, and give the most certain 
results of any with which we have experimented. They do 
well on the minute variety of Nitzschia closterium, but 
will feed upon almost any small diatom. Since the adults live 
in calcareous tubes attached to stones, and the tubes have to 
be broken open before the eggs can be obtained, it is not easy 
to get the latter free from infection of other organisms. If, 
therefore, the eggs are fertilised and placed in sterilised 
animal- charcoal water with only moderate precautions, 
sufficient growth of diatoms or other organisms will generally 
take place in the jar to feed the larve and bring them to the 
adult state. When once fixed to the glass the worms are very 
hardy and healthy, and a stream of ordinary aquarium water 
can be run through the jar. They then grow rapidly and 
attain a size equal to any found on the shore. The following 
experiment may be given in detail to illustrate the time 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 423 


occupied in development. On August 29th, 1907, eggs of 
Pomatoceros triqueter were fertilised in animal-charcoal 
Berkefeld water, and some pure culture of Nitzschia clos- 
terium var. added. The larve fed well, and on October Ist 
(i.e. thirty-three days after fertilisation), a great number 
had fixed on the sides of the jar and made quite normal tubes. 
A constant stream of the ordinary aquarium water was then 
allowed to run through the jar, and the worms continued to 
grow and flourish, reaching a large size, and are still alive and 
healthy (November, 1909). A similar result was obtained 
from the same batch of eggs by feeding on a pure culture of 
a flagellate infusorian. Temperatures during these two 
experiments were between 15° and 19° C. 

Chetopterus variopedatus.—Four experiments were 
made with this species. The food which gave most promise 
of success was the diatom Nitzschia closterium var. 
Larve from eggs fertilised on July 20th, 1908, fed on this 
material lived until October 30th, and reached an advanced 
stage. They did not, however, adopt the adult habit and 
form tubes. Two larve were also reared to an advanced 
stage by using flagellates, and, in later stages, the diatom 
Skeletonema costatum as food. 

Sabellaria alveolata.—One experiment only was made 
with this species, on eggs fertilised on July 19th, 1908. ‘The 
eggs were fertilised in “outside” water, and the larve 
subsequently transferred to jars containing animal-charcoal 
Berkefeld aquarium water. ‘They were fed ona pure culture 
of Nitzschia closterium var., and kept healthy and active, 
and developed well until nearly the end of October, when, 
simultaneously with a sudden drop in temperature from 15° 
and 16°C. to 12° and 9 C., they sank to the bottom of the 
vessel, and in about three days were all dead. ‘Temperatures: 
During July and August the temperature kept fairly constant 
at about 17° C., with a range from 15° to 19° C. During 
September it was generally about 15°C., and continued at 
about this level until the fall in the middle of October. 

Archidoris tubercnulata.—A good many trials have 


4.24, EK. J.. ALLEN AND E, W. NELSON. 


been made to rear the larve of nudibrunchiate molluscs, but 
up to the present not much success has been achieved. The 
best experiment was one made with larve of Archidoris 
tuberculata. A number of veligers of this species hatched 
out on May 8th, 1906, from some spawn which had just 
been collected from the shore. Some of these were put in a 
flask containing 1000 ¢.c. of sterilised animal-charcoal water, 
and about 1 c.c. of fine plankton was added. On May 14th 
a few veligers were transferred to another flask of sterilised 
animal-charcoal water and some pure culture of the green 
alga, Pleurococcus mucosus, was added. Whereas the 
larvee in the original flask did not live long, those provided 
with the green alga fed well and developed for some con- 
siderable time. A number of them were active and vigorous 
on July 4th,i. e. fifty-one days after. hatching, and several 
were still swimming at the end of July. On August 15th 
none could be seen moving, but two of those which lay on 
the bottom, when examined with the microscope, showed no 
sign of decomposition. The animal was retracted in the shell, 
but the tissue looked healthy, and the eye-spots and otoliths 
could be seen. ‘the growth in the flask seemed to be a 
quite pure culture of Pleurococcus. Larvee were examined 
again on September 14th, and appeared much as in August, 
the tissue still showing no sign of disintegration. ‘The flask 
was not again examined microscopically until July 25th of 
the following year (1907). No sign of the larvee could then 
be seen, but the culture of Pleurococcus remained pure 
and healthy. 

Subsequent experiments were made with spawn, which 
was deposited by the females in confinement. Although the 
spawn hatched and gave apparently healthy larve, these did 
not live for more than a few days. 

Calanus finmarchicus.—A single experiment is perhaps 
worth recording, as showing that it ought to be possible to 
rear this species without great difficulty. On August 8th, 
1905, to a flask containing 1000 ¢.c. of outside water 
(unsterilised) there was added } ¢.c. of Miquel’s solution B 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 425 


and $ c.c. of a 15 per cent. solution of anhydrous sodium 
carbonate. A few Calanus finmarchicus and some decapod 
Zoeeas were put in, together with a quantity of a culture 
containing mixed diatoms. On September 8th all the Zoeas 
were dead, but three Calanus were alive, and Nitzschia 
and a number of bottom diatoms were very plentiful. On 
September 17th the three large Calanus were alive and 
vigorous, and a considerable number of Nauplii were seen 
in the flask. By September 22nd two of the Nauplii had 
developed into young Calanus. ‘These, however, did not 
live for more than a week or ten days, and the adults also 
died. ‘The flask was abandoned on November 13th, the 
water in it not having been changed: since the commencement 
of the experiment. 

Hchinus hybrid.—A successful experiment on crossing 
H. esculentus and EH. acutus was carried out by Mr. 
W. De Morgan, who was working at the Plymouth Laboratory. 
We provided him with treated water and diatom cultures for 
food, and he followed our methods. We are indebted to him 
for allowing us to publish these results. Some eggs from a 
ripe EH. esculentus were fertilised by active sperm from an 
E. acutus, in sterilised water, on March 29th, 1909. Healthy 
larvee were obtained, and were transferred two days later to 
tank-water, which had been treated with animal charcoal and 
filtered through a Berkefeld filter. A culture of Nitzschia 
closterium var. was added as food, and the larve developed 
rapidly, feeding well. Several were completely metamor- 
phosed on May 7th, or thirty-nine days after fertilisation, 
In all thirty young hybrids were obtained, and:a number of 
these are still alive and feeding on red weeds. 

Sacculina carcini.—Mr.. Geoffrey Smith has recorded 
the fact (‘Quart.. Journ. Micr. Sci,,’ vol. 51, 1907,. p. 625) 
that he was able to rear the larve of Sacculina up to the 
Cypris stage, when they attached themselves to their host, 
Carcinus menas. These larve were kept in aquarium 
tank-water treated with animal charcoal and filtered through 
a Berkefeld filter. In this case the question of food did not 


426 E. J. ALLEN AND EK. W. NELSON. 


arise, as the larvae do not feed after hatching. It must be 
noted, however, that these larve had previously been reared 
by Miiller and by Delage. 

Summary of Method for Rearing Larve.—We have 
found that the best results in rearing marine larvee have been 
attained by taking the following precautions : 

(1) The eggs of the female selected must be really ripe, and 
the spermatozoa of the male active. 

(2) The smallest quantity of sperm necessary to fertilise 
the eggs should be used. 

(3) Sterile sea-water, treated in such a way that diatoms 
etc., will grow well in it, should be used. No frequent change 
of water is then necessary. 

(4) All dishes, jars, instruments, and pipettes, should be 
carefully sterilised before use. Every possible effort should 
be made to prevent the introduction into the rearing-jars of 
any organisms other than the larve to be reared, and 
organisms on which they feed. The jars should be covered 
with loosely fitting glass covers. 

(5) The eggs after fertilisation must be separated from all 
foreign matter, pieces of ovary, or testis, etc. As soon as 
the larvee swim up they should be pipetted off into fresh 
vessels of treated water, so as to leave behind any unseg- 
mented eggs, etc. 

(6) The food organisms should be small in size, so that the 
larvee can draw them into the mouth by ciliary currents. 
The food should distribute itself through the body of the 
liquid, and not settle too readily on the bottom of the vessel. 
(This is one of the great advantages of the diatom Nitzschia 
closterium, forma minutissima.) 

(7) The food should be abundant early, so that the larve 
may commence feeding as soon as they are able to do so. 
The food, however, must not be allowed to get excessively 
thick in the water. It can be kept down by diminishing the 
light, or by changing some of the water. 

(8) The temperature should be kept as constant as 
possible. Within limits the actual degree of temperature 


ARTIFICIAL CULTURE OF MARINE PLANK'TON ORGANISMS. 427 


is not so important as the avoidance of rapid changes of 


temperature. 


(9) A good north light, not exposed to direct sunlight, is 


most suitable for the rearing-jars. 


(10) In determining the amount of water to be used in any 


particular vessel, regard must be had to the amount of water 


surface exposed to the air, which should be large in propor- 
tion to the volume of the water. 


10. 


ii, 


(11) A change of food is generally required after the meta- 
morphosis of the larvee. 


BIBLIOGRAPHY. 


Cultures. 


. Beijerinck, M. W.—* Das Assimilationsprodukt der Kohlensaure in 


der Chromatophoren der Diatomeen,” ‘ Rec. Trav. Bot. Neerland,’ 
i, 1904, p. 28. 


. Benecke, W.—‘ Uber farblose Diatomeen der Kieler Féhrde,”’ 


‘Pringsh. Jahrb.,’ xxxv, 1900, p. 567. 


. Breazeale, J. F.—* Effect of Certain Solids upon the Growth of 


Seedlings in Water Cultures,” * Botanical Gazette,’ xli, 1906, p. 54. 


. Drew, G. H.—*The Reproduction and Early Development of 


Laminaria digitata and L. saccharina,” ‘Annals of Botany,’ 
acim L9LO, p. 177. 


. Gill, C. Houghton.—See Van Heurck, H., 21 and 57. 


. Karsten, G.—* Die Formanderungen von Skeletonema costatum 


Grun und ihre Abhangigheit von dusseren Faktoren,” ‘ Wiss. 
Meeresunters. Kiel.,’ N.F. iii, 1898, p. 5. 
“Ueber farblose Diatomeen,” ‘ Flora,’ Ixxxix, 1901, p. 404. 


. Lockwood, $8. M.—* Raising Diatoms in the Laboratory,” ‘ Journ. 


New York Micr. Soc.,’ 1886, p. 153. 

“ Aberrant Forms in Cultivated Diatoms,” ‘ Amer. Monthly 
Micr. Journ.,’ 1893, p. 259. 

——— ‘“ Formes anomales chez les Diatomées cultivées artificielle- 
ment,’ ‘Le Diatomiste,’ ii, 1893-96, p. 9; ‘ Ann. de Micrographie,’ 
x, 1898, p. 1. 

Miquel, P.—‘‘ De la Culture artificielle des Diatomées,” ‘Le Diato- 
miste,’ i, 1890-93, pp. 73, 93, 121, 149, 165; ‘Le Micro. Pré- 
parature,’ v, 1897, p. 69. 


428 ES J. ALLEN AND ZB. W. NELSON, 


12. Miquel, P.—‘ Recherches expérimentales sur la physiologie, la 
morphologie, et la pathologie des Diatomées,” ‘ Ann. de Micro- 
graphie,’ iv, 1891-2, pp. 273, 321, 408, 529; v, 1893, pp. 437, 521; 
x, 1898, pp. 49, 177, 182. ‘Le Micro. Préparature,’ xi, 1903, 
p. 174; xii, 1904, p. 32. 


13. “Du rétablissement de la taille et de la rectification de la 
forme chez les Diatomées,” ‘ Le Diatomiste,’ ii, 1893-96, pp. 61, 88. 

14. -~ “Des Spores des Diatomées,” ‘Le Diatomiste,’ ii, 1895-96, 
p. 26. 

15. “ Du noyau chez les Diatomées,” ‘Le Diatomiste,’ ii, 1893-96, 


p. 105; ‘Le Micro. Préparature,’ xii, 1904, p. 167; xiii, 1905, p. 
83. 

16. Richter, O.—“ Reinkulturen von Diatomeen,” * Ber. deut. bot. Gesell.,’ 
xxi, 1903, p. 493. 


a7, “Uber Reinkulturen von Diatomeen und die Notwendigkeit 
der Kieselsiiure fiir Nitzschia palea (Kitz) W. Sm.,” ‘ Verh. 
d. Gesell. deut. Naturf. u. Arzte, Breslau,’ ii, 1904, p. 249. 

18. “Zur physiologie der Diatomeen,” ‘S.B.K. Akad. Wiss. 
Wien.,’ exv, 1906, p. 935. 

19. “Ueber die Notwendigkeit des Natriums fiir eine farblose 


Meeresdiatomee,” ‘ Wiesner-Festschrift. Wien.,’ 1908, p. 167. 

20. Senft, E.—* Ueber die Agar-Agar Diatomeen,” * Zeit. d. Allgem. ost. 
Apotheker- Vereines,’ 1902, n. 9, figs. 1-9. 

21. Van Heurck, H.— Notice biographique sur C. Houghton Gill,” 
‘Le Diatomiste,’ ii, 1895-96, p. 125. 

“Culture des Diatomées,” ‘ Zeit. f. angew. Mikrosk.,’ iii, 

1897, pp. 195 and 225. 


22. 


Rearing. 


28. Agassiz, A.— Revision of the Echini,’ Cambridge, Mass., 1872-74. 

24. Cowles. R. P.—‘t Notes on the Rearing of the Larve of Poly- 
gordius,” ‘Johns Hopkins Univ. Circulars,’ xxii, No. 161, 1903. 

25. Doncaster, L.—‘* On Rearing the Later Stages of Echinoid Larve,” 
‘Cambridge Phil. Soe.,’ xii, 1903, p. 48. 

26. Grave, C.—* A Method of Rearing Marine Larve,” ‘Science,’ N.s. 
xv, 1902, p. 579. 

27. Lillie, R. S—* The Structure and Development of the Nephridia 


of Arenicola,” * Mittheil. Zool. Sta. Neapel.,” xvii, 1904-06, 
p. d41. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 429 


28. 


29. 


30. 


31. 


32. 


33. 


34. 


35. 


36. 


37. 


38. 


39. 


40. 


41. 


42. 


MacBride, E. W.—* The Rearing of Larve of Echinide,” ‘ Reports 
Brit. Assoc.,’ Dover, 1899, p. 458. 

“Notes on the Rearing of Echinoid Larve,” ‘Journ. Mar. 

Biol. Assoe.,’ N.S., vi, 1900-03, p. 94. 

“The Development of Echinus esculentus, together 
with some Points in the Development of E. miliaris and EK. 
acutus,” ‘ Phil. Trans. Roy. Soc.,’ B. exev, 1903, p. 285. 

Theél, H.—* On the Development of Echinocyamus pusillus,” 
‘Nova Acta R. Soe. Sci.,’ Upsala, 1892. 

Zeleny, C.—* The Rearing of Serpulid Larvee, with Notes on the 
Behaviour of the Young Animals,” ‘Biol. Bull. Woods Holl.,’ 
vill, 1905, p. 308. 


. Chemistry. 


Dittmar, W.—“On the Alkalinity of Ocean Water.” ‘Rep. 
Challenger Expdt.,’ 1873-76, Chem. 1, p. 124, London, 1884. 

Fox, C. J. J— On the Co-efficients of Absorption of the Atmos- 
pheric Gases in Distilled Water and Sea-Water.’ Part IT: 
“Carbonic acid.” ‘Publ. de Circonstance, Conseil Internat. 
pour l’expl. de la mer.,’ No. 44, 1909, Copenhagen. 

van ’t Hoff, J. H—‘ Zur Bildung der ozeanischen Salzablager- 
ungen, Braunschweig, 1905. 

Knudsen, M.—“ Hydrography,” ‘ Danish Ingolf-Expedtn., I, Part 
2, p. 21, 1899, Copenhagen. 

Krogh, A.—‘‘ On the Tension of Carbonic Acid in Natural Waters 
and especially in the Sea.” ‘Meddelelser om Grénland,’ xxvi, 


Raben, E.—* Uber quantitative Bestimmung von Stickstoffverbind- 
ungen in Meerwasser,” ‘Wissensch. Meeresunts.,’ Kiei, viii, 1905, 
pp. 81, 279. 

“Quantitative Bestimmung der im Meerwasser gelosten 
Kieselsiiure,” ‘ Wissensch. Meeresunts.,’ Kiel, viii, 1905, pp. 99. 
286. 

Ringer, W. E.—* Die Alkalinitiit des Meereswassers,” ‘ Verh. uit. 
Rijksinstituut v. h. onderzoek d. zee,’ iv, 1909, Helder. 

and Klingen, F. I. M. P—‘* Uber die Bestimmung von Stick- 

stoffverbindungen im Meereswasser,” ‘ Verh. uit. Rijkinstituut v. 
h, onderzoek d. zee,’ i, 1907-08, Helder. 
Salm, E.—‘‘ Studie tither Indikatoren,” ‘ Zeitsch. f. Phys. Chemie.,’ 
lvii, 1906, p. 471. 
VOL. 55, PART 2,—NEW SERIES. 28 


430 


48. 


44. 


45. 


46. 


47. 


48. 


49. 


50. 


51. 


52. 


53. 


54. 


55. 


56. 


57. 


E. J. ALLEN AND FE. W. NELSON. 


Tornée, H.—‘ On the Carbonic Acid in Sea-water,” ‘ Norwegian 
North-Atlantic Expdt.’ (1876-78), Chem. IT, p. 24, Christiana, 
1880. 


General. 


Baldwin, H. B.,and Whipple, G. C.—“ Observed Relations between 
Oxygen, Carbonic Acid, and Algw Growths in Weequahie Lake, 
Newark, New Jersey,” ‘Papers and Repts. Amer. Public Health 
Assoc.,’ xxxii, 1906. 

Brandt, K.—‘ Beitriige zur Kenntniss der chemischen Zusam- 
mensetzung des Planktons,” ‘ Wissensch. Meeresunters., Kiel, 
iii, 1898, p. 43. 

“Ueber den Stoffwechsel im Meere,” ‘ Wissensch. Meere- 

sunters.,’ Kiel, iv, 1899, p. 215; vi, 1902, p. 25. 

“Ueber die Bedeutung der Stickstoffverbindungen fiir die 

Produktion im Meere,” ‘ Beihefte z. Bot. Centralblatt.,’ xvi, 1904. 

On the Production and Conditions of Production in the 

Sea,” ‘ Rapports et Proces- Verbaux. Conseil internat. pour l’expl. 

de la mer.,’ iii, appdix. D, Copenhagen, 1905. 


Gran, H. H.—‘‘ Diatomeen,” ‘ Nordisches Plankton Kiel u. Leipzig, 
1908, Botanischer Teil, xix. 

Jérgensen, E.—‘* The Protist Plankton and Diatoms in Bottom 
Samples,” ‘Hydrographical and Biological Investigations,’ 
Bergen, 1906. 

Klebahn, H.—* Ein Uberlick iiber die neuere Diatomeenlitteratur,” 
‘Arch. fiir Protistenkunde,’ i, 1902, p. 421. 

Lemmermann, E.—* Flagellate, Chlorophycex, Cocecosphxrales und 
Silicoflagellatz,’’ ‘Nordisches Plankton, Kiel u. Leipzig, 1908; 
Botanischer Teil, xxi. 

Oltmanns, F.—* Morphologie und Biologie der Algen (with biblio- 
graphy), two vols., Jena, 1905. 

Ostenfeld, C. H.—* On the Immigration of Biddulphia sinensis 
Grev., and its Occurrence in the North Sea during 1903-1907, 
etc.,” ‘ Meddelelser fra Komm. for Havunderségelser.’- Plankton 
Series, i, No. 6, Copenhagen, 1908. 

Pfeffer, W.— The Physiology of Plants.’ English edition, three 
vols., Oxford, 1900. 

Smith, G. P. Darnell—‘ On the Oxidation of Ammonia in Sea- 
water,” ‘ Journ. Mar. Biol. Assoc.,’ N.S. iii, 1893-95, p. 304. 

Van Heurck, H.—‘ A Treatise on the Diatomacez’ (translated by 
W. E. Baxter), London, 1896. 


ARTIFICIAL CULTURE OF MARINE PLANKTON ORGANISMS. 491 


58. Vernon, H. M.—“‘The Relations between Marine Animal and 
Vegetable Life,” ‘Mittheil. Zool. Sta. Neapel,’ xiii, 1898-99, 
p. 341. 


59. Whipple, G. C.—‘‘ Some Experiments on the Growth of Diatoms,” 
‘Technology Quarterly,’ ix, 1896, p. 145, Boston. 


60. ‘The Microscopy of Drinking Water’ (and bibliography), 
New York, 1908. 

61. and Jackson, D. D.—‘Asterionella, its Biology, its 
Chemistry, and its Effect on Water-supplies,” ‘Journ. N.E. 
Waterworks Assn.,’ xiv, 1899, p. 1. 

62. and Parker, H. N.—*On the Amount of Oxygen and 


Carbonic Acid dissolved in Natural Waters, and the effect of 
these Gases upon the Occurrence of Microscopic Organisms,” 
‘Trans. Amer. Micr. Soc.,’ xiii, 1901-02, p. 103. 


ene 8 gy 7 = Ae : 
2 a > Z > 7 
. i, : ; ae 
vit! a he 
iw . @ 7 . _ 


Se eet Aen Arai ee ee 


NOTES ON THE FREB-LIVING NEMATODES. 433 


Notes on the Free-Living Nematodes. 


By 
F. A. Potts, M.A., 


Fellow of Trinity Hall, Cambridge, and Demonstrator of 
Comparative Anatomy in the University. 


With 11 Text-figures. 


I._The Hermaphrodite Species. 


Con'rENTS. 


PAGE 
Introduction . 433 
Summary of Sexual tehenomena in Bie Heivcapht baite 
Species . : , é , Pie 
Systematic Part : 437 
Biology of the Soil Nem: stodas in Beet in to the Methods 
of Experiment : : : . 443 
The Males of Hermaphrodite ereeres ; : . 446 
Structure and Organisation ; . 446 
(2) Proportions of Males in Hermaphrodite eres . 452 
(3 ) Sexual Instincts of the Males. : . 461 
The Hermaphrodites in Hermaphrodite Species — . 462 
(1) The Hermaphrodite Glands in Rhabditis andl 
Diplogaster ‘ : . 462 
(2) The Fertility of the Soil- Newmatodes : . 468 
(8) Partial Hermaphroditism ; . A473 
(4) The Nature of Hermaphroditism in the Wematoda . AUS 
(5) Self-fertilisation in Animals 479 
Summary of Results . ; : : . 483 


INTRODUCTION. 


Orley divided the Nematoda into three groups, roughly 
corresponding to differences of habitat found in the phylum. 
(1) Nematozoaembracing all parasitic forms, (2) Rhabditi- 
forme which live free in “ decomposing organic substances 

VOL. 55, PART 3.—NEW SERIES. 29 


45 ¥F. A, POTTS. 


or in earth saturated with such substances” ; and (3) Anguil- 
lulidee, the rest of the free-living nematodes, found in soil 
or water. Such a classification, grounded on cecology, pays 
no attention to the facts of morphology, and is naturally out 
of place in zoological arrangement, which aims at expressing 
the relationship of animals by descent. ‘The methods of life 
of an animal are, moreover, largely ruled by the mode of 
procuring nutriment which has been adopted. The first two 
groups of Orley are parasites and saprophytes respectively, 
but in the Anguillulide we have a heterogeneous collec- 
tion of forms varying greatly in their habits of life. Little 
is known of their sources of nourishment save in the case of 
a very definite division (e.g. Tylenchus, Dorylaimus), 
which live on the juices of plants, and for that end are 
provided with a small protrusible spear for piercing tissues 
and suctorial pharynx for absorbing sap thus set free. The 
vast majority of this family, however, possess an unarmed 
buccal cavity ; but in all the muscular pharynx is constantly 
at work, now dilated, now collapsed, constantly pumping 
fluid through the alimentary canal. ‘There is 10 morpho- 
logical distinction to be observed between such a free-living 
nematode as is found in the mud of a lake or amongst the 
alow of the marine littoralanda Rhabditisor Diplogaster 
of the soil. But the latter class can be kept in a culture 
fluid which swarms with bacteria, in which individuals of 
the former class would speedily succumb. ‘The tissues of a 
Rhabditis must be resistant to bacterial action and unharmed 
by the toxins which such organisms produce, and the worm 
is, In fact, capable of building up protoplasm from the 
bacteria themselves or from the products of their action. 
These are the most prominent physiological characteristics of 
the soil nematodes, Orley’s Rhabditiforme, and account 
for the peculiarities of their distribution, for they are 
apparently absent from dry soils and those with a small 
admixture of organic matter, and even in soils rich in humus 
are only detected in quantity by allowing some animal or 
vegetable substance to putrefy on the sample. Sufficient 


NOTES ON THE FREE-LIVING NEMATODES. 435 


attention has not been paid to the part which nematodes play 
in the economy of the soil,! but an investigation of this 
problem may well reveal results of as great interest as those 
which have been put on record by Maupas, working on the 
sexual organisation. In the present paper it is proposed to 
confine attention to the reproductive phenomena in certain 
hermaphrodite species, but it is hoped in a subsequent 
research to return to the nutrition and distribution of the 
class. 

Cultures of free-living nematodes in connection with this 
work were first started at the Stazione Zoologica, Naples, in 
1906, and continued at intervals in the next two years at the 
Zoological Laboratory, Cambridge, using for the most part 
Diplogaster linstowi. In 1909 I spent July to September 
at the Sutton Broad Laboratory, Norfolk, and procured from 
the neighbourhood the two forms, Rhabditis gurneyi and 
Diplogaster maupasi, the study of which enables me to 
amplify in one or two particulars Maupas’ account of the 
free-living hermaphrodite species of nematodes. I wish here 
to express my sense of the value of the opportunities for 
research afforded by the Sutton Broad Laboratory, and to 
thank Mr. Robert Gurney for his great kindness to me while 
working there. 


SUMMARY OF SexuaL PHENOMENA IN THE HERMAPHRODITE 
SPECIES. 

Guido Schneider, in his ‘ Monographie der Nematoden’ 
(1866), first discovered and put beyond doubt the existence of 
sell-fertilismgy hermaphrodite species of free-living uematodes. 

' The importance of the protozoan fauna of soil has but recently 
been realised. Like that of the nematodes their nutrition is composed 
of bacteria, and the place they take as a limiting factor in the increase 
of nitrifying forms has the closest possible bearing on the fertility of 
the soils they inhabit. It is, however, probable that these protozoa are 
more widely distributed in soil and so exercise a more important 
influence. (See E. J. Russell and H. B. Hutchinson, ‘Journ. Agric. 
Sci.,’ vol. iii, 1909, * The Effect of Partial Sterilisation of Soil in the 
Production of Plant Food,” especially p. 141.) 


436 F, ‘A. POTTS. 


In 1900 Maupas,! in a brilliant paper, drew attention to many 
striking features in the reproductive phenomena of such 
species. A full description of all prior work relating to 
hermaphroditism in the Nematoda is given by Maupas, and I 
shall here content myself with a short resumé of his own 
results, which later will be quoted more in extenso in 
connection with my own observations. 

The species of the free-living nematodes Rhabditis and 
Diplogaster fall into one or other of three categories : 

(1) Bisexual species, in which male and female individuals 
are produced in equal numbers. 

(2) Hermaphrodite species, in which, besides the self- 
fertilising protandrous hermaphrodites which form the great 
mass of the species, there are occasional male individuals, 
perfectly developed apparently, but taking no part in repro- 
duction. 

(3) Parthenogenetic species, in which males have not been 
found. 

It is reasonably supposed that each hermaphrodite species 
is derived from a bisexual form by the development of 
spermatozoa in the ovary of the female individuals, which 
thus become self-fertilising. The males are now useless, and 
have even to a large extent lost their sexual instinct. Their 
number dwindles in most cases to an almost imperceptible 
figure, but final disappearance does not appear to be reached 
in any species, and this persistence of apparently useless forms 
is one of the most curious facts recorded in biology. 

The hermaphrodite species appear even more numerous than 
the bisexual. There is, indeed, some evidence that the con- 
version of females to hermaphrodites in the bisexual species is 
a present-day process, furnished by the examples of partial 
hermaphroditism described by Maupas. An intermediate 
condition is shown in some hermaphrodite species by the 
occasional occurrence of pure females, or in the production of 


1 &. Maupas, “ Modes et Formes de Reproduction des Nématodes,” 
‘Arch. de Zool. Exp. et Gen., Sér. 3, it. 8, 1900, pp. 463-624, Pls. 


XVI-XXVI. 


NOTES ON THE FREE-LIVING NEMATODES. 437 


spermatozoa in one half of the genital gland only, the other 
producing eggs alone. Maupas emphasises the significant 
fact that these species with an incipient hermaphroditism 
yield the highest proportion of males he was able to chronicle, 
This conclusion that the more complete development of 
hermaphroditism and the suppression of the male sex neces- 
sarily proceed closely together is discussed further below. 

It is also highly characteristic of the hermaphrodite species 
in general that the sperm each individual produces only 
suffices for the fertilisation of a limited number of eggs, so 
that the period of fertility is followed by one even more 
prolonged, during which unfertilised eggs are laid, which do 
not develop. Such a phenomenon marks the hermaphroditism 
of the free-living nematodes as a character comparatively 
recently acquired and as yet not shaped by natural selection 
in anything like its final form. 

Finally, a most interesting result was obtained by experi- 
ments with hermaphrodites which had exhausted their stock 
of spermatozoa and supplemental males of the same species. 
In the rare occasions in which fecundation took place the egos 
which were afterwards laid produced males and females in 
equal numbers.! 


Systematic Parr. 


Diplogaster M. Schultze. 


This genus includes representatives both from soil and 
fresh water. But while the former possess a weakly developed 
bursa, which indicates the relationship of the genus to 
Rhabditis, the latter are without this character, and this 
fact, according to Bitschli, affords a natural distinction 
between the classes. 

* A preliminary note published in 1908 (‘Sexual Phenomena in the 
Free-living Nematodes,’ F, A. Potts, ‘Proc, Camb. Phil. Soc.,’ vol. xiv, 
Pt. IV, pp. 373-5) gave a general confirmation to Maupas’ results, 
founded on observations on Diplogaster linstowi which was kept in 
cultures for over a year and then died out. 


438 F. A. POTTS. 


The soil-nematodes belonging to this genus differ widely 
among themselves, particularly in respect of such definite 
characters as the number and arrangement of the papille on 
the tail of the male. The typical number is nine or ten pairs, 
but D. gracilis Biitschli and others have eight, and D. 
robustus Maupas, eleven. The arrangement of the papille 
is more variable than their number, but in a small group of 
species, with which | am more specially coucerned here, the 
relative positions are fairly constant and characteristic. 

The arrangement of the papilla follows the scheme given 
below. ‘The numbers correspond to those given in the various 
diagrams (see Text-fig. 4). 

(1) A pair of papille opposite the anterior end of the 
copulatory spicules. D. robustus Maupas possesses au 
extra pair, situated far in front of the spicules. In D. mau- 
pasi sp. n., as a frequent variation one of this pair may have 
been shifted forward to a markedly pre-spicular position. 

(2, 3) Two pairs of papilla opposite the posterior end of 
the copulatory spicules. 

In D. robustus Maupas shows three pairs in this position. 

(4) One pair slightly post-spicular. 

(5, 6) Two pairs, the anterior situated about half-way 
between the root of the tail and the anus, and the posterior at 
the root of the tail. 

(7-9) Three small pairs at the root of the tail, more ventral 
than the last-named. 

Since, then, there is so much similarity between the members 
of the group, the species are best distinguished by differences 
in size, proportions and biology, to which they are remark- 
ably constant. 

Common Characters of the Group.—Buccal cavity 
surrounded by lips with short sete. Within two! chitinous 
teeth. Vulva situated in middle of body.” 


‘Some species of Diplogaster, for instance D. fictor Bastian, 
possess only one buceal tooth. 

2D. gracilis Bitschli has a ‘“ monohysterous” 2 organ with the 
vulva a short distance anterior to the anus. 


NOTES ON THE FREE-LIVING NEMATODES. 439 


Male with bursa and nine (in one case eleven) pairs of 
papilla arranged in manner described above. Spicules 
-slender, with accessory piece. 


Synopsis of Group. 


(1) Bursa with nine pairs of papille: D. longicauda 
Claus. Bisexual species. Length of 2 1000-1500 nw; cesopha- 
gus fairly long (one sixth to one seventh of whole length) ; 
tail long (one third to one fourth of whole length). Germany. 

D. linstowi sp.n. Hermaphrodite species. Length of 
hermaphrodite 17604; cesophagus short (one ninth of whole 
length) and tail short (one-seventh). Oviparous at first, but 
soon became viviparous. Naples. 

D. maupasi sp.n. Hermaphrodite species. Length of 
hermaphrodite 1024-1252 4; cesophagus (one seventh to one 
eighth of whole length), tail short (one sixth to one seventh). 
Oviparous throughout life ; 150-3500 fertile eggs always laid at 
early stage of cleavage, and then about as many unfertilised 
egos. Norfolk Broads. . 

(2) Bursa with eleven pairs of papillae. D. robustus 
Maupas. Hermaphrodite species. Length of hermaphrodite 
2488 wu; cesophagus short (one ninth body length) ; tail very 
short (one ninth body length). First oviparous, then vivi- 
parous, after laying 150-230 fertile eggs. 

In addition to the summary diagnosis above the following 
characters are distinctive of the two new species. 

Diplogaster maupasi sp.n. (Text-figs. 1, 4, 5, 6, 8). 

Typical measurements of old ¢: 
Head to end of 


Total Head to ; SN Anus to Leneth 
; second bulb of as oe 
length. vulva. tail. of ege. 


cesophagus. 
Py C08 Gy, 12 ni). Wen @) F6u 


Buccal cavity small, with three indistinct lips, each with a 
slender seta, often distinguished with difficulty. Herma- 
phrodite at first lays eggs at long intervals, more frequently 
later. Males often fairly common. Spicules short, slender, 


44.0 PF, Ai POTTS, 


and almost colourless; accessory piece small, in lateral view 
generally a right-angled triangle, but frequent departures 
from this type by the rounding of the angles. Number and 
arrangement of the bursal papille strikingly variable. 


Abiup-caueayites, alp 


im ff 


7] 


I. dors. vent. 


It was at first thought that the shape of the buccal cavity 
was distinctive of species. The accompanying diagram of 
D.maupasi shows how greatly the state of contraction of 
the mouth affects the buceal cavity. 


TEXT-FIG. 2. 


D. linstowi sp.n. (Text-fig. 2). 


NOTES ON THE FREE-LIVING NEMATODES. 44,1 


Typical measurements of old ¢: 
Head to end of 


Total Head to 5) 5) ; Anus to 
tech alan second bulb of tail 
= * cesophagus. rs 
1760 pe 840 pe (4) 200 w (4) 240 we (+) 


Buceal cavity large, as broad as deep, with six papillar 
lips, each with a slight seta not easily seen. 

Males with long and slender copulatory spicules and stout 
accessory piece, elongated and pointed distally (contrast 
triangular piece of D. maupasi). 


Rhabditis Dujardin. 
(eh. gurneyi sp.n. (Text-figs. 9, 10). 
Measurements: 


Head to end ot Neat 
second bulb of ; 
tail. 
cesophagus. 


Length. Head Be 
5 vulva. 
Old herma- 


phrodite 14564 709 (3) 243 u (4) 149, (4-35) 


Diaguosis.—Hermaphrodite rather long and slender, 
tail short. Lips of buccal cavity indistinct, with very minute 
sete; buccal cavity narrow and deep. First division of 
cesophagus thick. Vulva median. Hermaphrodite gland with 
alternating production of spermatozoa and ova.  Sperma- 
tozoa of large size. Number of fertile eggs laid up to 800. 

Male unknown; probably never produced. 

Locality.—In peaty soil, Longmoor Point, Sutton Broad, 
Norfolk. 

(2) R. sechellensis, sp.n. (Text-fig. 3). 


Measurements . 
Leneth Head to me a Anus to 
ore vulva. g : tail. 
cesophagus. 
Old herma- 


phrodite 680u 3844u(4) 128u(2) 120p (4-4) 


A male measured 496 in total length. 
Diagnosis.—Small Rhabditis of pale, transparent 


442 F. A. POTTS. 


appearance. Lips of buccal cavity indistinct, surmounted by 
minute sete, only made out with greatestcare. Buccal cavity 
narrow and deep. ‘l'ail of moderate length. In herma- 
phrodite vulva median. Number of eggs produced small 


TEXT-FIG. 3. 


ace piece 


(150 or less), mother dying before exhaustion of sperma- 
tozoa. Males rare, inert. Copulatory spicules short and 
thick, accessory spicule small and inconspicuous. Bursa 
supported by nine rays, arranged as in Text-fig. 3. 
Locality.—Found in. moss from Seychelles; brought 
back by Professor J. Stanley Gardiner. 


NOTES ON THE FREE-LIVING NEMATODES. 44.3 


Briotocy in Repnarion to Merruops or EXPERIMENT. 


To obtain soil-nematodes in large quantities, it is only 
necessary to place scraps of flesh on samples of rich soil or 
mould kept moist and warm, and wait till decay has set in. 
Though the normal nutriment of these animals is presumably 
associated with the decay of vegetable products rather than 
decomposing animal matter, the latter prove exceptionally 
attractive. When once putridity commences, five or six days 
more suffice for the appearance of very large numbers of 
rhabdites or diplogasters, generally belonging to one or two 
species. Before, however, the last remains have vanished, 
it is probable that other species will have appeared and 
become dominant, entirely replacing the first kinds, so that 
an alternation is obtained somewhat similar to the succession 
of Protozoa in putrefying broth. It seems that the soil 
contains scattered throughout it numerous encysted larve, 
for, as Maupas has pointed out, when insufficient nutriment 
is supplied to soil-nematodes, the young larve envelop 
themselves in a thick cuticle, and become rigid and immobile. 
They are capable, however, of violent contortions, as if for 
the purpose of freeing themselves from the cyst, and by 
these movements migrate easily through the soil. The 
cuticular protection enables them to live uninjured in a dry 
environment, so that soil, etc., which has been subjected for 
long periods to fairly high temperatures, will yet yield large 
numbers of nematodes when treated in the way described 
above. The power of encystment, and consequently of 
resisting prolonged desiccation, is confined to the larve. 
Adult worms at once die when a liquid culture in which 
they are contained is allowed to dry up, and the eggs of 
these forms are provided only with a thin cuticular envelope, 
aud are incapable of resisting the vicissitudes to which the 
eges of parasitic forms like Ascaris are successfully exposed. 
When, then, animal-matter putrefies on a sample of soil, it is 
the encysted larvee which are attracted to its neighbourhood, 
where they emerge from their cysts and commence to feed 


4.4.4, EF. A. POTS. 


and grow rapidly. The rate of increase is very great: a 
single individual when once it has become mature will in 
five or six days give rise to one or two hundred, the eldest of 
which will be beginning to lay eggs. But a short interval 
then elapses between the migration of encysted larvee toward 
the putrid meat and the appearance of the swarms of young 
worms of the second generation. 

It is perfectly easy to keep free-living nematodes in drops 
of a nutrient fluid, and observe under the microscope every 
stage of their growth and reproduction. Each of these drop- 
cultures is contained in a solid watch-glass and secured against 
evaporation by a vaselined glass cover. Solutions of peptone 
were adopted as convenient culture media, and used almost 
exclusively in these experiments. The solutions were first 
allowed to putrefy till a cloudy growth of bacteria had 
developed throughout the liquid. So favourable an environ- 
ment for growth does a peptone solution in this condition 
constitute, that in four days the eggs laid by a mature 
hermaphrodite nematode have themselves produced mature 
individuals. It is only in the presence of great numbers of 
bacteria, or the substances formed by them, that the nema- 
todes thrive so well. In sterile solutions growth is suspended, 
and eggs are only laid at long intervals, for apparently 
nematodes find it difficult or impossible to assimilate peptones 
in an unaltered condition. It has not been discovered whether 
digestion takes place by the secretion of juices dissolving 
the protoplasm of the bacteria, or is merely confined to the 
absorption of soluble substances present in the culture fluid 
and prepared by the action of bacteria. If the second 
alternative be correct, then « parallel is established with the 
parasitic nematodes which nourish themselves on the dissolved 
and broken-down food of their host. An easily observable 
phenomenon of nematodes in culture is the rapid pumping 
action of the second cesophageal bulb and the rectum, and 
it may be argued from this that the nutriment obtained from 
the stream of fluid so constantly passing through the alimen- 
tary canal is in the form of easily abstracted soluble substances. 


NOTES ON THE FREE-LIVING NEMATODES. 445 


The insignificant development of glandular cells (which 
are found only in the cesophagus) may be cited against an 
intra-intestinal digestion of the bacteria, and whatever else 
its significance may be, the chitinous layer which lines the 
alimentary canal throughout must prevent an ingestion of 
bacteria by the endoderm cells themselves in such a way as 
Colpidium preys upon the bacteria of the soil. 

Besides peptone solutions other culture media have been 
used in the course of experiment. It was found possible to 
raise two or three successive generations in a saturated solu- 
tion of gelatin in water, and free-living nematodes matured 
from the egg in solutions of amides like tyrosin and leucin, 
but in these cases the growth was so much retarded and the 
production of fertile eggs so curtailed that oniy peptone 
solutions were used for extended experiments. 

The temperature at which the cultures were kept varied 
from about 18° C. in the summer to 12° C. in the winter, 
though at one period it fell within three or four degrees of 
zero. ‘lhe effect of a temperature approaching freezing-point 
was very marked, and showed itself in the almost entire 
suspension of growth. Sterility was not induced, but only a 
very few eges were laid every day. 

Experiments were also made to find the highest tempera- 
tures under which life and reproduction could continue. The 
cultures were placed in a water-bath which could be kept 
down to 25-30° C. Several individuals of the sixth genera- 
tion were isolated with the temperature of the bath at 26° C., 
going up to 28°C. One of these laid forty-three egos on 
September 8th. By September 11th these had developed 
into hermaphrodites of mature size, but although they lived 
for several days and were apparently in a quite healthy con- 
dition, they never produced mature eggs or spermatozoa. 
The ovary was distinctly seen with small nuclei, but there was 
no aggregation of yolk. Changes of this kind occurred in the 
other cultures. 

In addition individuals just ready to lay eggs were isolated 
from the cultures at the temperature of the room and placed 


4.4.6 F. A. POTTS. 


in a bath at 26-28° C. Under these conditions the ovary con- 
tinued to produce large-yolked eggs, and at first these were 
fertilised and laid, but after they had completed a few divi- 
sions they became disorganised. With eggs which later 
passed from the ovary into the uterus fertilisation did not 
apparently take effect. No egg-shell was formed, and the 
uterus became full of an amorphous, yolky mass. 

It seems, then, that the limits of reproduction lie in Diplo- 
gaster maupasi between 19° C. and 25° C., though life 
may be continued at slightly higher temperatures. It was 
found impossible, however, to keep cultures at a constant 
temperature of over 30° C. The individual worms became 
rigid and after a short exposure died. It is seen that the 
free-living nematodes are most sensitive to increased tempera- 
ture in the egg stage, when they can hardly endure high 
summer heat. The adult is also likely to succumb at 
temperatures which must be common in tropical countries 
at least. The encysted larve are probably the most resistant 
stage, and it must be supposed that these animals depend 
for their existence in periods of exceptional heat to their 
‘apacities for survival in this condition. 


‘’ae Mates oF HERMAPHRODITE SPECIES. 


(1) Structure and Organisation. 


The male sex in Rhabditis and Diplogaster, as in all 
nematodes, 1s sharply discriminated by the relation of the vas 
deferens to the alimentary canal, and by the well-defined 
secondary sexual characters, including a membranous bursa 
for adhesion to the female during copulation, and an arrange- 
ment of spicules for insertion into the vulva to facilitate the 
transference of the spermatozoa. 

The males of hermaphrodite species occurring in such 
small numbers, and apparently taking no part in reproduction, 
might naturally be expected to show some marked signs of 
degeneracy in organs other than the reproductive system. 


NOTES ON THE FREE-LIVING NEMATODES. 4.4.7 


In the Cirripedes we have another clear case of the successful 
establishment of hermaphroditism in a group in which the 
sexes were originally separate. Here, too, in hermaphrodite 
species there is a survival of the male sex, but the individuals 
which represent it are so degenerate in form and structure as 
to be described as little more than a bag of spermatozoa, and 
so reduced in size as to well merit the title of ‘ dwarf 
males.” 

It is, however, a surprising fact that in no particular of 
structural organisation do the males of hermaphrodite species 
appear to fall behind those of bisexual nematodes. ‘The 
conclusions which Maupas reached on this subjects are summed 
up in the following quotation : 

fees males . . . noffront rien de particulier et 
@anormal. On ne remarque rien dans leur structure et dans 
leur organisation générale qui puisse ies faire considérer 
comme des animaux mal venus ou mal constitués. Par leur 
taille, par les proportions de leur corps et par tous les details 
de leur organisation, ils répondent de tous points au type 
male ordinaire des Rhabditides dioique. Leur testicule 
luiméme est constitué dune fagon absolument normale et, 
ses produits, les spermatozoides, sont palreur forme, leur 
volume et leur structure absolument identiques a ceux que la 
glande génitale des femelles produit pendant sa periode 
d’activité protérandrique.” 

My own observations show that there is no imperfection of 
development in the residual wales of such species as I was 
able to study. The spermatozoa were always produced in 
vast quantities and exactly like those formed by the 
hermaphrodites. When liberated by pressure from the body 
of the male, they could be observed to put out amoeboid 
processes like those which Ziegler figures taking up their 
position in the uterus of Diplogaster lougicauda after 
fertilisation. This observation tends to show that the 
spermatozoa are physiologically active though the individual 
which carries them is prevented from playing its part in 
reproduction, possibly by a defect in nervous organisation, 


4.48 ¥F..A, POTTS. 


The experiments of Maupas with Rhabditis elegans 
showed that on the rare occasions when males do fertilise 
hermaphrodities, the spermatozoa are perfectly efficacious in 
the production of embryos. The curious change in the sex- 
proportions of the offspring of such unions may, however, be 


TEXxtT-FIG. 4. 


eventually traced back to some essential difference in the 
spermatozoa of males and hermaphrodites respectively, which 
might be revealed by a thorough examination of the spermato- 
genesis in the two cases. 
But though there is no manifest imperfection of organisa- 
tion in the males of hermaphrodite species, they appear to be 


NOTES ON THE FREE-LIVING NEMATODES. 449 


sometimes distinguished by extreme variability of the 
secondary sexual characters. In such specific characters as 
size and proportions of various parts the males are fairly 
constant, but the arrangement of the papillae supporting the 
copulatory bursa and the shape of the accessory piece ot 
the copulatory spicules show wide differences. When 
Diplogaster maupasi was first obtaimed from various 


TEXT-FIG. 4. 


samples of soil round Sutton Broad, the differences existing 
between the males found in separate cultures made me 
conclude that I was dealing with a number of nearly related 
species. It soon became clear that distinct types of male 
were not characteristic of each culture, but that even 
brothers from the same family often exhibited wide 
differences. 

The typical arrangement of the bursal papillz in Diplo- 

VOL. 5D, PART 5.—NEW SERIES. 30 


450 F. A. POTTS. 


gastermaupasi is shown in A, T'ext-figure 4, Departure from 
this type was found, however, in almost every other specimen 
examined. Below are given some of the clearest cases of 
variation observed in dealing with a comparatively small 
number (about forty) of males. 

(1) There should be normally a pair of papille situated 
exactly opposite the anterior end of the copulatory spicules. 
One of the most frequent and easily demonstrated variations 
occurs when one of the pair (or very rarely both) is shifted 
forward a smaller or greater distance. So marked a case as 
fig. c was observed two or three times. 

(2) A pair of papille (4-4’) occurs a short distance 
posterior to the anus. Only small variations in position were 
recorded here, but on one occasion a duplication of the 
papilla of one side was observed (fig. p). (The papilla of 
the other side was seen on altering the focus, so it was quite 
evident that the twin papillee belonged to the same side.) 

(3) In the position of papille 5 and 6 there is rather 
frequent variation; they are sometimes nearer together, 
sometimes further apart. Occasionally it may be seen (when 
the animal is lying on its back) that the papille of the two 
sides (5, 5’, and 6, 6’) have a tendency to alternate in position 
(fig. B shows this, but not very well). An example like fig. B 
was observed once, in which one of the papillz, either 5 or 6, 
was duplicated on both sides, and the twin papille then 
shifted apart. 

(4) he three small papille at the root of the tail (7-9) 
are rarely replaced by two. 

It is only occasionally on examining these animals that a 
frontal view is obtained, showing the rays of the bursa on 
both sides. In side views it is often difficult to correctly 
observe the position of the papillz. On this account only a 
few definite cases of variation are referred to above. They 
were observed in dealing with forty to fifty males. 

The accessory piece of the spicular apparatus varied in 
form in nearly every individual. Three types are figured. 
The first shows the most typical, in the shape of a right- 


NOTES ON THE FREE-LIVING NEMATODES. 451 


angled trangle, with an indentation at the anterior angle. In 
the other two the angles become more and more rounded. 

In Rhabditis sechellensis variations in the secondary 
sexual characters are occasionally found, but are much less 
numerous than in Diplogaster maupasi. Such variability 
as was observed was manifested in (1) inequality of the 
copulatory spicules, and (2) occasional asymmetrical dis- 
position of the rays of the bursa, 

The only reference to analogous phenomena which occurs 
in Maupas’ paper is found in his description of Rhabditis. 
guignardi (p. 525). He obtained only two males, but in 
one of these the copulatory bursa possessed on each. side nine 


TEXT-EIG. 9. 


A. B. C. 


supporting rays, in the other only seven. In the latter the 
remaining rays showed a disposition to fuse with each other, 
a phenomenon, it may be remarked in passing, which was 
responsible for the asymmetry of the bursal rays in R, 
sechellensis. The entire disappearance of two rays is a 
variation as great as any recorded above for Diplogaster 
maupasi, 

The position and number of bursal papillee or rays is looked 
upon as clearly diagnostic of species of Rhabditis or 
Diplogaster, and as faras I know no striking variation has 
ever been observed in the bisexual species. The connection 
of such a variability in the males with their disappearance 
from the economy of the species is no doubt significant, but 
it is impossible to offer any explanation of the facts, 


4.52 F. A. POYTS. 


(2) Proportions of Males in Hermaphrodite Species. 


Another remarkable feature of the males of hermaphrodite 
species studied by Maupas is their extreme rarity. In only 
one out of eleven species investigated was he unable to find a 
male; but in others males were only discovered by organising 
cultures of very considerable size, containing several thousand 
mature worms. So while in the majority of species the males 
were less than 0°1 per cent. of the whole number of adults, 
the proportion of 4 per cent. to which they rise in Rhabditis 
marionis affords quite a striking contrast. In Diplogaster 
maupasi, one of the species obtained from the Norfolk 
Broads, the ratio of male to female is very much more 
notable than anything which Maupas records, and does 
occasionally approach, though remotely, that equality of the 
sexes which is characteristic of the majority of animal forms. 
In one large culture the males reached 10 per cent. of the 
whole (377¢, 38¢ ¢), and in batches of eggs laid by the 
-same individual up to 30 per cent. (16 eggs, 11¢,53 9; 29 
eggs, 234,699). These instances are, of course, specially 
favourable, and picked from amongst scores of cultures which 
did not yield a single male. It is very unlikely that a species 
will be discovered uniformly consisting of equal numbers of 
males and hermaphrodites. Southern! supposed that in 
Rhabditis brassice he had discovered such a species, but 
in a culture with which he kindly supplied me I have been 
only able to find males and females, but no hermaphrodites. 

No illustrate the manner of oceurrence of the males, I give 
here an analysis of cultures of Diplogaster maupasi 
carried on over twenty-five generations, from August, 1909, 
to January, 1910. ‘lhe whole series of cultures commenced 
with a single individual. In every subsequent generation at 
least one hermaphrodite was isolated just before maturity to 
carry on the succession. When such an individual had com- 
menced to lay eggs it was removed every day to another 


1 Rowland Southern, ** On the Anatomy and Life-History of Rhab- 
ditis brassicx n.sp.,” ‘Journ. Econ. Biol.,’ vol iv, 1909, pp. 90-95. 


NOTES ON THE FREE-LIVING NEMATODES. 453 


watch-glass, so that the batch of eggs laid during the pre- 
ceding twenty-four hours was kept isolated. Hach batch 
was carefully counted to compare with the actual number of 
individuals attaining adolescence, and in this way records of 
cultures which gave the actual sex-proportions were distin- 
guished from others in which mortality before maturity 
obscured the true figures. In any drop culture which con- 
tained more than about thirty eggs the crowding which 
ensued was distinctly unfavourable to the chances of 
survival. 

Precautions were adopted in these experiments to prevent 
absolutely an association of mature males and hermaphrodites, 
and so remove any suspicion of cross-fertilisation in the line 
of descent here followed out. ‘To this end the individual 
destined to give rise to the next generation was separated 
before any male had become mature, or else the males them- 
selves were removed from the culture before the last moult, 
when they were perfectly recognisable as males, but had not 
yet assumed the spicular apparatus necessary for internal 
fertilisation. 

Both sexes become easily distinguishable a considerable 
time before maturity by the position of the developing gonad 
and its duct. In the majority of species of Rhabditis and 
Diplogaster, the vulva opens at the middle of the body of 
the female, and the gonad is paired, so that the immature 
hermaphrodite may be recognised by the symmetrical disposi- 
tion of the clear ovarian rudiments round the middle point of 
the body. In the male the rudiment of the testis is situated 
in the posterior half of the body, so that with a little experi- 
ence it is easy to distinguish a male, even among a 
ceaselessly twisting mass of other individuals, by the clear 
transparent testis running alongside the posterior part of the 
gut. Sperm-formation begins, it is true, before the last 
moult. But though the body of the male may contain mature 
spermatozoa, these can only be conveyed to the hermaphrodite 
individual by the co-operation of the copulatory spicules and 
bursa. A young male just before the last moult, at which 


454 FA, POTS, 


these latter are developed, is shown in Text-fig. 6. The 
proximal part of the vas deferens leading into the cloaca does 
not appear to be yet fully formed. The cloaca is spacious, 
and is produced on its dorsal surface into a pair of definite 
pouches in which the chitinous copulatory spicules are formed 
at the time of the last moult. 


TEXT-FIG. 6, 


™ intestine 


‘he history of the cultures may be divided into alternating 
periods, which are distinguished respectively by the frequent 
occurrence of males and their entire absence. During the 
first six generations, while these experiments were being 
prosecuted in Norfolk, the percentage of males was often 
quite high in batches of twenty or thirty eggs, and the off- 
spring of the majority of individuals contained at least one or 
two. In addition, the total number of eggs laid by each 
parent seldom exceeded 130 (150 in one case), and the 
spermatozoa were not exhausted before death. The seventh 
and eighth generations were reared away from a laboratory, 


NOTES ON THE FREE-LIVING NEMATODES. 455 


under conditions which made careful recording difficult. On 
removing the cultures to Cambridge a new kind of peptone! 
was used for the preparation of a culture-medium, and the 
behaviour of the nematodes altered considerably with this 
change. In five generations, from the ninth to the fourteenth, 
not a single male was produced. ‘The interval elapsing 
between the arrivals at maturity of successive generations 
decreased from seven days to four, and the number of fertile 
egos laid by each parent rose to between 150 and 300. In 
every case the life of the individual was prolonged under 
these more favourable (?) conditions, the period of fertile 
production being succeeded by another at least as long, 
during which sterile eggs were laid. 

Later, in the fifteenth generation, the peptone used in 
Norfolk was again tried, and at once males appeared 
sparingly in the cultures. Later the individuals raised from 
certain batches of eggs showed a fairly high ratio (e.g. in 
the nineteenth generation [25] 19 ¢ 4 ¢ @), but in general 
males were rarer than in the early cultures of August. After 
another removal at Christmas, 1909, the second period of 
male production was terminated like the first. It may well 
be supposed that the alteration of conditions, slight or other- 
wise, which ensues on changing the place of experiment was 
directly responsible for the disappearance of the males. 

It is not probable, however, that the proportions are 
controlled by nutrition, for though at first circumstances 
seemed to indicate that the use for a culture-medium of white 
peptone acted as a stimulus to male production, from the 
fifteenth generation onward four series of cultures were 
maintained, two in white peptone and two in brown (which is 
the more favourable medium for growth). As mentioned 
above, males first appeared in the former medium, but in the 
seventeenth generation they were also observed in brown 
peptone, and there was no sufficient difference in the 
figures to suggest which peptone was the better material for 
the production of males. 


' Tn dark brown crystals completely soluble in water. 


456 Hy tes OTS) 


In the second table a fuller analysis of the experiments 
lasting over the first six generations is given. An attempt 
was made to isolate strains, constantly producing high 
proportions of males, by breeding from a large number of 
individuals of the same generation. Thus in the third genera- 
tion a batch of 44 eggs produced 32 4 and 12 gg (about 
28 per cent.) did not, with one exception, maintain those 
high proportions. One, however, though giving at first 
hermaphrodites only, laid a batch of 16 eggs of which 11 
became ¢ and 5 ¢ ¢ (31 percent.). Nearly all these herma- 
phrodites were kept for an examination of their progeny, but 
five individuals, whose records were kept separate, furnished 
strikingly retrograde results, though males occurred in every 


case but one. ‘he male ratio was greater in a culture con- 


sisting of the offspring of three individuals, reaching 11 per 
cent. of the whole number. Further selection for the next 
generation proved equally indecisive. 

In the third generation a control series was also established 
by taking sister individuals from a culture in which only 
hermaphrodites were represented. ‘The total number of off- 
spring of the five parents selected was 319, of which 302 were 
$and 17 gg. This is exactly comparable to the total of 
262 ¢ and 15 ¢ ¢ produced by the five individuals from a 
culture with 28 per cent. of males. The individual details 
are closely similar in the two series. 

A brief inspection will serve to show how extraordinarily 
irregular is the distribution of males in the progeny of any 
single worm. ‘here is no rule that they should appear at 
stated intervals or restricting their production to a period or 
periods of maturity, but on the contrary the appearance of a 
few males from an early batch of eggs may be followed by 
a succession of hermaphrodites only and vice-versa; the 
last eggs may produce males when there have been only 
hemaphrodites hitherto, or, again, males may occur in several 
successive batches. 


TABLE 


457 


Nore. The figures enclosed in circles represent the 
number of eggs lad tn each batch: those to the 


right the ndariduals counted on arrwal amaturty 


or before. 


1$*Generation. Offspring of a single isolated hermaphrodite. 


no males observed. 


rd 
a . 26 


7” 


es 32 ¥ 
gh @ 269 


@ 


Total 7150 

6h 
in © 

Ve <2 

iq Total 127 
Re. 
Z eo 
SRK , @Z) 
gh >, 
GO: 
Ga) : 
. 
a Ct, Go : 
@ 
@ : 


Total 


ca 2 do". 


ows 


Bet 2 ih 


(aS 


These first. stc.generations 
were bred in the Sutton Broad 
Laboratory, Norfolk for the 
first generation anwtisiwon 
of Beef was used. Ahterwards 
two or three varveties of 
Peptone Ory, Albumen, Witte’) 
supplied by HarringtonBros, 
all of which had substantially 
the same value as atood stulf- 


The 10%-719%generations 
were bred irthe Loological 
Laboratory at Cambridge. 


WO: 593d 
Oe Ss 
ee 
MED ee 
— : 51 ¢208¢. 
P— : 26 ¢'. 
G0) : 31 ¥. 
48): 37 J. 
@ +72 ¢. 


TABLE 1! (Cont?) 


11*” Generation. 6): Sy Ge) +: 4a 

5 25 ¢ 

@ — @: «¢ 

OM: -—- OB: = 

182 : 60g WS: Re 

Ig (8) : 8 ¥ se 

ee @ wg 

@) : 3897 OD :uw¢ 

@ : 77 @ : 23 

@:— @: x7 

26 9g 198 : BY 

jae GO): we : ao 

@ nr¢(@: e¢ 

QB: 8¢7|\@Q: wg 

@): If |: we 

©: RY|O: B¢ 

250: Bg |@%: 4#¢ 

74 @ : 4 ¢ \257 : T59¢ 

@): 67 BO: we 

G2) : s777¢@) : Beg 

Qf @: we 

> — 39g 

@: 2¢ : 34 

TER? Ga) 

White Peptone. A I53 
BrownPeptone. @2) : 2° Brown Peptone. WhatePeptone. 
gen @) » 2¢ QD: OL OD: WY The one 
@ : 27 @: w¢ 8 ft 
2) u Oae @o) : a 0) : 50 ¥ entirely sterile. 

: 70 36 £187 @) 52¢ 

Gi) car et @ = 36 £ 2) : 30 

SO. Be eee 

285: Ble WE: WII : Woes 


TABLE 1} (Cont?) 
Brown Peptone. White Peptone. BrownPeptone. White Pepione . 


16*Generation G6): 249 GB: 359307@) : 249 @ : 6 


Vikan 


18h 


kag 2? 


6): 29 @:8f DW: #H9 &: —. 
@ : 54$ GA:19f 188 OD: — 
Pri ext Ore 
Qi wi @2¢ QO: @ 8 3 
249: 162¢ G6: Be 2e 1G: 
234:1309485 292 122 f 27 138 § 
@:i 2B — Be OQ: — 
@Qisr¢g @ Te we BD: —. 
@) eee GD :39 ¢ 


7 6) Bg es oe oe 4 G) cs ear 

WI 9 QO > See Gess® 7 
@Q : 239 IW5SIBIIMZ 4 F 
(254: Wag oat ae 


G9) :36 ¢§ GD: af @:4f @:age2se 
(‘Dr ws Cte ws GB) = 
@B:v¢g @:n¢g @):90 ¢ ‘ORs 

@ : 2g1ds@s): — GO:% 1odd2 23 G2SS 
@: ng @& :28~187@:2 ¢ 


150 M2 g71s3@) +12 ¢ W7:97FTIS 


126 : 96 G1 38 
@: 97 @O: we OM: Be Q:iwssse 
@2): 0 ¢¥ G): 15 ¢ @:e¢ 
@Q: 6 ¢ iv ee QA :a3gidd 
@: 39 G:9¢ G8 : 74g 
6): s1¢ @): 8 ¢ @) : 13 f73é 
136: Weg QD: 2 Ff 145 1378683 


199:169 § 


TABLE 2, 


460 


26 ¥ 2dd' BS f 2 bid 


ond ion 
2 ee ia Calture of reg! hetniap iron produced 
os bia 377 $38 i3 inthis generation. 
Bier Ee GO): 36 YY: 32 f12dd 
gth @9): 23 ¢68082):30924 ve 2343 29G3):3. gf 2¢¢@) WBS LB) 25 C6): 22g3IKED):22 f(D): 23 f 
" @3):__. Gd):30¢ 336 @9):14 61 SA @9) 134250 Ga)9 f2t8@3):34¢ (A): 74 
G3): 209202 Ga):54¢ @): __. @) 7 
AON g586 
a Se 
“haa RE B24 Ft (8:17 F700 A): 37 g Que “@: 19 ¢ 
(3 hermaphrodites gave G3): 32 g7 0 @3):37 f2¢¢@2): 41 ¢ @i):19 g7 dS 9): 27 G2 S¢ 
183 f 2bSS) / @):30¢ Gé:a6g G2): 37 ¢ G6): 26¢ GA): 37¢ 
Ga): 34% 6): 43 gage (—o): __ G5): 44 
@2):16 g1a8 ae 
ae J aba LEE! WBE :127$538132: 15 F lg Vy g/ FS130: 127 $2 IL 
” 4): 20 ¢ OO): 55 ¢38898): 12 G1 od : 
Qs): 24 gids 
GD: 56 g1de 
WE E237 


NOTES ON THE FREE-LIVING NEMATODES. 461 


Sexual Instincts of the Males. 


Maupas’ conclusion that the residual males could not take 
any part in the production of offspring is expressed in the 
following words: “ Mais si ces animaux examinés dans leur 
structure et leur morphologie, représentent des males vrais 
et complets, il n’en plus de méme lorsqu’on les étudie au point 
de vue de leurs facultés et de leur activité sexuelles . . . ces 
males ont a peu prés totalement perdu tout instinct et 
tout appétit sexuels. . . . Nous trouvousen présence d’une 
decadence psychique non concomitante avec une regres- 
sion morphologique.” 

This conclusion is supported by the inert behaviour of the 
males, the fact that they are never seen in copulation with 
hermaphrodites, but principally by the results of a fairly full 
series of experiments which Maupas made with males and 
hermaphrodites which had exhausted their own stock of 
spermatozoa. These conclusively showed that the males have 
almost, but not quite, lost their sexual instinct. One species 
alone stands as an exception. In Rhabditis marionis at 
various times cultures containing in the aggregate 28 herma- 
phrodites and 42 males were kept under observation. Since 
all the spermatozoa of the hermaphrodites were exhausted, 
any production of developing eggs was plainly due to the inter- 
vention of the male, and thus a measure of the activity of this 
sex was obtamed. Fertile eggs were laid by 15 individuals 
to the total number of 150-200, and all these produced 
hermaphrodites. This species is one of those for which 
Maupas described a partially developed hermaphroditism, 
and the author himself regarded it as specially significant 
that in such a form the male should be less degenerate. 

The most’complete series of experiments was made with 
Rhabditis elegans. Here, in twelve cultures, a total of 
159 hermaphrodites with their own sperm exhausted and 
males were associated. Only six of the hermaphrodites were 
actually fertilised, a proportion which illustrates exceedingly 


462 KF. A. POTTS. 


well the sexual inactivity of the males. ‘The chief point of 
interest lies in the constitution of the offspring of these six 
individuals. The young produced numbered 274, and of these 
147 were hermaphrodites and 127 males. So numerical 
equality of the sexes is secured in this species by cross- 
fertilisation, a result in striking contrast to that obtained when 
R. marionis was the subject of investigation. No permanent 
effect was produced on the heredity of sex, for when 38 of the 
hermaphrodites obtained by fertilisation by males were em- 
ployed as parents for the next generation, 2964 individuals 
were produced, of which only 7 were males, but the rest 
hermaphrodites. 

Further evidence of the psychical decadence of the males 
was secured in other species. Though nearly 100 males were 
employed belonging to five species only a single successful 
case of re-fecundation was observed, and in this (Rhabditis 
duthiersi) the fertilised eggs gave 70 hermaphrodites and 
1 male. 


‘HE HERMAPHRODITES IN HERMAPHRODITE SPECIES. 


(1) The Hermaphrodite Glands in Rhabditis and 
Diplogaster. 


In Rhabditis sechellensis the structure and develop- 
ment of the reproductive glands exactly correspond to the 
description which Maupas gives of R. elegans and R. 
dolichura. ‘hough no new details can be given, it will be 
convenient to summarise the changes which the hermaphrodite 
gland goes through before oviposition commences in any of 
the above three species. The three diagrams which illustrate 
the description are partly after my own drawings for R. 
sechellensis, but closely follow Maupas’ sketches of R. 
dolichura in Plate XX1I, figs. 7a, 7B, and c, 

The hermaphrodite organ is double, its two divisions being 
of equal development, and joining at the short and indefinite 
common vagina. Each division is U-shaped, and consists of 
a uterus, which extends from the vagina to within a short 


NOTES ON 'THE FREE-LIVING NEMATODES. 463 


distance of the bend of the tube, and an ovo-testis, occupy- 
ing the proximal part of the ventral limb and the whole ot 
the dorsal limb. In individuals examined some hours betore 
the first egg is laid the whole of the ovo-testis appears to 
consist of cellular elements of nearly equal size, which possess 
definite boundaries near the bend, but merge intoa syncytium 
distally. The anterior testicular region is indicated by the 
more regular polygonal form of a comparatively narrow belt 
of spermatocytes which succeed the uterus. ‘The young egg- 


= 


TEXT-FIG. 7. 


yy 
<O) 
a on 
loxce2 
(Ome) 
oOo 
OSV Receplaculam [cS 
OCS Seminis Se ans 
[o) (Spits) ect 
Ovary rx aon jets 
Siro 
mokome) 
(e) SLs 
aS exone) 
o°_ 3 (oye) 5 
ofa OOO; 
ola oo Ok 
of rerone) 
ose [feje) S| 
os oo 
one foxon®) 
SS9 Qs 
ese eos 
S ° 
A B. Cc 


cells which come next are all of small size, and can hardly be 
distinguished from the male cells. ‘Text-fig. 7, 4 represents a 
stage where the testis has begun to function, and several 
spermatozoa have been formed in the anterior part of the 
testis. 

In the second stage (B) sperm formation is in full activity, 
or may even be completed by the conversion of all the 
spermatocytes into spermatozoa. The female part of the 
gland now begins to show functional activity by the growth 
of the odcytes most anteriorly situated. The width of the 


4.64. F. A. POTTS. 


gonadial tube is so small in comparison with the size of the 
eve that the growing odcytes are arranged in a linear series. 
The odcyte nearest maturity is just posterior to the sperm- 
forming region, and behind it is a line of developing 
ege-cells showing the stages of growth from the scarcely 
differentiated odgonia. The spermatozoa as fully formed are 
small circular discs, capable of amoeboid movements when 
effecting fertilisation. They remain in the region of the 
gland where they were formed, so that what was testis in 
the first stage becomes receptaculum seminis in the 
second. In its formation, since the spermatozoa occupy @ 
much smaller bulk than the spermatocytes, the recepta- 
culum seminis shortens considerably ; its epithelium is of 
course the investing layer of the testis. The spermatozoa 
are now so disposed that the ripe ovum can pass out of the 
ovary and through the receptaculum seminis without its 
motion being impeded. During its passage a single sperma- 
tozoan fuses with the egg-cell and brings about fertilisation. 
The fertilised egg immediately becomes enveloped by a 
cuticular shell, and lies for some time in the uterus under- 
going segmentation before itis finally ejected to the exterior 
by the pressure of eggs from behind (Text-fig. 7, c.). The 
formation of ripe eggs after the first is perfectly regular, and 
fertilisation occurs in every case. Since, then, the whole 
quantity of spermatozoa is formed before the first egg is 
ready for fertilisation, it follows that a limit is set to the 
number of fertile eggs it is possible to produce, and as a 
matter of fact this limit is reached at a comparatively early 
point in maturity. When.the receptaculum seminis is 
completely emptied of its spermatozoa eggs still continue to 
be laid at a uniform rate, though they never develop to larve. 

In Diplogaster maupasi (Text-fig. 8) events follow a 
very similar course. here is, indeed, one difference in detail 
during the early periods of egg-laying which may be briefly 
mentioned. ‘The proximal limb of the gonad is shorter, the 
distal longer than usual. The former is entirely occupied by 
the uterus and testicular region, and the ovary is confined to 


NOTES ON THE. FREE-LIVING NEMATODES. 465 


the distal limb. Possibly in accordance with this shortening 
there is no linear succession of eggs increasing regularly in 
size in the anterior part of the gland, but each egg grows and 
reaches its full size before the one next in order begins to 
differentiate itself in size from the other odgoma. After an 
ego has passed out of the ovary and been fertilised, a period 
of some length elapses before the next finishes its growth in 
the ovary and travels through the receptaculum in its turn. 
It is only in the early stages, however, that oviposition is a 
slow process, for as the period of maturity advances, the 
zone of ege-maturation increases in length, and odgonia are 
able to start their growth long before the ovum in front is 


TEXT-FIG. 8. 


ready to be fertilised. The deliberate character of egg- 
production in D. maupasi is responsible for the fact that few 
individuals are seen with more than a single pair of eggs 
contained in their uteri. 

Rhabditis gurneyi.—When this species was first 
examined large numbers of adult individuals were obtained 
from cultures of decaying flesh. Amongst these a few were 
seen which, judging by their size, had only just attained 
maturity, but whose uteri and vagine were occupied by dis- 
organised eggs, as in hermaphrodites, which have exhausted 
their stock of spermatozoa. It was at first supposed that this 
was such a species as Rhabditis marionis (cf. Maupas, 
p- 512),in which a small number of females producing eggs 
only occur together with the hermaphrodites. When, how- 
ever, young immature worms were isolated, they were often 


466 F, A. POTTS. 


seen to a stage, sometimes extending over several days, 
during which eges passed into the uterus and degenerated. 
Later, however, the amorphous egg material was expelled 
and its place taken by fertile eggs which continued to be 
produced in large numbers. In this species, one could easily 
see, the hermaphroditism was not protandrous, but the 
formation of spermatozoa was sometimes delayed tilla number 
of eggs had ripened. In some cases, it is true, fertile eggs 
are produced from the first onset of maturity, and at first 
sight there is nothing to distinguish such forms from the 
typical protandrous hermaphrodite found in other species. 
But beside such an introductory period of infertility, there 
may be later interruptions of egg-production, which indicate 
a failure of the stock of spermatozoa. Frequently this is 
but temporary, and the worm begins again to lay fertile 
eggs. So short sometimes is the duration of sterility that 
it is indicated only by the ejection of one or two disorganised 
eggs, and very often only one gonad contains a supply of 
spermatozoa while they are lacking in the other. 

It is, then, saggested by the culture observations, and fully 
borne out by examination of the glands under high powers of 
the microscope, that eggs and spermatozoa come to maturity 
more or less alternately throughout the period of reproductive 
activity. 

Structure of the Gland.—In the general form of the 
reproductive glands of R. gurneyi there is no departure 
from that described above for other species of the genus. At 
various periods of development the arrangement of the histo- 
logical elements differs rather widely from the typical pro- 
tandrous gland. ‘'Text-fig. 9 shows part of the reproductive 
organ of a hermaphrodite which has just attained maturity. 
Tt will be seen that reproductive activity commenced with 
the formation of a very small number of spermatozoa (sp.?). 
And after the maturation of a single egg (ov.'!) a more 
numerous succession of spermatozoa (sp.” and sp.*°) Was pro- 
duced, only briefly interrupted by the appearance of another 
single egg (ov.*) which has not yet reached the limit of its 


NOTES ON THE FREE-LIVING NEMATODES. 467 


growth. After this, a prolonged period of egg-formation 
appears likely, for posterior to the spermatozoa there is a 
single row of developing egg-cells (ov.*) gradually diminishing 
in size and quantity of yolk, till in the middle of the limb the 
ovary becomes an undifferentiated syncytium. In this gonad 


TEXT-FIG. 9, Trxt-FIG. 10. 


Ov.2 


Sp. 3 sp.3 = 
OvV.3 
Spee 
Sp.l. 
ooeal| | 
° 
83 
oak 
a 
Xe) 
oo 
10. 


we have at one time the evidences of three alternations of 
male and female activity within a very limited period, 

In the second individual figured (Text-fig. 10) maturity is 
rather further advanced. The results of the early activity 
of the gonad are large numbers of spermatozoa and a few 
eggs. A series of developing ova now promise a long period 
of female productivity. There is an interesting departure 


4.68 Yr, <A, -FOUES. 


from the appearance of developing sperm-cells aud egg-cells 
in successive belts, tor here cells lying side by side may give 
rise respectively to spermatozoa and eggs. In one case the 
sperm-cells seem to have been actually formed at the expense 
of theovum. The early maturation of the spermatozoa will be 
noticed here, which terminates while young egg-cells forming 
from a mother-cell of the same age have only completed the 
first stages of their growth. 


(2) The Fertility of the Soil-nematodes. 

The hermaphrodite species of Rhabditis and Diplo- 
gaster are distinguished from the bi-sexual, as Maupas points 
out, by their lesser fertility, acharacter which indicates the in- 
completeness of the hermaphroditism. In eleven of the twelve 
species investigated by Maupas the number of fertile eggs 
laid by a single hermaphrodite individual varied between 200 
and 250, while in the twelfth (Rhabditis guignardi) the 
limit of production rose to 500 or 520. Maupas states that 
the female of a bi-sexual species is, on the other hand, capable 
of laying 700 to 800 fertile eggs. The low fertility of the 
hermaphrodites is due to the imsufficieucy of the supply of 
spermatozoa, for if to the number of fertilised eggs be added 
that of the unfertilised eggs laid when the male gametes are 
exhausted, it may be seen that a hermaphrodite produces as 
many eggs as the female in a bi-sexual species. Individuals 
producing 200-250 fertilised eggs will afterwards lay two or 
three times as many unfertilised,! so that the total equals the 
figure given for the bi-sexual species. 

Fertility, then, in these hermaphrodites is entirely controlled 
by sperm-production, and probably the actual number of 
spermatozoa formed in an individual is given or very closely 
indicated by counting the eggs laid which develop into larve. 
In these experiments the eggs laid by each parent were 
counted every twenty-four hours from the beginning of 
maturity onwards, and the mother then removed to a fresh 
drop of peptone. Usually after about six days of active ovi- 


! Maupas, loc. cit., p. 587. 


NOTES ON THE FREE-LIVING NEMATODES. 469 


position the spermatozoa become exhausted, but it is difficult 
to observe exactly when the limit has been reached, because 
the first laid untertilsed eges undergo a kind of incipient 
parthogenetic development. Such eggs possess a shell like 
fertilised eges and they complete a few divisions, but the 
blastomeres are more regular and equal than in normal seg- 
mentation ; the egg-substance appears greatly shrunk, so that 
a wide space oceurs between it and the egg-shell. 

An examination of the table of descent of Diplogaster 
maupasl will show how widely the fertility varies in a 
single species even under apparently uniform conditions. A 
few entries may be specially quoted here for comparison, each 
pair of individuals being taken from the same generation of 
nearly related strains and supplied with the same nourishment : 

(1) 12th generation October 20th—25th, 257 eggs. 

12th generation October 18th—22nd, 153 eggs. 
(2) 14th generation October 25th—31st, 143 eggs. 

15th generation November Lst—5th, 285 eggs. 

In this case a parent with low fertility gave in the 
next generation exceptionally prolific offspring. 

(3) 14th generation October 25th—-31st, 192 eges. 

15th generation November Ist—6th, 229 eggs. 

Other cases fall within the wide limits indicated above, so 
that it may be concluded that under favourable conditions a 
hermaphrodite individual of D. maupasi will lay 140-290 
egos. It is not pretended that such figures as these prove that 
it is impossible to select strains characterised by high and low 
fertility respectively, but as far as my observations go, there 
is a fluctuating variability, not governed by the laws of 
descent nor always directly traceable to minor changes in the 
environment. 

The influence of external conditions is, however, very great, 
and especially is this the case with nutrition. In peptone 
solutions of every kind, the number of eggs laid depends 
upon the development of bacteria in the culture-medium. 
When the peptone is fairly sterile the nematode only lays 
eggs at long intervals, and eventually dies when only a score 

VOL, 09, PART 3.—NEW SERIES. 31 


4.70 F. A. POTTS. 


or so of eges have been expelled from the uterus. In such a 
case of course the diminution in fertility is due to the small 
amount of nourishment supplied to the ovary, which is only 
enabled to produce a limited number of eggs. When a 
cloudy film of bacteria is seen at the bottom of the culture- 
drop the conditions are exceptionally favourable for the 
erowth of the nematodes, and fertile eggs are laid rapidly till 
the spermatozoa are exhausted. If, instead of peptone, a 
saturated solution of gelatin be used as a culture-medium, a 
very different effect is produced. For the first day or so after 
a worm is moved from a peptone solution into gelatin the 
‘ate of ego-production is fairly maintained, but afterwards it 
sinks very low indeed, though the life of the parent and the 
period of fertility is much jonger than that of individuals in 
peptone. Thus, for instance, for two hermaphrodites of the 
same generation bred in peptone but kept during maturity in 
peptone and gelatin respectively, the following figures were 
obtained : 


(1) Peptone. (2) Gelatin. 
Sept. 2nd—4th, 28 eggs. Sept. 2nd—4th, 19 eggs. 
»  Ath-d5th, 32 ,, » Ath-15th, 17 ,, 


5th—6th, 21 __,, 
6th—/th, 20 __,, 


3) 


3) 


Total for 5 days 101 eggs ‘Total for 15 days 36 eggs 

When a second generation of Diplogaster maupasi is 
raised in gelatin, when about twenty fertile eggs have been 
produced the uterus contains sterile disorganised ova. It 
appears from this that the effect of the substitution of gelatin 
as a foodstuff is not merely to curtail the formation of eggs 
in the ovary, but also to very considerably limit the number 
of spermatozoa produced. 

Though under favourable conditions the average fertility 
varies between two and three hundred in the majority of 
species now known, there are undoubtedly some which 
normally produce a very much smaller number of offspring. 
In the summer of 1907 I had under observation a species of 


NOTES ON THE FREE-LIVING NEMATODES. 471 


Rhabditis from the neighbourhood of Cambridge which I 
cannot adequately describe from the notes taken at the time 
It was remarkable for the very small proportion of fertilised 
egos laid by each individual. In one family six hermaphro- 
dites were selected before maturity, and their fertility com- 
pared. In each case the separate numbers represent the 
egos laid in a day, and those in brackets the total of fertile 
eggs: 

Pao, I v= (24) BO Oe ea On [22 

ieee 2S. 2—(39 | Hi9,.7,7,56,38=[31] IF 7, 6,1,1=[15) 

These cultures were carried on in July. Others, began 
later in August, gave rather higher numbers, e.e.: 

A 1,7, 13 (and 2 unfertilised eges), 12, 1=[34] 
B 14, 3, 8, 6, 5=[86] 

Oals, 17; 11=[43] 

Hes, LOeS, 5, 2= [38 

In A of this second series it will be noticed that the succes- 
sion of fertilised eggs was interrupted temporarily, but 
whether this was due to a retarded production of spermatozoa, 
asin Rhabditis gurneyi, or to some other cause, was not 
discovered. Itis much to be regretted that no trustworthy 
observations on the occurrence of males were made, for a 
species like this in which the hermaphroditism is of such an 
apparently recent and inefficient type, should, according to 
Maupas’ conclusions, possess a very large proportion of 
males, which was not, however, observed. It is hoped that 
the species may be rediscovered and this point investigated 
again. 

Rhabditis coronata Cobb, which was investigated by 
Maupas (pp. 037-541) and shown to be a protandrous herma- 
phrodite, is probably a similar form with very low fertility. 
No figures are given of the total of eggs laid, but it is 
mentioned that an isolated hermaphrodite only laid six eggs 
in twenty-four hours, and that in general eges were laid very 
slowly. An interesting feature shown in Maupas’ drawing 
of the species (Pl. X XI, fig. 8) is the small size of the ovarian 
part of the gland, which might well account for a restricted 


4.72 F. A. POTTS. 


ego-production. In the Cambridge species of Rhabditis, on 
the other hand, the early sterility was certainly due to the 
extremely small number of spermatozoa. ‘The length of the 
ovary was proportionately as great as in other species of the 
oenus. 

Rhabditis gurneyi, in contrast to the two species last 
discussed, is a free-living hermaphrodite nematode which has 
departed from the protandrous hermaphroditism, which we 
regard as the earliest development from the bisexual state. 
In consequence it far surpasses others of its kind in fertility. 
The spermatozoa are of unusual size, and possibly because of 
the difficulty of providing sufficient space to store a sufficient 
number at once, they are produced alternately with eggs 
throughout a great part of the period of reproductive activity. 
Asa result of this adaptation each individual is capable of 
laying as many eggs as a bisexual female, which frequently 
has its supply of spermatozoa replenished by copulation. 

It must be remembered that in many cases the hermaphro- 
dites of this species only produce unfertilised eggs in the 
initial period of oviposition which represent a total loss to 
the organism. When once this critical period has been 
passed, and a sufficient supply of spermatozoa established, 
fertile eggs are produced at the rate of 60-80 each day, or 
distinctly faster than in the case of Diplogaster maupasi 
and others. 

For figures to illustrate the fertility of Rhabditis gurneyi 
the following case is given. From the offspring of a single 
individual six immature hermaphrodites were selected. 
When maturity was reached the eggs laid every twenty-four 
hours were counted, and the parent removed to a fresh 
culture drop in the manner described above for Diplo- 
gaster maupasi. The dates in each case mark the period 
over which oviposition continued. 

(a) September 6th—17th, 525 fertile eggs. 
(B) . 7th—-17th, 686 e 

The figures here are not complete, for the culture dried up 

while the parent was still laying fertile eggs. When 343 had 


NOTES ON THE FREE-LIVING NEMATODES. 473 


been produced a prolonged failure of spermatozoa, lasting 
twenty hours, occurred in one of the glands, so that 16 
unfertilised eggs were laid with egg-shell, and the uterus 
beside blocked by disorganised egg material, while the other 
produced 40 fertilised eggs. After this interval developing 
egu's were counted to the number of 300. 

(c) September 7th—-12th, 168 fertile eggs. 

(D) 2 9th—20th, 730 . 

(x) 5 7th-17th, 362 S 

(F) 45 7th-10th, 81 : 

Out of the six individuals two laid about 700 eggs each, 
and though the figures obtained from the others show a high 
variability, this is partly to be explained by the very marked 
influence which even a slightly unfavourable change in the 
conditions can exert on sperm production. In cultures where 
several individuals are crowded together, it is noticeable that 
very few eggs are laid, and that the uterus of the worms 
speedily becomes crammed with disorganised eges, showing 
that the sterility is caused by the failure of the male, not the 
female gametes. 

In conclusion, it must be stated that the hermaphrodite 
species are apparently as successful as the bisexual species in 
the struggle for existence, for they are found in equal, or 
sometimes in greater abundance in nature. Evidently, though 
the means of dispersal of the species is limited by their 
generally low fertility, an advantage which more than counter- 
balances is secured by the self-fertilising capabilities of each 


individual. 


(3) Partial Hermaphroditism. 

It is here proposed to examine the description of certain 
species which are said to form a genuine link between the 
bisexual and hermaphrodite species. The species which 
Maupas deals with are Rhabditis marionis, R.duthiersi, 
and R. viguieri. 

(1) R. marionis.—A single hermaphrodite kept under 
observation was found to lay only 129 fertile eggs, while 


474, F. A. POTTS. 


other individuals of the same species produced about 250 
before their spermatozoa became exhausted. A closer exami- 
nation of a similar hermaphrodite led to the discovery that 
spermatozoa were only produced in one genital gland ; from 
the other only unfertilised eges were traced. In half its 
reproductive system the animal was hermaphrodite, in the 
other female. A few individuals were also noticed in which 
both genital glands apparently gave rise to eggs alone and 
never sperm. ‘I'he species is thus constituted of—(l) pure 
females (occurring very rarely) ; (2) individuals with one 
ovary and one ovo-testis ; and (3) full hermaphrodites forming 
the majority of the society. No mention is made of any 
variation in fertility among this latter class, but we are led 
to believe that all individuals fall into one or other of three 
sharply marked categories, according to the condition of their 
gonads. In the light of the results recorded above for other 
species this seems so remarkable that I think this case should 
if possible be re-examined. 

(2) Rhabditis duthiersi.—Three hermaphrodites were 
observed, each producing fertilised and sterile eggs simul- 
taneously, and it is suggested that these were possibly semi- 
hermaphrodites of the type described as occurring in R, 
marionis. It may, however, be pointed out that in 
R. gurneyi individuals are found with a similar appearance 
when the formation of spermatozoa is retarded and does not 
commence simultaneously on the two sides. 

(8) Rhabditis viguieri.—In this species the proportion 
of males was the largest met with by Maupas (though falling 
far short of some of the records for Diplogaster maupasi). 
Males formed 4 per cent. to 5 per cent. of the total in large 
cultures, and it is almost certain that the proportion would 
have been larger if single individuals had been selected for 
cultures. 

Of the other individuals some were females, which, when 
isolated, never produced offspring, but when united with 
males laid fertile eggs. The larve from such unions, it is to 
be regretted, were not kept. Hermaphrodite forms were in 


NOTES ON THE FREE-LIVING NEMATODES. A475 


a substantial majority, and it may be useful to quote Maupas’ 
words as to the relative frequency of the three kinds of forms : 
“ Les females non-hermaphrodites mais simplement unisexuées 
sont également trés fréquentes. I] ne suffisait, en effet, de 
placer sous le microscope une dizaine de femelles prises au 
hasard pour en recontrer une ou deux unisexuées. Les 
femelles simplement unisexuées y sont méme plus nombreuses 
que les males qui les fécondent sans difficulté. Hn résumé, 
chez cette espece les males encore relativement nombreux 
paraissent avoir conservé leur instinct sexual intact.” 

It is evident that this species, could it be re-discovered, 
would form a most interesting subject of study. A precise 
investigation of the comparative frequency of females and 
hermaphrodites, and in particular of the relative effects of 
self- and cross-fertilisation on the sexual constitution of the 
offspring, would prove of the utmost value.! 


(4) The Nature of Hermaphroditism in the 
Nematoda. 


The evidence that the hermaphrodites described by Maupas 
and myself represent the females of bisexual species, in which 
a part of the gonad has been given over to the formation of 
spermatozoa, is, indeed, overwhelmingly strong. Hermaphro- 
dites and females are identical in general anatomy, and the 
arrangement and form of their gonads are strikingly similar. 
Then, too, there exist a series of species showing the develop- 
ment of hermaphroditism from small beginnings in species 
where the ratio of fertilised eggs to unfertilised is very small, 
untilin Rhabditis gurneyi the number of spermatozoa is 
almost equal to that of the eggs they are required to fertilise. 
Lastly, there are, apparently, species like Rhabditis 
viguieri which have not yet decided between bisexuality 
and hermaphroditism, and present an assemblage of pure 


‘In Diplogaster maupasi, though careful watch was kept only 
one hermaphrodite was found which failed to develop spermatozoa (see 
Table I, fifteenth generation). 


4.76 ¥, A. POTTS. 


females, males and hermaphrodites, in cultures probably 
derived from nearly related individuals. 

Similarly, the various species may be arranged in gradation, 
to show the suppression of the male sex. In Diplogaster 
maupasi the males occur occasionally in such proportions as 


Trext-pre, 11. 


= 


Teceplaculum 


=) 


| | ‘J (Ga | } Seminis 
PAS) Ba 
Ot) 
ny | | 
00 Uf} 
) ) aly 
{) Be 
+ 
a 
, iS 
Rudiment of | 
altm. canal 423\5 
}) | ovary . 
Rudiment S| and @/ 
of Ovary Q ( Oviducl : 
PO 2) ! oo 
lterus ! (5 S| j o fa 9 
with fertilise 1) RAs [9 
eaqs IO 9fo/ 
99 ON oe) Ovary jo? © 6/ 
One! | and J | [alse 
or Oviduelh | |e} 
A S | Oe 
io some 4 
Nate \\ 4 
~~ ) 


= 


ABs 
Cyc 
1 


» & 


co 
5. ¢ 


to recall their original numerical equality with the female sex. 
But this species, in the majority of cultures and most others, 
at all times produces males in exceedingly small numbers. 
Finally, in Rhabditis gurneyi the male has possibly 
entirely disappeared, though of this it is difficult to adduce 


positive proof. 


NOTES ON THE FREE-LIVING NEMATODES. 477 


There are, however, some indications that it is not the 
female alone which is capable of developing a hermaphrodite 
gonad. In Rhabditis elegans Maupas records (pp. 491— 
492, Pl. XVII, fig. 2) the occurrence of large egg-like cells 
in the testis. A similar phenomenon has frequently been 
recorded as characteristic of the normal male gonad in 
Crustacea (Orchestia), and in other Crustacea the appear- 
ance of eges in the testis, without doubt to be attributed to 
the indirect action of parasites, is so definitely associated 
with the development of female secondary sexual characters 
as to indicate a change to hermaphroditism. In Rhabditis 
elegans the phenomenon is very slightly manifested, but 
there are indications that a very much more complete change 
is imposed on the male of Bradynema rigidum, a nema- 
tode parasitic in the body-cavity of the beetle Aphodius 
fimetarius.! This animal is so adapted for its parasitic life 
that mouth and anus have disappeared, and the alimentary 
canal, in the larva represented only by a single column of 
cells, has left not the slightest trace in the adult. In the 
autumn immense numbers of larve (up to *51 mm. in length) 
are found in company with one or two adults in each host. 
These larvee may be divided equally into females, whose 
genital glands, paired and situated in the middle of the body, 
have only attained to a rudimentary development, and males 
(Text-fig. 11, a), in which the testis, situated posteriorly in 
the body, often contains mature spermatozoa. When in this 
stage the larve bore through the walls of the alimentary canal 
and disappear. No intermediates are known between these 
forms and the adults 3-5} mm. in length, with a single excep- 
tion to be described later. In the adult condition there is only 
one class of individual with a long and vastly convoluted gonad 
opening to the exterior in the very posterior position which 
is occupied by the anus, in nematodes with a functional 
alimentary canal. It is this circumstance which led zur 
Strassen to derive the adults from the male larve, for if they 
are developed from female larve there must have occurred a 

1 Zur Strassen, ‘ Zeitschr. f. wiss. Zool.,’ t. 54, 1892, pp. 656-747. 


478 F. A. POTTS. 


shifting of the gonad during growth from a median to a 
posterior situation, and the conversion of a double rudiment 
into a single mature organ. In the most advanced male 
larvee the gonad is completely occupied by a brownish mass 
of spermatozoa save for an apical cluster of indifferent cells 
(inset to Text-fig. 11, a), and zur Strassen supposes that 
when the larvee begin to grow rapidly these cells proliferate 
and form an ovary. Ina single example of *75 mm. length 
(‘Text-fig. 11, B) the testis was represented by a receptaculum 
seminis full of spermatozoa, and this was succeeded by an 
ovary still slehtly developed and only posteriorly situated. 
In the adult (Text-fig. 11, c) the growth in size of the 
gonad has been so enormous that the whole of the body- 
cavity is occupied by it. The ovary and oviduct together 
form a narrow tube running twice the length of the body. 
Then succeed the receptaculum seminis, and lastly, the uterus, 
with a diameter nearly equal to that of the animal itself, runs 
from near the anterior end to the genital aperture. The 
great difference between this and the intermediate stage has 
been effected by the growth of the uterus with the fertilisa- 
tion of the eggs. 

Though in the absence of other intermediate forms it is 
impossible to produce clear proof that events take their course 
as indicated above, yet it is probable that the female sex, 
though represented by larvee, disappear without functioning, 
while in the males, after the spermatozoa have been formed, 
ova are produced in large quantities by the residual cells of 
the gonad. he evidence for the derivation of hermaphro- 
ditism in Rhabditis and Diplogaster from the female, and 
in Bradynema from the male, is in both cases of essentially 
the same nature, and depends on— 

(1) The recognition both in the original sex and the 
hermaphrodite derived from it, of a constant pattern of 
reproductive organ. 

(2) The discovery that the gonad of one sex is capable of 
developing the gametes of the other sex. 

If zur Strassen’s explanation is accepted, then in the limits 


NOTES ON THE FREE-LIVING NEMA'TODES. 479 


of the Nematoda it is found that now the female, now the 
inale, carries the characters of the other sex in a latent state, 
and when these are wakened to activity secondary herma- 
phroditism is developed. In Mendelian terminology either 
sex may be heterozygous. Moreover if the cytological phe- 
nomenon described by Maupas (p. 491) for Rhabditis 
elegans really shows tbat the male in that species 1s 
heterozygous, we are then forced to the hypothesis that both 
sexes are heterozygous in one and the same species, and at 
the same time. The phenomena of cytology and heredity as 
at present known in other groups, e.g. the Insecta, are 
capable of such diverse interpretations that it is impossible to 
say whether such a case as this suggested above is anomalous 
or no. 
(5) Self-fertilisation in Animals. 

Among hermaphrodite animals authentic cases of self- 
fertilisation are by no means common. In the Trematoda 
the rule of cross-fertilisation may occasionally be departed 
from, but only possibly in cases where the spermatozoa dis- 
charged into the body-cavity of the host find their way back 
into the female aperture of the same individual. Very little 
is known about the methods of fertilisation in the Cestoda. 
The evidence for self-fertilisation rests upon two observations, 
one by Leuckart of a penis inserted in the vagina of the same 
proglottis, and the other by Pagenstecher of similar relations 
between penis and vagina of adjacent proglottides.! 

In the Mollusca it is easier to prove by the isolation of 
individuals the possibility of reproduction without cross- 

1 In the Rhabdocel Turbellaria self-fertilisation is a very widely 
spread phenomenon and often the usual method of reproduction. Its 
existence has been put beyond doubt by the observations of individuals 
raised from the egg, but such experiments have not apparently been 
continued over several consecutive generations. In some forms the 
penis effects self-impregnation, in others there is no copulatory organ 
or female aperture and the spermatozoa migrate through the body tissue 
to the ovary (see Bresslau, * Verh. deutsch. zool. Gesell.,’ 1903, p. 126, 
and especially ‘Sekera Zool. Anz.,’ Bd. xxx, 1906, pp. 142-153). It must 
be noticed that in the three chief cases, the Turbellaria, the Nematoda, 


480 F, A’ POTTS. 


fertilisation. A. H. Cooke quotes two cases in the Cambridge 
Natural History, volume “ Mollusca.”’ In both Arion ater and 
Linnea auricularia, individuals isolated from birth pro- 
duced fertile spawn, although in somewhat limited quantities. 

In the Annelids a case has recently been described by 
Pierantoni! in Protodrilus. Ova are developed in the ante- 
rior segments, spermatozoa in the posterior, and a large pro- 
portion of the former are fertilised while still in the body- 
cavity. There is, however, a second method of reproduction, 
when by the rupture of the body-wall of the hermaphrodite 
the whole number of the eges is discharged into the sea. At 
the same time certain male individuals commonly occurring 
in the species emit their spermatozoa, which unite with such 
egos of the hermaphrodite as have escaped self-fertilisation. 

In the Crustacea hermaphroditism is largely developed in 
two groups, the Isopoda and the Cirripedia. In the former, 
the production of the spermatozoa in each individual precedes 
that of the ova, and the absorption of surplus spermatozoa 
ly phagocytes may preclude the possibility of self-fertilisation 
(e.g. Danalia?’). In the cirripedes adjacent individuals 
normally cross-fertilise ; a single case of self-fertilisation was 
recorded in Pollicipes (Gruvel). In the curious parasitic 
group, the Rhizocephala, both Sacculina and Peltogaster, 
invariably practise self-fertilisation.® 

Great interest attaches to the restriction of sperm-produc- 
tion accompanying the condition in this group. A small part 


and the Rhizocephala, the self-fertilisation which they practise is 
evidently a secondary and adaptive phenomenon. In the first two cases 
it has been developed as a means by which the actual existence of the 
race may be safeguarded, for both classes of creatures are liable to 
sudden extinction by the desiccation of the pool or moist soil, where 
they respectively live, and it is a manifest necessity that an isolated 
survivor should be capable of independent reproduction when conditions 
again become favourable. 

1* Fauna u. Flora Golfes von Neapel,’ t. 31, “ Protodrilus,” 1908, 
pp. 117-119. 

2G. W. Smith, ‘ Fauna u. Flora Golfes Neapel,’ Mon. 29, * Rhizo- 
cephala,’ 1906, p. 101. 

3G. W. Smith, loc. cit., pp. 21-24. 


NOTES ON THE FREE-LIVING NEMATODES. 48 L 


of the testis only is used for the formation of spermatozoa, 
and to prevent squandering of the slender stock the matura- 
tion of the spermatozoa is completed punctually just after 
a brood of eges enters the mantle-cavity. 

Both the Rhizocephala and the Nematoda, the two best cases 
of self-fertilisation, show one advantage obtained by the 
animal which adopts this method of reproduction, and that 1s 
the need for a reduced number of spermatozoa. In Saccu- 
lina the economy has been effected by a special change, to be 
looked upon in the hight of an adaptation, but in Rhabditis 
and Diplogaster, as we have seen, the small and markedly 
insufficient quantity of spermatozoa shows a tecent entrance 
into the hermaphrodite condition, and only because every 
spermatozoon fertilises an eye do these forms succeed in 
maintaining themselves. 

In the 'unicata, a group in which hermaphroditism has 
established itself completely, the ova ripen before the sperma- 
tozoa, and cross-fertilisation appears to be general. In 
Ciona ripe ova and spermatozoa are found in the ducts at 
the same time, and Castle! found that if the products from 
the same individual are mixed, as a rule fertilisation did not 
occur. ‘This result is so significant that it is not surprising 
that the experiment should have been repeated. Morgan? 
found some variation in the degree of self-sterility, but 
generally endorsed Castle’s results. In experiments which I 
carried out at Naples on the same tunicate in the early part 
of 1906 (and in which every care was taken to avoid contami- 
nation with foreign sperm), the eggs of an individual were 
found to be as fertile with their own spermatozoa as with 
those of other individuals, yielding in both cases nearly 100 
per cent. of embryos. ‘he pathological development which 
Castle found characteristic of self-fertilised embryos did not 
occur in my experiments. In conclusion, it seems possible 


1 Castle, W. E., “ The Early Embryology of Ciona intestinalis,” 
‘Bull. Mus. Comp. Zool., xxvii, 1896. 

2 Morgan, T. H., ‘ Journ. Exp. Zool.,’ i, 1904, p.137, ‘ Biol: Bull.’ vin; 
1905. 


482 F. A. POTTS. 


that the American form of Ciona intestinalis differs 
markedly, at least in its physiology, from the Mediterranean 
type species, and that, as is illustrated in plants, species which 
differ but little from each other in external appearance may be 
respectively easily capable of self-fertilisation and entirely 
restricted to cross-fertilisation. 

The free-living nematodes easily lend themselves to an 
investigation of the effects of continued self-fertilisation. 
Maupas organised cultures for this purpose, taking great care 
that the eight hermaphrodites chosen in each generation as 
the parents of the next should in no case have come into 
contact with mature males. With Rhabditis elegans, the 
period of experiment lasted from the beginning of December 
to the end of June, and in these seven months fifty-two con- 
secutive generations were reared. During the whole of this 
time no decline in vigour or productivity could be ascribed to 
the continuance of self-fertilisation. It is true that imme- 
diately afterwards the race became extinct owing to the 
onset of sterility, but the cause of this may well be traced to 
a sudden rise of temperature in the month of June (Maupas, 
p- 493). That this is the true explanation is indicated by 
the fact that Rhabditis duthiersi, another hermaphrodite 
species, which had only been isolated from the possibility of 
cross-fertilisation for a few weeks, became sterile at exactly 
the same time when its cultures were subjected to the same 
conditions. 

In my own researches Diplogaster maupasi has existed 
in cultures with no possibility of a cross through twenty-five 
generations, and that with not the slightest deterioration of 
the strain. It is hoped that under temperature conditions 
more equable than those of Maupas’ laboratory at Algiers it 
will be possible to prove that self-fertilisation may continue 
through a longer period and larger number of generations 
than was the case in R. elegans.! 


1 The cultures have now (June 21st, 1910) been carried over forty- 
six generations without cross-fertilisation with no observable diminution 
in fertility. 


NOTES ON THE FREE-LIVING NEMATODES. 483 


SuMMARY OF REsuL"s. 


In the preliminary summary on page 436, a short statement 
is given of Maupas’ results alone. In the present paper these 
are completely confirmed where the material allowed, and in 
some of the following details the study of hermaphroditism in 
Rhabditis and Diplogaster has been pursued further, 

(1) In one hermaphrodite species, Diplogaster maupasi, 
the residual males are much more numerous than in any other 
yet studied, and in small cultures may reach 50 per cent. of 
the whole number of individuals. 

(2) he male secondary sexual chararacters, i.e. bursal 
papille and accessory copulatory spicule, show great varin- 
bility. 

(3) The production of males is cyclical, periods (each lasting 
a few generations) when males are frequent alternating with 
others in which only hermaphrodites are produced. 

(4) Attempts to affect the sex-ratio artificially proved un- 
successful. It was also found impossible to increase the pro- 
portion of males by selection from favourable cultures. No 
rule could be discovered governing the constant fluctuations 
of production. 

(5) Even when males were most common there was no 
tendency to find female or partially hermaphrodite individuals, 
and the males were sexually inactive. ‘his contrasts with 
the conclusions reached by Maupas on Rhabditis. 

(6) The number of fertile eggs laid by D. maupasi is 
subject to wide variation. 

(7) In Rhabditis gurneyia far greater number of fertile 
egos may be produced by single individuals than in any other 
hermaphrodite species. he fertility is probably as great as 
the average bisexual species. 

(8) The formation of spermatozoa is not confined to the 
anterior end of the gonad as in other species, but may occur 
in any part and at any time throughout maturity. Frequently 
a number of sterile eggs were laid at the onset of maturity 
owing to the retarded production of the spermatozoa. 


484. Re AL LO MiSs: 


(9) No males have been observed in this species, so that 
they are either excessively rare or extinct. R. gurneyi, 
then, represents a much more complete and sufficient type 
of hermaphroditism than has litherto been recorded in the 
free-living nematodes. 

(10) Self-fertilisation has formed the exclusive means of 
propagation throughout twenty-five’ generations of Diplo- 
easter maupasi without any deterioration in the character 
of the stock. 


1 Now forty-six. (See note on preceding page. 
D 5 


OBSERVATIONS ON 'TRYPANOPLASMA CONGERI. AS5 


V 


Observations on Trypanoplasma congeri. 
Part I.—The Division of the Active Form. 


By 


Cc. H. Martin, B.A., 
Demonstrator of Zoology, University of Glasgow. 


With Plate 21, and 1 text-figure. 


ConrENTS. 

PAGE 
1. General Introduction . : : p . 485 
2. Methods : . : 487 
3. Morphology of the active Trypanoplasm. ? . 487 
4. Division : ‘ : ; : . 490 
5. Conclusions . : : : , . 492 
6. Summary of Results . : : . 494 
7. Literature. . : : ; . 494 
8. Explanation of Plate . ; ; : 495, 


GENERAL INTRODUCTION. 


In the ‘Zoologischer Anzeiger,’ Bd. xxxv, Nos. 14 and 
15, Mr. Elmhirst and I published a short note on a trypano- 
plasma parasitic in the stomach of the conger eel (Conger 
niger). Up to the present, as far as I am aware, no 
satisfactory account has been given of the division of any 
trypanoplasma, and the only point in connection with this 
process, on which previous workers have been unanimous, 
seems to be the extreme rarity of dividing forms. By what 
I must now regard as rather a fortunate accident, the second 
conger which I chanced to examine was so heavily infected 


VOL. 55, PART 3,—NEW SERIES. 32 


4.86 YH. MARTIN. 


that frequently two dividing forms have been found in the 
same field. As such infections, however, seem extremely 
rare, I have decided to publish my observations on the 
division of the active form of Trypanoplasma congeri 
at once, reserving the notes we have at present made on the 
changes into the resting form for a later paper. In the later 
paper we hope also to deal more fully with the general litera- 
ture of the group. I shall only mention in this paper the 
previous accounts of intestinal trypanoplasma, and, in « later 
section of the paper, I shall refer to the descriptions of 
division given for this genus by Keysselitz and Friedrich. 
The first description of an intestinal trypanoplasma was 
given by Léger in 1905 for a form, Trypanoplasma 
intestinalis, which he found in the stomach of Box boops. 

The second intestinal trypanoplasma, hitherto described 
Trypanoplasma ventriculi, was found by Keysselitz 
in the stomach of Cyclopterus lumpus, and is figured 
on p. 37 of his paper on Generations- und Wirtswechsel von 
T'rypanoplasma borreli. 

I do not propose to enter into any details as to the con- 
ditions under which Trypanoplasma congeri occurs in 
this paper, as these notes will be reserved for our later 
paper. The active form of the parasite is, however, always 
found in sections of the conger’s stomach in the mucus 
lining the surface of the wall, and it never seems to spread 
into the deep glandular pits. Up to the present no sign of 
the active trypanoplasma has been found in any part of the 
intestine or rectum, and, in fact, 1f active trypanoplasma are 
mounted in the intestinal juice they almost immediately 
become age 
have disappeared entirely at the end of a couple of hours. 
Up to the present forty-seven congers have been examined, 
and of these only ten have been found to be infected. ‘The 
parasite has been found in small numbers in some congers 
in which the stomach and intestine were full of food, but the 


only really heavy infections have been obtained from fasting 


lomerated by their posterior extremities, and 


congers. 


OBSERVATIONS ON TRYPANOPLASMA CONGERI. 487 


I should like to take this opportunity of thanking Mr. 
Elmhirst, the director of the Marine Station at Millport, for 
his assistance in getting material, and Miss Robertson for 
help in the drawing of the figures. 


Meruops. 


The stages figured in this paper were all obtained on wet 
smears from the stomach wall, fixed either in Flemming or cor- 
rosive acetic. Both of these methods gave excellent results. 
The films were stained in Giemsa, T'wort, iron-hematoxylin 
and eosin, and Mayer’s acid hemalum and eosin. All these 
stains gave satisfactory results, but the figures were all drawn 
from preparations made either with hemalum and eosin, or 
iron-hzematoxylin and eosin, 


MorruHonocy or tHE Active Form. 


As there seems to be a certain amount of discrepancy 
amonest different authors in regard to the nomenclature of 
the various structures in trypanosomes and trypanoplasmas, 
I have indicated in the following diagram the nomenclature 
I have decided to adopt. It is practically that used by 
Minchin in his paper on the structure of Trypanosoma 
lewisi in relation to microscopical technique (‘ Quart. Journ. 
Mieros.Sei.; vol..53, 1909, p. 799). 

The normal active Trypanoplasma congeri has rather 
an elongate body, measuring roughly 18 by 2°7 u. The two 
flagella arise apparently from a single basal granule near the 
anterior end of the kinetonucleus; the anterior flagellum 
passes up the mobile beak to end freely, while the posterior 
flagellum passes transversely across the body of the animal, 
and running down in connection with the narrow undulating 
membrane, projects freely for a distance of about 10 « beyond 
the animal’s posterior end. As regards the basal granule, 
most previous observers seem to have been of the opinion 
that each flagellum in T'rypanoplasma arises. from a 


488 Cc. H. MARTIN. 


separate basal granule, although it is evident that they do not 
regard the matter as absolutely certain, e.g. Minchin, in his 


TExtT-FIG. 1—Active form of Trypanoplasma congeri. 


7 


~ 


A. f. Anterior flagellum. Be. Beak. B.g. Basal granule. 
In. chr. Intra-nuclear chromatin. Ka. Karyosome. Ki. 
Kinetonucleus. P. 7. Posterior flagellum. Tr. Tropho- 
nucleus. 


pauper on the blood-parasites of fish, remarks as regards 
Trypanoplasma keysselitzi, p. 28, “In front of the 


OBSERVATIONS ON TRYPANOPLASMA CONGERI. 489 


kinetonucleus are situated the two minute blepharoplasts, 
from which the flagella arise. I believe them to be always 
two in number, but in iron-hematoxylin preparations they 
are so minute and often so close together that it is impossible 
to resolve them as two granules, and they may appear as a 
single dot.” 

In the active Trypanoplasma congeri the two flagella 
always appear to me to arise from a single basal granule, and 
from what I have seen of the dividing and resting forms I am 
certain that if the flagella do not arise from a single basal 
granule, the connection between the two granules must be so 
intimate that the flagella always behave as though they arose 
from a single point. Passing down the side of the animal 
under the membrane a row of very faintly staining rounded 
granules are frequently seen; these may correspond to the 
structures described in Trypanophis, or possibly to the far 
more strongly staining granules seen in some forms of 
Trichomonas. ‘The trophonucleus in the elongate form of 
Trypanoplasma congeri lies about one third of the 
animal’s length from the anterior end, and usually consists 
of a conspicuous membrane containing a darkly staiming 
elliptical karyosome, which is usually surmounted at its 
anterior end by a cap of chromatin granules. In some cases, 
however, the karyosome is central and the granules are 
arranged round it. ‘hese appearances recall Schaudinn’s 
figure of the chromosomes in the resting nucleus of Try pano- 
morpha and Leger’s description of the chromosomes of 
Trypanoplasma intestinalis. It will, however, I think, 
become abundantly clear from the behaviour of the dividing 
trophonucleus described below that it is impossible to regard 
the chromatin granules of Trypanoplasma congeri as 
chromosomes. | 

The kinetonucleus is usually a very darkly staining carrot- 
shaped structure lying laterally near the animal’s anterior 
end, the narrow posterior end of the kinetonucleus passing 
down the animal’s body to end in the region of the tropho- 
nucleus. In some cases the kinetonucleus presents an almost 


490 Cc. H. MARTIN. 


seomented appearance, and apparently this appearance has 
in many cases been taken as an early indication of division, 
though I believe this interpretation to be erroneous. 


Division. 


All of the preparations here figured are taken from films of 
the stomach of a fasting conger which had been kept in the 
tanks at Millport for four months, and was killed at 5.30 p.m., 
November 27th. In the early stage of division (PI. 21, fig. 2) 
the body of the animal becomes slightly shorter and thicker. 
The basal granule of the flagella divides, and this is followed 
by a splitting, first of the anterior flagellum along its whole 
length, and then of the posterior flagellum with its membrane. 
The trophonucleus and its contained karyosome become 
larger, and I believe that the intra-nuclear chromatin granules 
(? the “ chromosomes”’ of Schaudinn) at this stage become 
condensed on to the karyosome. ‘lhe kinetonucleus at this 
stage becomes slightly thicker, but shows no distinct indica- 
tion of division. In the next stage (PI. 21, fig. 5) the flagella 
have split along their whole length, aud it is important to note 
that, in marked distinction to the state of affairs found by 
Friedrich in T'rypanoplasma helicis, I have never been 
able to find the slightest evidence of the growth of uew 
flagella in any stage of division. ‘The trophonucleus now 
assumes a spindle shape, and the karyosome divides; the two 
halves, however, remaining connected by a rod, which persists 
until a very late stage of division. It might have been 
expected that some sign of the so-called chromosomes would 
be found at this stage lying around the dumb-bell-shaped 
karyosome in the spindle-shaped nucleus, but no trace of them 
has been detected. It is, of course, possible that this may be 
due to faulty technique, but so many of these dividing stages 
have been found lying near resting forms with nuclei clearly 
showing these granules that I believe this hypothesis is 
untenable. The relation of the axis of the trophonucleus 


OBSERVATIONS ON TRYPANOPLASMA CONGERI. 491 


spindle to the longitudinal axis of the animal’s body seems in 
these early stages to be rather variable, but in the later 
Stages the long axis of the spindle seems always to be 
arranged in direction transverse to the animal’s original 
longitudinal axis. The kinetonucleus now becomes very much 
enlarged, and gradually (P1.21, figs.4—7) pushes out a posterior 
hmb, which comes to lie at right angles across the dumb-bell- 
shaped trophonucleus. This relation seems very characteristic 
of this stage of division, which is a very common one on these 
films. It is rather interesting to note that the stages of 
division up to this point in the films from this particular 
conger are very common, the latter stages being comparatively 
rare. As these films were taken from various points all over 
the surface of the stomach, this would seem to point either to 
a cyclical epidemic of division in this parasite or (a view 
which seems to me rather more improbable) to an extremely 
short duration for the later as compared with the earlier 
stages of division. ‘lhe basal granules have now moved some 
distance apart, and as the animal shortens and thickens the 
membranes and flagella become shifted round till in the later 
stages they pass down the opposite sides of the body. The 
trophonucleus now is completely dumb-bell shaped, the 
handle of the dumb-bell being formed by the strand connect- 
ing the two karyosomes. In its early stages the dividing 
trophonucleus has presented a very superficial resemblance, in 
outline, at any rate, to the mitotic spindles found in the 
metazoan cell, but in the succeeding stages, in which the new 
trophonuclei have become definitely rounded, and their con- 
nection is limited to the bar joming the two karyosomes, this 
resemblance is completely lost. In Pl. 21, fig. 8,a late stage of 
division is figured in which the two products of division are 
still connected with each other by a narrowing band of preto- 
plasm, through which, even at this stage, the kinetonucleiand 
trophonuclei are still connected. In PI. 21, fig. 9, a form is 
shown which has evidently just divided. It is characterised 
firstly by its small size and rounded shape, secondly by the 
length of the kinetonucleus, and thirdly by the remains of 


492 Cc. H. MARTIN. 


the strand of the karyosome which had connected the two 
trophonuclei, and which has not yet been withdrawn. 

l'inally, the kinetonucleus becomes shortened and denser, 
the last remains of the karyosome strand are absorbed, and 
the animal elongates and regains its normal aspect. 


Conciusions. 

Lhave thought that it might be of some interest to compare 
shortly the above account of division of T'rypanoplasma 
congeri with that given by previous workers for other 
species of Trypanoplasma. As far as I am aware, thie 
only accounts of division in a trypanoplasma hitherto pub- 
lished are those by Keysselitz, in his paper, “ Generations- und 
Wirtswechsel im Try panoplasma borreli” (1906), and by 
Friedrich, in his paper, “Uber Bau und Naturgeschichte der 


Trypanoplasma helicis” 


(1909). Keysselitz gives on page 
28 of his paper five figures of dividing active forms from the 
blood of the fish, i.e. figs. 12, 14, 22, 23, 24. From these 
figures it would appear that the process of division in 
Trypanoplasma congeri shows some difference from 
that of Trypanoplasma borreli, though, as his series of 
division seems far from complete, it is quite possible that 
these differences may be more apparent than real. 

(1) As regards the behaviour of the flagella, Keysselitz 
seems inclined to believe that one of the products of the 
division keeps the old flagella, and that the other at a com- 
paratively late stage grows out new flagella. 

(2) In 'l’. borreli, according to Keysselitz, the tropho- 
nucleus divides, showing an internal division centre derived 
from the karyosome and eight chromosomes, at a stage at 
which there is no sign of division in the flagella, blepharo- 
plast, or cell body. 

(8) The kinetonucleus is said to divide transversely. 

The difficult feature in this account of division seems to 
me the extraordinary amount of variability in the time factor 
for all these processes ; in fact, Keysselitz himself states on 
page 31: “Den Verlauf der Teilung habe ich bisher in allen 


OBSERVATIONS ON TRYPANOPLASMA CONGERI. 493 


seinen einzelnen Phasen im Leben nicht verfolgen kénnen. 
Wie ich schon oben angegeben habe, trifft man relativ selten 
sich vermehrende Individuen an. Vorzugsweise sind es 
Tiere, bei denen die Teilung des chromatischen Apparates 
und des Plasmas, sowie die Bildung der lokomotorischen 
Organellen bereits beendet sind und die nur noch mit ihrer 
hinteren Enden zusammenhangen, eine Phase, die zeitlich 
laneste im Laufe der Teilung zu sein scheint.” It is parti- 
cularly over this last point, however, that a great deal of 
caution shonld be exercised. In well-infected smears it is 
an exceedingly common occurrence to find two trypanoplasma 
lying in a position which suggests division, but unless there 
is some absolutely distinctive feature, e.g.as regards the 
structure of the nuclei, which can be definitely connected 
with a corresponding structure in an undoubted dividing 
form, I feel that it is always most hazardous to interpret 
these appearances as division stages. On the other hand, 
the differences between the division of Trypanoplasma 
helicis, as described by Friedrich, and that of Try pano- 
plasma congeri, seem to be of an absolutely fundamental 
character. In the first place the karyosome, which is so 
characteristic a feature of the trophonucleus of most trypano- 
plasmas, is entirely absent in Trypanoplasma helicis,and 
in correlation with this fact the division of the trophonucleus 
appears to consist in a simple constriction of the large 
vacuolar trophonucleus with its scattered chromatin granules 
(p. 387). The division of the kinetonucleus is said to be 
longitudinal (p. 385), but the figures of this process seem 
hardly convincing. ‘The behaviour of the flagella, again, 
seems to be very complicated, since it is said on p. 390: 
‘““ Nachem die fiir die neue Zelle notwendigen ‘l'eile entwickelt 
sind oder der Anlage nach vorhanden sind, riicken die Kerne 
und Blepharoplasten auseimander.”’?  ‘ Dasselbe geschicht 
mit den Geisselursprungsstellen, die alsdann im die Nihe 
des Blepharoplasten verlagert werden. Dabei bildet sich die 
der alten undulierenden Membran zuniichst gelegene Geissel- 
anlage sur vorderen Geissel eines neues Tieres aus, wahrend 


4.94, C. H. MARTIN. 


die der urspriinglichen vorderen Geissel benachbarte zur 
undulierende Membran des neuen ieres wird.” 

It would be seen from the above that there is hardly a 
single point of agreement between the division of Try pano- 
plasma congeri and T'rypanoplasma helicis, and it 
would seem almost doubtful whether the two forms can be 
profitably united in the single genus. It would, I feel, be 
premature to enter here into a discussion on the comparative 
morphology of Trypanoplasma congeri and the trypano- 
somes proper until the rather complicated changes leading 
up to the resting-stage in the former have been more fully 
worked out. This I hope to do in a succeeding paper. 


REsuULTS. 


In the division of the active elongate Trypanoplasma 
congeri the following features are to be noted : 

(1) The basal granule divides. ‘his is followed imme- 
diately by a splitting of the anterior flagellum, and later, by 
the splitting of the posterior flagellum and membrane. 

(2) The trophonucleus in the first stage enlarges, the 
intra-nuclear chromatin condensing on the karyosome. The 
trophonucleus assumes first a spindle and later a dumb-bell 
shape, which persists to quite a late stage in division. The 
karyosome appears to act as an internal division centre, and 
no trace of individual chromosomes can be seen at any 
stage of division. 

(3) The kinetonucleus increases in size and divides by a 
simple transverse constriction. From its behaviour during 
division it is, I think, abundantly clear that, at any rate as 
far as Trypanoplasma congeriis concerned, the kineto- 
nucleus cannot be regarded as a centrosome. 


LITERATURE. 
Brumpt, E.—*‘ Trypanosomes et Trypanosomoses,” ‘ Rey. Scient.,’ vol. 
iv, 1908. : 


Dofiein, F —‘ Lehrbuch der Protozoenkunde, Jena, 1909. 


OBSERVATIONS ON TRYPANOPLASMA CONGERI. 495 


Elmhirst, R., and Martin, C. H.—‘*On a Trypanoplasma from the 
Stomach of the Conger Eel,” ‘ Zool. Anz.,’ Bd. xxxy, 1910. 
Friedrich, L.—*‘ Uber Bau und Naturgeschichte des Try panoplasma 
helicis,” ‘ Arch. f. Protistenk.,’ Bd. xiv, 1909. 

Hartmann, M., and Prowazek, S.—‘‘ Blepharoplast Karyosom und 
Centrosom,” ‘ Arch. f. Protistenk.,’ vol. x, 1907. 

Keysselitz, G.—‘ Uber Trypanophis Grobbeni,” ‘ Arch. f. Protistenk.,’ 
Bd. exi, 1904. 

* Generations und Wirtswechsel von Trypanoplasma Bor- 

reli,” ‘Arch. f. Protistenk.,’ Bd. vii, 1906. 

Laveran, A., and Mesnil, F.—‘ Trypanosomes et Trypanosomiasis,’ 
Paris, 1904. 

Léger, L.—‘‘Sur la presence dun Trypanoplasma intestinal chez les 
poissons,” ‘C.R. Soc. Biol.,’ vol. Ivii, 1905. 

Minchin, K. A.—* Investigations on the Development of Trypanosomes 
in the Tsetse flies and other Diptera,” ‘Quart. Journ. Mier. Sci.,’ 
vol. 52, 1908. 

“Observations on the Flagellates Parasitic in the Blood of 

Freshwater Fishes,” ‘ Proc. Zool. Soc.,’ 1909. 

Schaudinn, F.—** Generations und Wirtswechsel bei Trypanosoma und 
Spirochete,” ‘ Arb. a. d. Kais. Gesundheitsamte,’ Bd. xx, 1904. 

Woodeock, H. M.—“ The Hemoflagellates,” Lankester’s * Zoology,’ 
vol. i, fase. ‘1. 


EXPUANATION (OF  PLATHS21; 


I}iustrating Mr. C. H. Martin’s paper on “ Observations on 
Trypanoplasma congeri,” Part I. 


| All the figures were drawn with the camera lucida at table level under 
a Zeiss 15 mm. apochromat. and 18 compensating ocular. For the 
nomenclature of the structures compare text-figure. | 

Fig. 1.—Normal active Trypanoplasma congeri showing flagella, 
single basal granule, kinetonucleus and trophonucleus with its karyo- 
some and intra-nuclear chromatin granules. A row of faintly marked 
cytoplasmic granules may be seen passing under the membrane. 
Flemming, iron-hematoxylin, and eosin. 

Fig. 2.—Early stage of division. The whole body of the animal is 
shorter and stouter. The basal granule has divided, the anterior 
flagellum is split along about a quarter of its length, and the beginning 


4.96 CG. H. MARTIN. 


of the splitting of the posterior flagellum is shown. The kinetonucleus 
is slightly thicker and the trophonucleus is distinctly enlarged. The 
intra-nuclear chromatin granules have probably condensed upon the 
karyosome, which no longer presents the hard outline characteristic of 
the resting nucleus. 


Fig. 3.—The flagella have now split along their whole length. The 
karyosome has become drawn out into the characteristic dumb-bell- 
shape within the nuclear membrane. Corrosive acetic, iron-hzma- 
toxylin, and eosin. 


Fig. 4.—The body of the animal has become still shorter. The kineto- 
nucleus is becoming enlarged and losing its intense capatity for 
nuclear stain. The dividing trophonucleus is almost parallel to the 
longitudinal axis of the animal's body. Corrosive acetic, iron-hzema- 
toxylin, and eosin. 

Fig. 5.—The body of the animal has become still more deformed. The 
basal granules with their flagella have shifted apart. The kinetonucleus 
has become thickened and has now lost its intense capacity for nuclear 
stains, its lower border is crossed by the trophonuclear dumb-bell. 
Flemming, hemaluwn, and eosin. 

Fig. 6—The basal granules with their flagella now lie at opposite sides 
of the dividing animal. The lower limb of the enlarged kinetonucleus 
has adopted its characteristic position at right angles to the tropho- 
nuclear dumb-bell. Flemming, hemalum, and eosin. 

Fig. 7.—A slightly later stage than the previous figure, showing the 
characteristic relations of the enlarged kinetonucleus and the tropho- 
nuclear dumb-bell. Flemming, hemalum, and eosin. 

Fig. 8.—A late stage of division. The two products of division are 
still united by a broad band of cytoplasm, through which the kineto- 
nucleus and trophonuclei still retain their connection. Flemming, 
hemalum, and eosin. 

Fig. 9.—A recently divided form showing the characteristic rounded 
shape, the elongate kinetonucleus, and the unabsorbed strand which had 
connected the trophonuclei. The full length of the flagelia are not 
shown. Corrosive acetic, hemalum, and eosin. 


Luart Guirn. Mrcr Sct. Vol SE NSEPI I 


SS 


/ 
MARTIN, Fig 8 ee Gece 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 497 


The Development of Aplysia punctata. 


By 


A. M. Carr Saunders and Margaret Poole. 


With Plate 22 and 20 Text-figures. 


Turis work was begun by one of us in the spring of 1909, 
at the Zoological Station at Naples, wheu holding the Oxford 
biological scholarship. Owing to various reasons, the chief 
of which was ill-health, little more was done there than to 
collect material. It has been completed with assistance in 
reconstruction of sections and illustrations at Oxford in the 
department of Comparative Anatomy during the winter 1909— 
10. We must here express our gratitude to Prof. Bourne 
for the opportunity he has afforded us, and the encourage- 
ment he has given us to complete the work. 

The bionomics of Aplysia have been described with great 
eare by Carazzi and Mazzarell. The former deals at length 
with the deposition of the eggs and their early development, 
while the latter, in lis monograph on Aplysia (14), describes 
the general bionomics of the genus. The three common species 
found at Naples are punctata, limacina, and depilans. 
Carazzi (5), in his work on the cell-lineage of Aplysia, made 
observations on all three, and found little difference between 
them. Our results refer entirely to Aplysia punctata. 


' T wish to take this opportunity of expressing my thanks to the staff 
at Naples for their continual kindness during the time I was there, and 
especially to Professor Meyer and Professor Hisig for their valuable 
advice with regard to methods.—A.M.C.S. 


498 A. M. CARR SAUNDERS AND MARGARET POOLE. 


A number of stages of A. limacina were also examined, 
but the difference is insignificant. 

Carazzi states that Aplysia punctata disappears in May 
to reappear again in the winter. We were able, however, to 
obtain this species in large numbers until the middle of June. 
No difficulty was experienced in keeping Aplysia in the 
aquarium, and they laid eggs in great quantities. The eggs 
develop normally, and equally well if kept in jars or in the 
tanks with circulation, provided only that the water be 
changed every two days or so. Karly in the summer the eggs 
were at times attacked by bacteria, but if enough spawn was 
kept it was always possible to have some at the stage required 
in a healthy condition. Later in the year the eggs were 
attacked by alg, and the embryos destroyed long before the 
free-swimming stage was reached. ‘‘l'his was a more serious 
trouble than the bacteria, but the difficulty can be avoided by 
keeping the spawn in filtered water in the dark, where the 
alge do not develop. he rate of development varies with the 
temperature of the water. Thisis described by Carazzi for the 
different species. In April some fifteen days elapsed between 
the deposition of the eggs of A. punctata and the emergence 
of the free-swimming larve from the capsules. It is possible 
to keep the larvee in jars for some time, but even though they 
be kept in circulating water, they always die within a short 
time without exhibiting any change of structure. Mazzarelli 
states that he kept some larve of Bulla striata alive for 
twenty days, which is far longer than we ever succeeded in 
keeping Aplysia larvee, but even these showed no change during 
that period. No one has yet raised any Opisthobranch larve 
through the metamorphosis, aud there is therefore a large gap 
in our knowledge of the embryology of the group, for not only 
in the free-swimming larva are certain adult organs, such as 
the heart and pericardium and the gonads and genital ducts, 
entirely undeveloped, but the interpretation of some orgaus 
in the larva also must remain doubtful until the further 
development is known. Our failure to rear the larve of 
Aplysia beyond the free-swimming stage renders the present 


THE DEVELOPMENT OF APLYSIA PUNOTATA. 499 


work very incomplete, and it is therefore the intention of one 
of us to attempt to continue it and follow the metamorphosis. 
There would seem to be some hope of success if the methods 
of prepared sea-water and special feeding were used, such as 
have been employed so satisfactorily at Plymouth in rearing 
Hehinoderms. 

The living embryos are very opaque, and little can be seen 
of their organisation. As was the case with the work done 
previously on the cell-lineage, our observations were all made 
from preserved material. The eggs are enclosed in gelatinous 
capsules, and these are suspended in a lone thread of jelly. 
Carazzi made the following calculations :—there are on an 
average seven eggs in each capsule in A. punctata and 
fifty in A. limacina; the whole thread, or “nest,” as he 
ealls it, will therefore contain on an average 80,000 eges in 
the former species and 2,000,000 in the latter; this last 
number may at times be as high as 3,000,000. All the eggs 
develop with the exception of a few, which are not fertilised 
or are abnormal from some other cause. In the later stages, 
when movement is active and the muscles fully developed, the 
embryos will contract very considerably on the addition of the 
fixing agent, and this renders them difficult to interpret. To 
avoid this a 2 per cent. solution of cocaine in sea-water was 
used, which narcotises them in a few minutes and makes it 
easy to obtain preparations of fully-extended embryos. It is 
troublesome, and takes much time to extract the embryos alive 
from their capsules, and the great majority @et injured in the 
process. Most fixing agents do not harden the jelly, and it 
is therefore equally difficult to extract the embryos when 
fixed by most of the common means.  Formol, however, lias 
the effect of hardening the jelly, and it is on this account 
extremely useful. Alone it makes a good fixing agent, but 
subsequent staiming is rendered easier if it is used in combina- 
tion with some other fixative. At the suggestion of Prof. Meyer 
a solution of formol and picric was used, made up in the 
following way :—ten parts 40 per cent. formol, ten parts 1 per 
cent. picric, eighty parts sea-water. ‘This proved to be by 


500 A. M. CARR SAUNDERS AND MARGARET POOLE. 


far the best of all the fixing agents which were tried, though 
for special purposes others were used, as, for example, 
Hermann ’s fluid to show up the liver. 

When a thread of spawn was laid it was taken and sus- 
pended in the tank or jar by means of a string. When the 
embryos had reached a stage which it was desirable to 
preserve, a piece an inch or so in length was cut off the end 
of the thread, divided into a number of fragments a few 
inillimetres long, and put in the picric and formol solution for 
wbout half an hour. At the end of that time it was easy to 
break the capsules with a needle and extract the embryos, 
the greater number of them entirely uninjured. Various 
stains were used, but paracarmine gave the best results for 
whole preparations ; sections were stained on the slide with 
borax-carmine, followed by picro-indigo-carmine. 

The eggs of Aplysia are small, being less than 100, in 
diameter, and this makes orientation before section-cutting 
practically impossible. In the end, therefore, it was found 
more convenient to embed large numbers close together which 
could all be cut at the same time, for in this way one could 
be certain of getting a few embryos cut in the plane that was 
desired. In order to embed a large number of eggs in a 
small area of paraffin the following method was employed—a 
watchglass was filled with paraffin and allowed to cool; a 
small round hole, reaching at least half way through the 
paraffin, was then made; the embryos were trausferred into 
this by means of a fine pipette, and as much xylol as possible 
drawn off. ‘lhe watch glass was then placed for half an hour 
ona stand on the warm bath, for half an hour on the bath 
itself, and finally inside the bath until the paraffin melted 
completely, when it was cooled. 


More points remain undecided in the ontogeny of Molluscs 
than perhaps in any other group in the Animal Kingdom. 
The cell lineage has been worked out in numerous cases 
among the various groups, but our knowledge of those stages 
which follow upon the end of segmentation is very incomplete. 


THE DEVELOPMENT OF APLYSIA PUNCTATA. ° 501 


This work was undertaken to throw light, if possible, on the 
origin of kidney, heart and pericardium, about which the 
most diverse statements have been made. It will suffice here 
to point out that at least six different types of excretory 
organs have been described in Molluscs, and that the origin 
and homology of all of them is disputed; while as regards 
the ccelom, opinions differ equally widely. The bearing of 
our results on these questions will be discussed at the end of 
this paper. Owing’, however, to the fact already mentioned, 
namely, that it has not yet been found possible to rear the 
larve through the metamorphosis, they remain for the present 
inconclusive. 

For the purpose in view, Aplysia was chosen for two 
reasons, firstly because it 1s easy to obtain material at any 
period of the year, and secondly, because a very careful and 
complete account of the cell lineage has been given by 
Carazzi. It was hoped that by beginning at the point where 
Carazzi left off, it would be possible to follow the develop- 
ment of the organs, and definitely to ascertain from which 
cells they arose. 'I’o the excellent account of the cell lineage 
referred to we have nothing to add; every cell has been 
followed up in it to a time when there are more than one 
hundred, and the history of the endoderm and mesoderm has 
been traced further. His last description is of an embryo 
consisting of two hundred and fifty cells, with the velum 
already developed. 

‘he development as described by Carazzi may be sum- 
marised as follows: Segmentation is spiral, dexiotropic and 
unequal, the endomeres 4 and B being far larger than C and 
D. ‘he great size of these cells makes the cleavage look at 
first very irregular, but as a matter of fact their destinies 
show no exceptions to the scheme which has come to be 
recognised as normal in eges the segmentation of which is of 
the spiral type. ‘The first three quartettes give rise to all the 
ectoderm, 4d entirely to mesoderm, 'hereis no larval meso- 
derm arising from the ectoderm as has been described in 
some forms, The endoderm is derived from 34, 3B, 30 and 

VOL, 90, PART 3,—NEW SHRIES, 33 


502 A.'’M: CARR SAUNDERS AND MARGARET POOLE. 


TrExtT-FIG. 1, 


Surface view from the animal pole of an egg in the 12-cell stage. 


TEXxT-FIG. 2, 


Surface view from the animal pole of an egg in the 24-cell stage: 


TEXT-FIG. 3. 


Surface view from the animal pole of a later stage; the apical 
ectoderm is now formed from 8 cells of the first quartette 
of micromeres, 16 of the second, and 6 of the third; 2a?!, 
207, 2c?!, and 2d?! are the tip cells of the apical cross. 


TEXT-FIG. 4. 


Egg seen from the vegetative pole at a stage corresponding to 
Text-fig. 2. The mesoteloblast M is already formed by the 
division of D, 


TEXT-FIG. 95, 


Egg seen from the vegetative pole at a later stage. M has 
divided to give rise to the paired mesoteloblasts M and M’. 


TEXxT-FIG. 6. 


Optical section of an egg at a somewhat later stage at the level 
of the mesoblasts, seen from the vegetative pole. The two 
cells, M and M!‘, have each given rise to a small anterior 
mesoderm cell, m and m!. 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 505 


Trext-Fic. 7.) 


Diagrammatic optical sagittal section, seen from the vegetative 
pole of an egg in the 170-cell stage. The derivatives m, m', 2m, 
Yin!, of the mesoteloblasts M, M!', are seen spreading in an 
anterior direction from the region of the anal cells Ac. The 
macromeres A and B have diverged from one another to form 
a segmentation cavity. 


1 This and the preceding six text-figures have been modified from 
Carazzi’s drawings. The notation of the various blastomeres throughout 
the segmentation follows the system, now almost universally adopted 
for the description of cell-lineages, of Wilson (“ The Cell-lineage of 
Nereis,” ‘Journ. Morph.,’ vi), slightly modified by Conklin (* The 
Embryology of Crepidula,” ‘Journ. Morph.,’ xiii). A, B, C, D are the 
macromeres from which the successive quartettes of micromeres are 
divided off ; the quartettes being distinguished by the co-efficients 1, 2, 3. 
Thus the first quartette will consist of la—ld, and its derivatives 
will be la', 1la?, 1d!, 1d?; while the descendants of the latter generation 
will be la'!, la'?, la®', la?#-1d""!, 1d'?, 1d?!, 1d. By the division of 
D 4d is formed, which, since it contains the material for the production 
of the mesoderm, is designated by the letter M. This later divides to 
form the mesoteloblast Mand M', and from these, after the separation of 
a mesentoblast from each, small cells m and m!, etc., are budded off 


506 =A.. M. CARR SAUNDERS AND MARGARET POOLE. 


4D; gastrulation is of the epibolic type, and the blastopore 
is formed at the vegetative pole. It narrows to a slit-like 
opening, diminishing in size by the continual growth of the 
anterior and lateral parts of the ectodermal sheath, but does 
not close completely, but persists as the mouth. At the end 
of segmentation, owing to the large size of the endomeres 
A and B the embryo becomes somewhat heart-shaped. Between 
the large endomeres a small space appears, the segmentation 
cavity, which is more or less triangular in shape in optical 
section, the broadest end being towards the posterior end of 
the embryo. ‘'T'wo ectoderm cells, 2d”***! and 2d?***, increase 
greatly in size and come to project from the surface. These 
are known as the anal cells. During segmentation there 
is a shifting of the embryonic axis, and these cells come, 
in consequence, to mark the posterior end of the larve. 
At the opposite end the velum is formed as a simple ring in 
the region of the B quartette. By the time the cilia are deve- 
loped the embryos begin to rotate within their capsules. At 
this stage there are about two hundred and fifty cells. 

Text-fig. 7 represents an embryo with about one hundred 
and seventy cells, seen in optical section, from the vegetative 
pole. The blastopore is now reduced to a narrow slit, and 
posteriorly the anal cells project from the surface. Anteriorly 
the. polar bodies were present still adhering to the embryo, 
but are not represented. Internally the two large endomeres 
diverge from one another to leave the segmentation cavity 
between them, while the derivatives of the much smaller 
endomeres C and D are shown. The mesomeres, which are at 
this stage eight in number, stretch across from the anal cells 
towards the position occupied by the blastopore, which is not 
represented, 


anteriorly to give rise to the mesodermal bands. After the formation 
of the three quartettes of micromeres a fourth generation is produced 
by A, B, C; this consists of 44, 4B, 4C, which go towards the forma- 
tion of the endoderm. The above lineage is given in tabular form by 
Robert (26), to which the reader is referred for the detailed analysis of 
the later segmentation stages. 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 507 


We propose first to describe in some detail the earliest 
stage which we have investigated, and then to follow the 
development of the various organs separately up to the time 
when the larva becomes free swimming. 


STRUCTURE OF THE EBryo av THE END or SEGMENTATION. 

Pl. 22, fig. 1 is an external view of an embryo shortly before 
rotation begins, and fig. 2 in the same plate shows the cells 
which have sunk below the surface at the same stage. The 
total number of cells is more than 300. The ectoderm forms 
a thin and uniform layer covering the surface of the embryo. 
Round the anterior end there is a ring of somewhat larger 
cells derived from the B quartette, which bear long cilia and 
form the velum. At the opposite pole are the two anal cells 
2d and 2d?°, which are very prominent and project 
markedly from the surface, thus forming a convenient means 
of orientating the embryo with certainty. Near their bases 
are small nuclei which are sometimes difficult to see, and were 
not noticed by Blochmann (2). Their cytoplasm is very much 
vacuolated. They presumably function as temporary excre- 
tory organs. They are characteristic of Opisthobranch 
larvee, though in some cases, as in Fiona, described by Casteel 
(6), they are small and but little differentiated from the 
other ectoderm cells. It is well known that in certain other 
Gastropod larvee ectoderm cells of considerable size are found 
projecting from the surface ; Glaser (8) has described them 
iv Fasciolaria, where they occur singly or in groups of two 
or three together. ‘These would appear to be comparable 
to the anal cells of Aplysia, but in Fasciolaria their position 
is variable, being almost anywhere on the surface. In the 
region of the A quartette there is a slight projection, the 
cells being somewhat enlarged. This is the rudiment of the 
foot. Between the foot and velum is the blastopore on the 
ventral surface, and round it the ectoderm cells are beginning 
to sink in. Though the blastopore is at this time very 
small, we have always found it perfectly distinct, and in 
this we agree with Carazzi in contradiction to Mazzarelli and 


508 A.» M. CARR. SAUNDERS AND. MARGARET POOLE. 


Blochmann, who assert that it closes and then reopens to 
form. the mouth. ‘The stomodzum does not at this stage 
communicate with the space between the endomeres; it is a 
blind sac lined by ectodermal stomatoblasts and cesophago- 
blasts. ‘he former, according to Carazzi, are derived from 
3a! and 3b', and the latter from 3a* and 3b*?; together they 
number between twenty-five and thirty. On the dorsal 
surface of the embryo and posterior is the shell-gland. 
It consists of a deep and narrow invagination, formed by a 
large number of cells, which are slightly differentiated from 
the adjacent ectoderm by their more elongated shape and 
rounder nuclei. 

On each side of the embryo, just ventral to a line joining 
the anus and the mouth, a small ectodermal invagination, ot., is 
seen to bein process of formation. ‘These are the pair of oto- 
cysts. Fig. 3 shows the same structures at a slightly later 
stage. A little anterior to the anal cells on the right side are 
four large ectoderm cells identified by Carazzi as 3c!!!, 3c, 
8c!!! and 3c’, These cells are at this time clearly in the 
ectodermal layer, but they soon sink below the surface and 
give rise to the secondary kidney. ‘Their nuclei are of great 
size, and generally each contains one prominent darkly 
staining plasmosome. 

The greater part of the interior of the embryo is occupied 
by the two large endomeres A and B (not lettered in the 
plate). They diverge somewhat from one another, and thus 
enclose between them an irregularly triangular segmentation- 
cavity (marked st. in figs. 2 and 3). The broad end abuts upon 
the shell-gland posteriorly, while the narrow end reaches to the 
bottom of the still blindly-ending stomodeum. ‘The nuclei of 
the endomeres are large and oval in shape, lying to the inner 
side of the cells near the segmentation cavity. ‘The cytoplasm 
is heavily laden with large yolk-granules, and some of the 
yolk is often found in the nuclei also, causing the latter to 
stain very deeply with plasma stains. A large. vacuole is 
generally present in each of the endomeres, which is very con- 
spicuous in the living embryo, and persists for a long time. 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 509 


At this period there are about twelve other endoderm 
cells, the derivatives of 4a lying close against the stomo- 
dzeum in the anterior end of the segmentation cavity, those 
of 4b at the opposite end rather dorsal to the shell-gland, 
and C and D, with their descendants, also lying at the posterior 
end against the wall of the shell-gland. The latter are 
already beginning to form a fairly definite row, which will 
become the posterior well of the stomach. At this stage the 
greater part of the cavity, which will be the stomach, is 
bounded only by the endomeres A and B, but this soon ceases 
to be the case. A and B gradually take less and less part in 
the formation of the wall and give rise to the left liver. 

There are between fifteen and twenty mesoderm cells.in the 
embryo at this stage. They form an irregular band, which 
arises at the posterior end near the anal cells, and stretches 
forwards to the blastopore. The band lies chiefly on the 
right side, but certain cells are already beginning to pass 
dorsally and ventrally into the foot. 

We here see that it is possible to speak of a mesoderm 
band in Aplysia, though it is never clearly defined and soon 
breaks up. ‘The conditions are very much like those described 
in Fiona and Umbrella, though the great size of the 
endomeres in Aplysia has forced the mesoderm chiefly on to 
the right side. 

There is no secondary mesoderm either in this or later 
stages. It is present, however, in Fiona. Unfortunately, 
Heymons worked on Umbrella at a time when the existence 
of secondary, or ecto-mesoderm, was not recognised, so that 
its presence or absence in that form is unknown. 


DrVELOPMENYT OF THE ORGANS. 


The embryo rapidly assumes the appearance of the free- 
swimming larva, and from the beginning of rotation onwards 
there is but slight alteration in shape and very little increase 
in size until just before the embryo emerges from the capsule. 

The Velum.—'The velum is, as we have seen, originally a 


510 A. M. CARR SAUNDERS AND MARGARET POOLE. 


simple ring of cilia round the anterior end of the body (PI. 
22, fig. 4). As the anterior end, however, grows out within 
the velar area, and then becomes flattened and expanded late- 
rally, the circular shape of the ciliated band is soon lost, aud 
the latter comes to surround the widely extended anterior pro- 
longation of the body (Pl. 22, fig. 6). The velum then 
becomes notched in the mid-dorsal line and bilobed, but the 
latter characteristic is not so well marked as is generally the 
case in Opisthobranch larvee. 

In the free-swimming veliger the full extension of the velum 
is reached, but it can always be contracted completely within 
the shell. The cilia are long and prominent. Inside the 
circle of these cilia-bearing cells is a second row of cells of 
rather larger size with three or four cilia each, and in the 
middle, a cell with a single long and prominent flagellum. 

The Foot.—The rudiment of the foot is at first broad and 
blunt, projecting from the ventral surface of the embryo 
between the blastopore and the anal cells. ‘There is no sign 
of a division into two, as has been described in early stages 
of Patella. Between the stages represented in PI. 22, figs. 4 
and 5, the foot has undergone cousiderable change in shape, 
becoming elongated in an antero-posterior direction and 
flattened dorso-ventrally, and the operculum has been secreted 
on the lower surface. In the free-swimming larva it is still 
longer and covered with short cilia, and the operculum is 
capable of closing the opening of the shell completely when 
the animal is retracted. 

The Shell-gland.—In Pl. 22, figs. 2 and 3, the shell- 
gland is invaginated to form a narrow pit. It soon afterwards 
becomes everted, and fig. 4 of the plate shows the posterior 
end of the embryo covered with a thin shell. The cells that 
were invaginated now form a cap, which secretes the shell, 
the edge of the former becoming the edge of the mantle. 
The mantle-cavity in the free-swimming larva is fairly deep, 
and into it on the right side open both the anus and the 
secondary kidney. 

The Shell.—This is secreted directly the shell-gland is 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 511 


everted. It is at first very thin and transparent, and even at 
its fullest development in the free-swimming larva never 
becomes thick or resistant enough to interfere with section- 
cutting. It grows at once into its ultimate exogastric form, 
and is always perfectly symmetrical. In the free-swimming 
larva it is marked by a number of fine lines, forming an 
irregular network. 

The Anal Cells.—These have been already described (Ae. 
in Pl. 22). As development proceeds they decrease in size, 
this reduction being probably correlated with the growth otf 
the secondary kidney, which takes on the function of excre- 
tion. In the free-swimming larva they are still prominent 
features, though they are neither figured nor described in this 
stage by Mazzarelli. They presumably disappear towards the 
end of the larval period. 

The Otocysts.—These arise as ectodermal invaginations 
of about six cells, one on each side of the rudiment of the foot 
(ot. in Pl. 22). Later some ten or twelve cells sink well below 
the surface and form closed vesicles of some size, which are 
very obvious in the living larva lying at the base of the foot, 
below and to the sides of the cesophagus. At first these 
vesicles seem to be empty, but towards the end of embryonic 
life a large spherical otolith is very conspicuous inside each. 

The Nervous System.—We have seen no trace of the 
nervous system before a stage corresponding to Pl. 22, fig. 8. 
In such anembryo there are visible rudiments of both cerebral 
and pedal ganglia (c.g. and p.g.). Our preparations do not 
make the mode of origin of the nervous system very clear. It 
would appear to arise as a cell-proliferation from the ecto- 
derm, as there is no evidence of an ectodermal invagination 
to form the ganglia, as occurs in some forms, Dentalium for 
example. When the ganglia first appear, they take the form 
of slight thickenings in close contact with the ectoderm. 
The cerebral ganglia lie just above the mouth, the pedal 
ganglia to the outer and ventral sides of the otocysts, and 
slightly anterior to them. The ganglia become more definite 
and larger, but in the free-swimming larva they-are still near 


512 A. M. CARR SAUNDERS AND MARGARET POOLE. 


the surface. The two cerebral ganglia are close together and 
are united by a broad commissure. Mazzarelli states that 
cerebro-pedal and pedal commissures are present. We have 
been unable to discover these. ‘he velum and foot are at 
this stage full of connective tissue, and it would be difficult 
to trace a fine commissure if it did exist. Visceral ganglia 
are absent at this stage of development. 

The Secondary Kidney.—In making use of the term 
‘secondary kidney” we are following the nomenclature of 
Mazzarelli. In his study of the free-swimming larve of 
Opisthobranchs he gives this name to the unpaired right 
kiduey, which he has shown to be characteristic of all these 
larve. The term “ primitive kidney” he reserves (and we 
follow him in doing so) for the smaller paired kidneys, the 
nephrocysts of Trinchese, which lie anteriorly to the ‘ secon- 
dary kidney ” at the base of the velum. Itis necessary to make 
this clear, since owing to the nomenclature used in Carazzi’s 
recent work confusion may arise. In the earliest stage with 
which we deal we have already described four large ectoderm 
cells, and have said that they would give rise to the secondary 
kidney. Now Carazzi mentions these cells and identifies 
them as 3c!!, Scl!l?, 8cl?4, and 3c!#, We have no doubt that 
these four cells are the same as those which we describe and 
figure. But Carazzi in his table of the cell lineage marks 
these cells as giving rise to the “reni primitivi.’ In the 
text all he says with reference to the fate of these cells is: 
“Una parolo devo aggiungere sul desterio delle grandi 
cellule 3c!" 3c!" ; esse constituiranno uno dei primi organi 
emissionali, civé il rene primitivo.” It is quite impossible 
from this to understand whether, as one would incline to 
think from the quotation cited, Carazzi calls primitive kidney 
what we call secondary kidney, or whether, from the fact 
that in the table of cell lineage he says these cells give rise 
to “reni primitivi” (in the plural), he has-not made the 
mistake of thinking that the cells in question are the rudi- 
ment of what we call primitive kidneys, and not of what we 
call secondary kidneys, 


2) 


THE DEVELOPMENT OF APLYSIA PUNCTATA, 513 


These four cells become differentiated while in the ecto- 
dermal layer on the right side of the embryo and slightly in 
front of the anal cells. Even in the earliest stages, which 
we have examined, their nuclei are clearly to be distinguished 
from all the other nuclei in the embryo, not only by their 
much larger size, but by the presence in almost every case, 
of a conspicuous deeply-staining plasmosome, ‘'ext-fig. 15 is 
a section of a stage where these cells have just begun to sink 
below the surface. Of the two cells shown one is still in the 
outer ectodermal layer, while the other has already sunk 
below. ‘This process has gone further in Text-fig. 16 and the 
four cells are covered by a thin ectodermal layer. ‘They con- 
tinue to sink in further, and gradually give rise to a compact 
pear-shaped organ, the apex of which is directed towards the 
surface. ‘lext-fig. 17 is a section taken at a stage when the 
organ is first becoming definite. We have never seen any of 
the four cells in the process of division; but the kidney in the 
free-swimming larva consists of eight cells, and therefore each 
original cell must divide once. The cytoplasm is at first 
finely vacuolated, but as development proceeds the small 
vacuoles become confluent, and form in the external part 
of the kidney several large cavities, the narrow ends.of which 
converge to a point where they open into the mantle cavity. 
Here two small ectodermal cells form a short duct (Text-fig. 
19). In the living larva drops of coloured liquid are seen to be 
contained within the vacuoles, but as a rule they are dissolved 
out by the reagents used in the course of preservation. ‘he 
whole organ is clothed with a thin mesodermal epithelium. 

This single excretory organ has long been known in 
Opisthobranch larve, but the most diverse statements have 
been made both as to its originand function. With regard to 
the latter point, there can be no doubt that it is an excretory 
orgau since it is easy in the living larve of some Opistho- 
branchs to observe the process of excretion. ‘lhe view put 
forward by Lacaze-Duthiers and Pruvot (18) that it was an 
‘anal eye” must be held to be one of the most curious and 
unwarranted of zoological speculations. Two investigators, 


514 A.. M. CARR SAUNDERS AND MARGARET POOLE. 


Casteel working with Fiona and Heymons with Umbrella, 
have come to the same conclusions as we have regarding 
the origin of this organ. In Umbrella it is originally paired. 
There are large ectodermal cells, 3c'! and 3d'!, on either side 
of the embryo which divide and sink below the surface, one 
cell in each group remaining especially large. The cells 
on the left later disappear, while the right group forms the 
kidney. In Fiona the secondary kidney is unpaired from 
the beginning, as it is in Aplysia. It consists, however, of 
a single cell, 3c!!. ‘This closely resembles the cells which 
form the organ which we have described in Aplysia, the 
cytoplasm being much vacuolated, and the nucleus large 
and containing nucleoli. .In this case there are also other 
ectoderm cells near by, which seem to function in the same 
way. Clearly, we are dealing with a very similar organ 
in these three forms, but in Aplysia it is better developed, 
forming a definite organ with a duct and an enveloping 
epithelium, 

The ectodermal origin of this kidney was first recognised 
by Lacaze-Duthiers and Pruvot. Mazzarelli is the only 
recent writer who upholds the view that it is mesodermal. He 
has worked on Aplysia and a number of other Opistho- 
branchs, and has come to the same conclusion for them all. 
We find his observations difficult to reconcile with our own. 
The organ in question, according to his account, is derived 
from two mesodermal cells at the aboral pole, which repre- 
sent a paired rudiment of the kidney, as in Umbrella; in 
the course of torsion, however, both cells get pushed round 
on to the right side and form the single unpaired structure. 
They divide, become surrounded by other smaller meso- 
dermal cells, and finally come to communicate with the 
exterior by an ectodermal invagination. It would seem that 
he took for the rudiment of the kidney two of the large 
mesoderm cells, which lie, at the stage he describes, on either 
side of the aboral pole; the large ectodermal cells, still 
lying at the surface, he has apparently overlooked. But 
why at a later stage he should describe two cells, when 


THE DEVELOPMENT OF APLYSIA PUNCTATA, 515 


there are never less than four present, it is not easy to 
explain. 

The Primitive Kidneys.—These organs, described by 
Trinchese as nephrocysts, consist each of a large, much- 
vacuolated cell with a small nucleus, lying one on either side 
of the body at the base of the velum. In the living embryos 
they are very obvious on account of the brightly-coloured oily 
globules which they contain (Pl. 22, K1, in figs. 4 to 9). 
They appear to be characteristic of Opisthobranch larve, and 
presumably constitute temporary excretory organs. Mazza- 
relli ascribes to these cells a mesodermal origin, and when 
first seen they certainly appear to he well inside the body in 
both Fiona, according to Casteel, and Aplysia. All previous 
attempts to trace them back to their origin in segmentation 
have failed, and in spite of employing various methods of 
preserving and staining we have been equally unsuccessful. 
We consider it probable, however, that they are of ectodermal 
origin, of the same nature as the anal cells, which sink below 
the surface and lose their excretory function at a time when 
the secondary kidney is developed to take it on. 

The Alimentary Canal.—We left the segmentation 
cavity at a stage when it was largely bounded by the two 
large endomeres, 4d and B. PI. 22, fig. 3 represents a slightly 
later stage. The number of endomeres has increased, not by 
the division of the large blastomeres, A and B, but of the 
smaller endoderm cells. In the anterior region of the seg- 
mentation cavity the two large endomeres do not, therefore, 
contribute to form its boundary to such an extent as before. 
The cells which give rise to the wall of the cavity in this 
region are chiefly derived from 4a and D. The posterior 
wall of the cavity is still more complete, and its constituent 
cells are the derivations of 4b, C, and D. This corresponds 
very closely with the condition of things in Fiona, where the 
posterior wall is formed by 5B, 5b, 4c, 5C, 5c, 4D, and 
5A. Umbrella agrees very nearly in this respect with Aplysia 
and Fiona. 

At a slightly later stage (PI. 22, fig. 4) the stomodeum 


516 A. M. CARR SAUNDERS AND MARGARET POOLE. 


breaks through and comes into communication with the seg- 
mentation cavity. When this occurs, the anterior wall of the 
latter is still in part formed by the two endomeres, A and B, 
This, however, soon ceases to be the case. At this stage 
the intestine grows out as a tube-like evagination from the 
right posterior portion of the stomach, and reaches the surface 
just behind the anal cells. The anus is formed at once, and 
there is but a very slight ectodermal invagination, forming 
only the lip of the aperture. In this respect Aplysia agrees 
exactly with Umbrella. In this form the intestine arises 
from the derivatives of 5¢ and 5d; and this is probably the 
case also in Aplysia. 

The cesophagus is long and narrow, and almost entirely 
ectodermal, From an early period it is ciliated, in the free- 
swimming larva the cilia being very long and numerous, 
and in sections fillmg up almost the whole lumen of the 
tube. Mazzarelli describes a cuticle which lines the cavity 
of the cesophagus of Opisthobranch larve ; our preparations 
of Aplysia certainly do not show this structure. 

The stomach wall is formed by rather small, clearly defined 
ciliated cells, constituting a columnar epithelium. In the 
embryonic stages the cilia are all alike throughout the lining 
of the cavity, but in the free-swimming larva, in the posterior 
region they are replaced by stiff hair-like structures, which 
Mazzarelli calls “ bastoncelli.” They are probably fused 
cilia, and serve as a staining apparatus. ‘The fact that they 
are not developed until the larva becomes free-swimming 
supports this view, for until that stage is reached the embryo 
feeds upon the yolk stored in the liver, and does not take in 
food through the mouth. No epithelium can be seen covering 
the wall of the stomach externally. The intestine is asimple 
ciliated tube, and also appears to lack an epithelial investment. 
At first (Pl. 22, fig. 4) it arose from the right ventral poste- 
rior region of the stomach, but in the course of the torsion, 
which affects the whole of this part of the body, it becomes 
carried up on to the right side (PI. 22, figs. 6 and 7) ; and 
finally, in the free-swimming larva it is seen to pass from the 


THE DEVELOPMENT OF APLYSIA PUNOTATA. 517 


dorsal surface of the stomach, slightly to the left of the 
middle line. The anus opens into the mantle cavity a short 
distance below the secondary kidney. 

It will here be convenient to say a few words about the 
torsion which the embryo undergoes in the course of its 
development. This involves two processes, perfectly distinct 
from one another, though in Aplysia they take place simul- 
taneously, one being the oro-anal flexion, so characteristic 
of mollusean organisation, and the other the rotation of the 
anus and adjacent organs through more than 120°, round 
an axis coinciding with the antero-posterior axis of the 
embryo. By the former process the anus is carried forward 
to open anteriorly, and the intestine to le ventral and parallel 
to the cesophagus; while by the latter, torsion, properly so- 
called, the intestine, liver, kidney, and ccelom are carried 
from the ventral surface up on to the right side of the body. 
This movement is marked externally by the change in position 
of the anal cells (PI. 22, figs. 4, 5, 6, 7, 8). The shell alone 
appears not to be affected by the torsion, for before this 
process is complete it has in mimature assumed its final 
shape, which it retains, while the organs inside it are being 
twisted in the manner described. 

The Liver.—The endomeres A and B which form the left 
liver remain very large, and for a long time do not divide. 
At a stage corresponding to fig. 5 the nuclei divide, and 
thenceforward multiply slowly ; but for some time no corre- 
sponding cytoplasmic divisions are to be distinguished. The 
nuclei are at first large and filled with yolk, scattered amongst 
irregular fragments of chromatin. Later they become reduced 
m size after repeated divisions and lose their yolk contents. 
The endomeres are full of yolk-granules, and their cytoplasmic 
structure is thereby entirely obscured. Of the two, B is 
approximately dorsal and A ventral. The effect of torsion is 
to move Bmore over to the left and A slightly to the right. 
They remain perfectly distinct from one another for some 
time, but eventually become fused together on the left side 
to form a single organ, the left liver. We have spoken 

VOL. 55, PART 5.—NEW SERIES. 34 


518 A. M. CARR SAUNDERS AND MARGARET POOLE. 

of the two large irregular spaces generally to be seen in the 
two endomeres. These coalesce and form a cavity within 
the liver, which communicates with the stomach by an 
irregular gap in the wall of the latter on the right side and 
rather ventrally. T'ext-fig. 20 is a section taken through 
this gap. On the left side is seen the cavity of the stomach 
communicating with the cavity of the liver; on the right 
side is the right liver, to be described later on. The 
nuclei have divided and are reduced in size. The size and 
shape of the liver in the free-swimming larva can be made 
out from Pl. 22, figs. 8 and 9. It may roughly be considered 
to consist of two lobes, of which the left is derived from B and 
the right from A. The jeft lobe is much the largest; it covers 
the left wall of the stomach, projects anteriorly to it, and 
rises dorsally. ‘lhe right lobe lies ventrally, projecting 
beyond the stomach on the right side. ‘The liver is clothed 
externally by a fine epithelium of flattened cells. During 
the whole of the embryonic period the yolk is being con- 
sumed, and when the larva emerges from its capsule it is 
entirely used up. ‘The liver now takes on a new function, 
presumably one of secretion and digestion, for we have often 
observed algze and other food material in the cavity, and 
drops of secretion are at times to be seen in the cells. 

The right liver is formed in an entirely different manner. 
At a time between the stages represented in PI. 22, figs. 4 
and 5, certain cells of the right anterior wall of the stomach, 
rather nearer the dorsal than the ventral surface, become 
pushed out to form a flat, knob-like process (R./.). This 
gradually takes on a rounder shape, and the contained cavity 
increases in size, while it becomes constricted off from the 
stomach to form a definite organ, which is the right liver. 
In the free-swimming larva it is nearly round, and com- 
municates with the stomach only by a small and somewhat 
irregular aperture. When fully formed it lies almost entirely 
dorsal to the stomach, the smaller retractor muscle passes 
above it and the intestine below and to the right. The cells 
composing it are of large size, the cytoplasm consisting of a 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 519 


network enclosing uniform small vacuoles which never contain 
yolk, The nuclei are not very different in appearance from 
those of the stomach wall, but are usually larger in size. Ex- 
ternally it is clothed with a thin epithelium. The right liver 
undergoes no apparent change when the larva emerges from 
its capsule. It is presumably always secretory in function. 
Our account of the development of the liver differs from 
that of Mazzarelli, for he states that the right and left liver 
arise each from one of the two large endomeres, and this we 


TExtT-FIG. 8. 


Diagrammatic view of an embryo at a stage when the large 
retractor muscle is first formed ; seen from the dorsal surface. 
F. Foot. K1. Primitive kidney. K2. Secondary kidney. L.l. 
Left liver. M. mouth. Of. Otocyst. St. Stomach. 


have shown to be incorrect. ‘lhe peculiar mode of formation 
of the left liver is evidently correlated with the great size of 
the endomeres. All the yolk is contained within them, 
whereas, in such a form as Fiona, certain cells of the anterior 
portion of the stomach wall derived from 4b* are heavily 
laden with yolk and become gradually evaginated to form 
the liver. This mode of formation of the liver in Fiona is, in 
faet, not unlike that of the right liver in Aplysia, but the 
latter, as we have pointed out, never contains yolk. 

The large liver is very characteristic of Opisthobranch and 


520 A. M. GARR SAUNDERS AND MARGARET POOLE. 


Pteropod larvae, and serves to distinguish them from other 
Molluscan embryos. ‘There may be present only a single 
liver, as in Fiona, or there may be also a smaller right lobe, 
as in Aplysia. 

Muscles.—These are two in number, one large and con- 
spicuous and the other much smaller. The former makes its 
appearance ata stage slightly earlier than that represented 
in Pl. 22, fig. 6, when it consists of two or three fibres only, 
which arise from scattered mesoderm cells descended from 4d. 


TrxtT-FIG. 9. 


Similar view of an embryo at a slightly later stage. RJ. Right 
liver; other lettering as in Text-fig. 8. 


There is no larval mesoderm in Aplysia derived from the 
quartettes of ectomeres, which in many cases give rise to the 
larval musculature. The fibres are at first attached to 
the body-wall dorsally and to the left of the middle line at 
about the level of the mantle-cavity (Text-fig. 8). At a 
later stage they are found to have increased in number and 
to be attached further back and rather more ventrally (Text- 
fig. 9), while in the free-swimming larva their attachment 
to the body-wall is posterior, but slightly to the left of the 
middle line and still nearer the ventral surface (Text-fig. 10). 


THE DEVELOPMENT OF APLYSIA PUNCTATA. Pall 


The fibres are by that time fairly numerous, each one con- 
sisting of a single spindle-shaped cell, showing longitudinal 
striations. They pass dorsally to the left liver and are distri- 
buted to the velum and foot, and some are attached to the 
cesophagus near the mouth. The smaller retractor muscle 
appears later in development and always consists of three or 
four fibres only. They are similar to those of the larger 
muscle, and are no doubt of the same origin, This muscle is 
attached posteriorly to the dorsal body-wall on the right side, 


Trxt-FiG. 10. 


L) 
7. 


ye 
Vy 


Similar view of a free-swimming larva. C.g. Cerebral ganglion, 
P.g. Pedal ganglion. Other lettering as in Text-figs. 8 and 9. 
and thence appears to pass into the velum. As Mazzarelli 
has noticed, there are no other muscular elements of any 

kind to be seen in the larva. 

The change in position of the large retractor muscle 
noticed above is to be attributed to the fact that the left side 
of the embryo, or more correctly, a particular zone in the left 
side, grows more quickly than the corresponding zone on the 
right side. This excess of growth on one side is a familiar 
feature in discussions of the question of torsion. It was first 
brought to notice by Biitschli. There is little evidence of its 


‘a ays A. M. CARR SAUNDERS AND MARGARET POOLE. 


occurrence, but it is most interesting to note that Casteel has 
stated that in Fiona a portion of the left anterior wall of the 
stomach can be observed to grow more quickly than the 
corresponding portion in the right. It is not necessary 
for us to discuss the theories of gastropod torsion, but we 
may point out that the excess of growth on one side is 
merely the ontogenetic cause of torsion. What the original 
phylogenetic cause of torsion may have been we do not know, 
and possibly never will know. It is not only possible, but 
it is probable, that the phylogenetic cause was something 
totally different from the actual ontogenetic cause. In the 
more modified members of a group it often happens that 
certain of the older features in the organisation of the larva 
get thrown back in development. ‘To some extent this seems 
to have happened in Aplysia. with regard to the torsion, 
organs seem, that is to say, to develop already twisted. We 
have indicated the manner in which this occurs with regard 
to the development of the shell. It must also happen when 
the visceral loop is developed, for there is no sign of it in the 
larva when torsion is complete. It must therefore develop 
already twisted. 

The Coelom.—The mesoderm, as we have seen, appears at 
first in the form of an ill-defined band, but this arrangement 
is quickly lost. In Pl. 22, fig. 3 the cells are becoming 
uregularly scattered, and a little later are to be found every- 
where lying between the large yolk-laden endomeres and the 
ectoderm. In PI. 22, fig. 2 two small mesodermal cells are 
seen posteriorly, close to the anal cells; these are e and e! of 
Carazzi, which, he suggests, may possibly give rise to the 
rudiment of the genital organ. It is probable that these cells 
become involved in the formation of the ccelom, and that 
from the wall of the latter at a much later stage the germ- 
cells arise, but as it is quite impossible to follow these two 
cells through the larval development, their destiny must 
remain purely conjectural. As the foot and velum grow out, 
mesoderm cells pass into them, and there constitute a loose 
connective tissue; they also form thin epithelial investments 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 525 


to the right and left liver and to the kidneys, and later give 
rise tothe muscles. When the ectoderm cells of the secondary 
kidney have just sunk below the surface, and before they 
have become grouped together to form a definite organ, an 
irregular aggregation of mesoderm cells appears in this 
region, just anterior to the anal cells (‘ext-fig. 16). In Text- 
fig. 17, slightly later, these cells have formed a definite little 
mass close beside the now clearly developed secondary kidney, 
and in the next stage (Text-fig. 18) they are seen to bound a 
narrow slit-like cavity. This is the ccelom. In Text-fig. 19 
it has begun to extend anteriorly and dorsally so as to cover 
the dorsal wall of the secondary kidney and the right and 
antero-dorsal surface of the right liver. In the last stage, 
before the emergence of the embryo from the capsule (PI. 22, 
fig. 7), the coelom forms two lateral sacs (coloured red in the 
figure), that on the right being the larger, connected with one 
another by two transverse passages—one lying in front of, 
and the other behind, the right liver, which thus projects 
dorsally between them. In the free-swimming larva (Pl. 22, 
figs. 8 and 9) this subdivision of the ccelom has disappeared 
by the union of the anterior and posterior passages on the 
dorsal surface of the right liver. The body-cavity now con- 
sists of a considerable thin-walled sac, lying on the dorsal 
side of the larva and covering the stomach, intestine, the 
right, and a great part of the left liver, and the posterior half 
of the secondary kidney. Its ventral extension, however, is 
nowhere very great. 

Text-figs. 11-14 show the development of the ccelom 
in diagrammatic transverse sections through the region of the 
right liver and the secondary kidney. 

Hitherto the existence of the ccelom in Opisthobranch 
larvee has passed unnoticed. Mazzarelli, it is true, mentions 
a pericardium, which he describes and figures as a small oval 
sac, but in Aplysia, as we have shown, the ccelom is of con- 
siderable extent and irregular shape. It would seem that 
Mazzarelli only observed the ccelom in whole preparations, 
which would account for his describing it as a small sac, for 


Text-Fia. 11. 


A.c. Anal cells. C. Celom. Int. Intestine. K2. Secondary 
kidney. Z.l. Left liver. Cs. Esophagus. #.1. right liver. 


Trxt-Fie. 12. 


R.L. 


Lettering as in Text-fig. 11. 


Text-figs. 11-14 are diagrammatic transverse sections through the 
region of the right liver and the secondary kidney, in order to show 
the development of the ccelom, which is lettered C. 


TExt-FIG. 13.) 


QE s. 


C.d.e. Dorsal extension of celom. C.v.e.l. Ventral extension of 
ceelom on left side. Cv.e.1. Ventral extension of ccelom on 
right side. Remaining letters as in Text-fig. 11. 

Taxt-rie, 14: 

R.L. C.d.e 


«\. oof 
M\E oS mitt ant 


Qs. 


Lettering as in Text-figs. 1] and 13. 


1 The thick broken line indicates the posterior and the thin broken 


line the anterior extension of the celom; the x marks the spot at 
which the anterior and posterior portions unite dorsally to the right 
liver. 


526 A. M. CARR SAUNDERS AND MARGARET POOLE. 


only the deeper central portion is rendered visible by this 
method, and does appear somewhat as he figures it, while the 
processes which spread and pass among the organs can only 
be reconstructed from sections. 


GFENERAL CONSIDERATIONS. 


Although a large amount of work has been done on 
molluscan embryology, there still remain a number of im- 
portant questions about which there is no general agreement. 
One reason for this is that, of the many cases investigated, 
only a few stages of each are as a rule known. In the 
majority of cases where the cell lineage has been worked out 
there is no account of the later stages ; and conversely where 
these stages are well known, the cell lineage has not been 
traced. Such is the case with Paludina, about the later 
stages of which form more has been written than aboutany other 
molluscan genus. No satisfactory conclusion is likely to be 
reached until our knowledge of the earlier stages of Paludina 
is more complete. It was with the object of completing the 
account of the development in a single genus, in which the cell 
lineage is known, that we undertook this work. As yet it is 
not complete for reasons that have been mentioned. But the 
results that we have so far obtained make it necessary that 
we should consider their bearing upon certain theoretical 
points in connection with molluscan ontogeny. 

The most important facts in our description of the develop- 
ment of Aplysia are the large extent of the ccelom and 
the ectodermal origin of the kidneys. And the first of these 
points bears directly on the question of the relation of 
Annelids to Molluscs. The resemblance of the trochophore 
to the veliger larva has been long recognised; there 
cannot, indeed, be shown to be any essential difference 
between them. But there are two points in the development 
of both trochophore and veliger which concern us here—the 
type of segmentation and the development of the mesoderm. 
That form of segmentation which is known as spiral cleavage 


THE DEVELOPMENT OF APLYSIA PUNCTATA. BON 


is typical in both Aunelids and Molluscs. Whenever in these 
groups it does not occur, it is easy to account for its dis- 
appearance; in Nassa, for example, and in the Cephalopoda 
the form of cleavage is clearly correlated with the large 
amount of yolk presentin the egg. ‘There is often a strikingly 
close resemblance between the cell lineages that have been 
worked out in the two phyla, but too great stress should not 
be laid on this point, since the resemblances between the 
cleavage patterns may be taken to indicate a similarity of 
physical and mechanical conditions in the ege, rather than of 
any close phylogenetic relationship. However that may be, 
Carazzi’s work on the cell lineage of Aplysia only adds 
another to the already long list of remarkable parallels in 
_ this respect between the two groups. 

Our own work is more directly concerned with the origin 
and fate of the mesoderm in Annelids and Molluscs. In all 
Annelids in which the cell lineage has been investigated, the 
cell known as 4d gives rise to the most important part of the 
mesoderm, and sometimes to all the mesoderm, asin Aplysia. 
In some other forms the so-called larval mesoderm derived 
from the ectoderm contributes to a greater or less extent to the 
structure of the larva. In Annelids the subsequent history 
of the mesoderm is well known; and it is interesting to find 
that in many Molluscs well-developed mesodermal bands are 
found, in all ways comparable to those in Annelids. As 
might be expected, the best examples of mesodermal bands 
occur in the more primitive groups of Molluscs. Kowalevsky 
has described them in Chiton polii, and Heath in Ischno- 
chiton. Among the Solenogastres also we findin Dondersia 
and Proneomenia that the bands are unusually distinct, and 
Patten figured them clearly in Patella. But in the more 
modified groups of the Mollusca one could hardly speak of 
mesoderm bands except on the analogy of the less specialised 
forms. ‘This is the case in Aplysia, where the bands are 
never clearly defined and soon break up into scattered cells. 

Turning to the development of the ccelom, it is here that 
we find the first essential difference between Annelids and 


528 A. M. CARR SAUNDERS AND MARGARET POOLE. 


Molluscs. For in the latter phylum there is never any trace 
of sezmentation, whereas in the former, as is well known, the 
bands become split up into blocks, in each of which a 
coelomic eavity is formed. A further difference is the reduc- 
tion of the ccelom in Molluscs, but this is by no means so 
great as is usually supposed. Though the evidence is as yet 
scanty owing to the small amount of work that has been done 
on the later stages on development, nevertheless there is 
reason to believe that the ccelom is, at a late period in the 
metamorphosis, of considerable size, and that even in the 
adults of some of the more primitive groups it remains large. 
Kowalevsky long ago described the development of the ccelom 
in Chiton polii, and some of his figures, which show the 
coelom surrounding the gut, would pass well for a transverse 
section of an Annelid larvaat alate stage. Itis very probable 
that a similarly extensive development of the ccelom will be 
found in the Solenogastres, where, as we have seen, the 
mesoderm bands are of considerable size. Among the 
Aspidobranchia, as the most primitive of the Gastropoda, we 
might expect a larger body cavity than in the more special- 
ised forms, if one regards the extensive ccelom as a primitive 
factor preserved from an Annelid ancestor ; but unfortunately 
nothing is known about the later larval stages. In a recent 
account of the structure of the Neritide, Bourne, however, 
has lately described a very large ccelom, more extensive than 
has been described in any other gastropod. ‘l’o find a parallel 
to it we must refer, he says, to the Cephalopoda. There are 
a number of descriptions of the development of the Pectini- 
branchia, in which the ccelom is extensive; all the authors 
who have worked on Paludina agree upon this. Other cases 
are Vermetus, where Salensky describes a somatopleur and a 
splanchnopleur, though unfortunately he gives no figures ; and 
Bithynia, described and figure by Erlanger. Among Pulmo- 
nates and Lamellibranchs, as one might expect in such special- 
ised forms, we find no evidence of the existence of a large 
ccelom, this structure being reduced in every case to a small 
sac-like pericardium and a reno-pericardial duct. These 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 529 


examples will, however, serve to show that a well-developed 
ccelom is of frequent occurrence in the Mollusca, and that it is 
probable that when the later stages in other Molluscan groups 
have been more thoroughly examined, a large ccelom like 
that which we have described in Aplysia will be found to be a 
normal feature in the organisation of the Molluscan larve. 
But it is to be observed that the ccelom in Aplysia is developed 
at a stage when in both Annelids and Molluses the mesoderm 
bands are still intact-and a cavity has not yet been developed. 
In very few forms among the Mollusca has the development 
of the ccelom been traced from the segmentation period 
onwards. Among the forms which have been worked out, 
Aplysia and Physa follow what we may call the normal 
Annelid type, that is to say, the mesoderm, all of it in 
Aplysia, and the greater part of it in Physa, is developed 
from 4d, and from it the coelom arises. In the others there 
is a departure from this type of development; there are 
Dreissensia, Limax, and Cycias, which have been described by 
-Meisenheimer, and Paludina, according to Otto and Ténniger. 
In the first three cases the cell lineage is known, and 4d 
develops in the usual way and givesrise to bands, which split 
up and form mesenchymatous tissue, and thus seems to corre- 
spond to the larval mesoderm described by Lillie in the 
Unionidz, where it arises from 2a”, but is believed to give 
rise to the adductor muscle. The ccelom is stated to arise, 
not from the descendants of 4d, but from cells which proliferate 
from the “ectoderm” at a comparatively late stage when 
seomentation is complete. ‘lhe same is said to be the case 
with Paludina; the cell lineage is not known in this form, but 
it is distinctly stated that there are no pole mesoderm cells. 
Such a marked departure from the typical mode of develop- 
ment was hardly to be looked for; it is to be noticed that it 
occurs in widely separated members of the phylum, and 
further, that there are no peculiar bionomic conditions 
common to them. So far as our knowledge goes, it would 
seem to be an alternative mode of development, which may 
occur anywhere in the Mollusca. At first sight it might 


530 A. M. CARR SAUNDERS AND MARGARET POOLE. 


seem to involve serious difficulties as to the homology of the 
organs formed by these very different processes, but it would 
be clearly absurd to argue from this want of resemblance in 
the method by which the cells giving rise to the ccelom are 
segregated during development that the ccelom and _ its 
derivatives are therefore not homologous throughout the 
Mollusca. Evidently the heart, pericardium, and kidneys of 
adult molluscs are all homologous. It might thus seem that 
the evidence of embryology was worthless in this case; but 
these two modes of development are not so different as might 
seem at first sight. For, although a superficial examination 
of Molluscan cell lineages leads one to expect that meso- 
dermal structures are always formed from the descendants of 
4d at a parallel stage in development, closer inspection shows 
that this is by no means invariably the case. The period at 
which the mesoderm becomes segregated from the other 
embryonic elements varies considerably; it takes place in 
Planorbis marginatus whenthere are only twenty-four cells 
present; in Planorbis trivolvis when there are forty-nine ; 
and in Trochus magnus when there are 145. Statements 
have also been made that in Tethys and Teredo 4d does not 
give rise to mesoderm at all. Differences in the mode of 
segregation are thus to be found in closely allied genera, and 
we cannot lay down any hard and fast rule to govern 
developmental processes even in the same phylum. All that 
we are justified in saying in the present state of our know- 
ledge is that there are certain definite organ-forming 
substances present in the egg before segmentation begins 
which are homologous throughout the group. As this process 
takes place these may be separated into definite cells or groups 
of cells, from which the corresponding organs, or complex of 
organs, are subsequently developed; but this is by no means 
necessarily the case. The factors for the formation of 
certain organs, as, for example, the ccelom and related struc- 
tures above mentioned, instead of being aggregated at an 
early stage into a single cell, may be localised in many 
different cells with a totally different destiny, and only at a 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 531 


comparatively late stage become finally segregated out, in 
the present case by proliferation. 

That the way in which organ-forming substances present 
in the egg are finally separated from one another is quite 
immaterial in affecting the homologies of the organs to which 
they give rise is very clearly demonstrated by certain 
experimental work, as, for example, that of Wilson on the 
ego of Nereis. In this case the cleavage pattern was 
totally changed by subjecting the egg to pressure, and yet 
the larva produced was normal. 

These facts, taken together with what we know of the 
movements of the cytoplasm before and during segmentation 
in Cynthia, Dentalium, Cerebratulus, etc., show that the organ- 
forming substances often shift their position, and are segre- 
gated at different periods. Meisenheimer’s results have 
demonstrated a remarkable instance of this, but provide no 
evidence concerning the homologies of the organs. 

Before we discuss the larval excretory organs in Aplysia, 
we may briefly describe the types found among Molluscs. 

I. Flame-Cells: 

a. The flame is borne by one cell only ; the duct is intra- 
cellular. 

(1) Organ consists of two cells: Lamellibranchs. 
(2) Organ consists of four cells: Fresh-water Pul- 
monates and Basommatophora. 

3. The flames are borne by more than one cell; the duct 
is inter-cellular: ‘Terrestrial Pulmonates; Stylom- 
matophora and Paludina. 

II. Ectoderm Cells which enlarge, become vacuolated 
and project from the surface. 

a. Position variable, but near the base of the velum: 
Marine Prosobranchs. 

(3. Position definite, slightly anterior to the anus: 
Opisthobranchs. 

III. Nephrocysts (primitive kidneys). A single cell some 
distance beneath the surface, and without a duct: Opistho- 
branchs. 


532 A. M. GARR SAUNDERS AND MARGARET POOLE. 


IV. Secondary kidneys. Several large vacuolated ectoderm 
cells opening to the exterior by a short duct: Opistho- 
branchs. 

The origin of Type I has been differently described by 
several investigators, and though there seems to be much 
evidence of its arising from the ectoderm, yet a mesodermal 
origin has been ascribed to it by Erlanger in Paludina, 
Stauffacher in Cyclas, and Rabl and Holmes in Planorbis. 
If the ectodermal origin of these larval excretory organs 
should be proved beyond dispute, we should fairly be able 
to compare them with the Annelid nephridia. Nevertheless, 
if these organs are taken as representing ancestral nephridia, 
and thus indicating a relationship between the Annelid and 
Molluscan phyla, it is remarkable that they have never been 
found in the primitive groups of the latter, the Amphineura 
and the Aspidobranchia, and yet are present in the highly 
specialised Pulmonates. The second type of larval excretory 
organ is of no special significance. It has obviously been 
developed to meet some special need during the early stages 
of ontogeny. ‘'l'ype Il] may possibly be of a similar nature 
to the preceding, but it certainly stands apart from the others, 
and until we know whether it is ectodermal or mesodermal, 
as has been asserted by Mazzarelli, it is impossible to compare 
it with any other form of excretory organ. The fourth type 
of kidney, which we have called the secondary kidney of 
Opisthobranchs, offers some difficult problems both as regards 
its homology and its ultimate destiny. The position it occupies 
is very similar to that of the definitive kidney in the adult. 
Mazzarelli, when he originally described it as arising from 
the mesoderm, believed it to be the rudiment of that organ. 
This idea he has abandoned in his Jater work on the free- 
swimming larve of Opisthobranchs, and is inclined to believe 
that the organ disappears in the metamorphosis, and is in no 
way connected with the adult organ. He produces no evidence 
for this view, though, as will appear later, we think it has 
much justification. Casteel thinks it probable that the 
secondary kidney persists through the metamorphosis and 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 533 


becomes the kidney of the adult in spite of its ectodermal 
origin ; he supports his view by referring to Meisenheimer’s 
account of the “ ectodermal” origin of the common rudiment 
of heart, kidney, pericardium, and gonad in Dreissensia 
and other forms. Heymons considered the kidney to be 
merely a larval organ ; he further compared it to the external 
ectodermal kidneys of Prosobranchs, our Type II. This 
suggested homology seems to us very far fetched. To begin 
with, these Prosobranch kidneys are variable in position in 
the same species, while the Opisthobranch kidneys are derived 
from almost the same cell in the three forms, Aplysia, Fiona, 
and Umbrella. Further, the Prosobranch kidney is an 
external protruding organ, while the Opisthobranch kidney 
sinks well below the surface epithelium. 

Our own conclusion is that the secondary kidney of 
Opisthobranchs cannot be homologised with any of the other 
various molluscan kidneys. We have already given our 
reason for believing that it cannot be homologised with 
our Type II. The fundamental difference between it and 
our ‘ype I is that cilia are absent. It is possible that 
in such advanced forms as the Opisthobranchs the cilia might 
have been lost and the nephridium reduced to some such 
condition as that which we find in the secondary kidney. 
But the posterior position of the organ makes it unlikely that 
it has anything to do with the Annelid nephridium, the 
representative of which in Molluscs is always found close 
up under the velum. In the embryo of terrestrial Pulmonates, 
also, the nephridium is preserved, although they are more 
modified than the Opisthobranchs. 

The position of the secondary kidney suggests at first sight 
that it is the rudiment of the definitive kidney. But in all 
those cases in which the origin of the definitive kidney is 
known for certain, and in which it has been traced from the 
embryo to the adult, it has been found to arise as an 
evagination from the ccelomic epithelium, which joins an 
ectodermal invagination and so reaches the exterior. A 
communication between the ccelom and the kidney is present 

VOL. 9D, PART 3.—NEW SERIES. 35 


534 A. M. CARR SAUNDERS AND MARGARET POOLE. 


from the earliest stage in the formation of the latter, and 
persists as the reno-pericardial aperture in the adult of all 
Molluscs with the exception of Nautilus. Now in Aplysia 
there is never any connection between the ccoelom and the 
secondary kidney; there is no reno-pericardial duct in con- 
nection with it; the two organs, both in origin and later 
development, are perfectly distinct. We consider it therefore 
probable that the secondary kidney of Aplysia is a larval 
organ which degenerates and disappears during the meta- 
morphosis, and that the definitive kidney arises as an evagi- 
nation of the ccelomic epithelium as it does in Physa. 


List or WorKS REFERRED TO IN THE T'Ext.? 
1. Barbieri, C.—* Forme larvati del Cyclostoma elegans,” ‘ Zool. 
Anz.,’ Bd. xxxii, 1908, p. 257. 
2. Blochmann, F.—* Beitrige zur Kenntniss der Entwicklung bei 
Gastropoden,” ‘ Zeit. f. wiss. Zool.,) Bd. xxxviii, 1883, p. 392. 
8. Bourne, G. C.—‘* Contributions to the Morphology of the Group 
Neritacea,”’ ‘ Proc. Zool. Soc..’ 1908 (2), p. 810. 
4. Carazzi, D.—*L’embriologia dell’ Aplysia limacina, L.,” ‘Anat. 
Anz.,’ Bd. xvii, 1900, p. 77. ; 
“T’embriologia dell’Aplysia,”’ Archiv Ital. di Anat. e 
Embriol.,’ vol. iv, 1905. 
6. Casteel; D. B.—* Cell Lineage and Early Larval Development of 
Fiona marina,” ‘ Proc. Acad. Nat. Sci. Phil:,’ vol. lvi, 1904. 
7. Erlanger, R. von.—* Beitrage zur Entwicklungsgeschichte der 
Gastropoden,” ‘Mitth. Zool. Stat. z. Neapel., Bd. x, 1892, p. 375. 
8. Glaser, O. C.— ‘Correlation in the Development of Fasciolaria,” 
‘Biol. Bull.,’ vol. x, 1906. 
9. Heath, H.—‘* Development of Ischnochiton,” ‘ Zool. Jahrb. Abt. f. 
Anat., Bd. xii, 1899, p. 567. 
10. Heymons, °R.—**Zur Entwicklungsgeschichte von Umbrella 
mediterranea,” ‘ Zeit. f. wiss. Zool.,’ Bd. lvi, 1893, p. 245. 


11. Holmes, 8. J.—* Early Development of Planorbis,” ‘Journ, Morph..’ 
vol. xvi, 1900, p. 369. 
! For a complete bibliography of the whole subject the reader is 
referred to Mazzarelli’s monograph (14) for list of literature up to 1892, 
and to Carazzi’s paper (5) for more recent work up to 1906. 


12. 


13. 


14. 


15. 


16. 


17. 


18. 


19. 


20. 


21. 


22. 


23. 


24. 


25. 


26. 


27. 


28. 


29. 


30. 


THE DEVELOPMENT OF APLYS[A PUNCTATA. 535 


Kowalewsky, A.—‘* Embryogenie der Chiton politi,” ‘Ann. Mus. 
d’Hist. Nat. Marseilles,’ tom. i, 5, 1883. 

Lacaze-Duthiers, H., and Pruvot, G.—‘“Sur un ceil anale larvaire 
des Gastéropodes Opisthobranches,” ‘Comp. Rend. Acad. Sci. 
Paris,’ tom. ev. 1887, p. 707. 

Mazzarelli, G.—‘* Monografia delle Aplysiidse del Golfo di Napoli,” 
‘Mem. Soe. Sc.,’ tom. ix, (3), 1893. 

“Tntorno al rene secondario delle larve degli Opisto- 

branchi,” ‘ Boll. Sei. Nat. Napoli.’ vol. ix, p. 109. 

“ Contributo allo conosenza delle larve libere degli Opisto- 
branchi,” ‘ Archiv Zool.,’ vol. ii, 1904, p. 19. 

Meisenheimer, J.—‘‘ Entwicklunesgeschichte von Limax maxi- 
mus, I,” ‘ Zeit. f. wiss. Zool.,’ Bd. Ixii, 1897, p. 415. 

—— ‘“ Entwicklungsgeschichte von Limax maximus, II,” 
‘Zeit. f. wiss. Zool.,’ Bd. Ixiii, 1898, p. 573. 

“Zur Morphologie der Urniere der Pulmonaten,” ‘ Zeit. f. 

wiss. Zool.,’ Bd. Ixv, 1899, p. 709. 

* Entwicklungsgeschichte von Dreissensia poly- 

morpha,” ‘ Zeit. f. wiss. Zool.,’ Bd. lxix, 1901. 

“ Die Entwicklung von Herz, Pericard, Niere und Genital- 
zellen bei Cyclas,” ‘ Zeit. f. wiss. Zool..’ Bd. lxix, 1901, p. 417. 
Otto, H., and Tonniger, C.—‘* Untersuchungen iiber die Entwicklung 

von Paludina vivipara,” ‘ Zeit. f. wiss. Zool.,’ Bd. Ixxxv, 1906. 

Patten, W.—“ Embryology of Patella,” ‘Arb. Zool. Instit. Wien,’ 
tom. vi, 1886. 

Poetzsch, O.—‘ Entwicklung von Niere, Pericard und Herz bei 
Planorbis corneus,” ‘ Zool. Jahrb. Abt. f. Anat., Bd. xx, Heft 
3, 1904. 

Rabl, C.—* Die Ontogenie der Siisswasserpulmonaten,” ‘Jena Zeit. 
f. Naturwiss.,” Bd. ix, 1875, p. 206. 

Robert, A.—‘‘ Recherches sur le Développement des Troques,” 
‘ Arch. de Zool. expér.,’ 3¢ sér. x, 1902, p. 269. 

Salensky, W.—‘“ Zur Entwicklungsgeschichte von Vermetus,” ‘Biol. 
Centralbl., Bd. v, 1885. 

Stauffacher, H.—‘* Die Urniere bei Cyclas cornea,” Zeit. f. wiss. 
Zool.,’ vol. Ixiii, 1897, p. 43. 

Trinchese, 8.—‘‘ Per la fauna maritima Italiana, Molilide e famiglie 
affini,” ‘ Atti R. Acad. Lincei Mem..,’ vol. iii, Roma, 1880. 

Wierzejski, A.—* Embryologie von Physa fontinalis,” ‘Zeit. f. 
wiss. Zool.,’ vol. Ixxxiii, 1905. 


Trxt-FIG, 15.! 


\ | } 
age 
Section through the right posterior region of an embryo in the 
same stage as figs. 1 and 2 of Pl. 22, showing the ectoderm 
cells, K. 2, which will form the secondary kidney, still lying on 
the surface. 


TrExt-FIG. 16. 


Section through the same region of an embryo a few hours older. 

1 Text-figs. 15-20 are drawn from sections 5, thick, with an Abbe 
camera with 5), oil-immersion and a compensating eye-piece, Zeiss, No. 8, 
except Text-fig. 20, which is much less highly magnified. N.B—For 
signification of the lettering of figs. 15 to 20 see the explanation of the 


same letters in Pl. 22, given on p. 559. 


TEXT-FIG, 17. 
LE: 


Section through the same region of an embryo in the same stage 
as in fig. 4 of Pl. 22, showing the first signs of the accumulation 


of mesoderm cells in which the celom is formed at a later 
stage. 


TExT-FIG. 18. 


Section through approximately the same region of a slightly older 
embryo. The celom is now a definite cavity bounded by the 
mesoderm cells. The secondary kidney is here shown cut across 
its long axis. 


Trxt-FiGc. 19. 


Section through the same region of an older embryo, but taken almost 
at right angles to the section drawn in Text- fig. 18. The celom has 
increased in size and is beginning to extend dorsally between the body- 
walland the stomach. The secondary kidney is cut longitudinally to 
show its vacuolated structure and the duct opening into the mantle- 
cavity. 

TEXT-FIG. 20. 


Transverse section through an embryo ata stage corresponding to fig. 
showing the openings of the right and left livers into the toma 


THE DEVELOPMENT OF APLYSIA PUNCTATA. 539 


EXPLANATION OF PLATE 22, 


Illustrating Mr. A. M. Carr Saunders and Miss Margaret 
Poole’s paper on ‘“'Ihe Development of Aplysia 
punctata.” 


EXPLANATION OF ABBREVIATIONS IN FIGURES. 


A. Anus. A.c. Anal cells. C. Celom. C.c. Cerebral commissure. 
C.d.e. Dorsal extension of celom. C.g. Cerebral ganglion. C. v. e. 1. 
Ventral extension of ccelom on left side. C.v.e.r. Ventral extension of 
celom on right side. F'. Foot. Int. Intestine. K.1. Primitive kidney. 
K.2. Secondary kidney. JZ./. Left liver. M. Mouth. M.c. Mantle 
cavity. Ms. Mesoderm. (is. Hsophagus. O. Otolith. Ot. Otocysts. 
P.g. Pedal ganglion. &./. Right liver. Sh.gld. Shell-gland.  S¢é. 
Stomach. V. Velum. 


N.B.—This explanation of lettering applies to Text-figs. 15 to 20 
(pp. 536-538), as well as to the figures on PI. 22. 


Fig. 1—Embryo at the stage immediately before the beginning of 
rotation, seen from the right side. 


Fig. 2—Same embryo seen in diagrammatic optical section; the 
ectoderm is represented as peeled off from the right half of the embryo 
and thus seen in section. 


Fig. 3.—Slightly older embryo represented as in fig. 2. 


Fig. 4.—Slightly diagrammatic view of an embryo about twenty-four 
hours older than that shown in fig. 3, seen from the right side. 

Fig. 5.—Similar view of older embryo, showing first appearance of the 
celom (coloured red). 

Fig. 6.—Similar view of still later stage. The embryo has now 
assumed its characteristic veliger form. 

Fig. 7.—Similar view of an embryo a few hours before its emergence 
from the capsule. 

Fig 8.—Similar view of a free-swimming larva. 


Fig. 9.—Free-swimming larva seen from the dorsal surface. 


\j 4) 
i 
H 
~~ 
ie 
- ae 
~ 
- ’ 
7 
| e 
‘ ' 
‘ 
: 
\ 
” 
. 
= 
’ 


S29 & 
8 @.8 oe. 
@ @ & 


& 


huart Journ. Mic rmSct, Ul 5ENSPL 22. 


SS 


Fg * ANDRE & SLEIGH 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 541 


The Relation between Light and Pigment-Forma- 
tion in Crenilabrus and Hippolyte. 


By 


F. W. Gamble, F.R.S., 
Mason Professor of Zoology, University of Birmingham. 


With Plate 23. 


ContEN's. 
PAGE 

I. The Influence of Surroundings (algal backgrounds) on 
the Colour of Crenilabrus melops ; . 642 
(1) Introductory . : : . 042 
(2) Pigments and Colour Ghanees : 544. 

(3) The Influence of Daylight Reflected fr om. igal 
Backgrounds : . 346 
(4) The Influence of Light Teeenitied thr duck Alger. 548 
(5) Note on the Coloration of Larval Gobies . . 590 

(6) Summary ens )5) 

Il. The Colour-physiology of young H ippelyte varians . 9502 
(1) Methods : : . dod 
(2) Variability of the Teer Pigment F 555 


Ili. (1) The Influence of Transmitted Monochromatic Lig nt 
on the Formation of Pigments in Hippolyte 


varians : 557 

(2) The Influence of White ‘Bue fiero nde and of Tone 
chromatic Backgrounds in White Light . . 560 
IV. The Food of Hippolyte as a possible Source of Pigment . 9561 
V. Analysis of the Coloured Light Experiments —. . 564 
VI. Summary of Results , , J ote 
VII. Experimental Tables (I— VI) : : . 574 
Literature : : . 582 


Explanation of lage : : . 583 


542 KF. W. GAMBLE. 


INTRODUCTORY. 


Tis paper is a continuation of the series hitherto published 
in conjunction with Professor Keeble (1900-1905), and con- 
tains a further instalment of experimental results of a 
research upon the colour-physiology of the prawn Hippolyte 
varians, and the wrasse Crenilabrus melops. ‘The work 
was carried out by the author during the last three years, in 
part at the Plymouth Laboratory of the Marine Biological 
Association, in part at the Millport Marine Station, and also at 
Manchester University. ‘T’o the directors of these laboratories 
and to the staff of the Plymouth and Millport Stations, the 
special thanks of the author are due for the unstinted help 
which they have always been ready to give. His former 
colleague, Professor Hickson, has given the author ever- 
ready assistance and much helpful criticism. 


{. Tue Lyetuence or Surrounpines (ALcAL Backerounps) 
ON THE CoLouR OF CrENILABRUS ME ops. 


(1) Introductory. 


The immediate object with which this experiment was 
undertaken was to ascertain whether in the young stage of 
the fish there was a sensitive period at all comparable to 
that which is possessed by Hippolyte. I had thought that 
by exposing the young fish to backgrounds of diversely 
coloured weeds it would be possible to obtain some light 
as to the origin of the colour varieties which the wrasse 
exInbits. 

As in the case of Crustacea these colour varieties may be 
classed under two heads: First, the individual colour forms, 
which show a series of more or less marked vertical bars on 
a variable body colour, and second, the colour phases 
exhibited by any given individual. 

The Labride offer an exceedingly rich field of research for 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION, 543 


experiment on both these lines. It is well known that the 
ballan wrasse varies through a series.of monochrome, barred 
and spotted types of coloration from deep red to blue, and it 
has been ascertained by Holt that a given individual is 
capable of passing through colour phases from a spotted to a 
uniform livery with accompanying changes of colour. -It is 
also asserted (Noé and Dissard), that these colour varieties 
are associated with the substratum over which the fish range. 
Thus Gourret, in his beautifully illustrated memoir on the 
wrasses of Marseilles, describes the varieties of several species, 
associated respectively with Zostera and with Nullipore- 
grounds, and the seasonal changes which they undergo. 

The existence of a close relation between the coloration of 
many animals and that of their surroundings is a_ well- 
established conclusion. In the particular and striking case 
of Hippolyte varians, the development of this relation has 
been shown by Professor Keeble and the author to take place 
rapidly if young transparent animals are placed with algee in 
a strong light. ‘Vhus an experiment conducted in bright 
sunshine at Tregastel showed that out of fifteen colourless or 
pale red lined Hippolyte, eleven became red after two days’ 
association with red-brown weed; and that eight out of 
twelve became green on green weed in the same time. 
Analogous but much slower changes have been established 
by entomologists for sensitive geometrical moths during their 
early larval stages. Hxcept for this group, however, the 
amount of experimental evidence on the factors that deter- 
mine this colour sympathy is very limited. Certain insects 
excepted, Crustacea are apparently the only class in which 
the action of the environment has been tested; and even 
here the light-factors that determine the development and 
distribution of one or more pigments so as to produce an 
effect in harmony with the coloration of the environment, 
are quite unknown. 

With a view of determining these factors I undertook in 
1907 a series of experiments with one of the wrasses, the 
common gold sinny (Crenilabrus melops). 


544, I. W. GAMBLE. 


Young specimens of ballan wrasse were unfortunately not 
available at Plymouth, but this species would be an even 
more suitable one for such an investigation. 


(2) Pigments and Colour Changes. 


In the case of Crenilabrus melops the coloration is of 
a barred type. The head is marked with streaks of colour 
associated with the brain and with the lateral line organs on 
the operculum and jaws. ‘Ihe tail is usually marked by a 
central black spot, and the greenish or yellowish trunk is 
traversed by six or more vertical dark brown bars which 
extend from the dorsal fin to the anal, but do not cover the 
coelomic region. This species in its young state is the most 
abundant of the wrasses in Plymouth Sound. 

The arrangement of the pigment is as follows: Four 
colouring matters contribute to this result—blue, black, 
yellow, and red. Contrary to the statement by Krukenberg 
that the blue colouring of wrasses is due to a special 
pigment but is an optical colour merely, I find that in 
Crenilabrus melops a blue substance is associated with 
the skeleton in such a way as to give the young animal a 
transparent pale blue tone when the chromatophores are 
contracted. The nature of this substance, which, so far as I 
know, has not been previously recorded, is probably not 
pigmental, nor has it yet been determined. The green 
skeleton of Belone and the “ vivianite ” associated with some 
old red sandstone fishes possibly contain allied substances. 
Around the blood-vessels there is also a diffused . blue sub- 
stance, which is most easily noticeable in the fins and the peri- 
‘toneum, and forms a blue line along the aorta. The yellow and 
red pigments form a network derived from yellow or orange 
chromatophores scattered over the back and flanks and along 
the fin-rays. The combination of this yellow network with 
the underlying diffused blue pigment and the blue skeleton 
gives a green tinge to the young fish, whilst the expansion of 
red pigment gives a ruddy colouring; when both red and 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 545 


yellow chromatophores are expanded a dull yellowish-brown 
ground colour results (Plate 23). 

The vertical stripes are due to the development of black 
and red chromatophores along six somewhat irregular bands, 
beginning just in front of the dorsal fin and ending at the 
base of the tail fin. These bands of chromatophores are of 
considerable interest. They occur in their most marked form 
in the superb cross-striping of coral reef Labroids and other 
families, but they also appear under stimulation as a series of 
evanescent banded markings on the skin of unstriped fish. 
The common Crenilabrus rupestris shows this very well. 
When at rest it is of a nearly uniform brown or dull reddish 
colour, but on being handled or when transferred to con- 
trasting surroundings the body is seen to be overspread in a 
wave-like manner by bars of a deeper colour, which may con- 
tinue to come and goin blushes. Again, Holt has recorded 
the appearance and disappearance of dark transverse bars in 
the common ballan wrasse (Uabrus maculatus). In this case 
the fish had exactly the same property of expanding and con- 
tracting the metameric tracts of chromatophores without 
altering the body colouring. Sometimes, indeed, the bands 
disappeared almost entirely and the fish became of a uniform 
green colour. Crenilabrus rupestris has, at least in its 
younger stages, the same property. It may, and usually does, 
exhibit a banded appearance, but the bands may be extin- 
guished and the body assume an almost uniform green tinge.! 

The presence of these bars of colour is b¥ no means wholly 
dependent on the nature of the surroundings. In Creni- 
labrus melops they tend to appear under conditions that 
favour expansion of pigments, but they also appear instantly 
if a fish is transferred from white to dark vessels. ‘Tactile 
stimuli are especially effective in bringing about alternate 
flushing and pallor along these tracts. It is clear that they 
are more or less metamerically arranged tracts along which 

' Since this was written the observations of Townsend (1909) and of 


Tate Regan (1909) have revealed an unexpectedly wide range of rapid 
colour-changes in tropical fishes. 


546 lV. W. GAMBLE. 

nerve impulses act on the chromatophores. The phenomena, 
in fact, recall the pilo-motor or goose-skin reflex in man. 
Recent physiological researches (Van Rynberk) have shown 
in certain Pleuronectids that the ganglia of the sympathetic 
system supply each a definite transverse band-like region on 
the upper side, and that these regions overlap one another 
to the half of their width. Stimulation of these regions by 
induction currents produces contraction of the chromato- 
phores. Section of the spinal nerves and of the rami com- 
municantes of the sympathetic, leave the regions in question 
dark and their chromatophores permanently expanded. 

This power of localised colour change is still very imper- 
fectly understood. The development of the affected regions 
has not been undertaken, nor is the heightened coloration 
of the breeding season as yet in any way explained. It is 
therefore of some interest to note that in the case of the 
uniformly coloured adult Ctenolabrus rupestris I have 
been able to observe the banded pattern appearing in the 
post-larval stage 10 mm. long. The pattern at that stage 
differs but slightly from the livery of ‘‘melops,’’ but the 
difference is that in the former the pigmentation arises in the 
form of these transverse bars separated by clear areas, whereas 
in melops, so far as I have observed the species, the barred 
pattern is inter-connected by diffused chromatophores. The 
subsequent monochrome pattern of “rupestris ” is evidently 
derived from this earlier-barred one by development of 
interstitial pigment: but the presence of the bars even in the 
adult is revealed at the moment when under stimulation the 
skin becomes traversed by dark segmental bars alternating 
with areas of pallor. 


(3) The Influence of Daylight reflected from the 
Algal Backgrounds. 


Experiments on this problem were carried out. as follows : 
The young wrasses, obtained by hand-netting over Laminaria 
fringes, were placed in clear glass vessels, and these, in turn, 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 547 


were immersed in bell-jars filled respectively with Laminaria 
saccharina, Nitophyllum, and Ulva. Similar batches 
of wrasses were also placed in dark-bottomed and porcelain 
vessels, and in complete darkness. A double circulation was 
maintained, and the weeds were renewed twice a week. ‘The 
young fish were fed with tow-nettings and with amphipods. 
The bell-jars stood on the south side of the laboratory, and 
received diffuse daylight on all sides. 

Table I, pp. 574-575, gives the result of this experiment, 
which lasted for about three weeks. 

The light reflected from the weed backgrounds is a most 
important factor in the case, and in previous experiments has 
not received sufficient attention. Different as the three weeds 
are to the naked eye, their spectroscopic examination reveals 
little diversity ; indeed, the important differences in the hght 
reflected from their surfaces (or transmitted through them) is 
the preponderance of one or more of the parts of the spectrum 
they transmit in common. ‘Thus the green Ulva (in more 
than one layer) transmits from red to green, the green being 
somewhat more vivid than the red, but with no great 
difference of intensity. The red Nitophyllum also reflects 
red to green, but whereas the red is bright the green is exceed- 
ingly dim. Laminaria also transmits from red to green, 
but here the whole spectrum is very faint. 

The results show that brown weed backgrounds produce 
the same effect on the coloration of young Crenilabrus 
melops as does a black background. The fish may undergo 
temporary flushing and pallor under the conditions of 
examination, and there is a tendency for the dark bands to lose 
their distinctness, but the result (P].-23, fig. 1) is decisive. The 
amount of red pigment is greater than in similar specimens 
exposed to light reflected from red or green weed. The 
reflected light is more dim and is diffused over the whole 
spectrum in the case of black backgrounds than in that of the 
brown weed, and it is probably this difference which explains 
a tendency to greenness in some of the records. 

The results with green and red weeds en masse are some- 


548 F. W. GAMBLE. 


what surprising. ‘The fish in both cases become green or 
greenish, with brown bands. There is no well-marked 
differential result, such as we shall find in dealing with 
transmitted light. The yellow pigment is well developed and 
well expanded; the red pigment, however, showed more 
expansion in green backgrounds than in red. This coloration 
is one intermediate between a white and a black background 
result. In the case of red weed the effective rays are the red 
or the red-orange, and so far from these encouraging the 
develop ent and expansion of the red pigment they seem to 
have the contrary effect, for from August 14th to 20th the 
records all run green, and though there is a subsequent period 
of darkening the red colour is not noticeable. ‘The inference, 
therefore, is that in the case of red weed the red end of the 
spectrum is concerned in the formation and expansion of 
yellow pigment. In the case of green weed the results are so 
similar as to leave the specific action of the green rays 
uncertain. ‘he red and orange rays, both of green and of red 
weed, appear to act alike, while the bright green rays of Ulva 
or the dull green of Nitophyllum does not exert any very 
definite action. 


(4) Influence of Light transmitted through Alge. 
(Table II, p. 576, and Pl. 23, figs. 2 and 3.) 


When, however, the experiment of transmitting light 
through a thin layer of algal tissue is made, the results are 
not only more definite but also help to mterpret the former 
experiment. 

Table II gives the records obtained by exposing young 
wrasses of the same species as those employed for the preced- 
ing work, to light transmitted through two fronds of green, 
brown, and red weed respectively. 

For this purpose two rectangular museum jars were 
employed. ‘The inner one contained the fish, and was 
separated on three sides by a chink about 1 cm. wide from 
the outer, the space being filled with water and fronds of the 


THE RELATION BETWEEN LIGHT AND PLIGMENT-FORMATION. 549 


weed. The whole stood in strong diffused light and had a 
double circulation. 

These experiments give a much more definite result. In 
green weed surroundings three of the fish lost in a week 
their initial greenness, and, together with the remainder, 
became entirely brown. Not only so; out of four, three 
showed considerable amounts of red pigment, and, as we shall 
see, contrasted very markedly with the other experimental 
batches (fig. 3). 

‘he brown weed experiments gave a curious result. The 
fish were initially green and so remained, but in the red weed, 
out of five specimens, two of which had been originally 
greenish, three were now green and the remaining two 
showed only tinges of brown, although at first they had 
been barred with that colour. 

In this experiment, therefore, green and red weed acted 
quite differently from each other; the green light-filter 
encouraged the brown colour and red pigment, whereas 
red encouraged green colour and yellow pigment. Brown 
surroundings resembled red ones in maintaining the green 
tint. The contrast between this result with transmitted light 
and the former with reflected light is so striking and puzzling 
that I, at this point, undertook the experiments (referred to 
on pp. 952-561) on Hippolyte with a view to clearing up 
the discrepancy. 

The explanation, I believe, is to be sought in the spectro- 
scopic analysis of the light transmitted or reflected by thin 
and by thick masses of the respective weeds. hus Ulva, 
two fronds in thickness, transmits red to green, but as the 
thickness is increased it transmits orange, yellow, and green 
only. Single fronds of Nitophyllum transmit red to 
green, but several transmit almost pure red with a trace of 
green. Brown weeds transmit from red to green, the general 
intensity being low. Taking, therefore, the red as the purest 
screen, the remarkable feature about it is that along with the 
greater purity and intensity of the red light there is in the 
fish submitted to its action a green result due to the eombina- 

VOL. 50, PART 3.—NEW SERIES, 36 


550 F. W. GAMBLE. 

tion of many and well-expanded yellow pigmented chromato- 
phores with an underlying blue pigment. The light, it is true, 
is not monochromatic, but in the succeeding section it will be 
seen that a similar result obtains even when monochromatic 
light is used. 

Here, then, the result is that in strong red-orange light 
yellow pigment is well developed, but that red pigment is 
not. ‘Turning to this green-weed experiment we have the 
converse conditions and result. In these fish a brown colour 
and red pigment are strongly developed (Table IT). 

Considering that the contrast of green weed to red weed 
lies in the extension and greater brightness of the green part 
of the spectrum, the inference is that the development of red 
pigment is due to the green light and that the strong red 
light encourages the formation of yellow. The two together 
give a brown coloration. 

Brown weed in a thin film transmits from red to light blue, 
but only the red end is of fair intensity. Under a brown 
screen the fish maintain their green colour and the contracted 
condition of the red chromatophores. 


(5) Note on Coloration of Larval Gobies. 
Before passing on to these experiments on Hippolyte, I 
may interpolate a short statement of results obtained by sub- 
jecting certain larval fish to varying illumination. It is to 
be expected that the discovery of sensitive and responsive 
species will prove a difficult matter, and these notes may 

serve to help future workers in their choice of material. 
During the past summer I obtained the eggs of two species 
of Gobius (G. paganellus and G. minutus) with a view 
to determining the rate and direction of pigmentary response 
to cultural conditions. ‘The larval chromatophores are either 
black or of that “yellow” colour which is only seen by 
reflected light. Specimens developed and hatched in dark- 
ness showed normal pigmentation. On green backgrounds 
(obtained as explained on pp. 561, 562) the appearance of the 
larvee after a week’s exposure and again after a fortnight 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 551 


was not perceptibly different from that of those kept on red 
backgrounds, nor from control specimens kept in clear vessels 
uniformly illuminated. Larve of Lepidogaster gonanii 
and of L. bimaculatus were equally intractable. There 
can be no doubt that long-continued experiments are neces- 
sary owing to the slow pigmentary changes in these animals. 


(6) Summary. 
(Crenilabrus melops.) 


Darkness (for five days) produces extreme contraction of 
all chromatophores. 

Black and white backgrounds in white light (three weeks’ 
exposure) gave respectively the usual dark brown and the 
light (green) colouring associated with this illumination. 

Brown-weed background acts like black. Green and red 
algal backgrounds produce a greenish tint intermediate 
between the effects of black and white backgrounds. 

Weed light-filters produce an entirely different effect from 
the weed backgrounds (Table II). Daylight transmitted 
through green weed induced brown coloration and considerable 
amount of new red pigment. Daylight transmitted through 
red weed produced green coloration and yellow pigment. 
Brown weed is too opaque for any differential effect to show 
itself. 

These results, then, go to show that the action of algal 
backgrounds is complicated by the impurity of the colours 
transmitted or reflected from them. The nearer these approach 
to the complexity of white light the more does the background 
resemble a black or a white one, 1.e. general contraction or 
expansion results. The purer the colour or the more intense 
the particular part of the spectrum, the greater is the develop- 
ment of pigment of complementary colour. These results, 
however, were too few to establish the relation, and I there- 
fore undertook the following experiments on a more con- 
venient subject, Hippolyte varians. 


552, F. W. GAMBLE. 


II. Tae Cotour-rHysioLocgy or HirpoLtyre VARTANS. 


Previous analysis (1904) of the factors that determine the 
wide range of sympathetic coloration in Hippolyte. varians 
has revealed : 

(1) That at the time of birth the chromatophore system is 
constant in organisation, and offers such slight variations in 
the amount of the only true pigment (red) as to suggest that 
the colour of the parent does not influence that of the 
offspring. 

(2) That the next known (adolescent) stage (45 mm. long) 
presents three colour patterns: Red-lined pattern (by far 
the most common), barred pattern (rare), and monochrome 
pattern. 

(3) That these young, transparent, adolescent animals 
become green on green weed, or red on red weed within 
forty-eight hours if placed amongst a mass of weed strongly 
illuminated by direct sunlight. 

‘(4) That having assumed the tint of their surroundings 
the young animals can be persuaded to change it without 
difficulty, but that in later stages this elasticity is lost and 
colour-change is only effected slowly or not at all. 

(5) That the adolescent colour-pattern may become the 
adult one if the environment is kept constant, but that lined 
and barred patterns are in all probability transformed into a 
monochrome by the filling up of the interstices npon exposure 
to a more uniformly coloured background. ‘These results 
place the chief efficacy of colour-development in Hippolyte 
upon external factors. The eye and nervous system control 
the response to background, but do not determine it. 
Inheritance provides paths along which pigment develops, 
but does not settle the colour or pattern. The young animal 
appears plastic, but the old.one is a creature of habit. 

The need for.a more careful analysis of these responsible 
external factors and of their continued working has led to the 
following results. The variability of broods born of similarly 
and of diversely coloured parents and the prolonged action of 


THE RELATION BETWEEN LIGHT AND PIGMEN'T-FORMATION. 553 


monochromatic light are the two chief problems dealt with 
here. The conclusion is drawn that when monochromatic 
light is made to fall upon all sides of the experimental animals, 
so as to obviate a strong background effect, the result is a 
pigmentation complementary to the colour of the incident 
light and also to that obtained in Hippolyte by the use of 
coloured backgrounds and white lheht. 


(1) Methods. 


Although none of the methods employed for rearing the 
larve of Hippolyte were thoroughly successful, the record 
of the attempts made on this very difficult problem may be 
of assistance to other workers. A means of obtaining a 
satisfactory solution is one of the most pressing needs of 
experimental biology. 

The vessels used consisted of large bell-jars, supplied with 
an air- or water-current or stirred by a glass plunger. 
Seasoned vessels as well as sterilised ones were used ; filtered, 
“outside,” and tank-water were respectively employed ; 
diatoms (Nitschia) and algal cultures (Pheocystis mixed 
with other green flagellates) were used as food. ‘lhe vessels 
were shaded, exposed to diffuse light, and kept in darkness ; 
the backgrounds were translucent, absorbent, and reflecting ; 
the incident light used was monochromatic (red and green) 
as well as white light. ‘he temperature was kept down to 
16° C. by a water-jacket, and in other cases allowed to rise to 
18° C. or over, but in spite of all these variations of treatment 
the larvee only survived about ten days. It is possible that 
some means of removing the first sickly specimens would be 
a great improvement, and it is, of course, also hkely that a 
better diet could be found. ‘The larve, however, readily eat 
green flagellates and seemed to digest them. 

The monochromatic screens used in the case of larve con- 
sisted of selected pieces of coloured glass (ruby or green) 
combined with coloured gelatine films. These were placed 
over the inverted bell-jars, the sides of which were converted 
into absorbing or reflecting backgrounds. 


554 F. W. GAMBLE. 


A continuous air-current was led into the water, and the 
coloured screen was cut so that its halves embraced the air- 
tube, which was blackened at this point. ‘The junctions of 
the screen with the bell-jars consisted of black velveteen so 
as to cut out any oblique white rays, but it was found that 
great care is needed to avoid liquefaction of the gelatine films. 
A trial was made with Schott’s coloured glass, but except the 
red the samples submitted were not monochromatic. 

In order to observe the prolonged effect of monochromatic 
light, and to obviate the dominant influence of the back- 
ground, fluid screens were constructed. 'I'o insure a fairly | 
strong light the screen was made of one cell only, and not, as 
in the case of Landolt’s original design, of two or more. A 
double glass vessel was employed consisting of two beakers or 
of two large cuvettes, the inner one standing on glass supports, 
so that its rim just cleared that of the outer vessel. ‘lhe inner 
vessel was then provided with young, transparent Hippolyte 
in filtered water, and finely divided Ceramium was used as 
food. The space between the two was then filled with the 
colour filter until the level exceeded that of the water in the 
inner vessel, the top inch or so of which was rendered opaque. 
A cover of glass, or of glass and gelatine, was then placed 
over the double vessel, and the whole was then transferred to 
a shallow aquarium in a strong light. In one case a circula- 
tion of tank-water was maintained in the inner vessel. ‘The 
main point of the apparatus is to provide a means of flooding 
the animals (which remain in mid-water attached to their 
weed) with transmitted coloured light, and thus largely to 
avoid the affect of light reflected from an absorbent or 
reflecting background, such as has been generally employed 
in previous experiments. ‘I'he surfaces on which the vessels 
stood were either slate or dull white brick, but there was 
always a layer of the fluid, some 2 cm. in thickness, btween 
the bottoms as well as between the sides of the two vessels. 

The coloured solutions employed consisted of the following: 
For red a strong solution of erythrosin in distilled water, the 
strength being increased until a 2 cm. layer cuts out all the 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION, 555 


orange. Weak lithium carmine solution in a 2 mm. layer was 
used in 1909. For green a 60 per cent. solution of copper 
chloride with a trace (34, of the volume employed) of 6 per 
cent. potassium chromate gave a good result in 1°5 and 2 cm, 
thickness. For blue, ammoniacal solution of copper sulphate 
was used, a concentrated solution to which strong ammonia 
was added until the precipitate was thrown down and could 
be filtered off. Unfortunately this blue screen, probably 
owing to the ammonia exhaled, is very toxic. 

The light employed was direct, or direct and diffuse, day- 
light. In the former case the vessels stood for more than half 
their depth in a tank placed on the south side of the Plymouth 
Laboratory. In the latter the vessels were placed about 
10 ft. from the south window on the slate base of the table 
tanks. In 1909 the vessels stood opposite a north window 
on a glass shelf, and were illuminated from below by a mirror 
as well as from above and laterally. The temperature main- 
tained by a flow of water around the outer vessel was 16°5° to 
17°5° C. even in direct light; that of the inner vessel with a 
continuous water-current was 16° to 17° C, Other experi- 
mental batches were maintained in clear glass, and under 
white or black background influence as well as in darkness. 


(2) Variability of the Larval Pigment. 

The chromatophores of Hippolyte varians at the time 
of hatching usually contain a single granular pigment of a 
red (scarlet) colour. No true yellow pigment is present, but 
there is a substance in the chromatophores that is yellow by 
reflected light and brownish by transmitted ight. This is 
very constant in all broods. A variable amount of diffuse 
blue pigment is associated with the red. 

Previous investigations (Gamble and Keeble) have shown 
that “ the progeny of females (in Hippolyte varians) with 
much red pigment have more of this substance in each 
chromatophore than have those derived from green parents 
in which red pigment is less abundant. 

The question is of some importance since the initial amount 


556 . F. W. GAMBLE. 


of this substance might conceivably influence the subsequent 
colour-history. It seemed, therefore, advisable to obtain 
more records of this varying proportion of red pigment, and 
also to determine the conditions which favour or inhibit its 
development. 

The results obtained are shown in the adjoining Table, and 
are derived fromastudy of the offspring of some twenty parents. 
Larve from those Hippolyte which are pink, red, or brown, 
possess a fair amount of the red pigment in their chromato- 
phores at the time of hatching (the greatest amount in the 
samples examined being in the red-lined female broods). 
Larve from colourless (extremely pale pink) female varians 
are devoid of red pigment; whilst larve of green parents 
occupy an intermediate position, some batches being coloured 
like those of brown or pink parents, but not so deeply ; others 
from equally good green parents exhibit no red; and others, 
again, exhibit, unlike all the preceding cases, an inconstancy, 
and show traces of the red pigment in only 36 per cent. or so 
of the offspring. 


Hippolyte varians. 


Number and colour of female parent. Amount of red pigment in the just-hatched 


larve. 
2. Red-lined forms : : . | Much; constant. 
3. Pink 2 f ‘ ‘ . Fair amount; constant. 
ie estiile : ; - : . | Traces only in some, fair in rest. 
Ss.) Brown 5 : > . Fair amount; constant. 
6. Green E i é é - | Fair amount; constant. 
2. Green ‘ ; ‘ : . | Traces in 36 % of larvee examined. 
2. Green : : : é . | Absent. 
1. Almost colourless (pale pink). Absent. 


These results are of interestin several ways. They confirm 
on the whole the earlier conclusion that excessively red colour 
in the parent is associated generally with excess in the early 
larvee; but they also show that the offspring of green female 
varians show two types of coloration, namely, that with some 
red pigment and that with none; and further that the two 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 557 


types may be combined in a single brood in the proportion of 
36 per cent. dominant or pigmented forms. 

This result at once suggests that green in the parent is of 
-twofold origin; and the facts of earlier experiments support 
the suggestion. It has been shown (Gamble and Keeble) 
that green is both an independent stable colour-form aud 
also a colour assumed by brown specimens on a transfer to 
green weed. Further experiments are necessary to decide 
whether the green parents with recessive red colouring are 
of the former type, and those with more dominant red 
pigment are of the latter colour-history. ‘lhe new points 
that emerge are the absence of red pigment in certain 
broods, and its presence in only a percentage of others. 

Repeated attempts were made to experiment with broods 
from an isolated parent under diverse conditions of light, 
food, and temperature, but without much success after the 
first week or ten days. ‘The chief .results obtained were 
(1) that zoeze developed and hatched in darkness (from brown 
parents that became green under these conditions) possess 
the normal pigmentation, thus showing that light is not 
essential to pigment development, and also confirming the 
suggestion just made, that it is those green parents which had 
been previously brown that give rise to larve with red pig- 
ment; and (2) that there is a steady increase in the amount 
of red pigment in broods of green parents. For example, the 
tint of zcee of green parents approximated after a few days 
to the colouring of the larve of red parents. It is, there- 
fore, doubtful whether the initial differences in pigmentation 
between the broods of similarly or diversely coloured parents 
are of any moment in determining the ultimate coloration. 


Tit. (1) Tse Invtuence or Transmrvrep Monocuromatic Ligur 
ON THE Formation or Pigments 1x Hiprotyre VARIANS. 


Previous work on the influence of monochromatic light 
(1900, p. 619, 1904, p. 356) upon Crustacea concerned itself 


558 F, W. GAMBLE. 


chiefly with short exposures made upon an absorbing or re- 
flecting background. ‘lhe results showed that the light acted 
irrespective of its colour according to the nature of the back- 
ground, almost as though it were white light of low intensity. 
Moreover, experiments with coloured backgrounds of weed, 
against which young, transparent,almost colourless Hip polyte 
were exposed to direct sunlight, showed (1905, see Tables) 
that in two days, eleven out of fifteen prawns became red on 
red weed, and eight out of twelve became green on green 
weed. he coloured backgrounds, when flooded with white 
light, produced sympathetic colouring. The red was a 
mixture of red and yellow, the former predominating, the 
green a mixture of the same two pigments but yellow pre- 
dominating. In both cases a diffuse blue pigment occurs also. 

This result appeared to lend some support to the view of 
Wiener (1895) (which has since undergone elaboration 
[ Bachmetjew, 1903] ), and to suggest that the dominant rays 
of the background evoked especially that pigment or that 
group which agreed in colour with the reflected light. 

In order to ascertain more fully the effect of monochromatic 
light, I determined to eliminate, as far as possible, this 
dominant influence of background, and to ascertain the 
result of exposure to incident light of one colour, So far as 
I am aware, the experiment in this form has not hitherto 
been undertaken. ‘The starting-point for this experiment 
was furnished by young transparent Hippolyte varians 
taken by netting over Zostera beds and Laminaria-fringes, 
These fall into two groups: typical faintly red-lined forms 
provided with red and yellow chromatophores along the gut 
and nerve-cord, and with red ones at segmental intervals in 
the integument; and more uniformly coloured specimens 
with similar pigments, but with chromatophores more evenly 
distributed. In both cases the amount of pigment is not 
enough to give the specimens a decided tinge. They are 
similar to those used for the weed background experiments 
quoted above, and are figured on PI. 23, fig. 4. 

The vessels which were used are described above (p. 554), 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION,. 559 


and the conditions of the experiment were such as to flood 
the animals with monochromatic light on all sides, The 
weed chiefly used for food was the natural food-plant, 
Ceramium; a little fine green weed was used in one of the 
red light experiments. . The vessel was surrounded on three 
sides by the fluid colour screen and rested on a faintly re- 
flecting surface, so there was no strong background effect. 
The light employed was direct and diffused sunlight, and the 
effects of heat and of ultra-violet rays were largely obviated 
by the conditions of the experiment. 

The results of the experiment are given in Table III, pp. 
577, 578, and show that whilst the Hippolyte, in white light, 
developed into brown forms containing both red and yellow 
pigment in about equal proportions, those in red light passed 
through a brown stage, but ultimately (three weeks) becaine 
green, some remaining, however, reddish-yellow in 1909, 
whilst the survivors in green light became bright carmine, 
In other words the ultimate colour in this experiment is the 
complement of that of the incident light. 

The details of the end-result show clearly that the green 
Hippolyte produced in red light and the crimson Hippo- 
lyte produced in green light are peculiar and distinctive. 
The former possess yellow pigment in a maximally expanded 
state, and such little red as they possess is of a vermilion tint. 
Moreover, the yellow is of a distinctly greenish tinge and is 
accompanied by either very little diffuse blue or none. Thus 
the green colour in these experimental specimens under red 
hight is largely due to an increase in the amount and quality 
ot the yellow pigment accompanied by contraction of the 
formerly dominant red pigment (PI. 23, figs. 6 and 9), 

The crimson Hippolyte produced in green light is no less 
distinctive (Pl. 23, fig. 5). In contrast to the usual type of 
red forms, the yellow pigment has completely disappeared 
and the chromatophores are entirely filled with a deep carmine 
pigment suffused with a bluish tinge. The general deep 
carmine colour was new to me. Moreover, the chromatophores 
on the surface of the eye-stalks were abnormally developed. 


560 Ir. W. GAMBLE. 


In view of this very decided complementary colour-change 
the regrettable mortality that occurred in vessels exposed to 
green light in 1908 does not seriously diminish the value of 
this result, though larger numbers would add to its cogency. 
These were obtained in 1909. The experiments of 1908 and 
1909 are compared on Table V and with the other experiments 
of this paper on ‘lable LV. 


(2) The Influence of White Backgrounds and of 
Monochromatic Backgrounds in White Light. 


The effect of short exposures to the influence of white 
(porcelain) and of black (cloth or paint on glass) backgrounds 
on the colouring of young and old Hippolyte has been fully 
treated ina previous memoir.' It was there shown that what- 
ever the quality or quantity of the hght employed (within the 
experimental limits), the background effect dominated, pro- 

‘ducing contraction if: white and expansion if black. It 

occurred to me, however, to see whether the same results 
would follow a long exposure made with young specimens in 
which the pigments were rapidly developing. 

The results of a month’s trial are of considerable interest. 
The Hippolyte on black surfaces simply followed the usual 
procedure under such conditions, and deveioped maximal 
amounts of red and yellow pigments, which gave them a deep 
reddish tint. On the white background, however, after a 
first phase of transparency, they began to develop red pigment 
along the nerve-cord, and finally became uniformly marked 
with a ventral red stripe, whilst over the rest of the body the 
pigments were reduced to microscopic dots or disappeared. 
This remarkably adaptive result was obtained in diffuse light, 
the top of the deep porcelain vessel being covered with muslin, 
through which a stream of water was maintained from a tank 
above (Table IV, p. 579). 

In 1909 these background effects were extended so as to 
include the results of red and green. ‘Ihe vessels employed 
were large museum jars, painted, except for a large rectangular 

1 (1904), p. 353. 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION, 561 


window, with several coats of pure paint. Spectroscopic 
tests showed that the red was pure and that the green paint 
reflected only a trace of blue in addition to the whole of the 
green light. These vessels were kept under a water-circula- 
tion and faced a south light. Finely divided pieces of 
Ceramium were employed as food. . The Hippolyte used 
were smail, almost colourless specimens, similar to those 
employed in the other experiments on coloured light. 

The results of exposure to these monochromatic back- 
grounds was very decisive (Table IV, p. 579). Uponthe green 
one the development of pigment was arrested. ‘The Hippo- 
lyte assumed a semi-nocturnal (green) tint, and remained with 
the red pigment contracted throughout the experiment. This 
green colour is, however, not retained if the background is 
changed. Under these circumstances the animals revert to 
the pale red-lined colour variety which they exhibited 
initially. Upon the red background, on the other hand, the 
red and yellow pigments had considerably developed, and after 
a month’s exposure gave a bright orange-red tint to the 
specimens, and this persisted after change to other back- 
grounds. It would be of interest to know whether Minckie- 
wicz (1907-8), who has also obtained results of this kind 
with Hippolyte, tested the permanent or transient nature 
of the induced colouring. 


IV. THe Foop or Hrerotyre as A Possiste Source or 
PIGMENT’. 

The relation of Hippolyte varians to the alge of its 
choice is a distinctive one. The peculiar features of this 
species, the range and cryptic character of its variable colora- 
tion, its choice of, and tenacity of hold upon its weed, its 
distribution, and its food are all bound up with the presence 
of these plants. It is possible that Idothea and some 
Amphipods are equally intimately related to their habitat, 
but among macrurous Decapods Hippolyte varians is 


probably unique in this dependence upon its algal environ- 
ment, 


562 F. W. GAMBLE. 


In former papers on the subject, the relation existing 
between the pigments of Hippolyte and the coloration of its 
surroundings was explained as due to light effects, as if the 
weed backgrounds in virtue of their disposition, of their 
luminous character, and colour, acted as stimuli to the chroma- 
tophores of the prawn. However, before we accept that 
explanation, the influence of two other factors must be con- 
sidered: First, the effect of darkness on pigment-formation, 
and second, the source of these pigments, whether derivative 
or not. The first factor—darkness—is discussed on pp. 577— 
579, and it is there shown that the red (vermilion) pigment 
does not require the stimulus of light for its development, and 
that it increases in amount when the Hippolyte are kept in 
darkness. ‘The yellow pigment, however, is more dependent 
on light for its formation and increase, diminishing in amount 
in specimens kept in darkness, especially if little or no food is 
supplied to them. ‘The crimson pigment and the diffuse blue 
colouring matter are not at present investigated from this 
point of view. There is evidence, however, that light is 
essential to the production of all varieties of Hippolyte, 
except the reddish-brown ones. ‘The other factor—the source 
of pigment itself—is less known than are the conditions which 
determine each particular tint. The colouring matter of the 
food is one possible source, and this has to be briefly con- 
sidered, since, if proved, it would simplify the problem of 
sympathetic coloration. ‘That the sub-hypodermal colours of 
caterpillars and beetle larvee are due to diffusion of fatty 
pigments from the food-contents of the gut is a conclusion 
reached both by Poulton and Towers, though the physio- 
logical details of this remarkable process have never been 
ascertained. But the hypodermal colours of these animals are 
of an entirely different nature from those of Hippolyte, 
and appear to be determined by enzymes, elaborated by this 
layer acting upon the ‘primary ” cuticle or retained within 
the hypodermal cells. In Hippolyte and in Crustacea 
generally (as in the insect larve), the first formed pigments 
are developed independently of the plant-food present in the 


THE RELATION BETWEEN LIGHT AND PIGMEN'T-FORMATION. 563 


mother, and it would be of great interest to know how they 
were formed. 

In order to test the influence of food-pigments on the 
development of pigment in Hippolyte, the following experi- 
ment was carried out at Millport, N.B. A series of double 
olass vessels were prepared, the Hippolyte being placed in 
the inner chamber and a mass.of weed in the outer one. ‘I'wo 
series of pressure-bottles, one in diffused light, the other in 
darkness, were set up for isolated specimens. The food 
employed was chosen from the following: The natural alga 
chopped up into fine pieces so as not to act as a massive back- 
ground ; etiolated Laminaria, also subdivided ; the muscle of 
Hyas, the colourless ovary of Hyas, and the scarlet, mature 
ovary of the same crab. The specimens of Hippolyte em- 
ployed were 5-7 mm. in length, colourless, and tending on 
a black background to assume a faint brown-lined colour 
pattern. 


Taste A.—Feeding Experiment. 

Colour of Hippolyte after exposure to contrasted colours in 
food and surroundings, Millport, 1909. Colourless foods 
employed are crabs’ muscle, etiolated Laminaria, and 
colourless ovary of Hyas. The Hippolyte used were 
from 6-8 mm. long and colourless to the naked eye. 
Experiment lasted seven to ten days. 


Colours of foods employed. 


Colour and nature of 
surroundings. 
1 a Red (scarlet a 
Colourless. Biases Brown (alge). | Green (alge). 
| ; | aah 
Darkness . - | Pale brown-| Reddish- — Reddish or 
lined brown colourless, 
| 1 green. 
Green alge - | Green Green 2 green, =: 
f 1 brown 
Red alge . 1 = uals — — — 
; Greyish 
Brown alge = .5 | ea — -- — 
| 1 grey 
Parti-coloured | 
. A ere J 5, 4 . 
yellowish oilcloth |B lack-lined | Brown-lined | Brown-lined | Brown-lined. 
| 
\ 


564 F. W. GAMBLE. 


‘he results are shown on Table A, and at once bring out 
the fact that colourless muscle, white or red, ova are greedily 
taken up, but that the background is the dominating factor in 
the resultant coloration in daylight. Thus against a back- 
ground of green weed Hippolyte fed with colourless food, 
with red ovary, and with fine brown weed became green. In 
darkness, however, the amount of pigment in the food has 
a rough relation to the resulting colouring that will need 
further experimental testing, but there is no good evidence 
that the colour of the food determines that of the prawn. 


V. ANALYSIS OF THE CoLoureD LicHr EXpeRIMEN’s. 


(1) In green light, and amongst red weed, Hippolyte 
develops crimson and deep, not superficial, colouring. 

‘The presence in the experimental vessels of a fair quantity 
of finely branched red weed (Ceramium) would, under the 
action of diffused, strong green light act as a black back- 
ground, and this, as we know, in the presence of white light, 
encourages the formation of vermilion and yellow pigments, 
and these are most notably absent. 

‘he crimson effect in green light cannot, therefore, be 
merely due to dim light acting on a dark background. It 
must be due to a distinctive factor not present in the other 
experiments, and that factor can only be the green rays. In 
the presence of these rays, not only is the crimson pigment 
developed, but the vermilion and yellow pigments are dis- 
missed. Whether a similar result would follow if a colour- 
less food were used is of course a subject for further research. 

The most striking feature of this crimson colouring obtained 
during exposure to green light, is the fact that it is comple- 
mentary in colour to that of the incident hght. This relation 
may have a considerable significance. In an earlier paper 
(1905) it was pointed out that strongly insolated Hippolyte 
showed mobile fat in their chromatophores, and as this fat 
disappeared in specimens transferred to darkness there was 
some ground for the inference that the production of this 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 565 


fat was associated with the presence of light. If that were 
so the assumption of a complementary colouring would be 
obviously the best means of absorbing the maximum amount 
of coloured light, and of obtaining any other benefit which 
light might confer upon metabolism. Under the conditions 
of deep water, where the green or green-blue rays have 
filtered down from the surface, such a colouring would be the 
most efficient absorbing pigmentation, and it is well known 
that in hauls made from the deeper water of the English 
Channel the Hippolyte are uniformly of a crimson colour. 

The facts as to these crimson Hippolyte produced in green 
light would be most comprehensively explained by saying 
that the red Ceramium acted merely as an excitement to 
coloration, but that the carmine pigmentation is produced 
under the direct stimulus of the green light employed. 


2) Green weed . Green coloration . Superficial and deep. 
ep Red weed . Yellow or brownish- chromatophores. 
Red light 

yellow 


The action of red light is less easily analysed. The constant 
effect associated with it, is the production of yellow pigment 
and the maximal expansion of that pigment into networks 
producing a grand colour. ‘Then, according to the absorbent 
or reflecting nature of the background (i.e. green weed or 
red weed), we have a green or a brownish tint, in the latter 
case accompanied by a development of scarlet chromato- 
phores both at the surface and along the lines of the 
alimentary tract and of the nerve-cord. 

In the case of red light, therefore, it would seem that the 
direct action of the rays lies in the production of yellow 
pigment, and that the nature of the background, indirectly 
modified by the further action of the red lheht, modifies this 
yellow coloration less or more. If the background be red, 
the action of the rays is intensified, and a red background is 
thus instituted. Probably this is the factor that gives the 
scarlet chromatophores, for, as will be seen subsequently, 
that is the effect of a red background in white light; the 
resultant colour is then brownish-yellow ; but where, as in 

VOL. 55, PART 3.—NEW SERIES. 37 


566 TY. W. GAMBLE. 


the case of green weed, the background is of a less Iuminous 
character, the red colour contracts in the Hippolyte and 
the resultant coloration is then green, owing, in some cases, 
to the presence of diffuse blue mingling with the yellow 
network, and in the longest experiment to an apparent 
change in the pigment from yellow to green. ‘The most im- 
portant and most clear influence of red light, however, is the 
spread of the yellow pigment. 

These results are so strikingly dissonant from those obtained 
by subjecting Hippolyte to green or to red backgrounds that 
an explanation is clearly called for. They differ not only in 
being totally opposed to the sympathetic colouring so charac- 
teristic of the latter, but also in being slowly acquired. It 
may fairly be asked, if red light reflected from red surround- 
ings gives red Hippolyte, why does red light diffused give 
green or yellow ones? The same contradictory relation 
obtains between the action of green surroundings and diffused 
green light. 

In answer to this objection attention may be drawn to the 
double nature of the light affecting Hippolyte under natural 
conditions. here is the light reflected from the background 
and there is also the general diffuse light. 

The rapid sympathetic background colour-relations obtained 
experimentally have been made in strong daylight, and as the 
depth of water is increased or as the red end of the spectrum 
is cut off the conditions of the experiment are materially 
altered. A strongly colonred background becomes black in 
every light except that of its own colour, and in the presence 
of it we should expect the usual black background effect 
(brown, i.e. red and yellow pigments) to be produced in 
Hippolyte in any light except that with which it agreed in 
colour. But whilst this background effect is an undoubted 
factor, its potency is determined by another factor, namely, 
the definite action of diffused monochromatic light. The 
action of many rays has yet to be determined, but from the 
foregoing account a case has been made out for the action of 
green and of red light. This action, though slow, is very 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 567 


precise, and it would certainly help to account for the 
crimson and yellow colouring found in deep-waterand shallow- 
water Hippolyte respectively. 

The results, then, of these two factors, the action of diffused 
coloured light and that of backgrounds in white or mono- 
chromatic light, are not contradictory. They are the two 
factors which, so far as we yet know, are associated in the 
production of pigmentation in Hippolyte. The green speci- 
mens on Zostera are green, not only because they are ona green 
background in bright or fairly bright light, but because at or 
near the surface of the sea the red rays are miost potent, and 
their action is to produce that network of yellow pigment in 
Hippolyte, which is the basis not only of green tints but of 
those yeilowish tints that this animal assumes on the etiolated 
parts of Zostera, and of the brown specimens on various 
brown weeds so characteristic of the Plymouth littoral flora. 
The diffuse red hght, on penetrating to more densely absorb- 
ing backgrounds, such as coarser brown weeds, is checked in 
its action upon Hippolyte by the tendency for such back- 
grounds to produce red pigment in them. Hence the absence 
in such cases of that more precise colour-relation to the inci- 
dent light. The brown pigmentation contains many red and 
yellow chromatophores, but the red is scarlet and not the 
crimson of the deeper zones. 

Passing out of the range of the action of the red rays, the 
characteristic zone of the Floridez is encountered, and it is 
in this zone that the green rays are more potent. Their effect 
in producing crimson pigmentation is seen in parti-coloured 
specimens of the red-lined variety and in occasional pink 
specimens of the Laminarian zone, but it is not until a fair 
depth is encountered that their action is made clear by the 
dominance of this peculiar carmine pigment, which has 
hitherto been confused with the vermilion or scarlet one 
under the confusing term “ red.” No doubt there are similar 
effects of yellow, orange, and blue rays to be analysed before 
a full analysis of the coloration of Hippolyte can be given. 
The main conclusion derived from these experiments is that 


568 


i. W. 


GAMBLK. 


Influence of Light on the Colours of Lepidopterous Pupe. 
(After Poulton, Petersen, etc.)! 


Light. 


None 


Red (pure) 


Red (red and 
some yellow) 


Yellow (red to 
green) 


Green (pure), 
green glass 


Green (some red, 
yellow, and 
green) 


Blue. (General 
absorption least 
| in blue ; 
some red, yellow 
green and blue 
rays are 
transmitted) 


Colour of 
background. 


Red 


Orange 
Yellow 


Light green 


Dark green 
Blue 
White 


Light wood 
Orange 


White 
| Light wood 
Dark 


_ Light 


| 
| 
; 


Dark 
Plain wood 
White 


Dark 
Green 
White 

Red 
Orange 


Blue 


Light 


| Dark 


Spectrum of back- 
ground, 


| 
| 


Red 


Red to yellow 
Red to green 


Red to green (red | 
to yellow largely 
absorbed in some | 
experiments) 
General absorp- | 
tion least in green | 
General absorp- 
tion least in blue 


ee 


Resulting colouring. 


Vanessalo.|Pierisrape. 


Irregular 


(dark and | 


light) 


Darkest) - | 


l| @ 


(green) 
Very light 
(green) 
Light 
green ) 


Dark 
(brown) 
Dark 


Light 
(green) 
Ditto 


Light 
(green) 


Unknown 
Green 


Green 
Light 
(green) 


Darkish 


(V.urtice)| 


Darkish 
Ditto 

2 light, 3 
dark 
Dark 


Ditto 


Ditto 


Dark 


(Poulton), 


Light 
Green 


Light 
(green) 


eterson) 


1 The references to these papers are given fully in Bachmetjew’s 
quoted on p. 582. See especially ‘ Trans. Entomol. Soc. London,’ 1892. 


P.brassice. 


iI Dark 
| green. 


Green. 


Darker. 


work 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 9069 


the pigments developed in Hippolyte, when kept in diffused 
monochromatic light, are not the same as those which appear 
in specimens kept in daylight on a background reflecting 
these rays. Ona red background in white light, Hippolyte 
becomes reddish-orange; in pure red light it becomes 
yellowish or green. On a green background in white light 
Hippolyte becomes pale green. In pure green light it 
becomes crimson. On backgrounds of weeds, young colour- 
less specimens speedily acquire the corresponding tint. Mono- 
chromatic light, then, when saturated, has an entirely different 
effect from the same light diluted with daylight. As we pass 
from the surface to the deeper waters of the sea this dilution 
becomes less marked. The “ background effect,” so potent 
in producing the more littoral colour varieties, becomes less 
overwhelming as the red and yellow rays are absorbed by the 
surface waters. Further down, in British coastal waters, the 
blue end of the spectrum is said to be absorbed, so that at 
eight fathoms the dominant light rays are greenish or 
bluish-green (Oltmanns!). Consequently the effect of satu- 
rated monochromatic light is most probably felt in the region 
below the eight-fathom line. 

If this distinction between the effects of coloured back- 
grounds in white light and of diffused monochromatic light 
on pigment production is well founded, it should be supported 
by analogous results in other animals. Fortunately the work 
by Poulton and others upon Lepidopterous pup give a 
closely comparative result. As will be seen from the appended 
table extracted from their papers, the effects of monochromatic 
light are very different according as to whether the dominant 
rays are or are not diluted by white light. Although these 
experiments have not been made with a view to excluding 
background results so completely as those given in this 
paper, yet the distinction between the effect of red light, for 
example, when concentrated and when diluted, is quite 
analogous in the case of larval pigmentation in insects to its 
effect on pigment-production in Crustacea. As a pure 


1 * Jahrbuch. Wiss. Botanik.,’ 1892, p. 420. 


SA) F; W. GAMBLE. 


concentrated light, both red and green rays act like orange- 
yellow ones in suppressing pigment. When diluted, however, 
with white light,red rays produce pigment and pure green rays 
do likewise. As a background in daylight, therefore, the 
monochromatic rays act in one way ; as a pure incident light 
they act in an‘opposite fashion. ‘This apparently contradictory 
result is therefore supported by the evidence from experiments 
on two widely different groups of animals, Crustacea and 
Insecta. 

What exactly, then, are the factors that determine the 
extraordinary close sympathetic colour-rendering of the 
environment in the pigmentation of these animals? First of 
all in both groups, light is not essential to the production of 
pigment. Poultou’s results, as well as my own, show that 
dark-kept animals become dark coloured, though somewhat 
irregularly. In the case of Hippolyte darkness does not 
induce the formation of all the pigments. Red (vermilion), 
the dominant one, and yellow to a less extent (giving a brown 
coloration), are the only colours formed in the absence of 
light. In the insect larve, brown pigment is likewise formed 
in darkness, and develops as a sheath upon the green sub- 
epidermal layer. The action of light, then, in both groups is 
‘ather directive or inhibitory than effective. In the case of 
insects, the orange-yellow rays are apparently those which, 
when reflected from backgrounds, inhibit this brown pigment 
and allow the subjacent green pigment to confer its full 
value on the colour of the larva or pupa. In Crustacea the 
case is different; the action of these rays upon them is at 
present quite unknown. ‘The colours are pigmentary, con- 
tained in chromatophores and not “hypodermal” as in 
insects, but the production of the well-known green, brown, 
and reddish varieties of Hippolyte is due mainly to manipu- 
lations of a reddish-yellow coloration which is formed in the 
absence of definite stimulation. 

The light reflected from natural aigal backgrounds is of a 
mixed character, but with some yellow, some green, and 
varying amounts of red in it. All we have to imagine is 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 571 


that in the production of a green Hippolyte on Ulva the 
yellow and blue pigments are encouraged, the red discouraged. 
We do know that this effect occurs in pure red light, but 
in this case few red rays are reflected. We are driven to the 
conclusion that in daylight the yellow of Ulva directs the 
expansion and development of yellow pigment, and the green 
the expansion and development of blue pigment. In other 
words, we have here Wiener’s effect or conclusion confirmed. 
But when the water deepens, the red (vermilion) pigment, no 
longer inhibited by hght rays, develops more strongly, and 
yellow and brown, and even blackish, Hippolyte occur in 
response to the diffused background of brown weeds, the hght 
from which contains chiefly red and yellow-green rays. At this 
depth the incident light has lost some of its red and yellow 
rays, and is of a more bluish-green colour. From this depth 
onwards the action of diffused light becomes more and more 
apparent, that of the background less so. In the dominantly 
green water the crimson and diffuse blue pigments of 
Hippolyte develop to the exclusion and repression of the 
red and yellow ones, thus giving the various shades of 
carmine, purple and violet, that characterise Hippolyte 
taken in deeper water and in deep, shady crevices near the 
shore. Ina greater depth than that to which light extends, 
Hippolyte varians is not found. Indeed, it does not 
appear to extend beyond the range of some ten fathoms. In 
deeper water the genus is represented by Spirontocaris, the 
colour problems of which have not yet been investigated. 

If we accept this conclusion, that carmine, purple, violet, 
are colour effects, related directly to the diffuse green light 
in which many animals of deeper water live, an explana- 
tion may be found for the prevalence of these colours in many 
other groups. For example, carmine is a tint acquired by 
some fish, Crustacea, many echinids, starfish, and corals. 
Violet or purple is an even more characteristic pigment ot 
the deep-sea fauna. This purplish tint is complementary to 
green, and the relation has given rise to much speculation, 
but, so far as I am aware, the above experiments with 


or2 F., W. GAMBLE. 


Hippolyte give the first indication that the purplish colour 
is actually developed in a few weeks when the animal is 
exposed to green light. 

The significance of the scarlet colouring, so characteristic 
of abyssal Crustacea and of certain more shallow-water forms, 
e.g. Hemimysis lamorng, is still obscure, but the obser- 
vations made above as to the development of red (vermilion) 
pigment in young specimens kept in darkness may throw 
some light upon the subject. With regard to Hippolyte 
varians, the facts so far ascertained are these : 

The red pigment is the first to appear. It arises in the 
larva, even if this is reared in darkness, and the amount at 
the time of hatching is roughly proportional to that in the 
mother. In adolescent specimens subjected to darkness the 
scarlet pigment increases in amount. 


VI. Summary or REsutts. 


Crenilabrus melops. 

(L) The colouring of young specimens is due in part to the 
blue endo-skeleton and in part to chromatophores. 

(2) On backgrounds of weeds these fish assume varied 
coloration. On brown weed they become brown, on green 
weeds green, on red weed green. 

(3) In light transmitted through weeds, Crenilabrus 
assumes a colour, the complement of that which is most 
strongly represented in the incident light. ‘Thus, in light 
mainly green, a brownish red colour (due largely to red 
pigment) develops. In light mainly red, a green colour (due 
largelv to yellow pigment) develops. 


Hippolyte varians. 

(1) In any brood the amount of larval pigment (which is 
always red) is constant, and is correlated with the amount of 
red pigment present in the female parent in all colour- 
varieties except green. 

(2) A given green Hippolyte throws one of three kinds 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 573 


of young; red, colourless, or a mixed brood, containing red 
and colourless individuals in the proportion of nearly 3: 1. 

(3) This result suggests what is probable on other grounds 
—that green Hippolyte are of two, and possibly of three 
kinds: (1) Brown forms that have become green; (2) green 
forms that have undergone no change of colour; and (3) a 
cross between these two. In the absence of knowledge ot 
the male parentage of the broods, the last suggestion needs 
confirmation. 

(4) Light is not essential to the production of red pigment 
in the larva. Darkness does not prevent the continued pro- 
duction of red pigment in young forms. 

(5) The action of monochromatic light upon the pigment- 
formation of Hippolyte is entirely different from that of a 
monochromatic background in white hght. 

(6) In pure red light, yellow pigment develops. In some 
cases this leads to a green coloration: in others the colour 
remains yellow. 

(7) In green light a carmine pigment is produced, and any 
red or yeliow pigment existing in the experimental batch is 
either destroyed or disappears almost completely. 

(8) On a red background in white light, Hippolyte becomes 
reddish-orange. 

(9) Ona green background in white light, Hippolyte be comes 
green, but the colour is not retained if the batch is transferred 
to an absorbing dark background. 

(10) Continued exposure to daylight and a white back- 
ground produces hypertrophy of the red pigment along the 
nerve-cord and a disappearance of the red and yellow 
pigment elsewhere. 

(11) The production of sympathetic colouring in the 
shallower zones of the coast is explained as a background 
effect, in which the incident diffused lhght plays little part. 
The influence of background is predominant. The prod uction 
of crimson colouring in deeper water is explained as due to 
diffused green light. 

(12) There is no evidence that the pigments of the food 
(algze) are the sources of the pigments of Hippolyte. 


AMBLE. 


ri 


W.G 


F. 


) 


~ 
e 


Suol | -xe Wo 90U0 Ye SuTysNy 
| ‘UM FZ puv ‘TZ | kq dn poaoys quq “qureyz 


‘ST ‘poppe suowto | 4say ge spurgq ‘uooad | | Y90% 
‘aSuryo on | -ods popuvq-oped g | jo oSury yLH YstuMoag | -- — — ysneny 
oovjans 
[erjzuoA WO YST 
WOOL TWMOTG Curt TI- | 


JIM popueq |OT) poppe suo 


UMOIG | TAAIH JFYSraq spleq poyreut-[pomM | [Joa  pormoyoo | -to0ds popuuq | YET 
-ystppot spre, Smee) “qjgep ouo ATUG  yaIM TM OA - YSU) : -yunp Aaa [py | UMorg ATPUIETE) GsLoy 


‘(qeT wo poppe ody 
potieqg Apurey peusn 


jo F) “pappe poo (qa¢T (IST Jsu.sny wo 
poy [ep UMOAG-Ystp | Jsusny uo peppy | peppy atom F) “Woot 
-por YJIM ouo TTeUs T | aOM F) “pomreq  -YystUMoAq Zi ‘Woot WO9LG YUOL Url | 
‘uoato §=6ATQOULSIP [LV | -UMOIG T “Wood Z JO oSury YIM UMOIG Z — UMOLET -vdsirty o[vg | qsnony 


*poarireq A[IUTBT [[B SSOSSRIM F 


‘up ApouIeayxo !190LT 07 pat | “past iter | 
‘snommoads patrmq APUIEy F | {aurey ie esos M & vuSeee aban ar ae suommods g *37uR4) gnd strouroads be “116 
atrp ATA “AOI WULTLA “aur ere aetna es od Nita Maeda pee ga on “Spey OL! i qs 
fenimo"ebes a ae Ree BeOS Supe HOM YES oO Tbe pate an a ae e 4 cakant etek 
ap pee Lass |\—"pogooaTsod Fy sty | B YALM poprtaoad o10M S[assoA TTIIM ILRI OY RIS He ie Sea Sh Ee Se iu 
SUT “paom UMOAG 1OF : : : : yO wo puvavl| ssvpo avo fel | guow | 
sR ISUBIIY “BASU pue “poomM UMOAG OJ VRTYI OF TIO “PAM ICM pos LY awel-[paq V—‘SSouy1ecy Bnitewenne b aaa aia ee 
ails ¢ydo4arn) incite 9 ALOIS JUOUTOS URL WU POSMOUIUAT SUM SLIT puR (Lod I ce ta out U1 ante ae hearts ae 
oY TN) POM PON | _s(eat A) poom uaoty) ‘aul tmosnur ve ut paord aon = a wer ; i x Pa are ow | 
| SOSSBIM OIL —"(RUTIBTIOUS “IMAI aia }R ECO PINUS 
| BILBVULMLBT) POOM WMO 
| 
“STOO NT 


pue spodiydmy yt pez otam Ao, “punog oy} Ul purjsy seyxVaq punosl opty MOT Fv Uday a.10M 
sosswim ounod asol{y, ‘“alayMosfa URY} SUOLSaA OSat UL S1oquINU JayeaIs UL payesoisse soroyd 
-OFLWMOLYI Youlq puv pos OF onp 9.1B savq asaaAsuRly UMOIG OYJ, ‘So1oydoqemowyo MO[[Ox FO ¥.1OM 
-jaM VB HNO] Laas Sttiaq (Sapeos yuadsozedo of} 07 os[e Ayqavd pure) uoyo[aysopus onjq of) 07 pur 
skul-ug eq} Suisnyns quowsid onpq oqy 0} onp udots yuotedsurcy-1iuos “pup B st Inofod punoss 
ey, ‘(FFG ‘d uo uorjditosop pur Z sy) uoretofoo jo ody paasreg Ajurey [ensn oy Fo a0 
Aoy, “Ypousl seuladjxe Ul UL ZT PUB “WU OQ] SULSRADAR ‘SAZIS OA\Y JO OAOM OSSVAIM SLY} FO SMauTtOads at I, 


OGL “YQnowdé|q ‘spunoas 
-youg SNOIIVA WOAT pOZOOTFor yoy :]T guowmodxy ‘sdOTAN SAUAVIINGNAD “OSsVI,A—'|[ ATAVY, 


N'T'-FORMATION. 575 


7. 
By 


sHT AND PIGM 


EEN LI¢ 


THE RELATION BETW 


“quout 

-std mopped Jo WoryvUt 

-10F oonput sAvt pat 

Aygueurumtop ey 4eqy 

MOYS 0} sutees 4[NSet 
aed Appoqyoodxoun sty], 

“ATV LO 

qos Quoutsid poy 

‘popuvdxe pur padoy 

-oAop TOM Apes {tout 


-o1d MOTTOX “spurred 
qUIvE ATA Woado 


MOT[OA SpuUNoALS-taats 


uo potreq-WMo0.aq 

Ayquiey Zz “puno.s 
-yoeq yovyq wo yng 
‘ot 

“‘spueq 


UMOIG ITM “USTIIdLy 


‘popuvdxe 
Ajtaerodute}  saxtoyd 
-OJVULOAYD por ‘ spueq 
YStuUMorg, pamoys 


‘toumoaq ATJOUTAST(T 


SoUO DATPOTO 


ayy Ayquered 
-de sAvi o5ue10 


-pat oy, punoas 
-yoeq pot Fo yyy 
ayIyUN jou 4[Nsey 


spurq WMO TTA 
MoTTAA Z “YAO 
-jou MOT[AA poos 
pomoys worneurut 
-exo dIdOdsO.0Or TAT 
“MOT[OA - YSTtto90.a.5 
Z punoroyovg 


yoelq wo Aang 


potreg-WMo.aq 
Zz ‘UMOIG ST 
-“Md0ls T “Waal T 


/UMOIG G “WooLs T 


pepueq 
-UMOIG & "WAIL T 


asuvyo ON 


pUNoASyoV WERT ITT 
| ov 07 SMlees Poo UMOA 


poo Waa.ts 
10 pat YP yuorttod 
-xo UT UeYy yuoutoId pot 
AMO “PpUNoOLS YSTILIALS 
UO UMOIG TYITM pa.red 


Tk “TOTSURAX: 

| -TT9AA worsuvdxa 
oinstea OF punodrs 
-oVd YOVL uo Jug 


aor) ON 


YSTWoa.1.5 
Zz ‘anopoo-Apoq wWee.to 
(uo spuvq jury ygIm Z 


IMopoo puno.xs (TST 
-ue0ds) roped rayyet Z 
‘papurq-[Tom pure yarep g 


—aouURYD ON 


ayAjoddryy 
jo aseyd-anojoo 
[eumyoou 
| OF apqvaeduoo 
| ‘satoydozvut 
OLD FO WOTPORAY 
“OD OUT. XA 


| 
| 
| 
| 


UMOIG 
| Jo a0ury YIM 
| ONL “UsTTg 
quomdsuety, 
(q48zZ JSnony 
/ wo worsuvdxea 
Aaetoduta J, ) 
‘soroydoq 
|-BUTOATLD JO WOTy 
-OBAIPTOD OUTA.TY 
“XW “ony gue 
/-edsuviy Ato A 
(yQ9ZISony Ut 
qnd ouoy ‘urut 
| O&-0Z suout 
|-toods MMO. Q) 


| 


ysy 
[ey uourLtedxe 
UTP ILE gk BIBS) 
-yarq “puno.s 
| -youq yortq Fo 
qooyfo [Busn ot], 


SIV ALeF FTA 
onTq juored 
-SUB.Iy “peo 


out 


oUt 


0491 
popurq 


[Tom pure 


yap Agoa [Ty -edsuvay opeg | ysusny | 


pund.azcyorq 
OFIYM JO qooyo 
Teusn oy, 


| 
| 
| 
| 
| 
| 


| (peom | 
weeds) JuoUt 
|-1todxo Toyjour, 
| OF patty 
-SUBIY MOTLOTOM 
asoy, “OUT 


Wao FUuat 


soouo 
-LofUuy 


106 
4ysn.ony 


ILE 
4ysn.ony 


W4S6 


PUS 
qsnony 


ISTE 


576 


F. W. GAMBLE. 


Taste II.—Experiment II. Creniiasrus metors. Light 


transmitted through Weed. 


A. Green weed transmitting | 3 soee 
orange to green and a little | er a Le 00d te 
red hight.—T wo rectangular aaaition a % aa “bere f 
vessels fitted one inside the». Brown weed trans- red licht, Bite 
Date.— | other, with a space of lem. miutting orange and a Tae ‘taclon een D 2 a 
August |between the two. This space trace of redand green, nd Mito ‘i ites 
23rd, was filled with Ulva (3-4 light of low intensity.— an facainy a M9 a 
1908. fronds). ‘The inner con- Similar receptacles = te rs a 2 ieee 
tained the fish and was = with Laminaria. Sreenieh; 4 See 
Without weed. 3 specimens, 17 mm., faint barred; 1 
ask, ? 
| 8-10 mm., greenish ; 2 speci-| ae 10 ee brown 
mens, 8-10 mm., brownish. it 
| 
August 2green,3 brown-barred. 5 green specimens 2 greenish; 3 faint- 
26th (1 small green exa- | barred. 
mined: — pale blue 
pigment diffused 
roundthe gut. Yellow | 
pigment well ex- 
panded) 
August 1 green, 3 pale grey- | All green = 
27th brown barred 
August All brown, consider- Ditto 3 green (2 with tinge 
31st able amount of red of brown). 
pigment in 3 
Infer-| Conversion of green The amount of Under the  infiu- 
ence colour to brown by light was pro- ence of red light 
development of red! bablyinsufficient the complementary 
pigment. The effect) todo more than colour is retained 
of orange-green light | act as dimness, or developed. 
favours this change which produces 
a greenish color- 
ation 


TWEEN LIGHT AND PIGMENT-FORMATION, 577 


1 
vy 


THE RELATION BE 


“SSoTINOTOO 
Z@ iO [ ‘swqoy 
peult + uwMOAq-YsTp 
-pod IIB} IO poosgg | 
(yggTgsusny wo. 7) 

“U48SE 
qysnony wo praq 

| ‘peutt-poet q{urey 

|z ‘“peuly-pet poos 

z ‘4uotedsursy Ata A | 


| 


| SOTRT 


|B sypeqs oAn 


(ua8T 
qsnony wo peqyirys | 
OZ jo Jo[T «aoygo | 
|-Uy) ‘peur, pet [Ty 


inchicys) 


queustd pat oy 
punore peyreut [pea 
onjq  “jues 
-qe quourctd Molla x 
‘UOSUULIO «=PIATA ATOA 
eee 
-tadns puv deep yy40q 


| pedoyaaop-T[eM “yueur 


-std wosmi1 aand 
B Jo JIB satoydozeut 


qysty 
por UI osoyy URYy pea 
aSMzUL 910 — “yoo yo 
“pod ALey T 
‘pat poos G ‘pert [NF 


‘soroydoqzeut 
-oayo daop pure [eIoy 


_-aeadns ut mopped pur 


pet youyT “qs. 
ZU OT-O7FIp OF VpTUITS 


SUILLOF Paty 
por ystmozjoad @ Spurd 


|p ‘pour-pex Atoa [TY Z ‘pour, por peordAq g 


Ystusea.6 Y40q “Jol Z 


soTey 
aN] ALA purw poqzorary 
-0N—pey ‘pesva.ro 
“ul youtm ‘ynzryue,d 
puv pepurdxe TOA 
—'MOT[OX ‘(per outos 


| pue Moyjed Yonur) yst 


| OID ‘OUTUTIV ATESIAG T | SULMOF pouUT[-uMOAq 9 | 


-MOT[OA T *(ULLOF poury 


| -WeeI8 [ “Weds IVT T 


WdeIs POoOs Z) Wda.L0 F 
popurd 
-xo pur podorasop 
yon yuourstd Moyo 
‘moo. - yStMoTToA 
‘sumqOJ PoUT[-Ystppea 9 


ULIOF PoUl][-pod JUIVF T “ULAOF pouUry 


-Ystumorq [ ‘Sotoydozyeutorys ocuv10 ddvzns aoe] [T “poptoood ota 


atoyy Worsuedxe WNUWTXeUL UIL{GO 07 FYSTL WIP Ul punotoyoug youyTq 


0} otnsodxe 41oys vB to4yyy 


‘A[UO pot T ‘mopfe owos pur pet ITM 


z  ‘satoydozeutorya 
MOT[OA puV pat TZIM CG 


“W409 FOvIq 
YIM palaaoy "poof oF 
[TINTUERIIQ “pezyRpNoaTd 
TOBA TOL T.SnoOIy4 
‘rel tanesnur SurmL1eyu09 
YURy ayv[s y—"ssouyrvqd 


*yUR9 UL SUIpURIg 
*posuyip pus yo11q@—" I svyt 
‘aqRuoaryd “4od JO 9dB.14 
TIE OL) Jo (ory “WLd 
G.[) WOTNTOS "411990 zed 09 B 
Aq poyeiedas saaxzvaq OMY, 
—'YIeT usny —"4y.sTT W8eLH 


‘O o€ ZT-LT YUV] UT 
surpuryg 
10 JOO —"IYHvT “pooy roy 

WUtIMVvAIY “You wd 
G.T ulsoryyAsd *0°d QOF AG 


poyeredoas siaxyRoq 9.1] OM J, 


—TYIET ISNSNY— "FSI poy 


“qSI[UNS pesnytp 


“qILorpuns 
pesuyiq—Iysvy “suo, ‘uu 
8-9 suauttoads pauiy-ared OT 
1190z OY UO poppe BY daz0m 
-OLOIUM OFT Y “pooyx 
IOJFUINTMBAIID “Lassa 
IOUUT UT UOLYBTNIATO 194BM 
-yURyT, ‘WoAMJaqY WISsOItyy Ao 
jo LOART “UO ZB YIM soystp 
SSBLS ODAIVl Z—"4UST poy 


| SUMOF poulT-UMOIG | 


-op-]JomM = sjzuoursrd 
ames [AIM soroyd 
-oyeutoayo = daa 
‘govduavo m0 sa.coyd 
-ogeutoayo «= MOTTO 
pur port acaey Ato A 


podoyoa | 
| 
| 
| 


(ese INN. ‘939T § | 


yutd oped T 


‘SuUIIO} poul[-pot cG 


A | 


*q.SI[UNS posnuyip 
puByDotIp + 4SvT *.5U0T | 
“mu QT-J Suatateds pouty | 

-aywd § “pooy 1oOf poam 
UMOIG JUG SuMsypR pur 
UIN[T} Bp«royH ‘yuey a1 
YO YStuseds [[Mp Wo sur 
-pUuUBIS LOYROC SSBLS TBATO 
0°09 008 UWY— IAT] ONT MA 


206] ‘SNVINVA HLATOddIWR ‘JUOWMIIedxyY 4qoly poinojog—]]] Wavy, 


WP 
‘qdag 


W966 
qsnony 


1406 
4qsnony 


YI6L 
4gsnony | 


WALT 
qsnony 


“78 
qsnsony 
—"ayu 


sAMBLE. 


EY yea 


xD 


“pt0o 
-aAIOU PUL JS punos 
sesseut = Aysnq = ut 
MOTLOA PUL WOLPTULAL A. 
‘sotoydozeutoayo oovVyz 
“INS MAT “WAOF pouly 
-umoaq doaqy “Ava 
SUIALS PIS Surzpod 


“SULLOJ POULL- Por JFULLT 
lO aez [TV 
“SULLOF POUL[-Pot PULLF 


i 


“SULLOF POUT] 


0} SUIMO 4JoT T ALO 


'4yeL ZI 


‘quoredsuet oped Ano 4 


“pot poytvul-]poa [Ly 


*poqiwgs 
cz InNoqy 
Pat) WULUER19:) “AagVAL 

JO VOTPRI DOL JUBISUOL) 
“YURI OBIS UL Pesopoua 
PUB YS Suypoq WLM 

pataAoo wef umasnyy 


ouNE peWweEIg—ssouyaug 


“pooy a0y (pao 


“pag 


pAod-OAOU PUL JG 
oy} punot quourord 
OUTULIWO JO SosSvUT 
qUooSatogau — osMA(T 
soto dozvutoay [eIoy 
-radns oyg Jo seovry 


popurdxe A[png 


“OTRULOATLOIG, 
“god yo sontg + 8 Ou Jo 


ATU “UOSUILLD FILS & 


aaXBl “ud | BAG poguandes 
Soy BOG OMT —"IUSIL Uae) 


JF MoToq pure Apo eyy 
Jo aoujAns oy to YOR 
pedopoaop [pon soaoyd 
-OFWULO.LYD —- PaO Oo 
UOTPLULIO A YOM JOU 


Bw Suro; quoursid 


MOTTO “yurd T ‘pouty 
-pot T ‘mop yst 
“UMOLY [ SPOUL-UMOAG, T 
(4y9LI—pue 
oun?) JULOIT OPTI Ut 
Spooa pod WO poul[-pot 
AULD = PRL YOTIpAL 


poy ‘SULIOF pout[-por g  stoUtLoeds YALM popteyy 


Salrguir eat) 
UMIYIVT = “FULorL pay 


PUOULS TL WOT[LUOA 


poepuvdxe — pony 
‘quounstd = Moppoa 
popyepoorjot — You, 


“SMOT[OAP *YSTUOOLS | 


SSvT 
smmopoo T ‘ysppurd 
z ‘sour mopped 
YF -YSTUMOUd ZT. 


LOMB NOA1D-ALy 
“LoyBoq LAU punosre 
OULULIBO LUNIA JO 
MOAR “UL ZW “4ULoI] poy 


potoqyle 
oouEgsq ns 
OULPOT 
“SULLOF 
oped ZT 


-up 
MOT[oA 
Ol YOU] 
poutp-pot 


Oy 
“eZ “[q uo uMOTS 
| ad.44 og Jo Suouttdeds 67 
qQnoqy “poeM UMOAG OUT] 
“UOLPR[NOALO ALY “Layne 
9°02 OOSL V—"FU51T OU M. 


QOL 
Ayne 


YLT 
oun 
U9ST 
oune 
WIOL 
oun 


“6061 
‘W948 ouny 
—'97Bq 


WULULB IIL) 


‘pooy toy (paaa. paul) 


"606 ‘SNVIUVA ULATOddI}R—'(penutqyuod) [J] Alavy, 


‘(ecg ‘d aas) oAOqe WOT SU [JAM SB MO[OG MOAT poloqzUo FYSIT SpEssoA osoyy [[v UT 


Taste IV (1908-9).—Summary of Influence of Light on the 


Starting point for these experiments: Small specimens (43-54 
transparent colourless (on black background), 
and red 


| 


THE RELATION BETWEEN LIGH 


T AND PIGMENT-FORMATION. 


579 


Development of Pigments in HipponyrE varians. 


mm. 


long), 


faintly red-lined (on white background). 
chromatophores present, but not in sufficient quantity to give 


rise to a definite colour (Pl. 25, fig. 4). 


Weed (for Length of 


uniform 


Yellow 


chromatophores 


Light. Background. food). pepe ‘Resulting colour. eae ee 
& | 
| 
Re ( 1% ‘Reddish-brown Both yellow and red 
: edand Z 
None (Darkness) | },own 2 Red-lined increased, espe- 
5 6 | Brown-lined cially in the deeper 
| layers. Surface pig- 
ments disappear. 
Avoided by Ditto 13-8 That of weed | Surface and deep | 
uniform | pigments well de- 
ee iation | | veloped. 
| f 1 | Almost colour-| Surface pigments 
White | Ditto |- less absent. 
aka | 2 | Deep crimson | Deep crimson on gut 
below | and nerve-cord. 
| Green weed | Green 2 days Green in 66 | Evenly distributed. 
| | per cent. 
Red weed Red 2days | Red in 80 per Deeppigments better 
| | cent. | developed. 
Glass resting Red and) ( 2 | Reddish | Red disappeared, | 
| on slate then |; 4 |Greenandyel-| yellow developed, 
Red | green \ _ lowish green | blue developed. 
ee Avoided by Red | 4d Yellow and | Red present, yellow 
| uniform | greenish developed, blue in 
illumination | one. 
Glass resting, Red 3 Crimson Carmine and blue 
on ivory- | | only. 
“Green . glazed brick ee | : 
Avoided by | Ditto As Ditto | Carmine. Surface 
| 
| 


illumination 


have almost dis- 


appeared. 


580 


Kf. 


Wie 


GAMBLE, 


Taste V.—Summary of Results showing the Colouring 
obtained by subjecting Young, almost Colour- 
less, Hippotyre varirans to Diffused Transmitted 


Monochromatic Light. 


Red light. 


Period in 
weeks. rae So... 
1908. 
1 Reddish 
2 Reddish 
and brown 
3 Green 
| and yellow 
4. Greenish 
Food Ceramium 
and (latterly) 
a little fine 
green weed 
Final pig- | Much yellow. 
‘mentation of) Trace of red 
chromato- | (vermilion). 
| phores. Much blue or 
| (Figs. 8-9) green 


See figs. 5-7. 


Green light. 


1909. 


Brownish- 
yellow. 


Brownish-red 


Yellow and 
greenish 
Ceramium 


Much yellow. 
Some red (ver- 
milion). 


Blue 
in one 
(greenish) 
specimen 


1908, 


Red 
Carmine 


(pure ) 


Ceramium 


Carmine abun- 
Blue 


dant. 
(fair). No 
yellow. No 
vermilion 


Carmine. 


Ceramium | 


Carmine. 
Trace yellow. | 


THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 581 


Taste VI.—Showing Hffect o 


f Background on the 


Development of Pigments in young HippoLyre 
VARIANS. 
1908. 1909. 


| White porcelain vessels with an air- 
circulation. Fine red weed used for 
food. One vessel covered with a sheet 
of green glass and two layers of 
Baker’s green gelatine giving pure 
green light. 


T:me in White light. ee 
weeks. | White background. eround. 
it Allremained very 
faint red-lined parent faint 
forms. red-lined 
| | forms; trans- 
ferred to 
white light 
2 Superficial chro- — 


matophores had 
disappeared. Deep’ 
carmine ones clus- 
tered round the | 
gut and nerve-cord 
\(see Pl. 23, fig.11). 
| Specimens | 
appeared trans- | 
parent witha 
narrow crimson 
line down the 
| centre 


Infer- | Remarkably pro- | On reflecting | 
ences | tective develop- | backgrounds, 
ment of crimson green hight 
pigment in bright inhibits forma. | 
white light tion of pig- 
ments when 
employed for | 
a short time. 


VOL, 09, PART 3,—NE 


W SERIES. 


20. Pale trans- Colourless (7), Colourless (2), 


2 fect. It merely 


_ Large museum jars painted with 
| several coats of flatting, a clear) 

space being left in the front. The 
red flatting reflected red light only, 
the green flatting reflected green 
lightand atrace of bme Finered 
weed was used for food. Water- 
circulation employed. 25specimens. 


— — | 
White light. 


Green back- 
ground. 


White light. 
Red background. 


reddish (4), 


red (4), 
ereenish (4) 


ereenish (4). 


Colourless 
or faint 
ereenish (8) 


Orange (7). 


| 
deft; all faint| Bright 
ereen, but re- reddish-orange | 
verting to ned (2). 
lined forms on 
exposure to 
dark back- 
ground | 
Green light | Red light suf- 
suffused with | fused with 
bright white white light has) 
light has no adefinite effect, 
distinctive encouraging 
the develop- 
ment of red 
and yellow 
pigments. 


acts like dim 
white heht 


582 


1894. 


1904. 


1905. 


1906. 


1907. 


1909. 


1909. 


F. W, GAMBLE. 


LITERATURE, 


Noé, J., and Dissard, A.—‘‘ Déterminisme de ’homochromie chez 
les poissons,” ‘C.R., Soc. de Biol. de Paris,’ sér. 9, vol. v, pp- 
100-101. 


_ Wiener, O.—‘ Wiedemann’s Annalen,’ vol. lv, p. 225. 
_ Holt, E. W. L.—‘‘ An Observation of the Colour-changes of a 


Wrasse, Labrus maculatus, Donovan,” * Journ. Marine Biol. 
Assoc.,’ N.S., vol. iv, pp. 193-194. 


. Nagel, W. A.“ Ueber fliissige Strahlenfilter,” ‘ Biol. Centralbl.,’ 


1898, p. 654. 


. Gamble, F. W.,and Keeble, F.— Hippolyte varians: a Study 


in Colour-change,” ‘ Quart. Journ. Micr. Soc.,’ vol. 43, pp. 589-698. 


. Bachmetjew, P.—‘ Experimentelle Entomologische Studien,” 


Leipzig. (Gives full list of literature on experiments with lepi- 
dopterous larvee and pupx, and also a summary of these results 
and of the bearing of Wiener’s conclusions upon them.) 


_ Gourret P.— Monograph on the Labride Annales Museum, 


Marseilles,’ 1893, plate iv. 

Gamble, F. W., and Keeble, F.—* Colour-physiology of Higher 
Crustacea,” ‘Phil. Trans. Roy. Soc.,’ B., vol. elxxxix, pp. 195-388. 
“ Colour-physiology,” Part III, ‘ Phil. Trans.,’ B., 
vol. exeviii, pp. 1-16. 

Van Rynberk.—* Ueber den durch Chromatophoren bedingten 
Farbenwechsel,” ‘ Ergebnisse der Physiologie,’ v, pp. 347-571. 
Minckiewiez.—“ On the Range of Coloration acquired by Hippo- 
lyte,” ‘ Arch. de Zool. Expér. et générale.’ 

Tate Regan.— Observations on the Colour-changes of Certain 
Fish,” ‘ Proc. Zool. Soe. London,’ p. 130. 

Townsend, C. H.—‘ Observations on Instantaneous Changes in 
Colour among Tropical Fishes,” ‘Thirteenth Annual Report 
New York Zoological Society.’ 


THE RELATION BE'TWEHEN LIGHT AND PIGMENT-FORMATION. 583 


EXPLANATION OF PLATE 23, 


Illustrating Professor Gamble’s paper on “ The Relation 
between Light and Pigment-formation in Crenilabrus 
and Hippolyte.” 


Fig. 1—Young Crenilabrus melops (x 5) in the dark-banded 
phase induced by exposure to dark backgrounds. 

Fig. 2.—The green phase in the same fish induced by exposure to red 
light transmitted by red weed, and also by exposure to backgrounds 
of red weed for one week. 

Fig. 3—The reddish brown banded phase assumed by exposure for 
a week to light transmitted through green weed. The red colour is 
a shade too pronounced in the figure. 


Fig. 4.—Young Hippolyte varians in the almost colourless con- 
dition in which it is taken among weeds when 4-5 mm. long. (xX 24.) 
These colourless Hippolyte formed the starting-point for the experi- 
ments recorded in this paper. 

Fig. 5.—The brilliant carmine colouring induced in Hippolyte by 
exposure to pure green light for three to four weeks. (x 22.) Food- 
plant, Ceramium. 

Fig. 6.—The green colouring induced in Hippolyte exposed to red 
light for four weeks. Food-plant, fine green weed. 

Fig. 7.—The yellow colouring induced in some Hippolyte exposed 
to red light for four weeks. Food-plant, Ceramium. 

Fig. 8.—Chromatophores from fig. 7, highly magnified. (x 390.) 

Fig. 9.—Chromatophores from fig. 6, highly magnified. (x 390.) 

Fig. 10.—Chromatophores from fig. 5, 

Fig. 11.—Chromatophores from Hippolyte exposed to white 

reflected light for one month. 


mah 


a 
«ue 
: 
ee 
= 


Over brown weed, u 


Under red weed 


Quant.ourn Mien Sev. Wl, b5 NS&ZE. 


Red light 


Original Stock 


‘Huth, Lith? London 


IS THE TROPHOBLAST OF HYPOBLASTIC ORIGIN ? 585 


Is the Trophoblast of Hypoblastic Origin as 
Assheton will have it ? 


By 
A. A. W. Hubrecht. 


With 7 Text-figures. 


In the ‘ Quarterly Journal of Microscopical Science’ there 
has lately appeared (vol. 54, part 2) an article by my friend 
Assheton, in which he points out certain objections which he 
feels inclined to raise against some of the views that, were 
developed by me in a contribution to the fifty-third volume 
of this Journal, entitled ‘ Karly Ontogenetic Phenomena in 
Mammals.” 

Although I regret that he has not seen his way to comply 
with the invitation which I addressed to my fellow-embry- 
ologists in October, 1901 (it was published on p. 5 of my 
article on ‘“ Tarsius” in the ‘Verh. Kon. Akad. v. Weten- 
schappen te Amsterdam,’ vol. viii, No. 6, 1902), and which 
was intended to minimise printed disputes, where personal 
inspection of the preparations might bring about consensus 
of opinion, still, I accept his challenge (loc. cit., p. 221), 
and will now “ discuss more fully the difficulties which have 
arisen in the minds of some who are unable to accept (my) 
theoretical conceptions.” In doing so I wish to remind my 
readers that I am not going to treat all the objections raised 
by Assheton one by oue. Many of them will remain sub lite 
until new facts have been discovered, settling the point in 
dispute either one way or the other. I will on this occasion 


586 A. A. W. HUBRECHT'. 


restrict myself to a point of very fundamental importance on 
which Assheton’s and my own views are diametrically 
opposed to each other, ever since 1898. If the new facts 
which I bring forward in this paper should be convincing 
enough to change the minds of those who feel inclined— 
following Assheton’s example—to look upon the trophoblast 
as hypoblastic, I have no doubt that my proposal to exclude 
from the phylogeny of Eutherian mammals any ancestor who 
deposited megalecithal eggs, like the Sauropsids and the 
Ornithodelphia, will find a more easy acceptance on their 
part. 

Assheton’s reasons for considering the trophoblast as an 
essentially entodernial foetal envelope were first developed in 
1898, in his article on ‘‘The Segmentation of the Ovum of 
the Sheep” (‘Quart. Journ. Micr. Sci.,’ vol. 41). Plate 18 
of that article presents us with a series of diagrams most 
delicately shaded in red and blue, which were meant to 
explain the mutual relations of trophoblast, epiblast, and 
hypoblast in ten different genera of mammals, and to compare 
them with the Sauropsidan arrangement. 

These diagrams have not found favour with later authors 
on this subject, and have been taken no notice of in Hertwig’s 
extensive ‘ Entwickelungsgeschichte,’ in three volumes. At 
that time I refrained from entering into any polemical dis- 
cussion, considering that later observations would show the 
untenability of Assheton’s ingenious but unsatisfactory 
generalisation. In writing his latest article Assheton has, 
however, allowed himself to come too strongly under the 
influence of his own hypothesis of twelve years’ standing. I 
see no necessity for entering upon any detailed discussion 
concerning the numerous and different arguments which have 
led other embryologists as well as myself to reject that 
hypothesis of Assheton’s now that new facts have come to 
light concerning the very earliest segmentation stages of 
Galeopithecus. This very archaic genus may be looked upon 
as a derelict representative of a group that in earlier 
geological epochs gave rise to the modern bats. There are 


—_— 


IS THE TROPHOBLAST OF HYPOBLASTIG ORIGIN? 587 


certain points of agreement between its early development 
and that of Pteropus, whilst Leche’s anatomical work (‘ Kgl. 
Svenska Vet. Akad. Handl.,’ Bd. xxi, 1886) upon Galeopithecus 
points in the same direction. Of this genus I have now in my 
possession several series of sections made through segmen- 
tation phases, some of which I have here figured. 

These sections leave no doubt that the trophoblast of 
Galeopithecus originates by delamination at as early an age 
as the two- and four-cell segmentation stage, and render it 
utterly futile to try and explain the Galeopithecus tropho- 
blast as “due to an overflow of the yolk or hypoblast cells 
over the epiblastic rudiment ”’ (Assheton, |.c., p. 228). 

If we look more closely at the three stages of Galeopi- 
thecus here figured and begin with the one that is the furthest 
developed (Text-fig. 1), we find full coincidence with a similar 
stage described by Assheton for the sheep (I.c., 1898, PI. 16, 
figs. 14, 15), by Keibel for the stag (‘ Arch. f. Anat. and 
Phys. Anat. Abt.,’ 1902, p. 292), by Weysse (‘ Proc. Amer. 
Acad.,’ vol. xxx, p. 283) for the pig, by van Beneden for 
the rabbit and bat (‘Archives de Biologie,’ vol. i), by 
myself for the hedgehog, for the shrew (‘Quart. Journ, 
Mier. Sci.,’ vol. 30, Pl. 17; vol. 31, Pls. 36, 37), for Tupaja, 
for Tarsius (‘ Verh. Akad. Wetensch. Amsterdam,’ vol. iv, 
1895, Pls. 1, 2; vol. viii, 1902, Pls. 1, 2), and for Nycti- 
cebus (‘ Keibel’s Normentafeln,’ 1907), as well as by other 
embryologists for various other mammals. This is the 
common starting-point in which there is a trophoblast and 
an embryonic knob with a cavity below it, and in which a 
hypoblast is not as yet distinctly developed, although just 
beginning to make its first appearance. It should be borne 
in mind that this very stage is thus characteristic for genera 
of mammals so diverse as those mentioned above. The 
way in which Assheton attempts to prove from yet earlier 
stages of the sheep that the outer trophoblastic layer is in 
reality a derivate of the hypoblast appears to me to be so 
pre-eminently artificial (c.f. l.c. his figures 9-14) and the 
argumentation so weak, that I must ascribe to a similar 


588 A. Av W. HUBRECH'. 


incredulity on van Beneden’s part that this latter author in 
the important article which appeared one year later than 
Assheton’s (‘ Anat. Anzeiger,’ 1899, p. 305), does not take 
the slightest notice of the English author’s view that the 
trophoblast (van Beneden’s “couche enveloppate”) should be 
looked upon as an entodermal derivate. 

If we now return to T'ext-fig. 1 of this paper and inquire how 
this stage in the ontogeny of Guleopithecus has been reached, 


Trxt-Fic. 1. 


Section of a blastocyst of Galeopithecus with embryonic knob 
and enveloping trophoblast, just before the establishment of 
the continuous hypoblast. 


we see that it has been preceded by the stages of which 
Text-figs. 2 and 3 are the representatives. 

In Text-fig. 2 the centre of the different sections is occupied 
by comparatively large nuclei, evidently belonging to a central 
group of cells—the mother cells of the embryonic knob. 
Outside this embryonic knob and forming the peripheral 
layer in these sections is protoplasm in which distinct cell- 
boundaries are not visible, but in which a certain number of 
nuclei (smaller than those of the embryonic knob) clearly 
indieate that in the live blastocyst a peripheral cell-layer was 
differentiated in addition to the embryonic knob. 


IS THE TROPHOBLAST OF HYPOBLASTLG ORIGIN ? 589 


Going back yet one stage earlier, in which the two first 
cleavage-cells are just on the point of splitting up into four (as 
is distinctly indicated by the karyokinetic figures of fig. 3C), 
we notice, besides the two cleavage-cells, three polar bodies 
of comparative large size, as they are known for mammals in 
general. Moreover, at the periphery of the two cleavage- 
cells we find separate nuclei, indicating the very first origin 
by an early delamination process of the cells which in 
Text-figs.2 and 3 constitute the continuous layer of trophoblast. 


TEXT-FIG. 2. 


Fig? 


Galeopithecus. A series of five sections through a stage of 
cleavage preceding the blastocyst of Text-fig. 1. Embryonic 
knob with bigger nuclei contrasting with the smaller peripheral 
trophoblast-nuciei. No central space as yet developed. 


In the mammalian genera Iitherto examined with respect to 
the origin of the trophoblastic layer (Tarsius, Tupaja, rabbit, 
sheep, pig, stag, dog, mouse, guinea-pig, etc.), the tropho- 
blast undoubtedly makes its appearance at a somewhat later 
stage of cleavage, or rather the distinction in the morula 
stage between the mother-cells of the embryonic knob and 
those of the trophoblast is not so soon evident as it is in 
Galeopithecus. However, the karyokinetic processes by 
which in this latter genus the trophoblastic nuclei separate 
from the segmentation nucleus (which in its turn owes its 


590 A. A. W. HUBRECHT. 


origin to the union of the male and female pronucleus) at so 
early a moment are not revealed by my preparations, and we 
cannot for the present come to any sound conclusion as to 
which of the two modes of formation of the trophoblast is 
the more archaic one. 

Recognising that the definite answer to this question can 
only be given when a number of new observations will be at 
our disposal, I may still be allowed to call attention to the 
fact that in Galeopithecus the spot where the polar bodies are 


TEXT FIG. 3. 


Galeopithecus. A series of six sections through a cleavage stage 
just preceding the formation of the second pair of cleavage- 
cells. In C karyokinetic figures indicate this. In B—E the 
polar bodies are visible. Apparent trophoblast nuclei are 
situated peripherally. 


applied against the egg (see lext-fig. 3 B—H) remains without 
trophoblast nuclei somewhat longer than other parts of the 
egg’s surface. ‘The question presents itself—supposing the 
process 18 more primitive in Galeopithecus—whether this 
particularity might have led (in such mammalian genera that 
should be considered as phylogenetically younger) to the 
arrangement which has induced van Beneden, Duval, and 
Assheton (in his later publications) to consider the cleavage- 
process of those mammals as revealing epibolic characteristics. 
In case this question will later have to be answered in the 


IS THE TROPHOBLAST OF HYPOBLASTIC ORIGIN P 591 


affirmative, the so-called blastopore which van Beneden (1875) 
described in the rabbit?s morula-stage might correspond to 
the spot referred to in the three figures (Text-fig. 3 B—D), 
where the polar bodies lie. 


Having thus shown that Assheton’s hypothesis of the 
hypoblastic nature of the trophoblast is irreconcilable with 
the phenomena in Galeopithecus, I emphatically repeat my 
conclusion that we are not justified in accepting it for- any 
other vertebrate. He himself will admit that, such being 
the case, the comparison of the trophoblast of mammals with 
the ‘‘ deckschicht ” of fishes comes to the foreground with 
increased validity. 

I have already stated above that it is not my intention in 
this paper to follow Assheton’s criticism step by step. A 
more extensive article on the ontogeny of Galeopithecus will 
appear in the course of this year. I shall there find occasion 
to reply more fully to other parts of Assheton’s criticism. 
There is, however, one point on which I feel bound to 
apologise, viz. that I have not allowed enough space for the 
recognition of the fact that my kephalo- and notogenesis had 
already been partly forestalled in several of Assheton’s papers, 
and had by him been termed proto- and deuterogenesis. I 
ought to have particularly mentioned these names in my 
paper of 1908. Still, I must maintain my terminology now 
that Assheton himself states (l.c., p. 240) that his and my 
names “ signify a different interpretation,” and now that he 
maintains that mine ‘‘does not represent the actual facts.” 
As matters stand I feel that the important issue which is at 
the base of the whole question of gastrulation in vertebrates 
(very fully treated in Keibel’s contribution to vol. x of the 
‘Ergebnisse der Anatomie und Entwickelungsgeschichte,’ 
but since then looked upon in a somewhat different hght 
after his and my own short papers in the ‘Quart. Journ. Mier. 
Sci.’ [vol. 49] and in the ‘Anat. Anzeiger’ [vol. xxvi] had 
appeared) renders any polemics about the nomenclature that 
should be adhered to untimely. Very numerous investigations 


592 A. A. W. HUBRECHT. 


are yet necessary, and will undoubtedly soon be undertaken, 
before we dispose of the comparative material which is 
necessary for settling this important point in Vertebrate 
ontogeny, and for finally deciding which nomenclature ought 
to be adhered to. I gladly leave the latter decision to others, 
but would not let this paper see the light without recognising 
that until lately I have not sufficiently been aware that 
Assheton already in 1894 expressed opinions to which Keibel 
and myself have come along other roads, and which, though 
far from identical, still overlap each other in many respects. 


APPENDIX. 


While this paper was in the press, attempts were made b 
, y 


TEXT-FIG. 4. 


Part of a section through the blastocyst of Manis. The ectoderm 
(ec.) and endoderm (en.) of the embryonic knob are transversely 
cut. The trophoblast cells appear darker in this figure. 


me to ascertain whether other genera of mammals might 
perhaps exist which furnish evidence concerning the early 
phases of the trophoblast that might further corroborate the 
facts such as they are presented by Galeopithecus. I was all 
the more anxious to obtain information concerning the earliest 
stages of the scaled ant-eater (Manis), as, by a regrettable 
lapsus calami, which disfigures both the English and the 
German version of my “early ontogenetic phenomena in 
mammals, ete.,” a gastrula stage of Manis is erroneously 
attributed to Galeopithecus. 

Tt is fig. 18 on Pl. C, in vol. 53. of the ‘Quart. Journ, of 
Mier. Sci.,’ and fig. 46 in the German publication. I here 


—— 


IS THE TROPHOBLAST OF HYPOBLASTIC ORIGIN P 593 


reproduce the misnamed figure of this early Manis, and 
have since had the good fortune of obtaining sections of yet 
earlier cleavage stages of the same animal. 

Sections of early blastocysts of two specimens of Manis are 


TExT-FIGS. 54 AND B. 


Fig.5a.- 


Two consecutive sections of very early blastocysts of Manis, which 


show what is presumably the earliest trophoblastic covering of 
the mother-cells of the embryonic knob. 


reproduced here in 'l'ext-figs. 5 a,b, and 6. The stage of 'ext- 
fig.5 is presumnably a two-cell, the other (as far as I can follow 
it up in the consecutive sections of the series) a four-cell 
cleavage stage (purposely but incorrectly not counting the 


TEXT-FIG. 6. 


Fig.6. 


Another section through another blastocyst of the same genus 
in the same stage. 


trophoblast cellsas such). In both the differentiation between 
the mother cells of the embryonic knob on the one hand, and 
the already so much more numerous trophoblast cells, leads 
to the inevitable conclusion that the phenomenon of the 
separation of the larval trophoblast from the remaining 


594, A. A. W. HUBRECHT, 


embryonic cells takes place at quite as early a moment as we 
have above described it for Galeopithecus, and that’ also 
in Manis it is perfectly excluded to look upon the tropho- 
blast cells as hypoblastic. And so the early Manis may be 
joined to the early Galeopithecus as fatal to Assheton’s 
interpretation of the trophoblast. I cannot yet say with 
certainty, but I have reason to believe that also in the very 
young hedgehog similar peculiarities occur, 

At the same time it is very suggestive that the quaint and 
aberrant mode in which the trophoblast cells of Galeopithecus 
and Manis arise offers so many points of mutual resemblance 
between these two genera, and differ not inconsiderably from 
what we find in Primates, Rodents and Carnivores. 

Later investigations will have to decide whether the 
phenomenon, as it presents itself in Galeopithecus and Manis, 
is one of precocious segregation, 


FIBROUS TISSUE PRODUCED AS A REACTION TO INJURY. 595 


The Origin and Formation of Fibrous Tissue 
Produced as a Reaction to Injury in Pecten 
Maximus, as a type of the Lamellibranchiata. 


By 
G H. Drew, B.A., 
Beit Memorial Research Fellow; and 


W. De Morgan, F.Z.S. 


With Plate 24. 


ConrENTs. 


PAGE 

Introduction ; : : ; F . O95 
Methods : . 596 
Description of the Tissues of Pee ber maximus inv oly ed in the 

Experiments, and the Normal Process of the “ Clotting” of 

the Blood. . 598 
The Formation of Fibrous eau at fie site of the Truplantatinn 

of a Mass of Gill-Tissue : . 600 
The Formation of Fibrous Tissue at the site of the Implantation 

of Digestive Gland Cells : . 604 
The Reaction of the Tissues to the Taplantation of a Mass of 

Sterile Agar Jelly : : ‘ , . 606 
Summary of Results : ; ; : . 608 
References ; : : ; . 609 

INTRODUCTION. 


‘HE experiments described in this paper were performed 
on Pecten maximus at the Laboratory of the Marine 
Biological Association at Plymouth. 

The object of our work was to investigate the histology of 
the reaction of the tissues to the presence of a foreign body, 
and to determine the origin and method of formation of the 
fibrous tissue formed around it. 


596 G. H. DREW AND W. DE MORGAN. 


As one type of foreign body we chose sterile agar jelly, 
which has little or no irritative or toxic action on the tissues, 
and is not removed by phagocytosis. As another type we 
chose masses of gill-tissue and of the tissue of the digestive 
gland, taken from an animal of the same species. Neither 
of these could be injected under aseptic conditions, and both 
were capable of removal by phagocytosis. Considerable 
irritation was set up by the implantation of these tissues, 
especially in the case of the digestive gland. This produced 
marked degeneration of the neighbouring tissues, possibly 
owing to the liberation of ferments and consequent digestive 
action. 

Pecten maximus was selected for these experiments on 
account of the large size of its adductor muscle, which 
presents a homogeneous mass of tissue particularly suitable 
as a site for implantation of foreign bodies. Before making 
this choice, experiments were tried on several other animals, 
but it was found that in most cases the technical difficulties 
encountered in endeavouring to make implantations into 
small masses of tissue, and in determining the exact relation 
of the underlying organs to the superficial anatomy, were 
too great to render these animals suitable subjects for 
experiment. 

Such experiments were tried on Carcinus menas, 
Pagurus bernhardus, and others of the smaller species of 
crabs, on Palemon serratus, Ligea oceanica, Aphro- 
dite aculeata, Patella vulgata, Aplysia punctata, 
Archidoris tuberculata, and many Lamellibranchs, but 
none Offered such promise of success as Pecten maximus. 


Meruops. 


Pecten maximus can be readily obtained in the Saleombe 
Estuary. It was found necessary to allow these animals to 
become acclimatised to living in the laboratory tanks before 
proceeding to the experimental work. When first placed in 
the tanks the mortality was heavy, often amounting to 30 


FIBROUS TISSUE PRODUCED AS A REACTION TO INJURY. 597 


per cent. in the first three days, but after the lapse of about 
a week the survivors appeared to be fully acclimatised to the 
changed conditions, and often remained healthy tor some 
mouths. 

Experiments on animals whose health was doubtful were 
of no value, both because the shock consequent on the 
injection of the foreign body frequently caused death, and 
also because the reaction of the tissues was not normal in 
unhealthy specimens. When a Pecten is healthy it lies 
with the valves of the shell shghtly apart, the tentacles are 
expanded, and it responds rapidly to any stimulus by closing 
the shell; when held up in the air, the water which drains 
away is Clear and contains no slime. An unhealthy specimen 
lies with the valves of the shell wide open, there is little or 
no response to stimuli, and the valves only close under 
pressure. ‘lhe tentacles are retracted, and the gonads, gulls, 
and tissues generally, look flabby and unhealthy. ‘lhe water 
which flows out between the valves is shmy and viscid, and 
this is generally the first sign of deterioration. 

All instruments used in the experiments were carefully 
sterilised in boiling water. 

The transplanting needle resembles a large hypodermic 
needle about 1 mm. in diameter and 6 cm. long. Into the 
hollow needle a somewhat longer stylet fits closely and works 
like a piston. Any material taken up in the point of the 
needle is sucked in by drawing the stylet back, and again 
ejected by pushing it forward. 

For injecting into the muscle, a solution of agar in sea- 
water, coloured by a little hamatein, was used. The agar jelly 
was liquefied by heating in boiling water, and was drawn up 
into the transplantation needle. On cooling it forms a 
cylinder, of the diameter of the needle, which is easily intro- 
duced into the muscle. 

The adductor muscle of Pecten maximus 1s so large that 
there is no difficuly in selecting a spot at which to bore the 
shell. The apex of an equilateral triangle, having for its 
base the line of junction of the posterior auricula with the 

VOL. 55, PART 3.—NEW SERIES. 39 


598 G. H. DREW AND W. DE MORGAN. 


right valve, marks roughly on the surface a point at which 
the shell may be bored without damage to any organ. But 
as the animal gapes when removed from its tank, it is easy 
to slip a cork between the valves and select a spot by 
inspection. 

The holes were drilled in the convex or right valve by an 
ordinary dentist’s drill, the head of which was prevented from 
penetrating too deep by a lapping of thread. 

The spot selected for drilling was sterilised with a saturated 
solution of corrosive sublimate, washed off with a solution of 
hydrogen peroxide (30 vols.) or distilled water, care being 
taken not to allow any of the sublimate to run between the 
valves. ‘he transplanting needle was then introduced to the 
required depth, slightly withdrawn, and its charge projected 
into the channel. The hole was then thoroughly dried, and 
stopped with sealing-wax. If the drying is thorough the wax 
will adhere after the animal has been returned to the tank. 
It would, of course, have been possible to implant directly 
into the muscle through the opening of the valves, but the 
risks of sepsis would have been greater. 

When required for examination, the shell was opened by 
cutting the adductor muscle at its attachment to the right or 
convex valve, and a portion of the muscle containing the 
implanted material removed. ‘This was fixed by three or four 
hours’ immersion in Gilson’s fluid, then thoroughly washed, 
passed through the alcohols, cleaned in xylol, and embedded 
in paraffin wax. It was then cut into serial sections eight 
pe thick. 

Delafield’s hematoxylin, followed by Van Gieson’s stain, or 
Benda’s iron mordant and hematoxylin were used for staining. 


DescrieTioN OF THE TISSUES OF PECTEN MAXIMUS INVOLVED 
IN THE EXPERIMENTS, AND THE NorMAL PROCESS OF THE 
“‘CLoTTING”’? OF THE BLOoD. 


The adductor muscle of Pecten maximus consists of 
two portions, bound together by the same sheath of connec- 


FIBROUS TISSUE PRODUCED AS A REACTION TO INJURY. 599 


tive tissue, but differing in structure. The larger, semi-trans- 
parent and whitish, consists of striated fibres. The fibres of 
the smaller, which is opaque and dead white, and lies against 
the posterior surface of the larger mass, are non-striated. It 
was into the larger mass that all material in our experiments 
was introduced. 

There is a large blood supply to the muscle from the 
adductor artery (Dakin, 2), and it contains numerous lacunar 
spaces. Scattered through if are numerous strands of con- 
nective tissue. These contain fibroblasts with deep staining 
nuclei and long fibrillar processes. 

The digestive gland has a tubular structure and com- 
pletely surrounds the stomach, mto which its ducts open. 
The ducts break up into numerous alveoh, which ramify and 
ultimately form cca. The ducts are lined with ciliated 
epithelium, and the alveoli with secreting cells. These secre- 
ting cells are said to degenerate and become filled with a 
eranular pigment, and are ultimately shed into the lumen of 
the ducts (Dakin, 2). Thus in their younger stages they 
appear to have a secretory, and in their later stages an excre- 
tory function. In addition to these glandular cells, fibrous 
connective tissue and unstriated muscle-fibre are present. 
The ducts contain particles of food material, alge, diatoms, 
and bacteria, and consequently as a rule septic conditions 
prevail in the experiments. 

The blood of Pecten maximus is a slightly cloudy, 
colourless fluid. It does not coagulate, but when shaken a 
number of small, white, floccular masses appear, which soon 
fall to the bottom of the tube, leaving the supernatant fluid 
clear and transparent. ‘hese masses consist of blood-cor- 
puscles agglutinated to form plasmodia. 

The corpuscles, although varying in size, appear to be only 
of one kind. ‘They are amceboid bodies, which when expanded 
protrude a number of slender pseudopodia. When contracted, 
they are ovoid or spherical. There is a single compact 
nucleus, staining readily with methylene-blue. The cyto- 
plasm is finely granular, and stains with eosin, but there are 

39 § 


600 G. H. DREW AND W. DE MORGAN. 


no large eosinophile granules. According to Cuénot (1), they 
originate in a ‘ glande lymphatique ” situated at the base of 
the gills. 

One of us (Drew, 4) has shown in the case of Cardinm 
norvegicum that when the corpuscles come in contact with 
a rough foreign body, or with injured tissue, they possess the 
power of agelutinating and forming a compact plasmodial 
mass. In this way bleeding from a small wound is stopped. 
When the edges of a wound are covered with this mass of 
agolutinated corpuscles, protoplasmic strands are formed 
across the wound, connecting the plasmodia; these strands 
thicken and contract and so approximate the edges of the 
wound. So far as our observations go, there is no reason to 
suppose that the blood of Pecten maximus differs in any 
of these particulars from that of Cardium norvegicum. 

That Lamellibranch blood-corpuscles are capable of a 
phagocytic action towards degenerated cells has been shown 
by De Bruyne (8) in the case of Mytilus edulis, Ostrea 
edulis, Unio pictorum, and Anodonta cygnea. Sir 
Ray Lankester (5 and 6) has shown that certain corpuscles of 
Ostrea edulis have a phagocytic action on diatoms and 
minute green alge, and it has been shown by Drew (4) that 
the corpuscles of Cardium norvegicum have a phagocytic 
action on bacteria, and are attracted towards extracts of dead 


tissues. 


Tae Formation or Frprous Tisstk IN THE SITE OF THE 
IMPLANTATION OF A Mass or GiLL-TIssuE. 


As bacteria are normally present on the gill-filaments, the 
conditions when gill-tissue is implanted differ totally from 
those obtaining when sterile agar is used. 

The implantation soon produces an intense inflammatory 
reaction on the part of the animal. The blood-spaces in the 
immediate neighbonrhood of the implanted tissue become 
distended and crowded with corpuscles, which escape from 
the lacunar spaces and migrate towards the source of irrita- 


FIBROUS TISSUE PRODUCED AS A REACTION TO INJURY. 601 


tion, travelling in all directions between the muscle-fibres. 
On reaching the gill-tissue the corpuscles come to rest, and 
form a dense, agglutinated, plasmodial mass, completely sur- 
rounding and shutting off the gill-tissue from the neighbour- 
ing muscle (fig. 1). They soon appear as if they had under- 
gone some degree of pressure and the nuclei are slightly 
flattened, probably owing to the contraction of the plasmodial 
mass as it tightens round the implanted gill-tissue (Drew, 4). 
In time the corpuscles show signs of degeneration; the 
nuclei become irregular in outline, and the chromatin is 
represented by numerous granules staining darkly with 
hematoxylin. The degenerated mass of corpuscles is then 
invaded by fresh blood-cells, and is more or less completely 
removed, apparently partly by a process of phagocytosis and 
partly by autolysis. 

While this is going on, the cells of the gill-filaments have 
degenerated, their outlines are ill-defined, and the nuclei no 
longer discernible ; the bacteria present multiply consider- 
ably. 

The degenerated gill-tissue is then invaded by blood-cor- 
puscles which have penetrated through the surrounding mass 
of agglutinated cells, and in most cases the bacteria and 
epithelial débris are removed by phagocytosis, leaving only 
the chitinous supporting-rods of the gills. 

In the course of this process many of the invading cells 
also are destroyed, and appear in their turn to be removed by 
other phagocytes. In time the whole space originally occupied 
by the gill-tissue becomes filled with a loosely packed mass 
of blood-cells, among which the chitinous supporting bars are 
the only relics of the original implanted mass. In many of 
our experiments bacteria multiplied so rapidly that the phago- 
cytes were unable to cope with them. Consequently the 
bacteria invaded the neighbouring tissues, entered the blood- 
spaces, and rapidly caused death. 

In preparations from obviously unhealthy animals, it was 
commonly found that the bacteria had penetrated beyond the 
protecting mass of agglutinated cells and had invaded the 


602 G. H. DREW AND W. DE MORGAN. 


muscular tissue, which showed signs of degeneration in its 
somewhat swollen fibres and faint striation. 

When a blood-space had been entered, bacteria were often 
seen ingested by the blood-corpuscles, but in later stages it 
was obvious that the number of bacteria was so out of pro- 
portion to the number of corpuscles that they could not all 
be removed by phagocytosis, and were of necessity distributed 
all over the body in the blood-stream. 

During these processes the fibroblasts in the walls of the 
blood-spaces, and in the intermuscular connective tissue in the 
neighbourhcod of the implanted mass, undergo rapid division. 
This rapid division, resulting from the reaction of the tissues 
to the irritation caused by implantation, appears to be always 
amitotic. Mitotic division was only observed in much later 
stages, when the source of irritation had been removed by 
phagocytosis, and the rate of division of the fibroblasts was 
much slower. 

Before amitotic division the fibroblasts lose their spindle 
shape and become oval; a split then appears at one end, and 
progresses in the plane of the long axis of the nucleus until 
two daughter nuclei are formed, attached to each other at one 
extremity, and inclined at an acute angle to one another. 
These gradually straighten out until they form an hour-glass- 
shaped mass of nuclear material. Finally the two nuclei are 
separated at the constriction and become almost circular in 
shape. 

As a result of this active multiplication of the fibroblasts, 
the strands of connective tissue bounding the blood-spaces 
and forming the intermuscular connective tissue become 
crowded with nuclei. The bodies of the fibroblast cells 
become very indistinct, and little beyond rows of elongated 
nuclei is discermble. As the multiplication becomes more 
‘apid the typical spindle shape of the nuclei is lost, and they 
become first oval and finally circular. 

There appears to be a constant migration of these cells, 
with round and oval nuclei, towards the site of implantation. 
They have very little cytoplasm, and from this, and their 


FIBROUS TISSUE PRODUCED AS A REACTION TO INJURY. 603 


smaller size, are easily distinguished from the blood-corpuscles 
(figs. 2and 3). These fibroblasts largely follow the course of 
the strands of fibrous tissue bounding the blood-spaces, and 
they appear to travel along in the spaces, being most plentiful 
near the walls. At the same time, when they multiply very 
rapidly, many migrate in ali directions between the muscular 
fibres towards the implanted tissue, and are not confined to 
travelling only in the proximity of pre-existing connective- 
tissue strands. 

Onreaching the degenerating layer of agglutinated corpuscles 
surrounding the implanted tissue, they arrange themselves in 
rows, and their nuclei elongate in such a direction that their 
long axes form arcs of a circle surrounding the implanted 
tissue. Some fibroblasts penetrate among the degenerating 
cells of the gill-tissue, which are being removed by phago- 
cytes, and in this position start the formation of fibrous 
tissue. 

The surrounding layer of fibroblasts gradually thickens, 
and presents a somewhat stratified appearance. At first this 
layer contains a number of blood-corpuscles, but these even- 
tually are removed, probably by autolysis, leaving only the 
fibroblasts, which can now be seen to be connected with each 
other by numerous fine processes of the cytoplasm, the whole 
presenting a somewhat reticulated appearance. In time this 
tissue becomes more compact, and the reticulation vanishes. 
It would appear that this has been caused by the contraction 
of the processes of the fibroblasts, with consequent approxi- 
mation of the cells. Finally, the nuclei become long and 
spindle-shaped, the amount of cytoplasm slightly increases, 
and a layer resembling normal fibrous tissue results. 

In our experiments the great variation in the rapidity with 
which the various changes described took place was very 
noticeable. The health of the animal after the experiment 
seems an Important factor m accounting for this, for the slow 
rate of fibrous tissue formation in unhealthy, as compared 
with healthy animals, was very marked. 

Unfortunately none of the animals into which gill-tissue 


604. G. H. DREW AND W. DE MORGAN. 


was implanted lived long enough for all the elements of the 
gill-tissue to be completely replaced by fibrous tissue, but in 
healthy specimens most of the signs of inflammation had 
vanished, and the implanted tissue was surrounded by a wall 
of apparently healthy fibrous tissue, in four or five days. 


Formation or Frprous Tissuz AROUND HE SITE OF IMPLANTED 
Digestive GLAND CELLs. 


After the implantation of portious of the digestive gland, 
a marked degeneration of the muscular fibres in its neigh- 
bourhood is noticeable.  ‘l'hey swell slightly, all trace of 
striation is soon lost, and they stain less intensely. The area 
of degeneration gradually extends, and the muscular fibres 
in the immediate neighbourhood of the gland tissue are slowly 
dissolved. ‘This action is presumably due to the presence of 
ferments in the digestive gland, which digest and render 
soluble all tissues in the immediate neighbourhood. At the 
same time the cells of the gland itself degenerate aud appear 
to undergo auto-digestion, so that eventually only the brown 
pigment-granules originally contained within the secreting 
cells remain. Under these conditions bacteria do not seem 
to multiply, though they must have access to the ceeca of the 
digestive gland, as these are in direct communication with 
the alimentary canal. In none of our sections have we been 
able to find bacteria, though it is quite common to find the 
siliceous skeletons of diatoms in the ceca. It seems, there- 
fore, probable that the presence of digestive ferments inhibits 
the multiplication of bacteria. 

Asa result of the implantation of this tissue a condition 
of intense inflammation is set up, and all the blood-spaces in 
the neighbourhood become distended with blood-corpuscles. 
‘here appears to be an endeavour on the part of the orga- 
nism to shut off all the implanted gland, together with the 
area of muscular tissue which has undergone degeneration, 
from the general blood-stream. ‘This is effected by the 
formation of a layer of agglutinated blood-corpuscles around 


FIBROUS TISSUE PRODUCED AS A REACTION TO INJURY. 605 


the whole of the affected area (fig. 4). It was very notice- 
able in our preparations that the degenerated area was 
always larger in specimens that had been implanted with the 
digestive gland for some time (up to six days), than in those 
implanted for shorter periods, and thus it would seem that 
the range of action of the digestive ferments gradually 
increases. The degenerated area was always found sur- 
rounded by a layer of agglutinated corpuscles, though in 
different specimens this layer varied considerably in thickness. 
It would seem that while the degenerative process is spread- 
ing the layers of corpuscles must be continually dissolved, 
and others formed a little further back by the spread of the 
digestive ferments. During this process the fibroblasts 
undergo division as in the case of the gill-tissue, but while 
the inflammation is much more acute, the multiplication of 
fibroblasts is not so rapid, and they are not nearly so notice- 
able a feature in the sections. In the form of rounded cells, 
with oval or spherical nuclei, they migrate in small numbers 
towards the layer of agglutinated blood-corpuscles. Here 
they share the fate of the corpuscles, being dissolved by the 
digestive ferments, and accordingly there is no formation of 
fibrous tissue. 

We were never able to keep the animals alive for more 
than six days. At the end of this time all that remained of 
the digestive gland was the brown pigment-granules and a 
little epithelial débris. This was surrounded by a space 
from which most of the muscular tissue had been dissolved, 
aud this again by a relatively large area of degenerated 
muscle-fibres. Finally, the whole was surrounded by a layer 
of agglutinated blood-corpuscles, into which a few fibro- 
blasts were making their way. 

These experiments show that the protective layer of cor- 
puscles must very completely shut off the space it encloses from 
the neighbouring tissue. If this were not the case the digestive 
ferments, once they had gained access to the blood, would 
rapidly become disseminated over the whole body. Instead 
of this, we have distinct evidence that there is a slow and 


606 G. H. DREW AND W. DE MORGAN. 


steady invasion of the tissues by the ferments, and that the 
area of their action is always contained within a protective 
layer of agglutinated blood-corpuscles. It seems probable 
that the digestive gland, when implanted, contains little or 
no free enzyme, and quickly becomes surrounded by the 
protective layer of corpuscles, and that later the enzymes 
are slowly evolved from the zymogens contained within the 
cell. The vitality of these cells has been impaired by removal 
from their normal connections and by implantation into the 
muscle tissue, and accordingly they are dissolved by the 
enzymes they have themselves evolved. 


Tae Reaction or tHe Tissurs to ImMeLANTED AGAR JELLY. 


Sterile agar jelly has no irritative action on the muscle, 
and so differs from the tissues previously described. 

Agar jelly may be regarded as a physiologically inert 
substance, and as in these experiments it was made from sea- 
water in which the Pecten were living, it was approximately 
of the same salinity as their blood (Dakin, 2), and so was of 
the same osmotic concentration. Further, the cylindrical 
rods of agar are remarkably smooth, and if unbroken present 
no rough surface, except possibly at the extremities. 

One of us (Drew, 4) has shown that in the case of 
Cardium norvegicum, the agglutination of the blood- 
corpuscles (in vitro) is much influenced by the nature of the 
substance on which they impinge, and that it occurs very 
much more readily when they come in contact with a rough 
surface from which a large number of small points may be 
imagined to project, than when they impinge on a smooth, 
polished body. It seems probable that similar conditions 
obtain in the case of the blood of Pecten maximus. 

In accordance with these properties of the agar jelly, it 
was found that absolutely no inflammation resulted from its 
implantation in the muscle. No layer of agglutinated 
corpuscles was formed round it, and there was no sign of the 
collection of unusual numbers of the corpuscles in the 


FIBROUS TISSUE PRODUCED AS A REACTION TO INJURY. 607 


vicinity, nor of any distension of the blood-spaces. ‘The fact 
that the rod of jelly was always implanted as far as possible 
parallel to the long axes of the muscle-fibres, and that they 
were usually rather separated from each other, than cut by 
the insertion of the transplanting needle, probably contributed 
towards this result. 

After a period of about seven to eight days there were 
signs of division of the fibroblasts in the neighbourhood of 
the implanted mass, and a slow migration of the new-formed 
cells towards the agar took place. By about the tenth day 
these cells had arranged themselves so as to form a thin and 
delicate ensheathing layer. The process presents marked 
differences from that which occurs after the implantation of 
a substance which causes an inflammatory reaction, with the 
consequent development of a protecting layer of agglutinated 
corpuscles. he division of the fibroblasts, instead of being 
rapid and amitotic, is comparatively slow, and frequently, 
though not always, mitotic. The nuclei of the young fibro- 
blasts retain their elongated shape, and though the nuclei of 
the dividing cells lose their typical spindle-like appearance 
and become oval, they do not become round, as in the case of 
rapid division after inflammation. The layer of fibrous tissue 
formed is thinner and less compact, the proportion of cyto- 
plasm to nucleoplasm is greater, and the nuclei assume their 
typical spindle shape more rapidly. ‘The process seems to be 
complete by the tenth day, and the appearance is almost 
identical with that shown in fig. 5, which represents the 
condition after seventeen days. 

In some of our experiments the sealing-wax with which the 
drill holes were closed became detached in the tank. ‘The 
holes were re-sealed as soon as this was noticed, but the 
animals seldom survived long. On sectioning, an area of 
inflammation was usually found surrounding the agar, and 
rapid division of the fibroblasts in the vicinity was in pro- 
gress. In specimens that survived longer a complete sheath 
of fibrous tissue had formed round the agar, and the con- 
dition resembled that resulting from implantation of gill- 


608 G. H. DREW AND W. DIS MORGAN. 


tissue. It seems that in these cases bacteria must have 
entered through the drill-hole, and, travelling between the 
agar and muscle, have caused an inflammatory reaction. In 
one other case, in which the hole had not come unsealed, in- 
flammation and formation of fibrous tissue occurred, but as 
this ouly took place once out of twenty-six implantations made 
with sterile agar, it is probable either that the sealing-wax 
plug leaked at the edges or that bacteria found their way in 
when the agar was introduced. 


SumMMARY or Resutts. 


Our experiments show that the implantation of a tissue, 
such as that forming the gills, accompanied by the bacteria 
which adhere to it, produces an intense inflammatory reaction. 
This is characterised by the active migration of blood- 
corpuscles, which form a plasmodial mass around the im- 
planted tissue, shutting it off from the general circulation. 
This protective layer is gradually removed by phagocytosis 
and autolysis, and at the same time the gill-tissue is invaded 
and removed by phagocytes. - While this is going on, rapid 
amitotic division of the fibroblasts in the neighbourhood 
occurs ; they lose the typical spindle-shape of the nuclei, and 
the new-formed cells consist of rounded or oval nuclei, with a 
scarcely perceptible amount of cytoplasm. ‘These rounded 
cells migrate towards the implanted tissue, aud arrange 
themselves in layers around it, the nuclei become elongated, 
and the proportion of cytoplasm increases. Finally, a layer of 
typical “scar ” fibrous tissue is formed, enclosing the chitinous 
skeletons of the gill-bars. 

In the case of the implantation of digestive gland tissue a 
similar protective layer of agglutinated corpuscles is formed, 
but this is continually dissolved up and reformed, as the 
sphere of action of the enzymes in the cells of the digestive 
gland extends, All the muscle-fibres within this protective 
layer soon lose their striation, swell, and are partially dis- 
solved, presumably by the digestive enzymes. The fact that 


FIBROUS TISSUE PRODUCED AS A REACTION 'TO INJURY. 609 


there is a progressive extension of this digestive action shows 
that the layer of agglutinated corpuscles performs its pro- 
tective function very completely, as otherwise the enzymes 
would escape into the general circulation. Simultaneously 
the fibroblasts in the vicinity multiply and migrate, as in the 
case of implanted gill-tissue, but the multiplication does not 
seem to be so rapid. No permanent layer of fibrous tissue is 
formed, as the migrated fibroblasts are dissolved in the course 
of the extension of the sphere of action of the digestive 
ferments. 

In the case of the implantation of sterile agar jelly, made 
with sea-water, no inflammation results, and for some time 
there is no sign of any reaction of the tissues 1f absolute 
asepsis has been ensured. After seven or eight days there 
is a Slow and often mitotic division of the neighbouring fibro- 
blasts; they migrate and rearrange themselves to form a 
thin layer of fibrous tissue around the agar. 

It is noteworthy that though the tissues and the blood, 
especially in its manner of forming a “ clot,” present marked 
differences from those in Vertebrates, yet the formation of 
fibrous tissue, as a reaction to injury, does not differ in any 
essentials from the process which takes place in the higher 


types. 


REFERENCES. 

1. Cuénot, L.—* Etudes sur le Sang et les Glandes Lymphatiques,” 
‘Arch. de Zool. Expér. et Gen.,’ Deuxieme serie, tome ix, Paris, 
1891. 

2. Dakin, W. J.—* Pecten,” ‘ Liverpool Marine Biological Comunittee 
Memoirs,’ xvii, London, 1909. 

3. De Bruyne, C.—* Contribution A l'étude de la Phagocytose (1),” 
‘Arch. de Biol.,’ tome xiv, Paris, 1896, p. 161. 

4. Drew, G. H.—*Some Points in the Physiology of Lamellibranch 
Blood-Corpuscles,” ‘Quart. Journ. Mier. Sci., vol. 54, part 4, 
February, 1910, p. 605. 

5. Lankester, Sir E. Ray.—* On Green Oysters,” ‘Quart. Journ. Micr. 
Sci.,’ vol. 26, 1886, p. 71. 

6. ——— “Phagocytes of Green Oysters,” ‘Nature,’ vol. xlviii, 1893, 


Ley) 


VOL. 59, PART 3.—NEW SERIES. 4.0 


610 ‘ H. DREW AND W. DE MORGAN. 


DESCRIPTION OF PLATE 24, 


Illustrating the paper by Messrs. G. H. Drew and W. de 
Morgan on “ ‘The Origin and Formation of Fibrous 
‘Tissue produced as a Reaction to Injury in Pecten 
maximus, as a type of the Lamellib ‘anchiata.” 


REFERENCE LETTERS. 
ag. Agar. agg.lyr. Agglutinated layer of blood-corpuseles. — .c. 
Blood-corpuscles. deg. gill. Degenerated gill-tissue. deg. msl. Degene- 
rated muscle. dig. gl. Digestive gland-tissue. div. fbl. Dividing fibro- 
blasts. fbl.lyr. Fibroblast layer. mig. fl. Migrating fibroblasts. 
msl.-fbr. Muscle-fibres. 


[N.B.—In the figures the bundles of muscle-fibres are shown as a 
whole: the individual fibrils and their striations are not differentiated. 
The size of the muscle-hundles differs considerably in different parts of 
the adductor muscle. | 


Fig. 1—x 400, Gill-tissue which has been implanted for sixteen 
hours. A layer of agglutinated corpuscles divides the degenerated gill- 
tissue on the left from the muscular tissue on the right. Corpuscles 
ure making their way between the muscle-fibres to join the agglutinated 
layer. 

Fig. 2.— x 300. A later stage of fig. 1, tuken seventy-two hours after 
implantation. A definite layer of fibrous tissue has been formed round 
the gill-tissue, which is completely degenerated and invaded by phago- 
cytes. The fibroblasts are dividing and migrating towards the lesion. 

Fig. 3—x 700, A more highly magnified portion of one of the 
blood-spaces drawn from the same section as fig. 2. The fibroblasts 
are undergoing amitotic division, and migrating towards the gill-tissue, 
where they arrange themselves to form a layer of fibrous tissue. 

Fig. 4—x 450. Digestive gland-cells (on the left) which have been 
implanted for ninety-six hours. External to them is a region of 
degenerated and partially dissolved muscle-fibres, which is divided from 
the normal muscle by a thin layer of agglutinated corpuscles. These 
are also rapidly degenerating, but are reinforced by the continued 
arrival of fresh corpuscles. The cellular structure of the alveoli of the 
digestive gland has been lost, leaving little beyond traces of the original 
cell walls and the brown pigment-granules. 

Fig. 5—x 450. Agar jelly (to the left) which has been implanted 
for seventeen days. It is divided from the muscle-tissue by a delicate 
layer of fibroblasts. 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 611 


The Division of the Collar-Cells of Clathrina 
coriacea (Montagu): A Contribution to the 
Theory of the Centrosome and Blepharo- 
plast. 


By 
Muriel Robertson, M.A., and E. A. Minchin, M.A. 


With Plates 25 and 26. 


INTRODUCTORY. 


Ar the present time there is a great deal of confusion in the 
use of the words “ blepharoplast” and “‘ centrosome.” Two 
distinct questions arise with regard to the significance of these 
bodies; the first is the question of the homology of blepharo- 
plasts and centrosomes; the second is that of the nature of the 
centrosome, and more particularly whether or not it is to be 
regarded as equivalent primarily to a nucleus. 

With regard to the first of these questions, it is now 
generally admitted that blepharoplasts and centrosomes are 
essentially bodies of the same nature, for reasons that will 
presently be considered at greater length. The difference 
between a centrosome and a blepharoplast, on this view, is 
entirely a matter of divergence of function. A centrosome 
may be briefly characterised, in a general way, as a body 
which exerts or governs kinetic functions in relation to the 
division of the nucleus; a blepharoplast may be defined as 
a centrosome which governs the movements of motile 
organs, such as flagella, which arise from it and are in 
direct or indirect connection with it. 

With regard to the second of these questions, namely, the 
nature of the centrosome, two opposite views are current, 

VOL. 55, PART 4.—NEW SERIES. 4] 


612 MURIEL ROBERTSON AND E. A. MINCHIN. 


which may be summarised as follows: (1) The centrosome 
is to be regarded as primarily a body of achromatic! nature, 
elaborated and evolved, in all probability, in the nucleus 
or in connection with it, but not itself equivalent to a 
nucleus ; (2) the centrosome is regarded as the equivalent 
of a nucleus, and as representing primarily a nucleus which 
has become modified and specialised both in function and 
structure. These two theories may be termed conveniently 
the achromatic and the nuclear theory of the centrosome 
respectively. According to the second of these views, which 
has recently been revived and advocated by Hartmann and 
Prowazek (6), every cell is to be regarded as primarily and 
essentially binucleate; the two nuclei, at first, doubtless, 
equivalent and similar in all respects, became modified in two 
directions respectively, the one becoming specialised for 
trophic, the other for kinetic functions, with corresponding 
differentiation of structure. In the metazoan cell, aecording 
to this theory, the nucleus represents the original trophic 
nucleus deprived of all kinetic structures, while the centro- 
some represents the kinetic nucleus deprived of all “ vegeta- 
tive” functions and of its chromatic apparatus. On this 
interpretation of the centrosome, the minute granules which 
are the centre of kinetic functions are termed ‘‘centrioles,”’ in 
order to distinguish them from the centrosomes as a whole. 
In fact, from the point of view of the nuclear theory of the 
centrosome, the centriole requires to be defined in exactly the 
same way as the centrosome itself on the achromatic theory. 

The confusion produced by these two theories of the cen- 
trosome reaches its height in the nomenclature of the different 
parts of the body of a trypanosome. In these organisms, and 
in allied genera of flagellates, there are three distincts parts 


of the nuclear apparatus to be reckoned with. First, there 

! Meaning by the term “achromatic ” something which is not com- 
posed of chromatin, not necessarily something which is not coloured 
by stains. All is not chromatin that stains, even with a so-called 
nuclear stain. In our opinion a great deal of error and misconception 
has arisen by identifying as “ chromatin ” all bodies in the cell that are 
coloured black, for instance, by the iron-hematoxylin method. 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORTIACEA. 613 


is achromatic body, which may be denoted temporarily by 
the symbol N, situated usually in or near the middle of the 
cell-body, and in no special connection with the flagellar 
apparatus. Secondly, there is a second chromatic body, 
which may be denoted by the symbol m, distinctly con- 
nected with the flagellum or flagella, when they are present, 
and apparently kinetic in function. In the genera Try- 
panosoma, Herpetomonas, Leishmania, and Cri- 
thidia, N is always much larger than n, but in Trypano- 
plasma the reverse may be the case. Finally, the flagella 
arise, probably in all cases, from basal granules, which are 
often very minute and exhibit staining reactions quite 
different from either N or 1. 

According to the nuclear theory of the centrosome advo- 
cated by Hartmann and Prowazek, these three parts of the 
trypanosome body are to be interpreted and named as 
follows: N is the trophic nucleus, while n represents the 
second nucleus of kinetic function, in other words, the cen- 
trosome, which, since it controls the activities of the flagellar 
apparatus, is to be termed a blepharoplast. The basal granule 
is a mere thickening of the proximal end of the flagellum, of 
no special significance, or at most representing a centriole. 
Thus a trypanosome would represent the ideal binucleate cell 
of Hartmann and Prowazek in a very primitive state. 

An interpretation of the trypanosome body, quite different 
to that of Hartmann and Prowazek, has been advocated by 
one of us (12), which may be briefly stated as follows: N is 
a trophic nucleus, which contains its own centrosome or 
division-centre in itself; 2 is a distinct kinetic nucleus, a 
specialisation of the nuclear apparatus for a particular func- 
tion ; it has nothing to do with a true centrosome, though it 
may, like the trophic nucleus, contain a body of this kind, nor 
is it to be regarded as a blepharoplast, a body which is repre- 
sented by the basal granule of the flagellum.! 

' It is not our purpose here to summarise the various views that 


have been put forward with regard to the morphological interpretation 
of the trypanosome-body, but only to select two which show in sharp 


614 MURIEL ROBERTSON AND E. A. MINCHIN. 


In consequence of these divergent theories and interpreta- 
tions, a great confusion in nomenclature has arisen, especially 
with repard to n, which is always termed the blepharoplast 
in German works, the centrosome in French works, aud in 
this country is sometimes named the micronucleus, but more 
usually the kinetonucleus, 

In Flagellata other than the trypanosomes and their allies 
there is usually only one structural element other than the 
principal nucleus (N) to be reckoned with in the nuclear 
apparatus, namely, a deeply staining grain or set of grains, 
from which the flagellum or flagella take origin, and to 
which the name ‘ blepharoplast,” or the synonymous term 
‘“‘diplosome,”’! is commonly applied. ‘The question at once 
arises, How is the arrangement seen in a trypanosome to be 
compared to that of other flagellates, and to which element 
in the nuclear complex of a trypanosome should the blepharo- 
plast of an ordinary simple flagellate be compared? Does it 
represent the basal granule (true blepharoplast, on our view) 
or the kineto-nucleus (x)? In our opinion, the bodies in 
question are true blepharoplasts, comparable to the basal 
granules of the flagella of trypanosomes, and the kineto- 
nucleus or German blepharoplast of the trypanosomes and 
their allies isa nuclear body peculiar to them, and not 
found in ordinary flagellates. To this extent, at least, we 
are in agreement with the idea expressed by Hartmann, 
who has placed the trypanosomes and forms regarded as 


contrast opposed views with regard to the nature of the blepharoplast 
and the proper application of this word. Thus Layeran and Mesnil in 
their well-known work on trypanosomes use the term * centrosome” for 
n; so also Moore and Breinl, who contrast the extra-nuclear centrosome 
(n) with the intra-nuclear centrosome (karyosome of 1). 

1 “The term “diplosome,’ meaning literally and etymologically a 
double body, is commonly applied, by an abuse of language, to the 
single grain from which a flagellum arises. It should, of course, be 
used only for those cases where twin granules give origin to two or 
more flagella, that is to say it should not be regarded as synonymous 
with blepharoplast or basal granule, but as implying a condition in 
which such bodies are doubled. 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 616 


allied to them in a separate order of Flagellata termed the 
Binucleata. (The question as to whether or not the Heemo- 
sporidia should be included in the Binucleata is one which, in 
the present memoir, we do not wish to raise or discuss.) A 
trypanosome is, in our opinion, a_ binucleate organism, 
possessing a trophic nucleus (N), a kinetic nucleus (nz), and 
a blepharoplast (basal granule). 

In order to settle these disputed points, more knowledge is 
required regarding nuclear and other structures connected 
with the locomotor apparatus in other organisms, and in the 
hope of throwing some light on these questions we have 
studied the division of the collar-cells of a caleareous sponge, 
of which preserved material was in the possession of one of 
us. <A collar-cell, although occurring as tissue-element of a 
Metazoan organism, is essentially a flagellate organism, com- 
parable in every way with an individual of the Choano-flagel- 
lata. It has recently been pointed out by one of us (18) 
that there are two types of collar-cells in calcareous sponges. 
In one type, characteristic of the family Clathrinide, 
amongst Ascons, the nucleus lies at the base of the cell, far 
removed from the origin of the flagellum, which arises from a 
distinct basal granule or blepharoplast situated at the apex of 
the cell. In the other type, characteristic of the Leucoso- 
leniidz amongst Ascons, and of the Heteroccela generally, 
the flagellum arises directly from the pear-shaped nucleus, 
which is usually situated in the upper part of the cell, close 
to the point at which the flagellum emerges from the body of 
the cell. These two differences in arrangement are also 
paralleled amongst free-living Flagellates, for instance 
amongst the two genera of Mastigamcoebe described by 
Goldschmidt (4), and there can be no doubt that the con- 
ditions are perfectly comparable in the two cases—that is to 
say, that when the flagellum arises from a basal granule 
distinct from the nucleus, the basal granules are homologous 
structures. As the result of our investigations we have ob- 
tained, as will be apparent in the sequel, evidence of a most 
convincing kind as to the identical nature of centrosomes and 


616 MURIEL ROBERTSON AND E. A. MINCHIN. 


blepharoplasts ; but before proceeding to the detailed account 
of our observations it will be useful to give a brief resumé 
of previous work on this subject. For this we have relied 
chiefly on the excellent summaries given by Wilson (20) and 
Erhard (8). 

The most convincing and abundant evidence of the identical 
nature of blepharoplasts and centrosomes has come from the 
study of spermatogenesis in animals and plants. ‘These re- 
searches have been summarised by Wilson and Erhard, and 
it will be sufficient here to refer specially to the memoirs of 
Henneguy (7) on the spermatogenesis of Bombyx mori, etc., 
and of Belajeff (1) on that of Gymnogramme and Marsilia 
spp. Henneguy found, as we have done, the blepharoplast 
(in this case a diplosome in the true sense of the word) 
acting as a centrosome in the mitosis while still preserving its 
function as a blepharoplast. Similarly Belajeff found that 
the body which acted as a centrosome in the mitosis became 
subsequently the blepharoplast. 

In the case of tissue-cells other than spermatocytes in 
Metazoa, the relation of flagella and cilia to bodies of cen- 
trosomic nature has been studied by Joseph (94), in whose 
memoir will be found very full references to the work of 
others. Joseph’s researches have led him to support very 
definitely the theory of Lenhossek and Henneguy, according 
to which the basal corpuscles of the cilia arise from the cen- 
trosome ; and in his conclusions he states (l.c., p. 71) : “ Viele, 
vielleicht alle eingeisseligen Zellen sind Centralgeisselzellen, 
d. h. ihr Geisselfaden steht in Verbindung mit dem Cen- 
trosom.” Erhard (8) has reviewed the whole question in the 
hight of renewed investigations, and comes to the following 
conclusions: “ Das Diplosom in Flimmerzellen als Teilungs- 
organ wirkt, also ein echtes Centrosom darstellt. . . . Die 
ausserordentliche Seltenheit von Mitosen in Flimmerzellen 
darauf schliessen liisst, dass die Diplosomen in allgemeinen 
eine andre Rolle als die der Teilung auszufiillen haben. 

Zwischen Basalkérpern und Centrosomen kein erlei 
Beziehungen bestehen . . . die Basalkérpern an der 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 617 


Teilung der Metazoenflimmerzellen keinerlei aktiven Anteil 
nehmen, so kann fur diese Zellen die Henneguy-Lenhos- 
seksche Theorie nicht mehr aufrechterhalten werden.” Thus 
while maintaining the centrosomic nature of the diplosome, 
Erhard denies it for the basal granules of the cilia in ciliated 
cells. 

As regards the basal granules of the flagella in Protozoa, 
evidence bearing on their nature is scanty to a disappointing 
degree. ‘he majority of investigators appear to ignore these 
bodies. Schaudinn (17) found in Paramceba the “ Neben- 
kérper” acting as a centrosome in the mitosis; the flagella 
of the swarm-spore appear to arise quite independently of the 
Nebenkérper, a body which, from Schaudinn’s investigations, 
gives the impression of being rather of the nature of a kineto- 
nucleus than of a centrosome (pace Hartmann and Prowazek), 
and which very probably contains its own centrosome (or 
centriole), which acts also as the centrosome of the principal 
nucleus in the mitosis. Prowazek (15) points out that the 
flagellum of Flagellata may arise within the nucleus (“ Kern- 
endogener Ursprung” ) or outside it; in the latter case the 
flagellum may terminate in a ‘‘ diplosome,””? which again may 
be quite free from the nucleus (as in the collar-cells studied 
by us) or may be connected with the nucleus by a “ rhizo- 
plast.” In the nuclear division of Entosiphon, Prowazek 
(16) found a “ basalkérperartige Verdickung” at the origin 
of each flagellum, and from this body a rhizoplast passing 
back tothe nucleus. At the division of the nucleus a ‘ Centro- 
nukleolusspindel”’ is formed. The basal granules do not 
appear to influence the division of the nucleus in any way ; 
they divide, and two new flagella grow out from each pair. 

In his famous investigations on the trypanosome of the 
little owl, Schaudinn (18) gives the following account of the 
origin of the flagellar apparatus. The nucleus of the odkinete 
contains a karyosome in whicha “central grain” is surrounded 
by eight chromosomes. By heteropolar division the single 
nucleus divides into a larger nucleus, the trophic nucleus, and 
a smaller, the kinetonucleus (“ blepharoplast”’). The kineto- 


618 MURIEL ROBERTSON AND E. A. MINCHIN. 


nucleus is “a complete nucleus with centrosome and eight 
chromosomes, not merely a centrosome, karyosome, nucleolus, 
ora simple ectoplasmic thickening.” (The contrast drawn 
between a nucleus and a centrosome in this sentence is 
instructive.) he kinetonucleus then divides by another 
heteropolar mitosis and gives rise to a third nucleus, the 
smallest of the three; this third nucleus forms a nuclear 
spindle composed of eight mantle-fibres and a ‘central 
spindle” or centrodesmose connecting the two centrosomes 
situated at the two poles of the spindle. The central spindle 
becomes the flagellum and the eight mantle-fibres the eight 
myonemes. By growth and elongation of the flagellum and 
myonemes, one centrosome is carried out at the tip of the 
flagellum, while the other remains asits basal granule. From 
these statements of Schaudinn, it may at least be said without 
expressing any opinion as to the accuracy of the details in 
the development described by him that he regarded the basal 
granule of the flagellum as a centrosome, and that he dis- 
tinguished clearly between a centrosome and a nucleus, and 
in particular between the kinetonucleus and the centrosomic 
body from which the flagellum arises, although he used, in 
our opinion quite wrongly, the term “‘blepharoplast” to denote 
the kinetonucleus, instead of applying it to the basal granule 
of the flagellum. This mistake, as we consider it, in 
Schaudinn’s terminology is the more remarkable, since he 
seems to have understood so clearly the true centrosomic 
nature of the basal granule of the flagellum, and to have 
realised its existence independent of the kinetic nucleus. 

The most important contribution to the question of the 
blepharoplast in the Protozoa is the memoir of Jahn (8) on 
the swarm-spores of one of the Mycetozoa, Stemonitis 
flaccida. He finds that at division the centrosomes at the 
poles of the nuclear spindle give rise to the daughter-flagella 
while still actually engaged in their centrosomic functions ; 
a state of things entirely parallel to that which we have found 
in the collar-cells we have studied. 

Hamburger (5) found in Dunaliella the paired flagella 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 619 


arising from a basal granule which is connected with the 
nucleus. At division the basal granules divide and each 
gives off two flagella; though they do not appear to control 
the division of the nucleus inany way, nevertheless each basal 
granule is connected with the dividing nucleus by two streaks, 
giving an appearance very similar to that figured by us on 
Plate 25, figs. 4 and 5. (Jahn also figures a very similar 
condition.) Dobell (2), in his investigations on T'richo- 
monas, etc., appears to support a view similar to our own. 
Lastly, Yamamoto (21), who has studied the locomotor 
apparatus of various organisms by methods which seem to us 
unduly violent and severe, describes the flagellum of a trypa- 
nosome as arising from a basal granule (“ proximal centriole’’) ; 
his statements, in matters of fact, simply confirm those of 
Schaudinn.! 


OBSERVATIONS ON THE DIVISION OF THE COLLAR-CELLS.” 


The material on which this work was done consists of a 
number of specimens of Clathrina coriacea preserved by 
one of us at Roscoff, and embedded in paraffin at the time. 


1 Yamamoto states that he has obtained preparations of trypanosomes 
(species not stated) showing myoneme fibrille, of which he states I 
deny the existence. This is a glaring misstatement on his part, seeing 
that I have described and figured the myonemes of Trypanosoma 
peree and T. granulosum in full detail (vide ‘ Proc. Zool. Soc. 
Lond.,’ 1909, pl. v, figs. 84, 96, 97) —E. A. M. 

> I greatly regret that in my account. of the Sponges in Lankester’s 
‘Treatise on Zoology’ (Part II, 1900, p. 56) I gave an entirely erroneous 
account of the division of the collar-cells of Clathrina coriacea, 
stating that after division of the nucleus the cell divides transversely 
to its long axis, and then the basal portion forms a new collar and 
flagellum. I have re-examined the figures and preparations on which 
these statements were founded, and see that I was misled by sections 
passing obliquely through the epithelium, so that the top part of a 
dividing cell, with the nucleus at the apex, appeared superposed on the 
base of an ordinary cell, with its nucleus in the usual position. The 
account given in the present memoir will show clearly the error of my 
former statements.—H. A. M. 


620 MURIEL ROBERTSON AND E. A. MINCHIN. 


Most of the sponges were preserved in osmic acid followed by 
picrocarmine, a good method for showing clearly the cyto- 
plasmic structures, especially the collar and flagellum, but 
not suitable for demonstrating the finer details of the nuclear 
apparatus. Some of the material, however, had been pre- 
served in Hermann’s fluid, and it is on this that we base the 
results set forth in this memoir. Sections cut from sponges 
preserved in this way were stained with various stains, more 
particularly by Heidenhain’s iron-hematoxylin method, and 
counter-stained with eosin or Lichtgriin, the latter being 
found to be of great assistance in making out the details of 
the collar and flagellum, since these parts are tinged by it. 
(1) The Resting Collar-cell._—In Clathrina coriacea, 
as in all sponges of the family Clathrinide, the nucleus lies 
invariably, in the ordinary ‘‘ vegetative” or resting condition, 
at the base of the columnar collar-cell, that is to say, at the end 
which is furthest from the collar and flagellum. At the apex 
of the cell, in the centre of the area enclosed by the base of 
the collar, lies a minute granule—the blepharoplast—from 
which the flagellum takes origin. These structures, no less 
than the general form of the collar-cell and its position in the 
epithelium, of which it forms a part, give a definite orienta- 
tion to the cell; any direction parallel to an imaginary axis 
continuing that of the flagellum and passing through the 
blepharoplast and nucleus may be termed vertical, while any 
plane at right angles to the vertical axis may be termed 
horizontal. Fa 
The form of the collar-cell and the dimensions of their 
different parts vary considerably with the condition of the 
sponge, whether expanded or contracted, and may be different 
also in different parts of the same sponge. In specimens in 
which the pores are fully open, and in which all appearances 
indicate that the collar-cells are in full functional activity, 
the bodies of the cells are fairly broad, and about 12-13 4 
in height by 5-6 w in breadth ; the collar reaches a length of 
10-11 uw, and the flagellum some 25-27 uw. When, on the other 
hand, the pores are closed up and the sponge is partially con- 


ON 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 621 


tracted, the collar-cells become taller and narrower and the 
collar much shorter. In each cell the basal three fourths of 
the body is broader and more or less cylindrical in shape ; 
this part of the cell is in contact with the neighbouring cells, 
and constitutes the main body of the cell. We have not found 
processes connecting the bodies of the cells with one another. 
It has been shown by Minchin and Reid (14) that when the 
collar-cells are carefully brushed away and the wall of the 
sponge is stained with picro-nigrosin, a delicate blue-stained 
network is visible in surface view, representing a honeycomb- 
like structure, the spaces in which were originally occupied 
by the bodies of the collar-cells. Hence in life the bodies of 
the collar-cells are probably not in actual contact, but are 
separated by a delicate extension of the gelatinous ground- 
substance of the body-wall of the sponge. If, as would seem 
probable on theoretical grounds, the bodies of the collar-cells 
are connected across this intervening substance by proto- 
plasmic fibrils, such connections have escaped our notice, 
possibly on account of their being of extreme tenuity and 
requiring, perhaps, other methods of technique, in order to 
demonstrate their existence, than those employed by us for 
the study of the mitosis. It is well known that in other 
sponges the collar-cells may be connected by protoplasmic 
processes, as, for instance, in Hexactinellids, where such 
processes are extremely obvious, forming the so-called 
membrana reticularis. 

The cylindrical basal portion of the cell ends in a distinct 
rim or flange, and from this level arises a narrower portion, 
which may be termed the ‘neck,’ and which is quite free 
from any contact with neighbouring cells. The summit of 
the neck is rounded off, forming a convex lens-like area 
enclosed by the base of the collar, and giving off centrally 
the flagellum. The so-called collar has more the form of a 
cuff or sleeve when fully expanded. It is distinctly thicker 
and more rigid in its basal portion, becoming very delicate at 
its distal end, which is usually more or less shrunk and dis- 
torted in preparations. The uppermost limit of the collar is 


622°, MURIEL ROBERTSON AND E. A. MINCHIN. 


often very difficult to make out. It is best preserved in the 
osmic-picrocarmine preparations ; after Hermann’s fluid it 
appears collapsed and shrunk or frayed out. A short way 
above its origin the collar usually shows a distinct thickening, 
visible as a horizontal hoop-like structure, especially when 
the collar is a little contracted ; when it is expanded to its 
fullest extent the hoop is difficult to make out as a horizontal 
line, but its presence is marked by the fact that all the part 
of the collar below it stands out stiff and firm, and is not 
creased and folded like the part above. It is evident from 
the appearances seen both in the resting and the dividing 
cell that the collar for about 2 4 from its origin is thickened 
and strengthened as compared with its distal portion. 

The nucleus of the collar-cell is about 5 « in diameter and 
more or less spherical in form, sometimes slightly flattened in 
the vertical direction. ‘he most conspicuous element in its 
structural composition is a large grain, which stains deeply 
with iron-hematoxylin, and appears to be of the nature of a 
karyosome. This structure is always present, and sometimes 
double (figs. le, 3,7). The karysome is sometimes lodged 
in a clear space (fig. 36, d, e, f) ; its position in the nucleus 
varies. The remainder of the nuclear contents appear granular, 
but in thin sections of the nucleus a fine network can be made 
out (figs, 36, e, f, g), in the nodes of which the granules of 
chromatin are lodged. These granules vary very much in 
different nuclei in the same preparation, being sometimes so 
fine as to be scarcely visible individually, while in other 
nuclei they are coarse and irregular in size and shape (figs. 
30 and 36, a, b,c). All transitions can be found between the 
finely and the coarsely granular condition, but the two 
extremes form two: well-marked types, which may be 
characterised as the light and the dark type respectively. 
It is worthy of note that nuclei of cells about to divide are 
always of the light type, as will be pointed out in the next 
section. 

The above description of the nuclei refers to preparations 
stained with iron-hematoxylin. In material preserved and 


i ,, 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 623 


stained by the osmic-picrocarmine method the nuclear 
structure is not shown at all as a rule, but the nucleus simply 
stains evenly pink. Sometimes the karyosome can just be 
made out, sometimes not. A peculiar feature of the prepara- 
tions is that the red stain often does not extend up to the 
nuclear membrane; the stained portion forms a mass lying in 
the centre of the nucleus, and between this stained mass and 
the nuclear membrane a clear space remains, which can often 
be seen to be traversed by delicate radiating lines, as if fine 
filaments started from the membrane to support the central 
stained mass. Comparison with nuclei stained with iron- 
hematoxylin shows in many of the latter a distinct alveolar 
border to the linin-framework; sometimes the alveolar 
border is relatively very broad (fig. 36), and shows the 
radiating partitions of the alveoli very distinctly. It would 
appear as if the action of the osmic-picrocarmine method was 
to cause a shrinkage within the alveolar border, with the 
result that this inner portion of the nuclear framework 
contracts and appears as a homogeneous mass, which contains 
all the chromatin and stains deeply, leaving the alveolar 
border unstained. It should be noted that by no means all 
the nuclei of the collar-cells show the clear border within the 
membrane ; many of them stain evenly up to the membrane, 
and this is always so in those cells which are about to 
divide, 

The blepharoplast and flagellum stain black with iron- 
hematoxylin, but by the osmic-picrocarmine method they are 
not stained. 

The cytoplasm of the collar-cells is finely granular and 
usually very vacuolated. The neck is free from vacuoles as 
a rule, but in many cases a round vacuole-like structure, 
which differs in appearance from the other vacuoles, can be 
seen in the neck region. ‘lhe ordinary vacuoles in the body 
of the cell are clear and appear as empty spaces, doubtless 
representing drops of fluid in the living condition, but in the 
direct line between the nucleus and blepharoplast there is 
generally to be seen a vacuole, which has finely granular 


624. MURIEL ROBERTSON AND E. A. MINCHIN. 


contents and sometimes a minute central granule (fig. 30, cell 
on the extreme right). This body is sometimes nearer the 
blepharoplast, sometimes nearer the nucleus, but usually it 
lies at a level midway between the neck and the main body 
of the cell or in the neck itself; its significance is doubtful. 

In addition to the vacuoles, the cytoplasm almost always 
contains one or more coarse refringent granules of irregular, 
angular form and yellowish-brown colour. They are lodged 
in any part of the cell and are often present in the vicinity of 
the blepharoplast. They probably represent excretion-grains. 
After the iron-hematoxylin stain they become darker, but 
still retain their characteristic yellowish-brown tint, and can 
be easily distinguished from chromatin grains. No other 
enclosures, as a rule, are to be found in the collar-cells, but 
occasionally they contain large rounded bodies (figs. 31-35 
and 50, 51), which stain deeply with iron-heematoxylin and 
appear to be of the nature of organisms, though whether they 
represent parasites or food ingested by the cells is difficult to 
say. In some parts of the sponge they are found more com- 
monly than in others, and in one case (fig. 34) no nucleus 
could be made out in the cell; it may, however, have been 
cut off in the section. 

(2) Preparations for Division.—Before the nucleus 
begins to show any of the changes in its minute structure 
which initiate mitosis certain events take place in the cell, 
namely, the migration of the nucleus bodily from the base to 
the summit of the cell, the disappearance of the flagellum, 
and the division of the blepharoplast. As a general rule 
these three events take place in the order named, but not 
invariably, so that a number of different combinations arise 
in different cases. 

The migration of the nucleus is always the first sign that a 
collar-cell is about to divide, and this peculiarity is a great 
aid to the study of the division, since in a section of the 
sponge which shows the collar-cells cut vertically those that 
are dividing or preparing to divide arrest the attention at 
once, even with a comparatively low power of the microscope, 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 625 


owing to the fact that the nucleus is no longer in its usual 
position at the base of the cell, but has either migrated to the 
apex or has been preserved in the act of doing so, and is 
found in some position between the base and the apex (figs. 
1-5, etc.). Such cells are also characterised by being much 
broader and stouter than the ordinary resting cells, but they 
do not increase in height to an appreciable extent. 

By this process of migration the nucleus comes to lie 
immediately under the blepharoplast, and at this stage a 
curious appearance has been observed in two instances (figs. 
4 and 5); the nucleus is seen to be flattened on the side 
nearest to the blepharoplast, and from the blepharoplast itself 
two streaks appear to radiate to the two ends of the flattened 
side of the nucleus. Careful examination of each of these 
preparations gives the impression that these two streaks are 
in reality the optical section of a cone-shaped mass of proto- 
plasmic substance, the base of which rests on the flattened 
side of the nucleus, and which is, perhaps, the cause of the 
flattening. A comparison with the resting cell suggests that 
this conical mass is derived from the peculiar vacuole with 
granular contents, which was described in the last section as 
situated in the direct line between nucleus and blepharoplast, 
and that by the upward migration of the nucleus the vacuole 
in question is pushed up until it is caught, so to speak, 
between nucleus and blepharoplast, when, coming under the 
influence of the forces of attraction or repulsion exerted by 
the blepharoplast, it assumes the conical form seen. If this is 
a correct interpretation of the phenomena, the vacuole should, 
perhaps, be regarded as an archoplasmic vesicle, such as has 
been described in other cases, and which supplies some part of 
the material of the achromatic spindle in the mitosis. In fig. 4 
it is seen that the flagellum is still present, though short, 
while in fig. 5 the flagellum has entirely disappeared and the 
blepharoplast has divided. 

The disappearance of the flagellum and the division of the 
blepharoplast are two events which take place independently 
so far as their relative sequence in time is concerned, that is 


626 MURIEL ROBERTSON AND E. A. MINCHIN. 


to say, the flagellum may disappear completely before the 
blepharoplast divides or may persist until after this has 
taken place. In either case the two daughter-blepharoplasts 
migrate inwards and place themselves on opposite sides of the 
nucleus in order to become, as will be seen, the two centro- 
somes in the mitosis. If the flagellum persists during this 
process of events it remains attached to one of the two 
blepharoplasts (figs. 6 and 10), and becomes drawn into the 
body of the cell, as seen in figs. 7-9; in each of these 
three specimens the flagellum, though greatly shortened, is 
still persistent, and can be seen passing into the body of the 
collar-cell and terminating in one of the two blepharoplasts, 
while the other blepharoplast can be seen on the other side of 
the nucleus quite independent of the flagellum. On the other 
hand, figs. 5, 11, and 12 show the two blepharoplasts very 
close together at the apex of the cell and apparently very 
recently separated from one another, with no trace of a 
flagellum. 

The exact method in which the flagellum disappears is 
difficult to determine simply by comparison of different stages 
in sections; it could only be made out satisfactorily by 
watching the process in the living cell. In collar-cells in 
which the upward migration of the nucleus is taking place, 
the flagellum almost always appears much shorter than in 
the surrounding cells, an appearance too constant in occur- 
rence to be explained simply as due to artificial curtailment 
of the flagellum in the process of section-cutting, especially 
when the collar is intact and the flagellum does not project 
beyond it (figs. 9 and 10). But a remarkable feature of 
this stage is the frequent occurrence of a protoplasmic 
projection, like a small pseudopodium, from the apex of the 
cell round the base of the flagellum (figs. 7, 9, 39, 40) ; this 
process persists for atime after the flagellum has completely 
disappeared (figs. 14, 41). The appearances suggest that 
the cell throws out a pseudopodial process, by the help of 
which the flagellum is retracted and absorbed at its base ; 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORLACEA. 627 


in all cases the protoplasmic process in question is very 
short in proportion to the length of the original flagellum. 

The division of the blepharoplast takes place with forma- 
tion of a distinct centrodesmose connecting the two daughter- 
blepharoplasts (figs. 6, 7, 18). 

During these changes the collar remains practically unal- 
tered, except that it begins to show more or less clearly the 
appearance of shrinkage and degeneration characteristic of 
the succeeding stages of the division. 

(5) The Mitosis.—The general course of the mitosis in 
the collar-cell is similar to that known to occur in the cells of 
other Metazoa generally, and described for sponges by Maas 
_(10, 11) and Jorgensen (9). It is unnecessary, therefore, 
to do more than describe its most characteristic features. 

As already stated in a previous section, the nucleus of a 
collar-cell about to divide, but before any changes prepara- 
tory to division have begun in the chromatin contents, is of a 
pale type—that is to say, the granules of chromatin dis- 
tributed over the general framework are very fine and 
scattered evenly, so as to give the nucleus an almost homo- 
geneous appearance relieved only by the karyosome, stained a 
deep black, after iron-hematoxylin, in contrast with the pale 
grey tint of the remainder of the nucleus (figs. 1, 10,11). The 
nucleus at this stage is also distinctly larger than the average 
nucleus of a resting cell. 

The first changes to be observed in the chromatin contents 
of the nucleus are that they stain darker and become more 
blotchy and uneven in appearance, apparently as the result. of 
the minute granules of chromatin being clumped together to 
form coarse grains or masses. Figs. 6 and 7 show early 
stages in this process; the masses of chromatin still stain 
faintly, appearing to be loose in texture and ill-defined in 
ontline, and the karyosome stands out sharply. In later 
stages (figs. 8, 12) the chromatin masses become more definite 
in outline and somewhat smaller, and the deep stain they 
take gives the impression that they are more closely knit and 
of denser texture ; the karyosome, however, is still distinct. 


VOL. 55, PART 4,—NEW SERIES. 4.2 


628 MURIEL ROBERTSON AND E. A. MINCHIN, 


Finally, the chromatin masses become very definite and stain 
very deeply, and no distinct karyosome can be made out ; this 
body seems to break up and to contribute by doing so to the 
general store of chromatin, At first the chromatin masses, or 
chromosomes, as they may now be termed, appear to be con- 
nected together by delicate filamentous junctions (fig. 9) ; 
this stage corresponds apparently to the spireme stage. Next, 
the connections between the chromosome disappear, and they 
are seen lying separately from one another as irregular 
rounded masses, showing more or less distinctly indications of 
division, each into two (tig. 15). In spite of much searching 
we have not been able to find any stages other than those 
described, and, in particular, nothing more nearly resembling 
an ordinary spireme stage than the specimen shown in 
fig. 9. 

These changes in the interior of the nucleus also go on quite 
independently of the changes in the flagellum and blepharo- 
plast described in the previous section. Thus the flagellum 
may have vanished, and the two daughter-blepharoplasts may 
have taken up their definitive position when the nuclear con- 
tents are at the beginning of their changes (fig. 14); or the 
nucleus may be comparatively far advanced when the 
blepharoplast has only just divided (fig. 12), or betore the 
flagellum is absorbed (figs. 8,9). Finally, however, a stage 
is reached when the nucleus has resolved itself into a mass 
of separate chromosomes, and the two blepharoplasts, or, as 
they may now be termed, the centrosomes, are placed on 
opposite sides of it, indicating the two poles of the future 
nuclear spindle (fig. 15); when this stage is reached the 
nuclear membrane is absorbed and cannot be discerned. 

The formation of the nuclear spindle is seen in the two 
stages drawn in figs. 16 and 17. After the absorption of the 
nuclear membrane the chromosomes arrange themselves to 
form an equatorial plate, to which delicate rays can be seen 
to pass from the centrosomes, forming the characteristic 
achromatic spindle. The two centrosomes appear to be pushed 
further apart by the formation of the spindle, so that they 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 629 


come to lie at the extreme surface of the cell. The spindle 
is lodged in that portion of the cell which we have termed 
the neck in a previous section, and the centrosomes are 
situated about midway between the origin of the still per- 
sistent collar and the flange. The chromosomes appear 
massed together, and are difficult to distinguish individu- 
ally when the equatorial plate is seen in side view (figs. 17, 
18), but can be seen better in cells cut parallel to the plane 
of the equatorial plate (fig. 22). The number of chromosomes 
appears to be about sixteen. 

At this period, while the equatorial plate is still simple and 
undivided, an important event takes place. From the cen- 
trosomes at the two poles of the spindle the two daughter- 
flagella grow out, appearing as two minute hair-like projec- 
tions from the surface of the cell (figs. 18-21). This stage 
is a very common one, and it is, in fact, rare to find a collar- 
cell with a mitotic spindle without the two daughter-flagella 
projecting from the two centrosomes ; this indicates that the 
first formation of the flagella must be an extremely rapid one. 
Sometimes only one daughter-flagellum is to be seen, but in 
such cases the cell is usually shghtly oblique, and the missing 
flagellum has probably been cut off by the knife in cutting 
the section. ‘The two new flagella are formed entirely outside 
the original collar, which is still persistent. The condition of 
the collar is best studied in osmic-picrocarmine preparations 
(figs. 42—4.5) in which it is seen that the formation of the nuclear 
spindle causes the cell to become much broader, with the 
result that the base of the collar is greatly stretched. The 
thicker portion of the collar, below the hoop, retains its form 
more or less, but the portion above the hoop tends to collapse 
and fall together. 

From the stage with the single equatorial plate the diaster- 
stage arises in the usual way (figs. 23, 24). It is remarkable 
that we have succeeded in finding but few specimens of the 
diaster-stage, and, unfortunately, most of those have been cut 
obliquely or horizontally, and hence do not show well the 
relation of this stage to the cell as a whole. Figs. 23 and 24 


630 MURIEL ROBERTSON AND E. A. MINCHIN. 


show the two best diaster-stages we have found. Fig. 23 
shows the spindle well, but the cell is cut almost horizontally, 
and the collar and one daughter-flagellaum are sliced off; in 
fig. 24 the cell is cut more vertically, and shows the collar, 
but the plane of the spindle lies obliquely, and only one cen- 
trosome and daughter-flagellum can be made out. The scarcity 
of the diaster-stage indicates that it is passed over very 
rapidly, and this conclusion receives further support from 
the fact that in the subsequent stages, when the daughter- 
nuclei are being reconstituted, the daughter-flagella are 
scarcely longer than they were in the stage with the undivided 
equatorial plate. 

After the diaster-stage, and with the reconstitution of the 
daughter-nuclei, the cell-body begins to divide (figs. 25-28a). 
Between the two daughter-nuclei there are seen for a time 
streaks, the remains of the achromatic spindle, stretching 
across from one to the other (figs. 25-27); these streaks per- 
sist until the division of the cell-body is far advanced. The 
details of the reconstitution of the nuclei are difficult to make 
out clearly ; the chromosomes appear to fuse together into a 
compact mass in which their individuality is masked, if not 
lost. The division of the cell is effected by means of a con- 
striction in the vertical plane, producing a cleavage which is 
much more marked at the upper than at the lower end of the 
cell. The cleavage goes right through the old collar, and 
leads to its destruction and disappearance ; it appears to 
break down into a granular mass which disintegrates and 
vanishes. 

When division of the cell-body is complete the new collars 
of the daughter-cells grow out round the short but growing 
flagella. At their first origin the new flagella projected in 
an oblique direction from the dividing cell, as figs. 18-28 
show clearly; they took origin from that portion of the 
surface of the parent-cell which lies between the flange and 
the base of the collar. When the division is nearly complete 
(figs. 28a and 47), the point of origin of the flagella becomes 
slightly shifted so as to be placed at the uppermost level of 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 631 


the cell, with the result that the young flagella come to point 
vertically upwards. After complete division the form of the 
two daughter-collar-cells undergoes a change, becoming 
elongated in the vertical direction, so that the cell as a whole 
acquires a slender columnar form, with a shallow collar sur- 
rounding the short flagellum at the upper end (figs. 29, 30, 
48). <A curious feature of these stages, both those in which 
cleavage of the cell is taking place (figs. 25-284) and those in 
which division is recently completed (figs. 29, 30, 48), is that 
they are found in the sections at a higher level than the rest 
of the epithelium, as shown in figs. 30 and 48; the bases of 
the young cell are on a level with the flanges of the ordinary 
resting collar-cells. This peculiarity is very marked when 
the recently divided cells have assumed the columnar form ; 
they project so much above the general level of the collared 
epithelium that they become very conspicuous objects in the 
sections of the sponge, and are consequently very easy to 
find. Later they appear to push their way down amongst the 
other epithelial cells, and so find their normal level (fig. 49). 

The nuclei of the young collar-cells, at first compact masses, 
soon become looser in texture; the karyosome reappears and 
nucleus acquires the structure of the ordinary resting nuclei, 
from which it differs only in its smaller size. In osmic- 
picrocarmine preparations the young nuclei show the marginal 
clear zone very distinctly (figs. 48, 49). Immediately after 
division the nucleus is at the apex of the collar-cell (figs. 29, 
30, 48), but it now begins to migrate towards the base of the 
cell (fig. 49), and so resumes the position characteristic of the 
resting cell. ‘he collar and flagellum grow to their full 
leneth, and the latter arises froma basal granule or blepharo- 
plast which, as is clear from the development that has been 
described and depicted, is one of the two centrosomes of the 
nuclear spindle in the mitosis, derived from the division of 
the resting cell. 


632 MURIEL ROBERTSON AND E. A. MINCHIN. 


SuMMARY AND CoNCLUSIONS. 


The course of events that take place in the division of 
the collar-cells may be summarised briefly as follows, omitting 
the details of the mitosis, since they present no special 
peculiarities. 

The nucleus of the collar-cell migrates from the base to 
the apex of the cell, and so comes to lie immediately under 
the blepharoplast. The flagellum then disappears and the 
blepharoplast divides. The two daughter-blepharoplasts 
travel to opposite sides of the nucleus and take on the function 
of centrosomes. ‘The nucleus breaks up into chromosomes, 
its membrane disappears, and a mitotic spindle is formed in 
the ordinary way, with the two centrosomes at its poles. 
The two new flagella then at once begin to grow out from 
the two centrosomes, outside the original collar, before the 
equatorial plate is divided. The mitosis is completed, and 
as the cell-body divides the original collar breaks down and 
disappears. ‘lhe centrosomes become the blepharoplasts of 
the two daughter-cells, the flagella continue to grow out from 
them, the new collars grow up round the new flagella, the 
reconstituted daughter-nuclei migrate back again to the 
bases of the cells, and the two daughter-cells resume the 
structure and appearance of the ordinary resting collar-cells. 
Thus it is seen that the blepharoplast-centrosome is a 
permanent cell-organ, which multiplies with the cell; but 
that the collar and flagellum are formed afresh at each cell- 


division, quite independently of the collar and flagellum of 


the parent cell. 

In this process of division the feature to which we wish 
to draw special attention is the fact that the bodies which 
have the function of blepharoplasts in the resting-cell have 
that of centrosomes in the dividing cell. In fact, it is seen 
that during a certain stage in the division, the stage, namely, 
of the nuclear spindle, when the daughter-flagella are growing 
out from the centrosomes at the poles of the spindle, one and 


DIVISION OF COLLAR-CELLS OF CLATHRINA CORIACEA. 633 


the same body functions at one and the same time as a 
blepharoplast and a centrosome, thus furnishing a decisive 
proof of the identical nature of these bodies, at least in the 
class of cells that we have been studying. 

We are therefore in entire agreement with those authors 
who regard blepharoplasts as bodies of centrosomic nature. 
It is very obviousin the case which we have studied that the 
terms “ blepharoplast ” and “centrosome”? denote merely two 
different functional activities of the same body. It may well 
be that in other cases division of labour may lead to structural 
differentiation, and that two distinct and independent classes 
of bodies occur, centrosomes controlling nuclear division and 
blepharoplasts giving rise to locomotor cell-organs. But in 
all cases alike we regard centrosomes and blepharoplasts as 
organs similar in nature and identical in phyletic origin. 

It only remains to discuss how far the results we have 
obtained throw light on the state of things in other cases, 
and more particularly with regard to the vexed question of 
the true blepharoplast in trypanosomes, that is to say, whether 
thename “‘blepharoplast”’ should be given to the kinetonucleus, 
or to the basal granule of the flagellum in these organisms. 
With regard to this point, it may be stated at once that 
there is nothing whatever in the structure or behaviour of 
the centrosome-blepharoplast of the collar-cells to justify a 
comparison between it and the kinetonucleus of a trypano- 
some, or, indeed, a nucleus of any kind. We are fully in 
agreement with those who, following Schaudinn, regard the 
kinetonucleus of trypanosomes as a body of the nature of a 
nucleus, and it is precisely on this ground that we regard it 
as a body of a different nature from a true blepharoplast, 
such as that which is seen in the collar-cells, and which 
cannot, in our opinion, be identified as a nucleus by any 
stretch of the imagination. On the other hand, the body, 
which in a trypanosome corresponds in every way to the 
true blepharoplast, is the basal granule or centriole of the 
flagellum. 

Our position, therefore, with regard to the nuclear apparatus 


634 MURIEL ROBERTSON AND KE. A. MINCHIN. 


of a trypanosome is that the basal granule of the flagellum 
represents the true blepharoplast, a body of the nature of a 
centrosome, and that the kinetonucleus or German blepharo- 
plast is an accessory nucleus which is not represented in the 
economy of a collar-cell or in flagellated organisms generally, 
but which is a special feature of the genus Trypanosoma 
and its allies, especially the genera T'rypanoplasma, 
Herpetomonas, Leishmania, and Crithidia, a nucleus 
which doubtless possesses its own centrosome or centriole. 
With regard to the function of the kinetonucleus, its close 
association with the blepharoplast and the flagellar apparatus 
has generally been held sufficient to justify the assumption 
that it possesses a kinetic function, that is to say, that it is 
a nucleus specially concerned with the regulation of the 
function of locomotion. We require, however, more know- 
ledge with regard to the relatious of the kinetonucleus to 
the life-cycle as a whole, and more particularly to the 
phenomena of sex and sexual conjugation in these flagellates 
before this point cau be decided. We may refer in this 
connection to the interesting experiments of Werbitzki (19), 
who was able to obtain trypanosomes without a kinetonucleus 
(termed by him “ blepharoblast’’), and found that such 
individuals showed no difference, as regards their movements, 
from the trypanosomes of normal structure. This result 
seems to us to indicate that the flagellar apparatus of a 
trypanosome is not so dependeut on the kinetonucleus as is 
generally supposed, and also to be strongly in favour of our 
view that the basal granule of the flagellum, and not the 
kinetonucleus, represents the true blepharoplast. Werbitzki 
seems, in fact, to have reduced his trypanosomes artificially 
to the more primitive condition found in other flagellates aud 
also in collar-cells, a condition in which the organism 
possesses a nucleus and a true blepharoplast, but no kineto- 
nucleus. 

It may be objected to our conclusions that they are based 
only on analogy, and that a collar-cell is too far removed 
from a trypanosome in phylogeny and affinities to permit of 


DIVISION OF GCOLLAR-CELLS OF CLATHRINA CORIACKEA, 635 


conclusions being drawn with regard to the homologies of the 
flagellar apparatus of trypanosomes. It is, of course, possible 
that the conclusions drawn from the one do not strictly apply 
to the other, and it is certainly very desirable that these 
points should be studied in flagellates generally, and in forms 
allied to trypanosomes particularly, more than has been 
done at present. On the other hand a collar-cell, although it 
forms part of the epithelium of a sponge, is as much a 
flagellate organism in all points of structure and function 
as any free-living flagellate; and the study of cytology tends 
rather to demonstrate the essentially uniform nature of 
permanent cell-structures throughout the whole range of 
living organisms, whether animal or vegetable. 


ListER INSTITUTE, CHELSEA, S.W. 
April 26th, 1910. 


REFERENCES. 


1. Belajeff, W.—* Ueber die Centrosome in den spermatogenen Zellen,” 
‘Ber. Deutsch. botan. Ges.,’ xvii, 1899, pp. 199-205, pl. xv. 

2. Dobell, C. C.—* Researches on the Intestinal Protozoa of Frogs and 
Toads,” ‘Quart. Journ. Micr. Sci.,’ 53, 1909, pp. 201-277, pls. 
2-5, 1 text-fig. 

3. Erhard, H.— Studien titber Flimmerzellen,”’ * Arch. f. Zellforschung,’ 
iv, 1910, pp. 309-442, pls. xxii, xxiii, 16 text-figs. 

4. Goldtschmidt, R.— Lebensgeschichte der Mastigamében Masti- 
gella vitrea n. sp. und Mastigina setosa n. sp.,” * Arch. 
Protistenk.,’ suppl. i, 1907, pp. 83-168, pls. v—-ix, text-figs. A-U. 

5. Hamburger, C.—* Zur Kenntnis der Dunaliella salina, ete.,” 
‘Arch. Protistenk.,’ vi, 1905, pp. 111-180, pl. vi, 7 text-figs. 

6. Hartmann, M., and Prowazek, S. v.— Blepharoplast, Caryosom, und 
Centrosom,” ‘ Arch. Protistenk.,’ x, 1907, pp. 3U6-335, 8 text-figs. 

7. Henneguy, L. F.—“Sur les rapports des cils vibratiles avec les 
centrosomes,” ‘Arch. d’ Anat. Micr.,’ i, 1898. 

8. Jahn, E.—‘ Myxomycetenstudien 3,” ‘Ber. Deutsch. botan. Ges.,’ 
xxii, 1904, pp. 84-92, pl. vi. 

9. Jorgensen, M.—* Beitrage zur Kenntnis der Hibildung, Reifung, 
Befruchtung und Furchung bei Schwammen (Syconen),” * Arch. 
f, Zellforschung,’ iv, 1910, pp. 163-242, pls. xi-xv, 1 text-fig. 


636 MURIEL ROBERTSON AND E. A. MINCHIN. 


9a. Joseph, H.—* Beitrige zur Flimmerzellen und Centrosomenfrage,” 
‘Arb. Zool. Inst. Wien,’ xiv, 1903, pp. 1-80, 3 pls., 3 text-figs. 

10. Maas, O.—“ Ueber Reifung und Befruchtung bei Spongien,” * Anat. 
Anzeiger,’ xvi, 1899, pp. 290-298, 12 text-figs. 

“Die Weiterentwicklung der Syconen nach der Metamor- 
phose,” ‘ Zeitschr. wiss. Zool.,’ xvii, 1900, pp. 215-240, pls. ix—xii. 

12. Minchin, E. A.—‘ Investigations on the Development of Trypano- 
somes in Tsetse-flies and other Diptera,” ‘Quart. Journ. Mier. 
Sci.,’ 52, 1908, pp. 159-260, pls. 8-15, 2 text-figs. 


1 


138. “The Relation of the Flagellum to the Collar-cells of Cal- 
careous Sponges,” ‘Zool. Anzeiger,’ xxxv, 1909, pp. 227-231, 6 text- 
figs. 

14. - and Reid, D. J.—* Observations on the Minute Structure of 


the Spicules of Caleareous Sponges,” ‘ Proc. Zool. Soc., London,’ 
1908, pp. 661-676, pls. xxxiv-xxxvii. 

15. Prowazek, S. v.—* Flagellatenstudien,” ‘ Arch. Protistenk.,’ ii, 1903, 
pp. 195-212, pls. v, vi. 

16. ——— “Die Kernteilung des Entosiphon,” ‘ Arch. Protistenk., 
ii, 1903, pp. 325-328, 12 text-figs. 

17. Schaudinn, F.—** Uber den Zeugungskreis von Parameeba eil- 
hardi,” ‘ Sitz. ber. Akad. Wiss. Berlin,’ 1896, pp. 51-41, 12 text- 
figs. 


“ Generation- sund Wirtswechsel bei Trypanosoma und 
Spirochete,” ‘Arb. k. Gesundheitsamt., Berlin,’ xx, 1904, pp. 
387-459, 20 text-figs. 

19. Werbitzki, F. W.—*‘ Ueber blepharoblastlose Trypanosomen,” ‘CB. 
Bakt. Parasitenkunde,’ 1 Abth. Orig. lili, 1910, pp. 303-315, pls. 
i, ii, 6 text-figs. 

20. Wilson, E. B.—‘ The Cell in Development and Inheritance,’ 2nd 
edition, 1904. 

21. Yamamoto, J.—‘* Ueber den Lokomotionsapparat der Protisten- 
zellen,” ‘CB. Bakt. Parasitenkunde,’ 1 Abth. Orig. liii, 1909, pp. 
38-42, 1 pl., 1 text-fig. 


18. 


DIVISION OF COLLAR-CELLS OF GLATHRINA CORIACEA. 637 


EXPLANATION OF PLATES 25 anp 26. 


Illustrating Miss Muriel Robertson and Mr. E. A. Minchin’s 
paper on “ The Division of the Collar-cells of Clathrina 
coriacea (Montagu): A Contribution to the Theory 
of the Centrosome and Blepharoplast.” 


[All the figures are drawn from sections of material fixed with 
Hermann’s fluid and stained by Heidenhain’s iron-hematoxylin method ; 
the outlines were traced with the camera lucida at a magnification of 2000 
linear, with the exception of figs. 36 and 37, which are magnified 5000 
linear. | 

PLATE 25. 


Fig. 1.—Six collar-cells in their natural arrangement ; five of them 
are in the resting state; the sixth (d) shows the nucleus in its migration 
towards the blepharoplast, and has a very short flagellum. 

Fig. 2—Early stage in the upward migration of the nucleus; the 
blepharoplast in the act of division, with shortened flagellum. 

Fig. 3.—Similar stage to the last, the blepharoplast distinctly divided, 
the nucleus with two karyosomes. 

Figs. 4 and 5.—Stages showing the nucleus in close proximity to the 
blepharoplast and distinctly flattened on the side nearest the blepharo- 
plast, from which two streaks are seen to come down to the two ends of 
the flattened border of the nucleus; these two streaks appear to be 
the optical section of a cone-shaped figure. In fig. 4 the flagellum is 
seen to be still present, but shortened ; in fig. 5 no flagellum is seen and 
the blepharoplast is divided. 

Fig. 6.—Stage showing the divided blepharoplasts connected by a 
centrodesmose ; the flagellum is still present and of fair length. 

Figs. 7, 8, and 9.—Stages showing complete division of the blepharo- 
plast with persistent flagellum in each case; the two daughter-blepharo- 
plasts in each dividing cell have travelled inwards and placed them- 
selvesat opposite sides of the nucleus, drawing in withthem the root of the 
flagellum, and the free portion of the flagellum has its base surrounded 
by an upgrowth from the apex of the cell. In fig. 7 the adjacent rest- 
ing cell is drawn for comparison ; in the dividing cell a centrodesmose 
is seen between the two blepharoplasts, and there are also indications of 
a streak running down from one of the blepharoplasts to a granule in 
the body of the cell, but this streak appears to be due merely to the 
arrangement of vacuoles in the cytoplasm, and not of the nature of a 


638 MURIEL ROBERTSON ANID E. A. MINCHIN. 


centredosmose. These three figures also show three different conditions 
of the nucleus preparatory to mitosis. In fig. 7 the karyosome is very 
distinct, while the remainder of the chromatin is pale, but beginning to 
aggregate into larger masses. In fig. 8 the karyosome is also distinct, 
but the rest of the chromatin is darker and the coarse granulation is 
more distinct. Fig. 9 shows the stage which appears to correspond 
to the spireme-stage; the chromatin is in darkly staining masses 
(c hromosomes), connected by fainter lines, and no karyosome can be 
made out. All three cells are from the same slide. 


Fig. 10.—Blepharoplast divided, remnant of flagellum still present ; 
nucleus not showing any preparation for mitosis. Cell cut somewhat 
obliquely. 

Fig. 11.—Blepharoplast divided, flagellum entirely absent; nucleus 
as in last. 

Fig. 12.—Blepharoplast and flagellum as in last; nucleus showing 
beginning chromosome-formation, but karyosome still distinct. 

Fig. 13.—-Cell cut obliquely, showing two blepharoplasts connected 
by a centrodesmose. 


Fig. 14.—Cell showing the flagellum completely withdrawn, and 
represented only by a little upgrowth from the body of the cell; the 
two blepharoplasts (centrosomes) have placed themselves on opposite 
sides of the nucleus, which is still in a very early stage of preparation 
for mitosis, with distinct karyosome and pale chromatin. 

Fig. 15.—Similar stage, but with the chromatin of the nucleus com- 
pletely broken up into chromosomes. No karyosome is to be made out. 
One centrosome is seen on the right at the side of the nucleus, the 
other on the left, rather low down and almost under the nucleus. 


Figs. 16, 17.—Stages showing the formation of the nuclear spindle. 
In fig. 16 the chromosomes are still irregular in arrangement, while in 
fig. 17 they are arranged to form a definite equatorial plate. No 
flagella have as yet grown out from the centrosomes. 


Figs. 18-21.—Stages with the nuclear spindle and with daughter- 
flagella growing out from the centrosomes (blepharoplasts). In fig. 
18 the spindle lies slightly obliquely, and only one daughter-flagellum 
is seen. In fig. 21 the cell is cut obliquely. 

Fig. 22.—Nuclear spindle cut in the plane of the equatorial plate, 
which is seen from one of its flat surfaces. 

Figs. 23, 24.—Diaster-stages. Fig. 23 shows a cell cut obliquely, and 
only one of the daughter-flagella is seen. In fig. 24 the nuclear spindle 
lies obliquely, and only the left-hand centrosome and daughter-flagellum 
can be seen. 


DIVISION OF GCOLLAR-CELLS OF CLATHRINA CORIACEA. 639 


Fig. 25.—Late diaster-stage, with beginning reconstitution of the 
daughter-nuclei. Slightly oblique; only one daughter-flagellum to be 
seen. 

Figs. 26-28, 28a.—Stages in the division of the cell-body, with recon- 
stitution of the daughter-nuclei. In all figures, except 28a, the remains 
of the original collar can be seen clearly. In figs. 26 and 27 the remains 
of the achromatic spindle can be seen between the two daughter-nuclei. 
In fig. 284 the division is practically complete. 

Figs. 29, 30.—Pairs of young, recently divided collar-cells. In fig. 
30 some of the adjacent cells are drawn to show the way in which the 
cells at this stage are raised up above the surrounding cells. 


Figs. 31-35.—Collar-cells showing enclosures of various kinds, some 
of them perhaps of parasitic nature. In the cell shown in fig. 34 the 
nucleus seems to have disappeared, but may have been cut off. 

Fig. 36.—Nuclei of resting collar-cells, magnified 3000 linear. a, b. 
ce, dark nuclei; d,a light nucleus; e, f,g, thin sections of nuclei showing 
the reticular structure; in g the karyosome does not come into the 
section. 

Fig. 37.—Transverse sections of collar-cells in the region of the 
collar. a passes through the base of the collar, and b just above this 
level; both show the blepharoplast centrally. In ¢ the collar is cut 
transversely with the flagellum in the centre. 


PLATE 26. 


[ All the figures are drawn from sections of material fixed with osmic 
acid and stained with picro-carmine; magnification throughout 2000 
linear. | 


Fig. 38.—Two collar-cells, one of the normal resting type (on the 
left), the other with the nucleus migrating towards the apex of the cell 
preparatory to division. 

Figs. 39, 40.—Collar-cells showing the nucleus at the apex of the cell, 
and the flagellum in process of retraction by means of a pseudopodium- 
like process from the cell. 


Fig. 41.—On the left a normal resting cell; on the right a cell with 
the nucleus at the apex and the flagellum completely retracted, but 
represented by the still persistent pseudopodium-like process seen in 
the two preceding figures. 

Figs. 42, 45.—Stages with the daughter-flagella growing out from the 
poles: of the nuclear spindle, and with the collar beginning to collapse. 
The achromatic elements, namely, spindle and centrosomes, are not 
stained and are not visible in the preparation, but the equatorial plate 


640 MURIEL ROBERYSON AND E. A. MINCHIN. 


is seen. In fig. 43 the collar contains a foreign body, as in the right- 
hand cell in fig. 48. 

Figs. 44-46.—Diaster-stages with daughter-flagella. In fig. 44 a 
resting cell is drawn for comparison; in fig. 46 the cell is cut obliquely 
and does not show the collar. 

Figs. 464, 47.—Stages in the division of the cell-body. In fig. 46a the 
collar is still seen; in fig. 47 it has disappeared. 

Fig. 48—Two young, recently divided collar-cells, drawn with three 
ordinary resting collar-cells to show the manner in which the young 
cells project above the level of the epithelium. The collar-cell on the 
extreme right shows a foreign body lodged in the lumen of the collar. 

Fig. 49.—Four collar-cells, of which the two middle ones are evidently 
a pair of sister-cells, the product of recent division, showing the nuclei 
in the act of migrating down to the base of the cell. 

Figs. 50, 51—Two collar-cells showing bodies (parasites?) in the 
cytoplasm. 


M.R..adnar del 


DIVISION ¢ 


a 


Huth, Lith? London 


BLLAR-CELLS. 


MR.ad nat dei. 


Quart. Sourm.cMicn Sei. V0.5. NEM, 26 


Tiuth Tath? Landas 


DIMES TON MOF “C Oa eA R= riaigs= 


nd 


STUDIES ON AVIAN HAMOPRO'TOZOA. 641 


Studies on Avian Hemoprotozoa. 


I, On certain Parasites of the Chaffinch (Fringilla celebs) 
and the Redpoll (Linota rufescens),! 


By 
H. M. Woodcock, .D.Sc.(Lond.), 
Assistant to the University Professor of Protozoology. 


With Plates 27—31. 


ConvreNtTS 


PAGE 
1. Introductory . : : . 641 
2. Experimental Work and mectedue : F . 645 
3. The Parasites in Relation to their Hosts . , . 657 
4, Description of Trypanosoma fringillinarum, n.sp. . 664 
(A) As found in the Birds : : . 664 
(Bs) As found in Cultures ; 680 
The Significance of the Cultural Hoes of Tr ypano- 
somes in Relation to the Question of an Alternate, 
Invertebrate Host ; ; end 
(c) Notes on Nuclear Cytology and Divinion : ATG 
(D) Comparison with other Avian Species ; me alts: 
5. Note on Halteridium fringille (Labbé) : . toe 
6. Note on Leucocytozoon fringillinarum, n.sp. 728 


1. InTRODUCTORY. 


My reason for taking up the study of Avian Heemoprotozoa 
has been the desire to obtain, if possible, some definite 
enlightenment on the important question of their life-cycle. 
The far-reaching conclusions bearing upon this subject, to 

' This research was carried out as Mackinnon Student of the Royal 
Society during the year 1907-1908. The publication of the results has 
been delayed for several months owing to a long stay at Rovigno in the 
endeavour to supplement this work by the study of the actual parasites 
described by Schaudinn in Athene noctua. 


642 H. M. WOODCOCK. 


which the celebrated protozoologist, the late Fritz Schaudinn, 
was led as the result of his well-known researches (27) 
on certain parasites of the little owl (Athene noctua), have 
been largely discredited by many subsequent workers in this 
field. This is chiefly due to the suggestion, first put forward 
by the American workers, Novy and McNeal, that there is 
nothing in Schandinn’s description to show that the author took 
sufficient precaution against the liability of confusing the life- 
histories of what were really separate and independent para- 
sites. Novy and McNeal, in their endeavour to confirm 
Schaudinn’s views, investigated the trypanosomes of various 
birds (14), and also made a study of the flagellates occurring 
naturally in mosquitoes (15). As a result of their work they 
have maintained that Schaudinn was entirely wrong in regard 
to all his main conclusions. They consider, on the contrary, 
that the trypanosomes of birds are quite distinct from 
intra-cellular parasites (such as Halteridium), and further, 
that they do not undergo any part of their life-cycle in an 
insectan host, the flagellates occurring in the latter having 
no connection with the trypanosomes. 

I chose avian forms on which to work for the following 
reasons: In the first place, a considerable amount of research 
has now been done on various trypanosomes parasitic in other 
vertebrates, e.g., fishes and mammals, which will be referred 
to in due course. Secondly, it is from a study of avian forms, 
if any, that one may reasonably expect to learn how far 
Schaudinn’s views and statements were justified. As a 
matter of fact, at the present time the trypanosomes of birds 
are those about which the least is positively known, for 
Novy and McNeal’s work, while it has undoubtedly re- 
opened the entire question, does not, on the other hand, 
contribute much to its definite settlement. In my opinion, 
many of the conclusions reached by these authors are equally 
open to criticism. They themselves have certainly not 
brought forward adequate or sufficient evidence to justify 
the negative views adopted by them. 

Hosts Selected to Work upon.—It was my intention 


STUDIES ON AVIAN HAMOPROTOZOA. 643 


to study first the parasites of the “little owl” itself. In 
spite of all my efforts, however, I could not obtain a supply 
of these birds here at home, so that I was obliged to turn 
my attention to other birds. Recent observations have 
shown that many kinds of birds harbour trypanosomes, and 
it is probable that their infection with these parasites is 
fairly widespread in nature (cf., for instance, the numerous 
American species which Novy and McNeal. found to be 
infected). The only worker, to my knowledge, who has pub- 
lished any notes relating to the occurrence of avian 
trypanosomes here in England is Petrie (21), who 
observed the parasites in the blackbird, swallow, house- 
martin, song-thrush, chaftinch, and yellow-hammer; he 
failed to find them in the crow, sparrow, starling, or 
jackdaw. 

Had it been my object to find trypanosomes in as many 
different birds as possible and to content myself with 
noting their presence, it would have sufficed to shoot 
various kinds of wild birds and examine them at once. This 
habit of describing and naming trypanosomes from one or 
two casual observations is unfortunately far too prevalent ; 
it is one which adds little or nothing to our knowledge of the 
really essential points on which light is needed. For the 
purposes of my investigation I felt it was best to restrict 
myself to birds which could be obtained without much 
difficulty, and which were hardy and would live well in 
captivity. Hence, with a few exceptions at the commence- 
ment of the work, when I was endeavouring to “lay a 
course”? as it were, I have used small native cage-birds, 
obtained from various dealers. Mentioning the exceptions 
first of all, in order to give a complete list, I began with 
some Java sparrows (Padda oryzivora), from which host a 
trypanosome, T’. paddz, has been described by Thiroux. 
But after spending some time fruitlessly in attempts to find 
this parasite, which was not present, and my limited supply 
of these birds giving out, I relinquished the search. In spite 
of great efforts to trap common birds, the only result was a 

VOL. 55, PART 4.—NEW SERIES. 43 


644 H. M. WOODCOCK. 


blackbird caught for me at Elstree, which died two days 
after receiving it. Neither in this, nor in another blarkbird, 
purchased, were any trypanosomes found. A_ barn-owl 
(Strix flammea), which was kindly given me by Dr. Dean, 
also proved negative.! I may add here that in one of the 
Paddas and in one blackbird Halteridia occurred, but 
sparingly ; I thought it best, however, not to take up this 
aspect of the question at first, but to continue my search 
for hemoflagellates and concentrate my attention on them 
in the first place, turning to the Hemosporidia later, as 
should appear desirable. 

The small birds, of which I have examined most, are 
closely allied members of the finch family (Fringillide, 
sub-fam. Fringilline), namely, greenfinches (Chloris 
chloris), chaffinches (Fringilla ccelebs), redpolls (Linota 
[Acanthis] rufescens), and linnets (L. [A.] cannabina). 
Trypanosomes were found only in the chaffinches and red- 
polls, so that for the greater part of the time I have occupied 
myself entirely with these. Unfortunately during the spring 
these birds also were very scarce and difficult to procure, 
and I was unable to replenish or augment my stock when 
I particularly wished to do so. 

The occurrence of the parasites in these two hosts cannot 
be considered as at all rare. Out of twenty-two chaffinches 
examined, five were found to be naturally infected, sixteen 
birds were certainly uninfected, and one was doubtful. 
Neglecting this last,” the percentage works out at about 
24. As regards the redpoles, trypanosomes occurred in 
three out of fifteen; eleven were uninfected, and one, again, 
was uncertain. This gives an approximate percentage of 
21:5, which is not very different from that in the case of 
the chaffinches. As far as they go these proportions are 
reliable, because they are exhaustive—that is to say, the 


1 In the case of blackbirds this was not conclusive as to the absence 
of the parasites, for no cultures were made (cf. below, p. 658). 

2 Also in the case of the first chaffinch and redpoll no cultures were 
taken, as I had no tubes ready at the time. 


STUDIES ON AVIAN HAIMOPROTOZOA. 645 
negative side also can be relied upon, for reasons which are 
given below ; in this respect they differ from most previous 
tables and estimates of trypanosome-infections of birds. The 
figures suffice to show that,so far as occurrence is concerned, 
the birds with which I have worked do not bear out the dismal 
statistics given by many of the researchers (e. g. Ziemann, 
the Sergents, Dutton and Todd, etc.) 

Intra-cellular Parasites in the Chaffinch.—In 
several of the chaffinches I noticed, when looking for trypano- 
somes, the presence of Halteridia; except in one case, which 
I shall describe shortly, these were only scanty in number. 
I have also observed,in three cases, an interesting leucocytic 
parasite, which is quite different in appearance from the 
celebrated Leucocytozoon ziemanni of owls. 

What is undoubtedly a similar parasite has been observed 
independently by Dr. Stevenson, of University College, in 
smears of the blood of a greenfinch, which he has kindly 
shown me for comparison. 


2. Merrnops or Work; Arremprs at ‘TRANSMISSION BY 
Mosquitors ; TECHNIQUE. 


Fresh blood was always taken, in the living bird, from a 
fairly large marginal vein of the wing, prominent where it 
crosses the arm on the inner side, immediately below the 
elbow-joint. A fine-pointed surgical needle of the triangular- 
bladed kind was used. It is essential that the point be 
sharp. Unless a clean prick is obtained, the blood does not 
exclude freely in a good drop, but suffuses beneath the skin, 
raising a swelling from which blood cannot be got satisfac- 
torily. Asarule bleeding stops quickly. Should it give any 
trouble, a swab of cotton-wool, dipped in lysol, is applied to 
the wound and the wing closed up over it and held to the 
side of the body for a few minutes. The vein soon recovers 
from this little operation, and can be used again, if desired, in 
a couple of days or so. 

Culture-tubes.—The use of culture-tubes has been of the 


646 H. M. WOODCOCK. 


greatest service to me. I have developed and extended 
Novy and McNeal’s method, making use of it not only on the 
dead bird, but also—what is much more difficult—on the 
living bird. In taking drops of blood for culture-tubes, the 
great desideratum is to get the region of the arm above- 
mentioned sterile if possible. The part is very well washed 
and gently rubbed first of all with cotton-wool soaked in 
lysol, particular attention being paid to the skin near the 
base of the feathers. The lysol must then be washed away 
with distilled water, which has been well boiled. Lastly, the 
water is absorbed as well as possible with more cotton-wool, 
which has been boiled along with the water, and from which 
the hot water is quickly pressed out. ‘This is preferable to 
using loose wool and serves to take up most of the water, 
the warmth also helping in drying the part. It is most 
important to have the arm as dry as possible before pricking 
the vein, otherwise the blood spreads and runs over the 
surface. As it exudes, the blood is taken up by a sterilised 
Pasteur pipette, the drawn-out tube of which is long enough 
to pass into the expression-water of the culture-tube. 

It is, of course, a much easier matter to get sterile inocula- 
tions from the bone-marrow, heart, etc., if the ordinary 
precautions are adopted. 

If a culture-tube can be successfully inoculated with four 
or five drops of blood, I have found that in a few days 
(usually five to seven, sometimes fewer) one can generally say 
with confidence whether the bird was infected, according as 
the tube develops trypanosomes or not. Unfortunately, even 
with the greatest care, the inoculated tubes are sometimes 
badly contaminated before that time has elapsed. In such 
circumstances I never rely upon a negative indication, 
though I may add that now and again a positive result has 
been obtained where the medium had become contaminated. 
When I have been unable to get any cultures to develop in 
two or three sterile tubes taken from a bird, subsequent 
examination and culture of the bone-marrow after death 
have also proved negative. Hence I have regarded the 


STUDLES ON AVIAN HAMOPROTOZOA. 647 


above as a reliable test of the presence of the trypanosomes 
in the living bird. 

Culture Media.—The parasite from the chaffinch and 
redpoll lives and multiplies readily in a blood-agar medium, 
prepared either after Novy and McNeal’s recipe, or according 
to Mathis’ modification. At first I followed the American 
authors (see 14, p. 265), but added only an equal volume 
of defibrinated rabbit’s blood to the sterilised meat-agar, 
as I found this to be quite sufficient. ‘lubes so prepared 
always have an ample quantity of expression-liquid, in which 
the parasites thrive at any temperature from 20° to 25°C. A 
temperature of 28° to 30°C. was found to be too high, if it 
was desired to keep the tube for any length of time, as the 
try panosomes soon die off, owing to their too rapid multiplica- 
tion and exhaustion of the nutrient material. At the lower 
temperature the tube is all right for about twelve or fourteen 
days, and some of the trypanosomes will remain alive longer 
if a little salt-citrate solution is added to replenish the 
medium. If it is desired to keep the culture going for some 
time, however, it is necessary to make a sub-culture, after 
ten or twelve days, by transferring a drop of the. medium 
containing the parasites to a fresh tube. By this means I 
have kept a continuous series of cultural forms, both from 
the chaffinch and from the redpoll, thriving and multiplying 
for six and a half weeks, the one having been transferred 
(sub-cultured) four times, the other, I think, only thrice. 
Had it not been for the accident of the temperature of the 
incubator rising to nearly 30° 0. for two or three days, 
whereby the trypanosomes were all killed off, the cultures 
could apparently have been kept for as long as I wished. 

The great drawback to this method is that, where, as in my 
case, a large number of the tubes are used, too much time 
and labour are involved in obtaining sufficient rabbit’s blood. 
Mathis’ modification (10), which I have now followed for 
some time, avoids this difficulty. In this method, ox-blood, 
which can be readily got from a slaughter-house, is used 
instead. A quantity is allowed to fall direct into a sterilised 


648 H. M. WOODCOCK. 


receptacle, and at once defibrinated. As before, equal 
volumes of blood and agar are mixed. The tubes, when 
prepared, must be sterilised by the fractional method at a 
temperature of about 100° C. (under rather than over), for 
an hour or so on two successive days. This is necessary to 
ensure sterility. 

‘Owing to this process, however, tubes prepared thus are 
often deficient in expression-liquid ; to remedy this | or 2 c.c. 
of boiling salt-citrate solution (‘75 per cent. salt + 1 per cent. 
sodium-citrate), are added to each tube, which is then left 
for a day or two before being inoculated; the liquid absorbs 
nutrient material from the solidified part. ‘lhe trypanosomes 
will not live in salt-citrate solutions alone. I have tried various 
combinations of salt, sodium-citrate, and (or) citric acid, 
similar to those used in cultivating the Leishman-Donovan 
bodies, but with no success. For the practical purpose of 
ascertaining whether a bird is infected or not I have found 
these tubes to be, as a rule, as serviceable as the others; 
but I do not think they suit the parasites quite so well. 
The culture does not start quite as easily, and multiplication 
is often somewhat slow at first. It is at least four or five 
days before the trypanosomes can be found at all readily ina 
small drop taken for examination, whereas in the case of the 
other tubes three or four days usually suffice. Again, after 
a week or nine days the parasites tend to become very 
granular and altered, and large agglomeration-clusters form 
sooner. In short, the trypanosomes do not live “ healthily” 
so long in this kind of culture as in the other. 

I may point out, with regard to the macroscopic appearance 
of infected tubes, that in the case of the parasites with 
which I have been working, there is normally. nothing 
indicative of their presence to be seen. A culture (if free 
from bacteria) looks just like an uninoculated tube. Even 
when the parasites are very abundant, the expression-liquid 
remains clear and unaltered in colour. Not once have I 
found the parasites on the solid part of the medium. They 
never form visible colonies or masses there. The only 


STUDIES ON AVIAN HAMOPROTOZOA. 649 


instances where anything unusual is to be noticed are in old, 
used-up tubes, in which the liquid is full of clumps of agglo- 
merated parasites, and many are degenerating and dying. 
‘hese masses tend to settle to the bottom of the liquid, and 
may be apparent as a small quantity of whitish-yellow scum. 

Inoculation of Birds with Trypanosomes.—I en- 
deavoured to produce an infection with trypanosomes in birds 
which I had found to be uninfected. So far, the only means 
at my disposal of doing this has been by inoculating; and 
most, certainly, of my attempts in this direction failed. In 
all about twenty-five inoculations were performed, and only 
in three cases was any positive result afterwards observed, 
which might be due to the inoculation. Many of the failures 
resulted from attempts to inoculate other (uninfected) 
birds with the trypanosome of the chaffinch and redpoll. 
Thus, a couple of linnets, one of them inoculated twice, 
proved negative. Also a barn-owl was tried with no more 
success. I was rather surprised, however, to find that a 
canary, which I thought would be very likely to prove 
susceptible, refused to become infected. It was inoculated 
three times, twice from cultures, and once from fresh 
(infective) blood, mixed with a little salt-citrate solution. 

A few words in connection with the modus operandi. 
To begin with, I inoculated the birds intra-pleurally, as 
recommended by Novy and McNeal, but I lost two or three 
redpolls straightway as a result of the operation. It was 
very cold weather at the time, and this may have conduced to 
their collapse. Since then, J] have always found it much 
more satisfactory to do the birds intra-peritoneally or intra- 
muscularly (in the pectoral muscles). None of the birds so 
inoculated suffered any ill-effects, even though, occasionally, 
they were done in both ways at once. The “dose” was 
generally four or five drops(from one eighth to one sixth of 
a cubic centimetre) of the liquidin the tube. ‘his contained, 
of course, numbers of parasites. 

With regard to the three cases in which the trypanosomes 
were observed subsequently, I may point out that I had made 


650 H. M. WOODCOCK. 


sure, by means of good cultures, that all three birds had no 
trypanosomes in the blood prior to the inoculation, and 
therefore I considered them to be free from those parasites.! 
Hence these are in all probability instances of successful 
inoculation. One case was that of a chaffinch inoculated with 
the parasites from a redpoll; another was that of a redpoll 
inoculated with a culture from a chaffinch. With regard to 
the third case, that of a chaffinch inoculated with a culture 
from another chaffinch, I have been very uncertain, owing in 
part to the different course the infection took, whether the 
appearance of the trypanosomes in this instance was really 
due to the inoculation, or was connected with the presence in 
this bird of Halteridium. I now think this was also a 
case of successful inoculation, for reasons which are discussed 
below (see p. 678). 

Attempts to Transmit the Parasites by Mos- 
quitoes.—lIt was a great disappointment to me that all my 
efforts to get mosquitoes infected with the trypanosomes from 
the birds have been fruitless. Both from Schaudinn’s descrip- 
tion of the infection of Culex with the trypanosomes from 
the ‘little owl,” as well as on account of the known réle 
of this insect as alternate host of the Proteosoma 
(Hemoproteus) of birds, I thought it most likely that 
mosquitoes would prove to be the transmissive agents of the 
parasites—at any rate, the trypanosomes—of the chaffinch 
and redpoll. 

Unfortunately I was baffled in the very initial stage of all 
the experiments. I was never able to get the mosquitoes to 
bite the birds. I have tried at different seasons of the year, 
late spring, summer, and early autumn, and at periods when 
the temperature has been quite high for this country. Most 
of my attempts were made with females which were bred out 
from larve. None of them, however, showed the slightest 
inclination to bite. Nor would they feed on a guinea-pig, 
with which I tried them occasionally. They would only take 


' T have worked throughout on the assumption that if trypanosomes 
are present, they will occur, if sparingly, in the general circulation. 


STUDIES ON AVIAN HA#MOPROTOZOA. 651 


such things as sugar-water, banana-juice, or mashed date. 
And if they were not provided with something of this kind 
they soon died off. 

I also obtained several batches of ‘ wild”? mosquitoes 
(females), thinking these might at any rate bite. Indeed, 
Prof. Minchin, who sent me some from Norfolk, said they 
were biting the horses in the open fields at the time. But 
here again I had no better luck. In fact, the Culex seemed 
to starve instead of feeding on the bird. I have kept batches 
under observation without food,! and seen their bodies gradu- 
ally become attenuated, until, although placed for a couple of 
nights consecutively with a bird, and without other food, by 
the fourth or fifth day (since they last took food) many of 
them would be dead. The mosquitoes were nearly always 
placed with the bird in the late afternoon, and left with it all 
night. Care was taken, of course, that they should be per- 
fectly able to get to it and feed if they wished. Now 
and again, also, [ held a tube containing a few hungry- 
looking insects to the bird’s body for a little time, displacing 
the feathers so as to expose the skin; and similarly with the 
guinea-pig. I tried keeping the mosquitoes in a biological 
incubator at a temperature of about 25°C. (77°-78° F.), fora 
day or two before using them, but this did not make any 
difference. Even small pieces of organs containing blood 
from freshly killed rats remained untouched so far as I could 
see. In short, all my efforts to induce Culex to take blood 
were unavailing. 

What is the probable explanation of this unwillingness 
experienced of the insects to bite? Such a total failure in 
this respect was quite unexpected. Taking into consideration 
the results in this connection—fortunately more successful— 
since gained at Rovigno, I think that there is probably more 
than one reason for the above negative results. In the first 
place, the question of temperature and moisture in the air is 
very important. I found this to be the case at Rovigno. 


' But not without water, a small dishful of which was always kept in 
the cage. 


652 H. M. WOODCOCK. 


Until the beginning of June I had the same difficulty there. 
As soon, however, as the regular summer weather set in—a 
moist, sweltering warmth—there was no difficulty in getting 
the Culex to bite (once, at any rate). It must be re- 
membered that all the research done on Culex hitherto, in 
this connection, from which it is known both to transmit 
‘certain hematozoa and to harbour flagellates (which in many 
cases are most probably hemoflagellates), has been done in 
countries where a much higher average summer temperature 
is experienced than in England. And I do not think that I 
succeeded in getting sufficiently favourable environmental 
conditions in my laboratory attempts in London. 

There is another probably equally essential point, of which 
I was not aware at the time of my (the above) experiments. 
According to Mr. EH. H. Ross, in a report on the prevention 
of fever on the Suez Canal (Cairo : National Printing Depart- 
ment, 1909),' the mosquitoes (females) apparently desire to 
suck blood only after having been fertilised. As it happened, 
in my early work I kept the bred-out females separate from 
the males, of which I took no account, thinking they were 
not required (as, of course, they do not take blood). Hence 
those females used were certainly not fertilised. As regards 
the caught “wild” ones, however, it is just as likely that 
they were fertilised as not, so that some of these ought to 
have bitten, had other conditions been suitable.” 

Another Possible Insectan Host.—Owing to my lack 
of success in this essential preliminary, I was left in the dark 
as to whether Culex was the alternate host of the Heematozoa 
of the chaffinch or not. I may point out in passing that a 
study of the cultural forms of the Trypanosome which I have 


1 See ‘ Nature,’ vol. lxxix, 1909. 

2 In working at Rovigno, where I was able to breed out the Culex 
in greater abundance, I left the two sexes together, for the sake of con- 
venience in dealing with the insects. In this case many females were 
fertilised, for I frequently noticed the little “ egg-rafts ” floating on the 
dishes of water in the cage. Probably those females which sucked 
blood had been fertilised. 


STUDIES ON AVIAN HAMOPROTOZOA. 653 


obtained, and their comparison with various flagellates 
described in blood-sucking Invertebrates (cf. below), leaves 
no doubt whatever in my mind that these bird-trypanosomes 
have some alternate (doubtless insectan) host. But it is 
quite possible that, in the present instance, some other 
insect than Culex performs this role. I endeavoured to 
ascertain what other biting insect was likely to be concerned. 
Mr. Austen, of the- British Museum, very kindly informed 
me of a small hippoboscid fly, of the genus Ornithomyia, 
which is an ectoparasite of various birds, especially to be 
found on nestlings.! Up to the present, however, I have been 
unable to obtain a supply of these insects. 

It seems to me not at all unhkely that it is in this 
direction one must look for the alternate host. If this be 
the case, it is very probable that infection usually occurs 
while the birds are quite young, and before they leave the 
nest. 

Early in the autumn I obtained a young redpoll, infected 
with trypanosomes, which could not have been more than two 
months old, if that, when bought ; and as most of these little 
cage-birds are caught, I am told, as soon as they can look 
after themselves and before they finally leave the nest, this 
may very well be a case in point.”, Unfortunately, owing to 
the hampering restrictions of wild birds’ protection acts, etc., 
I could not get hold of any nests containing fledgelings for 
examination. Towards the end of the close season a bird- 
seller did procure a chaffinch nest for me, from which the 
young birds only flew away as he approached. ‘This was well 
searched for insects, but contained none. I may add that I 
have never noticed any insects (fleas, lice, etc.) on my birds 

1 The Sergents have recently found (80) that a hippoboscid fly 
belonging to the genus Lynchia is most probably concerned in the 
transmission of the Halteridium of the pigeon. Lynchia, however, 
is not met with in Britain. 

? An interesting observation noted by Danilewsky of trypanosomes 
being present in a young roller-bird only a week old also supports this 
view. The only alternative would be that of hereditary infection, which 
is extremely doubtful. 


654 H. M. WOODCOCK. 


when examining or inoculating them; they always seemed to 
be free from anything of this kind. 

Technique.—All my permanent preparations are in the 
form of smears made on slides. Asarule, the thinner the smear 
the better the result. In the case of very stout trypanosomes it 
happens occasionally that they are rather flattened out if the 
smear is too finely drawn ; but in thick smears the parasites 
are often not well stained by the Romanowsky method, being 
too blue in appearance. As regards smears of the cultural 
forms, I experienced some difficulty at first, on account of the 
expression-liquid (the medium containing the parasites), of 
which the drop to be smeared consisted. This was quite clear 
in the fresh condition, but formed a sort of coagulum after 
fixation, which stained very readily. Hence the trypanosomes 
appeared to lie in a layer of substance, stained reddish, which 
was often somewhat dense immediately around them. This 
coagulated layer was much more noticeable in smears made 
from the first kind of tubes than it was when I used the 
second kind, to which salt-citrate solution was added. The 
only means of obviating the trouble was to make the film as 
thin as possible and to take care that no stain was deposited 
on the slide. 

Fixation.—Most of my preparations have been fixed 
with osmic acid vapour; the few smears not so fixed were of 
little value as regards the trypanosomes. I make use of a 
4 per cent. solution of osmic acid, placed in the bottom of a 
stain-tube, to which two or three drops of acetic acid are 
added. The slide to be fixed is placed in the tube as quickly 
as possible after the film has been drawn. A fairly deep or 
thick glass ring in the liquid at the bottom of the tube 
prevents the slide itself from getting wet. Slides are left in 
contact with the vapour from twenty seconds to half a minute, 
the shorter time particularly in the case of a smear from a 
culture. After fixing, the slide is placed in absolute alcohol 
for fifteen to thirty minutes,according to convenience. If the 
smear is to be stained by the Romanowsky method, it is not 
advisable to leave the slide in absolute alcohol for much 


STUDIES ON AVIAN HAMOPROTOZOA. 655 


longer than half an hour; I have always found a longer 
period to be detrimental to the staining. I found this 
method of fixation to be the best for giving a correct idea of 
the size*and general appearance and morphology of the 
parasites, whether trypanosomes or intra-cellular forms; and, 
for the sake of uniformity, all my figures are of individuals 
so fixed, so that one may be compared at once with another, 
without any ulterior considerations having to be taken into 
account. 

Staining.—Nearly all my preparations are stained by 
some variety of the Romanowsky method. I have made use 
of two stains (or stain mixtures): one of them is the 
ordinary Giemsa solution, the other is a combination which I 
have found particularly good for culturalforms. ‘he Giemsa 
solution was always used in the customary proportion of one 
drop of the stain to 1 c.c. of water. The length of time for 
which slides were allowed to stain varied in different cases. 
The period required to give the best results varies con- 
siderably at times, even when the smears have been fixed, so 
far as can be told, in exactly the same manner. For one 
thing, the temperature made considerable difference. I used 
the stain at the laboratory temperature, and whereas in the 
winter and spring forms in the blood required to be stained 
for twelve to eighteen hours to be successful, in the summer 
they would be excellently stained in three or four hours. 

Cultural forms stain much quicker than the parasites in the 
blood, and need only about fifteen to twenty minutes in the 
stain ; but the Giemsa solution was found to be not nearly so 
suitable for smears of cultural forms as the other method 
which I adopted; by this latter method the parasites them- 
selves are more sharply stained, while the coagulated layer, 
which is often unpleasantly prominent as a reddish ground- 
substance, after Giesma, hardly stains at all. 

In my particular method three solutions are made use of, 
as follows : 

(1) A 1 per cent. solution of azure I, in equal parts of 
glycerine and methyl-alcohol. 


656 H. M. WOODCOCK. 


(2) A 1 per cent. aqueous solution of methylene-blue 
(Héchst—an essential point), to which 5 per cent. of pure 
sodium-carbonate is added. This solution is kept warm at a 
temperature of 40° to 45° C. for a couple of days or so, 
when it is made up, after which it is ready for use. 

(8) A 2 per cent. solution of eosin (also Hochst). 

In using the stain, I have found that a mixture made up in 
the following proportions gives very good results! : four drops 
of each of the three solutions are added to 10 c.c. of distilled 
water. The different liquids are poured from small drop- 
bottles of equal size, the drop-bottles being the same as are 
generally used for Giesma. (The drops themselves of the 
different liquids are not, it may be noted, of the same size.) 

By this method cultural forms are excellently stained in six 
to eight minutes ; and if any stain is deposited in the ground- 
substance it comes away readily with orange-tannin after- 
wards. In fact, on a good smear of cultural forms thus 
stained, it is often scarcely apparent macroscopically that 
there is anything at all on the slide. For staining trypano- 
somes in the blood, only forty to fifty minutes is required. 

In all cases, whichever method of staining was used, the 
slide was well rinsed with tap-water after staining, and then a 
few drops of orange-tannin were poured on the slide for half 
a minute or so, to remove the excess of stain. If, after 
further washing with water, the parasites still appeared to be 
over-stained, either more orange-tannin or else acetone was 
added. The latter must be used extremely cautiously and 
quickly rinsed off, for though at first it only extracts the blue, 
it soon begins to take out the red from the flagellum. 
Eventually the slide was washed with distilled water and 
allowed to dry. 

I have since regretted that, owing to the great scarcity 


1 These proportions can be varied, of course, as is found most suitable, 
in other cases. I may mention that I experimented some time using 
either (1) or (2) alone in combination with (3), in various proportions, 
but I never obtained anything like the good results that I did after 


using both (1) and (2) together. 


STUDIES ON AVIAN HAMOPROTOZOA. 657 


of the trypanosomes in the blood, I was not able to make 
use of the iron-hematoxylin method of staining. For there 
is one distinct drawback to the Romanowsky method and 
its variations. While it may be regarded as giving, after 
fixation with osmic, a perfectly reliable presentation of the 
form and general structure of the body, it is now quite clear 
from the most recent research (see, for example, Minchin 
[12] and Minchin and Woodcock [13]) that the nuclear struc- 
ture and details cannot be interpreted correctly by the aid 
of stains of this kind alone. This is owing to the invari- 
able tendency of Romanowsky stains to deposit the red 
colour in excess around certain organelle, especially small 
granules, which are thus overloaded with stain and arti- 
ficially enlarged to many times their real size, often with 
the result that other cytological features are quite obscured. 

Nevertheless, this characteristic behaviour of the 
Romanowsky stains being now proved and recognised, due 
allowance can be made therefor, and hence one is not 
hikely to be seriously misled in the case of a study such as 
is here described, which deals chiefly with the compara- 
tive morphology and behaviour of different types of form. 
Further, it may be pointed out that results obtained by the 
use of the same methods throughout may be compared with 
confidence. 


3. THE Parasites IN ReELation to THEIR Hosts. 


Numerical Scantiness of the Trypanosomes.—Asa 
rule, the trypanosomes are extremely scarce in the peripheral 
circulation of an infected host. This fact renders it often 
an excessively slow and wearisome process to get hold of the 
parasites at all in a living bird, and hampers any work 
upon them more than can be imagined until such research 
has been attempted. Unfortunately, there is all but 
unanimous agreement among observers upon this point,! 


1 The only exception of which I am aware is indicated by a statement 
of Vassal (86) in describing a trypanosome from an Annam pheasant. 


658 H. M. WOODCOCK. 


which it would be tedious to cite in detail (cf. the remarks 
by the Sergents [29], Novy and McNeal [14], Laveran [6], 
Dutton and Todd [4], and others). I will only add that Petrie, 
in the note already referred to, states that he could not find 
the trypanosomes in the blood of any of the infected birds, 
but only saw them in the bone-marrow. With respect to this 
numerical scarcity, birds are certainly the most trying of all 
vertebrate hosts. There can be no doubt that, owing to this 
factor, an erroneous idea has often been obtained of the pre- 
valence of trypanosome infections among birds. This has 
been well shown by Novy and McNeal, whose adoption of 
the culture method is of very great value in this connection. 
It will sufficiently illustrate this to give the statement of 
these authors that, in the case of forty-three various birds 
where microscopic examination had failed to reveal try- 
panosomes, nineteen, or 44 per cent., were proved by means 
of cultures to have been infected. 

‘lo give now my particular experiences. Out of five 
naturally infected chaffinches only in one were trypanosomes 
ever seen in freshly drawn peripheral blood ; in this case, I 
once saw an individual in a cover-slip preparation. The 
same bird was examined at intervals during three months 
subsequently, but I never saw any living parasites again. 
That they were still present in the general circulation, how- 
ever rare, was proved nevertheless on three occasions by 
means of cultures. Once, determined to find this elusive 
parasite if possible, I took a few drops of blood and made 
several smears, which were fixed and stained. In six good- 
sized films, which were minutely and thoroughly searched, 
representing a labour of several days, only one trypanosome 
was seen! It is important to note that these observations 
were made during the early spring, from January to April. 
In the case of the trypanosome parasitic in the redpoll I was 


This writer was in the happy position of being able to say that the 
parasites were not infrequent in the peripheral circulation. An indivi- 
dual could be found in every two or three fields (of an oil-immersion 
lens). 


STUDIES ON AVIAN HAMOPROTOZOA. 659 


not able to see it in the peripheral blood at all during the 
first five months of the year, although in two cases I knew 
by means of cultures that the birds were infected. During 
the early autuinn, however, I was able to find it in smears from 
a very young bird, which had probably not been long 
infected. ‘he number of parasites on a fair-sized film 
varied from six to ten in September, but only from four to 
eight in films made in October. 

Principal Habitat.—In general, the trypanosomes are 
most numerous in the bone-marrow; this is certainly their 
principal habitat. Two or three parasites can usually be 
found in a fresh cover-slip preparation from one of the long 
bones of an infected bird. But even here, at times, consider- 
able search is necessary,! since the parasites are apt to be 
hidden by clumps of leucocytes, erythroblasts, etc. However, 
there is generally no difficulty in finding the trypanosomes 
in a carefully made smear of a small, teased-up fragment of 
bone-marrow. ‘hus, when the chaffinch above alluded to 
was killed, some of the smears from the bone-marrow con- 
tained twenty trypanosomes or more. 

Artificial Infection.—Only in a couple of instances up 
to the present have I had the pleasure of finding trypano- 
somes at all plentiful in the peripheral circulation. One of 
these cases, at any rate, was certainly the result of success- 
ful inoculation. ‘his was a chaffinch which was infected 
with a culture of the form from the redpoll. Hxamined 
previously, no parasites had been found in this bird. On 
December 19th it was inoculated intra-peritoneally with a 
fifteen-day culture. On December 21st, twelve days later, 
examination of the blood showed at least five trypanosomes 
in two fresh cover-slip preparations, which were not ex- 
haustively searched; and permanent smears made at the 
same time proved to contain quite a considerable number of 
parasites—twenty to twenty-five or more on a good-sized 


' Certainly in one instance, where I failed to find any parasites in a 
careful search of the bone-marrow, the trypanosomes subsequently 
appeared in a culture taken from this organ. 


VOL. 55, PART 4.—NEW SERIES. , 44 


660 H. M. WOODCOCK. 


film. On New Year’s Day also, two parasites were found in 
a living preparation without much difficulty. When next 
examined, however, on January 10th, only one trypanosome 
was seen in two cover-slip preparations, which were 
thoroughly searched; this indicated a marked diminution in 
numbers. And in one permanent smear taken at the same 
time I could not find a trypanosome at all. This bird was 
not looked at again until the beginning of February, when no 
trypanosomes were seen in a living preparation. Neverthe- 
less, the parasites were still present, for a tube inoculated 
subsequently developed a culture; evidently the parasites 
had by this time diminished in number to their customary 
scantiness. Unfortunately, this chaffinch accidentally escaped 
soon afterwards, flying away through an open window. 

A Strong “Mixed” Infection.—I have left to the last 
a consideration of my most interesting case. On March 20th 
I inoculated a chaffinch with a seven-day culture of the 
chaffinch form. Three good (i.e. sterile) tubes had been 
inoculated from this bird previously, and had not developed 
any parasites. Hence I was practically certain that there 
were no trypanosomes present in this bird. Examination of 
the blood at intervals from March 26th until April 3rd, that 
is, until fourteen days had elapsed since inoculation, proved 
negative, no cover-slip preparations showing any parasites, 
so that I was very doubtful whether the inoculation had been 
successful. About three weeks afterwards the bird was 
again examined with a like result, but to make the matter 
certain, a tube (the first)' was then taken. To my surprise 
this developed a culture, the presence of the trypanosomes 
being thus proved, although I had never seen them in the 
fresh blood. I propose to leave aside, for the present, the 
question of whence these trypanosomes had come. 


1 [T had not made a culture on the occasions of the earlier examina- 
tions, thinking that if the inoculation had been successful the parasites 
would have been readily observed in the circulation, as in the other 
instance described. 


STUDIES .ON AVIAN HAMOPROTOZOA. 661 


This bird was then left alone for some weeks,! until with 
the approach of summer I decided to look at it again and 
see if the oncoming season appeared to make any difference 
in the number or condition of the parasites. Hxamining a 
cover-slip preparation on the afternoon of June 16th I was 
surprised to see numerous microgametocytes of Halteri- 
dium, The stimulus of cooling was causing many of them 
to rupture the red blood-corpuscles, and rapidly form and 
liberate the active male gametes, I had never seen any 
Halteridia in the preparations or smears made previously 
from this chaffinch; if this parasite was present then it must 
have been extremely scarce in the peripheral circulation. 

I was so occupied with watching this process of the libera- 
tion of the gametes and in endeavouring to see actual 
conjugation stages (unfortunately without success) that I did 
not search these fresh preparations for trypanosomes. In 
permanent smears made at the same time, however, trypano- 
somes occur, but they are not numerous (half a dozen or so 
on a slide). 

Having this abundant Halteridium-material, and know- 
ing the bird to be infected with trypanosomes also, I 
determined to examine it in the night-time to see if I could 
obtain any phases connecting these two types of parasite. 
Blood taken at 1.30 a.m. on June 18th showed the same 
condition as regards the Halteridia, and, in addition, Trypano- 
somes were easily found, three and four respectively being 
seen in two cover-slip preparations without any difficulty ; 
and there were probably several more in each, The trypano- 
somes seen were manifestly much larger than the Halteridia, 
and I saw no indications of a rapid transformation of the 
Halteridia into trypanosomes, or vice-versa; indeed, the 
only Haiteridia observed free in these living preparations 
were the adult gametocytes, male or female, behaving in the 


1 The bird was not made use of during this period because I had now 
given up making permanent preparations when a living drop failed to 
show the parasites. I had learnt that the probability was so much 
against my finding any trypanosomes in a reasonable time. 


662 H. M. WOODCOCK. 


usual manner. Many smears of the blood were made, some 
at once, others after waiting a moment or two, and with or 
without the addition of a drop of salt-citrate solution. 

The bird was again examined on the afternoon of June 
22nd, when one trypanosome was seen in two cover slip pre- 
parations after some searching. Another night examivation 
was made about 1 a.m. on June 50th. Compared with the 
previous night examination there appeared to be as many 
trypanosomes present, but the mature Halteridia did not 
seem to be quite so numerous as before. After a similar 
procedure I at length killed the chaffinch (about 2.30 a.m.) 
in order to obtain smears from the internal organs—heart, 
liver, spleen, bone-marrow, kidneys, etc. Most unfortunately, 
I omitted to make any preparations from the lungs—an over- 
sight which I have since greatly regretted. I need only 
mention here that the trypanosomes were afterwards found to 
be comparatively few in number in preparations from the 
bone-marrow, while in smears from the liver, etc., they are 
very scarce. As regards the peripheral circulation, the 
parasites are certainly more numerous in these night-slides 
than they are in those taken (from the same situation) in the 
daytime (afternoon). Hence there would seem to be, to 
some extent, a wandering of the trypanosomes from the 
internal organs (probably chiefly from the bone-marrow, 
which is their principal “ internal” habitat) into the peri- 
pheral circulation during the night-time. 

Halteridium in Relation to the Corpuscles.—As 
already indicated, the Halteridial infection of this bird was a 
very strong one, and the parasites were very numerous at 
this time; in fact, in some smears, for instance, from the 
liver, they are almost abundant. ‘The Halteridia are of all 
sizes, from minute forms up to fully grown adults. Nearly 
all the parasites are intra-cellular. Until recently the only 
cases in which I observed any forms free from the corpuscle? 


1 Of course, ripe sexual individuals, which have become rounded off 
and liberated themselves from the corpuscles, are not included in this 
statement ; neither are distorted or irregular individuals, which have 


STUDIES ON AVIAN HAMOPROTOZOA. 663 


—in spite of much searching—were four or five instances in 
which a special kind of individual, with peculiar features, 
was found free in the plasma. Having been led, however, 
as a result of my observations at Rovigno, to again examine 
very carefully certain of my preparations made at night, I 
have now found here and there a few individuals of small or 
intermediate size, and apparently of normal appearance, free 
in the blood. It is noteworthy that these free individuals 
have been seen only in smears from the peripheral blood, 
and not, for instance, in preparations from the liver, where 
the parasites are most numerous. Hence I do not think 
that the first impression I formed, namely, that the Halteridia 
do not leave the blood-corpuscle in the course of their 
growth, can be sustained. 

Occurrence of the Leucocytozoon.—The new 
leucocytozoon which I have observed occurred in three 
chaffinches. In two it was very scanty, only one or two 
isolated individuals having been noticed, and they were 
small. In one bird, however, which happened to be that 
which was successfully inoculated with Trypanosomes from 
the redpoll (see above, p. 659), the Leucocytozoon is not 
at all infrequent. The parasites are nothing like so 
numerous as the Halteridia are in the case just described, 
but there are certainly as many or more Leucocytozoa than 
there are trypanosomes on any smear. On one film more 
than twenty-five have been marked, and the slide has not 
been exhaustively searched for all the minute forms. 
Unfortunately, I did not detect this parasite in living, 
cover-slip preparations. For one thing, I was examining 
the chaffinch in which it occurred for Trypanosomes, which 
can be readily seen; further, as this species does not 
produce the characteristic spindle-like appearance of the host- 
cell, as in the case of nearly all other Leucocytozoa so far 
described, there was nothing about the parasites to catch 


obviously been accidentally set free from a ruptured corpuscle in making 
the preparation, such as are occasionally met with. 


664 H. M. WOODCOCK, 


the eye. If I passed over one in my search I doubtless took 
it merely for a large leucocyte. 


4, Description or TRYPANOSOMA FRINGILLINARUM, N. SP. 


(a) As Found in the Birds. 


The trypanosomes from the chaffinch (Fringilla coelebs) 
and the redpoll (Linota rufescens) most probably belong 
to one and the same species. The trypanosome once noted, 
but not described, by Ziemann, in 1898, was most likely this 
form; and the same applies doubtless to Petrie’s observations 
(21) in 1905. The occurrence of trypanosomes in the redpoll 
has not been known hitherto; this bird is a new avian host 
for the parasites. I regard the trypanosome from these two 
birds as a distinct and new species, for which I propose the 
name |’. fringillinarum. 

I discuss below the question of the specificity of different 
trypanosomes, with reference particularly to avian forms. I 
will merely give here the chief reasons which lead me to 
consider all the different types met with in the chaffinch and 
redpoll as belonging to one species. In the first place the 
ordinary, or definitive form of the parasite, the type, that is, 
which affords in the existing state of our knowledge the 
chief basis of morphological comparison in a systematic 
study of different Trypanosomes, appears to be essentially 
the same, as regards form and structure, both in the chaffinch 
and in the redpoll (cf. for instance figs. 4 and 31 of individuals 
froma naturally infected chaffinch with figs. 3 and 32 respec- 
tively of parasites from a naturally infected redpoll). Again, 
the forms which appeared in the blood of a chaffinch as the 
result of inoculation with a culture of the redpoll-parasite 
are also of a similar type (cf. figs. 1 and 28).! Secondly, 
although considerable polymorphism is shown, transition 
forms occur, which are intermediate between the more 

’ The fact that the inoculation of the parasites from the redpoll into 


the chaffinch was successful itself points to the specific identity of 
the two forms. 


STUDIES ON AVIAN HASMOPROTOZOA. 665 


extreme types noticed and serve to connect them. Lastly, 
it may be added that the various cultural forms to which the 
parasite from the chaffinch gives rise. are quite similar to, 
and cannot be distinguished from, those developed from the 
trypanosome of the redpoll. 

The ordinary or definitive type of T. fringillinarum is 
elongated and slender in appearance (figs. 1-4, 27, and 28) 
The aflagellar end is long and finely tapering, at times being, 
indeed, extremely attenuated (fig. 27).! The free flagellum 
is usually comparatively short. The trypanosome possesses 
a well-developed undulating membrane, which has three or 
four folds or pleats, broad and deep. The average dimen- 
sions of a full-sized ‘ adult” individual are as follows: 


Total length, including flagellum . . Al to 45 
Greatest width, including undulating 

membrane ‘ : ‘ . 4d to dp 
Greatest width of undulating mem- 

brane : 13 


Length of tapering aflagellar portion of 
body, i.e. the distance from kine- 
tonucleus to extremity. : . Sto7m 

Length of free flagellum ; ipaeeebOnd 

The trophonucleus (nucleus) is situated near the middle of 
the body, often slightly in the aflagellar half. It hes gene- 
rally somewhat nearer to the undulating membrane than to 
the opposite side. ‘The nucleus is more frequently ovoid in 
shape, but it may be approximately round (figs. 1 and 29) ; 
in the former case it may measure as much as 3p by 2 p, 
and in the latter case it may have a diameter of 24; but 
these dimensions are not always attained. 

The kinetonucleus appears as a relatively large body, 


1 This aflagellar prolongation is very delicate and liable to be broken 
off and lost in the preparation of the specimen; hence, now and again a 
parasite is seen which appears to have no “snout” at all, and where the 
body appears to be terminated by the kineto-nucleus; this is certainly 
an artificial condition, for it is characteristic of the fully grown ordinary 
individuals to have this long attenuated process at the aflagellar end. 


666 H. M. WOODCOCK. 


ovoid or rather oblong, which occupies the entire width of 
the parasite at the point where it is situated. Its apparent 
size isabout 1} tol4 by ly. It is nearly always intensely 
stained after Romanowsky stains, and shows no structural 
details. 

The flagellum, at its proximal end, nearly always stops 
short of the kinetonucleus; only very exceptionally does it 
appear to come into contact with the latter organella. In 
this connection it may be emphasised that my specimens 
are all from films properly fixed with osmic-acid vapour— 
none from air-dried smears. Moreover, at the point where 
the flagellum terminates, a definite granule, staining rather 
more deeply, can sometimes be made out quite clearly 
(figs. 4, 28). Unfortunately in many cases the root portion 
of the flagellum, which is probably intra-cytoplasmic, is not 
well stained, and in these the granule cannot be made out. 

The cytoplasm stains pale blue, and is of fairly uniform 
structure, appearing in sume instances finely alveolar. 
Occasionally a few small vacuoles or spaces are to be seen 
in the cytoplasm, but I have not observed anything that 
could be regarded as a definite, regularly occurring 
organella of that kind. In some of these forms the cyto- 
plasm is free from granules; in others, however, granules 
which stain bright red, and are of varying size, occur in 
greater or less number (figs. 1, 3, and 27). These granules 
are most probably of a chromatoid nature, derived from the 
nucleus, 

The structure of the undulating membrane shows an 
interesting feature. Running longitudinally in the broad 
folds or pleats, usually about the middle, is a prominent 
line, which stains blue—not red, like the flagellar border 
(figs. 1, 2, 4, 28-382). With a good light it is not difficult 
to make out that the part of the fold nearer to the body 
appears slightly denser than that on the outer side of this 
line, and stains faintly but distinctly blue, whereas the outer 
part is practically colourless. The explanation of this 
structure is that it represents a delicate intrusion of the 


STUDIES ON AVIAN HAMOPROTOZOA. 667 


endoplasm, running part of the way into the pleat of fold, 
between the two (otherwise) closely apposed ectoplasmic 
layers which constitute the membrane. The longitudinal 
line about the middle of the fold is the edge or limit of this 
inner endoplasmic layer. Laveran, in his account of T. 
avium (6), calls attention to a “rib” or longitudinal 
striation in the membrane. ‘This striation corresponds, in all 
probability, to the limit of an endoplasmic intrusion similar 
to that just described. 

Apart from the undulating membrane, I have never seen 
indications of an ectoplasmic layer. The trypanosomes I 
have studied show no sign of a well-developed, red-stain- 
ing “periplast,” such as has been described by several 
workers in the case of T. lewisi, for example. As a matter 
of fact, I should not expect to see any such appearance here, 
since the ectoplasmic part of the folds of the membrane is 
itself generally quite colourless, as already mentioned, and 
at most shows in one or two instances the faintest possible 
tinge of pink colour, which would be quite lost against the 
stronger blue of the body. Nevertheless, there is no reason 
to doubt that the parasites have a delicate ectoplasmic sheath, 
investing the body generally. 

I will leave until later the consideration of the minute 
structure of the trophonucleus. 

The above type of the parasite is the form which I have 
found in the blood of the host—at any rate, in the chaf- 
finches—during the winter and early spring months, when the 
numerical factor is low, the infection being, as it were, per- 
sistent, but in a quiescent and somewhat scanty condition. 

Young individuals, not yet full-grown, which belong to 
this ordinary definitive type, can be readily recognised. 
They are, of course, somewhat smaller, but their form and 
general appearance agrees in most respects with that of the 
adult parasites. The chief point of difference is that the 
“snout” is usually not so elongated and drawn-out; it is 
more conical, but still sharply pointed (figs. 31-33). ‘This 
aflagellar part of the body attains the extreme degree of 


668 H. M. WOODCOCK. 


attenuation only in the fully grown forms. An intermediate 
condition is seen in figs. 29 and 30. It will be noticed that 
there is often considerable variation in the size of the nuclei 
in these young or intermediate-sized individuals (cf. figs, 29— 
33), even where the parasites appear very similar in size and 
form. ‘I‘his feature is met with also in other series of forms 
to be described (see below, p. 672). I do not think much 
stress need be laid on apparent differences in size of these 
organelle in comparing parasites otherwise similar. 

Unfortunately, as already mentioned, I could not obtain 
any stained specimens of the parasite in the blood of the red- 
poll during that period, owing to its scarcity, although I 
had obtained cultures on two or three occasions. It was 
early autumn before I could obtain series of permanent 
preparations showing the trypanosomes in this bird; and in 
these smears, parasites which belong to the type above 
described are relatively scarce and outnumbered by another 
type. I have not found in this host at this period any 
ordinary forms which have attained quite the dimensions of 
the fully grown individuals occurring in the chaffinch in 
the early part of the year. The individuals observed, how- 
ever, correspond closely to the slightly smaller forms of the 
parasite, which have been described above (cf., for instance, 
figs. 33 with fig. 4, and, again, fig. 32 with fig. 31). Hence 
I have little doubt that they represent that phase of the same 
species, bearing in mind also the other considerations stated 
already. Itis probable that if I could have obtained examples 
of the trypanosome in the blood of the redpoll in the early 
part of the year I should have found “adult” definitive 
forms similar to those in the chatfinch. 

The predominating form of the trypanosome in the blood of 
the redpoll in the autumn (September, and again in October), 
is a very large parasite. Some of the individuals of this new 
type are, in all respects, the largest trypanosomes I have 
observed in the birds, being not only as long as the longest 
ordinary individuals, but also much stouter. ‘he individual 
drawn in fig. 37, for example, measures 48 yu in total length 


STUDIES ON AVIAN HAMOPROTOZOA. 669 


and 63 in total breadth, while that in fig. 35 is 44. by 6$ p. 
Even the rather smaller forms of this kind (figs. 34, 36, and 
38) are distinctly wider than the full-grown definitive 
parasites, their breadth varying from 5} to 6 yu. Hence, in 
general appearance these trypanosomes differ considerably 
from those of the first type. 

The aflagellar end is prolonged for some distance (6 to 
8 ») beyond the kinetonucleus; it may be fairly wide and 
somewhat blunt (fig. 36), or slender and tapering (fig. 37), 
but it is never so finely drawn-out and attenuated as in the 
case of the definitive individuals. The free flagellum is 
usually short, only about 4 to 43 mw long. The undulating 
membrane is well developed, but the folds or pleats are not 
usually so sharply separated from each other as in the case 
of the other forms. 

The cytoplasm of these massive forms stains blue, deeply 
and intensely. In structure it is quite different from that 
of parasites belonging to the other type. As a whole it 
is much coarser in texture and more granular. In the 
majority of cases it does not appear to be of uniform character 
throughout the body (figs. 34 to 36). In the aflagellar third 
or so of the body it is loose and spongy, with large granules 
more or less uniformly distributed ; but in the other two thirds 
or so, 1.e.1n the region from the trophonucleus to near the 
flagellar end, it is more compact, and the granules tend to be 
closely arranged in longitudinal rows, of which there are usually 
five or six. Thus the cytoplasm in this part of the body 
appears made up of narrow dark bands (composed of more 
prominent granules, packed together), with between them 
paler bands or zones of more finely granular (and hence less 
deeply staining) cytoplasm. The extent to which this serial 
arrangement of the larger granules is developed varies in 
different individuals. In some they extend through two 


1 There is no question of this difference being due merely to acci- 
dental variations in the staining; individuals representing the two 
types of form have been found on the same smear, and within a short 
distance of one another. 


670 H. M. WOODCOCK. 


thirds or more of the length of the body, while in others they 
occupy only the middle portion (fig. 35). Now and again 
these bands appear very narrow, but in no case can they be 
considered as lines or striations; I do not think they have 
any connection with, or themselves indicate, actual myo- 
nemes. Dutton and Todd (4) have described what is pro- 
bably a similar cytoplasmic differentiation in Trypano- 
soma mega and JT. karyozeukton. They distinguish 
the loose, spongy aflagellar region as ‘‘spongioplasm,”’ and 
the region of the longitudinal bands as “hyaloplasm.” The 
chief difference in their cases is that the dark bands are 
very broad and very compact, showing less obviously their 
granular structure, while the alternating, less granular zones 
are very narrow and pale, and appear as clear stripes. 

I have never seen any indications of division in any 
parasites belonging to either of the above types. 

The next series of forms of Trypanosoma fringilli- 
narum to be described consists, on the whole, of small 
parasites, some of which are extremely small. These forms 
have been found in two cases. The first instance of their 
occurrence noted was in the bone-marrow of a naturally 
infected chaffinch, which was killed about the middle of 
March. ‘This bird had the usual scanty number of ordinary 
definitive trypanosomes in the general circulation, and these 
are also present in the bone-marrow, along with the para- 
sites of small type. ‘lhe other case was in the chaffinch 
which was found to have a mixed infection of Halteridia as 
well as trypanosomes towards the end of June (cf. p. 660). 
In this bird the trypanosomes were comparatively numerous 
in the blood; but no individuals of the ordinary large type 
have been found in any of the preparations, whether from 
the blood or organs. As I shall frequently have to dis- 
tinguish between these two cases, it will be convenient, and 
will, I hope, render the description clearer, to refer to them 
as case A (the former, earlier case), and case B (the second, 
later case), respectively. 

T will begin the account of this small type of form by 


STUDIES ON AVIAN HAMOPROTOZOA. 671 


describing the parasites which occur in the later case (B). 
The smallest individuals have been found in the bone- 
marrow. The trypanosomes are distinctly Jess frequent in 
the bone-marrow than they are in the general circulation, 
and the individuals which do occur in this situation are 
nearly all small or minute in size. One of the smallest forms 
seen is drawn in fig. 40. Its total length is 15 uw, that of the 
free flagellum alone being 4 4; hence the length of the body 
itself is 11 wu. The width is a trifle under 24.4. It is only 
necessary to compare this parasite with some of those above 
described to realise the great difference in size which may 
be shown by different individuals of the same species of 
avian trypanosome. Another very small individual (fig. 5) 
has a total length of 18, partly accounted for by the 
rather longer flagellum of & mw, and its greatest breadth 
is Om. 

On the other hand, the largest individuals belonging to 
this series of forms which I have observed are seen in figs. 
44 and 45. The parasites are of only medium size; they do 
not really come inthe category of largeforms. The trypano- 
some of fig. 45 has a length of 354, its flagellum alone is 
84, and the greatest breadth is 514. The dimensions of the 
other individualare rather less. Between these two extremes 
of this type parasites of all intermediate sizes occur— 
forming, indeed, a regular gradation. This is illustrated by 
figs. 6, 42, and 43. The trypanosome in fig. 6, for 
instance, 1s 23 in total length, of which the flagellum is 
64 «, and has a width, including the undulating membrane, 
of 3%; again, the individual of fig. 43 is 27 long, the 
flagellum alone 6m, and the breadth 43 u. 

As will be noticed, there is a general similarity in form 
between all these parasites. ‘lhe body is fusiform or spindle- 
shaped, and fairly wide in proportion to its length; it is 
quite distinct in appearance from the body of a definitive 
individual. The aflagellar end is drawn out and pointed, 
but it is not so elongated and attenuated as in the case of the 
definitive parasites described above. In the smallest indi- 


672 H. M. WOODCOCK. 


viduals the undulating membrane is narrow and incon- 
spicuous (figs. 40 and 5), but with the increase in size of 
the body it becomes wider and more prominent. The kineto- 
nucleus may be relatively large, more particularly in the small 
individuals; in the parasite of fig. 5 it appears to be four- 
lobed, as if it were composed of four small masses. The free 
flagellum is fairly long, varying from 64 to 94 4. A modifica- 
tion of this type occurs, but it is very uncommon in this 
series; certain parasites are relatively very wide, and have 
the aflagellar end very short and abruptly conical, which 
gives the trypanosome a stumpy appearance (fig. 41). The 
dimensions of this individual are: Total length, 18}; of 
the flagellum alone, about 34,1; while the width is as much 
as 53 pu. 

Comparing now the small forms present in the earlier 
case (case A), the parasites are quite numerous in the: bone- 
marrow, and to this situation they appear restricted. They 
are of varying size, but I have not found individuals quite so 
minute as the smallest of those above mentioned, Parasites 
which are fairly small, nevertheless, are shown in figs. 46 
and 47, The former is 25 uw in length and 3} wide, the 
flagellum alone being as much as 9}; these two trypano- 
somes correspond fairly closely with that of fig. 6 from the 
other series, the chief difference being the longer flagellum. 
Here, again, it will be seen that there is considerable difference 
in the size of the kinetonucleus in the parasites compared. 
But on the same slide as the parasite of fig. 46, actually only 
two or three fields away, is another individual almost identical 
except that its kinetonucleus is nearly twice as large. Com- 
pare also figs. 44-and 45, and again, figs. 52 and 53. 

Rather larger forms are seen in figs. 49-51. Most of the 
parasites in this earlier case, however, are comparable rather 
with the wide, stumpy form alluded to above, than with the 
fusiform individuals. Typical examples are seen in figs. 
52-54, The parasite in fig. 54 has a total length of 27, 
the flagellum being 9, and its breadth is 53 to6m; the 
corresponding dimensions of the trypanosome in fig. 53 are 


STUDIES ON AVIAN HAIMOPROTOZOA. 673 


29u, 10u, and 644 respectively. The flagellum of these 
trypanosomes is usually comparatively long (from 9 to 114), 
being often longer than in the largest individuals of the 
fusiform kind. The kinetonucleus is always very near the 
aflagellar end, which is short and conical. The trophonucleus 
varies in shape; it may be more or less round, but it is often 
considerably elongated in a direction transverse to the longer 
axis (figs. 52-54). 

It is noteworthy that in this earlier case no forms have 
been observed which correspond to the larger fusiform 
trypanosomes of the other series (case B). The parasites, 
which are no longer very small—which are becoming inter- 
mediate in size—such as the individual drawn in fig. 50, are 
obviously approaching in character the wide, stumpy forms, 
and differ appreciably from the intermediate-sized indi- 
viduals of the fusiform variety in the features already 
indicated, namely, the broader body, the longer flagellum, 
and the abruptly terminating aflagellar part (cf. with figs. 
45,44, from the other case). 

Many of the individuals in the above-described series of 
“small”? parasites, including both fusiform and stumpy 
ones, show a cytological peculiarity which is at first some- 
what puzzling. This feature is a row or chain of granules, 
which take up the red stain strongly, and which are very 
closely apposed to each other, giving the idea of a thick, 
beaded line (figs. 42-44, 47-50, and 54). This chain runs 
approximately parallel to the flagellar border of the undulat- 
ing membrane, often following its curves closely, and it is 
frequently more deeply staining and prominent than the 
flagellar border itself. It begins near the origin of the 
flagellum, and always ceases with the limit of the body, at 
the opposite end, i.e. it never becomes free, as anything 
corresponding to a free flagellum. At first sight this line 
might be regarded as representing a new flagellum, formed 
either de novo or bya splitting of the old one, the parasites 
showing this appearance being therefore in the act of com- 
mencing division. After studying several of these individuals, 


674 H. M. WOODCOCK. 


it is clear, I think, that this structure has really nothing to 
do with a flagellum. ‘The line is usually most prominent 
in parasites which show numerous red-staining (probably 
chromatoid) granules in the cytoplasm; and, in suitable 
instances, it can be seen quite well that it is situated at the 
edge of the endoplasmic intrusion in the membrane (figs. 
42, 47,49). Further, when present, it can usually be traced 
right along the course of the membrane from end to end. 

If we had to deal here with a case of division or formation 
of a new flagellum, individuals showing either an earlier or 
later phase in the process might be expected to occur, for 
this appearance is not at all infrequent; but I have not 
found any such. Again, in most cases, there is not the 
least indication of nuciear division. Lastly, in one of the 
exceptionally few instances where any indications of division 
are present, in addition to the kinetonucleus having divided 
into two, the true flagellum can be seen to be itself double 
for a short distance near its proximal end, probably as a 
result of splitting (fig. 54)... The granular chain is also 
present, and, as before, quite separate from the flagellum. 
Hence there is no reason for regarding this structure as in 
any ,way connected with a flagellum, much as it simulates 
one at times. 

The small stumpy trypanosome in fig. 41 shows what is 
probably an early stage in the development of this line. 
Here there is a row of red-staining granules, quite separate, 
and not closely apposed to constitute a chain, which run 
parallel to the flagellar border, doubtless at the limit of the 
endoplasm. ‘The granules are apparently quite similar to 
others which are seen in the general cytoplasm. I| have no 
idea what is the explanation of this aggregation of chroma- 
toid granules into a compact chain, lying in the position 
described. I have never seen it either in the ordinary 
definitive trypanosomes or in parasites of the other large type. 
I may add that I have observed the same feature in the case 
of a trypanosome from a blackbird (l'urdus merula), at 


1 Cf. also the micro-photograph reproduced in fig. D. 


STUDIES -ON AVIAN HAIMOPROTOZOA. 675 


Rovigno, the parasites which showed it being also of the 
same type of form. 

There is still another variety of form to be mentioned, 
which occurs in case A (in the bone-marrow). This is a 
fairly small trypanosome (figs. 55 and 56), which is very 
narrow in proportion to its length. The aflagellar end is 
comparatively long and finely drawn out, and may approach 
the attenuated condition. The flagellum is fairly short, and 
the undulating membrane has well-developed folds. The 
dimensions of the individual in fig. 55 are: total length, 27 n, 
breadth (including membrane), 3, and length of flagellum 
65. The kinetonucleus is relatively large. These parasites 
strongly resemble in appearance young ordinary or definitive 
trypanosomes. 

With regard to the multiplication of these small forms the 
only evidence I have been able to obtain is very slight. 

I have observed three or four individuals (and not more) 
of the wide stumpy kind from case A, in which the kineto- 
nucleus is in two parts (figs. 48 and 54) ; and in one solitary 
instance, just alluded to, the flagellum is partially doubled. 
In no case have I seen two trophonuclei. The condition in 
fig. 48 is the nearest approach to trophonuclear division that 
I have observed ; this may represent commencing division 
because other organelle of this parasite are dividing, The 
flagellum has not yet begun to divide, but asa prelude thereto, 
the centrosomic granule at its proximal end (“blepharoplast ”) 
is clearly double. So far as the fusiform series (of the other 
case) is concerned, I have observed absolutely no signs of 
division at any phase. 

General Remarks.—The significance and relation to 
each other of all these manifold forms of the trypanosome is 
a somewhat difficult question. Where transitional forms 
or division phases occur they afford, of course, consider- 
able help. Beginning with the small forms, the stumpy 
parasites of case A, in which indications of division can be 
found, probably give rise, as a result of that process, to small 
individuals like those in figs. 46 and 47, which grow into 

VOL. 55, PART 4.—NEW SERIES. AD 


676 H. M. WOODCOCK. 


somewhat larger individuals of the fusiform type (figs. 49 
and 51). The stumpy trypanosomes themselves are best 
regarded, I think, merely as division-forms of young to 
medium-sized individuals of fusiform type. Hence, in this 
case, it may be said that the fusiform parasites present are 
of small to medium size and tend to multiply, by passing 
into the stumpy division form, rather than grow, at any 
rate at this period, into large trypanosomes. Next, with 
regard to the very thin, slender forms (e.g. figs. 55 and 56) : 
when first seen they appeared in such sharp contrast to the 
prevailing stout type of parasite that I was somewhat dis- 
posed to think they represented male forms. As above 
mentioned, however, [am now more inclined to look upon 
them as young definitive parasites, which -would grow into 
medium-sized ones, such as those in figs, 4 and 31, and so 
to full-grown adults, as in figs, 2, 28 (all from this series). 
‘Turning again to case B (the later case), we find no 
ordinary forms present. Fusiform individuals of mediuin 
size are not uncommon, and between these and very small 
forms parasites of all intermediate sizes occur. There are 
very few stumpy forms, and none of those found show any 
actual signs of division.! Hence, the main condition here is 
undoubtedly a series of steadily growing fusiform individuals. 
There remain two or three interesting questions in con- 
nection with the different type or phase of the infection 
occurring at different periods, in regard to which I can only 
put forward those surmises which seem to me the most 
probable. In the first place, comparing the condition found 
in a chaffinch (case B), in the summer, with that obtaining in 
a redpoll in the early autumn, where the parasites are mostly 
of the large massive type (e.g. figs. 34-36), I think it is 
most likely that the fusiform parasites of the former case 
(such as those of figs. 44, 45), would grow ultimately into 
individuals corresponding to those of the latter. The body- 
form is essentially similar in the two cases. The size of the 


1 Tt is possible, however, that the two or three small stumpy indi- 
viduals seen in this case may be about to divide. 


STUDIES ON AVIAN H@#MOPROTOZOA, 677 


parasites found in the autumn is of course greater, but the 
difference is not relatively more than that between the larger 
and the smaller fusiform individuals in the summer. A 
difference which might appear of more importance is that in 
the character of the cytoplasm in the two cases, This can 
probably be explained, however, by supposing that the 
cytological features shown by the large massive individuals 
in the redpoll have become more developed and consequently 
more prominent, as a result of the increase in size, And, on 
the other hand, there is no evidence whatever that the 
fusiform parasites will pass directly into the characteristic 
ordinary type. 

Assuming, then, this connection between these two 
sets of forms, how are we to explain the condition met 
with in the winter and early spring, when the only type 
of individual in the blood is the ordinary definitive form ? 
The answer to this depends largely, I think, on what 
significance is to be assigned to the large massive forms just 
referred to. Are they to be considered as sexual individuals 
—of the female type? This is, of course, possible, but more 
than that cannot be said. And if this is the case, I certainly 
do not know which are the individuals of male sex; there do 
not appear to be any forms present at the same time which 
could be so regarded. On the other hand, I think it is at 
least quite as probable that the massive individuals have 
grown to this size prior to multiplication; they may later 
undergo some process of multiple fission or segmentation, 
occurring in one of the internal organs, and so give rise to 
the small forms. This supposition would fit in very well with 
the condition found, for instance, in case A (in the spring), 
where, as we have seen, small parasites are numerous in the 
bone-marrow, along with the ordinary forms, tlie latter being 
probably to some extent replenished from them. And here 
there are no signs of the large massive individuals. At all 
events, in view of Chagas’ recent important work (2), showing 
that a new human trypanosome, Schizotrypanum cruzi, 
has a method of multiplication by multiple fission or 


678 H. M. WOODCOCK. 


schizogony, I think it is not at all unlikely that naturally 
occurring trypanosomes—about whose life-cycle in the Verte- 
brate host very little is yet really known—may show some 
such schizogonic process more commonly than has hitherto 
been supposed.' In default of such a process in the present 
case, I have no idea how the small forms are developed, since 
they certainly do not appear to be derived from the adult 
ordinary individuals. 

Another question is, What becomes of the ordinary, definitive 
forms of the trypanosome? As I have obtained many suc- 
cessful cultures from birds where this was the only type 
present in the blood, the natural inference would be that 
this form can be transmitted to the insectan host; but the 
same applies equally, it must be noted, to the fusiform 
parasites of case B, since I obtained cultures from them also. 
And I cannot be certain that both these types would develop 
naturally in the insect. Some of the ordinary forms, later 
on in the season, may pass into the large, massive type; this 
is not at all unlikely, if the latter is really a multiplicative 
form. The individual drawn, for instance, in fig. 39 may 
perhaps represent an intermediate stage in such a transition. 
Another possibility, of course, is that this definitive type 
disappears altogether in the summer, its place being taken 
by the fusiform type; the condition of the infection would then 
correspond with that of case B. I do not think this is likely. 
Case BK most probably represented a recent infection (see 
below) ; in such the condition may quite likely differ from that 
found in an old established infection. Moreover, in the earlier 
case A (about the middle of March), parasites of the ordinary 
type are quite numerous, and do not look like disappearing; 
and further, in the autumn, in the redpoll, this type is also 
present. 

It remains for me to say a few words with regard to the origin 

1 A most interesting piece of evidence bearing upon this point is 
supplied by Minchin (12), who mentions and figures the occurrence of 
a large individual of T. perce, which is apparently in an eneysted 
condition. Such a form might very well be about to undergo schizogony. 


STUDIES ON AVIAN HAMOPROTOZOA. 679 


of the infection in this later case B, the chaffinch in which 
there was also an abundant halteridial infection. As I have 
stated in my note (88) on this interesting Halteridium, | 
was at the time inclined to think that the very small trypano- 
somes might have been developed directly from the 
Halteridia. Paying attention, for the moment, only to the 
trypanosome side of the question, in addition to the faet that 
in this case we have certainly to do, not with division, but 
with growth and increase in size from the minute forms up to 
comparatively large ones, there were other reasons which led 
me to take this view. This chaffinch, originally free from 
trypanosomes, was inoculated with cultural forms, but the 
subsequent course of events was very different from that in 
the case of the other successful inoculation described. Inthe 
latter case the parasites soon became comparatively numerous 
in the blood, whereas in the former they were not found at 
all at first, and only after some weeks were they shown to 
be actually present, by tubing (for further details, cf. p. 660). 
When at length they did become sufficiently numerous to 
be found without difficulty in stained preparations, they 
proved to be, as we have seen, quite different in form from 
the ordinary individuals developed in the other case. Hence, 
taking all things into consideration, I considered that the 
trypanosome infection was probably not due to the inocula- 
tion (which, in several cases, it must be remembered, did fail), 
but to the presence of Halteridium. 

I admit now that I have changed my opinion about this 
case since writing my former note. In spite of the many 
features which seemed either to point strongly to this view, 
or at least to favour it, I think after all the trypanosome 
infection was not really connected with the Halteridial one, but 
was due to the inoculation (for further discussion of this subject, 
see under Halteridium). There remains the question, Why 
was the course of the infection so different in the two cases? 
Of course, in the one case where the parasites developed 
quickly, the inoculation was made with cultural forms which 
had come from a redpoll, while in the other they came from a 


680 ; H. M. WOODCOCK. 


chaffinch; but I do not think this sufficiently explains the 
difference, because everything points to the species being the 
same in both birds. Since I have been able to study my 
cultural forms, I have come to the conclusion that the pro- 
gress of the infection may have been so different, on account 
of a difference in the condition of the two cultures. The 
chaffinch-culture, from which resulted, we must suppose, the 
slowly developing infection, was one of six days’ age, and 
certainly coutained the characteristic trypaniform individuals 
to be subsequently described (cf. below, p. 690) ; for permanent 
preparations were made at the same time which showed this 
type. On the other hand, the redpoll culture used in the 
other (earlier) case was a fairly old original one of fifteen 
days; preparations were not made from this culture actually 
on the day when it was used for inoculating the bird, but im 
smears taken a couple of days before, none of these forms had 
been seen; the culture appeared quite healthy, and consisted 
almost entirely of the usual trypanomonad forms, to which, 
presumably, the infection must be ascribed. 

It is an interesting question in which of these cases the 
course of the infection, so very different in the two, more 
nearly resembles that occurring naturally, i.e. by the inocu- 
lation of the right developmental forms from the insect. 
As will be seen on reference to one or two papers discussed 
below (p. 709), the remarkable trypaniform type alluded to 
is thought to be probably the true propagative form, which 
produces the infection of the vertebrate host. If this is so, 
it would seem to follow that the later case (case B), where 
the infection developed slowly, agrees most with the course 
of events in a natural infection. 


(s) The Trypanosomes as Found in Cultures. 


Before beginning an account of the cultural forms, one 
or two introductory remarks are necessary. When I com- 
menced to make use of the cultural method, I did so solely 
because, from Novy and MecNeal’s work (14), it was evident 


STUDIES .ON AVIAN HAMOPROTOZOA. 681 


that it is of very great service in ascertaining whether a 
bird is infected with trypanosomes or not. I think now 
that I must have been unusually fortunate in my _ first 
experiences of the culture method.! [ had no difficulty in 
getting the parasites to develop in my cultures, and, more- 
over, in a perfectly healthy manner. I soon had no trouble 
in distinguishing between what could be regarded as normal 
types, of regular occurrence, and what were abnormal, irre- 
gular forms. Hence, I admit that I modified my former 
attitude towards this method, and came to the conclusion 
that the cultural forms were probably, for themselves, well 
worth studying. I claim some excuse for my earlier opinion, 
since at that time this method had only begun to be adopted 
for trypanosomes, and in the early descriptions of cultural 
forms most of the figures depict what can only be described 
as altered appearances, which certainly belong to the cate- 
gory of abnormal phases. Asa result of my own work, the 
view I now hold, and which I have expressed in my article 
in Lankester’s ‘ Protozoa’ (39), is that the cultural forms of 
trypanosomes may afford indications of value as to the 
developmental phases of the parasites occurring in the 
invertebrate host. 

As I have already indicated, the chief cultural forms 
developed from the trypanosomes in the redpoll are quite 
similar to, and practically indistinguishable from, those to 
which the parasite from the chaffinch gives rise. I have 
had, however, a much greater number of successful cul- 
tures from the latter bird than from the former; hence I 
have found a greater variety of intermediate phases in my 
cultures from the chaffinch, and have had the good fortune, 
moreover, to observe one or two particular phases which I 
have not seen in cultures from the redpoll. This is doubt- 
less due, however, merely to lack of sufficient material in the 


1 T may mention incidentally that I have since had a full measure of 
the trials and troubles which may attend the cultural method, for at 
Rovigno, in connection with the trypanosomes of the little owl, I had no 
success at all with it. 


682 H. M. WOODCOCK. 


latter case, and I have no reason whatever to think that one 
set of cultural forms shows any intrinsic differences from the 
other, which would imply that the trypanosomes from the 
chaffinch and the redpoll, respectively, are distinct parasites. 

The predominating type of the trypanosome in the cul- 
tures is a well-defined and characteristic form, which may be 
termed the trypanomonad form of the parasite, deriving 
this convenient general designation from one of the various 
alternative (synonymous) names (viz. ‘I'rypanomonas) 
given by Danilewsky to certain parasites described by him. 
This type is elongated and slender, the width usually varying 
but slightly in the middle of the body, and diminishing more 
or less gradually towards the aflagellar end. The essential 
diagnostic characters are: (1) ‘he two nuclei are always 
close together, and situated either about the middle of the 
body, or else distinctly in the aflagellar half; and (2) the 
flagellum is attached tor some distance to the side of the 
body, forming a distinct undulating membrane. The mem- 
brane may be at times fairly prominent, and possess a wavy 
edge, indicating a slight development of pleats or folds. 
The kinetonucleus is never near either end of the body. It is 
important to note that the flagellar end of the body is drawn 
out with the flagellum, as it were, and ultimately thins away, 
leaving the flagellum free. ‘lhis condition is of very general 
occurrence, of course, among trypanosomes (as seen in the 
blood), and is the natural consequence of the presence of an 
undulating membrane. In respect of all the above features, 
therefore, the trypanomonad type differs essentially from a 
herpetomonad form. 

‘Typical examples of the trypanomonad form, showing para- 
sites of medium to large size, are seen in figs. 7, 8, 71-75, and 
figs. 13, 77-79, from preparations of cultures from the chaf- 
finch and redpoll respectively. To give an idea of the size 
of these forms, three principal measurements may be taken: 
(4) length of body alone, (8) greatest width of body, and 
(c) length of free flagellum. These dimensions, in the case 
of some typical individuals, are as follows (in uz) ; fig. 72—(a) 


STUDIES ON AVIAN HAMOPROTOZOA. 683 


21, (B) 3, (c) 9; fig. 73—(a) 25, (8) 24, (c) 10; fig. 7—(a) 29, 
(pg) 24, (c) 29; fig. 75—(a) 26, (B) 34 (opposite nucleus) ; (Cc) 
11; and again, fig. 79—(a) 21, (B) 24, (c) 15; fig. 77—(a) 
23, (B) 3, (c) 19; fig. 88—(a) 26, (B) 34 (opposite nucleus), 
(c) 14. The measurements are given in a slightly different 
manner from that adopted in the case of the parasites when 
in the bird. In the cultural forms the length of the body 
by itself affords a better means of comparing the size of 
different individuals than the length of the body plus that 
of the flagellum. This is because of the great and 
apparently indiscriminate variation in the length of the 
flagellum, which cannot be said to bear any relation to that 
of the length of the body. ‘his is well seen by contrasting 
figs. 80 and 81, from a redpoll culture, with figs. 84 and 83, 
respectively, from achaffinch culture. This diversity is chiefly 
due to the manner of division, as will be explained shortly. 

Smaller forms, very similar in appearance to some of the 
larger ones indicated, are seen in figs. 85 and 86; the 
former is 17 by 1? and its flagellum 73. ‘The smallest 
parasites observed, however, belong to, or result from, a 
slightly modified variety of the above type. This is somewhat 
different in appearance (figs. 8, 97), but it really represents 
only another facies, as it were, of the same trypanomonad 
type, from which it is derived by the gradual drawing back 
of the nuclei well into the aflagellar half of the body, and by 
a somewhat modified manner of division which is then 
found (concurrently). 

As the process of multiplication plays an important part in 
the development of these various forms, it may be as well to 
give a general morphological description of it here before 
proceeding farther. ‘I'he mode of division by which the 
long, slender trypanomonad forms are produced is that of 
equal or subequal fission of the body. Sometimes the two 
daughter-flagella are practically equal (figs. 11, 96), but in 
the majority of cases one of the flagella is distinctly longer 
than the other (figs. 91-95). In all the instances I have 
noticed, the division of the cytoplasm begins. at. the flagellar 


684. H. M. WOODCOCK. 


end, It generally happens that, as the split extends, the 
parasites tend to separate from one another, turning out- 
wards, away from each other as it were (figs. 92, 93); eventu- 
ally the two daughter-individuals come to lie in one line (which 
may be more or less curved), with the flagella,waving freely 
at opposite ends, the parasites only remaining connected by 
what is actually the still undivided aflagellar end (figs. 
94,95). The fact, therefore, that we may find either equal 
or sub-equal cytoplasmic division in which the daughter- 
flagella differ considerably in length, explains the great 
variation in this respect which is met with among the 
ordinary trypanomonad individuals. 

In many cases the division of the two nuclear bodies does 
not take place in a direction quite transverse to the long 
axis of the body, but in an oblique direction, one pair of 
daughter-nuclei lying somewhat nearer to the aflagellar end 
than the other pair. In this manner are produced forms 
such as are seen in figs. 8,97, and 100. These individuals 
in which the nuclei have progressed into the aflagellar 
part of the body have the undulating membrane very well 
developed; it may be said that the trypanomonad condition 
is here accentuated. In such forms of the parasite the 
mode of division is also distinct, being markedly unequal in 
character (figs. 98-100). The two resulting individuals are not 
of the same type (cf. figs. 12, 103, and 107). One, the larger 
parasite, is of the same type as the parent individual, and 
possesses from the first a conspicuous membrane, but the 
other, the smaller daughter-individual, is at first pear-shaped 
and stumpy, and has only a short, inconspicuous membrane. 
This mode of division presents a general resemblance, it will 
be noted, to one of the types of division characteristic of 
T. lewisi. Indeed, in the present case, the process might 
also be regarded as a “budding-off” of a daughter- 
individual from the parent. I have never observed, how- 
ever, more than one bud formed, i.e. the process appears 
always to retain its character of binary fission and never 
to be of the multiple type. When set free, the smaller 


STUDIES .ON AVIAN HAMOPROTOZOA. 685 


daughter-individual elongates a little and becomes spindle- 
shaped instead of pyriform; the membrane also becomes 
more conspicuous. I have not seen any transitional phases 
between these fusiform individuals and the type represented 
by the parent form, and have therefore no indications as to 
whether they (the former) grow or otherwise pass into the 
accentuated trypanomonad type again. By successive multi- 
plication according to this manner the size of the parasites 
become considerably reduced. In fig. 102 is seen a very 
small couple of the kind described. Examples of free 
parasites, of different sizes, representing accentuated trypano- 
monad daughter-individuals are given in figs. 105 and 
108-110, 108 being from a redpoll culture, the others from 
a chaffinch one. The smallest form (fig. 110) is 10}, long, 
its flagellum is.13 4, and its breadth is 2}. The small fusi- 
form parasite of fig. 111, representing a pyriform daughter- 
individual, is 9 u long, its flagellum is 7}, and its width 23 mu. 

The great majority of the parasites in thriving cultures 
belong to the above-described types, After a fresh culture- 
tube has been inoculated (from a bird) about five days, 
by which time the trypanosomes have generally multiplied 
sufficiently to ensure that there will be a fair number of 
parasites on a permanent smear—in other words, that. an 
individual can be found without much searching—practically 
all the parasites present conform to the trypanomonad type. 
And up to the end of a week or so this type persists with 
great constancy, notwithstanding the rapid multiplication. 
The only variations that are numerically important are those 
already indicated, in the direction, that is, of an accentuated 
trypanomonad type and of a fusiform one. Further, if a 
sub-culture of these normal forms is made (preferably not 
later than the seventh or eighth day) the development of 
similar forms continues steadily in the sub-culture. Thus 
the parasites drawn in figs. 74, 109, are on a preparation 
from a.second sub-culture, and the total interval that had 
elapsed since the blood was originally taken from the bird 
was twenty-six days, or over three weeks. 


686 H. M. WOODCOCK, 


Certain other phases or developmental forms of the trypa- 
nosomes, however, have been encountered in cultures which 
were in a normal healthy condition, but these have been, 
as a rule, scanty in number, contrasting markedly with 
the abundance of the prevailing types. In cultures of 
six or seven days’ age or more a small percentage of the 
individuals—and usually only a very small percentage—show 
a tendency to lose the fusiform or more active type of form, 
and to develop a pear-shaped or rcunded, more passive type 
of form. In most of my culture-series (including sub- 
cultures), these pyriform or ovoid forms are very infrequeut 
and have to be carefully searched for, even on slides where the 
ordinary parasites are most abundant. The individuals of 
this character are generally of medium, or less than medium 
size, but occasionally are large and massive. Pear-shaped 
forms are seen in figs. 112-114, that of fig. 115 being from a 
redpoll culture,the others from different chaffinch ones. ‘The 
dimensions of these parasites (flagellum excluded), are, for 
example, 8a by 5 (fig. 113), and 634 by 3% (fig. 114). 
Medium-sized ovoid forms are 8 by 6 (figs. 116 and 117). 
The large ovoid individual of fig. 118 is 13. by 7. and has 
a very long flagellum of 24 4; the small corresponding form 
(fig. 115), is 6u by 4. Although I have distinguished 
these parasites as more “ passive” forms, it is difficult to 
know whether to regard them as being about to enter ona 
‘‘ resting-phase,” for in all cases where | have observed them 
in what were normal, healthy cultures, these individuals 
possessed a flagellum. I may mention here that the only 
instance where I have found rounded-off parasites which 
lacked a flagellum was in a culture (original), nineteen days 
old, which was full of atypical, altered forms (cf. below, 
p. 696). 

This type of parasite, whether pyriform or ovoid, to 
rounded, is almost certainly to be derived from forms in 
which the alteration in nuclear position has occurred, and in 
which the modified method of multiplication, by unequal fission, 
has made its appearance. Pear-shaped individuals, such as 


STUDIES ON AVIAN HASMOPROTOZOA. 687 


those of figs. 112 to 114, are probably simply the smaller 
daughter-individuals which have retained the pyriform shape, 
instead of taking on the fusiform, more active one. On the 
other hand, most of the ovoid or rounded forms, especially 
where they are of medium to large size, would seem to 
arise from the accentuated trypanomonad type of daughter- 
individual, ‘Transitional phases can be found, showing 
different degrees in the retraction of the drawn-out flagellar 
end and the concurrent reduction or disappearance of the 
undulating membrane. Thus, both the large and the small 
ovoid individual (figs. 115 and 118) have still a delicate but 
distinct continuation of the body along the proximal part of 
the flagellum, which doubtless corresponds, for the most part, 
to undulating membrane. And in others of these rounded 
forms, indications of the original membrane are still afforded 
by the attachment of the flagellum to the side of the body for 
some distance, the flagellum curving with it—at times partly 
curling round it, as it were—before becoming free (fig. 119). 

In general these rounded forms of the parasite do not, 
apparently, undergo division, In most instances where I 
have observed these forms, they are, as I have mentioned, of 
small or only medium size, and these never show indications of 
division. One of my culture-series, however, for some reason 
or other for which I was unable to account, but which was 
probably due to some variation in the condition of the culture 
medium, behaved differently from the usual manner. In this 
culture a pronounced tendency in the development of the 
parasites was the production of large, massive forms, which 
are sometimes ovoid or rounded inshape. Examples are seen 
in figs. 120-123. The parasites in my preparations of this 
series (taken when the culture was seven days old) are 
certainly not degenerate or abnormal; this is clearly shown 
by a comparison of their structure with that of. distinctly 
atypical or degenerate forms (cf. below, p. 693). ‘There is 
none of the irregular multiplication of organelle, nor of the 
alteration in the cytoplasmic constituents which is apparent 
in the latter. I consider that the unusually large proportion 


688 H. M. WOODCOCK. 


of broad or ovoid massive forms in this series was probably 
due to a greater growth activity than was usually met among 
the cultural parasites. And just in this case, it is interesting 
to note, I have found not infrequently various stages of 
division in ovoid or rounded individuals (ef, figs. 121, 123- 
125). Making allowance for slight differences due to the 
more massive form, the process appears to follow,in the main, 
the unequal method of fission. In all these forms, whether 
dividing or single, it may be as well to state, the flagellum 
was present; none of them showed any signs of absorbing or 
otherwise losing this organella. 

The next type of cultural form of the parasites which I 
have to describe is quite distinct from the preceding ones, 
being markedly trypaniform. By the term trypaniform is 
understood the condition characteristic of a trypanosome, 
where the kinetonucleus lies much nearer to the aflagellar 
end of the body than does the trophonucleus, and where, 
consequently, the flagellum is attached by an undulating 
membrane along the greater part of the length of the body. 

In my cultures I have found trypaniform phases, differing 
slightly in character, at two different periods of the develop- 
ment. As regards one case, | came across this type of the 
parasites rather accidentally as it were, in the following 
manner. I inoculated culture-tubes from the chaffinch which 
had a strong halteridial infection, in addition to small forms 
of Trypanosoma fringillinarum, in the peripheral circu- 
lation: These culture-tubes were examined much earlier 
than it was my custom to do, namely after forty hours had 
elapsed. This was not on account of the trypanosomes, as 
I knew from former experiences that at this early period 
they would probably not have multiplied sufficiently for me to 
be able to find an individual on a smear without prolonged 
searching ; it was because I wished to see what development, 
—if any—was undergone by the halteridia in the culture! In 
examining a good living drop to see if I could find any halte- 
ridial odkinetes, I noticed one or two trypanosomes which 


1 See below, p. 727. 


STUDIES ON AVIAN H#MOPROTOZOA. 689 


were very active, travelling much more rapidly than was 
customary in the case of these cultural forms. In the course 
of looking for halteridia on a permanent smear (made at the 
same time), I happened very fortunately to come across a 
trypanosome, and this was so different from the usual trypa- 
nomonad type that I subsequently examined my preparations 
of this series thoroughly to ascertain whether this was the 
prevailing type. Unfortunately the trypanosomes are very 
scarce, only three or four on a large film. It is noteworthy, 
however, that all the parasites seen as a result of systematic 
searching are in the same trypaniform phase, and show only 
slight individual variations. 

The typeis extremely thin and slender, the parasite having 
a distinctly vermiform appearance (figs. 10,126, and 127). The 
body is from 21 to 25, in length, excluding the flagellum, 
and its greatest breadth only from 1} to 1}. The aflagellar 
region is very long and finely tapering. The kinetonucleus 
is far removed from the trophonucleus, and generally hes 
about midway between the latter and the aflagellar extremity. 
Its actual distance from this end varies from 6 to 9 4, depend- 
ing upon the degree of attenuation. ‘lhe undulating mem- 
brane is in most cases very narrow, and practically distin- 
guishable only by its flagellar border. In some individuals 
the flagellar border originates, not in close proximity to the 
kinetonucleus, as is usually the case, but from a point some 
little distance beyond,i.e. on the aflagellar side of the 
kinetonucleus (figs. 10, 127). A distinct granule (blepharo- 
plast or basal granule) can often be made out at its com- 
mencement. ‘I'he length of the free flagellum is trom 8 to 
11. The trophonucleus, instead of being the usual shape, 
namely, oval or rounded, is considerably elongated in the 
long axis of the body, this being in relation, in all probability, 
with the narrow form, 

‘The other instance of the occurrence of parasites of a 
trypaniform type in my cultures wasin a series from a six-day 
(original) ‘culture of the chaffinch-form, taken when the 
trypanosomes, of the ordinary, definitive type, were very 


690 H. M. WOODCOCK. 


scanty in the blood. The parasites are numerous, nearly all 
being, of course, in one or the other variety of the trypano- 
monad phase. Exceptionally, however, individuals occur 
which show the trypaniform condition; for example, on a 
smear containing between two and three hundred parasites 
there are four or five such, three of which are drawn in 
figs. 129-131. I have not found any which correspond 
exactly to the individuals of this type just described. The 
parasite in fig. 129 approximates fairly closely to those of 
figs, 126 and 127, but it is distinctly shorter and relatively 
not quite so slender. ‘The two other individuals, on the other 
hand, while altogether much larger, are still very slender in 
proportion to their length ; and in these the aflagellar part is 
very prolonged and vermiform. While agreeing in general 
form and character with the parasite, for instance, of fig. 10, 
they represent, it would seem, an older, later condition. The 
individual of fig. 130 has attained, probably, the fullest deve- 
lopment of this type, at least as far as the culture is concerned ; 
it constitutes, I consider, a most important phase. 

The length of the body alone is 36, and its greatest 
width 2; the distance of the kinetonucleus from the 
aflagellar extremity is 1l;. The free flagellum is only 84 
long. The trophonucleus of this individual presents a 
remarkable appearance (fig. 130). The chromatin is arranged 
in a series of short transverse bars, forming a longitudinal 
row—hence the description “ladder-like.” I have found a 
quite similar condition in two other examples of this type; 
but in the other jarge vermiform individual I have figured 
(fig. 131) the chromatinis not arranged in suchadefinite ladder- 
like manner, but appears to form a fairly regular double row 
of grains. 

None of the trypaniform parasites which I have found—in 
either case—showed any indications of division. 

The types above described include all the cultural forms of 
the trypanosome observed, which I have no hesitation in 
regarding as perfectly normal and regular. As I shall 
mention more particularly later, they are closely paralleled 


STUDIES ON AVIAN H#®MOPROTOZOA. 691 


by flagellate forms known to occur in various blood-sucking 
invertebrate hosts. 

I may now contrast with them certain other cultural forms 
found, most of which I have equally little hesitation in con- 
sidering as abnormal or atypical forms, developed by the 
parasites asa result of unfavourable conditions in the medium. 
These forms are found in old, original cultures of, say, 
twelve days or more, in which multiplication has gone on toa 
very great extent. It must be borne in mind that such a 
medium no longer corresponds at all to any condition met 
with in an insectan host. In an insect, the digestion of the 
imbibed blood—the medium of the parasites—and its absorp- 
tion are completed in the course of a few days at most; by 
this time the parasites remaining in the digestive tract have 
passed into the resting, attached phase. In an old culture, 
on the other hand, the fluid medium is still present, presumably 
containing a certain amount of nutriment of a kind, but now 
considerably altered in character by the addition of waste 
products of the metabolism of the parasites, which have 
doubtless a deleterious action on the trypanosomes. In sub- 
cultures made at sufficiently short intervals, these abnormal 
forms are usually not found at all. In this case it is as if the 
transferred parasites remained continuously ina pure medium, 
which may be looked upon as a substitute for the medium 
in the stomach of the insect—at any rate during the early 
period of digestion. 

A most interesting feature of the morphology of these 
forms is that very few of them show the trypanomonad phase; 
nearly all the parasites have passed into a more or less 
herpetomonad-like condition. The earliest indication of an 
alteration in the character of a culture is afforded by the 
appearance of such forms. They are to be met with in 
cultures of ten or twelve days and onwards. At first, of 
course, these individuals are very few in number. 

Examples of this “ pseudo-herpetomonac 
propose to term it, are seen in figs. 140-146; figs. 140, 145, 
and 146 are from a chaffinch culture of twelve days; fig. 141 

VOL. 00, PART 4,—NEW SERIES. 46 


” condition, as I 


692 H. M. WOODCOCK. 


is from a redpoll culture of nine days, and figs. 142-144 from 
one of nineteen days. The body is fusiform to long and 
slender in shape. The two nuclei are situated distinctly in 
the flagellar half of the body; they lie usually fairly close 
together. The appearance of the flagellar end of the body 
and its relation to the flagellum is in general intermediate 
between that found in the trypanomonad type and that in a 
typical herpetomonad form. The flagellum itself is only 
connected with the body for a comparatively short distance, 
and is usually not obviously attached along one side of the 
body to any extent (figs. 140, 141, 143, and 144) ; hence there 
are no indications of an undulating membrane. This proximal 
portion of the flagellum is, in the majority of cases, chiefly 
intra-cytoplasmic, constituting simply a rhizoplast, and corre- 
sponding to the rhizoplastic part of the flagellum in the 
try panomonad forms (before it passes to the surface to become 
the border of the membrane). On the other hand, the flagellar 
end of the body, while sometimes fairly sharp and acute, 
approximating to the condition in an ordinary herpetomonad 
(cf. figs. 140, 141, and 147), may taper more or less gradually 
(figs. 142, 144,and 145); hence, in these cases, where it is drawn 
out a little with the flagellum, the latter may be regarded as 
“attached” for a short—or very short, distance. For this 
reason, and because the two nuclei are closer together than is 
customary in a herpetomonad, this condition is preferably 
distinguished as pseudo-herpetomonad. The difference will 
be readily understood when it is remembered that all these 
individuals are derived from trypanomonad forms by the 
more or less complete loss of the undulating membrane and 
its attached flagellar border; hence, of course, parasites 
showing all manner of intermediate stages in the process are 
to be met with. 

In the early formed individuals of this pserdo-herpeto- 
monad variety there is nothing about them to indicate that 
they are actually abnormal or unhealthy. As I shall discuss 
subsequently, however, I think it is very probable that the 
occurrence itself of this unusual condition is the consequence 


STUDIES ON AVIAN H#MOPROTOZOA. 693 


merely of the unusual environment; [I am very doubtful 
whether it can be regarded as representing a normal phase 
of the life-cycle. In any case, however, as the age of a 
culture increases, and these forms multiply and predominate 
—the trypanomonad phase as quickly declining—numerous 
irregular forms of the parasites are met with, which are 
manifestly unhealthy. As might be expected, the form and 
size of these individuals varies considerably (cf. figs. 145— 
154, taken either from a twelve-day chaffinch culture or from 
a nineteen-day one from a redpoll). Some of them are long 
and narrow, others pear-shaped, while others are large and 
massive, ovoid, or of ill-defined shape. 

The abnormal condition of these forms is_ particularly 
indicated by certain cytological characters, which I have 
never observed in normal individuals. A common feature is 
the occurrence of a peculiar altered appearance in the neigh- 
bourhood of the rhizoplastic part of the flagellum. Some- 
times there is a cluster of red-staining granules in this region 
of the cytoplasm (figs. 145, 146). In the more massive 
forms there is usually a greater or less amount of a diffuse, 
indefinite substance, which also stains red. This substance 
is often more or less streaky in form, one or more streaks 
commencing in the neighbourhood of the rhizoplast and 
running backwards in the cytoplasm for a short distance 
(figs. 150, 151, and 153). Ina few individuals the streaky con- 
dition is combined with the occurrence of the granules (fig. 152). 
I am unable, unfortunately, to offer any certain explanation 
of this interesting character, owing to the fact that I have 
only had material stained with Giemsa in which to observe 
it ; very likely the appearance is different after other methods 
of staining. So far as the granules are concerned, they do 
not differ in their staining reactions from the ordinary 
chromatoid granules which are often found in normal 
trypanomonad types; the latter, however, are scattered more 
or less generally throughout the body, whereas the particular 
granules under consideration are always concentrated near 
the rhizoplast. Hence, it is not certain that the granules 


694 H. M. WOODCOCK. 


have the same significance in the two cases. With regard to 
the curious streaky substance, its position in relation to the 
basal part of the flagellum certainly suggests some association 
with this organella; it seems to me not at all unlikely that 
its presence is connected with the disappearance of the 
trypanomonad character, and, indeed, a comparison of figs. 
119, 149, and 150 prompts the query whether it may not 
possibly represent the remains of a flagellar border which 
has been actually absorbed by the parasite in the case of 
some of these massive forms. 

Another cytological character often apparent in fairly old 
cultures is vacuolisation. One or two small vacuoles in the 
cytoplasm may be seen occasionally in individuals of quite 
regular form; but, on the whole, in my cultures parasites 
belonging to the definite types recognised above are free 
from vacuoles. The occurrence of a few small vacuoles in an 
individual doubtless signifies nothing very abnormal ; when, 
however, the cytoplasm either appears practically full of 
vacuoles, or else contains one or two huge ones (fig. 154), this 
ought most probably to be considered as an unhealthy sign. 

Very marked indication of a disturbance in the mutual 
balance of the various cell-constituents is frequently seen in 
an irregular distribution of the nuclear organelle. Parasites 
with two trophonuclei and a single kinetonucleus are not 
uncommon (fig. 156). These are not to be interpreted as 
individuals which are in an early stage of division, the process 
having been begun by the trophonucleus. On the contrary, 
they are the result of a division in which the nuclei have 
been unequally apportioned between the two daughter-para- 
sites. This is clearly shown by fig. 157, where the cytoplasm 
is splitting in such a manner that one daughter-individual has 
both the trophonuclei and the other only a kinetonucleus. 
The remarkable feature is that these forms without a tropho- 
nucleus can live alone, at any rate for a certain length of 
time, for I have observed four or five examples in the course 
of examining my slides of this series (fig. 155). I have 
never found an active, flagellated form with a trophonucleus 


STUDIES ON AVIAN HASMOPROTOZOA. 695 


but without a kinetonucleus. In some of the large massive 
parasites numerous nuclei and flagella are present (figs. 162 
and 163), the number of the different organelle not by any 
means corresponding. Successive multiplication of the latter 
has taken place without concurrent division of the cytoplasm ; 
later, the cytoplasm would probably split into three or four 
portions, and it might very well happen as a result that 
one of the individuals thus formed would be happy in the 
possession of three trophonuclei (fig. 158). 

Another interesting irregularity in division is met with 
rarely. This consists in the unequal splitting, longitudinally, 
of the cytoplasm of certain large individuals, a thin form, 
with (fig. 159), or possibly without (fig. 161), a flagellum 
being cut off from the side of the parent. An important 
point is that these forms have no definite nucleus of either 
kind—i.e. they are apparently without both tropho- and 
kinetonucleus. In fig. 161 the individual—if such that 
portion of the cytoplasm can be termed—about to be cut off 
has a clump of granules, but that in fig. 159 has nothing at 
all. I have not observed a narrow form of this kind actually 
free ; in fig. 160, however, an active pear-shaped individual 
is drawn which also has no definite nucleus, but which possesses 
many red-staining granules. I have no doubt whatever that 
these forms are purely “ freaks,” the result of a degenerative 
mode of division, and die off quickly after being set free. 
There is a general resemblance, it will be noted, between 
this production of enucleate forms, in my cultures, and the 
formation of sickle-like (so-called ‘“‘spirillar ”) forms in 
cultures of Leishmania donovani, described by Leishman 
and Statham (8). It is highly probable that, in that case, 
too, the process is due to an abnormal condition of the 
Leishmania parasites (which, of course, ultimately degene- 
rate and die off in cultures), and that such forms have nothing 
todo with any natural developmental phase in the insectan host. 

Reference has been made already to the occurrence of 
rounded forms lacking a flagellum. ‘These have been seen 
only in an old culture of nineteen days, in which they are not 


696 H. M. WOODCOCK. 


infrequent. A few are medium in size (fig. 135), but most 
of them are small (figs. 137, 138, and 139). It is quite 
obvious from their appearance that these forms of the 
parasite, in the culture at any rate, are not merely “ resting,” 
persistent phases, but are degenerating and dying. And it 
is interesting to note that the process of degeneration takes 
place by a gradual disappearance of the nuclear elements. 
These no longer stand out, sharply stained, in the cell. They 
lose their distinctive affinity for the stain and become less 
and less distinguishable from the general substance of the 
body ; at the same time they tend to diminish in size, as if 
they were being dissipated in the cytoplasm. ‘lhe last stage 
of the parasite is an indefinite body, which stains a dull or 
faint red. Hence, so far asthe cultural forms are concerned, 
all the evidence I have goes to show that tle loss of the 
flagellum means approaching degeneration and death (con- 
trast, for example, the parasite of fig. 136 and that of 
fig. 138, which are on the same slide and within a few fields 
of each other). 

The above description includes all the different types and 
the chief varieties of form which I have observed among the 
trypanosomes in cultures. 

Agglomeration.—I have, next, a few observations to 
make upon the characteristic feature known as agglomeration. 
I have seen many instances of this occurrence in my cultures. 
I have never found it in early original cultures (i.e., of less 
than six or seven days), nor in subcultures. Agglomerated 
clusters are only met with when the parasites have become 
abundant in the medium. The clumps are of all sizes, from 
small ones composed of a few individuals (a dozen or less) 
up to large masses containing hundreds of parasites. Now 
and again, in these large aggregations, the parasites are 
clustered round more than one centre, i.e. in these cases 
there is anapproach to the condition of secondary agglomera- 
tion, distinguished by Laveran and Mesnil from primary 
(single) clusters. In all the clusters seen the parasites have 
their flagella directed towards the centre of the rosette. 


STUDIES ON AVIAN HASMOPROTOZOA. 697 


On more than one occasion I have noticed the commencing 
formation of a clump in a cover-slip preparation of living 
parasites, where every field contained numerous individuals. 
Here and there are small numbers of parasites, which have 
become entangled by their flagella, the distal portions of 
which appear to be inextricably intertwined. 

Once started, the increase in size of cluster may take 
place in two ways: (1) by the addition of fresh individuals 
from the surrounding medium, which are continually being 
attracted; and (2) by the multiplication of forms already 
present. The increase is undoubtedly due much more to the 
former method than to the latter ; during the early stage, at 
any rate, it is probably almost entirely due to the accession 
of more individuals. In short, these clusters are formed 
mainly by agglomeration. Asa matter of fact, dividiug forms 
are comparatively rare in all the clusters I have examined 
(cf. figs. p-G, Pl. 5). I once left a cover-shp preparation 
containing a great many free, active prrasites for two or three 
hours; when | returned to it I found several large clumps 
which had not been there before. It was impossible that these 
rosettes could have arisen otherwise than by agglomeration ; 
they all had their flagella centrally directed and resembled 
the cluster of fig. G, except for the fact that some were even 
larger. 

An early stage in the formation of a cluster is seen 
in the micro-photograph reproduced in fig. £. ‘The individuals 
composing it differ appreciably in form and size; some of 
them, at the periphery, had apparently only recently been 
attracted, and were not yet firmly attached. Only two 
individuals are undergoing division. The beginning of a 
secondary agglomeration is instructive. Parasites continue 
to be attracted to the clump, but owing to the number already 
present the newcomers are unable to penetrate in between 
thein and become firmly attached. Hence they tend to form 
a subsidiary cluster for themselves (figs.nandr). The large 
agelomeration-cluster of fig. @ is apparently made up of 
individuals attached around three centres, two of which, the 


698 H. M. WOODCOCK. 


older two (in the upper right-hand part of the figure), are 
partially confluent. 

It is important to note that agglomerations are formed 
of individuals which are of a quite normal type. Nearly 
all the parasites of the clusters figured, for example, are 
definitely trypanomonad in character, either fairly long and 
fusiform, or belonging to the pyriform variety of individual. 
Agglomerations of less typical forms, pseudo-herpetomonad 
in character, also occur, but I have not met with them to any 
extent, even in old cultures. 

Novy and McNeal, in their account of cultures of avian 
trypanosomes (14), make a great point of distinguishing 
between multiplication rosettes and true agglomeration clus- 
ters. They regard all rosettes in which the parasites are 
joined by their flagella, corresponding, that is, to those I 
have just described, as arising by successive multiplication 
from a single individual, which starts the culture. Only 
those cases, on the other hand, where the parasites are 
united by their aflagellar ends, are considered to be true 
agglomeration clusters. Until I myself came to work with 
cultures, I had no idea but that the view of these authors was 
correct, and that these two opposite kinds of clusters resulted 
from quite different processes. Studying Novy and McNeal’s 
description and figures in the light of my own work, I feel 
sure that these authors have given an entirely wrong 
interpretation of the clusters, which they regard as multi- 
plication rosettes. Novy and McNeal consider that the whole 
process starts from a single cell, which is more or less 
rounded off, and has no flagellum. This gives rise, by division, 
to a few cells, which now possess flagella; by further multi- 
plication, a typical rosette of spindle-like forms is pro- 
duced. 

Novy and McNeal’s figures on Plates 8 and 9, which 
are from excellent micro-photographs, are most instructive, 
and are, in my opinion, convincing evidence that the view 
these authors put forward is incorrect. Most of the figures 
represent simply clusters, large or small, of different forms 


STUDIES ON AVIAN HAMOPROTOZOA. 699 


of the parasite, certain of which appear distinctly unhealthy. 
The authors state that all the figures on the plates to 
which I am now referring (as well as others) are of para- 
sites from a culture in the seventh generation, grown for 
seven days, by which I understand them to mean a sixth 
subculture, itself of seven days’ age. This long-continued 
cultivation doubtless accounts both for the varieties of form 
present, as well as for the number of clusters. Their fig. 2, 
Pl. 9, supposed to represent an early stage in rosette- 
formation, shows a large indefinite-shaped parasite, in which 
irregular multiplication of the nuclei is going on. There 1s 
no indication of the development of any flagella, and I have 
no hesitation in regarding this individual as an abnormal, 
degenerating form. That it would ever give rise to a rosette 
of active, flagellate parasites is most improbable. Again, 
fic. 3, Pl. 9, represents an agglomeration cluster of four or 
five somewhat similar forms, three or four of which, however, 
are not quite so degenerate, as they still possess flagella; but 
the same irregular multiplication of the nuclei is shown. 

Phases such as these have, I venture to say, no connection 
whatever with the rosettes of more typical parasites figured 
on Pl]. 8. Fig. 2, here, is a small cluster of a dozen pyriform 
individuals, each with a single, centrally directed flagellum. 
Not one of the individuals shows the least sign of division. 
Similarly in fig. 1, Pl. 8, there is a cluster of about eighteen 
parasites. Hence, in neither of these rosettes is there any 
evidence that they are going to give rise to one of many 
more individuals, such as that of fig. 4, Pl. 8, by multiplica- 
tion. And, from my own experience, I know that such 
rosettes can be formed very cuickly indeed. In this cluster 
of fig. 4 there are several individuals at the periphery, which 
are manifestly only loosely attached, and whose flagella 
cannot be connected with the central core (cf. my own 
figures). There can be no doubt that these are the individuals 
which have been most recently attracted to the cluster. 

A point in favour of this view of Novy and McNeal’s would 
be furnished by evidence which went to show that two 


700 H. M. WOODCOCK. 


typical daughter-parasites often remain entangled by their 
flagella after division. Now, as I have stated, the flagellar ends 
of the two individuals resulting from division (i.e. longitudinal 
fission) always become widely separated, and I have never 
seen any instance of such an occurrence. Even in the rare 
cases where multiple (quadruple) longitudinal fission is pro- 
ceeding, the flagella are all distinctly free from one another, 
and when the cytoplasmic division was completed, the 
daughter-individuals would doubtless separate. Moreover, 
from Novy and McNeal’s figures, it is obvious that the divid- 
ing forms in their cultures behaved in a similar way (cf. figs. 
i°2)and 5. Pl: 7), 

Hence, to conclude, I regard Novy and McNeal’s rosettes, 
in which the parasites are attached by their flagella, equally 
with those in my own cultures, as true agglomeration clus- 
ters, originating, and in the main increasing, by the coming 
together of independent individuals. ‘here can be no doubt, 
it may be pointed out, that agglomeration of trypanosomes 
by the flagellar end does occur in the invertebrate host; the 
process has been described, for instance, in the case of T. 
lewisi, when in a louse, by Prowazek (22), and when in a 
flea, by Swingle (88). 

On the other hand, there is no reason to doubt that in 
certain types or phases of the parasite agglomeration in cultures 
may take place by the aflagellar end; this is stated by Novy 
and McNeal to occur in the case of their “ spirochetes.” I 
have never had cultures which showed a sufficient number of 
parasites belonging to this type for agglomeration to occur, 
and so am unable to say more upon this point. It is interest- 
ing to note, however, that agglomeration of trypanosomes 
in the blood of the vertebrate hosts takes place by the 
aflagellar (kinetonuclear) end, and these ‘‘ spirochetes” are 
also definitely trypaniform ; in contra-distinction to these, para- 
sites of the trypanomonad type form rosettes which have 
their flagellar ends attached. 


STUDIES ON AVIAN HAMOPROTOZOA. 701 


Summary and General Remarks on the Cultural 
Forms. 


From my observations on the cultural forms of T’. frin gil- 
linarum a few interesting and important data have been 
obtained, relative to the course of the development of the 
parasites on passing into the culture-medium. The earliest 
type of form which I have found is a slender, trypaniform 
phase. This is soon replaced by the characteristic trypano- 
monad phase, into which most of the trypaniform individuals 
pass. This trypanomonad phase is the predominating 
cultural form, and it is persistent, apparently, so long as the 
condition of the medium remains healthy. During this period, 
however, in a culture of six days’ age, trypaniform individuals 
have also been seen, though they were extremely few in 
number. Further, rare instances of another form have been 
found, which is distinguished by its vermiform appearance, 
and by the remarkable ladder-like character of its trophonu- 
cleus. This phase is doubtless simply a further development 
of the ordinary trypaniform type. Whether these later try- 
paniform individuals represent forms of this character which 
have been persistent from the commencement of the infection, 
or whether they indicate a second development of this phase 
from the trypanomonad type, I have not sufficient evidence to 
decide. J am rather inclined to think, however, that the 
latter may be the case; for one or two individuals have been 
found which might correspond to transition-forms in such a 
passage (fig. 128). 

Since the above research was carried out, I have been 
studying, in conjunction with Prof. Minchin, the parasites of 
Athene noctua, andI have observed the early developmental 
phases of a trypanosome (most probably T’. noctue) from 
this bird, in the stomach of the mosquito (Culex pipiens). 
We hope to publish in due course a full account of this work, 
but 1 wish to refer here to one or two general facts. In the 
first place, to answer any possible criticisms, it may be stated 


702 H. M. WOODCOCK. 


expressly that the flagellates which I am about to mention 
were derived, beyond all question, from the little owl. 

The parasites occur both in the trypanomonad and in the 
trypaniform phase. Some of the latter individuals resemble 
the vermiform type of figs. 10 and 137 closely, the only differ- 
ence being that the attenuation may be even more pronounced. 
In fig. 132 is drawn such an example, which shows the extra- 
ordinary slenderness of the body. Hence, so far as I am able 
as yet to compare the two cases, this elongated trypaniform 
type develops to a much wore marked extent im natural 
conditions than was the case in my cultures; in the latter, for 
some reason or other, it was soon almost eutirely superseded 
by the trypanomonad type. 

The occurrence of anything approaching a herpetomonad 
phase has only been seen in cultures of a certain age, in 
which there is every reason to believe the condition of the 
medium must be becoming abnormal and unhealthy for the 
parasites. Even then, it is only very seldom that an individual 
is found which corresponds at all closely to a true herpeto- 
monad (fig. 147); most of the parasites assume what I have 
called a “pseudo-herpetomonad”’ condition, which is readily 
distinguishable from that of an ordinary herpetomonad. 
With regard to the occurrence of rounded-off “resting” 
phases, forms of this kind without a flagellum were seen also 
only in old cultures, full of altered forms, and the individuals 
which were in this condition were manifestly degenerating 
and dying. Hence, from such individuals no conclusions can 
be drawn respecting the occurrence of rounded, aflagellar 
phases as a normal part of the life-cycle in the insectan host. 
Such a phase may occur or it may not. 

What may be regarded as highly probable, however, is the 
occurrence in natural circumstances of forms which corres- 
pond to the small fusiform or pyriform individuals of the 
culture (cf. fig. 111) in an attached condition, i.e. with the 
flagellum more or less shortened or retracted, and serving as 
fixative organella. The predilection that such forms have for 
forming groups or clusters in the cultures (cf. fig. G, Pl. 31, and 


STUDIES ON AVIAN HAMOPROTOZOA. 703 


also Novy and McNeal’s figures of so-called multiplication- 
rosettes on P]. 8) is probably to be regarded, indeed, as indi- 
cating the tendency of these forms to become attached, when 
in the natural insectan-medium. In the culture-medium, how- 
ever, there is nothing for them to attach themselves to, ex- 
cepting these commencing clusters of their fellow-individuals. 
Hence, the probable explanation—in great measure, at any 
rate—of the clumps or clusters which have their flagella 
centrally directed, is that they represent the attached phase 
in the insect. This is of well-known occurrence, both among 
trypanosomes (cf. Prowazek [l.c.], figs. 53 and 54), and 
among insectan flagellates (cf. especially Patton [16, Pl. 9, 
fig. 22], where a number of Crithidia sp.,in Gerris are 
clustered around a food-particle, and again, Swingle [32], 
who states that a rosette of Crithidia in the sheep-ked, 
Melophagus, may be formed around a free epithelial cell). 
In the case of parasites in cultures, when one, two, or tnree 
individuals have become entangled by their flagella, the inter- 
locked ends furnish doubtless the “nucleus” for the attach- 
ment of many other parasites, with the result that a large 
cluster is soon formed. 

An important point brought out decisively by my cultures is 
that this avian trypanosome does not proceed to form 
rounded-off, resting phases immediately on passing from the 
vertebrate host into the cold medium. And further, I may 
mention, there is not the least indication of any such behaviour 
in the case of the trypanosome of the little owl when it 
passes into the stomach of the mosquito. 

Up to the present only one or two accounts of cultural 
forms of trypanosomes have been published which describe 
and make any attempt to distinguish between the different 
types of form and phases developed at different periods in 
the culture. Of these, the most important for purposes of 
comparison with my own results is the paper of Novy and 
McNeal, to which reference has been made. In this connection 
it must be emphasised that most of the authors’ figures of 
cultural forms (and apparently their descriptions also) are 


704 H. M. WOODCOCK. 


based upon the parasites present after cultivation has been 
continued for some time, i.e. in sub-cultures of the sixth or 
seventh generation, when the culture was fully developed 
and ‘‘enormously rich in flagellates.” In such cultures of a 
trypanosome, regarded by the authors as T’. avium, the 
great majority of the trypanomonad forms were found in 
clusters, some of which were large enough to be visible 
macroscopically as patches in the medium. The interesting 
point is that parasites in the form of “ spirochetes” were of 
common occurrence, sometimes abundant; “‘spirocheete,” it 
may beas well to state, is the term applied by Novy and McNeal 
—somewhat misleadingly—to individuals of the slender, 
trypaniform type, similar to those seen in my figs. 10, 126, 
and 127.1. These trypaniform individuals were mostly free, 
very active, and some were undergoing division. 

Hence the condition found by Novy and McNeal obviously 
represents a much later period in the development of the 
culture than any I have described above, and I cannot find any 
account of the early course of the development, i.e. during 
the first five or six days or so. The authors do not say at 
what intervals of time their sub-cultures were made, but it is 
evident, from the number of the “ generation ” given, that 
the trypanosomes must have been cultivated for at least 
some weeks. In the case of T. fringillinarum, I was 
unable to obtain any development in my cultures corre- 
sponding to that found by Novy and McNeal in T. avium. 
If I did not sub-culture frequently enough the parasites 
become abnormal and degenerative, so that a preparation 
would show nothing but altered, pseudo-herpetomonad forms 
and so forth, and when I sub-cultured frequently the try- 
panosomes retained, for the most part, the trypanomonad 
phase. I never continued subculturing for so many gene- 
rations as Novy and McNeal did; it is only since I have 
come to study carefully my preparations and to compare 

1 Although in one or two cases these parasites show indications of an 
extended nucleus, in no case is a definite ladder-like appearance figured 
or described. 


STUDIES ON AVIAN H#MOPROTOZOA. 705 
my results with those obtained by Novy and McNeal that 
I realise some additional knowledge might have been gained 
by continuing to cultivate longer. In one case I subcultured 
four times at fairly slow intervals; this was done chiefly 
with a view to seeing how long I could keep a culture of 
the trypanosomes alive (cf. above, p. 647). Unfortunately, 
being kept away for a few days by ill-health, I missed an 
opportunity of examining this fourth subculture at a time 
when the parasites would have been very numerous; and 
before my return an unfortunate accident had terminated 
their career. Possibly this subculture might have shown 
more trypaniform individuals. 

Novy and McNeal go to the length of founding two new 
species of Trypanosoma upon the different behaviour and 
appearance of certain of their cultural forms. In fact they 
distinguish several types or varieties chiefly or entirely upon 
a basis which is most inadequate and misleading, namely, on 
a comparison of the multiplication-rosettes (really the 
agglomeration-clusters) and the free ‘‘ swarming ” parasites 
in the different cases. I only wish to point out here that, in 
the case of both their new species, viz. T. laverani and 
T. mesnili, the free-swarming forms which they compare 
with the slender, trypaniform type of the other species dealt 
with (T’. avium) and contrast with the rosette-forms, are in 
reality not trypaniform (“spirochetes”) at all, but are ordi- 
nary trypanomonad forms, which do not differ essentially from 
those constituting the rosettes. This is perfectly obvious 
from a comparison of their figures on Pls. 5-7. 

The matter amounts simply to this: In the case of these 
two species, the authors have not got a development of the 
trypaniform type at all. Many of Novy and McNeal’s figures 
of these forms, especially of T. mesnili on Pl. 6, are of 
individuals which show pronounced vacuolisation, and which, 
in my opinion, appear distinctly unhealthy; also the cluster 
of individuals of T. laverani, reproduced in fig. 3, Pl. 7, 
I regard as partly composed of abnormal forms. In short, 
from a comparison of the figures given of T. laverani and 


706 H. M. WOODGOGK. 


T. mesnili with most of those of T. avium, I am 
strongly inclined to say that the cultural development of the 
former parasites was not proceeding so successtully—at any 
rate, when the preparations concerned were made—as that of 
the last-named species. 

Slight differences in the constitution of the medium may 
certainly influence the rapidity of growth of these cultural 
forms, as I have stated above, and probably also, toa certain 
extent, the manner of their development. Further, it is 
quite likely that different species of trypanosomes, when 
cultivated in the same medium, may also differ in their rate 
of growth and in the development of the different types of 
form. Hence, I think we may agree with Novy and McNeal, 
although on quite different grounds, that the parasites which 
they name 'I'. laverani and mesnili are at any rate 
different from the other (IT. avium). Moreover, it may 
be reasonably inferred that under slightly different condi- 
tious—in one way or another—of the medium, these forms 
would also develop a trypaniform phase. For it will be seen 
from the subsequent context of this paper that there is every 
reason to suppose such a phase is of regular occurrence at 
some period in the development of a trypanosome outside the 
vertebrate host. As a matter of fact, T. laverani itself 
appears to be very closely allied to the trypanosome with 
which I have been working. 

The only other paper dealing with cultural forms, to which 
I need refer is a note by Thomson (35), on the cultivation of 
trypanosome (probably T. danilewsky1), from the goldfish, 
which gives instructive indications of the course of develop- 
ment of that parasite in cultures. It is most interesting to 
find that there is a general resemblance between the course 
of events in the case of that piscine form, as outlined by 
Thomson, and in the avian parasites discussed above. 
Thomson does not describe any developmental forms occurring 
earlier than the seventh day. By this time the parasites are 
in a phase corresponding to my accentuated trypanomonad 
type; and division by a quite similar method of unequal 


STUDIES ON AVIAN HAMOPROTOZOA. 707 


fission is taking place, a smail fusiform or pyriform individual 
being cut off from the large, more or less club-shaped parent- 
form. Several of 'Thomson’s figures are, indeed, almost 
identical with some of my figures. Another important point 
is that distinctly trypaniform individuals were present, and 
such forms were found to be more frequent later on, for 
instance in a culture of the forty-second day. 

As Thomson says, it is probable that earlier phases in this 
development might have been found before the seventh day. 
It is interesting to note that Thomson figures an unaltered 
trypanosome (as it left the blood of the fish) in the culture of 
seven days. Thomson’s view is that the large, club-shaped 
trypanomonad individuals are derived directly from such 
trypanosomes by an alteration of the body-form, most of the 
protoplasm becoming concentrated in the aflagellar part of 
the parasite, which thus becomes greatly swollen in appear- 
ance. According to Thomson, there is no prior multiplication 
of the parasite in an ordinary trypanomonad condition. 
Hence in this case a type of form very similar to that which 
I have found in my cultures (cf. figs. 97, 98) is attained by a 
quite different process ; in the culture of the avian parasites, 
the trypanosome-phase is quickly lost and active multiplica- 
tion in the ordinary trypanomonad phase goes on. 

It is evident from this that the development of the piscine 
type in cultures proceeds much slower than that of the Avian 
form, and this bears out, in an interesting manner, the facts 
so far known relative to the development of the two types in 
the true invertebrate hosts (leech and insect) respectively. 


The Significance of these Cultural Forms of Try- 
panosomes in Relation to the Questiom of an 
Alternate Invertebrate Host. 


When we come to compare the chief types of form described 
above as occurring in cultures of trypanosomes from different 
vertebrates with the flagellates described by various authors 
from blood-sucking invertebrates, which they have considered 

VOL. 55, PART 4,.—NEW SERIES. 47 


708 H. M. WOODGOCK. 


as being phases in the life-cycle of some vertebrate trypano- 
some, we find at once a fundamental resemblance, while in 
one or two particular cases there is a strikingly close 
similarity in detail. It would occupy too much space to 
follow out this comparison at length. I must content myself 
with a reference to various papers, and with a few indications 
as to the chief points of agreement. 

It may be noted, as a preliminary, that I follow Patton’s 
definition of, and distinction between, a herpetomonad form 
and a crithidial or trypanomonad form ; the terms “crithidial ” 
and  trypanomonad” are practically interchangeable, but I 
prefer to use the latter, at all events when referring to this 
phase in connection with a vertebrate trypanosome.' Fur- 
ther, it is necessary to emphasise the fact that the characteri- 
sation of these two types is based upon their structure when in 
the active, extended, flagellate condition ; in other words, the 
diagnostic form of the parasites is only seen when they are in 
this condition. Rounded, resting phases, whether possessing 
a flagellum or lacking one, cannot be regarded by themselves 
as representing either a herpetomonad or trypanomonad 
phase, simply because, when the parasites are in this con- 
dition, the features used for distinguishing between the two 
types are not present. It is certainly due to Patton that we 
are at last able to realise that there are these two perfectly 
definite types, a herpetomonad and a crithidial or trypano- 
monad one, and to distinguish clearly between them. Until 
Patton separated the two types upon the above basis, the 
greatest confusion often prevailed as to whether a given 
parasite belonged to one or the other; and it must be 
admitted this confusion was chiefly due to the unsuitable 
diagnostic characters used by Léger in his earliest descriptions 
of these forms. . 

The memoirs in question are those by Miss Robertson (28, 
24, and 25), Minchin (11), Prowazek (22), Stuhlmann (81), 
and Roubaud (26). In all the parasites described, namely, T. 


1 There have been, hitherto, two quite different meanings attached to 
the term ‘ crithidial.” (ef. also below). 


! 


STUDIES ON AVIAN H#MOPROTOZOA. 709 


raiz and T. vittatz (Miss Robertson), T. grayi (Minchin), 
T. lewisi (Prowazek), T. brucii (Stuhlmann), and T. 
gambiense, cazalboui, and congolense (Roubaud), a 
trypanomonad phase occurs, and is usually prominent. In 
all of them a definite trypaniform phase (i.e. one in which 
the kinetonucleus is some distance on the aflagellar side of 
the trophonucleus), is also met with. And in two cases, 
namely, T. brucii in Glossina fusca (Stuhlmann), and T. 
raiz in Pontobdella (Miss Robertson), the occurrence of 
a greatly elongated trypamform type with an extended, 
ladder-like nucleus is described. These are the only cases 
of which I know where this characteristic type of form has 
been seen in an invertebrate ; and it is highly significant, I 
think, that a similar form occurs, beyond all question, as a 
developmental phase of more than one avian trypanosome. 
Unfortunately I am not yet able to add anything to our 
knowledge of the purpose or meaning of this interesting 
form, which has been variously considered as possibly a male 
form, and—more likely—as a propagative individual infecting 
a vertebrate host. 

The same close agreement holds good also for another 
important point, namely, the absence—apparently the entire 
absence—of anything corresponding to a true herpctomonad 
phase in these parasites when in the Invertebrate host. Out 
of a total of some hundreds of figures in the above memoirs, 
there is not one which shows a typical herpetomonad indi- 
vidual, such as, for instance, Herpetomonas musce-domes- 
tice, lygei, jaculum, etc.,or Leishmania. There are only 
one or two figures, e.g.in one of Miss Robertson’s accounts 
(24, figs. 12, 21, and 22), which could be regarded as in any 
way approaching a herpetomonad condition; and it is precisely 
in such a case, moreover, that the essential proviso noted 
above must be borne in mind. ‘he individuals figured are 
manifestly intermediate stages in the development from a 
rounded resting-phase to an active flagellate type of form. 
Further, they are all dividing, and one of the daughter- 
individuals (fig. 21, right-hand side) is already acquiring the 


710 H. M. WOODCOCK. 


trypanomonad condition. Hence these cannot be regarded as 
representing in themselves determinative phases, but are rather 
only transitory stages in the development of a trypanomonad 
(or it may be a trypaniform) type, such as is exemplified in 
most of Miss Robertson’s figures of active, flagellate indi- 
viduals. On the other hand, what is far more important is 
that none of the numerous elongated “ monadine”’ forms 
figured by Roubaud (26) show any indication of herpeto- 
monad affinity. Last, but not least, the so-called herpeto- 
monad forms of 'l’. grayi—the extremely slender ones, which 
proceed to encystment—have nothing whatever to do with 
the herpetomonad type, as indeed Patton has already pointed 
out, but are unmistakably of the trypanomonad type. This 
inistake arose, of course, simply by following Léger’s mode 
of distinguishing between the two types chiefly by means of 
the body-form. 

There can be no doubt, I think, that this briefly outlined 
comparison enhances the probability that the various accounts 
to which I have alluded do actually relate to phases of the 
life-cycle in an invertebrate host of the different vertebrate 
trypanosomes which they purport to do; in my own opinion, 
andin that, I venture to say, of most other people, the matter 
is certain. 

I should like to offer a few further remarks upon the still 
disputed question of a vertebrate trypanosome in its alternate 
host versus a natural flagellate of the invertebrate. In the 
first place, two classes of invertebrates are principally con- 
cerned, namely leeches and insects. The former I intend to 
leave altogether out of account, as up to the present not the 
slightest evidence has been brought forward of the occurrence 
of any flagellate parasites in this class of hosts, which are not 
developmental forms of some vertebrate trypanosome. In 
the case of insects the subject is much more complicated ; 
since in many non-blood-sucking insects flagellates occur 
which can be only parasites of the one host. 

As a result of the above comparative observations, one 
general proposition can be stated, I believe, which ought to 


STUDIES-ON AVIAN HAMOPROTOZOA. FAL 


prove of considerable help in this connection. It is this: 
Parasites exhibiting a trypaniform condition ina 
blood-sucking insect must be consideredas belong- 
ing to the life-cycle of a vertebrate trypanosome, 
until the contrary is definitely established ; and the onus 
probandi lies with those who maintain the opposite view. 

Another conclusion which appears indicated is that, in 
general, such parasites do not pass into a true herpetomonad 
condition; in other words, they have not a definite herpeto- 
monad phase in the life-cycle. Bearing in mind that many, 
at any rate, of the vertebrate trypanosomes which have an 
insect as their alternate host are almost certainly to be 
derived from a herpetomonadine form, which was originally 
a parasite solely of the insect, it will be understood, of course, 
that in certain circumstances the parasites may revert, as 
it were, to a pseudo-herpetomanad condition, or even to a 
herpetomonad one, as I have found in the case of my avian 
trypanosomes in cultures. But with this qualification, all the 
observations so far recorded point to the above conclusion. 

As a matter of fact, the occurrence of typical herpetomonad 
forms in blood-sucking insects has not been described in 
nearly as many cases as would appear, at first sight, to be 
the case. In many of the papers that I have seen which 
profess to describe such forms, a study of the figures shows 
that the authors have been dealing really with trypanomonad 
(crithidial) forms; these are merely further instances of the 
confusion formerly existing in regard to the diagnosis of 
these two types. Thus the Herpetomonas algeriense 
described by the Sergents (28) from Culex pipiens does 
not appear to have anything in common with a true Herpe- 
tomonas; from the figures given it must be regarded as a 
trypanomonad form.! 


1 Instances, on the other hand, of what are apparently true herpeto- 
monad forms occurring in mosquitoes and restricted to this host are 
given by Patton (‘ Brit. Med. Journ.,’ 1907, ii, p. 78) and also by the 
Sergents (1.c.); but there is not likely to be any difficulty in distinguish- 
ing such parasites from phases of a vertebrate trypanosome. I may 


712 H. M. WOODCOCK. 


Again, Novy, McNeal, and Torrey, in their paper on the 
flagellates of mosquitoes (15), distinguish two parasites, 
namely, Crithidia fasciculata and Trypanosoma 
(Herpetomonas) culicis. These authors also followed 
Léger’s unfortunate definition of a Crithidia, restricting the 
name to small oval or pyriform parasites with a truncated 
flagellar end and a short flagellum. ‘The whole objection to 
this definition lies in the fact that such forms are merely 
resting or attached phases (in natural conditions) of either 
crithidial (trypanomonad) or herpetomonad forms. However, 
in the case of their Crithidia, the figures given show that, 
in a more elongated condition, it conforms on the whole to 
the trypanomonad type. Similarly, their other parasite, 
Trypanosoma (Herpetomonas) culicis, also hasa well- 
marked trypanomonad phase, as, indeed, is implied by the 
generic position which the authors assign to it; apparently 
it is placed in the sub-genus Herpetomonas because of its 
monadine form. I may observe here that these papers by 
the American authors have been most difficult for me to 
comprehend, because the indications afforded or suggested by 
their plates often appear to be opposed to the account given 
in the text. I have only really grasped the significance of 
their first paper on avian parasites and their cultural forms 
since working on my own birds and cultures ; and I am sure, 
from the interesting plates of mosquito-parasites in the 
authors’ second paper, that a further study of the phases and 
forms which they figure is essential to a correct understanding 
of their significance. Hence I do not propose to criticise 
them further at present. 

This much, however, must be said in regard to all these 
cases of the occurrence of trypanomonad forms in mosquitoes. 
It is at least quite as likely that the flagellates observed 
were phases of vertebrate trypanosomes—say of avian forms 
—as that they were purely insectan parasites. I have referred 


say here that in the development of T. noctuz in Culex pipiens 
I have not come across the slightest indication of a herpeto monad 
phase. 


STUDIES ON AVIAN H#MOPROTOZOA. 713 


above to the undoubted indications I have obtained that a 
trypanosome of the little owl undergoes developmental phases 
in Culex pipiens. ‘There is, therefore, no reason whatever 
to doubt any longer that some, at all events, of the flagellate 
phases described by Schaudinn in mosquitoes which had fed 
on infected birds were also actually phases of ''rypanosoma 
noctue. Moreover, in regard to Crithidia fasciculata 
itself, the type-species of that unfortunate genus, no one has 
yet shown that it is solely an insectan parasite. In first 
describing it, Léger very wisely admitted the possibility that 
it was only a phase of a vertebrate trypanosome, and this 
still remains the most logical assumption with regard to it. 

Similarly with regard to crithidial forms in other blood- 
sucking insects, e.g. C. tabani, Patton (18), C. melo- 
phagia, Swingle (82), etc., by far the most hkely and 
reasonable view is that these parasites are merely the trypano- 
monad forms of a trypanosome.! One or two cases have 
been described, however, of the occurrence of crithidial 
forms in what are alleged to be non-sanguivorous insects, 
e.g. C. gerridis from Gerris fossarum, Patton (16) ; 
such parasites may apparently be regarded as true Crithidia, 
by which we may understand flagellates that have developed 
a trypanomonad condition, but which are restricted to an 
invertebrate host. 

‘Two or three parasites have recently been described, and, 
moreover, from non-biting insects, which have been regarded 
as “trypanosomes.” ‘They are Trypanosoma drosophile, 
Chattou and Alilaire (8), and two peculiar herpetomonad 
forms termed Leptomonas mirabilis, from Pycnosoma 
putorum and L. mesnili, from species of Luculius, which 

1 As regards C. melophagia, I have quite recently obtained evidence 
which makes this almost certain. After prolonged examination of the 
blood of a sheep on which were “ keds” infected with this parasite, I 
had the good fortune to find a typical, active trypanosome. This is 
the first occasion, so far as I know, of a (natural) trypanosome having 
been found in this domestic animal. There can be little or no doubt 


that the “ Crithidia melophagia” is simply a developmental phase 
of this sheep-trypanosome in its alternate, insectan host. 


714 H. M. WOODCOCK. 


have been described by Roubaud (26). In the case of the 
first-named, the individuals figured certainly appear to be 
in a definite trypaniform condition, possessing a distinct, 
though narrow, undulating membrane. The two other para- 
sites are very remarkable, in that typical herpetomonad 
forms appear to have also a “ trypanosome ” phase in their 
life-cycle, and all intermediate conditions between these two 
extremes are figured. So far as I can judge from the figures 
given, however, the so-called “ trypanosome” phases do not 
representa true trypaniform condition in the sense in which 
it has been understood in the above pages. To begin with, 
the flagellar end of the body is not drawn out at all, but the 
flagellum emerges straightway from it. The kinetonucleus 
is, indeed, near the aflagellar end of the body; but in all 
cases the course of the flagellum, from the point where it 
comes into contact with the cytoplasm up to the kineto- 
nucleus, is shown running through the middle of the cyto- 
plasm ; it is never drawn lying at the side, still less as showing 
any undulations. I think this is an important point, and one 
which tells very much against the presence of an undulating 
membrane in these Leptomonas. For in the great majority 
of preparations of trypaniform parasites, however attenuated 
they may be, and however narrow the membrane, the attached 
flagellum lies nevertheless at one side (cf. my figs. 10, 126-132, 
and also Minchin’s figures of T. grayi). I think, therefore, 
that in these peculiar phases a considerable part of the 
flagellum is intra-cytoplasmic, forming, as it were, a long 
rhizoplast, consequent on the passage of the kinetonucleus to 
the opposite end of the body. These forms appear to be 
quite distinct both from ordinary herpetomonad parasites 
and from the true trypaniform type. ‘‘T’’ drosophile, 
on the other hand, appears to exemplify the trypaniform 
condition. 


The above summary represents, in my opinion, the present 
position of this difficult problem of the flagellates occurring 
in blood-sucking invertebrate hosts. My view on the subject 


STUDIES ON AVIAN HASMOPROTOZOA. tio 


is the same as that I have maintained in my article on the 
Hemoflagellates in Lankester’s Protozoa (39), as will be 
seen by anyone who cares to compare that account with the 
above pages. As a matter of fact, there is now no doubt 
whatever that one of Schaudinn’s far-reaching conclusions 
was correct, namely, that vertebrate trypanosomes undergo 
a definite part of their developmental cycle in an invertebrate 
host, and that true cyclical infection occurs by means of the 
latter ; for conclusive experimental proof has been recently 
brought forward by Kleine, Bruce and others, Minchin and 
Thomson. ‘l'o indicate the work of these authors, however, 
would be going outside the scope of this paper; moreover, in 
this discussion, I have preferred to limit myself to the above 
comparative observations, since most of them provided 
material on which I relied for support in my article (I.c.). 
Patton has of late occupied himself in reiterating his view 
that in all those instances considered above, as well as in 
every other case where an author has purported to describe 
phases of a trypanosome in an invertebrate, the parasites in 
question were merely natural flagellates of the invertebrate, 
which had no connection with a vertebrate host. Patton’s 
view is that of scarcely anyone else; even Novy and McNeal 
have not gone quite so far in this wrong direction. I do not 
intend to argue the matter with Patton; a perusal of his 
recent papers suggests that he is unable to appreciate argu- 
ment. In his latest review (20), Patton has adversely 
criticised, in somewhat forcible terms, my article in Lan- 
kester’s treatise, chiefly because I have maintained the 
opposite view to himself. I do not think it necessary to 
reply at length to Patton’s remarks; itis obvious that Patton 
is hopelessly biassed, and in one or two places I consider he 
oversteps the boundary of legitimate criticism. I venture to 
say, however, in justice to my editor as well as to myself, 
that if a student of tropical medicine and protozoology 
follows Patton’s judgments on our knowledge relating to the 
hemoflagellates and their allies, as set forth in his “‘critical”’ 
review, he will obtain a distinctly erroneous and misleading 


716 H. M. WOODCOCK. 


impression of the group, and one which is further from the 
truth than the views expressed in my article. 


(c) Notes on Nuclear Cytology and Division. 


My material, having been all stained by the Romanowsky 
method, has not proved very suitable for a study of the 
minute structure of the nucleus (trophonucleus). Neverthe- 
less, in the light of the interpretation which Minchin and 
Woodcock (18) have shown is to be placed upon the 
“ Giemsa-picture ” of the nucleus of a trypanosome, I am 
able to say that, in the case of many, at any rate, of the 
parasites observed, the type of nuclear structure certainly 
agrees with that described in that paper. Unfortunately, in 
the parasites figured from the blood of the bird, the nucleus 
often shows the usual granular appearance; now and then, 
however, the definite clear region can be seen, corresponding 
to the central, plastinoid part of the karyosome, which con- 
tains a deeply staining granule in the middle—the intra- 
nuclear centrosome (figs. 30, 34,and51). Forsome reason or 
other cultural forms show this appearance, which is to be 
regarded as the typical one, much more frequently, indeed 
quite regularly (figs. 7, 8,72, etc.). The trophonucleus of the 
individual in fig. 5 is in an interesting condition; it is more 
faintly stained than usual, the nuclear sap apparently con- 
taining little or no chromatin (cf. the numerous chromatoid 
granules scattered in the surrounding cytoplasm). Whether 
the deeply-stained central body represents in this case a small 
karyosome or a greatly enlarged central granule, it is 
impossible to say. Other instances of an unusual appearance 
of the trophonucleus are seen in the parasites of figs. 38 and 
39; here there appear to be a certain number of separate 
chromatic masses, of varying size. This condition perhaps 
represents a fragmentation of the single large karyosome 
usually present. 

The blepharoplast, or basal granule, at the proximal end 
of the flagellum is sometimes visible in the parasites from the 


STUDIES ON AVIAN H#MOPROTOZOA. iB by, 


blood (figs. 4, 28); but frequently the proximal, rhizoplastic 
portion of the flagellum is not well stained, and then the 
blepharoplast cannot be made out. In preparations of cultural 
forms it is generally conspicuous, and now and again very 
prominent (figs. 10, 71, 81, etc.). 

As regards the details of commencing division, the try- 
panosomes in the blood have provided me, as already stated, 
with hardly any indications at all. On the other hand, I 
have obtained a nice series of stages among the cultural 
parasites. The first act in the process is apparently the 
division of the blepharoplast at the base of the flagellum 
(fig. 120). ‘This is followed by the splitting of the flagellum 
for some distance, which may be fairly short or fairly long 
(figs. 100, 104, 121, and 128); the splitting never extends, 
however, throughout the whole of the attached part of the 
flagellum. In the case of this avian parasite, the splitting-off 
of a portion of the old flagellum to form the foundation of the 
new daughter one appears to be of general occurrence. I 
have observed nothing which would indicate that the daughter- 
flagellum is formed as an entirely independent outgrowth 
from the second blepharoplast. Fig. 89 shows a flagellum 
caught in the act of dividing, the proximal portion being 
drawn out transversely, as a broad band, prior to splitting. 
In figs. 88, 100, 104, and 123, the newly formed part is still 
connected at its tip with the old tlagellum; and in fig. 121 
the new portion, in this instance only short, has just separated. 
Of course, once the rudiment, as it were, is cut off, its further 
growth is quite independent. 

The division of the nuclei may begin while the splitting of 
the flagellum is proceeding (figs. 104, 123), or it may be 
delayed until the latter process is completed (figs. 90, 121) ; 
there is apparently considerable variation in this respect. 
The first stage in the division of the trophonucleus is most 
probably the division of the intra-nuclear centrosome, which 
acts as a division-centre; this is clearly shown in fig. 88. 
All that can be said from Giemsa-stained preparations as to 
the rest of the process is that the nuclear substance becomes 


718 H. M. WOODCOCK. 


extended in a direction more or less transverse to the long 
axis of the body, this being doubtless brought about by the 
separation of the daughter-centrosomes (cf. fig. 99); the 
two centrosomes remain connected by a fibril, which at a 
later stage may become considerably drawn out (figs. 124, 
125). The nuclear material becomes aggregated around these 
two division centres ; as the latter continue to separate, it is 
pulled out more or less into the form of a dumbbell and finally 
constricted into two halves, the daughter trophonuclei. With 
regard to the division of the kinetonucleus, the process, so far 
as can be judged from the phases seen in figs. 101 and 104, 
appears to be similar to that occurring in the division of the 
other nucleus. A distinct thread or band connects the 
separating halves; this probably indicates a fibril, corres- 
ponding to the other, which may also have its terminations 
in two intra-nuclear division-centres. If this is really the 
case, not only the trophonucleus, but also the kinetonucleus, 
possesses an intra-nuclear centrosome. 


(p) Comparison of Trypanosoma fringillinarum 

with other Avian species. 

The reasons which have led me to consider all the manifold 
forms of the trypanosome met with as belonging to one and 
the same species have been given at the commencement of 
the description of the parasites, and also alluded to elsewhere 
in the account, so that I need not recapitulate them here. 
This illustration of the very great polymorphism which may 
be shown by one species is most instructive. If, for instance, 
only two types of form, at opposite extremes as regards size, 
had been observed, it might readily have been supposed that 
two different trypanosomes were concerned. And there can 
be no doubt that many observers, not only of avian parasites 
but also of others of cold-blooded vertebrates, who have based 
their descriptions on casual observations of the parasites, 
have fallen into such an error. So long as the mammalian 
forms, and among these chiefly the lethal ones, with their 
comparatively modest variations in form and size, remained 


STUDIES ON AVIAN H#MOPROTOZOA. 719 


those with which research was principally occupied, the 
possibility of such striking polymorphism was insufficiently 
recognised. It is evident, I think, that the safer plan for 
workers on these naturally occurring trypanosomes will be to 
regard all the forms met with in any one host as belonging 
to one species until they have satisfied themselves that this is 
not the case.' 

On the other hand, for the purpose of distinguishing 
different species of trypanosomes, I certainly continue to 
think that what may be called the biological consideration 
is, in the present state of our knowledge, the most reliable 
and useful guide. By this I mean that the less closely related, 
zoologically, two hosts are, the greater the probability that 
their trypanosomes are distinct species. As a general indica- 
tion it may be said that the same parasite may, in certain 
cases, be parasitic in different species of host, or even in 
closely allied genera,” but where the hosts in question belong 
to different families, or still more, to different orders, it may 
be safely assumed, as a working rule, that their trypanosomes 
are distinct species. The best practical test for this criterion 
is, of course, the production or non-production of cross- 
infection. 

In making use of resemblances or differences in morphology 
in comparing two trypanosomes, | think the ordinary adult 
form of the parasite furnishes the best indications. 'I'ake the 
case of 'l’. lewisi, for example; neither the young daughter- 
individuals resulting from multiple fission, nor the large, 
stout, multiplicative individual itself is regarded as the 
definitive form, the form of every-day occurrence, as it were. 
Now I think we can carry this comparison very usefully to 
other cases. Small, fusiform, or stumpy individuals are more 


1 T consider, for instance, that Wenyon (37) has done wisely in includ- 
ing the quite different types of form found, on the one hand, in the 
guinea-fowl (Numida) and, on the other hand, in a lizard (Mabuia), 
under one species in each case, viz. T. numide and T. mabuie. 

* In this connection attention must be paid to the question of distri- 
bution. 


720 H. M. WOODCOCK. 


likely to be young forms; these may, perhaps, themselves 
undergo division, as in Il’. lewisi, and, moreover, in many 
cases, owing to a slow rate of growth and increase in size, 
these small forms may give the impression of being distinct 
parasites. On the other hand, very large, massive forms are 
likely in many cases to be essentially multiplicative individuals. 
Of course the possibility must not be overlooked that, in 
some cases, large, stout forms may be sexual (female) indi- 
viduals, but up to the present evidence pointing to the 
occurrence of sharply differentiated sexual forms is only 
forthcoming in a few instances. At any rate, so far as T. 
fringillinarum is concerned, I think there is a general 
parallel with 'l’. lewisi in regard to the different types. 
In the case of many of the avian species so far described, 
the account has been based in all probability upon the ordinary 
adult type, e.g. T. avium, as emended by Laveran, T. 
padde, Thiroux, etc. But in other cases, where only 
stumpy forms have been described, such as T. hanne, 
another T. sp. from Senegambian birds,’ and T. laverani, 
these probably do not represent the definitive type. Passing 
on now to compare T. fringillinarum with certain other 
trypanosomes, we may begin with the type-species, T. avium. 
This name was originally given by Danilewsky, who followed 
his own methods of nomenclature, to trypanosomes found 
both in owls (sp. indet.) and in a roller-bird (Coracias 
garrula). Laveran (6) has rightly restricted this specific 
name to a parasite from an owl (Syrnium aluco), which he 
considers to be the same formas that observed by Danilewsky; 
the other trypanosome, from the roller-bird, is in all proba- 
bility a different species. T. fringillinarum, while 
showing a general similarity in size and form with T. 
avium, as described by Laveran, differs in two respects, 


1 This parasite, described by Dutton and Todd (4), occurred in a 
bird (Estrelda) in which the very different form T. johnstoni was 
found. It is not at all improbable, I think, that T. johnstoni is the 
ordinary form, and the broad, stumpy parasite a multiplicative form, 
of one and the same species. 


STUDIES ON AVIAN HAMOPROTOZOA. PAL 


namely, in the length of the free flagellum, which is much 
shorter, and in the appearance of the aflagellar end, whichis 
more elongated and attenuated. In addition, the hosts are, 
of course, quite different in the two cases. 

Novy and McNeal have included in the species T. avium 
a number of parasites they have found in various North 
American birds. They distinguish two chief forms, viz. large 
and small parasites, each of which shows considerable varia- 
tions in size. How Novy and McNeal have been able to 
ascertain any details with regard to form and size, if they 
had not better preparations to study than those from which 
their excellent photos have been taken, it is impossible to say. 
From their photos of the parasites in the birds, it is obvious 
that the trypanosomes were wretchedly fixed and stained ; in 
scarcely any can the length of the flagellum or the true 
nature of the aflagellar end be made out. Hence, any real 
morphological comparison is out of the question. In any 
case, on the grounds of occurrence and distribution, it is very 
improbable that any of the parasites represented the true T. 
avium. ‘This has been recognised by Lihe (9), who has 
placed all these forms identified by the Americansas T. avium 
in a new species, T’. confusum—a very apt name. I do not 
for a moment suppose, however, that all the forms described 
belong to one species. Novy and McNeal rely partly on the 
cultural characteristics shown, which they say were similar in 
allthese cases. All their photos of cultural forms of this 
group of trypanosomes are taken from preparations of a single 
culture, from one bird only. I should prefer to see figures of 
cultural forms from the other birds first of all. 

The trypanosome which Novy and McNeal distinguish as 
T. laverani, from an American goldfinch, Astragalinus 
tristis, is most probably closely related to I’. fringilli- 
narum, although I am hardly inclined to think the two forms 
are identical. The authors only figure a solitary example 
from the blood, which, from the size given, and from what can 
be made out from the photo, agrees very well with the small, 
fusiform individual of T. fringillinarum. There is a 


722 H. M. WOODCOCK. 


general agreement also, both in regard to appearance and 
size, between the trypanomonad forms in cultures. The 
reason which weighs most with me in keeping the two 
parasites distinct is the different hosts and their different 
distribution. Unfortunately Novy and McNeal do not 
describe, as I consider, the definitive type of the parasite, and 
so I am unable to compare it with that of T. fringilli- 
narum. Other reasons are that T. laverani is said to 
have a very sparse and slow growth in cultures, and the 
cultural forms themselves show very generally a peculiar 
rod-like structure near the aflagellar end of the body. I 
have certainly never seen this feature in any of the cultural 
forms of T. fringillinarum. 


3. Nore on HALTERIDIUM FRINGILLZ (LABBE). 


I have already published a short paper (38) relating to the 
chief features of interest which I have observed in connection 
with this parasite; and I do not propose to repeat in detail 
the description there given. I wish, rather, to add here a 
Tew general remarks and comments. 

I am now able to publish many of the actual drawings 
from which the text-figures in my previous note were made ; 
and these—especially the coloured figures—bring out certain 
distinctive points very clearly. It is particularly in such a 
case as this, I may say, that the value of the different tints 
and depths of colour, produced by the Romanowsky (Giemsa) 
stain, isapparent. Firstly, inregard to the dimorphism of the 
nuclear constituents (cf. especially figs. 14,15,and 17). The 
smaller nuclear body, representing the kinetonuclear element, 
is seen to be quite distinct in its staining reactions from the 
larger body, the ordinary nucleus. These two nuclear 
portions correspond closely in appearance (leaving out of 
account the marked difference between them as regards size) 
to the trophonucleus and kinetonucleus of a trypanomonad 
parasite, where these two organelle are close together or in 
contact. 


STUDIES ON AVIAN HAIMOPROTOZOA. (23 


Again, with respect to the so-called ‘‘ indifferent” indi- 
viduals, which are very scanty in number, compared with the 
female or male forms, figs. 15, 17, and 64, show the character- 
istically clear cytoplasm, not at all granular, and staining very 
faintly, of these individuals—readily distinguishable from 
the granular, deeply staining cytoplasm of female forms.! 
Further, in most of the parasites of this kind which I have 
found, the kinetonuclear element is relatively large, and 
may approximate in size to the other nucleus (cf. fig. 64). 
What exactly is to be understood by the term “ indifferent” 
as applied to these forms, and what their significance is, it is 
difficult to know. If they are neither male nor female they 
are not gametocytes; that much is obvious. At the time 
when I wrote my earlier note on this Halteridium, I was 
strongly inclined to think that these neutral individuals 
passed, in certain conditions or circumstances, directly into 
small trypanosomes. Unfortunately I have not been able to 
obtain any more evidence in support of this view, either from 
a renewed study of my own preparations of the chaffinch 
parasite, nor—which is even more important—from the study 
undertaken of Halteridium noctue, so far as this has yet 
progressed. Hence the meaning of these “indifferent” 
individuals, which certainly appear to be quite distinct from 
the forms of male or female character, has still to be ascer- 
tained. I have never found indications of division in them, 
any more than in the other types. 

In fig. 16 is drawn one of the two or three instances I have 
observed of the remarkable form of individual occurring free 
in the blood-plasma, which shows a conspicuous line running 
down the greater part of the body, near one side. This line 
stains distinctly red, like a flagellum; it appears to start in 
close proximity to the nuclear masses, and ends in a definite 
granule. The pigment-grains in this parasite are all aggre- 
gated together near one end of the body—that farther away 
from the nuclei. I regarded the halteridia in this phase 

Of course there is no possibility of confusing these forms with male 
gametocytes, which havea large, diffuse, pale-staining nucleus. 

VOL. 55, PART 4,—NEW SERIES. 48 


724, H. M. WOODCOCK. 


as being about to pass actually into little active trypanosomes, 
in a manner similar to that described by Schaudinn, that is 
to say, by getting rid of a portion of the cytoplasm con- 
taining the effete pigment-grains and by the development of 
a flagellum, the proximal, attached part of which constituted 
the flagellar border of an undulating membrane. In spite of 
much searching, I have not succeeded in finding any further 
stages in this developmental change. I cannot suggest any 
other satisfactory explanation of this peculiar structure, how- 
ever, and I still continue to think it has some connection with 
a flagellar development, as will be seen in a subsequent 
paragraph. 

The halteridial parasites of small or intermediate size, 
which I have now found to occur occasionally free from the 
corpuscles (cf. p. 663), seem to be quite ordinary in character 
and show nothing unusual. I have seen nothing at all in 
these to indicate that they undergo any transition to a 
trypaniform phase. ‘The same observation applies equally, I 
am sorry to have to say, to Halteridium noctue, where, 
in one or two cases of very strong infection, I have found 
free individuals, of varying size, to be quite numerous. 

As I pointed out in my note, the possession by an intra- 
cellular parasite of nuclear dimorphism, in the sense in which 
I have used this term, is very significant and important 
evidence in favour of a flagellate affinity or connection of the 
parasite exhibiting this feature. Indeed, ona priori grounds, 
the undeniable occurrence of this distinctive character in 
Halteridium is, even regarded by itself, a very weighty 
argument in support of Schaudinn’s view of the ontogenetic 
connection of this intra-cellular form with a trypanosome. 
When, in addition, the other evidential points to which I 
alluded were taken into account, such as the occurrence, 
now and then, of individuals attempting (as I consider) to 
develop a flagellum, and the occurrence of very small trypano- 
somes at the same time, which were no larger than the full- 
grown Halteridia, the most reasonable conclusion did appear 
to be that the two forms of parasite were indeed connected. 


STUDIES ON AVIAN HAMOPROTOZOA. 725 


I admit, nevertheless, that I am now doubtful of such an 
actual connection, especially since I have been working at 
Rovigno. Iam more inclined to think that an intra-cellular 
parasite may exhibit nuclear dimorphism, in certain conditions 
or phases as a result of a close phylogenetic relationship with a 
parasitic flagellate (say a trypanosome), without necessarily 
being any longer ontogenetically connected with one. Put 
into other words, this is to say that a parasite, such as 
Halteridium, which shows this feature, is probably derived 
from a trypanosome which has become adapted entirely to a 
resting, intra-cellular condition, and has coincidently lost, 
more or less completely, the ability to develop an active 
trypaniform phase. 

Berliner, in a recent paper entitled ‘ Flagellaten-Studien ” 
(1), has incidentally corroborated my account of the occur- 
rence of nuclear dimorphism in Halteridium by describing 
it in the case of H. noctue, i.e. in the very parasite in 
which Schaudinn first maintained it was present. Berliner’s 
figures are very striking and interesting. His preparations 
were stained with iron-hematoxylin, and another most im- 
portant point brought out by this method of staining is the 
close correspondence between the structure of the (chief) 
nucleus in the Halteridium and that of the trophonucleus 
of a trypanosome. I need not dwell upon this point here, as 
Professor Minchin and myself have already referred to it in 
our paper (13), showing the essential difference which exists, 
on the other hand, between the nuclear structure of a hemo- 
gregarine and of a trypanosome; and we shall have more to 
say about it in our own account of the parasites of Athene 
noctua. 

This fact furnishes, however, strong additional evidence in 
support of the (modified) view of a close relationship between 
Halteridium and the hemoflagellates, which I am inclined 
to prefer. On this view the gradual “ Riickbildung ” of the 
kinetonucleus—which is associated principally, of course, with 
the locomotor activities—can be readily understood, and is, 
indeed, to be expected. It accounts, further, for the com- 


726 H. M. WOODCOCK. 


paratively small size of the kinetonuclear element, as well as 
for the fact that it is not always distinguishable as a separate 
organella, differentiated from the main nucleus. On the other 
hand, such a phylogenetic connection of Halteridium with a 
trypanosome would also render it quite possible that, in certain 
cases, such as the incidence of an unusual stimulus or under 
some other special circumstances, the parasites might attempt 
to pass into—to revert to, as 1t were—a trypaniform condition. 
Thus would be explained the peculiar form of individual I 
have above described, which appears to have developed a 
flagellar thread. 

This view agrees in substance, it will be seen, with 
Hartmann’s ideas (5) of the Heemosporidia as a whole, which 
he has united with the hemoflagellates in one group—the 
Binucleata—the common character being the possession of a 
binuclear condition, i.e. of nuclear dimorphism. So far as 
the hemogregarines are concerned I do not think they show 
any evidence at all of this feature (ef. Minchin and Woodcock, 
l.c.), and therefore consider that these forms, at any rate, 
should be kept separate.! With regard to the malarial para- 
sites (e.g. Plasmodium and Proteosoma), Hartmann 
considers that these show indications of nuclear dimorphism ; 
apparently, however, the kinetonuclear element is in a more 
“yriickgebildet ” condition than is the case in Halteridium., 
Hartmann thinks, further, that these forms show other 
evidences of a hemoflagellate ancestry, such as the presence 
of a delicate, narrow, undulating membrane, with flagellar 
border in the microgametes. This opinion was maintained 
also by Schaudinn in the case of the microgametes of Hal- 
teridium. 

Not having personally studied the finer structure of the 
malarial parasites, I cannot say much about Hartmann’s 
opinion. If the above view is correct, as I consider it to be, 


1 In a later paper on this subject, which I have seen just as my MS. 
is about to go to the press, Hartmann and Jollos (‘ Arch. Protistenk.,” 
xix, p. 81, 1910) have apparently come to the same conclusion, and 
remove the hemogregarines from the Binucleata. 


STUDIES ON AVIAN HAMOPROTOZOA. ad 


in the case of Halteridium, there is nothing inherently 1m- 
probable in supposing that it holds good for the malarial 
parasites as well; this was, it will be remembered, Schaudinn’s 
idea also. The first essential point, however, is to show that 
these parasites possess a nucleus (trophonucleus) of the true 
hemoflagellate type (such as is shown by the trypanosomes 
and Halteridium), as revealed by a stain hke iron- 
hematoxylin. 

As regards the finer structural details of the microgametes 
of Halteridium, I have been unable to assure myself of the 
presence of an undulating membrane and flagellar border. I 
have examined both faintly stained and -intensely stained 
individuals, which, for all I know to the contrary, were as fully 
developed and mature as if they had been taken from the 
stomach of the insect; I have studied them with the best 
objectives and with the best possible illumination. I think the 
photos reproduced give very accurate representations of these 
delicate and minute organisms; and neither my friend, Dr. 
Reid, who has most kindly taken these photos for me, nor I 
myself, can make out such a structure. It may be there or it 
may not; I must leave the point unsettled. 

Certain of the microgametes in the photos show clearly the 
centrosomic granule at one end. The opposite end is finely 
tapering, and comparable to acytoplasmic tail; as Schaudinn 
pointed out, it does not appear to be of flagellar nature. The 
end possessing the centrosomic granule is to regarded as the 
anterior end; it is by this end that the microgamete pene- 
trates the female element, as can be distinctly seen in fig. J. 

As I mentioned in a former section, I examined particularly 
cultures inoculated with blood containing these ripe gametes, 
with a view to finding stages in the development of the 
odkinetes. Somewhat to my surprise, I could find no indica- 
tions of any developmental changes in the halteridia in the 
cultures. I saw no odkinete-like phases, and, indeed, only one 
or two halteridia which had become liberated from the cor- 
puscles, and these appeared to be degenerating and dying. 


728 H. M. WOODCOCK, 


6. Nore on LervucocyrozooN FRINGILLINARUM, N. SP. 


Habitat.—There has been considerable discussion with 
regard to the exact nature of the host-cell in which these 
Avian leucocytozoa are parasitic, some authorities stating 
that it is a leucocyte, while others regard it as an erythro- 
blast, or else an altered red cell. I have been able to assure 
myself that in the case of this species the host-cell is 
undoubtedly a uninucleate leucocyte, and not an inmature 
red cell or erythroblast.! After once carefully comparing 
them there is little difficulty in distinguishing between these 
two types of cell. Examples of immature red cells are seen 
in figs. 22 and 57, and of uninfected uninucleate leucocytes 
of about the same size, or a little larger, in figs. 23 and 58. 
The nucleus of the leucocyte is relatively larger than that of 
the other type of cell, occupying, indeed, most of the body ; 
moreover, it is nearly always eccentric in positiou, with the 
result that the cytoplasm lies chiefly on one side, whereas the 
nucleus of the erythroblast is central. ‘he appearance of 
the two nuclei is also different. The latter contains many 
small chromatic masses ; that of the leucocyte, on the other 
hand, appears to have a few large masses, which by the 
Romanowsky method of staining do not stand out so sharply 
from the general nuclear substance as in the other case. 
Further, the cytoplasm of the leucocyte is always distinctly 
paler than that of the other kind of cell. 

From the immature red cell all transitional stages occur to 
the ordinary full-sized red blood-corpuscle ; but I have seen no 
connection whatever between such cells and the others—the 
uninucleate leacocytes—which are entirely distinct. More- 
over, in no case have I found the parasites occurring in the 
former type of cell, but always only in the leucocytes. 
Wenyon, in his account of L. numide (87), figures unin- 
fected cells belonging to this type of immature red cell, above 


’ From the observations which I have so far been able to make upon 
L. ziemanni, in the little owl, I am strongly inclined to think that 
the same is true for this parasite also. 


STUDIES ON AVIAN HAMOPROTOZOA. 729 


described. He also figures a young Leucocytozoon in a 
cell which obviously corresponds to the uninucleate leucocytes 
(cf. his fig. 4 with my figs. 24 and 60). But he does not 
figure the true type of host-cell (uninfected) at all; this, I 
gather, he considers to be an immature red cell, such as he 
figures. I have no hesitation in saying—what, indeed, is 
apparent from my figures—that the uninucleate leucocytes 
(fig. 23) are the host-cells, and not immature red cells or 
erythroblasts (fig. 22). 

Effects on the Host-cell.—The young Leucocyto- 
zoon always penetrates the leucocyte on the side where 
there is most cytoplasm. It never becomes actually intra- 
nuclear, but it often has a curious position in relation to the 
nucleus during its early growing phases, appearing to be 
lodged in a deep depression or pit in the side of the nucleus 
(fig. 62). At times the parasite is almost entirely enclosed 
by the nucleus (fig. 19). This result is probably due partly 
to a tendency of the Leucocytozoon to push or sink further 
inwards, and partly to the growing out or extension of the 
nucleus, which undergoes a certain amount of hypertrophy, 
in the form of a wide crescentic or semi-circular mass, at 
the sides of the parasite. Coincidently, the nucleus under- 
goes an alteration in character, losing all indications of large, 
separate chromatic masses, and taking up the stain quite 
uniformly. As the parasite grows and expands, the free ends 
of the semi-circular nucleus are pushed outwards, and no 
longer enclose the Leucocytozoon. When the latter is 
full grown the nucleus of the containing host-cell is seen as a 
thick, curved mass at one side (figs. 20, 21, 25, and 26). 

In my preparations all the leucocytozoa are intra-cellular. 
I have never observed more than one parasite in one host- 
cell. 5; 

My observations, as also those of Wenyon (l.c.), of young 
and intermediate-sized gametocytes, intra-cellular in habitat, 
and manifestly growing into the adult individuals in a 
similar situation, do not support in the least Schaudinn’s 
view with regard to the origin of the adult gametocytes. 


730 H. M. WOODCOCK. 


Schaudinn considered that these were simply the resting- 
phases of large, sexual trypanosomes, which had come into 
relation, in a peculiar manner, with the leucocytes, causing 
the host-cell to become greatly extended and altered in form. 
I agree with Wenyon that this view cannot be sustained. 

Structure of Gametocytes.—In stained prepara- 
tions the parasites occur in two well-marked and distinct 
forms, which represent without doubt male and female game- 
tocytes, since they agree very well with these types in other 
leucocytozoa. The parasites occur in all sizes, from very 
young forms up to what are probably fully grown, mature 
individuals (figs. 19, 20, 24, and 25). Even in fairly young 
individuals the male or female character can be often recog- 
nised (figs. 19 and 24). The diameter of a rounded individual 
averages about 84 to 94; the ovoid parasite of fig. 26 is ll u 
by 63. Female forms appear to attain a slightly larger 
size than male forms. 

Comparing a male gametocyte with an individual of female 
sex, the cytoplasm of the former stains much paler than that 
of the latter, and appears to be more homogeneous in structure. 
The cytoplasm of a female individual is distinctly granular. 
The nucleus of a male form is large and somewhat diffuse ; 
it appears to contain a number of small chromatin granules 
(probably really chromatin “ dust,’ which stain pinkish. 
The female nucleus is small, compact and dense; its chromatin 
grains stain darker and more intensely than in the other case. 
Both in the male, as well as in the female form, a definite 
small chromatic body is sometimes found outside, but close 
to the chief nucleus (figs. 20, 25, and 26); it has also been 
seen in small parasites (figs. 19 and 60). This small body 
corresponds to that associated with the nucleus of L. 
ziemanni, where it was first described by Schaudinn. As 
I hope to have something to say subsequently, in conjunction 
with Professor Minchin, upon the nuclear structure of the 
latter parasite, I will not discuss this point at present, 
especially as my material is limited and all stained by 
the Romanowsky method. 


STUDIES ON AVIAN HAMOPROTOZOA. 731 


One feature about this new Leucocytozoon is of great 
interest and importance, the fact, namely, that in no instance 
observed is the cytoplasm of the host-cell extended in the 
form of a spindle at both sides. Kven where the body of 
the parasite is oval in shape, and more comparable in 
form to the deeply stained portion of the body in other 
leucocytozoa, there is no sign of any extension of the proto- 
plasm of the leucocyte. If in the case of other species, e. &. 
L. ziemanni, L. numide, this great drawing out of the 
ends of the host-cell is due merely to the parasitic influence 
of the Leucocytozoon, why does the parasite not produce 
the same effect here? I certainly think it is quite as 
probable that, in those cases where the spindle-like appear- 
ance 1s shown, there is some more material cause for this 
constant shape, and that there is really a prolongation of the 
body of the parasite,'!in the nature, perhaps, of a faintly 
staining ectoplasmic layer, for some distance at the two 
sides, to which is chiefly due this extension of the cytoplasm 
of the host-cell. Upon this point, also, I shall be able to say 
more when I have studied the preparations of L. ziemanni. 
If this is the correct explanation, it is evident that the 
Leucocytozoon of the chatinch has lost. its ectoplasmic 
layer, at least so far as can be made out. This develop- 
ment would indicate a closer adaptation to the intra- 
cellular condition, which is also seen, perhaps, in the rounded 
form of the parasite, the other species known being much 
more fusiform. 

I propose the name L. fringillinarum for this new 
species of Leucocytozoon from the chaffinch; the parasite 
found by Stevenson in the greenfinch probably also belongs 
to this same species, since, so far as I can ascertain from the 
preparation kindly given me by Stevenson, it also has the 
rounded form and does not cause the host-cell to become 
spindle-shaped. 

Of the many species of Leucocytozoon now known, only 
two or three, so far as I am aware, have been described as 
having the rounded form, and with the host-cell lacking the 


732 H. M. WOODCOUK. 


spindle-like prolongations. The descriptions of these forms 
are to be found in a series of notes by Mathis and Léger (10a— 
10p). I wish to point out that as regards one at any rate, 
and possibly more than one, of their parasites, the authors, in 
describing the gametocytes (and their host-cells) as rounded, 
appear to have been dealing simply with individuals which 
had begun the active process of rounding themselves off pre- 
paratory to rupturing the host-cell and becoming liberated as 
ripe gametocytes. Now, in preparations of the fusiform 
species (L. ziemanni and others), which show gametocytes 
caught in this act, it is generally impossible to recognise any 
longer the typical fusiform shape, the cytoplasm of the host- 
cell having been quickly disorganised. 

In the case of Mathis and Léger’s forms L. caulleryi (a 
rounded form) and L. sabrazesi (spindle-like), both trom 
the same host, namely a fowl (‘l'onkin), I teel sure that the 
latter parasite is the typical intra-cellular form of the former. 
Thanks to the authors’ kindness in sending some of their 
preparations of these parasites to the Lister Institute, I have 
been able to compare them. On a slide containing L. 
caulleryi all the individuals found are quite rounded-off, 
and, moreover, there is no sign of the host-cell in connection 
with them, i.e. the latter has been ruptured and dis- 
organised, and the parasites are seen as ripe, free gametocytes. 
A slide containing L. sabrazesi, on the other hand, shows 
the parasites still within their host-cell, the latter having the 
usual spindle-like prolongations. Mathis and Léger them- 
selves say, in their note on L. caulleryi (10a), that only 
exceptionally did they see the nucleus of the host-cell— 
evidence that the latter had been ruptured and disorganised. 
Hence I myself have no doubt, especially when the fact of 
these two parasites being found in the same host 1s considered, 
that L. sabrazesi is only a synonym for L. caulleryi, and 
that this species (L. caulleryi) belongs really to the fusi- 
form group. 

On the other hand, in the case of the species I have 
described, L. fringillinarum, there is no doubt that it is 


STUDIES ON AVIAN HY MOPROTOZOA. 733 


quite distinct from the fusiform group, since in all stages— 
from young forms right up to large gametocytes—the parasite 
and its host-cell retain the rounded form. Apparently Mathis 


and Léver’s form, L. marchouxi, from Turtur humilis 
(10c), also agrees with this type, for in this case the authors 


find the host-cell intact, the whole appearance of parasite 
and leucocyte being, so far as can be judged from the account, 


similar to that of L. fringillinarum. 


THe LISTER INSTITUTE, 


10 


April, 1910. 


BIBLIOGRAPAY. 


. Berliner, E.—‘* Flagellaten-Studien,” ‘Arch. Protistenk.,’ xv, p. 297, 


pls. 28 and 29, 1909. 


. Chagas, C.—* Studien titber Morphologie und Entwickelungscyclus 


des Schizotrypanum cruzi, n. g., n. sp., Hrreger einer neuen 


Krankheit des Menschen,” ‘Memorias Inst. Oswaldo Cruz 
(Rio de Janeiro),’ i, pp. 159, pls. 9-13, 1909. 


. Chatton, E., and Alilaire, E.—‘* Co-existence d’un Leptomonas 


(Herpetomonas) etd’un Trypanosoma chez un Muscide non 
vulnerant, Drosophila confusa,”’ ‘C.R. Soc. Biol.,’ lxiv, 
1908, p. 1004. 


. Dutton, J. E., and Todd, J. L.—* First Report of the Trypanoso- 


miasis Expedition to Senegambia (1902),” ‘Mem. Livpl. Sch. 
Trop. Med., No. 11, 1903. 


. Hartmann, M.—“ Das System der Protozoen,” ‘ Arch. Protistenk.,’ 


O07, p. 139. 


. Laveran, A.—‘*Sur un Trypanosome d'un chouette,” ‘C.R. Soe. 


Biol.,’ lv, 1909, p. 528. 


. Léger, L.—‘“Sur un Flagellé parasite de /Anopheles mac uli- 


pennis,’ ‘C.R. Soc. Biol.,’ liv, 1902, p. 534, 10 figs. 


. Leishman, W., and Statham, J. C.—* The Development of the Leish- 


man-Donovan Body in Cultivation,” ‘Journ. Army Med. Corp.,’ 
iv, 1905, p. 321, 1 plate. 


. Lithe, M.—‘‘ Die im Blute schmarotzenden Protozoen,” in Mense’s 


‘Handbuch der Tropenkrankheiten,’ iii, pt. 1, p. 69 (Leipsic: J. A. 
Barth, 1906). 

Mathis, C._—‘ Sur une modification au milieu de Novy-McNeal pour 
la culture des Trypanosomes,” ‘C.R. Soc. Biol.,’ lxi, 1906, p. 550. 


734 H M. WOODCOCK. 


10a. Mathis, C., and Léger, M—‘ Leucocytozoon de la poule,” 
‘C.R. Soe. Biol.,’ lxvii, 1909, p. 470. 


10B. “Sur un nouveau Leucocytozoon de la poule,” 
‘C.R. Soe. Biol.,’ Ixviii, 1910, p. 22. 
10c. “Leucocytozoon dune touterelle (Turtur 


humilis), ete.,” C.R. Soc. Biol.,’ Ixviii, 1910, p. 118. 

“Leucocytozoon de la perdrix du Tonkin, 
‘Ann. Inst. Pasteur,’ xxiii, 1909, p. 740, pl. 19. 

11. Minchin, E. A.—* Investigations on the Development of Trypano- 
somes in Tsetse-flies, etc.,” ‘Quart. Journ. Micr. Sci.,’ 52, 1908, 
p. 159, pls. 8-13. 


10p. 


12. “ Observations on the Flagellates parasitic in the Blood of 
Fresh-water Fishes,” ‘ Proc. Zool. Soc.,’ 1909, i, p. 2, pls. 1-5. 
13. and Woodcock, H. M.—* Observations on Certain Blood- 


Parasites of Fishes Occurring at Rovigno,” ‘ Quart. Journ. Mier. 
Sci.,’ 55, 1910, p. 113, 3 pls. 

14. Novy, F. G., and McNeal, W. J.—*On the Trypanosomes of 
Birds,” ‘Journ. Infect. Dis.,’ ii, 1905, p. 256, 11 pls. 

and Torrey, H. N.—* The Trypanosomes of Mos- 
quitoes,” ‘ Journ. Infect. Dis.,’ iv, 1907, p. 225, 7 pls. 

16. Patton, W. S.—* The Life-Cycle of a Species of Crithidia para- 
sitic in the Intestine of Gerris fossarum,” ‘ Arch. Protistenk..,’ 
xii, 1908, p. 131, pl. 9. 


15. 


Wie “Herpetomas lygei,”’ ‘ Arch. Protistenk.,’ xiii, 1908, p. 1, 
Blak: 

18. “The Life-cycle of a Species of Crithidia parasitic in the 

' intestine of Tabanus bilarius, ete.,” ‘ Arch. Protistenk.,’ xv, 

1909, p. 333, pl. 30. 

19. (with Strickland, C.)—‘* A Critical Review of Blood-suck- 
ing Invertebrates in Relation to the Life-cycles of the Trypano- 
somes of Vertebrates,” ‘ Parasitology, i, 1908, p. 522. 

20. “A Critical Review of our present Knowledge of the 


Hemoflagellatés and Allied Forms,” * Parasitology,’ 11, 1909, p. 91. 

21. Petrie, G. F—** Observations Relating to the Structure and Geo- 
graphical Distribution of Certain Trypanosomes,” * Journ. of 
Hygiene,’ v, 1905, p. 195. 

22. Prowazek, S.—‘‘Studien uber Saugethiertrypanosomen,” * Arb. kais. 
Gesundhtsa.,’ xxii, 1905, p. 1, 6 pls. 

23. Robertson, M.—** Studies on a Trypanosome found in the Alimen- 
tary Canal of Pontobdella, ‘Proc. Roy. Physic. Soe. Edin.,’ 
xvii, 1907, p. 83, 4 pls. 


24. 


25. 


26. 


27. 


28. 


29. 


30. 


dl. 


32. 


33. 


34. 


35. 


36. 


37. 


STUDIES ON AVIAN HAMOPROTOZOA. 735 


Robertson, M.—‘ Further Notes ona Trypanosome . . . from 
Pontobdella,” ‘Quart. Journ. Micr. Sci.,’ 54, 1909, p. 119, 
pie. 

“Studies on Ceylon Heematozoa: The Life-cycle of Try- 


panosoma vittate,” ‘Quart. Journ. Mier. Sci.,’ 53, 1909, p. 665, 
pls. 16 and 17. 

Roubaud, E.—“ La Glossina palpalis; sa biologie son réle dans 
Vétiologie des trypanosomiases” (‘Thése, Paris,’ 1909).  Ex- 
tract from ‘Rapport Miss. d’études maladie du sommeil au 
Congo francais (1906-1908). Laval (Barnéoud et Cie), 1909. 
[Brief notes on several of the points treated in this Memoir 
have been published in the ‘C.R. Acad. Sci.,’ exlvi, p. 423; and 
in ‘C.R. Soe. Biol.,’ lxiv, p. 1106, and lxv, p. 39 (on new Lepto- 
monas spp.), 1908. | 

Schaudinn, F.—‘ Generation- und Wirthswechsel bei Trypano- 
soma und Spirocheta,” ‘Arb. kais Gesundhtsa.,’ xx, 1904, 
p. 387. 

Sergent, E. and E.—‘Sur un Flagellé nouveau de l’intestin des 
Culex et des Stegomyia, Herpetomonas algeriense,” 
*C.R. Soc. Biol.,’ 1x, 1906, p. 291. 

— “Observations sur les Hématozoaires des oiseaux d’ Algerie,” 
‘C.R. Soe. Biol.,’ Iviii, 1905, p, 56. 

Sergent, E. and E.—‘ Etudes sur les Hématozoaires de oiseaux,” 
‘Ann. Inst. Pasteur,’ xxi, 1907, p. 251, pls. 6 and 7. 

Stuhlmann, F.—‘ Beitrage zur Kenntniss der Tsetsefliegen (G1. 
fusca, etc.),” ‘ Arb. kais Gesundhtsa.,’ xxvi, 1907, p. 83, 4 pls. 
Swingle, L. D—‘ A Study on the Life-history of a Flagellate 
(Crithidia melophagi, n. sp.), in the Alimentary Canal of the 
Sheep-tick (Melophagus ovinus),” ‘Journ. Infect. Dis.,’ vi, 

1909, p. 98, 3 pls. 

“Some Studies on Trypanosoma lewisi,” ‘Trans. Amer. 
Micr. Soc.,’ xxvii, 1907, p. 111, 1 pl. 

Thiroux, A.—* Recherches morphologiques et éxperimentales sur 
Trypanosoma padde,” ‘ Ann. Inst. Pasteur,’ xix, 1905, p. 65, 
pl. 4. 

Thomson, J. D.—‘ Cultivation of the Trypanosome found in the 
Blood of the Gold-fish,” ‘Journ. Hygiene,’ viii, 1908, p. 75, pl. 3. 


Vassal, J.—‘Sur un nouveau Trypanosome aviare,’ ‘C.R. Soc. 
Biol.,’ lviii, 1905, p. 1014. 

Wenyon, C. M.—* Report of Travelling Pathologist and Protozoolo- 
gist,” ‘ Report Wellcome Res. Lab..,’ iii, 1908, p. 121, pls. 9-16. 


736 H. M. WOuDCOCK. 


88. Woodcock, H. M.—* On the Occurrence of Nuclear Dimorphism in 
a Halteridium Parasitic in the Chaffinch, ete.,” ‘ Quart. Journ. 
Mier. Sci.,’ 53, 1909, p. 339 (text-figs.). 

“The Hemofilagellates and Allied Forms,” in Lankester’s 

‘Treatise on Zoology,’ vol. i, ** Protozoa,” fase. i, p. 193. London : 

A. and C. Black, 1909. 


39. 


EXPLANATION OF PLATES 27—381, 


Illustrating Dr. H. M. Woodcock’s paper “I. On certain 
Parasites of the Chaffinch (Fringilla ccelebs) and 
the Redpoll (Linota rufescens).” 


[All the drawings on Pls. 1—4 are drawn to a uniform magnification of 
2000 diameters. For several of the coloured figures on Pl. 1 and for 
two or three of the drawings on each of the other plates I am indebted 
to Miss Rhodes, who has kindly done them for me. | 


Plates 27 and 28. With the exception of figs. 7-15, 22, 25,57, and 58, 
all the figures relate to the parasites as found in the birds. 


PLATE 27. 
Figs. 1-6.—Trypanosoma fringillinarum, n. sp. 

Fig. 1.—Adult, ordinary individual from the blood of a chaffinch 
inoculated from a redpoll-culture. 

Fig. 2.—Ditto, from the bone-marrow of a naturally infected chaffinch. 

Figs. 3 and 4.—Slightly smaller forms ; 3, from the blood of a red- 
poll; 4, from the bone-marrow of a chaffinch. 

Figs. 5 and 6.—Small forms of the fusiform type (case B), from the 
bone-marrow of a chaffinch. 


Figs. 7-13.—Cultural forms of the trypanosome; 7-12 from chaffinch 
cultures; 13 from a redpoll one. 
Figs. 7-9, and 13.—Trypanomonad forms (6 and 7 days). 
Fig. 10.—Early trypaniform type (40 hours). 
Figs. 11 and 12.—Examples of equal and unequal binary fission. 


Figs. 14-18.—Halteridium fringille (Labbé). 


Fig. 14.—Female individual. 
Figs. 15 and 17.—“ Indifferent ” individuals. 
Fig. 16.—Special form, free in the blood-plasma, with chromatic 


STUDIES ON AVIAN HAHMOPROTOZOA. 737 


line. (Unfortunately the terminal granule has not come out in the 
plate.) 
Fig. 18.—Very young form. 


Figs. 19-21, 24-26, Leucocytozoon fringillinarum, n. sp. 


Figs. 19 and 24.—Young gametocytes, female and male. 
Figs. 20 and 26.—Large female gametocytes. 

Figs. 21 and 25.—Large male gametocytes. 

Fig. 22.—Immature red blood-corpuscle. 

Fig. 25—Leucocyte (uninfected). 


PLATE 28. 
Figs. 27-56.—T. fringillinarum. 

Figs. 27-33.—Ordinary definitive forms of the parasite of varying 
size; 27 from a chaffinch inoculated with redpoll culture, 28-31 from 
naturally infected chaffinch, 32 and 33 from naturally infected redpoll. 

Figs. 34-38.—Large, massive forms, from a redpoll. 

Fig. 39.—? Transitional form, intermediate between ordinary type 
and that last mentioned, from a redpoll. 

Figs. 40-45.—Series of fusiform parasites from very small to a 
moderate size, from a chaffinch (Case B). 

Figs. 46-54.—Small forms from a chaffinch (Case A), fusiform or 
broad and stumpy ; 48 and 54 show indications of division. Many of 
the individuals in both these series show the granular chain or line. 

Figs. 55 and 56.—Remarkably slender individuals (? young, definitive 
forms). 

Figs. 57 and 58.—Immature red blood-cell or erythroblast and unin- 
fected leucocyte, respectively. 


Figs. 59-62.—Leucocytozoon fringillinarum., 
Fig. 59.—Male individual. 
Fig. 61.—Female individual. 
Figs. 60 and 62.—Young forms, probably female individuals. 
Figs. 63-70.—Halteridium fringille. 


Figs, 63, 65, and 66.—Medium-sized to large female forms. 
Fig. 64.—“ Indifferent ” individual. 

Figs. 67-69.—Small or intermediate-sized individuals. 
Fig. 70 a and b.—Male gametes. 


PLATE 29. 
Figs. 71-111.—Cultural forms of T. fringillinarum. 


[All the figures are from original cultures of 6-8 days, except figs. 


738 H. M. WOODCOCK. 


74 and 109, which are from a second sub-culture of 26 days, specially 
for comparison. | 
[(c) indicates chaffinch culture ; (R) redpoll-culture. | 

Figs. 71-86.—The ordinary trypanomonad type, showing variations 
in size and in degree of development of the membrane. 

Figs. 71-76, 83-86 (c) ; figs. 77-82 (R). 

Figs. 87 and 88.—Individuals in which the kinetonucleus is a trifle on 
the aflagellar side of the trophonucleus ; in fig. 88 division is just being 
inaugurated. Both (c). 

Figs. 89-95.—Stages in equal binary fission. All (c) except fig. 85, 
which is (R). 

Fig. 96.—Division-form of sub-equal character, giving rise to indivi- 
duals of the accentuated trypanomonad kind. 

Fig. 97.—Accentuated trypanomonad individual (Cc). 

Figs. 98-104.— Various stages in the unequal division of the accen- 
tuated trypanomonad individuals. Figs. 100 and 105 are (R), the rest 
are (C). 

Figs. 105-111.—Illustrative of the two kinds of individual which 
result from unequal fission. Figs. 105, 107 (upper half), 108-110, 
accentuated trypanomonad forms, often more or less club-shaped, with 
nuclei far back and well-developed membrane; Figs. 107 (lower 
half), 106 and 111 a and B, fusiform individuals, with only slightly 
developed membrane; note the comparatively short flagellum. Figs. 
106 and 108 (R), rest (C). 


PLATE 30. 


Figs. 112-131, 153-163.— Cultural forms of T. fringillinarum 
(contd.) 

Figs. 112-114.—Pear-shaped forms, probably derived from the 
smaller halves of unequal divisions, which have not become fusiform. 
Figs. 112 and 114 (c), 113 (R). 

Figs. 115 and 118.—Small and large individuals of the accentuated 
trypanomonad kind. passing into the ovoid or rounded condition. 
Both (c). 

Figs. 116, 117, and 119.—Medium-sized rounded forms. (All ¢). 

Figs. 120-125.— Individuals from the (c) culture which showed a 
pronounced tendency to develop large massive forms. Many of them 
are undergoing division. 

Figs. 126 and 127.—Early trypaniform individuals. (c) forty hours. 

Fig. 128.—? Transition form from trypanomonad to trypaniform type. 
(c) 6 days. 

Fig. 129.—Small trypaniform individual. (c) 6 days. 


STUDIES ON AVIAN HAMOPROTOZOA. 739 


Figs. 150 and 131.—Greatly elongated trypaniform individuals. (c) 6 
days. 

Fig. 132.—Trypaniform phase of a trypanosome of Athene noctua 
from the stomach of Culex pipiens. 

Figs. 133, 134, and 136—Rounded forms still possessing a flagellum, 
but lacking any signs of an undulating membrane. In the two first a 
large vacuole is present. Fig. 135 (c), figs. 134 and 136 (rR). 

Figs. 135 and 137.—Rounded forms without a flagellum (R). 

Figs. 138 and 139.—Small rounded forms in a dying condition; the 
two nuclei are gradually disappearing (R). 

_ Figs. 140-146.—*“ Pseudo-herpetomonad” forms, illustrating various 
degrees in the loss of the membrane and attached part of the flagellum. 
Figs. 140, 145, and 146 (c), 141-144 (R). 

Fig. 147.—Herpetomonad form (c). 

Figs. 148 and 149.—Pear-shaped forms, with little or no attached 
part to the flagellum (R). 


Figs. 150-163.—Al]l] these forms are from a (R) culture of 19 days. 


Figs. 150-153.—Large, altered, unhealthy parasites, with a develop- 
ment of granular substance in the region of the base of the flagellum. 

Fig. 154.—Parasite showing two large vacuoles. 

Fig. 155.—Individual with a kinetonucleus, but no trophonucleus. 

Fig. 156.—Individual with one kinetonucleus and two trophonuclei. 

Fig. 157.—Dividing parasite, showing how the unequal distribution of 
the nuclei, as found in the two last forms, is brought about. 

Fig. 158.—Parasite with three trophonuclei for one kinetonucleus. 

Fig. 159.—Showing the splitting off of an individual witha flagellum, 
but with no nuclear substance at all. 

Fig. 160.—A free, active individual, with no definite nucleus of either 
kind, but with scattered granules. 

Fig. 161.—Showing the splitting-off of a portion of the cytoplasm 
containing only a few granules. 

Figs. 162 and 163.—Forms showing irregular multiplication of the 
different organelle. 


PLATE 31. 


[The micro-photographs on this plate were all taken for me by my 
friend Dr. D. J. Reid, to whom I wish here to express my deep sense of 
his kindness and to offer my sincere thanks. It is as well to point out, 
perhaps, that the more deeply stained parts have come out, in most 
cases, relatively far too dark. 

The magnifications are as follows (approximately): Figs. a—p 1630, 
fig. © 620, fig. F 500, fig. @ 550, figs. H and s 1630, fig. x 1840, 
fig. L. 1220.] 

VOL. 0D, PART 4,—NEW SERIES. 4.9 


740 H. M. WOODCOCK. 


Figs. A-D.—Trypanosoma fringillinarum, as found in the birds. 
For description of these figures see under figs. 2, 3,28, and 54, which are 
of the same individuals respectively. [In the reproduction the whole 
length of the delicate aflagellar prolongation, which is visible in the 
actual photos, cannot be made out. Unfortunately there are two small 
pieces of débris lying on the parasite of fig. B, which are, of course, 
reproduced. One lies about one third of the distance from the kineto- 
nucleus to the trophonucleus ; the other on the fold of the membrane 
opposite to the nucleus. In the drawn figure (fig. 3) these particles are 
omitted. | 

Figs. E-G.—Agglomeration clusters of various sizes of T. frin- 
gillinarum in cultures. [The parasites of the first two clusters 
are not so nicely stained, unfortunately, as those of the third, but they 
show the manner of formation of the cluster. | 

Fig. H.—Halteridium fringille; female individual showing 
nuclear dimorphism (the same is drawn in fig. 14). 

Fig. J.—Fertilisation of a macrogamete by a microgamete. Note 
that the latter is penetrating by the end which has the centrosomic 
granule. 

Figs. kK and t.—Microgametes. 


Duarrt, ousrr Mier Sci Vol, 55, WERT 


Ath Lith? London. 


. 


AVIAN HA MOPROTOZOA. 


IJuant.Sourn Mire Vol. bI NSA. 


Soh eon eat 
I en a eS 


.. } f 


Quart.cfounn. Mier Sei. Vl. 5 NEGLI 


i 
H 


i 
hee 


DPROTOZOA. 


———- >. “a 


Iuant, Sunm.cber Sei. Vol, §5 NSB 3O 


Se 


me 


Luar. Journ. Mure Su Vol, 65. NS G4 34 


Huth,coll 


MP ROTOZOA. 


STUDIES ON CEYLON HAMATOZOA. 741 


Studies on Ceylon Hematozoa. 


No. II.—Notes on the Life-Cycle of He mogregarina nicorie, 
Cast. and Willey. 


By 
Muriel Robertson, M.A. 


With Plates 52-41 and 1 Text-figure. 


CoNTENTS. 


, PAGE 
T. OccURRENCE OF THE PARASITE } : . 42 
Il. Brier SuMMARY OF THE LIFE-HISTORY ; 742 


Ill. PHASES OF THE HM#MOGREGARINE IN THE BLOOD OF 

THE TORTOISE 2 : : . 143 
IV. STAGES OF THE H#MOGREGARINE IN THE LEECH a aol! 
V. GENERAL REMARKS AND CONCLUSIONS 758 


Iy 1904 Drs. Castellani and Willey (8) described a hemo- 
gregarine from the blood of the common lake-tortoise of 
Ceylon, Nicoria trijuga. ‘They named the parasite H 2mo- 
gregarina nicorie after its host. Shortly afterwards 
these authors gave a somewhat fuller account of their obser- 
vations in the‘ Quarterly Journal of Microscopical Science’ (4). 

While in Ceylon in 1907-08 I was able, largely through 
the kindness of Dr. Willey, to collect the material described 
in the following pages. My observations agree in the main 
with those of the earlier observers already cited. I have, 
however, been able to supplement their results and to give 


742 MURIEL ROBERTSON. 


an account of some of the processes which take place in the 
intermediate host, the leech Ozobranchus shipleyi. 


I. OccURRENCE OF THE PARASITE. 


Nicoria trijuga occurs in very large numbers all over 
Ceylon. It generally frequents ponds, lakes, and rivers, but 
specimens are sometimes found living a semi-terrestria! exist- 
ence in places removed from water. The tortoises which 
have adopted the drier habitat occasionally show ticks, but I 
have never found them infected with the hemogregarine. I 
did not, however, examine a sufficiently large number of 
individuals to be able to draw the conclusion that the dry- 
dwelling tortoises are never infected. 

The Nicorias from the usual aquatic habitat are very often 
infected with the hemogregarine. It does not seem to 
produce any pathogenic effects even when present in large 
numbers. No other blood-parasites were ever observed in 
association with the hemogregarine. The intestinal parasites 
were not investigated, but it may be noted in passing that a 
Bodo-like flagellate was found on two occasions in the gall- 
bladder. 

The only ectoparasites present were ticks on dry-land 
tortoises and leeches on the water-dwelling tortoises. The 
leeches belonged to a species of Ozobranchus; only once 
was an isolated Glossiphonia found upon a Nicoria. I 
found that tortoises from all parts of Ceylon showed the 
hemogregarine. I never, however, investigated individuals 
from more than an elevation of 1500 feet. Generally speak- 
ing I found the up-country reptiles were free from blood- 
parasites. 


IJ. Brrer SuMMARY OF THE LIFE-HISTORY. 


For the sake of clearness it is, I think, advisable to give a 
brief account of the life-history of the form under discussion, 
in so far as it has been made out, before treating the various 
points in detail. 


STUDIES ON CEHKYLON HAMATOZOA. 743 


The hemogregarine in the blood of the tortoise shows the 
usual two types, a bean-shaped and a recurved type. Certain 
of the bean-shaped individuals, namely, the large forms with 
a nucleus in which the chromatin is rather loosely arranged, 
give rise by a process of schizogony in the lung to a great 
number (about seventy) of large merozoites. Another type 
of schizogony is found in the circulating blood-corpuscles, 
and arises also from bean-shaped individuals, This results 
in the formation of a small number (six to eight) of mero- 
zoites of quite small dimensions. It appears that the form 
which gives rise to this second type of merozoite is itself 
derived from the schizogony in the lung. It is probable that 
the small merozoites give rise to the gametocytes. ‘I'he reason 
for this assumption is given in another part of the paper. 
When the hemogregarines are taken into the crop of the 
leech, Ozobranchus shipleyi, together with the blood of 
the tortoise, certain of the hemogregarines pass into the 
intestine, and are there found as motile vermicules. They 
penetrate into the intestinal wall, where the differentiation of 
the hitherto indistinguishable gametes takes place, culminating 
in a process suggesting anisogamous conjugation. ‘The 
zygote breaks up to form eight sporozoites, which pass 
through the intestinal wall into the blood-spaces. ‘The 
hemogregarine is probably passed into the blood of the 
Nicoria through the contamination of the wound by the leech 
while feeding. 


III. PHaAsts oF THE H@&MOGREGARINE IN THE BLOOD OF THE 
‘TORTOISE. 


In the living state the hemogregarine may easily be dis- 
tinguished as a clear sausage-shaped inclusion in the red 
blood-corpuscles. The protoplasm is slightly more granular 
at one end than the other, and the nucleus can be seen as a 
sharply defined clear area. ‘The parasites do not show any 
sign of movement when they are observed upon a sealed 
slide, but free vermicules are very occasionally found in 


744 MURIEL ROBERTSON. 


serous blood that has been allowed to stand exposed in 
the air. The addition of salt-solution to the blood some- 
times causes the hemogregarines to quit the corpuscle, but 
never in large numbers. Altogether, it may be said that 
H. nicoriw shows far less tendency to become motile 
in the blood than the majority of the species of hamo- 
gregarines, 

The greater part of the blood-films made were preserved 
by the drying method and stained with Giemsa; a few were, 
however, fixed while still wet in sublimate-acetic, and treated 
by wet methods throughout. ‘l'hese wet films were stained 
with iron hematoxylin, and it has been clearly shown that 
wet fixation followed by hematoxylin, hemalum, or other 
suitable stain, gives far truer pictures than those obtained by 
the Giemsa method. All the detail of structure, etc., described 
were worked out on the wet films. 

Parts of the various organs, such as the spleen, liver, and 
lungs, were also preserved (in Flemming, corrosive-acetic, 
and Bles’s fluid) and sections made.  Bles’s fluid was found 
to give an exceedingly good fixation of the blood-corpuscles 
and of the parasites they contained, especially in the tissue - 
from the lung. When stained with hemalum a very clear 
and precise picture was obtained, and the results derived 
from a study of the films could thus be corroborated and 
criticised by means of the section material. 

In the stained films it can be seen that the parasite is sur- 
rounded by a delicate sheath or capsule. ‘The nature of this 
capsule shows the greatest possible variation in different 
members of the genus Hemogregarina. In some species 
it is a thick refractile envelope, which opens to let out the 
enclosed parasite when the motile phase is adopted. Even 
when the capsule is more delicate it is often capable of 
persisting for a time after the hemogregarine has escaped. 
This has been observed by many workers; Castellani and 
Willey (4) have shown it in H. mirabilis, Dobell, in a form 
from Boa constrictor (6); I have myself seen the same 
thing in H. triedrus. In H. nicoriez the capsule is rather 


STUDIES ON CEYLON HAMATOZOA. 145 


difficult to demonstrate; iron-hematoxylin films or those 
counter-stained with eosin are the best for this purpose. The 
capsule never persists after the parasite has escaped, in fact 
it seems that in this case the envelope may be said to dis- 
integrate rather than to be shed in the usual way. The 
capsule is to be seen quite clearly in the live state in in- 
dividuals from the crop of the leech, especially at the time 
when the blood-corpuscle has been already digested away, 
but before the parasite has passed down to the intestine, 
where it becomes motile. 

The protoplasm is delicately alveolar, and is sometimes 
slightly granular ; chromatoid particles outside the nucleus are 
very rare, and this form does not show the curious eosinophile 
inclusions found, for instance, in H. vittate (11). ‘The 
nucleus consists, as a general rule, of a number of isolated 
chromatin granules arranged, often rather symmetrically, 
round a small central body (see figs. 1-3). The peripheral 
grains of chromatin may be connected by strands with the 
central granule. ‘I‘his central granule cannot be called a 
karyosome in anything approaching the same sense in which 
this word is applied in protozoan literature generally. In the 
nucleus of this hemogregarine it is only the position that 
marks off the central body from the peripheral chromatin 
granules ; it is in no way distinguished from them in size or 
staining reaction, and in those cases where the chromatin 
granules are less regularly arranged (fig. 9A) it is quite impos- 
sible to pick it out with certainty. Nevertheless, it appears 
to me to be of a diiferent nature from the other uuclear 
elements, in so far that, in the very primitive nuclear division, 
it seems to form a kind of centrodesmose. Not infrequently 
the peripheral chromatin granules are joined to one another, 
a chromatin ring being thus formed all round (see figs. 2, 5, 
and 8). It must not be supposed that this chromatic ring 1s 
truly a nuclear membrane; it takes the chromatin stains 
deeply, and assumes a bright red colour with ‘I'wort’s stain. 
It is thus in sharp contrast to the green membrane found by 
this method round the nucleus in, for instance, some trypano- 


746 MURIEL ROBERTSON. 


somes and certain amoebze. Iam inclined to think that it is 
simply formed by the running together of the grains of 
chromatin. Forms are sometimes found in the blood of the 
tortoise which show the chromatin arranged in an inner and 
outer ring; this type is shown in fig. 5. Finally, forms are 
also seen in which the chromatin is in the shape of a large 
number of irregularly disposed granules, which may at times 
give the appearance of a kind of reticulum (figs. 6, 17, and 
18). In FH. nicorie, as in almost all the species known, there 
are, in addition to the young forms, two types in the blood of 
the vertebrate host, the one a bean-shaped organism with an 
approximately central nucleus, the other a long recurved 
creature with its nucleus situated in the broader limb near 
the bend (see figs. 1, 3, 8, 9, 17 and 18). ‘he bean- 
shaped form is always present in far greater numbers than 
the fully developed vermiform individuals, but specimens are 
very common where the more slender recurved limb is only 
about half as long as the broad limb (figs. 2and 9a). This is 
one of the points in which the wet fixation method is so much 
superior to the dried films. In the latter the great majority 
of these specimens, where the recurved limb is shorter than 
the broad one, appear simply to be bean-shaped, the drying 
having artificially obscured the recurved limb. ‘l'here are no 
very marked or constant nuclear differences in these types ; 
it may, however, be observed that generally speaking the 
larger bean-shaped forms show the more scattered arrangement 
of the chromatin. The smaller bean-shaped individuals and 
the half recurved creatures have usually the more symmetrical 
circular type of nucleus, while the large vermiform specimens 
have a slightly elongated nucleus, with a tendency for the 
chromatin masses to run together at their edge. A glance 
at the figures will make these points clear. 

Two main theories as to the significance of the bean-shaped 
and vermiform (fully recurved) creatures have been put 
forward: (1) That the bean-shaped individuals are macro- 
gametes or macrogametocytes, and the recurved ones micro- 
gametes or microgametocytes. These two different types or 


STUDIES ON CEYLON HAMATOZOA. 747 


their immediate derivatives are by this view expected to con- 
jugate in the intermediate host and give rise to the sexual 
cycle. (2) ‘lhe second view considers that the bean-shaped 
creatures are responsible for the endogenous cycle within the 
vertebrate, while the recurved vermiform type carries on the 
life-history in the intermediate host (15). It appears that in 
H. nicorie, at all events, the schizonts (in both types of 
schizogony) are beau-shaped when they enter upon the process 
of schizogony (see figs. 10 and 19). I am inclined to think, 
however, that too much importance has been attached to the 
difference in shape between the recurved and bean-shaped 
individuals. ‘lhe recurving is an appearance caused by the 
growth in length of the parasite inside its capsule, and there 
seems to be evidence (fig. 8a) which goes to show that the 
recurved part is capable of being reabsorbed as the parasite 
increases in width. It is therefore not improbable that 
certain of the schizonts are really derived from the vermiform 
individuals. 

Although doubly and trebly infected corpuscles are to be 
seen, I have never come across any trace of binary fission nor 
of any process that could reasonably be interpreted as conju- 
gation within the corpuscle. Hahn (8) has recently described 
this process, but I have not been able to corroborate his 
results. 

Schizogony.—Two quite different types of schizogony 
occur in the vertebrate host. The one takes place in the 
lung, each schizont giving rise to a very large number (about 
seventy) of large merozoites. The other takes place in the 
circulating blood-corpuscle, each schizont producing six to 
eight quite small merozoites. 

Schizogony in the Lung.—tThe first stage is shown in 
fig. 19, and is from a section of the lung; it represents a 
bean-shaped hemogregarine, rather larger in size than those 
found in the blood-stream. There is a delicate envelope 
round the creature, the protoplasm is rather granular, and 
there is a single nucleus with the chromatin arrauged in 
small irregular grains. The hemogregarine is not contained 


748 MURIEL ROBERTSON. 


in a blood-corpuscle, but is apparently lying free in a capillary 
of the lung. The schizont now increases immensely in size, 
and the nucleus multiplies by successive divisions. The 
mitosis is of a very simple type ; the amount of chromatin 
seems to augment by division of the granules, the nucleus 
becomes slightly elongated, the central body divides, and the 
strand of staining material which connects them appears to 
play the réle of a simple spindle. The chromatin granules 
now become loosely grouped about each new central body, 
and the connecting strand disappears. From the scarcity of 
division-figures one is inclined to think that this primitive 
mitosis must take place very rapidly (figs. 16, 21). 

During these processes of growth and nuclear multiplica- 
tion the shape of the body is inaintained, and there results a 
very large bean-shaped or sausage-shaped organism sur- 
rounded by a membrane. It is circular in section (figs. 
21-23), and contains a large number of nuclei; I have 
counted about seventy, but the number appears to vary. <A 
point of some interest is that very little, if any, diminution 
takes place in the size of the nuclei; it will be observed, also, 
in the figures that they are evenly distributed through the 
cell-body and not arranged at the periphery. 

The protoplasm finally segregates round the nuclei, and 
there are formed a corresponding uumber of merozoites, 
which still lie within the envelope. ‘They are presently set 
free as sausage-shaped hemogregarines of 6 to 75 in 
length, that is to say, only little below the average size 
(8 to 10)! of the hemogregarines seen in the blood. ‘They 
have usually rather regular nuclei of the rounded or slightly 
elongated type. 

Schizogony in the Blood-corpuscle (figs. 10-16).— 
In the blood of practically all the infected tortoises examined 
multinucleate hemogregarines were found in greater or less 
numbers. ‘hese forms may show any number of nuclei up to 
eight ; generally, however, they do not show more than six. 
From a study of the early binucleate phases it is clear that 
these speciniens arise from bean-shaped hemogregarines (figs. 


STUDIES ON CEYLON HAMATOZOA. 749 


10 and 11).!' The parasite remains inside the blood-corpuscle 
(fig. 11 is a case where the creature has been liberated 
mechanically in the making of the film), and does not undergo 
any increase in size. Finally, the protoplasm segregates 
round the small, slightly elongated nuclei, and a correspond- 
ing number of little falciform merozoites are formed inside 
the original envelope (figs. 14 and 15). This stage is rather 
difficult to find and must be of short duration, as it is some- 
what rare in comparison with the number of multinucleate 
creatures to be found in the blood. In the cases I‘ have 
found the number of merozoites is six, but I should expect 
that eight may sometimes be formed, as rare stages with 
more than six nuclei are to be seen (see fig. 15). Schizogony 
stages of this type occur in blood from any part of the 
tortoise. ‘he merozoites, which are much smaller (4) than 
those formed in the lung, finally escape and penetrate into 
another blood-corpuscle, where they proceed to grow. It is 
unfortunately almost impossible to trace the subsequent 
career of these young forms with any satisfying measure of 
certainty. ‘There are, practically speaking, no distinctive 
features to lay hold of, and once they have increased in size 
there is nothing to distinguish them from other forms. The 
impression I have gained in my attempts to follow their 
development is that they grow into a compact bean-shaped 
creature of no great size (see figs. 4. and 7). The nucleus is 
inclined to stain deeply, and is composed of separate granules, 
which may be arranged irregularly or in a circle—the latter 
is on the whole the more common. Beyond this pot I have 
not been able to trace these forms; I was always working 
with natural infections, which appeared to be ot a chronic 

1 The question arises as to whether these bean-shaped forms which 
give rise to the schizogony in the peripheral blood are derived from the 
vermiform type. The evidence to be drawn from the infections of 
H. nicoriz which I examined is very inconclusive. In H. vittate, 
a form parasitic in the tortoise Emyda vittata, however, the recurved 
type appears only relatively late in the infection, and I am therefore 
inclined to think it is associated with the later periods of schizogony 
and possibly with the process as it occurs in the peripheral blood. 


750 MURIEL ROBERTSON. 


type and generally of long standing. It is obvious that only 
by following the successive stages of the infection in a 
previously clean tortoise can points like this be really con- 
clusively determined. 

Interpretation of the 'l’wo Types of Schizogony.— 
There are three views which might be put forward in explana- 
tion of the facts: (1) That the schizogony in the lung with 
the large merozoites gives rise to the female gametes, and the 
schizogony in the blood-corpuscles to male gametes. This 
view is, I think, inadmissible, as it is very unlikely that the 
small male gametes should be produced in such small numbers, 
namely six to eight to one parent individual, while the female 
gametes are produced in large numbers—about seventy to 
one parent individual. 

(2) The second, and I think more probable, oxplanaianll is 
that the schizogony in the lung is the endogenous asexual 
multiplication, and that certain of the merozoites thus formed 
proceed in turn to form gametocytes by the schizogony in the 
blood-stream. 

(3) A third quite plausible explanation is that the schizogony 
in the lung is brought about by the newly injected parasite— 
that is to say, it is the first activity of the hemogregarine 
upon arriving in the vertebrate host. Miller’s (9) account of 
Hepatozoon perniciosum, Chagas’ (5) work on Schizo- 
trypanum, and Aragao’s (1) on Hemoproteus columbe 
furnish parallels for such an interpretation. On this view the 
schizogony in the blood-corpuscle would be the later, and, so 
to speak, chronic process of multiplication, which would at 
some period culminate in gamete formation. 

I think the evidence is strongest in support of the second view 
(2) put forward, namely that the schizogony in the lung is the 
asexual multiplication, and that in the blood gamete-forma- 
tion. A somewhat important point against view (3) is the 
fact that the schizont in the lung does not appear to penetrate 
a lung-cell, which one would expect it to do did it arrive in 
the lung as a free vermicule (sporozoite). Moreover the 
possession of an envelope in so early a stage as that shown 


STUDIES ON CEYLON HAMATOZOA. TOL 


in fig. 19 strongly suggests that it has had an endo-corpuscular 
existence. I have never seen any sign of the parasite reach- 
ing the lung by being engulfed by leucocytes, and I am there- 
fore inclined to think that the schizonts in the lung must 
have come from the blood-corpuscles. 


IV. Sraces In THE LEECH. 


Before giving an account of the stages of Heemogregarina 
nicoriz observed in the leech, it will, I think, be well to 
describe the more important features of the leech itself. 

The form in question belongs to the Rhynchobdellid genus 
Ozobranchus. Mr. W. A. Harding, to whom the leech was 
sent for identification, found that it belonged to a new species, 
and called it Ozobranchus shipleyi. It isa small aquatic 
form carrying a row of feathery gills on each side of its body. 
The creature rarely reaches more than about one third of an 
inch in length even when fully extended. Generally speaking, 
it is fonnd attached to the tortoise at the back of the neck, 
round the sockets of the limbs, and more rarely upon the 
ventral side near the throat. The leeches have a tendency to 
assemble together in groups—a habit they preserve even when 
keptin a glassdish. The gills of the Ozobranc hus are kept 
in constant motion, and the animal dies if left out of water for 
any length of time. I was not very successful in getting the 
leeches to live for long in captivity, nor was I able to discover 
exactly what was amiss in the conditions to which they were 
exposed. Possibly the smaller quantity of water rose to too 
high a temperature. Leeches are usually very hardy and live 
well in captivity. I had no difficulty in keeping Pcecilob- 
della alive in Ceylon for months. I have often observed, 
however, that newly fed specimens are much less resistant 
than fasting individuals, and this seems true of a number of 
different species of leech. Almost all the Ozobranchus I 
got were either in the act of feeding or newly fed, and there- 
fore in the least favourable condition. This leech seems to 
show a much closer adaptation to its host than generaliy 


7oe MURIEL ROBERTSON. 


obtains among the group. Thus it was never found upon 
Hmyda (the milk tortoise) living in the same lake with the 
Nicoria, nor upon the siluroid fish Saccobranchus, nor upon 
the water-snakes which shared the same habitat. Even in 
an area so restricted as a well, these leeches were only found 
to infest the Nicoria. Moreover, Ozobranchus lays its eggs 
upon the carapace of the tortoise; they are of a dark brown 
colour, closely resembling that of the tortoise, and are so 
firmly cemented on that it requires a knife or some fairly 
sharp instrument to detach them. It appears that the leeches 
move readily enough from one tortoise to another, but it is 
difficult to make out exactly how they are adapted to the 
terrestrial night-wandering of their host. The Nicoria spends 
all the day sleeping in the water and comes to land to prowl 
around at night, so most likely the leeches feed during the 
day and drop off at night. Generally speaking I got more 
leeches from nicoria caught in the evening, but there were, 
however, some exceptions to this; presumably these were 
cases where the tortoise had spent the night either in the 
water or in a damp place. Ozobranchus is capable of 
executing rather feeble swimming movements, and, in addi- 
tion, can creep around upon its suckers in the usual way. The 
time taken to digest a meal seems to vary from about three 
to seven days, according to the size of the leech. : 

In Ozobranchus shipleyi the proboscis leads into the 
crop, which is a wide, very extensible sac dividing into two 
large lobes at its lower end. The intestine opens from the crop 
at the point where the division takes place. The upper end of 
the intestine, which is rather wide, shows four long diverticula 
on each side (see fig. in text, p. 753). This wide part of the 
intestine terminates in a kind of chamber which opens by a 
narrow communication into a simple coiled tube, which leads 
to the exterior attheanus. For some reason the most infected 
part of the gut wall is almost always this chamber at the end 
of the wide intestine. The accompanying diagram, which was 
made from reconstructions of sections by the glass-plate 
method, shows the relations of the various parts of the ali- 


STUDIES ON CEYLON HAMATOZOA. ie 


mentary tract. The cells lining the intestine are very large 
and richly ciliated; their protoplasm has a strong affinity 
for all nuclear stains, including the red element in Twort’s 
stain. ‘lhe nuclei are very large and reticulate, often showing 
several karyosomes. 

The stages of the parasite in the leech had to be studied 
for the most part upon section material; sublimate acetic and 
Flemming’s fluid were the fixatives used. The leeches were 
usually placed between two slides, so as to prevent undue 


Diagram of the alimentary tract of Ozobranchus shipleyi. 


retraction. As regards staining, Delafield’s hematoxylin, 
Twort’s stain, thionin, methyl-blue eosin, and Mayer’s hema- 
lum were all used with good effect, hamalum and Delafield 
being the most generally useful. Heidenhain’s iron-hema- 
toxylin was quite impossible, as it darkened the whole 
intestinal region so intensely that, long before that region 
was sufficiently colonrised, the remainder of the section was 
completely bleached. 

I am indebted to Mr. Peter Jamieson for the skill with 
which he has cut the many sections required. 

I may mention in passing that the intestine of the leech is, 


754 MURIEL ROBERTSON. 


as a general rule, extraordinarily free from bacteria, schizo- 
mycetes, etc., and although a total of about 150 leeches were 
examined, I never found them to contain any flagellates, or, 
indeed, any protozoan parasites other than the hemogregarine. 
A large number of observations upon live material from the 
leeches were made in the hope that the sequence of the 
processes might be followed by direct observation, but this 
proved to be impossible, as the development occurs in the 
tissues of the leech. 

The blood upon being taken up by the leech is stored in 
the large crop, where the blood-corpuscles undergo a gradual 
degeneration. ‘The blood passes in small quantities into the 
intestine, where it is digested and absorbed. Blood-corpuscles 
are never found, nor even their nuclei, in a recognisable state 
in the intestine, and this holds good even in the case of a 
newly fed leech. A large number of live observations were 
made, but no motile hemogregarines were ever found in the 
crop. ‘his particular haemogregarine appears to be digested 
out of the corpuscle (figs. 25 and 26), and only to become 
motile when it passes into the intestine. I am persuaded 
that this cannot be universal amongst hemogregarines ; so 
many species react almost instantly to the mere shedding of 
the blood that I expect in other cases the parasites will be 
found to become motile at once upon being taken into the 
intermediate host. Motile hemogregarines are to be found 
in the intestine at intervals all through the digestion, but 
except in cases where the blood is very rich in parasites, 
there are never a very large number present at one time. 
The hemogregarine never makes any attempt to attack the 
wall of the crop. 

A number of hemogregarines seem to degenerate in the 
crop (fig. 27), but degeneration stages are only rarely found 
in the intestine; it seems to fare with these, as with the 
blood-corpuscles, that they disintegrate before reaching the 
intestine. So far as my observation goes, neither the large 
bean-shaped forms nor the completely recurved individuals 
are to be recognised in the intestine. The individuals which 


STUDIES ON CHYLON HAIMATOZOA. 759 


are met with in this situation have a round or oval nucleus 
with the chromatin grains fairly regularly arranged (figs. 
28-34), and seem, as far as morphological features are con- 
cerned, to be the motile phases of such types as are shown in 
figs. 1-3 and 9a from the blood of the Nicoria, and figs. 
21 and 26 from the crop. The protoplasm of the hamo- 
gregarines have very little affinity for most stains, and this 
is particularly true of the stages in the leech. 

The motile creatures carry out movements of flexion and 
also of contraction and extension; in addition to this they 
can glide by means of very shallow undulations passing down 
the body. This constricting motion, as in analysis it really 
is, is most strikingly seen in H. leschenaultii (a heemo- 
gregarine from Hemidactylus leschenaultii), but the 
difference is purely one of degree. 

The faculty of contracting and extending the whole body 
shown by the motile forms of H. nicoriz is a disturbing 
factor when an attempt is being made to divide the parasites 
into different categories. After much searching, I have 
come to the conclusion that the only distinction between the 
parasites while still in the lumen is one of size, and [ 
consider this to have practically no value when one remembers 
the capacity of the creature for stretching, and the great 
difficulty in getting a correct idea of bulk in an animal of 
this type. The drawings have been made from sections, and 
here one has the additional danger of not always getting the 
animal in a perfectly horizontal position. 

It was noticed not infrequently in the live specimens from 
the intestine that two equal individuals ranged themselves 
side by side, but complete fusion was never observed. In the 
sections this association in couples was again found (see 
fig. 35), and the individuals showed no differentiation. Here, 
also, stages indicating complete fusion were not seen; only 
the two cases figured (figs. 56 and 37) were observed, and 
as both these are cut obliquely they are not particularly 
convincing. I therefore think that if appearances such as 
those shown in figs. 35-37 relate to conjugation at all, they 

VOL. 55, PART 4.—NEW SERIES, 50 


706 MURIEL ROBERTSON. 


are only instances of (perhaps precocious) association. After 
a time (figs. 38-41) the hemogregarines penetrate the intes- 
tinal wall, where appearances quite different from those just 
described suggest conjugation of a type closely resembling 
that found by Siedlecki (14) in Adelea ovata, and by 
Perez (10) in Adelea mesnili. Fig. 44 gives a picture of 
an early stage; the macrogameta has become differentiated 
as a large rounded organism with a nucleus in which the 
chromatin is beginning to form a rather diffuse mass instead 
of being arranged in the definite granules seen in the motile 
phase. The nucleus of the microgametocyte is very com- 
pact, and stains deeply, the protoplasm has not fused with 
that of the microgamete, nor does it appear to do so sub- 
sequently. 

From appearances such as fig. 45, the microgametocyte 
nucleus seems to divide into three or four, of which two or 
three, as the case may be, remain outside and degenerate ; 
they sometimes persist for a long time, and are to be seen 
forming a dense mass of chromatin on the edge of the 
sporocyst (see figs. 49, 51, 52). The division of the nucleus 
of the microgametocy te into four is probably the more normal 
condition, the cases where three are formed being most likely 
due to the suppression of one of the divisions. One of the 
four microgamete nuclei thus formed appears to pass into 
the protoplasm of the macrogamete; unfortunately quite 
clear pictures of the fusion of the gamete nuclei and the first 
division of the zygote nucleus were not found. Fig. 478 
shows a condition suggesting the latter stage, but in view of 
certain reactions on the part of the host-cell to be noted 
later, I do not feel perfect confidence in this interpretation. 

It is quite impossible to pass over these appearances without 
noting their very probable significance as conjugation and 
their close resemblance to the fertilisation of Adelea; at the 
same time I am fully aware of important gaps in the series. 
Great caution is required in interpreting these appearances, 
as degenerating hemogregarines are occasionally found in 
the gut wall. Moreover, the host-cell seems sometimes in 


STUDIES ON CEYLON HAMATOZOA. 757 


strong infections to react to the presence of the parasite 
by the formation of internal masses resembling the mucoid 
globules described by Leger and Duboscq (8 4). 


Formation of Sporozoites. 


The further development of the parasite culminates in the 
formation of eight sporozoites. A membrane is secreted 
round the protoplasm, forming a kind of cyst-wall, but it 
appears to be thin and not very resistant. Fig. 46 shows an 
early stage where there are only two nuclei present. Subse- 
quent divisions occur, and appearances such as fig. 48 are 
produced, where the larger nucleus at one end of the creature 
is preparing for division. Finally (see figs. 49-53), the 
protoplasm segregates round the nuclei, and there are pro- 
duced eight individuals; these when fully developed show 
considerable resemblance to the free motile forms found in 
the lumen of the intestine, and are of much the same size. 
‘The sporozoites are set free in the wall and pass out into the 
blood-spaces (see fig. 47c, 54-56), where they can be dis- 
tinguished from the corpuscles of the leech by their shape 
and characteristic nuclear appearance. 

There is a well-marked correlation between the processes 
of digestion in the leech and the condition of the parasite. 
In a recently fed leech the free motile forms are numerous in 
the intestine but no multiplicative stages are to be seen in 
the wall. Later on the hemogregarines have penetrated the 
wall, but only the earlier stages are present. Still later ripe 
cysts with fully formed sporozoites are found in considerable 
numbers in good infections. Quite late towards the end of 
digestion, when the crop is empty, the sporozoites have for the 
most part escaped into the blood-spaces, and the intestinal 
wall is once more almost free from parasites. | 

I have not been able to carry my investigations beyond 
this point, and cannot say by what means the hemogregarines 
are passed back into the blood of the tortoise. In spite of 
much searching I have never found motile stages of the para- 


758 MURIEL ROBERTSON. 


site in the proboscis, nor do they appear in this region in the 
sections. 


V. GeneraL REMARKS AND CONCLUSIONS. 


When the foregoing account was all but complete, I 
received Dr. Reichenow’s (12) interesting preliminary note 
on H. stepanovi. The results I have obtained coincide in 
all essential points with those of Reichenow, and the evidence 
he has obtained upon the question of conjugation is much 
more conclusive than that brought forward by myself, as he 
has figured the first two divisions of the zygote-nucleus. The 
type of conjugation is clearly the same in the two cases. The 
only point of divergence in the two life-cycles is the schizo- 
gony in the vertebrate host; in H. stepanovi this takes 
place in the bone-marrow and always occurs inside the blood- 
corpuscle, the number of merozoites not exceeding twenty- 
four. This difference is the main justification for preserving 
the species name of H. nicorie. 

There is scarcely a single point in the development of H. 
stepanovi as described by Siegel (138) which is in agree- 
ment with the results obtained by Reichenow, or with what I 
have myself observed in H. nicoriz. I have never seen 
the formation of the minute microgametes, nor the sporu- 
lating stages in the blood-spaces of the leech, nor the worm- 
like sporozoites which he describes. It would appear that 
this worker must have been dealing with conditions differing 
widely from those presented by the leeches I examined. 

It will be observed that the life-cycle of H. nicoriz 
differs in one or two points from that of Hepatozoon 
perniciosum, the hemogregarine of the rat, described very 
completely by Miller. The most important divergence occurs 
in connection with conjugation and the formation of sporo- 
blasts, which in turn produce sporozoites. ‘The sporozoites 
never become motile in the mite, and the parasite returns to 
the rat by way of the alimentary tract when the rat eats the 
mite. 


STUDIES ON CEYLON HAMATOZOA. 759 


The life-cycle of H. nicoriz at once recalls the processes 
observed in Coccidia, but there are two points of difference 
which are, I think, important as diagnostic characters. 
Firstly, at no stage does H. nicoriz show in its nucleus the 
karyosome so characteristic of the coccidia; secondly, the 
sporozoites are not enclosed in a resistant cyst, and become 
motile within a relatively short time after they are formed 
without the stimulus of transference to another host-indi- 
vidual. In all the coccidia hitherto described the sporozoites 
remain dormant, until by one means or another they pass to 
the exterior, and are taken up by another individual of suit- 
able species where the sporozoites are set free. As regards 
the question as to whether the stages in the leech might not 
belong to an independent parasite, and have no connection 
with H. nicorix, the following points may be urged: The 
close correspondence between the stage of digestion and the 
development of the parasite, the strong morphological resem- 
blance between such stages as those figured in figs. 1, 2, 3, 
9a, 25, 26, 28-34, 38-41, 51, 54-56, derived respectively 
from the blood of the tortoise and different parts of the leech, 
and the apparent absence of the parasite in leeches taken 
from uninfected tortoises. Lastly, on the hypothesis that 
the stages in the leech are independent of those in the 
tortoise, the only other group in which the forms from the 
leech could be placed is that of the Coccidia. The points of 
divergence noted in the preceding paragraph are, I think, 
sufficiently important to distinguish them from any form 
belonging to that group. The point is, of course, one which 
could be determined experimentally when suitable material is 
available. 


LISTER INSTITUTE, 
April, 1910. 


760 MURIEL ROBERTSON. 


List or REFERENCES. 


1. Aragao, H. de Beaurepaire.—‘ Ueber den Entwicklungsgang und 
die Uebertrigung von Hemoproteus columbe,”’ ‘Arch. f. 
Prot.,’ vol. xii, p. 157. 

2. Balfour, H.—‘ A Hemogregarine of Mammals,” ‘Second Report 
of the Wellcome Research Laboratories, Khartoum,’ p. 97. 

3. Castellani and Willey —‘“ Observations on the Hawmatozoa of Ver- 
tebrates in Ceylon,” ‘Spolia Zeylanica,’ ii, p. 78, 1904. 

ibid., ‘Quart. Journ. Micr. Sci.,’ vol. 49, p. 583. 

5. Chagas, C.—‘ Ueber eine neue Trypanosomiasis des Menschen,” 
‘Inst. Oswaldo Cruz,’ i, 1909, p. 159. 

6. Dobell, C. C._—* Some Notes on the Hemogregarines Parasitic in 
Snakes,” ‘ Parasitology,’ vol. i, No. 4, p. 289. 

7. Doflein, F.— Lehrbuch der Protozoenkunde,’ 1909. 

8. Hahn, C. W.—‘ The Stages of Hwemogregarina stepanovi 
Danilewsky found in the Blood of Turtles, with Special Refer- 
ence to Changes in the Nucleus,” ‘ Arch. f. Prot.,’ vol. xvii, 1909. 


8a. Léger and Duboscq.—* Les grégarines et l'épithélium intestinal 
chez des Trachéates,” * Arch. de parasitologie,’ 1902. 

9. Miller, W. W.—“‘ Hepatozoon perniciosum (n.g., n. sp.), a 
Hemogregarine Pathogenic for White Rats: with a Description 
of the Sexual Cycle in the Intermediate Host, a mite Lelaps 
echidninus,” ‘ Bull. No. 46 Hyg. Lab. U.S.A. Pub. Health and 
Marine Hosp. Service, Washington,’ 1909. 

10. Perez, Ch.—* Le cycle evolutif de /Adelea mesnili,” ‘ Arch, f. 
Prot.,’ Bd. ii, 1903. 

10a. Prowazek, J. v.—‘* Untersuchungen iiber Heemogregarinen,” ‘Arb. 
kaiserl. Gesundheitsamte,’ vol. xxvi. 

11. Robertson, M..—‘‘ Preliminary Note on Hzmatozoa from some 
Ceylon Reptiles,” ‘Spolia Zeylanica,’ vol. v, p. 178, December, 
1908, 

12. Reichenow, E.—‘ Der Zeugungskreis von Hemogregarina 
stepanovi,” ‘Sitzb. der Gesell. nat. Freunde, Berlin,’ No. 1, 1910. 

13. Siegel, J—‘ Die geschlechtliche Entwicklung von Hemogre- 
garina stepanoviim Riusselegel Placobdella catenigra,” 
* Arch. f; Prot. Bad. 1; 19038. 

14, Siedlecki, M—‘ Etude cytologiqne de Adelea ovata Schn.,” 
‘Ann. Inst. Pasteur,’ xiii, p. 169. 

15. Sambon, L. W., and Seligmann, C. G.—‘“ The Hzemogregarines of 
Snakes,” ‘ Path. Soc. Trans.,’ vol. lviii, Part ITI, 1907. 


lor 
—_ 


STUDIES ON CEYLON HAMATOZOA. 7 


EXPLANATION OF PLATES 32—41, 


Illustrating Miss Muriel Robertson’s paper on “Studies on 
Ceylon Heematozoa.”’ 


[The figures are all drawn with the Abbé camera at a uniform magni- 
fication of 2400 diameters. | 


Figs. 1-24 represent stages from the vertebrate host Nicoria trijuga. 
Figs. 1-15 (with the exception of fig. 9) are stages from the blood 
treated by the wet method throughout and stained with Heiden- 
hain’s iron-hematoxylin. Figs. 16-24 are from sections of the lung 
stained with Mayer's hemalum. 

Fig. 1.—Bean-shaped hemogregarine with circular type of nucleus. 

Fig. 2.—Half-recurved specimen with circular nucleus; the chromatin 
is symmetrically arranged and the central granule is visible. 

Fig. 3.—Bean-shaped specimen, the stain further extracted. 

Fig. 4.—Small form derived from schizogony in the blood. 

Fig. 5.—Bean-shaped specimen showing the chromatin in the nucleus 
arranged in two rings, one within the other. This creature has been 
set free mechanically in the making of the film. 

Fig. 6.—Bean-shaped specimen with reticulate nucleus. 

Fig. 7.—Small compact specimen. 

Figs. 8 and 9.—Fully recurved (vermiform) specimens. Fig. 9 is 
from a dried film stained with Giemsa. 

Fig. 8a.—Recurved specimen where the recurved limb is being 
reabsorbed. 

Fig. 94.—Half-recurved individual, rather broad, and with a large 
nucleus containing irregularly arranged chromatin. 

Figs. 10-13.—Early stages of schizogony in the blood-stream. 

Figs. 14 and 15.—Final stages of schizogony in the blood-stream. 
Fig. 15 has been decolourised to a greater extent than fig. 14; both show 
six merozoites. 


Fig. 16.—Early stage of above type of schizogony from section of 
the lung; one of the nuclei is undergoing division. 

Figs. 17 and 18.—Bean-shaped specimens from section of lung. 

Fig. 19.—KEarliest stage of schizogony in the lung. 

Fig. 20.—Slightly later stage of schizogony; the schizont has 
increased in size and the nucleus has divided. 


762 MURIEL ROBERTSON. 


Fig. 21.—Still later stage; some of the nuclei appear to be preparing 
for division. 

Fig. 22.—Multinucleate schizont cut across in section. 

5 

Fig. 23.—Late stage of schizogony in lung, the protoplasm beginning 
to segregate round the nuclei. 

Fig. 24.—Fully formed merozoites; only a very few of the total 
number formed are shown in the section. 


Figs. 25—56 represent stages in the leech Ozobranchus shipleyi 
Harding. 


Figs. 25 and 26.—Non-motile stages from the crop. 

Fig. 27.—Degenerating stage from the crop. 

Figs. 28-32.—F ree motile stages in the lumen of the intestine of the 
leech. 

Figs. 33-37.—Association in the lumen of the intestine. 

Figs. 38—41.—Early stages in the cells of the intestinal wall. 

Fig. 42.—Precocious differentiation of microgamete. 

Fig. 43.—Early stage of macrogamete. 

Figs. 44 and 45.—Stages suggesting conjugation. In fig. 44 the 
microgametocyte is lying closely applied to the macrogamete. In 
fig. 45 the microgametocyte appears to be giving rise to the micro- 
eamete nuclei, one of which will fuse with the nucleus of the macro- 
gamete. 

Fie. 46.—Early stage of sporocyst showing two nuclei. 

Fig. 47—(A) Stage apparently representing a zygote; (B) stage show- 
ing what appears to be the first division of the zygote nucleus; (Cc) free 
sporozoites in the wall of the intestine. 

Fig. 48.—Stage of above showing five nuclei, of which one is prepar- 
ing to divide. , 

Fig. 49.—Sporocyst with eight nuclei; the protoplasm has not yet 
divided up; the rejected microgamete nuclei are still visible. 

Figs. 50-53.—Sporocysts showing sporozoites. 

Figs. 54 and 55.—Motile sporozoites escaping through the cells of 
the intestinal wall. 


Fig. 56.—Sporozoite in blood-space of the leech. 


Se 


QUART. JOURN. Micr. Scr., Vou. 55, n.s. PU. 


QuaART. JoURN. Micr. Scr., Vou. 55, N.S. 


Pity (Bi), 


» © 


oF 


abe 


12, Bis, 


QUART. JouRN. Micr. Scr., Vou. 55, Ns. 


Caley 


Journ. Mager SOL, VOR. oo. NS: Pr. 36. 


Z& 


Ya 


oe 


5a 


Quart. Journ. Micr. Scr., Von. 55, n.s. Pu. 37. 


; 
: 


i> hal 
Quart. JounN. Micr. Scr., Vou. 55, N.S. Pu. 38. — 


| 
. 
| 
| 


ae eee 


5 7 


ate, Se ——— oe y " 


Quart. Journ. Mrcr. Scr, Von. 55, ns. PL, 39. 


Quart. JouRN. Mier. Sctr., Vou. 55, n.s. Pr. 40. 


_—s 


ithe 


STINGING-CELLS IN CRASPEDOTE MBEDUS#. 763 


On the Origin and Migration of the Stinging- 
Cells in Craspedote Medusz. 


By 
Charles L. Boulenger, M.A.(Camb.), 
Lecturer on Zoology in the University of Birmingham. 


With Plates 42 and 43 and 5 Text-figures. 


ConreNTS. 


PAGE 
1. Introduction. ; : ; . 763 
2. The Stinging-cells of the Adult Medusa of Merisia . 765 
3. Migrating Stinging-cells in other Meduse . a eau 
4. The Development of the Medusa of Mcrisia 771 
5. General Conclusions : : : BA7ic) 
6. Bibliography . : ; Bigs! 


1. IyrropucTion. 


Lewis Murbach (1), in 1894, definitely established the fact 
that the stinging-cells of the Hydromeduse have the power 
of active movement in the tissues by the formation of pseudo- 
podial processes from the cnidoblasts. 

These observations were confirmed by K. C. Schneider! (2) 
who published a detailed account of the development of 
nematocysts in A galmopsis and other Siphonophora, and in 
his paper stated emphatically that: 

“Alle Nesselzellen der Siphonophoren enstehen an locali- 

' Schneider, as early as 1890, pointed out the fact that developing 
nematocysts were excessively rare in the tentacles of Hydra, and 
suggested that they might be formed on the body of the animal. He did 
not, however, pursue the subject any further (véde “‘ Bibliography,” 8.) 

VOL. 55, PART 4,—NEW SERIES. D1 


764 CHARLES L. BOULENGER. 


sierten Bildungsherden, von denen aus sie in einem bestimmten 
Entwickelungsstadium als Wanderzellen auf die Verbrauchs- 
stiitten tberwandern.” 

The subject has been recently revived by Jovan Hadzi (4) 
in a remarkable paper in which he records his observations on 
the thread-cells of marine hydroids. Hadzi’s results are of the 
greatest interest, as he was able to examine the living tissues 
as well as preserved material. His main conclusions are as 
follows : 

(1) The thread-cells of hydroids are not formed “in situ” 
but in the ectoderm of the ccenosarcal branches, where, on 
account of the thick perisarcal investment, they can obviously 
not become functional. 

(2) When completely developed, except for accessory 
structures such as the cnidocils and the stalks, they migrate 
to the important nematocyst batteries on the tentacles. 
This migration can take place in two different manners. In 
simple forms, e.g. Campanularia, the thread-cells move 
actively by means of their pseudopodia, making their way 
between the ectodermal cells of the colony. In Tubularia 
however, they adopt a quite different method of locomotion : 
from the ectoderm of the coenosare they force a way through 
structureless lamella and endoderm into the cavity of the 
hollow stem, whence they are carried by the current caused 
by the flagella of the endoderm cells to the hydranths. 
Here the thread-cells re-enter the tissues and migrate 
actively by their own movements to the ectoderm of the 
tentacles. 

In a recent paper, whilst describing the structure of the 
Egyptian lacustrine medusa, Mcerisia lyonsi (5), I called 
attention to the fact that large nematocysts were to be found 
in abundance among the endoderm cells of the manubrium. 
Being at a loss to account for their presence in this position 
I cut sections of a large number of specimens, careful 
examination of which convinced me that I was dealing with 
a case similar to that investigated by Hadzi in Hydroids. 
As this phenomenon has not been described previously in 


——ES 


STINGING-CELLS IN CRASPEDOTE MEDUSA. 765 


Medusze, I have endeavoured in this paper to give as complete 
an account as possible of the origin and distribution of the 
nematocysts of this form. 

The material used for this investigation was collected by 
Dr. Cunnington and myself in Lake Qurun, and was carefully 
fixed either with osmic acid or with hot corrosive sublimate. 
Sections were cut by the ordinary paraffin method and a 
number of stains were tried, the best results being obtained 
with hematoxylin followed by eosin; this produced an 
excellent double-stained effect, the eosin bringing out the 
nematocysts and rendering them most conspicuous. Borax 
carmine followed by picro-indigo-carmine was another good 
differential stain and iron-hematoxylin was useful when 
examining sections of the developing Medusee. The work in 
connection with this paper was carried out partly in the 
Morphological Laboratory at Cambridge and partly in the 
Zoological Laboratory of Birmingham University. I wish to 
express here my sincere thanks to Professor F. W. Gamble, 
who very kindly read through my manuscript and made 
many valuable suggestions. 


2. Tur STINGING-CELLS OF THE ADULT MeEpusA or Marista. 


As mentioned above, a striking feature of the anatomy of 
this medusa is the presence of numerous thread-cells! in the 
endoderm at the base of the manubrium. At first it seemed 
possible to account for their occurrence in this unusual 
position by assuming that these stinging-capsules were used 
ones taken in by the jelly-fish together with its food. On 
careful consideration this view was found to be quite un- 
tenable, for— 

1 The nomenclature of the different parts of the stinging-cells is 
somewhat cumbrous and complicated; moreover, the various names 
have been used very loosely. In this paper I have employed the terms 
thread-cell or stinging-cell for the whole structure comprising the 
nematocyst (the actual stinging capsule), and nematoblast (the 
cell in which the former is embedded, and of which the enidocil and the 
stalk are parts). 


766 CHARLES LL. BOULENGER. 


(a) The nematocysts found in the endoderm are always 
undischarged. 

(b) Favourable sections show them to be accompanied by 
their nematoblasts. 

(c) The nematocysts are never to be found near the free 
margins of the endoderm cells, but, for the most part, 
between the more basal portions of these cells near the 
structureless lamella. 

These thread-cells can, obviously, not become functional in 
this position, and the only possible explanation of their 
occurrence here is that they are making their way from their 
place of origin to some battery where they can be of use. 

At this point it may be well to review the distribution of 
stinging-cells in the ectoderm of the manubrium. The chief 
battery is situated around the mouth-opening; here the 
thickened ectoderm formsa circular lip crowded with nemato- 
cysts, and constitutes a powerful organ of offence (Pl. 42, 
figs. 3and 4). ‘lhe ectoderm of the remainder of the manu- 
brium proper consists of a single layer of low, closely fitting 
epithelial cells with occasional isolated nematocysts ; it is to 
be noticed that here, as well as on the oral lip, interstitial 
cells are completely absent. At the base of the manubrium 
is the broad stomach, the ectoderm of which is considerably 
thickened and forms the conspicuous gonad. 

Interstitial cells and developing thread-cells being absent 
from the more distal parts of the manubrium, the question 
arises— Where are the nematocysts of the oral battery formed, 
and how did they attain their position in this region? An 
answer is, I think, afforded by the study of the distribution 
and arrangement of the nematocysts in the manubrial endo- 
derm. ‘The greatest number of these are to be found just 
below the region of the gonad, where, in most specimens, 
numerous thread-cells are to be met with among the large 
digestive cells of the endoderm. In this position one can 
usually find a number of dark-staining interstitial cells, some 
of which contain rudiments of stinging-capsules, and are 
obviously nematoblasts (Pl. 42, figs. 1 and 2). 


STINGING-CELLS IN CRASPEDOTE MEDUSA. 767 


In the more distal parts of the manubrium we find nemato- 
eysts to occur less abundantly, and their position in the 
endoderm is very regular, the longer axes of the capsules 
being parallel with the structureless lamella and their broader 
ends directed towards the mouth of the medusa (PI. 42, fic. 
3). Previous authors have shown this orientation to be 
characteristic of migrating thread-cells, and we must come 
to a similar conclusion; namely, that they are making their 
way from the base of the manubrium to the oral battery. 
This view is confirmed by an examination of the tissues of 
the mouth region, where one can often find thread-cells 
actually forcing their way through the structureless lamella 
to the oral battery. Here they take up their definitive position 
and develop accessory structures, e.g. cnidocil and stalk, 
from the nematoblast. A  stinging-cell occasionally turns 
aside before reaching the oral region (Pl. 42, fig. 4), and 
passing through the lamella, forms one of the isolated 
nematocysts to be met with in the more proximal parts of the 
manubrial ectoderm. 

The route followed by the thread-cells of the medusa is 
readily explained. These structures, when the nematocysts 
are completely developed, are of considerable size, whereas 
the ectoderm of the manubrium is very low, and, moreover, 
forms a very definite epithelium of closely fitting cells, 
between which the large stinging-cells could scarcely force a 
passage. We need, therefore, not be surprised that they 
adopt the much easier way between the large and _ loosely 
packed cells of the endoderm. 

From the above account it appears, therefore, that in 
Meerisia the nematocysts of the oral battery of the medusa 
are developed in the endoderm at the base of the manubrium ; 
this does not necessarily imply that the nematoblasts are 
themselves endodermal in origin, as will be explained in the 
section of this paper which deals with the development of 
the medusa-bud. 

In addition to that surrounding the mouth opening, the 
main nematocyst batteries of the medusa are situated on the 


768 CHARLES LL. BOULENGER. 


four perradial tentacles suspended from the umbrella edge. 
These tentacles are slender and of great length when fully 
extended; at their bases they are swollen to form the very 
conspicuous ocellar bulbs, each of which bears on its ex- 
umbrellar surface a bright red eye-spot. The tentacles are 
hollow, their cavities being continuous with that of the 
circular canal; the ectoderm is thickened at regular intervals 
to form conspicuous transverse rings crowded with nemato- 
cysts, and becoming very noticeable and almost bead-shaped 
when the tentacles are fully extended. 

On examination of sections and maceration preparations of 
these organs, one is again struck by the almost complete 
absence of nematoblasts or other interstitial cells, and we are 
driven to the only possible conclusion, namely, that the 
stinging-cells have developed elsewhere and have migrated to 
the batteries on the tentacles. The large, eye-bearing bulbs 
at the bases of the tentacles immediately suggest themselves 
as possible nematocyst “factories,” and sections of these 
structures show that such a function must be assigned to them 
(Text-fig. 1). 

An ocellar bulb consists of a mass of thickened ectoderm 
crowded with small, irregularly shaped cells and nematocysts 
in various stages of development. The fully formed thread- 
cells are devoid of enidocils or other accessory structures, and 
the capsules are never orientated so as to lie at right angles to 
the surface; we must, therefore, conclude that they do not 
become functional in this region. In the centre of the bulb 
the nematocysts lie in all directions, but near the base of the 
tentacle we find a distinct tendency for these organs to be 
arranged with their longer axes parallel with the structureless 
lamella, a position, as mentioned above, characteristic of 
migrating thread-cells. 

The above-mentioned facts lead us to the conclusion that 
the stinging-cells of the tentacles, like those of the oral 
battery, are not developed “in situ,” but migrate into these 
organs from “ factories” situated ina more central position on 
the medusa, in this case from the ocellar bulbs, whence a 


STINGING-CELLS IN CRASPEDOTE MEDUSA. 769 


TEXT-FiG. 1. 


A longitudinal section through the ocellar bulb and the base of a 
tentacle of Mcerisia lyonsi.  tent.b. Ocellar bulb. nem. 
Nematocyst migrating into the tentacle. 


TEXT-FIG. 2. 


Section of the umbrella edge of Merisia lyonsi showing the 
velum (vel) and part of an ocellar bulb (tent.b.) nem. 
Nematocyst migrating towards the edge of the velum, 


770 CHARLES LL. BOULENGER. 


continual stream of thread-cells are being poured forth. ‘The 
majority of these are obviously on the way to their tentacular 
batteries, although occasionally one may wander into the 
velum, as shown in T'ext-fie. 2. 


3. MIGRATING STINGING-CELLS IN OrHeR Mepusm. 


In the preceding paragraph I have attempted to prove that 
the conspicuous bulbous swellings which occur so constantly 
as the bases of the tentacles of craspedote Medusz have an 
important function besides that of bearing the ocellar sense- 
organs. In such craspedote Medusz as are devoid of tentacle- 
bulbs, e.g. the Trachomedusez and Narcomeduse, we 
find that the edge of the umbrella is provided with a special 
thickened 1ing of ectoderm, containing stinging-cells, some- 
times known as the “ nettle-ring.” Further, those forms in 
which the tentacles take their origin some distance from the 
margin of the bell on the exumbrellar surface are provided 
with special bands of nematocysts, called peronia, which 
connect the above-mentioned uettle-ring with the bases of 
the tentacles. These facts make it very tempting to assume 
that the marginal ring of nematoblasts replaces the ocellar 
bulbs in function, and reference to the figures of this organ, 
given by various authors, seems to show that this assumption 
is probable correct. It is a point which requires special 
investigation, and I will at present merely refer to the 
evidence which is at my disposal. 

The Hertwigs’ most accurate figure of the umbrella edge 
of Carmarina (6, Pl. iv, fig. 5)! shows the nettle-ring to 
be packed with thread-cells without definite orientation ; at 
the base of the tentacle, however, a number of nematocysts 
are drawn arranged in such a manner that there can be 
little doubt that they are migrating from the marginal ring 
to the batteries on the tentacle. I have examined sections 
through the tentacles of a medusa of the same genus, and 


'T should like to express my indebtedness to Dr. 8. F. Harmer, F.R.S.. 
for calling my attention to this figure. 


STINGING-CELLS IN CRASPEDOTE MEDUS. rival 


these showed the same orientation of nematocysts as in the 
specimen figured by the Hertwigs. I have figured one of 
these sections (‘Text-fig. 3) chosen from a series in the Cam- 
bridge Morphological Laboratory ; comparison with that of 
Mecerisia (Text-fig. 1) is very instructive. 


TEXT-FIG. 3. 


A longitudinal section through the base of a tentacle of Car- 


marina sp. vel. Velum. nem. Nematocyst migrating through 
the ectoderm of the tentacle. 


Giinther’s figure of Limnocnida (7, fig. 6) shows that a 
similar migration of thread-cells must occur in that medusa. 


4. Tue DEVELOPMENT or THE MeEpuUsA oF Marista. 


As shown above, the nematocysts of the main stinging 
batteries of Mcerisia are formed in two quite distinct positions 
in the medusa: (a) The manubrial endoderm, (>) the ecto- 
derm of the ocellar bulbs. 

In order to properly understand the origin of these 

VOL. 05, PART 4.—NEW SERIES. 52 


Tae CHARLES L. BOULENGER. 


different situations of the stinging-cell factories it is necessary 
to examine the development of the medusa in some detail. 

Until recently the accepted view of the development of the 
gonophores of the Hydromedusz was based essentially on L. 
Agassiz’s observations on Syncoryne mirabilis, published 
in 1862 (8). His account of the process was confirmed by 
Hertwig (9), Weismann (10), and almost all later workers on 
the same subject, and is essentially that to be found in the 
majority of modern text-books. ‘The following description of 
the development of the medusa of Bougainvillea is taken 
from one of the latter (18), and represents the prevailing ideas 
on the subject : 

The medusa-bud makes its first appearance as a simple 
hollow bud formed by the evagination of the two layers of 
the mother-polyp. Multiplication of the ectodermal cells at 
the apex results in the production of a lens-shaped mass of 
small cells which sinks below the level of the superficial 
ectoderm, pressing the endodermal wall in front of it into 
the shape of a cup. This mass of ectoderm is called the 
entocodon (Glockenkern), and a cavity which appears in its 
interior is the rudiment of the subumbrella cavity. It is 
followed by an invagination of the superficial ectoderm, the 
wall between the new cavity thus formed and the subumbrella 
cavity being the future velum. Growth of this subumbrella 
cavity results in an approximation of the endodermal walls of 
the coelenteron, and these ultimately fuse into an endoderm 
lamella except where the circular and radial canals are to 
lie. The upgrowth of the manubrium from the floor of the 
subumbrella cavity, the formation of the tentacles and the 
perforation of the velum and manubrium complete the 
formation of the medusa. 

A. Goette (11) has recently made a thorough examination 
of the development of the gonophores of Podocoryne 
carnea and a large number of other hydroids, and has 
published a long and elaborate paper on the subject. As 
the result of his investigations this author concludes that the 
current views on the origin of these structures are quite 


STINGING-CELLS IN CRASPEDOTE MEDUSA. 773 


erroneous, and states that carefully cut series of sections of 
developing medusa-buds show that a double-walled eup of 
endoderm is not present at any stage; moreover, the four 
radial canals arise from four unconnected pouches of endo- 
derm which grow out separately, although simultaneously, 
from the ccelenteron of the bud, and are completely indepen- 
dent of the entocodon. The endoderm lamella is formed 
later by the lateral extensions of the solid edges of these 
pouches, which finally fuse with one another. Again, an 
invagination of the superficial ectoderm does not take place 
and the forecast of the velum is present at a quite early 
stage, and is then represented by the flattened apex of the 
bud, where the superficial ectoderm and the distal wall of the 
entocodon come into contact with one another. 

Goette’s paper has not received (at any rate in this country) 
the attention which so important a communication deserved, 
and the only confirmation of his results is that of his pupil, 
Walter Richter (18), who, acting on his professor’s advice, 
worked cut the development of the gonophores in Rhizo- 
phora, Physalia, and other Siphonophora and described a 
similar origin for these structures in this division of the 
Hydromedusz. 

In my account of the anatomy of the hydroid stage of 
Meerisia I did not go into this subject with any detail, but 
merely stated that the development of the medusa-buds 
seemed quite typical, the growth of the cavity in the ento- 
codon causing the approximation of the endodermal walls of 
the bud, 

The examination of a large series of sections during my 
investigation of the origin of the nematocysts has shown me 
that this statement was erroneous, and that the development 
of the medusa of this form agrees very closely with that of 
Podocoryne carneaas desribed by Goette. My error, 
like that of other writers on the same subject before Goette, 
was due to the use of optical sections, and partly to the 
examination of single sections of the buds instead of complete 
series. 


7174 CHARLES L. BOULENGER. 


In Meerisia lyonsi the medusa-buds are to be found 
scattered irregularly on the broadest region of the hydranth 
between the bases of the tentacles (Text-fig. 4), thus differing 
in position from the asexual lateral buds, which are restricted 
in the majority of cases to the more proximal parts of the 


body. 


TrxtT-FIG. 4. 


Outline sketch of a hydranth of Merisia lyonsi to show 
developing gonophores and a small asexual bud. x 30. tent. 
Tentacle. m.b. Gonophore with conspicuous ocellar bulbs. 


as.b. Asexual lateral bud. 


The ectoderm of this region is somewhat deeper than in 
other parts of the hydroid, the boundaries of the large 
musculo-epithelial cells are difficult to detect, and the whole 
tissue is crowded with interstitial cells, for the most part 
nematoblasts, containig nematocysts i various stages of 
development. The endoderm consists of large vacuolated 


a 
EE 


STINGING-CELLS IN CRASPEDOTE MEDUSA, 715 


digestive cells, between which are numerous characteristic 
gland-cells with coarse granular contents which stain deeply. 

The first indication of a developing medusa-bud is to be 
traced in the ectoderm, an accumulation of interstitial cells 
causing this layer to project slightly outwards. The endo- 
derm soon begins to take part in this bulging out of the 
tissues, and owes its increase in area chiefly to the prolifera- 
tion of the large cells, but partly also to the accumulation of 
interstitial cells, which are to be found in the endodermal 
tissue in the region of a developing bud. These cells I 
believe to be ectodermal in origin, for favourable sections 
show occasional interstitial cells to migrate from the ectoderm 
through the structureless lamella into the endoderm. In this 
way a hollow, double-layered bud is formed (Pl. 43, fig. 5) 
by a process which cannot be called one of simple evagination, 
but in some respects resembles that of the formation of the 
early stages of the lateral buds in Hydra, as recently des- 
cribed by J. Hadzi (14). 

As long ago as 1891, W. B. Hardy (15) showed that in the 
early development of the gonophores of Myriothela 
phrygia there was a certain mixing up of endodermal and 
ectodermal cells to form a kind of blastema, and it seems 
probable that further investigations will prove that the pro- 
duction of a bud from the body of a hydroid is by no 
means so simple a process as has been made out by some 
authors. 

The entocodon is next formed by the proliferation of the 
ectoderm at the apex of the bud, and consists of a small-celled 
plug of tissue between ectoderm and endoderm. Four pouches 
of endoderm are arising simultaneously from the ccelenteron ; 
from them the radial canals of the adult are to be derived. 
Reference to fig. 6 will show that there is nothing of the 
nature of a double-walled endodermal cup in the bud, one 
side of the obliquely cut section showing a radial pouch, the 
other the contact of the entocodon with the superficial 
ectoderm. 

It is to be noticed that this superficial ectoderm has not 


776 CHARLES L. BOULENGER. 


changed in character and is identical in structure with that 
covering the hydranth, consisting of large epithelial cells, 


interstitial nells, and nematoblasts, with occasional nemato- 


cysts. 

The independent origin of the four radial pouches of endo- 
derm is still more obvious in figs. 7, 8, and 9, which are 
three sections in different planes of a slightly later stage. 
In the transverse section (fig. 7) the entocodon is seen to be 
roughly square in section, being in contact with the super- 
ficial ectoderm at the four corners (interradii); the four 
perradial pouches are thus completely separated from one 
another. A median longitudinal section (fig. 8) through the 
perradii at this stage shows, of course, two of the endodermal 
pouches separated by the hollow entocodon. As pointed out 
by Goette, it is from the examination of such a section, 
independently of others of the series, that the idea arose that 
a double-walled cup of endoderm was formed by the growth 
of the entocodon. A tangential section taken a short way on 
either side of this median section will naturally show a single 
pouch only, as illustrated in fig. 9. In this stage the forecast 
of the manubrium is already conspicuous, and is, of course, 
clothed externally by the proximal wall of the entocodon. 

The four endodermic pouches continue their growth out- 
wards to the very tip of the bud, and at their terminations 
push out the ectoderm, causing the formation of four perradial 
buibous projections, which are the forecasts of the ocellar 
bulbs. A section, therefore, taken through a perradius gives 
rise to the false idea of an invagination of ectoderm towards 
the entocodon (Pl. 48, fig. 10). The four bulbs are very 
conspicuous features of the external anatomy of the medusa, 
even at this relatively early stage of development (Text- 
fig. 4). 

The formation of the endoderm lamella is exactly as 
described by Goette for Podocoryne carnea; the central 
part of each endodermal pouch becomes a radial canal, the 
large cells at the edges growing out to form two solid wings 
of endoderm, which meet similar projections from the other 


oa. 


STINGING-CELLS IN CRASPEDOTE MEDUSZ&. ray 


pouches at the interradiu (Pl. 43, fig. 7, r.p.e.!). The ring- 
canal is formed by the fusion of the distal ends of the radial 
pouches at the bases of the bulbous swellings referred to 
above. 

Up to this point the histology of the two layers has been 
quite constant; the superficial ectoderm has retained its 
original character and remains crowded with interstitial cells 
of all kinds, in striking contrast with the small-celled regularly 
arranged tissues derived from the entocodon. The endoderm 
lining both the ccelenteron and the radial pouches consists 
of large clear cells, with somewhat indefinite outlines and 
containing numerous large nutritive spheres, which stain 
deeply with iron-hematoxylin; a few irregularly shaped 
interstitial cells are to be found, most numerous between the 
endoderm cells hning the manubrium. 

In the last stage of the development described above we 
found all the organs of the adult medusa already well defined, 
with the exception of the tentacles. From this point onwards 
the more important changes are to be found in the structure 
of the umbrella, which now grows rapidly, especially in the 
region between the ocellar bulbs and the base of the manu- 
brium, so that the superficial ectoderm loses its characteristic 
features, as noticed above, and gives rise to a low, small- 
celled epithelium covering the external surface of the bell. 
The endoderm behaves in a somewhat similar fashion. The 
ocellar bulbs, however, remain unaltered; the endoderm still 
consists of large irregular cells with nutritive spheres; the 
ectoderm is still crowded with interstitial cells, thread-cells, 
and nematoblasts, the latter increasing rapidly and forming 
new nematocysts, both large and small (Text-fig. 5). 

The ocellar bulbs give rise to the tentacles, their main 
function being obviously that, already mentioned on p. 768, 
of supplying these organs with stinging-cells. 

In the preceding paragraphs I have tried to emphasise the 
fact that beyond an increase in the actual number of cells, 
the ectoderm of this region has remained practically un- 
changed throughout the development of the gonophore. The 


778 CHARLES L. BOULENGER. 


tentacular nematocysts of the fully-formed medusa thus arise 
in the interstitial cells derived from the ectoderm of the 
parent hydroid, 

The ocellar bulbs are, of course, retained throughout the 
life of the medusa, and, as mentioned above, keep on supply- 
ing the tentacles with stinging-cells; they are no doubt 
especially active during the regeneration of these organs. 
This explains the constant presence of such swellings at the 


THXT-FIG. 0. 


A longitudinal section of an ocellar bulb of Mceerisia lyonsi 
just before the development of a tentacle. vel. Velum. nem. 
Small nematocyst. nem.’ Large nematocyst. 


bases of the tentacles of the Hydromeduse, as well as their 
early appearance and relatively enormous size in the develop- 
ing medusa-buds. The function of bearing the ocellar sense- 
organs must bea secondary one, for such bulbs are conspicuous 
in the formation of medusze which do not possess ocelli, e. g. 
Podocoryne carnea, concerning which Goette (11, p. 19) 
remarks: 

“ Bald nach der Fertigstellung des Velum verdickt sich 
das Ectoderm jedes Randwulstes dicht iiber dem Velum zu 


| 
= 


STINGING-CELLS IN CRASPEDOTE MEDUSA. 779 


einem vorspringenden Polster, das den Ocellarbiidungen 
andrer Medusen entspricht, aber, wie schon die Alteren 
Beobachter (Allman, 16; Grobben 17) feststellten, keine 
Ocellen entwickelt.” 

Ina young medusa of Mcerisia a short time before its 
liberation the manubrium is still without a mouth opening, 
and is clothed externally by a single layer of low ectodermal 
cells (Pl. 43, fig. 11), the internal lining consisting of large 
clear endoderm cells containing nutritive spheres and occa- 
sional irregularly shaped interstitial cells. The latter become 
more numerous as development proceeds, and some can be 
clearly recognised by their enlarged nuclei to be sex-cells. 
These at a later stage, no doubt, become transferred to the 
ectoderm of the stomach region, and by their further division 
form the gonad. Owing to the absence of individuals of the 
right age, [am unable to state how the transference of sex- 
cells from one layer to another takes place. I have never 
met with them migrating through the structureless lamella, 
and it is quite possible that the transference is a passive one, 
similar to that described by Goette in the male gonophores of 
Hydractinia (11, p. 70). In the youngest free-swimming 
medusz examined by me the endoderm of the slightly 
swollen stomach had lost its small cells, and was separated by 
a very thin lamella from the ectoderm, which contained a few 
rows of developing sex-cells. 

The endoderm slightly distal to this region had retained 
a number of interstitial cells, some of which prove to be 
obvious nematoblasts and contained developing nematocysts. 
These are, of course, the rudiments of the fully formed 
stinging-cells, which, as described in the first part of this 
paper, are to be found in the endoderm, just below the 
stomach of the adult medusa, and which later migrate to the 
battery at the oral extremity of the manubrium. 

From this we must infer that the nematoblasts of the 
manubrium arise in the endoderm of the developing gonophore 
in exactly the same way as do the sex-cells; like the latter 
they are able to migrate through the tissues of the medusa. 


780 CHARLES L. BOULENGER. 


When we remember the similar origin of the two kinds of 
cells from undifferentiated interstitial cells, we need not be 
surprised that they both possess the same powers of active 
movement. 

That the thread-cells are identical in origin with the sex- 
cells is further emphasised by the fact that in exceptional 
cases part of the testis of Moerisia can give rise to a nemato- 
cyst battery instead of producing sperm-cells, as shown in 
Pl. 43, fig. 12. 

Both kinds of cells are first to be recognised in the endo- 
derm of the medusa-bud; this does not necessarily imply 
that they originate in that layer ; in my account of the early 
development of the gonophore, I showed that interstitial 
cells of the ectoderm occasionally migrate through the 
structureless lamella of the hydranth and become incorpor- 
ated among the proliferating cells of the endoderm. It is 
probable that these cells or their derivatives give rise to the 
sex-cells and nematoblasts. 

In my description of the anatomy of Mcerisia lyonsi (8), 
I mentioned that exactly the same types of nematocysts 
were to be found in the medusa as in the hydroid; in this 
paper I hope to have proved that they are not only identical 
in structure, but actually originate from the same cells. 
This fact is one which might be of use in systematic work on 
the Hydromeduse, where the assignment of Medusz to 
hydroids is often only a matter of inference ; a careful com- 
parison of the nematocysts of the two stages should be of 
great value in this connection. 


5. GENERAL CONCLUSIONS. 


(1) The stinging-cells of the medusa of Mcerisia lyonsi 
are not developed “in situ” on the principal batteries, but 
migrate to their final positions on the oral lip, or on the 
tentacles. 

(2) The stinging-cells of the oral battery are formed in the 
endoderm of the manubrium, just below the stomach ; those 


ts 


STINGING-CELLS IN CRASPEDOTE MEDUSA. 781 


of the tentacles in the ectoderm of the conspicuous ocellar 
bulbs at the terminations of the radial canals. 

(3) There is reason to believe that the bulbous swellings 
at the bases of the tentacles have this function throughout the 
craspedote Medusz. In the sub-divisions Trachomeduse 
and Narcomedusex, they are probably replaced by the 
thickened ring of thread-cells on the margin of the bell. 

(4) The development of the gonophores of Mcerisia takes 
place in the manner described by Goette for other hydroids, 
There is no double-walled cup of endoderm at any stage, the 
radial canals and the endoderm lamella being derived from 
four separate pouches of endoderm, which grow out simul- 
taneously from the ccelenteron of the simple bud. 

(5) The stinging-cells of this medusa are developed from 
cells, which, like the sex-cells, arise directly or indirectly from 
the ectoderm of the parent hydranth. 


BIRMINGHAM, 
June 19th, 1910. 


BIBLIOGRAPHY. 


This bibliography includes only those works actually mentioned in 
the text of my paper; for a more complete list of literature I must 
refer the reader to the memoirs of Hadzi (4) and Goette (11). 


1. Murbach, L.—* Beitrage zur Kenntnis der Anatomie und Entwicke- 
lung der Nesselorgane der Hydroiden,” ‘ Arch. f. Naturg.,’ Jahrg. 
60, Bd. i. 


2. Schneider, K. C.—‘* Mittheilungen itber Siphonophoren: V, Nessel- 
zellen,” ‘ Arb. Zool. Inst. Wien,’ Tom. xii, 1900. 

“Histologie von Hydra fusca,” ‘Arch. f. Micr. Anat.,’ 
Bd. xxxv, 1890. 

4. Hadzi, J—‘Ueber die Nesselzellwanderung bei den Hydroid- 
polypen,” ‘ Arb. Zool. Inst. Wien,’ Tom. xvii, 1907. 

5. Boulenger, C. L.—“ On Merisia lyonsi, a new Hydromedusan 
from Lake Qurun,” ‘ Quart. Journ. Mier. Sci.,’ vol. 52, 1908, 

6. Hertwig, O., and R.—‘ Das Nervensystem und die Sinnesorgane 
der Medusen,’ Leipzig, 1878. 

7. Ginther, R. T.—‘A further Contribution to the Anatomy of 
Limnocnida tanganyice,” ‘Quart. Journ. Micr. Sci.,’ vol. 
36, 1594. 


782 CHARLES L. BOULENGER. 


8. Agassiz, L.—‘ Contributions to the Natural History of the United 
States of America,’ vol. iv, Boston, 1862. 

9. Hertwig, O., and R.—‘ Der Organismus der Medusen und seine 
Stellung zur Keimblattertheorie,’ Jena, 1878. 

10. Weismann, A.—‘ Die Enstehung der Sexualzellen bei den Hydro- 
medusen,’ 1883. 

11. Goette, A.— Vergleichende Entwicklungsgeschichte der Gesch- 
lechtsindividuen der Hydropolypen,” * Zeit. f. wiss. Zool., Bd. 
Ixxxvii, 1907. 

“Ueber die Entwicklung der Hydromedusen,” ‘ Zool. Anz.,’ 
Bd, xxvii, 1904. 

13. Richter, W.—‘ Die Entwicklung der Gonophoren einiger Sipho- 
phoren,” ‘ Zeit. f. wiss. Zool.,’ Bd. lxxxvi, 1907. 

14. Hadzi, J—‘ Arb. Zool. Inst. Wien,’ Tom. xviii, 1909. 

15. Hardy, W. B.—* The Histology and Development of Myriothela 
phrygia,” ‘Quart. Journ. Mier. Sci.,’ vol. 32, 1891. 

16. Allman, G. J.—‘ Monograph of the Gymnoblastic or Tubularian 
Hydroids,” ‘ Ray Soc.,’ 1871-2. 

17. Grobben, C.—‘ Ueber Podocoryne carnea,” ‘Arb. Zool. Inst. 
Wien,’ Tom. ii, 1875. 

18. Fowler, G. H.—‘ Hydromedusee” in Lankester’s ‘Treatise on 
Zoology,’ vol. ii, 1900. 


EXPLANATION OF PLATES 42 anp 43, 


Illustrating Mr. C. L. Boulenger’s memoir ‘‘On the Origin 


and Migration of the Stinging-cells in Craspedote 
Meduse.” 


PLATE 42. 


EXPLANATION OF LETTERING. 


ect. Ectoderm of the manubrium. end. Endoderm of the manubrium. 
gl.c. Gland-cell. nem. Endodermal nematocyst. nem.' Nematocyst of 
the oral battery. nem.2 and nem.s Migrating nematocysts.  s.l. 
Structureless lamella. test. Testis. 


Fig. 1—A longitudinal section of the proximal part of the manubrium 
of the medusa, Merisia lyonsi, to show the nematocysts in the 
endoderm. 


STINGING-CELLS IN CRASPEDOTE MEDUSA. 783 


Fig. 2.—A transverse section through the same region. 

Fig. 3.—A longitudinal section of the distal part of the manubrium 
showing the oral battery and a stinging-cell (nem.’) migrating through 
the endoderm towards it. 

Fig. 4.—A similar section showing a stinging-cell (nem.*) making its 
way through the structureless lamella to the ectoderm. 


PLATE 48. 


EXPLANATION OF LETTERING. 


c.b, Cavity of the medusa-bud. c.e. Cavity of the entocodon, i.e. 
subumbrella cavity. c.m. Cavity of manubrium. ect. Superficial ecto- 
derm of the developing gonophore. end. Endoderm of the same. ent. 
Ectoderm of the entocodon. g.c. Gland-cell. 7.c.e. Interstitial cell of 
the endoderm. xem. Small nematocyst. nem.' Large nematocyst. 
n.s. Nutritive sphere of the endoderm. 7.p.e. Radial pouch of endo- 
derm. r.p.e.! Lateral solid entension of the same, which later forms the 
endoderm lamella. s.c. Sex-cell. tent.b. Tentacle-bulb. fest. Testis. 
v. Velum. 


Fig. 5.—Longitudinal section of an early stage in the formation of the 
gonophore of Merisia lyonsi (cf. text, p. 775). 

Fig. 6.—Tangential longitudinai section of a young bud showing the 
entocodon and a single radial endoderm pouch. 

Fig. 7.—Transverse section of an older gonophore to illustrate the 
complete independence of the four radial pouches. The entocodon 
already has a large cavity (subumbrella cavity), and at 7.p.e.' can be seen 
the solid extension of the edge of a pouch which later forms the endo- 
derm lamella. 

Fig. 8.—Radial longitudinal section through a similar (slightly 
younger) bud, showing two radial pouches of endoderm separated by the 
entocodon. 


Fig. 9.—Tangential longitudinal section of the same medusa-bud ; 
only asingle pouch is shown. 

Fig. 10.—Longitudinal section of an almost completely developed 
medusa to show the bulbous swellings at the termination of the radial 
pouches. 

Fig. 11.—Manubrium of the same medusa under a higher magnifi- 
cation. 

Fig. 12.—Section through the testis of an adult medusa, part of which 
has given rise to a stinging-cell battery. 


IJuant.Lourn. Mier Se. tb, 55 NS Cee. 


59 Sie 


alo) 2 


AANA, 


eRe 


Hoth Lith? London. 


C.L.Boulenger, del. 


a 
« 


a ee oe 


Se a ee =a ae ek aw 


> 
er 


nn vat a i ee 8 a i Si 


C.L.Boulenger, del 


fi 2 
rey, 


ae 


&, SEOCOos 


(m) 
ren 


Oe 
es 
en 
ay 
als 


D 
ce) 


aD A 
O09 


is 
» 
ES 


Quart. Sourn Micr:Ser.VA.b5 NSM. 


Fhith Lith? London 


Ay. 


MUTATIONS IN CRUSTACEA OF THE FAMILY ATYIDA. 785 


The Researches of Bouvier and Bordage on 
Mutations in Crustacea of the Family Atyide. 


By 
Ww. TT. Calman, D.Sc., 
of the British Museum (Natural History). 


With 4 Text-figures. 


Some six years ago Professor H. L. Bouvier (?04, 705)! called 
attention to the remarkable dimorphism of certain tropical 
river-prawns of the family Atyide, which he compared with 
the phenomenon of mutation described by de Vries in the 
vegetable kingdom. He pointed out that the case was espe- 
cially noteworthy, not only because of the marked discon- 
tinuity and constant occurrence of the variations, but also 
because they affected characters regarded as distinctive of 
genera; and he drew the conclusion that these genera had 
originated by a process of mutation. M. EH. Bordage has 
recently published (’08, 09a, 7098) the results of some obser- 
vations and experiments on the living animals which seem to 
support Bouvier’s views, and to indicate, at all events, a 
promising field for further investigations. At the suggestion 
of Sir Ray Lankester the following account has been pre- 
pared in the hope that it may induce some naturalists, who 
have the opportunity of studying the animals under natural 
conditions, to give attention to the matter. 

The Atyidz (see Text-fig. 1) are a family of Decapod Crus- 
tacea belonging to the tribe Caridea (whichincludes most of our 
common prawns and shrimps), and are widely distributed in 
fresh waters in the warmer regions of the globe (see Ortmann 

1 The numbers refer to the list of papers on p. 796. 


786 W. T. CALMAN. 


94 and Bouvier 705). Some.of the members of the family 
show very primitive characters, having, for instance, swim- 
ming branches or exopodites on all the thoracic limbs, as im 
the so-called “Schizopods.” In this and in other features 
they resemble the deep-sea Hoplophoride, from which, or 
from some allied forms, most authorities are agreed in con- 
sidering them to have been derived. 

Other members of the family, however, are considerably 
specialised. In some characters this specialisation has pro- 
ceeded along lines parallel to those followed in other series of 
the Caridea—for example, in the progressive disappearance 


Trxt-FiG. 1. 


Atya bisuleata. Ovigerous female of the Atya-form. x3. From 
a specimen in the ‘* Challenger” collection from Honolulu. 


of the exopodites and, later, of the epipodites of the legs, and 
a diminution in the number of the branchiz. In other 
characters specialisation has followed lines peculiar to the 
family, and thisis especially the case with the modifications 
of the chelate first and second pairs of legs. In nearly all 
Atyidee these limbs are comparatively small, not dissimilar in 
size, and have the fingers each tipped with a brush of long 
hairs (Text-fig. 1). Fritz Miiller (’92) has described how these 
brushes are used in collecting pellets of mud on which the 
animal feeds.! Among the more specialised members of the 

1 T do not understand Bordage’s statement that the chele are used 


for excavating burrows in the mud, for which their structure would 
appear to be ill-adapted. 


MUTATIONS IN CRUSTACEA OF THE FAMILY ATYIDE. 787 


family the characters used as distinctive of the genera are 
chiefly drawn from the modifications of the chelipeds, and 
some of these may now be considered in fuller detail. 

In the very numerous species of the genus Caridina (Text- 
fig. 2) the chele themselves do not differ greatly, except in 
carrying brushes of set, from the typical form found in many 
other Decapods. The dactylus (d.) or terminal segment of 
the limb, forming the “ movable finger,” is opposed to a 
thumb-like process (‘ immovable finger ’’) of the penultimate 


TEXT-FIG. 2. 


Caridina nilotica var. 1, 2, first and second chelipeds. 
¢., carpus; d., dactylus ; p.. palmar portion of propodus. x 40. 
- From a specimen collected by Dr. W. A. Cunnington in the 
Victoria Nyanza. 
segment or propodus. ‘he proximal part of the propodus, 
expanded to contain the muscles moving the dactylus, forms 
what is known as the “ palm” (p.) of the chela. In Caridina 
the two pairs of chelipeds differ in the form of the segment 
which supports the propodus, the “ wrist” or carpus (c.). In 
the second pair it is more or less elongated and slender, and 
the propodus articulates with its distal end; in the first pair, 
on the other hand, it is short and broad, its distal margin is 
more or less concave (cf. Text-fig. 2,1, and Text-fig..4, 4’), 
and the propodus articulates with its lower corner. 
VOL. 55, PART 4.—NEW SERIES. 53 


788 W. T. CALMAN. 


The species of the genus Ortmannia (formerly known as 
Atyoida) differ from those of Caridina chiefly in the fact 
that the carpus of the second pair resembles that of the first 
pair (Text-fig. 3, B’, B”’), being short and broad, with its distal 
margin excavated and articulating with the propodus at its 
lower corner. It is to be noted that these characters are not 
equally well marked in all the species referred to Ortmannia; 
in some the second carpus is still, as in Caridina, somewhat 
longer than the first, and the excavation of its distal margin 
is shallow (asin Text-fig. 4, B”) ; in other species the carpus is 
nearly similar in the two pairs and so deeply excavated as to 
assume an almost crescentic form (as in Text-fig. 3, B’, BY’). 
Associated with this excavation of the carpus is a shifting 
(already begun in Caridina) of the carpo-propodal articulation 
from the proximal end to the lower border of the propodus. 
Further, while in some species the chele themselves are quite 
is much 


“ce ? 


similar to those of Caridina, in others the “ palm’ 
shortened, or, in other words, the articulation of the movable 
finger is carried backwards towards the base of the propodus. 

These modifications lead towards the conditions found in 
the genus Atya, which includes the largest and most highly 
specialised members of the family. In these the two pairs of 
chelipeds (‘l'ext-fig. 3, A’, A”) are quite similar, and the carpus 
is reduced by the excavation of its distal border to a narrow 
crescent, with the lower limb of which the propodus articulates. 
The propodus itself assumes a form unlike that of any other 
Decapod; the backward shifting of the articulation of the 
dactylus has been carried so far that the palm has entirely 
disappeared, and the chela is composed of two similar parts, 
hinged together at one end, like the legs of a pair of 
compasses. 

Although, within each of the genera, there is some varia- 
tion in the degree to which these characters are developed, 
this variation is so far discontinuous that all the known 
species could, prior to Bouvier’s researches, be referred 
without much difficulty to one or other of the genera. If it 
be objected that such apparently trivial differences should 


MUTATIONS IN CRUSTACEA OF THE FAMILY ATYIDM. 789 


not be regarded as of generic value, it may be pointed out 
that, as a rule, though not in every case, they are coincident 
with other features which help to characterise, although 
they do not define, the generic groups; and further, there is 
no criterion by which the generic value of a character may 
be estimated, except that of its constancy throughout a 
croup of species. 

Bouvier’s discovery may be shortly expressed by saying 
that certain species were found to be dimorphic and to 
oscillate, as it were, in a state of unstable equilibrium between 
one generic group and the next. Thus, Miss Rathbun (’01) 


THXT-FIG, 3. 


Atya bisuleata. <A’, A", First and second chelipeds of the 
Atya-form. 8B’, B", First and second chelipeds of the 
Ortmannia-form (Ortmannia Henshawi). x 7. From 
specimens in the ‘‘ Challenger ” collection from Honolulu. 


had described a new species, Ortmannia Henshawi (Text- 
fic. 3, B’, B’), found in association with Atya bisulcata 
(d’, A’), on the island of Hawaii; Bouvier pointed out that 
this association was not accidental, but constant, that the two 
forms were indistinguishable, except by the characters of the 
chelipeds, and that they should be regarded as constituting 
a single dimorphic species. He found a similar phenomenon 
in the case of Atya serrata, described by Spence Bate 
from specimens obtained by the ‘‘ Challenger”? Expedition at 
the Cape Verde Islands, and since found in many localities 
on the islands of the Indian and Pacific Oceans. To the 
Ortmannia-form of this species Bouvier gave the name 


538 


790 W. T; CALMAN,. 


O. Alluaudi. In both species the two forms were sharply 
distinguished, although in the Ortmannia individuals 
(especially in O. Alluaudi) a considerable amount of varia- 
tion was observed in the relative proportions of the fingers 
and palm of the chele; the Atya-form, on the other hand, 
presented no noteworthy variation. In both species Bouvier 
found that the dimorphism was independent of age and sex ; 
both forms were found through a wide range of size, although 
the Atya individuals were, on the whole, somewhat larger, 
and females of both were observed carrying eggs. In the 
case of A. bisulcata (O. Henshawi) both forms occurred 
in about equal numbers; in A. serrata (O. Alluandi) 
there was some evidence that the relative proportions varied 
in different localities.' 

In one species of Caridina Bouvier found evidence of the 
existence of an analogous mutation leading to the genus 
Ortmannia. Among eleven examples of C. apiocheles 
(Text-fig. 4) (probably from the Seychelles), he observed 
one in which the carpus of the second pair of chelipeds (Text- 
fig. 4, B”’), instead of being long and slender as in the typical 
individuals, was short, broad, and excavated distally, re- 
sembling that of the first pair, so that the specimen, had it 
occurred alone, would have been referred to Ortmannia. 
In this case, however, it remains to be seen whether the 

1 It may be of interest to give here the results of a preliminary 
examination of the material of these two species in the British 
Museum collection. In one lot of Atya bisulcata obtained by the 
‘Challenger’ Expedition at Honolulu forty-two specimens are of the 
Atya-type and forty-six of the Ortmannia-type. Only one speci- 
men cannot be referred to either, having three chele of the Atya-shape, 
while the fourth is distinctly of the Ortmannia-shape. There is a 
considerable amount of variation in the chele of the Ortmannia- 
individuals, and their terminal brushes of sete are always much 
shorter than in the Atya-individuals. In a second lot of specimens 
from Hawaii only nine Ortmannia-individuals are found among 
thirty-eight Atya-individuals. Of the two type-specimens of Atya 
serrata from the Cape Verde Islands in the ‘Challenger’ collee- 
tion, the larger is of the Atya-type while the smaller is a distinet 
Ortmannia, 


MUTATIONS IN CRUSTACEA OF THE FAMILY ATYIDH. 791 


occurrence of the mutation is a normal and constant feature of 
the species.! 

Professor Bouvier discusses at length the possible explana- 
tions of these curious phenomena. He points out that it is 
impossible to continue to regard Atya bisulcata and Ort- 
mannia Heushawi, for instance, as distinct and independent 
species; their constant association and their identity in all 
characters except those of the chelipeds forbid their separa- 
tion, and it may be added that Professor Bouvier’s proved 
skill and experience as a carcinologist give special weight to 
his opinion on this point. He also dismisses, and no doubt 


TEXT-FIG. 4. 


Caridina apiocheles. A’, A”, First and second chelipeds of 
the typical Caridina-form. Bb’, B”’, First and second chelipeds 
of the Ortmannia-form (O. Edwardsi). After Bouvier. 


rightly, the suggestion that the phenomena are due to 
hybridisation ; and he concludes that the facts he describes 
have their closest analogy in the ‘ mutations” of de Vries. 


! The question whether the genera implicated in these phenomena of 
mutation are to be retained as valid is of secondary importance, and 
hardly concerns more than the convenience of the systematist. If they 
are to be retained, however, it would seem that a good case exists for 
the re-instatement of the name Atyoida in place of Ortmannia. 
Miss Rathbun displaced Atyoida on the ground that the surviving 
type-specimens of Randall’s Atyoida bisulcata, the type-species of 
Atyoida, have chelw of the Atya-type. If, however, O. Henshawi, 
the type-species of Ortmannia, is only a form of A. bisuleata, the 
two genera are synonymous and the older name should be used, 


792 WwW. T. CALMAN. 


Instead of being limited to comparatively trivial characters 
and giving rise to varieties or ‘‘ petites espéces ” as in de 
Vries’s examples, the mutations of the Atyidze affect characters 
of generic importance. Bouvier believes that the course of 
evolution from the more primitive Caridina to the specialised 
Atya has been discontinuous, proceeding at a single step 
from Caridina to Ortmannia and again from Ortmannia 
to Atya, and that the species mentioned remain in the con- 
dition of instability accompanying the transition from one to 
the other. It is also implied, although Bouvier does not dwell 
on the point, that these genera are polyphyletic and have 
originated independently in several regions of the globe. 

There is still another possibility, not alluded to by Bouvier, 
that deserves mention here, namely, that the apparent di- 
morphism is due to heteromorphic regeneration of the chelipeds 
after mutilation. Many cases are now known among Arthro- 
poda in which regenerated appendages depart from the normal 
type, and not infrequently revert to a simpler and more 
primitive form (“‘régénération hypotypique” of Giard). 
Although the chelipeds of many Atyide readily break off 
from the body in preserved specimens, it seems very impro- 
bable that this mutilation should happen so frequently in 
nature that 50 per cent. of the specimens collected would 
have regenerated limbs; nor is it less improbable that all four 
chelipeds would be removed simultaneously!; and the experi- 
ments of Bordage, described below, lend no support to this 
suggestion. 

Professor Bouvier pointed out the desirability of testing 
his conclusions by observation and experiment on the living 
animals, and it was at his suggestion that Bordage undertook 
the researches of which the results are presented in his recent 
papers (708, 7094, ’09B). On the island of Réunion Ort- 
mannia alluaudi, with its mutation Atya serrata, occurs” 
abundantly in mountain streams at altitudes above 300 
metres. Owing to the high temperature prevailing at the 


1 Only one case has been noticed in which one of the chelipeds 
differed from the others (see above, p. 790, footnote). 


MUTATIONS IN CRUSTACEA OF THE FAMILY ATYID%. 793 


coast (St. Denis), where the experiments were carried on, it 
was impossible to keep the animals alive in small aquaria, but 
after several failures Bordage succeeded in keeping living 
specimens in a small tank of masonry through which a current 
of water from the town supply was kept flowing. The inflow 
and outflow were guarded by fine wire gauze covered with 
muslin to prevent the escape of adults or larve, or the acci- 
dental introduction of additional specimens. A_ single 
ovigerous female of the Ortmannia form was placed in the 
tank, and in a few days numerous zoea larvee were observed 
in the water. Only seven individuals survived to assume the 
perfect form a fortnight later, and these proved to be all, like 
the parent, of the Ortmannia-type. A second experiment, 
however, was more successful. Another ovigerous Ort- 
mannia was placed in the tank (which had been emptied and 
cleaned out between the experiments) and the larve were 
hatched in due course. When they were about to pass into 
the final stage of their metamorphosis some weeks of torrential 
rain rendered the water-supply muddy and opaque, so that the 
young prawns were lost sight of. On cleaning out the tank, 
however, sixteen specimens were discovered among the mud, 
and of these ten were like the parent, while six were of the 
Atya-type. Bordage assures us that the precautions he 
took absolutely exclude the possibility of these young 
prawns having come from any source other than the eges 
carried by the original female. In another experiment two 
females of the Atya-type produced twenty-seven young, 
all of which resembled the parents. Bordage states that he 
was unable to obtain fecundation of Ortmannia females 
by Atya males, while they bred readily with males of their 
own type. 

These results are somewhat surprising, and can hardly be 
accepted as final without a good deal more experimental 
evidence. If the two forms do not interbreed, and if, as 
Bordage considers probable, the Atya-form always breeds 
true, it is evident that the Ortmannia-form would disappear 
(in the absence of a selective death-rate operating in its 


794 W. T. CALMAN. 


favour) even more speedily than is required by the “ loi de 
Delboeuf” to which Bouvier refers. 

Bordage also made some experiments on the regeneration 
of the chelipeds. He found that after amputation of the 
chelipeds of an Atya, the regenerating limbs had at first the 
Ortmannia-form—that is to say, the propodus showed a 
distinct palmar portion. At the first moult after the opera- 
tion, however, the Atya-form was assumed, the articulation 
of the dactylus having shifted to the proximal end of the 
propodus. It is not clear from the account given whether 
the chela were perfectly formed and movable before the first 
moult. Bordage regards this as a typical case of atavistic 
regeneration (régénération hypotypique), and he also cites a 
case described by Fritz Miller (92) as showing that the 
regenerated second pair of chelipeds in Ortmannia poti- 
mirim have an elongated and slender carpus like that of 
Caridina.! 

While Bordage’s results are highly interesting and sug- 
gestive, they rest upon a very narrow basis of experimental 
evidence. ‘There seems to be no reason to doubt his statement 
that young of the Atya-type were hatched from the eggs of 
an Ortmannia female, but it is based on the result of a 
single experiment carried out under unfavourable conditions, 
and no figures of the young prawns are given. The supposed 
inability of the Atya females to produce Ortmannia young 
rests also on the negative result of a single experiment and 
the simple statement that the two forms do not interbreed 
deserves to be examined in greater detail. It would be of 
interest to have further particulars as to the normal course 
of development in the two forms, and to know whether there 
is any trace of an Ortmannia stage in the development of the 
Atya-form of cheliped. The phenomena of regeneration also 
require more thorough investigation ; it is possible that, as is 


1 It may be mentioned that a comparison of Miller's original figure 
with the copy given in Bordage’s paper does not increase our confidence 
in the diagrammatic drawings which the latter author gives to illustrate 
his own observations. 


MUTATIONS IN CRUSTACEA OF THE FAMILY ATYIDM. 795 


known to be the case in other Decapods, the form of the 
regenerated limbs may differ according to the age of the 
individuals experimented on. While it is very improbable, for 
the reasons stated above, that the whole appearance of 
dimorphism can be due to regeneration, it remains to be 
tested whether the form of the chelipeds does really remain 
constant throughout the life of the individual. Apart from 
the possibilty of further experiments with the living animals, 
it would be of importance to get together sufficient material 
for a biometrical investigation into the degree of discontinuity 
in the variation and its incidence in relation to age, sex, and 
locality. 

One of the most interesting features of these mutations, if 
Bouvier’s interpretation of them be confirmed, is the direct 
way in which they bear on the problems suggested by a 
study of the Atyidez from the systematic standpoint. ‘his 
may be illustrated by an example. In Lake Tanganyika 
(Calman, 799 and ’06) the collections of Mr. J. E. 8S. Moore, 
and, more especially, of Dr. W. A. Cunnington, have revealed 
the existence of numerous peculiar species of Atyidex, which 
differ from all the other members of the family (with one 
exception to be mentioned presently) in having a reduced 
branchial formula. Thus the Tanganyikan Caridella re- 
sembles Caridina in most of its characters, except that it 
has no pleurobranchia on the last somite of the thorax, and 
Atyella differs in the same character from Ortmannia. 
I have pointed out elsewhere that while the reduction in 
the number of branchiw may have occurred independently 
in each of the Tanganyikan genera, so that Caridella 
may be supposed to be derived from Caridina, and 
Atyella from Ortmannia, Bouvier’s results suggest as 
a possible alternative that Atyella may have originated 
from Caridella by a mutation parallel to that by which, 
in other parts of the world, Caridina has given rise to 
Ortmannia. ‘The latter hypothesis has recently received 
the support of Prof. Bouvier himself (09a, ’098), in conneec- 
tion with his very interesting discovery that Atya Poeyi 


796 W. T. CALMAN. 


of the West Indian Islands has the same branchial formula 
as Caridella and Atyella, and in fact only differs from 
the last-named genus in having chelw of a distinctly Atya 
type. He refers the West Indian species to a new genus, to 
which he gives the name Calmania. He supposes it to 
have been derived from Atyella in the same way as Atya 
from Ortmannia, and he concludes that Atyella (and 
Caridella also) must formerly have existed in America. 
From this view I would venture to dissent. Even if the 
phenomena of mutation lead us to believe that similar forms 
of chelipeds may have been acquired independently in 
different localities, there is no greater difficulty in supposing 
that a simple suppression of the posterior pleurobranch may 
also have occurred more than once in the evolution of the 
family. In all the groups of animals composing the remark- 
able fauna of Tanganyika, there is reason to believe that 
many of the endemic genera and species have been differen- 
tiated within the limits of the lake itself; and until the 
Atyide with a reduced branchial formula are shown to 
have a much wider geographical distribution than is at 
present known, it seems impossible to believe in a direct 
affinity between the Tanganyikan Atyella and the West 
Indian Calmania. 

It may be freely admitted that these phylogenetic specula- 
tions rest upon much less solid ground than do the conclusions 
drawn directly from experiment or based upon statistics ; 
but unless we are to abandon all hope of rationalising the 
facts of systematic and geographical biology, some such 
hypotheses are, for the present, indispensable. 


List OF PAPERS REFERRED TO. 


Bordage, E., 08.—‘ Recherches expérimentales sur les mutations 
évolutives de certains Crustacés de la famille des Atyidz,” ‘C.R. 
Acad. Sci. Paris,’ exlvii, pp. 1418-1420, fig. 

09a.—* Sur la régénération hypotypique des chélipédes chez 
Atya serrata Sp. Bate,” ‘C.R. Acad. Sci. Paris,’ exlviii, pp. 
47-50, 


MUTATIONS IN CRUSTACEA OF THE FAMILY ATYIDA, 797 


Bordage, E., ’098.—‘ Mutation et régénération hypotypique chez 
certains Atyidés,” ‘ Bull. sci. France Belgique,” xliii, pp. 95-112, 
7 figs. 

Bouvier, E. L., ‘04.—* Sur le genre Ort mannia Rathb. et les mutations 
de certains Atyidés,” *C.R. Acad. Sci. Paris,’ exxxvili, pp. 446— 
449 ; translated in ‘ Ann. Mag. Nat. Hist.’ (7), xiii, pp. 377-381. 

—— '05.—* Observations nouvelles sur les Crevettes de la famille des 

Atyidés,” ‘ Bull. sci. France Belgique, xxxix, pp. 57-154, 26 figs. 

*09a.—* Sur Vorigine et l’évolution des Crevettes de la famille 

des Atyidés,” ‘C.R. Acad. Sci. Paris,’ exlviii, pp. 1727-1731. 

‘09B.—** Les Crevettes d’eau douce de la famille des Atyidés qui 
se trouvent dans Vile de Cuba,” ‘ Bull. Mus. d’Hist. nat. Paris,’ 

1909, pp. 327-336. 

Calman, W. T., °99.—* On Two Species of Macrurous Crustaceans 
from Lake Tanganyika,” * Proc. Zool. Soc. London,’ 1899, pp. 
704-712, 2 pls. 

—— °06.—* Zoological Results of the Third Tanganyika Expedition 
conducted by Dr. W. A. Cunnington, 1904-1905: Report on 
the Macrurous Crustacea,” ‘ Proc. Zool. Soc. London,’ 1906, pp. 
187-206, 4 pls. 

Miller, Fritz, °92—*O Camario miudo do Itajahy, Atyoida poti- 
mirim,” ‘ Arch. Mus. Nacion. Rio de Janeiro,’ vili, pp. 155-178, 
2 pls. 

Ortmann, A. E., 94.— A Study of the Systematic and Geographical 
Distribution of the Decapod Family Atyide, Kingsley,” ‘ Proc. 
Acad. Nat. Sci. Philadelphia,’ 1894, pp. 397-416. 

Rathbun, Mary J.,’01.—*“ The Brachyura and Macrura of Porto Rico,” 
‘Bull. U.S. Fish. Comm.,’ 1900 (2), pp. 1-127 and 129*-137*, 2 
pls. 


INDEX 


LO 


VOL. 55, 


NEW SERIES. 


Allen on the artificial culture of | Craspedote meduse, nettle-cells of, 


marine Plankton organisms, 361 
Anaspides tasmanie,a Gregarine 
from, by Julian Huxley, 155 
Aplysia punctata, development 
of, by Saunders and Poole, 497 
Assheton and _ trophoblast, 
Hubrecht, 585 


by 


Atyide, mutations in, by Calman, | 


785 


Blepharoplast and centrosome, 611 

Blood-parasites of fishes, by 
Minchin and Woodcock, 113 

Boulenger on the origin and migra- 
tion of the  stinging-cells 
craspedote medusz, 763 

Bourne on the anatomy of Incisura, 
1 


Bouvier and Bordage, account of | 


their researches on mutations in 
Atyide, by Calman, 785 


Calman on the researches of Bouvier 
| Eye of Pecten, by Dakin, 49 


and Bordage on mutations in 
Crustacea of the family Atyide, 
785 

Centrosome and blepharoplast, 611 


Cercomonas, by Wenyon, 241 


of | 


by Boulenger, 763 

Crenilabrus, pigment formation in, 
54d. 

Crithidia melophagia, an endo- 
parasite of the sheep-ked, by 
Porter, 189 


| Crustacea of the family Atyide, 


mutations in, by Calman, 785 
Culture of marine Plankton, by 
Allen, 361 


Dakin on the eye of Pecten, 49 

De Morgan and Drew on fibrous 
tissue as a result of injury in 
Pecten, 595 

Development of Aplysia, by Saunders 
and Poole, 497 


| Drew and De Morgan on fibrous 


tissue as a result of injury in 
Pecten, 595 


| Duke on a new Gregarine, Meta- 


mera schubergi, 261 


Fishes, blood- parasites of, by 
Minchin and Woodcock, 1138 
| Flagellate of the genus Cerco- 


Clathrina, division of collar-cells of, | 


by Robertson and Minchin, 611 
Collar-cells of Clathrina, 611 


VOL. 90, PART 4.——-NEW SERIES. 


monas, by Wenyon, 241 
Hoetal membranes of Vertebrates, by 
Hubrecht, 177 
o4. 


800 INDEX. 
Ganymedes anaspidis, a Gre- | Minchin on collar-cells of Clathrina, 


earine, by Huxley, 155 


Gregarine from Anaspides, by 
Huxley, 155 

Gregarine, on a new species to be 
called Metamera 


by Duke, 261 


schubergi, 


Hematozoa from Ceylon, by Robert- 
son, 741 

Hemogregarina Nicorimw, by 
Robertson, 741 

Hemoprotozoa of birds, by Wood- 
cock, 641 

Hippolyte, pigment formation in, 
541 

Histriobdella 
Shearer, 287 

Hubrecht on Assheton and tropho- 
blast, 585 

Hubrecht on the foetal 
of Vertebrates, 177 


Huxley on 


Homari, by 


membranes 
Ganymedes anas- 
pidis, by Huxley, 155 


Incisura, anatomy of, by Bourne, 1 
Injury producing fibrous tissue in 
Pecten, 595 


Ked of the sheep, a parasite of, 189 


Light, influence on pigment forma- 
tion in Crenilabrus and Hippolyte, 
541 

Marine Plankton, 
Allen, 361 

Martin on Trypanoplasma con- 
geri, 485 


eulture of, by 


Medusex craspedotex, stinging- 
cells of, by Boulenger, 763 
Melophagus ovinus, a parasite of 
189 
Metamera 
Gregarine, by Duke, 261 
Minchin 


schubergi, a new 
and Woodcock on blood- 
parasites of fishes occurring at 
Royvigno, 113 


ADLARD AND SON, IMPR., 


611 


Nematodes, the free-living, by Potts, 
433 


Pecten maximus, fibrous tissue 
in, as a result of injury, 595 

Pecten, the eye of, by Dakin, 49 

Pigment formation in Crenilabrus, 
and Hippolyte, by Prof. Gamble, 
541 

Plankton, culture of, by Allen, 361 

Poole and Saunders, development of 
Aplysia, 497 

Porter on Crithidia melophagia 
an endo-parasite of the sheep-ked, 
189 

Potts on the free-living Nematodes, 
133 


Robertson, Muriel, on collar-cells of 
Clathrina, 611 

Robertson on 
Ceylon, 741 

Rovigno, blood-parasites of fishes 
occurring at, 113 


Hematozoa from 


Saunders and Poole, development of 
Aplysia, 497 

Sex, experimental analysis of, by 
Geoffrey Smith, 225 

Shearer on Histriobdella 
Homari, 287 

Smith, Geoffrey, on the experimental 
analysis of sex, 225 


Trophoblast, by Hubrecht, 585 
Trypanoplasma congeri, by 
Martin, 485 


Wenyon on a flagellate of the genus 
Cercomonas, 241 

Woodcock and Minchin on blood- 
parasites of fishes occurring at 
Rovigno, 113 

Woodcock on Hemoprotozoa of birds, 
641 


LONDON AND DORKING. 


New Series, No. 220 (Vol. 55, Part 4). Price 10s. net. 


Subscription per volume (of 4 parts) 40s. net. 


NOVEMBER, 1910. 


THE 


| QUARTERLY JOURNAL 


MICROSCOPICAL SCIENCE. 


| Str RAY LANKESTER, K.C.B., M.A., D.Sc., LL.D., F.R.S., 


HONOKAKY FELLOW OF EXETER COLLEGE, OXFORD ; 
MEMBER OF TUE INSTITUTE OF FKANCE (assucifii BTRANGER DE L’ACADEMIE DES SCIENCEs) ; 
CORRESPONDENT OF THE IMPERIAL ACADEMY OF SCIKNCES OF ST. PETEKSBUKG, AND OF THE 
ACADKMY OF 8CIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIENCES 
OF TURIN; FORKIGN MEMBER OF THE ROYAL SOCIETY OF SCIENCES OF 
GOTTINGEN, AND OF THE ROYAL BOHEMIAN SOCIETY OF SCIENCES, AND 
OF THE ACADEMY OF THE LINCEL OF ROMER, AND OF THR AMERICAN 
ACADEMY OF ARTS AND SCIENCKS OF BONTON } ASSOCIATE OF THK 
ROYAL ACADEMY OF BELGIUM; HONORARY MEMEBR OF TUE 
NEW YORK ACADEMY OF SCIKNCES, AND OF THR 
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OF 
THE ROYAL PHYSICAL SOCIETY OF EDIN- 
BURGH, AND OF THE 
BIOLOGICAL SOCIKTY OF PAKIS, AND OF THE CALIFORNIA ACADEMY OF SCIENCES OF SAN FRANCISCO, AND 
OF THE ROYAL ZOOLOGICAL AND MALACOLOGICAL SOCIETY OF BELGIUM; 
CORRESPONDING MEMBER OF THE SENKENBERG ACADEMY OF FRANKFURT-A-M} 
FORKIGN ASSOCIATE OF THK NATIONAL ACADEMY OF SCIENCES, U-S., AND MEMBER OF THE 
AMERICAN PHILOSOPHICAL SOCIETY 3 
HONORARY FELLOW OF THE ROYAL SOCIETY OF EDINBURGH; 
LATE DIRECTOR OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSKUM: LATE PRESIDENT oF THK 

BRITISH ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE: LATE FULLKKIAN PKOFKSSOR OF 

PHYSIOLOGY IN THK ROYAL INSTITUTION OF GREAT BRITAIN; 

LATER LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFORD} 
EMERITUS PROFESSOR OF ZOOLOGY AND COMPARATIVE ANATOMY IN UNIVERSITY COLLEGE, UNIVERSITY OF LONDON. 


WITH THE CO-OPERATION OF 


ADAM SEDGWICK, M.A., F.R 


FELLOW OF TRINITY COLLEG 


Ctr) 
S, CAMBRIDGE, AND PROFESSOR OF ZUOLOGY IN THE IMPEKIAL COLLEGE OF 
SCIENCE AND TECHNOLOGY, LONDON; 


SYDNEY J. HICKSON, M.A., F.R 


BEYER PKOFKSSOK OF ZOOLOGY IN THE UNIVERSITY OF MANCHESTER, 


KE. A. MINCHIN, M.A., 


PROFESSOR OF PROTOZOOLOGY IN THE UNIVERSITY OF LONDON; 


AND 


GILBERT C. BOURNE, M.A., D.Soi, F.B.S., 


LINACRE PROFESSOR OF COMPARATIVE ANATOMY, AND FELLOW OF MERTON COLLEGK, OXFORD, 


WITH LITHOGRAPHIC PLATES AND TEXT-FIGURES 


TGOWEDO N 
J. & A. CHURCHILL, 7 GREAT MARLBOROUGH STREET. 
1910. 
ee ee a eee 


Adlard and Son, Impr.,] 


(London and Dorking. 


5 PB so 


CONTENTS OF No. 220.—New Series. 


MEMOIRS : 
PAGE 
The Division of the Collar-Cells of Clathrina coriacea (Montagu); 
a Contribution to the Theory of the Centrosome and Blepharoplast. 
By Muriet Roserrson, M.A., and E. A. Mincuin, M.A. (With 
Plates 25 and 26) . 4 5 : . 6H 
Studies on Avian Hiswapestornk: 1.—On Certain Parasites of the 
Chaffinch (Fringilla celebs) and the Redpoll (Linota 
rufescens). By H. M. Wooncock, D.Se.(Lond.), Assistant to the 
University Professor of Protozoology. (With Plates 27-31) ~, G4 
Studies on Ceylon Hematozoa. No. IIl.—Notes on the Life-Cycle of 
Hemogregarina nicorie, Cast. and Willey. By Murier 
Rozserrson, M.A. (With Plates 32-41 and 1 Text-figure) 741 
On the Origin and Migration of the Stinging-Cells in Craspedote 
medusa. By CuHartss L. Boutencer, M.A.Camb., Lecturer on 
Zoology in the University of Birmingham. (With Plates 42 and 
43 and 5 Text-figures) : . 763 
The Researches of Bouvier and cde: on Mstaties in Crastectee 
of the Family Atyide. By W. T. Catman, D.Se. (With 4 Text- 
figures) : , . ; : : 


=I 
leo) 
or 


1 AND OF THE IMPERIAL ACADEMY u™, SCI 
a 


New Series, No. 218 (Vol. 55, Part 2). 


Price 10s, net. 
Subscription per volume (of 4 parts) 40s. net. 


} JUNE, 1910. 
5 THE 


QUARTERLY JOURNAL 


OF 


MICROSCOPICAL SCIENCE. 


EDITED BY 


Sim; RAY LANKESTER, K.C.B., M.A., D.Sc., LL.D., F.R.S., 


HONOKARY FELLOW OF EXETER COLLKEGK, OXFORD}; CORRESPONDENT OF THE INSTITUTK OF FRANCE, 
CES OF ST. PETERSBURG, AND OF THE ACADEMY 


AND OF THE ROYAL ACADEMY OF SCIKNCES 
OF TURIN}; FOREIGN MEMBER OF THE ROYAL SOCIETY OF SCIENCES OF 


GOTTINGEN, AND OF THE «ROYAL BOHEMIAN SOCINTY OF SCIENCES, AND 
f OF THE ACADEMY OF THE LINCEIL OF ROMER, AND OF THR AMERICAN 
ACADEMY OF ARTS AND SCIENCES OF BOSTON 3 ASSOCIATE OF THE 
ROYAL ACADEMY OF BELGIUM: HONORARY MEMBER OF THR 
NEW YORK ACADEMY OF SCIENCES, AND OF THR 
CAMBKIDGE PHILOSOPHICAL SOCIETY, AND OF 
THE ROYAL PHYSICAL SOCIETY OF EDIN- 


BURGH, AND OF THE 
BIOLOGICAL SOCIETY OF PARIS, AND OF THE CALIFORNIA 


OF THE ROYAL ZOOLOGICAL AND MAL 
CORRESPONDING MEMBER OF THE SENK 


FOREIGN ASSOCIATE OF THK NATIONAL ACADKMY OF SCIENCES, 
AMER “AN PHILOSOPHICAL SOCIETY 5 
HONORARY FELLUW OF THE ROYAL SOCIETY OF EDINBURGH, 


OF THE NATUKAL HISTORY DEPARTMENTS OF THE BRITISH MUSKUM, LATE PRESIDENT OF THE 
BRITISH ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE; LATE FKFULLKRIAN PROFESSOR OF 


PHYSIOLOGY IN THE ROYAL INSTITUTION OF GREAT BRITAIN 3 
LATER LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFORD) 
EMERITUS PROFESSOR OF ZOOLOGY AND COMPARATIV ANATOMY IN UNIVERSITY COLLEGE, UNIVERSITY OF LONDON 


WITH THE CO-OPERATION OF 


ADAM: SEDG WICK, M.A. ERS, 


FELLOW OF TRINITY COLLEGE, CAMBRIDGE. AND PROFESSOR OF ZOOLOGY IN THE IMPERIAL COLLEGE OF 
SCIENCE AND TECHNOLOGY, 


OF BCIENCES OF PUILADELPISIA, 


ACADEMY OF SCIENCES 
ACOLOGICAL SOCIETY Or 
NBERG ACADEMY OF 


OF SAN FRANCISCO, AND 
BELGIUM ; 


FRANKFURT-A-M; 
U-S., AND MEMBER OF THE 


LATE DIKRKCTOR 


LONDON; 


SYDNEY J. HICKSON, M.A., F.R.S., 


BEYRR PROFESSOR OF ZOOLOGY IN THE UNIVERSITY OF MANCHESTER, 


H. A. MINCHIN, M.A., 


PROFESSOR OF PROTOZOOLOGY IN THE UNIVERSITY OF LONDON; 
AND 


GILBERT C. BOURNE, M.A., DSc, F.R.S:, 


LINACRE PROFESSOR OF COMPARATIVE ANATOMY, AND 


FKLLOW OF MERTON COLLEGH, OXFORD. 


WITH LITHOGRAPHIC PLATES AND TEXT-FIGURES. 


LONDON. 
J. & A. CHURCHILL, 7 GREAT MARLBOROUGH STREET. 


SOR 


| EERNNDATE a Fe Ce TAD SL A TT PBSC EE 
Adlard and Son, Impr.,] 


{London and Dorking, 


CONTENTS OF No. 218.—New Series. 


MEMOIRS: 
PAGE 


The Structure and Life-History of Crithidia melophagia (Flu), 
an Endo-parasite of the Sheep-Ked, Melophagus ovinus. By 
Annie Porrzr, B.Sc.Lond., Zoological Research Laboratory, 
University College, London. (With Plates 12 and 13 and 15 Text- 
figures) . , : : ' : - 189 
Studies in the Experimental Analysis of Sex. By Gzorrrey 
Smiru, Fellow of New College, Oxford. (With Plate 14). . 225 
Some Observations on a Flagellate of the Genus Cercomonas. By 
C. M. Wenyon, M.B., B.S., B.Sc., Protozoologist to the London 
School of Tropical Medicine. (With 19 Text-figures) : . 241 
Some Observations on a New Gregarine (Metamera schubergi 
nov. gen., nov. spec.). By H. LynpHursr Duxs, B.A., B.C. 
Cantab. (With Plates 15 and 16) ; : ; - 261 
On the Anatomy of Histriobdella Homari. By CRESSWELL 
SuuareR, M.A., Trinity College, Cambridge. (With Plates 17— 
20 and 5 Text-figures) ; ; : : . 287 
On the Artificial Culture of Marine Plankton Organisms. By E. J. 
Autrn, D.Se., Director of Laboratories and Secretary of the 
Marine Biological Association, and E. W. Netson, Assistant 
Naturalist : : : : ‘ . 861 


& 


New Series, No. 219 (Vol. 55, Part 3). Price 10s, net. 


Subseription per volume (of 4 parts) 40s. net. 


SEPTEMBER, 1910. 
THE 


QUARTERLY JOURNAL 


OF 


MICROSCOPICAL SCIENCE. 


EDITED BY 


Sir RAY LANKESTER, K.C.B., M.A., D.Sc., LU.D., F.R.S., 


HONOKAKY FELLOW OF EXETER COLLEGE, OXFORD, 

MEMBER OV THE INSTITUTE OF FRANCE (associé STRANGER DE L’ACADEMIE DES SCIENCES) ; 
CUKKESPONDENT OF THE IMPERIAL ACADEMY OF SCIENCES OF ST. PETERSBUKG, AND OF THE 
ACADKMY OF SCIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIKNCES 
OF TURIN; FOREIGN MEMBER OF THE ROYAL SOCIETY OF SCIENCES OF 
GOTTINGEN, AND “OF THE ROYAL BOHEMIAN SOCIETY OF SCIENCES, AND 
OF THE ACADEMY OF THE ILINCEI OF ROMER, AND OF THR AMERICAN 
ACADEMY OF ARTS AND SCIENCES OF BOSTON, ASSOCIATE OF THE 
ROYAL ACADEMY OF BELGIUM; HONORARY MEMBER OF THE 
NEW YORK ACADEMY OF SCIENCES, AND OF THR 
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OF 
THE ROYAL PHYSICAL SOCIETY OF EDIN- 

BURGH, AND OF THE 
BIOLOGICAL SOCIKTY OF PARIS, AND OF THE CALIFORNIA ACADEMY OF SCIENCES OF SAN FRANCISCO, AND 
OF THE ROYAL ZOOLOGICAL AND MALACOLOGICAL SOCIETY OF BELGIUM; 
CORRESPONDING MEMBER OF THE SENKENBERG ACADEMY OF FRANKFURT-A-M; 

FOREIGN ASSOCIATE OF THE NATIONAL ACADEMY OF SCIENCES, U-S., AND MEMBER OF THE 
AMERICAN PHILOSOPHICAL SOCIETY 35 
HONORARY FELLOW OF THE ROYAL SOCIETY OF EDINBURGH, 

LATE DIKRKCTOK OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSKUM, LATE PRESIDENT OF THE 
BRITISH ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE; LATE FULLEKIAN PROFRSSOR OF 
PHYSIOLOGY IN THE ROYAL INSTITUTION OF GREAT BRITAIN, 

LATER LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFORD) 
EMERITUS PROFESSOR OF ZOOLOGY AND COMPARATIVE ANATOMY IN UNIVERSITY COLLEGE, UNIVERSITY OF LONDON. 


WITH THE CO-OPERATION OF 


ADAM SEDGWICK, M.A., F.RS., 


FELLOW OF TRINITY COLLEGE, CAMBRIDGE. AND PROFESSOR OF ZOOLOGY IN THE IMPERIAL COLLEGE OF 
SCEENCE AND TECHNOLOGY, LONDON; 


SYDNEY J, HICKSON, M.A., ER.S., 


BEYER PROFESSOR OF ZOOLOGY IN THE UNIVERSITY OF MANCHESTKR, 


Ee -AS MENCEHIN, MA: 


PROFESSOR OF PROTOZOOLOGY IN THE UNIVERSITY OF LONDON; 
AND 


GIBbERT CC. BOURNE, M.A,  DiSc,o2. RS. 


LINACRE PROFKSSOR OF COMPARATIVE ANATOMY, AND FKLLOW OF MERTON COLLEGE, OXFORD. 


WITH LITHOGRAPHIC PLATES AND TEXT-FIGURES. 


J, {TO =542\ 
Wwonceast 


LONDON. 
CHURCHILL, 7 GREAT MARLBOROUGH STREET. 


1910. 


Adlard and Son, Impr., ] {London and Dorking. 


CONTENTS OF No. 219.—New Series. 


MEMOIRS : 


Notes on the Free-living Nematodes. I. The Hermaphrodite Species. 

By F. A. Porrs, M.A., Fellow of Trinity Hall, Cambridge, and 

Demonstrator of Comparative apes 2 in the University. (With 

11 Text-figures.) . : ; . 433 
Observations on Ty panoplasma, Congeri, Part I. The Division of 

the Active Form. By C. H. Martin, B.A., Demonstrator of 

Zoology, University of Glasgow. (With Plate 21 and 1 Text- 

figure) : ; : . 485 
The Development of. PE ee A By A. M. Carr 

SAUNDERS and MarGarer Pootr. (With Plate 22 and 20 Text- 

figures). 497 
The Relation between iste ae Piaman-foe aie in C conte rus 

and Hippolyte. By F. W. Gamstg, F.R.S., Mason Professor of 

Zoology, University of Birmingham. (With Plate 23) . 541 
Is the 'lrophoblast of Hypoblastic Origin as Assheton will have it? 

By A. A. W. Husrecut. (With7 Pext-figures) 585 
The Origin and Formation of Fibrous Tissue produced as a Beaeaie 

to Injury in Pecten Maximus, as a Type of the Lamelli- 

branchiata. By G. H. Drew, B.A., Beit Memorial Research Fellow, 

and W. Dr Morean, F.Z.S. (With Plate 24) . at . 595 


PAGE 


te! on es al 


New Series, No. 217 (Vol. 55, Part 1). Price 10s. net. 
Subscription per volume (of 4 parts) 40s. net. 
APRIL; 1910. 


THE 


QUARTERLY JOURNAL 


OF 


MICROSCOPICAL SCIENCE. 


EDITED BY 


Sir RAY LANKESTER, K.C.B., M.A., D.Sc., LL.D., F.R.S., 


HONOKARY FELLOW OF EXETER COLLEGE, OXYORKD}; CURKKSPONDENT OF THE INSTITUTE OF FRANCE, 
AND OF THE IMPERIAL ACADEMY OF SCIENCES OF 8T. PETERSBURG, AND OF THE ACADEMY 
OF 8CIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIENCES 
OF TURIN; FORKIGN MEMBER OF THE ROYAL SOCIETY OF SCIENCES OF 
GOTTINGEN, AND OF THE ROYAL BOHEMIAN S8OCINTY UF SCIENCKS, AND 
OF THK ACADEMY OF THE LINCEI OF ROMR, AND OF THR AMERICAN 
ACADEMY OF ARTS AND SCIENCES OF BOSTON { ASSOCIATE OF THK 
ROYAL ACADEMY OF BELGIUM; HONORARY MEMBER OF THK 
NEW YORK ACADEMY OF SCIENCES, AND OF THR 
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OF 
THE KOYAL PHYSICAI SOCIETY OF EDIN- 
BURGH, AND OF THE 
BIOLOGICAL SOCIKTY OF PARIS, AND OF THE CALIFORNIA ACADEMY OF SCIENCES OF SAN FRANCISCO, AND 
OF THE ROYAL ZOOLOGICAL AND MALACOLOGICAL SOCIETY OF BELGIUM; 
CORRESPONDING MEMBER OF THE SENKENBERG ACADEMY OF FRANKFURT-A-M; 
FOREIGN ASSOCIATE OF THE NATIONAL ACADEMY OF SCIENCES, U.S., AND MEMBER OF THE 
AMERICAN PHILOSOPHICAL SOCIETY 3 
HONORARY FELLOW OF THE ROYAl SOCIETY OF EDINBURGH; 

LATE DIKRCTOR OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSKUM, LATE PRESIDENT OF THE 
BRITISH ASSOCIATION FORK THE ADVANCEMENT OF SCIENCE; LATE FULLKRIAN PROFESSOR OF 
PHYSIOLOGY IN THR ROYAT, INSTITUTION OF GREAT BRITAIN 5 
LATR LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFORD 3 
EMERITUS PROFESSOR OF ZOOLOGY AND COMPARATIVE ANATOMY IN UNIVERSITY COLLEGE, UNIVERSITY OF LONDON 


WITH THE CO-OPERATION OF 


ADAM SEDGWICK, M.A., F.RS., 


FELLOW OF TRINITY COLLEGE, CAMBRIDGE. AND PROFESSOR OF ZOOLOGY IN THE IMPERIAL COLLEGE OF 
SCIENCE AND TECHNOLOGY, LONDON, 


SEUNEY« J cHICKSON,. MAS ERS, 


BEYER PROFESSOR OF ZOOLOGY IN THE UNIVERSITY OF MANCHESTKR, 


AND 


ie A. MINCHIN, M.A; 


PROFESSOR OF PROTOZOOLOGY IN THE UNIVERSITY OF LONDON, 


WITH LITHOGRAPHIC PLATES AND TEXT-FIGURES. 


OND ON. 
J. & A. CHURCHILL, 7 GREAT MARLBOROUGH STREET. 


2910: 


Adlard and Son, Impr., ] (London and Dorking, 


YATES 


thioaD 


UAB lif 4 shia 


MEMOIRS : 


PAGE 
On the Anatomy and Systematic Position of Incisura (Scissu- 


rella) Lytteltonensis. By Ginserr C. Bourne, Fellow of 

Merton College, Oxford, and Linacre Professor of Comparative 

Anatomy. (With Plates 1-5) 1 
The Eye of Pecten. By W. J. Tier: M.Sc., Demonstantie ae 

Assistant Lecturer in Zoology, University of Belfast. (With 

Plates 6 and 7, and 2 Text-figures) . 49 
Observations on Certain Blood-Parasites of Fishes eneevaae ae 

Rovigno. By Prof. E. A. Mincury, M.A., and H. M. Woovcocx, 

D.Sc. (With Plates 8-10) . : : a 
On Ganymedes anaspidis (nov. gen., nov. a a Grecunne from 

the digestive tract of Anaspides tasmanie (Thompson). 

By Juxian 8S. Huxuey. (With Plate 11, and 5 text-figures) . 155 
The Foetal Membranes of the Vertebrates. By A. A. W. Husrecutr. 177 


With Ten Plates, Royal Ato, 5s. 
CONTRIBUTIONS TO THE KNOWLEDGE OF RHABDOPLEURA 
AND AMPHIOXUS. 
By E. RAY LANKESTER, M.A., LL.D., F.R.S. 
London: J. & A. SSE: 7 Great IES EL paar 


Quarterly Journal of Microscopical 


Science. 

The SUBSCRIPTION is £2 for the Volume of Four Numbers ; 
for this sum (prepaid) the JourNaL is sent Post Free to any part 
of the world. 

BACK NUMBERS of the Journat, which remain in print, are 
now sold at an uniform price of 10/- net. 

The issue of Supprement Numbers being found inconvenient, 
and there being often in the Iditor’s hands an accumulation of 
valuable material, it has been decided to publish this Journal at 
such intervals as may seem desirable, rather than delay the appear- 
ance of Memoirs for a regular quarterly publication. 

The title remains unaltered, though more than Four Numbers 
may be published in the course of a year. 

Hach Number is sold at 10/- net, and Four Numbers make 
up a Volume. 

London: J. & A. CHURCHILL, 7 Great Marlborough Street. 


TO CORRESPONDENTS. 


Authors of original papers published in the Quarterly Journal 
of Microscopical Science receive fifty copies of their communica- 
tion gratis. 

All expenses of publication and illustration are paid by the 
publishers. 

Lithographic plates and text-figures are used in illustration. 
Shaded drawings intended for photographic reproduction as half- 
tone blocks should be executed in “‘ Process Black”’ diluted with 
water as required. Half-tone reproduction is recommended for 
uncoloured drawings of sections and of Protozoa. 

Drawings for text-figures should nor be inserted in the MS., 
but sent in a separate envelope to the Hditor. 

Contributors to this Journal requiring extra copies of their 
communications at their own expense can have them by applying 
to the Printers, 

Messrs. ApLARD & Son, 22}, Bartholomew Close, E.C., on 
the following terms: 

For every four pages or less— 


25 copies : : 5/- 
oO 255 : : : f 6/- 
(see : : : ; 6/6 
100 7/- 


Plates, 2/- per 25 if uncoloured ; if coloured, at the same rate for 
every colour. 
Prepayment by P.O. Order is requested. 
ALL CoMMUNICATIONS FOR THE EDITORS TO BE ADDRESSED TO THE CARE 
or Messrs. J. & A. CourcHitt, 7 Great Mariporougs Srreer, 
Lonpon, W. 


THE MARINE BIOLOGICAL ASSOCIATION 


OF THE 


UNITED KINGDOM. 
Patron—HIS MAJESTY THE KING. 


President—Sir RAY LANKESTER, K.C.B., LL.D., F.R.S. 
20; 


‘THE ASSOCIATION WAS FOUNDED “ TO ESTABLISH AND MAINTAIN LABORATORIES ON 
THE COAST OF THE UNITED KINGDOM, WHERE ACCURATE RESEARCHES MAY BE CARRIED 
ON, LEADING 10 THE IMPROVEMENT OF ZOOLOGICAL AND BOTANICAL SCIENCE, AND TO 
AN INCREASE OF OUR KNOWLEDGE AS REGARDS THE FOOD, LIFE CONDITIONS, AND HABITS 
OF BRITISH FOOD-FISHES AND MOLLUSCS.” 


The Laboratory at Plymouth 
was opened in 1888. Since that time investigations, practical and scientific, have 
been constantly pursued by naturalists appointed by the Association, as well as by 
those from England and abroad who have carried on independent researches. 


Naturalists desiring to work at the Laboratory 


should communicate with the Director, who will supply all information as to 
terms, etc. 


Works published by the Association 


include the following :—‘ A ‘Treatise on the Common Sole,’ J. ‘I’. Cunningham, M.A., 
4to, 25/-.. ‘The Natural History of the Marketable Marine Fishes of the British 
Islands,’ J. I. Cunningham, M.A., 7/6 net (published for the Association by 
Messrs. Macmillan & Co.). 


The Journal of the Marine Biological Association 


is issued half-yearly, price 3/6 each number. 

In addition to these publications, the results of work done in the Laboratory 
are recorded in the ‘Quarterly Journal of Microscopical Science,’ and in other 
scientific journals, British and foreign. 


Specimens of Marine Animals and Plants, 


both living and preserved, according to the best methods, are supplied to the 
principal British Laboratories and Museums. Detailed price lists will be forwarded 
on application. 


$n 


TERMS OF MEMBERSHIP. 


ANNUAL MEMBERS : : . £1 1 Oper annun. 

Lire Members . : : : . 15 15 0 Composition Fee. 
FOUNDERS . : ; Z 2 =i 200340 2G ~ os 
Governors (Life Members of Council) 500 0 0 


Members have the following rights and privileges:—They elect annually the 
Officers and Council; they receive the Journal free by post; they are admitted to ; 
view the Laboratory at any time, and may introduce friends with them ; they have the } 
first claim to rent a table in the Laboratory for research, with use of tanks, boats, etc. ; / 
and have access to the Library at Plymouth. Special privileges ure granted to Governors, 
Founders, and Life Members. | 
Persons desirous of becoming members, or of obtaining any information with 
regard to the Association, should communicate with— 


The DIRECTOR, 
The Laboratory, 
Plymouth. 


ays 


With Ten Plates, Royal 4to, 5s. 
CONTRIBUTIONS TO THE KNOWLEDGE OF RHABDOPLEURA 
AND AMPHIOXUS. 
By E. RAY LANKESTER, M.A., LL.D., F.R.S. 
London: J. & A. CHURCHILL, 7 Great Marlborough Street. 


Quarterly Journal of Microscopical 
Science. 

The SUBSCRIPTION is £2 for the Volume of Tour Numbers ; 
for this sum (prepaid) the Journal is sent Post Free to any part 
of the world. 

BACK NUMBERS of the Journat, which remain in print, are 
now sold at an uniform price of 10/- net. 

‘he issue of Suppremenr Nomeers being found inconvenient, 
and there being often in the Hditor’s hands an accumulation of 
valuable material, it has been decided to publish this Journal at 
such intervals as may seem desirable, rather than delay the appear- 
ance of Memoirs for a regular quarterly publication. 

The title remains unaltered, though more than Four Numbers 
may be published in the course of a year. 

Each Number is sold at 10/- net, and Four Numbers make 
up a Volume. 


Authors of original papers published in the Quarterly Journal 
of Microscopical Science receive fifty copies of their communica- 
tion gratis. 

All expenses of publication and illustration are paid by the 
publishers. 

Lithographic plates and text-figures are used in illustration. 
Shaded drawings intended for photographic reproduction as half- 
tone blocks should be executed in ‘‘ Process Black” diluted with 
water as required. Half-tone reproduction is recommended for 
uncoloured drawings of sections and of Protozoa. 

Drawings for text-figures should nor be inserted in the MS., 
but sent in a separate envelope to the Editor. 

Contributors to this Journal requiring eatra copies of their 
communications at their own expense can have them by applying 
to the Printers, . 

Messrs. Aptarp & Son, 224, Bartholomew Close, H.C., on 
the following terms: 

For every four pages or less— 


25 copies ; 5/- 
a0 5 : , 6/- 
Lose oe Cae : 6/6 
LOO Eo 7/- 


9 , , . . 
Plates, 2/- per 25 if uncoloured; if coloured, at the same rate for 
every colour. 
Prepayment by P.O. Order is requested. 
ALL COMMUNICATIONS FOR THE EDITORS TO BE ADDRESSED TO THE CARE 
or Mussrs. J. & A. Cuurcuitt, 7 Great MarLBorouGh SrreeEv, 
Lonpon, W. 


THE MARINE BIOLOCICAL ASSOCIATION 


OF THE 


UNITED KINGDOM. 
Patron—HIS MAJESTY THE KING. 


President—Sir RAY LANKESTER, K.C.B., LED, FR Se 
sO; 


THE ASSOCIATION WAS FOUNDED “ TO ESTABLISH AND MAINTAIN LABORATORIES ON 
THE COAST OF THE UNITED KINGDOM, WHERE ACCURATE RESEARCHES MAY BE CARRIED 
ON, LEADING TO THE IMPROVEMENT OF ZOOLOGICAL AND BOranIcat SCIENCE, AND TO 
AN INCREASE OF OUR KNOWLEDGE AS REGARDS THE FOOD, LIFE CONDITIONS, AND HABITS 
OF BRITISH FOOD-FISHES AND MOLLUSCS.” 


The Laboratory at Plymouth 


was opened in 1888. Since that time investigations, practical and scientific, have 
been constantly pursued by naturalists appointed by the Association, as well as by 
those from England and abroad who have carried on independent researches. 


Naturalists desiring to work at thé Laboratory 


should communicate with the Director, who will supply all information as to 
terms, etc. 


Works published by the Association 


include the following :—‘ A Treatise on the Common Sole,’ J.T. Cunningham, M.A,, 
4to, 25/-. ‘The Natural History of the Marketable Marine Fishes of the British 


Islands, J. I. Cunningham, M.A., 7/6 net (published for the Association by 
Messrs. Macmillan & Co.). ‘ 


The Journal of the Marine Biological Association 
is issued half-yearly, price 3/6 each number. 
In addition to these publications, the results of work done in the Laboratory 


are recorded in the ‘Quarterly Journal of Microscopical Science,’ and in other 
scientific journals, British and foreign. 


Specimens of Marine Animals and Plants, 


both living and preserved, according to the best methods, are supplied to the 


principal British Laboratories and Museums. Detailed price lists will be forwarded 
on application. 


TERMS OF MEMBERSHIP. 


ANNUAL MEMBERS ; : - £1 1 Oper annun. 
LIFE Memsers . : ‘ , - 15 15 O Composition Fee. 
FOUNDERS . -.. £O0F FO 218 


>” ” 


GOVERNORS (Life Members of Council) 500 0 0 


Members have the following rights and privileges :—They elect annually the 
Officers and Council; they receive the Journal free by post; they are admitted to 
view the Laboratory at any time, and may introduce friends with them ; they have the 
first claim to rent a table in the Laboratory for research, with use of tanks, boats, ete. ; 
and have access to the Library at Plymouth. Special privileges ure granted to Governors, 
Founders, and Life Members. 

Persons desirous of becoming members, or of obtaining any information with 
regard to the Association, should communicate with— 


The DIRECTOR, 
The Laboratory, 
Plymouth. 


ees ya > mA 


With Ten Plates, Royal Ato, 5s. 
CONTRIBUTIONS TO THE KNOWLEDGE OF RHABDOPLEURA 
AND AMPHIOXUS. 
By:E. RAY LANKESTER, M.A., LL.D., F.R.S. 
London: J. & A. CeCe, & Great See eoren Street. 


Quarterly . Journal of Microscopical 


Science. 


The SUBSCRIPTION is £2 for the Volume of Four Numbers ; 
for this sum (prepaid) the JouRNAL is sent Post Free to any part 
of the world. 

BACK NUMBERS of the Journat, which remain in Bret are 


now sold at an uniform price of 10/- net. 
The issue of SuppremMenr Noumpers being found inconvenient, 


and there being often in the Kditor’s hands an accumulation of 
valuable material, it has been decided to publish this Journal at 
such intervals as may seem desirable, rather than delay the appear- 
ance of Memoirs for a regular quarterly publication. 

The title remains unaltered, though more than Four Numbers 
may be published in the course of a year. 

Kach Number is sold at 10/- net, and Four Numbers make 
up a Volume. 


London: J. & A. CHURCHILL, 7 Great Marlborough Street. 


TO CORRESPONDENTS. 


Authors of original papers published in the Quarterly Journal 
of Microscopical Science receive fifty copies of their communica- 
tion gratis. 

All expenses of publication and illustration are paid by the 
publishers. 

Lithographic plates and text-figures are used in illustration. 
Shaded drawings intended for photographic reproduction as half- 
tone blocks should be executed in “‘ Process Black” diluted with 
water as required. MHalf-tone reproduction is recommended for 
uncoloured drawings of sections and of Protozoa. 

Drawings for text-figures should nor be inserted in the MS., 
but sent in a separate envelope to the Editor. 

Contributors to this Journal requiring eatra copies of their 
communications at their own expense can have them by applying 
to the Printers, 

Messrs. ADLARD & Son, 22}, Bartholomew Close, E.C., on 
the following terms : 

For every four pages or less— 


25 copies : ‘ ; i d/- 
aOR S35 , : 6/- 
718 ae ‘ ; : é 6/6 
100. ,, 2/- 


Plates, 2/- per 25 if uncoloured ; if colour ed, at the same rate for 
every colour. 
Prepayment by P.O. Order is requested. 
ALL COMMUNICATIONS FOR THE EDITORS TO BE ADDRESSED TO THE CARE 
or Messrs. J. & A. Courcaiti, 7 Great Marizoroucs Srreet, 
Lonpon, W. 


THE MARINE BIOLOCIGAL ASSOCIATION 


OF THE 


UNITED KINGDOM, 
Patron—HIS MAJESTY THE KING. 


President—Sir RAY LANKESTER, K.C.B., LL.D., F.R.S. 
505 


THE ASSOCIATION WAS FOUNDED ‘“ 10 ESTABLISH AND MAINTAIN LABORATORIES ON 
THE COAST OF THE UNITED KINGDOM, WHERE ACCURATE RESEARCHES MAY BE CARRIED 
ON, LEADING TO THE IMPROVEMENT OF ZOOLOGICAL AND BOTANICAL SCIENCE, AND TO 
AN INOREASE OF OUR KNOWLEDGE AS REGARDS THE FOOD, LIFE CONDITIONS, AND HABITS 
OF BRITISH FOOD-FISHES AND MOLLUSCS.” 


The Laboratory at Plymouth 


was opened in 1888. Since that time investigations, practical and scientific, have 
been constantly pursued by naturalists appainted by the Association, as well as by 
those from England and abroad who have carried on independent researches. 


Naturalists desiring to work at the Laboratory 


should communicate with the Director, who will supply all information as to 
terms, etc. 


Works published by the Association 


include the following :—‘ A 'l'reatise on the Common Sole,’ J.'I'. Cunningham, M.A., 
4to, 25/-. ‘The Natural History of the Marketable Marine Fishes of the British 
Islands, J. TI. Cunningham, M.A., 7/6 net (published for the Association by 
Messrs. Macmillan & Co.). 


The Journal of the Marine Biological Association 
is issued half-yearly, price 3/6 each number. 
In addition to these publications, the results of work done in the Laboratory 
are recorded in the ‘Quarterly Journal of Microscopical Science,’ and in other 
scientific journals, British and foreign. 


Specimens of Marine Animals and Plants, 


both living and preserved, according to the best methods, are supplied to the 
principal British Laboratories and’-Museums. Detailed price lists will be forwarded 
on application. 


TERMS OF MEMBERSHIP. 


ANNUAL MEMBERS . : . £1 1 Operannum. 
LIFE MremBErs . : : ; . 15-15 0 Composition Fee. 
FOUNDERS . 100 0 O a 


Governors (Life Members of Council) 500 O 0 


Members have the following rights and privileges:—They elect annually the 
Officers and Council; they receive the Journal free by post; they are admitted to 
view the Laboratory at any time, and may introduce friends with them; they have the 
first claim to rent a table in the Laboratory for research, with use of tanks, boats, ete. ; 
and have access to the Library at Plymouth. Special privileges ure granted to Governors, 
Founders, and Life Members. 

Persons desirous of becoming members, or of obtaining any information with 
regard to the Association, should communicate with— 


The DIRECTOR, 
The Laboratory, 
Plymouth. 


With Ten Plates, Royal 4to, 5s. 
CONTRIBUTIONS TO THE KNOWLEDGE OF RHABDOPLEURA 
AND AMPHIOXUS. 
By E. RAY LANKESTER, M.A., LL.D., F.R.S. 
London: J. & A. CHa, a Great Sean Street. 


Quarterly Journal of Microscopical 


Science. 


The SUBSCRIPTION is £2 for the Volume of Four Numbers ; 
for this sum (prepaid) the JournaL is sent Post Free to any part 
of the world. 

BACK NUMBERS of the Journat, which remain in print, are 
now sold at an uniform price of 10/- net. 

The issue of Suppnement Nomsers being found inconvenient, 
and there being often in the Hditor’s hands an accumulation of 
valuable material, it has been decided to publish this Journal at 
such intervals as may seem desirable, rather than delay the appear- 
ance ot Memoirs for a regular quarterly publication. 

The title remains unaltered, though more than Four Numbers 
may be published in the course of a year. 

Kach Number is sold at 10/- net, and Four Numbers make 
up a Volume. 


en: J: & A. CHURCHILL, 7 Great eee street. 
TO CORRESPONDENTS. t 


Authors of original papers published in the Quarterly Journal 
of Microscopical Science receive fifty copies of their communica- 
tion gratis. 

All expenses of publication and illustration are paid by the 
publishers. 

Lithographic plates and text-figures are used in illustration. 
Shaded drawings intended for photographic reproduction as half- 
tone blocks should be executed in ‘‘ Process Black” diluted with 
water as required. MHalf-tone reproduction is recommended for 
uncoloured drawings of sections and of Protozoa. 

Drawings for text-figures should nor be inserted in the MS., 
but sent in a separate envelope to the Hditor. 

Contributors to this Journal requiring evtra copies of their 
communications at their own expense can have them by applying 
to the Printers, 

Messrs. ApiarD & Son, 224, Bartholomew Close, E.C., on 
the following terms : 

For every four pages or less— 


25 copies : ; f j 5/- 
5). ey eye ; , : 6/- 
hie. 5 : ; ; : 6/6 
100 We 


Plates, 2/- per 25 if uncoloured; if coloured, at the same rate for 
every ‘colour. 
Prepayment by P.O. Order is requested. 
ALL COMMUNICATIONS FOR THE EDITORS TO BE ADDRUSSED TO THE CARE 
or Messrs. J. & A. Cuurcuint, 7 Great Mariporouca Srreet, 
Lonpon, W. 


THE MARINE BIOLOGICAL ASSOCIATION 


OF THE 


UNITED KINGDOM. 
Patron—HIS MAJESTY THE KING. 


President—Sir RAY LANKESTER, K.C.B., LEZD: 7k R.S: 
20; 


THE ASSOCIATION WAS FOUNDED “ 10 ESTABLISH AND MAINTAIN LABORATORIES ON 
THE COAST OF THE UNITED KINGDOM, WHERE ACCURATE RESEARCHES MAY BE CARRIED 
ON, LEADING 10 THE IMPROVEMENT OF ZOOLOGICAL AND BOvraNIcaAL SCIENCE, AND TO 
AN INCREASE OF OUR KNOWLEDGE AS REGARDS THE FOOD, LIFE CONDITIONS, AND HABITS 
OF BRI’ISH FOOD-FISHES AND MOLLUSCS.” 


The Laboratory at Plymouth 
Was opened in 1888. Since that time investigations, practical and scientific, have 
been constantly pursued by naturalists appointed by the Association, as well as by 
those from England and abroad who have carried on independent researches, 


Naturalists desiring to work at the Laboratory 


should communicate with the Director, who will supply all information as to 
terins, etc, 


Works published by the Association 


include the following :—‘ A ‘Treatise on the Common Sole,’ J.T. Cunningham, M.A., 
4to, 25/-. ‘The Natural History of the Marketable Marine Fishes of the British 
Islands,’ J. T. Cunningham, M.A., 7/6 net (published for the Association by 
Messrs. Macmillan & Co.). 


The Journal of the Marine Biological Association 
is issued half-yearly, price 3/6 each number. 
In addition to these publications, the results of work done in the Laboratory 
are recorded in the ‘Quarterly Journal of Microscopical Science,’ and in other 
scientific journals, British and foreign. 


Specimens of Marine Animals and Plants, 


both living and preserved, according to the best methods, are supplied to the 
principal British Laboratories and Museums. Detailed price lists will be forwarded 
on application. 


TERMS OF MEMBERSHIP. 


ANNUAL MeMBEns : - £1 1,0 per annum. 
LIFE Mempers . : : : - 15 15 O Composition Fee. 
FOUNDERS . - —100°40%:0 = 2 


Governors (Life Members of Council) 500 0 0 


Members have the following rights and privileges:—They elect annually the 
Officers and Council; they receive the Journal free by post; they are admitted to 
view the Laboratory at any time, and may introduce friends with them; they have the 
first claim to rent a table in the Laboratory for research, with use of tanks, boats, ete. ; 
and have access to the Library at Plymouth. Special privileges are granted to Governors, 
Founders, and Life Members. 

Persons desirous of becoming members, or of obtaining any information with 
regard to the Association, should communicate with— 


The DIRECTOR, 
The Laboratory. 
¢) Plymouth. 


ee . = 


’ pee 


tapes ors Journal of micy 


| N.S. vol 55. 910, 
| 


1O- 46556 
Date Loaned | R 


Borrower’s N ame 


4 ? Ve —— Tp 


) 

aC) e) 

JUN 1 | b, [JE Si) | 

i ya stra University Library : 

irae 
a 


Uma J i) a% t “a Pm 
‘ 4 ‘ ‘ 
’ wy , ‘ ; 
r , 5 “ 
‘ ‘ { | ’ 
f 
fy 7 
‘ 
, 
wig li ( , 
H \ ' : 
' 
A 
¥ 
‘ A HO 
de , y " WENA 
? 4 r ‘i 
y whey LY 
r . 
a & : 
wk 
. 
“ 
« 
. 
fan. 
“ 
‘ 
‘ 
. 4 
ay 
ne 
+ 
> 
+ 
® 
‘ 
* 
* 
; 
® 
‘ 
wie 
+ . al 
, ae 
7) 
. . 
’ * 
A 
* * 
* « 
. am