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