HARVARD UNIVERSITY.

PibR AY

OF THE
MUSEUM OF COMPARATIVE ZOOLOGY.
aa) k 2|
| Pes
eer Cee

‘

M en ‘\ a: b a hee “, \4 Ob.

< ioe
ier,

QUARTERLY JOURNAL

OF

MICROSCOPICAL SCIENCE.

EDITED BY

EH. RAY LANKESTER, M.A., LL.D., F.R.S.,

NONORARY FRLLOW OF EXETER COLLEGE, OXFORD; CORRESPONDENT OF THE INSTITUTE OF FRANCE
AND OF THE IMPERIAL ACADEMY OF SCIENCES OF ST. PETERSBURG, AND OF THE ACADEMY
OF SCIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIENCES

OF TURIN; FOREIGN MEMBER OF THE ROYAL SOCIETY OF SCIENCES OF
GUTTINGEN, AND OF THE ROYAL BOHEMIAN SOCIETY OF SCIENCES, AND

OF THE ACADEMY OF THE LINCEI OF ROMER, AND OF THK 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 THE
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OF
THE ROYAL PHYSICAL SOCIETY OF EDIN-

BURGH, AND OF THE

BIOLOGICAL SOCIETY OF PARIS, AND OF THR CALIFORNIA ACADEMY OF SCIENCES OF SAN FRANCISCO),
FOREIGN ASSOCIATE OF THK NATIONAL ACADEMY OF SCIENCES, U-S., AND MEMBER OF

THE AMERICAN PHILOSOPHICAL SOCIETY 5
DIRECTOR OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSKUM, LATE
PROFESSOR OF PHYSIOLOGY IN THE ROYAT INSTITUTION OF GREAT BRITAIN;

FULLERIAN

LATE LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFORD.

WITH THE CO-OPERATION OF

ADAM SEDGWICK, M.A., F.RS.,

FELLOW AND TUTOR OF TRINITY COLLEGE, CAMBRIDGE}

AND

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

BEYER PROFESSOR OF ZOOLOGY IN THE OWENS COLLEGE, MANCHESTER,

VOLUME 50.—New Series.

With Mithographic Plates and Cext-Higures

f T9ID35

ib unoven

rn

“LONDON:
J. & A. CHURCHILL, 7, GREAT MARLBOROUGH
1906.

STREET.

CONTENTS.

CONTENTS OF No. 197, N.S., APRIL, 1906.

MEMOIRS :

The Life-cycle of “Cystobia” irregularis (Minch.), together
with Observations on other ‘‘ Neogamous”’ Gregarines. By
H. M. Woopcock, D.Se.Lond. (With Plates 1—6)

The Anatomy of Oncholaimus vulgaris, Bast., with Notes on
two Parasitic Nematodes. By F. H. Stewart, M.A., B.Sc.,
M.B., Lieut. Indian Medical Service. (With Plates 7—9)

The Hemoflagellates : a Review of Present Knowledge relating to
the Trypanosomes and allied forms. By H. M. Woopcock,
D.Se.Lond. (With Text-figures.) (Zo be continued)

CONTENTS OF No. 198, N.S., JUNE, 1906.

MEMOIRS:

The Hemoflagellates: a Review of Present Knowledge relating
to the Trypanosomes and allied forms. (Continued from p, 231.)
By H. M. Wooncock, D.Se.Lond. (With Text-figures.)

Notes on the Development, Structure, and Origin of the Median
and Paired Fins of Fish. By Epwin 8. Goopricu, F.R.S.,
Fellow of Merton College, Oxford. (With Plates 1O—14)

Preliminary Account of a New Organ in Periplaneta orien-
talis. By Rurn M. Harrison, Lady Margaret Hall, Oxford.
(With Plate 15) : : :

CONTENTS OF No. 199) NS: AUGUST, 1906:

MEMOIRS:
On the Development of Nebalia. By Marearer Rosinson,

Zoological Research Laboratory, University College, London.
(With Plates 16—21) .

PAGE

101

151

233

3033

377

383

1V CONTENTS.

On the Early Stages in the Development of Flustrella hispida
(Fabricius), and on the Existence of a “ Yolk Nueleus” in the
Egg of this Form. By R. M. Pace (née Crank), Late Scholar
of Girton College, Cambridge. (With Plates 22—25)

Researches on the Origin and Development of the Epiblastic
Trabecule and the Pial Sheath of the Optic Nerve of the Frog,
with illustrations of Variations met with in other Vertebrates,
and some Observations on the Lymphatics of the Optic Nerve.
By J. T. Grapvon, M.A., St. John’s College, Oxford. (With
Plates 26 and 27) ‘ 4 : F

Piroplasma muris, Fant., from the Blood of the White Rat,
with Remarks on the Genus Piroplasma. By H. B. Fantuam,
B.Se.Lond., A.R.C.S., Derby Research Scholar, University Col-
lege, Loudon; and Demonstrator in Biology, St. Mary’s Hospital
Medical School. (With Plate 28) : :

CONTENTS OF No. 200, N.S., NOVEMBER, 1906.

MEMOIRS:

On the Structure of the Nephridiaof Dinophilus. By Cress-
WELL SHEARER, Trinity College, Cambridge. (With Plates 29
and 30)

Contributions to our Kaueledee of the arsine of Mictur sate
typhlops, Stirling; Part I1].—The Eye. By Grorerna
Sweet, D.Sc., Melbourne University. (With Plate 31)

Structure and Origin of Canker of the Apple Tree. By James
E. Bromriecp, M.A., M.D.Oxon. (With Plate 32) .

Review of Dr. Richard Goldschmidt’s Monograph of Amphi-
oxides. By A. Wrttey, Hon.M.A.Cantab., D.Se.Lond.,
F.R.S. (With seven text-figures)

The Modification of the Sexual Characters of the Ei cvsal Crab
caused by the Parasite Peltogaster (castration parasitaire of
Giard). By F. A. Ports, B.A., aan Hall, Cambridge.
(With Plates 33 and 34) : °

On the Medusa of Microhydra ryderi, ay on the Known Borne
of Meduse inhabiting Fresh Water. By Epwarp lorts, of
Philadelphia, U.S.A. (With Plates 35 and 36)

On the Freshwater Medusa liberated by Microhydra eyileri,
Potts, and a comparison with Limnocodium. By Epwarp
T. Browne, B.A., Zoological Research Laboratory, University
College, London. (With Plate 37)

TitLe, INDEX, AND CONTENTS.

PAGE

435

479

493

517

599

623

635

New Series, No. 197 (Vol. 50, Part 1). Price 10s, net.

APRIL, 1906.
THE

QUARTERLY JOURNAL

OF

MICROSCOPICAL SCIENCE.

EDITED BY

KH. RAY LANKESTER, M.A., D.Sc., LU.D., F.R.S.,

HONORARY FELLOW OF EXETER COLLEGE, OXFORD}; CORRESPONDENT OF THE INSTITUTE OF FRANCK
AND OF THE IMPERIAL ACADEMY OF SCIENCES OF ST. PETERSBURG, AND OF THE ACADEMY
OF SCIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIENCES
OF TURIN} FOREIGN MEMBER OF THE ROYAL SOCIETY OF SCIENCES OF
GOTTINGEN, AND OF THE ROYAL BONEMIAN SOCIETY OF SCIENCES, AND
OF THE ACADEMY OF THE LINCEI OF ROMR, AND OF THK 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 THE
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OF
THE ROYAL PHYSICAL SOCIETY OF EDIN-
BURGH, AND OF THE
BIOLOGICAL SOCIETY OF PARIS, AND OF THE CALIFORNIA ACADEMY OF SCIENCES OF SAN FRANCISCO)
FOREIGN ASSOCIATE OF THE NATIONAL ACADEMY OF SCIENCES, U.S., AND MEMBER OF
THE AMERICAN PHILOSOPHICAL SOCIETY j
DIRECTOR OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSKUM, LATE FULLERIAN
PROFESSOR OF PHYSIOLOGY IN THR ROYAD INSTITUTION OF GREAT BRITAIN;
LATE LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFORD.

WITH THE CO-OPERATION OF

ADAM SEDGWICK, M.A., F.RS.,

FELLOW AND TUTOR OF TRINITY COLLEGE, CAMBRIDGE 4

W. F. R. WELDON, M.A., F.BS.,

LINACKK PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLKGK, OXFORD;
LATE FRLLOW OF ST. JOHN’S COLLEGE, CAMBRIDGE 3

AND

SYDNEY J. HICKSON, MA] E-ES.,

BEYER PROFESSOR OF ZOOLOGY IN THE OWENS COLLEGE, MANCHESTER.

WITH LITHOGRAPHIC PLATES AND TEXT-FIGURES.

HON DON:
J. & A. CHURCHILL, 7 GREAT MARLBOROUGH STREET.

1906.

Adlard and Son, Impr,, | [London and Dorking.

CONTENTS OF No. 197.—New Series.

MEMOIRS:

PAGE
The Life-cycle of ‘“‘Cystobia” irregularis (Minch.), together with

Observations on other ‘“‘Neogamous”’ Gregarines. By H. M.

Woopcock, D.Se.Lond. (With Plates 1—6) : : a:

The Anatomy of Oncholaimus vulgaris, Bast., with Notes on two
Parasitic Nematodes. By F. H. Stewart, M.A., B.Sc., M.B.,
Lieut. Indian Medical Service. (With Plates7—9) . - Jo

The Hemoflagellates: a Review of Present Knowledge relating to
the Trypanosomes and allied forms. By H. M. Woopcocx,
D.Se.Lond. (With Text-figures.) (Zo be continued) . « phon

LIFE-CYCLE OF “ CYSTOBIA”’ IRREGULARIS (MINCH.). 1

The Life-Cycle of ‘‘Cystobia” irregularis (Minch.),
together with Observations on other ‘ Neo-
gamous’”! Gregarines.’

By
H. M. Woodcock, D.Sc.Lond.

With Plates 1—6.

ConrTENTS.
PAGE

Preface. : ; ; : : : ]

1. Introduction : F : ; : - 3
2. Methods . ; : : : 6
3. Habitat and Mode of Tete) : : : : 11
4, Form, Size, and General Appearance ; : 19
5. Minute Structure . , : : : 24
6. Commencement of Sporulation . : ; : 34
7. Nuclear Multiplication . : : 40
8. Formation of Gametes and their Gonnenien , , 48
9, Spore-formation and Systematic Position . : : 55
10. Precocious Association or ‘* Neogamy” . : : 61
11. General Significance of Association : : : 70
Detailed Summary : . . : : 83
General Summary : ‘ : . : 89
Analysis of Contents ; : : : 90
Bibliography ‘ ‘ : 3 92
Explanation of Plates ; ; : : : 95

PREFACE.

The Gregarines of Holothurians were first brought to my
notice by Professor Minchin, who suggested that I might suit-

1 From veoc, young, and yapoc , marriage, on the analogy of Neosporidia.
I am indebted to Prof. Minchin for suggesting this appropriate and convenient
term.

2 Thesis approved for the Degree of Doctor of Science in the University
of London.

vot. 50, PART 1.—NEW SERIES. 1

2 H. M. WOODCOCK.

ably devote a portion of my time as Derby Scholar in endeavour-
ing to obtain further stages in the life-history of Cystobia
(Gregarina) irregularis. This Gregarine, which is para-
sitic in Holothuria forskali,! the “ cotton-spinner” of our
South-Western coast, was originally described by him (25),
but, in view of the recent important advance in our know-
ledge of the life-cycle of many Gregarines, it appeared
desirable to try and ascertain more fully to what extent
Cystobia agrees with, or diverges from, other members of
the order.

I applied for, and obtained, the use of a British Associa-
tion table at Plymouth, which I occupied in the spring and
again during the summer of 1902. Some of my time, how-
ever, was taken up with other Sporozoan work. Apart from
“free” or unencysted parasites, which were usually examined
fresh, most of the material collected was preserved at once in
different ways and cut, stained, and examined either there or
on my return to University College, where all the drawings
were made.

I experienced, unfortunately, great difficulty in procuring
the hosts, especially in the spring and early summer, when
some very important stages in the life of the parasite are
undergone. Only too often, moreover, the animals, when
obtained, were found to be uninfected. Nevertheless, I have
been successful in learning many interesting additional facts
with regard to this species of C'ystobia—sufficient to
enable me to give a fairly complete account of its life-cycle.
In addition, I have found another Gregarine, inhabiting
Cucumaria pentactes and C, planci, which shows marked
peculiarities in its habitat and trophic life and is certainly a
distinct species. Owing to the even greater scarcity of this
parasite I am unable, in spite of great efforts, to give an equally
complete description of this new form. Whereas in C. irreg-
ularis, the trophic stages are the more uncommon, in this

1 This species is generally known as H. nigra, Kinahan, but Koehler (15)
states that it has been shown to be identical with H. forskali, De C., and
that the latter name takes priority.

”

LIFE-CYCLE OF ‘* CYSTOBIA”’ IRREGULARIS (MINCH.). %

case it is the ripe cysts, containing spores, which are ex-
tremely rare. For this new species I have already proposed
(40) the specific name minchinii, after Professor Minchin,
in recognition of his earlier work on C. irregularis.

Since publishing my preliminary note on the trophic
phases of C. irregularis and C. minchinii, I have been
able to examine the spores of the latter. Mainly as a result,
I have come to the conclusion that these two parasites must
be separated from the other well-known species (C. holo-
thuriz), and placed in a distinct genus. The reasons for
this step are best reserved, however, until we are in a position
to consider a full and revised definition of the two former
species (see below, p. 58). Meanwhile, in describing the life-
history, confusion will be avoided by retaining the old name
of Cystobia for all of them.

I have also come across Diplocystis schneideri,
Kunstler, in a new host, Periplaneta orientalis. The
fortunate rediscovery of this interesting Gregarine has enabled
me to compare its trophic phase with that of other known
species of the genus, and also with that of Cystobia.

(1) Inrropuctrion.

For a complete account of the literature relating to
Cystobia up to 1892 the reader is referred to Minchin’s
article (25) already quoted. Since that date I know of no
paper dealing with this Gregarine except the systematic
enumeration of Labbé (1899). This author, in his ‘Sporo-
zoa’ (17), gave the characteristics of the genus and its known
species, which may be summarised as follows:

Cystobia, Mingazzini, 1891.—A monocystid Gregarine
of large size, oval or irregular in shape. The adult stage
always has two nuclei, probably arising from the early asso-
ciation of two individuals. Spores with dissimilar poles, the
epispore at one end being always turned outwards as a
funnel-like projection. Hight sporozoites in the spore.

(1) C. holothuriz (Schn.).—Cysts attached to the wall

4 H. M. WOODCOCK.

of the blood-vessels by a delicate stalk, and also, when ripe,
floatmeg free in the body-cavity. Spores ovoid, with the
epispore having the funnel at one end, and, in addition, at
the other end a flat process like a lance-head. Sporozoites
with rounded nucleus. Habitat: H. tubulosa, at Naples.
(Also in Chiridota pellucida, according to Sars.)

(2) C.irregularis (Minchin).—Cysts always attached to
the wall of the blood-vessels. Spores ovoid; epispore pro-
longed into a cup-like expansion, open to the exterior. ‘The
sporozoites possess an elongated nucleus. Habitat: blood-
vessels of Holothuria forskali (H. nigra).

(5) C. schneideri, Ming.—Of smaller size, and much
less resistant to sea-water than the foregoing. [Thisis rather
a slight diagnosis on which to base a new sp.] Habitat:
H. poliiand H.impatiens, Naples.

It will be seen, therefore, that practically nothing has been
described for Cystobia with regard to such important ques-
tions as the processes of nuclear division preceding sporulation,
the formation of primary sporoblasts and their conjugation,
whether isogamous or anisogamous, etc. Nor was Minchin
able to make out at all satisfactorily the extent and intimacy
of association in C. irregularis, a point which is of great
interest and importance. A knowledge of these various stages
in the life-history was eminently to be desired, since during
the last few years much has been learnt respecting them in
the case of other Gregarines. It will be useful, first of all, to
outline briefly the results of recent research in this direction.

Siedlecki (86) was the first to correctly work out the life-
eycle, which he did for Lankesteria ascidiw, a member
of the sub-order Ase ptata or Haplocyta, parasitic in the
gut of Ciona intestinalis. Association occurs between two
equal-sized adults. The syzygy rotatesand gradually becomes
spherical, and an outer and inner cyst-membrane are suc-
cessively laiddown. After elimination of much of the nuclear
material, the remainder gives rise by successive divisions,
the earlier of which exhibit well-marked mitosis, to a number
of daughter- or micro-nuclei. Meanwhile the bodies of the

LIFE-CYCLE OF ‘ CYSTOBIA ” IRREGULARIS (MINCH.). 5

two Gregarines have become interlobed in a complicated
manner, though each still remains quite distinct from, and
not in any way united with, the other. Little unimuclear
protuberances appear on the surface of both, and these are
finally cut off as primary sporoblasts or gametes. According
to Siedlecki, the gametes formed from one associate or
parent-individual are perfectly similar to those formed from
the other, and therefore isogamous. Soon after hberation
the gametes throughout the whole of the cyst are seen to be
in a state of rapid motion, the result of which is to get them
all thoroughly mixed. They next conjugate in pairs, each
member of a pair coming, in all probability, from a different
half of the cyst. The definitive sporoblast becomes a spore,
containing eight sporozoites, in the usual manner.’

On the other hand, in Stylorhynchus and Ptero-
cephalus, two members of the sub-order Septata, Léger
(22) and Léger and Duboscq (28) find a differentiation of the
gametes into male and female, with, consequently, anisoga-
mous conjugation.” All the elements arising from the same
chamber of the syzygy are of the same sex, so that we may
consider the two sporonts or associates themselves as respect-
tively male and female. The male gametes are motile, elon-
gated or fusiform in shape, with a minute rostrum anteriorly
and a long flagellum posteriorly. Those of Pterocephalus
are much smaller than the massive female elements of that
parasite; while in Stylorhynchus both kinds are about

1 Cuénot (10), soon after, described very similar facts for different species
of Monocystis in the earthworm, and also for Diplocystis spp., celomic
parasites of the cricket (Gryllus). As regards the latter parasites, there is
no reason to doubt that conjugation is completely isogamous, as in Lankes-
teria, While, however, both Cecconi (8) and Prowazek (81) have confirmed,
in the main, Cuénot’s version of the process in Monocystis, Brasil, in a
recent note (4) states that conjugation in this form is not completely isogamous,
distinct, though slight, differences between the gametes being observable (see
below, pp. 51 and 76.)

2 It is important to notice that anisogamous conjugation is not universal in
the Septata, both Berndt (1) and Paehler (29) having described isogamy in
various species of Gregarina.

6 H. M. WOODCOCK.

the same bulk, the chief differences between them being in
respect of form and movement.

In other words, sexual differentiation in the latter genus is
less marked. There can be little doubt, it seems to me, that
this condition represents an early step in the direction of
isogamy. I hold that isogamy in the Gregarines is secondary,
and derived from anisogamy, being, in fact, closely correlated
with the phenomenon of association; this subject, however,
will be fully discussed later (see p. 73, et. seq.).

(2) Murnops.

(a) Examination.—In examining both Holothuria and
Cucumaria for the parasites, the dissection was rendered
much easier by leaving them for some time previously in
a jar of sea-water, to which a few crystals of menthol had
been added, This stupefied the animals, and always in an
expanded condition. If they were taken out after a few
hours, before death had occurred, the Gregarines were in no
way affected, as they do not inhabit the gut. By this precau-
tion I avoided the disagreeable extrusion of the Cuvierian
organs and the violent contractions of the body which other-
wise take place.

On opening a Holothurian it is easy to ascertain, by
scrutinising the vascular network and blood-vessels gener-
ally, whether any parasites are present. They show up dis-
tinctly either as little white, oval spots if in the lumen, or as
spherical cysts if attached to the wall. In the former case
it is a very delicate matter to obtain the Gregarines free
without injuring them. I followed Minchin’s procedure,
suipping the vessel close to the parasite on either side and,
by gently pressing the wall, causing the Gregarine to pass
out. Sometimes, however careful I was, when at length I
got the Cystobia safely on a slide, it would be distorted
and irregular in shape, although touched as little as possible
during the manipulation (for the reason, see p. 25).

In the case of ©. minchinii in the Cucumarie the

LIFE-CYCLE OF “* CYSTOBIA””’ IRREGULARIS (MINCH.). 7

parasites are usually in the respiratory trees or else attached
by a stalk to the ccelomic epithelium. Those in the latter
position are always adults, and invariably more or less en-
veloped by a double layer of transparent epithelium. Hence
they are much easier to remove and mount without damage
than are the unencysted adults of C.irregularis, for the
body-form is preserved unaltered, with the exception, occasion-
ally, of the free extremity. All that has to be done is to take
firm hold of the stalk of invagination with a fine pair of
forceps, and carefully break it away at its point of attach-
ment to the general epithelial layer. The parasite comes
away, of course, with the stalk and need not, itself, be
touched during the operation. It can thus be readily brought
either into a watch-glass or on to a slide, as required for
fixation.

(6) Fixation and staining.—Adults were usually fixed
on the slide with osmic vapour by holding the slide inverted
over the mouth of a bottle containing a | per cent. solution
of osmic acid for five minutes. They were then washed well
with water, stained with dilute Ranvier’s picro-carmine, dehy-
drated, cleared, and finally mounted in balsam. This pro-
cedure gave very satisfactory results. As an alternative
method, a saturated aqueous solution of corrosive sublimate
was used, to which had been added 5 per cent. of glacial
acetic acid. ‘The objects were afterwards stained either with
carm-alum or with alcoholic para-carmine. Both these stains
differentiated the nuclei well. Some adults obtained from
Cucumariz, which had been previously preserved in 90 per
cent. spirit, were rather shrunken in outline. The nuclear
contents also were retracted away from the membrane, but
the cytoplasm was well preserved. I sectioned a few, and
stained on the slide. Thionin, followed by orange, was very
successful.

For Gregarines in the blood-vessels, respiratory trees, or
elsewhere, and for all cysts, either the corrosive sublimate
and acetic mixture referred to, or strong Flemming, was
generally best. Perenyi’s fluid and picro-sulphuric were

8 H. M. WOODCOCK.

also tried, but only with indifferent success. Material was
left in sublimate and acetic twenty to forty minutes, and in
Flemming two to four hours. That fixed by the former
method was usually stained in bulk with borax- or para-
carmine, but material by the latter generally first on the slide.
If evaginated Gregarines were still in the trophic condition
(see below, pp. 15 and 20) the nuclei were easily visible ; if,
on the other hand, the parasites had commenced to sporulate,
it was usually necessary to cut the cysts before they could be
examined. In sections of advanced cysts stained whole the
nuclei were usually nicely differentiated ; in earlier stages,
however, the nuclei, for some reason or other, were indistinct
and the sections had to be restained.

Sections were generally cut 3-4 « thick, but occasionally
more. In experimental cutting of the blood-vessels and
of the gut, undertaken in the hope of finding very young
and minute forms, the usual thickness was 10-12 p. As
a rule the cutting presented no difficulty ; only when the
parasites were in the retractor muscles did I find it advisable
to paint the block with a solution of collodion and gum-
mastic.

For staining on the slide the combination most frequently
used, and the one which gave the best results, was Heiden-
hain’s iron-hzematoxylin method, followed by orange or eosin.
Thionin and orange rendered good service, as did also Kleinen-
berg’s hematoxylin, the slides being immersed in the latter
for a long time, forty-eight hours or so. Safranin, although
made up in two or three different ways, was rather diffuse,
and never stained the nuclear reticulum, but only the karyo-
some. Neither Auerbach’s mixture (methyl-green and acid
fuchsin) nor Romanowsky’s stain was of much service as a
“double” stain. I find that methyl-green cannot be relied
on as a chromatin stain unless used fresh on the living animal,
and this was quite unsuitable for Cystobia. Romanowsky’s
stain also, though well adapted for blood-films and smear
preparations, is most unsuitable for tissue parasites or where
sections are required. Neither for Gregarines nor for Myxo-

LIFE-CYCLE OF ‘‘ UYSTOBIA””’ IRREGULARIS (MINCH.). 9

sporidia is it of the slightest use so far as nuclear detail is
concerned.

Special effects of fixatives or stains on different structures
will be mentioned when dealing with them in detail.

(c) Attempts at artificial infection.—I endeavoured
to keep the Holothuriz alive in the aquarium, but from
some cause or other they would not settle; it may have been
owing to the great difference between the pressure in the
tanks and that to which they are accustomed in their natural
surroundings. Invariably after a short time the skin became
broken and patchy, showing evident signs of maceration, and
this was the sure prelude to general evisceration.) The
animals were always kept quite separate from anything else
hkely to injure them, so the ill-health was not the result of
being bitten. For this reason alone the prospect of successful
infection and of obtaining early stages of the parasites
appeared to be very doubtful. When, moreover, I came to
try it, 1 soon found artificial infection at all to be a most
difficult matter.

It was necessary to actually convey the cyst containing
spores inside the Holothurian’s mouth, not only to be certain
that the animal really swallowed the spores, but also to know
approximately the time, in order to have any chance (by
killing the animals after different intervals) of subsequently
observing the liberated sporozoites and their passage through
the gut-wall. A Holothurian’s mode of eating is to sweep up
particles of sand, shell, etc., with whatever organic material
may be amongst them, from the surface on which it happens
to be crawling by means of its tentacles. These, of the
Aspidochirote type, are furnished with an expanded, brush-
hke, distal end to which the particles adhere. The tentacles
are in turn stuffed into the mouth and then withdrawn, the
food having been sucked up into the cesophagus.

I could not trust to such a haphazard method as this
with much likelihood of success, so I endeavoured to feed

1 Ludwig (24) mentions this curious fact, saying that, after irritation, “ die
Haut sich ziemlich rasch in formlosen Schleim auflost.”

10 H. M. WOODCOCR.

the animals directly. As soon as they were touched, however,
they withdrew their tentacles into the buccal cavity, and
closed this tightly up, leaving nothing visible externally. They
evinced, moreover, a decided reluctance to be turned on their
backs, which was necessary since the buccal cavity is slightly
ventral, or at any rate normally turned ventrally. Sometimes
after holding a Holothurian gently, yet firmly, with my hand,
often so long that my arm was quite cramped, it would unfold
its tentacles and allow me a beautiful view of its mouth. I
then cautiously steered a tiny piece of blood-vessel with a
cyst attached’ to the open mouth. I have seen it pass in and
apparently fall some distance down the gullet, but it was
nearly always pushed up again and out into the buccal cavity,
usually to be lost among the tentacles. Whether this expul-
sion took place actively or passively I am not quite certain,
although I think probably the latter, since there generally
seemed to be a current alternately of inspiration and expira-
tion when the animal was not feeding. If the cyst stayed in
as long as I retained my hold, it invariably fell out when I
allowed the animal to resume its normal position. Now and
again, by way of diversion, the Holothurian would suddenly
eject the whole contents of its body about my hand, which I
had then to extricate from the Cuvierian organs.

I also tried mixing some cysts with particles of sand and
mud, making a small accumulation on the floor of a dish and
placing the animal with its buccal cavity near to it. The
heap would be disturbed by the tentacles, to one of which a
cyst, or part of one, might even adhere, only to be brushed
off by something a moment later ; in short, the Holothurian
would not eat to order. Once or twice I thought a few spores
might possibly have succeeded in getting well down and
staying there, and the animals were killed after a certain
time had elapsed. I fixed and sectioned different portions of

1 Since there is no true cyst-wall to a ripe cyst of C. irregularis (the
thin eyst-membrane having before this broken down and disappeared), the
spores are only held together by the delicate evaginated epithelium of the
blood-vessel.

LIFE-CYCLE OF “ CYSTOBIA ? IRREGULARIS (MINCH.). 11

the gut, but, after laborious examination of a great many
sections, no traces of Cystobia could be found.'

(3) Hasrray and Moper or Lire.

(a) ©. irregularis.

Minchin (loc. cit.) has described the general habitat of this
parasite and its relation to the vascular system. This is well
shown in his figures 9, 10, and 19, and also in my figures 5, 20,
and 41. Some additional observations, however, may be noted.

Trophozoites—as both young forms and adults are termed
while they are still taking in nourishment—occur chiefly in
the complex vascular meshwork, known as the “‘rete mirabile,”
which is attached to the second loop of the gut, the intestine
proper. ‘This is essentially the region of absorption, and the
fluid circulating in these finely anastomosing vessels is doubt-
less very rich in nutrient solutions. Hence it is particularly
suitable for the growing Gregarines, which probably pass
into it directly from the intestine.

Histology of the vascular network.—tThe structure
of this network differs considerably from that of the more
anterior part of the vascular system, which runs along the
first, or what may be called the stomach, loop. In the latter
the longitudinal vessels are much more distinct and of larger
calibre, and the vascular plexus is much simpler. The chief
difference, however, is in the wall. That of a main vessel or
connecting branch in this anterior region is very thin and
transparent-looking, and pale or slightly reddish in colour.
In section (fig. 7, pl. 1) it is seen to consist of loose, spongy
cells (sp. c.), bordering? and traversing the lumen irregularly,

1 A reason for these unsuccessful infections is perhaps to be found in the
fact that, in the majority of spore-containing cysts, subsequent examination of
stained material showed that the spores themselves were not ripe, having only
four nuclei. This preponderance of the four-nuclear stage was noticed by
Minchin (loc. cit.). Probably the enclosing epithelium often breaks down
before the sporozoites are fully formed, and the spores ripen either in the

body-cavity or after expulsion from the host (see below, in text).
? The vessels have no true epithelial lining internally.

12 H. M. WOODCOCK.

then a delicate muscular layer (m. l.), and, externally, a
single layer of coelomic epithelium (c. ep.).

The wall of the complicated “rete mirabile,” on the other
hand, always appears yellow to yellowish-brown and opaque,
and in section (fig. 8) is much thicker and of a firmer texture,
with a corresponding reduction in the lumen. The loose,
spongy tissue is less developed, and only forms a thin layer
most internally (sp. c.). Next comes the layer of muscle-
fibres (m. 1.), and, finally, making up nearly the whole thick-
ness of the wall, a great development of coelomic epithelium
(emejaye

Relation of the parasites to the blood-vessels.—
To return to our parasites. As they grow they are carried
about passively (for they lack all power of movement, see
below, p. 19) along with the blood-stream, and at last pass
into the larger thin-walled vessels—the distributing part.
Obviously it is much easier for them to evaginate the walls
here than it would be if they remained in the “ rete mirabile” ;
in fact, I never once saw a cyst attached to this latter. A
curious place of occurrence of the evaginated cysts is in
connection with the membrane which stretches from the ring-
canals to the ossicles in the body-wall and in which run the
radial canals. Lying in this membrane are numerous vascular
cross-connections between the radial vessels, and to these the
Gregarines are often attached in considerable numbers, having
evaginated their walls.

I never noticed any cysts free in the body-cavity of the
Holothurian, a point in which C.irregularis differs from
C. holothurix, in H. tubulosa, where even sporulating
cysts—not yet ripe—occur free. ‘The fact that out of a great
many cysts none were loose, although several had formed
spores, and also the delicate nature of the cyst-envelope (see
below, p. 36), points to the conclusion that in this species the
cyst ruptures in situ, liberating the spores into the ccelome ;

1 This condition affords an interesting parallel to the layer of yellow cells

around the intestine of Lumbricus, and it is not improbable that this great
development of ccelomic epithelium serves a similar purpose in the two cases.

LIFE-CYCLE OF “ CYSTOBIA”’? IRREGULARIS (MINCH.). 18

they most likely escape thence when some Cuvierian organs
are extruded. Probably the stalk of attachment is here
stronger, in order to prevent the Gregarines breaking loose
and being carried away before spore-formation has proceeded
far; in H. tubulosa, which does not possess Cuvierian
organs, such a precaution is not necessary.

There is no hard and fast line to be drawn with regard to
the condition the parasites are in when they evaginate the wall
of the blood-vessel. They may, as sporonts, have already
commenced nuclear division, or they may be quite young
trophozoites. In fact, the smallest example I obtained was
in the latter condition. Such cases are frequent towards
autumn, when the lowering of the temperature and other
factors proclaiming the end of the season may perhaps induce
precocious evagination, in the endeavour to sporulate before
the approach of winter. About one third of the adults met
with were “out,” and had become rounded off ; the remainder
were, of course, in the lumen, and these possessed a perfectly
definite body-form.

(b) C. minchinii.

The habitat of C. minchinii in Cucumaria is very
different from that of C. irregularis. In the first place,
I never once came across the parasites in the vascular net-
work attached to the gut, strong & priori evidence that
they do not penetrate through the wall of this latter. The
most general situation is inside the respiratory trees, in the
wall of the branching and blindly ending diverticula. Here
they occur as white, opaque little spheres. PI. I, fig. 4, gives
an idea of part of a small diverticulum, which contained seven
Gregarines. When present, the parasites are sometimes very
numerous and all in a practically similar stage, only varying
in size. They range from 17, to ‘20 mm. in diameter, and
between these limits all gradations in size are to be met with.
The smallest individuals I obtained were in this situation.
Figs.31 and 382 show two of these tiny ones in section, both being

14 H. M. WOODCOCK.

in the loose connective tissue of the wall. In each ease Juin.
is the lumen of the diverticulum, 2.2. are the nuclei in the
wall, m.f. muscle-fibres, and ¢.n. nuclei of the ccelomic
epithelium externally. Two much larger examples are shown
in figs. 18, 19, pl. 2, and it will be seen that the connective
tissue of the wall tends to arrange itself in layers around the
parasite.

Probable mode of infection.—This situation, together
with the non-occurrence in the vascular network, leads me
to think that the infection is not by way of the mouth and
gut, as in Holothuria, but through the cloacal aperture
and into the respiratory trees. Probably the nature of the
tentacles, which are here of the Dendrochirote type, and the
slightly different manner of feeding account for this. Whereas
a Holothuria shovels up sand and shell, together with any
accompanying organic matter, into its mouth, Cucumaria
spreads its branched tentacles, like a beautiful net encircling
its anterior end, and waits for living organisms to be entangled
in the trap. It then conveys its tentacles with their prey, one
after another, to its mouth. Hence the passive spores lying
about on the ground are far more likely to be sucked up with
the respiratory current through the cloaca and so into the
“trees.” The fact, too, that the Cucumarize-are more
sedentary than the Holothurie, not moving about so much
in search of food, may explain the comparative scarcity of
the parasites. This is, I believe, quite a unique instance
of the “‘ casual” or accidental method of infection, which is
usually accomplished by the spores being taken in at the
mouth when the host is feeding. ‘The excretory acids known
to be present in the trees probably perform in this case the
function—elsewhere allotted to the gastric or pancreatic
juice—of opening the spores.

Relation of the parasites to the cclomic epi-
thelium.—Adults of the typical gregariniform shape never
occurred in the trees, but in a very curious position, trying,
apparently, to penetrate the coelomic epithelium, either of the
body-wall (most frequently) or of one of the vascular strands

LIFE-CYCLE OF “ CYSTOBIA ”? TRREGULARIS (MINCH.). 15

connecting the posterior part of the gut with the body-wall.
Fig. 6a shows two Cystobie, each attached to the body-
wall by an epithelial stalk, which itself is partially invaginated
by the parasite. Fig. 6b represents one of the two drawn
with a lens; the slight constriction about one third of the
length from the free end marks off the portion of the Gregarine
still left uncovered by the epithelium from that which is
already surrounded by it. ‘here can be no doubt that they
are going in and not coming out; of this I have quite assured
myself. The process is exactly the opposite to that of
evagination which occurs in C. irregularis. One must
perforce suppose something like the following to take place:
After a successful infection some of the young parasites in
the trees, instead of becoming rounded off and growing
in situ, pass straight through the wall into the ccelome,
assuming the typical ovoid form. (Unfortunately, I have
never observed any free in the body-cavity.) Probably the
parasites are not long carried about here and there by the
movements of the coelomic fluid before touching a suitable
place. To this the Gregarine would at once adhere, doubtless
by means of a little secretion, which at the same time starts
the ccelomic epithelium at that point proliferating, with the
result that a stalk is formed into which the parasite is mean-
while pushing. ‘he process undoubtedly takes considerable
time before being completed, during which the Gregarine
continues growing, for I have seen parasites of very different
size in this position.

A typical C. minchinii thus endeavouring to penetrate
the epithelium is seen in outline and optical section in fig. 9.
The stalk is, of course, broken off from its attachment to
the body-wall. The dotted lines represent the limit of the
epithelial investment, where it is reflexed internally. Fig. 10,
drawn in surface view, shows the covering of epithelium ™
more distinctly. The Gregarine drawn on a large scale in
fig. 12 was similarly attached to one of the more or less
vascular strands crossing the body-cavity. It is viewed
whole, in optical section, and the lumen of the stalk (in

16 H. M. WOODCOCK.

communication with a blood-capillary) contains some of the
peculiar amoebocytes (am.), laden with excretory pigmented
granules. The bending inwards of the wall at b, and the
proliferation of nuclei where it is applied to the Gregarine
are also well shown. The cytoplasm of the part still un-
covered has mostly shrunk away from the limiting-membrane;
this is probably due to the fixing. Fig. 17, Pl. 2, shows a small
portion of another invaginated parasite and its covering,
highly magnified. The cytoplasm is dense and granular,
and closely surrounded by the inner layer of epithelium.
Here, also, the lumen of the stalk (/) contains many amcebo-
cytes, which are very common in the blood-capillaries of both
Cucumaria and Holothuria; the peripherally-situated
nucleus (x) of each is deeply stained.

Finally, in Pl. 2, fig. 13, we have a Cystobia, which has
thrust its anterior end, as a long finger-like process, into a
rather narrow vascular cord; into this the animal appears to
be trying to penetrate. This parasite was found in a spirit-
fixed Cucumaria, and its outer, free end is rather shrunk
and irregular, as also are the outlines of the nuclei. The
section along c—p (fig. 14a) shows the Gregarine every-
where enclosed by the double epithelial layer. X is a portion
of the process in the strand, which is also cut through, e
being the outer and e’ the inner inyaginated wall. Lum.
is the lumen, more or less filled up with spongy tissue, with
here and there a few amcebocytes (am.). Fig. 140 is a
section drawn along the line a—s, and passing through
the tip of the protuberance (p). This is still enveloped by a
cellular layer, the cells being thickly aggregated at one
side.

The above is the only instance I obtained showing such a
modification of the Gregarine’s shape, and I doubt whether
it frequently occurs. I am more inclined to think that the
parasites, after a preliminary invagination of the coelomic
epithelium, are rather enclosed and overgrown by this and
the underlying connective tissue of the immediate neighbour-
hood than that they themselves actually penetrate deeply

LIFE-OYCLE OF “ CYSTOBIA””’ IRREGULARTS (MINCH.). 17

when about to encyst. They are probably able to induce
proliferation of the surrounding tissue to a considerable
extent. Nor do I think that they usually break through
the epithelial layer which they first invaginate, for in sections
of encysted stages this is always present next to the body
of the animal, and completely enclosing it.

Situation in which the parasites encyst.—The re-
maining position in which I found the parasites bears out
these suppositions. Hncysted Cystobie often occurred
firmly attached to—sometimes, indeed, almost in—the strong
retractor muscles which work the buccal mass. PI. 1, fig. 11,
shows three so attached, m. being a piece of the retractor
muscle. The two larger oneg are fairly separate from the
muscle, connected to it by the short thick stalk (st.), but
the smaller one is more completely imbedded. Fig. 16 is a
section through a portion of the muscle with a Gregarine
firmly enclosed in the proliferated epithelium and connective
tissue lying external to the muscle-fibres. A is the originally
invaginated epithelium equivalent to that seen at b, figs. 9,
10, and 12. Hp. is the normal ccelomic epithelium, which
at ep’ is flattened out and in some places wanting, evidently
having been unable to keep pace with the proliferation of
connective tissue. C., c. are concentric layers of connective-
tissue around the parasite, and m., m. are the muscle-fibres.
Fig. 15 shows another one so encysted in section; it is more
separate from the muscle, which in this case is cut trans-
versely ; nevertheless the parasite is firmly enough attached.
A thick layer of connective tissue (c.) surrounds it, especially
fibrous nearest to the inner epithelial cells, which are here a
little disorganised. The nuclei (w.) of the outer epithelial
layer can be seen in places, extending over the surface.

Conclusions.

It will, perhaps, be suggested, may not these Cystobize
in the connective tissue surrounding the muscle have arrived
in this situation by passing along a blood-capillary as sporo-

vou, 50, part 1,—NEW SERIES. 2

18 H. M. WOODCOCK.

zoites, or very young trophozoites (starting originally from
the vascular system of the gut), then grow in situ and so
attain to the sporont phase? I think this is conclusively
negatived by the following considerations:

(a) I have never seen a single C. minchinii in the vas-
cular system proper (i.e. the vessels and network attached
to the gut), or in any way connected with it.

(b) I have never seen small forms in the connective tissue
around the muscle, and indeed, j

(c) I have never found tiny forms anywhere but in the
respiratory trees.

(d) The distinct epithelial layer next to the parasites when
in this situation. No such cellular aggregation is ever found
around them while in the trees.

(ec) The frequently seen process of attempted invagination
already described represents the commencing stage of en-
cystment. No other explanation than the one above given,
namely that the animals are endeavouring to push into, and
are becoming enclosed by, the host’s tissue, is possible.!

These facts lead one inevitably, I think, to the conclusion
that the parasites enter the respiratory trees directly from
the outside, and that many of them pass thence into the
coelome,” and as soon as possible become attached to the
coelomic epithelium, either of the body-wall or of a connective-
tissue or muscle strand, where, after further growth, they
finally become encysted.

I never observed any sporulating stages in the respiratory
trees, and so cannot say whether the Gregarines sporulate in
this position or not. Ido not think those which have become
large and spherical in this position pass into the ccelome.

1 Prof. Minchin kindly looked at several of my preparations, which show
the process most clearly, and he is equally of the opinion that the parasites
are entering and not emerging.

? Presumably the young forms always pass through the wall of the tree into
the body-cavity, since any internal (ceelomic) openings of the former are
usually denied.

LIFE-CYCLE OF “ CYSTOBIA”” IRREGULARIS (MINCH.). 19

(4) Form, Size, anD GENERAL APPEARANCE.
(a) C. irregularis.

Observed living and “ free,” the parasites never showed
the least tendency to displace themselves ; neither did I notice
anything comparable to euglenoid movements causing flexion
or constriction of the body,! such as Brasil (2) figures in the
case of Urospora lagidis. Changes in shape, e.g. the
rounding off of the Gregarines, on encysting are probably
more passive than active—that is to say, impressed, as it were,
upon them by their relation to the surrounding tissue; and
the same applies equally to C. minchinii. In short, the
Gregarines appear to_be perfectly motionless ;* this peculiarity
will be readily understood when the minute structure of the
peripheral region of the body has been described.

The typical form of an adult trophozoite, really of a double-
adult or “couple,” is that of a beautifully symmetrical ovoid.
Fig. 1, pl. 1, represents a medium-sized parasite, its length
being °5 mm. and its greatest width ‘25 mm. Around the
middle of the body is a slight, V-shaped constriction, causing
a break in the contour of the side. ‘his marks the plane of
junction of the two members of the couple, which is further
indicated by the distinct septum running transversely across
the body.*® The two halves are always, so far as can be seen,
of equal size. In each is a perfectly spherical nucleus with a
single large karyosome. Another rather larger trophozoite,
obtained free of the vessel and scarcely so perfect in outline,
is shown in fig. 2. Here the union is complete, there being
no perceptible break in the contour, and the couple might be

1 Trregularities in contour, such as Minchin (loc. cit.) mistook for the
normal appearance, are undoubtedly the result of deformation.

2 Brasil, in a quite recent paper (3), characterises Gonospora varia as
“completely immobile.” This ccelomic parasite is, indeed, probably closely
related to C. irregularis (see under Systematic Position, p. 60).

3 The reasons why this septum is considered to represent a plane of union,

and not a separating partition or plane of division, will be discussed later (see
below, p. 65).

20 H. M. WOODCOCK.

taken for one septate Gregarine, with equal protomerite and
deutomerite, were it not for the important fact of there being
a nucleus in each half. The two clearer spheres denote the
position of the nuclei as they showed up in life. In fig. 3 is
drawn a much younger stage, where the two associates, though
firmly attached to one another and not separable by gentle
touching, are not yet so completely united.

On the other hand, this precocious association or neogamy!
may be still more intimate in character. In many cases,
espec‘ally in trophozoites which had early evaginated the
wall of the vessel, but also sometimes in adults still in the
lumen, there was no septum at all, and the two nuclei were often
in contact. Figs. 5a and b show couples of this nature. In
the latter 7 is part of the second nucleus, just beginning to
be cut through. Here complete fusion of parasites has taken
place, with the result that one appears to be looking at a
single binuclear Gregarine. Such unions evidently occur
extremely early in the life-history (see below, p. 64). I may
here say, indeed, that I have never succeeded in finding an
isolated individual, either of this species or (still less) of
C. minchinii. I have searched hundreds of sections in
the endeavour, but up till now have not seen a single uni-
nuclear form. Fig. 30, pl. 3, shows the smallest specimen of
C. irregularis which I obtained, and this has a diameter
of barely 20. Yet it has two relatively large nuclei, in
this case touching each other and with no sign whatever of
any dividing septum between.

(b) C. minchinii.

The extreme degree to which neogamy attains in C. irre gu-
laris is, so far as I am aware, the only condition to be
observed in C. minchinii. The smallest examples of this

1 I propose to use the terms neogamy and neogamous in describing
this phenomenon, since, besides implying the early occurrence of the process,
they also indicate its essential meaning, to which I attach great importance
(see below, p. 74, et seq., where this interesting question is fully discussed).

LIFE-CYCLE OF “ CYSTOBIA’”’ IRREGULARIS (MINCH.). 21

parasite which I have seen are drawn in figs. 31 and 32.
The latter individual is only 18m by 12,, and from its
elongated form (not yet having become rounded off) looks as
though it had but recently passed from the lumen into the
wall of the respiratory tree. Its general appearance is
similar to that of the C. irregularis of fig. 30. The Gre-
garine in fig. 31 shows only one nucleus, but the other is
only one or two sections distant. Other, larger, trophozoites
in the same situation are seen in figs. 18 and 19, and in
neither is there any trace of a septum. The parasite in the
latter figure has a diameter of 0°15 mm. and the two nuclei
are far apart from each other, but in the other trophozoite,
which is rather smaller, they are again in contact. This
nuclear contiguity ‘is purely accidental, and does not indi-
cate either nuclear union or separation; in all cases the
nuclear membrane is quite intact. As in C. irregularis,
each nucleus has a single large karyosome. The body of
the parasite, when in this situation, is always practically
spherical, this being, probably, because it thus offers more
or less uniform resistance at all points to the surrounding
tissue-layers of the host.

The true “gregariniform” shape, however, is here also that of
aregular ovoid. This is exhibited by all the individuals which
are endeavouring to penetrate the ccelomic epithelium. ‘Two
of these are seen in figs. 9 and 10, the larger one being
0°34 mm. in length by 0°2 mm. in width. Besides the invari-
able absence of any septum, another very constant feature of
this species, and one which distinguishes it from C. irregu-
laris, is the position of the nuclei. In a typical gregarini-
form adult the nuclei are always placed transversely (instead
of longitudinally) with respect to the long axis of the body,
and generally about one third of the animal’s length from the
anterior end, designating thus the end which is farthest in.

The transverse position of the nuclei in the gregariniform
adults of C. minchinii strongly suggests that the association

1 The same thing also occurs not infrequently in C. holothuria, from
H. tubulosa (see Minchin, loc. cit., fig. 21).

22 H. M. WOODCOCK.

is lateral (i. e. that the members of a pair originally join side
to side after the manner in Gonospora sparsa, Léger
[19] ), and not terminal (end to end) as in C. irregularis.!
This arrangement is, of course, masked while the animals are
in the trees.

Trophozoites finally encysted are seen in figs. 15 and 167;
the former of these represents a very large parasite *56 mm.
by ‘43 mm. in diameter, which should rather be termed a
sporont, asit was probably ready to begin sporulation. Both
individuals show only one nucleus in the section drawn, the
other being some sections further on.

(c) Diplocystis schneideri.

I have been able, fortunately, to re-examine this interesting
parasite, originally described by Kunstler (16). Although in-
cluding an account of it in this paper chiefly for the purpose
of comparing neogaiy in this form with the same occurrence
in other species of the genus and also in Cystobia, I may
here indicate why I thus identify the Gregarines which have
come under my notice. ‘The parasites occurred in a single

1 It is interesting to note, in this connection, that Sars (82), so long ago
as 1861, figured Gregarines associated in couples and, moreover, laterally, in
the Holothurian, Chiridota pellucida, ‘The parasites apparently adhere to
the outside of the blood-vessels, which have here little pear- or flask-shaped
diverticula. They appear only loosely attached to each other, with, of course,
a distinct wall or partition between the two members, separating the trans-
versely-placed nuclei. Judging from the magnification given, their length was
about ‘125 mm. and the breadth ‘06 mm. It is impossible to say whether they
were adults or not, as this is all Sars describes of the parasites. In “ pairing”
side-to-side, and in being apparently free in the body-cavity (they do not seem
to be attached by any stalk to the vessel, i. e, to have evaginated its wall),
they exhibit a certain resemblance to C. minchinii; the association would
appear, however, to be much less intimate in character. It would be very
interesting if these Gregarines could be re-discovered in Chiridota.

2 The irregularity in outline at times shown, unfortunately, by these
encysted stages (e. g. fig. 15) is due to the difficulty I experienced in fixing
them, owing to the thick connective-tissue layer through which the fixative
had to penetrate. Corrosive sublimate, well sharpened with acetic acid, served
best.

LIFE-CYCLE OF “‘ CYSTOBIA”” IRREGULARIS (MINCH.). 23

specimen of the common cockroach, Periplaneta orien-
talis, and were all in a practically similar condition, viz. that
of trophozoites of varying age and size.

The animals agree, on the whole, with Kunstler’s descrip-
tion as regards size, shape, and general appearance. I found
the length of a couple to vary from 1°2 mm. up to nearly
1:7 mm., while the greatest breadth varied from ‘6 mm. to
Imm. None of those measured were quite as large as the
largest Kunstler found, which attained 2 mm. in length, but
since the size of these coelomic Gregarines varies consider-
ably, not much stress need be laid on that point. Figs. 21
and 22 show couples of large and small size respectively, that
in fig. 21 being seen whole, while fig. 22 represents a section
which happens to pass through the nucleus of one member of
the pair. An adult syzygy is distinctly bilobed ; and this was
the case with all the individuals examined. Moreover, a well-
marked septum is invariably present.1. The couple is always
longer than broad, more or less dumbbell-shaped in fact, and
never rounded or spherical as in D. minor (vide Cuénot, loc.
cit... Further, the much larger size to which they grow and
the structure of the nucleus—which is, in general, quite like
Kunstler’s description and figures—also tend to remove my
Gregarines from this species of Cuénot’s. Hach couple has a
general or “common” membrane completely investing it
(v.m., fig. 21) , which is quite unrepresented in the tropho-
zoites of Cystobia. In this respect my parasites agree
equally with D. minor and D. schneideri, and differ from
D. major (see below, p. 62). The above facts prove without

1 Unfortunately, many of the specimens showed the appearance drawn in
fig. 23. Whether fixed by corrosive and acetic, Flemming, or absolute alcohol,
most of the protoplasm (the endoplasm) was usually found more or less
retracted, leaving only the ectoplasm and delicate cuticle in close connection
with the investing membrane to show the true size and shape of the animal
when alive. Under a low power, of course, these together appear extremely
thin, compared with the relatively huge mass of endoplasm, ‘lhe apparent
thickness of the septum in this fig. is due to its being viewed slightly out of
the plane in which it lies; I generally found this appearance in such shrunk
specimens.

24 H. M. WOODCOCK.

doubt, I think, that these Gregarines belong to Kunstlev’s
species.

Triple association.—Not infrequently I came across
instances where the trophozoites possessed three nuclei. In
none of these cases was there any septum visible; they
evideutly belonged to the category of very early associations.
Figs. 20, 29, and 28 show examples of such, the two first
being C. irregularis and the latter one C. minchinii.
The C. minchinii figured is in the gregariniform stage, as
usual, partly enclosed by epithelium. The middle nucleus is
not quite in the transverse plane of the other two; neverthe-
less the three associates have, manifestly, all joined side by
side. Inall cases the nuclei were of equal size and quite
similar in structure; there was nothing to suggest that two
were formed by one of the original ones dividing.

I have also found a couple of instances of triple association
in Diplocystis Schneideri (figs. 24 and 25). In fig. 25
two members of the triplet are slightly smaller than the third,
but in fig, 24 all three are about the same size.! Similarly in
C. irregularis the three members are sometimes apparently
equal (fig. 20) and sometimes not (fig. 29), but the inequality
in the latter case is never pronounced. In C. minchinil, on
the other hand, the members of all the triplets I observed were
all quite equal. ‘This equality is, indeed, only what one might
expect, since it may be said that the individuals cannot live
alone long enough to become appreciably unequal in size.

(5) Minure Srrvcrurs.

(a) Nature of the peripheral region and composi-
tion of the septal plane.

The minute structure of the peripheral part of the body

1 Kunstler (loc. cit., fig. 3) has drawn a triple association, and rather
insists upon the inequality of the members constituting it, regarding the two
smaller as resulting from the division of one large’one. In my fig. 24, how-
ever, scarcely any inequality is noticeable.

LLFE-OYCLE OF “ CYSTOBIA”’ IRREGULARIS (MINCH.). 25

in Cystobia is fairly simple, and practically similar in the
case of both the species I examined. The body is always
limited by a very delicate membrane or cuticle, in which I
could never see any striations. When, as sometimes happened,
the fixative employed has caused the cytoplasm to shrink
away slightly from the membrane, the latter is easily seen
(fig. 12, J.m.).

The parasites do not show any obvious differentiation of
the cytoplasm into ecto- and endoplasm. I could never
assure myself, by any method of fixation, of a well-marked
peripheral layer constituting a definite ectoplasm. The some-
what clearer border round the Gregarine depicted in fig. 2 is
simply due to the lessening thickness of the body approaching
the edge (the specimen, it should be remembered, was drawn
when living). Correlated with this absence of any ectoplasmic
supporting layer to the body is the extreme delicacy of the
latter, and the readiness with which I found it became
irregular or distorted in shape if touched or placed in an
uncongenial environment. In two instances, I should add,
in sections of a couple of Gregarines which were fixed with
Flemming and stained with iron-hematoxylin and orange, a
more finely granular, compacter structure of the cytoplasm
can be discerned near the margin (fig. 37 b) denoting a slight
alteration in its character, but hardly amounting to a distinct
layer. I only observed it in these two examples of C. irregu-
laris, and never in C. minchinii. A strong contrast is
afforded by the distinctness of the ectoplasm in Diplocystis
schneideri, where it forms a thick, well-marked layer,
finely granular in character (fig. 27).

A word or two with regard to the composition of the septal
plane. The septum is constituted by an extremely thin and
delicate membrane, which is, however, remarkably persistent.
This runs straight to the limiting membrane at the periphery
(figs. 37, 39), with which, indeed, it corresponds. There is no
sign of anything in the nature of ectoplasm in the dividing
partition, and, moreover, where the marginal cytoplasm shows
the slight alteration just mentioned, this is not continued into

26 H. M. WOODCOCK.

the plane of junction, which would be the case were it a true
ectoplasm (cf. D. schneideri).

Loss of the power of movement.—Even less is there
anything corresponding toa layer of myocyte fibrille to be
noticed in Cystobia!; and the same appears to be equally
true of Diplocystis. The loss of mobility which is exhibited
by these coelomic parasites is undoubtedly correlated with
their more confined situation, since they tend to remain in
close relation with their host’s tissue, instead of early becoming
“free” in the lumen of the gut, as is the case with most
motile Gregarines (compare, e.g., C. minchinii and a
Coccidian).

Consideration of ‘‘Diplocystis schneideri.’’—
After careful examination, I have to differ in many points
from Kunstler’s interpretation of the structure of the peri-
pheral region and the nature of the septum in this parasite.”
In fig. 27 I have drawn on a large scale a small portion of
the peripheral region of two syzygies, showing the com-
mencement of the plane of junction in each. Passing from
without inwards, there is first a general investing membrane
(v.m.) secreted by the parasite. If this represented a serous
sac, it would certainly show nuclei or other traces of cellular
structure, of which, however, there is not the least sign. It

1 The fact that this lack of muscle-fibrils is accompanied by a complete
absence of mobility in these Gregarincs is a strong confirmation of Crawley’s
theory (9), that, “in general, throughout the Sporozoa, the possession of
muscle-fibres and the power of moving from place to place go hand in hand,
while those forms which are not known to move lack muscular elements,”
which thus attributes actual progression to the myocyte fibrille rather than to
the extrusion of gelatinous threads posteriorly.

2 Kunstler (loc. cit.) regarded the common or investing membrane as
consisting of two layers, the outer being derived from the host, the inner
corresponding to a cuticle or limiting membrane. Internal to this came the
ordinary ectoplasm, which alone (in his opinion) formed the partition between
the two ‘‘halves.” For Kunstler considered that each couple represented a
single individual in process of division, the septum being a transverse continu-
ation, through the endoplasm, of the peripheral ectoplasm, There can be little
doubt that this interpretation was entirely erroneous,

LIFE-CYCLE OF “‘ CYSTOBIA”? IRREGULARIS (MINCH.). 27

is only necessary to compare my figs. 5, 29, etc., showing a
genuine peritoneal investment enclosing C. irregularis, to
seo the difference. The peritoneal sac by which young
Diplocystis-couples appear to be first attached to the gut-
wall (see Kunstler’s figs. 18 and 19) undoubtedly breaks and
liberates the parasites into the ccelome.

The investing membrane.—The investing membrane
is homogeneous and, for the most part, fairly thin but firm
and tough. Around the greater part of the body it closely
surrounds the next internal layer, which is a delicate limiting
membrane (/.m.), quite comparable to that of Cystobia.
About the plane of junction, however, where the regular
contour is broken by the V-shaped groove, the investing
membrane or ectocyst leaves the body proper, as is shown
in the figures. Its middle portion just here is invariably
thickened and usually of a triangular shape. Between the
ectocyst and the limiting membrane there is a space (sp.) of
varying size. This I regard as being due to the contraction
of the investing membrane. In the fresh condition the
thickened ectocyst probably fills up this space and is in con-
tact with the limiting membrane, and this part is represented
in the section by the triangular tongue (t), the shrinkage
having been doubtless caused by fixation.

In other words, the investing membrane in D. schneideri
is evidently secreted at all points where the limiting-mem-
brane and ectoplasm remain free after association of the
parasites. This thickening of the ectocyst in the plane of
junction must considerably strengthen the union of the
couple, just at the point where it is most required, and thus
minimise the risk of dissociation. So there is a slight dis-
tinction between the formation of the ectocyst here and that
of the membranes surrounding an ordinary syzygy when it
becomes encysted. While, in the latter case, the ectocyst is
only secreted by the posterior ends of the Gregarines and
forms an approximately spherical cyst (owing to the rotation
of the animals), here it is secreted at all points of the body
equally, and is laid down in the form of the couple. It

28 H. M. WOODCOCK.

probably undergoes slight alterations in shape as the asso-
ciates increase in age and size, but I feel certain that in life
it remains contiguous to the body throughout, at any rate,
the trophic period. For, although the bilobed appearance may
become very slight (cf. fig. 21), yet in section the V-shaped
inturning of the limiting-membrane and ectoplasm is always
noticeable to a greater or less extent, and there is always this
thickening of the investing membrane.

The limiting membrane cannot be traced through the
septum as a distinct layer, having apparently fused with the
ectoplasm to form a single homogeneous partition ; it certainly,
however, has entered the septal plane (compare my figures
with Kunstler’s fig. 8). Where it bends inwards, and also fora
short distance after the junction, the ectoplasm of each asso-
ciate is frequently somewhat thick and loose in character ;
more internally, however, throughout the greater part of the
septum, it is narrow and dense.

(b) General cytology.

The cytoplasm.—The general nature of the cytoplasm in
Cystobia is well seen in figs. 12, 18, and 19, Pl. 2, the first
example being viewed whole, but more or less in optical
section, and the others being actual sections. It consists of
innumerable numbers of paraglycogen spherules of various
sizes imbedded in a semi-fluid matrix, often with minute,
highly-refringent granules of a different nature in between.
In sections, which were always drawn as they appeared under a
Zeiss apochromatic lens, the appearance varies rather, due
chiefly to the method of fixing and staining employed. After
staining with iron-hematoxylin and orange, one usually getsa
reticular appearance, the meshes being of varying size. This
is shown, for instance, in figs. 37a and 52, and is due to the
fact that the stain has been extracted from the paraglycogen
spherules much more readily than fromthe ground-substance.
Hence the apparent spaces of the reticulum are really occu-
pied by the unstained grains. In fig. 87) the reticulum is

LIFE-CYCLE OF “ CYSTOBIA ”” IRREGULARIS (MINCH.). 29

very close and fairly dense, and there are besides numbers
of minute highly refractile granules (7.g.) which have stained
deeply and give the cytoplasm a granular look. On the other
hand, in sections stained with Kleinenberg’s hematoxylin,
followed or not by eosin, the spherules have retained the
stain and stand out distinctly from the cytoplasmic matrix,
which is in these cases only faintly stained (fig. 45, Pl. 5, and
fig. 36, Pl. 4).

Sometimes, however, these paraglycogen grains are very
small and not prominent (figs. 38a and 38 Db, Pl. 2), but there
are besides numerous larger, irregular, more flattened gran-
ules (a.g.) ; these apparently correspond to the lenticular
plates, which Cuénot (loc. cit.) figures, and which constitute
albuminoid reserve material. They are abundant in fig. 38 },
which represents part of the cytoplasm of a sporulating
Cystobia, fixed with corrosive sublimate and acetic, and
stained with thionin and orange. The refringent granules
(r.g.) are also very numerous, especially in fig. 38 a, a simi-
larly stained section. These granules have a purple tinge
owing to their retention of the thionin, the other constituents
having only kept the orange.

In Diplocystis schneideri the paraglycogen spheres
attain a relatively enormous size compared with those of
Cystobia (vide fig. 27b, which shows a portion of the
cytoplasm of the former). Between and around the spheres
is seen the cytoplasmic matrix. In my sections of this Gre-
garine I have not observed any irregular or lenticular grains
of any kind.

The nucleus.—An adequate idea of the nuclei of a
Cystobia is best obtained when they have been treated with
iron-hematoxylin. No other stain demonstrates so well the
fact that there is a definite chromatic reticulum, although
Kleinenberg’s hematoxylin succeeds to a certain extent.
Safranin and thionin, however, while staining the karyosomes
and also the nuclei of the surrounding tissue well, come out so
quickly from the remaining parts of the Cystobia nuclei,
that one might imagine there was nothing more in these save

30 H. M. WOODCOCK.

diffuse nuclear sap. Neither did picro- and para-carmine, in
staining entire adults, seem sufficiently powerful chromatic
stains to reveal the whole structure. They served to give a
general idea of the position, etc., of the nucleus in relation to
the cytoplasm, and of the structure of the karyosome, and
that was all. Now, the chromatin is by no means all confined
to the latter. Whether the parasites have been fixed with
sublimate and acetic or with Flemming, hematoxylin re-
veals a distinct, well-marked (linin) reticulum, impregnated
with chromatin,' the latter occurring sometimes as local
thickenings of the network, and at other times as numerous,
distinct, but small granules and dots. This is shown in
figs. 18 and 19, Pl. 2, also on a larger scale in figs. 385
and 36, Pl. 4.

The nuclear membrane is generally well marked, and, in
perfectly fixed nuclei, of evenly-rounded contour. I have
never seen, in any of my preparations, the least sign of a so-
called “ geflammte Kern.” In one or two instances, e.g.
fig. 36, the membrane appears irregular and shrunk, but this
is entirely due to contraction on fixation. As a rule, it stains
deeply, and probably itself contains chromatin; moreover,
the reticular threads often start from it.

The karyosome (always single) is more or less vacuolated
in structure, and, certainly in some cases, is slung in position
by very distinct threads of the reticulum (figs. 19 and 35)
which appeared to end in it. In other cases this is not so
marked, and the karyosome seems more suspended in the
network—as if, to use a borrowed expression, “it was a
football lodged in. the branches of a tree.’ In all the
trophozoites, however small, which I have examined, the
nucleus has invariably a karyosome of some sort. The size
and number of the vacuoles it contains is very variable, and
to a certain extent dependent upon the age of the nucleus ;

1 Also in the case of Lankesteria ascidig, Siedlecki (loc. cit.) has
pointed out that the nucleus has a distinct chromatic reticulum, in which are
suspended as well some comparatively large grains of chromatin (see his figs.
2 and 3).

LIFE-CYCLE OF ‘‘ CYSTOBIA’’ IRREGULARIS (MINCH.). 31

in very young parasites there are none, and the karyosome
appears homogeneous.

The nucleus of Diplocystis schneideri differs from
that of D. minor (and also of Cystobia) in possessing,
usually, several karyosomes (figs. 22 and 26) of various sizes
and of the usual vacuolated structure. The nucleoplasm
here, as in Cystobia, is in the form of a chromatic network,
not, however, so deeply staining. Sometimes it does not
appear to be of the same character throughout, a certain por-
tion, in which all the karyosomes are imbedded, being denser,
more granular, and with a stronger affinity for the chromatic
stain (2, fig. 26).

(c) Formation and probable function of the
karyosome.

In the tiny specimens seen in figs. 30-32 the karyosome is
evidently in process of formation. In all three cases it is in
contact at one side with the nuclear membrane, and much
paler in colour than the chromatin, which is here in the form
of deeply-staining grains and lumps in the nucleoplasm.
Hence it is most likely that at this time it consists only of
the plastinoid portion (of extra-nuclear origin ?) forming a
kind of basis, to which will be added later some of the chro-
matin fragments, the remainder contributing to the reticulum.
In fig. 80 each karyosome is drawn out and elongated in
shape, and projects inwards into the midst of the cluster of
chromatin grains, many of which probably come into relation
with it. As the karyosome increases in size, fluid (?) vacuoles
are formed in it, whose contents stain slightly, and these are,
as it were, imbedded in the less fluid, darkly stained, chro-
matic ground-substance. Different degrees in this vacuo-
lisation are seen in the nuclei figured in fig. 33, Pl. 2, which
were drawn whole from a preparation stained with picro-
carmine. .

1 A somewhat similar differentiation of the nucleoplasm, in the form of a
zone or halo around the karyosome, is described by Berndt (loc. cit.) in the
nucleus of Gregarina cuneata,

82 H. M. WOODCOCK.

Expulsion of karyosomatic material into the
nucleoplasm.—Some of the smaller vacuoles probably run
together and unite to form larger ones, for one often gets two
or three large ones—sometimes one very huge one—and a
number of little ones besides (fig. 33, bandc). Fig. 54 is a
section of the nucleus seen in Fig. 33 ¢ (the Gregarine con-
taining it having been unmounted and cut’), and it shows an
occurrence by no means infrequent. In the karyosome lies a
huge vacuole, which is obviously just ready to have its con-
tents expelled into the surrounding nucleoplasm, either by
diffusion through or by the actual rupture of the denser,
more deeply staining, portion, which is here very thin near
the surface. (I should add that, in some instances, the edge
or border of the karyosome stains deeper, and appears as a
thin, dark line, perhaps constituting a definite wall.) I
certainly consider there is an actual discharge, in some such
manner, of the contents of these huge vacuoles, for in the
karyosomes of ripe trophozoites about to commence sporula-
tion the vacuolisation is much more uniform. Fig. 36 shows
the nucleus of an encysted C. minchinii (a full-grown
sporont attached to the muscle), and the vacuoles in the
karyosome are all comparatively small and of about equal size.

A similar elimination of karyosomatic material, though by
a rather different process, is described by Cuénot (loc. cit.)
in the nucleus of Diplocystis. The karyosome there buds
off little portions of itself, each containing a vacuole, into the
surrounding nucleoplasm,? where they become eventually
dissolved. In Cystobia, on the other hand, the karyosome

1 The nucleus is somewhat flattened, owing to the original preparation
having been compressed by the cover-slip, and the nuclear contents are all at
one side, around the karyosome; the karyosome itself is, however, perfectly
normal,

2 The same process, in one form or another, is of frequent occurrence
outside Gregarines. Ina Coccidian of the cuttle-fish, Eucoccidium eberthi,
Siedlecki (87) describes a budding of tle karyosome prior to the formation of
the gametes. As many as twenty secondary or daughter-karyosomes are thus
set free, many of which at length dissolve in the nuclear-sap. An analogous
budding of the “nucleoli” of eggs also often takes place.

LIFE-CYCLE OF “CYSTOBIA”’ IRREGULARIS (MINCH.). 33

never loses its spherical contour, and the material to be
expelled is either diffused out at the pomt where the vacuole
is nearest the surface, or else squeezed out, as it were, through
a minute rupture in the thin wall.

Significance of the process.—Opinions differ as to the
meaning of the process. Some authors, e.g. Cuénot, attri-
bute to it an excretory function, holding that the vacuole
contains waste material (“un produit de déchet”); others,
including Siedlecki, maintain that the karyosome, which they
consider to be a storehouse of reserve chromatin, is giving
back by this means some of its chromatin to the nucleus in
readiness for division. Bearing in mind the essential difference
between true nucleoli or plasmosomes on the one hand, and
karyosomes, where chromatin is intimately bound up with the
ground-substance, on the other hand, it seems to me that
the latter view has much in its favour.

I do not mean to imply, of course, that there is never
anything in the nature of elimination from the karyosome.
Speaking generally, it may be said that, where some of
the contents of the karyosome are passed out into the
cytoplasm and there become altered, and either expelled
or re-absorbed (probably in certain cases being of use to
the formative cytoplasm), we have to deal with such a
removal of unrequired chromatic material. Instances of
this process are seen in Monocystis and Diplocystis,
described and figured by Cuénot (loc. cit.), in Lankes-
teria ascidie, according to Siedlecki (loc. cit.), and again,
in the case of many eggs, where the expelled grains or
spherules can be traced right to the periphery of the cyto-
plasm.

On the other hand, where a portion or all of the karyoso-
matic material becomes ultimately incorporated with the rest
of the nuclear material, whether by direct dissolution or by
fragmentation, it is much more probable that we have to do
with a reinforcement of the chromatin of the nucleoplasm.
This is almost certainly the case when the dissolution is
followed by an increase in the general chromaticity of the

VOL. DU, PART 1.—NEW SERIES, 3

34 H. M. WOODCOCK.

nucleus, as Siedlecki found in Eucoccidium,! and as also
occurs in Cystobia. In the nucleus of the ripe sporont
shown in fig. 36, for example, notwithstanding its large size,
the chromatic reticulum is, if anything, denser and more
marked than in the case of younger nuclei, which still have
large vacuoles in their karyosomes (cf. figs. 19, 34, and 35).

In conclusion, I regard the contents of the vacuoles in the
karyosome of Cystobia as also containing chromatin, but in
a more liquid or “ storage” form, with, at present, no affinity
for chromatic stains. Only the chromatin united with the
plastinoid basis stains up, and the liquid spherules are to be
regarded as being imbedded in this matrix or ground-substance.
There is not the least evidence in favour of the excretory
nature of this vacuolar expulsion in Cystobia; in none of
my sections of adult trophozoites with uniformly vacuolated
karyosomes is there any sign of chromatoid grains or spherules,
either in the nucleus or in the cytoplasm. Moreover, the
subsequent history of the karyosome supports the view I have
taken here.

(6) CoMMENCEMENT OF SPORULATION.

(a2) Encystment.

Encystment in Cystobia is much simpler than in most
Gregarines. In fact, it is often difficult to speak of any real
encystment at all.

C. irregularis.—lIt is no unusual thing for sporulation in
C.irregularis to begin before the animal has evaginated
the wall of the blood-vessel, and while it still has the adult
ovoid form. The Gregarine drawn in Minchin’s fig. 2 (loc.
cit.), which remains, unfortunately, the earliest stage with
more than two nuclei that I have seen, was still in the blood-

1 Cuénot, again, describes a sudden appearance of intensely staining
chromatic grains (“ chromosomes ”’) in the nucleus of Monocystis, after a
dissolution of some of the karyosomes had taken place in the nucleoplasm.
Although this author does not appear to have seen anything significant therein,
I should say the two facts stand in close relation to each other.

LIFE-CYOLE OF “‘ CYSYOBIA”” IRREGULARIS (MINCH.). 35

vessel and had eight nuclei. I have also obtained several
examples from this situation which possessed numerous
nuclei (figs. 39 and 40) ; in the latter case nuclear multipli-
cation was already far advanced. In all these instances the
partition separating the two associates is persistent, and the
animals had not, so far, made the slightest attempt to encyst.

Taking the majority of cases, where the Gregarines have
become evaginated and spherical, they are surrounded, of
course, by the peritoneal epithelium, which shuts them off
from the ccelome and serves, indeed, as the outer wall of the
cyst. The parasites are bounded internally by a fairly thin
membrane, corresponding to the originally limiting membrane,
which can now be spoken of as an endocyst (en., figs. 41, 44).
I doubt whether any other membrane, equivalent to an ecto-
cyst such as we find in D. schneideri, is secreted as a rule
in C. irregularis; at all events, I have only rarely seen
anything resembling one. A thick, protective cyst-wall is
not necessary, the evaginated wall of the blood-vessel serving
equally well for the purpose of enclosing the developing
sporoblasts. In rare instances, however, when the parasites
are in the swollen evaginations described below, and not
closely surrounded by the peritoneal epithelium, the cyst-
membrane does appear to consist of two parts, there being a
pale, homogeneous layer outside the endocyst, which perhaps
corresponds to an ectocyst (ect., fig. 44a and b.)' It is not
nearly so well marked, however, as in C. minchinii, and in
typical cysts attached to the blood-vessels there is no sign of
such a layer.

In sections of sporulating cysts which only contain, as yet,
numbers of nuclei scattered throughout the cytoplasm (fig. 41),
there is, between the endocyst and the ccelomic epithelium, a

1 Gregarines in which the cytoplasm is becoming segregated (i. e. the out-
lines of the sporoblasts becoming visible) are rather liable to shrinkage,
especially when in this unusual position in the membrane anteriorly. In these
cases the cyst-membrane is often folded on itself in places (fig. 444), the
folds (f) appearing somewhat like strings of attachment under a low power ;
the membrane has been unable to shrink equally with the cytoplasm as the
latter became retracted.

36 H. M. WOODCOCK.

rather fibrous layer, staining with the plasma stain (f. L.,
figs. 41, 44c). This is not in any way comparable to an
ectocyst, but is, on the contrary, a layer of the spongy tissue
of the blood-vessel, somewhat altered in character, being
drawn out and tending to disappear. Later on, when the cyst
is full of sporoblasts and spores, it is no longer recognizable.
By this time the enclosing wall is very thin; it consists only
of the peritoneal covering outside, and, next internally, the
endocyst, which has now become extremely delicate and diffi-
cult to make out where it is applied to the epithelial layer.
Sometimes, however, the endocyst has shrunk slightly away
from the latter, and it is then seen more readily (fig. 44 d) ;
at other times, again, especially in ripe cysts, it has quite
broken down and vanished. Indeed, in many of my sections
through spore-containing cysts, however carefully cut, the
delicate peritoneal wall itself is ruptured in places.

The cyst shown in fig. 42 was in an evaginated blood-
capillary in the membrane already mentioned (p. 12) as run-
ning from the ring-canals to the body-wall, and conveying
the radial vessels. It appears to have very thick walls, but
this is due to the fact that these evaginations are swollen and
expanded and partly filled with vascular fluid ( jl.) more or
less coagulated by fixation. ‘There is no question whatever
of this being a thick, gelatinous ectocyst, for scattered here
and there are seen amcebocytes and blood-corpuscles (#., @.).
The wall itself (w.) is very faintly stained, and consists of the
usual loose tissue, here excessively spongy and having an ill-
defined limit internally ; indeed, it is difficult to say in places
where the wall ends and the cavity begins. ‘The Gregarines
in this position are much freer than those in the ordinary
evaginated cysts, which are closely invested by the fibrous
layer above described; and this fact probably accounts for the
development of an ectocyst, albeit only slight, in such cases.

C. minchinii.—In encysted adults of C. minchinii
attached to the muscles there is, as yet, no sign of any
cyst-wall apart from the delicate limiting membrane. ‘This
is usually difficult to make out owing to the thick nuclear

LIFE-CYCLE OF “ CYSTOBIA”’ IRREGULARIS (MINCH.). 37

ageregation (a., fig. 16) surrounding the parasite; it is seen,
however, in fig. 38 a (/. m.) between the cytoplasm and the
layer of nuclei. On the other hand, when the cyst is sporu-
lating, I find a much more distinct ectocyst than ever occurs
in C. irregularis.. External to the endocyst, which corres-
ponds, in this case also, to the original limiting membrane, is
a pale but firm and homogeneous-looking layer (ect., fig. 44 e) ;
this represents a true outer cyst-membrane. Outside, again,
is the nuclear layer, which has now become rather broken
down.

The spore-containing cyst, which is viewed whole in fig. 43,
is attached to one of the vascular strands mentioned above
(p. 14); it is surrounded by an enormous number of cells
which have migrated to the locality. Hp. is the coelomic
epithelium attaching it to the strand (st.), and c.a. is the
cellular aggregation around the parasite, the nuclei being
closely packed just outside the cyst-wall (c.w.). Whether
these are phagocytic cells or not, they do not appear to have
done any injury to the Gregarine, as the contents of the cyst
are quite normal,

(b) Fragmentation of the karyosome.

Owing to the difficulty of obtaining material at the time
of the year when sporulation generally begins and the rela-
tive scarcity of ripe trophozoites (sporonts) in the material I
examined, I have, unfortunately, considerably fewer stages
showing the nuclear changes at the commencement of multi-
plication than I should have liked.' Still, thanks to a few
fortunate preparations, which I have not the least reason to
consider as otherwise than perfectly normal, and which fit in
quite well with each other, I have been able to obtain a fairly

1 Many recent writers have commented upon their inability to obtain the
earliest. phases in nuclear division (compare Brasil [3], Cuénot [10], Léger
[22], and Paehler [29]); it will be readily understood, however, that the
difficulties are enhanced in working with marine hosts, and especially where
it is impossible to follow the process in vivo.

38 H. M. WOODCOCK.

connected idea of the earlier processes in nuclear division in
C.irregularis; as to those in C. minchinii, however, I
am without any information.

In fig. 45a is seen the earliest stage I obtained in this
nuclear preparation for sporulation, and one which is most
important. Only one nucleus is shown in this section, the
other being further on; a section of it is drawn separately
at b. Both nuclei are in exactly the same condition. Their
outline is slightly retracted and irregular, attributable to the
fact that the animal was fixed in a piece of the vascular net-
work, being still inside the lumen of the vessel.! There is
no septum visible in this specimen, and it is evidently an
instance of precocious and complete union.

The important point to notice is that the large, uniformly
vacuolated karyosome, as we saw it, for example, in fig. 36, is
no longer present as such. It is represented instead by the
numerous small fragments of slightly varying size seen at /. ;
they are fairly well stained, more deeply so at the periphery,
and stand out distinctly from the rest of the nucleus. The
karyosome has undoubtedly broken up or separated into these
little, more or less spherical, pieces, each of which probably
corresponds to one of the spherules of the original karyosome.?
The nucleoplasm itself is distinctly chromatic, but not so
obviously reticular, being of a more granular nature, with
the granules of practically uniform size. There is not the
slightest appearance of any expulsion of karyosomatic material
into the surrounding cytoplasm, and the nuclear membrane is
perfectly entire.

Further stages in the history of these karyosomatic frag-
ments are shown by the nuclei drawn in fig. 46. These are
daughter-nuclei of the fourth generation, all belonging to the

1 The relatively large size of the nucleus, as compared with the apparent
length of the body, is due to the fact that the sections passed obliquely
through the Gregarine.

2 Compare Gregarina blattarum, whose karyosome forms a chain
(“‘chapelet ”) of numerous tiny ones by successive buddings or divisions (see
Cuénot loc. cit., fig. 33).

LIFE-CYCLE OF ‘‘ CYSTOBIA””’ IRREGULARIS (MINCH.). 39

same Gregarine. Different degrees in the process of incor-
poration are exhibited by different nuclei. I have arranged
them in a series. Each of the daughter-nuclei seen at a
possesses one little karyosome, corresponding, in all proba-
bility, with one of the spherules of the original karyosome,
which became divided up in the parent-nucleus as seen in
fig. 45. With successive divisions of the nucleus, these
daughter-karyosomes have been passed on and, as it were,
apportioned out among the daughter-nuclei, till by this time
many of the latter have only one, or the fragments resulting
from one. The periphery of each stains deeply, and the
interior part also stains rather more than hitherto (compare,
for example, fig. 45), as if the chromatin were becoming pre-
cipitated or re-converted into the customary staining-form.

Nuclear incorporation of the fragments.—These
little karyosomes next divide up and gradually become in-
distinguishable from ordinary chromatic grains. In fig. 48 b
the first stage in this division is shown; each of the daughter
karyosomes has divided into two of about half the size.
Further division takes place more or less irregularly, as in ¢,
followed by d and then e. Sometimes the fragments may
remain together after division and form a ring or chain of
grains or rodlets, as in f and g (compare again the nucleus
of G. blattarum). As the process goes on the grains tend to
stain up more homogeneously with the chromatic stain. From
either e or g—in both of which the nucleoplasm, as a whole,
is becoming chromatically denser—it is but a slight step, on
the one hand, to fig. 47, or on the other to fig. 46. In the
former are drawn two nuclei from different multinuclear
sporonts in about the same phase, a being taken from one
still in the vessel and b from one that was evaginated.! Some
of the chromatic grains, particularly in a, are large and
prominent.

1 It will be noticed that, although both @ and @ are nuelei of the fourth
generation, @ is distinctly the larger ; there is, of course, a certain variation in
the size of different sporulating individuals, which is manifest also in their
nuclei, karyosomes, etc.

40 H. M. WOODCOCK.

There can be little doubt that the above series represents
what has been gone through in the nuclei shown in fig. 46.
These are two nuclei of the eight-nuclear stage to which
reference has already been made; that is to say, they are
nuclei of the second generation only of daughter-nuclei.
They are very similar in constitution to those of fig. 47. The
chromatin grains on the reticulum are not all of equal size;
some are larger than others, and many of these are probably
directly derived from the fragments of the karyosome. The
process in this case, therefore, has evidently been more rapid
and has taken place earlier, the fragmentation of the daughter
karyosomes perhaps commencing before the parent-nucleus has
divided. Hence there is a certain variability in the time
elapsing between the first break-up of the karyosome and its
final incorporation with the rest of the nuclear material as a
part of the chromatic reticulum."

(7) Nuctzar MULrIeLication.

(2) Nature of the early nuclear divisons.

With regard to the manner in which the first nuclear
divisions take place, I have not the least doubt that they are
purely direct or amitotic. Unfortunately, I have no prepara-
tions showing the earliest ones, but in a section through a
sporont containing about thirty daughter-nuclei of the fourth
generation there is still one of the nuclei of the third genera-
tion dividing directly into the two of the fourth, in a manner
that is perfectly unmistakable. It is shown in fig. 47 c, and
as to its absolutely amitotic character there can be no
question ; the nucleus has become constricted in the middle
and the two halves are now being cut off from each other,
half the chromatic grains going to one portion and half to
the other. There is not the slightest sign of any attraction-

1 A similar fragmentation of the karyosome with subsequent dissolution in
the nucleoplasm, followed by increased chromaticity of the latter, is also

described by Berndt (loc. cit.) for Gregarina cuneata, and by Caullery
and Mesnil (5) for Selenidium sp.

LIFE-OYCLE OF “ CYSTOBIA”’ IRREGULARIS (MINCH.). 41

spheres or nuclear spindle, and the former are equally absent
from the resting nuclei in other parts of the preparation.
Neither is there any indication of karyokinetic apparatus
in the eight-nuclear stage, and as division-centres are very
easy to see later on, it is not likely that I have overlooked
them here. Hence, I maintain there is a very great
probability that the preceding divisions are amitotic also.
Another strong point in favour of this view is the fact that,
when asters and nuclear spindle appear at first, and the
division of the “segmentation”’-nucleus! is mitotic, that of
the resulting daughter-nuclei is so too, and the division-
centres in connection with them are very apparent (Mono-
cystis, Diplocystis, Lankesteria, see Cuénot, Siedlecki,
Prowazek [loc. cit.]). I think this fact and the obviously
direct division of a daughter-nucleus of the third generation
(fig. 47 c.) are sufficient to warrant my saying that the first
nuclear divisions in C. irregularis are completely ami-
totic.”

We may now take a bird’s-eye view of one or two sporu-
lating Gregarines at about this period, in order not to forget
the whole in considering a part. Figs. 59 and 40 show two
such couples, both from the blood-vessel, and neither in the
least indicating, by cyst-like shape or the possession of cyst-
membranes, any appearance of encystment. In both Gre-
garines the persistent septum entirely separates the nuclei
derived from the parent-nucleus of each associate. Fig. 39
is a section and so only few nuclei are seen ; altogether there
are about thirty. Fig. 40 is drawn whole from a stained
preparation, and in this the multinuclear condition is well

1 That is, the functional portion of the original sporont-nucleus remaining
after the unnecessary constituents have been expelled.

2 1t would seem, moreover, as if this fact stood in some relation with the re-
tention of the karyosomes in the nucleus. For, in G. cuneata and in Seleni-
dium, the two instances above referred to as resembling Cystobia in the
latter respect, the first nuclear divisions are equally amitotic, although differing

somewhat from those in C. irregularis in being of the ‘ multiple” type
jnstead of by simple binary fission.

42 H. M. WOODCOCK.

advanced; the septum is also still present. In figs. 41 and 42,
on the other hand, more of an encystment is recognisable,
the exact character of which has been described above. The
former is a section showing many nuclei, and no septum is
visible! Already it can be seen that the nuclei are of two
kinds, some (NV) being much larger than the others (n). In
fig. 42 this difference in size has become accentuated.

(b) Distinction of the multiplying nuclei into two
classes, somatic and germinal, and their
further history.

In fig. 49 are drawn some of the nuclei from sections of
another Gregarine in a condition very similar to that in fig.
41. They show the earliest stages in this differentiation of
the nuclei into two kinds. <A considerable change has
occurred in the nuclear constitution since we last saw it in
fig. 47. Some of the nuclei (the larger ones) have divided
less often than others, and are of an irregularly ovoid to
rounded shape; the smaller ones are more uniformly ovoid.
In the large nuclei there is, in place ofa dense, finely meshed,
chromatic network, either a very loose reticulum (a) or none
at all; in the smaller, the chromatin is situated peripherally,
and is in the form of fairly even-sized rodlets or grains in
contact with the nuclear membrane. In all cases the rest of
the nucleus is filled with a non-staining nuclear sap. ‘The
chromatin in the larger nuclei is in the form of irregular,
deeply-staining masses or lumps, with sometimes fine granules
or chromatic dust besides, unevenly distributed throughout
the extensive nuclear sap.

These two sorts of nuclei represent the division of the
nuclear material into a somatic portion, which eventually
breaks down and is absorbed by the cytoplasm, and a func-
tional or germinal portion which alone will form the nuclei

1 Tt may either have disappeared or never have been formed; in sections

through another cyst ata similar stage there are traces of one, not so distinct,
however, as that in fig. 42.

LIFE-CYCLE OF “* CYSTOBIA”? IRREGULARIS (MINCH.). 48

of the gametes (sporoblasts).!. Whether there has been a
qualitative as well as a quantitative separation of nuclear
material, and how the selection has been brought about in the
absence of karyokinesis, I am quite unable to say.

The subsequent course of events in the two kinds is very
different. Fig. 49a represents the least modified condition
of one of the somatic nuclei; it differs chiefly from a germinal
one in having its chromatin aggregated in larger masses and
disposed more in the interior. At b, c, and d,and in the large
nuclei in fig. 50, are seen further stages in the alteration.
The sterile nuclei do not at present divide any more, but grow
considerably, chiefly by an increase of nuclear sap ; at the same
time, there is an increase of the chromatoid granular matter
(cf. figs. 49d and 50D).

Turning now to the smaller or germinal nuclei, I have no
doubt that division in these is henceforth indirect or mitotic.
Unfortunately, I could not see any divisions actually occur-
ring in these two preparations, as all the nuclei were in a
resting condition. Nevertheless, karyokinetic apparatus is
already quite distinct in connection with many of them, and at
a slightly later stage mitotic nuclear division is unmistakably
in evidence. At one or both ends of certain of the smaller
nuclei, and quite at the periphery,” is seen a little thickening
which gradually assumes the form of a rounded grain. It
stains deeply with iron-hematoxylin, and is probably of
chromatic nature. From each of these bodies a few delicate
rays diverge into the surrounding cytoplasm. The whole
structure is better developed later on, the central grain
especially becoming large and prominent (fig. 51 aand b).

Nature and origin of the division centres.—I have
felt somewhat uncertain whether to regard these large con-

1 It is interesting to note that Léger (22) has also described an early dis-
tinction of the multiplying nuclei in Stylorhynchus into sexual ones, ulti-
mately forming the nuclei of the gametes, and somatic ones, which are left
behind in the “somas”’ or cystal residue.

2 It is difficult to say whether the thickening is at first inside or outside the
membrane. .

44. H. M. WOODCOCK,

spicuous grains as centrospheres! or centrosomes. Notwith-
standing the size to which they attain, I am now? inclined to
consider them as more comparable to true centrosomes than
to archoplasmic structures. They are entirely homogeneous
in appearance, and even when by chance less intensely
stained than usual, in no case can any distinct centriole or
granule be perceived inside them; moreover, they have a
sharply defined outline or limit. They are very different
from the centrospheres in Monocystis, as figured by
Cuénot and Prowazek (loc. cit.); nor do they appear to have
anything in common with the acidophile structures (probably
of archoplasmic nature, corresponding to attraction-spheres)*
which Brasil (3) describes im connection with the nuclei of
Urospora.

On the other hand, the centrosomes in my fig. 51 recall
the prominent grains at the heart of the division-centres in
Diplocystis (Cuénot, loc. cit., fig. 55). Again, in their
deeply-staining property and practically homogeneous ap-
pearance they agree with the centrosomes figured by Léger
(loc. cit.) in nuclear mitoses of Stylorhynchus, and also
with the “ centrioles”? of Urospora (Brasil, loc. cit.), the
chief point of difference being that in both these cases the
size of the granules is less.*

With regard to the origin of the centrosomes in C, irregu-
laris, they certainly arise in intimate connection with the
germinal nuclei, and probably with these only. Hach is
formed just at the periphery of the nucleus, and is itself most
likely of nuclear (chromatic) origin. Probably the centro-

1 IT follow Wilson (89) in considering as centrospliere the inner, central
region of the aster (consisting of the so-called “ medullary ” and “ cortical”
zones), restricting the term “centrosome ” to the grain or granule, situate, when
present, actually at the centre of the astral system (Boveri’s *‘ centriole’’).

2 In my preliminary note (40) I termed them ‘‘centrospheres,”

3 The term “ attraction-sphere ” may be used in a general sense, as including
the whole division-centre.

4 Hertwig (18a), however, has described relatively large centrosomes in
connection with certain nuclei in Actinospherium (vide his figs. 2 and 38,
pl. 5); these show a general resemblance to those of Cystobia,

b)

LIFE-CYCLE OF “ CYSTOBIA”? IRREGULARIS (MINCH.). 45

some, when differentiated, causes the appearance, in the
cytoplasm of the immediate neighbourhood, of the astral
fibres which become united with it to form the complete
karyokinetic apparatus.!

Indirect nuclear division.—In fig. 51 are drawn many
nuclei of both kinds, taken from a section similar to that seen
in fig. 42; the differences between them are now considerable.
A typical sexual nucleus in the resting condition is seen at a;
at b is another, probably resulting from a recent division,
since it has only one attraction-sphere. Dividing stages are,
here again, very rare, this being due, undoubtedly, to the
relatively slow rate at which the life-history normally pro-
gresses in Cystobia. ‘The commencement of division is
shown in c, where the chromatin has left the periphery and
occupies a central position, taking the form of grains or little
masses; some of these perhaps become dissolved in the
nuclear sap, for at d there are only two or three visible. The
next stage is seen at e, where the chromatin is stretched out
in a thick line or band of grains between the centrosomes,
preparatory to breaking up and becoming divided between
the two daughter-nuclei. One of the two halves resulting
from division is seen at f!. The chromatin is in the form of
three or four deeply-staining grains in close association with
the division-centre. ‘The nuclear membrane next reappears,
the chromatin becomes more spread out, and as the nuclear
Sap increases in amount, the usual vesicular appearance is
regained (f*). he daughter-nucleus is now completely re-
constituted.

Mitosis itself is therefore of a comparatively simple nature
in C. irregularis, but whether this condition is primitive
or secondary I am unable to say. ‘The process most resembles
that described by Cuénot as occurring in the nuclei of Diplo-
cystis. I have seen no signs of a well-developed nuclear
spindle, on which are arranged beautiful Y-shaped chromo-

1 Compare Hertwig’s suggestive ideas (18) respecting the constitution and
mutual relationships of the different constituents of living protoplasm.

46 H. M. WOODCOCK.

somes, as described in the case of Monocystis, and also by
Léger for Stylorhynchus and Brasil for Urospora.

Independent division of the centrosomes.—A very
interesting peculiarity of the division centres in C. irregu-
laris must be mentioned—namely their occurrence alone and
quite separate from the nuclei. For it is certain that isolated
centrosomes do at this period occur, and, indeed, in many of
my sections they are not at all uncommon. One of them is
figured at g, and it will be observed that it is quite similar
to those in relation with the sexual nuclei. Equally certain
is it that these isolated centrosomes also divide by themselves;
h, i, and j show this distinctly. Inall cases, whether of single
or dividing ones, I have carefully examined neighbouring
sections of the series, but without finding any trace of a
nucleus in connection with them. This independence of
certain division-centres is an unusual circumstance, and, so
far as I am aware, has not hitherto been described in the case
of Gregarines.!

It may, perhaps, be suggested, Does not the appearance seen
at 7 represent, rather, the final stage in nuclear division, the
two halves remaining attached only by a delicate, drawn-out
thread? If this were so, it would imply that practically all
the chromatin of the two daughter-nuclei had become (tem-
porarily) united with the two centrosomes. This is exactly
what Cuénot describes and figures (loe. cit., fig. 55 d) in the
mitoses of Diplocystis minor; and, asa result, the grain
at the centre of each attraction-sphere (consisting of centro-
some plus the chromatin of a daughter-nucleus) is very greatly
enlarged as compared with the normal size of a centrosome
when associated with a resting-nucleus (fig. 55 a—c). The
centrosomes in my fig. 517, on the contrary, exhibit no per-
ceptible differences, whether of size or staining properties,
from those above described in close connection with the sexual
nuclei. Moreover, the aspect of a daughter-nucleus immediately
after separation (fig. 517’, already described), is quite different.

1 Tt is not an uncommon feature in spermatogenesis, however, for well-
differentiated kinetic centres to divide by themselves.

LIFE-CYCLE OF * CYSTOBIA”’ IRREGULARIS (MINCH.). 47

And, lastly, h and 7% represent, indubitably, earlier stages in
this centrosomic division.

Disintegration of the sterile nuclei. — Turning
again to the large or sterile nuclei, most of these have
increased greatly in size (all the nuclei are drawn to the
same magnification) and have now very large and apparent
centrosomes in connection with them. Of these, there are
generally three or four in relation with each nucleus, but
they are not always equally closely attached. In &, for
instance, while one is touching the nuclear membrane and
appears to have pulled it out more or less, another at the
left side is not actually in contact. Many of these centro-
somes are about double the size of the solitary ones, as if
they had not divided for some time. I am inclined to think
that the attraction-spheres in connection with these large
somatic nuclei do not originate from the latter, in the manner
in which the others are formed at the periphery of the sexual
nuclei, for I have never seen any, as it were, commencing to
be developed with a small centrosomic portion (compare figs.
49 and 50). It is more likely that they correspond to some
of the loose division-centres in the cytoplasm which have
come into relation secondarily with the large nuclei, in order
to bring about their break-up and ultimate dissemination.

For there is another variety of nuclear figure to be met
with, different aspects: of which are seen in fig. 52; these
represent stages in the further alteration of the sterile nuclei,
leading to their complete resolution into chromatoid grains
and granules, which eventually become scattered throughout
the cytoplasm. ‘The granular chromatin or chromatoid matter
becomes more abundant and dispersed through the nuclear
sap, which is thus divided up into a number of vacuole-like
meshes often containing one or more chromatic lumps. These
vacuolar meshes (which somewhat resemble a number of vesi-
cular nuclei closely aggregated) are at first variable or large in
size (aand part of b), but become subdivided into smaller and
more uniform ones (c), until they are gradually obliterated,
leaving only the granular chromatic material, as in the other

48 H. M. WOODCOCK.

part of b; from this to d is a slight transition. In all these
cases there is no longer any sign of kinetic apparatus and
the nuclear membrane has also quite vanished. A portion of
the cytoplasm around each has been drawn as well, and, while
in 6 and ¢ the delimitation of the nuclear material is still fairly
well marked, in d there is no distinct boundary between the
two constituents, the granules at the periphery spreading out
into the cytoplasm.!

This separation of the nuclei into two kinds represents the
only nuclear purification, or elimination of unnecessary chro-
matic material, which occurs in Cystobia irregularis.
The germinal nuclei, after one or two more divisions, become
the nuclei of the primary sporoblasts, the formation of which
we have now to consider.

(8) ForMATION OF GAMETES (SPOROBLASTS) AND CONJUGATION.

Fig. 56 is-a section through a cyst of C. irregularis in
which the sporoblast nuclei (7) are all formed. The nuclei
have now a certain definite arrangement, being situated at
or near the periphery of irregularly shaped lobes or portions
of the cytoplasm, which, indeed, they mark out and delimit.
In other words, we find in Cystobia the same intertwining
and interlocking of the two cytoplasms (one belonging to
each half or member of the “couple”) that Siedlecki describes
for Lankesteria, and Cuénot for Monocystis. In Cys-
tobia, however, especially in C. minchinii, the process
is carried to a much greater extent, which is easily to be
understood when the far more intimate nature of association
in this genus is remembered.

Seen in section,’ this mutual interlobing is here of such a

1 The general cytoplasm up till now has been entirely free from chroma-
toid or nuclear matter, not staining with iron hematoxylin; in later stages,
however, this is no longer the case, chromatoid granules being very much in
evidence and persisting more or less conspicuously throughout sporoblast- and
spore-formation.

2 Unfortunately, nothing whatever is visible in an advanced cyst stained and
mounted whole until the definitive sporoblasts (zygotes) or spores are formed.

LIFE-CYCLE OF ‘* CYSTOBIA”” IRREGULARIS (MINCH.). 49

complicated character that it is quite impossible to say to
which half any given lobe or portion belongs. Asa rulea
very narrow space separates any two adjoining segments, but
sometimes this is not easy to make out; in these cases the
nuclei help to define their limits. ‘The segments are very well
marked out by this means in fig. 55, which is a section through
such a stage in C. minchinii (one of the very few sporulating
cysts of this species that I have come across). Here the
nuclei are unusually numerous and close together, and nearly
everywhere in paired rows, each row belonging to a different
lobe and, in all probability, to a different associate. In
fig. 44. a small region at the periphery of the cyst is seen
highly magnified, and shows a portion of two lobes with the
narrow space (sp.) between. All along the margin of each
segment are arranged the sexual (sporoblast-) nuclei.!
Segregation of the sporoblasts.—The later develop-
ment of the sporoblasts follows a slightly different course
from that generally described in the case of other Gregarines.
When the lobes and processes are fully developed and com-
pletely intermingled the sporoblast-nuclei throughout the
cyst take up a position of equal distribution, each being
surrounded by about the same area of cytoplasm. In a low-
power view (fig. 53a) they appear, in any section, evenly
arranged throughout the mass and no longer define the limits
of the segments, the outlines of which have ceased to be dis-
tinct. A portion of the cytoplasm next becomes segregated
around each nucleus, simply by the appearance of narrow,
irregular spaces which sufficiently divide up the whole.
Practically the entire cytoplasm is thus used up*; in other
words, there is no large cystal residue or “ gregarinoid
soma”’ left over in the cyst. I can say this with absolute

1 Neither this section nor that of C. irregularis in fig. 56, both of which
were stained with thionin and orange, shows the chromatoid granules in the
cytoplasm, as thionin is soon extracted from them; for this reason, in fact, L
occasionally used it, since iron-haematoxylin stains them so intensely that the
nuclei themselves tend to be obscured.

2 See below, p. 54, for instances of what are, probably, abnormal exceptions.

VoL. 50, part 1.—NEW SERIES. 4

50 H. M. WOODCOCK.

certainty in the case of C.irregularis. I have examined
serial sections through many cysts, but whether these contain
sporoblasts just about to be liberated, gametes copulating,
zygotes, or spores, they are always uniformly filled with the
same (compare fig. 54). I am strongly of the opinion that
C. minchinii agrees equally in this respect, judging from
sections like that shown in fig. 55 and also from sections
through a spore-containing cyst, in which the spores are
closely packed and completely fill the cyst, there being no
sign of any unused cytoplasm. I lay emphasis upon this
point because of its bearing upon the actual manner in which
copulation is brought about (see below, p. 52).

Fig. 536 shows a small portion of a section! similar to that
seen in fig. 53 a, in which this separation of the sporoblasts is
taking place. At this moment the cytoplasm has a very loose,
irregular, and ill-defined appearance, and it is difficult to
represent exactly; in some places the splitting up is rather
more pronounced than in others.” Besides the fact that the
process of segregation is going on practically uniformly
throughout the entire cyst, there is another, most important,
point to be noticed. All the nuclei are identical in size and
appearance. Brasil (loc. cit.) states that, in both Urospora
and Gonospora, the nuclei belonging to the two halves of
the cyst—to the two associates, that 1s—are very different in
character and easily distinguishable. That is certainly not
the case here. The nuclei in the rest of the section from
which fig. 536 is drawn, and equally in other sections of the
series, appear exactly like those shown.

The gametes.—As soon as the segregated uninuclear
portions are thus cut out and separated from one another
they become rounded off, and each forms a primary sporo-
blast. The primary sporoblasts or gametes in C. irre-
gularis are, so far as I can ascertain, quite equal and

1 This section was stained with iron-hematoxylin, and the chromatoid
granules are very obvious.

a The “cutting-out ” process is, perhaps, slightly more advanced near the
periphery than elsewhere.

LIFE-CYCLE OF ‘‘CYSTOBIA”’ IRREGULARIS (MINCH.). 51

similar; in other words, they are completely isogamous.
The nuclei certainly do not exhibit the distinctions in chro-
matic density, intensity of staining, ete., which Brasil (4)
finds in the case of Monocystis, and which, as he points
out, are, indeed, apparent both in Cuénot’s figs. 22 and 23
and in Prowazek’s figs. 11 and 13 (loc. cit.). On the
other hand, judging from the figures of Siedlecki, Cuénot,
and Berndt (loc. cit.) respectively, the gametes in Lankes-
teria, Diplocystis, and Gregarina approximate very
closely to those of Cystobia in their apparent similarity one
with another.

This perfect agreement (morphologically speaking) between
the sporoblasts is best seen in cysts which have been crushed
whole on a slide after staining. In sections through a cyst
at this stage it frequently happens that either a sporoblast
or its nucleus is not bisected, when a casual glance would
give the idea of differences in size and nuclear constitution ;
such an interpretation would be quite false. Primary sporo-
blasts are shown in fig. 57a, b, and c. They are very
simple in structure. Hach has a definite and fairly constant
shape, shghtly ovoid to round. The nucleus is spherical and
usually situated excentrically. There is no sign of any
centrosomic granule in relation with it, such as is figured by
Brasil in the case of Urospora. Practically all the chro-
matin lies at the periphery of the nucleus in the form of many
short rodlets, leaving the nuclear sap free. The cytoplasm
is full of round paraglycogen grains; other granules are not
very evident in preparations stained with para-carmine but
are numerous in iron-hematoxylin ones.

Conjugation.
As soon as the sporoblasts are definitely formed and
isolated they are ready for conjugation.’ Notwithstanding

1 T have never observed the least appearance of anything comparable to the
chromatic reduction which, according to Paehler (29), takes place in the
primary sporoblasts of Gregarina (Clepsydrina) ovata, and which he
suggests may indicate bisexual differentiation. The author could not perceive,

a2 H. M. WOODCOCK.

the apparent equality between the gametes, I regard this and
other cases of complete isogamy in Gregarines as constituting,
really, instances of binary sexuality, and not merely of exo-
gamy (using this term in the sense of Hartog [12]) ; that is
to say, the condition we find here is derived, in all proba-
bility, from a typical bisexual condition, although the binary
sex is, in these cases, rather implied—or, more correctly,
potential—than actually manifest (see also below, p. 76).
It is almost certain that each gamete copulates with one
developed from the other member of the couple—probably
that nearest available. Although I have been unable to
demonstrate this point in Cystobia, for reasons which are
apparent from the context, there cannot be the least doubt
that the statement is justified when the formation and consti-
tution of the zygote in cases of anisogamous conjugation is
borne in mind. ‘The origin of the two conjugates from
separate associates (parent-individuals) appears to be uni-
versal, whether the anisogamy be strongly marked (Ptero-
cephalus) or only slight (Urospora).

Absence of any movement.—With regard to the
mechanism of the process, I feel confident that in this case
there is no bustle, no tumultuous ‘‘ danse des sporoblasts”’ to
be rushed through; thanks to the remarkable extent to which
the intertwining is carried all is done easily and quietly. I
have carefully watched live cysts for long periods, both by
day and by night, but I have never seen the slightest move-
ment taking place. Now, if there had been any animated
movement I should certainly have observed it, at any rate at
the periphery ; and, moreover, in my sections through cysts
of this period all stages in conjugation are to be met with, so
that this was evidently the right time to look for the “dance.”
Hence I feel assured that if, at the moment of separation,
each gamete is not actually contiguous to a suitable fellow-

however, any appreciable difference between either the two associates or the
gametes to which they gave rise; nor does he state whether he observed this
nuclear purification in elements formed from both halves of the cyst, or only
in those from one, which would have helped to decide the question.

LIFE-CYCLE OF “ CYSTOBIA ”’ IRREGULARIS (MINCH.). 53

conjugant, a very slight movement or mutual attraction will
bring two such, when set free, into contact.

There are, besides, very good reasons for the absence of
this peculiar and characteristic phenomenon. In the first
place, since practically all the cytoplasm is brought into use,
there are neither lobes nor processes (as in Stylorhynchus
and Lankesteria), nor a large central residuum formed
by the two gregarinoid somata (as in Diplocystis and
Gregarina cuneata) to be, as it were, circumnavigated,
either actively or passively, before selective fusion can result.
On the other hand, again, there is no apparatus for pro-
ducing the movement.! Hence, considering all things, I
no longer wonder, since examining my fixed and stained
material, that I could not see any movement when observing
cysts at this stage alive, and I do not believe that any such
occurs.

The actual union.—In sections through a cyst in which
conjugation is actually taking place, various stages are to be
met with indiscriminately throughout the cyst—here the two
cytoplasms commencing to fuse, there a single definitive
sporoblast with the two nuclei still separate, and, lastly, a
fully-formed zygote. Hence there is a slight variation in
time of the process to be noticed, probably due to the primary
sporoblasts not all bemg cut out at the same moment. The
actual conjugation processes are, however, better seen in cysts
which have been fixed and stained whole and then crushed
ona slide. If this is done carefully, the exact stage reached
in each case can be readily observed. All degrees in the
union are seen in fig. 57, drawn from such a crushed prepara-
tion; d—h show successive stages in the fusion, and at 7 we
have the definitive sporoblast, copula, or zygote formed. The

1 Léger, in his interesting paper (22), shows that the chief function of the
fusiform motile gametes (sterile male elements) in Stylorhynchus is to bring
about, by their vigorous movements, the ‘‘ mélée sexuelle.” Again, Berndt
(loc. cit.) finds that, in the mealworm Gregarines, the unused cystal ‘ reli-
quats” become ameeboid and send out processes of various lengths, which
serve to drive the peripherally situated gametes round in the cyst and mix
them thoroughly together.

54 H. M. WOODCOCK,

gametes apparently unite by the pole or point farthest away
from the nucleus, and as the fusion becomes complete the two
nuclei gradually approach and, at length, also unite. The
nuclear union is quite simple. Asin the case of other Gre-
garines so far described, there is no sign of a fusion-spindle
such as is often met with in the zygotes of Coccidia (Cocci-
dium, Adelea, etc.).

The copule in fig. 58 were all drawn from sections, in
which it is easier to make out the subsequent behaviour of
the fusion-nucleus. The chromatin of the two constituents
becomes completely intermingled and assumes the form of fine
rodlets ; ultimately it appears more granular, the whole nucleus
meanwhile contracting and becoming denser. While this is
going on, the zygote has surrounded itself with a delicate
membrane, and has begun to change its shape, taking on an
ovoid form (f). The nucleus has also passed to one end,
which we may term the basal end. A section through a
cyst full of zygotes (at a rather earlier stage) is drawn in
fig. 54.

Abnormalities.—Before passing on to consider spore-
formation, there are one or two abnormal and degenerating
appearances to be noticed in the cyst at this time. In fig. 63a
are seen two sporoblasts which have not succeeded in copu-
lating. ‘They are considerably shrunken as compared with
normal ones, the cytoplasm having lost its granules and
become diminished and pale. ‘The nucleus, too, is rather
smaller, and stains up almost homogeneously. Two copule,
which for some reason or other are also unsuccessful, are
seen at b. (Possibly, in this case, for lack of better partners,
sister-sporoblasts have paired.) They exhibit practically the
same structure as the uncopulated ones, only being larger ;
in the right-hand one a few reserve granules are still left.!
Fig. 63 c shows an instance of polygamy; three sporoblasts
have evidently copulated, the nuclei of two having already
fused. Such cases are not rare; there are generally a few to

1 A similar degeneration of certain sporoblasts is described by Cuénot (loc.
cit.) in Monocystis,

LIFE-CYCLE OF “ CYSTOBIA”’ IPREGULARIS (MINCH.). 55

be met with in each cyst.! I do not think these polygamous
copule develop into anything; I have never come across
monstrous or abnormally large sporocysts.

I cannot be quite sure whether the larger mass with many
typical sporoblast nuclei seen in fig. 64a is an example of this
on a large scale,” or whether, on the other hand, it represents
a small portion of the cytoplasm of some lobe or other which
did not entirely segregate up, but remained in this compact
condition. ‘The latter view seems to me more probable, for as
many as twelve nuclei may occur in one of these masses, of
which there are occasionally one or two in a cyst. While in
the former case that would imply a very comprehensive poly-
gamy, in the latter the cytoplasmic residue with all its nuclei
would represent actually only a very insignificant portion of
the entire sporulating cyst. However that may be, a later
stage where such a mass is degenerating is seen in b, drawn
from a spore-containing cyst. ‘he nuclei have coalesced
together, the chromatin has become granular and stains up
densely, and the cytoplasm is also altered and full of
chromatoid matter.

(9) Spore-ForMATION AND SysTEMATIC PosrTIoN.

(a) The spores.

C. irregularis.—The zygote next commences to assume
the typical shape of the spore. While the basal end (where
the nucleus is situated) remains round and unaltered, the
other end, which is destined to form the funnel, becomes
elongated and then narrowed (fig. 59a and b). When it has
thus acquired the definite shape, the clear but firm exospore
or outer spore-membrane is laid down. The cytoplasm then
retracts itself somewhat temporarily, leaving the funnel free,

About this time the first division of the zygote- or spore-

1 Léger (loc. cit.) also figures exactly similar instances in Stylo-

rhynchus,
* Léger mentions such cases where five or six gametes have fused.

56 H. M. WOODCOCK.

nucleus takes place. After having become very finely gran-
ular, the chromatin is aggregated into four or five lumps or
pieces (c). A well-marked spindle then makes its appearance,
stretching transversely or slightly obliquely across the spore.
Hence it is usually best seen when the spore happens to be
cut transversely, as in d—g. At each end of the spindle
is a distinct centrosome, which is, however, minute when
compared with the large ones described above in connection
with the multiplying nuclei. The chromatin fragments (or
chromosomes ? I have not been able to make out any definite
number) are next divided and pass along the spindle-fibres
to either end, where they unite again to form the daughter-
nuclei, the spindle eventually disappearing (f—j). The
division of the two daughter-nuclei to form four is very
similar, and also takes place transversely to the length of
the spore; I have not seen the division of one of these four
nuclei to form one of the eight sporozoite-nuclei.

The cytoplasm possesses a marked affinity for chromatic
stains owing to the chromatoid granules which it contains
(see fig. 59 c—j inclusive). These are at first uniformly
distributed throughout the contents of the spore; gradually,
however, they become united in a mass centrally (c, h, and J),
and form, together with a certain amount of cytoplasm not
used in sporozoite-formation, the spore-residuum or “ reliquat
sporal.” Fig. 60 shows a couple of typical spores at this
binuclear stage. I think the endospore or inner spore
membrane is secreted about this time, as the cytoplasm seems
to fill up the funnel rather more, but at present it is very
delicate and difficult to make out. Later, when the four-
nuclear stage is reached (fig. 61), the cytoplasm has become
finally retracted from the funnel end, leaving the endo-
spore showing as an inner funnel (7.f.) inside the neck of
the outer one (formed by the exospore), both being open to
the exterior. ‘he inner funnel is closed at its base by a
broad, fairly thick, and deeply-staining plug or cap (c), which
effectually shuts off the cytoplasm from the exterior. This is,
doubtless, dissolved by the digestive juices of the fresh host

’

LIFE-CYCLE OF “ CYSTOBIA’? IRREGULARIS (MINCH.). 57

when the spore is swallowed, thus allowing the sporozoites to
be liberated. ‘The two membranes, which together constitute
the spore-wall or sporocyst, are now easily discernible (ea. and
end.). he spore-residuum (s.7.) 1s also very prominent,
occupying the central part of the spore. Now and again
(fig. 61a) the rim of the outer funnel shows a distinct tendency
towards angularisation or serration.

I have nothing to add to the observations of Minchin
(loc. cit.) regarding the eight-nuclear stage and the structure
of the sporozoites, and will therefore refer readers to his
description and figures.

C. minchinii.—After repeated and prolonged examina-
tion, I have at length succeeded in making out the general
appearance of the spores of ©. minchinii sufficiently well to
assure myself that, as regards form, they agree fundamentally
with those of C. irregularis. Aripespore of C. minchinii
(fig. 62) is, however, much smaller than one of C. irregularis.
Comparing fig. 62a with fig. 61, which shows spores of
C. irregularis drawn to the same magnification, it will be
seen that the lineal measurements of a spore of C. min-
chinii are approximately only half those of one of the latter
parasites. The shape of both is, nevertheless, strikingly
similar. The only difference that need be noticed is that
there appear to be two (possibly more ?) delicate, curved pro-
cesses in connection with the rim of the funnel. I do not
feel so certain about this point as I should like to be, owing
to the very few cases in which I have been able to see the
entire longitudinal outline of the spore in my sections.!

I have been quite unable to distinguish any division of the
sporocyst into exo- and endospore. In short, the delicacy of
the spore-membranes in C. minchinii is much greater than

1 The spores are so numerous, so small, and so closely packed in the cyst
that it is a most difficult matter to find one cut perfectly longitudinally.
Moreover, in all the material examined, only twice besides have I found spore-
containing cysts. These were both crushed instead of being sectioned.
Unfortunately, however, they were from spirit-fixed Cucumarie, aud the

spores proved to be so crinkled and warty in outline that they were not worth
drawing.

58 H. M. WOODCOCK.

in C. irregularis, where they are firm and do not easily
become distorted in appearance. Neither have I noticed any
deeply-staining plug at the base of the funnel, such as is
apparent in ©, irregularis, although the spores of
C. minchinii are similarly stained. Hence it seems to me
probable that in this species the delicate sporocyst itself is
dissolved. I have no doubt this difference in structure stands
in relation with the different method of infection, the dis-
solving power of the excretory acids, etc., in the respiratory
trees being, probably, much less than that of the gastric or
intestinal secretions of Holothuria.

The contents of the spore, on the other hand, are readily
made out. There are, as usual, eight sporozoites, in this
case arranged in two distinct groups of four—one near either
end of the spore; in the centre lies the sporal residuum (s.r.),
separating the two groups. ‘Two sporozoites are drawn
separately at b. Hach is small and spindle-like, and possesses
a deeply-staining, square to rounded nucleus, situated near
the centre.

(b) Systematic position.

We are now in a position to compare the principal species
hitherto included under the generic heading of Cystobia.
(For C. holothuriez see p. 3). It will be at once apparent
that C. irregularis and C. minchinii are more nearly
related to each other than is C. holothuriz to either. The
two former agree as regards the shape of the spore, which is
usually considered to be the principal classificatory character
in a Gregarine; moreover, their distribution, which is
apparently very limited, also narrowly coincides. In both
these respects C. holothuriz differs markedly. The fact
that association is terminal in C. irregularis and lateral in
C. minchinii is not one that can be allowed to outweigh
the more fundamental points of agreement just mentioned,}

1 There is the same difference in the manner of association in two closely
allied species of Gonospora, namely G, varia and G, sparsa.

LIFE-CYCLE OF “ CYSTOBIA ’ IRREGULARIS (MINCH.), 59

even though association should prove to be lateral also in
C. holothurie.

With regard to the variation in the shape and position of
the sporozoite-nucleus shown in the different cases it is a
little difficult to know what value, if any, to assign to these
differences, since it is somewhat uncertain whether the sporo-
zoites, as thus compared, are equally ripe. Minchin (25)
considers the sporozoites of C. holothuria, in the form of
the definition, as not, perhaps, quite mature, and appears to
think that their nucleus may ultimately become elongated.
If that is so, the same may also be true of C. minchinii.
Possibly, moreover, the sporozoite-nucleus in this species
subsequently takes up a position near one end, but in any
case the point is, at most, one of minor specific importance.!

It is otherwise with the difference in the form of the
spore exhibited by C. holothuriz, which is not sufficiently
emphasised by its present position, especially when we bear
in mind the importance attached to this character in dis-
tinguishing the different coelomic Gregarines. C. holothurie
has been hitherto united with C. irregularis, largely on
account of the similar habitat and precocious association.
Both of these features, however, may have been independently
acquired by another parasite. We see the latter condition
strongly developed in quite a different Gregarine—namely
Diplocystis—and, granting that another form also become
parasitic in the peculiar situation of C. irregularis, it would
be most likely to develop the same intimacy of association.
Again, the distribution of C. irregularis and C. holo-
thurie is, so far as is known, quite distinct. Hence, con-
sidering everything, it seems probable that these two species
are not very closely related (one being derived from or through
the other, as, for instance, C. minchinii almost certainly is

1 The position of the sporozoite-nucleus is not invariably fixed or constant
even in one and tle same species. In the case of the sporozoites of
Urospora lagidis, Brasil (loc. cit.) finds that, while the nucleus is
usually more or less central, it may be near either one end or the other
(compare his figs. 79 and 80).

60 H. M. WOODCOCK.

from C. irregularis), but ought rather to be placed in
different genera. As C. holothuriz is the type-species, this
form ought to retain the generic name of Cystobia, the
other two species being united in a new genus, for which I
propose the name Diplodina.

Relationships.—Diplodina irregularis and D. min-
chinii are, undoubtedly, closely allied to Gonospora,
G. varia, in particular, having a fundamentally similar spore.
On the other hand, the tailed spore of Cystobia holothuriz
resembles that of one or two genera placed by Léger (19), on
account of that very character, in a family distinct from the
Gonosporide—namely, the Urosporide. The affinities
of C. holothuriz seem to be, in fact, with Lithocytis
schuneideri, and, in a slightly less degree perhaps, with
Urospora (cf. lagidis).

Definition.—We may therefore summarise the systematic
position of the parasites above discussed as follows :

Family, Gonosporide; genus, Diplodina (nom. nov.).

A neogamous Gregarine of large size and regular shape,
parasitic in Holothurians. When in the “free” or gregarini-
form condition the adult couple is perfectly oval in outline,
when encysted it is spherical. Conjugation is completely
isogamous. The ovoid spores have dissimilar poles, the basal
end being rounded, while the exospore at the opposite end is
prolonged as a funnel- or cup-like expansion open to the
exterior.

(1) D.irregularis (Minch.). Association terminal (end-
to-end). Septum between the two members either present
or absent. ‘The cysts remain attached to the wall of the
blood-vessel. The size of the spore averages 26 w x 12 pn.
‘The endospore, also, is turned outwards, as an inner funnel open
to the exterior, its base being closed by a cap or plug. The
sporozoites possess an elongated nucleus near one end.
Habitat: blood-vessels of H. forskali.

(2) D. minchinii, Woode. Association lateral. Septum
never present. Sporulating cysts usually occur in the con-
nective tissue of the body-wall or retractor muscles. The

LIFE-CYCLE OF “ CYSTOBIA’’ IRREGULARIS (MINCH.). 61

size of the spore averages 14 x 6m. There is probably
no inner (endospore) funnel open to the exterior ; moreover,
no sign of a plug is visible. Sporozoites possess (in all
instances observed) a central nucleus, square or rounded.
Habitat : respiratory trees and body-cavity of Cucumaria
pentactes and C. planci.

Family, Urosporide; genus, Cystobia, Ming., 1891.

The exospore at the basal end of the spore is prolonged
into a flat process, like a lance-head.

C. holothuriz (Schn). Not having had an opportunity
of personally studying this parasite, I am unable to add to
the definition above given (p. 3).

The consideration of these neogamous Gregarines affords
an excellent opportunity for a general discussion of the
phenomenon of association, as it is known to occur in the
order.

(10) Precocious Association or Neoagamy.

(a) Comparative account of the principal varia-
tionsinthe mannerand time of association.

We may commence by following the gradual increase in

the precocity and intimacy of the “ pairing”

as exemplified
by different forms. Leaving out of account, for the present,
the “chains” formed by certain Gregarines, which are pro-
bably often broken and differently reconstituted, and whose
significance is not fully understood, we have, in the simplest
cases, a union of two adult trophozoites immediately prior to
encystment and the commencement of sporulation (Lankes-
teria, Monocystis, Stylorhynchus, etc.). The dividing
partition separating the two associates does not usually break
down until nuclear division is well advanced. Next, we get
the binary syzygies of most adult Gregarines, which proceed
to encystment after remaining attached to each other for
some time, as in Gregarina cuneata and other meal-worm

62 H. M. WOODCOCK.

Gregarines (Berndt [1]), etc. The first indication of pre-
cocity is perhaps seen in Zy gocystis, where the individuals
always associate as soon as they become adult trophozoites.

Diplocystis.—The development of neogamy to an ad-
vanced degree, along parallel but quite independent lines, is
well shown by Diplocystis on the one hand and Diplodina
and Cystobia onthe other. In the case of Diplocystis
major, a ccelomic parasite of the cricket (G@ryllus), Cuénot
(loc. cit.) found that the Gregarines can grow toa certain size
alone (up to ‘2 mm. in diameter) but not to their full extent,
i.e. to become sporonts; couples, however, attain to a diameter
of 1:3 mm. Still, the mutual adherence is very feeble—the two
individuals being easily separated—and, moreover, often in-
discriminate, several Gregarines of different ages being at
times grouped together (see Cuénot’s fig. 40, Pl. 20), some of
which have to be got rid of asde trop. A further advance
is seen in D. minor, in which the longest diameter of a
couple is, at most, "6 mm. Here “pairing” takes place
much earlier in life. Solitary individuals larger than 30 pu
are never met with, and, indeed, association is sometimes
accomplished before the parasites have passed from the gut-
wall into the coelom. Moreover, the union is closer and more
definite, the two individuals becoming enclosed in a common
membrane, which may be looked upon as an early-formed
ectocyst. Nevertheless, inside this, the double nature of the
syzygy is distinctly to be seen, there being an obvious
V-shaped constriction all round the middle, marking the
plane of junction, and only slightly less prominent than in
my fig. 3 (see Cuénot, fig. 41, or Minchin [26], fig. 22).

Next comes D. schneideri, the trophic phase of which I
have above re-described. While the association in this para-
site may be, apparently, sometimes more precocious and
intimate than in D. minor,! an advance upon the condition

1 In none of the individuals of D. schneideri which I examined was the
septum absent. Kunstler, however, in his fig. 16 (loc. cit.), has drawn a
small example, still attached to the intestinal wall by the peritoneal epithelium,
which appears to lack a septum. Cuénot (loc. cit.) is of the opinion that

LIFE-OYOLE OF “ CYSTOBIA”” IRREGULARIS (MINCH.). 638

found in that species is chiefly noticeable in the modification
of the encystment-process (see above, p. 27). Nothing of
this kind is mentioned by Cuénot in his description of
D. minor. Judging from this author’s figures, the formation
of the ectocyst there resembles the manner in which it occurs
in the ordinary process of encystment, as the spherical shape
of the couple is much more cyst-like and the ectocyst does
not appear to enter at all into the plane of junction. D.
minor represents, therefore, a condition intermediate be-
tween ordinary encystment and the secreting of an early
ectocyst at all points of the body, closely following the adult
form, which is thereafter retained, as in D. schneideri.

Thus we can conveniently arrange the different species of
Diplocystis ina series, as follows:

(1) D. major, in which precocious association is already
necessary if the trophozoites are to become full grown, but
its occurrence is haphazard and promiscuous, and the actual
attachment only feeble.

(2) D. minor, in which neogamy occurs early in life, and
its permanence is assured by precocious encystment.

(3) D. schneideri, in which the ectocyst-formation is
modified and distinctly adapted to the trophic phase of the
life-history.

Diplodina and Cystobia.—In these two genera we
find the most advanced condition of neogamy known among
Gregarines.

In D. irregularis this extremely early and intimate
joing appears to be in process of development. Many
individuals—probably the majority—are content to unite so
as to acquire at length the appearance of one septate but
Kunstler must have overlooked one in this case. ‘The septum when present
is, however, so obvious, and the structure of the particular specimen was, for
the rest, so clearly seen, that I do not think this view altogether certain. It
is not necessary, moreover, to suppose that the septum is invariably present.
Quite possibly in certain, perhaps exceptional, instances neogamy in D.
schneideri may be so precocious and intimate that no partition separates the

two associates, just as is often the case in Diplodina irregularis (see
next page).

64. H. M. WOODCOCK.

binuclear Gregarine. Fig. 3 represents the least intimately
joined couple that I have observed. In this case association
had probably taken place but a short time previously. With
growth, however, the “double” nature of the parasite (as
indicated by the V-shaped constriction) would, doubtless,
have become less and less obvious, until the typical adult
condition was reached (figs. 1, 2). On the other hand, many
individuals unite at the earliest opportunity and very inti-
mately ;1 in such cases the actual cytoplasm of the two
associates hecomes joined,? and there is no indication what-
ever of the plane of union.®

In D. minchinii complete union of the two associates is
found in all cases (cf. figs. 9, 10, 12, 18, and 19). This
species, so far as the cytoplasm is concerned, is absolutely
indistinguishable from an ordinary Monocystid Gregarine.
The only clue, indeed, to the plane of junction is the
invariably transverse arrangement of the nuclei (see above,
pal).

There can be no doubt that the presence or absence of a
partition between the two associates depends upon how soon
neogamy occurs. Where we find an intimate cytoplasmic
union it is almost certain that the association has taken place
between naked sporozoites, when there would be no lmiting

1 Associations belonging to this category appear to be more numerous as
the season advances. Probably such factors as temperature, change of season,
and the supply of nutriment available in the host at the time, exert an
influence on the parasites, and induce them to anticipate, and prepare for,
the ultimate phase of their existence—sporulation.

2 This fusion is not to be considered comparable to the complete and homo-
geneous union which occurs in actual conjugation (see below, p. 68).

5 The position of Cystobia holothurie with regard to this biological
feature is somewhat doubtful. Mingazzini (27) says that a septum divides
the Gregarine while it is attached to the blood-vessel, but Schneider (85)
only saw one in cysts which had fallen off into the body-cavity. Minchin
(25), again, did not see a septum in any case. Probably the explanation of
these conflicting observations is that this parasite exhibits the same variation
in this respect that D. irregularis does. Further study of the earlier
phases of the life-history, while the trophozoites are still in the vessels, is
necessary to decide the point.

LIFE-CYCLE OF “* CYSTOBIA’’ IRREGULARIS (MINCH.). 65

membrane to divide the two halves. For I have never seen
the least indication of absorption of the septum, during the
trophic phase, when it is present (see above, p. 25). Though
I have not had an ocular demonstration of the actual joining
so early, my fig. 52, showing in the wall of the respiratory
tree a D. minchinii only 18 w by 12 mw, not yet rounded off
since its entry, is strong evidence in support of this view.
The fact that I have never observed a uninuclear form, how-
ever small, in the wall of the tree, where the parasites are
readily distinguishable, also points to the same conclusion.
In addition, the universal occurrence of cytoplasmic fusion
in this species seems to imply that any unpaired individuals
very quickly die off, being probably quite unable to become
even young trophozoites alone.

The union itself, in D. minchinii, most likely occurs in
the lumen before the parasites have penetrated into the wall.!
In D. irregularis, in those cases where the association is
between young trophozoites, it probably takes place in the
lumen of the blood-vessel; where neogamy is most intimate
it doubtless occurs when the individuals are little more than
sporozoites, but in what situation I am unable to say with
certainty. ‘The passage of the parasites to their ultimate
situation may be very rapid, for I have often found intimately
associated trophozoites which have evaginated the wall of the
vessel while very young (see above, p. 13). Probably in such
cases the sporozoites have associated at once, before passing
through the gut-wall and into the lumen of the vessel.

The morphological condition found in Diplodina
is to be interpreted as one of union, and not of
imperfect division.—I am thus unable, it will be seen,

1 Unfortunately, searching for very minute forms in this situation is most
difficult work, owing to the numbers of loose cells of various kinds, which
constitute in places a spongy tissue, tending to block up the lumen. More-
over, since the solitary sporozoites would not remain long in the lumen before
associating and passing into the tissue, only by very good fortune would one
come across them in a section.

voL. 50, PART 1.—NEW SERIES. 5

66 H. M. WOODCOCK.

to bring forward any absolute proof that in Diplodina we
have to do with union at all, and not, on the contrary, with
division of varying incompleteness. It may be worth while,
therefore, to say briefly why the trophozoites of this genus
are to be regarded as couples, and not as individuals which
have undergone imperfect division.

The case of Diplocystis schneideri is of much im-
portance in this connection. Here, also, is an instance where
neither a uninuclear individual nor the actual formation of a
pair has yet been observed. However, in D. major and
D. minor, two closely allied species, we have certainly to do
with the union of distinct individuals. Now, the condition
found in D. schneideri differs only very slightly from that
of the adult trophozoites of D. minor, and, as explained
above, is easily derivable therefrom. Hence it is hardly
open to doubt that D. schneideri is also an example of
precocious association, and not of early partial division. We
have seen, moreover, that in very young forms the septum
between the two halves may, apparently, be absent.

The genus Diplocystis affords, perhaps, the strongest
argument in favour of a similar interpretation of the mor-
phology of Diplodina. One or two other corroborative
points may also be mentioned. Comparing trophozoites of
D. irregularis of different age, it is at once seen that, as
growth proceeds, the separation between the two halves of a
syzygy invariably becomes less instead of more pronounced.
It can confidently be said that the two members of a “ couple”
never separate, which is the reverse of what one would expect
if they resulted from the division of a single individual. If
that were so, moreover, we should have conjugation taking
place between the progeny (gametes) of the two halves of
what was originally one individual (derived from one sporo-
zoite), an occurrence for which there is no precedent or
analogy among the whole of the Telosporidia. For these
reasons, therefore, I think it most likely that the trophozoites
of Diplodina—and equally, of course, those of Cystobia—
are correctly regarded as neogamous.

LIFE-CYCLE OF ‘‘ CYSTOBIA’”’ IRREGULARIS (MINCH.). 67

(6) Biological Considerations.

The reason for neogamy.—The reason for the gradually
increasing precocity and intimacy of association, as we can
trace it in different spp. of Diplocystis and Diplodina,
is undoubtedly to be found in the absence of any power of
movement in these coelomic parasites. The whole process, in
short, is the result of an endeavour to insure a suitable and
durable association, since the animals are not able to move
about themselves in search of a partner when ready to com-
mence sporulation. The fact that neogamy is only known
to occur in these non-motile forms bears out this view.

Let us consider D. major, for example. It is purely a
matter of chance whether two individuals of the right age are
brought near enough—as they are passively washed about in
the body-cavity—to attract each other reciprocally and asso-
ciate. Besides this, there is the danger of unsuitable and
indiscriminate grouping, and also of a separation of two
partners (if not ripe for encystment) when violently dis-
turbed ; in such cases a second association might be impos-
sible. Probably, therefore, a considerable percentage of
individuals thus isolated from one reason or another is
doomed to phagocytic destruction. A method so haphazard
must prove very expensive, and in the other two species
association is modified as above described. Thus unequal
and multiple associations are, as a rule (though not entirely)
prevented, but the membrane probably interferes, to a
certain extent, with the absorption of nourishment by the
trophozoites.

At all events, this plan is not followed in Diplodina and
Cystobia. In these parasites the difficulties in the way of
association between adult trophozoites would be even greater,
because of the more confined situation. We find, corre-
spondingly, in these genera no stage equivalent to that seen in
D. major. Those races of individuals which did not adopt
early and permanent association appear to have all died off.

Effect on the individuals.—Notwithstanding the appa-

68 H. M. WOODCOOK.

rent intimacy of the two members of the couple in those
cases where the septum is absent, it may safely be said that
there is nothing approaching a true conjugation as yet
between the two halves. Though I have many a time found
the two nuclei in contact, I do not believe there is any
nuclear fusion or interchange of material; for I have never
seen the slightest sign of a break-down in the nuclear wall
or of alteration in the nuclear constituents.

It is more difficult to be certain with regard to the cyto-
plasm. Personally, I think the two cytoplasms remain (until
sporulation) as individually distinct in the cases where there
is no septum as in those where it is present. The fact that
in D. irregularis the septum may persist until nuclear
multiplication is considerably advanced, when the two cell
territories remain for the time being quite separate, renders
it highly probable that this is true when the septum is
absent, although the delimiting plane is not distinguishable.
The reason for the inability to see the plane of junction in
such acase is not, I consider, because it is no longer there,
but because it is not constituted by a layer (ectoplasm, or
limiting membrane) different from the general cytoplasm.

In all the other Gregarines whose life-history has been
investigated the complete morphological “separateness” of
the cytoplasm of the two associates is a recognised feature
(see Introduction, p. 5), and there is no reason to suppose it
is otherwise in Diplodina. That there is any complete and
homogeneous intermingling of the cytoplasm of the two indi-
viduals, resulting in one binuclear organism, I do not for a
moment believe; such an actual merging of two distinct
entities into one takes place first later, when the gametes
copulate. Association certainly gives a stimulus to further
development, and, as we have seen, neither growth nor life
can go on in Diplodina without neogamy. Nevertheless, 1
think that such stimulus is principally cytotactic (see below,
p- 73), and of a radically different character from that which
is imparted by true conjugation.!

' See Hertwig’s suggestive essay (14) on the significance of fertilisation.

LIFE-CYCLE OF “ CYSTOBIA”’? IRREGULARIS (MINCH.). 69

I do not even think that the intertwining lobes and _ pro-
cesses above described are developed in D. minchinii until
nuclear division is well advanced. The very definite and
constant position of the nuclei in the gregariniform adults of
this species (see figs. 9, 10, and 12) is strong evidence that
the “body” of each associate remains (except in the plane of
junction) quite as distinct morphologically as it does, for
instance, in Gonospora sparsa (see Léger’s figs. 3 and 4,
Pl. 20 [19], or Minchin’s fig. 20 [26]).

Is triple association successful ?—The Diplodina
triplets drawn in figs. 20 and 28 seemed perfectly healthy,
and possessed typical nuclei; the one drawn in fig. 29, how-
ever, has not the customary appearance and looks rather
abnormal. This is a large specimen, being *6 mm. by ‘4 mm.,
and its nuclei have stained up almost homogeneously, the
karyosomes being small and indistinct. The cytoplasm, too,
has an altered look, with deeply staining lumps and grains
(g), probably of an albuminoid nature, unevenly scattered
about; in addition, a few vacuoles (vac.) are present near
the margin. ‘his syzygy, in short, appears as if it were
already commencing to degenerate. Opinion differs as to
whether these triplets are capable of successful sporulation
or not; the weight of evidence, where their further develop-
ment has been followed, seems to point to their subsequent
degeneration.!. The reason for the binary nature of a typical
syzygy will be fully understood when the essential signifi-
cance of association has been discussed.

Probably, in any case, the associates require to be of about
equal age and size if the union is to be successful. Berndt
(loc. cit.) mentions that if the members of an ordinary
syzygy are of considerable difference in size, no true cyst-

1 Berndt (1) states that in Gregarina cuneata associations of more than
two individuals invariably come to nothing. Cuénot (loc. cit.) instances only
very rare cases of triple association in Diplocystis, one of which has
apparently produced sporoblasts. Judging from his fig. 47, however, the
sporulation would not seem to have been successful, the sporoblasts being

extremely minute and scarcely visible, very different from the well-developed
layer in the normal cysts figured.

70 H. M. WOODCOCK.

wall is formed, and the animals soon die off. Siedlecki
(loc. cit.), Léger [19] and Brasil [8] give examples where
one member of a couple has produced primary sporoblasts in
a normal manner, while the other, most likely because the
above condition was not fulfilled, has remained stationary
and, indeed, succumbed. One Gregarine of a couple can in
certain cases, apparently, exert sufficient influence upon its
associate to induce it to commence sporulation, although it
itself may not be ripe enough to do so, and as a result not
only does it not benefit by the stimulus or “Reiz” of the
other, but this, on the contrary, appears to harm it and it
succumbs instead.

(11) GenmRAL SIGNIFICANCE OF ASSOCIATION.

Descent of the Telosporidia.—We are thus led toa
consideration of the origin and significance of association in
general. We may commence by endeavouring to trace the
descent of the Telosporidia—that sub-class of the Sporozoa
which includes the Gregarines—from a Flagellate stock.
For Schaudinn (83) has recently demonstrated that a certain
malarial parasite of birds, namely Halteridium noctue, is,
in reality, only a phase in the life-cycle of a Trypanosome, and
it is becoming increasingly probable that the Heemosporidia
as a whole are derivable from, and in many cases still
closely connected with, Heemoflagellates. There can be
little doubt that the Coccidia and Gregarines have an
origin similar to that of the Heemosporidia, for the first-
named are closely allied to this order, and the differences
exhibited by the Gregarines are easily understood when their
somewhat different relation to the host is borne in mind,
Hence it appears most likely that the whole of the Telo-
sporidia have originated from a Flagellate ancestor.!

In the case of the Hmosporidia the original parasitic

1 This ancestor may be assumed to have been of a fairly generalised type,
such, for instance, as that whose life-history is outlined by Doflein (11),
p. 53.

LIFE-CYCLE OF “ CYSTOBIA’”’ IRREGULARIS (MINCH.). 71

Flagellate became specially adapted for life in the blood.
With the acquirement of an alteration of hosts the life-
history has become extremely complicated, but the details do
not concern us here. It is probably in this order, however,
that the Flagellate phase is retained to the greatest extent.

The Coccidia are to be regarded as having resulted from
the original parasite becoming entirely intra-cellular. With
this is correlated their non-motility and the great specialisa-
tion which we find in the sexual process. The numerous male
gametes are highly differentiated and constitute, so far as is
known at present, the only phase in the life-cycle where the
Flagellate ancestry becomes manifest. The female individual,
or megagametocyte,! no longer divides up to form many
gametes, but itself becomes one directly after nuclear re-
duction. In Cyclospora, however, Schandinn has recorded
(84) certain abnormal cases of the persistence and further
multiplication of the “reduction nuclei” of the female
element, followed by multiple fertilisation. This occurrence
points very strongly to the conclusion that there were ori-
ginally many female gametes also, a condition which would
agree with that found in Gregarines.

Lastly, we come to the Gregarines, which are by far the
most successful group. I do not agree with Minchin (26,
pp. 272, 273) in considering this order (any more, indeed,
than the other Telosporidia) as derived from an originally
intra-cellular form, which re-acquired, secondarily, an inter-
cellular habitat. It is much more likely that the Gregarines
have never been entirely intra-cellular, since, in that case, 16
is almost certain they would have completely lost their
mobility and any cytoplasmic differentiation they possessed.
Moreover, the female individuals would also have tended to
reduce the number of gametes, of which we find no trace.

Whereas, in the evolution of the Coccidia, the young para-
site at length penetrated completely into the host-cell and
stayed there, in no instance do we find such an advanced

1 In many Coccidia bisexual differentiation becomes marked in the adult
trophozoites, or even earlier.

fp: H. M. WOODCOCK.

condition in Gregarines. Speaking broadly, further develop-
ment in this order can be regarded as having taken place
along two distinct lines. In the one case, characteristic of
the great majority of Septate forms, the young growing para-
site penetrates only partially into the host-cell after attach-
ment,! and remains in that position only during its early
growth.’ For the rest of its trophic (and, of course, its sporo-
gonic) life the Gregarine is free in the gut. Concurrently
with the loss of the flagellum the parasite has developed the
peculiar gregarinoid movement (easily derivable from one of
a euglenoid nature) and, correlated with this, we find a marked
differentiation of the peripheral cytoplasm.

The other line, that of the Monocystids, branched off early,
before the septum characterising most intestinal or Polycystid
Gregarines became developed. ‘The young parasite, at the
beginning of its life, followed one of two courses. Lither it
penetrated entirely into the host-cell and remained there for
a long time before falling into the gut (e.g. Lankesteria
ascidiz) or, on the other hand, it passed inwards between
the cells of the mucous membrane and came into intimate re-
lation with the connective tissue, etc., of the sub-mucosa, being,
as a rule, intercellular. In this latter manner have originated
the coelomic forms. In many cases (above described) these have
now, for the first time, lost their motility; in other words,
these ccelomic Gregarines have developed a trophic condition
similar to that prevailing in the Coccidia, but quite apart and
independent from them. Neither on account of their habitat
nor because of the apparently simple nature of the sexual
process (see below, p. 79, et seq.) are the Monocystids to be

1 This question of attachment is very interesting as regards its bearing upon
the Flagellate ancestry of the order. Léger (21) has recently described
various parasitic Flagellates (Herpetomonas, Crithidia), characterised
by the possession of a gregariniform phase, in which the flagellum either
becomes greatly shortened and reduced, serving as an organ of attachment to
the epithelial cell of the host, or else disappears.

2 In one or two instances (e. g. Stenophora) the sporozoite penetrates
completely into the cell, and the parasite remains intra-cellular until growth
is finished. This is, undoubtedly, a secondary and not a primitive condition.

LIFE-CYCLE OF ‘‘ CYSTOBIA”’ IRREGULARIS (MINCH.). 73

regarded as more closely representing the ancestral Gregarine
than the Polycystids. One primitive character which has
been preserved, however, is the unseptate condition of the
body.

Origin of association.—With regard to its sexual
reproduction! our ancestral Gregarine, doubtless, greatly
resembled at first the parasitic Flagellate from which it
was derived, numerous gametes being developed inside a
membrane (probably lttle more than the unused parental
residue), and sexuality being by this time marked; at any
rate, we may assume that the male elements were smaller
and more active than the females. It was at this stage that
association made its first appearance, and to this adaptation
one cannot doubt that the highly successful evolution of the
order is chiefly due. Association, to commence with, would
be quite accidental and indiscriminate, at first only occurring
to any purpose between individuals which were dividing up
to form gametes, when we may suppose them mutually
capable of exerting an attractive or cytotactic influence.
Such a coming and reinaining together would obviously
allow of a far greater percentage of the gametes copulating
than if the male elements had to wander about trying to
find the female ones. In other words, successful reproduc-
tion—ample provision for which is especially necessary in a
parasite—would be much facilitated and increased. ‘The
advantage thus gained would lead to the development of
this beneficial modification, and reciprocal attraction would
not only be strengthened, but would tend to be manifested
earlier, before the gametes were formed. In this way we
get the cytotactic (chemiotactic ?) influence thrown back to,
and becoming apparent in, the parent individuals—that is,
the adult Gregarines themselves. It is but a slight step in
advance to suppose that this gradually led to the formation
of the gametes being regulated by, and dependent on, the

1 Asexual reproduction, it may be mentioned, has hecome completely lost
in most Gregarines.

74: H. M. WOODCOCK.

occurrence of such association ;’ in other words, the process
acquired a distinct physiological import, in that by it a
developmental stimulus was imparted to each associate. We
thus arrive at the present-day condition in the group.
Essential importance of the process.—Association
is necessary in the great majority, if not in all, Gregarines for
sporulation to take place ; in other words, we may say that an
interacting stimulus such as is mutually exerted by the two
associates is requisite to produce the formation of gametes or
primary sporoblasts, and without it the animals are unable to
fulfil their complete life-cycle.” A very interesting parallel
case, proving that association may undoubtedly become neces-
sary for the production of gametes, is seen in the Coccidia.
Here, with the exceptions to be mentioned, the phenomenon
is unknown, the gametocytes giving rise to the gametes
apart and quite independently. In Adelea mesnili, how-
ever, we generally find association of the male and female
gametocytes either in the same or in neighbouring cells.
The male individual (microgametocyte) may, however, still
develop microgametes apart from the female individual
(megagametocyte), but in that case they come to
nothing (see Perez [80]). In Adelea ovata and Klossia
helicina, on the other hand, unless the male gametocytes
become associated with the female ones, neither can they
form male elements nor can the female cells matu-
rate and become ready for fertilisation (see Siedlecki
[38]). That is to say, we have here an attempt on the part
of certain Coccidia to imitate the highly successful expedient
of the Gregarines; and just as in the above-mentioned para-
sites association is obviously necessary to induce the for-
mation of gametes (equivalent to sporoblasts), so it has
undoubtedly become in the vast majority of Gregarines.

1 There are many examples of ripe solitary sporonts dying off instead of
producing sporoblasts, and this is, in all likelihood, because they have been
unable to associate.

2 Siedlecki (loc. cit.) has already expressed a similar opinion with regard to
the stimulating effect of association, in the case of Lankesteria ascidie.

LIFE-CYCLE OF “ CYSTOBIA”’ IRREGULARIS (MINCH.), 75

As the life-cycle of any Gregarine becomes completely
known, it will almost certainly be found that each individual
is, at some time or other prior to the commencement of
sporulation, associated with a fellow one, such association being
in nearly all cases permanent, and followed by encystment
and true conjugation. I have not been able to find a single
instance where it has been definitely proved that a single
Gregarine, which has never associated, can sporulate,
and the more life-histories that are thoroughly investigated
the more unlikely does such an occurrence become. Siedlecki,
Cuénot, Berndt, and Léger, in their recent researches, respec-
tively emphasise the fact that they never observed single
encystment at all. Brasil observed solitary encystment occa-
sionally in Gonospora and Urospora, but it was unsuc-
cessful; the nuclear multiplication wasabnormaland the cyst fell
a prey tothe phagocytes of the host. Still, one or two authors
have described the occurrence of normal solitary encystment,
and appear to consider that successful sporulation without
conjugation (i.e. parthenogenesis) results. ‘I'hese instances
are considered below. What I maintain, however, is that,
granting such cases do occur, in all probability there has
been a temporary association previously, when the develop-
mental stimulus was imparted; and, indeed, where solitary
sporulation is described, temporary association is usually
mentioned as also taking place.

Relation between associationand conjugation.—
We may now compare the varying relations between associa-
tion and conjugation which we find in the group. ‘The least
modified condition is seen in the Septate Gregarine Ptero-
cephalus, where there is markedly anisogamous conjuga-
tion, after association and common encystment of the couple
(see Léger and Duboscq [23]). Now, it is obvious, when the
formation of gametes became delayed until after association,
and their successful copulation was thus practically insured,
there was no longer the same necessity for the gametes to
be differentiated. Hence, as might be expected, we find a
reduction to isogamy taking place, of which Stylorhynchus

76 H. M. WOODCOCK.

already shows the commencement (see above, p. 5). A
most interesting stage in this reduction is exemplified by
Urospora lagidis (and also by Gonospora varia),
recently described by Brasil (loc. cit.). Both kinds of
gamete are here of about equal size and alike in shape.
There is, however, a sharp nuclear distinction apparent. The
nuclei of the elements proceeding from one half of the cyst
are smaller and chromatically denser than those belonging
to the gametes from the other half.! | The former are pre-
sumably male elements and the latter female, though this is
not explicitly stated by Brasil.

The final stage is that of complete isogamy, where there is
no discernible evidence of sexual differentiation whatever, at
any rate in the gametes themselves. Typical instances are
those of Gregarina, Lankesteria, and Diplodina, all of
which have been discussed above. Nevertheless one may, I
think, regard binary association as itself, in a measure,
physiological evidence of sexuality. And, indeed, in a few
instances (e.g. Aggregata vagans, Stenophora varians,
Pterocephalus, and Diplocystis clerci) there are mor-
phological differences between the two associates which
Léger (22) is inclined to consider as indications of sexual
differentiation. If, then, one may ascribe binary sex to the
parent individuals (associates), even where the gametes are
completely isogamous, this would explain why, in most cases,
association is binary, and why sporulation is unsuccessful in
the triple syzygies commented upon above.

Unusual and specially modified forms.—While in
most of the coelomic Gregarines actual copulation is evidently
still necessary, since the trend of development has been to
insure a lasting union, there are, on the other hand, one or
two forms known in which conjugation of the gametes—i. e.
the true sexual process—can be apparently dispensed with.
The starting-point of parthenogenesis in the order is perhaps
seen in instances where we find both conditions, copulation

1 An almost similar state of things is really presented also by Monocystis,
which has been lately re-examined by this same worker (4).

LIFE-CYCLE OF “ CYSTOBIA ”? [RREGULARIS (MINCH.). 77

and independent development of the sporoblasts, co-existent
(e.g. Ophryocystis, Léger [20]). Again, in Ceratospora
mirabilis from the body-cavity of Glycera, Léger (19) de-
scribes and figures the formation of (many) sporoblasts and
spores separately in the two associates, the septum appa-
rently remaining unabsorbed and preventing copulation. If
this can be relied upon as being a normal mode of spore-
formation, it would seem that we have here, also, the
beginning of this independent development of the sporo-
blasts.1

We have next to consider the case of two or three forms
which are said to be capable of encysting and successfully
sporulating alone, in addition to following the customary
course. The evidence brought forward so far is, however,
by no means conclusive. Léger has described (loc. cit.)
solitary encystment and sporulation in Gonospora varia,!
as well as the permanent encystment of a couple. Brasil,
however, in his paper on the same species (8) says nothing
about the independent (complete) sporulation of each member
of a couple.”

On the other hand, Brasil draws attention to the work
of Caullery and Mesnil on Gonospora longissima and
Selenidium, spp. Unfortunately, the observations of these
authors are too scanty to help much in deciding this question.
Concerning the former parasite they remark (7) that the cysts
seem formed either of one or two individuals, but nowhere do
they definitely state that the cases of solitary encystment
were successful. In addition the occurrence, now and again,

1 It must be confessed that this observation (and also that of Gonospora
below) stands greatly in need of corroboration. Léger’s early work was all
done, it must be remembered, before true conjugation was known. Probably
sometimes, at any rate, the dividing septum breaks down, allowing conjugation
to take place, for in some cysts of Ceratospora active movement of the
gametes (Schneider’s ‘‘ danse des sporoblasts”’) was observed.

2 Brasil notes an instance of a cyst of Urospora in which one individual
had died off and was degenerating, while the other, to judge from its healthy
appearance, seemed to be pursuing its normal course, with what ultimate
success, however, we are not informed.

78 H. M. WOODCOCK.

of very intimate association! in this species, renders it not
unlikely that such an encysted couple, when sporulating, may
have been mistaken for a single individual. Again, a species
of Selenidium from the gut of Spio martinensis appears
to encyst as often singly as in couples, nuclear changes taking
place in both cases (see C.and M.[5]). Actual spore-forma-
tion, however, was never observed, since sporulation is
not completed while the parasites remain inside the host.
S. echinatum, from Dodeceria, exhibits the same latitude
in the matter of encystment. Nevertheless, C. and M. state
(6) that in three hosts in which spore-containing cysts were
found they were undoubtedly normal, constituted in all cases
by associated couples.

To sum up, we may say that while, except in Ophryo-
cystis, successful independent sporuiation cannot be re-
garded as proved, it is not improbable that such a modifica-
tion of the typical life-history may be met with, in special
circumstances. Consider Gonospora varia, for example.
Here we have a ccelomic parasite which is non-motile, but
which has not developed neogamy. Hence association, just
when desired, must be more or less precarious. It would not
be surprising, therefore, to find that this form had acquired
the ability to develop also parthenogenetically. There can be,
however, little doubt, I think, when the whole line of argument
of this and the preceding section is considered, that if such
independent sporulation occurs, it is only after prior tem-
porary association.” Such association, although the tropho-

1 In certain cases, Caullery and Mesnil observed that the partition dividing
the two associates disappeared. Brasil (loc. cit.) points out the analogy between
this occurrence and the neogamy of Cystobia [Diplodina]. Lt probably also
represents a step taken to bring about permanent association ; it will be seen,
however, that this parasite is proceeding upon a slightly different line from that
taken by Diplodina, where the septum, once formed, never breaks down (in
trophic life).

2 Commenting upon association in the Selenidium from Spio, Caullery
and Mesnil (loc. cit.) say that the adherence of the two members is very
feeble, and they are easily and often separated. Heuce, it isextremely probable
that the individuals which sporulate alone have at one time or another been
associated,

LIFE-CYCLE OF “ CYSTOBIA”’? IRREGULARIS (MINCH.). 79

zoites are not ready to encyst permanently, may well suffice,
under certain modified conditions, to impart the essential
developmental stimulus.

Gregarine “chains.”—A word or two, finally, with
regard to the curious “chains” or aggregations of Gre-
garines, such as are known to occur, for example, in Hirmo-
cystis, Clepsydrina, etc. Their significance is very
problematical. It is possible that the cytotactic attraction
has become modified in function. Its primary object, to
induce the formation of gametes, being, as in Diplodina,
delayed or in abeyance, it may perhaps serve a subsidiary
purpose in trophic life.

Isogamy, in the Gregarines, is to be regarded as
the more modified, and not as the more primi-
tive, condition.

In conclusion, it may not be superfluous to point out some
of the reasons why the general course of evolution above out-
lined seems to me to have much more in its favour than what
is practically the opposite view. I should, indeed, hardly
consider it necessary to do so were it not that this contrary
view has been recently supported by Brasil (8), and Nusbaum
(28). Both these authors consider isogamous copulation to
represent the more primitive condition in the order. Brasil
regards the case of Urospora and Gonospora as marking
an early stage in the evolution of anisogamy, the highest
development in this direction being attained by Ptero-
cephalus. Nusbaum thinks that association at first was
well developed, and served as a stimulus to the two mem-
bers, which gave rise to isogamous gametes (example,
Lankesteria). Subsequently, the process became less
important, the two associates not entering into such inti-
mate contact; correlated with this we have the formation
of male and female elements (example, Stylorhynchus).
Lastly, the association has a quite transitory significance, and
is, in short, a reduced phenomenon ; the associates themselves

80 H. M. WOODCOCK.

are not necessarily of opposite sex, and the encystment is only
temporary, the gametocytes becoming free later and giving
rise to markedly anisogamous gametes (example, Schau-
dinnella).!

Now, the great success of the Gregarines is undoubtedly
due to their power of closely and permanently associating
during the period of formation of the gametes. Granting,
therefore, the possession by the parasites (in the first place) of
such a useful property, it seems most unlikely that they would
dispense with the same (as, according to Nusbaum, Schau-
dinnella has done) without very good reason. For in that
case we should have a tendency exactly the reverse of that to
which the prosperous development of the order has above
been attributed. Moreover, the primitive forms in the gut
(supposedly isogamous), being capable of movement, could
have no difficulty in associating ; it is not easy to see, there-
fore, why they should adopt a less certain mode of sexual
reproduction.

The ccelomic Gregarines are very interesting in this
connection. Here, if anywhere, association is difficult of
attainment (cf. above, p. 67). Hence, if anisogamy in
this order has been developed from an isogamous con-
dition, we might reasonably expect to meet with it here,
since by this means the necessity for permanent associa-
tion would be obviated. We find, on the contrary, how-
ever, that, speaking generally, it is these ccelomic forms
which are most nearly or entirely isogamous. So far from
acquiring anisogamous gametes (comparable to those of
Pterocephalus or Stylorhynchus), they have, in many
cases, developed the condition of neogamy—that is to
say, they have been at pains to insure association, And in
the few cases where, if anywhere, permanent association can
apparently be dispensed with this is certainly not by the
development of anisogamy, but by the very unusual modifi-

1 Nusbaum (loc. cit.) does not endeavour to explain why, as is apparently

the case, the male elements of this remarkable form are non-motile, which
is hardly what one would expect according to his view of the course of evolution.

LIFE-CYCLE OF ‘* CYSTOBIA”’ IRREGULARIS (MINCH.). 81

cation of parthenogenetic reproduction. It would seem
rather as if these isogamous forms could not become (again)
anisogamous.

The Coccidia are very interesting in this connection. From
what has been said above, it follows that, if permanent asso-
ciation together with isogamy is to be regarded as tlie more
primitive condition in the Gregarines, it was probably also
a feature of the ancestral Coccidian. We might, there-
fore, reasonably expect to find association persistent, in some
cases at any rate, and—equally probably—precociously de-
veloped. Now, in no instance is this occurrence known in
the Coccidia as a primitive condition. But we do find it
now and then (see above, p. 74) as a quite secondary
condition,! acquired later and independently. That
is to say, in spite of the extreme specialisation of the sexual
process, which (as Minchin, loc. cit., points out) is entirely
to facilitate sexual reproduction, rendered difficult by the
intra-cellular habitat and non-motile condition of the para-
sites, certain forms have found it profitable to associate pre-
cociously, thus insuring fertilisation and allowmg of an
economy to be effected in the number of (male) gametes.

We see, therefore that in this closely allied order no
support is to be found for the theory that the separate forma-
tion of anisogamous gametes is derived from a condition of
association and the formation of isogamous gametes.

One or two other points may be briefly noted. Anisogamy
is only known to a marked extent in intestinal forms. ‘The
ancestral Gregarine was, in all probability, an intestinal
form, from which the coelomic parasites have been derived.
As we have seen, the latter are modified by the greater

1 In the few cases where association is met with there is only one megaga-
mete formed as in the other Coccidia ; in these forms, moreover, and only in
these, we find few (often four) microgametes developed. Schaudinn’s work
on Cyclospora (see p. 71) tends to show, however, that the original condi-
tion in this order also was the production of many female gametes. If, there-
fore, association in Adelea mesnili, ete., was a primitive character, there
would be, in all probability, many female (and, ef course, male) gametes.

>

vot. 90, PART 1.—NEW SERIES. 6

82 H. M. WOODCOCK.

degree of parasitism attained. Brasil (loc. cit.) admits this
influence upon their trophic organisation, but thinks, appa-
rently, that the exact nature of the reproductive process is
uninfluenced by this feature. I do not think this view can
be maintained. If not directly, the influence can certainly
be traced indirectly. There is abundant evidence that modi-
fications in trophic life themselves influence the mode of
reproduction; examples of this have been numerous in the
preceding pages. Hence for this reason alone we ought to
be very cautious in assuming that the manner of reproduction
of these ccelomic Gregarines is primitive. Lastly, when we
remember that the Flagellate origin of the Gregarines is
almost certain, it seems most natural to regard a condition
where we have motile anisogametes as being more primi-
tive than the one we find in the Monocystids, which has most
likely resulted in the way I have described above.

Tabular comparison.—lIt may be useful to summarise,
in tabular form, the principal stages in this line of develop-
ment, as illustrated by known forms. Schaudinnella itself
is best left aside. ‘I'his is, undoubtedly, a very exceptional
parasite, primitive in some respects (e.g. in the character of
its association), but specialised in others. As Brasil says, it
cannot be very well compared with any known Gregarine.
In my opinion it represents a primitive Telosporidian parasite
which has endeavoured, as it were, to go in more than one
direction at once, combining, to a certain extent, both
Gregarine and Coccidian features.

(A) Intestinal forms.

(1) The gametes are highly differentiated. Example:
Pterocephalus.

(2) The conjugating elements, though readily distinguish-
able into male and female, are not so markedly differentiated :
certain very motile and spermatozoon-like male ones are
sterile, and have acquired a subsidiary function. Example:
Stylorhynchus.

LIFE-CYCLE OF “ CYSTOBIA”’ IRREGULARIS (MINCH.). 83

(3) The gametes are, to all appearance, perfectly similar
(isogamous). Examples: Gregarina, Lankesteria.

In all these cases there is permanent association between
ripe sporonts, followed by common encystment.

(4) The isogamous gametes (greatly reduced in number)
may develop parthenogenetically, the septum between the
two associates not breaking down. Hxample: Ophryocystis.

(? 5) The gametes may develop parthenogenetically, the
association being only temporary (Selenidium ?).

(s) Coelomic forms.

(1) The gametes are only very slightly differentiated, and
chiefly distinguishable by their nuclei. Permanent association
and common encystment are usual. Hxamples: Urospora,
Gonospora, and Monocystis. (This stage is easily deriv-
able from a 2.)

(2) The gametes are quite isogamous. Precocious asso-
ciation occurs. Diplocystis, Diplodina (probably also
Cystobia).

(? 3) The gametes may develop parthenogenetically, the
association being either permanent, in which case the septum
does not break down (Ceratospora ?), or else temporary,
the two associates afterwards sporulating separately (Gono -
spora varia ?).

DETAILED SUMMARY.

My investigations have related chiefly to Diplodina
(Cystobia) irregularis (Minch.), parasitic in Holothu-
ria forskali, and to D. (C.) minchinii Woodcock, from
Cucumaria pentactes and C. planci. All the material
was collected near Plymouth, in which neighbourhood these
Holothurians have a very localised distribution. he Cucu-
mari are scarce, and the percentage of infected individuals
is very small,

I have also examined the trophozoites of Diplocystis
schneideri, from a new host, Periplaneta orientalis.

84 H. M. WOODCOCK.

(3)! Diplodina irregularis lives either in the lumen of,
or attached to, the blood-vessels. There is no definite relation
between the growth, or period in the life-history, of the
animals and the time of their evagination of the wall of the
vessel. They may either come out when quite minute, or,
on the other hand, they may commence sporulation while still
in the lumen. ‘The parasites, whether as trophozoites or as
cysts, are never free in the coelome. The cysts are most
probably ruptured in situ, the lberated spores escaping
when some Cuverian organs are extruded.

The habitat of D. minchinii is very varied. The para-
sites are most numerous in the wall of the respiratory trees.
They also occur attached to the ccelomic epithelium of the
body-wall, of the retractor muscles, and of various more or
less vascular strands which cross the body-cavity, chiefly in
the hinder part. They are never in, or in any way related to,
the vascular system proper in connection with the gut, and
obviously do not reach the site of infection by way of the
mouth and intestine, as does the other species. All the
evidence points to the conclusion that the spores enter
the host through the cloacal aperture, being sucked up by the
inhalant current into the trees.

The parasites in the coelome are always partly covered by
a double layer of epithelium, the inner one being really an
invagination of the outer one. There is not the slghtest
doubt that the animals are passing in and not emerging.
The process is the reverse of the evagination process met with
in D. irregularis. The later stages constitute more an
overgrowing and surrounding of the parasite by the epithelium
and connective tissue of the host than an actual inpushing on
the part of the Gregarine itself, which only rarely occurs to
any extent. The parasite becomes at length completely
encysted and ready for sporulation. —

(4) The gregariniform adults of both species are perfectly
regular in form and typically ovoid. They are quite motion-
less. Hach adult is really a “couple,” Diplodina being a

' The numbers refer to the corresponding sections.

LIFE-CYCLE OF “ CYSTOBIA”’ IRREGULARIS (MINCH.). 85

neogamous Gregarine, or one in which precocious association
occurs. In D. irregularis the two associates are sometimes
separated by a distinct septum and sometimes not, this being
dependent upon the time of union. In either case, however,
the adult couple—when the union is completed—presents
superficially almost the aspect of a Monocystid Gregarine.
In D. minchinii there is never any septum, the union of
the two cytoplasms being intimate. In this species the
association is lateral, while in D.irregularis it is terminal.

The general appearance of Diplocystis schneideri
agrees with Kunstler’s description.

I have occasionally met with instances of triple association
in both species of Diplodina, and also in Diplocystis.

(5) There is a marked absence of differentiation in the
peripheral region of the body of Diplodina. The general
cytoplasm is limited by a delicate membrane, but there is no
appearance either of ectoplasm or of myocyte-fibrille. In
strong contrast is the firm and definite ectoplasmic layer
in Diplocystis schneideri. Moreover, in this parasite
the couple is always enclosed in an investing membrane,
comparable to an early-formed ectocyst, which is laid down
in the form of the couple and specially thickened around the
plane of junction.

The cytoplasm in Diplodina has a quite typical gregari-
noid structure. The nucleus possesses a distinct chromatic
reticulum, in which is slung or suspended a single karyosome.
With growth, the karyosome becomes very vacuolated, the
smaller vacuoles running together to form two or three large
ones, the contents of which are manifestly passed out into
the nucleoplasm. I regard this process, not as one of excretion,
but as a reinforcement of the chromatin of the nucleoplasm.

(6) The process of encystment in D. irregularis is very
slight. In the majority of cases there is no ectocyst-forma-
tion whatever, and the limiting membrane serves as a delicate
endocyst. The most obvious cyst membrane, indeed, is the
peritoneal epithelium enclosing the cyst, which persists
after the endocyst has broken down and disappeared. In

86 H. M. WOODCOCK.

D. minchinii the encystment prcecess is more typical, and
there is a well-marked ectocyst.

The nuclear changes at the commencement of sporulation
in D. irregularis are very important. The karyosome
becomes divided up in the nucleoplasm, and with successive
divisions of the sporont-nucleus the resulting fragments
(daughter-karyosomes) become apportioned out among the
daughter-nuclei. The karyosomatic fragments break down
still further, and ultimately become incorporated with the
chromatin of the nucleoplasm. I do not believe any nuclear
material is, at this period, eliminated.

(7) My observations tend to prove that the earliest nuclear
divisions are completely amitotic. The multiplying nuclei
become distinguishable into two kinds, germinal ones, which
subsequently form the sporoblast-nuclei, and large somatic
or sterile ones, which eventually become dissipated in the
surrounding cytoplasm. This process represents the nuclear
purification of D. irregularis. The germinal nuclei divide
by mitosis, and the attraction-spheres have very large and
apparent centrosomes. ‘These attraction-spheres can also
exist and divide independently, and are to be met with scat-
tered about in the cytoplasm. Some appear to come into
relation with the large sterile nuclei, and probably help to
bring about their disintegration.

(8) In both D. irregularis and D. minchinii, and
especially in the latter, the process of sporoblast-formation
is characterised by the remarkable extent to which the inter-
twining of the lobes and processes of the two associates is
carried. As a result of this practically all the cytoplasm is
utilised to form the gametes. ‘here is no cystal residue
(“gregarinoid soma’’) left over. ‘lhe primary sporoblasts
or gametes are quite simple and all alike morphologically,
and conjugation is, therefore, completely isogamous. I have
never observed the least movement in live cysts at this
period, and do not believe the so-called “ danse des sporo-
blasts ” occurs.

(9) Spore- and sporozoite-formation in D. irregularis

LIFE-CYCLE OF “‘ CYSTOBIA”’? IRREGULARIS (MINCH.). 87

follows the usual plan. To Minchin’s description of the spore
it may be added that there is a deeply-staining plug or cap
closing the mouth of the inner spore membrane at the funnel
end; this is doubtless dissolved by the digestive juices of the
fresh host, thus allowing the sporozoites to pass out.

The spore of D. minchinii is very similar in form to that
of D. irregularis, but only about half the size. The sporo-
cyst is very delicate, and is probably itself dissolved ; the fact
that I have not observed any plug at the base of the funnel
also points to this conclusion,

These two parasites are much more nearly related to each
other than is either to Cystobia holothuriez, and for this
reason have been placed in a distinct genus. Diplodina
is undoubtedly closely allied to Gonospora; the affinities of
Cystobia holothuriz, on the other hand, are rather with
Lithocytis and Urospora.

(10) The consideration of these neogamous Gregarines has
afforded a useful opportunity of discussing the phenomenon
of association as a whole.

Considering first the variations in time and manner of the
process, we see that precocious and intimate association has
been especially developed in the three genera Diplocystis,
Diplodina, and Cystobia. ‘These have proceeded along
rather different and independent lines, which reach their cul-
minating points in D. schneideri and Diplodina min-
chinii respectively. In the former genus the desired result
is attained by successive (phylogenetic) modifications of the
encystment process ; in the latter (and also in Cystobia) the
tendency is towards intrinsic cytoplasmic union between the
two associates.

Precocity of association is undoubtedly correlated with the
absence of movement in these ccelomic forms. It is simply
an endeavour to insure a suitable and durable association.
There is certainly nothing approaching true sexual conjuga-
tion, as yet, between the two members. Even where neogamy
is most intimate the nuclei remain, manifestly, quite distinct,
and I think the cytoplasm of each associate also retains its

88 H. M. WOODCOCK.

morphological separateness and individuality excepting in the
plane of union, although, from the nature of the case, this is
not apparent.

(11) Seeing the efforts made by these non-motile ccelomic
parasites to insure association, we are naturally led to try and
account for its obvious importance.

The result of recent research points strongly to the con-
clusion that the Telosporidia, as a whole, are descended from
a Flagellate ancestor. This ancestral Flagellate would possess
motile gametes, which were, probably, more or less aniso-
gamous. Such an ancestor we may regard as being common
to the three Telosporidian orders, namely the Gregarines, the
Coccidia, and the Heemosporidia, the differences exhibited by
them being easily understood when their different habitat and
degree of parasitism are borne in mind.

Of these orders the Gregarines are by far the most success-
ful, this being undoubtedly due to their possession of the power
of association. Hence, if this were a primitive condition, it
would almost certainly be apparent (as such) in the Coccidia,
since by it sexual conjugation is rendered practically certain.
The state of affairs in that order seems to show conclusively
that association is not a primitive, but rather an acquired, con-
dition, one which has been acquired to any extent only by the
Greearines.

I regard the power of cytotactic attraction as having
become so developed and specialised in this order that the
formation of gametes is now entirely regulated by and de-
pendent on such cytotactic influence; in other words, asso-
ciation is most probably necessary for sporulation to take
place. There is abundant evidence to show that, in most
forms at any rate, this is undoubtedly the case. Correlated
with this, differentiation of the gametes is no longer necessary,
and a beautiful series of stages exemplifying the gradual tran-
sition or reduction from anisogamy to isogamy is now known.

The above view seems to me much more probable than the
opposite one, namely that association is a primitive condition
tending to become less important with increasing anisogamy.

LIFE-CYCLE OF “ CYSTOBIA”’ IRREGUIARIS (MINCH.). 89

Tn this connection it is, I think, a significant fact that, speak-
ing generally, isogamy is most prevalentin the coelomic forms,
which are the most modified and which in certain instances
have developed neogamy.

GENERAL SUMMARY.

In conclusion, it may be said that Diplodina is a very
advanced and specialised Gregarine. Its principal modifica-
tions are those of non-motility, absence of cytoplasmic differ-
entiation, neogamy, and complete isogamy ; these are closely
correlated with one another, and are for the most part ulti-
mately traceable to the degree of parasitism attained by the
genus.

Diplodina and (also) Diplocystis are to be regarded as
forms which, following slightly different but parallel lines,
have pursued to the farthest extent what may be described
as the main or typical course of evolution of a ccelomic
Gregarine.

University Cotiece, Lonpon,
July, 1905.

Since the above was written only two or three papers
dealing with Gregarines have come under my observation.
There is only opportunity here to add a word or two with
reference to these, in so far as they bear upon the more
important points considered in my work.

Reference has been made to Brasil’s preliminary note on
the sexual reproduction of Monocystis. In ‘Arch. Zool.
exp.’ (4), vol. iv, p. 69, 1905, he describes the process in detail,
and it is apparent from his figures that the gametes resemble
those of Urospora and Gonospora. Besides the nuclear
differences between the two kinds of element, however, there
is also a slight inequality in size, one set (microgametes ?)
being rather smaller than the other (megagametes ?).

Schnitzler (‘ Arch. Protistenk.’, vol. vi, p. 309, 1905), in
describing the reproduction of Clepsydinia (Gregarina)

90 H. M. WOODCOCK.

ovala, confirms Paehler’s account with regard to the absolute
agreement in appearance between the conjugating elements.
He finds, moreover, that the process of nuclear maturation
described by that author (see footnote, p. 51) occurs equally
in both kinds. Hence in this form there is complete
isogamy. It is interesting to note that in Monocystis this
nuclear reduction is apparently delayed, Brasil having
observed it to take place in the zygotes. Another point
noticed by Schnitzler is the occurrence of two kinds of cyst,
containing respectively small and large spores (micro- and
mega-spores). As the author surmises, this fact may perhaps
stand in relation with the occurrence of solitary as well as
common encystment and sporulation.

Lastly, Crawley (‘Amer. Nat.,’ September, 1905, p. 607)
derives the Telosporidia (i.e. Coccidia, Hemosporidia, and
Monocystid Gregarines) from a Polycystid Gregarine. As
will be seen on referring to page 71 et seq. above, I do not
concur with this view. The ancestral Gregarine (derived
from a “ gregariniform” Flagellate) was probably not so
highly differentiated and specialised morphologically as the
Polycystids now are. ‘These are rather to be looked upon as
constituting one line of descent, the Monocystids another.
The Heemosporidia and Coccidia, again, have branched off in
other directions from the ancestral Flagellate.

January, 1906.

ANALYSIS OF CONTENTS.
Preface, p. 1.

(1) Introduction, p. 3.
(2) Methods.
(a) Examination, p. 6.
(6) Fixation and staining, p. 7.
(c) Attempts at artificial infection, p. 9.
(3) Habitat and mode of life.
(a) C. irregularis, p. ll.
Histology of the vascular network, p. 11.
Relation of the parasites to the blood-vessels, p. 12.

LIFE-CYCLE OF “ CYSTOBIA” IRREGULARIS (MINCH.). 91

(6) C. minchinil, p. 18.

Probable mode of infection, p. 14.

Relation of the parasites to the ccelomic epithelium, p. 14.

Situation in which the parasites encyst, p. 17.

Conclusions, p. 17.

(+) Form, size, and general appearance.
(a) Ce irrecularis:, p. 19:
(4) C. minchinil, p. 20.
(ec) Diplocystis schneideri, p. 22.
Triple association, p. 24.
(5) Minute structure.
(a) Nature of the peripheral region and composition of the septal
plane, p. 24.
Loss of the power of movement, p. 26.
Consideration of D. schneideri, p. 26.

The investing membrane, p. 27.
(4) General cytology, p. 28.

The cytoplasm, p. 28.

The nucleus, p. 29.
(c) Formation and probable function of the karyosome, p. ol.

Expulsion of karyosomatic material into the nucleoplasm, p. 82.

Significance of the process, p. 33.

(6) Commencement of sporulation.
(a) Eneystment, p. 34.
C. irregularis, p. 34.
C. minchinii, p. 36.
(6) Fragmentation of the karyosome, p. 37.
Nuclear incorporation of the fragments, p. 39.

(7) Nuclear multiplication.
(a) Nature of the early nuclear divisions, p. 40.
(6) Distinction of the multiplying nuclei into two classes, somatic and
germinal, and their further history, p. 42.
Nature and origin of the division-centres, p. 43.
Indirect nuclear division, p. 45.
Independent division of the centrosomes, p. 46.
Disintegration of the sterile nuclei, p. 47.
(8) Formation of gametes (sporoblasts) and conjugation, p. 48.
Segregation of the sporoblasts, p. 49.
The gametes, p. 50.
Conjugation, p. 51.
Absence of any movement, p. 52.
The actual union, p. 53.
Abnormalities, p. 54.

92 H. M. WOODCOCK.

(9) Spore-formation and systematic position.
(a) The spores, p. 55.
C. irregularis, p. 55.
C. minchinil, p. 57.
(4) Systematic position, p. 58.
Relationships, p. 60.
Definition, p. 60.

(10) Precocious association or neogamy.
(a) Comparative account of the principal variations in the manner and
time of association, p. 61.
Diplocystis, p. 62.
Diplodina and Cystobia, p. 63.
The morphological condition found in Diplodina is to be inter-
preted as one of union, and not of imperfect division, p. 65.
(6) Biological considerations, p. 67.
The reason for neogamy, p. 67.
Effect on the individuals, p. 67.
Is triple association successful ? p. 69.
(11) General significance of association.
Descent of the Telosporidia, p. 70.
Origin of association, p. 73.
Essential importance of the process, p. 74.
Relation between association and conjugation, p. 75.
Unusual and specially modified forms, p. 76.
Isogamy in the Gregarines is to be regarded as the more modified,
and not as the more primitive condition, p. 79.
Tabular comparison, p. 82.

Detailed Summary, p. 83.
General Summary, p. $9.

BipLioGRAPHY.

1. Bernot, A.—“ Beitrag zur Kenntniss der im Darme der Larve von
Tenebrio molitor lebenden Gregarinen,’,‘ Arch. Protistenk.,’ vol. i,
pp: 3875-420, Pls. XI-XILI, 1902.

2. Brasit, L.— Contribution a la connaisance de |’appareil digestive des
Annélides Polychétes; L’épithélium. intestinal de la Pectinaire,”’
‘Arch. Zool. exp.’ (4), vol. ii, pp. 91-255, Pls. IV-VIII, 1904.

10.

11.

12.

13.

LIFE-CYCLE OF ‘‘ CYSTOBIA”’ IRREGULARIS (MINCH.). 93

. Brasin, L.—“ Recherches sur la reproduction des Grégarines Mono-

cystidées,” op. cit., N. et R. (4), vol. iii, pp. 17-38, Pl. IT, 1905.
“La genése des gametes et l’anisogamie chez les Mono-
cystis du Lombric,” ‘C. R. Ac. Sci.,’ vol. exl, March 13th, 1905.

. Cautxery, M., and Mesni, F.—‘ Sur une mode particuliére de division

nucléaire chex les Grégarines,” ‘ Arch. Anat. Microse.,’ vol. iii, pp.
146-167, one plate, 1900.

—— “Sur quelques parasites internes des Annélides,” ‘ Misc.
Biol.’ (‘ Trav. Stat. Zool. Wiim.’), vol. ix, pp. 80-99, Pl. 1X, 1899.
“Sur une Gregarine. . . presentant. . . une phase
de multiplication asporulée,” ‘C. R. Ac. Sci., vol. exxvi, pp. 262-264,
1896.

. Ceccont, G.—* De la sporulation de la Monocystis agilis, St.”

‘Arch. Anat. Microsc.,’ vol. v, pp. 122-140, Pl. V, 1902.

. CrawLey, H.—‘*‘ Progressive movement of Gregarines,” ‘P. Ac. Philad.,’

vol. liv, pp. 4-20, Pls. I-II, 1902.

Curnot, L.—* Recherches sur lévolution et la conjugaison des
Grégarines,” ‘Arch. Biol.,’? vol. xvii, pp. 581-652, Pls. XVILI-
XXI, 1902.

Doruetn, F.—‘ Die Protozoen als Parasiten und Krankheitserreger,”’
(G. Fischer, Jena.). 1901.

Hartoc. M.—‘‘Some Problems of Reproduction—II,” ‘Quart. Journ.
Mier. Sci.,’ vol. xlvii, pp. 583-608, 1903.

Hertwic, R.—“ Die Protozoen und die Zelltheorie,”’ ‘Arch. Protistenk.,’
vol. i, pp. 1-40, 1902.

13a. —-— ‘Ueber Kerntheilung, Richtungsk6rperbildung und Befruch-

14.

15.

16.

iT:

18.

19.

tung von Actinosphaerium eichhorni,” ‘Abh. bayer. Ak.’ (2)
vol. xix, pp. 1-104, 1898. Hight plates.

“Ueber Wesen und Bedeutung der Befruchtung,” ‘S. B. Ak.

Munel.,’ pp. 57-73, 1902.

KoEHLER, —.—[On synonymy of different species of Holothuria],
‘Zool. Anz.’ vol. xx, pp. 507-509, 1897.

KunstiER, J.—‘‘ Diplocystis schneideri,” n.g., u.sp., ‘Tabl. Zool,’
vol. ii, pp. 42, Pl. 1X, 1892.

LasBsBE, A.—Article on “‘Sporozoa”’ in ‘ Das Thierreich,’ 5te Lief, Berlin,
1899.

Laveray, A., and Mesnit, F.—*‘Sur quelques particularités de l’évolu-
tion dune Grégarine et la reaction de la cellule-hote,” ‘C.R. Soe.
Biol.,’ vol. lii, pp. 554-557, 1900. Nine figs.

Licer, L.—“ Recherches sur les Grézarines,” ‘Tab!. Zool.,’ vol. iii,
pp. 1-182, 1892. ‘Twenty-two plates.

94.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.
33.

34.

35.

H. M. WOODCOCK.

Licer, L.—*‘ La reproduction sexuée chez les Ophryocystis,” ‘C. R.
Soe. Biol.,’ vol. lii, pp. 927-930, 1900.

“Sur quelques Cercomonadines nouvelles ou peu connues

parasites de l’intestin des Insectes,” ‘Arch. Protistenk.,’ vol. ii, pp.

180-189, 1903. Four figs.

‘*La reproduction sexuée chez les Stylorhynchus,’
vol. iii, pp. 304-367, Pls. XILL and XLV, 1904.

-———- and Dusoscg, O.—“ La reproduction sexuée chez Ptero-
cephalus,” ‘Arch. Zool. exp. N. et R. (4), vol. i, pp. 141-151
1908. . Eleven figs.

Lupwice, H.—Article on “ Echinodermata” in Bronn’s ‘ Klassen und
Ordnungen des Thierreichs’ (‘ Holothuroidea,’ vol. ii, pt. 3.)

Mincuin, E. A.—‘‘ Observations on the Gregarines of Holothurians,”
‘Quart. Journ. Micr, Sci.,’ vol. xxxiv, pp. 279-310, Pls. XX VII and
XXVIII, 1893.

Mivycuiy, E. A.—Article on ‘‘Sporozoa”’ in Laukester’s ‘Treatise on
Zoology,’ vol. i, pt. 2, pp. 150-360, 127 figs., 19038.

Mineazzini, P.—“ Le Gregarine delle Oloturie,” ‘Rend. Acc. Line.’ (4),
vol. vii, pp, 312-319, 1891.

Nuspaum, J.—‘‘ Ueber die geschlechtliche heterogame Fortpflanzung
einer im Darmkanale von Henlea leptodera Vejd. schmarotzenden

Gregarine—Schaudinnella henlee, mihi,” ‘ Zeitsch. wiss. Zool.,’
vol. Ixxv, pp. 281-3807, Pl. XXII, 1903.

Paruter, F.—‘‘ Ueber die Morphologie, Fortpflanzung und Entwickelung
von Gregarina ovata,” ‘ Arch Protistenk.’ vol. iv, pp. 64-87, Pls.
V and VI, 1904.

Perez, C.—“ Le cycle évolutif de 1?Adelea mesnili, coccidie celo-
mique parasite d’un Lepidoptere,’ op. cit., vol. ii, pp. 1-12, Pl. I,
1903.

Prowazex, S.—‘Zur Entwickelung der Gregarinen,” op. cit., vol. i,
pp. 297-305, Pl. IX, 1902.

Sars, M.—‘‘ Oversigt af Norges Echinodermer,” Christiania, 1861.

op. cit.,

Scuaupiyy, I'.—‘‘ Generations- und ,Wirthswechsel bei Trypanosoma
und Spirochete,” ‘Arb. a. d. kaiserl. Gesundheitsamte,’ vol. xx,
pp. 387-439, 20 figs., 1904.

Scuaupinn, F.—‘‘ Studien ueber krankheitserregende Protozoen. I.
Cyclospora caryolytica, Schaud,” op. cit. vol. xviii, pp. 378-
416, Pls. XII-XI1I, 1902.

ScunEIpDER, A.—(Various memoirs on * Gregarines nouvelles ou peu
connues’’), ‘ Tabl. Zool.,’ vols, 1 and 2, Poictiers, 1886—1892.

LIFE-CYCLE OF “ CYSTOBIA” IRREGULARIS (MINCH.). 95

‘36. Stepteckr, M.— Ueber die geschlechtliche Vermehrung der Mono-
cystis ascidie,” ‘ Bull. Ac. Cracovie,’ 1900, pp. 515-537, Pls. I
and IT.

37. - “Etude cytologique de la Coccidie de la Seiche,” ‘ Ann. Inst.
Past.,’ vol. xii, pp. 799-836, three plates, 1898.
38. ‘Etude cytologique de Adelea ovata, Schn.,” Op. cit.,

vol. xiii, pp. 169-192, Pls. I—III, 1899.

39. Witson, E.—‘‘The Cell in development and inheritance,” ‘Columb.
Univ. Biol. Ser.,’ vol. iv, 1900.

40. Wooncock, H. M.—“On Cystobia irregularis (Minch.) and allied,
‘neogamous’ Gregarines” (preliminary note), ‘Arch. Zool. exp.,’
N. et. R. (4) vol. ii, pp. 125-128, 1904.

DESCRIPTION OF PLATES 1--6,

Illustrating Dr. H. M. Woodcock’s paper on “'The Life-Cycle
of ‘ Cystobia’ irregularis (Minch.), together with Observa-
tions on other ‘ Neogamous’ Gregarines.”

All figures with a high magnification were drawn as the sections appeared
under either the Zeiss apochrom. 1°40 N.A. 3 mm. lens, or the similar 2 mm.
one, with different compensating eyepieces. All figures, excepting fig. 6, were
drawn with the aid of the camera lucida.

REFERENCE LETTers.

a. or an. Loose wandering cells, ameebocytes, etc. a. g. Albuminoid
granules (lenticular grains). 0/7. or d/.v. Blood-vessel. c., cl. Layers of con-
nective tissue. c¢. a. Nuclei: of cellular aggregation. ¢. e. or c. ep. Ceelomic
epithelium. c. 2. Nuclei of same. c. w. Cyst wall. cypl. Cytoplasm.
e. Outer epithelial layer. e!. Inner invaginated one. ect. Ectocyst. ex.
Endocyst (or endospore). ep. Attaching epithelium. ev. Exospore. f. Frag-
ments of karyosome. fl. Fluid contents of lumen, / J. Fibrous layer of tissue.
G. Gregarine. g. Albuminoid grains or lumps. 7. m. Common or investing
membrane. k. Karyosome. /./!. Limits of cup-like investment. lam. (or l.). Lumen
or cavity. DZ. M. Longitudinal muscle. l.m. Limiting membrane. m. or m. l,
Muscle layers. mes. Mesentery attaching gut. N. Nucleus (of parasite). . 2.
Nuclei of epithelial layer. x. of c. w. Nuclei of cyst-wall (vascular epithelium).
p. Tip of protrusion. p.g. Paraglycogen grains. 7. g. Minute refractile granules,
ret. Reticulum. s. Septum. s. 7. Sporal residuum. sp. Spores. sp. c.
Spongy cells (loose tissue). sp. x. Sporoblast nuclei. s¢. or A. s¢. Stalk of

96 H. M. WOODCOCK.

attachment (epithelial). ¢. Tongue of investing membrane. 1’. M. Trans-
verse muscles. vac. Vacuoles. v. s. Vascular strands. z. Special zone of
the nucleoplasm.

PLATE 1.

Mode of life and general appearance, “Cystobia” irregularis and
*C.” minchinii.

Fic. 1.—C. irregularis in the blood-vessel of Holothuria forskali;
outline drawn alive, the details filled in after mounting and staining. (The
nuclei of the wall rather diagrammatic). Osmic, picro-carmine. x 80.

Fic. 2.—Ditto, obtained free of the blood-vessel; drawn alive. x 100.

Fic. 3.—Ditto, younger stage. Mounted whole. Osmic, picro-carminc.
x 150.

Fic. 4.—Portion of diverticulum of respiratory tree of Cucumaria pen-
tactes, showing seven C. minchinii zz stéw~; viewed whole. (In two cases two
of the parasites overlap, but they are not actually in contact.) Flemming,
para-carmine. X 80.

Fic. 5.—Two young evaginated couples of C. irregularis (#. is the
second nucleus beginning to be cut through, ep. epithelial sac in which the
parasite lies), (A) Seen whole; corr. subl. + acetic, para-carmine X 80;
(B) in section; corr. subl. + acetic, iron-hem. + orange. xX 250.

Fic. 6.—(A) Portion of body-wall of Cucumaria pentactes, showing
two C. minchinii attached. x 38. (B) The upper parasite of (a). x 12.

Fic. 7.—Portions of the wall (a) of a large distributing vessel, and (b) of a
small cross-connection in H. forskali. Fron a section. Corr. subl. +
acetic, iron-hem. + orange. X 400.

Fic. 8.—Portion of a section through the wall of a vessel in the “ rete

mirabile.” The actual diameter of the lumen is much less, but it is less
obstructed by spongy tissue. Corr. subl. + acetic, iron-hem. + orange.
+ 400.

Fics. 9 and 10.—Adults of C. minchinii attached to the coelomic epi-
thelium, which they have invaginated; (4) shows where this is reflexed
internally, (a) being the part of the parasite still uncovered. Vig. 9 seen in
optical section; corr. subl. + ac., carm.-alum; Fig. 10, as a transparency ;
Flemming, para-carmine. Both x 150.

Fic. 11.—Portion of retractor muscle, showing three C. minchinii com-
pletely encysted. ‘The small one at (a) is most imbedded. x 80.

PLATE 2.
Mode of life and general appearance (continued). C. minchinii. (All
figures except fig. 20 relate to this parasite.)

LIFE-CYCLE OF “ CYSTOBIA ’’ IRREGULARIS (MINCH.). 97

Fig. 12.—Inpushing adult, to show especially the bending inwards of the
invaginated epithelium. (The cytoplasm has shrunk away a little from the
limiting membrane.) Flemming, picro-carmine. xX 250.

Iie, 13.—Ditto, with a protrusion (w) penetrating into a vascular strand.
(Nuclei of parasite are rather shrunk.) 90 % alc., para-carmine. x 150.

Fre. 14.—Two sections of preceding. (a) Along line c—p; (b) along
A—B. In the former the parasite is everywhere enveloped by the double layer
of epithelium. Thionin + eosin. x 150.

Fig. 15.—Section of parasite encysted in connective tissue around retractor
muscle. Corr. subl. + acetic, Klein. ham. x 80.

Fic. 16.—Section of another one, more imbedded. The inner layer of
epithelial nuclei distinctly seen at a. Flemming, iron-hem. x 120.

Fie. 17.—Section through part of an invaginated parasite (p.), showing the
peculiar amoebocytes with peripherally situated nucleus (z.) in the lumen of
the stalk. Tron-hem. + orange. xX 700.

Fires. 18 and 19.—Sections through parasites in the wall of the respiratory
tree. Flemming, iron hem. + orange. xX 325.

Fic. 20.—Evaginated C. irregularis; mounted whole. ‘Triple associa-

tion. (The cytoplasm is rather retracted at one side.) Osmie + picro-carmine.
x 50.

PLATE 3.

Structure of Diplocystis schneideri. Triple association in Diplo-
cystis and “Cystobia”; precocious association.’

Fig. 21.—Diplocystis schneideri, viewed whole. Flemming. x 50.

Fie. 22.—A section through a couple. Corr. subl. + acetic, iron-hem. +
orange. xX 50.

Fie. 23.—Another whole specimen. The endoplasm has shrunk away from
the outer layers (see text). Corr. subl. + acetic. x 50.

Fie. 24.—Ditto, triple association. The septal partitions are seen in two
places. Flemming. x 50.

Fre, 25.—Another triplet, the members of which are not quite so intimately
joined. Corr. subl. + acetic. x 59.

Fie. 26.—Nucleus of D. schneideri (froma section). m. Nuclear mem-
brane. Corr. subl. + acetic, Klein. hem. x 200.

Fie. 27.—Portion of the periphery of two couples of D. schneideri,
showing the plane of junction (see text). sp. Space. (a) Klein. hem., (b)
para-carmine, after corr. subl. + acetic. x 600.

Fig. 28.—C. minchinii. Triple association. (The outer, free part of
the parasite is rather irregular.) Flemming, para-carmine. x 120.

VoL. 50, PART 1.—NEW SERIES, 7

98 H. M. WOODCOCK.

Fic. 29.—C. irregularis. An evaginated triplet. This parasite has a
somewhat altered, unhealthy look. Osmic, picro-carmine. Xx 150.

Fie. 30.—The smallest example of C. irregularis observed; already eva-
ginated. The cytoplasm is scanty and without any granules. Corr, subl.
+ acetic, iron-hem. + orange. xX 1000,

Fies. 31 and 32.—Ditto, of C. minchinii. Both in the wall of the

respiratory tree. 2. Nuclei of tissue-cells. Flemming, iron-hem. + orange.
x 1000.

PLATE 4.

Minute structure. Encystment and sporulation,

Fic. 33.—Different nuclei of C. irregularis to show the different
degrees in the vacuolisation of the karyosome (seen whole). Osmic, picro-
carmine. X 400.

Fic. 34.—Section through the nucleus of c. in the last figure (for descrip-
tion see text). Iron-hem. + eosin. xX 700.

Fic. 35.—Section through a typical nucleus (C. minchinii). Flemming
iron-hem. + orange. X 500.

Frc. 36.—Large nucleus and portion of surrounding cytoplasm of an
encysted C. minchinii. Klein. hem. + orange. x 400.

Fic. 37.—Portions of the periphery of two couples of C. irregularis, to
show the plane of junction (see text). Corr. subl. + acetic, iron hem. +
orange. X 600.

Fig. 38.—(a) Periphery of C. minchinii to show cytoplasmic detail.
(b) Portion of cytoplasm of C. irregularis. (a) 90 % alc., (b) corr. subl.
+ acetic; thionin + orange. x 600.

Fic. 39.—Section through sporulating C. irregularis from the lumen of
the blood-vessel. Corr. subl. + acetic, iron-hem. + orange. x 80.

Fie. 40.—Sporulating C. irregularis from same situation (seen whole).
More advanced stage. (The nuclei of this specimen were better stained than
usual, probably because it was not an “ encysted” parasite.) Corr. subl. +
acetic, para-carmine. X 100.

Fie. 41.—Section through encysted sporulating C. irregularis. N.and 2,
Large and small (sterile aud germinal) nuclei. Corr, subl. + acetic, iron-
hem. + eosin. X 250.

Fie. 42. Ditto. This parasite was in a swollen, “ spongy ” evagination
(see text), of which w is the wall. 2. Amcebocytes and blood-cells. Flemming,
iron-hem. + orange. X 120.

Fic. 43.—Entire sporulating cyst of C. minchinii, attached toa fibrous
vascular strand. ‘he cyst (seen in optical section) is full of ripe unstained
spores, Corr. subl, + acetic, para-carmine. X 85,

LIFE-CYCLE OF ‘* CYSTOBIA”’ [RREGULARIS (MINCH.). 99

Fic. 44.—Varying character of the encystment process in “ Cystobia”;
e is a section of C. minchinii, the rest being of C. irregularis (for
description see text). (WV. Nuclei of inner epithelial layer and of connective-
tissue layers.) Iron.-hem. + orange or eosin (except e, which is thionin +
orange). All x 850.

PLATE 5.

Nuclear changes and multiplication. (All figures are from sections, and all
relate to C. irregularis, except fig. 55, which is of C. minchinii.)

Fig. 45.—Commencing nuclear changes: 4. is the other nucleus of the
parasite. Corr. subl. + acetic, Klein. hem. + eosin. X 300.

Fic. 46.—Two nuclei of the 8-nuclear stage. Iron-hem. + orange.
x 1150.

Fic. 47.—(a) and (4) Daughter nuclei of the {4th generation; (¢) one
of the third generation dividing. Corr. subl. + acetic, iron-hem. + orange.
x 850.

Fic. 48.—Stages in fragmentation of the daughter-karyosomes. All nuclei
of the fourth generation. Corr. subl. + acetic, iron hem. + orange. xX 700.

Fics. 49 and 50.—Commencing distinction of the multiplying nuclei into
large sterile ones and small germinal ones.
(49) Thionin + orange

(50) Iron-hem. + eosin } after corr. subl. + acetic. x 850.

Fic. 51.—Later stage. a.b. Resting sexual nuclei. c¢.—/. Dividing ones.
g-—j. Centrosomic division. &. Sterile nucleus. (The arrows point to
figures constructed from two or three sections.) x 850.

Fig. 52.—Stages in the disintegration of the large sterile nuclei. Flemming,
iron-hem. + orange. xX 830.

Fie. 53.—(a) Equal distribution of the nuclei and commencing segregation
of the cytoplasm into sporoblasts. x 175. (4) Part of periphery of similar
cyst. Cytoplasm loose and full of chromatoid granules. Both corr, subl. +
acetic, iron-hem. + orange. X 850.

Fre. 54.—Cyst containing zygotes or copule, about stage c-d, fig. 58. No
cystal residue. Corr. subl. + acetic, para-carmine. X 175.

Fig. 55.—Earlier stage in C. minchinii, showing the intertwining pro-
cesses. (The connective tissue, etc., surrounding the cyst is not shown [see
fig. 44¢.]). Corr. subl. + acetic, thionin + orange. x 175.

Fic. 56.—Corresponding stage in C. irregularis. Owing to the sporo-
blast nuclei (z) being less numerous and less closely arranged, the outlines of
the segments are not quite so sharply defined, Corr. subl. + acetic, thionin
+ orange. X 175.

100 H. M. WOODCOCK.

PLATE 6.

Conjugation and spore-formation. (All figs. except fig. 62 relate to
C. irregularis.)

Fic. 57.—The isogametes (primary sporoblasts) and stages in their con-
jugation. From a crushed preparation. Flemming, para-carmine. X 600.

Fic. 58.—Fusion of the sexual nuclei in the zygote; condensation of the
definitive nucleus and change of shape of the copula. From sections. Corr.
subl. + acetic, para-carmine. X 600.

Fie. 59.—a, b. Formation of outer cyst-membrane (exospore); d-j. first
division of zygote-nucleus. a, 4. Whole from crushed preparations ; corr. subl.
+ acetic, para-carm.; c-j. from sections (spore seen obliquely), corr. subl. +
‘acetic, iron-hem. + orange. X 600.

(The remaining figs. are from sections.)

Fic. 60.—Spores with two nuclei; in 4 only one is entirely in plane of
section. Corr. subl. + acetic, iron-hem. + orange. x 600.

Fic. 61.—Fully formed spores with four nuclei. (¢. Cup or plug, closing
mouth of inner funnel.) Corr. subl. + acetic, iron-hem. + orange.
x $850.

Fic. 62.—(a) Ripe spores of C. minchinii, with eight sporozoites.
(b) Two sporozoites. Corr, subl. + acetic, iron-hem. + orange. X 850.

Fie. 63.—(a) Degenerating primary sporoblasts; (4) degenerating copule ;
(c) polygamous copula. Flemming, para-carm. x 600.

Fic. 64.—Polynuclear masses from sporulating cysts. (a) About time of
conjugation; (4) during spore-formation. Corr, subl. + acetic, iron-hem.
+ orange. X 600.

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 101

The Anatomy of Oncholaimus vulgaris, Bast.,
with Notes on two Parasitic Nematodes.

By

F. H. Stewart, M.A., B.Se., M.B.,
Lieut. Indian Medical Service.

With Plates 7—9.

I pecan this work with the intention of trying to ascertain
if any light could be thrown on the comparative morphology
of the ccelom and Nephridia by the study of Nematodes. The
uncertainty as to the true nature of the body-spaces in this
class is well known. No one has been sufficiently daring to
identify any space in Nematodes with the ccelom. ‘The space
between the body wall and the gut is generally admitted to
be a schizcele.

Ray Lankester (11, p. 9) writes: ‘‘In some few groups of
Coelomoccela the cceloms have remained small, and limited to
the character of simple gonoccels. This seems to be the case
in the Nemertina, the Planarians, and other Platyhelmia.”
The Nematodes are not mentioned, since nothing is definitely
known in regard to them.

Thanks largely to the work of Jagerskidld (9), we know
that the excretory organs of Nematodes are unicellular, with
intra-cellular canals, and, physiologically at least, interchange-
able with skin glands, but they have not yet been homologised
with Nephridia. Ray Lankester, in the same article (11,
p. 14), writes: “True Nephridia are only found in the

VOL. 50, PART 1,—NEW SERIES. 8

102 F. H. STEWART.

Platyhelmia, Nemertines, Rotifers, Chaetopods, and embryonic
Mollusca.”

In all text-books the Nematodes are placed in a very
isolated position, and no attempt is made to compare them
with other groups. This, I believe, is largely due to the
fact that most work has been done on parasitic forms, and
that these have been held up as the types of the group. But
even although the great majority of a group were parasitic,
it is to the free-living minority we should go for the type,
and not to the secondarily modified majority.

I therefore selected for study a free form Oncholaimus
vulgaris (Bast.), a parasitic form Ascaris clavata (Rud.)
to compare with it, and an embryo of a parasite to connect
the two, if possible.

No member of the genus Oncholaimus has been investi-
gated by modern methods. I therefore have gone thoroughly
into its anatomy.

In regard to A. clavata the alimentary and excretory
systems have been fully described by Jagerskidld (9), I have,
therefore, limited my work to the reproductive system, while
in the embryo I have confined myself to the excretory organs.

The specimens were all obtained at St. Andrews, and a
large part of the practical work was done at the Gatty
Marine Laboratory there. My thanks are due to Professor
McIntosh for the use of the laboratory and for his kind
assistance in many other respects.

OncnoLtaimus VuuLGaris (Bast.)}.

Measurements:

Male. . Length . 10—15 mm.
4, Breadth(max.) ‘185—'221 mm,
Female . Length . 12—15 mm.
Breadth . ‘18—:225 mm.

9

The body is elongated and cylindrical, tapering very
slightly and slowly to the anterior extremity (PI. 7, figs. 1

' Bastian (2).

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 103

and 2), which is square in profile, rapidly to the posterior,
which forms a rounded and rather blunt cone (PI. 7, fig. 5).

The animal is quite translucent ; the only pigment present
is in the intestinal wall, golden-brown granules arranged in
patterns, which produce a tesselated appearance, and are
sometimes sufficiently abundant to cause the animal to
appear of a brown colour, even to the naked eye.

The mouth is terminal, and is surrounded by three very
flat papillee (Pl. 7, fig. 2, a).

Immediately behind the latter is a ring of short sete
(ibid., 6), while smaller hairs are scattered over the anterior
portion of the body, and arranged along the dorsal (c) and
ventral lines.

The anus (Pl. 7, fig. 5, cl. ap.) is subterminal, 153 mm.
from the tail, in the midventral line.

The cellular character of the longitudinal lines (PI. 7, fig.
1, mp1, Lc, MvL) and the striation of the muscle fields (Pl. 7,
fig. 5) are obvious. Prominent objects also are the cup-
shaped pharynx, with its three teeth (Pl. 7, fig. 2); the ceso-
phagus (Pl. 7, fig. 1); surrounded by the cesophageal
ring and collar of ganglion-cells (PI. 7, fig. 1); the pigmented
intestine; the hyaline ducts of the tail glands filling up the
post-anal region (PI. 7, fig. 5); the ventral gland in the male,
opening ‘112 mm. in front of the nerve-ring; and the large
cells of the body space which generally lie in the submedian
lines (Pl. 7, fig.-1, a).

The gonads occupy a large part of the body in either sex,
but are more conspicuous in the female; the large chalk-
white ova (text-fig. 2, ov., p. 124) in the uterus can be distin-
guished even with the naked eye. Under a low power the
central vulva(vw.) surrounded by the cellular mass of vulvar
and vaginal glands and the elongated ovaries (ovr.) running
alongside the uteri can be seen.

In the male the gonads appear as a fairly uniform cellular
cylinder opening posteriorly in common with the gut. Here
also are to be found the sabre-shaped spicules (PI. 7, fig. 5
sp.) with their central accessory piece.

]

104 F. H. STEWART.

It is convenient in describing the animal to divide the body
according to the regions of the alimentary canal. I shall,
therefore, frequently refer to pharyngeal, cesophageal, intes-
tinal, rectal, and post-anal regions. The cesophageal region
it is convenient to divide into anterior cesophageal in front of
the nerve-ring, posterior cesophageal behind it.

Habitat.—Oncholaimus vulgaris is very common
under stones between tide-marks. It is essentially a sociable
animal, twenty to thirty being often found together under one
stone; it is not usual to find individuals isolated.

THe CurTicLe.

The cuticle forms a continuous layer over the whole body,
as in all Nematodes, and passes in through all the apertures,
mouth, anus, cloacal opening, and vulva to become continuous
with the cuticular linings of the cavities into which these
apertures open.

It is a structureless membrane ‘00215 mm. thick, and with
certain stains shows a division into two layers (Pl. 7, fig. 15),
Hairs occur in certain localities: (1) in a circle round the
head at the base of the lips; (2) scattered over the anterior
cesophageal region, and (3) arranged along the mid-dorsal
and mid-ventral lines (Pl. 7, figs. 1 and 2). They are more
abundant in the dorsal than in the ventral line, and in the
cesophageal than in other regions of the body. The numbers
vary considerably in different individuals, but an average
would give about forty for the dorsal and twenty for the
ventral line. In the female several occur in front of and
behind the vulva.

Each hair is formed by a projection of cuticle, while at the
base the cuticle is perforated by a minute canal containing a
core of protoplasm passing out from the epidermis. The hairs
along the median lines spring from the centre of shallow de-
pressions in the cuticle.

The canals at the roots are, as pointed out by Jigerskidld
(10), identical with the integumental pores of Bastian (2).

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 105

In prepared specimens the hairs are very frequently broken
off.

The cuticle is also perforated by the openings of the
ventral and tail glands (PI. 7, fig. 5, tg.ap.) and of the glands
of the lateral lines.

From the base of the lips four flattened pouches (PI. 7,
fig. 2, e) extend outward and backward in the substance of
the cuticle as far as the oral circlet of hairs. They are semi-
circular in shape, the base directed toward the mouth, the
arc away from it. At the base of the lips the pouches are
continuous with each other, and here the four pore-like open-
ings to the exterior are situated.

The pouches contain a number of coarse granules with
amphophil staining reaction. I have not been able to trace
any connection between these granules and the epidermal
protoplasm.

From the nature of the contents I believe the function of
these structures to be glandular. They appear to correspond,
not with the “Seitenorgane” of other forms, but with the
circumoral “ patterns” and glands described by De Man in
Osfuse ws (12).

Toe EpmpERMIs AND NERVOUS SYSTEM.

It is hardly possible to separate these two. The only
structure which is definitely specialised as nervous is the
circumcesophageal ring’; other structures which are nervous
in function, such as the ganglionic mass which surrounds the
ring and the anal ganglion, shade off into the general
epidermis. I shall, however, describe the latter first.

The epidermis consists of four lines of cells—the longi-
tudinal lines (PI. 7, fig. 3, MDL, MVL, LL), which run from one
end of the body to the other, and which project into and
divide the muscular layer of the body wall, and of a thin
layer of protoplasm—the subcuticula or hypodermis (PI. 7,
fig. 3, sc.), which connects these four lines, and hes between
the cuticle and the muscular layer. ‘This consists merely of

~ 106 F. H. STEWART.

an outgrowth of protoplasm from the cells of the longitudinal
lines and contains no nuclei. The longitudinal lines are,
in fact, situations where the epidermal nuclei are aggre-
gated, and where the nutrition and general government of the
entire epidermis is carried on,

I have described the lines as cellular, but although cell-
walls do occur, they are often not complete, the protoplasm of
one cell being at one point continuous with that of another,
or several nuclei may appear in one mass of protoplasm, and
as no cell-walls are to be found in the subcuticula, all the
cells whose protoplasm is coutinuous therewith are presumably
in this way continuous with one another, although this layer
is so thin that divisions might exist in it but be undemon-
strable.

The four longitudinal lines lie, as usual, one mid-dorsal,
one mid-ventral, and two lateral, dividing a transverse section
of the body-wall into four quadrants. They extend, as I
have stated, from one end of the body to the other. Their
absolute and relative breadth varies in the different regions.
In the cesophageal regions all four lines are of fairly equal
breadth, and occupy each about one fourteenth of the cir-
cumference, each showing about four cells in transverse
section.

In the intestinal region (Pl. 8, fig. 24) the median lines
decrease in size, showing only one or two cells in section ; the
lateral lines, on the other hand, are considerably broader and
occupy each about one eighth of the circumference. Opposite
the gonads the lines are very much flattened from pressure.
Behind the anus they are of approximately equal breadth,
and each occupies one eighth of the circumference.

Two types of cells occur—viz. (1) cells whose shape varies
from square to oval, and whose greatest diameter varies from
00086 to 0°0107 mm. The protoplasm is fairly abundant and
is non-granular. The nucleus, spherical or oval, measures
from 0°0043 to 0°:00538 mm., and has a well-marked nuclear
membrane containing numerous minute chromatin granules
and a nucleolus (Pl. 7, fig. 9). (2) Rather smaller cells,

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 107

0:0043 mm., varying in shape, but more generally circular
in sections, the protoplasm scanty, nucleus spherical or
oval, 0°0032 mm., and staining diffusely with basic dyes
(Pl. 7, fig. 8, 7). In addition the lateral lines contain
gland-cells, but these will be described separately. Type 1
is most common.

Where there is not much pressure from the bulk of internal
organs, as in the cesophageal and post-anal regions, the cells
project into the space between the body-wall and gut. Oppo-
site the gonads, however, they are very much flattened. In
the posterior cesophageal region Type 1 cells often show
vacuolation ; while in the median lines in the intestinal region
they are generally triangular on section, the apex directed to
the cuticle, and the protoplasm shows fibrillee passing in from
the subcuticula to surround the nucleus.

The submedian lines (Pl. 7, fig. 18, s.m.l.) are not epi-
dermal, but are merely mesodermal partitions between groups
of muscle-cells.

Jaigerskidld (10) describes epidermal sub-median lines in
neighbourhood of the nerve-ring in Cylicolaimus magnus;
nothing corresponding is present in our animal. In Thora-
costoma acuticaudatum he mentions sub-median lines,
but does not make it clear whether they are cellular. The
subcuticula is an excessively fine sheet of protoplasm ; in fact,
in places it is so fine that it is impossible to demonstrate it. It
is continuous with protoplasm of the cells at the margin of the
longitudinal lines.

De Man (12) in reference to the genus Oncholaimus and
more particularly to the species O. fuscus, describes the
subcuticula as richly cellular and granular. Neither of these
statements applies to O. vulgaris.

I have used the term ‘‘ epidermis” freely. I see no reason
why this structure in Nematodes should be veiled under
terms such as “ hypodermis,” etc., Wandolleck (15), Jammes
(7), and Hamann (4), are agreed as to its ectodermal origin.
The only authority who maintains its meso-dermal nature is
Zur Strassen (14).

108 F. H. STEWART.

The nervous system consists of the cireumcesophageal
ring and ganglionic collar and of the ganglion in the wall of
the rectum or cloaca. These are the only structures which
are in any way separated from the general epidermis.

The nerve-ring (Pl. 7, fig. 4, 7) hes at the junction of the
first and second thirds of the cesophagus, and consists of fine
fibrillee united into a bundle. It contains a few nuclei, 4 to 6,
which resemble those of the Type 1 cells of the epiderm, but
are somewhat smaller. A sheath of fine connective tissue
derived from the fibrillar stroma which here fills the space
between the body-wall and the gut encloses the ring. The
sheath stains more deeply than the nerve-fibrils. Processes
which in places can be shown to be hollow pass off from the
sheath, and join the connective tissue surrounding the cells of
the collar and the muscle-cells. I have in a few instances
found nerve-fibrils in these hollow processes, but they are so
fine that it is not possible to follow them for any distance.
They probably connect the ring with the collar-cells and
the longitudinal lines.

The circumcesophageal collar (Pl. 7, figs. 1 and 3) extends
from nearly the commencement of the cesophagus to a short
distance behind the nerve-ring, and is composed of cells grow-
ing in from the longitudinal lines. It completely fills the space
between the body-wall and csophagus. In preparations of
the entire animal it can be clearly seen as a compact cylinder
of cells.

The cells are of three types, viz. cells of the same character
as 'l'ypes 1 and 2 of the longitudinal lines, and in addition
large oval cells (Pl. 7, fig. 3, bc!) ‘016 mm. in diameter, with
abundant protoplasm which may show very faint, irregular,
basophile markings, the nucleus ‘0064 mm. with a definite
nuclear membrane, finely granular acidophil contents, the
chromatin aggregated into a large spherical pseudo nucleolus.
These cells are, I believe, identical with the basophil cells
which occur in the space between the body-wall and the gut in
other regions of the body; they differ from them, however,
in some points, and as I can only speculate in regard to their

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 109

functions, I shall discuss them under the section dealing with
this space and its contents, rather than include them in the
nervous system, to which they may not belong.

Yet another type of cell, the coarsely granular acidophil,
occurs among the collar-cells as well as elsewhere, but is
certainly not nervous.

The fibrillar groundwork binds together all these various
units.

De Man (12) for the genus Oncholaimus describes
numerous cells lying in the body cavity in front of and behind
the nerve-ring.

Bastian (2) also mentions their presence.

Jagerskidld (10) for Cylicolaimus magnus and Thora-
costoma acuticaudatum describes and figures what is
evidently an cesophageal collar; the cells shown in his draw-
ings are, however, not so numerous as those in our subject.
He frankly states that he has not examined the matter fully,
and he does not connect them in any way with the nervous
system, but considers them to be the same as the phagocytic
“piischelformige organ” described by himself, Nassonow
(13), and others in parasitic forms. This opinion, which
I believe to be an error, accounts for the fact that he also
considers them to be identical with the “ floating gland-cells”
of Bastian and the “ fat cells” of De Man. These cells occur
isolated in the “ body cavity ” through all regions of the body,
and are entirely different from this localised compact struc-
ture.

In parasitic Nematodes also a cellular investment to the
nerve-ring is found; e. g. Hamann describes and figures it in
Lecanocephalus (5) as connected with all four longitudinal
lines, and the same holds good for the embryo (A. capsu-
laria) which I shall describe later.

The anal and cloacal ganglia (Pl. 7, fig. 6, ag.) are also
formed by ingrowths of cells from all four longitudinal lines.
Shortly before reaching the level of the anus in the female,
or of the cloacal opening in the male, cells begin to project
inwards from the dorsal and the two lateral lines, passing

110 F, H. STEWART.

towards the dorsal wall of the anal canal or cloaca. Here
they form a layer one cell deep, lying on the cuticular lining
of the space, and continuous at the sides with the protoplasmic
wall. Since the latter is continuous with the cells of the ventral
line, this line may also be presumed to take part in the forma-
tion of the ganglion. The cells are of Type 1.

The only sensory organs to be found are the hairs above
described. They are, as above noted, specially aggregated on
the head and in front of and behind the genital apertures.

To sum up, the nervous system is very imperfectly differ-
entiated. The circumcesophageal ring and collar form the brain
of the animal; the longitudinal lines, and possibly the sub-
cuticula as well, form the conducting paths, both motor and
sensory, in the latter capacity receiving stimuli from the
sensory hairs. I have not found processes from the muscle-
cells to the lines such as occur in other forms; the motor
mechanism is, therefore, obscure. ‘The anal and cloacal
ganglia doubtless control defeecation and copulation.

Tur ExcrETORY AND GLANDULAR APPARATUS.

Three sets of glands are included in this system :

(1) The single excretory ventral gland.

(2) The series of glands of the lateral lines.

(3) The three tail glands.

These glands resemble each other in being all unicellular.
The ventral and tail glands have ducts of considerable length,
formed by outgrowths of the protoplasm of the cells; the
glands of the lateral lines have no ducts but open directly
through the cuticle.

The ventral gland! occurs in males and immature
females; it is absent in mature members of the latter sex.
It is composed of one large cell, which lies in the body cavity
immediately ventral to one of the lateral lines, in males on

1 Golowin (8) has described this gland in O. vulgaris. I have not been
able to procure a copy of his paper, which is, I presume, in Russian, He
regards the three “ Keimdriisen ” of the tail as of the same nature as this
gland.

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. IL11

the right side, in immature females on the left. The cell
body, which forms the secreting portion of the gland, is found
at a level a little behind the commencement of the intestine ;
from it a duct runs forward to the level of the nerve-ring, and
bending towards the midline, opens by a minute pore in the
median ventral line, ‘112 mm. in front of the ring.

The cell (Pl. 7, fig. 12) is of a flattened oval shape. The
nucleus is central, and contains one large pseudo-nucleolus.
The protoplasm is hollowed out by large vacuoles, which in
some of the specimens contain numerous large basophil
granules arranged round their periphery.

The duct (Pl. 7, fig. 4, vgd.; Pl. 7, fig. 13) is narrow and
cylindrical. It has a fine protoplasmic wall, which is appar-
ently an outgrowth from the gland-cell. The contents are
homogeneous, and stain with basic dyes. Near its termination
the wall of the duct becomes continuous with the cells of the
median ventral line, which separate, continuing the lumen to
the pore in the cuticle.

The glands of the lateral lines le in series at the
margins of the lateral lines. They are found as far forward
as the posterior limit of the cesophageal collar and as far
backward as the commencement of the rectum.

They consist of large, pear-shaped cells (Pl. 7, fig. 7),
with the pointed ends directed to the cuticle. The outlines
are sharply marked off from the other cells of the lateral lines.

Kach cell is filled with large rounded granules, with ampho-
phil staining reaction. ‘The nucleus lies toward the rounded
extremity, has a nuclear membrane, chromatin granules, and
true nucleolus. From the pointed end a minute duct leads
through the cuticle, filled with a hyaline substance which
frequently projects beyond the surface as a small spike. The
glands are identical with those described by Jagerskidld (10)
in Cylicolaimus magnus,

The tail glands are three in number. In preparations
of the entire animal they are exceedingly conspicuous as
hyaline, club-shaped masses, extending from the extremity
to some distance in front of the anus (Pl. 7, fig. 5, tgd.;

112 F. H. STEWART.

Pl. 7, fig. 6), the middle gland extending further forward
than the other two.

In the posterior intestinal region they lie ventral to the
intestine, in males in the grooves between the intestine and
the ductus. They pass on either side of the rectum, and
behind the level of the anus occupy almost the entire space
within the muscular wall. They open by a single pore on the
tip of the tail. On section, the protoplasm shows a highly
vacuolated appearance, the contents of the vacuoles staining
very uniformly, suggesting a colloid. The nucleus resembles
that of the ventral gland.

The duct has a very fine membranous wall; the contents
are basophil, sometimes acidophil. The three ducts remain
separate until a point immediately before the external pore.
There is, however, only a single such pore.

MosciLe or tHE Bopy WALL.

The muscular layer is the thickest part of the body wall.
The cells which compose it are arranged in eight longitudinal
fields, four on each side, dorsal, dorsolateral, ventrolateral,
and ventral (Pl. 7, fig. 3, pM, DLM, VLM, vM), separated from
each other by the four epidermal longitudinal lines and by the
four submedian lines. ‘lhe latter, as I have stated above, are
not epidermal in O, vulgaris, but are merely mesodermal
partitions.

The entire eight fields extend forward to the anterior
extremity ; only four, the two dorsolateral and the two ventro-
lateral, reach the posterior, the two dorsal and the two ventral
ending at the level of the anus. In transverse section the
fields at the different levels show on the average the following
number of cells:

Buccal Region. @sophageal. Intestinal. Post Anal.
Dorsal : ats Og 2 : 12) 0
Dorsolateral . ae 7 : oo -. 7
_Ventrolateral . ars. 7 : py. 6
Ventral oe Bead 6 . 6 0

The muscle-cells are of the usual Nematode type. I have

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 113

not been able to detect any of those fibre-like projections
from the undifferentiated portions of the cells to the longi-
tudinal lines which occur in other forms, and no doubt act as
nerves.

THe Bopy Space.

The interval between the body-wall and the gut is, in
sections, found to be occupied by a substance the characters
of which vary in different regions. The gonads, glands, etc.,
are imbedded in it.

As the subject is a somewhat contentious one, I shall first
describe the substance in question and then discuss its nature.

It extends through almost the entire length of the body;
the only regions in which it is impossible to prove its presence
are at the level of the pharynx and behind the anus. It
naturally varies in development according to the space to be
filled, is most abundant around the termination of the
cesophagus (Pl. 7, fig. 18) and around the rectum, where
a considerable interval occurs between the alimentary tract
and the body wall. It is also fairly abundant where a large
organ such as the ovarian ceecum or testis ends.

In that part of the cesophageal region which lies behind
the nerve-collar it forms a fibrillar network (PI. 7, figs. 8, 10,
and 13 m.) The fibrillee are tortuous, but their general direc-
tion is from the muscular layer inward to the cesophagus,
In places the meshes between the fibrils are circular, as if
they had been occupied by globules of some substance. Over
the outer surface of the esophagus the fibrils interlace, forming
an irregular membrane, while the same occurs over the internal
surface of the cells of the body-wall, muscular and epidermal, the
interwoven fibrils applying themselves closely to these cells,
or passing in between the muscle fields at the submedian
lines, and to a lesser extent between individual muscle-cells.
At the submedian lines, indeed, they reach the epidermis.

They stain intensely with nigrosin, and also take up eosin,
but with less avidity.

Nuclei occur in the tissue (Pl. 7, fig. 8, mn.), but are not

114 F. H. STEWART.

common. ‘l'hey generally lie opposite the longitudinal lines,
but also occur opposite the muscle fields. T'hey resemble the
nuclei of the 'ype 2 epidermal cells, staining diffusely with
basic dyes, measure *00215 to °00522 mm. in diameter—that
is, rather smaller than the nuclei of I'ype 2. I have not been
able to find any protoplasm surrounding. ‘They are completely
isolated from the epidermis by the fibrils.

In the region of the collar, owing to the presence of the
ganglionic cells, the tissue is not so much developed (PI. 7,
figs. 3 and 4). Its characteristics are the same as already
described, but as it forms a stroma for the collar cells, the
general direction of the fibrils is rather parallel with the body-
surface than radial, since these cells are growing in from the
epidermal lines across the muscle fields.

In front of the collar the tissue is still less developed, but
other tissues, such as the body-wall, are also somewhat meagre
as they approach the pharyngeal region.

I believe that this fibrillar network has a very definite
function—viz. that of connecting the outer surface of the
cesophagus with the body-wall and affording a surface of
origin for the cesophageal muscle, so that when this contracts
the entire value of the contraction is devoted to widening the
lumen. In parasitic forms a similar surface is provided by
the thick cuticle which surrounds the cesophagus.

‘Throughout the greater part of the intestinal region, owing
to the close approximation of the gut and body-wall, and
‘to the presence of the gonads, the space is narrow. The
substance filling it is best studied in the interval, triangular
in cross section, between the gut, body-wall, and gonad tube,
or where the gonad tubes end, but it must not be imagined
that it is confined to these regions, since it forms a complete,
although narrow, cylindrical covering for the gut.

In sections through this region stained with thionin and
eosin (Pl. 7, figs. 14, 15) a dull-pink hyaline ground-work
occurs with a very fine, more intense pink, granulation. ‘The
very fine, obscure fibrillation which is found in the protoplasm
of the epidermis or of the gonad tube-wall is not present.

i. ee ie

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST, 115

In sections stained with nigrosin and in some stained with
eosin a few fibrilla can be detected. They are, however,
much finer than those of the vesophageal region, and stronely
resemble fibrin filaments,

Nuclei (ibid., mn.) occur in this matrix identical with the
nuclei in the cesophageal region—‘00215 to 00322 mm. in
diameter. Around them is sometimes a thin film of proto-
plasm which stains more distinctly than the matrix, but
which has not a sharp line of demarcation from it. It can
be demonstrated that these nuclei have no connection with
either epidermal or muscular cells, with gut or gonads, or
with either of the types of cell which will be described later
which he in the body-space. They are often to be found
lying completely isolated and free in the matrix.

At the commencement of the intestine the cesophageal
fibrillar network does not end abruptly, but the fibrille
become gradually more and more scanty.

In this situation, and near the termination of the intestine,
muscular fibres traverse the space, running almost longitu-
dinally. They are passing very obliquely from the body-wall
to the gut.

The interval between the narrow rectum and the body-wall
is considerable, but here the connective tissue is largely
replaced by muscular fibres and by the ingrowth of epidermal
celis to form the anual ganglion. There is, however, a basis
of fine fibrils.

In the post-anal region also the epidermal cells project
inward to such an extent that it is difficult to say how much
of the fine tissue surrounding the duct of the tail glands is
epidermal and how much, if any, is mesodermal.

Two types of cell occur imbedded in this substance:

(1) The coarsely granular acidophil cell (Pl. 7, fig. 3),
oval in outline, maximum diameter ‘01076 mm. with a dis-
tinct cellular membrane. The protoplasm does not stain, but
contains numerous large spherical acidophil granules. The
nucleus is central, and stains rather diffusely, although it
shows some granulation (chromatin). These cells occur most

116 F. H. STEWART.

frequently opposite the epidermal and submedian lines, but
also opposite the muscle fields. ‘They are to be found
throughout the entire length of the body with the exception
of the pharyngeal region.

(2) Rather flattened cells (Pl. 7, fig. 10, be*) generally
crescent-shaped in transverse sections of the animal, owing to
compression between the gut and the body-wall. Maximum
diameter, ‘(02152 mm. The protoplasm contains a basophil
substance. ‘lhe nucleus is more or less spherical, has a nuclear
membrane, and finely granular acidophil contents. The chro-
matin is aggregated into a large spherical pseudo-nucleolus.

These are the cells referred to above as possibly identical
with the third type of cell in the collar. They differ from this
type, however, in the fact that the basophil markings in the pro-
toplasm are much more extensive and constant, and that the
material filling the nucleusis definitely acidophil, not amphophil.

They occur most commonly opposite the submedian lines,
but also opposite the muscle fields. Commencing at the
posterior limit of the collar, they are found as far back as
the rectum, As to function, they do not appear to be phago-
cytic; they are possibly nervous.

The only previous reference to an organised tissue filling
the body space in Nematodes is on the part of von Linstow
(11 a), who describes a “ plasma cylinder” surrounding the
eut in Thoracostoma denticaudatum (Schn.) and Eno-
plus edentatus (v. Linst.) ‘von dem dorsal und ventral je
zwel Leisten ausstrahlen.” His figures, however, show no
nuclei in this “cylinder.” Jagerskidld (10) for Cylicolaimus
magnus writes: “Hie und da findet sich zwischen dem
Darme und der Muskulatur eine ganz homogene, bisweilen
sich stirkfarbende Schicht. Ich kann sie nur als eine Art
coagulierter Flussigkeit deuten, denn alle Kerne fehlen.
Vielleicht ist es sogar nicht allzu gewagt zu vermuten dass
diese Fliissigkeit, die, wie angedeutet wurde, nicht itiberall
oder gar immer vorhanden ist, nur als folge kraftigen Ergrei-
fens mit der Pincette oder dergleichen kleinerer Verletzungen
exsudiert worden. Denn dass die Thiere normalerweise

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 117

eine freistrOmende Leibes fliissigkeit zeigen, habe ich, wie
schon erwahnt, nie bemerkt.” He suggests that v. Linstow’s
“plasma cylinder” is such a coagulated fluid, the “leisten ”’
the submedian lines. As mentioned above, however, he
describes the fibrillar stroma supporting the collar, but he
apparently has not observed that it extends beyond the limits
of the cellular collar—if it does soin Cylicolaimus magnus,
or that there are any nuclei present apart from those of the
collar cells. He does not attach any morphological import-
ance to it.

To commence, then, I may repeat that there is no doubt
that the body cavity in the cesophageal region is occupied by
a nucleated fibrillar stroma.

The nature of the substance filling the rest of the body
cavity is debatable.

There are three views which may be taken in regard to it—

(1) That it is a pathological exudate coagulated by the
fixing fluid.

(2) That it is a physiological body cavity fluid, coagulated
in the same way, and containing cells.

(3) That it isa mesenchyme tissue of rather low organisation.

(1) This is the view taken by Jagerskidld, and his reasons
for it have been quoted above. But in dealing with speci-
mens preparing for sectioning I have never, until they were
securely fixed, used any coarser instrument than a camel-hair
brush, so that I can see no reason why a pathological exudate
should be present.

It is, of course, not at all necessary to suppose that C.
magnus and O. vulgaris are identical in this respect; but
as they resemble each other very markedly in other points, it
would be natural to expect that they should also resemble each
other in this highly important morphological point.

Jaigerskidld states that the substance in question is only
present in certain localities, and that it is non-nucleated. I
have found it present in almost every region of the body and
continuous, and it contains nuclei proper to itself.

(2 and 3) The substance when stained with eosin has a

voL. 50, PART ].—NEW SERIES. 9

118 F. H. STEWART.

certain resemblance to a coagulated albuminous fluid, but it
should be contrasted with, e. g., the coelomic fluid of annelids,
as seen in section. The latter, if stained with nigrosin,
appears as an exceedingly loose reticulum of very fine fibrils,
which do not stain very intensely; the former, in some
places, showsa very fine fibrillation, with a homogeneous back-
ground, in others stains intensely and evenly. I would not,
however, lay much stress on this, since I know of no test
which would enable us with the microscope to differentiate a
coagulum from a lowly organised jelly. On the other hand,
a coagulum formed in a fluid ought to contract, and not com-
pletely fill the space occupied by the fluid; the substance in
question completely fills the space, while the thickness of the
walls lining the space forbids the idea that they might have
contracted on a loose coagulum.

The presence of nuclei and their character is the strongest
argument in favour of the third view. Any nuclei occurring
in a fluid must belong to floating cells. Such cells would
probably be amoeboid, and would almost certainly have a
reasonable amount of protoplasm and definite cell outlines.
The nuclei which I have found, on the contrary, are either
entirely naked or have a very fine pellicle of protoplasm,
which shades off into the surrounding matrix. This, I think,
suggests that they are connective-tissue nuclei rather than
the nuclei of wandering cells. ‘l'hey are also identical with
the nuclei of the stroma in the cesophageal region.

For these reasons I would put forward the view that the
body cavity is filled by a mesenchyne, which in the ceso-
phageal region takes the form of a nucleated fibrillar net-
work in the regions behind the cesophagus of a jelly-like
substance, also nucleated. ‘The reason for the difference in
character of the tissue in the cesophageal and other regions
is, that in the former it has a special function, which I have
explained above. At the same time, I put this view forward
only tentatively. It would be rash to speak dogmatically on
so important a point without having thoroughly investigated
a number of allied forms.

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 119

ALIMENTARY SYSTEM.

The alimentary system is divided into pharynx or buccal
cavity, cesophagus, intestine, rectum, and cloaca in the male,
anal canal in the female.

The pharynx (Pl. 7, fig. 2, ph.) is cup-shaped, narrowing
at its hinder extremity ‘085 mm. long, ‘034 mm. in diameter.

The mouth is surrounded by the diaphragm-like ring of the
lips. In life the lips are in constant motion.

The cavity of the pharynx is shamrock-shaped in transverse
section. It has a cuticular lining, and from this lining there
project into it three large teeth (ibid., d.) composed of the
same substance, one lying in the dorsal line, the other two
subventral. At the tips of these teeth are situated the
openings of the cesophageal glands. Outside the cuticle is a
fine membrane connected with the cells of the longitudinal
lines.

The cesophagus (PI. 7, fig. 1, oes.) is 1:3 mm. in length.
Its general shape is that of an heraldic club. In diameter it
measures, at its commencement, ‘0525 mm., at the level of
the nerve-ring the same, and at its broadest part posteriorly
0862 mm.

The walls are thick and muscular, the direction of the
fibres radial. When at rest the internal surfaces are in
apposition and the lumen appears in cross section tri-
radiate—one radius in the midventral line, the other two
subdorsal. There is a cuticular lining considerably finer
than that of the pharynx.

The organ is essentially a strong suction pump, since the
radial muscle-fibres in contracting must open out the lumen
with considerable force. ‘This appears a somewhat anomalous
organ for a tree-living form.

The cesophageal glands occur in the posterior quarter of
the organ, as canals ramifying in the muscular substance.
They unite to form the ducts, which run forward and open
into the pharynx at the tips of the teeth.

120 F. H. STEWART.

The only evidence of a secreting epithelium is to be found
in the presence of cells with finely granular protoplasm in the
lumen of the ramifying canals. The lumen also contains
numerous sharp spherical granules which stain intensely with
acid dyes.

The ducts (Pl. 7, figs. 8-4, oes. d.) also lie in the muscular
wall, alternating with the radii of the cesophageal lumen, one
dorsal, two sub-ventral. They are elliptical in cross section,
and have a very fine protoplasmic liniug, They also contain
the same granular material. .

On leaving the esophagus, the ducts lie along the pharynx,
external to the cuticle, between two layers of the membrane
referred to above. This membrane also supplies them with a
linmg as they pass up the centre of the teeth to their
openings.

The intestine is about 11°5 mm. in length, cylindrical, except
where it is compressed by the reproductive organs. In the
living animal it has a tesselated appearance from the patterns
formed by numerous golden-brown globules contained in its
wall.

The wall is formed of columnar epithelium, from twenty-four
to thirty cells occurring in a transverse section. The cells vary
in depth from 0°015 mm. near the commencement, to 0°008
mm. near the termination. Their protoplasm does not stain,
and contains the golden-brown globules referred to above.
In certain of the cells numerous coarse acidophil granules
occur, strongly resembling those of the coarsely granular
acidophil cells of the body space, but, on the average, slightly
smaller. The cells containing these granules are distributed
in what appears to be an entirely capricious manner, gener-
ally wedged in between cells entirely free from granules.

I attempted to discover if there was any relation between
these cells and the granular cells of the body space, but was
unable to find any. ‘I'he two kinds of cells do not occur in
apposition, or even, as a rule, in close proximity, and there
is no correspondence between the frequency of occurrence of
the two in different animals or in different regions of the same

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 121

animal. As to the nature of the granules there is no clear
evidence. If they represented digested food in the process
of absorption, a more uniform distribution should occur.
They are probably either a digestive secretion produced in
cells which, although specialised, show no other signs of
specialisation, or excretory. There is no sign of a basement
membrane external to the epithelium.

The contents are very meagre, apparently portions of small
Algz, but one fact is very striking—viz. the presence in
the adult female of large masses of spermatozoa, which have
found their way in through the gonenteric canals.

The rectum is a short tube leading from the intestine to the
anal canal or cloaca. Its width, including walls, is 0°016 mm.
The wall has an outer layer of circular muscle-fibre and an
internal epithelium, while ganglion-cells which have grown in
from the longitudinal lines are to be found on the dorsal
surface. It terminates abruptly on joing the cloaca or anal
canal. The former I shall describe with the male repro-
ductive organs; the latter is a very short invagination of
epidermis with cuticle, in the dorsal wall of which the anal
ganglion lies.

Mate RepropuctivE ORGANS.

The male reproductive organs (text-fig. 1) consist of two
testes (at., pt.), an anterior and a posterior, lying in two
cavities, the testicular regions of the reproductive tubes
(atr., ptr.). The testicular regions open into a single ductus
ejaculatorius (de.), which unites with the rectum to form
the cloaca (cl.). The anterior testicular region runs in a
straight line backward into the ductus ; the posterior, on the
other hand, lies along the side of this structure, and it is
at its anterior extremity that it opens into it.

The symmetry, it may be assumed, was originally bilateral,
the reproductive tract consisting of a pair of cavities with a
common duct, and that, as the result of the narrow form of
the body, one of the cavities has been turned through an angle

129 F. H. STEWART.

of 180°. The entire system lies ventral to the alimentary
canal.

Each testicular region with its contents forms an elongated
mass, tapering at either extremity. In cross sections the
outline varies from semicircular to triangular in adaptation
to the other internal organs. The anterior mass is the larger ;
where it attains its greatest breadth it occupies as much as
two thirds of the body cavity.

In preparations of the entire animal the richly cellular
character of the testis and the mass of developing sperms can
be readily distinguished.

On examination by serial sections, the testicular region is
found to consist of a tube with epithelial walls, the commence-
ment occupied and closed by the germinal syncytium, the ter-
mination becoming continuous with the epithelial lining of the
ductus.

The epithelium of the wall is, throughout the greater part
of the length, very fine and flattened (Pl. 7, fig. 17, gw.).
It is only possible to demonstrate the protoplasm in places
where the contained sperms have been artificially separated,
but the flattened nuclei (ib., ngw.) are always readily de-
tected. Toward the junction with the ductus the epithelium
increases in depth, forming a well-marked layer still without
cell limits, but with fairly large oval nuclei. A fine layer of
muscular fibre appears on the outside. This portion is analo-
gous or, possibly, homologous with the vas deferens described
by Jagerskidld (10) in Cylicolaimus magnus and Thora-
costoma acuticaudatum.

The germinal syncytium (Pl. 7, fig. 16) which occupies
the fundus consists of a mass of nucleated protoplasm, which
is in continuity with the protoplasm of the wall: in other
words, it is a specialised portion of the epithelium lining the
gonocel. From this syncytium the sperm mother-cells are
developed, growing down and filling the lumen of the tube,
multiplying and undergoing development as they pass down.

The nuclei of the syncytium at the commencement are
spherical, or oval, ‘0043 mm.in greatest diameter, have clearly

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 128

marked chromatin granules, and one or two pseudo-nucleoli.
On passing further from the fundus, the nuclei increase in
size, the protoplasm becomes relatively less, then cell outlines
appear; the nuclei have increased to ‘00645-00753 mm., and
the chromatin takes the form of a network (Pl. 7, fig. 17,
sg.). About the commencement of the last quarter of the
region the sperm mother-cells begin to divide, the chromatin
is aggregated in larger granules, the nuclear membrane dis-
appears ; still further on the chromatin takes on a star-shape.
The cells (mature spermatozoa) are here more numerous and
smaller, ‘0043 mm. in diameter, are not so closely packed,
and are surrounded by some residual protoplasm. Spermato-
zoa are found in the anterior portion of the ductus. ‘The rest
of this organ is generally found to be empty.

The ductus ejaculatorius is a straight cylindrical tube,
although at its commencement slightly flattened. ‘The pos-
terior testicular region opens into it a short distance behind
its origin from the anterior.

Its wall consists of two layers—internal epithelial, external
muscular (Pl. 7, fig. 17, pg.). ‘The epithelium is cubical,
becoming columnar towards the termination of the tube. 'The
cells are highly vacuolated, and the free surfaces present a
frayed-out appearance, as if a secretion had been discharged.
In places this secretion can be made out as numerous acido-
phil granules.

The muscular wall consists at first only of a layer of circular
fibres. Flattened nuclei with a fine film of protoplasm can
be made out on the outer surface, no doubt the nuclei of this
layer. In the posterior quarter there is also an external layer
of longitudinal fibres.

At its termination the ductus unites with the rectum to
form the cloaca (text-fig. 1, cl.). As this latter structure is
intimately connected with the copulatory apparatus, I shall
describe it here rather than under “The Alimentary System.”

It is a short chamber, at its commencement roughly
cylindrical in section, but becoming flattened towards its
external opening, which is slit-like.

124 F, H. STEWART.

The wall (Pl. 7, fig. 6, cl.) is continuous with the epithelium
of the ductus and rectum, and with the epidermis in the
ventral line. It consists of protoplasm, which shows a circular
striation as if it might fulfil a muscular function. Nuclei are

Text-Fic. 1. TEXT-FIG.2.

fairly numerous. Cells from the dorsal and the two lateral
lines extend inward through the body-space towards the
cloaca and form a layer of cubical epithelium on its dorsal
wall, the cloacal ganglion (ibid., cl., ag.). There is a cuticular
lining continuous with the external cuticle.

The accessory sexual apparatus consists of the two spicules

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 125

(Pl. 7, fig. 5, sp.) with their accessory piece, with the
musculature which governs them and assists in expelling
the sperm. Lying dorsal to the cloaca a muscular mass
(Pl. 7, fig. 6, ps.) is to be found, imbedded in which are the
spicules and accessory piece. The fibres run in the same
direction as the long axis of the body and parallel with the
spicules. In front they envelop the anterior ends of the
spicules, behind they are attached to the wall of the cloaca.
This muscle obviously acts as protractor of the spicules.

A second set of muscles (ibid., rs) passes from the body-
wall to the posterior portion of the ductus, to the cloaca, and
to the protractor muscle. It consists of a series of muscular
bundles lying in the coronal plane of the animal and on the
side of the body-wall attached mainly to the lateral lines.
By enclosing the posterior portion of the ductus it assists in
expelling the sperm; by the attachment to the protractor it is
enabled to act as retractor of the spicules, and by its attach-
ment to the cloacal wall as dilator of the cloaca.

The spicules are curved, pointed at their free ends, °13 mm.
long. They consist apparently of a chitinous material, are
hollow at their upper portions, containing protoplasm. The
accessory piece is triangular, grooved at the sides, the spicules
fitting into the grooves.

FremMALe REPRODUCTIVE SYSTEM.

The female reproductive organs (text-fig. 2) consist, as is
usual in Nematodes of two tubes, uniting before reaching the
external aperture. ‘The tubes do not lie side by side, but one
is found in front of, the other behind, the external aperture.
This aperture is situated in the middle of the mid-ventral
line of the body, the entire gonads being included in the
middle two quarters of the body.

Hach tube consists of ovarian region (text-fig. 2, ov. r.),
including an ovarian cecum (ov.c.), uterus (wt.), and vagina.
The ovarian cecum lies furthest from the external aperture,
the rest of the ovarian region being bent back on the oviduct.

126 F. H. STEWART.

The first part of the ovarian region (ov. r.)—that is, that
part which is not ovarian cecum—has the form of an
elongated cone, the base being continuous with the ovarian
cecum. It measures 1:87 mm. in length. It is completely
filled by the ovary and by the mass of developing ova.

The ovarian cecum, or second part of the ovarian region,
is a blind prolongation of the first part. The openings into
it of the first part and of the uterus le side by side. It is
oval in shape, with a truncated end at its junction with the
first part and the uterus. It measures *85 mm. in length,
‘221 mm. in breadth.

The ovarian ceecum is the original gonad cavity, as will be
shown when describing the immature female organs. The
first part is a secondary outgrowth from it.

The wall of the ovarian region (PI. 8, figs. 19 and 20) is
formed of excessively fine flat epithelium. Indeed, so fine
is this epithelium, that except near the termination of the
organ, the only evidence of its existence consists in the pre-
sence of very much flattened nuclei (ngw.) closely apposed to
the sides of the ovary and of the column of ova. The pro-
toplasm of this epithelium cannot be distinguished even with
a magnification of 1000 diameters. ‘I'he state of matters in
the immature female which I shall describe later, and the fact
that this layer of flattened nuclei can be traced into the
epithelium lining the ovarian czecum, leave no doubt, however,
as to the real existence of a wall.

The wall of the ovarian ceecum (PI. 8, fig. 18, gw.) is com-
posed of flattened epithelium, in which no cell outlines are
discernible. The nucleiare flattened oval. The transition from
the first part to the czecum is, of course, gradual, not abrupt.

The germinal syncytium (Pl. 8, fig. 20, gs.) occupies the
fundus of the tube, and doubtless springs from the epithelial
lining. It consists of a protoplasmic mass, which shows an
affinity for basicstains. Nuclei are imbedded irregularly in it, at
the commencement 4 to 8 in a transverse section, increasing
up to 12, and then again decreasing until at the end of the
syncytium not more than two occur in any section. The

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 127.

nuclei are oval, have a nuclear membrane, distinct chromatin
granules, and nucleolus. At the commencement of the
syncytium they measure ‘0136 mm. in their greatest diameter,
at the termination ‘(0477 mm. ‘There is a narrow, more con-
densed, ring of protoplasm round each nucleus.

The syncytium occupies about one eighth of the length of
the ovarian region. It is followed by a column of disc-
shaped ova (PI. 8, fig. 19, ov.), which have become separated
from it. ‘These ova increase in size as they pass toward the
ovarian cecum, and become cylindrical rather than disc-
like in shape. About the middle of the column the ova
measures ‘0484 mm. in diameter.

The ovarian cecum contains about eight ova. A space is
left between the ova and the dorsal wall, through which the
ovum which occupies the fundus, and which is the most
mature, can pass to reach the oviduct. This mature ovum is
richly supplied with yolk-granules (Pl. 8, fig. 18, ov.).

The next division of the gonad tube is, in O. vulgaris,
physiologically merely an oviduct, but morphologically it is
identical with the subdivision which functions as an uterus and
receptaculum seminis in other free Nematodes—e. g. Cy lico-
laimus magnus and Thoracostoma acuticaudatum.
As this division has been described as an uterus by Jigers-
kidld, I shall retain the name for the sake of uniformity.
The uterus measures 2°56 mm. in length. Its calibre and shape
vary according to its contents. It may be distended by a
series of large ova or entirely empty. The latter condition is
the exception. Generally from four to twelve opaque white
ova can be seen in each uterus (text-fig. 2, ov.) lying end
to end like a short chain of beads. The uterus then adapts
itself, of course, to the shape of the ova. When empty it is
collapsed and flat.

At its commencement a glandular mass, ‘204 mm. in length,
projects into and fills the lumen. This is the shell-gland (text-
fig. 2, shg., Pl. 8, fig. 19, shg.). Ova, alter passing it, are
found to have acquired their shells. Its shape is very much
that of the ovarian cecum, resembling a conical bullet, the

128 F. H. STEWART.

pointed end directed towards the czecum. The base is concave,
and receives into its concavity one end of the first uterine
ovum. The gland is attached to the right wall of the uterus,
except at the base; here it projects freely into the lumen.

The gland is interesting histologically. It is formed of
protoplasm continuous with the wall of the uterus. Cell out-
lines are not present. At the apex the nuclei are large and
spherical, have a very promivent pseudo-nucleolus and a
nuclear membrane; at the base, however, the nuclear mem-
brane has disappeared, and the nuclear substance is diffused
into the protoplasm. Contiguous nuclei thus become con-
tinuous, and, as the nuclei are arranged at the periphery of
the mass, a basophil circle, in transverse sections, results. In
this basophil circle the pseudo-nucleoli stand out, and might
easily be mistaken for the nuclei themselves.

The wall of the uterus (PI. 8, fig. 20, wt.) for the greater
part of its length is composed of a cubical epithelium. This
is, of course, considerably flattened where the uterus is dis-
tended by the ova. Near the junction with the ovarian cecum
the outlines of the cells become indistinct. When, as some-
times happens, there is no ovum in contact with the base of
the shell-gland, the wall is here thick and the epithelium has
a glandular appearance.

The ova in the uterus (ib., ov.) are oval in shape and measure
‘27 mm. in length. The protoplasm is obscured by the mass
of yolk-granules. A single nucleus occurs in each ovum, and
this is invariably undergoing karyokinetic division, presum-
ably in preparation for the extrusion of the polar body, although
I have not observed this body. ‘he shell is ‘001 mm. thick
and is sculptured, narrow ridges running over its outer sur-
face.

Spermatozoa do not occur in the uterus.

The vagina (PI. 8, figs. 22, 28, 24, vag.) is a glandular and
muscular tube 595 mm. in length. The glands surrounding
it (vge.), with those which he around the vulva, give it a richly
cellular appearance in preparations of the entire animal. At
its origin from the uterus it is tightly constricted, but it soon

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 129

opens out although its lumen is not so wide as that of the
uterus. Shortly after its commencement, where the lumen
begins again to widen out, a narrow canal passes through its
dorsal wall and opens into the intestine. This is the gon-
enteric canal.

The wall of the vagina consists of three layers—an internal
epithelium, a middle muscular, and an outer glandular layer.

The epithelium is composed of cubical cells, the walls of
which are thick and consist of specially condensed protoplasm.
Within these walls the protoplasm does not stain, so that the
nucleus appears to lie in a vacuole. The muscular layer is
thick, the fibres circular. Where the two vaginze meet (PI. 8,
fig, 24) the fibres pass outwards to the body-wall, forming a
longitudinal layer around the short common terminal portion
of the tube. It is difficult to distinguish any definite epithelial
lining in this portion, but a fine film of cuticle is invaginated
through the external aperture. he aperture is surrounded by a
sphincter internal to the longitudinal fibres.

The vaginal and vulvar glands consist of cells lying in the
body space. Around the first half of the vagina they form a
single layer, and as they are pear-shaped give the appearance
of a rosette in transverse section. I have not been able to
demonstrate any openings from these cells into the lumen of
the vagina, but from the definiteness of their arrangement it
seems natural to suppose that their secretion is discharged
into the vagina. Beyond the middle of the vagina the cells
begin to arrange themselves around the vulva, and processes
pass from them to a circle of minute pores surrounding this
aperture (Pl. 8, fig. 24, vug. ap.).

The vagina always contains masses of spermatozoa (ibid,, s.).
Near the vulva these are spherical, with stellate nuclei, but
higher up at the uterine end they become elongated and the
nucleus almost thread-like.

Fertilisation must presumably take place in the vagina
during the passage of the ova. It is somewhat peculiar that
spermatozoa do not find their way up to a level at which the
ova are without their shell, as occurs in the forms described

130 F. H. STEWART.

by Jagerskidld. I have not detected any opening through
the shell, but such an opening would no doubt be very minute,
and it is difficult to obtain a complete series of sections of the
ova without some slight tearing of the tough shell.

The vulvar aperture is slit-like, the greatest diameter trans-
verse to the long axis of the body.

The gonenteric canal.—This is a minute duct which,
as stated above, opens into the vagina close to its com-
mencement (Pl. 8, fig. 23, gec.)—in fact, just below the
sphincter at the junction of uterus and vagina. It traverses
the dorsal wall, passing vertically through the muscular
layer. In this part of its course a few minute nuclei indi-
cate the presence of an exceedingly fine lining of epithelial
nature. Kmerging from the vaginal wall, it lies in the body-
space (Pl. 8, fig. 22, gec.) with nucleated protoplasmic walls
which conduct the canal to the gut in the midventral line
(Pl. 8, fig. 21, gec.) and becomes continuous with the ali-
mentary epithelium. Passing from the vagina to the gut,
the canal inclines slightly away from the mid-point of the
body—+.e. the anterior canal inclines slightly towards the
head, the posterior slightly towards the tail.

The function of this canal is to carry off superfluous sper-
matozoa. Spermatozoa are found in it and one of my sections
shows a sperm passing from it into the gut. The gut in the
adult female contains large masses of sperms.

No such canal has been described previously in any nema-
tode. Its comparative morphology will be discussed later.

REPRODUCTIVE ORGANS IN THE IMMATURE FEMALE.

I was fortunate enough to secure a specimen which was
full grown, but in which the reproductive organs were unde-
veloped. In this specimen the gonads occur in from one fifth
to one fourth of the length of the body. The vulva lies some-
what behind the centre.

They consist, as in the adult, of two tubes, an interior

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 131

and a posterior, uniting to open by a single aperture to
the outside. The tubes are straight, not bent back upon
themselves as in the adult, but, as will be explained further
on, the fundus does not correspond with the commencement
of the first part of the ovarian region in the adult, but with
the fundus of the ovarium cecum.

The tubes are only potentially tubes; towards the fundus,
where the oogonia occur, these cells completely fill the lumen;
in that part destined to become vagina the lumen is occupied
by a solid core of cuticle, and between these two regions the
“tubes” consist of solid rods of protoplasm, which show
their tubular nature only by the arrangement of the nuclei
in a single layer around the periphery.

If each tube were divided into five parts, the first part,
counting from the fundus, would contain the oogonia (PI. 8,
fig. 25). Here the wall (gw) is of flattened epithelium, not
showing cell outlines, the nuclei flattened oval, with chromatin
granules and nucleolus.

At the junction of the first and second fifths in the ventral
wall the germinal portion of the epithelium is situated. Here
the protoplasm of the wall grows into the lumen, the nuclei,
still retaiming the same characters, become larger and more
spherical. The oogonia (0g) as they develop from this proto-
plasmic projection pass, not towards the external aperture, but
towards the fundus. The youngest nuclei which can be
definitely recognised as oogonial resemble the nuclei of the
germinal projection, but are more oval, and are larger,
measuring 0:0016 mm. in diameter, and the chromatin granu-
les are smaller and fewer in number. A transverse section
at this point shows about eight such nuclei. The protoplasm
is scanty, shows no cell outlines, and stains with basic dyes.

In the later stage of development the nuclei have become
still larger, 0°0161 mm., and consequently there are fewer in
a transverse section. The chromatin granules have disap-
peared, the nuclear vesicle being filled by a finely granular
material which does not stain, the nucleolus is large. The
protoplasm is relatively greater in amount. The immature

132 KF. H. STEWART.

ovum which occupies the fundus has a nucleus of the same
character, but still larger, 0°0172 mm.

Comparing this with the adult, it is obvious that the pro-
jection of germinal epithelium corresponds with the com-
mencement of the ovary, and the fundus of the tube with the
ovarian cecum. In development to maturity the germinal
syncytium and the oogonial mass grow in the direction of the
external aperture, evaginating the epithelial wall, and forming
the reflected portion of the edult gonads.

Of the remaining four fifths of the tube, three fifths consist
of a solid protoplasmic rod, with nuclei arranged around
the periphery. In the last fifth the epithelium becomes
columnar (PI. 8, fig. 25). The cuticle of the body-wall passes
into the lumen of the gonads as a solid core, but extends
only a very short distance beyond the union of the two tubes.
Gonenteric canals do not occur.

The last quarter of each tube and the united portion are
surrounded by gland-cells (vgc.) lying in the body space.
Processes pass in the direction of the vulva, but I have not
been able to detect any openings through the cuticle.

ASCARIS CLAVATA (Ruvp.).

FEMALE REPRODUCTIVE ORGANS.

The female reproductive organs of A. clavata occupy one
half of the length of the body. If the body is divided into
ten parts, they extend from the end of the third to the end of
the eighth part. The external aperture lies at the end of the
third part.

They conform to the general Nematode type, consisting
of two germinal masses lying at the fundi of two tubes, the
two tubes uniting before reaching the external aperture.

Physiologically each tube may be divided into two regions —
viz. one in which the ova develop from the germinal syn-
cytium, and undergo maturation, the other strictly a passage

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST, 133

to the exterior, but in which fertilisation and a certain degree
of development also take place.

The first portion is, as usual, termed “ovary,” although this
term should strictly be applied only to the germinal syn-
cytium, and to the rhachis with the oogonia attached to it.

TeEXT-FIG. 3.

The second part is divided into oviduct, receptaculum seminis,
and uterus. The common portion is formed by the union of
the two tubes in the vagina.

The ovarian region (text-figs. 3, 4, 5, 6, ov. 7.) is by far the
largest; it extends from the end of the eighth to the tenth part

TExT-FIG. 4.

=e

of the body forward to the end of the fourth—i.e. it extends
through almost the entire reproductive region. It is, in addi-
tion, highly convoluted; in a specimen rendered transparent
with, e. g., cedar-wood oil the tightly-packed coils form a most
conspicuous object. The commencement is slightly conical,
the diameter increasing as we pass down, but after a compara-
tively short distance the diameter ceases to increase, the tube
being for the greatest portion of its length perfectly cylindrical.
The wall (Pl. 8, fig. 26.4, gw.) is excessively thin, consisting
VOL, 50, PART 1,—NEW SERIES. 10

134 F, H. STEWART.

of a layer of flattened epithehal cells. At the fundus this
layer becomes continuous with the germinal syncytium; in
other words, the germinal syncytium is a specialised portion
of the epithelium lning the gonoccel.

The germinal syncytium and the ova developing from it
extend from the fundus down the tube as an unbroken
column, completely filling the lumen, and this mass is, of
course, the true ovary. In the germinal syncytium numerous
nuclei occur, imbedded rather irregularly in the relatively
scanty protoplasm. As we pass down the tube the nuclei
begin to arrange themselves in a single layer at the periphery
of the column; the central, non-nucleated protoplasm is the
rhachis. Next, cell outlines appear around the nuclei, and
the rhachis becomes smaller with the increasing size of the
developing ova. Finally the rhachis ceases, the ova are
arranged in suecessive tiers, closely packed against each other
(Pl. 8, fig. 264, ov.) ‘The protoplasm of the ova at this point
consists of a meshwork. Before the ova becomes separated
from this continuous column and pass into the oviduct
eosinophil granules (yolk) make their appearance.

The division of the remaining portion of the genital tract
into regions is necessarily somewhat arbitrary. Hach region
shades off into the one preceding and the one following it,
and in different individuals the appearances at the same
level vary. The same portion morphologically may be empty
and constricted, or distended with ova or spermatozoa. In
the following description I have divided it, according to the
characters of the wall, into oviduct, receptaculum seminis,
and uterus; the vagina is the single terminal portion.

The oviduct and receptaculum (text-fig. 5, r.s.) follow the
ovarian region. They extend from about the middle of the body
backward to the end of the sixth to tenth, then turn forward
again for a short distance, and become continuous with the
uterus. ‘I'heir course is slightly tortuous, only slightly when
compared with the ovarian region. The appearance of the
tube varies with the contents ; at its commencement it con-
tains at most a few single ova and spermatozoa, and here it

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 1385

is fairly narrow, viz. °102 mm. This portion is the oviduct
(Pl. 8, fig. 27). Further on it widens out somewhat, to
become the receptaculum, measuring *255 mm. in diameter.
The wall of both oviduct and receptaculum consists of an
epithelium with an external basement membrane (ibid., bm.).
In the oviduct the epithelial cells are rounded, the lamina

TEXxtT-FIG. 5.

poorly marked. ‘The outer portions of the epithelial cells
show indistinct circular fibrillation, as if they might act as a
sphincter.

On passing to the receptaculum the lumen widens out, and

TrxT-FIG, 6.

the epithelia cells become less spherical, although they still
project into the lumen. ‘Their protoplasm is reticular. The
basement membrane becomes thick and distinct. he contents
consist of spermatozoa, which frequently form a single layer
on the surface of the epithelium, and of ova in small groups.
Fertilisation takes place here.

The uterus (text-fig. 6, ut.) extends from about the end of
the sixth tenth forward to the middle of the body, where it

136 F. H. STEWART.

unites with its fellow to form the vagina. It is the widest
and most distended portion of the tract. In a transparent
specimen the mass of ova which fills it is very prominent.
Its course is fairly straight, with the exception of an S-shaped
bend at one point, where it is constricted. Its wall (PI. 8,
fig. 28) is three-layered, consisting of epithelium and basement
membrane as in the oviduct and receptaculum, with the ad-
dition on the outside of a layer of flattened epithelioid cells.

The epithelium is more flat than in the oviduct. On the
internal surface the protoplasm is differentiated into a more
dense, almost cuticular, layer.

The outer epithelioid layer (ml.) is exceedingly fine, and it

Trxt-Fic. 7.

is difficult to demonstrate it except where the nuclei cause a
slight bulging. Its function I believe to be muscular. It is
continuous with the muscular layer of the vagina. Hamann
describes a similar layer in Lecanocephalus as muscular,

The uterus is always distended with masses of spermatozoa
and segmenting ova. Spermatozoa predominate towards the
commencement, ova toward the termination. Toward the com-
mencement the spermatozoa produce a remarkable appearance,
arranging themselves on the surface of the epithelium ina
closely-packed single layer, the nuclei lying at the free ends
(ibid.,s.). It is very easy to mistake them for a tall columnar
epithelium.

The vagina (text-fig. 7) extends from the junction of the
two uteri at the level of the middle of the body to the external

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 137

aperture which lies in the mid-ventral line at the end of the
third tenth of the body.

In preparations of the entire animal the appearance of this
portion varies according to whether it is distended or not.
At its commencement it contains large quantities of ova, and
is not to be distinguished from the uterus. Further on, how-
ever, it contracts, and forms a fairly thick walled muscular
tube, with here and there a single ovum in its lumen.

In sections this division is also clear, the upper portion
thin walled, the lower thick and muscular. Histologically,
indeed, for about the first quarter the wall is identical with
that of the uterus. ‘hen it begins to change, the epithelium
becomes more and more columnar, the basement membrane
more and more indistinct, merging on the one hand with
the outer wall of the epithelial cells and with the partitions
between them and on the other with the outer muscular layer.
This outer muscular layer is continuous with the outer epi-
thelioid layer of the uterus ; it consists of circular muscular
fibres.

Towards the termination there is also an internal cuticular
lining, continnous with the cuticle of the body wall.

The external aperture is puckered, and, if open, would be
circular. Around the aperture the epithelium of the vagina
becomes continuous with the subcuticular layer of epidermis.

Mate Repropuctive OrGans.

The male reproductive organs consist of a single tube, in
contrast to the double female tubes. In other respects
analogies, if not homologies, are easily demonstrable. The
germinal mass again is attached to the fundus, and projects
down the lumen. ‘The tube is divided into a portion in which
development and maturation of germs take place and into a
portion serving merely as a passage to the exterior, the
former consisting of a region containing the testis correspond-
ing with the ovarian region, the latter of a vas deferens cor-
responding with the oviduct, a seminal vessel corresponding

138 F. H. STEWART.

with the receptaculum and uterus, and a ductus ejaculatorius
corresponding with the vagina.

The testicular region occupies two thirds of the reproduc-
tive division of the body. It is, in addition, highly tortuous.
The wall is identical with that of the ovarian region, consist-
ing of an excessively flat epithelium (PI. 8, fig. 29, gw.).
The germinal syncytium again arises from this epithelium at
the fundus. The vesicular nuclei of the syncytium are at
first scattered through the protoplasm. They soon arrange
themselves around the rhachis, however, measuring ‘003 mm.
in diameter.

Further down, they invade the rhachis (Pl. 8, fig. 29),
running in lines through it (‘0045 mm.). Cell outlines begin
to appear in the protoplasm, leaving some residual protoplasm
between the cells. ‘he rhachis ceases, and the cells lie free
in the lumen, packed against each other. ‘lhe nuclei continue
to increase in size (0064 mm.), the protoplasm is scanty. In
the lower reaches the protoplasm again increases. The
spermatogonia divide by karyokinesis and form the spermato-
zoa, spherical bodies, ‘(0053 mm. in diameter, showing fine
amoeboid processes, the nucleus represented by a single
chromatin granule (Pl. 8, fig. 30, s.).

The vas deferens (Pl. 8, fig. 30, v.d.) is a short muscular
passage from the foregoing division to the seminal vesicle.
Its length is about ‘34 mm. It is exceedingly narrow,
‘068 mm. compared with ‘41 mm. of the vesicle. ‘The wall
consists of a cubical epithelium with an outer muscular layer.

The seminal vesicle (ibid., s.v.) extends through rather less
than one third of the reproductive region of the body. Its
course is straight. It is the widest portion of the genital
tract, ‘41 in diameter, the lumen distended with spermatozoa.
It is often constricted near the middle of its course. The
wall consists of an epithelial layer, which does not show cell-
outlines. Where the tube is distended the epithelium is
flat; where it is contracted the epithelium is thrown into
ridges. ‘lhe muscular layer consists of circular fibres.

‘The ductus ejaculatorius (Pl. 8, fig. 31) is the shorter,

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 139

terminal portion. Its lumen is narrow, but this is due to the
thickness of the walls, as the diameter over all measures
306 mm. The epithelium is the thickest layer; it is
columnar, and shows a remarkable intra-cellular structure.
The protoplasm is differentiated into two layers, an outer
which stains with hematoxylin, and an inner which stains
with eosin. The former projects into the body of the cell in
a fingerlike process which surrounds the nucleus.

The muscular Jayer is well developed, and consists, as usual,
of circular fibres. Muscular trabeculae pass from the body-
wall in the neighbourhood of the lateral lines, and converge
toward the midventral line; they are attached to the outer
surface of the ductus.

The ductus opens into the cloaca. This is formed by an
invagination of epidermis, the wall consisting of cuticle, and
a protoplasmic layer continuous on the one hand with the
epithelium of the ductus and on the other with the epidermis
(subcuticular).

The spiculze are attached to the dorsal wall of the cloaca,
and pierce this wall to project through the external aperture.

ASCARIS CAPSULARIA (Rup. ?).

Excretory GLAND.

In almost every cod numerous nematode embryos are to be
found encysted under the peritoneum. ‘They occur in largest
numbers on the surface of the liver and among the pyloric
ceca, but are also common in the mesentery. ‘They certainly
belong to an Ascaris, but to what species I am unable to say.
As usual with Nematode embryos, they occur coiled up like
a watch spring and surrounded by a capsule of badly-formed
fibrous tissue. They measure 22 to 28 mm. in length. The
head is blunt, with three papillze—one dorsal, two subventral—
around the triangular mouth. The body tapers more gradually
toward the head than toward the tail. There are no lateral
membranes. The anus is subterminal. There is no diverticulum

140 F. H. STEWART.

at the junction of cesophagus and intestine. The nerve-ring
lies shortly behind the mouth and is surrounded by a ganglionic
collar. Generative organs indistinguishable as such. One of
the most prominent internal organs is the gland which will
be described below.

I believe that this form is identical with A. capsularia
(Rud.) found in salmon and Gadide, but as there are very
few points to distinguish Nematode larve from one another,
this may not be the case, or several distinct embryos may be
included under the one name. Dajardin (2a) gives the length
as 27 mm., von Linstow (11a) i9 mm. I have therefore
given the above description.

In addition, A. capsularia is, of course, merely an embryo ;
it should therefore not have a separate specific name, except
for convenience, until the adult form is identified.

The gland which I propose to describe is interesting from
the fact that it is homologous with the ventral gland of free
Nematodes, and with, e.g., the poison glands of Strongylus
filaria (Rud.).

It extends through the anterior six tenths of the body,
lying ventral to the alimentary canal (PI. 8, fig. 32, vg.), and
its duct, after a short course, opens in the midventral line
between the two subventral oral papilla and immediately in
front of the ganglionic collar which surrounds the nerve-ring.

The body of the gland (Pl. 8, fig. 32 a) is composed of a
single gigantic cell, 15 mm, in length, ‘255 mm. in greatest
breadth, somewhat flattened between the body-wall and the
alimentary tract, tapering to the posterior extremity, and
rather blunt at the anterior.

The body of the cell is composed of finely granular acidophil
protoplasm, the outer layer of which appears somewhat more
condensed, and stains with hematoxylin,

The nucleus (”.) is a remarkable structure lying in the
anterior half of the cell. It is 6 to7 mm, in length. In
transverse section its outline varies from linear to circular or
biconcave. It has a wall identical in appearance with the
outer wall of the cell; this encloses a vacuole containing

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 141

a finely granular substance which stains intensely—almost
black—with haematoxylin.

A fine canal (can.), 0°0075 mm. in diameter, traverses the
entire length of the cell, and becomes continuous with the

duct. Its course is in places slightly tortuous. It receives

numerous smaller canaliculi which traverse the protoplasm.
Its walls are also composed of condensed protoplasm; its
contents when fixed are very finely granular and acidophil.

The short duct (Pl. 8, fig. 33) runs from the anterior pole
of the cell to the midventral line, and in the substance of this
line to the external aperture. Its wall is composed of pro-
toplasm continuous with the protoplasm of the cell; one or
two small nuclei occur in it. There is a very fine cuticular
lining continuous with the external cuticle.

Hamann describes a similar organ in the embryos found
in Zeus faber. It again consists of a single cell, in
front thread-like, at its greatest breadth extending from one
lateral line to the other, is always in connection with the
dorsal median line, contains a canal lned by a “ glashelles
membran,” which opens dorsally behind the lips. ['The italics
are mine. |

Cells occur in the body cavity identical with the basophil
cells of Oncholaimus vulgaris.

Lying between the ventral gland and the right lateral line
is a solid cellular mass, possibly the rudiments of the gonads.

CoMPARATIVE MorpHoLoGy oF THE ExcrreroRY GLANDS IN
NEMATODES.

This subject has been very fully worked up by Jagerskidld
in a masterly paper (9). I have, however, a few points
to add. I shall commence with a brief summary of his
results.

He finds that the excretory organs of Nematodes can be
classified into four groups:

(1) The ventral gland of the most free living forms (PI. 9,
fig. 35).

142 F. H. STEWART.

(2) The unilateral asymmetrical excretory organ, in its
anterior part flattened and band-shaped, and with a highly
modified nucleus, as in A. decipiens (Pl. 9, fig. 37).

(3) A similar organ, but without the band-like enlargement,
asin A. clavata (PI. 9, fig. 38).

(4) The bilateral organ of, e.g., A. megalocephala
(PI. 9, fig. 40).

In all these types the organ consists of a single large cell,
with an intra-cellular system of canals, and with a duct formed
in many cases by an invagination of epidermis. All four are
homologous, an intermediate type between (2) and (4) being
found in A. rotundata, in which a small limb crosses from
the main stem of the gland on the left side to the right
(Pl. 9, fig. 39). The cause of the change in type from (1)
through (2) and (3) to (4) is to be found in an increase of
work thrown on the gland, probably by change of habits in
the animal. The gland is compelled to enlarge, and adapts
itself to the narrow body-form by elongating, and following
the line of least resistance, applies itself first to one and then
to both lateral lines.

Bastian (1) had previously (and he is quoted by Jiigerskidld)
stated the same opinion, that the ventral gland of free Nema-
todes was homologous with the excretory organ of parasitic
forms.

The results given above supply another link in the chain.
‘The excretory gland in the embryo above described (Pl. 9,
fio. 36) is a typical ventral gland, inasmuch as its opening,
although situated close to the mouth, is also immediately in
front of the nerve-ring (corresponding with the situation in
Oncholaimus vulgaris), and as it lies free in the body
space, comes into contact only with the left lateral line, never
being in continuity with it.

On the other hand, in many points it resembles the excretory
organ of A. decipiens. Its duct is formed by an ingrowth
of cells from the ventral line ; its anterior portion is broad,
almost band-like, and contains the nucleus ; its posterior por-
tion is narrow and thread-like. It contains a central canal

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 145

with branches ramifying through the protoplasm. The nucleus
is highly elongated, in places band-lhke, and its structure
very strongly resembles that of A. decipiens and allied
forms.

We should expect in the embryo of a parasitic form to find
a transition between the type found in the free-living forms
and that in the adult parasite. As I have shown, our expecta-
tions are fulfilled.

A step further can be taken in pointing out homologies; it
appears probable that the excretory organ of Nematodes, in
whatever form it occurs, is a nephridium homologous with the
nephridia of, e. g., Platyhelmia or Chaetopods.

A nephridium is defined by Ray Lankester (11) as follows:
“Nephridia are distinguished by their independent origin,
each from a single superficially placed cell, which often is
seen to be derived from ectoderm, and probably must be
traced to that layer even when it appears as part of the meso-
blast. They are also distinguished by their structure, which is
primarily that of a number of perforated or drain-pipe cells
placed, as it were, end to end.”

The excretory organ of Nematodes is a cell perforated by
an intra-cellular canal. That it consists of a single cell and
not of a number placed end to end does not militate against
the homology suggested, since the chains of cells originate
from a single cell.

Jammes (7) and Jigerskiéld (9) agree that it is ectodermal
in origin. The condition as found in the ventral gland of
free-living forms by many observers, and as described for
Oncholaimus vulgaris by myself, certainly suggests an
ectodermal origin.

Hamann (5) is the only modern authority who regards it
as mesodermal.

We are accustomed to think of nephridia as paired organs,
but a single organ arising from the midline is as much
bilateral as two arising one on each side of the midline.

If the excretory organ when specialised resembles a
nephridium, in its most simple form, it bears a strong

144 F. H. STEWART.

resemblance to an unmodified skin-gland. Jagerskidld (10) has
pointed this out also, stating that physiologically at least the
ventral gland and the skin-glands are interchangeable. The
results given above agree with this; in the female of Oncho-
laimus vulgaris the functions of the ventral gland are
taken over presumably by the vulvar and lateral line glands,
as the animal reaches maturity.

It is hardly necessary to insist on the structural resemblance
between the ventral gland on the one hand and the tail-
glands and glands of the lateral lines on the other.

THE Ca@Lom.

It seems probable that the chief reason why a ccelom has
not been recognised in Nematodes is to be found in the
manner in which the reproductive organs have been described.
The terms “ testis” and “ovary” have been made to cover,
not only these organs themselves but also the spaces which
contain them, and although the spaces have of course been
recognised, sufficient morphological importance has not been
attached to them.

The condition may be summarised as follows: In all Nema-
todes hitherto described there is to be found either a single
median space or two bilateral spaces, lined by a characteristic
flat epithelium. These spaces with their walls I have de-
scribed as testicular and ovarian regions. <A specialised
patch of this epithelium forms the testis or ovary. From the
space, or spaces, a duct, or ducts, leads to the exterior: where
there are two ducts they unite before reaching the exterior.
In regard to origin, the cavity and, no doubt, the ducts as
well are mesodermal. This is the view of Jammes (7), and in
Hallez’s (6) figures it can be seen that the gonad tubes arise
from two groups of cells which in the early embryo lie in the
blastoccele, one on either side of the gut.

The cavities are, then, typical gonoccels or protocceloms
(11, pp. 35, 36), and the ducts (unless they should prove to be
formed by ingrowth of epidermis, which there is no reason to
expect) are typical gonoducts or ccelomoducts.

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 145

The fact that in parasitic Nematodes the cavities form
highly convoluted tubes of great length should not be allowed
to obscure this truth. In the forms which I have described
we have a series. It commences with the simple gonoccel
found in the immature female or mature male of Oncho-
laimus vulgaris; next, in the mature female the gonocel
is slightly more complex; there is a narrow conical outgrowth
from the original cavity, which is represented by the ovarian
cecum; finally, in A. clavata the entire gonoccel has
become enormously elongated and convoluted.

The cause for this latter development is obvious—-viz. the
call for an enormous increase in the reproductive products on
changing from a free to a parasitic life.

In the description of Oncholaimus vulgaris and
Ascaris clavata above I have refrained from the use of
the terms “ gonoccel”’ and “ gonoduct,” since I did not wish
unnecessarily to bring theory into the strictly descriptive part
of my article. I wish now, however, to repeat that I believe
that the “testicular” and “ovarian regions’? correspond
accurately with the male and female gonoceels respectively,
and that the remaining portions of the genital tracts (uterus
and vagina and ductus ejaculatorius in Oncholaimus
vulgaris; oviduct, receptaculum seminis, uterus and vagina,
vas deferens, vesicle, and ductus in Ascaris clavata)
correspond with the gonoducts. The wall of the former
consists of an exceedingly fine squamous epithelium only; in
the wall of the latter the epithelium is cubical or columnar
(except when the duct is much distended), and there are
other coats.

ComMPARATIVE MorpHoLoGy OF THE GONENTERIC CANALS.

The only organ in any way similar which has been pre-
viously described in Nematodes is the remarkable “ Rohren-
formige organ” with its ducts, found by De Man (12) in
Oncholaimus fuscus. This consists of a tube lined by
epithelium, lying dorsally in the body-space, communicating

146 F. H. STEWART.

by two canals with the exterior and by two canals with the

two uteri.
This condition and that found in O. vulgaris suggest at
once a comparison with Trematodes. The former reminds us

Text-ria. 8.
a ‘
R.0 oe

\
Ut
=

of Laurer’s canal as it occurs in Malacocotyleans, the latter of
the genito-intestinal canal of Heterocotyleans. The diagrams
(text-figs. 8, 9) illustrate the point better than much writing.
The diagrams of the Trematodes are copied from the article
in Ray Lankester’s ‘Treatise on Zoology,’ Pt. IV, p. 87.
Text-fig. 8a represents diagrammatically a transverse sec-

TEXT-FIG. 9.

tion through O. fuscus, The “ Rohrenformige Organ” (R. 0.),
with its ducts, is shown. Text-fig. 8b represents a Malacoco-
tylean, Germ is the germarium, L. c. Laurier’s canal. Text-
fig. 9a represents O. vulgaris, and shows the gonenteric
canals, to be compared with the genito-intestinal canal (g. 7. ¢.)
of the Heterocotylean (text-fig. 9 b).

In text-figs. 8a and 9a the two uteri and vagine are, for
fairness of comparison, represented as if they lay side by side.

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 147

LITERATURE REFERRED TO.

1. Bastran.—‘‘ Anatomy and Physiology of Nematoids, Parasitic and Free,”
‘Phil. Trans. Roy. Soc.,’ 1866.
2. Bast1an.—-“ Monograph of the Anguillulide,” ‘Trans, Linn. Soe.,’ vol.
xxv, 1866.
2a. Dusarpin.—‘ Systema Helminthum.’
8. Gotowin.—* Beobachtungen an Nematoden,” II, ‘ Excretorishes
Apparat.,’ Kasan, 1902. (Summary, ‘ Zool]. Centralblatt,’ ix, p. 449.)
4, Hamann.— Centralblatt f. Bakteriologie,’ vol. xi, p. 501.
5. Hamann.—‘‘ Die Nemathelminthen,”’ II, ‘ Die Nematoden,’ Jena, 1895.
6. Hatiez.—< Developpement de quelques Nematodes,” ‘ Recherches sur
l’Embryogenie,’ Paris, 1885.
7. JamMES. —‘ Ann. des Sciences Natur.,’ t. xiii, p. 321.
8. Jaumes.—‘ Recherches sur ]’Organisation et Developpement des Nema-
todes,’ Thesis, Paris, 1894. (Summary, ‘ Zool. Centralbl.,’ Jahrg. ii,
p. 398.)
9. JAGERSKIOLD.—* Beitriige zur Kentniss d. Nematoden,” ‘ Zool. Jalirb.
Abtheil. Anat. u. Outog.,’ Bd. vil, p. 449.
10. JAGERSKIOLD.—‘ Weitere Beitrige z. Kentniss der Nematoden,’ Stock-
holm, 1901.
11. Ray Layxester.—‘ Treatise on Zoology,’ part il.
lla. von Linstow.-—‘ Fauna Arctica’ (Schaudinn and Romer), Bd. i, p. 119.
12. pp Man.—‘ Anat. Untersuchungen tiber Freilebende Nordsee-Nematoden,’
Leipzig, 1886.
13. Nassonow.—‘ Zool. Anz.,’ xx, p. 203.
14. Zur Srrassen.—‘ Zeitschr. f. wiss. Zool.,’ vol. liv.
15. WanpotteK.—‘ Arch. f. Naturgeschichte Jahrg.,’ lviii, Bd. i, 1892.

take)

EXPLANATION OF PLATES 7—9,

Illustrating Dr. F. H. Stewart’s paper, “The Anatomy of
Oncholaimus vulgaris, Bast., with Notes on two Parasitic
Nematodes.”

REFERENCE LETTERS.

a.c. Acidophil cell. «.g. Anal ganglion. a.¢. Anterior testis. a. ¢. 7.
Anterior testicular region. 4. ¢.! Basophil cell in ganglionic collar. 6. ¢2
Basophil cell behind collar, d.m. Basement membrane. Caz. Canal, excre-

148 F. H. STEWART.

tory. Cl. Cloaca. Cl. ap. Cloacal aperture. cut, Cuticle. D. 4. Ductus
ejaculatorius. D. LZ. M, Dorso-lateral muscle-field. D. Mf. Dorsal muscle-
field. ea. ap. Excretory aperture. g.c. Ganglionic collar. G@. He. Gonen-
teric canal. g.s. Germinal syncytium. gy. w. Gonoccel wall. J. Intestine.
L. L. Lateral line. JZ. ZL. g. Lateral line gland. MM. D. LZ. Median dorsal
line. Jf. V. L, Median ventral line. m. ‘ Mesoglea.’ m.c. muscle cell.
m. J. muscular layer. m.2. ‘Mesogloea’ nucleus. 2. Nucleus. ». g. w.
Nucleus of gonoccel wall. 2.7. Cireumeesophageal nerve-ring. @s. Giso-
phagus. q@s.d. Hsophageal gland-duct. og. Oogonium: ov. ¢. Ovarian cecum,
ov. r. Ovarian region. ph. Pharynx. P.S. Protractor spiculorum muscle.
p. ¢. Posterior testis. p. ¢. 7. Posterior testicular region. rh. Rhaciiis.
Rec. Rectum. &. S. Retractor muscle. s. Spermatozoon. se. Subcuticle.
sg. Spermatogonium. s/g. Shell-gland. sm. 7. Submedian line. sp. Spicule.
S. V. Seminal vesicle. 7’! Epidermal cell or nerve-cell, type 1. 7.? ditto,
type 2. ég. ap. Tail-gland aperture. ¢g. d. Duct of tail-gland. 7. R. Testi-
cular region. w¢, Uterus. vag. Vagina. v.d. Vas deferens. o.g. Ventral
gland. v.g.c. Vaginal or vulvar gland-cell. ». gy. d. Duct of ventral gland.
V.Z.M. Ventro-lateral muscle-field. V.d/. Ventral muscle field. Ya. Vulva.
Vu g.ap. Aperture of vulvar gland,

PLATE 7.

Fie. 1.—Anterior extremity, O. vulgaris. Adult female. x 80. (a) A
large cell of the body space.

Fic, 2.—Head, O. vulgaris. x 350. (a) Cirenmoral papilla; (4) cireum-
oral ring of hair; (c) dorsal row of hair; (d) dorsal tooth.

Fic. 3.—Transverse section, O. vulgaris, male, shortly in front of nerve-
ring, x 350, Stain: Safranin, picronigrosin.

Fig. 4.—Transverse section, O. vulgaris, male, through the nerve-ring.
x 350. Stain: Safranin, picronigrosin.

Fic. 5.—Tail, O. vulgaris, male. xX 350.

Fic. 6.—Transverse section, male, through cloaca. x 350. Stain: Thionin,
eosin.

Fic. 7.—Gland-cell of lateral line. > 800. Stain: Safranin picro-
nigrosin.

Fic. 8.—Part of a transverse section in posterior cesophageal region,
x 800. Stain: Safranin, picronigrosin.

Fic. 9.—Type 1, epidermal cell. x 800.

Fic. 10.—Part of a transverse section in posterior cesophageal region,
x 800. Stain: Safranin, picronigrosin,

Fic. 11.—Coarsely granular acidophil cell of the body space. x 800,
Stain: Safranin, picronigrosin,

THE ANATOMY OF ONCHOLAIMUS VULGARIS, BAST. 149

Fic. 12.—Transverse section through ventral gland, showing nucleus.
x 3850. Stain: Thionin, eosin.

Ira. 13.—Transverse section at the level of the termination of the ceso-
phagus. Immature female. x 350. Stain: Safranin, picronigrosin.

Fic. 14.—Part of a transverse section in intestinal region. Mature
femate. X 800. Stain: Thionin, eosin.

Fic. 15.—Part of a transverse section, same as 14.

Fic. 16.—Transverse section through anterior testicular region, showing
the germinal syncytium. x 350. Stain: ‘Thionin, eosin.

Fic. 17.—Transverse section through posterior testicular region, ductus
ejaculatorius, and intestine. x 350. Stain: Safranin, picronigrosin.

PLATE 8.

Fie. 18.—Transverse section through ovarian cecum, containing ripe
ovum with yolk granules. x 350. Stain: Thionin, eosin.

Fic. 19.—Transverse section through ovarian region (first part) and shell-
gland. xX 350. Stain: Thionin, eosin.

Fra. 20.—Transverse section through ovarian region (first part), and
uterus. X 350. Stain: Heematin, picronigrosin.

Fic. 21.—Transverse section, showing gonenteric canal opening into
intestine. x 850. Stain: Borax-carmine, picronigrosin.

Fria, 22.—Gonenteric canal (between the sections represented in figs. 21
and 23). xX 350.

Fie. 23.—Gonenteric canal opening into vagina. X 350.

Fic. 24.—Transverse section through vulva. x 350. Stain: Borax-
carmine, picronigrosin.

Fie. 25.—Immature female. Transverse section through gonocel. x 350.
Stain: Thionin, eosin.

Fia. 26.—-Immature female. ‘Transverse section through gonoduct, not far
from vulva. > 850. Stain: Safranin, picronigrosin.

Ascaris clavata.

Fie. 26a,—Transverse section, ovarian region. X 350. Stain: Hema-
toxylin, eosin.

Fic. 27.—Transverse section, oviduct. X 350. Stain: Hematoxylin,
eosin.

Fie. 28.—Transverse section, uterus. xX 350. Stain: Hematoxylin,
eosin.

von. 50, PART 1.—NEW SERIES. 11

150 F. H. STEWART.

Fig. 29.—Transverse section, testicular region. »X 350. Stain: Heema-
toxylin, cosin.

Fia. 30.—Transverse section through vas deferens at its opening into
seminal vesicle; portion of wall of latter shown. x 350. Stain: Hama-
toxylin, eosin.

Fie. 31.—Portion of transverse section, ductus ejaculatorius. x 350.

Stain: Haematoxylin, eosin.

Ascaris capsularia.
Fia. 32.—Transverse section, showing ventral gland. x 80. Stain:
Hematoxylin, eosin.
Fie. 82a.—Transverse section through ventral gland, anterior portion,
x 350. Stain: Hematoxylin, eosin.

Fic. 33.—Oblique section through duct of ventral gland. x 350. Stain: ©
Hematoxylin, eosin.

PLATE 9.

Diagrams illustrating the development of the excretory apparatus in
Nematodes.

(Frias. 37-40 are based on the descriptions given by Jiigerskidld [9)).
Fie. 34,—Skin-gland of lateral line.

Fig. 35.—Ventral gland of eg., Oncholaimus vulgaris.

Fia. 36.—Ventral gland of Ascaris capsularia.

Fia. 37.—Excretory organ of A. decipiens.

Fria. 38.—Excretory organ of A. rotundata,

Fig. 39.—Excretory organ of A. clavata.

Fig. 40.—Excretory organ of A. megalocephala.

THE HAMOFLAGELLATES. fol:

The Hemoflagellates: a Review of Present
Knowledge relating to the Trypanosomes
and allied forms.!

By

H. M. Woodcock, D.Sc.(Lond.).

(With Text-figures.)

ConTENTS.
PAGE
1. Characteristics F : : . 152
2. Introductory. ; ; : ; : . 154
3. Historical : : ‘ : . 162
4. Mode of infection and habitat: effects on host . 165
5. General account of Trypanomorpha (Trypanosoma)
noctue (Celli and San Felice) . ; : . 179
6. Comparative morphology of Trypanosomes. 201
7. Biological considerations :—movement ; seolomeradiont ate
normal and involution forms é : : eal)

‘ All the more important papers published up till February Ist, 1906
have, so far as the author is aware, been considered in drawing up this
review.

152 H. M. WOODCOCK.

Class. —Mastigophora.
Sub-class.—FLAGELLATA.
Order.—Lissoflagellata.
Sub-order.—M onadina.
Family.—Try panomorphide, n. fam.
Genus.—Try panomorpha,! ng.
Sub-order.—Heteromastigina.
Family.—Try panosomatide, Dofl. emend.
Genera.—Trypanophis,* Trypanoplasma,
Trypanosoma.

Section I. CHARACTERISTICS.

The Heemoflagellates, although possessing in common a
uniform type of organisation, are probably not to be con-
sidered as all belonging to a single, well-defined group of
monophyletic origin. They constitute, rather, an assem-
blage of forms springing from at least two different stocks,
the resemblances which they exhibit being due to con-
vergence, brought about by the acquirement of similar adap-
tations in response to their similar and highly-specialised
mode of life. They are entirely parasitic, their characteristic
habitat being the blood of a Vertebrate. It is unlikely,
however, that, in the majority of cases, the whole life-cycle
is undergone in that host. The transmission of the parasites
from one Vertebrate individual to another is by means of a
blood-sucking Invertebrate, which, in several instances, is
now known to be a true alternate host, and not merely a
carrier; indeed, it is becoming more and more probable that
an alternation of hosts normally occurs in each life-history.

1 The name for this genus has been kindly suggested by Prof. Léger.
The writer desires, here at the outset, to warmly thank Prof. Léger for
much advice and assistance, especially in connection with the section on the
derivation and phylogeny of the Trypanosomes.

2 Although Trypanophis is most probably not a hxmal parasite, it is
included in this article since it is undoubtedly closely related to Try pano-
plasma,

THE HAMOFLAGELLATES. 153

The Heemoflagellates possess either one or two flagella,
inserted into the body, with few exceptions, at or near its
anterior end. When there are two flagella, one is free and
directed forwards; the other is attached for the greater part
of its length to the side of the body, by means of an undu-
lating membrane, and terminates ultimately in a free portion
directed backwards. When only one flagellum is present, it
is invariably attached in this manner, but the flagellum is

Fic. 1.—“* Undulina ranarum,” Lankester, 1871. In B,
the nucleus is shown. (From Lankester.)

probably not to be considered homologous in all these cases.
In certain forms, which are to be derived from a Monadine
ancestor, it is, of course, the single anterior flagellum that is
represented ; in others, however, which are rather to be re-
garded as descended from a Heteromastigine ancestor, it is the
trailing, posteriorly-directed flagellum that persists. There
are two nuclear bodies, one, the trophonucleus, regulating
the trophic life of the cell, the other, the kinetonucleus,
directing its kinetic activities.

The most general method of reproduction is by binary,

154 H. M. WOODCOCK.

longitudinal fission, but multiple division or segmentation is
also met with. The complete life-history, where known, is
very complicated. It includes true bi-sexual conjugation,
which takes place in the Invertebrate; and it appears very
likely that, in most instances at any rate, this host is to be
considered as the definitive and primary one, and the Verte-
brate as the intermediate or secondary one.

Section II. Intropucrory.

A Study of Recent and Rapid Growth.—Even more
marked than in the case of the Sporozoa has been the recent
great and rapid increase in our knowledge of the Trypano-
somes. The bulk of the important research on these organisms
has been accomplished, indeed, within the last four or five
years, culminating, for the time being, in the remarkable
and far-reaching discoveries announced by Schaudinn at
the beginning of 1904. The realisation of the extreme
economic importance of these parasites is mainly responsible
for this advance. Until almost the commencement of the
present century they had been very little studied from a
purely zoological point of view. Apart from the work of
Danilewsky in the eighties scarcely anything had been
previously done towards elucidating their morphology and
life-history. Reasons are not far to seek which explain, at
any rate, to a certain extent, this lack of interest.

Occurrence.—The minute size of the parasites, together
with their habitat in the blood, renders them, unlike the
majority of Sporozoa, very inconspicuous; and they are
consequently overlooked unless specially searched for. In
the light of recent investigation, however, it cannot be
maintained that Trypanosomes are at all limited in distribu-
tion. For, although they are restricted,! so far as is known, to
blood-sucking Insects and leeches among Invertebrates, they

! The allied form Trypanophis is an exception, being parasitic in
certain Siphonophora. *

THE H#EMOFLAGELLATES. 155

appear to be widely distributed among the principal classes
of Vertebrates; and, at the present time, hardly a month
passes without a new host being added to the list. It is more
difficult to be certain with regard to the frequency with which
individual species of the parasites occur, the data being, as
yet, somewhat scanty. In one or two instances, however, they
are known to be of fairly common occurrence, the Try pano-
soma lewisi of rats, for example, being quite as abundant
as many Gregarines. This form is met with in all parts of
the world, having accompanied the Rodents in their ubiquitous
migrations. The proportion of hosts infected varies usually
from 10 per cent. to 50 per cent., according to the locality,
but, in Berlin, Rabinowitsch and Kempner have found that it
may be as high as 41 per cent.

Parasitism in General.—Another reason accounting for
the comparative neglect of the ''rypanosomes has been the
fact that the forms prevailing throughout the greater part of
Kurope are non-pathogenic—that is, they do not, under
ordinary circumstances, give rise to any obviously harmful
effects in the animals which harbour them. Attention has
not therefore been directed to them by anything comparable,
for example, to the devastating epidemics of coccidiosis or
myxosporidiosis which sometimes occur.

Animals liable by their natural distribution to the attacks
of a given parasite may be divided into two classes according
to their behaviour towards it. Hither they are immune—this
term being used to signify that the attacked animal is actively
repellent ! to the parasitic organism, which is thus unable to
gain a footing—or they are susceptible. The reaction be-
tween any given parasite and its host, in the latter case, may
be regarded as the resultant of several factors. ‘The host, on
its part, in many, perhaps in most instances has become
accustomed or inured to the invader, and is, apparently,
practically indifferent to its presence. Again, to consider

1 The terms “repellent” and “tolerant” are suggested by Lankester

(‘Quarterly Review,’ July, 1904) in his interesting discussion of the bio-
logical relations between a parasite and its host.

156 H. M. WOODCOCK.

the matter from the point of view of the parasite, it may
not be advantageous for this to cause, by its ravages, the
functional disorganisation or premature death of its host.
For one group of parasites especially would such a procedure
be likely to have disastrous consequences, namely, the Heema-
tozoa, which are dependent for the completion of their life-
cycle upon being able to pass into an alternate Invertebrate
host at the moment when it sucks the blood of the Verte-
brate. In this case, therefore, we may say that such
mutual toleration! exists between the parasite and its host,
as, in ordinary circumstances, enables a proper balance to be
maintained on both sides. This equilibrium is disturbed only
when the situation is affected by adverse influences (e. g. an
unusually strong infection, or weakness of the host owing to
unfavourable seasonal or nutritive conditions, etc.).
Pathogenicity of Trypanosomes.—These considera-
tions may afford some explanation of the non-occurrence of
trypanosomosis,” or illness due to a Trypanosome, under
normal conditions in nature. It is a very different matter
when animals and parasites belonging to distinct regions are
brought into contiguity to one another. A Hematozoan,
and especially a Trypanosome, produces marked pathogenic
effects upon gaining an entry into organisms which, previously,
have never been, by their distribution, liable to its invasion.
As Lankester (l.c¢.) points out, such a state of affairs is con-

1 It is, perhaps, desirable to emphasise the fact that Haematozoa, whether
we regard Hemosporidia or Trypanosomes, do not, in natural conditions,
cause any serious injury to their hosts. Consider, for instance, the Hw mo-
sporea and the Trypanosomes of cold-blooded Vertebrates ; and, again, the
very great number of Avian hosts in which malarial parasites appear to be
more or less generally present (cf. the observations of Galli- Valerio [23],
Sergent [102], and others).

2 This term is adopted in preference to various others in use for the
following reasons: (a) it has the right of priority, having been proposed
by Brumpt in 1901; (6) it agrees with the nomenclature of all other
Protozoan diseases, e. g. coccidiosis, piroplasmosis, ete.; and (c) it is pure
and not a latinized hybrid, like trypanosomiasis, for example (see Blanchard,
‘ Arch. Parasit.,’ viii, p. 572).

THE HAMOFLAGELLATES. 157

stantly being brought about by the never-ceasing, restless
activity of man. With the march of civilisation into the
“hinterlands” of the various colonies, man, together with
the numerous domesticated animals which accompany him, is
brought into close proximity to big game and other wild ani-
mals, and, what is more important, into the zone of the par-

Fie. 2.—Various blood-sucking flies. «a and Bs. Glossina
morsitans (transmits the Trypanosome of Nagana, T. brucii)
x 2; c, Hippobosca rufipes (thought to transmit the parasite
of Galziekté,” T. theileri) x 13; p, Tabanus lineola (pro-
bably conveys the Surra parasite, T. evansi) x13; E, Stomoxys
calcitrans (suspected in connection with T. equinum, of Mal
de Caderas [see, however, under Systematic] ) x 23. (a and B,
from Laveran and Mesnil, after Bruce; c, after L. and M.; p and
gE, after Salmon and Stiles.)

ticular blood-sucking Insects which prey upon them. ‘The new
arrivals thus render themselves liable to infection by parasites
to which they, unlike the indigenous animals of the neigh-
bourhood, are quite unaccustomed. The new kind of host,
being, of course, totally unadjusted to the special environment
in which it finds itself and insufficiently supplied with reactive
or defensive powers, is unable either to exert a repellent in-
fluence on the parasite or to maintain a proper balance between

158 H. M. WOODCOCK.

itself and the latter. The parasite, on its part, enjoys, at
first, a more lusty and vigorous development than usual in
the new and fertile soil, rapidly gains an ascendancy and
overcomes its host, which, sooner or later, almost invariably
succumbs.

Thus it happens that the discovery of these minute and in-
conspicuous blood parasites has usually been the result of an
endeavour to ascertain the cause of various perplexing mala-
dies (malaria, trypanosomosis, piroplasmosis, etc.) to which
“civilised”? man and “ imported” animals in these regions
are subject. It follows, however, from what has been said
above, that the animals for which these parasites are markedly
pathogenic cannot be regarded as their true or natural hosts,
which are rather to be sought among the native, tolerant
animals of the locality concerned.

Abnormal and Involution Forms.—This method of
discovery, moreover, tends, unfortunately, to militate against
a thorough investigation of the morphology and life-cycle of
the parasites themselves. The medical authorities by whom,
in most of these cases, new disease-causing parasites have
been first discovered were, as of course was only natural,
chiefly intent upon an investigation of the malady and its
prevention and cure. The parasite was studied, more or
less incidentally and from a pathological point of view,
attention being focussed principally upon the relation of its
different forms to the different phases of the illness, ete.
But when the reaction between the attacked organism and
the invader is particularly severe (as in trypanosomosis, for
example) great care is required in determining the exact rela-
tion in which any given form of the parasite stands to the
typical life-history of the same. LHspecially is this so in the
case of a T'rypanosome, for here, at any rate, the parasite by
no means has matters all its own way. A strenuous fight for
life is made by the host’s cells and tissues in which many of
the organisms, notwithstanding their remarkable vitality,
certainly suffer. As Laveran and Mesnil, two prominent
researchers on this group, have pointed out, there can be no

THE H#MOFLAGELLATES. 159

doubt that many authors have not taken this factor sufh-
ciently into account in constructing what they regarded as
the normal life-history of the Trypanosome concerned, and
that much of the variety in form and mode of division which
has been described is due to abnormal and altered appearances
of the parasite. These involution forms in reality represent
the commencing degeneration of the Trypanosome and are to
be carefully distinguished from its typical phases.
Schaudinn’s Work on the Life-cycle of a Try-
panosome.—The normal life-cycle of a Trypanosome can be
studied much more easily and, it may be said, only thoroughly
when it is undergone in the host or hosts to which it is by
nature specially adapted, and which, on their part, have
become accustomed to that particular parasite. It is highly
significant that when an investigation under these conditions
has been carefully undertaken, as recently by Schaudinn (98),
in the case of two common, well-tolerated ‘Trypanosomes of
the Little Owl,! a more complete and comprehensible account
of the whole life-history is made known as the result than
had, up till then, been given for all the Mammalian Trypano-
somes put together. Especially noteworthy is Schaudinn’s
revelation of the part played by the “ carrier” of these para-
sites—a gnat. Bruce was the first to demonstrate, in a bril-
liant manner, the carrying function of the T'se-tse fly? in
Nagana or the ''se-tse fly disease, and he showed that this
Insect acts as the intermediary between wild game (tolerant
of the Nagana Trypanosome, and serving as a reservoir) on
the one hand, and domesticated animals on the other.
Following his methods, much has since been ascertained
by various workers concerning the bionomics of other species
—their probable source, mode of infection, carrying agents,
etc. Yet in no single instance had it been proved whether
the Invertebrate is a true alternate host, one, that is, in
which definite stages of the parasite’s life-history are passed
through, until the publication of Schaudinn’s work.

1 Athene noctua, 2 Glossina morsitans.
3 It is casting no reflection on this author’s brilliant work to say that,

160 H. M. WOODCOCK.

The main facts elucidated by this author’s epoch-marking
research may be here stated. His description relates to two
distinct organisms, ''rypanosoma (here called Try pano -
morpha) noctue and Trypanosoma ziemanni. The
latter Trypanosome is remarkable for the great resemblance
it offers in certain phases to the Bacterial parasites, possess-
ing the form of spiral threads, which constitute Ehrenberg’s
genus Spirocheta (allied to Spirillum).! In both these
Heemoflagellates, the general plan of the life-cycle shows a
fundamental agreement. The gnat is a true host, and,
indeed, the principal or definitive one, since in it the sexual
process is undergone. During much of the time spent in
the blood of the bird, the parasites are attached to or pene-
trate into a blood-corpuscle, acquire a resting, “ gregarini-
form” condition, and become, in fact, what have hitherto
been recognised as characteristic Heemosporidia. The first-
named ‘T'rypanosome passes into a species of Halteridium,
the latter into a species of “ Haemameceba,” or ‘ Leucocy-
tozoon.”’ In other words these two Heemosporidian parasites
represent, respectively, only a transient phase in the life-cycle
of a particular Hemoflagellate.

Is a Similar “Alternation” Common to the
Majority of Trypanosomes?—These facts will serve
to indicate the great gap at present existing between our
knowledge of these two parasites of the owl and that of
most Trypanosomes. In many cases, indeed, almost the only
facts with regard to the life-cycle as yet known with cer-
tainty are that the parasites possess the faculty of “ agglome-
ration,” and that they multiply by longitudinal division.
There is, however, already a certain amount of evidence to
hand

and such evidence is rapidly increasing—which tends

dealing as it does with revelations of such fundamental importance in the
study of the Hzematozoa, the corroboration recently afforded by Sergent’s
investigations (103) is highly welcome.

1 So marked is this similarity, indeed, that Schaudinu was at first inclined
to consider this Trypanosome as exhibiting the typical characters of a
Spirocheta (see Appendix).

THE HEMOFLAGELLATES. 161

to prove that other forms agree with the examples men-
tioned above, at least so far as regards the broad features of
their life-history. For instance, with respect to the question
of their unity in possessing an alternation of true hosts, the
‘Trypanosomes are at the moment! in a position quite similar
to that in which, until lately, the Heemosporidia were.
It has been customary hitherto to sharply separate the
Hemosporidia of cold-blooded from those of warm-blooded
Vertebrates, notwithstanding their close agreement in habi-
tat and morphology, on the ground that the former had
no alternation of hosts. Recently, however, Siegel (105)
demonstrated such an occurrence in the case of Haemo-
eregarina stepanovi, parasitic in a tortoise, the discovery
being at the same time extended by Schaudinn for another
member of the Heemosporea, namely, Karyolysus
lacertarum. ‘The alternate hosts, in which in both cases
the sexual process is undergone, are respectively a leech
and a tick.2, Hence it may be said with practical certainty
that a definitive Invertebrate host is common to all the
Hzemosporidia, and that being so the distinction between
the two sub-orders vanishes. ‘The arguments in favour of a
similar fundamental agreement in the case of the different
Trypanosomes may be discussed under two principal head-
ings:—(L) the important and, in fact, essential part of trans-
mitter of the parasites played by a blood-sucking Inverte-
brate, which is in some cases known, and strongly surmised
in others, to be a true alternate host; and (2) the evidence
which points unmistakably to a close connection between,
at any rate, certain Hemoflagellates and certain Heemo-
sporidia. The various observations, etc., coming under each
heading are, however, considered in detail below (see espe-
cially section 9).

1 Since this was written Prowazek (88) has described in detail the life-
eycle of T. lewisi. Many of the phases, including sexual conjugation, are
undergone in a louse (Hematopinus), which is a true alternate host for
this parasite. This important discovery helps to bring Mammalian Try-

panosomes into line with the rest.
* Placobdella catenigena and Ixodes ricinus.

162 H. M. WOODCOCK.

Section III. Hisroricat.

The first observation of a Trypanosome is probably to be ascribed to Valentin
(117), who, in 1841, announced his discovery of Ameba-like parasites
in the blood of a trout. In the two or three years immediately following,
Remak, Berg, and others recorded the occurrence of Hematozoa which
were undoubtedly Trypanosomes in various fishes. The observers usually
remarked upon the transparent, membranous portion of the body, with a
denticulate fringe or border,—the well-known appearance presented by the
undulating membrane when in motion. The parasite of frogs appears to have
been first seen by Gluge (1842), and in July, 1843, Mayer (73a) described
and figured certain corkscrew-like and ameeboid organisms from the blood
of the same animal, which he termed variously Ameba rotatoria, and
Paramecium costatum or loricatum. A few months later (November)
Gruby (25) also published an account of this organism, to which he gave
the new generic name of Trypanosoma. The same parasite was subse-
quently described and figured by Lankester (80) in 1871, who, unaware of
Gruby’s work, called it Undulina ranarum; this author was the first to
indicate the presence of a nucleus (fig. 18) in the organism. The next
discovery was that by Lewis, in 1878, of the form parasitic in Indian rats.
This Trypanosome was named Herpetomonas lewisi! by Kent, and has
since been shown to be of common occurrence in sewer-rats throughout the
world. Trypanosomes were first met with in cases of disease by Griffith
Evans, who in 1880 found them in the blood of horses suffering from
Surra. The organisms were thought by him to be Spirilla. Steel (110)
rediscovered the same form five years later in transport mules in British
Burmah which were suffering from an “ obscure and fatal disease.’ He
took a similar view with regard to its affinities, and named it Spirocheta
evansi. An early description, with figures, of this parasite was given by
Crookshank (1886).

To Mitrophanow (18838 to 1884) and Danilewsky (1885 to 1889) we owe the
first serious attempts to study the comparative anatomy of these Hematozoa.
The work of the latter in particular does not appear to have received as
much attention as it deserved. This author examined many birds and fishes,
and endeavoured to fit the various phases of the parasites met with in each
case into their proper place in the life-cycle. Considering the difficulties of
technique with which Danilewsky had to contend, his researches merit
great commendation. Some of his figures of a Trypanosome of birds are

‘reproduced in fig. 8. Unfortunately the complete absence from his writings

of any system of nomenclature, which leads to the same form being often

1 This form is new placed in the genus Trypanosoma for reasons which
will be given in the Systematic section.

THE H#@MOFLAGELLATES., 163

referred to under distinct names, renders it sometimes quite difficult to
follow him correctly. The thoughtful character of his work, however, is
well illustrated by the following passage taken from the ‘ Parasitologie com-
parée du sang des oiseaux’’ (18). The author draws attention to the analogy
between the spirilliform flagella of “ Polymitus” (i.e. the male gametes of
a malarial parasite) and the Spirocheta obermeieri found in the blood
in relapsing fevers, and goes on—‘‘ One may very likely suppose that S.
obermeieri also is, by origin, not a free Bacterial form, but in all proba-
bility represents only a stage [in the life history] of a Hamatozoan, more

Fic. 3.—a—c. Different forms of ‘Trypanosoma san-
guinis avium,’ Danilewsky; p, the same parasite dividing
longitudinally. 2 = nucleus; w. m. = undulating membrane ;
f. = flagellum. (After Danilewsky.)

complex than is yet known, which at some period may even be intracellular (a
Heemocytozoon). Although this surmise has not been verified (so far) for
that particular organism, it has been proved for one very Spirocheta-like
parasite (T. ziemanni), and it seems by no means unlikely that it will be
found to be true in the case of certain others also (cf. Appendix). It
shows, at any rate, that Danilewsky was fully alive to the manifold
possibilities in connection with these organisms.

The discovery by Bruce, in 1894, of the South African parasite
(Trypanosoma brucii) in the blood of cattle and horses suffering
from Nagana may be said to have inaugurated a rapid and consider-
able increase in the number of known forms, the knowledge of which has

164 H. M. WooDcock.

in many cases thrown light upon the etiology of maladies hitherto obscure.
Thus Rouget, in 1896, ascertained that a Trypanosome is the cause of the
illness known as Dourine, which afflicts horses, mules, ete., in Northern
Africa and the Mediterranean region. A very deadly malady of horses in
South America, known as Mal de Caderas (hip-paraplegia) was shown to be
due to one of these parasites by Elmassian, Sivori and Lecler, and Voges,
working independently, in 1901. Similarly Theiler showed (1902) that
another species, a very large one, causes a distinct disease of cattle in the
Transvaal, known as Galziekté (bile disease). Since then, moreover, other
varieties of trypanosomosis have been observed in different regions of
Africa, but the exact specific nature of the parasites causing them remains,
in many cases, problematical.

Finally, there is the discovery of the human parasite. The credit for
first recognising a Trypanosome in human blood, and describing it as such,
must undoubtedly be assigned to Dr. Nepveu (1898), although it is possible
that Barron, who some years earlier reported having found Flagellate organ-
isms in the blood of an anzemic woman at Liverpool, was in reality the first
to notice these parasites in man. His description of them is, however, much
too meagre to render this at all certain. Trypanosomes were next seen in
Senegambia, in 1901, in the blood of a European suffering from intermittent
fever. Forde first found the parasites, but was uncertain of their nature ;
he showed them to Dutton, who recognised them as Trypanosomes, and
gave this form the name of Trypanosoma gambiense. A year later
(1902) Castellani discovered a similar parasite in the cerebro-spinal fluid of
patients suffering from sleeping sickness in Uganda, and it has since been
conclusively proved by Bruce and Nabarro that this organism is the true
cause of that ghastly disease. In all probability the species is the same as
that investigated by Dutton (see below, under “ Effects on host,” p. 177,
and also in the Systematic section).

More important, from the standpoint of zoology, than these interesting
medical discoveries, have been the investigations by Laveran and Mesnil,
Léger, Schaudinn and others during the last two or three years upon
numerous “tolerated ” species (many of them new) which supply, indeed,
nearly all the material for the sections on morphology, life-history, and
taxonomy. At the present time, scarcely a month passes without some new
form being described by one or other of these indefatigable researchers, and
it may be confidently expected that with such a rate of progress our know-
ledge of the complete life-cycle will not for long remain dependent on so
few observations as is at present the case.

THE H#MOFLAGELLATES. 165

Section IV. Mope or Inrecrion anp Hasirat; Hrrects
on Host.!

A. Relation of the parasites to the Invertebrate
host.

Schaudinn (l.c.) has minutely described the manner in
which the infection of Athene noctua, on the one hand,

cul.

7\

Fie. 4.—Diagrammatical longitudinal section through Culex
pipiens to show the distribution of the parasites in the body.
The arrows indicate the direction of their movement, the clusters
of stars the places of agglomeration. wl. = upper lip; ll. =
lower lip; hp. = hypopharynx; ph. = pharynx; s.g. = salivary
gland; ws. = esophagus; od. = esophageal diverticula (gas reser-
voirs) ; prov.=proventriculus; st.=stomach; m.¢.= Malpighian
tubes; c.=junction of ileum and colon; aort.= aorta. (After
Schaudinn.)

and Culex pipiens (females”), on the other, is brought
about. For a detailed account of the complicated part played
by the different organs of the gnat in the act of biting, the

1 The habitat and effects upon its host of ‘“ Piroplasma” donovani
are discussed in Section X, it being thought preferable to consider all the
facts relating to this parasite at the same time.

* Only the females of gnats and mosquitoes suck blood.

VOL. 50, paRT 1.—NEW SERIES. 12

166 H. M. WOODCOOK.

reader is referred to this work. It must suffice here to give
some idea of the manner in which the passage of the Try-
panosomes to and fro, and their wandering through the body
of the Insect, is effected.

The first act after penetration of the proboscis or upper lip (wl., fig. 4)
is a sudden, particularly vigorous, respiratory contraction of the abdomen,
which causes the blood in the body of the gnat to rush ferwards. Pressure
is thus exerted on certain sac-like diverticula of the esophagus (od.) and
on the salivary glands (s. g.), the gaseous and liquid contents of which are
thereby expelled through the long tubular hypopharynx (hp.) into the
wound, carrying with them, in the case of an infected Culex, loose masses
of agglomerated Trypanosomes situated at the junction of the pharynx and
cesophagus (shown by the cluster of stars in the figure). In this way is
brought about the entrance of the parasite into the Vertebrate host. The
quantity of saliva secreted is small, and serves principally to digest the
blood. Schaudinn finds that the poisonous effects caused by the gnat’s bite
are due, not so much to the saliva, as to the irritant enzyme of a Fungus
(related to the Entomophthoree), which is a very common commensal
of the Insect, and lives in the esophageal diverticula (the so-called “ suck-
ing-stomachs’’), Here it gives rise, during the processes of metabolism, to
carbonic-acid gas, and when the contraction of these “ gas-reservoirs”’ takes
place, the secreted gas, together with a small quantity of the Fungus in a
pullulating condition, is injected into the wound. The author also thinks
that rapid coagulation of the blood (before it can pass into the Insect’s
stomach) is prevented by the gas rather than by the salivary juice.

With the cessation of the respiratory contraction, a small area of negative
pressure occurs at the junction of the gas-bubble in the wound and the
capillary-like hypopharynx, and the liquid (chiefly blood) rushes up the
narrow tube into the pharynx, carrying with it any Trypanosomes it may
contain. From the pharynx the blood is pumped through a valve into the
cesophagus and its reservoirs, which become filled before the next respiratory
contraction takes place. When this happens the blood is driven from these
esophageal diverticula into the stomach (sf.), where digestion goes on.
The whole process may be repeated until the stomach becomes filled, often
to overflowing.

The chances against a successful infection of the gnat
appear to be, however, considerable. The author found from
experiment that frequently infection did not occur at all,
many of the gnats either not biting the birds, or at other
times being unable to digest the blood, which was then
evacuated practically unaltered together with the parasites.

THE HAMOFLAGELLATES. 167

Again, many individuals, as well as races of gnats, appear to
have acquired immunity against the parasite, i.e., they are
repellent to it. Moreover, even if the Trypanosomes, in the
requisite phase of their life-history,! gain a footing in the
Insect, their further development may be hindered by the
existing presence of a different parasite. On the other hand,
if the infection is too strong and the development of the
organisms too lusty, the gnats are unable to withstand them,
and, instead, succumb. A like consequence, it is interesting
to note, may ensue if the commensal Fungus (normally, as
has been seen, of much utility to the Insect) obtain too great
an ascendancy.

The distribution of the Trypanosomes in the body of the
enat is intimately connected with the process of digestion.

As the imbibed blood passes through the posterior part of the cesophagus,
the cuticular lining of the latter becomes altered into a gelatinous layer,
which is cast aff and envelopes the blood in a kind of sheath, the whole
passing on into the mid-gut or stomach. Except during the actual time of
feeding the hinder region of the cesophagus is invaginated into the anterior
part of the stomach, the narrow neck (or proventriculus) of which is
expanded to receive it (fig. 4, prov.). The epithelial regeneration which
takes place at the conclusion of digestion begins in this region soon after
the reinvagination of the esophagus.

Towards the end of the digestion (which may take from two to six days)
the Trypanosomes,’ after a period of multiplication, enter upon a resting
phase, and are found either attached to or between the epithelial cells.
After a second meal, when the fresh quantity of blood has become digested
and ready to be assimilated, a second period of multiplication of the para-
sites takes place, and the organisms gradually collect in the anterior part
where, since the folds of the invaginated cesophagus are non-absorptive, the
nutriment remains longest unabsorbed. Here the parasites commence to

1 Tt must be remembered that, so far as is known in the case of the
Hemosporidia, only sexual forms are able to stand the transfer from the
Vertebrate to the Invertebrate (see the account of the malarial parasites by
Minchin [75]); and Schaudinn finds the same to be true in the case of the
Trypanosomes he examined.

* This account refers to the first of the two parasites described by
Schaudinn, namely Try panomorpha (Trypanosoma) noctue; certain
minor differences exhibited by the other one are mentioned below.

168 H. M. WOODCOCK.

cluster. This is an especially favourable position for them to become
attached, since the esophageal epithelium has only lately shed its cuticle to
form the gelatinous sheath around the second quantity of blood, and its
cells are being actively regenerated prior to secreting a fresh one. The
Trypanosomes, therefore, are able to penetrate the delicate surface of this
layer, to which, indeed, as many as possible cling. With the increasing
scantiness of nutriment elsewhere more parasites are drawn into the imme-
diate neighbourhood, and these press in between those already attached
until finally, at the close of the second digestive period, an enormous mass
of parasites has accumulated at this spot, arranged in rows and layers, and
all in a resting condition. By this time the new cuticle has become
firm and chitinous, and when, at the next meal, the esophageal invagination
is withdrawn out of the neck of the stomach, it leaves behind it the cuticle,
serving as a base of attachment for the mass of agglomerated Trypano-
somes.

The next (the third) inflow of blood drives this mass before it, in the
form of a rolled-up ball, until it reaches the junction of the ileum and colon
(fig. 4, c), the narrowest point of the intestine. The wall is here very thin
and easily ruptured on distension. In this way most of the Trypanosomes
are enabled to pass through it, into the vascular lacune around, whence
they are carried to the heart. From the heart, the Trypanosomes are borne
through the aorta into the sinus surrounding the pumping-organ of the
pharynx ; and between the latter and the pharyngo-cesophageal valve! they
become at length arrested. The parasites continue to slowly multiply and
gradually collect again into agglomerated masses, which surround this
region of the pharynx and press on its walls, owing to the narrow throat of
the Insect. By the close of the third digestive period, these clumps of
Trypanosomes have broken through, and partly block up the cavity of the
pharynx. In the next biting act they are forcibly ejected thence into the
blood of the owl, as above described.

Thus the parasites cannot leave the gnat until the fourth meal, including
that which effected their entry, or not until the third meal after infection
has taken place. Schaudinn found that the shortest time elapsing between
entrance and exit was seven or eight days; this is the case when the Insects
are maintained at the optimum temperature for digestion.

Not all the Trypanosomes, however, are able to leave the gnat. Those
which become attached to the epithelium of the stomach, instead of to that
of the wsophageal invagination, are not carried backwards to the colon, and
so into the circulation, but remain behind in the mid-gut. These are
chiefly females, and they can produce a general recurrence of the parasites

1 A ring of muscular tissue, by means of which the cavity of the pharynx
can be shut off from that of the esophagus.

THE H®@MOFLAGELLATES. 169

during the next digestive period. Hence if a gnat becomes successfully
infected, it remains so throughout life.

The other parasite described by Schaudinn, Trypanosoma ziemannl,
differs slightly in its behaviour in the gnat. Towards the end of the first
digestive period the active parasites, instead of collecting in the anterior
part of the stomach, press backwards and enter the Malphigian tubules,
where, after undergoing multiplication, they come into relation with the
excretory epithelium and assume a resting-phase. When the next supply of
food has entered the stomach, the Trypanosomes again become active, and,
after the process of epithelial regeneration, are carried along with masses of
disintegrated cells into the colon, where the subsequent course of events
agrees with that just related. The author adds that he has only twice
found this form in the salivary glands, and does not think this is a normal
habitat of the parasites.

Before leaving the consideration of the Trypanosomes in
their relation to the Insect, a very interesting discovery of
Schaudinn’s must be mentioned, namely, the occurrence of
true hereditary infection. Both the parasites described (and
also, indeed, the commensal Fungus) may be inherited by the
enats. After breaking through the wall of the colon some of
the T'rypanosomes, instead of being carried forwards, may
pass to the ovarian follicles, there penetrate into the eggs of
the youngest, and so infect a succeeding generation. Asa
rule only few parasites, mostly females, thus penetrate into
the ova; once there they assume a gregariniform resting-
condition, in which they remain throughout the development
of the embryo. When at length the imago first sucks blood
these female forms undergo parthenogenesis, and the body of
the gnat becomes overrun with Trypanosomes. Sometimes,
however (in cases of ample nutrition), the parasites multiply
in the yolk of the growing eggs, and may become so numerous
that castration results.

Schaudinn has also ascertained that a similar hereditary infection of
Anopheles with the tertian parasite occurs. Thus true hereditary infec.
tion in the Sporozoa is by no means limited to the case of Glugea
bombycis, and it is highly probable that the infection of the progeny of
ticks (Rhipicephalus = Boophilus) by Piroplasma also belongs in
this category, although the parasites have not yet been demonstrated
actually in the ova of the mother-individual,

170 H. M. WOODCOCK.

In the case both of Culex and Anopheles the number of individuals
inheriting the parasites appears to be limited. In contra-distinction to
Rhipicephalus, where this is, apparently, the only mode in which infec-
tion is transmitted, Schaudinn thinks that the infection of gnats by this means
is not common in nature. It probably occurs chiefly in the autumn, when
the Trypanosomes penetrate into the young eggs, there to pass the winter
in a quiescent condition.

Thanks to the elaborate and painstaking investigation of
Schaudinn, we are thus enabled to form a very good idea
of the manner in which the ‘Trypanosomes are transmitted
from the gnat to the owl, and vice versa. There can be
little or no doubt that the process is of an essentially similar
nature in the relations of Mammalian Trypanosomes to other
biting Insects. In fact, the elucidation of the many factors
upon which infection and re-infection are dependent, and of
the adaptive modifications of the parasite to the biology of
the Insect, goes far towards explaining why many previous
investigators, unsuccessful in ascertaining anything definite
of the parasites in various “carrying” insects, have con-
cluded that these are not true hosts.

These views are borne out by the recent work of Prowazek (1. ¢.), who
finds that the behaviour of T. lewisi in Hematopinus and its passage
through the body of the Insect agrees in the main with that above described
for the Avian Trypanosomes in Culex. Such differences as there are stand
in close relation, on the one hand, to the somewhat different mode of feed-
ing and of absorption of nutriment in the louse, and on the other hand to

the fact that this parasite appears to be more resistant to ‘‘ external”
influences.

B. Relation of Trypanosomes in general to their
Vertebrate hosts.

Once an entrance into the blood is effected the parasites
pass rapidly into the general circulation, and are thus carried
to all parts of the body. In considering the distribution and
numerical abundance or otherwise, of the Trypanosomes in
any given individual, it is necessary to bear in mind whether
they are in a tolerant host or in an unaccustomed one.
Dealing with the former case first, the general trend of

THE HAUMOFLAGELLATES. 171

observation so far is to show that the parasites are, as a
rule, never abundant, but, on the contrary, usually rare,—at
least, in the T'rypanosome form.’

One reason for this numerical scarcity is, undoubtedly, the fact that
multiplicative stages are hardly ever met with. Multiplication of the
Trypanosomes, as such (although in all cases, so far as is known, facultative),
appears to be generally held in abeyance, except for a short period at the
commencement of the infection. This has been well shown by Laveran and
Mesnil (40) in the case of T. lewisi of the rat. After inoculation of an
uninfected animal with these parasites multiplication goes on briskly for a
time, but then slackens and finally ceases. In ordinary, naturally infected
rats, multiplication forms are scarcely ever seen. Sivori and Lecler (106)
alone have described them from very young sewer-rats, doubtless only just
infected. Similarly in the case of the great majority of Trypanosomes of
cold-blooded Vertebrates, dividing forms are extremely rare in naturally
infected hosts. The two parasites of the owl appear, however (according to
Schaudinn, 1. ¢.), to behave differently in this respect. At stated intervals
rapid, successive multiplication goes on for a short time, after which a
period of rest and growth ensues. The reason for this is probably to be
sought in connection with the fact that neither of these forms undergoes
multiplication when in the Hemosporidian phase (i.e. asa Halteridium
or a ** Leucocytozoon”’), this absence of “ schizogony ” being a quite ex-
ceptional occurrence, up till now, among Hemosporidia.

The Trypanosomes in the active motile form are, of course,
always free in the blood-plasma (interglobular). But
it cau no longer be maintained that none are ever, at
any stage of life, ecto-* or endo-globular. Of the two
parasites in the owl, one (Trypanomorpha) comes into in-
timate relationship with the erythrocytes, and the other
(Trypanosoma ziemanni) with the leucocytes and also

1 It need hardly be again pointed out, that in unusual circumstances, such
as an unusually strong infection, mal-nutrition of the host, or captivity,
the Trypanosomes may overrun even a tolerant host, and give rise to (in
more or less degree) the general symptoms of a pathogenic case. Such
instances are noted by Léger (64), Plehn (84), and Hofer (‘ Allg. Fischerei
Zeitung,’ xix, p. 48, 1904).

2 See, however, below (Section IX 8B) relative to the possibility of their
occurrence in a Hemosporidian guise.

3 The term ectoglobular is here used to denote superficial attachment
to the blood corpuscle.

172 H. M. WOODCOCK.

with hematoblasts, which then fail to become red blood-
corpuscles. In the former genus young, non-sexual or
‘indifferent ” individuals attach themselves firmly to the ery-
throcytes by the flagellar end and come to lie parallel to the
surface of the corpuscle. They then enter on a resting period
and sink slightly or nestle into the latter, incurving its surface
but nevertheless remaining ectoglobular.! The only part of
the substance of the corpuscle used up by the parasite is that
represented by the space which it has hollowed out, and upon
which it pressed. After growing for a certain time the para-
site leaves the erythrocyte without having decolourised or
apparently injured it to any serious extent. Female forms,
on the other hand, penetrate into the corpuscle and become
endoglobular,! growing at the host-cell’s expense, and eventu-
ally absorbing all its hemoglobin, forming therefrom the
well-known pigment, just like the human malarial parasites.
A slightly different course is followed by Trypanosoma
ziemanni with respect to the white corpuscles. In the first
place the parasites become attached by the non-flagellate end,
and secondly, the sexual forms of this Trypanosome are so
large that it becomes, here, rather a question of the parasite
taking up the leucocyte than the reverse.” Fig. 33 B shows a
fully-grown female form just attached to a uninuclear leueo-
cyte. According to Schaudinn, the parasite (which proceeds to
enter upon a resting-phase as soon as attachment is effected)
draws up, as it were, into itself the leucocyte, so that this

1 Schaudinn finds an exactly parallel behaviour in the case of the indiffe-
rent (“schizont”’) and female forms of the tertian parasite; hence there is
equal truth both in his and in Argutinsky’s views with regard to this point.
See also Minchin (I. ¢., p. 240).

2 Both Danilewsky (1. c.) and Ziemann (121) agree with Schaudinn in
attributing to this parasite a leucocytic habitat (whence its original name
of “ Leucocytozoon’’), and the latter author, moreover, is inclined to admit
the possibility, at any rate, of the leucocytes being enveloped by the para-
sites. Laveran (87) on the contrary, in his recent description of this form
in the ‘‘Hemameba’”’-phase, regards the parasites as invading hematids
(red blood-corpuscles) which become greatly altered and fusiform owing to
their presence,

THE HA#MOFLAGELLATES. Zs

comes to lie between the ectoplasm and the endoplasm of the
Trypanosome,'’ becoming greatly distended and elongated,
and more or less spindle-shaped (fig. 33). ‘The nucleus of
the corpuscle also becomes drawn out and band-like. The
ectoplasm of the parasite is apparently then transformed into
a protective envelope, and finally cast off with the remains of
the leucocyte at the close of this period.

There are, as well, one or two important observations show-
ing that Mammalian Trypanosomes also may come into rela-
tion with the blood-corpuscles. Voges (119) often noticed
individuals of T. equinum attached by the non-flagellate end,
and also by the side (cf. Trypanomorpha noctue above),
to red blood-corpuscles. In some cases, moreover, it appeared
as if the parasite had actually penetrated the corpuscle and
was destroying it. Similarly Buffard and Schneider (14),
in the case of T. equiperdum, frequently observed the tem-
porary fixation or attachment of the parasites by the non-
flagellate extremity. On the other hand, Prowazek (l.c.)
could find neither endoglobular nor ectoglobular phases of
T. lewisi, and considers that the habitat of this parasite is
restricted to the plasma.

With regard to the distribution of the Trypanosomes
throughout the body, they are to be met with practically
wherever the blood circulates. They are frequently more
numerous in the spleen, bone-marrow, kidneys, and liver,
than elsewhere*; and Schaudinn finds, in the case of his
Avian forms, that it is when passing through the capillaries
of these organs (especially of the heematopoétic ones), where
the circulation is more sluggish, that the parasites usually
leave one host-cell, or seek a fresh one. Danilewsky, again,

1 The author does not explain further how this feat is accomplished.
Remarkable it certainly is; for the ectoplasm appears to be a well-defined
layer without anything in the nature of a mouth-orifice, and normally, of
course, the parasite absorbs food osmotically.

2 One or two other points in connection with the distribution are more
conveniently noticed below, when considering the pathogenic effects caused
by the parasites.

174 H. M. WOODCOCK.

says that the forms he examined in the Roller-bird (Coracias)
were numerous in the bone-marrow, where multiplication went
on actively, the physiological conditions there obtaining being
very favourable for the infection of new host-cells by young
parasites. Multiplication may also go on, of course, in the
general peripheral circulation, and, in the case of newly-
infected fishes or rats, dividing stages of Trypanosomes have
been usually described from this situation.

Before passing on to consider the T'rypanosomes as patho-
genic agents, one very important point may be mentioned,
namely, that hereditary infection of the Vertebrate does
not, so far as is known, take place in normal circum-
stances. In the case of Mammals, whether tolerant or
unaccustomed hosts, the parasites appear to be unable to
traverse the placenta unless this has been in some way injured.
Several instances of the delivery of perfectly healthy young
from infected mothers have been noted; and in no case have
the organisms been found in the blood of a foetus, where
gestation was being accomplished without unfavourable
incident.

A detailed description of the effects produced by the Try-
panosomes upon unaccustomed Mammalian hosts, into whose
blood they may pass, would be out of place in thisarticle, and
is, besides, unnecessary, since medical writers have paid great
attention to this side of the subject. For full particulars,
and also for lists of the various mammals for which a given
Trypanosome is pathogenic and their degree of suscepti-
bility, the works of these authorities should be consulted ;
here, it must suffice to give a general idea of the course of
events.

The parasites may either remain infrequent or rare in the blood, some-
times, indeed, being unnoticed until shortly before death, or, on the other
hand, they may soon become numerous (fig. 5), and go on increasing more

or less constantly until the end. Speaking generally, it may be said that
the former case usually occurs in those animals (Bovide, Equide, ete.)

1 It must also be remembered that no instance of the inheritance of a
Hemosporidian infection by a Vertebrate host has been recorded.

THE HA@MOFLAGELLATES. 175

which are especially liable to suffer naturally from the various maladies
(Nagana, Surra, Dourine, etc.), while the latter condition is more often met
with in small Mammals which have been artificially inoculated with one or
other of those Trypanosomes. While, in this case, the disease is of an
acute character and rapidly fatal, in the former it is more chronic and lasts
much longer (often several months), with, however, nearly always the same
termination. Even when microscopical examination of the blood is unsuc-
cessful in finding the parasites, their presence in it is proved by the fact
that, after the injection of a small quantity into another more susceptible
host, the Trypanosomes soon appear in the blood of the latter.

There is, moreover, often considerable variability (particularly in chronic
cases of Surra and Mal de Caderas, for example) with regard to the appear-~

Fie. 5.—Trypanosoma equiperdum (of Dourine), in the
blood of a rat eight days after inoculation. a= parasites; b=
blood-corpuscles. (After Doflein.)

ance and number of the parasites in the blood at any moment. Occasionally
and at irregular intervals, evidently following upon a period of multiplica-
tion, the Trypanosomes may be fairly numerous, their appearance often
(though not invariably) coinciding with an access of fever. At other times
they seem to vanish almost entirely from the peripheral circulation. Why,
exactly, this should be so is not certain. Some authorities attribute it to
the rise in temperature, as being unfavourable to the parasites; others
think it is due to the more potent operation of chemical and physiological
defensive agencies of the host at a higher temperature. It is supposed that
certain of the organisms, more resistant than the majority, and situated,
perhaps, in some more favourable region of the body, survive and give rise
later to a fresh succession of parasites in the blood.

The main features of the illness show a general agreement, whichever
variety of trypanosomosis is considered ; one symptom may be more marked
than another in any particular disease, but a fundamental similarity in type
is usually noticeable—so much so that it was, for instance, a long time

176 H. M. WOODCOCK.

uncertain whether Nagana and Surra were distinct diseases or only two
varieties of the same. It may be here mentioned, in passing, that the
morphological differences between the organisms themselves are sometimes
so slight that it is impossible to say, from these alone, whether or no one is
dealing with distinct species; the minute distinctions observed might be due
to the parasites being in different hosts, for it is known that the same form
often varies somewhat (e. g. in size), according to the host in which it is.
Laveran and Mesnil, however, have performed a series of instructive experi-
ments (see 52 and 53) tending to prove that an animal which has been
successively immunized against one Trypanosome and its disease is still liable
to, at any rate, certain others. Hence there is great probability that a try-
panosomosis of any particular region (when it is not, obviously, one which
has been transmitted thither from another locality) is produced by a distinct
species of parasite. This view is also supported by the specialised and
limited facilities for distribution which Trypanosomes possess.

The pathogenic effects are nearly all referable to disorganisation either of
the circulatory, or of the nervous system, or of both combined. Fever
always occurs, at some time or other, during the course of the malady. Its
manifestation is extremely irregular, both in character and in time of
occurrence, and it is therefore usually readily distinguishable from malarial
fever. It may be variable or continuous; in the former case it appears to
be generally remittent rather than intermittent, the temperature, although
varying considerably, remaining, for the most part, above the normal.
There are, however, often periods of apyrexia, and the temperature may
also fall below the normal, especially towards the close of the illness.
There is, particularly in chronic cases, marked and progressive anemia and
emaciation, leading to pronounced enfeeblement, which is, in fact, the most
characteristic symptom of naturally occurring trypanosomosis. The loss of
red blood-corpuscles is frequently great (the number may diminish by as
much as 50 per cent.), and hematuria is also met with, though never to the
same extent as in piroplasmosis. Another common feature is the occurrence
of cedematous swellings in various parts, chiefly in the neighbourhood of
the genitals, of the abdomen, and around the eyes. The parasites are often
more numerous in the bloody serosities bordering these places than in the
general circulation. This fact is of great importance in connection with the
transmission of Dourine. In this disease the parasites are extremely rare in
the blood, but are generally numerous in the immediate neighbourhood of
the cedematous excoriations on the penis, so that, in coitus, they come into
contact with the vaginal mucous membrane of a healthy mare, through
which they are able to pass. Among other externally visible symptoms
which are met with in certain instances and to varying degrees may be
mentioned the following :— staring of the coat,” or localised bristling of
the hair; appearance of small naked areas of skin owing to the falling out
of the hairs; occurrence of sanguineous subcutaneous clots, which usually

THE HAMOFLAGELLATES. Wiz.

furnish a rich source whence to procure the parasites, and are doubtless the
result of embolism of the capillaries or small vessels by the same.

Nervous symptoms may be only slightly noticeable (e.g. a dull and
lethargic tendency towards the close of the illness), or they may be strongly
in evidence, especially in Dourine, Mal de Caderas, and sleeping-sickness.
In the two former more or less general paralysis of the posterior part of the
body frequently sets in; Mal de Caderas of horses in South America is,
indeed, often called ‘‘hip-paraplegia.” In neither of these two diseases,
however, have the parasites been observed actually in the nervous system
itself, although the brain and spinal cord show considerable histological
alteration. But in sleeping-sickness the Trypanosomes penetrate through
the membranes surrounding the brain and spinal cord, and can usually be
found upon centrifugalising a sufficient quantity of the cerebro-spinal fluid;
they have also been seen, in post-mortem examination, in the lateral
ventricles of the brain. It is this invasion of the nervous system by the
parasites that marks the transition of the case from one of “Try panosoma-
fever’? (while the parasites are confined to the blood) to one of sleeping-
sickness. The results of the change are soon apparent in the onset of
apathy, lassitude, tremor, and the other associated nervous symptoms which
characterise this dreadful malady.

Death from trypanosomosis is generally due either to
weakness and emaciation (in chronic cases), or to blocking
of the cerebral capillaries by the parasites (where these are
abundant and the disease consequently acute and rapid), or
to the disorganisation of the nervous system (paraplegic and
sleeping-sickness forms). Laveran and Mesnil have expressed
the opinion that some factor in addition to the presence of
the parasites themselves—especially when these are rare—is
requisite to explain the severe effects produced, and suggest
that the Trypanosomes secrete a toxine. Neither they nor
other investigators have, so far, been able to discover traces
of any such substance. In post-mortem examination the
most obvious pathological feature is hypertrophy of the
spleen, which is generally met with, and sometimes to a
very considerable degree. Hypertrophy of the liver, and of
the lymphatic glands, also occurs; the glands in the neck,
inguinal region, etc., are occasionally greatly swollen, and
contain numerous parasites.

The spleen and lymphatic glands are, undoubtedly, the

178 H. M. WOODCOCK.

organs which react most strongly to the parasites, and their
enlarged condition is probably to a great extent due to
enhanced activity in elaborating blood-corpuscles, leucocytes,
etc., to cope with the enemy. Javeran and Mesnil (40)
frequently noticed, in the peritoneal exudations of rats arti-
ficially infected with T. lewisi, instances of phagocytic
action by leucocytes upon the parasites (Fig. 6). Bradford

Fic. 6.—Phagocytosis of T. lewisi. In a the leucocyte is com-
mencing to engulf the Trypanosome ; in B the latter is completely
intracellular; c—r show the gradual dissolution of the parasite,
the two nuclear elements remaining longest recognisable. p=
parasite ; » = nucleus of leucocyte ; c = ingested blood-corpuscles ;
v = vacuoles remaining after their dissolution. (After Laveran

and Mesnil.)

and Plimmer (6) describe the same taking place in the spleen
of rats and mice infected with T. brucii, and also in the
blood of “spleenless ” animals (i.e. those from which this
organ has been extirpated). These authors conclude that,
at any rate in the earlier stages of the disease, a good deal of
phagocytic action takes place in the spleen. Castellani (15),
again, has observed phagocytosis of T. gambiense.

But it is probable, also, that the haematopoétic organs or
lymphatic glands secrete some chemical or physiological
substance which exerts a harmful action upon the Trypano-
somes and causes them to undergo involution, assuming

THE HAMOFLAGELLATES. 179

“amoeboid ” and ‘ plasmodial’”’ forms.'' These evidences of
commencing degeneration or slow death of the parasites are
often numerous in the spleen, lymphatic glands, and the bone
marrow (especially of spleenless hosts), and, of course, are
also met with in the general circulation. Bradford and
Plimmer say that they have observed phagocytosis only of
such forms, and not of typical adults. This is in favour of
the view that these forms are abnormal, already weakened by
some agency of the host, and, therefore, less capable of
resisting ingestion by the leucocytes. It is these altered
forms which are especially liable to block the cerebral capil-
laries. Their morphology will, however, be more conveniently
discussed after considering that of typical Trypanosomes.

SecTION V. GENERAL Account or ''RYPANOMORPHA (TRyY-
PANOSOMA) NOCTU# (CrLii anp San Feticz).

(A scheme indicating the principal phases of the life-
history, and serving as a general summary, is given on
p. 180.)

In the life of Trypanomorpha noctue,’ parasite of
Athene noctua and Culex pipiens, the T'rypanosome
phase is so frequently lost sight of—the parasite passing into
the Heemosporidian phase, when it takes the form of a Halte-
ridium—that certain stages in development are most easily

' McNeal (74) would assign the destruction of the Trypanosomes to the
cytolytic agents in the peritoneal fluid, which bring about their immobi-
lization and gradual solution, rather than to phagocytosis.

? This parasite has been selected as an example of the complete life-cycle
of a Trypanosome, not as intending to imply that it is in every way typical
of the majority, but because, when the plan of this article was arranged,
choice was limited to one of the two forms described by Schaudinn. Try-
panomorpha (and still more, Trypanosoma ziemanni) has advanced
further in the direction of the Hemosporidia than, for instance, Mammalian
Trypanosomes probably have. Nevertheless, there can be little doubt that,
as regards the chief features of its biology, morphology, and life-history,
this parasite may be considered as a representative Hemoflagellate (see
below, under ‘‘ Comparative account of the life-cycle”).

WOODCOCK.

M.

Ee

180

‘YO satp soyrusis || — puv £ pazvadat oq Lvur posopoue snyy [[e seytusts [ ] £ uorzIzedar seqeorpur “**'"**** S saATFBUIARTR
SOYyTUsis i fuoyesnfuos sayeorput +{ Suorstatp Aq ‘07 OSIX SOAIS SoylUSIS x fozuI sassed ‘samtodaq seyiusts <—

—:sjoquids jo uorjeurdxg |

Ce]

[Z| p (stseuesouoyzaed £q) ——
lea Gorg a ‘oyAoojoMIVSVSoUl AB[NGO]|Sopus UMO.LS-[[NF
‘Ly ] zeus 044 oyur sassed
(amosoued fry, ayeuray
sunos) eyfo0jamesesayy [¢]

‘(Ly ] yeus oy} our Surssed jo yuvysur oyy ur {[uo ynq) sajouesoroim g X o4yfoojamvSo10r [Z]

— ([!jun seqzeutezje ATMoys)"***''**vurserd url vary <— are[nqo;sopue <—

‘Le ] ayAoogoureSes
ne cubedarrt Ga ee [— somosourd dry -1ayyonep x" x “yy, 4ynpe

(@AT}0R) (UIIOFIULIV SAIS) ({L°V] yeus oy} woz)
roreserptserd UL aaty aposnd.109-pooyq 0} peyor}4e awosourddry, guateyrpur Sunox [T]

:(BNq0OU eUAY4V) [MO VT O43 UT “g

<— (soul, 9 sa}vu.te4]e)

se -osvyd-Sulysat <—

‘Te “a) [x0 043

oyut ‘sunok jr ‘sassed

yorya ‘[g] uiese
WLLIOF O[VUIAT
(stsauasouayyaed Aq)<—(ur10j1ut1esai8)
: asvyd-sutjsat —— (e}A00j,0uIvSeSout)
Lig) macy vane DU emosourdd1y, ajeuey —— [¢]
"|| — (seyemesoszorm) sautosoued dry, apeut g X (0}LoojyommvSoso1m) ulA0F ayer —— [Z]
“(( [e] ee x “ce ‘e e[vuley ths z
"({@] yg snoSojouoy) sumo; oyem —— yoy “sy-cozySnep Axis Bae X aarpor
‘TT “q] [80 049 oyur ssed
——(,, WIOFIUIIVSaIS ,,) Sutysat <— samosourd Ary,-19}zySnep x | "+" (aoissy Areurq Xq) x euosourdAry, yuatayrpur —— [T]

[2] WILOF 9] BUI

fe, i 9] BULay a ajoSk7 { ayouBsBseuL <— ‘ 2 op foopouvseseayy
ike rT yeu “c ——— + [Avo 94} FO poolq
“[T] toyovaeyo quoxeytput Jo oyouryoo OTRO L ape ace Te ee Tolan rac a es Ae eto

:(suetdid xajnQ) zeus oy} uy “Vy

*(G0ITE,T NVQ GNV ITIGO) WOALOON VHdAYONWONVdAUY, FO AIONO-GdITT AHL JO ANAHOG

THE HMMOFLAGELLATES. 181

and fittingly described by means of the conventional Hemo-
sporidian terminology.

It is most convenient to commence the account of the
complicated life-cycle with the zygote or copula, which results
from the union of a microgamete with a megagamete in the
stomach of the gnat. Even before nuclear fusion is complete,
the copula has acquired the vermiform, motile condition, in
which it is known as an ookinete (fig. 7). Its hyaline and
more refractive anterior end is capable of considerable and
rapid changes of shape, now being extended and pointed,
now bluntly rounded off. Behind this follows, usually, a
region containing one or more larger or smaller clear
vacuoles, then a denser cytoplasmic part with the nuclear
spindle, and, lastly, the rounded posterior end containing
pigment and other grains. The movements of the ookinete
are identical with those of the corresponding phase in other
malarial parasites. The cokinetes next proceed to get rid of
unnecessary material, including the pigment-grains and the
reduction-nuclei left over in the cytoplasm after fertilisation.
These are expelled from the hinder end, enclosed in a portion
of the cytoplasm, which is cut off at the same time, and forms
a gelatinous investment to the mass (figs. 8B and 5, 9B, and
10 8).

While this is going on, the complete fusion of the male and
female pronuclei into the definitive nucleus (“synkaryon”’)
is being slowly accomplished. This, however, by no means
includes all the nuclear apparatus present, which is of an
exceedingly complex character. In T’. noctuz the nuclear
material is highly differentiated and organised. Not only is
there a definite and constant number of chromosomes, but
the individuality or ‘‘ separateness ”’ of the principal nuclear
constituents, according to their particular function, is very
marked. ‘This is revealed in two distinct ways: (a) in the
sharp resolution of the nuclear material into trophic and

1 In order to avoid recapitulatory explanation of terms, readers are

reminded that a fuil account of malarial parasites is to be found under the
heading “ Hemosporidia,” in Minchin’s article (1. c.) on the Sporozoa.

voL, 50, pART 1.—NEW SERIES. 13

182 H. M. WOODCOCK.

kinetic constituents, which are practically separate and inde-
pendent, at any rate, during the trypanosome phase; and
(b) in the looseness of the union between the male and female
elements. The fertilisation spindle or definitive nucleus is to
be regarded as representing the trophic portion, and it will
be convenient, therefore, to distinguish it as the tropho-
nucleus. When reconstituted and in the resting condition
this body possesses, as the normal nuinber, eight distinct

Fic. 7.—Ookinete of Trypanomorpha (or Halteridium)
noctuz (Celli and San Felice). The fertilisation spindle has not
yet become rounded off (see text). k, k’ = kinetonuclear elements
of the male and female gametes, not yet united; v = vacuole;
v.12. = residual nuclei and pigment grains. (After Schaudinn.)

chromatic aggregations or chromosomes (fig. 8 A, t.chr.).1
Close to either end of the spindle is another chromatic body
(fig. 7, k, k’). These two masses also come from the micro-
and megagamete respectively. They proceed to fuse, and
the resulting body, which may be termed the kinetonucleus,
passes into the now rounded trophonucleus, where it takes up
a central position. The kinetonucleus also possesses eight
peripherally situated chromosomes (fig. 8 a, k.chr.), embedded

in a plastinoid matrix; near its centre lies a centrosomic
granule (c), surrounded by a clear zone. As the name

' Strictly speaking, these ought perhaps to be regarded as half-chromo-
somes,

THE H#MOFLAGELLATES. 183

implies, the kinetonucleus represents that part of the nuclear
material which regulates the locomotor activities of the cell.

Different ookinetes, though, at first, essentially similar in
nuclear constitution, exhibit considerable variation in the
minute structure of the cytoplasm, the quantity of reserve
material this contains, and the size of the body as compared
with that of the nucleus. In fact, the ookinetes are differen-
tiated into three distinct types, leading to the formation of
indifferent, male, or female Trypanosomes, with widely different
subsequent histories; and these cytoplasmic variations afford
the earliest indications of the direction of further development
in any given case. Schaudinn considers that this variability
in character stands in intimate relation with the previous
history of the sex-cells (gametes) which may have been, as
will be seen later, extremely diversified.

(4) Ookinete of Indifferent Character and its
subsequent history in the Gnat.

The cytoplasm of an indifferent ookinete (fig. 8) is fairly
clear and faintly staining; it usually possesses one or two
large vacuoles in the anterior part, but little, if any, reserve
material is noticeable. The pigment grains in the hinder
region may not all be eliminated at the first attempt, a
second expulsion sometimes taking place (fig. 8 ©), which
leaves the cytoplasm free. Meanwhile, important nuclear
changes are occurring. The kinetonucleus becomes amceboid
and gives up its material to the trophonucleus (8). The
result is that the eight chromatic elements of the former
become united, by the aid of the plastin basis, with those of
the latter, leaving the above-mentioned granule in the
middle. This granule divides in a dumb-bell-like manner,
producing a small axial spindle (fig. 8 ©, a.s.), around which
the eight compound chromosomes arrange themselves.
These next split, and the halves pass to either end, forming
a diaster which is markedly heteropolar (c). The right

184. H. M. WOODCOCK.

(or dorsal) half, the kinetic’ portion, is perceptibly smaller,

Fie. 8.—Development of an indifferent Trypanosome from an
ookinete of indifferent character. ¢.chr. = trophonuclear chromo-
some; k.chr. = kinetonuclear do.; c= centrosomic granule; @.s.
= first axial spindle; a.s?, a.s* = second and third do.; ¢=
trophonucleus; 4= kinetonucleus; k.c. = kinetonuclear centro-
some; t.c. = trophonuclear do. ; m = myonemes; f.b. = flagellar
border of undulating membrane (third axial spindle) ; c* = its
proximal centrosome. (After Schaudinn.)

1 It is here inferred (though not explicitly stated by the author) that the
splitting merely separates again the trophic and kinetic constituents which
had united, male and female kinetic elements, on the one hand, and trophic
ones, on the other, remaining together. There is no evidence that, in the
indifferent ookinete, each of the two nuclei thus formed contains both
trophic and kinetic parts (i. e., that each is entirely unisexual, one having
male, the other female chromosomes only). This is the more unlikely,
since, in that case, the nuclei of the indifferent Trypanosomes would not be
comparable with those of the others (cf. below, pp. 188 and 191).

THE HAMOFLAGELLATES. 185

but denser, richer in plastin, and more deeply staining than
the other.

In this manner, therefore, two distinct nuclear bodies are
formed, of different size and structure. The larger one,
lying nearer the middle of the body, which rapidly recon-
stitutes itself (fig. 8p), is the trophonucleus; its chromo-
somes have become long, winding threads. The eight
chromosomes of the other, the kinetonucleus, are only to be
made out distinctly in maceration preparations. Situated
near the middle of each nucleus is a small centrosomic grain.
These two intra-nuclear centrosomes are connected by a fine
achromatic thread, which represents the remains of the axial
spindle.

While the trophonucleus now enters upon a resting phase,
the kinetonucleus proceeds to form the characteristic loco-
motor apparatus of the Trypanosome. It passes forward
slightly, and takes up a position at the periphery of the
endoplasm, lying, indeed, against the limiting ectoplasm."
Its centrosome divides again in a dumb-bell-like manner,’
and forms another axial spindle (1, a.s.”) at right angles, as
before, to the length of the organism. ‘The chromatin be-
comes aggregated round either end, again in a slightly hetero-
polar fashion, the dorsal half being somewhat the smaller.
There are now two daughter-kinetonuclei, the one on the
right being quite in the ectoplasm; they also remain con-
nected together by the drawn-out central spindle, which, as
before, joins the two centrosomes, The peripheral daughter-
nucleus forms yet another spindle (fig. 8 F, a.s.°), whose axis
is now, however, longitudinal. This assumes large propor-
tions, and spreads forward to the anterior end of the body

1 Schaudinn thus mentions the existence of a definite ectoplasmic layer
but does not describe it, nor do his figures indicate it (apart from the
undulating-membrane) at all clearly ; see, however, below, p. 209, et seq.

2 The behaviour of this organella strongly recalls that of the ‘‘nucleolo-
centrosome” of Euglena, during division, and also that of the karyo-
some of many Coccidia in schizogony; quite possibly, it should be con-
sidered in the double light of a “karyo-centrosome.”

186 H. M. WOODCOCK.

(a), the whole lying im the ectoplasm, which becomes greatly
developed to form the undulating membrane.

In T. noctuz the undulating membrane arises by the
anterior part of the body becoming much flattened laterally,
and, to a certain extent, drawn out dorsally by the spindle,
the two ectoplasmic surfaces thus coming close together.
The well-developed, sinuous, axial spindle has now become
excentric in position, and strengthens, or rather itself con-
stitutes the free (dorsal) edge of the membrane, forming a
flagellar border to the same (H, f.b.) A supporting frame-
work is formed by eight myonemes, representing the eight
elongated chromosomes, four of them being arranged on
each lateral surface. The flagellar spindle does not stop on
reaching the anterior limit of the body, but continues to
elongate, drawing out with it the undulating membrane,
which narrows and finally thins out. The myonemes then
unite with the spindle to form the free flagellum, which
eradually tapers away at its distal extremity. ‘The centro-
some at this end disappears, as such, but that at the basal
or posterior end of the spindle persists (c*). The other
daughter-kinetonucleus has now become rounded off as the
functional kinetonucleus (k) ; it remains connected with the
complicated locomotor apparatus by means of the delicate
thread which represents the second axial spindle. In older
stages the kinetonucleus may pass backwards behind the
trophonucleus (¢), pulling with it the associated structures,
which thus become even more extended (cf. fig. 13.4).

The characteristic Trypanosome-form is now attained, and
the “ indifferent”? parasite next enters upon a period of mul-
tiplication by binary fission. ‘The details of the process are,
Schaudinn says, too complicated for explanation without the
aid of numerous figures, so that only the main features can
be outlined here. The division of the nuclear apparatus is
the first to occur; either the tropho- or the kineto-nucleus
may lead the way. This is followed by the duplication of
the locomotor apparatus, which begins at its basal end,
starting either from the still undivided kinetonucleus or

THE HMMOFLAGELLATES. 187

from the new daughter-one. The new flagellum, myonemes,
etc., are laid down independently alongside and _ parallel
with the old organelle, arising, exactly as these did, by the
great extension of a kinetonuclear spindle. Lastly, the
general cytoplasm divides, and two _ practically equal
daughter-Trypanosomes result.

After active movement and multiplication (the latter taking
place without any loss of motility) have continued for some
time a resting condition succeeds. The parasites now
become gregariniform, and strongly recall the similar phase
described by Léger (63 and 68) in certain Herpetomonads.!
The Trypanosome bores into an epithelial cell of the stomach
by means of its flagellum, which becomes reduced to a short,
rod-lke organ, serving to anchor the parasite firmly. Binary
fission may go on during this gregariniform condition, and
this often leads to the formation of a dense layer of parasites
all attached to the epithelium. Besides this superficial attach-
ment they may also penetrate far in between the cells, when
they assume a rounded form, aud lose all traces of the
flagellum. Upon the 'l'rypanosomes again becoming active,
or trypaniform, the flagellar apparatus is re-constituted
by the kinetonucleus.

This alternation of resting and active periods, accompanied
by division, has a limit, dependent upon internal causes
(within the parasites themselves) and external ones (due to
the reactions of the host). The course of succeeding events
may be very varied. The indifferent forms may pass into
the blood of the owl, or they may, in certain circumstances
which Schaudinn was unable to ascertain, lose their indifferent
character and become sexual, either male or female, having
the same subsequent development as the male or female
Trypanosomes resulting from ookinetes of corresponding
character. Finally, if hunger ensue and the gnat is unable
to make another meal, the indifferent Trypanosomes gradu-
ally die off.

1 See below, in Section I1.

188 H. M. WOODCOCK.

(s) Ookinete of Male Character and its subsequent
development in the Gnat.

Ookinetes which will produce male forms are easily dis-
tinguishable from those of indifferent character. The cyto-
plasm (fig. 9) is almost hyaline, and much clearer than in
the indifferent ookinetes, which occupy, in this respect, a
position intermediate between the other two forms. Reserve
materials are completely lacking. The body is smaller than
that of either of the other kinds of ookinete; the nuclear
apparatus is, however, much larger relatively to the cytoplasm,
and very rich in chromatin. The earliest nuclear changes
which take place are, apparently, similar to those above
described, and lead to the union of the chromatin of the two
nuclear constituents. A heteropolar spindle is next formed,
and the chromatic elements divide, half being drawn to
either end.

It is clear from the subsequent development, however,
that this division is in a different sense from the corre-
sponding one in the indifferent ookinetes. Instead of the
trophic constituents being separated from the kinetic ones,
we must consider the male elements of both kinds as being
grouped together and separated from the female ones. The
smaller, more condensed half (9 8B) is entirely male in cha-
racter, while the larger, looser half is of female sex. This
latter nuclear body, which remains centrally situated (c and
D, f.n.), is not to be regarded, therefore, merely as a tro-
phonucleus, but as containing the female elements belonging
to both tropho- and kineto-nucleus. It rounds itself off, but
takes no further part in development, gradually disappearing
in situ and being finally left behind with the unused cyto-
plasm (fig. 9r). The male nucleus, on the other hand,
divides successively to form eight nuclei (c and p, m.n.),
which become uniformly distributed throughout the body of
the microgametocyte, as the male ookinete may now be
termed. Each of these nuclei is, moreover, double, the

THE HAMOFLAGELLATES. 189

kinetic and trophic portions of each having separated to form
a kineto- and a tropho-nucleus respectively, the former being,
in this case, almost as large as the latter (p, m.k. and m.t.).
The microgametocyte itself never becomes trypaniform.
Its cytoplasm assumes a rounded shape, and the eight double
nuclei pass to the periphery and there take up a radial

Fie. 9.—Development of microgametocyte and male Trypano-
somes from an ookinete of male character. m.n. = male nuclei;
fu. = degenerating female nucleus; m.f. = male trophonucleus ;
m.k. = male kinetonucleus; M.7. = Male Trypanosome; 7.0.
= residual body. (After Schaudinn.)

position (£), the kinetonuclei being nearest the surface. The
superficial cytoplasm opposite each forms a little prominence
or hillock; these eight elevations grow out (accompanied
by the eight nuclei, one to each), become narrower as they
lengthen, and gradually assume the typical Trypanosome
shape.! The kinetonucleus of each is producing, meanwhile,

1 This process, it is interesting to note, takes place in a manner quite
comparable to that by which the schizogony of Coccidia is effected.

190 H. M. WOODCOCK.

the locomotor apparatus exactly as in the case of an indifferent
Trypanosome, and finally the eight little male Trypanosomes
(r, M.7.) break away from the central residual mass. They
are easily distinguishable from the other forms by their
minute size and by their flagellar apparatus, which is, rela-
tively, much more strongly developed, and gives them an
unusual degree of activity.

Schaudinn finds that these male Trypanosomes (homo-
logous with microgametes) in the gnat are quite incapable of
further development. ‘They cannot divide and soon die off,
even though they pass into the blood of the owl. It should
be pointed out that the trophonuclei of the male Trypano-
somes have already undergone reduction during the second
nuclear division in the parent gametocyte,' and now possess
only four chromosomes each; an important difference in this
respect is shown by the kinetonuclei, which have not been
reduced (cf. below, p. 197). Schaudinn considers that the
reason for the inability of a male Trypanosome to live inde-
pendently is to be found in this early reduction of its tropho-
nucleus and the consequent derangement of metabolism.

(c) Ookinete of Female Character and its subse-
quent history in the Gnat.

The cytoplasm of a female ookinete (fig. 10 a) is fairly
dense, with dark staining bodies of reserve material. The
nucleus is somewhat smaller, relatively to the size of the body,
than in an indifferent ookinete. Nuclear changes bring about
the formation of a male and female nucleus, exactly as de-
scribed for the male ookinete, but in the present instance it
is the male nucleus which ultimately degenerates. As before,
this gives rise, by successive divisions, to eight daughter-nuclei,
which pass into the hinder region of the body (B and c, m.n.).
Each of them becomes distinctly separated into trophic and
kinetic portions, but after this the eight double-nuclei gradu-

1 The author leaves a minute description of the process for his detailed
memoir.

THE H#MOFLAGELLATES. 191

ally fade away and are eventually dissolved up in the cyto-
plasm. Concurrently, the large female nucleus has behaved
in a manner recalling the division of the original compound
nucleus of an indifferent ookinete, giving rise, by a hetero-
polar division, to a tropho- and a kineto-nucleus (c). The
latter proceeds to form a complete but somewhat feebly-
developed locomotor apparatus (D).

A female Trypanosome differs from an indifferent one by
its plumper shape and its denser, more deeply-staining cyto-

Fig. 10.—Development of a female Trypanosome from an
ookinete of female character. m.n.= degenerating male nuclei ;
a.sp. = first axial spindle of female nucleus; ft. = female tropho-
nucleus ; 7.4. = female kinetonucleus. (After Schaudinn.)

plasm, containing granular reserve nutriment. ‘The kineto-
nucleus is smaller and the flagellum shorter; hence the
movements of the parasites are feebler and slower, and they
soon pass, for a time, into the attached, resting-phase,
characterised, as before, by the retrogression of the locomotor
apparatus. These forms appear to have quite lost the capacity
of longitudinal division, either in the trypaniform or gre-
gariniform phase. Growth is accompanied by a considerable

192 H, M. WOODCOCK.

accumulation of reserve material, and the parasite’s dimen-
sions may attain thrice those of an indifferent Trypanosome.
The older forms are no longer able to pass from the passive
into the active condition, and can only perform slow movements
of contraction, flexion, and the like.

In consequence of their reserve stores the adult females, in
the gregariniform phase, are exceptionally resistant to ex-
ternal influences, and the most able to withstand unfavourable
circumstances. During long hunger-periods of the gnat all
the other stages of the parasite die off. Only the females,
deeply seated between or beneath the epithelial cells, remain
alive, slowly using up their supply of nutriment. With the
advent of fresh blood into the stomach they undergo partheno-
genesis, at the end of which the parasites are able to become
either indifferent, male, or female forms again, the result
being that once more fresh generations of Trypanosomes
overrun the alimentary canal. It is, moreover, the gregarini-
form females which bring about the infection of succeeding
generations of the gnat, remaining dormant in the ovaries
throughout the winter until the eggs are laid and the larvee
develop in the following spring.

Parthenogenesis.—The cytoplasm of a parasite about
to undergo parthenogenesis is poorer in reserve material, and
now more or less vacuolated. No trace of the locomotor
apparatus is visible, and the kinetonucleus hes in contact
with the trophonucleus (fig. 11 4). The centrosome of the
latter is now surrounded by a chromatic body which some-
what resembles the kinetonucleus when occupying this posi-
tion in the compound nucleus of an ordinary ookinete! (cf.
figs. 8 A, and 9 a); and the whole nucleus next undergoes a
process apparently similar to that which occurs in an in-
different ookinete at that time. The formation of a hetero-
polar spindle leads to the separation of a smaller (kinetic ?)
half from a larger half, the latter rapidly reacquiring a

1 Perhaps this body represents an increase, during the resting-period, of
kinetic nuclear constituents, in which a female form is apparently more
deficient than either a male or an indifferent form.

THE H#=MOFLAGELLATES. 193

resting condition. There are now two almost equal-sized
nuclei lying contiguous to, and on opposite sides of, the larger,
central, probably trophic one, namely, the old kinetonucleus
and the newly-formed body. Hach of them next divides
twice (the old kinetonucleus may have commenced before, as
in [B]), entting off successively two reduction-nuclei (c and
D, 7.2.), Which are gradually absorbed by the cytoplasm. ‘he
reduced kinetonuclear elements penetrate into the resting
trophonucleus from opposite sides (D, 7.k.e.), and fuse to form

Fie..11.—Parthenogenesis of a gregariniform female. 7.n. =
residual nuclei; 7.k.e. = reduced kinetonuclear elements; e.n. =
compound nucleus, equivalent to that of an ookinete. (After
Schaudinn.)

the central kmetonucleus (£ and F), exactly as do the reduced
kinetonuclei of the male and female gametes, after fertilisa-
tion. After this double process of parthenogenesis and
“self-fertilisation,’ the gregariniform parasite can develop
along any of the three lines above indicated. The actual
course taken is, doubtless, largely determined by the existing
condition of the cytoplasm as regards nutritive material, and
by its size relative to that of the nucleus.

The three types of parasite above described include all the
varieties of form met with in the gnat. From the standpoint
of reproduction the indifferent Trypanosomes are by far the
most important, the capacity for longitudinal fission being,

194 H. M. WOODCOCK.

indeed, limited to them. Hence these forms largely pre-
dominate in number, and many of them, especially small ones
resulting towards the end of a multiplication-period, go to
swell the ranks of male and female parasites. Correspond-
ingly, while all the types can pass into the blood of the owl,
the great majority of those which do so are indifferent forms.
Before considering the Trypanosomes in the blood of the
bird, however, there is one very important characteristic of
the parasites which must be mentioned.
Agglomeration.—In common with many other Trypano-
somes, ‘'rypanomorpha noctuz possesses the capacity
for agglomeration. Agglomeration takes place upon the
advent of unfavourable conditions (e. g. a period of hunger)

Fie. 12.—Cluster of agglomerated male Trypanosomes in the
intestine of the gnat. (After Schaudinn.)

in the environment of the parasites, and all three types
possess this faculty; it occurs to the greatest extent, how-
ever, among the male and indifferent forms. Agglomera-
tion consists in the grouping or union together of the
Trypanosomes around a common centre; this leads to the
formation of rosette-like clusters (fig. 12), or even large
masses. The parasites are invariably attached by the same
end, which is, in the case of Try panomorpha, the anterior,
flagellate end; so that all the flagella are directed inwards,
towards the centre of the rosette. If the unfavourable
conditions continue the parasites remain in this agglomerated
state until they die and disintegrate. If, however, a favour-
able change suddenly sets in, the Trypanosomes are able to
undergo the inverse process, namely, disagglomeration.

THE H#EMOFLAGELLATES. 195

The parasites free themselves from one another, and the
rosette is dissolved. The entire set of phenomena also occurs
in the blood of the owl. Schaudinn ventures no conjectures
with regard to its biological significance.

(vp) The Behaviour and Subsequent Development of
the Trypanosomes in the Vertebrate Host.

All the Trypanosomes met with in the blood of the owl are
easily recognised as belonging to one or other of the three
categories found in the gnat. Even though the parasites in
any given phase may not exactly agree in the two hosts so far
as minute detail is concerned, a study of their previous and
subsequent history in both cases renders it, nevertheless,
manifest that the two forms are the homologues of each
other; and, in fact, in all essential characters the 'l'rypano-
somes, aS seen in the owl, agree with the corresponding
type in the gnat. One distinction which may be noted
is the presence, in the former case, of pigment in the cyto-
plasm, produced as the result of the alteration of the hemo-
globin of the red blood-corpuscles.

(1) Indifferent Trypanosomes.—The larger forms
entering the blood continue to divide until the size of
their descendants is sufficiently reduced. The small ones
attach themselves directly to the erythrocytes, and enter
upon a period of rest and growth. The mode of attachment
has been described above (p. 171), and is shown in fig. 13 a.
The locomotor apparatus disappears, and the kinetonucleus
takes up a position in close contact with the trophonucleus
(Bp and c). The form of the body is now quite that of a
young Halteridium, and, after twenty-four hours, the
first pigment grains appear in the cytoplasm. By this
time the parasite has increased to about double its original
size. It now becomes vermiform and active, reconstitutes its
flagellum, etc., and leaves the host-cell (p), usually in the

196 H. M. WOODCOCK.

night time,' to become once more atypical Trypanomorpha
(x). After a (short?) period of movement free in the
plasma, the parasite again becomes attached, resumes the
Halteridium form, and grows until the next night, when
the above changes recur. This cycle is repeated for six
days, until the full size of the organism is attained (F and
Gg). The adult Trypanosome then undergoes, in the active,
trypaniform condition, a number of successive longitudinal
divisions, until the resulting daughter-Trypanosomes have

Fie. 13.—Stages in the growth of an indifferent Trypanosome
in the blood of the owl. = nucleus of red blood-corpuscles ;
p = young intra-cellular parasite. (After Schaudinn.)

at length reached a minimum size, whereupon the process of
attachment and growth is begun anew. It is important to
note that Schaudinn never observed any division of the
parasites when in the gregariniform (Halteridium) con-
dition by multiple fission or schizogony, such as occurs in
other Halteridia aud Heemosporidia generally.”

(2) Microgametocytes and Microgametes (male
forms and male Trypanosomes).—If any little male
Trypanosomes from the gnat do succeed in gaining an entry

1 The reason being, Schaudinn thinks, the fall in temperature of the
bird, which occurs at this time.

2 As the author remarks, this is probably of later phylogenetic develop-
ment (see below, in Section IT).

THE H#MOFLAGELLATES. 197

nto the blood they soon die off. The microgametocytes,
here, arise from very young indifferent Trypanosomes, which,
as they grow, reveal, by their pale, clear cytoplasm, coarse
pigment, etc., their assumption of male characters. A careful
examination of the adult microgametocyte shows that the
apparently single nucleus, which previous authors have
united in describing as, relatively, very large, is, in reality,
a highly complex organisation, and consists of eight groups
of double-nuclei (i. e. tropho- and kineto-nuclei in close
association, compare above p. 189) aggregated together.
As in the case of the corresponding forms in the gnat, the
trophonuclear elements have undergone reduction, and now
possess only four chromosomes each; the kinetonuclei still
have the normal number. The formation of the eight micro-
gametes and their separation from the parent-individual takes
place in the manner already described.

Each microgamete is a very specialised organism, as is seen °
from fig. 14 a, and the accompanying diagram. The body
is extremely slim and tapering, especially at the posterior
end, in marked contrast to that of a male Trypanosome in
the gnat; the anterior end, on the other hand, is not drawn
out, but acutely conical. The trophonucleus is greatly elon-
gated and has the form of a very long thread, extending
nearly the whole length of the body (fig. 14 4, ¢.), to which
it serves as an axis. he four chromosomes are strung
upon it like beads, at regular intervals. The kinetonucleus
(k.) is also somewhat elongated and shows distinctly eight
chromosomes and an intranuclear centrosome (k.c.). It
should be noted that the “tail”? end is not the flagellar
end, but a posterior, whip-like extension of the cytoplasm
and trophic nucleus, and it has no relation whatever to the
locomotor nucleus and apparatus. As a matter of fact there
is here no free flagellum; in other words, the kinetonuclear
spindle, which has formed in the usual manner the strongly
developed border of the undulating membrane (a.sp.*) and
the eight strengthening myonemes (m.), does not extend
anteriorly beyond the limits of the cytoplasm. It ends in a

voL. 50, paRT 1,.—NEW SERIES. 14

198 H. M. WOODCOCK.

distinct centrosome (a.c.), which, as above-mentioned, dis-
appears during the formation of the flagellum in the other
trypaniform types. Posteriorly, the locomotor apparatus also
terminates in a centrosome (p.c.) just in front of the last
trophonuclear chromosome. ‘This centrosome is connected

Fie. 14.—A. Structure of a microgamete. B. Schematic ground-
plan of the same. = trophonucleus (already reduced); k=
kinetonucleus ; k.c. = kinetonuclear centrosome; a.sp? = second
axial spindle; a.sp?= third do. (flagellar border of undulating mem-
brane); a@.c. = anterior centrosome of same; p.c. = posterior do. ;
m = myoneme; ¢ (below) = tail-like prolongation of the body
posteriorly. (After Schaudinn.)

with that of the kinetonucleus by a thread (a.sp.*) repre-
senting the second axial spindle (see above, p. 185). (The
mutual relations of the different nuclear constituents are
perhaps, more easily apprehended from the schematic ground-
plan).

(3) Megagametocytes (female Trypanosomes).—

THE H#MOFLAGELLATES. 199

Large female forms laden with reserve materials are unable
to pass through the proboscis of the gnat. Young females,
on arriving in the blood, at once penetrate into the erythro-
cytes. Growth is much slower than in the case of the in-
different forms. Moreover, the parasites change host-cells
less frequently, and the older ones appear to be no longer
able to assume the Trypanomorpha-form. Such indivi-
duals leave one red corpuscle, wander about in the plasma,
and then pass into another corpuscle while in the gregarini-
form condition. A ripe adult megagametocyte is incapable
of movement, and remains enclosed by the now pallid and
disorganised host-cell, the nucleus of which has been pushed
to one side. Its general structure is already well-known
through the older researches on Halteridium. Schaudinn
points out that, in contact with the relatively small nucleus
(trophonucleus), there can be seen a correspondingly small
kinetonucleus.

As in the gnat the female Trypanosomes are. the only
forms able to survive unfavourable circumstances, so here, in
the blood of the bird, the megagametocytes alone remain
when the indifferent Trypanosomes and microgametocytes
have all died off. Similarly, they are able to cause, at
intervals, a recurrence of the infection by undergoing the
process of parthenogenetic development above described, in
the same way that the recurrence of malaria is brought about
in the case of Plasmodium (see Schaudinn [97]).

(4) Maturation of the Megagametocyte, and Ferti-
lisation of the Megagamete by the Microgamete.—
Maturation and fertilisation do not take place until the
sexual forms are transferred, with the blood, to the ali-
mentary canal of the gnat. The main outlines of the process
have been well described by MacCallum (72) for another
species of Halteridium, so that our present author directs
attention more especially to the cytological details.

As soon as the megagametocyte leaves the warm-blooded
host it becomes rounded off, ruptures the delicate envelope of
the host-cell still surrounding it, and is thus set free. The

200 H. M. WOODCOCK.

chromatin of the trophonucleus becomes arranged in a long,
spirally-wound thread. Its centrosome has disappeared.
The chromatic thread is next segmented, both by longitu-
dinal and transverse divisions, so that four separate tetrads
result. Meanwhile the kinetonucleus has passed inside the
trophonucleus, and appears to serve as the spindle for the
reduction-divisions of the latter. After the first mitosis each
resulting half has, of course, four dyads instead of four
tetrads. In the second division of the germinal nucleus these
four dyads are split into monads, or single chromosomes, in

Fig 15.—A ripe megagamete liberated from its host-cell (lying to
the left). 2 = nucleus of disintegrated corpuscle; ¢ = reduced
trophonucleus; / = reduced kinetonucleus; 7. = one of the
degenerating reduction-nuclei. (After Schaudinn.)
the usual way. After this the kinetonuclear part resumes its
old position outside, but contiguous to, the reduced tropho-
nucleus. How, exactly, the kinetonucleus becomes reduced
Schaudinn was not able to determine. The end-result of the
process is seen in fig. 15. To the right lies the spherical
female trophonucleus (¢.), with its four thread-like chromo-
somes, and below it the kinetonucleus (k.), consisting of five
deeply-staining masses, namely, the four chromosomes and the
centrosome.' In the middle lie the two reduction-nuclei (7.7.).
1 It is important to note that, in both gametes, the two kinds of nuclear
element are present. Hence, on the one hand, neither is the trophonucleus
solely female in character nor the kinetonucleus solely male; and, on the
other hand, neither is the former merely somatic, nor the latter purely
sexual,

THE HAMOFLAGELLATES. 201

The female element is now a ripe megagamete or ovum
ready for fertilisation. The microgamete penetrates it at a
receptive cone, which arises from the cytoplasm on the side
where the female nuclei are situated. Its flagellar apparatus
disintegrates and disappears, and, in fact, the only parts
remaining distinct are the reduced male trophonucleus and
the still unreduced male kinetonucleus.!’ The latter next
undergoes two reduction-divisions (not clearly made out),
and then the two trophonuclei unite to form the well-known
elongated ‘ fusion-spindle,” the kinetonuclei taking up a
position at either end of it, as in fig. 7. With this act
the zygote or copula arrives at the stage with which
this description began, and the complete cycle is now
accomplished.

Ssecrion VI. Comparative MorpHoiocy or TRYPANOSOMES.

‘The body varies greatly with regard to size. Hven in one
and the same species this is frequently noticeable, particu-
larly under different conditions of life; and since, moreover,
different authors often give different estimates of the size of
a particular parasite, it is evident that any dimensions given
for purposes of comparison can only be considered as approxi-
mate. The common Trypanosoma rotatorium of frogs
(fig. 174 and 8) is, taking it all in all, one of the largest
forms so far described. Its length? varies from 40—60 p,°
while its greatest width dorso-ventrally * is from 8—80 p;
in the very wide examples breadth is gained more or less at
the expense of length. Conversely, T. gambiense, the
human parasite (fig. 16 c), is one of the smallest forms known.

See footnote on previous page.
The length is always inclusive of the flagellum, unless otherwise stated.
The forms known as T. mega and T. karyozeukton, which are
closely allied, but probably distinct, species are somewhat longer.

4 Adopting Léger’s convention by which the convex side, bearing the
undulating membrane, is distinguished as dorsal; the measurements of
width always include the undulating membrane.

w i> ~

202 H. M. WOODCOCK.

Its average length is about 21—23y, and its width 13—2 yp.
The majority of Mammalian Trypanosomes are fairly uniform

Fie. 16.—Representative Mammalian, Avian, and Reptilian
Trypanosomes, to illustrate the chief morphological characters.
The figures (excepting E) are all drawn to the original magnifica-
tion, given where stated. a, Trypanosoma lewisi, after Brad-
ford and Plimmer; B, T. brucii, after Lav. and Mesn., x 2000;
c, T. gambiense (blood, 7.-fever), after Bruce and Nabarro; p,
T. equinum, after L. and M., x 2000; 8, Trypanomorpha
(Trypanosoma) noctue, after Schaudinn; r, T. avium, after
L. and M.; eg, Hanna’s J. sp. from Indian pigeons; H, T. zie-
manni, after Schaudinn; s, T. damonia, after L. and M., x 2000.
c.g. = chromatoid grains; v. = vacuole; l.s. = fold or striation.

in size (fig. 16 a, B, D), the only noteworthy exception being
T. theileri (fig. 49), which is much larger than the rest,

THE HEMOFLAGELLATES. 208

varying, indeed, from 30—65 , in length. The Piscine Try-
panosomes, on the other hand, though possessing an equally
great range, exhibit amuch more regular gradation. Starting
with relatively small forms like T. remaki, var. parva, with
a medium length of 30 yp, parasites of all sizes are to be met
with up to T. granulosum (fig. 17k) and T. raje, which
are among the longest Trypanosomes known, attaining a
length of 80.

There is equally great variation in respect of form. Typi-
cally, the body is elongated, more or less curved and spindle-
shaped, and tends to be slightly compressed laterally. It
may be, however, anything from extremely slender or vermi-
form, to thick-set and stumpy; while, in some cases, the
parasites show little or no trace of the spindle-form, but
are squat and elliptical. Some authors are inclined to group
the parasites according to type of form; the writer does not
think, however, that anything is to be gained by so doing.
It is very difficult to draw any hard and fast distinctions,
because of the individual variation. Apart from the fact
that a fully-grown adult, ready to divide, is, in many cases,
very much plumper than a young adult (cf. T. lewisi, fig.
274 and B), there can be no doubt that considerable poly-
morphism! also sometimes occurs; illustrations of this are
given below. Some of the chief variations in form, as found
in the different groups of Vertebrates, may now be discussed
a little more fully ; it will be seen that no particular type can
be said to be peculiar to a given class of hosts.

On the whole the greatest uniformity is seen in the
Mammalian and Piscine Trypanosomes. Among the former
the typical fusiform shape prevails. The parasites may be
very slender (as in T. lewisi, fig. 16 a, and some forms of T.
gambiense, fig. 48 B), fairly so (as in the majority of cases,
fig. 16 B—p), or relatively thick-set (I. transvaaliense,
fig.50.c). The animals are usually either crescentic (fig. 16 B)
or sickle-like (fig. 16 p). Piscine Trypanosomes are nearly
always very elongated and often, relatively, quite as thin as,

1 This is, of course, quite apart from degeneration and involution forms.

Fic. 17.—Representative Amphibian and Piscine Trypanosomes. Mag-
nification as in Fig. 16 (except D). A and B, Trypanosoma rotatorium,
after Lav. and Mesn., x 1400; c, T. inopinatum, after Sergent, x 1000;
p, T. karyozeukton, after Dutt. and Todd, x 1000; 8, I. nelsprui-
tense, after L. and M, x 2000; F and a, Trypanoplasma borreli
(living and stained), after Léger; u, I’. cyprini, after Plehn; gs, Try-
panosoma solee, after L. and M., x 2000; x, T. granulosum, after
L. and M.; 1, T. remaki, var. magna, after L. and M., x 2000. h. =
clear zone or halo around kinetonucleus ; ch. = chain of chromatic rodlets
running from trophonucleus to kinetonucleus; afl. = anterior flagellum ;
pj. = posterior do.; Us. = longitudinal striations or myonemes; v. =
cytoplasmic vacuole.

THE H#MOFLAGELLATES. 205

or even thinner than T. lewisi; T. granulosum (fig. 17 x),
for instance, is extremely worm-like. In Piscine forms
owing, probably, to their great length, the body is frequently
coiled or rolled up on itself, as in T, solee (fig. 17 3), and T.
rajx (fig. 62s). Trypanoplasma, as exemplified by T.
borreli (fig. 17 F, a), differs from the majority of Piscine Try-
panosomes in being short and relatively broad ; the length
(of the body alone) is 20—22y, and the width 33—44 yp.

Coming to the Amphibian parasites, T. inopinatum (fig.
17 c) somewhat resembles a Mammalian Trypanosome, and
T.nelspruitense (17 u) a Piscine form; finally, there is the
well-known T. rotatorium, which, when most T'rypanosome-
like, has the form of a thick spindle (17 8). Frequently,
however, this parasite is greatly flattened out, and, conse-
quently, very broad and stumpy (figs. 17 a and 56). A
similar polymorphism of form is described by Léger (64) in
Trypanoplasma borreli. Certain individuals are more
massive than others, and often, indeed, very short and
squat (fig. 18 8).

There can be no doubt that, in some cases at any rate, this
variation indicates a difference in sexuality. We have already
seen that this is so in Trypanomorpha; and Léger inter-
prets, in the same manner, the broad distinctions between
different individuals of Try panoplasma just mentioned (see
also below, in Section IX 8B).

It is particularly necessary to bear this factor in mind in
considering the Avian Trypanosomes. For it is here, perhaps,
that we find the extreme types of form; and we are, unfortu-
nately, at present largely in the dark as to how far these
represent different parasites and how far polymorphism,
That one and the same species may appear entirely different
in different phases of the life-history is manifest on com-
paring the chief “forms” of Trypanosoma ziemanni
described by Schaudinn. The asexual or indifferent type is
extremely thread-like (fig. 16 H), and, indeed, greatly resem-
bles a Spirocheta. On the other hand both male and
female individuals (the latter especially) have the form of a

206 H. M. WOODCOCK.

wide spindle (fig. 334 and p). In the case of other Avian
Trypanosomes, whose life-cycle has not yet been worked out,
the parasites have generally been given distinct names
according to their appearance when observed. ‘Thus, the
attenuated or “spirochetiform” type is exemplified by T.
johnstoni (fig. 51) from a small bird (Estrelda) in Sene-
gambia; while the broad fusiform type is represented by
Hanna’s Trypanosome from Indian birds, and particularly by
Dutton and Todd’s parasite (also from Estrelda), which is
almost rhomboidal in shape (fig. 52). According to Laveran’s
figure, T. avium (fig. 16 F) occupies an intermediate position.
Novy and McNeal find, however, considerable polymorphism
in this species, and include herein some of the above forms.’

No one doubts that the anterior extremity of the body in
the biflagellate or Heteromastigine forms (Trypanoplasma
and Trypanophis) is that from which spring the two
flagella. In these the anterior end may be acutely or obtusely
conical (fig. 17 Fr) or bluntly rounded (figs. 17 cg, 18 a and c).
At the tip is often to be noted a little active (“ metabolic”),
sensitive beak or rostrum; although this points to one side
in Trypanophis, it is probably morphologically terminal.
Equally certain is it that in the uniflagellate form Try pano-
morpha (Trypanosoma) noctuz, whose life-cycle has
been described above, the end of the body bearing the free
portion of the flagellum is the anterior one. But with regard
to the correct orientation in the rest of the uniflagellate
Trypanosomes (collectively included in the genus Try pano-
soma), the greatest confusion exists. At present,’ for the
sake of clearness, the terms flagellate and non-flagellate end
will be used.

In Trypanosoma the non-flagellate extremity presents
considerable variation, often in different individuals of the

1 It is uncertain how far the conclusions of these authors are warranted,
since their “determinations’’ appear to have been based largely on the
different forms of the parasites observed in cultures. See, however, below,
in Systematic Section.

The whole question is fully discussed in Section 11.

THE HM@MOFLAGELLATES. 207

same species; as Laveran and Mesnil point out, it is par-
ticularly plastic. Hence, in the following examples, it must
not be inferred that the parasites always conform to that
particular description, but merely that such and such a mode
of termination is usually to be noted in their case. The non-
flagellate end may be much drawn out and pointed, as in
T. lewisi, T. ziemanni, T. avium, and T. inopinatum;
or shorter, and sharply acute, as in T. gambiense, T. granu-
losum, T. equinum (fig. 16D). In other instances it may
be obtuse or even rounded off at the tip, as in individuals of
T. brucii (fig. 42 B), T. equiperdum (fig. 42 c), and, some-
times to a marked degree, in IT. dimorphon (fig. 49 1);
although these are all relatively slender forms. Nor does it
always follow that the non-flagellate end is blunt in the thick,
fusiform parasites; it tapers very finely in T. padde (fig.
54), and in Hanna’s Trypanosome it is extremely long and
attenuated (fig. 16a). Lastly, figs. 17 a and B, 56 show the
different modifications in this respect in T. rotatorium. In
one or two instances, the extreme tip evinces acertain degree
of contractility,! as seen in fig. 42p of T. equiperdum,
where it is retracted, so that the extremity appears forked or
bifid. Dutton and Todd seem to have observed a similar
appearance in the case of IT’. rotatorium.

The flagellate end of the body is more uniform and nearly
always tapering; it sometimes thins away so gradually (e.¢.
fio. 16 F) that it is difficult to be quite certain of the exact
point where it passes into the flagellum. In one or two cases,
however, it ends rather abruptly (e. g. T. sole, fig. 173).

There are two flagella only in Trypanoplasma and
Trypanophis. ‘They are inserted into the body very close
to the anterior end, just to one side of the rostrum. The two
flagella are quite separate from each other, and, while one
(that most anteriorly situated) is entirely free and directed
forwards, the other at once bends backwards and is attached
to the convex (dorsal) side of the body, throughout the
greater part or all the length of the latter. Posteriorly this

1 Compare the active, sensitive, beak of Try panoplasma borreli.

208 H. M. WOODCOCK.

flagellum becomes free, and is directed backwards like a tail.
A comparison of the degree of development of the two
flagella in different cases is very instructive in a phylogenetic
connection.

In Trypanosoma there is only one flagellum, which is
invariably attached to the body in the manner characteristic
of the posterior flagellum of the biflagellate forms. The
point of insertion of the flagellum into the body is generally
near the non-flagellate end, but may vary considerably ; it is
in close relation with the position of the kinetonucleus, which
is discussed below (p. 216). Although there is generally a
free continuation of the flagellum, this may be short or
absent. It is very short in T. paddz and indifferent indi-
viduals of T. ziemanni, while in T. johnstonii (fig. 51)
and female forms of T. ziemanni (fig. 33.4) the flagellum
ceases with the limit of the body. Laveran and Mesnil
(54) maintain that the same occurs in the case of T. dimor-
phon (fig. 491), but Dutton and Todd, who first described
this parasite (21), figure a distinct, free flagellum, sometimes
short, sometimes long (fig. 4911). Further investigation is
necessary to decide this point.

Along the dorsal side runs a characteristic fin-like expan-
sion of the body, the undulating membrane. This always
begins proximally at the place where the attached flagellum
emerges from the body; hence, its distance from the non-
flagellate end is dependent upon the point of insertion of the
flagellum. ‘The free edge of the membrane is more or less
sinuous in outline, which gives the structure, even when at
rest, a wavy appearance. ‘The edge itself is really formed by
the attached flagellum. Distally the membrane thins away
concurrently with the body, and when it ceases the flagellum
becomes, with one or two exceptions, free. Probably, in fact,
in forms with a very tapering flagellar end, the last portion
of the body is constituted mainly or entirely by the undu-
lating membrane, which has been drawn out in some such
manner as in the case of the anterior end of Trypano-
morpha. The membrane may be only slightly developed,

THE HH MOFLAGELLATES., 209

narrow, and chiefly discernible by its thickened flagellar
border, as in IT. inopinatum (fig. 17c), sometimes in T.
lewisi, and particularly in certain individuals of T’. zie-
manni (fig. 32) and T. johnstonii (fig. 51); in these cases
it is comparatively straight, and of about equal narrowness
throughout. In most Mammalian Trypanosomes it is mode-
rately developed, and usually more or less curved and wavy.
Lastly, in T. rotatorium (fig. 17 a and 8), T. avium (fig.
16¥),! T. damonie (fig. 163), Trypanoplasma and most
Piscine Trypanosomes it is very well developed, and often
thrown into broad folds or pleats of varying number.

Minute Structure.

The body appears to be in all cases naked, without
distinct limiting membrane or cuticle. This is probably an
adaptation to the peculiar habitat, and would undoubtedly
facilitate nutrition, which, of course, takes place here solely
by osmosis. ‘I'he occurrence of any differentiation of the peri-
pheral cytoplasm in the form of an ectoplasmic layer has only
seldom been noted. Most writers simply ignore the point ;
Laveran and Mesnil (56) say that they have not succeeded
in differentiating any ectoplasmic layer, either in T’. lewisi
or in the other Trypanosomes they have investigated.
Wasielewsky and Senn (120),? however, observed such a layer
in the case of 'T. lewisi, distinguished from the rest of the

! According to Laveran and Mesnil there is a slight peculiarity in the
undulating membrane of T. avium, a well-marked line running in the
middle, parallel with its contour (fig. 16 F, l.s.); it does not appear to be
continuous, being interrupted at the narrow parts of the membrane. These
authors consider that it represents a fold. It rather recalls the strength-
ening filaments or ribs in the membrane of Trichomonas, and perhaps
serves a similar function.

2 These authors use the botanical term “ periplast”’ to denote this layer.
Not to enter here into a discussion of the various forms which Senn’s
“neriplast’’ assumes, it may be merely stated that, at any rate so far as
the Trypanosomes are concerned, the zoological designation of ectoplasm
undoubtedly best indicates its nature and character.

210 H. M. WOODCOCK.

cytoplasm by taking a slightly different tint; and Prowazek,
in his recent work on this parasite, remarks to the same
effect. In Trypanophis, Keysselitz (28) finds a prominent,
highly-refractive, and finely-alveolar ectoplasmic layer,
especially well developed near the anterior end, where it forms
a little cap (fig. 4], e.c.). Unfortunately, even now very few
direct references to the subject are available.

It appears most likely that the undulating membrane is
largely, if not entirely, an ectoplasmic development. ‘This is
usually clearer, more hyaline, and less granular than the rest
of the cytoplasm, and in these characters it agrees with a
typical ectoplasm. The fact, too, that it is so closely asso-
ciated with locomotion supports this view.! Similarly, with
regard to an ectoplasmic differentiation surrounding the body
generally, the occurrence of distinct, superficial, longitudinal
strie, probably comparable with myonemes (see below, p. 219),
points to the existence of such a layer, since myoneme
fibrillee, when they occur, are always situate in the ectoplasm.
Dutton and Todd (l.c.) found that, in injured individuals of
T. mega, a delicate membranous (?) envelope, continuous
with the undulating membrane, could be easily separated
from the rest of the body. In it were fine, pink-staining
lines, having a looped arrangement. There can be little
doubt this structure was an ectoplasmic sheath with myoneme
striations. In short, it seems probable that, in the majority
of Trypanosomes, there is such a layer around the body,
although, apparently, often poorly developed and ill-differ-
entiated when compared with the undulating membrane.

The general cytoplasm may be of a clear, finely granular,
or alveolar nature, presenting a fairly homogeneous appear-
ance, as in 'T’, lewisi, T. equiperdum, and T. gambiense
(some individuals), although even in these cases it rarely
stains up quite uniformly. It may be coarser and relatively
dense, as in T. avium, T. raje, T. scyllii (fig. 62), and
others. The cytoplasm of male forms is, in general, much
clearer and less granular than that of femaleforms. Accord-

Cf. also its formation in Try panomorpha (p. 186).

THE HHMOFLAGELLATES. Poa

ing to Dutton and Todd the cytoplasm in T. mega and
(though to a less extent) in T. karyozeukton shows marked
differences in different regions of the body. In the third of
the body in front of the nucleus (fig. 58) it is very spongy,
and appears loose and alveolar in character; behind the
nucleus it is arranged in alternating light and dense, deep-
staining bands (‘“hyaloplasm” and “spongioplasm”), run-
ning more or less longitudinally. In Trypanophis there
are one or two rows of highly-refractive, yellowish inclusions
running the length of the body (figs. 40 and 41). The larger
ones lie in the row nearer the convex side, close to the undu-
lating membrane, and these may be oblong in shape, arranged
at right angles to the length of the body. ‘These grains are
probably not comparable with those next described. Keysselitz
thinks they represent collections of fatty and oily substances.

In many forms deep-staining grains or granules, of a
chromatoid nature, and of varying size are to be observed in
the cytoplasm.! These are few and minute in Tl, danilewsky1
(fig. 60 a) and T. tince (60 B), somewhat more numerous in
T. equinum (16 p), and T. theileri (fig. 50), and relatively
large and numerous in T. brucii, and certain individuals of
Trypanoplasma borreli (fig. 17G@). In most instances the
granules are, if not confined to, chiefly distributed in the
flagellate half of the body (in the case of Trypanoplasma,
the posterior half). In T. nelspruitense (178) and T.
granulosum (17 xk) the grains are large and particularly
numerous, and, in the latter parasite, spread forwards almost
to the non-flagellate end.

In certain Trypanosomes a vacuole is often, though by no
means constantly, to be observed, situated at a varying dis-
tance from the non-flagellate end. This vacuole is well
defined, usually of oval shape, and sometimes very prominent,
especially in certain Mammalian forms, e.g. T. brucii (fig.
44), T. gambiense (figs. 16 c, 48), and T. evansi (fig.
45). Prowazek (1.c.) also describes, for the first time, the
occurrence of one in I’. lewisi. Laveran and Mesnil are

1 For their probable origin and nature see below, p. 229.

212 H. M. WOODCOCK.

not inclined to regard this structure as a normal constituent
of the cell,—comparable, for instance, with an excretory
vacuole. They only describe its occurrence very rarely (fig.
42 c); their preparations showing it are of ‘Trypanosomes
which were not in the blood, but in another medium, e.g.
serous, or cerebro-spinal fluid, and they consider that it
is an artifact caused by the imperfect fixation of such fluid,
in which, of course, the parasite is bathed. There are,
however, several considerations to be set against this view.
All the above figures are from blood preparations, and in the
Reports of the Sleeping Sickness Commission there are some
realistic figures of T. gambiense from the blood, in most of
which this vacuole is well marked.t Moreover, Dutton and
Todd’s figures (l.c.) of dividing stages of T. dimorphon
show clearly that an oval vacuole is present and also divides
(fig. 49 1). Again, it is important to note that a similar
structure has been described in the case of parasites in their
natural (tolerant) hosts, and in what there can be no question
were absolutely normal conditions. Thus Léger (l.c.) de-
scribes a clear, oval space near the anterior end of ''ry pano-
plasma, which, he says, probably represents a true vacuole,
and is not to be confused with certain rounded vacuoles that
sometimes appear in unhealthy individuals?; Hanna figures
a small vacuole in his Trypanosome from Indian pigeons
(fig. 16 g) ; and lastly, Schaudinn describes one or two large
ones in the indifferent forms of Try panomorpha noctue,
near the apparently opposite, flagellate end be it
noted. Hence it appears almost certain that a vacuole,
probably excretory in function, may occur normally in many
Try panosomes.®

1 The writer would add that he has seen some of these, and other
preparations of Mammalian Trypanosomes, in which the parasites appeared
perfectly normal and exhibited this vacuole.

2 Progressive vacuolisation on a large scale is often met with in atypical
or abnormal conditions of the parasites (see below, p. 229).

3 It may possibly be that the technique used by Laveran and Mesnil is
not suited for demonstrating this particular point (see below, p, 217).

THE HAHEMOFLAGELLATES. 213

There can be little doubt that much has yet to be ascer-
tained concerning the details of nuclear structure in most
forms. We have above described the complexity of the
nuclear apparatus in T'ry panomorpha noctue, its division
into two distinct parts, trophic and kinetic, and the inti-
mate connection of the latter with the locomotor apparatus.
It must be remembered, moreover, that T'rypanomorpha
is not the only Trypanosome in which this highly complex
condition exists. Schaudinn finds a complete parallel in
Trypanosoma ziemanni (the “Spirocheta”-form) ;
in this parasite the nuclear apparatus is even more com-
plicated, owing to the fact that the number of ‘ chromo-
somes”’ is sixteen, as compared with eight in the first-named
form. Again, according to Prowazek’s recent investigations,
T. lewisi and T. brucii also possess the same funda-
mental type of nuclear structure. Indeed, the system of
axial spindles produced by successive divisions of the karyo-
centrosome is even more elaborate in the former parasite
than in Trypanomorpha. In both forms, the number of
trophonuclear chromosomes was clearly seen to be eight;
those of the kinetonucleus were more difficult to make out,
owing to the compacter form of the latter body. The only
other case in which a definite number of chromosomes has
been made out is the trophonucleus of Trypanoplasma
borreli. Here Léger describes eight dumbbell-shaped chro-
mosomes? radially arranged around a central grain (karyo-
centrosome). ‘These instances would seem to suggest that,
with finer and more detailed investigation, the nuclear
apparatus of Trypanosomes will be found to show a greater
uniformity of complex organisation than is so far known.
Our present knowledge of the nuclear elements in the
majority of Trypanosomes relates chiefly, however, to their
position and general appearance, and is soon summarised.

The trophonucleus (nucleus) occupies a position usually
about the middle of the body (figs. 16 B—p, rp, 17 5 and x) ;
it may, however, be either in the flagellate half, as in 1.

1 They may be divided, forming 16 small chromatic masses.

vou. 50, PART 1,—NEW SERIES, 15

214 H. M. WOODCOCK.

lewisi (164), T. remaki(171), and T. raje (628), or in
the non-flagellate half, as in T. damoniz (163) and in
T. nelspruitense (178); in Trypanoplasma borreli it
is often comparatively far forwards (fig. 18). In some cases,
at all events, the position is by no means constant (cf. T.
equinum [figs. 16 p, 47] and T. evansi [figs. 42 a, 45]).
The trophonucleus presents no striking variations in size,
which, indeed, appears to be often independent of the size
and shape of the parasite. Thus in T. rotatorium it is
scarcely larger than that of many slender Mammalian or
Piscine forms, in which it occupies almost the entire breadth
of the body. The trophonucleus attains, perhaps, its greatest
size in the large T. theileri and 'T. rajz. In form it is
generally ovoid, the longer axis being directed longitudinally,
but in the Trypanosome described by Dutton and Todd from
Senegambian birds, and, similarly, in Hanna’s Trypanosome,
the long axis is transverse to that of the body (figs. 52, 53).
The shape of the trophonucleus in the latter instance is also
unusual, resembling an isosceles triangle. The minute struc-
ture is generally described as consisting of a more or less
compact aggregation of chromatin grains embedded in a
plastin-like base or matrix; these may be uniformly dis-
tributed throughout the nucleus, or more closely packed in
the peripheral part, leaving a clearer central area (figs. 19,
44, 56). No mention is usually made of a nuclear membrane
or distinct reticulum.’ In T. remaki, possibly also in TT.
solew, and in Trypanophis, there is a large, deeply-
staining granule in the centre of the nucleus, surrounded
by a clear area, which probably represents the tropho-
nuclear centrosome (karyocentrosome); in Trypanophis
this granule appears to divide by simple mitosis (fig.
41).

Of the highest importance is Schaudinn’s revelation of the
true nature of that hitherto enigmatical and much-discussed
chromatic body, which is situated near the root of the

1 Prowazek describes a nuclear reticulum in both T. lewisi and T.
brucii.

THE HAMOFLAGELLATES. 215

flagellum.! In view of its essential nuclear character and the
fact that it serves as the directive centre for the locomotor
activities of the cell, and since, moreover, none of the other

G:

Fie. 18.—Trypanoplasma borreli, Laveran and Mesnil.
a.f. = anterior flagellum; p,f. = posterior do.; m. = undulating
membrane; 17’. = trophonucleus ; K. = kinetonucleus; f. = fibril
(myoneme) ; ¢ = centrosomic granule at base of flagellum. (After
Léger.)

names given correctly indicates its real nature, the term
kinetonucleus has been adopted in this article to designate

1 According to Wasielewsky and Senn the “ Geisselwurzel” (to call the
body for the moment by a non-committal name) is a blepharoplast, i.e. a
purely superficial thickening or ectoplasmic differentiation (in botanical
language, a periplastic or kinoplasmic development), and bears in its origin
no relation to the nucleus. This view is now completely out of court.
Rabinowitsch and Kempner (89) considered it as a nucleolus, notwith-
standing the fact that it is nearly always extranuclear. The other two well-
known theories, namely, the centrosomic view of Laveran and Mesnil and the
micronuclear one of Bradford and Plimmer, have each a certain modicum
of truth. But, as Schaudinn points out, the kinetonucleus is much more
than a centrosome—possesses, in fact, a centrosome of its own—and, on the
other hand, it has not much in common with the “micronucleus” of
Infusoria, beyond the fact that it is of nuclear origin. A micronucleus is
essentially a sexual nucleus, its role being played when that of the mega-
or somatic nucleus is finished; whereas here the kinetonucleus has primarily
a kinetic function, the other, the trophonucleus, being just as important
sexually.

216 H. M. WOODCOCK.

this structure. For it can scarcely be doubted that this
characteristic organella of Trypanosomes is homologous in
all cases, and agrees, in origin and significance, with the kineto-
nucleus of Trypanomorpha above described. Although,
unfortunately, very little is known in most cases about the
minute structure of this body, owing to its propensity for
staining up deeply, the evidence afforded by the work of
Schaudinn and Prowazek is, we think, sufficient to justify
this view.

One or two additional confirmatory points may be men-
tioned. In the first place there is the well-developed nature
of the kinetonucleus in Trypanoplasma borreli, where
it may be almost as large as the trophonucleus, and the same
remark applies to the other species of this genus.! Again, in
the ookinetes of T. barbatule (see below, Section IX), Léger
has observed heteropolic division of the single large nucleus,
doubtless leading to the formation of tropho- and kineto-
nucleus. Lastly, Bradford and Plimmer (6) have themselves
observed the latter body given off from the larger, trophic
nucleus (the ‘‘ macronucleus ”’ of these authors) in Try pano-
soma brucil.

The kinetonucleus in an adult Trypanosome is always
situated normally in the non-flagellated half of the body.
Its distance from the end varies considerably, even in differ-
ent individuals of the same species. The more slender and
tapering the extremity, the farther away from it, usually, is
the kinetonucleus. In certain forms, e. g. IT. mega (fig. 58),
T. transvaaliense (fig. 50c), and some individuals of T.
rotatorium (178, 56 4) and of T. inopinatum (17 c), it lies
more centrally, and is often contiguous to the trophonucleus.
The kinetonucleus is usually rounded in shape, but may
be sometimes elongated or rod-like, as in T. inopinatum,
T. johnstoni (fig. 51), T. transvaaliense, and Try-
panoplasma borreli (fig. 18), with its axis either transverse
or longitudinal. It attains its largest size in certain Piscine
forms, e.g. ‘I’. solee; in these cases it frequently occupies

1 To judge, that is, from Plehn’s rather unsatisfactory figures (84).

THE H#EMOFLAGELLATES. Paley

the entire width of the body at that point (fig. 17,5 and x).
‘here is sometimes a clear zone or halo surrounding this
organella, as in figs. 17 a, B, 28c. Possibly this clear region
represents the cytoplasmic vacuole of other forms, which
usually lies close to the kinetonucleus (cf. figs. 52, 53).
In T. equinum the kinetonucleus is extremely small and
difficult to make out; according to Ligniéres (71) and
Laveran and Mesnil it is a dot-like thickening at the root-
termination of the flagellum (fig. 16 p)'; in this form it has
apparently become secondarily reduced.

In intimate relation with the kinetonucleus is the flagellum.
That portion lying in the general cytoplasm of the body is
distinguished as the rhizoplast. When the kinetonucleus
is situated near the non-flagellate end and, consequently,
nowhere far from the surface, the rhizoplast is very short,
and the flagellum leaves the cytoplasm almost at once,
becoming, of course, the flagellar border of the undulating
membrane. When the kinetonucleus is deeply situated, the
rhizoplast is somewhat longer, and (in Trypanosoma)
usually runs in an oblique direction to the point where it
emerges (figs. 17 a, 58). In Trypanoplasma, the rhizo-
plastic portion of each flagellum is well-developed ; the two
are quite distinct, and lie parallel to each other (fig. 18).

Apart from Schaudinn’s observations, only in very few cases
up till now has the occurrence been recorded of a distinct
centrosomic granule, from which the flagellum originates.
Léger describes a well-marked granule (“diplosome”’) at the
base of each rhizoplast in Try panoplasma (fig. 18 B, c); and
Keysselitz suspects the presence of similar ones in Try pano-

1 It is most easily demonstrated by using Laveran’s special stain. In
this connection, and a propos of the genuineness of the vacuole mentioned
above, it is interesting to note that Elmassina and Migone (22) describe
(in preparations made according to their own method) a clear, refringent
sphere, situated very near the point where the flagellum emerges from the
body. This structure is not rendered visible, these authors find, by Laveran’s
stain, although this alone reveals the kinetonucleus. From Elmassina and

Migone’s description and figures it seems most likely that this refringent
sphere is a cytoplasmic vacuole.

218 H. M. WOODCOCK.

phis, although owing to the intensity with which the kineto-
nucleus stains he is unable to be certain. Prowazek frequently
figures such a centrosomic granule, situated between the
kinetonucleus and the base of the flagellum, in both T. lewisi
and T. brucii. Sometimes this granule is closely attached
to the kinetonucleus, and appears separated by a short gap
from the flagellum; at other times it is attached to the
flagellum and separated from the kinetonucleus. As the
author points out, it is probably homologous with that at the
base of the flagellum in other cases, and, indeed, as already
stated, Prowazek finds the minute structure and development
of this region entirely comparable to that described by
Schaudinn in Trypanomorpha and ‘Trypanosoma
ziemanni.!

Bearing this in mind, it seems most likely that there is, as
a rule, actual organic union between the flagellum and the
kinetonucleus, even where there appears to be a gap between
the two. Probably the delicate connecting-thread (or axial
spindle of an earlier division, see above, p. 185) is not easily
demonstrated.? In the majority of Trypanosomes the flagellum
is described as joined to, and originating from, the kineto-
nucleus (figs. 16, 17). Even in those cases where there is a
clear zone around the latter, and the flagellum seems to begin
at the outer border of the halo, Laveran and Mesnil consider
that there is, nevertheless, an unbroken union between the
two organelle; for, in involution forms of the particular
Trypanosome concerned, when most of the organism has
perished, kinetonucleus and flagellum still persist, intimately
united.

1 An interesting point bearing on the view that the flagellum represents
the greatly elongated axial spindle of a nuclear division may be noted. In
Trypanosoma johnstoni, where there is no free portion of the flagellum,
this terminates (at the limit of the cytoplasm) in a small deep-staining
granule, which is perhaps comparable to the distal centrosome of the axial
spindle (compare also pp. 197-8).

2 In one or two maceration preparations of T. brucii, Prowazek was able
to see this connecting fibril.

THE HE MOFLAGELLATES. 219

The occurrence of prominent myonemes in the undulating
membrane of T'rypanomorpha, and their nuclear origin (as
“ mantle-fibrils ”’) has been above described. Schaudinn finds
a like development in Trypanosoma ziemanni, the fibrille
being particularly noticeable in the male and female forms
(fig. 83). In this parasite, however, they are not restricted
to the undulating membrane, but are arranged laterally, half
running down each side of the body, in the ectoplasm. ‘Their
total number is sixteen, corresponding to the number of
chromosomes. According to Prowazek an exactly similar
state of affairs exists in both T. lewisi and T. brucii,
although the myonemes (of which there are here eight) are
very delicate and difficult to make out. In two or three
other forms longitudinal striations, comparable to muscle-
fibrille, are well-marked ; nothing is known, however, with
regard to their origin. Thus in Trypanoplasma borreli
there are two, which start in front, sweep round, one on each
side of the body, and run backwards more than half-way,
finally joining ventrally (fig. 18 c, /f.). The myonemes are
very prominent in the ribbed! forms of T. rotatorium (fig.
17 8), in which the surface of the body is thrown into longi-
tudinal folds or ridges, often having a somewhat spiral course
(56 A); the striations appear to lie in the furrows between the
ridges. In 'l’. solez, again, the longitudinal striz are dis-
tinctly discernible (fig. 173); and Novy and McNeal (81)
record the occurrence of six or eight in T. avium.

Section VII. BrotogicaL ConsipERATIONS ; MovEMENT ;
AGGLOMERATION ; ABNORMAL AND INVoLUTION Forms.

(A) Movement.

In general Trypanosomes are extremely active, as would
naturally be expected from their powerful locomotor organs.

1 Other individuals, distinguished as smooth forms (figs. 17 a, 56 B), do
not show them.

220 H. M. WOODCOCK.

According to the manner in which they are produced, two
kinds of movement, broadly speaking, can be distinguished :
(1) displacement of the body, usually rapid, and (2) movements
of flexion, extension, and contraction, often comparable to
“euglenoid” movements. ‘The latter are brought about, in
all probability, by the superficial myonemes above described
(cf. the muscle-fibrillee of Gregarines).

In Trypanoplasma the anterior end always moves first
in displacement. According to Léger (I. c.) the principal
organ concerned is the undulating membrane, which, by its
rapid vibrations from side to side, is thrown into a series of
curves or folds; the effect produced is that of rapidly
succeeding waves, starting in front and running backwards."
The oscillations may be continued into the posterior flagellum,
which is then a subsidiary organ of locomotion, of the type
known as a “ pulsellum,’’—i. e. it acts in a driving sense, like
the tail of a spermatozoon. Léger thinks, however, that this
flagellum functions principally as a rudder. The anterior
flagellum is not greatly, if at all, concerned in locomotion.
During rapid displacement it is directed backwards, probably
passively carried along by the movement of the animal, only,
of itself, making slight wavy or circumrotatory movements.
At other times it seems to function rather as a sensitive
organ, being repeatedly thrust out, as it were, tentatively.

1 It may be noted that Keysselitz (1. c.) takes a somewhat different view
of the manner in which displacement occurs in Try panophis. Keysselitz
thinks that change of place is brought about by rapid, vibratile, or serpen-
tine movements of the whole body substance, the membrane acting, on the
contrary, as a curb or drag, which, of itself, would drive the animal in the
opposite direction. The author bases his opinion on the observations of the
parasites in the attached phase, when the typical swimming motions are
often continued, and contends that were it not for this antagonistic working
of the membrane the parasites would be driven against their base and
swollen out of all shape. It may very well be that the membrane works in
an opposite sense at such a time (cf. next page); but it is rather unlikely
that the delicate body-substance can produce the vigorous movements shown
by the actively swimming parasites, when its only muscular elements are
perhaps myoneme fibrils (cf. the slow progressive movement of Gregarines).

THE HE MOFLAGELLATES. Dall

The thick, stumpy parasites only accomplish jerky move-
ments of flexion, which scarcely serve to displace the animal.

The manner of locomotion in Trypanosoma differs in one
or two points from that in Trypanoplasma. In Try-
panosoma the flagellar extremity generally leads the way in
movements of displacement, though the parasites can, and
sometimes do, move with the non-flagellate end directed
forward. The movements may be very rapid and relatively
considerable, as in IT’. lewisi, for example, which quickly
darts across the field of the microscope and is lost to sight.
T. evansi, again, also easily traverses the field, although it
is somewhat slower; on the other hand, T. brucii scarcely
ever leaves the field of view, its powers of active displace-
ment being either insignificant or else very little used.
There is some difference of opinion as to whether the un-
dulating membrane or the flagellum plays the principal part
in these movements. Probably the former does, though the
flagellum doubtless acts to a certain extent as a “ tractellum,”
especially in cases of very rapid movement. All Trypano-
somes undergo, more or less continually, a vibratile or un-
dulatory motion, caused by the membrane. ‘This may be in
either direction, i.e. commencing anteriorly or posteriorly.
Movements of contortion are much in evidence in Piscine
forms, which, as above mentioned, are frequently coiled up
on themselves. In many Trypanosomes, especially the more
slender or spirochetiform ones, the undulating membrane
often appears spirally wound round the body, this being
really due to a more or less pronounced torsion of the latter,
which gives the animals a corkscrew-like movement. Hugle-
noid or semi-amceboid movements are common in the more
sluggish parasites, constituting in many individuals of T.
rotatorium, for instance, practically the only kind that
there is to be observed.

Before leaving the question of movement it is essential to
note that slow displacement of the body, occurring in quite a
different way, has been observed in certain cases. Thus
Léger (66) describes a creeping or crawling movement in T.

Dae H. M. WOODCOCK.

barbatule, which would seem to be quite comparable to a
“ ovregarinoid”? movement. Again, Gray and Tulloch (24a),
_in their description of T. gambiense in the fly (Glossina
palpalis), say that the parasites also progress in a zig-zag
manner, advancing by a series of contractions, which bend
first one side of the body and then the other (cf. the flexion
movements of sporozoites). In all such cases described, not
the flagellate end, but the non-flagellate end, goes first.’

(s) Agglomeration.

Before considering the process itself, a few words are necessary with
regard to its occurrence and causation. This characteristic phenomenon of
Trypanosomes appears to take place chiefly or only upon the advent of
unfavourable biological conditions in the surrounding medium. As said
above, increasing scarcity of nutriment brings about its occurrence in the
case of Trypanomorpha, when in the Insectan host. In the normal,
unaltered blood, or other humour, of Vertebrate hosts, agglomeration has
only seldom been observed. A tendency to it has been noticed in T.
lewisi, and Prowazek has found it to occur in T. brucii, in the inner
organs of guinea-pigs and rats. In these cases the phenomenon does not
appear to be very persistent. Schaudinn also describes its occurrence in
T. ziemanni in the owl. This variety is often termed self- or auto-
agglomeration.

On the other hand, when blood containing some of the parasites is
drawn off, defibrinated, and kept for some time at a low temperature?
(in the refrigerator), agglomeration usually sets in partially, and con-
tinues more or less persistently, until the death of the parasites super-
venes. It is readily and, one may perhaps say, typically produced, when
the serum of an animal which has been two or three times inoculated with
its specific Trypanosome (e. g. T. lewisi, in the case of the rat), and which
is thus becoming repellent to, or acquiring immunity against, that par-

1 For the importance of this fact see below, in Section XI.

2 It may be here mentioned that Trypanosomes appear to be much more
resistant to a lowering than to an increase of temperature. Their vitality,
in cold solutions, has a significant bearing upon the question of a cold-
blooded, alternate host.

3 Without entering here into the question of how immunity is acquired,
it may be mentioned that this does not seem to stand in any relation with
the property of agglomeration. For one thing, the action of heat upon any
serum differs considerably as regards the destruction of its agglomerative

THE HASMOFLAGELLATES. 223

ticular form, is added to blood containing it; in such a case agglomeration
is very rapid and frequently total—i.e. embracing the entire number of
individuals present. Laveran and Mesnil consider that this peculiar occur-
rence is intimately connected with the development, in the blood of the
host which is undergoing the immunisation, of a substance which has this
specific property towards that Trypanosome; in other words, of a specific
agglomerine, perfectly analogous to the agglutinines that cause the
agglutination of Bacteria. Probably this specific agglomerine is present,
or can be produced, in a rat infected for the first time, but only to a very
slight degree ; with successive reinfections its power is greatly increased.
Moreover, sera other than that of the particular kind of host, for the time

Fig. 19.

Fic. 19.—Binary union or agglomeration of T. brucii. (After
Bradford and Plimmer.)

Fic. 20.—a, ditto of T. lewisi; B, agglomeration cluster or
primary rosette of same parasite. (After L. and M.)

being, of a given parasite, may also possess, in a varying degree, this
agglomerative property towards that Trypanosome ; in such cases, of course,
the agglomerine concerned is not specific. Lastly, agglomeration has also
been produced by the addition of chemical solutions, and in artificial
cultures of the parasites. For fuller details and examples of the compara-

and immunising powers. Again, certain sera which exhibit the former
property towards a given Trypanosome may not be able to prevent infection
by that form. Laveran and Mesnil consider that the ‘“ preventivity” of a
serum is principally dependent upon its enhanced phagocytic activity.

224. H. M. WOODCOCK.

tive behaviour of different sera, the reader is referred to the works of
Laveran and Mesnil (40, 48 and 56), Ligniéres (71), Brumpt and Wurtz
(12), Thiroux (114, 115), and others.

Agglomeration commences by two Trypanosomes coming together and
joining (figs. 19,204). In all cases in which the phenomenon has yet been
witnessed in a natural (as opposed to an artificial’) medium, a particular
form of Trypanosome always unites by the same end. In Trypano-
morpha, as already described, the parasites join by the anterior,
flagellate end. Agglomeration has not, up till now, been observed in
Trypanoplasma, but there can be little doubt that, if it occurs, it will be
found to take place there also by the anterior end. . On the other hand, in
all the species of Trypanosoma (including T. ziemanni) for which the
occurrence has been so far described, the parasites unite by the non-
flagellate end. The union may sometimes remain only binary. In
other cases agglomeration is rapid and progressive, the union of two para-
sites being quickly followed by that of many others around the same
centre, the whole forming a “multiple union” or rosette (fig. 20 8B).
Such a rosette is termed a primary agglomeration, and may be composed of
as many as one hundred individuals. In many cases, especially where the
agglomerine is specific and very powerful, the rosettes themselves become
grouped together to form large tangled masses known as secondary agglome-
rations.

In “ cultures,’

el

rosettes or clusters are frequently observed in which the
arrangement of the parasites is different, and may vary even in the same
species; that is to say, in some cases the Trypanosomes have their flagella
at the periphery, while in others they are all attached by the flagella, which
are directed towards the centre (fig. 29). This has been considered as
indicating that the end by which agglomeration takes place cannot be

1 Much attention has lately been paid—particularly by Novy and McNeal
(79—81), Smedley (107), and Thiroux (1.c.)—to the cultivation of different
Trypanosomes in artificial media, in‘ the same way in which cultures of
Bacteria are obtained. The writer does not propose to give the details
of the composition of the various media tried, nor to discuss the technique
and the great (but quite natural!) difficulty experienced in persuading the
parasites to live and thrive [?]. For it cannot be too strongly insisted that
this is not a zoological method of research, and that the results obtained do
not add to our knowledge of the parasite’s real life-history and biology, but
must be accepted as normal phases only with the greatest caution. It is
not merely a question of obtaining a “ pure culture,’ as Novy and McNeal
consider: the Trypanosomes are not Bacteria. As will be seen later, some
of the opinions to which the authors are led, as a result of practically
limiting themselves to this method of investigation, are—to say the least—
not generally accepted.

THE HAMMOFLAGELLATES. 225

regarded as of importance in determining the orientation of the body. It
appears, however, that two entirely different processes are concerned. In
most cases, if not in all, the clusters which have the flagella pointing
centrally are not instances of agglomeration, but of rapid division (see
below, Multiplication), where the parasites remain in contact and form large
colonies (exs.: T. lewisi, L.and M.,Smedley, MeN., and others, T. avium
and other forms, N. and MeN., T. duttoni, Thiroux). On the other
hand, even in cultures, true agglomeration clusters, formed by the union of
independent parasites, are attached by the non-flagellate end, as in the
blood (exs.: T. brucii and T. lewisi, McN., Smedley and others, T
avium, N. and MecN., T. padde, Thiroux).

These agglomerations differ strikingly from an agglutination of Bacteria,
in that the Trypanosomes do not, in the slightest degree, lose their mobility.
Each individual continues active movements, its flagellum lashing away at
the periphery, and appears to be making strenuous endeavours to escape.

Another distinctive feature of the phenomenon is that of disagglome-
ration. The individuals constituting agglutinated Bacterial clusters are
never known to detach themselves, with the resultant dissolution of the
mass; and it is for this reason that Laveran and Mesnil use the distinct
terms here adopted. Disagglomeration is in consequence of the retention
of the power of movement by the parasites during the progress of agglome-
ration; thus the Trypanosomes are able to disengage themselves from the
cluster, and so to cause the complete break-up of the rosette. Sometimes
all the individuals, apparently quite unaltered morphologically, become thus
dispersed. At other times the break-up is only partial, a certain number of
the more feeble and less mobile parasites remaining together and slowly
dying off. Even the larger secondary masses may be thus dissolved. The
ability of the Trypanosomes to disagglomerate themselves stands in inverse
relation to the strength of the agglomerating serum; where the agglome-
rine is powerful the parasites appear unable to liberate themselves.

Not only normal and actively-living Trypanosomes undergo this process,
but also parasites which have been stupefied, paralysed, or even killed by
chemical reagents or strong doses of a serum become united together. In
such cases the Trypanosomes are quite irregularly and indiscriminately
arranged, forming more or less compact masses of varying shape. Hence
Laveran and Mesnil argue that the rosette-formation in typical agglomera-
tion is determined solely by the fact that the parasites possess unimpaired
mobility, and are actively striving to free themselves, the resulting figure
being that of equilibrium.

Agglomeration does not of itself seem to have any ill effects upon the
parasites. Unless disagglomeration occurs the rosettes and masses persist
unaltered for some time, the agglomerated individuals retaining their
vitality just as well as free individuals in the same surroundings, The
only exception is seen in the case of a central rosette, which has served, as

226 H. M. WOODCOCK.

it were, as the nucleus of a large secondary agglomeration ; here, if dispersion
does not soon take place, the individuals comprising it rapidly degenerate
and die, owing to their confined and unfavourable situation.

The significance of the process has yet to be ascertained. By some it is
considered as a purely involuntary proceeding on the part of the parasites,
and brought about mechanically, by the operation of external influences.?
The clusters of paralysed and dead Trypanosomes which may be formed are
adduced in support of this view. Prowazek’s explanation (I. c.) is that
nuclear substances resulting from the partial break-up of the kinetonucleus
are passed out, causing the surface of the body near that end to become
sticky and viscous; and this brings about agglomeration. McNeal (1. c.)
also considers that the agglomerating end of the parasites is sticky and
adhesive. Neither he nor other workers, however, have described such a
fragmentation of the kinetonucleus in Trypanosomes constituting typical
rosettes (cf. fig. 20), this organella usually appearing quite normal,? In
view, also, of the fact that the parasites may disagglomerate, it does not
seem probable that the disorganisation of the kinetonucleus is the cause of
agglomeration.

On the other hand some authors (e.g. Bradford and Plimmer [6], and
Stassano [108, 109]) have seen in the binary unions of T. brucii an oceur-
rence more or less comparable with the conjugation of Infusoria. There can
be no doubt that this is too extreme a view to take, there being certainly not
sufficient ground for supposing that conjugation in the strict sense, i.e.
with nuclear fusion, etc., is here taking place. What evidence there is, is
entirely against such a conclusion. In the first place agglomeration can
scarcely be considered, as an integral part of the life-cycle. ‘The process is
the more entire and lasting, the more unfavourable the conditions which
induce it. Indeed, complete fusion is not known to occur (even in binary
unions) except in those cases where the parasites are powerless to liberate
themselves, when they gradually coalesce and degenerate (see below).
Again, it is almost certain that a true conjugation, in the form of the union |
of differentiated gametes, occurs at quite a different stage of the life-cycle.

1 Laveran and Mesnil say that there is sometimes to be noticed, at the
centre of a rosette, a leucocyte or hematoblast which may, perhaps, have
served as a nucleus of attraction (cytotactic or chemiotactic) for the indi-
viduals of that cluster.

2 It seems to the writer uncertain how far these parasites with the
kinetonucleus broken up into fragments (as was frequently the case,
Prowazek says, in agglomerated clusters) ought to be regarded as normal
forms; for in addition, in one or two of Prowazek’s figures, vacuolisation
is much in evidence. Hence it is not unlikely the parasites were somewhat
altered and commencing to degenerate (see also below, under ‘‘ Chromato-
lysis’’).

THE HAEMOFLAGELLATES. poy

Lastly, it must be remembered that, even if, in some of the binary unions,
the kinetonuclei themselves join, as Bradford and Plimmer have supposed,
these are not to be regarded as sexual nuclei, comparable to micronuclei (see
above, p. 200, footnote).

Bearing in mind, however, the recent work of Calkins and others
upon the essential meaning of fertilisation, a remark of Ligniéres (71)
is not without interest in this connection. This author investigated the
phenomenon in T. equinum, where binary unions are very frequent but
fugitive, separation (disagglomeration) readily occurring. He considers it
quite probable that, as a result of the close intimacy, a molecular inter-
change goes on between the associates. The process may be stimulating or
recuperative, induced by the effect of the changes in the environment. It
is, moreover, somewhat suggestive that the agglomeration is, generally, at
first binary, and sometimes (though not often) tends to remains so. Can
the process perhaps be considered as affording hints of a plastogamic union ?
It is evident that much has still to be learnt respecting the biological
meaning of agglomeration.

(c) Abnormal and Involution Forms.

Involution and degenerative stages of Trypanosomes have received atten-
tion, and acquired an importance altogether undeserved, owing, chiefly, to
the fact that many of the parasites have been studied, so far, only in strange
and unaccustomed hosts, hosts to which they are unadapted, and for which
they, on their part, prove markedly pathogenic. That these forms are the
outcome of the unusual environment seems clearly proved by the fact that
they are rarely or never described in the case of the many tolerated para-
sites now known. For being the first to suggest the real significance of
the weird shapes often met with, and thus throwing light upon much that
had greatly puzzled previous investigators, students of the group are
indebted to Laveran and Mesnil, and this is, from the point of view of
zoologists, not the least important of the many contributions of these
authors to our knowledge of the Trypanosomes. Even as it is, the line of
demarcation between forms which are to be regarded as typical and repre-
senting a phase in the life-history, and those which are abnormal and
commencing to degenerate, is often sufficiently difficult to draw.

Trypanosomes appear to be, in most cases, able to support, for a longer or
shorter period, unfavourable conditions of environment, whether due to the
reaction of the host itself, or to the transference of the parasites to astrange
medium. Moreover, although the organisms, sooner or later, feel the effects
of such altered circumstances and show signs of involution, it by no means
follows that they rapidly die off. On the contrary, a great number of these
abnormal forms, on entering the blood of a fresh host, are able to infect it,

228 H. M. WOODCOCK.

although they may have been, in certain cases, kept for a long while in
artificial surroundings. Indeed, their vitality and ability to recuperate
themselves and give rise to a fresh succession of parasites,! if involution
has not proceeded too far, is sometimes nothing less than remarkable. For
instances and experiments showing this, reference should be made to the
works of Laveran and Mesnil, Ligniéres, Sivori and Lecler, and others.
The course which involution takes varies somewhat in different cases, and
the process may be considered as following one or another of three lines,

Fie. 21.—Involution and degeneration forms of different
Trypanosomes. For description see text. A—zE, T. gambiense
(a, c, and £ after Bruce and Nabarro; B and p after Castellani).
F, K—P, T. brucii (F after Bradf. and Plim.; x.—pP after L. and
M.). e—s, gandr, T. equinum (after Ligniéres), s, T. brucii,
plasmodial mass, from spleen-pulp (after Bradf. and Plim.).

which, though here dealt with separately for convenience, are, of course,
occasionally to be met with in combination in any given abnormal form.
These three directions are—(a) Chromatolysis, (b) vacuolisation, and
(c) change of form.

(a). In chromatolysis, either there is a more or less complete loss by the
nucleus (i.e. the trophonucleus) of its chromatic constituents, which in

1 Hence their virulence and power of infection.

THE HAMOFLAGELLATES. 229

some way pass out into the cytoplasm, leaving, finally, only the faintly or
non-staining plastinoid basis (fig. 214); or else direct fragmentation of the
nucleus occurs (F—J), this being probably a modification of the former
method. At other times it is apparently the kinetonucleus which under-
goes fragmentation. This isso in T. lewisi, according to Prowazek, and
from some of his figures of the parasites the process would certainly seem
to bear the interpretation of abnormality. In any ease, the result is much
the same. Chromatic lumps and grains, varying greatly as regards size
and number, become more or less generally disseminated throughout the
cytoplasm. Ligniéres (1. ¢.) considers that this process is simply an
abnormal development of what is a common occurrence in many Trypano-
somes. It has been pointed out that in several forms chromatoid grains
are frequently to be noticed in the cytoplasm. This author has followed _
the formation of such in T. equinum, and has seen them given off at
intervals from the nucleus.!. Moreover, it has been observed that these
chromatoid grains increase considerably beyond their normal number in
individuals placed in unfavourable surroundings, and in some cases the
involution process, at any rate in this direction, appears to stop here.

(s). Vacuolisation may also be regarded as a normal function carried
to excess. The frequent presence of a vacuole in many ‘Trypanosomes has
been mentioned above, and reasons adduced for considering that this
structure represents a normal, though not necessarily constant, cell-
organella. The first indication of abnormality in this direction is perhaps
afforded when the vacuole increases very greatly in size, as in figs. 21 £,
22x". This may be followed by the appearance of others in the cytoplasm
(figs. 21 c and 22 @) when the involution becomes pronounced in character.

(c). Change of form. This is the most obvious, and at the same time
the most far-reaching in the effects produced, of the chief lines of involu-
tion. Alteration in shape is presaged and accompanied by an increasing
loss of mobility until the parasites can no longer move. The manifold
varieties which abnormal Trypanosomes may exhibit in respect of shape
can be most easily understood, when arranged according to the normal
condition or phase which they represent, and of which they are the
degenerative results. (i). Single forms. Examples of these are seen in
fig. 21. The body becomes fat and stumpy (B—E), and may entirely lose
its trypaniform shape, becoming ovoid or spherical—in fact, like a ball
(c, 1). The flagellum is limp and inactive, and is often partially coiled
around the body (zs). The undulating membrane can no longer be made
out. (ii). Division forms. Certain individuals may have commenced
longitudinal division when the change of shape just described sets in,
and the process is not completed. Thus result more or less rounded

1 Probably some process of nuclear excretion or readjustment is con-
cerned.

VoL. 50, PART 1._—NEW SERIES. 16

230 H. M. WOODCOCK.

forms with duplicated kinetonucleus (and sometimes also trophonucleus)
and flagellum (a). At other times a quite irregular multiplication of the
locomotor apparatus takes place, leading to the formation of distorted
bodies, possessing three or four flagella at the corners, with or without
associated nuclei (fig. 22 a, E, G). Sometimes again, in more massive
forms, the cytoplasm becomes lobed and partly divided up, tending towards

Fie. 22.—Involution and degeneration forms (continued).
a—c, T. brucil, after Bradf. and Plim.; p—e, T. gambiense,
after Castellani; Hu, T. brucii, after Martini (interpreted as a
small degenerating agglomeration form); J—x, T. equinum,
after Voges; t, T. brucii, agglomeration-cluster, commencing to
form a plasmodium, after B. and P.

transverse division (J and K) or multiple segmentation (c and p). Perhaps
the appearance seen at B is to be thus interpreted, the two halves only
remaining joined by a thin cytoplasmic connecting bridge. (iii). Fusion
forms. These result either from the grouping together (partial agglomera-
tion) of individuals which had begun to show form involution (fig. 21 N),

THE H#MOFLAGELLATES. 231

or from the degeneration of more or less typical agglomeration rosettes.
In the latter case the individuals fuse up into a common mass. The process
begins in the centre and gradually extends to the periphery, the Trypano-
somes losing their independence and distinctness (fig. 221). Thus are
formed large plasmodial masses (so-called ‘“ plasmodia’’) consisting of great
numbers of nuclei embedded in a now more or less hyaline cytoplasmic
matrix (fig. 21s).

If the organisms remain subjected to the unfavourable influences, or if
involution has reached too advanced astage, death and disintegration result.
The cytoplasm is the first to disappear, becoming hyaline and colourless, and
refusing to stain up (Q, R). The nucleus rapidly follows suit. The most
resistant elements are the kinetonucleus and flagellum, which may persist
long after other traces of the organism have vanished (Pp), the former as
a little thickening at one extremity of the latter; sometimes the flagellum
alone is left.

(To be continued.)

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 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/-.

The issue of Supptemenr Numbers 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.

Each Number is sold at 10/-, 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 twenty-five copies of their
communication gratis.

All expenses of publication and illustration are paid by the
publishers.

As a rule lithographic plates, and not woodcuts, are used in
illustration.

Drawings for woodcuts should nor be inserted in the MS.,
but sent in a separate envelope to the Hditor.

Contributors to this Journal requiring eatra copies of their
communications at their own expense can have them by applying
to the Printers,

Messrs. ApLARD & Son, 224, Bartholomew Close, E.C., on
the following terms :

For every four pages or less—

25 copies : ; : : 5/-
SOW fs. ‘ ; : : 6/-
HOOs> es. : 7/-

Plates, 2/- per 25 if uncoloured; if coloured, at the same rate for
every colour.
Prepayment by P.O. Order is requested.
ALL ComMMUNICATIONS FOR THE EDITORS TO BE ADDRESSED TO THE CARE
or Mussrs. J. & A. Cuurcuitt, 7 Great Mariporouca Street,
Lonpon, W.

-————————_—_——

THE MARINE BIOLOGICAL ASSOCIATION

OF THE

UNITED KINGDOM.

70:
Patron—HIS MAJESTY THE KING.

President—Prof. E,. RAY LANKESTER, LL.D., F.R.S.

—_—_——():—_———_

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 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. 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 Oper annum.
LIFE MreMBERs . ; : : . 15 15 O Composition Fee.
FOUNDERS . : OOM OR 20 < n

GOVERNORS (Life Members of Council) 500 O O

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.

———— ee eee Oe

New Series, No. 198 (Vol. 50, Part 2). Price 10s, net.

JUNE, 1906.

THE

QUARTERLY JOURNAL

OF

MICROSCOPICAL SCIENCE.

BY

EDITED

E. RAY LANKESTER, M.A., D.Sc., LL.D., F.R.S.,

HONORARY FELLOW OF EXETER COLLEGE, OXFORD} CORRESPONDENT OF THE INSTITUTE OF FRANCE
AND OF THE IMPERIAL ACADEMY OF SCIENCES OF 8T. PETERSBURG, AND OF THE ACADEMY
OF SCIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIENCES
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 LINCEI OF ROME, 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 THE
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OF
THE ROYAL PHYSICAL SOCIETY OF EDIN-
BURGH, AND OF THE
BIOLOGICAL SOCIETY OF PARI8, AND OF THE CALIFORNIA ACADEMY OF SCIENCES OF SAN FRANCISCO;
FOREIGN ASSOCIATE OF THE NATIONAL ACADEMY OF SCIENCES, U.S., AND MEMBER OF
THE AMERICAN PHILOSOPHICAL SOCIETY 3
DIRECTOR OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSEUM, LATE FULLERIAN
PROFESSOR OF PHYSIOLOGY IN THE ROYAL INSTITUTION OF GREAT BRITAIN;
LATE LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW ©OF MERTON COLLEGE, OXFORD.

WITH THE CO-OPERATION OF

ADAM SEDGWICK, M.A., F.RS.,

FRLLOW AND TUTOR OF TRINITY COLLEGE, CAMBRIDGE S

AND

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

BEYER PROFESSOR OF ZOOLOGY IN THE OWENS COLLEGE, MANCHESTER,

WITH LITHOGRAPHIC PLATES AND TEXT-FIGURES.

cu ONDON:
J. & A. CHURCHILL, 7 GREAT MARLBOROUGH STREET.

1906.

Adlard and Son, Impr,,] (London and Dorking

CONTENTS OF No. 198.—New Series.

MEMOIRS:

The Hemoflagellates: a Review of Present Knowledge relating to
the Trypanosomes and allied forms. (Continued from p, 231.)
By H. M. Wooncock, D.Se.Lond. (With Text-figures.)

Notes on the Development, Structure, and Origin of ‘the Median and
Paired Fins of Fish. By Epwin 8. Goopricu, F.R.S., Fellow of
Merton College, Oxford. (With Plates 10—14)

Preliminary Account of a New Organ in Periplaneta orientalis.
By Rutu M. Harrison, Lady Margaret Hall, Oxford. (With
Plate 15) : ; ‘ ‘

PAGE

233

333

377

JUN .3

Sn Memory of

WALTER FRANK RAPHAEL WELDON, |
M.A., D.8c., F.BS.,

Linacre PROFESSOR OF COMPARATIVE ANATOMY IN THE UNIVERSITY
oF OXFORD, AND FELLOW OF MERTON COLLEGE, OXFORD,

WHO DIED APRIL 131u, 1906.

THE H#MOFLAGELLATES. 200

The Hemoflagellates: a Review of Present
Knowledge relating to the Trypanosomes
and allied forms.

(Continued from page 231.)

By
H. M. Woodcock, D.Sc.(Lond.).

With Text-figures.

CONTENTS.

PAGE

8. Multiplication : ; . 233
9. Comparative account of the life. agale, : , . 242
10. The ‘ Leishman-Donovan-Wright ” bodies. j . 254
11. Phylogeny and evolution ; : : : . 267
12. Systematic enumeration ‘ ; . 283
13. Appendix.—The question of the Spir ooliae ae , ow ol
14. Bibliography . : : : : : . 322

Section VIII. Muurrezication.

Binary longitudinal division is, probably, of universal
occurrence, and appears to be the usual method of multipli-
cation,! at any rate, in the Trypanosome phase. T. lewisi,

1 Certain authors (e. g. Rabinowitsch and Kempner, and Voges) have
described stages in T. lewisi and T. equinum, which they consider as
indicative of transverse division, but it is very unlikely that these represent
normal dividing forms; none of the more recent investigators of the same
(or other) parasites have observed such a method, and, in short, these stages
are to be placed in the category of involution forms showing irregular
segmentation, to which reference has just been made.

VoL. 50, PART 2.—NEW SERIES. Ay,

234. H. M. WOODCOCK.

at any rate,' possesses another method in addition, namely,
rosette-like segmentation, which is easily derivable from
the former. Longitudinal fission in general follows, in its
main outlines, the process above described in the case of
Trypanomorpha noctue; the chief differences to be
observed are slight variations in mode and order of procedure.
Full-grown Trypanosomes about to divide are, as a rule,
rather broader than the ordinary adults; in IT. lewisi, this
increase in size may be very marked, the parasites being not

A. B .

Fig. 23.—Stages in binary longitudinal fission of T. brucii.
(After Lav. and Mesn.)

only much wider, but also longer (fig. 2738). The kineto-
nucleus is frequently the first to divide (fig. 23 a), but some-
times either the kineto- or the tropho-nucleus may do so
indifferently; whichever leads the way, the other very soon
follows suit (B).

The duplication of the flagellum always begins at its
proximal end, which is in relation with the kinetonucleus.
Until recently the process has always been considered as an
actual longitudinal splitting of the flagellum, following upon the
separation of the two daughter-kinetonuclei. The splitting

1 See below, p. 240, for other possible instances.

THE HAMOFLAGELLATES. 235

has been described, either as extending to the distal end of
the undulating membrane (i. e. as far as the flagellum acts as
a border to the same [fig. 23 c]), after which the two halves
separate ; or, as being practically limited to the root-portion,
which becomes thickened and then divided, one half breaking
away as a new short flagellum, the further growth of which
is basal and centrifugal (fig. 27D). As above stated, however,
Schaudinn finds that, in Trypanomorpha noctue, the
whole of the flagellum, etc., is developed independently
from the daughter-kinetonucleus and laid down alongside
the old locomotor apparatus; moreover, Prowazek (I. c.)
maintains that this is also the case both in Trypanosoma
lewisi and T. brucii.! It appears uncertain, therefore,
whether splitting of the flagellum really occurs. However
this may be, one of the resulting flagella (the new one) is
often at first shorter than the other, either possessing only
a small free portion (fig. 23c) or none at all; more par-
ticularly is the latter the case when rapid successive multi-
plication has been going on.

Cytological details with regard to the behaviour of the nuclear apparatus,
such as are given by Schaudinn (1. ¢.), are only to hand in one or two cases.
For the most part, nuclear division has been, so far, described as consisting
simply of aggregation of the chromatin at either end, followed by con-
striction in the middle and subsequent separation of the two halves as
daughter-nuclei. Wasielewsky and Senn (120), however, mention and
figure a kind of simple mitosis in a case of multiple division in T. le wisi.
Prowazek (1. c.) has described the process in T. brucii more fully. The
kinetonucleus becomes thickened and more or less spindle-like (fig. 24 4)
Subsequently it assumes a dumb-bell-like appearance, and the two halves
become further separated, remaining connected only by a long thread (x) ;
this sometimes shows a delicate thickening (apparently divided in the fig.)

1 McNeal (74) is of the same opinion with regard to the multiplication
of T. brucii in cultures.

* The same applies equally, of course, to the formation of the undulating
membrane. If the flagellar border splits, the membrane doubtless divides
also; but where the flagellum is a new development, the undulating mem-
brane is so too. Certainly, to judge from many of the published figures
(cf. also some of Prowazek’s), one would conclude that actual splitting of
the flagellum is taking place.

236 H. M. WOODCOCK.

in the middle. The trophonucleus increases in size, and its chromatin
becomes arranged in eight rather elongated chromosomes, which next begin
to divide in a similar dumb-bell-like manner (fig. 24c). The trophonuclear
karyosome (karyocentrosome) has frequently divided by this time (c); but
in one case Prowazek observed it much drawn out and functioning as an
intranuclear division centre (D), the chromatin having become aggregated
around its two ends. In fig. 24.c the chromatoid grains in the cytoplasm
are also seen dividing.

The division of the general cytoplasm takes place last of
all. In the great majority of forms thisis equal or sub-equal
and the two resulting daughter-Trypanosomes are of approxi-
mately equal size (figs. 23, 25 c). Although the cytoplasmic
fission usually begins at the flagellar end, it is important to

Fig. 24.—Nuclear details in the division of T. brucii. A—B,
division of the kinetonucleus; c and p, of the trophonucleus.
(After Prowazek.)

note that it may commence instead at the non-flagellate ex-
tremity. This is the case (according to both Ligniéres and
Elmassian and Migone) in IT’. equinum, where the division
starts indifferently at either end (cf. fig. 25 c and p). It
would be very interesting to know at which extremity it
begins in Trypanoplasma, but, unfortunately, only an
early stage has, up till now, been described for this form
(fig. 26), and many more details are needed.

In some instances (e.g. T. brucii, ‘Il’. equinum [fig. 25 8],
‘T’. equiperdum [fig. 25 @]) the longitudinal fission is appar-
ently multiple, the original individual giving rise, simultane-

THE HEMOFLAGELLATES. 237

ously, to three or even four descendants. This is most likely
due to the successive division, before the common cytoplasm

Fig. 25.—a—p, stages in binary longitudinal fission of T.
equinum; £, multiple longitudinal division in same parasite;
F and G, binary and multiple fission in T.equiperdum. (After
Lignieres.)

has divided, of the organelle of one or both of the two
halves resulting from the first multiplication.

Fic. 26.—Early stage in binary fission of Try panoplasma
borreli. (After L. and M.)

T. lewisi differs from other Trypanosomes in that the cyto-
plasm divides ina most unequal manner (fig. 27). Indeed,

238 H. M. WOODCOCK.

the process is more comparable to budding, since the larger,
or parent-individual may produce, successively, more than
one daughter-individual?; moreover, the progeny may them-

Fria. 27.—Unequal division in ‘I’. lewisi. m. = parent-indi-
vidual; d. = daughter-individual; d’. = daughter-individual
dividing. x 2000. (A—z, after L. and M.; Fr, after Wasielewsky
and Senn.)

selves subdivide before separating, the whole family remain-
ing connected together by the non-flagellate end, which is
often much attenuated in each individual as a consequence

1 Léger (66) instances a somewhat similar unequal division or budding
in T, barbatule.

THE HM MOFLAGELLATES. 239

of the numerous divisions (in fig. 27, » represents a small
family, and Fr a large one where the parent-individual is
very distinct). It is important to note that the kineto-
nucleus changes its position during the commencing stages
of division and comes to lie alongside the trophonucleus, or
even passes to the other side of it, i.e. nearer the flagellar
end.

Frq@. 28.—a—D, rosette segmentation in T. lewisi; b, daughter-
individual; F, one dividing. 1750. (After L. and M.)

This variety of division forms a direct transition between
binary fission and the other characteristic method of ‘I’. le wisi,
viz. segmentation ; indeed, such a family as that just described
often greatly resembles a rosette, but is generally distin-
guishable therefrom by the presence of a parent-individual.
In rosette-formation, on the other hand, segmentation is
multiple and equal.' The body assumes an ovoid to spherical

1 McNeal (I. c.) doubts the occurrence of true equal segmentation. He
thinks that the ‘ budding”’ process is rather concerned in all cases, i. e. that
a parent-individual can always be recognised. The fact, however, that equal
multiple longitudinal fission occurs, in which there is no sign of a parent-
individual (cf. fig. 25 a), at least makes it possible that the process of equal
segmentation occurs. ‘The condition seen in fig. 28 B is easily derivable
from that in the former figure. See also next page.

24.0 H. M. WOODCOCK.

form, and then repeated division of the nuclei (both kinds)
and flagellum goes on, each daughter-kinetonucleus remaining
in contiguity to the corresponding trophonucleus, and all
taking up a position of uniform distribution near the margin
of the cytoplasm (fig. 28 a—c). The cytoplasm next becomes
lobulated peripherally, and gradually segregates around the
nuclei, forming as many little, radially-arranged daughter-
Trypanosomes as there are nuclear groups (D).'_ This method
of multiple division offers, it will be seen, considerable analogy
to the schizogony of Hemosporidia, and the latter is probably
to be regarded as a modification of it, adapted to an intra-
cellular and gregariniform stage.

These small Trypanosomes so formed (xz) differ from the
typical adults by their stumpy, pyriform shape, the position
of the kinetonucleus near the flagellar end of the body, and
the absence, during the first part of their youth, of an undu-
lating membrane. The parasites have at this period, it is to
be noticed, a very Herpetomonas-like facies, the import-
ance of which is discussed below (p. 276). These young indivi-
duals of T. lewisi can themselves multiply by equal binary
fission (fig. 28 r), and give rise to little fusiform ‘Trypanosomes.
With growth the latter gradually assume the adult appearance,
by the progression of the kinetonucleus past the trophonucleus,
almost to the other end of the body, and the concurrent
development of an undulating membrane as the extended
flagellum takes up its regular superficially-attached position.

Wasielewsky and Senn (l.c.) also describe and figure
multiple segmentation in this same parasite; again Rabino-
witsch and Kempner, in a recent paper (90), mention that
they have occasionally observed appearances and groupings
of Nagana and Dourine parasites (‘I’. brucii and T’. equi-
perdum) which strongly suggest that these also possess, in
addition to the usual method of equal binary fission, a
similar modification. Apart from these instances, however,

1 A multiplication-rosette is readily distinguished from a typical agglo-
meration-cluster by the different shape of the individuals, and the different
position of the kinetonucleus (cf. fig. 20 B).

THE H#MOFLAGELLATES. 241

the writer has not come across any other observations of
multiplication-rosettes being formed by the parasites while in
the blood.

On the other hand, as above indicated, the occurrence of such rosettes
has frequently been observed in artificial cultures of different Trypanosomes.
Fig. 29 shows two such clusters in the case of T. lewisi. The larger ones
are formed apparently by the successive divisions of the elements of the
smaller ones; concurrently the individuals gradually lose their pyriform
shape and become more elongate and fusiform. In this way colonies of
hundreds of individuals are formed.

The origin of these rosettes appears to be by the multiple division of a
single form. The radial arrangement and the general shape of the para-
sites in a small cluster suggests this (cf. fig. 29 a with fig. 28 p, showing

Fic. 29.—Multiplication-rosettes of T. lewisi from a culture.
(After L. and M.)

segmentation in the blood). Novy and McNeal, moreover, describe and
figure (81) early stages in such multiple division in T, avium and other
Avian parasites, which they consider will lead on to the formation of a
rosette. Probably, however, “segmentation” is soon replaced by rapid
binary division. A noticeable distinction from the multiplication-rosettes
of T. lewisi in the blood is that, in most cases described, the clusters of
Trypanosomes so formed in cultures have their flagella and kinetonuclei
centrally disposed.! It is by no means impossible that this results from the
strange medium in which the parasites are; in so far as this is the case,
the arrangement must be considered abnormal.

1 This is not so in ‘I’. brucii, which, according to both Smedley and
McNeal, preserves the “ blood-type,”—i. e. the flagella are outwardly
directed.

242 H. M. WOODCOCK.

Srectrion LX. CompaRATIVE ACCOUNT OF THE LIFE-CYCLE.

(a4) Life-history of Trypanosoma ziemanni com-
pared with that of Trypanomorpha.

Besides demonstrating a complex life-cycle in the case of
Trypanomorpha, Schaudinn has shown that another para-
site of Athene noctua, Trypanosoma ziemanni, also
has a similar history, and undergoes much of its development

Fig. 30.—Growth and metamorphosis of an indifferent ookinete
of Trypanosoma ziemanni. (After Schaudinn.)

in the same Invertebrate host, Culex pipiens. While
agreeing in its main features with that of the first-named
parasite, the life-cycle of this form presents certain important
differences, which may now be considered."

The most remarkable modification, and one which bears forcibly upon the
relationship of the Hemosporidia to this group, occurs almost at the com-
mencement of the life-history. Instead of passing at once into the Try-
panosome stage, all three types of ookinete first enter upon a period of

1 For the general account of the life-history the reader is referred to
the description of Trypanomorpha, and to the tabulated summary of
the principal stages given on p. 180,

THE H#MOFLAGELLATES. 243

growth and nuclear multiplication. Following the process in an indifferent
form (fig. 30), it is seen that growth results principally in an increase in
length, and the ookinete becomes serpentine-like and then rolled and coiled
up upon itself (B and c), finally assuming the appearance of a tangled ball
or skein. Meanwhile nuclear division is proceeding. The nucleus here
remains for some time in the compound condition. Nuclear division takes
place in the same manner as in the megayametocyte of Trypanomorpha
(Halteridium) noctua, and the kinetonucleus functions as a central
spindle. By successive divisions a great number of nuclei! are at length
produced, uniformly distributed throughout the coil (fig. 30c). To each
daughter-nucleus a small zone of cytoplasm is apportioned, and each of
the “cell-territories”’ thus segregated becomes a little Trypanosome in
the way above described. By a similar process male and female ookinetes

Fig. 31.—Liberation of young indifferent Trypanosomes from
the coiled ookinete in ‘I’. ziemanni. (After Schaudinn.)

give rise at length to numbers of male and female individuals. When fully
organised, the young Trypanosomes, be they male, female, or indifferent,
liberate themselves and move away, leaving behind a large residual mass
of unused cytoplasm (fig. 31).

When first set free, the young trypaniform parasites are very small, the
males being indeed, according to Schaudinn, hardly visible.?- The indifferent
ones become greatly extended in length, and somewhat spirally twisted,
soon attaining the adult form. In all fundamental respects the organiza-
tion of the different types agrees completely with that of other Trypano-

1 The number varies according to the type of ookinete.
? This is also true of the indifferent forms after repeated multiplication,
which can then only be made out when moving or agglomerated in clusters.
Schaudinn puts forward the interesting suggestion that there may possibly
be Protozoan parasites which, at certain periods of the life-cycle, ean no
longer be optically resolved. Such may perhaps be the case in yellow fever.

244, H. M. WOODCOCK.

somata; differences in detail have been commented upon in the section
on Morphology.

Longitudinal fission is peculiar because of the fact that the two resulting
daughter individuals do not at once separate, but remain united by the non-
flagellate ends and take up a position of alignment one with the other
(fig. 32 B), either end of the thread being formed by the flagellar extremity
of one of the two parasites. These double individuals or “couples”
constitute very thin, corkscrew-like, spirochetiform threads. Moreover,
further division takes place while the individuals of a couple remain
thus joined. Fig. 32 c and p shows successive stages in the longitudinal

Fig. 32.—a—p, formation and multiplication of “couples”
(‘Spirochexta’’-threads) in T. ziemanni; E, F, resting-phases
of same; @, agglomerated cluster of very minute forms. (After

Schaudinn.)

fission of a couple, and the separation of the two daughter couples from
each other. These couples move indifferently in either direction, now one
member leading, now the other. There is at present no evidence as to the
particular biological signification, if any, of this occurrence.

With the assumption by the parasites of a resting phase, the body
becomes pear-shaped in form.’ In fig. 32 £ is seen a couple in the gregarini-

2 The writer is not able to gather from Schaudinn’s account whether the
two members of a daughter-couple then separate before again dividing, or
whether further multiplication goes on in a similar manner.

3 The striking resemblance between such a stage and “ Piroplasma”
donovani (see below, p. 258, et seq.) hardly needs pointing out,

THE HA MOFLAGELLATES. 245

opposite ends are rounded and have lost the flagella. Multiplication also
goes on during this phase ; fig. 32 F shows four individuals, i. e. two couples
not yet separated.

The behaviour of the indifferent forms, on entering the blood of the owl,
is very much the same as in the case of Trypanomorpha. After alter-
nation of resting, attached phases with multiplication periods has proceeded
for some time, sexual individuals are developed in increasing numbers from
the young indifferent forms. Sexual forms (gametocytes), whether male

B.

Fic. 33.—Active and resting phases of the gametocytes of T.
ziemanni. A—C, megagametocyte (female Trypanosome) ; D, E,
microgametocyte (male Trypanosome). (After Schaudinn.) [In
D, the myonemes are rather too accentuated. |

or female, are easily distinguishable by the remarkable size to which
they grow, becoming, as they do, very much larger than the leucocytes.
An adult fully-grown male individual, or microgametocyte, in the try-
paniform phase is seen in fig. 33 p. Both trophonucleus and _ kineto-
nucleus are very prominent, and the undulating membrane and flagellum
are well-developed; the latter extends some distance beyond the posterior
end of the body. The sixteen myonemes are arranged in four double rows
or pairs on each side. A female Trypanosome (megagametocyte) in the
same phase (fig. 33 A) is even larger than a male form, but its nuclei are
relatively smaller, and there is no free prolongation of the flagellum. The
myonemes of each side are not arranged in pairs.

246 H. M. WOODCOCK.

Resting, intracellular phases (the ‘ Leucocytozoon”’ of Danilewsky and
the “Hamameba’’-stage of Laveran) of these sexual forms are shown
in fig. 33 c (male) and E (female). In both gametocytes the greater part
of the body (endoplasm) becomes retracted into an ovoid mass; the ecto-
plasm with its myonemes, in conjunction with the cytoplasmic envelope of
the host-cell, alone retains the original spindle-shaped outline of the para-
site. No trace of the flagellar apparatus is left, and the kinetonucleus
takes up an internal position, close to the trophonucleus (fig. 33 Cc).

Fic. 34. Fie. 35.

Fig. 34.—Formation of the eight microgametes from the micro-

gametocyte in T. ziemanni. The unused cytoplasm breaks up
into three or four residual masses.

Fig. 35.—Fertilisation of a megagamete by a microgamete.
The trophic and kinetic female pronuclei are seen on the left.
Near the centre lie the two reduction-nuclei.

In both figures the remains of the host-cell, together with the

cast-off ectoplasmic envelope of the parasite, are seen on the
right. (After Schaudinn.)

The gametocytes in this stage are, of course, to be met with in the peri-
pheral circulation, and when introduced (with the blood) into a gnat during
the act of biting they at once proceed to gamete-formation. A fully grown
microgametocyte may undergo the requisite nuclear changes and multipli-
cation while still in the blood of the owl, and it is this process actually
which is beginning in the individual seen in fig. 33 £2 Eight double nuclei

1 Laveran (37), it may be noted, figures also rounded or spherical forms
of the gametocytes and their host-cells, but Schaudinn does not mention
the occurrence of such.

2 In some cases apparently the microgametes may even be formed and
liberated from the parent-cell while in the blood. lLaveran (I. c.) figures

THE H#MOFLAGELLATES. 24.7

are formed, and the reduction of the chromosomes of the trophonuclear parts,

from sixteen to eight, occurs at this time. The formation and liberation of

the microgametes (fig. 34), the maturation of the megagametoeyte and the

fertilisation of the megagamete (fig. 35), all take place, according to the

author, in the same manner as in Trypanomorpha (Halteridium)

noctuz. The microgametes themselves are constituted on the same Try-

panosome plan, allowance being made for the different number of chromo-

somes. With fertilisation! and the subsequent formation of the vermiform, -
motile ookinete, the life-cycle of Trypanosoma (“Hemameba’”) zie-

manni is completed.

It only remains to add that “recurrence” is produced in the usual way,
by parthenogenesis of female forms remaining over in the blood of the bird.
The only point requiring notice is that the rejuvenated parasite, instead of
giving rise to a single individual of any type, undergoes multiple division
like an ordinary ookinete and produces many little Trypanosomes.

>

(s) Evidence in favour of a corresponding Dige-
netic Life-cycle in other Trypanosomes.

Since Schaudinn’s work was published, evidence has been
accumulating which tends to show that the two examples so
brilliantly investigated by this author are not isolated cases,
but are rather to be regarded as, if not in every way typical,

four such in the act of being set free in this same parasite. MacCallum
(72) also observed the same number liberated in a species of Halteridium,
and, moreover, the actual fertilisation of a megagamete. Compare also
the “ Polymitus’’-forms of earlier authors, which were simply detached
“flagella” or gametes. Whether the occurrence is due to unfavourable
circumstances (e. g. removal from the body) is not certain, but it does not
seem to be quite normal, the number of gametes, for instance, being, at any
rate in the first named case, only half the normal number produced in the
enat. Is this possibly consequent on the non-completion of the reduction
phenomena?

1 Details of the process are not given, but the writer would point out
that the penetration of the microgamete into the megagamete is, most
probably, by the opposite end to that which moves first in Trypano-
morpha, namely, the non-flagellate end. This would be in accordance
with the behaviour of the parasites as regards attachment, agglomeration,
ete. Probably also in this case (it appears so, indeed, from Schaudinn’s
figure) there is a true flagellar prolongation posteriorly to act as a steerage
organ.

248 H. M. WOODCOCK.

at least indicative of the general course of events in the
usual life of a Trypanosome. For one thing, there can be
now little or no doubt that most, if not all, Trypanosomes
possess an alternate, Invertebrate host, in which a definite
part of the life-cycle, including sexual conjugation, is under-
gone. Again, it is very probable that many are capable of
entering upon a resting, attached phase, at different periods
of the life-history, during which the parasites lose, for the
time being, their trypaniform nature, and become gregarini-
form. In certain cases, indeed, they have gone even farther,
and acquired a completely intra-cellular, or Heemosporidian
phase.!

The facts in support of the above-mentioned propositions
are most conveniently treated more or less separately.
Apart from the many Invertebrates which are known, or
with good reason suspected, to be, at least, “ carriers ” of
‘Trypanosome parasites (see below in the Systematic section),
there are certain other Trypanosomes, from different Verte-
brate hosts, for which the possession of an alternate In-
vertebrate host has been, practically speaking, proved. Per-
haps the most important instance, since it relates to a
Mammalian form, is that lately described by Prowazek (88).?
This author finds that T. lewisi undergoes an essential part
of its life-cycle in a louse, Heematopinus.

Soon after their arrival in the mid-gut the parasites
undergo reduction of the nuclear apparatus, preparatory to
conjugation. By this means the number of chromosomes is
reduced from sixteen to four. The differences between the
gametocytes of different sex (male and female Trypanosomes)
are not well marked. A noteworthy distinction from the
instances above described is that the male form (comparable
to a microgametocyte) does not give rise to several micro-

1 The derivation of the Hzmatozoa as a whole, and the direction in
which evolution appears to have tended, are considered in Section XT.

2 As the writer has only had access to Prowazek’s work since this article
was sent to the publishers, it is impossible to do more than summarise the
principal features,

THE HEMOFLAGELLATES. 249

gametes, but becomes itself a single one, in the same way
that a female form becomes a megagamete after maturation.
The body diminishes in size, and the nucleus (tropho-
nucleus), becomes elongated and band-like (fig. 36 A) ; also
the cytoplasm stains differently from that of a female gamete.
Hence, when fully formed, the two kinds of sexual element
are readily distinguishable, although there is not any pro-
nounced dissimilarity in type. Actual conjugation stages
are shown in fig. 3638 and c; and it is, in the writer’s

Fic. 36.—a, microgamete of T. lewisi; Band c, stages in con-
jugation; D, zygote (ookinete). (After Prowazek.)

opinion, a point of great importance that the gametes come
into contact by their non-flagellate, kinetonuclear ends.
The zygote becomes an ookinete (fig. 36D), quite similar in
constitution to those of Trypanomorpha and Trypano-
soma ziemanni; and this gives rise toa single Trypano-
some by the separation of kinetonucleus and locomotor appa-
ratus, also in the same manner as in those parasites.

Prowazek has also endeavoured to ascertain more of the life-cycle of
T. brucii. He was unable, however, to obtain the Tsetse-fly (Glossina
morsitans) in which, as he remarks, the sexual phases normally occur, and
only very rarely were maturation-processes, comparable with those described
for T. lewisi, to be seen in the blood. In one case, where the parasites
were in a guinea-pig which had just died, he was able to observe what was
undoubtedly an actual conjugation, although probably not altogether typi-
cal (cf. footnote 2, p. 246).

vou. 50, PART 2.—NEW SERIES. 18

250 H. M. WOODCOCK.

In addition the author describes various nuclear changes and divisions
undergone by the parasites (both forms) while in the blood of the
Vertebrate host. These include a regulatory process, characterised as
autosynthesis of the karyosome (karyocentrosome) and partheno-

genesis.!

Equally interesting is the evidence already to hand, which
tends to prove that the réle played by an Insect in connection
with the Trypanosomes of warm-blooded Vertebrates is per-
formed by a leech in the case of those of cold-blooded
Vertebrates.

To Léger (66 and 67) we owe certain instructive observations
relating to Trypanoplasma varium and Trypanosoma
barbatulz from the loach.2. This investigator distinguishes
indifferent and female forms of Trypanoplasma varium
in the blood of the fish. When a leech (Hemiclepsis
marginata) sucks blood containing the parasites, which
thereupon pass into its stomach, the indifferent forms dege-
nerate and perish, while the female ones become massive and
show nuclear changes (division of both tropho- and kineto-
nucleus), preparatory, Léger thinks, to a sexual process. At
any rate, after some days the intestine of the leech contains
little, narrow Trypanoplasmata, of which certain, very
filiform, ones represent, perhaps, male forms.*® Other stages
were also noticed whose interpretation remains at present
doubtful. In the case of Trypanosoma barbatule the
resemblance between the development of the parasites in
another leech (a Piscicola) and that of Schaudinn’s Avian
forms in the gnat is still more pronounced. Highteen hours

1 Tt may be pointed out that certain of the appearances depicted as
representing different stages in these processes do not suggest, quite so
readily as might be desired, the interpretation given of them. In one or
two cases, at any rate, the figures recall those of Bose (5), who describes
some wonderful phases in the development of a Trypanosome from the
rabbit which the writer, however, does not for a moment consider are really
normal.

? Only a short preliminary note without figures is as yet available.

3 Brumpt (10) has also noticed small, very motile Trypanosomes in this
leech,

THE HAMOFLAGELLATES. 251

after infection Léger observed, in the intestinal contents,
pyriform ookinetes without a flagellum ; some of these pos-
sessed a single large nucleus (i.e. a “ compound” nucleus),
either at rest or in process of heteropolic division, while
others had two nuclei, of which one was smaller than the other.
Four days later the intestine contained numerous Trypano-
somes which could be distinguished as belonging to one of
the three types described by Schaudinn in Trypano-
morpha, namely, indifferent, male and female. The indif-
ferent forms, it may be noted, multiply actively by longitudinal
fission (see p. 238). Other details of these various types
are given, but their further evolution and the manner of
their passage back into the loach was not followed. The
above-stated facts, however, hardly leave room for doubt
that both these piscine Trypanosomes have a true, alter-
nating, Hirudinean host.

Billet’s work (8) on Trypanosoma inopinatum is also
very important, and goes far towards proving both proposi-
tions for this form. The author brings forward evidence to
show that (1) the alternate, Invertebrate host of this para-
site is another leech, Helobdella algira, and (2) that it
has an actual ontogenetic relationship with a Drepanidium
or Lankesterella parasitic in the same Vertebrate, and
which is, most probably, its Hzmosporidian phase. The
investigation was somewhat complicated by the presence in
the same frogs of the common T. rotatorium, but, on the
other hand, the coincidence was instructive, since the result
obtained tends to show that that particular leech is not the
alternate host of the latter parasite.

The principal facts brought out by Billet’s experiments are as follows :—
(a) After the examination of a number of leeches which had been ectopara-
sitic upon frogs containing in their blood either Lankesterella (Drepa-
nidium) plus T. inopinatum plus T. rotatorium, or only Lankes-
terella, it was found that, in either case, the Helobdelle contained in
their intestine only T. inopinatum. Neither Lankesterella, as such,
nor T. rotatorium was met with in this host. (b) Moreover, Billet
several times observed in the intestine of the leeches, within twenty-four
hours after the time of infection, parasites which were more or less rounded

252 H. M. WOODCOCK.

and mobile, and with nucleus and centrosome (i.e. tropho- and kineto-
nucleus) distinct—stages intermediate, that is, between a Hemogregarine
and a Trypanosome.! (c) Frogs quite free from all Hematozoa and then
infected by placing Helobdellew, whose digestive tube contained T.
inopinatum, upon them were found afterwards to contain only Lankes-
terella.

Billet considers the Trypanosome phase to be very uncommon in the
frog, but of general occurrence in the leech, and, conversely, the Haemo-
gregarine phase to be absent, as such, in the Invertebrate host, but common
in the Vertebrate. T. inopinatum does of course occur in the frog—
Sergent (101) first described it in that host—but apparently it is only at rare
intervals that the parasites lose the Hemogregarine condition and become
trypaniform. It may be pointed out in this connection that Sergent (108)
who has corroborated Schaudinn’s researches, says that Trypanomorpha
in the Trypanosome-form is comparatively rare in the blood of the owl, but
common as Halteridium, and, vice versa, in the gnat the latter phase
is not represented.

Further, Billet in a previous communication (4) has described forms
which he regards as intermediate between merozoites of a Lankesterella
and the typical trypaniform phases of T.inopinatum. He also observed
the latter penetrate into the red blood-corpuscles, losing the flagellum in so
doing, and then multiply in this endo-globular situation, either by binary
longitudinal fission or by schizogony.

The study of this important question of the life-history of
a Heematozoan, with regard to its bearing upon the relation-
ship between Trypanosomes and Heemosporidia, may also be
approached from exactly the opposite direction. In other
words, it is sometimes quite as easy, or even easier, to work
from the latter to the former, to look for a trypaniform phase
in a recognised and well-known Heemosporidian. Some
very interesting instances of such a discovery are now to hand,
foremost among them being one for which we are again
indebted to that marvellous investigator, Schaudinn.

This author states that he has observed the development of
a motile trypaniform phase at two points in the life-cycle of

1 Brumpt (10), it is important to note, has also observed in the ookinetes
of Hemogregarina bagensis (from Emys leprosa) which he found in
the cesophageal and stomach diverticula of Placobdella catenigera, two
nuclear bodies, a large nucleus of the ordinary type, and a smaller, highly-
staining body much resembling the centrosome (ie. the kinetonucleus) of a
Trypanosome.

THE HAMOFLAGELLATES. 253

the tertian parasite (Plasmodium vivax), both the sporo-
zoites and merozoites evincing in their construction charac-
teristic Trypanosome features.' He further considers, and
hopes later to completely demonstrate, that this malarial
parasite agrees with Trypanosoma (“Hemamceba”)
ziemanni in other fundamental respects, as follows :—(a)
That the ookinetes of Plasmodium are really formed of a
compact coil: and (b) that the ‘“ sporozoites”’ are not all of the
same character, but that indifferent and female ones can be
distinguished ; those corresponding to the male forms, how-
ever, perish prematurely while still in the ookinete of that
type. Moreover, while the indifferent sporozoites are typical,
actively-motile ‘'rypanosomes, the female ones no longer
appear able to assume the trypaniform condition, but remain
gregariniform.

Schaudinn concludes his remarkable work with an appendix
in which he brings forward certain facts noticed by Kossel
and Weber, which led these investigators to consider that there
is a similar close relationship between Piroplasma and the
Hemoflagellates. With this view Schaudinn, who has exa-
mined the preparations, expresses himself in complete agree-
ment; and (it may be here mentioned) it has recently received
strong confirmation from the work of Rogers, on anew human
parasite, Piroplasma donovani.? So long ago as 1900,
Weber observed in preparations ® of the blood of a cow suffer-
ing from hemoglobinuria, besides the typical ‘Texas-fever
parasites (P. bigeminum), Trypanosome-like forms, much
smaller than those of Surra (Nagana). ‘These parasites showed
the general characteristics (shape, nuclear dimorphism, etc.)

1 In this connection the writer would call attention to the markedly
spirochetiform nature of the sporozoites of Hamogregarina stepanovi,
as described by Siegel (105) in Placobdella.

7 A full résumé of present knowledge relating to this remarkable para-
site is given in the next section.

3 It may be noted that the blood was drawn towards evening, and the
animal was confined in a stall into which scarcely any light came. Cf. the

time when “Halteridium” parasites leave the blood-corpuscles and
become free in the blood, as Trypanomorpha.

254 H. M. WOODCOCK.

of a Trypanosome. Besides this, Schaudinn has since also
noticed nuclear dimorphism (i. e. the presence of a larger and
a smaller nuclear body) in typical endoglobular individuals of
P. canis (cp. P. donovani below, p. 260) ; and Kossel and
Weber on re-staining and re-examining smears of the intestinal
contents of ticks which had fed upon cattle suffering from
piroplasmosis came across similar forms.

Lastly, the instances where authors mention the association,
concurrently, of ‘'rypanosomes with either Avian malarial
parasites, Piroplasmata, or Hemogregarines are _ too
numerous to specify (see below in Systematic). Of course,
in many cases, this may reasonably be set down as a mere
coincidence; it would be unduly straining that explanation,
however, to suppose that it is of universal application.

Section X. THe “ LersHman-Donovan-Whicur’’! Bopiss.

We may now consider, in some detail, the peculiar para-
sites which are generally held to be the cause of certain
tropical fevers, particularly prevalent throughout Indo-
Burmah, though not, apparently, by any means restricted to
that region. ‘These diseases, characterised by irregular
pyrexia, splenomegaly and cachexia,” are known by various
names (e.g., Dum-dum fever, Kala-azar, tropical spleno-
megaly, etc.) according to the slightly different features
and circumstances attending their occurrence in different
cases. ‘l'hese varieties are, however, most likely, all due to
one and the same specific form of parasite. Moreover,
organisms very similar to these parasites (morphologically,
indeed, the two sorts appear hardly distinguishable) are found
in certain superficial sores or ulcers, to which people in various

1 In order not to injure the delicate susceptibilities of medical investi-
gators, where priority rights are concerned (vide the pages of the ‘ Lancet’
and ‘B. M. J.’ during the last few years with reference to the discovery of
parasites in tropical diseases !), the fullest possible title is conferred for the
nonce upon these unhappy parasites,

2 Other prominent symptoms in different cases are ulceration of the
intestine, edema of the feet, and increase in pigmentation of the skin.

THE HMMOFLAGELLA'TES. 255

parts of the Kast are liable, and which are known by such
names as Delhi boil, ‘bouton d’Alep,” oriental sore, tropical
ulcer, etc. The latter type of disease is one of localised
infection, the organisms being restricted to the neighbour-
hood of the sore or ulcer, whereas in the former there is a
general infection, the parasites spreading to all parts of the
body, and being met with in the liver, spleen, bone-marrow,
etc., and (rarely) in the peripheral circulation. No actual
connection has yet been established between the parasites of
local and general infectivity!

Medical opinion on the whole is at present inclined to
regard both these types of disease as being due to different
manifestations of the same kind of parasite, the differences
in symptomatology being largely explainable by the different
habitat of the parasites in the two cases.? In view, however,
of this different habitat and behaviour it is very uncertain
whether the organisms, notwithstanding their apparent simi-
larity, really belong to one and the same species. Bearing in
mind the very sight morphological differences to be found with
any constancy among many Mainmalian 'l'rypanosomes, which
are, on other grounds (habitat, behaviour towards immu-
nising sera, etc.), regarded as belonging to distinct species,
it is not unlikely that the same holds good here also. But
the question is still far from being settled.®

1 In other words, the parasites, when limited to the neighbourhood of an
ulcer or sore, never seem to become generally distributed throughout the
body, and produce the febrile type of disease ; although persons suffering
from the latter may have skin eruptions or papule, leading to the formation
of small ulcers which somewhat resemble ‘ oriental sores.”

* James (126) is a little inclined to doubt whether the organisms are
really the cause of such a severe illness as Kala-azar, although recognising
the constancy of their occurrence. He hesitates because of the great
resemblance which they bear to the parasites of the other type of disease
(oriental sore, for example), and the very different effects to be accounted
for (see next footnote, however).

» Manson (182) puts forward the interesting hypothesis that the two
types of disease, of such different gravity, may bear to one another the same
relation that variola and vaccinia do. He suggests that the germ of Delhi
boil may be that of Kala-azar, which has become attenuated by passing

256 H. M. WOODCOCK.

The history of our knowledge of these organisms is soon stated. The
parasite of cachexial fever and splenomegaly was first discovered by Leish-
man in 1900 in a splenic puncture taken, post-mortem, from a soldier
who had contracted Dum-dum fever. His first account of it (129) was pub-
lished in the spring of 1903. About this time Donovan, in Madras, found
these bodies in the same situation, but he appears to have been dubious
about their nature until he learnt of Leishman’s discovery. Since then
many investigators have examined the parasites and published their
views concerning them, among others being Laveran and Mesnil (127),
Christophers (128), Donovan (124), and Marchand and Ledingham (1386).
Progress in this direction has culminated, for the time being, in the very
important discovery of Rogers (188) already mentioned. With regard to the
other type of disease, Wright (142) first published, at the end of 1903, an
account of tropical ulcer in which similar parasites were clearly recognised
and definitely described as such. Earlier writers on the Delhi boil malady
(e.g. Cunningham, Firth) may or may not have actually seen the same
organisins,! as distinct from altered leucocyte cells, etc., but from their
descriptions it is quite impossible to say. ‘lherefore the credit of the dis-
covery can, logically, be no more given them than can, say, that of first
finding a Trypanosome in the human blood be assigned to Barron, on the
strength of his loose description of a Flagellate met with in an anemic
woman—which may have been anything. Quite independently of Wright,
and before seeing his paper, two Russian workers, Martzinowsky and
Bogroff, also found the parasites in cases of “ bouton d’Alep,’’ but did not
publish their discovery till later.

After these preliminaries, we may pass to the parasites
themselves and their relation to the host. Considering first
Leishman’s form, which may be termed the splenic variety,
since it is always present in spleen punctures or smears ;
this is either free or intracellular. Jn the latter case the
organisms are usually parasitic in large uninuclear leucocytes
(fig. 37 11) or (and perhaps chiefly) in cells of the vascular
endothelium, particularly of the spleen (fig. 37 I M1), which are
often packed with the little parasitic elements, becoming
greatly enlarged and distended (macrophages). Parasite-
containing cells, both leucocytes and macrophages, are also
through camels (cf. footnote, p. 266), just as the smallpox germ is deprived
of its virulence by passing through the cow.

1 It seems very improbable that Cunningham saw the real parasites; the

‘“nucleoid’’ bodies he describes averaged nearly three times the lineal
diameters of those below described.

THE HE MOFLAGELLATES. 957

to be found, in greater or less number, in the other organs,
—liver, kidneys, mesenteric glands,—and in the granulation
tissue of intestinal ulcers and skin lesions.

In films or smears made from “ spleen-pulp ’
parasites seem to be “free ”—1. e. not definitely intracellular.
They are, however, embedded in a zoogloea-like matrix or
stroma, and frequently clustered together in groups. This
matrix 1s composed of rounded or irregular elements, of a
finely-granular or reticular nature, and varying greatly in
size, which have more or less run together to give the
stroma-like appearance. ‘lhe generally-accepted explanation
of this structure (which is not evident in sections of the
same material) is that of Christophers, who considers it to
be mechanically produced during the preparation of the film
by the rupture or fragmentation of the large macrophages.
These often possess cytoplasmic buds or outgrowths, easily
detachable from the main cell, and each usually containing a
larger or smaller number of the parasites.

There can be no doubt that a similar process goes on
normally. The organisms appear to be quite uninjured by
the leucocytic cells; but these, on the contrary, when
strongly infected, or after endogenous multiplication has
gone on for some time, become vacuolated and gradually
used up, and reduced to a mere skin or envelope, which at
length ruptures and liberates the enclosed parasites, just as
in the case of an ordinary Hemosporidian. It is not known
if the parasites remain free in the general circulation for any
length of time before invading a fresh host-cell; the life-
history has not yet been sufficiently ascertained for us to do
more than conjecture. It appears not unlikely, however,
that multiplication also goes on in this condition. Certain
workers (e.g. Laveran and Mesnil, Donovan, and Rogers)
figure multiplication forms free in the blood, and from
analogy, either with Piroplasma or a Trypanosome, such
an occurrence might be expected. It is most likely that, here
also, time and circumstance are largely responsible for the
behaviour of the organisms when liberated.

> many of the

258 H. M. WOODCOCK.

It is also uncertain whether the infection of the leucocytes
is an active or a passive one. The opinion has been
expressed that the host-cells, in their phagocytic capacity,
ingest the parasites—if so, verily a case of the “ biter bit.”
On the other hand, if the parasites enter the cells when in
a Flagellate phase (as e.g. ‘I’. ziemanni penetrates the
leucocytes of the owl) it is more likely an instance of active
infection, the organisms being especially adapted to that kind
of cell—‘ Leucocytozoa” in fact. Nevertheless, it appears
highly probable that the parasites are not exclusively limited
to such a habitat, but that they also attack red blood-
corpuscles as well, being also, therefore, in that respect, true
Hemosporidia. Both Laveran and Mesnil and Donovan
describe and figure them as endoglobular, and great
weight must be assigned to the view of the first-named
investigators. Moreover, the figures of Donovan (Il. c.) are
sufficiently convincing ; and, having regard to the unmistak-
ably Piroplasma-like facies which the organisms at times
possess, the rod or comma-torms depicted (fig. 37 Le) strongly
recall the “ bacillary ” type of P. bigeminum in a similar
situation (see Laveran [84]). Lastly, Donovan has observed
the parasites in the peripheral circulation, although only
rarely and during periods of high fever; and Laveran, who
has examined his preparations, confirms this observation.

There is little additional to be noted concerning the
habitat of the localised parasite (Wright’s ulcer form).
This also is either free or intracellular; in the latter case
it is parasitic in the ulcer cells and in the large migratory
corpuscles (phagocytes), which doubtless correspond, in part,
to the macrophageal cells of the other type. Mesnil,
Nicolle, and Remlinger (137) have seen, in sections, multi-
plication forms, both free and intracellular.

The parasites themselves are very minute, and appear
rounded, ovoid, or pyriform in shape (fig. 37); the typical
form may very well be that of a slightly flattened pear. ‘The
splenic form is somewhat smaller than Wright’s parasite,
and this is about the only visible distinction between the

THE HAMOFLAGELLATES, 259

two. The former is, when round or oval, from 2—8} in
diameter,’ when pyriform from 3$—4 pin length by 14 or

Fre. 37.—]. Piroplasma donovani (Lav. and Mesn.): a,
typical pear-shaped or oval forms; b, various stages in longitu-
dinal division; c, nuclear division preparatory to multiple fission ;
d, endoglobular forms, in red blood-corpuscles (p = pigment
grains); e, bacillary form of the parasite in a corpuscle; M
= large macrophageal cell with many parasites. (After Donovan.)

II. Uninuclear leucocyte (LZ) containing several parasites. (After
L. and M.)

Ill. P. (Helcosoma) tropicum (Wright). a, single indi.
viduals ; b, dividing forms. (From Mesnil, mostly after Wright.)

IV. P. donovani in cultures of different ages. a, ordinary
forms of varying size; b, c, stages in multiple division; d, binary
fission ; e, f, and g, flagellate forms. (After Rogers.)

slightly more in breadth. The forms from tropical ulcers
average about 4 by 3m (fig. 37 III). The body is, most pro-
1 Laveran and Mesnil describe those occurring in the peripheral blood as

being much smaller, and probably young forms; this may account for their
being frequently overlooked.

260 H. M. WOODCOCK.

bably, not limited by any distinct cuticle or membrane.) 'lhe
cytoplasm is finely granular and fairly uniform in character
(Rogers, Laveran and Mesnil); in some of Donovan’s
figures there is a faintly stained or clear area of varying
size, more or less centrally situated, which possibly repre-
sents a vacuole.” The most interesting point in the
morphology of these bodies is the fact that two chromatic
masses, of very unequal size, are invariably to be recognised,
except in very young forms. ‘The larger nuclear mass,
which, it may be at once said, is in all likelihood homologous
with the trophonucleus of a Trypanosome, is usually round
or oval,’ and varies in position; in the pyriform parasites it
is generally situated near the basal end, and in the oval ones
about the middle of one side. The smaller nuclear body
(representing, probably, a kinetonucleus) stains more in-
tensely, and takes the form either of a little rod (sometimes
curved) or of a round dot or grain. It is generally quite
separate from the larger nucleus. In the round or oval
parasites if is on the opposite side of the body and at the
periphery ; in the pear-shaped forms it is about the middle
of the length or nearer the narrow end (fig. 37 I). In some

1 Ross (140, 141) considers that the free forms, at any rate, possess a
distinct and resistant cell-wall or cuticle, comparable to a spore membrane.
Christophers at first thought so too, but in his later Report appears to be
doubtful, in view of the non-resistance of the bodies and their rapid break
up after the death of the host.

? Both Christophers and Wright represent the bodies as having only a
narrow peripheral border of stainable cytoplasm, all the central part
remaining unstained, and constituting (according to Christophers) one or
two huge vacuoles. These authors have probably been misled by deceptive
appearances due to staining peculiarities, which have, unfortunately, led
them to an incorrect interpretation of the bodies (see below, p. 263).

* In some cases it is heart-shaped or bilobed, probably indicative of
approaching division.

4 These different forms and appearances are explainable, as Christophers
points out, by regarding the parasite as viewed from different aspects,
e.g. end-on or sideways. This may, very likely, often be the case, the
typical form being that of a slightly flattened pear. There is no necessity,
however, to consider the body as having a fixed and unchangeable shape,

THE HEMOFLAGELLATES. 261

cases, however, the smaller nucleus is in contact with, or
attached by a delicate thread to, the larger one.

The parasites multiply in two ways—(a) by binary fission,
and (b) by multiple division or segmentation. The principal
stages in the first mode are well known—at least, their
general outlines—and certainly offer strong resemblances to
the process in Piroplasma. In the pyriform parasites the
division is evidently longitudinal (fig. 37 16), and, all things
considered, it is most probable that binary fission usually
takes place in the long axis (compare the figures of Rogers,
Laveran and Mesnil, and Donovan), the apparently transverse
division of oval forms being due to their being seen more or
less end-on. The large nucleus becomes bilobed and finally
constricted into two; the smaller nucleus becomes elongated
transversely, and also cut into two halves; lastly, the cyto-
plasm splits up, the cleavage furrow commencing either,
usually, at the broad or basal end, or, occasionally, at the
narrow, pointed extremity. Daughter-individuals, which
have evidently just separated, are seen lying side by side in
the figure.

The other form of multiplication—multiple division—is
probably largely responsible for the numbers of parasites
with which the host-cells are often packed and distended.
This process, however, has not yet been so satisfactorily
made out as that of binary fission. It appears to conform
more or less to the radial or rosette type, enlarged, rounded
parasites, with a varying number of nuclei (up to about
eight), equally arranged near the periphery, having been
frequently noticed (fig. 37 Ic). Different writers, however,
describe and figure this nuclear division somewhat differently.
While, according to Laveran and Mesnil and Donovan, the
nuclei are all of one size (the two kinds of nuclear element
having apparently united), according to Christophers and

as would be the case if it were rigidly limited by a spore membrane;
delicate, endocellular parasites are usually capable of change in form to a
certain extent, and the rod-like or bacillary form well instances such a
change here. Compare also other Hzemosporidia.

262 H. M. WOODCOCK.

Rogers the larger and smaller chromatic elements (in other
words, the trophic and kinetic parts) remain distinct and
divide up independently (fig. 37IVb). Laveran and Mesnil
consider that the very young forms observed, with only one
nuclear body, result from the segmentation up of such a
stage with several similar nuclei. Rogers (I. c.) has followed
the process in parasites kept in citrated blood, maintained
at a temperature considerably below blood heat. Fig. 371Ve
represents, according to this author, practically the end
stage in the multiple division of a parasite, a group of young
daughter forms, each with two chromatic masses, being seen
embedded in a faintly-staining ground substance, which,
there can be little doubt, is of the same nature as the
zoogloeal matrix above discussed.! Further investigation is
necessary to ascertain the particular circumstances which
bring about the occurrence of these two varieties—if they
are really distinct.

The above is a summary of all that was known concerning the morpho-
logy and development of this parasite prior to the quite recent announce-
ments of Rogers, and many and various have been the views expressed with
regard to its nature. Leishman, the discoverer of the splenic form, at

1 Rogers considers that this ground substance is derived from the organ-
ism itself, and consists of residual cytoplasm not used up in forming the
daughter individuals. In other words, the author, to judge from his
description and figures, regards the daughter parasites as being formed
within the body of the parent; i.e. this is an instance of endogenous
multiplication. The writer does not agree with this conclusion, as all our
knowledge, both of Trypanosomes and Hemosporidia—as well as of other
Sporozoa derived from a Flagellate ancestor—goes to show that multiple
division is uniformly exogenous or peripheral, and quite distinct in
character from the endogenous type (cf. the preceding pages and see
Minchin, 1. c.). Rogers figures another similar group of young forms—
still clustered together—in fresh blood from the spleen; in these there is
no sign of residual matrix. It is much more probable that Christophers’
explanation holds good here also. Rogers mentions that the parasites (in
the cultures) are in a slimy or zooglwa-like matrix, which develops in the
course of a day or two, and this is doubtless due to the alteration and
breakdown of splenic cells, blood corpuscles, ete., in which the parasites are
developing.

THE H#®MOFLAGELLATES. 263

first considered the organisms as representing involution or degeneration
appearances of Trypanosomes, being largely influenced by the two unequal
chromatin masses; in this view he has been supported by Marchand and
Ledingham. Later, Leishman has gone somewhat farther, and regards
the parasites as perhaps representing an actual stage in the life-cycle of
a Trypanosome. Laveran and Mesnil, taking more particularly into account
the general form and very suggestive binary fission, considered the parasite
to be a new species of Piroplasma, which they called P. donovani;
donovani is, therefore, the correct specific name of this form. Donovan
concurred in this view, and Mesnil, Mouton and Remlinger (1. ¢.), who have
studied Wright’s form ina case of “bouton d’Alep” also consider this as
a Piroplasma,—probably, however, a distinct species.!

Other authorities (e.g. Christophers, Ross and Wright) have gone some-
what wide of the mark, and have seen in this parasite an entirely different
kind of Sporozoan; or, rather, they (with the exception of Wright) have
regarded the parasitic bodies as being, themselves, only the spores of a new
Myxosporidian, the parent body or piasma of which has not yet been satis-
factorily made out.2 This view is at once put out of court by the fact
that Sporozoan spores never divide up in the way these bodies admittedly
do; a valve or lid may open, liberating enclosed germs, after which the
spore-case is cast empty aside, but the whole spore, membrane and all, can-

not possibly divide up into two or more “ daughter-spores ! ”

The above description of the parasites sufficiently justifies,
we think, Laveran and Mesnil’s opinion that they agree
closely enough with the known stages of Piroplasma to
be considered as belonging to that type of organism.? This
does not, of course, prejudice, in any way, the view that
these parasites represent, nevertheless, only a phase or part
of a complete life-cycle. Now, as mentioned above, there
is evidence that P. bigeminum is closely associated with

1 The specific name, in that case, will be tropicum, as Wright termed
his form Helcosoma tropicum.

? Ross, who has created the genus Leishmania for the parasites, thinks
the matrices above discussed represent “ relics of the parent organism.”

3 Since Schaudinn has observed nuclear dimorphism in a typical Piro-
plasma (namely, P. canis, see above, p. 254), any objection based upon
this feature of P. donovani is removed. The only difference of any
importanee, in fact, appears to be that of habitat, and, granting this (though
the new parasites do not appear to be, by any means, exclusively leucocytic),
at least one Hemosporidian, namely ““Hemameba” (Trypanosoma)
zieman ni, also has a leucocytic habitat.

264 H. M. WOODCOCK.

a Trypanosome, and the connection between these new Piro-
plasma-like forms and a Trypanosome—or, at any rate, a
Heemoflagellate—has been definitely established by the work
of Rogers, to which allusion has several times been made.

The parasites with which Rogers experimented were splenic
forms taken from cases of cachexial fever and Kala-azar.
As the author himself admits, the artificial conditions in which
the organisms were cultivated cannot be supposed to have
been as favourable to their further development as the natural
conditions, whatever these may be, which bring about the
same changes. Hence there must be, for the present, more
or less uncertainty as to how far the forms described accur-
ately represent typical evolutive stages of the parasites. Some
of Rogers’ figures certainly suggest the idea that the parasites
were unhappy at the time he portrayed them.

However, the great fact remains, that what were unmis-
takably Flagellate forms developed in the cultures at different
intervals. Fig. 37 IV shows two pear-shaped forms, lying
side-by-side after binary division, one of which has developed
a flagellum near one end. This was in a culture of the third
day. Another pair of pyriform stages with longer flagella,
belonging to a fourth day culture, is seen at IV e. The
most convincing stages (IV g) developed suddenly in a
one-day culture from another patient. Rogers accounts for
this by the condition of the blood being less altered than
after three or four days’ incubation. Probably, also, the
organisms, when they left the host, were in a more favourable
condition or phase for further development than in the other
cases. In nearly all instances, the flagellum originated from
that side or end of the body, near which the smaller nucleus

1 Among these are one or two rather indefinite forms which Rogers con-
siders represent stages in fusion (comparable to conjugation) ; earlier stages
in the process are, the author thinks, exemplified by many of the pairs or
couples (e.g. those in fig. 37 IV d). The writer does not consider this
very likely ; the couples much more probably represent individuals which
have not yet separated after division (cf. the figs. of Donovan, and Laveran

and Mesnil), the somewhat atypical form and arrangement being readily
accounted for by the medium in which the parasites are.

THE HAMOFLAGELLATES. 265

(kinetonuclear element) was situated, and the author mentions
that he was occasionally able to trace a connection between
the two. From the figures it certainly appears as if the para-
sites by successive divisions, became more fusiform and less
pear-shaped, that seen at g, being, perhaps, derived from
a form like that of f, which is in the act of dividing.
Even in the most slender and Trypanosome-like stage ob-
served, however, Rogers could not distinguish any indications
of an undulating membrane, and the kinetonucleus was never
far from the insertion of the flagellum.

These results have since been fully corroborated by Chatterjee (122),
Christophers (123 [8rd Rep.]), and Leishman and Statham (181). The
general appearance of the Flagellate stages figured by these workers quite
agrees with that seen in fig. 37 [V, f. and g. Leishman and Statham bring
forward interesting additional observations, and the illustrations given are
particularly good. The cytoplasm of the parasites is usually very vacuolated ;
this is most probably due to the effects of the artificial medium upon the
metabolism. Leishman and Statham (and also Christophers) describe the
actual formation of the flagellum, which is developed very suddenly, in a
remarkable manner, from a distinctive, vacuole-like structure, termed the
* flagellar vacuole”’; this arises at the anterior (?) end, in close connection
with the kinetonucleus (‘‘ micronucleus”’), Some of the contents of this
vacuole are expelled to the exterior in the form of a tuft or branched
process, and, at the same time, the flagellum appears. At present it seems
impossible to say exactly what occurs. Another remarkable process de-
scribed is unequal longitudinal fission. Very thin, sickle-like (“ spirillar”’)
portions of the body are split off from one side of the parent-individual.
More than one thread-like form may be thus separated off. The strange
feature about the process is that neither of the two principal nuclear
elements appears to be concerned. In a few of the parasites chromatin
grains are noticeable in the cytoplasm, and in one of the peculiar fission-
forms two of these grains are contained in the portion being cut off.
Whether these would become the definitive nuclear organelle of the daughter-
parasite is not certain. Anyhow, later on, when the sickle-like form has
developed a flagellum, two chromatic elements are present, apparently
corresponding to those in the ordinary (adult) forms.

All these accounts agree with that of Rogers with regard
to the entire absence of an undulating-membrane.
Nevertheless, bearing in mind the fact that cultural forms
of many Trypanosomes have either a very smal! membrane
VOL. 00, PART 2,—NEW SERIES. 19

266 H. M. WOODCOCK.

or none at all (cf. above), the possibility of these organisms
possessing an undulating membrane during certain phases of
the life-history, when this is undergone in normal conditions,
is by no means excluded. On the contrary, indeed, the
markedly Herpetomonad-like facies of the parasites, which
greatly resembles that of ‘ cultivated” Hemoflagellates,
strongly points to their being closely related to this group.
If so, there can be little doubt that Piroplasma (Leish-
mania) donovani (and, by inference, P. (L.) tropicum)
also has, at some period or other, a typical trypaniform phase.

Nothing is known with regard to the transmission of the
parasites, and the possible occurrence of an alternate host.
The superficial position of the localised form strongly points
to infection in this case by the bite of some blood-sucking
Insect.!. That being so, it seems most natural to infer that
the same holds for the splenic type, although what determines
the limitation of the parasites in the one case, and their
dissemination throughout the system in the other, remains a
mystery. It has been suggested that, in the case of the
latter type, the organisms leave the host by way of the
alimentary canal, since they have been found in ulcers of
the large intestine. At any rate, it is very likely that an
important part of the life-cycle is passed through outside
the human host, though whether in the free condition or in
an Insectan host, or in both combined, has still to be learnt.”

1 Crombie (‘ Brit. Med. Journ.,’ 1904, ii, p. 658) points out that persons
who attend upon camels are very liable to a form of oriental sore. The
camel is the host of a Trypanosome and, possibly, the “ camel-fly ” transmits
the parasites to man. There is, indeed, a close parallelism between the
distribution of camels and oriental sore.

2 Rogers’ experiments showing that the parasites live and develop in
cold solutions outside the body, but rapidly degenerate when kept at blood
heat, point to the first part of this sentence being correct. With regard to
the second part, this same worker, in a more recent note (189), finds that
cultures of the parasites (Leishman’s form) develop most rapidly and
successfully in an acid medium. In his opinion, this indicates the acid-
containing contents of the stomach of some blood-sucking Insect as the
place in which the extra-corporeal stages of the parasite’s existence are

THE HEMOFLAGELLATES. 267

Section XI. Puynogeny anp Evo.vurion.

The subject of the derivation of the Trypanosomes is one
of much difficulty, owing to our very insufficient knowledge
of the majority of the parasites. The views put forward below,
which involve an apparently different interpretation of the
orientation of the body in different cases, are, to a large
extent, based upon Léger’s important researches, on the one
hand, upon Trypanoplasma (64 and 65), and, on the other,
upon certain Herpetomonadine forms (61—63, 68, 69). ‘The
Trypanosomes, as a whole, are to be regarded as including two
entirely distinct families, in one of which the attached flagellum
becomes free at the true anterior end, and in the other at the
true posterior end. Before discussing the reasons for this
division of the Hemoflagellates into two groups, however,
we may consider the principal features in the structure of the
Herpetomonadine parasites to which reference has just been
made.

In a typical Herpetomonas (e.g., H. musce-
domestice,! H. jaculum, or H. gracilis [fig. 38 B]), the
kinetonucleus is situated near the anterior end; the flagellum
is not attached to the side of the body at all but straightway

undergone ; and, in this connection, he is inclined to suspect fleas or bugs.
The difficulty in the way of this view is the rarity of the occurrence of the
organisms in the peripheral circulation; the skin papule which sometimes
occur are suggested as furnishing a source of infection for the Insectan host.

1H. musce-domestice is here included as a typical uniflagellate
Herpetomonad. Léger has observed no signs of two flagella, either
in this species or in others of the genus (not considering, of course, indi-
viduals about to divide by longitudinal fission). Prowazek (87) inter-
preted this form as a bipolar Flagellate in which the body has been bent
up so that the two ends have come together and united, the flagella alone
remaining distinct. This view is the less tenable since Schaudinn’s concep-
tion of a primitive, bipolar ‘“‘ Urhemoflagellate,” based upon Laveran and
Mesnil’s unfortunate description of Trypanoplasma borreli, appears
to have no prospect of being realised; for, up to the present, there is no
evidence of this bi-polarity in any known Flagellate.

268 H. M. WOODCOCK.

becomes free, and, correlated with this, there is no undulating
membrane. ‘These forms are, for the most part, parasites of
Insects which do not suck blood. A stage in advance is seen
in H. subulata (fig. 38 5, F), parasitic in the digestive tube
of Tabanus glaucopis and Hematopota italica, which
are predatory on cattle and horses. This parasite, when in
the Monadine form, has still the usual acicular shape. The
kinetonucleus, however, lies much farther from the anterior

Fic. 38.—a, c, Herpetomonas (Crithidia) minuta,
Léger; ov, attached (gregariniform) stages of same; B, H.
gracilis, Lég.; £, r, H. subulata, Lég.; G, attached stages of
same. (All after Léger, x 1800.)

end, and may, in fact, be almost opposite the trophonucleus,
The flagellum, which has been, as it were, drawn back with
it, is, in the majority of individuals, attached, for the proximal
part of its length, to the anterior part of the body by means
of a delicate cytoplasmic border, which constitutes a rudi-
mentary undulating membrane. ‘The flagellum, it will be seen,
has its root in a diplosome (probably of centrosomic nature)
just in front of the kinetonucleus.

H. (Crithidia) minuta, Léger, parasitic in Tabanus
tergestinus, differs from the last-mentioned type in having
the posterior end thicker and more rounded (fig. 36 a and c);

THE H#MOFLAGELLATES. 269

it is intermediate, in short, between the acicular forms and the
genus Crithidia, characterised by its pyriform shape. ‘The
kinetonucleus may apparently be either near the anterior end
or near the trophonucleus; but L. does not mention the
occurrence, in the latter case, of any rudimentary membrane.
We come next to Crithidia fasciculata, which Léger
found in the intestine of Anopheles maculipennis
(females). A striking resemblance in form is offered by
certain phases of this parasite’ to those of Trypanomorpha
noctuz when in the gnat. One side of the body appears
more delicate than the other, possesses a distinctly wavy
border, and is prolonged anteriorly, attached to the flagellum,
gradually tapering away however before the latter ter-
minates.

There are one or two other important morphological points
to note in connection with these Herpetomonadine forms. In
many a vacuole can be easily demonstrated situated in the
cytoplasm near the anterior flagellate end. In H. muscea-
domestice and H. gracilis (fig. 36 B) this structure is con-
tractile or pulsatile, but in H. subulata (fig. 36 £), a more
specialised type, it appears constant (non-contractile), though
probably still retaining its original excretory function. <A
feature in H. gracilis is the occurrence of a number of
deeply-staining grains in the cytoplasm, doubtless comparable
to the chromatoid grains described in many Trypanosomes.
In this parasite they are all in the non-flagellate part of
the body behind the nucleus.

Lastly, it is to be remembered (1) that all the parasites
mentioned have also gregariniform resting-phases, in which
the locomotor apparatus may become reduced to a short
rostrum serving for attachment to a host-cell, or may entirely
vanish (see fig. 36 Dand G); and (2) that H. subulata, as
well as certain other species, becomes united in rosettes in
which the parasites are all attached by the anterior end.

From these considerations it appears scarcely open to doubt
that an intimate degree of relationship exists between

1 Unfortunately the only figures at present available are not very good.

270 H. M. WOODCOCK.

Try panomorpha (Trypanosoma) noctue# and Herpeto-
monas. Inall known respects thereis either entire agreement
between the two genera, or the condition found in the former
is easily derivable from that existing in the latter. Certain
species of Herpetomonas may well represent a phylogene-
tic stage in the evolutionof Try panomorpha; indeed, in some
cases—especially where they have been described from blood-
sucking Insects—parasites identified as Herpetomonads pro-
bably themselves constitute, in reality, only a phase in the
life-cycle of a particular Heemoflagellate. As Léger points
out, Crithidia fasciculata itself is very likely a Hemo-
flagellate, and closely allied to Trypanomorpha, though
exactly how closely cannot be said until its life-history in the
Vertebrate host has been investigated.

Many, perhaps most, authorities regard all uniflagellate
‘'rypanosomes as descended in this manner, and as agreeing
with I'rypanomorpha in having the flagellum at the
anterior pole of the body. ‘here are, however, in the writer’s
opinion, several cogent reasons against accepting this view,
reasons which, on the contrary, lend considerable support to
the alternative theory that, at any rate, most ‘l'rypanosomes
(sensu stricto) are derived from a type like Trypano-
plasma by the loss of the anterior free flagellum, so that the
non-flagellate end is really the anterior one.’ ‘I'hese reasons
are best discussed under two headings, dealing, respectively
(a) with biological or physiological considerations, and (b)
with certain morphological points. ‘l'o the former particular
weight must be assigned, and the indications which they
afford are borne out and, in fact, strengthened by a considera-
tion of the latter.

(a) It may, we think, be fairly assumed that the sensory
extremity of a Hemoflagellate is, as in non-parasitic Ilagel-
lates, the anterior end. By the sensory extremity is here
meant that region of the body which is most in touch with
the environment; the region specially concerned with the
adjustment of the voluntary relations of the parasite to the

This view was first outlined by Léger (65).

THE HASMOFLAGELLATES. 21

immediate surroundings and conditions. There can be no
doubt that such is the case in those forms for which the cor-
rect orientation of the body is known (cf. on the one hand
Trypanomorpha, and on the other the delicate tactile and
sensory beak of ''rypanoplasma and the allied Trypano-
phis), and there is no reason to suppose that the same is not
true for the numerous uniflagellate forms. In other words,
we may regard the sensitive or physiological end of the
body as homologous throughout the group.

Among the biological processes in which the sensitive ex-
tremity may be expected to take the lead are the following :

(1) Attachment.—Schaudinn’s discovery that Try pano-
soma (“Hemamceeba’’) ziemanni, when about to enter
upon a resting phase, attaches itself to and penetrates a
leucocyte by its non-flagellate extremity, is of the greatest
importance in this connection. ‘This manner of fixation is in
sharp contrast to that met with in Trypanomorpha and
other Herpetomonadine forms. Although the mode of
attachment in T'rypanoplasma is not yet known, there can
be no doubt that, if it occurs, fixation takes place by the
anterior end, as Keysselitz (l.c.) has found is the case in
Trypanophis, the short active beak or rostrum of which
itself probably serves this purpose. In the case of most 'Try-
panosomes attachment has not, so far, been described. ‘There
are, however, one or two observations on record with regard
to certain species of Trypanosoma, which tend to show that
fixation occurs in these also by the non-flagellate end (see
above, p. 173).

(2) Agglomeration.—Whatever its exact meaning, it 1s
most likely that this peculiar phenomenon is a_ biological
feature of definite import. ‘This seems indicated, for instance,
by the fact that—in what may be regarded as normal cir-
cumstances, if not, indeed, under all conditions '—the union,
im any particular case, occurs in a constant manner. More-
over, as already described, the mode of agglomeration and
that of attachment differ in exactly the same way in

See Section VII (B), with reference to “ rosette-formation ” in cultures,

My be H. M. WOODCOCK.

Trypanomorpha noctue and Trypanosoma ziemanni.
Hence, it may be taken for granted that agglomeration
normally occurs by the sensitive pole. It cannot but be
regarded as highly significant, therefore, that, in all species
of Trypanosoma for which this property has been ascer-
tained, the union is by the non-flagellate end; in the
writer’s opinion, this evidence to a large extent compensates
for the paucity of observations relative to attachment. Up
till now agglomeration has not been observed in Amphibian
or Piscine forms.

(3) Conjugation.—It is difficult to imagine a micro-
gamete, which exhibits definite cell-polarity, coming into
contact with the female element, prior to union or penetra-
tion, by any other than the sensitive, anterior pole. Hence,
in the case of the Heemoflagellates, the end by which a
microgamete (a male Trypanosome or its homologue, it is to
be remembered) conjugates with a megagamete is of great
interest. Unfortunately, very few instances of fertilisation
are, so far, to hand. From Schaudinn’s account, the process
in ‘'rypanomorpha noctue fully bears out the evidence
afforded by other considerations, and the same is probably
true of Trypanosoma ziemanni’; a noteworthy point in
the former case is that the flagellum of the microgamete,
being at the anterior, penetrative end, is not developed.
The only other case known is that of I’. lewisi (Prowazek
[l.c.]), to which reference has been made. Granting,
as is very probable, that this author’s description and
figures (see above, p. 249) represent the manner in which
true conjugation occurs, the question of the correct orienta-
tion of the body in this form appears conclusively settled in
favour of the view here taken.

(4) Movements of Investigation.—In this category
are to be placed the slow creeping or crawling movements
which have been observed in certain ‘Trypanosomes (see
Section VII (a) ). Both the apparently tentative or inves-
tigatory character of the movement, and the fact that it is

1 See footnote, p. 247.

THE HH MOFLAGELLATES. 278

essentially comparable in manner and origin to gregarioid
or euglenoid movement, would naturally lead us to expect
that the sensitive or anterior extremity goes first. _The fact,
therefore, that this mode of progression is with the non-
flagellate end in front in all the instances hitherto recorded, is
of considerable importance in the present argument. It
entirely negatives—in the writer’s opinion it distinctly out-
weighs—the opposing fact that the movement during rapid
displacement usually occurs with the flagellate end leading.
For, in these energetic movements, where the principal object
ot the parasite is to change its vicinity as quickly as may be,
it obviously cannot much matter which end goes first, the
determining factor being, doubtless, mechanical considera-
tions (cf. also the account of movement).

As a corollary may be mentioned the plastic or “‘amceboid ”
nature of the non-flagellate end in ‘l’rypanosoma—1.e. its
power of retraction and extension, commented upon by
Laveran and Mesnil (cf. fig. 42, of T. equiperdum). In
short, all the physiological data we possess tend to show that
this non-flagellate end in Trypanosoma is the sensitive ex-
tremity, and homologous with the anterior end of Herpeto-
monas, Trypanomorpha, and 'rypanoplasma.

(6) Morphological Points.—Reference has above been
made to the frequent presence of a vacuole near the non-
flagellate end in different species of Trypanosoma (e.g. T.
brucil, T. gambiense, Hanna’s Try panosoma of Indian
birds, etc.), here considered to be a normal cell-constituent
(though not necessarily one of invariable occurrence), which
probably represents the (originally) contractile vacuole of
an ancestral Hemoflagellate. or there is no reason to
doubt that this vacuole is homologous with that described
both in Try panoplasma (fig. 17 F) and in certain Herpeto-
monadine forms.! The important point of distinction is that

1 In Trypanomorpha, it will be remembered, Schaudinn describes
one or more cytoplasmic vacuoles of varying size, in the anterior part
of the body (cf. fig. 8) ; this perhaps represents a more advanced condition,
in which the formation of a distinctive organella is tending to be lost.

274. H. M. WOODCOCK.

in the last-named instances this structure is near the anterior,
flagellate extremity of the cell. If, therefore, Trypano-
soma sensu stricto is a Herpetomonadine form, why is
the vacuole, where known to occur, always near the non-
flagellate, apparently opposite end of the body ?

(1) Conversely the chromatoid grains, so frequently notice-
able in the cytoplasm (see above, p. 211), are, in Trypano-
soma,! mainly or entirely in the flagellate half of the body,
or, at any rate, on that side of the nucleus (figs. 16,17, 42, 47,
and 57). On the other hand, in Herpetomonas gracilis,
which also possesses a number of similar grains, they are all
in the non-flagellate part of the body, that is to say, behind
the nucleus (fig.58 8). Compare the cytological details
of this parasite with those of, say, I. nelspruitense
(fig. 178), and it appears unquestionable that the
flagellate end of the former corresponds to the non-
flagellate oneofthe latter. Again, in 'l'rypanomorpha,
pigment grains and other effete material collect near the non-
flagellate (posterior) end, and the cytoplasm near this ex-
tremity is frequently more granular and deeply staining (figs.
8, 10, and 13) than it is near the flagellate end, or just the
opposite to what is generally the case in Try panosoma,

It is surely needless to suppose the position of these two
cytological constituents in the body of Trypanosoma to
have become, for no obvious reason, absolutely reversed
(which one is bound to do if the Herpetomonadine view
be adhered to), when, on the one hand, no stage from
Herpetomonas to T'rypanomorpha inclusive shows any
indication of such a change, and, on the other hand, in
Trypanoplasma borreli, they have exactly the same
position relative to the attached (posterior) flagel-
lum that they have in Trypanosoma (cf. fig. 17 F and 4,
and especially Laveran and Mesnil’s amended Fig. LVI, 5,
[56], in which the chromatoid grains are entirely behind the
nucleus). Granted the loss of the anterior flagellum, and we

' In T. granulosum (fig. 17 K) they are uniformly distributed through-
out the cytoplasm, but this is, apparently, the only exception.

THE HE MOFLAGELLATES. 975

have a body which is in complete agreement with that of Try -
panosoma.

3. Lastly, what undoubtedly seems to be a transitional
stage in this disappearance of the anterior flagellum is seen
in Trypanoplasma cy prini (fig. 17), where this organella
is short and delicate, and evidently undergoing reduction.

While the evidence thus adduced points, in our opinion,
unmistakably to the derivation of at least some of the forms
included in the genus Trypanosoma from a _ Hetero-
mastigine ancestor, there is, it may at once be admitted,
one morphological feature which does, at first sight, seem to
oppose this theory. ‘The character in question is the varia-
tion in position of the kinetonucleus, especially during certain
developmental phases.

In most species of Trypanosoma, the kinetonucleus
occupies, in adult parasites, a position near the non-flagellate
end. Its position here, therefore, compares with that in the
simple Herpetomonadine forms (e. g. H. gracilis, jaculum,
and musce-domestice) and in Trypanoplasma in
exactly the same way as does that of the other cell-con-
stituents referred to above (where noticed) in the two sets of
cases.

Were the approximately terminal position of this organella
constant, therefore, this character would be in complete
agreement with the other morphological ones, and would be,
undoubtedly, a very strong point in favour of the view
here taken. As it is, however, the kinetonucleus varies
considerably with respect to its situation in the body. On
the one hand, in Herpetomonas subulata, some indivi-
duals of H. minuta and Trypanomorpha noctua,
it has passed backwards almost as far as, or sometimes
even behind (posterior to), the trophonucleus; and, on
the other hand, in certain species of Trypanosoma
(e.g. I. inopinatum, T. rotatorium, T. transvaaliense)
it occupies a similar position on one side or other of the
trophonucleus, in more or less contiguity to the same.
Hence, having regard to these facts alone, it might reason-

276 H. M. WOODCOCK.

ably be inferred that, in the latter forms also, the result is
attained in the same manner, by the passage backwards of
the kinetonucleus from the flagellate (anterior) end; and that
in the other species of Trypanosoma, in which the kineto-
nucleus is practically terminal, this body has merely passed
still further back, right to the non-flagellate (posterior)
end,! thus bringing about the great development of the
undulating membrane.

This is, in fact, the explanation given by the adherents of
the Herpetomonadine theory. ‘They consider that the con-
dition observed in young parasites of 'T. lewisi, formed by
“budding” or multiple division (see above, p. 238 et seq.,
and, figs. 27—29), furnishes strong support to this view;
and, indeed, they look upon the manner in which the young
forms become adult in this parasite as indicating the actual
course of morphological evolution in Trypanosoma. Did
these facts constitute the only evidence available, there could
be, it is needless to say, little doubt as to the correctness of
the monophyletic theory. When, however, we bear in mind
the many opposing facts already discussed, including certain
biological considerations of fundamental importance which it
is difficult, if not impossible, to reconcile with this view, we
are bound to conclude that the migration of the kineto-
nucleus cannot possess the significance attributed to it.

There is one point which may be suggested as perhaps
accounting, at any rate in part, for this feature. It may be
regarded as proved beyond question that the kinetonucleus
is of nuclear origin. Hence its wandering propensity in
general and frequent contiguity to the trophonucleus are not
difficult to understand. ‘This association of the two organelle
is particularly marked in cases of multiple division, whether
occurring in normal (figs. 27, 28) or in cultural conditions

1 Admitting, for the moment, that this explanation of the position of the
kinetonucleus in Trypanosoma is possible, it must be emphasized that the
argument relative to the other morphological characters is not thereby
invalidated; for no intermediate stages in the “reversal” of their position
are known (cf. H. subulata, and especially Try panomorpha).

THE HHMOFLAGELLATES. 277

(fig. 29). It is not found, on the contrary, in binary longi-
tudinal fission,! which, it must be remembered, is by far the
most prevalent and most typical mode of division in natural
circumstances. In such cases there is no indication whatever
of a Herpetomonadine facies (cf. figs. 25, 25). Hence it
seems to the writer that this association of the nuclear
elements bears some relation to their rapid multiple division
during segmentation, perhaps facilitating the complex me-
chanical processes which apparently occur.

Now the juxtaposition of the two nuclear bodies of itself
confers a semi-Herpetomonadine aspect on an individual ex-
hibiting it. When further, as frequently in this rosette type
of multiplication, the daughter-nuclei travel towards the peri-
phery, doubtless to assist in the segregation of the general
cytoplasm into the individual portions,” the resemblance to a
Herpetomonad is much increased. We have, in short, what
may be termed a “‘ pseudo-Herpetomonadine ” form developed
(cf. figs. 28, 29). The absence or rudimentary character
of the undulating membrane is sufficiently explained by the
position of the kinetonucleus and the very short portion (if
any) of attached flagellum present in consequence, and need
not be further considered.

To sum up, therefore, this difficulty with regard to the
variation in position of the kinetonucleus does not seem in-
superable; on the other hand the biological considerations,
as well as the other morphological characters above men-
tioned appear to offer far greater objections in the way of
the Herpetomonadine theory. Nevertheless, this does not
necessarily imply that all the species included below under
Trypanosoma are derived from a T'rypanoplasmatine an-
cestor. Some almost certainly are, e.g. T. brucii and T.

1 This refers, of course, to adult Trypanosomes as such ; parasites exhibit-
ing the ‘ pseudo-Herpetomonadine ” form may divide while retaining it
(see above, p. 240).

? With reference to the arrangement of the parasites in the multiplica-
tion-rosettes found in cultures, which cannot be regarded as entirely normal
see above, p. 241.

278 H. M. WOODCOCK.

lewisi among Mammalian forms, T. ziemanni among Avian
parasites, and 'T’. nelspruitense among Amphibian types.
And the same is probably true of the majority of the Piscine
forms; at any rate, their morphology—the only evidence
available at present—agrees more closely with that of Mam-
malian parasites than with that of Trypanomorpha. On
the other hand, in the case of one or two Amphibian forms,
it is much more doubtful which interpretation of their
morphology is the correct one. One of these, most unfor-
tunately, happens to be Trypanosoma rotatorium, the
type-species of this genus. It can at least be said, however
that this parasite is quite as probably a Trypanoplasmatine
form as a Monadine one (see below, p. 288).

Evolution.

In conclusion, one may, perhaps, venture to indicate what
appears to have been the general course followed in the
evolution of the Hzmatozoa, although the tentative cha-
racter of any hypothesis which can be at present advanced
must be fully recognised.

The Hemoflagellates are here considered to be, in general,
descended from forms originally parasitic only in Invertebrate
hosts (probably non-predatory), in which, as intestinal or
entero-ccelomic parasites, all their life-cycle was undergone.
Upon their becoming associated with a blood-sucking host,
either by the acquirement by the original one of this mode of
life, or by the adaptation of a particular parasite to a closely
allied predatory host, the Flagellates gradually became
specialised for life in the blood. In other words the Inver-
tebrate is here regarded as the primary host, and the
Vertebrate as the secondary or intermediate one.

There can be no doubt that this is the case in Try pano-
morpha, at any rate. In this form the development of the
characteristic undulating membrane appears to have been
brought about by this change in habitat. Transitional stages
are beautifully illustrated by Léger’s observations (1.c.) on Her-

THE HA@MOFLAGELLATES. 279

petomonadine forms, of which mention has above been made.
Moreover, in this parasite the primary host is also the defini-
tive one, or that in which the sexual process takes place.

Turning next to the Heteromastigine forms we are led to
the same conclusion, though on more general grounds, for we
lack, at present, an illustrative series like that just referred
to. For instance, with regard to the origin of the undulating
membrane. In Trypanoplasma, which is most probably
to be derived from a Bodo-like ancestor, the undulating
membrane very likely arose, as Doflein (19) points out, as
the result of the contiguity of the trailing flagellum to the
side of the body, which led to fusion without loss of motility
and thus brought about the development of a membranous
expansion. Now there are one or two forms which possess
an undulating membrane, but which are not, however (so
far as is known), Hematozoan parasites, namely, Try-
panophis grobbeni, which inhabits the ccelenteric cavity
of Siphonophores, and an enteric form parasite in Box
boops, which Léger (70) regards as a Trypanoplasma
(intestinalis), but which appears to the writer, if anything,
nearer T'rypanophis (see below, in Systematic). Hence it
would seem that the presence of an undulating membrane in
a Trypanoplasmatine form does not necessarily imply a hemal
habitat.

In connection with this derivation of the Heteromastigine
forms it is interesting to note the gradual change in the
degree of development of the flagella which can be traced.
Starting with Bodo lacertx, both flagella are of equal
length, and the trailing one does not reach the posterior
limit of the body. In Trypanophis grobbeni (fig. 41) the
posterior flagellum is more developed than the anterior one,
and, of course, attached to the side of the body. Its free ter-
mination, however, is very short. In T, (Try panoplasma)
intestinalis this is longer and more strongly developed
(fig. 40), though still shorter than the anterior flagellum itself.
In Trypanoplasma borreli (fig. 17 8, G) the anterior
flagellum and the free portion of the hinder one are of equal

280 H. M. WOODCOCK.

length. In T. cyprini (fig. 17 4) the former is much shorter
than the latter, and shows signs of reduction, while in Try-
panosoma it is entirely lost.

Hence, while it is very probable that the condition of the
locomotor apparatus in ‘'l'rypanophis represents a phylo-
genetic stage in the development of the condition found in
Trypanoplasma, this form itself is not of much assistance in
determining the original kind of host of the Heteromastigine
Heemoflagellates.

The reasons for thinking that this section of the Hale
flagellates is also derived from a primitively Invertebrate
type are, as a matter of fact, the considerations fully dis-
cussed in Section IX, which point, on the one hand, to the
possession by the parasites of an alternate, Invertebrate host,
in which conjugation takes place, and, on the other hand, to
a relationship between the Trypanosomes and the Hemo-
sporidia. For, on a priori grounds, we may reasonably
infer that the definitive host is the primary, original one.
Thus would be explained the necessity which exists in most,
if not all, Hematozoa for the return of the parasites to the
alternate host. Again, both sections of Hemosporidia, those
of warm- and of cold-blooded Vertebrates, have recently been
brought into line in respect of the occurrence of conjugation
only in the Invertebrate (see Introduction, p. 161). It follows,
therefore, that all evidence pointing to a connection between
the two “groups” of Hzmatozoa at the same time favours
the view here put forward.

While, of course, the possibility of a Heematozoan being
derived from a form primarily parasitic in the intestine of a
Vertebrate must not be ignored, the writer agrees, never-
theless, with Léger, who aptly remarks that one is not
justified in so deriving any Hemoflagellate until a ccelomic
form, known with certainty to undergo sexual conjugation in
the Vertebrate, has been found.!

1 Even in the case of T. equiperdum, the fact that the parasites are
transmitted directly from one Vertebrate to another, during the act of
coitus, does not imply that this form has never possessed a definitive Insectan

THE HH#MOFLAGELLATES. 28]

The course of evolution of the Heematozoa (as a whole)
in the blood of the Vertebrate appears to have been charac-
terised by the acquisition of an intra-cellular phase, together
with the capacity for multiplication while in that condition.
In such cases the subsequent tendency has evidently been to
develop the gregariniform or Hemosporidian phase at the
expense of the trypaniform or Hemoflagellate one. For, on
carefully reviewing the evidence set forth in Section IX, it
seems scarcely open to doubt that there is a very close
phylogenetic relationship between the Hemoflagellates and
the Heemosporidia, and that, moreover, in some cases there
is an actual ontogenetic connection between the two types of
form at the present day.!

host, or that it is not still obliged to return, if only at rare intervals, to
such a habitat, in order that, by sexual conjugation, the stock may be rein-
vigorated. The direct mode of transmission, most probably a secondary
acquirement, is readily accounted for by the peculiar superficial distribution
of the organisms (see above, p. 176). It must be borne in mind that this is
the only species which is known to be able to traverse healthy mucous
membranes, and this itself is against the derivation of Trypanosoma
from an intestinal or entero-ccelomic, Vertebrate form.

1 Novy and McNeal (81), basing their view upon the results of their
cultural methods of work, conclude that there is no connection between
Cytozoa (Hzmosperidia) and Trypanosomes. They disbelieve in Schaudinn’s
work, and think that this author was misled by working with ‘‘ impure
cultures,’ as they regard the conditions in which the parasites occur in
their natural hosts. But, as Mesnil says, the presence of the characteristic
Hemosporidian pigment in the trypaniform phase is a fact very difficult to
explain away. Again, Novy and McNeal consider that the appearance of
Flagellate forms in Rogers’ culture of “‘ P.” donovani was merely a coin-
cidence, indicating the existence of a variety of human trypanosomosis
hitherto unknown, the two parasites being quite distinct! The ample con-
firmation which both these important researches have already received
sufficiently disposes, we think, of the above view.

It is, of course, very probable that there are Trypanosomes of birds (as
of other Vertebrates) which have no Hemosporidian phase. It is not safe,
however, to argue that there is no connection between a given Hemoflagel-
late and a given Hemosporidian from the negative evidence obtained by
cultures and their injection. For it is most unlikely that the further deve-
lopment of the parasites can go on in an artificial medium as successfully
and as normally as in the Insectan host, to which their biology is specially

VOL, 50, PART 2,—NEW SERIES, 20

282 H. M. WOODCOCK.

It is, indeed, possible to construct what may be regarded
as stages in the direction of evolution thus outlined. The
first step is shown by those forms which possess resting,
attached phases ([rypanoplasma almost certainly, and
probably some species, at any rate, of Trypanosoma,
e.g. T. equinum and T. equiperdum). A stage in
advance is exemplified by Trypanomorpha noctue and
Trypanosoma ziemanni; these parasites become, for a
portion of their life, intracellular, but do not multiply
asexually in this condition. On the other hand, it appears
very probable from Billet’s work (p. 251) that certain Amphi-
bian parasites possess (as Hemogregarines) typical schizo-
gony, and this may be also the case in some Piscine forms.
Here the trypaniform phase is very likely much less in
evidence, and assumed (in the blood) only at rare intervals.
Probably about this level, too, come the Piroplasmata,
which are very interesting in that some of them retain the
nuclear dimorphism while in the intracellular phase. Lastly,
a trypaniform phase in the life-cycle can only be recognised
with difficulty, in the sporozoites or merozoites (Plasmo-
dium vivax, possibly other malarial parasites and certain
Hemogregarines [cf. p. 253]); or, in other cases, may be
qnite lost.

It will be gathered from the above remarks that, with the
advance of our knowledge, it may well become extremely
difficult to decide whether a given parasite should be looked
upon as a Hemoflagellate or a Heemosporidian, and, if so,
where to draw the dividing line. The writer is truly thankful
that he does not feel himself called upon, at present, to
attempt to answer such momentous questions.

>

adapted; or that the “ cultivated forms ” when injected into the Vertebrate
host, can give rise to the same phases as those which may develop after
natural infection (see also below, p. 301).

THE HAMOFLAGELLATES. 283

Section XII. Systematic ENUMERATION.

The reasons for the division of the Trypanosomes into two
distinct and entirely independent families have been fully
discussed in the preceding section. Apart from the funda-
mental diagnostic characters, it is quite likely that other
important features in which the parasites differ in the two
cases will become known as our knowledge of the complete
life-cycle increases.

Sub-order.—Monadina.

Family.—Trypanomorphidae, n. fam.!

Hemoflagellates derived from a uniflagellate, Herpeto-
monadine form, in which the point of insertion of the flagellum
into the body has travelled backwards from the anterior end
for a greater or less distance, the flagellum itself having be-
come, concurrently, attached to the body for a portion of its
length by means of an undulating membrane. At present
only one genus can be said to be known with certainty.

Genus Trypanomorpha, n.g. With the characters of
the family. ‘The only species yet known is the type species,
T. noctue (Celliand San Felice). [Syn. Trypanosoma n.
(C. and 8. F.), Schaud. = Halteridium n. (C. and 8. F.)]?
The full life-cycle of this parasite has been described above
(Section V). Vertebrate host: Athene noctua (Little
Owl) ; Invertebrate host: Culex pipiens.

1 It is considered best to remove Trypanomorpha, and other allied
Monadine forms which are true Hemoflagellates, from the old family of the
Cercomonadide or Oicomonadacee, in the same way that Doftlein (19)
has separated the Heteromastigine section (his inclusive genus Try pano-
soma) from the Bodonide.

* Schaudinn places this form in the genus Trypanosoma. ‘The writer,
however, inclines to the view that the type-species of that genus (T. rota-
torium) is a Heteromastigine form (see below, p. 288), in which case this
Avian parasite cannot be included therein. Moreover, it is altogether
uncertain whether the type-sp. of Halteridium (H. danilewskyi)
agrees generically with H. noctuz. Quite possibly it does not, since, for
one thing, it possesses typical schizogony. Hence the writer thinks it best
to place H. noctuze in a distinct genus, Try panomorpha.

284. H. M. WOODCOCK.

There are, in addition, two or three forms, which are most probably to be
placed in this family, but which are not yet sufficiently characterised for
their generic position to be settled. It is, for instance, quite likely that
Léger’s parasite Crithidia fasciculata (see above, p. 269) from females
of Anopheles maculipennis is sufficiently allied with Trypano-
morpha for the two forms to be united in the same genus, in which case,
of course, the name Crithidia will take priority.

Sub-order.—H eteromastigina.

Family.—Trypanosomatide, Dofl., emend.

Flagellates, in the great majority of instances hemal para-
sites, derived from a biflagellate, Bodo-like type, in which
the posteriorly-directed flagellum (the so-called trailing
flagellum) is always present and attached to the body by
means of an undulating membrane, of which it constitutes the
thickened edge. ‘The other, the anterior, flagellum may or
may not persist. ‘Three genera so far distinguished.

Genus 'rypanoplasma, Lav. and Mesn., 1902. The
anterior flagellum is present. Both flagella are inserted close
together at or near the anterior end of the body. Three species
certainly known, which can be arranged in two groups:

(a). The anterior flagellum is well developed, and the free portion of
both flagella are of about equal length.

T. borreli, L. and M., 1902. Length! from 20-22 m, of free flagella
13-15; breadth 33-43 p» (figs. 17 F, @, 18). Hosts: (V.)? Leuciscus
(Scardinius) erythrophthalmus, rudd, and Phoxinus levis,
minnow ; (I.) not yet known.

T. varium, Lég., 1904. Length (medium) about 25 /, of free flagella
18-20. Hosts: (V) Cobitis barbatula, loach; (1) Hemiclepsis
marginata, perhaps also Piscicola sp., leeches. This parasite differs
from T. borreli by the rather longer flagella, and by its not having the
pronounced cytoplasmic granules of the latter form, Léger considers
specific distinction’ is also shown by the fact that, in streams containing
both loach and minnows, only the former are infected with this parasite.

(s). The anterior flagellum is much shorter than the free portion of the
posterior one, and apparently tending to disappear.

T. cyprini, Plehn, 1903. Medium length about 20 » (fig. 17 8).
Host: (V) Cyprinus carpio, carp.

1 Of the body alone, independent of the flagella.
(V.) signifies vertebrate hosts; (I.) invertebrate ones.
3 See remarks on specific distinctions, pp. 238—9.

THE H#MOFLAGELLATES. 285

It is uncertain whether the Flagellate organism described by Labbé (29)
from the medicinal leech (Hirudo), which had probably sucked the blood
of a horse or ass, should be placed here or not. This parasite (fig. 39),
to which Labbé gave the name of Trypanomonas danilewskyi,!
was elongated almost filiform (15-20 » by 1»), with apparently along, thin,
more or less coiled flagellum at either end. It also possessed a delicate un-
dulating membrane. Labbé considered one of the flagella to be a very
attenuated prolongation of the body and membrane, on the analogy of the
spindle-like forms figured by Danilewsky (cf. fig. 16 @ of Hanna’s Trypano-
some from Indian pigeons). In that case, the parasite would really be a
spirochetiform Trypanosoma, but the figures, so far as they go, do not
convey that impression. Re-investigation of it is necessary in order to
settle its position.

Fie. 40.

Fie. 39.—“Trypanomonas” danilewskyi, Labbé, x 1200.
(After Labbé.)

Fie. 40.—Trypanophis (Trypanoplasma) intestinalis
(Léger). In 4, note the row of spherules down the side near the
undulating membrane; in B the kinetonucleus is in two parts, pro-
bably resulting from division. (After an unpublished drawing

kindly lent by Prof. Léger.)

Genus Trypanophis, Keysselitz, 1904. The body resem-
bles that of Trypanoplasma in form and general appear-
ance. ‘The locomotor apparatus does not appear to be so
well developed, however, especially in T. grobbeni. The

1 Even if this form is found to agree generically with Trypanoplasma,
Labbé’s name Trypanomonas could not be used, since this designation
was originally employed by Danilewsky for the young forms of a Try-
panosoma, with which, therefore, it is synonymous.

286 H. M. WOODCOCK.

anterior flagellum is longer than the free portion of the
posterior one. The species included are not, so far as is
known, hemal parasites.

T. grobbeni (Poche), 1903. Average length 60—65 »; width about 4p.
The undulating-membrane is relatively narrow and not expanded into wider
folds at intervals (fig. 41). The anterior flagellum is fairly long, but the
free part of the posterior one is short. This species is parasitic in
certain Siphonophora, Cucubalus kochii, Halistemma tergest-
inum, Monophyes gracilis, of common occurrence in the Gulf of
Trieste. Apparently the same parasite has also been observed in Abyla

Fig. 41.—Try panophis grobbeni (Poche). e.c., ectoplasmic
cap; e.l., delicate ectoplasmic layer, thinning out posteriorly ; m.,
undulating-membrane ; 7,, inclusions in the cytoplasm ; 2., nuclear
body of uncertain origin and significance. (After Keysselitz. )

pentagona, from the Gulf of Naples. The organisms are to be met with
in all the ramifications of the cclenteron, from the digestive cavity of the
gastrozoids to the radial canals of the medusoid buds.

Nothing is known regarding the transmission of the parasite from one
Siphonophoran colony to another. That is to say, we are quite in the dark
as to whether they have any alternate host or not. Of the two suppositions,
the latter seems much the more probable.

T. (Trypanoplasma) intestinalis (Léger), 1905. A much smaller
form than the last (fig. 40). The dimensions given are: Length of body
(without “tail”) 14; of anterior flagellum, 16 [judging from the

THE. HA MOFLAGELLATES. 287

figures it does not seem quite so long]; of tail and posterior flagellum (free
part) 16. Habitat: esophagus and anterior part of stomach of Box
boops.

This form is manifestly closely allied to T. grobbeni. At the same time
it also exhibits great resemblance to Trypanoplasma borreli, in which
genus Léger has included it. In the writer’s opinion, it appears to be, as
regards its morphology, intermediate between these two parasites, and it is
not easy to decide where to place it. The fact that the free part of the pos-
terior flagellum is not so long as the anterior one, the occurrence of a row of
spherules on that side along which runs the undulating membrane, and,
lastly, the fact that this form is not a hemal parasite point to its associa-
tion with Trypanophis. Léger is inclined to do away with this latter
genus altogether, but this seems premature, until the life-cycle is better
known.

Genus Trypanosoma, Gruby, 1843. Principal synonyms!:
Undulina, Lank., 1871; Herpetomonas, Kent, 1880 (only
in part, since the type sp. is H. musce-domestice) ;
Paramcecioides, Grassi, 1881; He matomonas, Mitrophan,
1883; Trypanomonas, Danil, 1885 (for young forms).
There is no anterior flagellum. The point of insertion of the
attached (posterior) flagellum into the body, and, consequently,
the commencement of the undulating membrane may be

1 The synonymy of this genus and of its different species is well discussed
by Laveran and Mesnil (47, 56), and Salmon and Stiles (96), from whom most
of the information here given is compiled. An explanatory note may be
added with regard to Herpetomonas, since until recently the parasite of
rats (T. lewisi) was called by this name. This parasite was referred by
Kent (27) to his new genus Herpetomonas, founded for H. muscex-
domestic (Burnett), the genus Trypanosoma being reserved for the
frog parasite and for Eberth’s form from the cecum of birds, which Kent
termed T. eberthi. Later, when H. lewisi became better known, Senn
(100) revised the diagnostic characters of the genus. As a result, the prin-
cipal feature left by which to distinguish between Herpetomonas and
Trypanosoma was the presence in the former of a thickened external
border to the undulating membrane, of flagellar nature, which was supposed
to be absent in Trypanosoma. With Laveran and Mesnil’s reinvestiga-
tion (45) of T. rotatorium (T. sanguinis), this difference was found not
to exist, and therefore the older name prevails. The name Herpetomonas
is retained for the original type (musce-domestice), which has no undu-
lating membrane.

288 H. M. WOODCOCK.

almost anywhere in the anterior half of the body, but is
usually near the extremity.1

The question of the sub-classification of this genus is one
of much difficulty. One is confronted by a great number of
forms, several of which have no very definite or constant
differential feature by which to characterise them, such as in
the case of the Gregarines and Coccidia, for example, is
naturally afforded by the spores. This is largely owing to
the fact that so little is yet known of the life-history of most

Fic. 42.—Different Mammalian Trypanosomes to show uni-
formity in size and shape. 4, Trypanosoma evansi; B, T.
brucii; ©, D, T. equiperdum; £, I. equinum, x 16500.
(After L. and M.)

that reliance has to be placed almost entirely upon the adult
size and form in any endeavour to classify the parasites on a
morphological basis. Now it has been already stated that,
in many cases, the variation in this respect is very slight,
and, in addition, in several instances a particular form may

1 The type-species is T. rotatorium (Mayer) of frogs. At present,
unfortunately, this parasite cannot with certainty be included in the above
diagnosis, owing to its unusual shape, position of kinetonucleus, ete. The
occurrence, however, of an allied form in Hyla, which is evidently inter-
mediate between T. rotatorium and the more typical, fusiform species
of the genus, strongly points to the agreement of the former (T. rotatorium)
with the majority of Trypanosomes in belonging to the Heteromastigine
section. At any rate, it would be premature to separate the Mammalian and
Piscine forms under a new generic name; the confusion which this course
would entail is to be avoided until it is shown to be necessary.

THE H#MOFLAGELLATES. 289

itself vary almost as much at different times and under
different conditions (see under Morphology). For the present,
at any rate, a very useful aid towards distinguishing dif-
ferent species is furnished by the biological relations of the
parasites. Speaking generally, it may be assumed that here,
as is known to be usually the case in the Sporozoa, a particular
species is restricted either to one particular host or, at most,
to a few closely allied ones. The greatest difficulty arises in
considering the Mammalian forms, many of which have never
been observed in the true natural hosts, but only in various
“foreign ” animals, for which they are all more or less patho-
genic. The immunisation experiments of Laveran and Mesnil
(see above, p. 176) render it most likely, however, that such
forms are, at all events, distinct varieties, and, quite probably,
distinct species.

On the other hand, having regard to the gross or extreme
differences in form which are met with (cf. T. rotatorium
and '’. inopinatum) one might be tempted to arrange the
different species into sub-genera, grouping together, for
example, parasites with a blunt or rounded anterior extremity,
or those with a filiform or attenuated one, and, again,
separating forms with no free prolongation of the flagellum
posteriorly, from those which have one. Bearing in mind,
however, the great polymorphism which is known to occur, any
such arrangement would be quite arbitrary and certainly pre-
mature in the existing state of our knowledge. It may very
well be that when more life-histories come to be revealed some
of the forms at present included in the genus Trypanosoma
will have to be transferred to one of the other genera of
Heemoflagellates or placed in a new one.

In this article the writer considers it best to arrange the
different species under the different classes of Vertebrate
hosts in which they are parasitic; this will, at any rate,
facilitate reference to any particular form.

290 H. M. WOODCOCK.

Mammalian Forms.

(4) Non-pathogenic.—T. lewisi, (Kent), 1879. (Syn. Herpeto-
monas l., Kent.) Length 24-25 p, breadth 13-13 .! This species (figs. 16 4,
27) is characterised by its thin, drawn out, and pointed anterior extremity,
and also by the position of the trophonucleus in the posterior half or third of
the body. The cytoplasm is very clear and free from granules. Hosts: (V.)
Mus decumanus, M. rattus, and M. rufescens; (I.) the life-cycle can
certainly be undergone in the rat-louse, Ha matopinus spinulosus, which
is therefore a true alternate host. Prowazek thinks, however, that fleas
may also serve the purpose of transmission. A closely allied species (by
some considered as only a race or variety of T.lewisi) is found in Cricetus
arvalis (frumentarius?), the hamster; this form is not inoculable into
rats, and, conversely, T. lewisi is not capable of living in the hamster.

T. duttoni,? Thiroux, 1905. Length 25—30p, of free flagellum 63—
10; breadth 23». This parasite resembles T. le wisi, and also, in a general
fashion, the Trypanosomes of other small Rodents mentioned below.
(V.) host: Mus musculus. This form is not inoculable into rats, and is
therefore distinct from T. lewisi. It was found in mice in Senegal, and
Thiroux wonders whether it is identical with the parasite described as a
Herpetomonad by Dutton and Todd (21), also from Senegambian mice.
In the latter, however, no undulating membrane was observed, and hence
it is doubtful whether it was really a Trypanosome.

In addition to the above forms, Trypanosomes have been casually observed
in various other Rodents, but as yet they are unnamed and not much is
known about them.

Petrie (82) recently found three rabbits spontaneously infected with a
Trypanosome (fig. 43.4). The parasites were quite numerous in the blood,
and, in one case, present for six months at least, without causing any ill-
effects. This form appears to be similar, as regards size and morphology,
to T. lewisi, but most likely belongs to a distinct species, since the latter
form is not inoculable to rabbits.

Another Trypanosome (fig. 438) has been recently observed by Donovan
in an Indian squirrel (Seiur us palmarum). Its total length was 18—20 p,

- ‘The (dimensions given are ee to jidoeate the average size of the
adult parasite, but, as above said, they can only be considered approximate.
Unless otherwise stated, the length is inclusive of the flagellum.

2 The writer commenced this article with the intention of having a
figure of every species; the number of those known has increased so
greatly, however, while it has been in progress, that this has proved
impracticable, and hence some of the most recently described species are
unfigured.

THE H#®=MOFLAGELLATES. 291

distinctly smaller, that is, than T. lewisi. This is probably also a
distinct species.

Galli- Valerio (24), again, has remarked upon the presence of a Trypano-
some-like Flagellate, about 22 in length, in the blood of a dormouse
Myoxus avellanarius).!

An earlier observation is that of Chalachnikow (15 a) of an elongated
form, 30—40 p by little more than 1p, from Russian marmots (Spermo-
philus guttatus and S. musivus).

T. pestanai, Bettencourt and Franca, 1905. Length given as 30—32
(somewhat uncertainly because the parasites are generally rolled up); of
free flagellum 4°5—5 yp. This is a relatively wide form, the breadth being
5-65. Anterior end long and fine; kinetonucleus some distance from
the extremity. (V.) host: Meles taxus, badger.

Fie. 43.—a, Petrie’s Trypanosome from a rabbit; B, Donovan’s
Trypanosome from a squirrel. Xx 2000. (After L. and M.)

Trypanosomes appear to be not uncommon in bats. Dionisi? first
reported the occurrence of the parasites in Miniopterus schreibersii;
Durham * noticed some ina Phyllostoma in Brazil—to be exact, in the
blood contained in the stomach of a Stegomyia which had fed on the
bat; and Donovan (vide L. and M. [56]) mentions having observed a large
form in Pteropus medius, in Madras. These were probably all distinct
species, but no account of them has yet been given. During the last year
or two, however, somewhat brief descriptions of certain species are to hand.

T. nicolleorum, Sergent HE. and E., 1905. Length 20—24 p, of free
flagellum 4—5 »; breadth 13. Anterior end tapering and pointed. The

1 Brumpt appears to have named a Trypanosome from the garden dor-
mouse (Myoxus nitela), T. blanchardi (vide Brumpt and Lebailly [11)) ;
the writer can, however, find no reference to the paper describing this form.

* Drontst, ‘ Atti Soc. Studi Malaria,’ i, p. 145, 1899.

* Duran, ‘ Rep. Yellow Fever Expedition, Para,’ Liverpool Sch. Trop
Med., mem. 7, 1902, p. 79.

292 H. M. WOODCOCK.

kinetonucleus is some distance from the extremity. (V.) hosts: Myotis
muruius, and Vespertilio kuhli, Algeria and Tunis. In certain of
the Vespertilio kuhli examined much larger forms were seen, 25—30
by 6, for which the name T. vespertilionis is proposed provisionally.
The possibility is recognised, however, that these forms may be individuals
about to divide [or perhaps sexual (female) individuals?]. Petrie (83)
has recently observed Trypanosomes in Pipistrellus pipistrellus, in
Hertfordshire, which he thinks may have been T. nicolleorum, although
they were shorter (apparently only about 16 p long).

T. dionisii, Bettencourt and Franca, 1905. Full dimensions not
given; length of free flagellum 6°54. Kinetonucleus quite at anterior
extremity. Hosts: Vesperugo pipistrellus, V. serotinus, and V.
nattereri, Portugal.

B. Pathogenic forms.—We come next to the so-called pathogenic
group. These parasites have been “successfully” inoculated into many
and various Mammalia, which cannot, however, in the majority of cases be
regarded as natural, tolerant hosts. In dealing with these disease-causing
forms, it is obvious that the more narrowly the original source of the para-
site is defined the closer do we get to the true host or hosts. Similarly with
the Invertebrate hosts, it is sometimes rather difficult to be certain which
is the natural one for the species concerned, for experiment has shown
that a biting-fly, other, in all probability than the true host, can, as it were,
accidentally convey the parasites, if, after feeding on an infected animal,
it is allowed to bite a fresh one within a limited time. One helpful factor
in this determination is the coincidence of the zone of a particular Insect!
with that of any disease.

T. brucii, Plimmer and Bradford, 1899. (Syn. T. brucei, Buff. and
Sch., followed by L. and M. and others.) Length 28-30 p, breadth 13-25 p.
The anterior end is usually bluntly rounded or truncated (figs. 42B, 44).
The cytoplasm often contains in the posterior half large deeply-staining
granules. Hosts: (V.) probably Antilopidw, such as Catoblepas gnu,
Strepsiceros capensis (‘Koodoo”), and Tragelaphus scriptus
sylvaticus (‘ Bushbuck”’), perhaps also buffaloes; (I.) Glossina mor-
sitans and G. pallidipes, Tsetse-flies. The cause of Nagana or Tsetse-
fly disease in South and South-east Africa among cattle, horses, ete. Most
domestic animals are susceptible.

Other trypanosomoses, more or less allied to Nagana, and perhaps caused
by different varieties or races of the same parasite, have been observed in
German East Africa and Togoland among cattle, horses, and other animals.

Again, the disease known as “ Aino,” which occurs in Somaliland among

1 A very useful map showing the zones of distribution of the different
species of Glossina is given by AusreEn, E., Rep. 8. 8. Comm. Roy. Soc.,
6, p. 278, 1905.

THE H#MOFLAGELLATES. 293

dromedaries, and which appears to be transmitted by another Tsetse-fly,
namely, G. longipennis (locally termed the “ Aino”), is probably also a
variety of Nagana (see Brumpt [9 a]).

T. evansi (Steel, 1885). (Syn. Spirocheta evansi, Steel.) Length
about 25 pw, breadth about 14”. This parasite (figs. 42 4, 45) is, morpho-
logically, very like T. brucii. It is generally rather more slender than
that form, and the anterior end rather more tapering and usually acutely
conical. Moreover, the free part of the flagellum is slightly longer in T.
evansi, and the cytoplasm lacks the prominent granules of T. brucili.
T. evansi also performs greater and more rapid movements of displace-
ment than the Nagana parasite. Natural hosts uncertain ; (V.) perhaps to

Fig. 44. Fie. 45. Fic. 46. lic. 47.

Fie. 44.—T. brucii (after Bradf. and Plim.).

Fie, 45.—T. evansi. X 2000. (Original, from a preparation
of the blood of a mule, kindly lent by Mr. Plimmer.)

Fie. 46.—T. equiperdum. (After Ligniéres.)

Fie. 47.—T. equinum. (After Ligniéres.)

be found among indigenous Bovide, and (I.) is probably a Tabanus,
T. tropicus and T. lineola having been suggested (Rogers [92, 92 a)) ;
in Mauritius the epidemic is thought to have been spread by Stomoxys
nigra. The cause of Surra in Indo-Burmah, which is particularly dan-
gerous‘to Equidz. The disease is less fatal to cattle than Nagana is.
Various other animals also liable. The malady has been recently imported
into Mauritius and the Philippines.

The illness known as ‘‘ Mbori,’” occurring among dromedaries coming
from the Sahara into the Soudan (Timbuctoo, etc.), which is apparently
also conveyed by a Tabanus, is considered both by Vallée and Panisset

294 H. M. WOODCOCK.

(117 a) and Laveran and Mesnil (54a) to be a milder form of Surra, the
parasite which causes it being a “race” of T.evansi.

T. equiperdum, Dofl., 1901. (Syn. T. rougeti, L. and M.) Length
25-28 p, breadth 13-2; slightly smaller than T. brucii. Also differs
from that species in not having prominent grains in the cytoplasm. The
kinetonucleus is relatively large and well-developed (figs. 42 c and p,
46). ‘This form has been the least studied of the better known patho-
genic ones. ‘The cause of Dourine in horses in Algeria and in certain parts
of Southern Europe (chiefly the Mediterranean littoral). The infection is
transmitted (invariably?) during the act of coitus, and this explains
why mules and geldings are exempt. In the case of roving (wild) asses
the illness is usually slight, and the parasites are apparently more or less
latent, but whether these animals constitute the true Vertebrate host or
not cannot at present be said. Moreover, it is not yet certain whether the
parasites in the natural conditions have any alternate Insectan host into
which they must pass at intervals in order to complete the life-cycle. (See
also p. 280, footnote.)

There appear to be one or two other varieties of trypanosomosis in
Northern Africa (Algeria). Thus, Sergent (101 a) announces a malady of
dromedaries, which is very similar to Mbori; and Rennes (91), and Rouget
(98, 98a) and others seem unable to decide whether there is a trypanoso-
mosis of horses, distinct from Dourine.

T. equinum, Voges corr., 1902. (Syn. T. elmassiani, Ligniéres.)
Length 22-25 p, breadth 13-2 #. A morphological character which sharply
distinguishes this species from the rest—and which may ultimately prove
to be of more than specific importance—is the very minute size of the
kinetonucleus (figs. 42 5, 47). Cytoplasmic grains present, but not so
numerous as in I’. brucii. In shape and details of form this parasite much
resembles ‘I. evansi. The Vertebrate host is most probably Hydro-
cherus capybara; the transmitting Insect perhaps a T'abanid (many
workers have considered a Stomoxys, either S. nebulosa or S. cal-
citrans, to be the carrier, but Ligniéres and Elmassian and Migone think
otherwise). ‘I’. equinum causes the destructive disease of horses known
as Mal de Caderas in Brazil, Argentina, and Central South America.

’. gambiense, Dutt., 1902. (Syn. 'l. ugandense,! Castellani, T. castel-
lanil, Kruse.) Length 21-23, breadth 13-2. This species (fig. 48) is,
according to its average size, one of the smallest yet found. The parasites in
the blood frequently exhibit slight morphological differences from those in
the cerebro-spinal fluid. ‘Che former are somewhat longer and more slender,

1 The specific name ugandense was the one first given to the parasite
found in sleeping-sickness cases, Dutton’s name having been previously
conferred on the form originally found in cases of human trypanosomosis
(Trypanosoma fever). See next page.

THE HMMOFLAGELLATES. 295

and, correlated with this, the kinetonucleus is situated farther from the
anterior end than it is in the more stumpy forms. Investigators are, how-
ever, at one in considering that these differences are due merely to the
different habitat, since both varieties, when inoculated into other animals,
give rise to the same kind of form. ‘The cerebro-spinal fluid would appear
to be less favourable a medium than the blood. T. gambiense is the
cause of human trypanosomosis in West and Central Africa, the earlier
stages of which, when the parasites are confined to the blood, are known as
Try panosoma-fever, the later ones, after they have penetrated into the
cerebro-spinal canal,' constituting the deadly malady of sleeping sickness.

B:

Fic. 48.—T. gambiense, from the blood. (a, after Bruce and
Nabarro; B, after Castellani.)

It seems most probable that the original Vertebrate source or “‘ reservoir ”
of this parasite is some indigenous tribe or race of natives in whose blood—
and in all probability only in their blood—the Trypanosomes live, as it were,
normally, parasite and host having become mutually tolerant.2 Whether
any animal other than man is also a natural host, is quite unknown. If

1 Plimmer (85) has lately expressed the view that the forms met with
in Trypanosoma-fever and sleeping-sickness are distinct species, basing
his opinion on the behaviour and appearance of the parasites after being
inoculated into the same host (rat), and also on the symptoms presented by
the latter in the two cases. The weight of evidence at present, however, is
decidedly against this view; see the Reports of the 8. S. Comm., also
Brumpt and Wurtz (18), Laveran (38), Thomas and Linton (116), and
several other workers.

2 For various factors which have helped to bring unaccustomed, un-
adapted tribes and individuals into the zone of the parasite and have thus
led to the spread of the infection and its invasion of fresh regions in the
character of a fatal disease, the reader is referred to Lankester’s instructive
article (81).

296 H. M. WOODCOCK.

there is no other host among the Primates it is improbable that there is
any other Mammalian one for this species. The Invertebrate host is,
undoubtedly, Glossina palpalis; the possibility of this being so appears
to have been first suggested by Brumpt (9). The distribution of this
Tsetse-fly narrowly coincides with that of the disease, and where there is no
fly, sleeping-sickness is not prevalent. It has not yet been proved whether
or no other species of Glossina also naturally transmit the parasites.

Nabarro and Grieg (S. 8. Rep., No. 5) mention, in addition, the occur-
rence of three or four cases of trypanosomosis among diseased animals
which came under their observation in Uganda. The parasites occurred as
follows :—(a) in sick transport oxen in Entebbe, which came originally from
East Africa (the illness caused being locally known as ‘‘ Mukebi”); (b) in a
herd of diseased cattle at Jinga, Busoga (“Sutoko”); (c) in an English
dog which had contracted the disease while with the Abyssinian Boundary
Commission; and (d) in a sick mule at Entebbe. From the observations
and animal experiments by these workers, and subsequently by Grieg and
Gray (24), the conclusions arrived at are as follows :—the Trypanosomes
concerned are probably all distinct from T. gambiense; (a) the parasite
in sick oxen at Entebbe (‘‘ Mukebi”) is a distinct species; (b) the ‘‘ Jinga”’
Trypanosome is most probably a variety of ‘I. brucii, the illness being an
acute form of Nagana; (c) and (d) the “ Abyssinian Boundary” parasites
and the Trypanosomes found in a sick mule appear to be identical, and con-
stitute probably another distinct species.

‘'. dimorphon, Dutt. and Todd, 1904. As implied by the specific name,
more than one “type” of this parasite is usually distinguished. One is
small and tadpole-like (fig. 49, 1b, Ila), the other long, fusiform, and more
resembling an ordinary Trypanosome (la, 11d). A third variety, wide
and stumpy (IIb and ¢) is also distinguished by Dutton and ‘Todd, but
Laveran and Mesnil (54) consider that this only represents an enlarged
tadpole form about to divide. The average dimensions are: of the tadpole
variety, length 10-15 by ‘7-15; of the elongated one, length about
25 by 14-2”. The exact relations of these two kinds of form to one
another have not yet been ascertained. Although Laveran and Mesnil
describe and figure a series of intermediate stages (1), they do not think
that the smaller type grows or passes into the longer one, basing this view
on the ground that each is capable of reproducing its like by equal binary
fission.! In one important respect the accounts of Laveran and Mesnil and
of Dutton and ‘odd are at variance. ‘he first-named authors maintain
that none of the parasites have any free continuation of the flagellum, this
terminating, in all cases, at the posterior end of the body, which is here

1 It is quite possible, however, that these large parasites possess some
other mode of division in addition, by means of which they give rise to a
fresh succession of tadpole forms.

THE HHMOFLAGELLATES. 297

attenuated and filiform (fig. 49, Ia). On their part, Dutton and Todd
fizure a distinct free portion of the flagellum in all their types (ID), it being
short in the tadpole-like and stumpy forms, but long and well-developed
in the large (adult?) ones. Reconciliation between these two views must
await further research. T. dimorphon causes marked trypanosomosis
of horses in Senegambia; these animals are few in number in that colony
and mostly imported. The short forms are met with in the earlier stages

Fie. 49.—T. dimorphon. I, after Lav. and Mesn.; a, long
form, b, short form. II, after Dutton and Todd; a, tadpole-
form, 6, stumpy form (from a rat), c, stumpy form dividing, d,
long form. x 1500.

of the malady, the long ones occurring later. The “natural” Vertebrate
host is not known. The transmitting Insect is, possibly, Glossina
palpalis, which is abundant in the district, but it should be added that
Dutton and Todd’s endeavours to artificially transmit the parasites to rats
by this means were unsuccessful.

T. theileri, Laveran,! and T. transvaaliense, Laveran, 1902. These
two forms are considered together, since it is uncertain whether they do
not really both belong to one species, which would bear the former name.
T. theileri (fig. 50 a and b) is the largest known Mammalian Trypano-

1 Laveran (32) named the parasite on the 3rd March, 1902, and Bruce
(7) on the 8th of the same month.

VoL. 50, PART 2.—NEW SERIES. 21

298 H. M. WOODCOCK.

some ; medium length about 50, breadth 33—4,. It is thus readily dis-
tinguished by its size from the other Mammalian forms, although it agrees
with most of them in being of the typical fusiform shape. It appears to be
confined to Bovide, occurring especially in the Transvaal, and causing the
disease of cattle known as “ Galziekté” or bile-sickness. Herds of cattle
imported from Argentina, Texas, etc., are particularly liable to suffer from
it. Theiler, who discovered the parasites, thinks that they are transmitted
by a biting fly, Hippobosca rufipes. Another species, H. maculata,
recently imported with cavalry from India, may also aid in spreading the
disease. T. transvaaliense (fig. 50 c—e) averages about 30 » in length
by 4 inwidth. From the dimensions given (though not, apparently, from
the figures) it would appear to be rather wider than T.theileri. Its
distinguishing morphological feature, however, is the position of the kineto-

B\_

Fie. 50.—a and b, T. theileri; c—e, T. transvaaliense.
e is asmall form dividing. x 1250. (After L. and M.)

nucleus ; this organella is rod-like and situated in contact with the tropho-
nucleus. Correlated with this, the undulating membrane is short and
poorly-developed. Moreover, this variety is quite capable of division
(fig. 50 d), even in the case of very young individuals (e). Nevertheless,
Theiler has recently observed transitional forms, intermediate between T,
theileri and T. transvaaliense, with a varying distance between
the tropho- and kineto-nucleus. This author has also found that, on
inoculating the latter variety into Bovidx, the former (T. theileri)
is produced in time as well. These facts would seem to show that the
two forms are not really independent and distinct species.

bo
Ne}
wo)

THE H#MOFLAGELLATES.

Forms Parasitic in Avian Hosts.

T. ziemanni (Layv.). (Synn. Spirocheta z. [Lav.], Schaud., “ Hama-
meba” z. Lav.; the “ Leucocytozoon”’ of Danil.). Schaudinn, who has
fully described the life-history of this form (see above, p. 242), does not give
the actual dimensions. According to Laveran (87) the size of the female
gametocytes in the resting, intra-cellular condition, varies from 12—21 p
by 4—7 yp, the male ones being slightly less. The length of the micro-
gametes (“flagella”) is from 20—25 pp. The ordinary indifferent indi-
viduals are characterised by their extremely Spirocheta-like facies and
their habit of remaining united together in pairs after division. Hosts :
(V.) Athene noctua, Little Owl, also Syrnium aluco (see below, under

Mivet, (yl Fie. 52. Fie. 53.

Fie. 51.—T. johnstoni. g = deeply-staining granule at
distal extremity of flagellum. x 1500. (After Dutton and
Todd.)

Fie. 52.—T. sp., from Senegambian birds. x 1500. (After
D. and T.)

Fic. 53.—T. sp., from Indian birds (Hanna’s T.). (After
Hanna.)

T. avium) and ‘‘a little white owl from Cameroun ” (Ziemann [121]), in
which sexual forms were seen; (I.) Culex pipiens.

T. johnstoni, Dutton and Todd, 1903. Length 36—38 py, width 1:4 to
16. This parasite resembles T. ziemanni in shape, being also markedly
spirochetiform (fig. 51). The undulating membrane is narrow and poorly
developed. The most interesting morphological point is the absence of
any free continuation of the flagellum. This terminates abruptly, at the
end of the body, in a distinct, deeply-staining granule (g), which is pro-
bably of centrosomic nature. This species was found in small birds

3800 H. M. WOODCOCK.

“ millet-eaters” (Estrelda estreld), in Senegambia. (I.) host probably
(for this and other Avian forms) some species of gnat or mosquito.

Another form, not named, is described by the same authors from the same
host and also from another (Crithagra). The birds were more frequently
infected with this parasite than with T. johnstoni. This Trypanosome
represents the other extreme of type, being relatively very wide and stumpy
(fig. 52). Its total length is about 32 p, and its greatest width 8y. The
free portion of the flagellum is from 10—12 4 long. The trophonucleus is
placed transversely across the body. The kinetonucleus is very close to the
anterior end, and immediately behind it is a vacuole.

Trypanosomes from Indian Birds.—Hanna (26) has recently noted
two forms, one parasitic in the crow, the other in the pigeon. The former

Fie, 54. Fie. 55.

Fie. 54.—T. padda, Thiroux. At the base of the flagellum
is separate from the kinetonucleus, and thickened prior to division.
x 1200. (After Thiroux.)

Fie. 55.—T. avium. Here the kinetonucleus is itself divided.
(After L. and M.)

is only briefly mentioned, since, owing to the preparation having lost its
staining colour, little beyond the general form could be made out. The
parasite appears to conform to the usual spindle-like shape. Its size is
given as from 40—56 pp by 83—4°8 uw. The other parasite (fig. 53), from the
pigeon, is more fully described, and possesses several points of interest. It
also is relatively very wide, but not at all stumpy. Its length is from 45
—60 p, breadth 6—8 ». The posterior part tapers away finely, and there is
a free prolongation of the flagellum. The characteristic feature, however,
is the long, drawn-out and extremely thread-like anterior end. The kineto-
nucleus is, correspondingly, a considerable distance from the anterior end,
and behind it isa vacuole. In this case, also, the trophonucleus stretches
transversely across the body, and is of a curious triangular shape. The
surface of the body exhibits longitudinal striations. This Trypanosome is
most probably quite distinct from Dutton and Todd’s parasite.

THE HA@MOFLAGELLATES. 301

T. padde, Lav. and Mesn., 1904. Length 30-40 p, breadth 5-7 p.
From the description this form (fig. 54) appears to be of a very similar type
to the last. ‘lhe undulating membrane is perhaps rather better developed,
however, and the free part of the flagellum is shorter. ‘The anterior end is
usually very attenuated, sometimes so much so that “it appears as though
there was a flagellum at each extremity.” The cytoplasm shows fine
longitudinal striations. (V.) Host: Padda oryzivora (from the Paris
markets), in whose blood also Halteridium danilewskyi is often
found.!

T. polyplectrum, Vassal, 1905. Length 46 p, of free flagellum 12 p ;
breadth 5. Anterior end extremely attenuated and flagelliform, even
more pronouncedly than in T. paddae. From a pheasant (Poly-
plectrum germani) in Annam.

T. avium, Danil., 1885 (Laveran emend., 1903). Length 35—46 p,
inclusive of flagellum. Body elongated and fusiform in shape (figs. 16 F,
55); anterior end tapering. Undulating membrane well-developed, with
a longitudinal striation or fold running in it. lLaveran found the para-
sites in Syrnium aluco, Tawny Owl. Danilewsky observed Trypano-
somes in various phases (fig. 3) in owls (sp. indet.) and also in “ roller-
birds” (Coracias garrula). Their size is given as from 20—60 4 by 7—8F.
Laveran (86) thinks it not unlikely that the Trypanosome in the rollers is
a distinct species. It is interesting to note that Laveran mentions the
occurrence, in the Tawny Owl which contained T. avium, of “ Hem-
amceba” ziemanni and “H.” (Halteridium) danilewskyi. Is it
possible that the latter Hzmatozoan represents the corresponding phase
of T. avium which the former does of T. ziemanni?

Novy and McNeal’s work (81) on the Trypanosomes of birds is most
difficult of exact estimation from a systematic standpoint for the reasons
already given (pp. 281,282). In only a single instance were their numerous
cultures of the parasites successful in giving rise to Trypanosomes, when
injected into the blood. ‘This, certainly, does not point to the organisms
having been at the time in a very healthy condition; on the contrary,
some of the authors’ excellent photomicrographs strongly suggest abnormal
and involuted phases (cf. their pl. 7)

Novy and McNeal consider they have investigated at least four different
species. A parasite identified as Danilewsky’s T. avium is most frequent.
Two types are recognised (corresponding to that worker’s majus and
minus). One is of less frequent occurrence, and measures 50 by 6p,
the flagellum being 15—20, in addition; the other (more common) is
20 # by 3—5 p, the flagellum 10} extra. This smaller variety is thought

According to Thiroux this association is purely accidental, there being
apparently no connection between these two parasites (see, however, under
T. avium.

3802 H. M. WOODCOCK.

to be the same as Laveran’s form in Syrnium aluco.. In this species,
moreover, the authors would include both Dutton and Todd’s Senegambian
T'rypanosome and Hanna’s Indian one, although keeping the very similar
type T. padde distinct. Cultures produced two forms of the parasites,
wide spindle-like ones and thin spirochetiform ones with very short
flagellum.' The hosts from which this parasite was obtained were:
Agelaius pheniceus, Colaptus auratus, Cyanocitta cristata,
Icterus galbula, Melospiza fasciata, Merula migratoria,
Passer domesticus, Sialia sialis, and Zenaidura macroura.

T. mesnili, Novy and McNeal, 1905. Length 50”; breadth 8 p.
Considered to be distinct by reason of its large size and peculiar shape and
behaviour in cultures. The latter grow very rapidly, and show two types
of cells: small ones (multiplication rosettes) very short and wide (l0—
12 by 6p): and larger ones (20—25 X 4—6) corresponding to the
spirochetiform type above, with, however, a very long flagellum. It is
the latter parasites which form the typical agglomeration rosettes. This
Trypanosome was obtained from Buteo lineatus.

T. laverani, Novy and McNeal, 1905. In size this parasite agrees with
the smaller variety of T. avium, being only a trifle wider. Its specific
distinction is based on cultural forms, which grow very slowly, and do not
give rise to such markedly diverse types as in the above cases. Host:
Spinus tristis.

A form from Cyanocitta cristata and Scolephagus carolinus
is regarded as distinct, and others from Dryobates crilosus, Harpo-
rhynchus rufus and Troglodytis edon gave rise to various sub-
types and strains, which might or might not be distinct.

It only remains to add that Trypanosomes have also been observed in
other Avian hosts, but, for the most part, their occurrence only is men-
tioned. Thus, Sergent, E. and E. (102), have noticed the parasites in various
Algerian birds, e. g. goldfinch (Fringilla (Carduelis) carduelis), linnet
(Sylvia atricapilla), and swallows. Donovan informs Laveran and
Mesnil (56) that he has seen ‘Trypanosomes in the blood of an owl (Athene
brama) in Madras; and the same worker has also observed them in
Milvus govinda (Indian Kite), the parasites being 34 x 3—33 #, free
flagellum 16 (see Thiroux [114]). Lastly, Ziemann? has found a Try-
panosome in a chaflinch (Fringilla celebs).

1 These forms strongly recall T. johnstoni; although Novy and McNeal
think that parasite, again, is distinct.
* «Ueber Malaria und andere Blutparasiten,’ Jena, 1898 (p. 106).

THE H#MOFLAGELLATES. 303

Reptilian Forms.

Trypanosomes have been but rarely found, so far, in Reptiles. The only
one described and figured is—

T. damonia, Lav. and Mesn., 1902. Length 32 p, breadth 4p. The
body (fig. 16 5) is fusiform and fairly wide in the middle, but in general
structure presents nothing remarkable. The parasite often appears rolled
up on itself. ‘I'he chromatic grains in the cytoplasm ‘are more or less
uniformly distributed throughout the body, the posterior end being, if
anything, freer.

In addition, Dutton and Todd (21) mention having observed Trypano-
somes at rare intervals in the blood of tortoises, and Gehrke! has noticed
one in a gecko.

Forms parasitic in Amphibian hosts.

T. rotatorium (Mayer). (Synn. Ameba rotatoria and Parame-
cium costatum or loricatum,? Mayer, July, 1843; Trypanosoma
sanguinis, Gruby, Nov., 1843; Undulinaranarum, Lankester, 1871).
—The great variation in form exhibited by this parasite has been already
discussed. ‘Two principal types are distinguished, one having the surface
of the body thrown into parallel ridges, which run either longitudinally or
with a slightly spiral course, the other having a smooth regular surface.
These two types and the manifold varieties of shape are best realised by a
comparison of figs. 17 4 and B, 56.

In size, the parasites vary from 40—60 - in length, by from 5—40 » in
breadth ; the two dimensions vary more or less inversely, the width being
greatest when the parasites are relatively short, which gives them often an
extremely broad and stumpy appearance. Correspondingly, the anterior
end may be either drawn-out and finely pointed, in the comparatively
narrow forms, or conical, obtuse, or even blunt and rounded, in the stumpy
ones. The undulating membrane is very well developed and thrown
into numerous folds. The free portion of the flagellum is usually com-
paratively short. The kinetonucleus is generally situated some distance
from the non-flagellate or anterior extremity,’ and may be quite close to
the trophonucleus (figs. 17 B, 56 a); sometimes, however, it is fairly

1 * Deutsche med. Wochenschrift,’ 1903, p. 402.

2 From the drawings given, “ Amcba rotatoria” is almost certainly a
‘Trypanosome, the other organism much more doubtfully so; hence the
first-named specific designation.

3 With regard to the bearing of this point upon the correct generic position
of this form see above, p. 288.

304 H. M. WOODCOCK.

near the anterior end. (V) hosts: Rana esculenta, R. temporaria,
R. trinodis (?), and Bufo viridis. (I) host, probably a leech (possibly
an Ixodes??),

The Trypanosome (fig. 57), unnamed, but perhaps a distinct species,
which Laveran and Mesnil (56) figure from Hyla arborea, is of interest,
since, while of the voluminous T. rotatorium type, it has the kineto-
nucleus close to the anterior end, and this occupies, in respect of its shape,
a position midway between the tapering and the bluntly-obtuse forms of
that parasite. The length of this Trypanosome is given as 75 p, its
breadth 7 p.

Dutton and Todd (1. c.) describe two Trypanosomes, both characterised by

Fig. 56. Pig. 57.

Fig. 56.—T. rotatorium (Mayer). 4, ribbed form ; B, smooth
form. x 1000 (about). (After L. and M.)
Fie. 57.—T. sp., from Hyla arborea. (After L. and M.)

their great length, from frogs (sp. incert.) in Gambia. The authors give
each form provisionally a new specific name, T. mega and T. karyo-
zeukton respectively. Both parasites strongly resemble the fusiform
type of T. rotatorium (which was also encountered in the frogs of that
district), so much so that Laveran and Mesnil consider them to be
hardly specifically distinct. Although in these new forms the anterior end
is very long and proboscis-like, these authors say that they have observed
individuals of T. rotatorium with an equally thin and extended anterior
extremity. While the parasites are undoubtedly closely allied to that

1 Durham (lc. [fn. 3, p. 291]) considers that an Ixodes is the Inver-
tebrate host of a Lankesterella (Drepanidium) of toads; see, how-
ever, T. inopinatum, below.

THE H#®MOFLAGELLATES. 305

species, they exhibit certain cytological differences, and T. karyozeukton
possesses one uncommon morphological feature which well differentiates it.
It is, therefore, considered preferable to retain the distinctive names for
both, pending the further investigation of their exact degree of relation-
ship to T. rotatorium.

T. mega, Dutton and Todd, 1903. Length from 82—87 p, the free flagel-
lum being from 10—15 »; the breadth, in the widest part, is about 8». The
kinetonucleus is immediately in front of the trophonucleus, about one third
of the length from the anterior end. ‘The longitudinal ridges and furrows
are well marked, but do not extend towards the anterior end of the body, as
in T, rotatorium (cf. fig. 58), becoming indistinct about opposite the

Fie. 58.—T. mega, Dutton and Todd. ¢ = trophonucleus.
x 1350. (After D. and T.)

nucleus. Thus, while the posterior two thirds of the cytoplasm appears
composed of parallel, alternating darker and lighter bands, the anterior
third seems of a spongy and alveolar nature.

T. karyozeukton, Dutt. and Todd, 1903. Length about 82°5 p, that of
the free flagellum being about 15»; breadth 63. The cytoplasm (fig. 17 D)
shows the same general division into anterior and posterior portions which
characterises the last form. In this species, however, the kinetonucleus is
situated about midway between the trophonucleus! and the proboscis-like

1 According to Dutton and Todd the trophonucleus itself is the pale,
structureless-looking body seen on the right in the figure, the large chromatic
grains in a compact mass being independent. It seems, however, equally
probable that the latter form an integral part of the nucleus, the pale
appearance being perhaps due to shrinkage.

306 H. M. WOODCOCK.

anterior extremity. A noticeable feature is the presence of a chain of
chromatic rodlets (perhaps really a chromatic thread) running from one
nuclear body to the other. The loops of the undulating membrane are not
so wide as in T. mega.

T. inopinatum, Sergent, 1904. Length 25-30 »; breadth 3». The shape
of this parasite is that of a typical Mammalian or Piscine Trypanosome
(fig. 17c). It is very like T. remaki, being slightly wider than T. lewisi,
which, however, it more resembles in having the anterior end finely taper-
ing. The kinetonucleus is well developed, and often stretches transversely
across the body, in a rod-like manner. It is usually situated about midway
between the trophonucleus and the anterior extremity, but may be nearer
either the former or the latter. (V) host, Rana esculenta; (I) host, a
leech, Helobdella algira. Billet’s very important work on the alter-
nation of hosts of T. inopinatum and the parasite’s relation to a Lankes-
terella is discussed above (p. 251).

T. nelspruitense, Laveran, 1904. This is another very distinct and well
characterised form. Average length from 55—60p; breadth 3. The
free flagellum is extremely long, about 25, or more, or almost as long as
the body itself. The body is slender and vermiform in shape. The posterior
and is relatively short and blunt. The trophonucleus lies well in the
anterior half. The cytoplasm in the hinder two thirds of the body is dense
and uniformly filled with deeply-staining grains; that in the anterior third
is much clearer and faintly-staining, with only a few small grains, and
around the nucleus itself there is usually a quite pale zone. This Trypano-
some somewhat resembles T. granulosum of fishes. The name of the
frog in which the parasite was found is not given.

Forms Parasitic in Piscine Hosts.

These Trypanosomes are, on the whole, very uniform in shape, being
typically fusiform and slender. The alternate, Invertebrate host is most
probably, in all cases, a leech. This has already been proved by Léger (66)
for T. barbatule (also for a Trypanoplasma, see above, p. 250);
moreover, various workers have remarked on the general occurrence of
Ichthyobdellids (Hemibdella, Piscicola, Pontobdella) on the skin
of infected fish (vide e.g. L. and M. [51)).

T. remaki, Lav. and Mesn., 1901. his parasite occupies a correspond-
ing position among Piscine forms to that of ‘I. lewisi among Mammalian
Trypanosomes. It is very slender, with tapering, pointed extremities, not
quite so drawn-out, however, as in T. lewisi. The trophonucleus is in the
posterior half of the body, and often shows a large, deeply-staining grain
centrally (centrosome ?j. Laveran and Mesnil distinguish two varieties,

THE H#EMOFLAGELLATES. 307

characterised by difference in size.!. They consider the morphological agree-
ment between the two kinds of form to be, otherwise, too close for them to
be assigned to different species, especially since both occur in the same host,
namely, Esox lucius, pike. T. remaki, var. parva (fig. 59). Length
(medium) about 30/4, of free flagellum alone 10O—12,; breadth 13m. Cer-
tain individuals may, however, reach a length of 42. The cytoplasm is
fairly homogeneous and faintly-staining. T. remaki, var. magna (fig. 171).
Minimum length 45 », of which 17—20 p is for the flagellum; breadth 2—

3H. In the largest forms the length may attain 57. The cytoplasm
appears more deeply-staining than in the other variety, but, as Laveran and
Mesnil point out, this may be partly due to the increased thickness.

Fic. 60.

B.

Fie. 59.—T. remaki, var. parva, Lav. and Mesn. x 2000.
(After L. and M.)

Fic. 60.—a, T. danilewskyi; B, T. tince. x 1500. (After
L. and M.)

We have, next, a series of forms, from closely allied hosts, many of which
are, doubtless, the same as those noticed by earlier observers, e. g. Chalach-
nikow and Danilewsky, but which, for the most part, have been left for
_ Laveran and Mesnil and Léger to rediscover and name.

T. danilewskyi, Lav. and Mesn., 1904. Length 35—45 mu by about 3 u
in width. Free portion of flagellum 15—17 ». The anterior end (fig. 60 4)
is shorter and more bluntly conical than in T. remaki. The undulating
membrane is rather better developed, with wider folds. The nucleus is
rather nearer the posterior than the anterior end. (V.) host, Cyprinus
carpio, carp (probably also in the minnow [Phoxinus], Laveran [39 a]).

1 Laveran and Mesnil say they have not noticed any stages which might
be regarded as intermediate between the two varieties. Nevertheless, as is
apparent from the dimensions given, the maximum size of the smaller
variety closely approaches the minimum of the larger one!

308 H. M. WOODCOOK.

T. tince, Lav. and Mesn., 1904. Average length 35; width 23—3p.
This form (fig. 608) is very similar to the last. The free part of the flagellum
seems to be rather shorter—judging from the figures, its length not being
stated. (V.) host, Tinca tinca, tench.

T. abramis, Lav. and Mesn., 1904. Named, bit not diagnosed. The
authors say the blood of the specimen in which this Trypanosome was seen
was in too bad a condition to permit of examination. From a bream,
Abramis brama.

T. barbatule Lég., 1904.—The body of this parasite is rather wider
(relatively), and more stumpy than usual. Inclusive length 30—40p;
width 4—6p. Free part of flagellum 11—12y. The undulating mem-
brane possesses large and deep folds. The anterior end terminates in a
small beak, about 14 in length. (V.) host, Cobitis barbatula, loach ;
(I.) host, Piscicola,sp., a leech, in the intestine of which Léger has observed

Fic. 61.—a, T. carassii, Mitr.; B, T. cobitis, Mitr. (After
Mitrophanow, from Doflein.)

important evolutive stages of the parasite (see above, p. 250). Léger regards
this form as being distinct from the following one.

T. cobitis, Mitrophanow, 1883. (Synn., Hematomonas c., Mitr,
Trypanosoma piscium and T, fusiforme piscium, in part, Danil.).—
This and the other Trypanosome originally described by Mitrophanow do
not appear to have been studied since, hence the insufficient description and
also the poor figures (fig. 618), which are the only ones available. The
length is given as from 80—40p, and the breadth 1—1} p [?]. The
flagellum alone is 10O—15y. (V.) host, Cobitis fossilis, another loach.

J. carassii, Mitr.1883. (Synn., Hematomonasc., Mitr.,T. piscium
and ‘I’. fusiforme piscium, in part, Danil.).—Larger, but more flattened,
than the preceding form, which, otherwise, it much resembles. This very
slight description is about all that one has upon which to rely for a
diagnosis of this form from Carassius vulgaris, Prussian carp.

T. granulosum, Lay. and Mesn., 1902, is readily distinguishable from the
preceding forms, being more sharply marked off from them by its morpho-
logical characteristics than they are from one another. There can be no

THE HAMOFLAGELLATES. 309

doubt that this parasite, at any rate, belongs to a distinct type. It is very
long and narrow, typically eel-like (fig. 17 kK), and attains a length of 70—
80 pw, with a width of only 2;—3 yp. The free flagellum is long, about 25 p,
and, indeed, the whole locomotor apparatus is well developed. The kineto-
nucleus is large and situated very near the anterior end, which ceases
abruptly in a short but pointed cone. Practically the entire cytoplasm is
filled with large deeply-staining grains, which may surround and obscure
the nucleus. (V.) host, Anguilla vulgaris, eel. The discovery of this
Trypanosome is due to Sabrazes and Muratet (94).

T. clarie, Montel, 1905. Length 60; breadth 4. Ribbon-like.
Anterior end short and sharply conical, with kinetonucleus close to the
extremity. This parasite appears not unlike the last, differing in being
somewhat wider. (V.) host Clarias (Silurus ec.), Cochin China.

Flat-fish appear to have a fair share of these Hematozoa, but only one
form has been figured. Moreover, in nearly every case, a Hemosporidian
has been often found associated with the Trypanosomes, the latter being
usually much rarer than the former.

T. solex, Lay. and Mesn., 1901.—Length about 40 »; width not given.
The free flagellum is only about 8 » long, or relatively short (fig. 175). The
anterior end is fairly tapering, more so than the hinder one, which is some-
what blunt. The spherical kinetonucleus is very large, and occupies the
entire width of the parasite. Fine longitudinal striations are noticeable in
the cytoplasm. (V.) host, Solea vulgaris, sole, which also contained
Hemogregarina simondi. (I.) host, probably Hemibdella solee,
a leech, which was common on all the fish examined.

T. platesswx, Lebailly, 1904—Length 52; width 3—33p. Free
flagellum 12 ». Anterior end finely tapering. Trophonucleus in the
posterior half of the body. (V.) host, Pleuronectes platessa (Pla-
tessa vulgaris), plaice. Associated with a new Haemogregarina,
H. platesse.

T. flesi, Lebailly, 1904.—Length 55; width 5, or rather wider than
T. platesse. Free flagellum 10, long. Kinetonucleus rather nearer
the anterior end than in the preceding species, and trophonucleus about in
the middle. (V.) host, Pleuronectes flesus (Flesus vulgaris),
flounder, which also contained a new Hemogregarina, H. flesi.

T. laterne, Lebailly, 1904.—Length 65 1; breadth 5—6. Flagellum
very short, only about 8 long. (V.) host, Platophryslaterne. Asso-
ciated with He mogregarina laterne, Lebailly.

These three forms of Lebailly, have, it will be gathered, much in common
with each other, and also with T. solew. The relatively short flagellum is
a feature in all.

T. bothi, Lebailly, 1905. Length 42; of flagellum alone 13 p;
width 3. Anterior end thin and tapering. Trophonucleus in the pos-
terior half of the body. This parasite much resembles the next one. (V.)

310 H. M. WOODCOCK.

host, Bothus rhombus (Rhombus levis), brill. Associated with a
new Hemogregarina, H. bothi.

Brumpt and Lebailly (11) have briefly described a number of new Piscine
Trypanosomes, and also, at the same time, several new Haemogregarines,
many from the same hosts as the Trypanosomes.

T. limande, Brumpt and Lebailly, 1904.—An extremely thin, mam-
malian like form, differing from T. platess# in having a much longer
flagellum. Length 45, of flagellum alone 20p. Breadth only 2—25 p.
Anterior extremity very pointed. (V.) host, Limanda platessoides.

T. delagei, Brumpt and Lebailly, 1904.—A shorter form than the fore-
going, but also thin and fusiform. Length 33,, flagellum alone 12 p.
Breadth 23. Anterior part pointed and rectilinear (?). Host, Blennius
pholis.

Fig. 62.—a, T. scyllii; B, T. rajew. x 1200. (After L. and M.)

The remaining forms are larger, and relatively much broader, and agree
with the majority of parasites from flat-fish above described in possessing a
very short flagellum.

T. gobii, Brumpt and Lebailly, 1904.—Length 66 p, of flagellum alone
10. Width 5—53,. Anterior extremity generally somewhat blunt or
rounded. (V.) host, Gobius niger.

T. cotti, Brumpt and Lebailly, 1904.—Length 53, of flagellum 8p.
Width about 5. Anterior end fairly short and rounded. This parasite,
especially in the case of the largest individuals, resembles T. gobii.
(V.) host, Cottus bubalis.

T. callionymi, Brumpt and Lebailly, 1904. Length 70 p, of flagellum
alone about 5 (in some cases up to 8). Breadth 5m. Anterior extremity
long and tapering. Correspondingly, the kinetonucleus is situated some
distance from the end. (V.) host, Callionymus dracunculus,

Lastly, both divisions of the Elasmobranchs furnish hosts for these
ubiquitous parasites.

T. scyllii, Laveran and Mesnil, 1902. This is a very large parasite,
being from 70—75 / long, by 5—6 broad; free flagellum about 14 long.
The body is generally rolled up on itself (fig. 62 4), often forming a com-

THE H#MOFLAGELLATES. Sit

plete circle. It is fusiform in shape, and the anterior end is shorter and
rather blunter than in (say) T. solew, while the posterior extremity is
drawn out and tapering. The undulating membrane has many well-
developed folds. (V.) hosts, Scyllium canicula and 8. stellare (catu-
lus), dogfish.

T. raje, Laveran and Mesnil. This form is, if anything, even larger
than T. scyllii, being from 75—80 » long, by 6 wide. The free part of
the flagellum is about 20 y. The shape and appearance of the parasite
agree in general with that of the last form, but the anterior extremity is,
usually, more tapering, and may, indeed, be very attenuated and proboscis-
like (fig. 6238). (V.) hosts, Rajaclavata, R. macrorhynchus, R.
mosaica, and R. punctata. Laveran and Mesnil consider that the Try-
panosomes found in these different rays all belong to the same species. (I.)
host, probably Pontobdella muricata, of frequent occurrence on infected
rays.

Trypanosomes, probably distinct species, have also been observed in other
Piscine hosts, but not adequately described. Thus Valentin, in 1841,
noticed a Hematozoan parasite in a trout (Salmo fario) which, to judge
from his account, was in all likelihood a Trypanosome, this being the first
recorded observation of such. In addition, Trypanosomes which have still
to be identified have been mentioned at various times as occurring in the
perch, gudgeon, and certain members of the Siluridew (e.g. Macrones
seenghala, M. tengara, Ophiocephalus striatus, and Trichogaster
fasciatus).

APPENDIX.

(a) Doubtful Trypanosomes.

There are one or two parasites which have been relegated to this group of
organisms which appear to be not really Trypanosomes. Thus, there is the
form originally described by Eberth in 1861 from the cecum and ileum of
poultry (hens, geese, ducks, etc.), which was named by Kent (27) T. eberthi.
This parasite is now generally thought to be a Trichomonas,

Much more important is the organism described by Certes (143) in 1882 as
Trypanosoma balbianii. This parasite occurs in oysters and other
bivalves (Ostrea edulis, O.angulata, Tapes decussata, and T, pul-
lastra), where it inhabits the digestive tube, including the crystalline style
(fig. 63). All who have written on this form have agreed that it has no
free flagellum, but possesses, apparently, an undulating membrane (figs. 63 4
and c). Its length may be relatively enormous, from 50 or less up to 150 p,
but it is extremely thin, only from 1—3p wide. The two most recent

312 H. M. WOODCOOK.

accounts are those of Laveran and Mesnil (145 a) and Perrin (146). According
to both, the nueleus consists of numerous small chromatic masses, having the
form of grains or rodlets (c) transversely arranged, and extending almost from
one end of the body to the other. These rodlets are about equidistant from
one another and separated by faintly-staining spaces. Further, according to
Perrin, these transverse bands are arranged in a single row upon a delicate,
spirally-wound thread or axis, the spiral being very flat where the chromatin
rodlets are.

The two accounts differ, however, concerning the other characteristic
feature of this parasite. Laveran and Mesnil consider that what appears to

Fic. 68.—Spirocheta (Trypanosoma) balbianii (Certes).
(B, a number of individuals associated with the crystalline style of
a cockle.) A and B after Certes, c after Perrin.

be an undulating membrane is, in reality, a wide, “ periplastic” sheath or
investment, which may be attached only at the two ends, the greater part of
the body of the organism being more or less free inside it. In certain cir-
cumstances, especially if flattened or ribbon-like, and having regard to the
spiral form of the body, this sheath would simulate the appearance of an
undulating membrane. Perrin, on the other hand, regards this structure as
a true undulating membrane, comparable to that of a Trypanosome, and
believes, in addition, that it possesses a thickened chromatic border, connected
to one end of the nuclear spiral by a delicate thread.

It seems most likely that multiplication is by longitudinal, rather than by
transverse fission. Perrin describes the process as commencing by the
division of the undulating membrane ; this is followed by the transformation

THE H#EMOFLAGELLATES. ole

of the chromatic band into rounded granules (chromosomes) first arranged in
one longitudinal row, and subsequently, by division, in two. The final division
of the cytoplasm is slow, and the two halves may remain united at one end
for some time before separating, thus giving the impression of transverse
division.

Perrin regards the above form of the parasite as representing the “ indiffe-
rent” type. In addition, he describes ‘‘ female” forms and “ male gametes ;”
the latter result after a kind of maturation-process, large hernias being formed
at the side of the body, by means of which an expulsion of chromatic material
takes place. Lastly, a process of encystment is described in the ‘indifferent ”
and ‘female’ forms.

Laveran and Mesnil came to the conclusion that this parasite is not a

B. Gc

Fie. 64.—(a) Spirocheta plicatilis, Ehrenberg; (s) Spiro-
cheta refringens, Schaudinn; (c) Spironema_ pallidum
(Schaud.). (After Schaudinn.) [In B the central axis (drawn
black) represents both endoplasm and nuelear core; in a, the spiral
axis (also black) is the nucleus alone, the surrounding endoplasm
not being distinctly indicated in Schaudinn’s figure. ]

Trypanosome but a Bacterium allied to Spirocheta,! in Ehrenberg’s original
sense of the term (145) ; and, it may be here mentioned, Léger, who has recently
studied this organism, in a note to the writer expresses the same opinion.
On the other hand, Perrin is confident of the essential Trypanosome nature of
the parasite, and, while recognising its resemblance to Bacteria in nuclear
structure, etc., sees in it a representative of the ancestral Hemoflagellate,
somewhat on the lines of Schaudinn’s bipolar “ Urhemoflagellate ’’ (see foot-
note, p. 267).

1 The correct way of spelling this name is Spirocheta, not Spiro-
chete; vide Ehrenberg (I. c.).

VOL. 50, PART 2,—NEW SERIES. 22

314 H. M. WOODCOCK.

Quite recently, Schaudinn (148) has published a brief note on certain
“Spirochete,” which is of great assistance in deciding between these two
views. Part of a long individual of S. plicatilis, showing its form and
structure, is reproduced in fig. 64 4; on comparing it with fig. 63 a and c,
the general agreement between the two forms appears undoubted. Schaudinn
describes a well-developed periplastic [ectoplasmic] undulating membrane en-
closing the endoplasmic axis of the body; in the latter lies the nuclear appa-
ratus, which, here also, has the form of a (fairly thick) thread on which is
suspended a single row of large chromatic grains. The type of nuclear
structure is manifestly the same as in “TT.” balbianii, but the spiral is much
more condensed. When, in addition, the rounded termination of the body
and the absence of any flagellum are noted, it seems obvious that in whatever
group of organisms we place Spirocheta plicatilis we must also include
<t. balbianit.

(s) Are the Spirochete Trypanosomes, ie. Fla-
gellates, or Bacteria?

This question has been much discussed during the last
year or two. In Schaudinn’s great memoir he regarded
Trypanosoma ziemanni as possessing, in certain phases,
the actual characteristics of a Spirocheta as then known ;
and, further, he was inclined to consider that other Spiro-
chete (e.g.S. obermeieri of relapsing fever) were also
really only phases in the life-cycle of a particular Heemo-
flagellate. Asa result of his more recent investigations on
Spirocheta, however, he finds that this is not the case,
but that the latter type of organism is essentially different
in character from a T'rypanosome; T. ziemanni, at certain
periods of the life-history, merely simulates a Spirocheta
to a remarkable degree (see above, p. 244).

The chief reasons for considering Spirocheta as a
Bacterial form rather than a Flagellate may now be briefly
discussed, since they bear upon the contrary view taken
below with regard to Spironema pallidum.

(a) The bluntly-rounded ends, and the absence of any
indication of a flagellum in Spirocheta are characters
whose importance hardly needs further emphasis.

(b) The “undulating-membrane” of a Spirocheta,

THE H®PMOFLAGELLATES. S115

although, doubtless, comparable functionally with the simi-
larly-named organella of a 'rypanosome, seems most probably
not homologous with the latter as regards origin and struc-
ture. Although it is generally well-developed (cf. fig. 64 4,
B), Schaudinn in no instance mentions the observation of any
thickened edge comparable with the flagellar border of the
membrane of a Hemoflagellate.?

On the other hand, again, the more Spirocheta-like a
Trypanosome is, the narrower and less developed is the
undulating-membrane ; indeed, in such cases the flagellar
border is the most conspicuous part of this organ-
ella, the membrane being often recognisable only with
difficulty (cf. T. ziemanni [fig. 32], T. johnstoni [fig.
51], other slender Trypanosomes, and, lastly, Spironema
pallidum, below).

The writer is more inclined to Laveran and Mesnil’s idea
of the “membrane” of a Spirocheta as a general ecto-
plasmic investment surrounding the body (see above, p. 312) ;
such a sheath might well appear, as in Schaudinn’s figures.

(c) The character of the nucleus in both 8. balbianii and
S. plicatilis differs considerably in two respects from that
of a Trypanosome. In the first place, it is (particularly in S.
balbianii) very diffuse, more akin, in short, to the “ distri-
buted”’ type associated with Bacteria, and, in the second
place, there is not the well-marked distinction between
kinetic and trophic chromatic elements which is so charac-
teristic of a Hemoflagellate. The writer does not think,
with Perrin and Schaudinn, that the axial thread represents
the locomotor nucleus of a Trypanosome. It may be kino-
plasmic, but it certainly appears to serve as a ground-work

1 As mentioned above, Perrin considers there is such a chromatic border
in “T.” balbianii; but neither Laveran and Mesnil nor Léger noticed such
a feature in their examination of this form. The writer cannot help thinking
to some extent with Mesnil that Perrin has been rather too preoccupied with
the idea of realising, in this parasite, the ancestral Trypanosome. Mesnil
considers that some of the stages described by Perrin (e. g. the formation of

“male gametes” and the encystment process) represent really involution
forms.

316 H. M. WOODCOCK.

or support (reticulum ?) for the general chromatin, which is
not the case with the kinetonucleus in a Trypanosome. In
this connection we may refer to the nuclear structure of a
microgamete of Try panomorpha (see above, p. 197) where
the trophonucleus is also greatly elongated, the chromosomes
being similarly suspended on an axial thread. But the
kinetonuclear elements are quite distinct and compactly
arranged. Up till now no sign of such a kinetonuclear
body has been observed in a true Spirocheta (cf. on the
other hand, Spironema, below).

To sum up, we may, it seems to the writer, agree with
Schaudinn (1) that the organisms exemplified by Spiro-
cheta plicatilis are to be widely separated from Try-
panosoma ziemanni (and equally, of course, from other
Trypanosomes), with which they have, at most, only a very
remote phylogenetic connection; and (2) that, at any rate,
certain other spirilliform parasites, e.g. S. refringens (fig.
64 8) and S. ('T.”) balbianii, agree fundamentally in struc-
ture with 8S. plicatilis, the type species. Where, exactly,
among micro-organisms Spirocheta is to be placed, e.g.
how close to, or distant from, Spirillum, need not be dis-
cussed here ; it is sufficient to say that the available evidence
appears to be against this form having anything to do with
the parasitic Flagellates.

(c) Spironema pallidum, Schaudinn= Trypanosoma
luis, Krzysztalowicz and Siedlecki.

In the spring of last year Schandinn and Hoffmann pub-
lished a note (150) on the discovery by the former of spirilli-
form organisms, greatly resembling Spirochete, in
syphilitic lesions. Two principal kinds of parasite are
distinguished. One, which Schaudinn called Spirocheta
refringens, is strongly refractive in life, and of a somewhat
coarse or rough appearance; the other, named by him §.
pallida, is extremely delicate and very feebly refractive.

THE HEMOFLAGELLATES. 317

From this, and later papers (149 and 151) by the same authors,
it appears that 8. refringens is not met with in pure
syphilitic lesions; S. pallida, on the other hand, occurs
regularly both in the primary affections and in the secondary
eruptive formations, as well as in the internal products and
developments associated with the disease.

It can, therefore, be said that there is more probability of
S. pallida having some causal relation to syphilis than there
is of 8. refringens.

Besides the above, other differences between the two
parasites, with regard to behaviour and structure, are to be
observed. The spiral! of S. pallida is much narrower, more
acute, and composed of many more turns” than that of 5.
refringens (cf. fig. 64.8, c) ; it is, in short, more like a cork-
screw than a series of curves, thus differing considerably
from the usual type of spiral ina Spirocheta. Moreover,
the former parasite is very difficult to stain by the methods
used in other cases, while 8. refringens stains readily.

In his later comparative note on Spirochwte, Schaudinn
(1. c.) shows that 8. pallida differs also in more fundamental
respects from other Spirochete. In fixed and stained pre-
parations he was unable to demonstrate the presence of the
undulating membrane which is so apparent in other forms
when similarly stained (cf. figs. 63, 64) ; although from the
movements of the parasites during life, he is inclined to sus-
pect the existence of one, doubtless extremely thin. Again,
the body tapers finely at both ends, and Schaudinn believes
there is really a long delicate flagellum at each extremity
(fig. 64c). At times he observed individuals with two flagella
at one end, probably indicative of longitudinal division, For
these reasons Schaudinn agrees with Vuillemin (152), who
separates 8. pallida both from Spirocheta on the one

1 Its dimensions are as follows: length, from 4-14 w; breadth extremely
thin, +»; the number of turns is given as from 6-14, but in the later notes
as from 10-26.

2 When in its “spirochetiform” phase, ought perhaps to be added (see
below).

318 H. M. WOODCOCK.

hand and from Spirillum on the other, as the type of a new
genus Spironema.

A few weeks subsequently to Schaudinn’s paper Krzysztalo-
wicz and Siedlecki published a remarkable account (147) of
this same parasite. While agreeing as to its finely tapering
and pointed appearance, which furnishes a ready means of dis-
tinguishing this organism from S. refringens, these authors
consider that this is not due to the presence of flagella, but
to the body itself being drawn out like a filament at both

Fic. 65.—Spironema pallidum (Schaud.) = Trypanosoma
luis, K. and 8S. a, indifferent (Spirocheta-) forms; ¢ and d,
stages in division of same; 4, an indifferent form shortened and
thickened, leading on to the ‘‘‘Trypanosome”’-phase (e); 4, elon-
gated compound (multinuclear) form, which divides into several
little sickle-shaped uninuclear individuals (g); 4, conjugation of
male and female elements. (After Krzysztalowicz and Siedlecki.)

ends. K.and&8. find that, in the course of its movements,
the parasite may contract itself considerably, and become
much shorter and thicker (fig. 65 b.) ; even in this condition,
however, one or both extremities remain attenuated and
pointed. At such times Spironema greatly resembles a
little Flagellate such as Micromonas. In both these
phases complete longitudinal fission takes place (fig. 65) ;
the parsites may remain for some time attached end-to-end

THE HAMOFLAGELLATES. 319

(fig. 65d), and even divide again while so united [cf. Try-
panosoma ziemanni].

So far these authors have been also unable to discern an
undulating-membrane. They have observed, however, cer-
tain other interesting details of minute structure. In the
most successful preparations of spirochetiform parasites the
body appears almost severed in two near the middle, this
being due to the occurrence of a small oval clear area (fig.
65 a,c,and d). ‘This structure is regarded as the nucleus,
apparently very poor in chromatin, and resembling an empty
space; K. and 8S. compare it with the nucleus of a Coccidian,
where the chromatin is contained mostly in the karyosome
and the nuclear membrane.

In grave or neglected syphilitic lesions forms were en-
countered which resembled the contracted spirals, but which
were relatively wider and more fusiform. In these indi-
viduals (fig. 65 ¢) one end of the body is drawn out as a long
fine filament, while the other is shortand sharply conical. In
a varying position between the latter extremity and the middle
of the body lies the relatively large ovoid vesicle, with well-
defined contours, which is considered to be the nucleus. At
one side of the nucleus is a distinct, deeply-staining grain or
corpuscle. The whole of the body stains uniformly, but,
nevertheless, in large “ Spirilla,”’ a deeply-staining filament
can be recognised, starting at the corpuscle just alluded to,
and running superficially in a somewhat sinuous manner
(cf. fig.). This structure is regarded by the authors as
entirely comparable with the flagellar border of the undu-
lating-membrane of a ‘l'rypanosome, the corpuscle represent-
ing a kinetonucleus.

From these facts K. and 8. conclude that, at certain periods
of its existence, Spironema pallidum possesses a try-
paniform phase. They consider that the ‘ Trypanosome-
forms” result from the growth of the spirocheetiform parasites
(in the retracted condition). Probably all the organelle
visible in the trypaniform phase are also present in the spiro-
cheetiform one, although mostly too fine to discern and study

320 H. M. WOODCOCK.

They propose the name Trypanosoma luis? for the para-
sites when in the former condition.

K. and 8. also bring forward observations pointing to the
occurrence of sexual forms and conjugation. The wider try-
paniform parasites are themselves considered to represent
female elements. In the same preparations showing these
are also to be found peculiar spirals differing from the ordinary
ones in being much longer and relatively wider, and in pos-
sessing several “nuclei” each having a corpuscle in relation
with it (fig. 65/.). By the side of these elongated multinuclear
forms may be seen also very fine and minute ones, about 3 pu
long (g.). Hach is falciform with pointed ends, possesses
one “ nucleus” and “ corpuscle,” and appears identical with
one of the turns of the compound spiral. These tiny forms
correspond to male elements and probably result from the
fragmentation? of the others. Boths kinds of gamete may
undergo longitudinal division, the males thus becoming ex-
tremely delicate and difficult to perceive (ef. T. ziemanni,
p- 243].

Only in one case were K. and 8. able to find stages which
resembled the process of conjugation, in a preparation froma
large primary ulceration which was commencing to cicatrise.
A male gamete becomes attached to a female gamete by one
of its extremities (fig. 65 h.), and the two gradually join to-
gether (laterally) and at length fuse completely. The subse-
quent development was not followed. The authors think that
probably a resting-period ensues in which the parasite may
become encysted, and this in turn gives rise to the indifferent

1 Kven if this parasite is ultimately shown to be a true Trypanosoma,
and the generic name Spironema is, in consequence, retired, the specific
name must still be pallidum. Pending further investigation, however, it
seems best to retain the generic name also, as although probably a Trypano-
some, this form may differ somewhat from the genus Trypanosoma.

2 The authors consider that this apparent transverse division is easily
derivable from the successive longitudinal fission of the “Spirochete,” by
supposing that the parasites remain all closely attached end-to-end, become
somewhat modified (condensed), and ultimately separate into the constituent
units.

THE H@!MOFLAGELLATES. 321

(“Spirochata”) typeof form again. Various enigmatical,
rounded, and kidney-shaped bodies were observed, which
might possibly belong to the life-cycle, but as no intermediate
stages were seen, this remains uncertain.

Brief comment may be made upon one or two points in the above descrip-
tion. The structure regarded by Krzysztalowicz and Siedlecki as a nucleus
cannot be said to resemble greatly that of a Trypanosome. In the trypani-
form parasites, at all events, it rather recalls the cytoplasmic vacuole of a
‘Trypanosome, both by its appearance and its relation to the kinetonuclear
grain.

Moreover, in one long irregular parasite figured (probably an abnormal “ com-
pound spiral’”?) some of the “nuclei” certainly have a distinctly vacuolar
appearance. However, Krzysztalowicz and Siedlecki think that in certain
dividing forms, where the “nucleus” is not apparent as a vacuole, it is repre-
sented by a deeper-staining region of the body, its enhanced chromaticity at
this time being due to nuclear changes. The solution of this point must
await further research.

Again, with regard to the occurrence of a flagellum, Krzysztalowicz and
Siedlecki do not think that the spirochetiform parasites have one. Schaudinn,
on the other hand, thinks there is one at each end of the body. In some of
the former authors’ figures of retracted and trypaniform individuals, the
appearance of the body at one end strongly suggests the presence of a
flagellum (cf. fig. 65 e); and this, moreover, at the end towards which the
sinuous, deeply-staining line runs, which is homologized with a flagellar
border, and which may very likely be continued (in some phases, at any rate)
as a free flagellum. It is more difficult to judge in the case of the “ Spirilla.”
In one or two very spirilliform Trypanosomes (e. g. T. ziemanni, I’. john-
stoni) the free flagellum is very short or absent. Supposing, however, there is
a flagellum at one end, it is hardly likely (Gf Spironema is a Trypanosome)
that there is also one at the other; as a matter of fact, in certain indubitable
Trypanosomes, the non-flagellate end of the body is often almost, if not quite,
as fine as the flagellum itself (cf. T. ziemanni, Hanna’s Trypanosome of
Indian pigeons [fig. 53], and also T. polyplectrum).

However this may be, there can be little doubt, from the
above account, that Spironema pallidum is allied, not to
the Spirochete, but rather to the Flagellates, possessing,
as it does, in certain phases at least, markedly trypaniform
characters. It is clear, moreover, that the parasites hitherto
grouped together under the general heading of “Spiro-
chet’ include, really, two quite distinct types of organism.

322 H. M. WOODCOCK.

Further careful investigation of all (including, e.g. “8S.”
obermeieri and “8S.” anserina) is necessary, with a view
to their correct sorting and classification.

In conclusion, the occurrence of Spironema (Trypano-
soma) pallidum in syphilis is extremely suggestive when
the very similar biology of Trypanosoma equiperdum
and its causal relation to Dourine (horse syphilis) are borne
in mind (cf. above, p. 176). The former parasite has un-
doubtedly been evolved along similar lines to the latter.
Probably, indeed, the two are not distantly connected,
although Spironema is apparently more modified, and
almost certainly now restricted to the Vertebrate host. —

BIBLIOGRAPHY.

[This list is intended to be a guide to the principal papers
in connection with the zoological and biological study of the
parasites published up to the end of January, 1906.]

A. Relating to the Trypanosomes.

1. Berrencourt and Franga, C.—“Sur un ‘Trypanosome du blaireau
(Meles taxus, Schreib.),”’ *C. R. Soc. Biol.,’ 59, p. 805, 1905.

“Trypanosome de la chauve-souris,” f. c., p. 806, 1905.

ivy)

. Bitvet, A.—“* Culture d’un Trypanosome de la grenouille chez une
Hirudinée ; relation ontogénique possible de ce Trypanosome avec une
Hémogregarine,” ‘C. R. Ac. Sci.,’ 139, p. 574, 1904.

“Sur le Trypanosoma inopinatum de la grenouille verte

d’Algerie et sa relation possible avec les Drepanidium,” ‘C. R.
Soc. Biol.,’ 57, p. 161, 16 figs., 1904.

5. Bosc, F. J.— Recherches sur la structure et l’appareil nucléaire des
‘Trypanosomes (a propos d’un Trypanosome observé cliez le lapin),”
‘Arch. Protistenk.,’ 5, p. 40, 68 figs., 1904.

6. BRrapForD, J. R., and Puummer, H. G.—“ The Trypanosoma brucii,
the organism found in Nagana or the Tsetse-fly disease,” ‘Q. J. Micr.
Sci.,’ 45, p. 449, 2 pls., 1902. [This species was originally named and
described in ‘P. R. Soc.,’ 65, p. 274, 1899. ]

7. Brucr, D.—‘‘ Note on the discovery of a new Trypanosoma,”

‘ Lancet,’ 1902, i, p. 664, and ‘ Nature,’ 65, p. 454, 1902.

THE H®MOFLAGELLATES. 323

8. Bruce, Nasarro, D., and Greic, HK. D.—[‘‘ Various Reports on
sleeping-sickness and other trypanosomoses in Uganda”’], ‘ Roy. Soc.
Comm.,’ Nos. 1, 4, and 5, 1903—1905.

9. Brumpr, E.—‘‘ Maladie du sommeil et mouche tsé-tsé,” ‘C. R. Soc.
Biol.,’ 55, p. 889, 1903,

9a. “La maladie désignée sous le nom d’Aino par les Somalis de
l’Ogaden est une trypanosomose probablement identique au Nagana
de |’Afrique orientale,’’ op. cit., 56, p. 673, 1904.

10. ‘Contribution a l’étude de l’évolution des Hémogregarines et
des Trypanosomes,” op cit., 57, p. 165, 1904. ;

ill and Lxesartiy, C.— Description de quelques nouvelles espéces
de Trypanosomes et d’Hémogregarines parasites des Téléostéens
marins,” ‘C. R. Ac. Sci.,’ 139, p. 613, 1904.

12. and Wurtz.—* Agglutination du Trypanosoma castellanii,
Kruse, parasite de la maladie du sommeil,” ‘C, R. Soc. Biol.,’ 55, p.
1555, 1903.

13. [‘‘ Experimental sleeping-sickness among various

animals], ‘C. R. Soc. Biol.,’ 56, pp. 567—573, 1904.

14, Burrarp, M., and ScunemDeR, G.—‘ Le Trypanosome de la Dourine,”
‘Arch. parasitol.,’ 3, p. 124, with pls., 1900.

15. Castretuani, A.—“ Trypanosoma and sleeping-sickness,” ‘ Reports
S. S. Comm. Roy. Soc.,’ Nos. 1 and 2, 1903.

15 a. CHALACHNIKOW.— Recherches sur Jes parasites du sang chez les ani-
maux a sang froid et a sang chaud,’ Kharkoff, 1888.

16. CrooksHank.—* On T, (Trichomonas) sanguinis evansi,” ‘J. R.
Micr. Soc.,’ 6, 1886.

17. Danitewsky.—‘‘ Zur Parasitologie des Blutes,” ‘Biol. Centrbl.,’ 5, p.
529, 1885.

‘Recherches sur la parasitologie comparée du sang des oiseaux,’
Kharkoff, 1888-89.

19. Dortern, F.—‘ Die Protozoen als Parasiten und Krankheitserreger
nach biologischen Gesichtspunkten dargestellt,’ Jena (G. Fischer),
1901.

20. Durron, E.—‘ Note on a Trypanosoma occurring in the Blood of
Man,” ‘ Brit. Med. Journ.,’ 1902, ii, p. 881, 1 fig.

and Topp, J. L.—‘ First Report of the Trypanosomiasis Expe-
dition to Senegambia (1902),” ‘Mem. Livp]. Sch. Trop. Med.,’ No. II
(‘ Thompson-Yates Lab. Rep.’ [5] 2), 1903.

22. EumasstaAn, M.,and Miconr, H.—‘‘Sur le Mal de Caderas ou Flagellose

parésiante des Equidés Sud-americains,” ‘Aun. Inst. Pasteur,’ 17, p.
241, 1 pl., 1903.

18.

2.

324 H. M. WOODCOCK.

23. Gatii-VALeRIO, B.—‘ Untersuchungen iiber die Hamosporidien der
Alpen- Vogel,” ‘ Centrbl. Bakter.,’ (1), 31, Orig., p. 162, 1902.
“Notes de parasitologie,” ‘ Centrlbl. Bakter.,’ (I), 35, Orig., p.
81, 1903.

24a. Gray, A. C., and Tustocu, F. M.—‘‘The Multiplication of the
Trypanosoma gambiense in the Alimentary Canal of Glossina
palpalis,” ‘ Rep. 8.8. Comm.,’ No. 6, p. 282, 1 pl., 1905.

24b. Grize, E. D., and Gray, A. C.—“ Continuation Report on Sleeping-
sickness in Uganda,” t. c., p. 1, with plates, 1905.

24.

25. Grusy.— Recherches et observations sur une nouvelle espéce d’ Héma-
tozoaire (Trypanosoma sanguinis),” ‘C. R. Ac. Sci.,’ 17, p. 11384,
1843, and ‘ Ann, Sci. Nat.,’ (3), i, p. 105, 7 figs., 1844.

26. Hanna, W.—“* Trypanosoma in Birds in India,” ‘Q. J. Mier. Sci.,’
47, p. 433, 1 pl., 1903.

27. Kent, SavitLE.—‘ A Manual of the Infusoria,’ vol. i, 1880-1.

28. Krysseitz, G.—‘ Uber Trypanophis grobbeni (Trypanosoma
grobbeni, Poche), ‘ Arch. Protistenk.,’ 3, p. 367, 3 figs., 1904.

29. Lapp, A.— Note sur un nouveau parasite du sang (Trypanomonas
danilewskyi),” ‘ Bull. Soc. zool. France,’ 16, p. 229, 1 fig., 1891.

80. Lanxester, E. R.—“ On Undulina, the type of a New Group of
Infusoria,” ‘Q. J. Mier. Sci.,’ 11, p. 387, 4 figs., 1871.!

31. “The Sleeping-sickness,” ‘Quart. Rev.,’ July, 1904, p. 113, 7

figs.
82. Laveran, A.—‘‘ Sur un nouveau Trypanosome des Bovidés,” ‘C. R. Ac.
Sci.,’ 134, p. 512, 1902.

33. “Au sujet de deux Trypanosomes des Bovidés du Transvaal,”
op. cit., 1385, p. 717, 5 figs., 1902.

34. “Sur la piroplasmose bovine bacilliforme,” op. cit., 136, p.
648, 18 figs., 1903.

35. “ Sur deux Hippobosques du Transvaal susceptibles de propager
Trypanosoma theileri,” ‘C. R. Soc. Biol.,’ 55, p. 242, 1903.

36. “Sur un Trypanosome d’une chouette,” t. c., p. 528, 2 figs.,
1903.

37. ‘Contribution a l’étude de Hemameba ziemanni,” t. c.,
p- 620, 7 figs., 1903.

38. “Sur lagent pathogéne de la trypanosomiase humaine, Tr.
gambiense, Dutton,” ‘C. R. Ac. Sci.,’ 138, p. 841, 1904.

39. **Sur un nouveau Trypanosome dune grenouille,”’ ‘O, R. Soc.

Biol.,’ 57, p. 158, 2 figs., 1904.

THE HM#MOFLAGELLATES. 325

39 a.

“Trypanoplasme et Trypanosome du vairon,” T. c., p. 250,
1904.

and Mzsnit, F.—“ Recherches morphologiques et expérimentales
sur le Trypanosome des rats, Tr. lewisi (Kent);” ‘Ann. Inst.
Pasteur,’ 15, p. 673, 2 pls., 1901.

40.

41. ‘Sur la morphologie et la systématique des Flagellés a
membrane ondulante (genres Trypanosoma et Trichomonas,
Donné),”’ ‘C. R. Ac. Sci.,’ 183, p. 181, 3 figs., 1901.

42. “Sur les Flagellés & membrane ondulante des poissons
genres Trypanosoma, Gruby, et Trypanoplasma, n. gen.),”
t.c., p. 670, with figs., 1901 [fuller account in 47].

43. **Sur le mode de multiplication du Trypanosome du
Nagana,” ‘C. R. Soc. Biol.,’ 53, p. 326, with figs., 1901 [see also
48].

44, “Sur la nature centrosomique du corpuscle chromatique
postérieur des Trypanosomes,” t. c., p. 329, 1901.

45. **Sur la structure du Trypanosome des grenouilles et sur
Pextension du genre Trypanosoma Gruby,’’ t. c., p. 678, 3 figs.,
1901.

47. “ Des Trypanosomes des poissons,” ‘ Arch. Protistenk.,’
1, p. 475, 15 figs., 1902. [For brief accounts see 42, 49, and 51. ]

48. “Recherches morphologiques et expérimentales sur le
Trypanosome du Nagana ou maladie de la mouche tsé-tsé,” ‘ Ann.
Inst. Past.,’ 16, p. 1, 18 figs., 1902.

49. ‘Sur le mode de multiplication des Trypanosomes des
poissons,” ‘C. R. Ac. Sci.,’ 184, p. 1405, 1902.

50. “Sur quelques Protozoaires parasites d’une tortue
d’Asie (Damonia reevesii),” op. cit., 135, p. 609, 14 figs. 1902.

51. “Sur les Hématozoaires des poissons marins,” t. c.,
p. 567, 1902.

52. “Le Nagana et le Mal de Caderas sont deux entités
morbides bien distinctes,” t. c., p. 838, 1902.

58. ‘Le Nagana le Surra et le Caderas constituent trois
entités morbides distinctes,” op. cit., 136, p. 1529, 1903.

54. “Sur un Trypanosome d’Afrique pathogéne pour les
Equidés, T. dimorphon, Dutton et Todd,” op. cit., 138, p. 732, 7
figs., 1904.

54a, * Sur le Surra et le différenciation des trypanosomiases,””
op. cit., 140, p. 831, 1905.

55. “ Infections naturelles de rats blancs par Trypanosoma

lewisi,” ‘C. R. Soc. Biol.,’ 57, p. 247, 1904.

326 H. M. WOODCOCK.

56. Laveran and Mesnit.—‘‘ Trypanosomes et trypanosomiases,” Paris
(Masson et Cie), 1904.

57. Lepartiy, C.—“ Sur quelques Hémoflagellés des Téléostéens marins,”
‘C. R. Ac. Sci.,’ 139, p. 576, 1904.

‘Sur des Hématozaires nouveaux parasites de la barbue

(Bothus rhombus L.),” ‘C. R. Soe. Biol.,’ 59, p. 304, 1905.

59. Lecrr, L.—‘“ Sur la systématique des Cercomonadines aciculées san
membrane ondulante,”’ ‘C. R. Ac. Sci.,’ 134, p. 665, 1902.

58.

60. “ Sur la structure et le mode de multiplication des Flagellés du
genre Herpetomonas, Kent,” t. c. p. 781, 7 figs., and ‘C. R.
Soc. Biol.,’ 54, p. 398, 7 figs., 1902.

61. “Sur un Flagellé parasite de ’Anopheles maculipennis,”
‘C. R. Soc. Biol.,’ 54, p. 354,10 figs., 1902.

62. “Sur la forme gregarinienne des Herpetomonas,” t. ¢., p.
400, 1902. [See 60 for figs. ]

63. “Sur quelques Cercomonadines nouvelles ou peu connues
parasites de l’intestin des Insectes,” ‘ Arch, Protistenk.,’ 2, p. 180, 4,
figs., 1903.

64. “Sur la morphologie du Try panoplasma des vairons,” ‘C. R.
Ac. Sci.,’ 138, p. 834, 1904. [For figs. see 65.]

65. “Sur la structure et les affinities des Trypanoplasmes,” t. c¢.,
p. 856, 5 figs., 1904.

66. “Sur les Hémoflagellés du Cobitis barbatula, L. L, Try-
panosoma barbatule, n. sp.,” ‘C. R. Soc. Biol.,’ 57, p. 344, 1904.

67. “Trypanoplasma varium, un. sp., parasite du sang de
Cobitis barbatula, L., f.¢., p. 345, 1904.

68. “Sur un nouveau Flagellé parasite des Tabanides,” t. ¢., p. 613,
6 figs., 1904.

69. “Sur les affinités de !Herpetomonas subulata et la phylo-
genie des Trypanosomes,” t. c., p. 615, 1904.

70. “Sur la presence d’un Trypanoplasma intestinal chez les

poissons,”’ op. cit., 58, p. 511, 1905.

A

71. Lignibres, J.—<Contribution & l'étude de la trypanosomose des
Equidés Sud-americains connue sous le nom de Mal de Caderas
(Trypanosoma elmassiani),” ‘ Rec. Med. vet., Bull. et Mem.’ (8),
10, pp. 51, 109, 164, 2 pls., 1903.

72. Mac Catitum, W. G.— On the Hemosporidian Infection of Birds,” ‘J.
Exp. Med.,’ 3, p. 117, 1 pl., 1898.

73. Martini, E.—“ Uber die Entwickelung der ‘'setseparasiten in Siuge-
thieren,” ‘ Zeitschr. Hygiene,’ 42, p. 841, 1 pl., 2 figs. 1903.

THE HE MOFLAGELLATES. 327

73a. Maver, A. F.—*“ Spicilegium observationum anatomicarum de organo

74.

75.

76.

LWe

78.

79.

80,

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

electrico in Raiis anelectrias et de Hematozois,” Bonn, 18738, 18 pp.,
3 pls.

McNzat, W. J.—“On the Life-history of T. lewisi and T. brucii,”
‘J. Inf. Diseases,’ 1, Nov., 1904.

Mincuin, E. A.—‘ The Sporozoa,” Section K in ‘ Lankester’s Treatise
on Zoology,’ 1, p. 150, 127 figs., 1903.

MitropHanow.—Beitrige zur Keuntniss der Hiamatozoen,” ‘ Biol.
Centrbl.,’ 3, p. 35, 2 figs., 1883.

Montet, R.—Trypanosome d’un poisson de Cochinchine,” ‘C. R.
Soe. Biol.,’ 58, p. 1016, 1 fig., 1905.

Nepvev, G.—‘‘Sur un Trypanosome dans le sang de l’homme,” op.
cit., 50, p. 1172, 1898.

Novy, F. G., and McNrat, W. J.—‘‘On the Cultivation of Trypano-
soma lewisi,”’ ‘Contrib. Med. Research, Michigan,’ p. 549, 1903.
**On the Cultivation of Trypanosoma brucii,” ‘J.

Inf. Diseases,’ 1, part 1, 3 tables, 1904.
“On the Trypanosomes of Birds,” op. cit., 2, p. 256, 11

pls., 1905. ;
Perris, G. F.— A Note on the Occurrence of a Trypanosome in the
Rabbit,” ‘Centrbl. Bakter.,’ (1), 35, Orig., p. 484, 1904.

‘Observations relating to the Structure and Geographical Dis-
tribution of Certain Trypanosomes,” ‘J. Hygiene,’ 5, p. 195, 1905.

Pienn, M.—“Trypanoplasma cyprini, n. sp.,” * Arch. Protistenk,’
2, p. 175, 1 pl., 19038.

Purmmer, H. G.— Note on the Effects produced on Rats by the Try -
panosomata of Gambia-fever and Sleeping-sickness,” ‘P. Roy.
Soc.,’ 74, p. 388, 1905.

Pocueg, k.—‘“‘ Uber zwei neue in Siphonophoren vorkommende Flagel-
laten,” ‘ Arb. Zool. Inst. Wien,’ 14, p. 307, 1 pl., 1903.

ProwazeEk, §8.— Die Entwickelung von Herpetomonas, einen mit
den Trypanosomen verwandten Flagellaten,” ‘ Arb. kaiser]. Gesundt-
heitsamte,’ 20, p. 440, many figs., 1904.

“Studien tiber Saugethiertrypanosomen,” op. cit., 22, 45 pp., 6
pls., 1905.

RasrnowitscuH, L., and Kempner, W.—‘‘ Beitrag zur Kenntniss der
Blutparasiten, speciell der Rattentrypanosomen,” ‘ Zeitschr. Hyg.,’
30, p. 251, 1 pl., 1899.

** Die Trypanosomen in der Menschen- und Thierpatho-

logie, so wie vergleichende Trypanosomenuntersuchungen,” ‘ Centrbl.
Bakter.,’ (I), 34, Orig., p. 804, 1903.

3828 H. M. WOODCOCK.

91. Rennes.—‘ Contribution a |’étude de la Dourine,”’‘ Ree. med. vet.,’ (8),
10, p. 424, 1903.

92. Rocrrs, L.— The Transmission of the Trypanosoma evansi by
Horse-flies, and other Experiments pointing to the Probable Identity
of Surra of India and Nagana or Tsetse-fly Disease of Africa,” ‘ P.
Roy. Soc.,’ 68, p. 163, 1901.

“On the Role of the Horse-fly in Trypanosoma-infection,”

‘Brit. Med. J.,’ 1904, ii, p. 1454.

x

93. Roucet, J.— Contribution 4 l’étude de la Dourine,” ‘ Rec. med. vet.,’
(8), 10, p, 81, 1903. :

93 a.— “‘Trypanosome de la Dourine; son inoculation aux souris
et aux rats,’ ‘C. R. Soc. Biol.,’ 56, p. 744, 1904.

94. SaBraziss, J., and Muratet, L.—“‘ Trypanosome de l’anguille (Anguilla

vulgaris),” ‘Pr.-verb. Soc. Bordeaux,’ 57, pp. 82 and 124, 1 pl.,
1908.

92a.

95. “Trypanosome de l’anguille. Processus de division,”

‘C. R. Soc. Biol.,’ 56, p. 66, 1904.

96. Satmon, D., and Stites, C.—“ Emergency Report on Surra,” ‘ Bull. U.S.
Dep. Agric. and Animal Ind.,’ No. 42, 152 pp., 112 figs., 1902.

97. Scuaupiny, F.—‘‘Studien iiber krankheitserregende Protozoen, II,
Plasmodium vivax (Gr. and Fel.),” ‘Arb. kaiserl. Gesundheits-
amte,’ 19, p. 169, 3 pls., 1902.

“ Generations- und Wirthswechsel bei Trypanosoma und
Spirochate (Vorl. Mittheilung),” op. cit., 20, p. 387, 20 figs., 1904.

99. Senn, G.—‘ Der gegenwartige Stand unserer Kenntnisse von den flagel-
laten Blutparasiten,” ‘ Arch. Protistenk, 1, p. 344, 1902.

“ Plagellata,”’ in ‘Engler und Prantl’s natiirlichen Pflanzen-
familien,’ I, ia, 1903.

101. Sergent, E. and E.—“Sur un Trypanosome nouveau parasite de la
grenouille verte,” ‘C. R. Soc. Biol.,’ 56, p. 128, 1 fig., 1904.

98.

100.

101 a. ‘“‘ Notes sur une trypanosomiase des dromadaires d’ Algérie,”
t.c., pp. 120, 914, 1904.

102. “Sur les Hématozoaires des oiseaux d’Algérie,” t.c., p. 132,
1904.

103. “ Hémamibes des oiseaux et moustiques. ‘Generations alter-
nantes’ de Schaudinn,” op. cit., 58, p. 57, 1905.

104. “Sur des Trypanosomes des chauves-souris,” t. ¢., p. 58, 2

figs., 1905. ;
105. Srmcen.— Die geschlechtliche Entwickelung von Hemogregarina

stepanowi im Riisselegel Placobdella catenigera,” ‘ Arch.
Protistenk.,’ 2, p. 339, 7 figs., 1903.

THE H#MOFLAGELLATES. 3829

106. Sivort, F., and Lecter, E.—‘ Le Surra americain ou Mal de Caderas,”
‘Ann. Min. Agric. Argentina,’ 1, p. 1, 7 pls., 1902.

107. SMeptey, R. D.—‘‘ The cultivation of Trypanosomata,” ‘J.
Hygiene,’ 5, p. 24, 2 pls., 1905.

108. Stassano, H.— Contribution a |’étude du Trypanosome,” ‘ C. R. Soc.
Biol.,’ 53, p. 14, 1901.

Sur la fonction et relation du petit noyau des Trypanosomes,”
t.c., p. 468, 1901.

110. Strex, J. H.— Report on his Investigation into an Obscure and Fatal
Disease among Transport Mules in British Burmah,”’ 1885.

109.

111. THrrRoux.—“ Sur un nouveau ‘T'rypanosome des oiseaux,” ‘C. R. Ac.
Sci.,’ 139, p. 145, 5 figs., 1904.

112. “Sur un nouveau Trypanosome de la souris domestique (Mus
musculus), ‘C. R. Soc. Biol.,’ 58, p. 885, 2 figs., 1905.

113. * T’infection du Padda oryzivora par le Trypanosoma
padde n’a pas de rapport avec |’infection de cet ciseau par l’Halteri-
dium danilewskyi,” ‘C. R. Ac. Sci.,’ 140, p. 109, 1905.

114. * Recherches morphologiques et expérimentales sur les Try-
panosoma padde,” ‘ Aun. Inst. Past.,’ 19, p. 65, 1 pl., 1905.

115. ‘Recherches morphologiques et expérimentales sur Trypano-

soma duttoni (Thiroux),” t.c., p. 564, 1 pl., 1905.

116. THomas, H., and Linton, 8.—‘‘ A Comparison of the Animal Reactions
of the Trypanosomes of Uganda and Congo Free State Sleeping-sick-
ness with those of Trypanosoma gambiense, Dutton,” ‘ Lancet,’
1904, i, p. 1837.

117. Vatentin.—Uber ein Entozoon im Blute von Salmo fario,” ‘ Muller’s
Arch.,’ 1841, p. 435.

117a, VALLEE and Panisset.—‘“ Sur les rapports du Surra et de la Mbori,”’
‘C. R. Ac. Sci.,’ 139, p. 901, 1904.

118. VassaL, J.—‘ Sur un nouveau Trypanosome aviare,” ‘C. R. Soe. Biol.,’
58, p. 1014, 1 fig., 1905.

119. Vocrs, O.— Mal de Caderas,” ‘ Zeitschr. Hygiene,’ 39, p. 3238, 1 pl.,
1902.

120. WASIELEWSKY and Senn, G.—“ Beitrage zur Kenntniss der Flagellaten
des Rattenblutes,” op. cit,, 33, p. 444, 3 pls., 1900.

121. Ziemann, H.—“ Uber ein neues Halteridium und ein Trypano-
soma bei einer kleinen weissen Eule in Kamerun,” ‘ Arch. Schiffs. u.
Tropenhyg.,’ 6, p. 389, 1902.

VOL. 50, PART 2.—NEW SERIES. 23

330

H. M. WOODCOCK.

z. Relating to the “Leishman-Donovan-Wright” bodies.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

133.

134.

135.

136.

137.

138.

Cuatrerser.—[“ On the Cultivation of Flagellate Forms from the
Leishman-Donovan bodies],” ‘ Lancet,’ 1905, i, p. 16.

CuristorHers, S. R.—‘ Reports on a Parasite found in Persons
suffering from Enlargement of the Spleen in India,” ‘Sci. Mem.
India,’ Nos. 8, 11, and 15, 1904-5.

Donovan, C.—‘ Human Piroplasmosis,” ‘ Lancet,’ 1904, ii, p. 744, 1 pl.

James, S. P.—‘‘ Oriental or Delhi Sore,” ‘Sci. Mem. India,’ No. 13,
1905.

“On Kala-Azar, Malaria, and Malarial Cachexia,” op. cit.,
No. 19, 1905.

Laveran, A., and Mesniz, F.—‘‘ Sur un Protozoaire nouveau (Piro-
plasma donovani, Lav. et Mesn.) parasite d’une fievre de I’ Inde,”
‘CO. R. Ac. Sci.,’ 137, p. 957, 17 figs., 1903.

“‘ Nouvelles observations sur Piroplasma donovani,
Lav. et Mesn.,” op. cit., 188, p. 187, 1904.

LrisumMan, W.—‘ On the Possibility of the Occurrence of Trypanoso-
miasis in India,” ‘ Brit. Med. J.,’ 1903, i, p. 1252, 2 figs.

“Note on the Nature of the Parasitic Bodies found in Tropical

Splenomegaly,” op. cit., 1904, 1, p. 303.

and StatHaM.— The Development of the Leishman Body in
Cultivation,” ‘J. Army Med. Corps,’ 3, 14 pp., 1 pl., Mareh, 1905.

Manson, P.—* Note on Two Cases of Febrile Tropical Splenomegaly
(Kala-Azar) and a Suggestion,” ‘ Brit. Med. J.,’ 1905, t, p. 1261.

and Low, G. C.—‘‘ The Leishman-Donovan Body and Tropical

Splenomegaly,” ‘ Brit. Med. J.,’ 1904, i, p. 183, 12 figs.

** The Leishman-Donovan Body in Ulcerated Surfaces ;
a Possible Route of its Eseape from the Human Body,” ‘J. Trop.
Med.,’ 7, p. 239, 1904.

Marcuanp, F., and Lepincuam, J.—“ On the Question of Try pano-
soma Infection in Man,” ‘ Lancet,’ 1904, i, p. 149, 1 fig.

“Uber Infektion mit ‘ Leishman’schen Kérperchen’
(Kala-Azar?) und ihr Verhaltniss zur Trypanosomen-Krankheit.,”
‘Zeitschr. Hyg.,’ 47, p. 1, 2 pls, 1904.

Mesnit, I'., Nicotxe, F., and Remiincer, P.—‘‘ Sur le Protozoaire du
bouton d’Alep,” ‘C. R. Soc. Biol.,’ 57, p. 167, 1904.

Rogers, L.—‘On the Development of Flagellated Organisms (l'ry-
panosomes) from the Spleen Protozoic Parasites of Cachexial Fevers
and Kala-Azar,” ‘Q. J. Mier. Sci.,’ 48, p. 367, 1 pl., 1904.

THE H#MOFLAGELLATES. ool

189. Rocers, L.—‘‘ The Conditions Affecting the Development of Flagellate
Organisms from Leishman Bodies and their Bearing on the Probable
Mode of Infection,” ‘ Lancet,’ 1905, i, p. 1484.

140. Ross, R.—‘ Notes on the Bodies recently described by Leishman and
Donovan,” ‘ Brit. Med. J.,’ 1903, i, pp. 1261 and 1401, 17 figs.

‘A New Parasite of Man,’ ‘ Thompson-Yates Lab. Rep.,’ (4),
2, p. 79, 1 pl., 1904.

142. Wricut, J. H.—* Protozoa in a Case of Tropical Ulcer (Delhi Sore),”’
‘J. Med. Research,’ Boston, 10, p. 472, 4 pls., 1903.

141.

c. Relating to the “Spirochete.”

143. Crertes, A.—“‘ Note sur les parasites et les commensaux de l’huitre,”’
‘ Bull. Soc. Zool. France,’ 7, p. 347, 1 pl., 1882.

“Sur le Trypanosoma balbianii, op. cit., 16, pp. 95 and
130, 1 fig., 1891.

145. Exrenperc.— Die Infusionsthierchen als vollkommene Organismen,”’
Leipzig, 1838. [Describes and figures Spirocheta plicatilis;
this form was originally diagnosed in ‘ Abh. Ak. Berlin,’ year 1833,
p. 313 (publ. 1835) ].

145a. Laveran and Mrsnin.—“‘Sur la nature bactérienne du pretendu
Trypanosome des huitres, T. balbianii,” ‘C. RK. Ac. Sci.,’ p. 883, 1901.

146, Perrin, W. S.—“ A Preliminary Communication on the Life-history of
Trypanosoma balbianii,” ‘P. Roy. Soc.,’ 76 B, p. 367, 4 figs.,
1905.

147. Krzysztatowicz, F., and SrepLecki1, M.—‘ Contribution a l'étude de
la structure et du cycle évolutif de Spirochaete pallida Schaud,”
‘Bull, Ac. Cracovie,’ 6th November, 1905, p. 713, 1 pl.

148. Scuaupinn, I.—“ Zur Kenntniss der Spirochete pallida (Vorl.
Mittheilung),”’ ‘Deutsch. med. Wochenschr.,’ No. 42, 19th October,
1905, p. 1665.

144.

149. and Horrmann, E.—“ Uber Spirochxtenbefunde im Lymph-
driisensaft Syphilitscher,” op. cit., No. 18, 1905, p. 711.

150. —— “Vorlanfiger Bericht tiber das Vorkommen von Spiro-
cheten in syphilitschen Krankheitsproducten und bei Papillomen,”’
‘ Arb, Kaiserl. Gesundheitsamte,’ 22, p. 527, 1905.

151. “Uber Spirochete pallida bei Syphilis und die

Unterschiede dieser Form gegeniiber anderen Arten dieser Gattung,”
‘ Berl. klin. Wochenschr.,’ No. 22, 1905.

152. Vuittemin, P,—“Sur la dénomination de l’agent présumé de la
syphilis,’ ‘C. R. Ac. Sci.,’ 140, p. 1567, 1905.

=

NOTES ON MEDIAN AND PATRED FINS OF FISH. 3338

Notes on the Development, Structure, and Origin
of the Median and Paired Fins of Fish.

By
Edwin 8. Goodrich, F.R.S.,
Fellow of Merton College, Oxford.

With Plates 10—14.

ConrTeENTS.

PAGE
INTRODUCTION 4 ‘ ; : . 334
The Lateral Fold Theors ; : : 5 O00
The Gill-arch Theory . : : . 340
Objections to the Gill-arch Theory 5 : . 342
The Apparent Migration of Fins. : ‘ . 3843
The Faithfulness of Muscle and Nerve 5 : . 348
The Shifting of Limbs Explained : : . 349

OBSERVATIONS ON THE STRUCTURE AND DEVELOPMENT OF THE FINS
or EKLASMOBRANCHS : - 852
Development of the Median Fins of Soylliam eamicals . 353
Development of the Skeleton of the Median Fins : . 3855
Development of the Paired Fins . . 9356
The Concordance between the Muscles and tlie Radials 5 ODY
Concentration : ; : . 9358
The Fusion of the Muscle: buds . : : . 059
The Position of the Fins i : ; - 360
On tHE NERVE SUPPLY oF THE Fins . : : 1 eouo
The Plexus of the Dorsal Fins. : , . 864
On the Real Nature of the Nerve-plexus . ‘ . 366
Experiments on the Nerve ee of the Fins ‘ . 369
SUMMARY AND CONCLUSION . : : « ofl

vot. 50, PART 2.—NEW SERIES. 24,

334 EDWIN S. GOODRICH.

INTRODUCTION.

A creat deal has been written in the last few years about
the structure, development, and origin of the paired fins of
fish, yet two rival and incompatible theories are still preva-
lent. According to the theory put forth by Gegenbaur (14,
15, 16), the paired fins have been derived from gill structures,
the gill-arch having been modified into the lmb-girdle, and
the fin itself, with its skeleton, having been derived from the
gill-flap or septum, and its supporting gill-rays. This may
shortly be called the “ gill-arch theory.” The second theory,
that of Balfour (1, 2), Thacher (85), and Mivart (23), holds
that the paired fins are of the same nature as the unpaired
median fins. According to this view, the limbs have been
derived from paired longitudinal fin-folds, in which skeletal
supports, the radials, or somactidia (Lankester), became
developed as in the median fins, and subsequently gave rise
to the lhmb-girdles. This may be called the “lateral fin-fold
theory.” Each of these theories may claim to have among
its numerous supporters the names of some of the most
eminent exponents of the morphology of the vertebrates.
Dohrn (10), Haswell (20), Rabl (31), Moller (24, 25, 26),
Harrison (19), Wiedersheim (36), A. Smith Woodward (387),
and Dean (9) have written in favour of the lateral fold theory;
Davidoff (8), Fiirbringer (12), Braus (8-7), and others have
supported its rival. It is unnecessary for me in these notes
to give a history of the discussions to which the question has
given rise; the literature has been recently reviewed by
Mollier and Braus, and the whole subject is familiar to
zoologists. But there are certain essential points which seem
to be in danger of being obscured from view in a cloud of
controversy, and it is in the hope of clearing up some of these
points and of filling up certain gaps in the evidence that
these notes have been published.

As Iam anxious to keep this paper within reasonable limits
and not to overburden the already very bulky literature on
the subject of the origin of the paired limbs, only some aspects

NOTES ON MEDIAN AND PATRED FINS OF FISH. 335

of the problem will be dealt with in detail. A brief and some-
what dogmatic statement of the case is made at the beginning,
followed by a description of my own observations, and ending
with a short summary.

THe Laterat Forp THeory.

Balfour’s conception of an originally continuous fin-fold,
reaching from the pectoral to the pelvic region (1) is dis-
credited because it has been found only (as an epidermal fold)
in those forms, like Torpedo, in which the pectoral fins reach
the pelvic fins in the adult, a condition which is probably
rightly considered as secondary. Moreover, the appearance
of an epidermal longitudinal fold, as a first indication of the
development of the paired fins, is considered to be of little
importance, and its presence between the paired fins is denied
in sharks (Mollier 24, Braus 4, ete.).

Now, the continuity of the pectoral with the pelvic fin-
fold is not an essential point. The important thing is to
recognise that the paired fins always arise as a longi-
tudinal fold or ridge, similar to that which gives rise to
the median fins. hat this is really the case is now admitted
by all (Braus 7). Even in Ceratodus, where the paired fins in
the adult are set at a pronounced angle to the long axis of
the body, they make their first appearance as longitudinal
ridges (Semon 33).

Possibly from the very first, in phylogeny, the paired fins
were discontinuous, and differentiated into pectoral and
pelvic. For conclusive evidence on this point we must look
to paleontology; and it has not yet been obtained. But
there is some evidence to be gathered from comparative
anatomy and embryology in favour of Balfour’s view, as has
frequently been pointed out (Dohrn 10, Mollier 24).

For instance, the musculature of the fins is developed in
Klasmobranchs, from buds given off from the ventral ends of
the myotomes, and these buds have been shown to be pro-
duced not only on themyotomes in the region of the fins, but

336 EDWIN S. GOODRICH.

on all the trunk myotomes situated between the pectoral
and the pelvic fins in such sharks as Pristiurus and Scyllium,
in which these fins are widely separated in the adult (Dohrn
10, Rabl 31, Braus 4, and p. 343 below). Many of the inter-
mediate buds seem to disappear entirely during development.
In those segments which are near the fins the buds become
better developed and more persistent, and a large number
pass into the fin-fold. Muscle buds are also found in front of
the pectoral fin and behind the pelvic fin, dwindling in size as
they are farther removed from the fin-base. Thus, in these
sharks, the muscles of the paired fins are formed by the great
development in two regions of a continuous series of muscle
buds, vanishing posteriorly behind the cloaca. The manner
in which these vestigial buds disappear by reduction at
either end of the fin rudiment, and in which the persistent
buds become concentrated at the relatively narrowing
base of the fin, has been admirably described and figured by
Mollier (24) and Braus.

The fin-base of the adult occupies much less space relatively
than the fin-fold of the embryo.

Now, the radial fin-muscles being developed from buds of
the myotomes, naturally receive their motor nerve-supply
from the ventral roots of the spinal nerves, and these corre-
spond in number to the myotomes which share in the forma-
tion of the musculature. Owing to concentration, the nerves
are found to converge toward the base of the fin. In front
and behind the nerves may be drawn together so as to form
a “collector” nerve or compound stem.

It is part of the lateral fold theory to suppose that the
endoskeleton of the paired fins has been derived from a series
of cartilaginous rods, radials, or somactidia, similar to those
of the median fins (Thacher 85, Mivart 28). The various
types of fin-skeleton, with their cartilage rays and basal pieces,
would have been developed from such originally segmental
radials by concentration and fusion. ‘To this contention it is
objected that in development the radials of Elasmobranchs
arise in a continuous rudiment—a plate of procartilaginous

>

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 33

mesenchyme (Balfour 2, Mollier 24, Ruge 32, and p. 357 below).
It may be answered (Dohrn 10, Mollier 24) that, the radials
being closely approximated, their procartilaginous rudiments
with indefinite borders necessarily merge together to a con-
siderable extent. As a matter of fact, the cartilage pieces
appear as islands in the vaguely-defined rudiment, which
correspond closely in position and number with the separate
elements of the adult fin-skeleton. Some shght indications
of recapitulation, some fusion of neighbouring radials, may be
detected, which bears out the views so convincingly advocated
by Thacher and Mivart. But it cannot be claimed that re-
capitulation is complete in this respect in the development of
the paired fins. It is obvious, however, that if its absence is
considered as evidence against the lateral fold theory it tells
with equal force against the gill-arch theory, since
the skeleton is, according to this view, also derived from
originally separate (branchial) rays.

But the whole argument against the lateral fold theory
collapses when we find that, as Balfour long ago showed,
the radials of the median fins likewise arise in a
continuous prochondral plate, in the median fins of
Elasmobranchs, even when they are separate in the adult
(p. 355 below). These median fins are much concentrated,
and nothing proves so clearly that the early continuity of the
rudiments is due to their approximation, for here the original
metameric nature of the radials will not be denied. The
most enthusiastic supporter of the gill-arch theory would
not suppose that the continuous plate represents an early
stage in the phylogenetic history of the skeleton of median
fins! Unfortunately, we know but little concerning the
development of the skeleton in unconcentrated median fins.
Doubtless, in such cases the radials arise separately; Harrison,
indeed, has shown this in his valuable paper on the salmon (19).

Yet other objections have been brought forward by Braus,
in the elaborate and beautiful memoirs which have of late
contributed so much to our knowledge of the structure and
development of fins (8, 4, 6, 7). It has been shown that two

338 EDWIN S. GOODRICH.

muscle-buds are given off by each myotome to the paired fins
in Elasmobranchs; that these pass outwards into the fin-fold,
dividing into upper and lower halves, which give rise to the
dorsal and ventral radial muscles. Between each pair of
corresponding upper and lower buds develops a cartilagimous
radial. Thus, as Rabl showed, since two radial muscles and
cartilages correspond to each segment, the relation between
the number of radials in the fin-skeleton, and the number of
trunk vertebra belonging to those segments which contributed

to the formation of the fin, may be expressed in the formula
Radials

= vertebrae. Braus has endeavoured to prove that

this formula does not hold good (p. 444, 3). But it is quite
obvious that, although in the main correct, it can only be
intended to give approximate results when applied to whole
fins. In most paired fins of Elasmobranchs the anterior and
posterior regions are much modified by excessive concentration
and reduction, and here the correspondence between muscles
and radials becomes much disturbed. The formula applies
perfectly over the greater part of a fin which is normally
developed, as is seen in Braus’s own figures (4,6). More im-
portant is the contention that the adult radial fin-muscles
do not correspond to the muscle-buds in the embryo. It is
urged that the muscle-buds become mixed and that the adult
muscles are no longer unisegmental and haploneurous, but
are compound and polyneurous, and, in fact, bear no definite
relation to the segments from which they arose.

It is true that, as Mollier has shown (24), the muscle-buds
in Elasmobranch fins may be connected together at their base,
at all events temporarily, by strands of tissue. It is also true
that the mixed motor and sensory nerves form a complicated
plexus at the base of, and round about, the radial muscles.
But it does not follow that these muscles are either compound
or polyneurous. So far as lam aware, it has never been proved
that muscle-forming substance actually passes from one bud
to another (p. 359); nor has it ever been proved that one
‘adial muscle is really innervated by more than one motor

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 339

root. In fact, it seems to be very probable indeed that, even
in the adult, the radial muscles are strictly segmental and
haploneurous (see below, pp. 364-3871). Some fusions may
take place, some disturbances of the metameric order may
occur, especially at the extreme anterior and posterior ends of
the fins; but it is quite firmly established that each adult radial
muscle develops from, and corresponds in position to, a single
muscle-bud. It may be asserted with confidence that a radial
muscle is derived, at least mainly, from that bud
whose position it later occupies; and that the radial
muscles in the normally developed region of the
paired fin of an Elasmobranch corresponds accu-
rately in number and position to the group of primi-
tive buds from which they have been formed.!

There is a last objection which Braus persistently reiterates
in his papers, and of which he makes a great deal. He alleges
that the “ concordance” which exists in the adult between the
radial muscles and the radial cartilages is not primitive, but
secondary. He states that in the early stages of development
there are “ discrepancies” between these elements, that the
muscle-buds do not correspond exactly with the rudiments of
the radials, and that the perfect correspondence, or concord-
ance, is gradually established in later stages. This subject
will be dealt with later on in greater detail (p. 357); but it
may here be said that the evidence on which Braus bases his
argument seems to be of the slenderest and most unconvincing
nature. Not even in the adult is the concordance perfect ;
marked disturbances occur at both the anterior and the pos-
terior extremities of the fins. The peripheral ends of the adult
muscles correspond exactly with the radials in the middle

1 Tf it is objected that in Ceratodus, where the adult paired fin has about
thirty radials and radial muscles, only about three segments have been shown
to contribute muscle-buds in the embryo (Semon 38), it must be answered
that this result is not trustworthy. Davidoff (8) and Braus (3) have found
twelve spinal nerves contributing to the limb-plexus. It is probable that
Rabl’s formula holds good in Ceratodus (Mollier 24), and that a large and
sufficient number of segments really contribute muscle-forming cells to the
limb, but not in the form of distinct buds.

340 EDWIN S. GOODRICH.

region; but as they pass inwards to become attached to the
base, or the girdle, the muscles no longer preserve the “ con-
cordance.” On the other hand, nothing is so striking on
examining sections through the developing fins of Hlasmo-
branchs, whether paired or unpaired, as the extraordinarily
regular “concordance”; it is obvious on the very first appear-
ance of the procartilaginous radial (p. 358, figs. 5, 8, 9, 18).

These attempts to undermine the lateral fold theory, by
showing that the adult muscles are compound and polyneurous,
and that the concordance is secondary, are not borne out by
the evidence. Moreover, even if it could be proved that the
metamerism of the fin elements has been lost, the lateral fold
theory would scarcely be affected, since it only claims that
the muscles and skeletal radials formed a longitudinal series
of metameric origin in the beginning. No one doubts that
the metamerism has been obscured, or lost, in the higher
vertebrates; it matters little, theoretically, whether it still
persists in modern fish.

THe Gitt-Arcu THEORY.

Let us now pass to the rival theory. It is claimed that the
initial stages in the phylogenetic history of the paired fins
are more easily accounted for on the gill-arch theory of
their origin. Now, according to the lateral fold theory
the paired fins appeared, as they do in ontogeny, as longi-
tudinal ridges, which, from their very first appearauce, may
have been useful as balancing and directing organs. Even
in modern fish the paired fins are used not so much for pro-
gression as for guidance and balancing.

On Gegenbaur’s theory the direction of the paired fins must
at first have been dorso-ventral across the long axis of the
body ; such folds would probably be a hindrance to progres-
sion, and both the pectoral and pelvic fins would have been
placed close together behind the head in a most unfavourable
situation,

The position of the pelvic fin is accounted for by supposing

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 341

that it has migrated backwards from the head region. Now,
there is no evidence of a more anterior position of the pelvics
in primitive fishes generally, either living or extinct. Indeed,
the only known fish in which the pelvics are far forward
(some Teleostei) are acknowledged to be specialised in this
respect. The presence in ontogeny of rudimentary muscle-
buds in front of the pelvic fins, is supposed to indicate
backward migration. ‘This is negatived by the fact that
similar rudimentary buds are found behind the pelvic fin
(Braus 4, Pl. 22, and in this paper, figs. 1, 4, 25). The fins
could not have migrated both ways at once, and there is
no reason to believe that they first migrated backwards to a
point behind the cloaca, and then forwards towards the head.

Davidoff (8), Gegenbaur (15), and others have held that
the presence in front of the pelvic fin of a collector nerve,
composed of branches of a number of spinal nerves, and the
greater extent of this plexus in the young than in the adult
(Punnett 29, 30), indicates backward migration. But, again,
both a similar plexus and extension are found on the pos-
terior side of the fin.

The question of the nerve supply of the fins will be dis-
cussed in greater detail later (p. 363) ; but in describing the
general nerve-plexus at the base of the fins one must be
careful to distinguish between the collector nerve formed
by the convergence and combination of branches of a series
of spinal nerves and the plexus proper, due to intertwining
secondary branches, made up chiefly, if not entirely, of
sensory nerve-fibres. The formation of a collector nerve
is simply and easily explained as the result of concentra-
tion. The mere presence of a connecting plexus (mainly
longitudinal) is due neither to concentration nor to migra-
tion (p. 367).

Moreover, both these arguments in support of the theory
of migration are sufficiently answered by the fact that
rudimentary buds are found both in front of and behind
the median fins (Mayer 22 and p. 353 below), and that a longi-
tudinal nerve-plexus may extend along their base even when

342 EDWIN S. GOODRICH.

the fin is continuous and there is no possibility of migration.
Longitudinal connecting nerves have long been known to
exist at the base of the unconcentrated fins of Teleostean
fish; I find them also at the base of the dorsal fin of
Chimera, which is scarcely, if at all, concentrated.

OpseECTIONS 'to THE GiiL-ARcH ‘THEORY.

We may now deal with some very serious difficulties in the
way of the gill-arch theory. Firstly, it offers no intelligible
explanation of the participation of a large number of seg-
ments in the formation of the paired fins. Yet it is always
the case that a considerable, and sometimes a very large,
number of spinal nerves and myotomes contribute towards
its development.

Secondly, if the skeleton of the paired fins were derived
from gill-rays we should expect the muscle supply to be
drawn, not from the myotomes at all, but from the unseg-
mented “lateral-plate,” or visceral, musculature, which is
innervated by the dorsal roots of the spinal nerves. It is true
that the trapezius muscle attached to the scapula is of
lateral-plate origin, and is supplied from the vagus nerve ;
yet it does not enter into the fin, does not, in fact, belong to
the fin musculature. At all events, in the pelvic region
there is no trace whatever of other than segmented muscles.

A third, and perhaps still more important, objection to
Gegenbaur’s theory is this: the position of the lmb-girdles
in relation to the nerves, blood-vessels, coelom, etc., is
exactly the reverse of what it should be if they
were derived from visceral arches. The ccelom, the subin-
testinal vessel (heart, etc.), the myotomes and their nerves,
all pass outside the visceral arches. ‘The limb-girdles, on the
contrary, lie morphologically outside these structures, so that
the nerves frequently pass through the girdles to reach the
fins. In fact, the girdles lie in the outer body-wall, while the
visceral arches lie in the wall of the alimentary canal. No
mere superficial resemblance in shape of the girdle to the

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 343

arch in a developing Hlasmobranch, such as is insisted upon
by Braus (5), no mere opinion, unsupported by evidence, that
the relative position of the girdle has been altered, such as is
expressed by Fiirbringer (12), can outweigh these facts.

The fourth and last objection which we shall urge against
the gill-arch theory is one which will probably seem to most
zoologists to be the most fatal of all: the theory gives no
explanation of the remarkable resemblance borne
by the paired fins to the unpaired fins. The resem-
blance is not vague and indefinite, it is minute; it can be
followed out in every detail both of their structure and of their
development. In no respect is this more striking than in the
development and differentiation of the dermal fin-rays in the
various groups of fishes.

All these facts, which clearly support the lateral fold
theory, are so many deadly blows aimed at the rival gill-
arch theory. Far from being difficulties which have to be
explained away, they become evidence actually in favour of
the fundamental likeness of the paired and unpaired fins.

THe APPARENT MIGRATION OF FIs.

We have now to account for the apparent migration of
linbs from one place to another on the body of vertebrates.
Every trunk segment may be said to be capable of
producing muscular, nervous, and skeletal “limb
elements” of a paired character. This “ potentiality” is
actually called into force in the case of the Rajidee throughout
the trunk region, with the exception of a few anterior seg-
ments (see Rabl 31, Mollier 24, and especially Braus 3). In
Torpedo, for instance, the 4th to the 30th spinal nerves
supply the pectoral fin, and the 31st to the 42nd the pelvic
fin. In Trygon the 3rd to the 59th supply the pectoral, and
the 60th to the 71st the pelvic fin (Braus). The same con-
clusion is indicated in the case of forms lke Pristiurus and
Scyllum, where the paired fins are widely separated, by the
development of muscle-buds on all the trunk segments (see

544. EDWIN S&S. GOODRICH.

figs. 1,25). It is also borne out by a comparison of the range
of extension of the fins in various genera; for instance, whilst
the paired fins occupy segments 5-23 and 47-65 in Zygoena,
they occupy segments 2-19 and 29-50 in Heptanchus and
segments 2-15 and 19-37 in Chimeera ( Braus).

The conclusion that every trunk segment is capable
of producing muscular, nervous, and skeletal ele-
ments of the median dorsal fin is likewise reached on
examining the structure and development of that organ. It
is well known that a more or less perfectly continuous dorsal
fin still exists in many modern Teleostei, and was present
in many extinct forms (Dipnoi, Pleuracanthus). I shall be
able to show below (p. 353) that the muscle-buds giving rise
to the widely separated adult dorsal fins of Scyllium form a
continuous series in the embryo.

Every trunk-segment, then, is potentially able to produce
paired and unpaired “ fin-elements.” But, even if the ances-
tral Gnathostome was provided with continuous-paired fin-
folds, the position of the paired limbs of vertebrates can not
be accounted for merely on the supposition that these folds
have survived in this or that region. The paired limbs have
certainly altered in position since they were first established
with regard to the numerical order of the segments they
occupy. In fact, it is clear that a perpetual shifting of
the position of the limbs has taken place in all classes of
Gnathostome vertebrates.

It seems to be often held that these changes of position are
brought about either by the actual shifting or migration of
the limb from one place to another, or by the excalation and
intercalation of segments. We cannot, in this paper, enter
into a discussion as to the origin and significance of meta-
meric segmentation in vertebrates; but something must be
said about the theory of excalation and intercalation, strongly
supported many years ago by v. Jhering (18). Already it
has been so severely and successfully attacked by Fiirbringer
(11) that it can be very shortly dismissed.

In the case of the pelvic fins of Teleosts, for instance, there

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 345

are fifteen trunk segments between the pectoral and the
pelvic nerve-plexus in [sox lucius, three in Cyprinus tinca,
and none at all in Gadus. To account for this by v. Jhering’s
theory, we must suppose that a new trunk, presumably also
new viscera, have developed behind the pelvic fins, while the
old trunk and viscera have disappeared in front! Moreover,
in Lepidoleprus and Uranoscopus the 3rd spinal nerve shares
in both the pectoral and the pelvic plexus,

Still more difficult to explain by excalation and inter-
calation is the case of the Elasmobranchs. There are twenty-
three segments between the pectoral and the pelvic plexus
in Zygena, only three in Pristis, and none at all im many
Rajidee ; yet, of course, the other parts remain unaffected.

The evidence of embryology is also thoroughly opposed to
such a theory. Comparing various forms, such as Rana
with Necturus, Lacerta with a snake, etc., we find large,
sometimes vast, differences in the number of segments; we
might, therefore, expect to discover in the embryo zones
where segments are either being formed or absorbed. Nota
trace occurs of such zones of growth or absorption.

‘The nerve-plexus of the pectoral fin of Spinax occupies ten
segments, that of Torpedo twenty-seven, that of Trygon fifty-
seven ; no sign whatever of zones of excalation or intercalation
has been found in their development. It is unnecessary to
multiply instances (Fiirbringer 11, Braus 8).

But if it is difficult to account for the varying position of
the paired limbs on the theory of excalation and intercala-
tion, the task becomes impossible if we attempt thus to
explain the varying position of both the paired fins and the
unpaired fins; for we find that the various fins alter in
position and extent independently of each other.
No scheme of excalation and intercalation, however ingeni-
ously devised, can ever account for the position of the first
dorsal fin opposite the pectoral in Lamna, between the
pectoral and the pelvic fins im Alopecias, opposite the pelvic
in Scyllium, and well behind it in Raja.

Returning, now, to the other explanation of the change of

546 EDWIN S. GOODRICH.

position of paired fins, we find that Gegenbaur seems to have
held that the whole girdle and fin-skeleton could move from
its place of origin, dragging to some extent the muscles and
nerves with it. He pointed to the collector nerves and rudi-
mentary buds as evidence of this actual migration of the
ready-formed pelvic fin. This argument has already been
dealt with above (p. 340), and will be further answered below.
Braus believes that he has proved that actual migration of
the paired fins takes place during the development of Acan-
thias. His excellent figures, however, afford convincing eyi-
dence to the contrary. It is obvious that if a fin, in ontogeny,
moves as a whole, no one part of it can remain in its original
position. If now we compare his figure of the earlier with
that of the later stage in the development of the pelvic fin (figs.
1, 2, 3 and 4, Pl. 22), we find that the muscle-buds and nerve
belonging to segment 36 remain throughout in approximately
the same position. The neighbourhood of segment 36, there-
fore, represents a fixed point. It is true that the fin-fold
extends further forward in the earlier stages than it does in
the later, and further back in the later than it does in the
earlier ; but this is due to the fact that the fin develops, on
the whole, from before backwards, and undergoes more reduc-
tion in front than behind. The apparent migration of the
fin from segments 21-30 to segments 30-39, during develop-
ment, is brought about, not by the actual motion backwards
of the whole fin structure, but by the concentration of the
fin towards a central region, and by the great reduction of
its anterior border.! ;

A fin-fold will appear to move, during development, back-
wards or forwards, according as there is concentration and
reduction, more in the one direction than in the other.

In agreement with this, it is found that a fin-fold, and
its contained muscular, nervous, and skeletal ele-
ments, are derived from that region of the trunk
which is occupied by the adult fin (see further, p. 360).

1 T am inclined to doubt the correctness of the enumeration of the segments
in Braus’ figures. No such extensive apparent migration occurs in Scyllium.

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 347

The exact origin of the muscles of the paired fins is rarely
as easily traceable as in the Hlasmobranch; but, as far as is
known, the above rule holds good for all Gnathostomes.
Unfortunately, in many forms distinct muscle-buds are not
produced, and the muscle-producing cells are budded off
separately from the Myotomes. Nevertheless, the derivation
of the limb muscles has been distinctly traced in the case of
various Elasmobranchs, of Salmo (Harrison 19), of Acipenser
(Mollier 26), Cyclopterus (Guitel 17), and of Lacerta
(Moller 26). In all cases where the development has been
followed it has been shown that the nerve-supply (“limb-
plexus”) in the adult is a sure guide to the identifica-
tion of the segments from which the muscles have
been derived. Segments before, and behind, those of the
limb-plexus may have ceased to contribute, owing to reduc-
tion during development, but adult nerve-supply shows which
segments have contributed most.

Unfortunately, with regard to the skeletal element the
facts are not so well established. From the very nature of
the case, it is much more difficult to deal with. The carti-
laginous radials are merely local differentiations in continuous
connective tissue, or mesenchyme. And although probably
this tissue has itself been derived from segmental sclerotomes,
yet the limits of the segments have long ceased to exist when
the radials develop. There is, however, no valid reason for
believing that radials are less constant than the muscles
with which they are related. Nor is there any evidence that
the skeleton of the pelvic limb, for instance, is formed of
tissue derived from any other segments but those belonging
to its muscles.

Of course, limb elements may undergo relative displace-
ment in the course of ontogeny. In the development of fins
the anterior muscle-buds are relatively displaced backwards,
and the posterior buds are relatively displaced forwards—
this is the process of concentration. It may also happen,
in the higher vertebrates, that a limb may be shifted a
seoment or two up, or down, the vertebral column with which

348 EDWIN &. GOODRICH.

it becomes connected. Inthe case of the Gadide, with jugular
pelvic fins, it is clear that these have moved to their position
in front of the pectorals. But—and this is the important
thing to remember—these limbs do not really lose their
original connections, the displacement can be traced
in ontogeny, and the nerve supply in the adult in-
fallibly betrays its course.

Tue FAttHFuLNESs OF Muscie AND NERVE.

That in a series of metameric myotomes and nerves
each motor nerve remains faithful to its myotome,
throughout the vicissitudes of phylogenetic and ontogenetic
modification, may surely be considered as established. That
a motor nerve is unable to forsake the muscle in connection
with which it was originally developed to become attached to
some other seems to be in the highest degree probable, both
on physiological and on anatomical grounds. As a matter of
fact, this appears always to be the case in the development
of lmbs.

Now, the paired limbs, and also the median fins, are supplied
by branches from a number of segmental nerves forming a
“limb plexus.” In such a plexus the branches may fuse to
common stems, or become joined together by connecting twigs,
so that the nerve-fibres appear to become inextricably
mixed; at all events, they form a network of mixed fibres
(motor and sensory). The motor “plexus” of a limb, so far
as it can be said to exist (see p. 366), is brought about, not
by the nerve deserting one muscle for the sake of another,
but by the combination of muscles derived from
neighbouring segments. (I venture to make this dogmatic
statement in spite of the fact that the embryological evidence
is still, unfortunately, very incomplete because it seems to me
to result inevitably from what has been ascertained concerning
the anatomy and development of muscles and nerves generally.)

We may thus get compound muscles formed which receive
motor branches from more than one spinal nerve. Strictly

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 349

speaking, even in this case the nerves in all probability
remain faithful to the muscle substance of their own segment,
for it has been proved that each motor root supplies its own
special muscle-fibres, which are merely bound together in the
same muscle (Sherrington 384).

It seems to me very doubtful whether such compound
muscles are ever produced in the fins of fish, and I shall show
later (pp. 359 and 569) that there is good reason for believing
that the adult radial muscles are both unisegmental and
haploneurous. However, compound polyneurous muscles may
perhaps be found in fish, as they are in higher vertebrates.
Thus segmental nerves, involved in a limb-plexus, may
apparently, but only apparently, become connected with
muscles belonging to other segments than their own.

THe Sarrrinc oF Limes EXpLAINeD.

Briefly we may repeat, the muscle and nerve-supply is
drawn in the embryo from the segments of the region
occupied by the limbs in the adult; in cases where the
development is unknown, the nerve-supply indicates to which
segments the limbs belong. The size of the nerves compos-
ing the plexus may be considered as proportional to the im-
portance of the share the several segments take in the
formation of the muscles. The muscle-buds and adult muscles
in fins are usually better developed in the central regions of
the fins than at their two ends. So the nerve components of
a limb-plexus are usually stouter in the middle than in front
and behind. Just as the muscular elements dwindle or
increase in size, owing to the backward or forward extension
of the base of a limb, just so far may the nerves increase or
diminish in thickness.

The position of a limb-plexus may shift backwards or
forwards in all Gnathostomes ; no one would suppose that the
nerves actually pass up or down through the vertebra, etc.
Fiirbringer has clearly shown how the shifting may take place
in his important and beautiful works on the anatomy of birds

VoL. 90, PART 2.—NEW SERIES. 25

300 EDWIN S. GOODRICH.

and reptiles (11, etc.). As may be seen in the diagram
(fig. 27), the alteration in position of a limb is due to the
contribution made to the limb-muscles, etc., of certain seg-
ments at one end becoming less and finally ceasing altogether,
while the contribution made by certain segments at the other
end becomes correspondingly large. Thus new segments
may be taken in at one end and old segments may drop out
at the other, or the number of segments contributing may be
merely increased or diminished.

A limb may in this way undergo change of position without
necessarily undergoing any change of form or structure.
The only change involved in the process is that the limb,
instead of being derived from a certain set of segments in one
region, 1s derived from a similar set of segments farther up or
down the trunk. This is Fiirbringer’s principle of imitative
homodynamy, or parhomology, accompanying the progressive
metameric modification of a plexus. To borrow Professor
Lankester’s illustration, it may be compared to the trans-
position of a tune from one key to another on the piano.
The tune remains the same, but it is played on different
notes.

We conclude, then, that the change of position of limbs is
not due to the actual migration of the lhmb-rudiment,

limb-substance, but to reduction on one side and
growth on the other. The migration is apparent, not
real. It is, if one may be allowed the expression, the calling
forth of the potentiality of the segments, which shifts, passing
up or down like a wave. This view might be called “the
theory of the transposition of the limbs.”

The same argument applies to the girdles. In some
Hlasmobranchs (Braus 8), for instance, the pectoral girdle is
pierced by thirty-six nerves belonging to the limb plexus
(Trygon), in others by twenty (Torpedo), or by three
(Laemargus), in Ceratodus by none at all. These diazonal
nerves may each pass through separate foramina, or several
may pass through the same foramen. It seems probable,
therefore, that the material (scleromere) of a varying number

_ NOTES ON MEDIAN AND PAIRED FINS OF FISH. 351

of segments may share in the formation of the girdle, and
that when no diazonal nerves are present only one cartilaginous
segmental element is fully developed, at all events at the
point where the nerves pass outwards to the limb. When
several nerves pass through the same foramen we may
suppose that the cartilagmous elements between them have
been suppressed. It is interesting to note that in the case of
the Chondrostei (Thacher 35, Wiedersheim 36, Mollier 26) the
pelvic girdle still shows distinct traces of segmentation. Since,
however, the girdles are structures which grow inwards, en-
veloping the nerve-plexus, with which they only come into
secondary connection, it is quite possible that all strict
metameric concordance between the two has been modified or
lost in most cases. But a limb-girdle may be trans-
posed, like a plexus, by the addition of new elements at one
end and their disappearance at the opposite end. And thus
is brought about the apparent backward or forward motion
of a girdle through a number of segmental nerves, or, in other
words, the passage of nerves through a girdle.

To this theory of transposition it may be objected that, if
true, the limbs and girdles of the Gnathostomata are not
strictly homologous. Now, if by the homology of two structures
we mean that they are produced by the same number of
segments, occupying in both cases the same place in the
metameric series, the hmbs and girdles are certainly not
always homologous. Jn this strict and narrow sense they are
often not homologous amongst closely allied species, nor in
individuals of the same species, nor even on the two sides of
the same individual. Fiirbringer, Braus, and Punnett have
clearly demonstrated the great variability of the nerve-plexus
supplying the paired limbs. So long as a distinct individuality
and persistence are attributed to each segment, so long as
segment « of one animal is considered to be represented
only by the same segment @ in another animal, the term
“homology” can only be applied in a general sense
to the limb and its nerve-plexus, etc., as a whole. And let it
not be imagined that we can escape from this conclusion by

aoe EDWIN S. GOODRICH.

calling in the aid of the theory of excalation and intercalation
(see above, p. 344). The pectoral fin of Spinax, with its ten
segments, and that of Trygon, with its fifty-seven segments,
cannot be strictly homologous on any theory, whether the
extra forty-seven segments have been added in the latter
genus, or withdrawn in the former.

OBSERVATIONS ON THE STRUCTURE AND DEVELOPMENT OF THE Fins
or ELASMOBRANCHS.

The material used was obtained chiefly from the Plymouth
Laboratory of the Marine Biological Association; but I also
have to thank Prof. Dohrn and Mr. Adam Sedgwick for the
generous gift of valuable embryos.

The lateral fold theory is founded on the similarity between
the median and the paired fins, yet comparatively little has
been published on the development and structure of the
median fins of Klasmobranchs since the pioneer work of
Thacher (85) and Mivart (28).

Balfour (2) studied their development, and described the
origin of the cartilaginous radials from a continuous pro-
chondral plate. An epoch in our knowledge of the median
fins dates from the appearance of an important paper by
Mayer (22). He there describes the development of the
skeleton in Pristiurus, and of the radial muscles from muscle-
buds, which had already been noticed by Doran (10). Atten-
tion is drawn to the presence of abortive buds behind the
dorsal fins, and the collector nerves and general nerve-plexus
is described in many adult forms. But Mayer was unable to
trace accurately the relation borne by the buds to the myotomes,
nor did he follow out the process of concentration in detail.

Harrison (19) has published an excellent account of the
development of the median fins in Salmo; in this fish, how-
ever, the conditions are somewhat different, and the concen-
tration much less pronounced.

Finally, Braus has lately described some stages in the

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 35553

ontogeny of the dorsal fin of Acanthias (5). But this fin is
too much modified to yield much for our purpose.

Development of the Median Fins of Scyllium
canicula,

Fig. 1 is a careful reconstruction from longitudinal serial
sections of a portion of an embryo about 18 mm. long.
Unfortunately, this specimen was cut short at the forty-ninth
segment, so that only the first dorsal fin is included.

At this stage is very well shown the origin of the muscle-
buds from the myotomes. One bud only is given off by each
myotome, not two as was surmised by Mayer. Already the
first steps in concentration are discernible in the convergence
of the buds towards a central region (about the forty-third
segment). The buds dwindle in size on both sides from this
region. They can be traced with certainty to the thirty-
second segment, and, more doubtfully, even beyond to the
twenty-eighth. There appear to be some eighteen buds in all.
The rudiment of the fin-fold itself, with its ridge of mesenchy-
matous tissue indicated by shading in the figure, extends over
at least a dozen segments, passing off gradually in front.

Fig. 25 is drawn with a camera from a specimen, 19 mm.
long, mounted whole in Canada balsam. The median fins are
here shghtly more advanced, but only the largest muscle-buds
can be made out clearly on this preparation owing to the
smaller ones being hidden below the edge of the myotomes.
The hinder edge of the first dorsal fin is about at the level of
the forty-third ganglion, and that of the second dorsal at the
fifty-seventh ganglion. The buds are rather more concentrated.

In fig. 4 are reconstructed the buds of two dorsal fins of
an embryo 19 mm. in length. It is important to notice that
there is, at this stage, no gap between the two fins. The
first bud passing towards the second dorsal lies immediately
behind the last given off towards the first dorsal fin.

The two dorsal fins of an embryo 24 mm. long are re-
constructed in fig. 2. In the second dorsal the origin of each

354 EDWIN S. GOODRICH.

bud from its myotome can still be traced, for the most part,
with ease, but in the first dorsal, which is a little more
advanced in development, concentration is more pronounced.
Here some of the posterior and anterior buds are seen to be
breaking up into irregular masses of cells, and are rapidly
losing their connection with, and becoming separated from,
the myotomes from which they arose.

Fig. 3, a reconstructed first dorsal of an embryo 26 mm.
long, shows a slightly different case of the same process of
concentration. The anterior buds have become separated off
in irregular masses, leaving slender stalks, probably nerve
rudiments, attached to the myotomes.

An embryo 28 mm. long (fig. 6) shows the muscle-buds
beginning to acquire their definitive structure. At their
peripheral ends they are still buds of embryonic epithelial
tissue; but towards the base of the fin they are becoming
changed into muscular tissue (indicated by a paler tint in the
reconstruction). A more detailed view of these growing
radial muscles is given in fig. 19.

At this stage we can already distinguish twelve well-marked
developing radial muscles, corresponding to twelve original
buds. That these become gradually converted into twelve
adult radial muscles, and were derived from the buds of
twelve myotomes, there can be no possible doubt. An ex-
amination of numerous intermediate stages proves it.

A mass of tissue derived from muscle-buds is becoming
converted into radial muscle at each end of the series.
Whether each of these masses is derived from a single bud
or from several it is extremely difficult to determine. As
already noticed, the buds at the extreme anterior and posterior
ends of the fin become irregular in shape and heaped up close
together, so that it is impossible to make certain how many
persist to form the adult muscles at these two points. ‘That
some few of the buds disappear altogether in the course of
development seems to be almost certain ; but possibly two, or
even three, persist at this stage.

The above description applies also to the development of

NOTES ON MEDIAN AND PAIRED FINS OF FISH. B00

the second dorsal, which differs only from the first in being
rather smaller, and in developing a little later.

Passing now to later stages (figs. 7, 11, 16, and 13), the
radial muscles are seen to become thoroughly differentiated,
retaining all the while their individuality. The little mass
described above at each end develops into a bundle of radial
muscle-fibres, which in some cases appears to represent only a
single segment. lor instance, in the dorsal fins of a Seylium
canicula, about a foot long, dissected in Naples, and shown
in fig. 23, there are only twelve muscles altogether. Possibly,
however, even here the anterior muscle is compounded of two
at least. As a rule the anterior and posterior bundles of
muscle-fibres show traces of subdivision in the adult.

The adult dorsal fins vary in their extent and in the
number of their constituent parts. Neopolitan specimens
generally have twelve radial cartilages, with ten or eleven
clearly-defined radial muscles (figs. 22 and 23). Dog-fish
from Plymouth usually have thirteen radial cartilages, with
eleven or twelve distinct muscles, not counting the complex
muscle bundle at each end. Sometimes there are fifteen
radials and sixteen muscles (fig. 26).

Development of the Skeleton of the Median Fins.

The first indication of the endoskeleton of the median fins
is seen in embryos about 30 mm. long. By this time the
muscle-buds are concentrated, but are only just beginning to
become converted into muscle-fibres. A slightly darker zone
appears near the base of the fin (fig. 17). Here the nuclei of
the mesenchymatous layer filling the fin-fold are rather more
closely crowded together. ‘This denser zone spreads a little,
and soon between each pair of right and left muscle-buds is
seen a dark streak of crowded nuclei, the first rudiment of
the radials (fig. 5); this embryo, 28 mm. long, is, however,
more advanced than the previous one. In an embryo
33 mm. long the procartilaginous rudiments of the radials
are clearly shown (fig. 20). The whole future skeleton is

356 EDWIN S. GOODRICH.

now faintly indicated; the radials have no definite outlines,
and merge together above and below. At this stage the
radial muscles are well differentiated (fig. 7).

Cartilage begins to appear near the middle of each radial
when the embryo is about 36 mm. in length (fig. 10), thence
it spreads upwards and downwards to the joints, where it
stops. The proximal and the distal elements are separately
chondrified. The dorsal, or distal, edge is the last to become
cartilage (fig. 15).

The skeleton of a first dorsal fin drawn in figs. 15 and
16 is of some interest, for unusual concrescence has taken
place. Not only are the first two and the last two radials fused
ventrally, as is usually the case, but also the 10th and the
11th, and the 10th has fused dorsally with the 9th.

With regard to the very origin of the skeleton—whether
the radials appear as separate rods or not—the evidence
is somewhat obscure. It is true that a patch of denser
mesenchyme is first seen (fig. 17), but it can hardiy be
called a plate of procartilage. It is merely a cloud of more
densely packed nuclei, and the radials, as such, make their
appearance as a series of denser zones separate from each
other, and extend outwards into a region not previously
occupied by the “plate.” This stage may, perhaps, more
justly be considered as representing their first appearance.
Their apparent continuity I believe to be due to their close
approximation, as Mollier has suggested (24).

The Development of the Paired Fins.

Little need be said about the paired fins, which have been
so thoroughly studied by others.

In Scylium each muscle-bud, having divided into an upper
and a lower secondary bud, develops into a pair of radial
muscles. These are regularly formed from their corresponding
buds throughout the greater part of the fin (figs. 1, 25, 6, and
9). But at each end, especially the anterior end, is an indis-
tinctly segmented mass of muscle-fibres, probably derived from

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 307

several buds. Asa rule twenty to twenty-two pairs of radial
muscles can be made out in the adult pectoral and pelvic fins.

A denser region of closely-packed nuclei at the base of the
fin-folds is the first sign of a skeleton, as in the case of the
median fins. hen the girdle, basals, and radials make their
appearance as procartilagimous tracts, all in continuity with
each other (fig. 9). The radials, however, arise as streaks of
denser mesenchyme, which are separate along the greater
part of their course. ‘They are continuous only at their base,
where they join the girdle or the basipterygium (meta-
pterygium). Even this basal shows faint indications of
segmentation as if it had been formed by the concentration
and fusion of the radials.

Cartilage is first formed in the girdle and basipterygium
(fig. 10), then inthe radials. The joints remain unchondrified.

The Concordance between the Muscles and the

Radials.

In the adult fins there is almost perfect “ concordance ” of
these elements in the peripheral regions, one cartilage being
lodged between two corresponding muscles. But, especially
at the anterior edge of the paired fins, and both at the front
and hind ends of the dorsals, the agreement is imperfect.
Here excessive concentration has taken place ; the cartilages
are possibly compound, and the muscles are indistinctly sub-
divided into bundles, which do not agree exactly with them
either in number or in shape (figs. 11, 12, 16, 15, and 26). The
muscles also generally extend beyond the cartilages. There
can be little doubt that most fins contain more muscular than
skeletal metameres. Not every segment which contributes a
muscle-bud necessarily contributes a skeletal radial.

Now, on tracing the development both of the paired and
of the median fins, we find that, so far as the two elements
co-exist, they are in exact correspondence. This concordance
of the muscles with the radials, which is, indeed, more perfect
in the young than in the adult, can be clearly demonstrated

358 EDWIN S. GOODRICH.

by sections, reconstruction, and whole preparations. Horizontal
sections through the dorsals, at a stage when the procarti-
laginous radials are just beginning to appear, show the con-
cordance quite plainly (fig. 8). In figures of longitudinal
sections, passing obliquely through two pelvic fins (fig. 18)
the correspondence is obvious. A reconstruction of one of
these pelvic fins, in which the ventral muscle buds are drawn
in the anterior half and the dorsal buds in the posterior half,
is no less conclusive (fig. 9). The same may be said of a
reconstruction of the first dorsal of an embryo 33 mm. long,
in which the radials are beginning to appear (fig. 7), and
also of reconstructions of both a pelvic and a dorsal fin, in
which the cartilage is developing (figs. 11 and 12).

The contention of Braus (4, 7, see p. 339 above), that the
concordance is secondarily established late in development,
is utterly at variance with all my observations.

Concentration.

The process of concentration in the median fins can be
followed on comparing the whole series of stages from the
earliest appearance of the buds to the adult condition (figs.
1-4, 6, 7, and 11).

At first the fin rudiment extends over as many segments as
produce buds—about sixteen to eighteen. ‘l'his ideal first
stage is, however, never perfectly recapitulated, since the
central buds develop first and become slightly concentrated
before the other buds have appeared (figs. 1 and 25).

In later stages the body segments lengthen much faster
than the fin rudiment, so that the buds have the appearance
of actively growing towards the base of the fin from both
sides. As a matter of fact, they probably remain passive
during the process. It is possible, however, that they may
grow to a slight extent towards the fin, but such a growth
would be very hard to prove. On the other hand, they
undoubtedly grow outwards into the developing fin-fold.

Both the fin-fold, with its contained muscle-buds, and the

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 309

body are growing rapidly; but they lengthen at a very
different rate, and it is to this fact that “concentration” is
due. An embryo 19 mm. long has the first dorsal fin
muscles extending over some fourteen segments (fig. 1), an
embryo 26 mm. long over about ten segments, an embryo
28 mm. long over about four and a half segments. Finally,
in the adult dog-fish, the base of the muscles occupies about
the length of three segments. A minimum of fourteen fin
segments has, then, been relatively concentrated into the
length of three trunk segments.

During concentration the hinder limit of the dorsal fins
remains approximately at the same place—about the level of
the forty-second and fifty-sixth ganglion respectively. The
two dorsal fins retain their relative distance during the whole
of development; the second dorsal is always fourteen segments
behind the first. Concentration takes place very much more
at the front than at the hind end of the fins. Their anterior
edge, then, moves backwards relatively to the body. Each
fin, as a whole, remains throughout in the same position.

In the case of the paired fins concentration takes place in
exactly the same way. But here it is not so pronounced, and
the apparent motion of the fins is less.

The Fusion of the Muscle-Buds.

It has been stated above that in the normally developed
region of the fins each muscle-bud gives rise to one radial
muscle. Mollier (24) describes and figures certain strands of
cells which at a certain stage unite the bases of the developing
radial muscles, suggesting that the adult muscle may con-
tain cells derived from several buds (p. 338, above). I find
similar strands joining the bases of the muscles in both the
paired and the median fins of Scyllium canicula (figs. 6 and
19). They are most conspicuous in embryos of about 28 mm.
in length. In the early stages of development, when the
buds are still of embryonic tissue, no such connecting bands
are present. On the other hand, when the radial muscles are

360 EDWIN &S. GOODRICH.

histologically differentiated, the connecting strands can no
longer be clearly seen, In late embryos the radial muscles
appear to be quite as distinctly separated from each other as
in the adult. But in these later stages one can find in
sections the network of nerve-fibres which in the adult fin
runs along the base of the muscles, and passes from one to the
other in a complicated intermuscular plexus.

It seems, therefore, highly probable that the connecting
strands of embryonic tissue found by Mollier, Braus (4), and
myself are really the rudiments of the nerve-plexus. I am,
unfortunately, unable actually to prove this; but there is no
doubt that the strands occur before the nerve-plexus can be
found, and at about the stage when we should expect it to
develop. At all events, there is no evidence that any muscle-
forming cells pass from one muscle-bud to another.

‘The Positien oe: the. Wine:

In estimating the exact position of the fins at various
stages in development only approximate results can be
obtained. It is not possible to compare different stages in
the growth of one individual, and there is considerable
variation amongst several. Moreover, it is probable that, in
the course of growth, segments may become incorporated
into the occipital region of the head, where myotomes and
their nerves may be reduced or obliterated. We can, there-
fore, never make quite certain that a given segment, say
the twentieth, in one adult dogfish corresponds to the
twentieth segment in another adult, or to the twentieth in
an embryo.

The position of the fins in an adult Scyllium canicula is
shown in text fig. 1, and in an embryo, about 19 mm. long,
in text fig.3. In the first the myotomes are not represented ;
in the second the nerves are omitted, the ganglia only being
indicated.

In enumerating the nerves of the adult the spinal nerve
issuing immediately behind the skull was counted as the first.

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 361

The thirteenth nerve is generally the last to contribute to the
pectoral fin plexus. Occasionally the fourteenth also sends a
twig, while sometimes the twelfth is the last of the plexus.
The second and third nerves generally send branches which,
together with the fourth nerve, pass through a foramen in
the girdle to reach the fin muscles. The nerves 4-15 pass
behind the girdle.

About eleven nerves supply the pelvic fin. Of these the
last usually belongs to the thirty-fifth segment and the first
to the twenty-fifth segment. The first three nerves may form
a collector passing through the girdle. The twenty-fourth
and twenty-third may also contribute some fibres in front,
and the thirty-sixth and thirty-seventh behind.

The plexus of the first dorsal fin is made of branches from
about the twenty-seventh to the forty-third nerves. Very
small twigs possibly enter into it from the twenty-sixth and
twenty-fifth nerves, but it is probable that these, and perhaps
also those of the twenty-seventh and twenty-eighth, are
merely sensory. The plexus of the second dorsal spreads
from about the forty-fourth to the fifty-seventh nerves.

In the embryo the first nerve was taken to correspond
to the first ganglion. Several small myotomes, some four or
five, occur in front of the first ganglion. They appear to be
represented in the adult by those small myotomes which lhe
in front of the first spimal nerve, and are supplied by the
spino-occipital nerves. Text figure 2 represents the condition
of the fins in the adult if concentration had not taken place.
The fins have here been deconcentrated.

Now, it is only necessary to compare these three diagrams
to see that the position of the fins has remained approximately
the same throughout development. Concentration, however,
has brought about considerable apparent shifting of the
pelvic fins, but there is a fixed point in the neighbourhood of
nerves 28-30. In the case of the pectoral fin the drawing
back of the anterior margin of the fin has been almost entirely
compensated by the drawing forward of the posterior margin,
so that in spite of great concentration the position of the fin

ee te,

, AA

i=

“ou

E
= !
: rst
oO / :
ro)
o | |
a | i) :
s TTT TALL pin) b Ach ab | :
e : z) “Us LL a yer veNyyyngmngae cru ju
S| 4ou y : ; : ere z

+490 tou ar

seu
d ay
e : i
7 an AARNE NNT ee rt ; Ty) AAACN
“or ne Me \S
Po. HIN) \) SN }).)) Wut :
g LL iy) Se ULE yp)
oO “ar Midd £6 CIAG (Ld bs bbbhett CAT
soi pz panied eee seu a as ¢

Pp!

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 363

isunchanged. Much more pronounced is the apparent shifting
backwards of the dorsal fins. While their hinder margin is only
slightly carried forwards, the anterior margin retreats over
some eight or nine segments. Here, again, there are fixed
points about the fortieth and fifty-fourth nerves which do
not move at all (pp. 345-352).

On tHE NERV#-SUPPLY OF THE FINS.

We must now more closely examine the structure of the
plexus of nerves which supply the fins, dealing more particu-
larly with the median fins.

Mayer (22), to whom I am much indebted for many useful
hints on the best methods for this purpose, has described and
figured the nerve-plexus of the median fins of many Elasmo-
branch fish. But he did not follow out in detail the relation
of the nerves to the fin muscles and to the body segments.
With the object of continuing and extending his researches, I
have dissected the nerve-plexus in a large number of speci-
mens. [or this purpose material has been used after treat-
ment with hot water, or after maceration in weak nitric acid.
Osmic acid added to these preparations brings out the nerves
most distinctly. Owing to the delicate and complex nature
of the plexus of the median fins and to the very brittle state
of the nerves, it is very difficult indeed to obtain a perfectly
complete dissection of the plexus in a single specimen. The

EXPLANATION OF TEXT-FIGURES.

Text-figure 1.—Diagram of an adult Scyllium canicula, showing the
nerve-supply of the fins.

Text-figure 2.—Diagram of an adult Scyllium canicula. The fins are
expanded, and their nervous, muscular, and skeletal seemental elements are
distributed as if concentration had not taken place. The nerve foramina in
the girdles are indicated by shaded oval areas; the girdles themselves are not
shown.

Text-figure 3.—Diagram of an embryo Scyllium canicula about 19 mm.
long, in which are shown the ganglia, the myotomes, and the muscle-buds.
a, Anal fin; ae, anterior collector of first dorsal fin ; ev, cartilaginous radial
projecting beyond the radial muscles; ~ 1-57, spinal nerves and ganglia;
pe, collector nerve of second dorsal fin; p/, pelvic fin; pé, pectoral fin; 7m,
radial muscle; 1d and 2d, first and second dorsal fins.

364 EDWIN S. GOODRICH.

plexus supplying the paired fins is stronger and much less
difficult to expose. The pectoral and pelvic plexus have been
admirably described and figured by Braus (8, 6) in a
large number of Elasmobranchs, while Punnitt (29, 30)
has studied the pelvic plexus in Mustelus and Acanthias.

The nerve-plexus of the paired fins is very variable, both
as regards the number of nerves which contribute towards it
and the exact course of its secondary branches. Such is also
the character of the nerve-plexus of the dorsal fins; but here
it is less easy to decide as to the exact number of nerves which
enter into its composition. In minor details no two specimens
correspond, and even the two sides of the same individual
may differ considerably. However, on the whole, there is
great constancy in the character and metameric value of the
plexus of the median fins, as is shown by comparing a large
number of specimens. Unfortunately, it is so complex, and
the nerve branches are so fine, that I have not found it pos-
sible to trace out its formation in ontogeny.

The plexus of the dorsal fins.—We have seen above
that a dorsal fin contains some fourteen muscle segments.
We should, therefore, expect at least fourteen spinal nerves
to join in its formation. Moreover, smce concentration takes
place to a much greater extent in front than behind, we should
expect the longitudinal collector, formed by the gathering
together of various nerve components, to be situated chiefly in
front of the fin-base. Now this is just what dissection reveals.

Figs. 21 and 22 show the general nerve-supply of the two
dorsal fins. A comparatively stout collector is seen to run
forwards from the base of each fin (alc.). It is composed of a
number of twigs derived from the rami dorsales of some dozen
seoments. The collector increases in bulk as it passes back-
wards, and more nerves enter into it. Where the collector
begins is often very difficult to determine, in the case of the
first dorsal especially ; for its first components are so extremely
slender that they are very hard to discriminate from the inter-
crossing plexus of nerve-twigs which are present all along the
median dorsal septum.

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 365

The collector gives off branches to the fin as soon as it
reaches its base, and often ceases about half way down the
fin. Then come one or two nerves which give off branches
independently to the fin. In many fins all the nerves passing
to the fin are joined together by communicating branches,
continuations of the collector (fig. 26).

At the hind end of the fin are one or two nerves with a
short, and often ill-defined, posterior collector (ple.).

When pterygial nerves, passing out from the collectors,
reach the base of the radial muscles, they run in amongst
them and branch repeatedly. A plexus of extraordinary
complexity is thus formed round and through the muscles
and along the cartilages outwards to the web of the fin.

We find, then, that some fourteen to sixteen spinal nerves
undoubtedly contribute to the innervation of the first dorsal
fin, and that the rami pterygiales of those situated in front of
the middle of the fin-base always combine to a longitudinal
collector. The collector clearly shows that the radial muscles
derived from these segments have been displaced backward.
The one or two rami pterygiales gathered into a posterior
collector indicate a similar but very much less extensive con-
centration forwards.

An examination of the nerve-supply of the second dorsal
fin yields the same results. The anterior collector of this fin
begins immediately behind the posterior collector of the first
dorsal.

The anatomy of the adult fully bears out the conclusion
arrived at from a study of the development ; the dorsal fins
are made up of a large number of greatly concentrated
segmental elements—muscular, nervous, and skeletal. The
lateral fold theory is, then, strongly supported by our know-
ledge of the structure and development of the median and
paired fins, since the paired fins have long been known to be
constructed and developed on exactly the same principle.

But there remains one important, though not essential,
question to discuss: How far is the original metameric struc-
ture preserved in the adult ?

vou. 50, PART 2,—NEW SERIES. 26

366 EDWIN S. GOODRICH.

With regard to the musculature, we have already decided
that there is no definite evidence that the metamerism is lost
(p. 359). The skeleton is still obviously sezmentally divided
in the dorsal fins of Scyllinm, in spite of the slight con-
crescence of some of the radials at their base. The radials
of the paired fins have undergone much greater fusion and
modification. But there is nothing in their structure or de-
velopment which precludes the idea that even the basals were
once metamerically segmented. In modern sharks, however,
this segmentation of the skeleton of the paired fins is to a
ereat extent lost. It is to the nerves that appeal is generally
made for evidence against metamerism (pp. 338-340) ; let us,
therefore, examine further the nerve-supply of the fins.

On the real nature of the nerve-plexus.—Many
anatomists seem to consider that the nerve-plexus is formed by
a combination of several nerves, which lose their individuality,
and are then redistributed to the linb, somewhat as a number
of blood-vessels may anastomose and supply a gland. Insuch
a case the nerves would be so mixed in the plexus that even
their motor fibres might lose all trace of metamerism.

But such is not really the case, even in the highest verte-
brates, as Herringham (21), Patterson (27, 28), and others
have shown.

Now, we may well ask whether in the Elasmobranch there
is really any motor plexus at all, if by plexus is meant a
mixing of nerve-fibres bringing about a disturbance or
destruction of the original metamerism. If the nerves could
be traced to each radial muscle of a fin, it would be easy
enough to prove whether or not it isthe case. Unfortunately,
dissection can help us but little in settling this point. Most
of the nerves to the paired fins pass directly to the fin-base ;
but as soon as they reach it they become joined together by a
complicated system of connecting nerves, even before they
enter the muscles. When they reach the latter they become
involved in such a complex network that it becomes impos-
sible to determine for certain whither the nerve-fibres lead.
That on the whole each nerve supplies two radial muscles in

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 367

regular order can be fairly well established; but it cannot
be asserted that they do not also supply others. In fact, I
have found it impossible to prove by mere dissection that
these muscles are haploneurous. Nevertheless, it can be shown
that the muscles of the paired fins are innervated in regular
order from before backwards by the spinal nerves, each of
which supplies a pair above and below.

Turning to the dorsal fins, we find that not only do the
rami pterygiales form a longitudinal collector, in which it is
impossible to follow out for certain the nerve-fibres from indi-
vidual segments, but also that the branches running to the
fin from the collector form a plexus of even more complicated
structure. Over and over again have I tried in vain to follow
the nerve-fibres from a spinal nerve to a radial muscle. It
must be remembered that the rami pterygiales are nerves of
mixed character, containing motor and sensory fibres. The
real difficulty is, not to trace a branch to a muscle, but to
make sure that no motor fibres from that nerve pass on
elsewhere to other muscles along the ramifying twigs of the
plexus.

Having failed to analyse the plexus of the dorsal fin by
dissection, it remained to be seen whether any fin could be
found in which the motor fibres are distinguishable from the
sensory. Such a condition I discovered in the ventral lobe
of the caudal fin of Scyllium.

The small radial muscles with which this lobe is provided,
unfortunately, do not develop from regular muscle-buds, so
they cannot be traced in ontogeny to the myotomes. They
are subdivided into a large number of small bundles, much
more numerons than the segments, and are developed from
cells which come off from the proliferating lower edge of the
myotomes. The same thing occurs in the anal fin.

Fortunately, in the tail of Scyllium the nerves from the
ventral motor roots do not combine into mixed trunks with
those from the dorsal sensory roots. Both the motor and the
sensory branches pass obliquely downwards to the base of
the fin. Here they form an elaborate plexus (fig. 24), in

368 EDWIN S. GOODRICH.

w hich can be distinguished a large, longitudinal “ collector”
and twigs running outwards to the radial muscles and skin of
the fin. Now, by careful dissection, under the high powers
of the binocular microscope of Zeiss, one can follow out the
motor and the sensory fibres to their destination. It soon
becomes evident that, while the latter combine to form the
longitudinal trunk and the plexus of anastomosing nerves,
which send branches at intervals to the skin (fig. 24), the
motor fibres pass through the plexus without really becoming
involved in it. Each spinal nerve sends down motor branches
supplying a considerable number of radial muscles. It is, of
course, by no means easy to follow out every twig to its
ending; but from a careful and minute study of several tails
I have satisfied myself that the motor branches of one seg-
ment do not anastomose or mix with those of another segment
—the area supplied by one motor root begins where that of
another ends, Inthe specimen figured there is one twig (marked
with an *), in two segments, which seems to join one segment
to the next behind; but I am inclined to believe that the
fibres do not mix peripherally. In other tails investigated
since, I have found no such junction. At all events, the
facts are quite compatible with the view that no mixture
takes place.

There appears, then, to be no such thing as areal motor
plexus in the caudal fin. Whether there is a true sensory
plexus, or whether it is more apparent than real, I am unable
to determine for certain, as the fibres cannot be disentangled.

Seeing that the so-called motor “plexus” in the caudal is
probably only apparent, we may well ask whether a true
motor plexus exists in any of the dorsal or paired fins. May
it not be that here also the motor fibres pass through a sen-
sory network and do not lose their original metameric order ?

IT am strongly of opinion that this is the case, and that the
radial muscles are haploneurous, the original metamerism of
the fin being preserved in the adult. Since this question
cannot be answered by anatomy, we must appeal to experi-
ments on the living tissues.

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 369

Experiments on the nerve-supply of the fins.—
While occupying the British Association table at the Naples
Aquarium last winter, I had the opportunity of conducting
some experiments with a view to tracing out the nerve-supply
of the radial muscles. I have to thank Prof. Gotch for advice
as to the best way of stimulating the nerves, and Mr. G. W.
Smith for helping me to carry out the experiments.

If in the limb-plexus the motor fibres of various segments
were crossed or mixed, and if, as some authors contend, the
radial muscles were polyneurous, we should expect to produce
a general, or at least an extensive, contraction of the fin-
muscles on the stimulation of one nerve. ‘To test these views
experiments were made on the pectoral fin of Raja. Owing
to its great size, the muscles and nerves of this fin are
peculiarly well adapted for the purpose.

The first series of experiments were made by directly
stimulating individual spinal nerves, and watching the con-
traction of the radial muscles. ‘lo insure definite and correct
results, the nerves were first severed near the spinal cord, and
were then stimulated in various ways at their proximal end,
the necessary precautions having been taken to keep the
tissues in good condition. The electric stimulus is the easiest
to use ; but a mechanical stimulus, applied either by snipping
the nerves with scissors, by higaturing with a thread, or by
pinching with ivory forceps, gives, perhaps, the surest
result.

It was found that the series of radial muscles of one side
could be made to contract regularly, in pairs, from before
backwards, by stimulating the successive nerves of the plexus,
beginning at the anterior end. Similarly the muscles contract
regularly in pairs from behind forwards on stimulating the
nerves in the reverse direction. It was determined with
absolute certainty that the stimulation of one nerve does
not produce a general contraction of the muscles of the fin,
but only of a limited portion of the musculature corresponding
in position to the nerve.

Owing to the fact that the radial muscles lie very near

370 EDWIN S. GOODRICH.

together, closely bound to each other and to the skeleton by
the connective tissue which surrounds them, it is difficult to
establish, without the possibility of doubt, that the contraction
is restricted to two radial muscles only. Nevertheless, after
repeated trials, I am quite convinced that such is really
the case. There can be no doubt whatever that the con-
traction does not spread to several neighbouring
muscles.

If all the nerves, excepting one or two, are severed from
the spinal cord, and if then a general stimulation be induced
through the cord, only those pairs of muscles contract which
correspond to the one or two nerves left intact.

The following experiment was also repeatedly made to
determine whether the nerve-supply of neighbouring segments
overlaps. ‘Three consecutive nerves of the plexus, A, B, and C,
were severed from the spinal cord. The two outer ones, A
and C, were then excited by constant application of the
electric stimulus until the corresponding pairs of radial
muscles scarcely, if at all, responded. Then the middle nerve
B was stimulated, and its muscles were found to respond in
perfectly normal fashion. They contracted equally well
whether the outer nerves were still being stimulated or not.
This seems to me to prove, without the possibility of doubt, if
not that there is no overlap whatever, at all events that it can
only be very slight.

So clear and definite was the evidence derived from these
and other experiments of a like nature, that I have no hesita-
tion in stating my opinion that each pair of radial muscles
(containing two dorsal and two ventral elements) derived
from a single segment, is supplied exclusively with motor
fibres from the ventral root of the nerve belonging to that
same segment. In fact, no plexus exists in the pectoral fin
of Raja in the sense of a mixture or overlapping of the areas
supplied by the segmental motor nerves.

So far as experiments were conducted on the pelvic fins
they gave the same results.

Unfortunately, the dorsal fins of Scyllium do not lend them-

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 371

selves so readily to experiment. I have not yet been able to
apply so decisive a set of tests to the delicate “plexus” sup-
plying these fins. But it can be easily shown that the suc-
cessive stimulation of the rami dorsales sharing in the
“plexus” induces the successive contraction of the series of
corresponding radial muscles.

It results from these experiments that the metamerism of
the fin elements may remain undisturbed in the paired, and
probably also in the unpaired, fins of Hlasmobranchs. In
this they agree with the evidence of embryology.

SUMMARY AND CONCLUSION.

The chief observations described above may be summarised
as follows: The development of the median dorsal fins is
essentially similar to that of the paired fins. They arise as
longitudinal folds, into which grow buds from the myotomes.
Some fourteen or sixteen myotomes contribute to the fin each
one muscle-bud. Concentration sets in almost from the first
appearance of the buds; it is chiefly, if not entirely, due
to the body growing faster than the fin. Along the greater
part of the dorsal fin each muscle-bud becomes converted
into one radial muscle. At the extreme ends of the fins the
exact metameric origin of the muscles is difficult to trace and
is somewhat obscured. Only here fusion of neighbouring
segmental buds perhaps takes place. At certain stages slender
strands of embryonic tissue connect the bases of the radial
muscles; these are probably rudiments of the nerve-plexus.
Neither the study of development nor of the adult structure
affords any definite evidence that the primitive metamerism
of the musculature is lost. Experiments seem to establish
that the radial muscles remain haploneurous, retaining their
primitive connection with the nerve belonging to that myotome
from which they have been developed. ‘The nerve- “ plexus”
of the fins is composed of intertwining sensory fibres, along
or through which the motor fibres proceed to their destina-
tion without mixing with those of other segments. There is

ale EDWIN S. GOODRICH.

probably no real motor plexus, but the motor nerves may be
gathered together into more or less longitudinal collectors,
and become again sorted out on reaching the musculature.
Such collectors are found at the base of the dorsal fins, com-
pounded of some fourteen to sixteen segmental rami pterygiales.
All the fins remain throughout development in approximately
the same position. Apparent change of place may be brought
about by concentration being greater in the one direction
than in the other. ‘This is especially the case with the dorsal
fins, the anterior edge of which may undergo a relative shift-
ing over some ten segments.

The general bearing of these results has been sufficiently
discussed in the Introduction (p. 334), and need not again be
dealt with here. But it may be pointed out how completely
they support the lateral fold theory of the origin of the paired
fins.

List oF REFERENCES.

1, Batrour, F. M.—“ A Monograph on the Development of Elasmobranch
Fishes,” ‘Journ. of Anat. and Phys.,’ 1876-8.

2. Bateour, F. M.—‘On the Development of the Skeleton of the Paired
Fins of Elasmobranchii,” ‘ Proc. Zool. Soc. London,’ 1881.

3. Braus, H.— Ueber die Innervation der Paarigen Kxtremititen bei
Selachiern,” ‘Jen. Zeitschr. fiir Naturw.,’ Bd. xxix, 1895.

4. Braus, H.— Beitrage zur Entwick!. d. Muskulatur u. d. periph. Nerven-
system d. Selachier,” ‘ Morph. Jahrb.,’ Bd. xxvii, 1899.

5. Braus, H.— Thatsachliches aus der Entwickl. d. Extremitatenskelettes,”
‘ Haeckelfestschrift Deukschr. Med.-Nat. Ges.,’ Jena, 1904.

6. Braus, H.—‘“‘ Die Muskeln u. Nerven der Ceratodusflosse,” ‘ Zool.
Forsch. Denkschr. Med.-Nat. Ges.,’ Jena, 1901.

7. Braus, H.—* Die Entwickl. d. Form. d. Extremitiaten u. d. Extremi-
\atenskeletts,”’ ‘Hertwig’s Handbuch d. vergl. u. exp. Kutwick.
der Wirbeltiere,’ 1904.

8. Davivorr, v. M.—“ Beitrage z. vergl. Anatomie der hinteren Gliedmasse
der Fische,” ‘ Morph. Jahrb.,’ Bd. v, 1879.

9. Dean, B.—“ Historical Evidence as to the Origin of the Paired Limbs
of Vertebrates,” ‘ Am. Nat.,’ v. xxxvi, 1902.

LO.

Ly.

12.

13.

14.

15.
16.

17.

18.

19.

20.

al.

22.

23.

24,

25.

26.

27.

28.

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 373

Dourn, A.—‘ Studien z. Urgeschichte des Wirbelthierkérpers: VI. Die
Paarigen u. Unpaaren Flossen der Selachier,”’ ‘ Mitth, Zool. Sta.
Neapel.,’ Bd. v, 1884.

Firprincer, M.—‘‘Zur Lehre von den Umbildungen des Nervenplexus,”
‘Morph, Jahrb.,’ Bd. v, 1879.

Firprmncer, M.—“ Ueber die Spino-occipitalen Nerven,” ‘ Festschrift
fiir Carl Gegenbaur,’ Bd. iti, 1897.

FUrprincer, K.—“ Beitrage z. Morph. des Skeletes der Dipnoer,”
‘Zool. Forsch. Deukschr. d. Med.-Nat. Gesell.,’ Jena, 1904.

GucenBaur, C.—‘ Untersuchungen z. vergl. Anatomie der Wirbelthiere,’
II, Leipzig, 1865.

GGeGENBAUR, C.—“ Zur Gliedmassenfrage,” ‘Morph. Jalrb.,’ Bd. v, 1879.

GecenBaur, C.—“ Das Flossenskelet der Crossapterygier u. das Archi-
pterygium der Fische,” ‘ Morph. Jahrb.,’ Bd. xxii, 1895.

Guiret, I'.—“ Recherches sur le développement des nageoires paires du
Cyclopterus lumpus, L.,” ‘Arch. u. Zool. exp. et genér.,’ 3rd
série, v, 1896. i

Jugrine, v.—‘ Das Periph. Nervensystem d. Wirbelthiere,’ Leipzig,
1878.

Harrison, K. G.—‘‘ Die Entwickl. der Unpaaren u. Paarigen Flossen
der Teleostier,”’ ‘Arch. f. Mikr. Anat.,’? Bd. xlvi, 1895.

Haswei, W. A.—*“‘ On the Structure of the Paired Fins of Ceratodus,”
‘ Proc. Linn. Soc. New South Wales,’ vol. vii, 1883.

Herrincuam, W. P.—‘ The Minute Anatomy of the Brachial Plexus,”
‘Proc. Roy. Soc.,’ vol. xli, 1886.

Mayer, P.—‘‘ Die Unpaaren Fiossen der Selachier,’ ‘ Mitth. Zool. Sta.
Neapel.,’ Bd. vi, 1886.

Mivart, St. G.—‘‘ Notes on the Fins of Hlasmobranclis,” ‘Trans. Zool.
Soc. London,” vol. x, 1879.

Mouurer, 8.—“ Die Paarigen HExtremitaten der Wirbelthiere: I. Das
Ichthyopterygium,” ‘ Anat. Hefte,’ Bd. i, 1898,

Mo.uter, 8.—“‘Die Paarigen Extremitaten,” ete.: LI. “ Das Cheirop-
terygium,” ‘ Anat. Hefte,’ Bd. v, 1895.

Mo.uier, §.—‘ Die Paarigen Extremitaten,” ete.: II]. “Die Entw. d.
Paar. Flossen des Stérs,” ‘ Anat. Hefte,’ Bd. viii, 1897.

Patrerson, A, M_—‘‘ The Position of the Mammalian Limb,’’ ‘Journ.
of Anat. and Phys.,’ vol. xxii, 1889.

Patterson, A. M.—“ The Origin and Distribution of the Nerves of the
Lower Limb,” ‘Journ. of Anat. and Phys.,’ vol. xxviii, 1893-4.

374 EDWIN S. GOODRICH.

29. Punnett, R. C.—“On the Formation of the Pelvic Plexus, ete., in
Mustelus,” ‘Phil. Trans, Roy. Soc. London,’ vol. B excii, 1900.

30. Punnett, R. C.—*On the Composition and Variations of the Pelvic
Plexus in Acanthias vulgaris,” ‘Proc. Roy. Soc. London,’
vol, Ixix, 1901.

31. Razi, C.—‘‘ Theorie d. Mesoderms,” ‘Morph. Jahrb.,’ Bd. xix, 1892.

32. Ruce, E.—* Die Entwickl. Skeletes der vorderen Extremitatem von
Spinax niger,” ‘ Morph. Jahrb.,’ Bd. xxx, 1902.

33. Semon, R.—‘‘ Die Entwickl. d. Paarigen Vossen des Ceratodus
forsteri, ‘Zool. Forsch. and Jen. Denkschr.,” Bd. iv, 1898.

34, SuERRINGTON, C. S8.—‘ Notes on the Arrangement of some Motor
Fibres in the Lumbo-sacral Plexus,” ‘Journ, of Phys.,’ vol. xiii,
1892.

35. Tuacner, J. K.—‘‘ Median and Paired Fins,” ‘ Trans. Connecti. Acad.,’
vol. iii, 1877.

86. WiepersurimM, R.—‘ Das Gliedmassenskelet der Wirbelthiére,’ Jena,
1892. ;

87. Woopwarp, A. S.—‘* The Evolution of Fins,’ ‘ Natural Science,’ 1892.

EXPLANATION OF THE PLATES.

Illustrating Mr. E. 8. Goodrich’s paper, “ Notes on the Deve-
lopment, Structure, and Origin of the Median and Paired

Fins of Fish.”

LETTERING OF THE FIGURES.

a. f. Anal fin. a.l.c. Anterior longitudinal collector nerve. az. Con-
necting strand of tissue. dp. Basipterygium. 47. ar. Branchial arch. ca.
Nerve canal. c. 7. Cartilaginous radial. c. ¢. Connective tissue. d. /.
Dorsal fin. d.7. Dorsal root. g. Ganglion. 4. m. Hypoglossal muscula-
ture. 7. s. c. Longitudinal sensory collector nerve. m. Myotome. a, 6.
Muscle-bud. m. f. Motor fibres. 2. Nerve. a. a. Neural arch. x. ed.
Nerve-cord. p. g. Pelvic girdle. p. 7. c. Posterior collector nerve. p. r.
Procartilaginous radial. pt. Pectoral fin. pv. 7 Pelvic fin. 7. Radial or
somactid. 7. m. Radial muscle. 7. pf. Ramus pterygialis. s. First indica-
tion of the skeleton. s. Sensory fibres. v. Vagus. v. 7. Ventral root.
z. is placed in front of a number which could not be accurately determined.

NOTES ON MEDIAN AND PAIRED FINS OF FISH. 370

All the figures refer to Scyllium canicula, and the arrows point towards
the head. In several of the figures the myotomes, their muscle-buds, and the
radial muscles are drawn in red. Blue in fig. 24 represents sensory nerves.

The spinal nerves in the embryo are numbered from the first ganglion, in
the adult from the first which issues behind the skull. About five myotomes
are found in embryos in front of the first ganglion. When portions of the
embryos were cut off before imbedding, the number of the ganglia was esti-
mated, and an & inserted before it to denote its uncertainty, as in fig. 4.

PLATE 10.

Tic. 1.—Reconstruction from serial longitudinal sections of a portion of an
embryo about 18mm. long. It is cut off behind the first dorsal fin. The
pectoral fin has been cut off near its base.

Fic. 2.—Reconstruction of the first and second dorsal fins of an embryo
24 mm. long,

Fic. 3.—Reconstruction of the first dorsal fin of an embryo 26 mm. long,

Vic. 4.—Reconstruction of the first and second dorsal fins and of a portion
of the pelvic fin of an embryo 19 mm. long.

Fic. 5.—Longitudinal vertical section of the hinder region of the first
dorsal fin of an embryo 28 mm. long. ‘The radials are beginning to appear ;
the extremities of several muscle-buds are seen above. ‘Cam. Ob. Z.,’
a a, OC. 3.

PLATE 11.

Fic. 6.—Reconstruction of a portion of an embryo 28 mm. long (same as
that in fig. 5), with the first dorsal and pelvic fins.

Fic. 7.—Reconstruction of the first dorsal fin of an embryo 33 mm. long.
Both the muscles and the procartilage radials are represented.

Fie. 8.—Longitudinal horizontal section of the first dorsal fin of an embryo
382 mm. long. Z. A. oc 3, Cam.

Fic. 9.—Reconstruction of the pelvic fin of an embryo 28 mm. long. ‘The
procartilage skeleton is represented complete; the anterior twelve ventral
muscle-buds and ten posterior dorsal muscle-buds are indicated. At this
Stage the base of the muscle-buds is being converted into contractile tissue,
which is not represented in the figure.

Fic. 10.—Recoustruction of a portion of the vertebral column, the skeleton
of the first dorsal, and of the pelvic fin of anembryo 37.mm. long. The cartilage
is beginning to develop.

Fics. 11, 12.—Reconstructions of the first dorsal fin (fiz. 11), and the

pelvic fin (fig. 12), of the series represented in fig. 10. The muscles are out-
lined in red and the skeleton in black.

376 EDWIN S. GOODRICH.

PLATE 12.

Fic. 13.—The skeleton and muscles of the first dorsal fin of an adult.

Fic. 14.—Oblique longitudinal section of the pelvic fin reconstructed in
fig. 9. Traces of segmentation extend into the basipterygial region.

Fic. 15.—Reconstructed skeleton of the first dorsal fin of an embryo
55mm. long. Procartilage extends along the dorsal edge.

Fic. 16.—Portion of the vertebral column, skeleton of the first dorsal fin
(without the procartilage), and radial muscles of the embryo 55 mm. long
represented in fig. 15.

Fic. 17a.—Reconstructed muscle-buds of the first dorsal fin of an embryo
30 mm. long.

Fic. 178.—Longitudinal vertical section of the first dorsal fin, drawn to the
same scale as fig. 17a and taken from the same series, showing the first indi-
cation of the skeleton.

Fie. 18.—Longitudinal section of the pelvic fin of an embryo, 33 mm. long.

Fic. 19.—The base of three radial muscles, showing connecting strands of
tissue (az.). Embryo 28 mm. long. ‘Cam. Ob. Z. D.,’ oc. 2.

Fic. 20.—Longitudinal vertical section of the first dorsal of an embryo
33 mm. long, reconstructed in fig. 7.

PLATE 13.

Fig. 21.—Dorsal branches of the spinal nerves 24-57 of an adult, showing
the longitudinal collectors near the dorsal fins.

Vig. 22.—Similar figure of an individual about 25cm. long. The skeleton
of the fins is indicated.

Fic. 23.—First dorsal fin of the same individual showing the radial
muscles.

Fic. 24.—Nerve-supply of the anterior region of the ventral lobe of the
caudal fin of an adult. The muscles are shown only in one part. The motor- .
nerves are in black and the sensory are drawn in blue. * indicates twigs
which appear to mix with neighbouring nerves.

PLATE 14.
Fie. 25.—Embryo, 19 mm. long, mounted in Canada balsam. Cam.

Fic. 26.—Skeleton (indicated by a dotted line), and muscles of the first
dorsal fin of an adult. The nerve-plexus at the base of the fin is shown and
the beginning of the collector of the second dorsal fin.

Fic. 27.—Diagram to illustrate the transposition of limbs.

NEW ORGAN IN PERIPLANETA ORIENTALIS. B77

Preliminary Account of a New Organ in
Periplaneta orientalis.

By

Ruth MM. Harrison,
Lady Margaret Hall, Oxford.

With Plate 15.

Wuite dissecting a cockroach in the Zoological Laboratory
at Oxford I noticed a pair of small oval pouches lying below
and to either side of the nerve-cord between the fifth
and sixth abdominal ganglia. Being unable to find either
description or figures of any such structure, I proceeded to
examine it further by means of dissection and serial sections.

The position of the organ is shown in fig. 1. This is a
dissection of the posterior segments of a male cockroach
from which the tracheal system has been removed. In a
freshly-killed specimen it has a yellowish, transparent appear-
ance which renders it somewhat inconspicuous; but it was
present in every full-grown cockroach that I examined, both
male and female. In the male it measures about 2 mm. in
length, but in the female it is very much smaller, never
being more than about half the size of that of the male;
sections, however, show no histological difference. If the
nerve-cord is stretched to one side, the pouches, which before
appeared to be separate, are seen to be two lobes projecting
upwards and forwards from a median structure, which opens
below to the exterior by a single aperture on the ventral
surface of the animal between the sixth and seventh sternites.
This is clearly seen by a comparison of figs. 4 and 5, which

378 RUTH M.. HARRISON,

are transverse sections through the same specimen. The
section represented in fig. 4 is taken just behind the fifth
abdominal ganglion, and shows that anteriorly the pouches
are quite separate ; fig. 5 represents a more posterior section,
showing that the two lobes unite to form a single pouch.
The median portion of the anterior border of the seventh
sternite is hollowed out in the form of a crescent, the edge of
which forms a distinctly thickened rim of chitin; this is seen
in longitudinal section in fig. 3 and in transverse section in
fig. 5. Fig. 3 also shows the opening to the exterior in its
natural position; the sternites overlap to a considerable
extent, so that only half of each plate is exposed to a surface
view; this gives the appearance in section of a long duct
leading to the exterior, but the limit of the organ is marked
by the thickened chitinous rim on sternite 7.

A dorsal view of the structure has the appearance repre-
sented in fig. 2. In this figure the nerve-cord has been re-
moved altogether, but the tracheal system has been preserved
intact. The paired main ventral longitudinal vessels are seen
to lie above the organ, and a complicated system of smaller
vessels is distributed over its entire surface.

In preparing specimens for microscopic examination various
methods of softening or dissolving-the chitin-were attempted,
but the most. satisfactory results. were obtained by painting
with collodion and cutting through the chitin, thus preserv-
ing the soft structures in their natural position. . Corrosive-
acetic and Perenyi’s chromo-nitric proved the best fixatives,
and borax carmine and hematoxylin were used as stains.

In section the structure of the wall of the organ at once
suggests some kind of gland. Round the anterior lobes and
onthe dorsal surface is a layer of elongated cells with their
long axes vertical, and having large rounded nuclei placed
towards the periphery; the cytoplasm. shows a very fine
granular consistency, and in each cell is a stoutish tube, one
end of which appears to open near the nucleus, and the other
is directed towards the lumen of the gland. Round this tube
the cytoplasm is denser, the granules being here more closely

NEW ORGAN IN PERIPLANETA ORIENTALIS. 379

accumulated. Ina surface view of the wall these cells are
seen to be hexagonal in shape and to fit together with great
regularity. Towards the external opening these cells pass
imperceptibly into the epithelial layer below the cuticle.
Within this layer isa mass of tissue where all cell outline has
become entirely obliterated ; the nuclei are quite irregular in
shape, and the cytoplasmic portion of the cell is represented
by scattered fragments bounding large vacuoles. Lining the
whole cavity internally is a thin transparent membrane, con-
tinuous with the chitinous skeleton of the insect. ‘This appears
to be of the nature of chitin since it remains unaffected by
the action of potash. It is very much folded and crumpled,
and attached to it, projecting away from the lumen of the
gland, are numerous bunches of extremely long and fine hair-
like processes, which also seem to be chitinous, remaining
unchanged by maceration in potash. ‘These are not very
obvious in sections except at the point where each hair is
attached to the lining membrane, for here a distinct minute
circular spot is visible. In a preparationin which the proto-
plasmic structures have been dissolved in potash these pro-
cesses become clearer. They are attached irregularly in little
groups of two or three to about a dozen or more, while some
occur singly.

From the above description the glandular nature of the
organ becomes apparent. It would seem that the secretion
is stored in the form of granules in the cells of the outer
layer, and that these cells migrate inwards—i. e. towards the
lumen of the gland (figs. 6, 7), where they disintegrate.

In a young specimen the gland consists of a single layer
of epithelial cells of a non-granular character with large
oval nuclei, lined by a comparatively broad belt of chitin
continuous with the cuticle of the animal. The non-granular
condition of the epithelial layer, and the entire absence of
any intermediate tissue like that present in the adult organ,
is doubtless due to the fact that the gland has not yet become
actively secretive. An inspection of the sections represented
in figs. 4, 5 seems to afford evidence that the cells do actually

380 RUTH M. HARRISON.

migrate inwards and disintegrate; the animal from which
this gland was taken was a fine large male, and the scarcity
of granular cells may possibly be explained by the fact that
the organ is ceasing to be functional, and is in a degenerate
condition. The nature of the secretion and the function of
the whole organ I have not so far investigated; but this I
hope to do, together with a further study of the individual
elements and their relations to each other.

The curious tubes in the outer granular layer bear a strong
resemblance to certain tubes described and figured by Claus
in a stink-gland of the larva of the Coleopteron Chrysomela.
But the hairs are quite unlike those described by Minchin in
the dorsal glands of this same species of Periplaneta, or
those described by Kraase in a dorsal stink-gland of the Blattid
Aphlebia, in that they project inwards, away from the lumen
of the gland. Whether there is any definite relation between
these tubes and the bunches of hairs I have not been able to
ascertain. The tubes seem to come to an abrupt end near
the nucleus, and I cannot trace them beyond the limit of the
granular cells; the hairs are comparatively stout and
cylindrical at their bases, but appear gradually to taper
away at their free extremities. But it seems doubtful
that this should actually be the case; for the chitinous
lining membrane is quite continuous, and it is difficult to
see how the secretion reaches the exterior unless these hairs
serve in some way as a means of communication between the
lumen of the gland and the secreting cells.

In the above description it has been shown that Peri-
planeta orientalis possesses a glandular organ on its
ventral side, lying in the sixth abdominal segment, between
the fifth and sixth abdominal ganglia, and opening to the
exterior between the sixth and seventh sternites. ‘This
opening is median, and from it the gland extends upwards
and forwards as two distinct lobes. It is composed of a
layer of modified epithelial cells lined by a chitinous mem-
brane continuous with the external chitinous skeleton of the
insect. The epithelial cells of the adult are finely granular,

NEW ORGAN IN PERIPLANETA ORIENTALIS. 3881

and in each a denser patch of granules surrounds astout tube
which extends from the region of the nucleus to the inner
edge of the cell. Attached to the lining membrane, some-
times singly, sometimes in groups, are immensely long hairs
with free extremities, directed towards the epithelial layer.
The cells of this outer layer appear to migrate inwards, and
there degenerate, leaving only irregular nuclei very unlike
the large round nuclei of the granular cells. In a young
specimen the epithelial cells are not granular, and there is
no intermediate tissue between them and the chitinous lining.
This suggests that the granules represent the secretion of
the gland, and their absence in the young indicates that the
organ is immature.

This work has been carried on in the Department of Com-
parative Anatomy in the Oxford University Museum, and I
wish to express my warm thanks to Mr. Goodrich for much
kind help throughout the course of my research.

REFERENCES.

Criaus.—‘ Zeit. f. wiss. Zool.,’ Bd. xi, 1861.
Kraask.—‘ Zool. Anz.,’ Bd. xiii, 1890.
Mincuin.—‘ Quart. Journ. Micr. Sci.,’ vol. xxix, 1888.

EXPLANATION OF PLATE 15,

Illustrating Miss Ruth M. Harrison’s paper, “ Preliminary
Account of a New Organ in Periplaneta orientalis.”

List oF REFERENCE LETTERS.

G. The gland. ec. Cerca. epi. Epithelium of body wall. epi.g. Hpithe-
lium of gland. f.d. Fat body. gr.c. Granular cells. 4. Hairs. 7. Lumen
of gland. /.m. Lining membrane. m.g/. Mushroom-shaped gland. 2. Nuclei
of disintegrating cells. 7. Rectum. ss. Style. ¢h.7.¢e. Thickened rim of
chitin on sternite 7 round the external aperture. ¢r.v. Tracheal vessel. ¢z.
Tubes in the granular cells. 5¢h abd. g. and 6th abd. g. Fifth and sixth abdo-

vot. 50, PART 2,—NEW SERIES. 27

382 RUTH M. HARRISON.

minal ganglia. s/,—s¢,. Sternites five toseven. 7¢,, ¢,—t,). Tergites four, five
to ten.

Figs. 3—12 have all been drawn with the aid of a camera Incida.

PLATE 15.

Fig. 1.—Dissection of the six posterior segments of a large male cockroach
6 ec

Fie. 2.—The gland removed with the tracheal vessels associated with it.
x 22.

Fie. 3.—Longitudinal sagittal section through a female cockroach in the
region of the gland. The nerve cord has been cut just below the fifth abdo-
minal ganglion. The external opening between the sixth and seventh sternites
is shown, also the thickened chitinous rim at the anterior end of the seventh
sternite. x 43.

Fig. 4.—Transverse section through a male cockroach at the anterior end
of gland, showing that two lobes project forward. x 70.

Fia. 5.—Transverse section through the same cockroach a little further
back, showing that posteriorly the lobes unite. The thickened chitinous rim
of sternite 7 has been cut near the posterior limit of the crescent. x 70.

Fie. 6.—Portion of the wall of a gland from a male cockroach more highly
magnified. ‘The tubes in the cells of the epithelial layer have been cut in a
good many cases. X 380.

Fic. 7.— Portion of the wall of the gland from another cockroach under a
higher magnification. The tubes are not visible in this preparation, but the
hairs attached to the lining membrane are shown. ‘his section also shows
some of the cells of the outer layer migrating inwards. x 400.

Fic. 8.—A small portion of the gland after maceration in potash so that
only the chitinous hairs remain. The magnification in this figure is so high
that it was impossible to get the whole of a single hair within the field of the
camera. Only the attachments and beginnings of each hair were therefore
traced with a camera, and the free extremities had to be drawn in separately.
Oil imm. oc. 12.

Frias. 9 anp 10.—Longitudinal sections through a young male cockroach
about 1°5 em. long. Fig.9 isa median section showing the external openings ;
Fig. 10 is a section taken to one side of the middle line, showing the extent
of one lobe of the gland. x 105.

Fie. 11.—An enlarged drawing of a portion of the same gland. The com-
paratively enormous breadth of the chitinous lining in the last three figures is
probably artificial, due to a breaking apart in the section-cutting. x 400.

Fria. 12.—A small portion of the wall of the gland highly magnified, showing
the tubes in the epithelial layer, x 960,

With Ten Plates, Royal Ato, is.
CONTRIBUTIONS TO THE KNOWLEDGE OF RHABDOPLEURA
AND AMPHIOXUS.
By EK. 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 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/-.

The issue of Suprprement NumpBers being found inconvenient,
and there being often in the Editor’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/-, 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 twenty-five copies of their
communication gratis.

All expenses of publication and illustration are paid by the
publishers.

As a rule lithographic plates, and. not woodcuts, are used in
illustration.

Drawings for woodcuts should nor be inserted in the MS.,
but sent in a separate envelope to the Hditor.

Contributors to this Journal requiring ewtra copies of their
communications at their own expense can have them by applying
to the Printers,

Messrs. ApLtarp & Son, 224, Bartholomew Close, E.C., on
the following terms :

For every four pages or less—

25 copies : : : 5/-
Ue Pe, : ‘ : : 6/-
TOOK as 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 Mussrs. J. & A. CuHuRcHILL, 7 GREAT MarLporoucs SrreeEt,
Lonpon, W.

THE MARINE BIOLOCGIGAL ASSOCIATION

OF THE

UNITED KINGDOM.
Patron—HIS MAJESTY THE KING.

President—Prof. E. RAY LANKESTER, LL.D., F.R.S.
BLO

‘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 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.

TERMS OF MEMBERSHIP.

ANNUAL MEMBERS . : . £1 1 Oper annum.
Lire MemMBERs . ‘ 2 : . 15 15 0 Composition Fee.
FOUNDERS . ; OOS OUR “fA a

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.

New Series, No. 199 (Vol. 50, Part 3). Price 10s. net.

AUGUST, 1906.
THE

QUARTERLY JOURNAL

OF

MICROSCOPICAL SCIENCE.

EDITED BY

E. RAY LANKESTER, M.A., D.Sc., LL.D., F.B.S.,

HONORARY FELLOW OF EXETER COLLEGE, OXFORD; CORRESPONDENT OF THE INSTITUTE OF FRANCK
AND OF THE IMPERIAL ACADEMY OF SCIENCES OF ST. PETERSBURG, AND OF THE ACADEMY
OF SCIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIENCES
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 LINCEI 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 THE
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OF
THE ROYAL PHYSICAL SOCIETY OF EDIN-
BURGH, AND OF THE
BIOLOGICAL SOCIETY OF PARI8, AND OF THE CALIFORNIA ACADEMY OF SCIENCES OF SAN FRANCISCO;
FOREIGN ASSOCIATE OF THE NATIONAL ACADEMY OF SCIENCES, U.S., AND MEMBER OF
THE AMERICAN PHILOSOPHICAL SOCIETY 5
DIRECTOR OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSEUM; LATE FULLERIAN
PROFESSOR OF PHYSIOLOGY IN THK ROYAL INSTITUTION OF GRRAT BRITAIN;
LATE LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFORD.

WITH TILE CO-OPERATION OF

ADAM SEDGWICK, M.A., F.RS.,

FELLOW AND TUTOR OF TRINITY COLLEGE, CAMBRIDGE s

AND

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

BEYER PROFESSOR OF ZOOLOGY IN THE OWENS COLLEGE, MANCHESTER,

WITH LITHOGRAPHIC PLATES AND TEXT-FIGURES.

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

1906.

Adlard and Son, Impr., ] (London and Dorking,

CONTENTS OF No. 199.—New Series.

MEMOIRS:

On the Development of Nebalia. By Marcaret Rosinson, Zoologi-
cal Research Laboratory, University College, London. (With
Plates 16—21) . : ;

On the Early Stages in the Development of Flustrella hispida
(Fabricius), and on the Existence of a ‘‘ Yolk Nucleus” in the Egg
of this Form. By R. M. Pace (née Clark), Late Scholar of Girton
College, Cambridge. (With Plates 22—25) . = .

Researches on the Origin and Development of the Epiblastic Trabe-
cule and the Pial Sheath of the Optic Nerve of the Frog, with
illustrations of Variations met with in other Vertebrates, and some
Observations on the Lymphatics of the Optic Nerve. By J. T.
Grapon, M.A., St. John’s ee Oxford. (With Plates 26 and
27) : ‘ :

Piroplasma muris, Fant., from the Blood of the White Rat, with
Remarks on the Genus Piroplasma. By H. B. Fantuam,
B.Se.Lond., A.R.C.S., Derby Research Scholar, University College,
London ; a Demonstrator in Biology, St. tee s Hospital Medical
School. (With Plate 28). ;

PAGE

383

435

479

493

ON THE DEVELOPMENT OF NEBALIA. 383

On the Development of Nebalia.

By

Margaret Robinson,
Zoological Research Laboratory, University College, London.

With Plates 16—21.

INTRODUCTION.

Ir was at the suggestion of Professor W. F. R. Weldon
that I began this investigation, and he kindly gave me the
material which he had taken from the brood pouches of speci-
mens obtained by dredging at Naples. Unfortunately it had
been preserved some time when it came to me, and part of it
(including some of the later stages) had been rendered of
little use by the action of tannin from the corks of the bottles.
I have, however, been able to work at six stages which are
fairly consecutive, and seem important.

I may mention that I have tried to supplement my material
by means of visits to places where Nebalia is abundant,
namely, Jersey and Roscoff. At the latter place I received
much kindness and assistance from Professeur Yves Delage and
Monsieur A. Robert. Numbers of adult Nebalia were obtained
from decaying crabs and lobsters, which had previously been
placed under stones on the shore at low tide. None of these
had eggs in the brood pouch, and though I spent a good deal
of time and trouble in trying to make the animals breed in

VOL, 50, PART 3,—NEW SERIES, 28

384, MARGARET ROBINSON.

dishes, I could never obtain a fertilisedegg. I had previously
tried to do this in Jersey, where Mr. James Hornell kindly
provided me with specimens. This absence of females with
eggs in the brood pouch from the shore leads me to believe
that the animals go into deeper water to breed. All the
people who have worked at the development of Nebalia—
Metschnikoff, Claus, Butschinsky—have obtained their
material from tideless seas.

From many practical hints and much other help I am in-
debted to the kindness of Professor E. A. Minchin. I wish
here also to thank my many friends and fellow-workers for
their assistance, particularly Dr. E. J. Allen, who has read
through this manuscript and made many valuable suggestions
and criticisms.

HISTORICAL.

The earliest notice of Nebalia is that by Otho Fabricius
(1780) in his Fauna Greenlandica. This is a not very
exact description of the external features, and is accompanied
by a little figure. The next account is that of Herbst (1796).
It is a translation of the description given by Fabricius with
a copy of his figure. Both of these authors call the animal
Cancer bipes.

Montagu (1815) found Nebalia on the coast of Devon,
and described it under the name of Monoculus rostratus.
The first person to call it Nebalia was Leach, who described
it in the ‘ Zoologist’s Miscellany ’ (1813), and noted it as a
very distinct genus belonging to the Crustacea Malacostraca.

Latreille, in his text-book (1831), places Nebalia with
Cuma in an order which leads from the sessile-eyed crusta-
ceans and those with stalked eyes, especially from Mysis, to
the Cyclops. He calls the order Diclapoda ‘Ce nouvel
ordre comprendra les genres Nébalie, Pontie, Condylure, et
Cume, qui se lient d’une part avec les Mysis et de l’autre
avec les Cyclopes.” He lays stress on the fact that Nebalia
carries its eggs in a brood pouch, as does Mysis.

ON THE DEVELOPMENT OF NEBALIA. 385

Milne-Edwards (1828 and 1835) who found Nebalia on
the coast of Brittany, and described it under the name of
Nebalia Geoffroyi, after Geoffroi St. Hilaire, gave as his
reason for not placing it among the Malacostraca the fact
that the thoracic legs with their lamellate appendages do
resemble those of Branchipus, and these gills are in no way
like the gills of Decapods. In 1840 he writes of the Nebalide
“ Hlles semblent 4 plusieurs égards établir le passage entre
les Mysis et les Apus,” but places them in the family “Les
Apusiens”’ among the Phyllopods.

Kroyer (1847) (of whom I know only through Claus and
Metschnikoff) stated that Nebalia could not be a Phyllopod
because it carried its embryos in the brood pouch till uney
were practically adult.

Metschnikoff, in 1868, published a long account of the
development of Nebalia. As this is an important paper of
which unfortunately there is no published translation from
the Russian, | give here a somewhat lengthy abstract of it.
I must first thank Miss Zelda Kahan for translating the paper
for me.

Metschnikoft begins his account of the development by
describing the formation of the blastoderm. In this part I
will give his words as nearly as possible: “ Before total seg-
mentation there is another division in the formative yolk, and
there appears a polar vesicle. Both of these appear on the
lower portion of the egg. Here there is formed a small
accumulation of colourless protoplasm containing a large
number of granules. ‘This protoplasm, which is nothing more
than generative yolk, separates itself from the egg envelope,
and thus there comes into existence asmall space between the
shell and the yolk. In this space there appears a small
globule of protoplasm which plays no part in development
and soon disappears.

“The further development depends on the increase in quan-
tity of the generative yolk. When it has increased so as to take
up about one fifth of the volume of the entire egg it divides
longitudinally into two parts oval in shape. These divide

386 MARGARET ROBINSON.

meridionally, resulting in four cells. These again divide
longitudinally. At first the cells only cover the lower pole
of the egg forming two rows of cylindrical cells. Later they
spread over the whole egg.”

Then follows a description of seven stages in development.
In the last of these only does the embryo become a true larva
—for here alone it is free swimming; the other stages being
passed through in the brood pouch of the mother.

First Stage.—lIn the first of these stages he notes a
thickening on the ventral surface and the first appearance of
the papilla.

Second Stage.—He describes an embryo with nauplius
appendages, and an abdominal papilla which is bent over the
ventral surface. In this stage he also notes the appearance
of the endoderm and the way in which it begins to grow round
the yolk.

Third Stage.—He notes two pairs of maxille and two
thoracic legs, as well as the nauplius appendages. Here, too,
he observes mouth and stomodezum, as well as anus and
proctodeeum.

Fourth Stage.—He describes the change in position of the
' appendages (they are now directed backwards instead of out-
wards), and the appearance of the third thoracic hmb. He
also notes the possession of dorsally directed palps by the
maxille and thoracic appendages. In this stage, too, there
is an increase in the size of the optic lobes and brain, and
here he first describes the labrum. In each of these two last
stages he describes the further circumcrescence of the yolk
by the endoderm. He remarks, too, on the different appear-
ances of the yolk in different places; that which is still not
enclosed by endoderm being dense and granular, while the
enclosed portion seems to be liquid and contains but few
granules.

Fifth Stage.—The embryo now bursts the vitelline mem-
brane, but is still enveloped by a cuticle, and the abdomen
springs back so that now instead of lying on the ventral sur-
face it is slightly curved in a dorsal direction. The cuticle

ON THE DEVELOPMENT OF NEBALIA. 387

invests all the first five appendages (nauplius appendages and
maxillze) closely following all their curves and outgrowths.
The rest of the appendages are covered by an unbroken skin
which forms a general sac over them and the back. This
skin separates itself gradually from the parts it covers, and
is finally thrown off. In this stage the shell first appears as
a dorsal crease. Some embryos, which closely resemble that
last described, show the head flap, further lateral developments
of the shell, and pigment in the eyes; but in these, as in thie
last described stage, the front part of the yolk is still unen-
closed by endoderm.

Sixth Stage.—lIn this stage he notes:

(a) The shell valves covering the body of the embryo as
far back as the fifth thoracic appendage.

(b) The head flap with its front end more rounded than in
the adult.

(c) The body now segmented and covered with chitin, the
latter forming rows of teeth on the two segments before the
last, whilst in the adult these teeth surround all the six seg-
ments before the last in the form of belts.

(d) Segmentation of several of the appendages.

(e) All the yolk now enclosed by endoderm. Liver out-
growths which diverge in the shape of blind conical sacs lying
at the sides of the intestine and containing grey matter (de-
rived from the yolk) and some yolk. Also two shorter out-
erowths directed forwards.

Seventh and Last Stage, which differs only from the
adult in having three instead of four abdominal legs, in having
fewer segments in the antenne, and in having no vertical
split in the middle lamine of the gills of the thoracic legs.

In one place he regrets that he saw so httle in the embryos.
To us the wonder must be that he saw so much without the
help of all our modern appliances and contrivances. He
refers to the work of a previous observer, Kréyer, who found
twenty-four appendages in the adult. Kroéyer was of the
opinion that Nebalia could not be a Phyllopod, as it did
not leave the brood pouch of the mother until practically an

388 MARGARET ROBINSON.

adult. He suggested further investigations to find out if it
resembled the Decapods.

Metschnikoff himself is strongly in favour of the inclusion
of Nebalia among the Malacostraca. He says that its most
phyllopod-like feature, viz. the thoracic legs, differ really
from those of the Phyllopods in number and development.
Further, that the mouth-appendages resemble those of the
decapods, that the digestive organs are not like those of the
Phyllopods, and that the openings of the oviducts are in
the wrong place for a Phyllopod. He concludes by saying
that it has only a general similarity to the Branchiopoda,

He then points out that Nebalia greatly resembles the
Schizopods in the number and arrangement of the append-
ages, many of these having five to eight thoracic limbs with
gills, though in their case the change of leg into gill does
not go so far as it does in Nebalia. Then he remarks on
the great likeness between the gill of Nebalia and the
swimming leg of Euphausia. In conclusion Metschnikoff
suggests that Nebalia be removed from the Phyllopods,
where Milne-Edwards had placed it, to the Decapods, and
that a special group be made for it side by side with the
Schizopods. (One must suppose that by Decapods he means
Malacostraca. )

Claus (1872) wrote concerning the anatomy and systematic
position of Nebalia. He could not agree with Metschnikoff
in placing it among the Decapods, as he thought it had no
true Zowa stage. However, he expressed the opinion that
Nebalia must be very nearly allied to the Malacostraca.

v. Willemoes-Suhm (1875), in describing a new species of
Nebalia found by the Challenger Expedition, places the
Nebaliadee among the Schizopods. In 1876 Claus, in his
‘Crustacean System,’ gave an account, with some figures, of
Nebalia. He there drew attention to the likeness between
Nebalia and the Myside as regards heart, digestive canal,
nervous system, antennary gland, and genital organs.

One might infer, from the words of Milne-EKdwards and
Latreille, that they believed that the Malacostraca were

ON THE DEVELOPMENT OF NEBALIA. 389

descended from the Phyllopoda through Nebalia. Boas
(1883), while considering Nebalia as transitional, is careful
to point out that, in his opinion, this is not a case of direct
descent. “Sie steht nicht auf dem geraden Weg von den
Phyllopoden zu den Malacostraken, sondern etwas seitlich.”
He considers Thysanopus (an Huphausid) to be the most
primitive Malacostracan, and the nearest to Nebalia, but it,
like Nebalia, does not lie on the direct route from Phyllo-
poda to Malacostraca. His views are founded mainly on
differences and likenesses in appendages and other external
features. He expresses the result as regards Nebalia thus:
Malacostraca.

__——Nebalia.
Phyllopoda.

Sars (1885), on the other hand, in his Report on the
Schizopoda of the Challenger Expedition (1885), expressly
states that he canuot agree with Boas as to placing Nebalia
either among or near the Schizopoda. He then regarded it
as a Phyllopod.

In 1885 Claus again wrote on Nebalia. After alluding
to Metschnikoff’s paper he proceeds to give his own reasons
for the assertion that Nebalia is “no Phyllopod.” As to
external features he says that there are only the following in
which Nebalia resembles the Phyllopods—the shell, the
thoracic legs, and the tail.

The form of the shell, he justly remarks, is one which is by
no means confined to the Phyllopods. He thinks that, as in
the case of the Huphausia larva, it is the original mala-
costracan shell which has been retained. Further, he says,
it shows that the carapace of the Malacostraca and the shell
of the Entomostraca had the same starting point. The
thoracic limbs of Nebalia are, he continues, intermediate in
character between those of the Phyllopoda and those of the
Schizopoda. In internal structure Nebalia differs still more
from the Phyllopoda, and is especially like the Myside.
With regard to the abnormal number of abdominal segments,

390 MARGARET ROBINSON.

he considers that, as the two hindmost have not special
ganglia, they are not true segments, but, rather, joints of a
segment; and here he alludes to the jointing of the sixth
abdominal segment in Gnathophausia, which he says is
not accompanied by any corresponding division of the
ganglion.

He regards the extra joint in Gnathophausia and the
two extra joints in Nebalia as being representatives of the
telson. In this paper he gives a schematic tree of the
Crustacea. In the tree the Leptostraca, the Protoschizopoda
and the Stomatopoda are made to come off together from the
Protomalacostraca.

In 1887 Sars, in his Report on the Phyllocaride of the
Challenger, following Dr. Packard (in ‘Phyllopod Crustacea
of North America’), is inclined to derive the Nebaliade from
copepod-like ancestors. He says that the Podophthalmia are
in no way related to them, but that the Branchiopoda proba-
bly came from the same stem, and have become altered to
suit conditions of life, whereas the Nebaliadz have preserved
many primitive features.

In 1889 Claus published his last paper on Nebalia. In
this he gives a very full and detailed account of the anatomy,
and puts forward at some length his views as to the sys-
tematic position of the animal. He mentions several fresh
pieces of evidence in favour of its being a Malacostracan.
The chief of these are:

(a) That the openings of the genital ducts in both sexes
are in the Malacostracan position. He had found the male
ducts previously.

(b) The great complexity of the brain.

(c) The structure of the eyes and the optic ganglion, which
resemble those in Mysis.

(d) The rudimentary shell gland and the eight pairs of
ectodermal excretory glands on the thoracic legs. These
function as well as the antennary gland.

(ec) The two last abdominal segments, as representing the
Malacostracan telson. The last but one of these he says has

ON THE DEVELOPMENT OF NEBALIA. 391

a ganglion in the late larval stage which disappears in the
adult.

Claus places the Leptostraca as one of the three orders into
which the Malacostraca are divided by him, namely, Lepto-
straca, Arthrostraca, and Thoracostraca.

In 1892 Grobben, in a paper on the classification of the
Crustacea, pointed out the several points of resemblance
between Branchipus and Nebalia, but finally came to the
conclusion that it would be harder to join Nebalia to the
Phyllopoda than to the Malacostraca. He divided the Mala-
costraca into two main divisions, the Leptostraca and the
Eumalacostraca, the latter division including the three orders
Stomatopoda, Thoracostraca, and Arthrostraca. His paper
contains a tree of the Crustacea, similar to that given by
Claus (1885), in which the Leptostraca are made to come off
from the Crustacean stem with the Protoschizopoda.

In 1893, Hansen suggested a new classification of the
Malacostraca. Like Grobben he divided them into two
groups, Leptostraca and LHumalacostraca, which latter
consists of three orders (unnamed). The first of these
orders contains the Myside, the Cumacea, Isopoda and
Amphipoda; the second, the Euphausids and the Decapods,
while the third exists for the Stomatopoda only. He con-
siders that the Leptostraca are decidedly the most primitive
Malacostraca, and that of the Humalacostraca the most nearly
related to them are the Myside. He found in the shaft of
the second antenna of Nebalia five joints of which the last
showed a tendency to consist of two pieces. This makes the
shaft resemble that of the second antenna of Mysis. Other
points of likeness which he mentions are (1) the develop-
ment of the larve; (2) the form of the heart; (8) the fact
that young embryos of Mysis have at the hindmost end of
the body two small hard processes fairly well chitinised
which must be homologous with the furce in Nebalia; (4)
the presence of conical outgrowths at the openings of the
male ducts.

In 1897, Butschinsky published in the ‘Zool. Anzeiger’ a

392 MARGARET ROBINSON.

short account of the formation of the blastoderm, and 1900,
a longer paper on the development in the same journal. His
account of the formation of the blastoderm differs slightly
from Metschnikoff’s. He says that the cleavage which he
observed was really intermediate between a discoidal cleavage
and a superficial one. In his second paper he again alludes
to this difference between his account of the formation of the
blastoderm and Metschnikoff’s, and gives a short summary
of the development.

Meruops.

The eggs were taken from the pouches and fixed in a hot
concentrated solution of corrosive sublimate to which a little
acetic acid was added. They were then washed and taken
very gradually through alcohols of increasing strengths up to
80 per cent. For their most excellent fixation and preserva-
tion I am very grateful to Professor Weldon.

The shells of the early stages were removed by teasing with
very fine sewing needles.

As yolk preserved in sublimate is extremely brittle, in
order to cut it without its breaking I had recourse to the two
following methods :—(1) Embedding in celloidin. (2) Paint-
ing each section with a mixture composed of equal parts of
gum mastic and celloidin.

The orientation was done in the first case by cutting the
celloidin into the required shape, and in the second by fasten-
ing the embryo in position on a piece of lardaceous liver
before embedding in paraffin. The sections were cut with
Jung’s microtome, and are 4, in thickness.

The embryos from which the surface views are taken were
stained with Delafield’s hematoxylin; the sections with
Kleinenberg’s hematoxylin, and orange.

The sections were all drawn under Zeiss’s objective D D.
and ocular No. 3 with Abbé’s camera lucida, the details being
filled in with the help of a ;4,” immersion objective.

ON THE DEVELOPMENT OF NEBALIA. 393

THE DEVELOPMENT OF THE Embryo.
The Formation of the Blastoderm.

Metschnikoff describes a discoidal segmentation; but a
series of sections made through a stage which is earlier than
that shown in Plate 16, fig. a, leads one to suppose that we
have in Nebalia not an instance of discoidal segmentation
proper, but rather a case of Korschelt and Heider’s (1893),
type IIIb, in which the increase of the blastoderm is aided
by the accession of new elements from the inside of the egg.
It seems that the blastoderm, certainly at first, increases both
by division of the cells already on the ventral surface and also
by the reception of additional cells which come from within
the ege itself (Plate 16, fig. 1). According to Butschinsky
(1897) the first two divisions take place in the protoplasm
while it is still lying in the centre of the egg within the yolk.
The four cells resulting from these divisions travel to the
ventral pole of the egg and there divide, thus forming a cap
of eight cells. This cap, by subsequent divisions of its cells,
gradually surrounds the yolk. He also describes a few cells
which are smaller than the other cells of the blastoderm.
These smaller cells have escaped my observation.

Stage A. (Plate 16, fig. a.)

Figure A shows an egg in which the yolk is still incom-
pletely covered by the blastoderm. Sections through 1t
contain none of the large cells within the yolk shown in
fig. 1. Therefore it may be supposed that only in a very
early stage is the blastoderm increased by the reception of
additional cells coming from within the yolk.

The blastoderm cells are more or less oval in shape and
their protoplasm is granular. The yolk is broken up into

394 MARGARET ROBINSON.

small angular lumps some of which stain more deeply than
the others. All through the development, until the yolk is
completely liquefied, this irregular staining can be noticed.
Even at this very early stage some cells, which, as will subse-
quently appear, may be regarded as vitellophags, are budded
off from the blastoderm into the yolk (fig. 2, vp.).

Stage B.

External View.—The external appearance of the embryo
closely resembles that of the next stage (Stage B’). On the
ventral surface the blastoderm shows three areas of thicken-
ing which are arranged so as to form the apex and two
adjoining sides of a triangle. This triangle, however, is much
shorter than that outlined by the ventral thickenings in
Stage B’. It is,in fact, an area which barely covers the
posterior two thirds of the ventral surface of the blastoderm.
It is outlined by two thickened strands which converge, with
a thickened patch uniting their convergent ends.

These lateral thickened strands consist of large, rounded,
granular cells which are thicker and rounder than the cells
of the rest of the blastoderm. In some places these cells
appear to give rise to others by a kind of tangential division
(figs. 3 and 4). That some of the new cells formed in this
way are vitellophags there can be little doubt, others are pro-
bably some of the first cells of the ordinary mesoderm.

The median posterior thickening (the thickened apex of
the triangle) has in its centre a very narrow groove running
in a longitudinal direction. This groove is short, extending
through eight sections (each only 4 in thickness).

Of these sections I have drawn two, one (fig. 6) through
the middle region of the groove, and the other (fig. 5) near
its anterior limit.

Behind the region shown in fig. 6 the groove widens out
alittle. A section taken immediately behind the groove is
shown in fig. 7.

The section through the middle region shows an inpushing

ON THE DEVELOPMENT OF NEBALIA. 395

of five cells of the blastoderm, and on each side of this in-
pushing there is a cell (fig. 6, mes.).

There can be little doubt that the entrance to the groove
represents a blastopore, and that there is here an invagination
of blastoderm cells to form the endoderm. The cells on
either side of the invagination I take to be mesoderm, and to
have been budded off from the blastoderm in situ.

In the section close to the anterior limit of the groove the
lumen is narrower. This makes one think the closure begins
at the front end of the groove. Just dorsal to the narrow
depression there are a few large rounded cells (fig. 5). These
I take to be endoderm cells which have been invaginated or
have arisen by proliferation from invaginated cells.

The section behind the blastopore also shows two layers of
cells in the ventral region (fig. 7). I find it hard to say
whether the inner layer consists entirely of mesoderm or not.
The large almost central cell may be endodermal, the others
are certainly mesoderm.

All the cells in this and in the next stage contain a very
granular protoplasm.

Stage B’ (fig. B’.).

External View.—The embryo lying on what will be the
ventral surface of the yolk shows three distinct regions of
thickening, namely, the two optic thickenings at the anterior
end, and the caudal thickening at the posterior end (c.t.).
These three thickenings are connected by strands of cells in
which strands there are again thickenings foreshadowing the
antenne.

One can also see in this external view an indication of the
caudal groove between the caudal thickening and the blasto-
derm in front of it. Many dividing nuclei can be seen in the
central region of the ventral surface of the blastoderm.

Internal Structure.—Transverse sections through the
optic thickenings show them to consist of cylindrical cells
(fig. 8), and of only one layer of these, though here and there

396 MARGARET ROBINSON.

the beginnings of a second layer can be noticed. As one
follows the sections farther back one notices that the cells
gradually lose their cylindrical shape becoming flattened
(figs. 9 and 10). Sections through the strands show that all
along them mesoderm cells are apparently being budded off
from the blastoderm (figs. 9 and 10, mes.). In the anterior
region these cells are few and far between, but the farther
back we trace them the more numerous do they become, so
that on reaching the caudal thickening we find the strands to
consist of two definite layers of cells—ectoderm and mesoderm
(fig. 11).

Besides the above-mentioned mesoderm cells there are
budded off from the blastoderm more cells of the type men-
tioned as occurring in the earlier stages (figs. 9 and
10). These cells have large much-vacuolated nuclei (about
twice as large as the nuclei of the other cells), and a
small amount of cytoplasm which is spread out in processes
resembling pseudopodia. In fact each cell has the appear-
ance of an amoeba with an immense nucleus. They are to be
found throughout the whole length of the embryo, but are
not numerous. In the embryo from which the figured sec-
tions were taken there were about ten of these cells. Though
they come to lie in the yolk which is ultimately surrounded
by endoderm, these cells are distinctly mesodermal in origin,
i.e. they are budded off from the blastoderm in the same
manner, and in the same regions as the rest of the mesoderm.
It has been suggested by Kowalewsky (1886), and Nusbaum
(1886) that these cells help in some way to soften the yolk,
and so render it easy of absorption by the protoplasm. That
the yolk is of a different consistency in different stages can
be seen even in preserved specimens, and in the later stages
the vitellophags are first diminished in numbers and then
gradually disappear. These facts taken together certainly
support the above-mentioned view as to the function of these
cells.

Sections through the posterior end of the embryo pass
through the hind ends of the lateral thickenings as well as

ON THE DEVELOPMENT OF NEBALIA. 397

the caudal thickening. They show, in the region occupied
by the groove in Stage B, a band of large rounded cells, the
endoderm, resulting from the invagination. On either side
of this band in its anterior region the body wall consists of
two layers of cells, ectoderm and mesoderm, since there are
here the posterior convergent ends of the thickened lateral
strands (fig. 11).

Unfortunately the sections of the series here drawn are a

Diacram ]1.—Transverse section through the caudal thickening
in Stage B’, showing the remains of the groove “roofed in”
by the closure of the blastopore. He. Hctoderm. Hx. Endoderm.
Mes. Mesoderm.

little oblique, or they would show more distinctly than they
do the remains of the groove which is now roofed in, as it
were, by the closing of the blastopore, as shown in the
above diagram.

It can be seen from figs. 12 and 13 that the endoderm cells
have already begun to proliferate ; and from figs. 13 and 14
it would appear that their numbers were also increased by
cells being pushed in from the ectoderm. Further back it
becomes rather difficult to distinguish between the endoderm
and the mesoderm ; but there is a difference in the shape of
the cells, those of the mesoderm being oval or almost almond-
shaped, whilst the endoderm cells are much more nearly
spherical.

All the protoplasm in this stage is very granular, and the

398 MARGARET ROBINSON.

cells in the caudal thickening are many of them much vacuo-
lated.

Putting together the facts concerning these two stages B
and B’ one is forced to the conclusion that the cells which are
invaginated at the blastopore give rise to the endoderm. I
cannot help thinking that they give rise to that only, and
that the mesoderm arises quite independently of this invagi-
nation. I think that fig. 6 points strongly to this. That
the vitellophags are budded off from the blastoderm figs. 3
and 4 show, but unfortunately I have not been able to find
in any of the embryos at this stage a blastoderm cell in
process of division to form mesoderm, except in fig. 3, div.,
and the results of these two divisions may be vitellophags
after all.

Bergh (1893) in his account of the formation of the germ
layers in Mysis describes two stages which almost corre-
spond as to age with Stage B described above. They seem
both to be a little younger than it, and also a third stage
which is perhaps a little younger than Stage B’. He is care-
ful to state that in none of these does he find any vitello-
phags. In the proliferation in the region of the blastopore
he finds two kinds of cells, large and small, the large being
mesoderm and the small endoderm.

In a stage a little older than the oldest of these three
Bergh finds some vitellophags for the first time; and in
summing up the results of this part of his work he states
that the cell mass resulting from the invagination at the
blastopore differentiates itself further into three regions of
germination (Anlagen).

(1) The vitellophags,

(2) The entoderm disc,

(3) The eight original cells of the mesoderm.

The likeness between these stages in Mysis and those in
Nebalia described above is very great; but there are
differences.

(1) I find vitellophags in both Stage B and Stage B’, and
indeed in an earlier stage than B, i.e. in Stage A, where the

ON THE DEVELOPMENT OF NEBALIA. 399

blastoderm has not as yet completely surrounded the yolk.
Therefore they must arise quite independently of the invagi-
nation.

(2) From fig. 6 it would seem that the mesoderm in this
region arose also from the blastoderm independently of the
invagination. This mesoderm Bergh calls metanauplial. The
mesoderm of the nauplius appendages he calls nauplial, and
says it is probably budded off from the ectoderm in situ.
This obliges him to admit of two origins for the mesoderm.

(a) The invagination at the blastopore for the metanauphial

mesoderm.

(b) Budding off from the blastoderm in situ for the

nauplial mesoderm.

Now in Nebalia it seems to me that there are two possi-
bilities, either

(1) All the mesoderm cells are budded off in situ from
the blastoderm, i.e. the nauplial mesoderm in the places
where the appendages are about to be folded off, and the
metanauplial mesoderm from the blastoderm in the region of
the blastopore at about the same time as the invagination
takes place but quite independently of it—or

(2) All the mesoderm is budded off from the blastoderm in
the region of the blastopore at about the same time as the
invagination takes place (but quite independently of the
invagination), and the cells which will form the nauplial
mesoderm travel forwards.

Hither of these possibilities seems a little more probable
than the possibility that in the same animal part of the
mesoderm should arise as the result of an invagination and
part by division in situ.

Butschinsky (1900) is of opinion that both endoderm and
mesoderm originate from a mass of cells to which in this stage
the blastoderm gives rise at the caudal thickening ; and this
is the only point of any importance in which our accounts of
the development differ.

Putting aside the difference with regard to the origin of
the metanauplial mesoderm, there is a striking similarity

VOL. 90, PART 3.—NEW SERIES. 29

400 MARGARET ROBINSON.

between the development of Mysis and that of Nebalia in
these early stages which show the origin of the endoderm
from an invagination which results in the formation of a

solid band of cells.

Stage C (fig.c). Embryo with Nauplius
Appendages.

External view.—In an external view of the ventral sur-
face of the embryo at this stage one can distinguish :

(1) The eye thickenings.

(2) The first and second antenne.

(3) The mandibles.
(4) The abdominal papilla which is bent forward.

(5) The mouth which is here a crescent-shaped depression
lying between the first and second antenne.

Internal structure.—As is usual among the Crustacea,
embryos of approximately the same external appearance vary
not only in size but also in internal development. I have
therefore made drawings from three series of transverse
sections taken from three embryos, each of which has the
nauplius appendages. It is chiefly in the ectoderm of this
stage that differences occur.

All three series show the optic thickenings in the anterior
region, though in the very front these thickenings are not
much pronounced, and indeed show little or no advance in
size on those of the preceding stage (fig. 16). All three too
show at a little distance behind the most anterior part of the
optic thickenings on either side an invagination of ectoderm
cells (op. in. fig. 17). This very closely resembles the optic
invagination described by Reichenbach (1886) in the cray-
fish embryo with nauplius appendages.

In each of the three series, almost immediately behind the
optic invagination on either side, there are a few ectoderm
cells which are much larger than the rest, almost double the
size, with large nuclei (fig. 18 g.1). These I take to be the

ON THE DEVELOPMENT OF NEBALIA. 401

first rudiment of the optic ganglion. They are continuous
with similar cells lying near the base of the first antenna,
and forming the rudiment of the antennulary ganglion (fig.
20, g. 11). At the point of junction of these two rudiments
there is a distinct groove or fold in the cells composing them
(seen in transverse section, fig. 19, and in longitudinal
section, fig. 23, between g.1 and g.t1). This fold makes it
appear as though the optic ganglion cells were being nipped
off anteriorly and laterally from those of the antennulary
ganglion, these last lying nearer the middle line. Now, it
might be argued that there is here a case of the optic
ganglion being budded off from the antennulary ganglion;
but one must bear in mind that these rudimentary ganglion
cells are at present merely slightly differentiated ectoderm
cells which still form part of the outer wall of the body of
the embryo, and also that all the evidence we have seems to
show that the differentiation has taken place simultaneously
in front of, and just behind the groove, and in the cells of
the groove itself. This groove or fold marks the division
between the optic region and that of the first antenna. It is
in fact almost the earliest trace of segmentation.

There are similar groups of specialised ectoderm cells near
the bases of the second pair of antennez and of the mandibles
fies. 21 and 22, g. Il, g. IV).

There are then in this embryo with nauplius appendages
four pairs of rudimentary ganglia—(1l) optic, (2) antennu-
lary, (5) antennary, (4) mandibular (fig. 23).

Since the optic ganglia appear simultaneously with the
other pairs of ganglia but independently of them; since too
experiments made by Herbst (1896) and by others before
him have shown that eye-stalks in Decapods may be replaced
by antenna-like appendages, the optic region possesses two
of the main essentials of a segment, and there seems to be
little reason for not considering it the first segment of the
body.

It was on account of the replacement of the eye-stalk of
Palinurus by an antenniform palp that Milne-EHdwards,

402 MARGARET ROBINSON.

who was followed by Huxley, considered the eyes as the first
pair of appendages. (Lankester, 1904.)

Reichenbach (1886) and Nusbaum (1887) found first traces
of a nervous system in the crayfish and in Mysis similar to
those above described, and they too looked upon the optic
tract as the first body segment.

If this be so, then the eye region forms the first segment of
the body not only in all the Malacostraca, but also in all the
Crustacea, and in the rest of the Arthropoda.

Evidence against this opinion has been given by Claus in
his papers on the development of Branchipus (1873 and
1886). Since he referred to the facts there recorded at
great length in his last paper on Nebalia (1889) one feels
bound to mention them, but it seems that his observations
were not made on sufficiently early stages.

My observations on the development of the nervous aystoal
in Nebalia are merely additional evidence in favour of the
opinion that in the Crustacea the optic region forms the first
segment of the body. ‘This is the view taken by Heymons
(1901), and the one to which, lately, Prof. Lankester has
given his support. (Lankester, 1904.)

In this paper I am using the terminology of Viallanes and
Heymons (1892) for the different parts of the brain. Proto-
cerebrum for that in the first or optic segment, deutero-
cerebrum for that belonging to the second segment, and
tritocerebrum for the pair of antennary ganglia in the third
segment.

To continue the description of the nervous system at this
stage. The cells of the four pairs of rudimentary ganglia,
although they form part of the general ectoderm of the body-
wall, differ from the other ectoderm cells in size, and also in
the size and colour of their nuclei. They are larger than the
cells of the rest of the ectoderm, and have larger and paler
nuclei, which stain very slightly.

The ganglia of each pair are separated from each other by
a median narrow band (one cell deep) of the cells of the
ordinary ectoderm. The ganglia of the first pair are in

ON THE DEVELOPMENT OF NEBALIA. 403

continuity with those of the second on either side, and these
two pairs of ganglia lie in front of the mouth, while those of
the antennary and mandibular segments lie behind it.

In one specimen a section through the first antenna shows
the ganglion near its base to consist of two kinds of cells,
as it does in Stage D. This is only worthy of notice as an
instance of a difference in internal development between
embryos which have the same external features.

Other Ectodermal Structures.—I find in this stage
the rudiment of a labrum (fig. 24), and also a well-marked
stomodzum, which runs a little forward from the mouth (fig.
26), but, though I have looked through several series of hoth
transverse and longitudinal sections, I can find here no
beginnings of either anus or proctodzum.

Endoderm.—tThe solid band of cells resulting from the
invagination at the blastopore now spreads out laterally, 1. e.
its cells multiply laterally, so as to gradually enclose the
yolk. The enclosure seems to take place in this way. Start-
ing from the anterior end of the caudal thickening, just
behind the depression, between it and the thorax, the cells
grow backwards and dorsalwards, so that the first part of the
yolk to be enclosed is that lying in the papilla (fig. 26).
When this is done, i.e. when the yolk within the papilla is
enclosed, the cells seem to increase more rapidly on the
dorsal than on the ventral surface. A transverse section in
front of the papilla through the farthest advanced of the
embryos with nauplius appendages shows endoderm lying on
the dorsal surface of the yolk, but none on the ventral
surface. Similarly, a longitudinal section through an embryo
belonging to the next stage shows endoderm stretching much
farther forward on the dorsal side than on the ventral (see
Diagram 2, p. 408). The endoderm cells in this stage begin
to assume a columnar shape.

Mesoderm.—The mesoderm can be traced lying between
the ectoderm and the yolk in chains of three or four cells on
the ventral side, from the abdominal papilla to the optic
invaginations. It is most abundant in the region just in

404. MARGARET ROBINSON.

front of the papilla. The mesoderm cells can be seen being
carried along with the ectoderm as it is being folded off to
form the nauplius appendages (figs. 20—22).

Vitellophags.—tThese cells are now more numerous than
they were in the preceding stages. Wherever one of them
is seen the yolk round it has a granulated appearance. This
granulation may be a step onwards towards liquefaction. It
seems that the vitellophags, while transforming the yolk,
undergo disintegration themselves. They certainly appear
to lose their own protoplasm, and their huge swollen nuclei
look as though they were ready to disintegrate (fig. 26, v.p.).
Fig. 26, v.p., shows a number of vitellophags, some with a
little protoplasm, and others consisting of a nucleus only.
This disintegration certainly militates against any idea that
the vitellophags become blood corpuscles or take any part
in the development of the embryo other than that of render-
ing the yolk more easy of absorption by the protoplasm.

Stage D (fig. p).

External Features.—The chief advances noticeable in
an external view are the increase in the number of the
appendages, and the growth farther forward of the abdominal
papilla.

The embryo now has, in addition to the nauplius appendages,
two pairs of maxilla and rudiments of the first three pairs of
the thoracic appendages. The mouth has moved farther back,
so that now it is in the segment which bears the second pair
of antennee. Behind the mouth there is a distinct median
ventral depression. In optical section the endoderm sur-
rounding the hinder two-thirds of the yolk can be seen dis-
tinctly. This stage, Claus notwithstanding, bears a likeness
to a Zova, and might certainly be called the stage with Zoaa
appendages.

Internal Structure. EHctoderm; the Nervous
System.—There is here a great advance in the complexity
of the nervous system. In each segment there is a well-

ON THE DEVELOPMENT OF NEBALIA. 405

marked pair of ganglionic swellings, and the first three of
these pairs are now (since the second antenna and its ganglion
have moved forward) joined to help in forming the syncere-
brum. Of these three pairs of swellings those in the optic
and antennulary segments are much larger than those in the
segment bearing the second antenne. Hach ganglion now
consists of two kinds of cells which may be distinguished by
their nuclei.

(a) Cells with large clear nuclei, each nucleus being sur-
rounded by an appreciable quantity of protoplasm, though
cell outlines can now no longer be made out. These cells are
seen to divide, hence it may be reasonably inferred that they
give rise to—

(b) Cells with small darkly-staining nuclei. These lie for
the most part on the inner side of the large cells. The small
nuclei are closely packed together, and the protoplasm sur-
rounding each of them is very small in amount, almost in-
appreciable. Some of these small nuclei are to be seen in
each of the three pairs of appendages behind the eyes (first
and second antenne and mandibles) as well ag in the ganglia.

The rudimentary brain here consists mainly of three pairs
of ganglia, and may be spoken of as consisting of three parts
—the proto- deutero- and trito-cerebrum.

In their most anterior region the two optic ganglia are
separated, as they were in the last stage, by a narrow band
of ectoderm cells only one cell deep (fig. 29), but posteriorly
there lies between the optic ganglia a central mass of nervous
cells which, from the appearance of the section shown in
fig. 31 and others, seems to have arisen by an invagination
from the cells of the band of ectoderm which, in the preceding
stage, separated the ganglia of the first pair in this region.

In his work on the Crayfish, while describing the state of
the nervous system in an embryo which is a little more ad-
vanced than this, Reichenbach (1886) figures a central mass
lying between the two optic ganglia. ‘This central mass con-
sists of three parts, two large outer ones which he considers
to be parts of the supra-cesophageal ganglion of either side

4.06 MARGARET ROBINSON.

and a small median strand of cells which is clearly an invagi-
nation.

Nusbaum (1887) describes a median double cord of nervous
tissue in the brain of a Mysis embryo, and in his account of
the development of Oniscus (1886) he states that the brain
in the optic region consists of four lobes, the two outer ones
only being the optic lobes. This, indeed, is the state of things
in this stage. ‘The median cord here more closely resembles
that in Mysis than it does the tiny invagination figured by
Reichenbach (1886).

In neither of his papers does Nusbaum say how this median
cord originates. That it does arise from the band of cells
which connects the optic ganglia at their first appearance is,
I think, certain, and from its appearance in figs. 31 and 32
it seems probable that it arises by invagination.

The central mass is itself double, and the optic ganglia are
slightly bilobed, so that we may say that each ganglion here
consists of two parts—a small inner one which is single, and
a large outer one which is bilobed (fig. 31). But it seems to
me that the real state of things is better expressed by saying
that the protocerebrum in this region consists of a pair of
large lateral bilobed optic ganglia, and a central mass of
nervous tissue which is double.

Tracing the nervous system farther back we find that the
central mass increases in size, and the two lobes of each optic
ganglion gradually fuse into one (figs. 32—34). In a section
just anterior to the first antenna (fig. 85) one can see a still
further separation of the central mass from what is here
the rest of the deuterocerebrum, while in the region of the
first pair of antennee the median mass is partly separated
from the ganglia by a layer of ectoderm cells (fig. 36, f.).
These cells have fusiform nuclei which look as though they
might belong to connective-tissue cells, but the mesoderm at
this stage is not sufficiently advanced for that to be the
case.
On tracing the median mass backwards one finds that it
first becomes single (fig. 37), and finally, in the region of the

ON THE DEVELOPMENT OF NEBALIA. 407

second antenna and the mouth, disappears altogether (figs.
38, 39).

Summing up our knowledge of the brain at this stage one
may say that it consists of three pairs of lateral ganglia and
a median mass of nervous tissue which extends from the pos-
terior region of the first pair of ganglia, where it begins as a
double cord, to just in front of the third pair of gangha,
where it ends as a single cord.

The ganglia of the mandibles and first maxille at this
stage are much spread out and flattened (figs. 40, 41), while
the second maxilla has the merest rudiment of a ganglion.

The Eyes.—Sections through the anterior regions of the
optic lobes show a slight depression in the ectoderm cells
forming the outer wall (figs. 28,29, d.). These cells are pro-
bably the forerunners of the crystalline cone cells or of the
cells of the corneal hypodermis, or, as I believe, of both these
sets of cells. Very slightly behind this depression lies the
anterior limit of the optic ganglion, and on a level with this
the optic invagination (fig. 30, op. iv.). This has now the form
of a solid cup. On the inner side of this now solid invagina-
tion there can be seen a few of the smaller nerve nuclei.
Reichenbach (1886), in his account of the development of the
eye in the Crayfish, says that the invagination first becomes
solid and then divides into two layers, the inner of which
furnishes cells to the optic ganglion, while the cells derived
from the outer layer become retinule. The state of things
shown in fig. 80 is very ike that described and figured by
him in a Crayfish embryo at about this stage.

Other Ectodermal Structures.—The stomodeum still
runs forward from the mouth, and in this stage there is the
first appearance of the anus and proctodeum (fig. 42, proc.).
In a transverse section through the caudal papilla of an
embryo with nauplius appendages I noticed two exceptionally
large ectoderm cells. In this stage also I find two very large
cells in similar sections just behind the anus (fig. 43, l.c.), but
I have not succeeded in tracing these cells in later stages.

Nusbaum (1887) found similar cells in a very early stage in

408 MARGARET ROBINSON.

Mysis. He believed them to be genital cells, and found
them later in the abdomen, later again in the thorax in a
ventral position, and later still in a dorsal position between
the digestive canal and the heart. However, he is not very
convincing in this part of his paper, and I am inclined to

SS
“N

SY

Diackam 2.—Median longitudinal section through embryo at
Stage D. Ze. Ectoderm. Zz. Endoderm. Mes. Mesoderm. Proc.
Proctodeum. S¢. Stomodeum.

think that these large cells, in Nebalia at least, are simply
part of a zone of growth similar to the Knospungs-zone
described by Reichenbach. This zone of growth is mentioned
by Nusbaum, but has been much more fully described in
Mysis by Bergh (1893). As I unfortunately did not notice
it while making surface views of the embryos I have been
unable owing to want of more material to describe it.

The Dorsal Organ.—I have not found a trace of a
dorsal organ in any of my stages, and in this my account
of the development differs from that of Butschinsky
(1900).

ON THE DEVELOPMENT OF NEBALIA. 4.09

In this stage in the region of the first antenna there is a
decided thickening of the ectoderm on either side of the
body-wall dorsal to the antenne. These thickenings are
lateral, but more ventral than dorsal in position. I take
them to be the earliest traces of the shell valves (fig. 37, s.v.).

Mesoderm.—There is no great advance in this layer upon
its condition in the embryo with nauplius appendages. We
find mesoderm as far forward as the optic lobes, though not
in the lobes. It is to be found in the appendages (1. e. in the
first and second antennez and mandibles (figs. 35—39). As
yet there is no sign of a split in the mesoderm except in the
second antenna (fig. 39) where there is the appearance of a
split which may, however, be accidental and due to reagents.
Vitellophags are still fairly abundant.

Endoderm.—This consists of large columnar cells with
large nuclei. It is still confined to the hinder part of the
body, extending on the dorsal side from the proctodeeum to
the region of the optic lobes. Ventrally it does not extend
so far forward. Only in the abdominal papilla does it
completely surround the yolk as is shown in the above
diagram (2).

Stage EH (fig. z).

External View.—The chief new features to be noticed in
an external view of the embryo at this stage are the increase
in the length of the appendages, the appearance of the median
head flap, and the great increase in the size of the labrum.

There is also a change in the position of the appendages.
They now more or less follow the outline of the body and are
directed backwards instead of standing out at an angle to
the body as they did in the last stage. The shape of the
body itself too has altered a little. It is now longer and
more oval than in Stage D. The posterior half is, however,
much narrower than the front end, and the papilla stretches
farther forward than it did in the last stage.

Internal Structure. Hctoderm; the Nervous

410 MARGARET ROBINSON.

System.—The arrangement of the nervous elements which
form the brain is very similar to that described in the last
stage. The large ganglion cells still form part of the general
ectoderm of the body. The protocerebrum still consists of
two large lateral and two small median lobes (fig. 44), but
the central mass reaches farther forward than it did, only the
most anterior region of the protocerebrum being now without
this central mass between the optic ganglia.

The most anterior parts of the optic ganglia resemble those
in Stage D, i. e. they consist merely of the two kinds of nerve-
cells there described.

The deuterocerebrum has grown considerably, and now
consists of two parts:

(a) A median mass of nervous cells lying in front of the
first antennee.

(b) A pair of ganglia lying on a level with, and inner-
vating the first pair of antenne.

The anterior of these parts is, in reality, formed by that
part of the central mass of nerve-tissue which les imme-
diately behind the optic region.

Reichenbach (1886) in describing the brain of a similar
stage in the Crayfish says that the anterior parts of the supra-
cesophageal ganglia are separated from their posterior parts on
the one hand, and from the optic ganglion on the other by
strands of connective tissue. Now, in a transverse section
through the first antenna in Stage D (fig. 36, f.) there are to
be seen cells which partly separate the median mass from the
ganglion on either side, but these cells have the appearance
of ectoderm cells. They do not resemble the mesoderm cells
of Stage D, and, as yet, there is no mesoderm in the ventral
part of the brain. In Stage E it is only just beginning to
enter the brain from the dorsal, i.e. the yolk side.

Further, the cells lining the slight groove which separates
the two lateral halves of the deuterocerebrum ventrally are
certainly ectodermal. ‘They can be seen forming a slight
inpushing in fig. 47, and also as wedge-shaped cells in figs.
50, 46, and 34 (wed.).

ON THE DEVELOPMENT OF NEBALIA. 411

In sections through developing Mysis Nusbaum (1887)
showed ectoderm cells separating the different parts of the
brain from each other. In my longitudinal sections through
the brain at this stage I have not been able to demonstrate
such cells, though fig. 48 shows the different parts of the
brain fairly well.

In the optic lobes, and in both parts of the deutero-
cerebrum, there are now to be seen fibres, as well as the two
kinds of nerve-cells.

In the optic lobes the fibres are present in the posterior
half only. Fig. 44 shows a transverse section taken about
midway through the optic lobes. In it one can see the two
lateral optic ganglia, each slightly bilobed, and the central
mass, which is also bilobed, the lobing in both cases being
much more distinctly marked on the dorsal than on the
ventral surface. The median mass is, as I have said above,
in direct continuity with the anterior part of the deutero-
cerebrum. In this latter region it is widely bilobed on its
dorsal side, and transverse fibres can be seen passing between
its lateral parts (fig. 47). Fibres can also be traced from the
antennulary ganglion into the antennule (fig. 50, 7.f.).

The smaller nerve-cells are grouped very definitely and
symmetrically all through the brain, forming more or less
geometrical designs (figs. 44—46), but the main arrangement
of the nervous elements is that large ganglion cells lie
outside, within these smaller ganglion cells, and within these
again, fibres (figs. 44—51).

Behind the region shown in fig. 47 the anterior part of the
stomodzum pushes up, as it were, between the two latero-
dorsal lobes of the central nervous mass, so that at the level
of the first antenna there are no more transverse fibres to be
seen, and the ganglia of the antennules appear to be in
direct continuity with the ventral and dorso-lateral portions
of the central mass, which here comes to an end (fig. 49 and
50). As in the preceding stage, the tritocerebrum, which
consists merely of the pair of ganglia which innervate the
second antenne, is not nearly so large as either of the two

412 MARGARET ROBINSON.

parts of the brain which le in front of it (fig. 48, t.c.).
These ganglia of the third pair are separated from each other
by the labrum (fig. 51). In some specimens here, as in the
first antenne, one can see fibres running from the ganglion
to the antenna. ‘These fibres seem to run from the smaller
dark-staining nuclei. This fact, taken with the arrangement
of these small nuclei in the patterns alluded to above, leads
one to think that the fibres originate from these small nuclei.

To sum up. The brain in this stage, as in the last,
consists of three pairs of ganglia and a central mass of
nervous tissue. ‘This central mass extends farther forward
than it did in the last stage, and in it, and in each of the
ganglia, nerve-fibres have made their appearance. ‘The deu-
terocerebrum consists of two parts lying one behind the
other, but the anterior of these parts is not the anterior
portion of the second pair of ganglia, but the posterior
portion of the central mass of nervous tissue mentioned
above. As has been stated above, this central mass does
not originate from the primitive ganglia first seen in the
embryo with nauplius appendages, but from the median band
of cells lying between the ganglia of the first and of the
second pairs.

In the stage at present under consideration it extends
from almost the most anterior region of the space between
the optic ganglia to the place where its two lateral lobes join
the ganglia of the antennules. It consists now of two parts
—(a) lying between the optic ganglia, (b) lying behind this
and in front of the ganglia of the first pair of antenna.

Between the antennary ganglion and that of the mandible
on either side there is a narrow chain of nerve cells, repre-
senting a future commissure.

There is, as in the last stage, a deep groove between the
two ventral halves of the mandibular ganglion. In the
anterior part of this ganglion fibres can be seen in the
middle of each half (fig. 52 n.f.). It is difficult to say whether
the ganglion is formed by the fusion of merely the ganglia
of either side, or by the fusion of three elements, the ganglia

ON THE DEVELOPMENT OF NEBALTA. 413

of either side, and a median strand. I incline to think that
it is formed from the fusion of two ganglia only, though the
central part looks very like the central mass described above,
and Reichenbach describes a median strand as forming part
of the ventral chain in the Crayfish.

Posteriorly the mandibular ganglion is flattened and spread
out as it was in the last stage, but the ganglion of the first
maxilla has now assumed a more compact shape. It has no
fibres as yet.

The Hyes.—Fig. 46 shows the optic invagination fairly
well. Other sections in the same series show a closer
contiguity between the cells with large colourless nuclei and
the small dark-staining nuclei of the nerve-cells which here
abut on them. It seems not improbable that these small
nerve-cells have originated from the optic invagination,
though I can find no more definite indication of it than that
shown in fig. 30.

In contra-distinction to the other cells at this stage the
cells of the outer wall of the invagination show distinct cell
outlines. Their nuclei, too, are larger and paler than those
of the other cells.

Other Ectodermal Structures.—The labrum (fig. 52
and fig. £)is much larger than it was in Stage D, and in some
specimens there is a great advance in the growth of the
stomodeeum, the future stomach being distinctly foreshadowed
(figs. 53, 54 st.).

The lateral ectodermal thickenings (future shell valves)
have increased in size (figs. 51 and 53). In one or two
specimens I have found a distinct median dorsal thickening
of the ectoderm. This I take to be the first stage in growth
of the dorsal part of the bi-valve shell (figs. 53, 54 s.).

Mesoderm.—The mesoderm cells have increased in
number, and are now pretty evenly distributed along the
whole length of the embryo, i.e. they are now no longer
more abundant in the papilla than elsewhere. Fig. 54 shows
several mesoderm cells lying round the stomodzum, some of
them very closely attached to it.

4.14, MARGARET ROBINSON.

In fig. 54 I have shown seven mesoderm cells (h.c.) lying
just under the dorsal thickening of the ectoderm. These
I think are probably the first traces of the heart. It is in
just this way, according to Nusbaum (1887), that the heart
first appears in Mysis.

At the base of the second antenna the rudiment of the
antennary gland can be distinctly seen in some specimens

Jsleletelle

a ON

Dracram 3.—Median longitudinal section through embryo at
Stage E. Lettering as in Diagram 2.

(fig.55). Itis, roughly speaking, triangular in shape with a
triangular lumen. ‘The cells, of which there are not more
than six or seven, have a finely granular protoplasm with
here and there a Jump or rather globule of something which,
since it stains in the same way as the yolk, I take to be the
excretion. There are still some vitellophags to be seen.

Endoderm.—On the dorsal side the endoderm now extends
along almost the entire length of the embryo, but ventrally
still only a little way in front of the curve formed by the
folding over of the papilla (Diagram 3).

ON THE DEVELOPMENT OF NEBALIA. 415

Stage F (fig. r).

Kxternal View.—The embryo now has burst its shell,
but is still enclosed in a cuticle. Metschnikoff says of this
cuticle that ‘it invests all the first five appendages (nauplius
appendages and maxillz) closely following all their curves
and outgrowths. The rest of the appendages are covered
by an unbroken skin which forms a general sac over them
and the back.” In looking at the embryo under a dissecting
microscope I have not succeeded in making out this cuticle,
much less its disposition, but in many of my sections [ have
noticed pieces of it loosely enveloping the embryo.

In breaking the shell the embryo has become uncurled.
It seems, in fact, to have sprung back with a rebound, so that
now there is a slight dorsal curvature of both head and tail.
It has now, besides the eyes, antennz, mandibles, and maxille,
seven, sometimes eight, thoracic appendages, sometimes also
the rudiments of three abdominal legs. ‘The optic lobes are
now almost completely nipped off from the yolk. All the
thoracic appendages are bilobed so that if the embryo were
free-swimming one might call this the Mysis stage.

Internal Structure. Nervous System.—The nervous
cells are now leaving the surface, i.e. they now no longer
form part of the ectoderm of the body wall. This can be seen
in fig. 56 e.c., where there is shown a thin layer of ectoderm
lying outside the large dividing cells of the ganglion of the
antennule.

Protocerebrum.—Concomitant with an increase in size
of the optic ganglia there is an increase in the number of
fibres which they contain. ‘The nerve cells in these ganglia
have a very regular arrangement resembling that shown in
Nusbaum’s figures of the brain of the Mysis embryo
(Nusbaum, 1887, fig. 80).

The central mass of nervous tissue between the optic
ganglia has grown considerably. It is now very distinctly
bilobed, and, in section, can be seen to be clearly marked off

VoL. 50, PART 3,—NEW SERIES, 30

416 MARGARET ROBINSON.

from the central mass which forms the anterior part of the
deuterocerebrum (fig. 57).

The Deuterocerebrum now consists of three distinct
regions :

(a) An anterior, median, bilobed mass lying directly be-
hind the central mass of the protocerebrum (fig. 57, ¢c.m.).

(b) Two large outer lobes which lie outside of and slightly
behind the hinder part of (a), and which are in direct con-
tinuity with—

(c) The ganglia of the antennules. These are slightly
larger than they were in Stage E.

Tritocerebrum.—This has grown considerably and each
ganglion now consists of two parts :

(1) A ventral mass of nerve cells lying close to the
stomodzeum, and—

(2) A larger dorso-lateral mass lying farther from the
middle line (fig. 60).

These masses are in continuity with each other except at
the point on either side where fibres pass from the ganglion
to the antenna. The bay between these masses of cells is
filled with fibres, and, posteriorly, each ventral mass is con-
tinued as a thin cord or chain of cells which connects the
antennary ganglion with that of the mandible on the same
side.

There is more mesoderm in the brain than was present there
in the last stage. This mesoderm consists of chains of large
cells which run more or less obliquely from the stomodzeum
to the dorsal body wall (figs. 59 and 60). These chains are
the forerunners of the bands of connective tissue and muscle
which connect the stomodeeum with the dorsal body wall in
the later stages.

The brain here closely resembles that of Mysis at a like
stage of development.

In the mandibular ganglion the two dorso-lateral clumps
of ganglion cells present in the adult are beginning to make
their appearance, and, as in the adult, there are ganglion cells
in the centre showing the double origin of this ganglion. The

ON THE DEVELOPMENT OF NEBALIA. 417

space between the central mass of ganglion cells and the
dorso-lateral mass on each side is spanned by fibres (fig. 61,
g. Iv, also Claus 1889, fig. 3, Taf. 1x).

The ganglia of the first and second maxille both consist,
at their widest parts, of a ventral mass of ganglion cells
which is continued dorso-laterally into two rounded masses of
cells, one on each side, the space between these masses being
spanned by fibres. In shape these maxillary ganglia are very
much like those of the adult, and, as in the adult, the fibres
from them to the maxillz go off between the dorso-lateral
humps on each side and the ventral mass.

The first, second, and third thoracic ganglia are similar to
those of the maxillz. Between the ganglia there runs a
double cord of nerve cells. The centre of each half of this
cord is filled with fibres (fig. 64, nf.).

The posterior ganglia of the ventral chain are as yet tri-
angular masses of nerve cells, each triangle having a slightly
double appearance, and lying with its apex directed ventrally.

One cannot at this stage speak with certainty as to the
future part to be played by the central mass of nervous tissue.
It seems not improbable that it furnishes fibres in the later
stages of brain development.

As will be seen from the above I have not been able to find
a central mass in any of the ganglia of the ventral chain.

The Hyes.—The separation of the optic lobes from the
yolk, which, indeed, began in the more advanced specimens
of Stage EH, is here almost complete.

On the outer and lower edge of each optic lobe the outer
ectodermal layer can be seen in some places to be more than
one cell deep (figs. 59 and 60), and this leads me to believe
that this layer gives rise not only to the corneal hypodermis,
but also to the cone cells.

The large cells with pale nuclei can be easily recognised
again here (figs. 58, 59, and 60, p.n.), though cell outlines can
now no longer be distinguished. Nor can one at this stage
see any connection between these large cells and those of the
optic ganglion.

418 MARGARET ROBINSON.

Unfortunately this stage is not sufficiently advanced to
enable one to write definitely about the future of the different
parts of the eye mentioned here; still it seems almost certain
that the outer cells of the optic thickening first seen in Stage
B, seen also in Stages C and D, to lie in front and outside of
(i.e. more lateral than) the optic invagination and the optic
ganglion, and to be traced in the later stages, do really fur-
nish the cone cells and the corneal hypodermis, while the
optic invagination furnishes the retinule, and in the early
stages certainly gives off some cells for the increase of the
optic ganglion.

I have not been able so far to find any mesoderm cells
between the optic ganglion and the future retina.

Since Reichenbach (1886) a great deal has been written
about the development of the crustacean eye as a whole, and
the optic mvagination in particular. Kingsley (1887) seems
to derive the whole eye, corneal hypodermis excepted, from
the invagination which he says never becomes solid. Parker
(1891) was of opinion that when an invagination occurs it is
concerned with the optic gangliononly. Inalaterpaper (1895),
however, he seems to have veered round to Reichenbach’s
view as to the Crayfish eye, viz. that the outer wall caused
by a division of the primary invagination forms the retina,
and the inner furnishes cells to the ganglion. ‘There can be
little doubt as to the homology of the proliferation in
Branchipus, the Lobster and Mysis, with the invagina-
tionin the Crayfish and Nebalia. Parker (1895) says that
eventually in the Lobster, and probably in the Crayfish
the ganglion loses its connection with this centre of growth,
which continues as a growing area for the retina only. But
Claus (1889), on the other hand, considered that the prolifera-
tion in the Branchipus larva was continued as the zone of
growth in the adult, and that this zone furnished ganglion
cells in the direction of the ganglion, and retina cells in the
direction of the retina. He believed the zone of growth in
the eye of the adult Nebalia to be homologous with that in
the eye of Branchipus, and that in the Crayfish, In the

ON THE DEVELOPMENT OF NEBALIA. 419

main he was in agreement with Reichenbach for he considered
the proliferation in Branchipus to be virtually the same
thing as the invagination in the Crayfish. Also, it is quite
clear that he looked upon the zone of growth in the adult
Nebalia as having originated in either a proliferation or an
invagination.

Other Ectodermal Structures.—The stomodzum has
grown considerably, and the shape of the future stomach is
foreshadowed more definitely than it was in the last stage.
Its two lateral walls are now composed of deep columnar cells
arranged so as to form two curved pads which nearly meet in
the middle (fig. 61, st.).

The labrum has grown larger, and in sections one can see
that jointing has begun in the antenne (figs. 57—59).

In the region of the mandibles the thickening of the dorsal
ectoderm of the body wall mentioned in Stage E can again
be seen (fig. 61). This, I think, is precedent to the formation
of the dorsal part of the shell.

The two lateral shell thickenings have greatly increased in
size, and fig. 64 f. shows a fold beginning to form between
one thickening and the rest of the ectoderm in its region.

Endoderm.—tThe yolk is now surrounded by endoderm
except in the most anterior part of the embryo. In the
thoracic region two latero-ventral outgrowths of the endoderm
(which here completely surrounds the yolk) can be seen
(figs. 63 and 64). These I take to be the beginnings of the
liver lobes. Itis in this way that the hepatic lobes in Mysis
first make their appearance (Nusbaum, 1887).

Metschnikoff (1868) in describing this stage states that
the yolk which is surrounded by endoderm is a coherent fluid
mass, while that in the anterior region which is unsurrounded
is broken up into cone-shaped lumps. My sections show a
similar difference between the surrounded and unsurrounded
yolk. Also, in the surrounded yolk I find no vitellophags,
while in the unsurrounded portion some of these cells are
still to be seen. This, I think, goes to prove that the
vitellophags do in some way soften the yolk, and that when

420 MARGARET ROBINSON.

it is ready for absorption by the endoderm they (having done
their work) disappear. The appearance of yolk in preserved
embryos is, of course, to a certain extent influenced by the
fixative used.

Mesoderm.—There is an increase in the number of
mesoderm cells, but as yet they show no very definite
arrangement except in the region of the stomodeum where
they are beginning to form a layer one cell deep round the
stomach (fig. 61, mes.).

Figs. 59 and 60 show mesoderm bey running in oblique
lines dorsalwards from the dorsal wall of the stomodzum.
These lines doubtless represent the bands of connective tissue
and muscle which in later stages run from the stomach and
cesophagus to the dorsal body-wall.

In the region of the second maxilla, beginning at the
posterior end of its ganglion, there are on either side
stretching outwards from the ganglion towards the shell-
thickening, three mesoderm cells which form a rudimentary
muscle (fig. 62). The figure represents a transverse section
taken slightly behind the ganglion. I have not been able to
trace fibres from the ganglion to the muscle, but have little
doubt that it is mnervated from this ganglion (that of the
second maxilla). Therefore it does not foreshadow the great
transverse shell muscle of the adult which is innervated from
ganglion of the first maxilla (Claus, 1889).

The Heart.—Of the seven mesoderm cells which were
present under the dorsal ectodermal thickening in the last
stage, I can find but one here (fig. 61). However I have
little doubt that those cells do represent the heart in its
earliest stage, for in a stage which is but very little older
than this (Stage I") one sees in the same region in which this
band of cells was present in Stage EH (i. e. the region near the
posterior end of the stomodeum) a few mesoderm cells
arranged round a definite lumen just under the dorsal shell-
thickening (figs. 55a, 55b,55c). This must be the anterior
end of the heart or dorsal aorta. I imagine this to have been
formed by a bending inwards of the mesoderm cells so as to

ON THE DEVELOPMENT OF NEBALIA. 421

form a space between themselves and the dorsal ectoderm.
It is much in this way that the heart in the Crayfish arises
(Reichenbach, 1886).

The Antennary Gland.—In the stage with nauplius
appendages mesoderm cells can be seen being carried into
the appendages as they are folded off from the rest of the
blastoderm (figs. 20—22, mes.). In Stage D mesoderm cells
can be seen lying in the antennz between the two layers of
ectoderm, though there is there very little histological
difference between ectoderm and mesoderm (figs. 37—39, mes.).
Still, having traced mesoderm into the antennee and finding
no sign of an ectodermal invagination of any kind I have
little doubt that the glands which appear at the bases of the
second antenne in Stage EH are mesodermal in origin. ‘hese
glands in this stage (F) are in the same condition as in
Stage Hi (figs. 55, 59). There is no duct as yet, nor is there
any sign of a gland in the second maxilla. Claus (1889)
showed that in the adult Nebalia both glands function, and
bis observations have lately been confirmed by Bruntz (1904).
In the adult the antennary gland is very rudimentary, and
that in the second maxilla still less developed. Claus (1889)
uses these facts as evidence in favour of Nebalia’s belonging
to the Malacostraca; for, he says, that while in Phyllopods
the antennary glands appear first, and function, while the
shell glands, if present, are insignificant and functionless, in
the adult the autennary glands have dwindled, and it is the
shell glands which function. In the Malacostraca, he
continues, the relative importance of these two glands at the
different ages is reversed, and in the adult itis the antennary
eland which functions while the maxillary gland, if present,
is comparatively insignificant. These statements have
received further confirmation in Dr. Allen’s work on the
Nephridia of Decapods (1893). But it seems that logically
speaking this piece of evidence, though it certainly helps to
remove Nebalia from the Entomostraca, does not help to
place it among the Malacostraca; for though in the adult
Nebalia the antennary gland is the larger and the more

422 MARGARET ROBINSON.

important it is also the first to appear in the embryo, and
apparently functions before there is any sign of a gland in
the second maxilla. I must add that I have found antennary
glands in two stages which are more advanced than Stage F,
and maxillary glands in one of those stages. There is no
sign of degeneration in the glands of the adult as compared
with the glands in those late embryonic stages.

It seems to me that these facts about the excretory glands,
if they show anything at all as to the systematic position of
Nebalia, point rather to its having come off from the original
Crustacean stem (from a form in which both glands were
equally developed) before either the Entomostraca or the
Malacostraca, than to its being descended from either a
primitive Phyllopod (Entomostracan) or a primitive Malaco-
stracan.

Genital Cells.——In the thoracic region close to the
hepatic outgrowths there is on each side, lying almost
between the outgrowth and the mesenteron from which it
has arisen, a small group of three or four mesoderm cells
(figs. 63, 64). Nusbaum (1887) found what he considered to
be genital cells in this place in a similar stage in Mysis,
though he somewhat unsatisfactorily traced them from the
abdomen to this position, and considered them to be ecto-
dermal in origin.

Wagner (1894) states that in Mysis the genital cells
appear very early as lateral outgrowths of Bergh’s (1893)
“entoderm disc,” and that they travel from a ventral to a
dorsal position, which last they reach at a very late stage,
having on the journey become surrounded by a layer of flat
mesoderm cells. It can be seen that neither of these authors
is very convincing as to the origin of the genital cells in
Mysis, and therefore it is with much hesitation that I suggest
the possibility that these cells are genital cells. In my figure
they certainly appear to be mesodermal, though they are
darker, and more granular, as well as slightly smaller than
the other mesoderm cells. If they are not genital cells they
may possibly be future connective-tissue cells.

ON THE DEVELOPMENT OF NEBALIA. 425

The vitellophags are much reduced in number. They are,
in fact, only to be found in the unsurrounded yolk at the
front end of the embryo.

CoNCLUSION.

Claus’s last paper is, among other things, a great summing
up in favour of the Malacostracan position of Nebalia. The
history of the development, as far as I have taken it, can do
little but make his position still stronger. The thoracic limbs
are perhaps the most Phyllopod-like feature that Nebalia
possesses, and Claus has shown these to be intermediate in
character, between those of a Phyllopod and those of the
Schizopoda.

A Malacostracan feature in the development of Nebalia,
to which, I believe, notice has not yet been drawn, is the
sharpness of definition with which the embryonic stages are
marked off from one another.

Among the Malacostraca the form which appears to be
most nearly related to Nebalia is Mysis. The organs
which are alike in the adults are alike also in their develop-
ment. To begin with external points. The very definite
early embryonic stages in Nebalia resemble very closely
the early embryonic stages in Mysis, and in both animals
the young stay in the brood pouch of the mother till they
are practically adult. The brood pouches in the two animals
are formed in the same way by spiny ontgrowths on the
coxopodites of the thoracic legs.

The peculiar form of gastrulation, the development of the
endoderm in Stage B, the subsequent formation of the mid-
gut by circumcrescence, and the development of the liver
lobes in Nebalia all closely resemble the facts as recorded
for Mysis (Nusbaum, 1887, and Bergh, 1893).

If the heart arises and develops as I have suggested above,
then its development resembles that in Mysis more nearly
than that in any Phyllopod. Another point of resemblance

424. MARGARET ROBINSON.

lies in the development of the brain with its large central
mass of nervous tissue (Nusbaum, 1889). There is also a
likeness in the early stages of the development of the eyes,
assuming that the proliferation in Mysis represents the
invagination in Nebalia. If the cells mentioned above be
indeed genital cells, there is still another point of resem-
blance in the development of the genital organs which in
the adult Mysis and Nebalia are very much alike. The
accounts of the development of these organs in Mysis given
by Nusbaum (1887) and Wagner (1894) are certainly not
very lucid, but they agree in one point. According to each
of these accounts the cells travel from a ventral to a dorsal
position, and this is what the cells shown in fig. 66 will have
to do if they be really genital cells. They certainly resemble
very closely those shown by Nusbaum in an embryo of
Mysis.

It is very unfortunate that we have at present no account
of the development of the excretory glands in Mysis. I
have, in a somewhat circuitous way, found out that its
antennary and maxillary glands develop in the order which
obtains for these glands in Nebalia. Nusbaum (1887),
though he carried his investigations on to late stages in
development, says that he could find no trace of excretory
glands in his specimens. I have cut sections through two
Mysis embryos which, since their genital organs are slightly
more dorsal in position than those shown in Nusbaum’s
figures, I take to be a little older than his latest stages. In
each of these I found a well-developed antennary gland, but
no true trace of a gland in the second maxilla.

In a paper on excretion Kowalewsky and Metschnikoff
(1889) state that shell glands have been found by Claus in
the larvee of Myside. Unfortunately I have been unable to
verify this assertion, as they give no reference, and, though
I have looked through numbers of papers by Claus, I can
find no record of the observation. Assuming Claus’s obser-
vation to have been correct, J am forced to the conclusion
that in the embryos, in which I found antennary glands but

ON THE DEVELOPMENT OF NEBALIA. A425

no maxillary glands, these latter had not yet appeared,
and that, therefore, the antennary gland in Mysis, as in
Nebalia, develops before that of the second maxilla. A
maxillary gland has not yet been recorded for the adult
Mysis. There can be no doubt that the glands on the
thoracic legs in Nebalia, in spite of their alkaline reaction
(Kowalewsky and Metschnikoff, 1899), are morphologically
equivalent with the glands in the same position in Mysis,
Squilla, and other Malacostraca.

As far back as 1868 Metschnikoff suggested that, in a
classification of the Crustacea, Nebalia should be placed by
the side of the Schizopods. Since then, as can be seen in
the historical section of this paper, other observers have
from time to time noted the likeness of Nebalia to the
Schizopods or the Myside. Claus (1886) and Grobben (1892),
in their schematic trees (which differ only as regards the
Stomatopoda) make the Lepostraca come off from the Proto-
malacostraca with the Protoschizopoda. Inthe present state
of our knowledge this arrangement perhaps shows the rela-
tionship between Nebalia and the other Malacostraca better
than any other, for the Mysidz are undoubtedly the most
primitive Schizopods.

The classification of the Malacostraca by Grobben (1892)
and Hansen (1893) into Leptostraca and Humalacostraca, and
the division of the Schizopoda by Boas (1883) and Hansen
(1893) also make apparent the nearness of the Myside to the
Leptostraca; but Calman (1904) in his amplification of
Hansen’s classification of the Malacostraca seems to have
masked this nearness. He has done this by the introduction
of Anaspides into the Humalacostraca. We know very
little about this animal, and nothing of its development.
Judging from the account of its internal organs given by its
discoverer (Thomson, 1894) it would seem to differ so much
from the other Crustacea as to warrant its not. being given a
place at least until some well preserved specimens have been
examined.

426 MARGARET ROBINSON.

Since writing the above I have read a paper in the
‘Quarterly Journal of Microscopical Science,’ for December,
1905, in which the author, G. H. Carpenter, while suggesting
that the Leptostraca are the most primitive Crustacea, and
admitting their nearness to the Malacostraca, to which he
says ‘‘ they may be, to some extent, ancestral,’ states that he
considers their nearest relations among the Entomostraca to
be not the Phyllopods but the Copepods. This leads me to
think that, since it has been my aim to emphasise the Mala-
costracan position of Nebalia, and its nearness to the
Myside, I have, perhaps, rather imphed than explicitly stated
(as I should have done), that though one cannot now think
that Nebalia is descended from the Phyllopods, or, indeed,
from any of the Entomostraca, yet I believe that its nearest
allies among the Entomostraca are the Phyllopods. Itseems
to me that, leaving aside other points of resemblance between
Branchipus and Nebalia, the likeness between the thoracic
limbs of Nebalia and those of Branchipus and Apus
cannot be accounted for by homoplasy. Thinking over this,
I have been led again to Professor Lankester’s illuminating
papers (1881—1904), the reading of which has only
strengthened my previous convictions.

It seems not improbable that Nebalia is the most ancient
Crustacean of which we know at present. Perhaps the
strongest piece of evidence for this view of its position lies
in the fact that the adult animal has three pairs of ccelomo-
duets. If, however, I were to speculate as to the Ancestral
Crustacean I should be inclined to imagine it as possessing
the most primitive features not of Calanus, Nebalia, and
Triarthrus, but of Apus and Nebalia.

March, 1906.

REFERENCES TO LITERATURE.

Auien, BE. J.—1893. ‘On the Nephridia and Body Cavity in some Decapod
Crustacea,” ‘ Quart. Journ. Mier. Sci.,’ vol. 35.

ON THE DEVELOPMENT OF NEBALIA. 42'°7

Bereu, R. S.—1893. ‘Zur Bildungsgeschichte des Keimstreifen von Mysis
Zool. Jahrbiicher’ (Anatomie), Bd. vi, Jena.

Boas, J. E. V.—1883. ‘Studien ueber die Verwandtschafts beziehungen
der Malakostraken,” ‘ Morph. Jahrb.,’ Bd. viii.

Bruntz, L.—1904. ‘Contribution a Etude de ’Exerétion chez les Arthro-
podes,” ‘ Archives de Biologie,’ tom. xx.

Butscutnsky, P.—1897. “Die Furchung des Eies und die Blastoderm-
bildung der Nebalia,” ‘Zool. Anzeiger,’ Bd. xx.

—— 1900. “Zur Entwicklungsgeschichte der Nebalia Geoffroyi,”
‘Zool. Anzeiger,’ Bd. xxiii.

Cauman, W. T.—1904. ‘On the Classification of Crustacea Malacostraca,”
‘Annals and Mag, Nat. Hist.,’ ser. 7, vol. xii.

Carpenter, G. H.—1905. “Notes on the Segmentation and Phylogeny
of the Anthropoda, with an account of the Maxille in Polyxenus
lagurus,” ‘Quart. Journ. Mier. Sci.,’ vol. 49, part 3, N.S., No. 195,
Dec., 1905.

Craus, C.—1872. ‘Ueber den Bau und die System Stellung von Nebalia

nebst Bemerkungen iiber das seither unbekannte Mannchen dieser
Gattung,” ‘ Zeitschr. f. wiss. Zool.,’ Bd. xxii.

1873. ‘Zur Kenntniss des Baues und der Entwicklung von
Branchipus und Apus,” ‘Abhand. Gesell. d. Wissensch.,’ Gottingen,
Bd. xviii.

1876. “ Untersuchungen zur Erforschung des Crustaceen Systems.’
5D to) Yi

1886. ‘Neue Beitrage zur Morphologie der Crustaceen,” ‘ Arbeit.
Zool. Instit.,? Wien, Bd. vi.

1886. ‘ Untersuchungen tiber die Organisation und Entwicklung
von Branchipus und Artemia,” ‘ Arbeit. Zool. Inst.,’ Wien, Bd. vi.

1889. “ Ueber den Organismus der Nebaliden und die Systematische
stellung der Leptostraken,” ‘ Arbeit. Zool. Instit.,’ Wien, Bd. viii.

Fasricius, O.—1780. “Fauna Groenlandica,” Hafnie et Lipsie.

(ROBBEN, C.—1892. ‘Zur Kenntniss des Stammbaues und des Systems der
Crustaceen,” ‘Sitz. Kaiser]. Akad. d. Wissench.,’ Wien, Bd. ci.

Hansen, H. J.—1893. “Zur Morphologie der Gliedmassen und Mundtheile
bei Crustacen und Insecten,” ‘Zool. Anzeiger,’ Bd. xvi.

Hersst, J.—1796. “Versuch. einer Naturgeschichte der Krabben und
Krebse,”’ Berlin.

Hersst, C.—1896. “Ueber die Regeneration von Antennenihnlichen
Organen,” ‘ Archiv f. Entwickelungs Mechanik.,’ Bd. ii.

428 MARGARET ROBINSON.

Heymons, R.—1901. ‘Die Entwicklungsgeschichte der Scolopender Zoo-
logica,’ Bd. xiii, Heft xxxiii.

Kinestry, J. S.—1887. “The Development of the Compound Eye of
Crangon,” ‘Journal of Morphology,’ vol. i.

Korscuett, E., und Herper, K.—1893. ‘Lehrbuch der vergleichenden
Entwicklungsgeschichte der wirbellosen Thiere,’ Bd. ii.

Kowatewsky, A.—-1886. “Zur embryonalen Entwicklung der Musciden,”’
‘ Biol. Centralbl.,’ Bd. vi.

und Merscunixorr, H.—1889. ‘Beitrag zur Kenntniss der Ex-
kretionsorgane,” ‘ Biol. Centralbl.,’ Bd. ix.

Kroyer, H.—1847. ‘Karcinologische Bidrag. Naturhist.,’ Tidskrift ii,
Raekke, Bad. iv.

Lanxester, FE, R.—1881. “Observations and Reflections on the Append-
ages and on the Nervous System of Apus cancriformis,” ‘ Quart.
Journ. Mier. Soc.,’ N.S., vol. 21.

1904. “The Structure and Classification of the Arthropoda,”
‘Quart. Journ. Mier. Soc.,’ N.S., vol. 47.

— 1904. ‘The Structure and Classification of the Arachnida,”
‘Quart. Journ. Micr. Soc.,’ N.S., vol. 48.

LatREILLE, P. AA—1831. ‘Cours d’Entomologie,’ Paris.

Leacu, W.—1814. ‘ Zoologist’s Miscellany.’

MetscunikorF, E.—1868. ‘‘On the Development of Nebalia” (Russian),
‘Acad. Imp. d. Sci.,’ St. Petersburg, tom. xiil.

Mitne-Epwarps, H.—1828. ‘‘ Mémoire sur quelques Crustacés nouveax,”
‘Ann. Sci. Nat.,’ Paris, tom. xiii.

1835. ‘Note sur le genre Nebalia,” ‘Ann. Sci. Nat.,’ sér. 2,

tom. ill.

1840. ‘Histoire naturelle des Crustacés,” tom. ii, Paris.

Monvtacu, G.—1818. ‘‘ Descriptions of several New and Rare Animals dis-
covered on the South Coast of Devonshire,” ‘Trans. Linn. Soc.,’ Lond.,
vol. ix.

Nussaum, J.—1886. ‘L’Embryologie d’Oniscus murarius,” ‘Zool.
Anz.,’ vol. ix.

1887. ‘L’Embryologie de Mysischameleo,” ‘ Archives de Zoologie,
Expé. et générale, Deuxiéme sér., tom. v.

Parker, G. H.—1891. “On the History and Development of the Eye in
the Lobster,” ‘ Bull. Mus. Comp. Zool.,’ Harvard, vol. xx and xxi.
1895. ‘The Retina and Optic Ganglion in Decapoda,” ‘ Mittheil.

Zool. Stat. Neapel,’ Bd. xii,

ON THE DEVELOPMENT OF NEBALIA. 429

ReicuenBacu, H.—1886. ‘Studien zur Entwicklungsgeschichte des Fluss-
krebses.’

Sars, G.O.—1885. ‘Challenger Reports,’ vol. xiii.

1887. ‘Challenger Reports,’ vol. xix.

Tuomson, G.—1894. “Ona Freshwater Schizopod from Tasmania,” ‘Trans,
Linn. Soe.,’ London.

Viattanrs, H.—1892. “Systeme nerveux des Articulés.” ‘Annales des
Sciences naturelles,’ 7e série, Zoologie, 13, 14.

v. WitLemogs-Suum, R.—1875. “On some Atlantic Crustacea from the
‘Challenger’ Expedition.” ‘Trans. Linn. Soc.,’ London, ser. 2, Zool.,
vol. 1.

Waener, Jut.—1894. “Zur Entwicklungsgeschichte der Schizopodon,”
‘Zoologischer Anz.,’ Bd. xvii, p. 437.

EXPLANATION OF PLATES 16—21.

Illustrating Miss M. Robinson’s paper “On the Development
of Nebalia.”

REFERENCE LETTERS.

ant. Antenna. a.g. Antennary gland. a.p. Abdominal papilla. 4d.
Blastoderm. 4p. Blastopore. cué. Cuticle. eg. Caudal groove. .c.
Crystalline cone cells. c.f. Caudal thickening. e¢.m. Central mass of
nervous tissue. ad. Depression in optic thickenings. div. Dividing cells.
de. Deuterocerebrum. ec. Ectoderm, ex. Endoderm. f. Fold between
shell thickening and body wall. g.c. Genital cells. gy. Ganglion. hac.
Heart cells. hep. Hepatic lobes. 4,f. Head flap. /.c. Large cells of the
ectoderm. Jab. Labrum. m. Mouth. md. Mandible. ma. Maxilla. mes.
Mesoderm. mus. Muscle. x,f. Nerve fibres. o0.¢. Optic thickening. op.iz.
Optic invagination. 0.7. Optic lobe. pe. Protocerebrum. p.n. “ Pale
nuclei” of the optic invagination. proc. Proctodeum. s. Shell. sé.
Stomodeum. s.v. Shell valve. ¢e. Tritocerebrum. vp. Vitellophag. wed.
Wedge-shaped cells. y. Yolk. y.s. Yolk sac.

Exigencies of space have forbidden the drawing of more than the ventral
half of each section in most cases.

The magnification of the sections is 475, except where otherwise stated.

430 MARGARET ROBINSON.

PLATE 16.
Fic. A.—Stage A. Young embryo showing the blastoderm, which does
not as yet completely surround the yolk.

Fie. B’.—Stage B’. Late gastrula stage in which the blastopore is closed,
and the optic and antennal thickenings are definitely apparent.

Fic. C.—Stage C. Embryo with nauplius appendages.
Fic. D.—Stage D. Embryo with zoxa appendages.

Fic. E.—Stage E. Embryo in which the appendages more or less follow
the outline of the body.

Fic. F. Stage F. Earliest stage in which the embryo is free from the
shell. On account of its bifid appendages this might be called the Mysis
stage.

Fic. 1.—Longitudinal section through the ovum, younger than Stage A,
showing cells within the yolk as well as on the ventral surface. X 238.

Fie. 2.—Longitudinal section through Stage A, showing the yolk as yet
incompletely surrounded by the blastoderm. X 238.

Fic. 3.—Transverse section through the lateral thickenings in Stage B.
div. Dividing cells.

Fie. 4.—Transverse section through the lateral thickenings (behind Fig. 3)
in Stage B.

Fic. 5.—Transverse section through the anterior region of the caudal
thickening (front end of the groove) in Stage B. x 238.

Fie. 6.—Transverse section through the posterior thickening at the middle
of the groove in Stage B. x 238.

Fic. 7.—Transverse section through hind end of the caudal thickening in
Stage B (behind the groove). x 238.

PLATE 17.

Fic. 8.—Transverse section through the optic thickenings in Stage B’.

Fic. 9.—Transverse section through Stage B’, just behind the optic
thickenings.

Fic. 10.—Transverse section through Stage B’, taken midway between the
optic thickenings and the caudal thickening.

Figs. 11—15.—Transverse sections through the caudal thickening in
Stage B’.

Fie. 16.—Transverse section through the anterior part of the optic seg-
ment in Stage C. of. Optic thickening.

ON THE DEVELOPMENT OF NEBALIA. 431

Fic. 17.—Transverse section through the anterior part of the optic seg-
ment in Stage C, behind Fig. 16. op.cz. Optic invagination.

Fic. 18.—Transverse section through the optic segment in the region of
the ganglion in Stage C.

Fic. 19.—Transverse section through the posterior end of the optic gan-
glion in Stage C, at its junction with the ganglion of the first antenna.

Fic. 20.—Transverse section through the first antenne in Stage C.

PLATE 18.

Fic. 21.—Transverse section through the second antenna in Stage C.

Fic. 22.—Transverse section through the mandibular segment in Stage C,
showing ganglion.

Fic. 23.—Longitudinal section through Stage C, showing the first four
appendages and the first three ganglia.

Fig. 24.—Transverse section through Stage C, just in front of the mouth,
showing the labrum.

Fic. 25.—Transverse section through the mouth of Stage C.

Fic. 26.—Longitudinal (nearly median) section through Stage C.

Fic. 27.—Section through vitellophag in Stage C.

Fic. 28.—Transverse section through the anterior region of the optic
thickenings in Stage D, in front of the optic ganglia. d. Depression.

Fic. 29.--Transverse section through the optic thickenings in Stage Ds
going through the most anterior region of the optic ganglia.

Fic. 30.—Transverse section through the optic thickenings in Stage D,
behind Fig. 29, showing the optic invagination and the ganglion.

Fic. 31.—Transverse section through the posterior region of the optic
thickenings in Stage D. c.m. Central mass.

Fries. 32—34.—Transverse sections through the hindmost part of the optic
segment in Stage D.

Fie. 35.—Transverse section through the most anterior part of segment II
in Stage D, showing the ganglia of the antennules and the central mass of
nerve tissue lying between them.

PIGATE, 19:
Fic. 36.—Transverse section through the middle region of segment II in
Stage D.
Fic. 37.—Transverse section through the hindmost region of segment I1
n Stage D.
VOL. 50, PART 3,—NEW SERIES. 3]

432 MARGARET ROBINSON.

Fic. 38.—Transverse section through the anterior region of segment ILL
in Stage D.

Fic, 39.—Transverse section through the middle region of segment III in
Stage D. ;

Fig. 40.—Transverse section through segment LV in Stage D.

Fie, 41.—Transverse section through the ventral part of segment V in
Stage E.

Fies. 42, 43.—Transverse sections through hindmost end of the embryo in
Stage D.

Fics. 44—46.—Transverse sections through the optic segment in Stage E.
Fig. 44 is the most anterior.
Fie. 47.—Transverse section through the anterior part of the deutero-

cerebrum in Stage EK. The posterior ends of the optic lobes are cut through
in this section.

PLATE 20.

Fie. 48.—Longitudinal section through the brain in Stage E.

Fie, 49.—Transverse section through the posterior region of the deutero-
cerebrum in Stage H, showing the two ganglia of the antenuules with the
stomodeum between them.

Fic. 50.—Transverse section through the ventral part of the segment IL
in Stage EK, showing the antennules, the antennulary ganglia (posterior part
of the deuterocerebrum) and the stomodeum.

Fic. 51.—Transverse section through the ventral part of segment III in
Stage E, showing the antennary ganglion, the shell thickenings, labrum, and
mouth.

Fie. 52.—Transverse section through the ventral part of segment 1V
(mandibular segment) in Stage KE, showing mandibular ganglion with its
cleft, labrum, ete.

Figs. 53, 54.—Transverse sections through the antennary segment (behind
that shown in Fig. 52) in Stage E, showing the thickening of the dorsal
ectoderm and mesoderm cells (future cardiac cells) lying just below it.
x 238.

Fig. 55.—Transverse section through the ventral part of the antennary
segment in Stage E, showing the first trace of the antennary gland.

Fic. 55, a, 6, c.—Three consecutive transverse sections through the
dorsal side of a stage a little older than Stage F, showing mesoderm cells
turning inwards to form the heart.

Go

ON THE DEVELOPMENT OF NEBALIA. 43

PLATE 21.

Fic. 56.—Transverse section through segments I and IL in Stage F,
showing shell flap, optic and antennulary ganglia, central mass of nervous
tissue and labrum. X 238.

Fic. 57.—Transverse section just behind that shown in Fig. 56, showing
divisions of the central mass of nervous tissue, the stomodeum, and the first
appearance of jointing in the antennules. x 238.

Fie. 58.—Transverse section through Stage F, behind that shown in
Fig.57. x 238.

Fic. 59.—Transverse section through Stage F, behind that shown in
Fig. 58. x 238.

Fic. 60.—Transverse section through Stage F, behind that shown in
Fig. 59. This figure shows the cuticle. x 238.

Fic. 61.—Transverse section through stomodeal region in Stage F.
xX 238.

Fic. 62.—Transverse section through ventral side of Stage F, between
segments VI and VII, showing rudimentary muscles. x 238.

Fies, 63, 64.—Transverse sections through the first thoracic segment in
Stage F, showing the ventro-lateral out-pushings of the mesenteron, genital
cells, and shell thickenings. Fig. 63 shows the first thoracic ganglion at its
largest part. x 238.

Ne

DEVELOPMENT OF FLUSTRELLA HISPIDA. 435

On the Early Stages in the Development of
Flustrella hispida (Fabricius), and on the
Existence of a “Yolk Nucleus” in the Egg
of this Form.

By

R. M. Pace (née Clark),
Late Scholar of Girton College, Cambridge.

With Plates 22—25.

ContTENTSs.
PAGE

Introduction . : : ; 5 . 435
Material and Methods : : ‘ . 436
Spermatogenesis and Oogenesis : : . 441
The Maturation of the Egg: the Yolk Nasleus : . 442
The Nature and Function of the Yolk Nucleus . . 446

The Yolk Nucleus and Yolk Formation in Bryozoan
Eggs : 449

Fertilisation, and the Barre of the Ege fifo the Tentacle
Sheath . : a bail
Segmentation, and the Tarn of Te Geena Thess - 451
Summary : : 3 . 458
Comparison =k other Sioa : : . 459
The Development of the Larval Organs : : . 461
The Mature Larva. : : ; : . 464
The Degenerating Larva : : : . 467
The Alimentary Canal in Larval Beesneon : : - 468
General Summary : : : : . 469
References. ; : ; : . 470
Explanation of Plates : - : : - 472

INTRODUCTION.

In the following paper an attempt has been made to trace
out the earlier stages in the development of Flustrella
hispida. The larva of this familiar Bryozoan has previously

436 R. M. PACE.

been studied by Hincks (18), Redfern (22), Joliet (14),
Barrois (2), and Prouho (20), but the first four authors
studied only the living larva, of which Barrois has given an
excellent account with numerous figures, while Prouho, who
has described the later stages of the larval history in detail,
has paid but little attention to its earlier stages.

The research was undertaken at the suggestion of Dr.S. F.
Harmer, with the view to determine whether that structure
overlying the internal sac in the mature larva of Flustrella
is to be regarded as a stomach comparable to that which he
had described (11) as present in Alcyonidium. As the
work proceeded it has seemed expedient somewhat to extend
its scope, and to follow out the history of the egg from its
first appearance ; and the presence of a “ yolk-nucleus” being
detected, this structure has also been studied in some detail.

The work, which has been conducted partly at Cambridge,
partly at Plymouth, and partly at Brighton, has been greatly
assisted by a grant from the Government Grant Committee of
the Royal Society, to whom my best thanks are due. I would
also take this opportunity to express my thanks to Dr.
Harmer for his kindly interest and criticism, and for the loan
of some of his own preparations of later larval stages, and for
permission to reproduce one of his drawings (PI. 25, fig. 65 a),
to Dr. E. J. Allen for granting me the use of a table at the
Plymouth Laboratory of the Marine Biological Association, to
the authorities of Newnham College, Cambridge, for per-
mission to work at the Balfour Laboratory, and to Prof. J.
Graham Kerr, Dr. EK. G. Gardiner, and Mr. W. Wallace, for

advice on technical points.

MateriAL AND Mernops.

Collecting Material.—The material on which this paper
is based was collected at the following places and dates:
Swanage, March, 1902; Totland Bay, Isle of Wight, April,
1903; Plymouth, February to April, 1903, February to July,
1904; Brighton, May to July, 1903.

DEVELOPMENT OF FLUSTRELLA HISPIDA. 437

At each of the above places Flustrella hispida grows
abundantly between tide marks on Fucus, and occasionally also
on other Algze. The colonies form characteristic, dark, mossy-
looking patches encrusting the algal fronds: the Fucus,
when growing near low water mark, is often almost entirely
covered with this Bryozoan, but nearer high water mark the
Flustrella is not nearly so abundant, nor is it so well deve-
loped. Young colonies occur mainly on Fucus of the same
season’s growth.

For the study of the larval development, colonies of one or
two seasons’ growth taken from close to low water mark have
proved the most suitable. Such colonies contain abundance
of spermatozoa or of ova and larvee, according to the season.
Older colonies contain larger masses of dead zocecia, larvee
being relatively less abundant, while these latter are wanting in
such colonies as are presumably of the current year’s growth.
Again, in colonies taken near high water mark, larve and
egos are comparatively scarce, thus suggesting that the
conditions of life are not so favourable as at a lower tide
level, possibly because the colonies are uncovered by the
water for a considerable part of the day. Flustrella
colonies containing ripe reproductive elements or larvae may
be recognised by the presence of numerous dark-brown
blotches.

The reproductive period commences early in February and
continues until the beginning of August. Generally speaking,
it appears that spermatozoa are abundant in February and
March, and that they are not in evidence after the latter
month. Young ova are scarce early in February and
increase in number in March; during this latter month
seomenting eggs and young larve are also abundant and
some mature larve are present. Sections of the younger
portion of a colony taken in March have shown the
presence of segmenting eggs and larve in the older
zocecia, and of spermatozoa, together with a few young
ova, in the younger zocecia near the apex of the colony
(PI. 22, fig. 1), The maximum development of young embryos

438 R. M. PAOE.

is from April to June, while in July and August larve in
advanced stages are still abundant, but at this time young
stages and ova are rare. On the other hand, even in March
and April, colonies may be found which contain only mature
larve. It should be stated that in any one colony the
majority of the larve are at approximately the same stage of
development.

Flustrella hispida will readily live in standing water
for from one to five days according to the time of year, while
in running water it has been kept in good condition for over
a week,! but it is best whenever practicable to work with quite
fresh material, as the larve appear less healthy in colonies
which have been kept even a few days.

Methods.—The work has been done partly by the study
of entire larve and partly by means of sections. The larve
were examined both in the living state and after fixation.
For the latter purpose large numbers of eggs and larve were
removed from the colonies and preserved and in addition
portions of colonies were fixed entire in order to study the
larvee in situ.

Fixing Reagents.—The preservatives used were :—

(1) Cold saturated solution of corrosive sublimate, with the
addition of 5 per cent. glacial acetic acid.

(2) 100 parts 5 per cent. chromic acid, with five drops of
glacial acetic acid.

(3) Flemming’s solution.

(4) Hermann’s solution.

(5) Dr. Allen’s chromo-nitro-osmic acid mixture.

(6) Acetic alcohol containing sublimate to saturation.

(7) Kleinenberg’s solution.

Preserving Larve in situ.—For preserving larve in
situ it has been found best to cut the colonies with the
seaweed on which they are growing into small portions, and
to immerse these in the fixing solution for some time to allow

1 At Millport, owing to the exceptional purity of the water, there appears

to be no difficulty in keeping Flustrella alive in the tanks for a quite in-
definite period —R. M. P.

DEVELOPMENT OF FLUSTRELLA HISPIDA. 439

complete penetration. As soon as possible after fixation the
colonies were removed from the Fucus and after washing
transferred to 70 per cent. alcohol. Chromo-acetic acid and
corrosive acetic have given the best results when dealing
with material fixed in bulk.

Isolated Larve.—The removal of larvee from the colonies
is best effected by slicing off the front wall of the colony with
a sharp razor; the larve le immediately below this wall,
enclosed in the tentacle-sheath of the polypides, and they can
then be readily removed by means of a scalpel. Before
attempting to preserve the larve a considerable amount of
washing is necessary in order to free them from a mucus-like
substance in which they lie imbedded.

The best fixing reagents for the isolated larvae appear to
be corrosive acetic, and acetic alcohol saturated with sub-
limate ; chromo-acetic acid sometimes gives good results; and
the fixing reagents containing osmic acid have proved useful,
especially in the study of entire eggs and larve before
clearing. Material preserved in chromo-acetic requires
very prolonged washing and frequently proves difficult to
stain.

Entire Eggs and Larve.—The external characters of
isolated eggs and larve of all stages have been studied
during life. After fixation the larve were again examined
unstained in 70 per cent. alcohol, and were then stained either
in borax carmine followed by acid alcohol, or in safranin,
and re-examined. After clearing with cedar-wood oil or clove
oil—both of which reagents gave good results—the larve
were either mounted entire in Canada balsam or imbedded in
paraffin for sectioning. Staining the larva with borax carmine
after acetic alcohol and corrosive sublimate or corrosive
acetic brings out the nuclear spindles and also the yolk
nucleus very clearly, and so greatly assists in the interpreta-
tion of the external appearance of a segmenting egg. In a
few cases the embryo was removed from Canada balsam
after having been examined and drawn, and was imbedded
in paraffin for sectioning.

44.0 R. M. PACE.

Preparation of Sections.—Sections were made both
of the isolated larvee and of portions of colonies containing
larve. The watch-glass method of imbedding was found the
most convenient, especially when dealing with isolated larvae.
Groups of from twenty to thirty isolated larvee were imbedded
en masse and sectioned, sections thus being obtained in a
variety of planes. The larva at nearly all stages has a definite
axis, which renders it possible to orientate it and so to obtain
sections in any desired plane. To serve as a guide in deter-
mining in which direction unorientated larve had been sec-
tioned, a set of standard sections was prepared by carefully
orientating single larve which had been first studied entire.

Finally, portions of colonies were imbedded and cut with
the larvee in situ. To insure thorough impregnation, it was
found best to soak the material in xylol for about a week,
then to leave it in a mixture of xylol and paraffin for about
six hours in a warm place, such as the tray of the water-bath,
and finally to transfer to pure paraffin for about an hour. In
cutting such material great difficulty has been experienced
owing to the fact that the larvee he close under the front wall
of the colony: this wall, bemg beset with chitinous spines,
renders it difficult to imbed and to cut in such a manner that
the razor encounters none of the spines when passing through
the larve, since the chitin is sufficiently hard to notch the
razor, thereby of course causing the section to tear. This
difficulty is less marked in the case of transverse than of
longitudinal sections.

Staining.—The most useful stain for sections appears to
be Heidenhain’s iron hematoxylin, followed by eosin dissolved
in 90 per cent. alcohol. By this treatment the structure of
the yolk nucleus and of all nuclear bodies is brought out very
clearly. Borax carmine and safranin have given good results,
and double staining with methyl blue and eosin has also been
found useful. Mayer’s alcoholic cochineal, picronigrosin,
hematoxylin with a few drops of Kleimenberg’s solution,
hematoxylin and methyl orange have also been utilised;
Mayer’s mucicarmine was used for the detection of mucus.

DEVELOPMENT OF FLUSTRELLA HISPIDA. 441

Spermatogenesis and Oogenesis.

In Flustrella hispida the zocecium is hermaphrodite,
but the spermatozoa are chiefly developed earlier in the year
than the ova. In February and early in March, however,
ova and spermatozoa are found to occur simultaneously in the
zocecium ; the spermatozoa are in such cases fully developed,
while the ova are immature. PI. 22, fig. 1, shows a section
through a very young colony taken early in March; ova are
seen to be present in one zocecium, and in the anterior portion
of the same zocecium spermatozoa also occur.

Early in February the colonies assume a very puffed and
spotted appearance, large dark-brown patches becoming
visible. On cutting a section of such a colony these brown
patches are found to be due to the presence of an immense
number of spermatozoa, which can be removed in the same
way as the ova by slicing off the front-wall of the colony.
As has already been described by Calvet (8), the spermatozoa
are developed from the mesenchyme lining the lateral walls
of the zocecium, and the mother cells lie in masses close to the
front wall in the region of the tentacle sheath (Pl. 22, fig. 1, 71).
When ripe the spermatozoa have the typical flagellate form.
Frequently masses of spermatozoa are seen to be lying
with their heads imbedded in a central mass of protoplasm,
and with their tails vibrating at the periphery. No attempt
has been made at present to work out the details of spermato-
genesis. The spermatozoa decrease in number towards the
middle of March, and they are not in evidence after the end
of that month.

The ovary les at a point to the rear of and at a lower level
than that at which the spermatozoa are developed. It is
situated on a funicle passing from the mesenchymatous lining
of the lateral zocecial wall to the intestine (Pl. 22, fig. 1, Ov.).
At first the young ovary shows no indication of cell-walls, but
consists merely of a protoplasmic mass containing numerous
large nuclei (PI. 22, figs. 1 and 4, N.). Cell-walls subsequently
arise in this protoplasmic mass, four or five of the ovarian

442 R. M. PACE.

cells being differentiated in this manner (PI. 22, fig. 5) and
developing into ova, while the remainder give rise to the
follicle cells. As maturation proceeds, the follicle cells in-
crease in number and appear to grow in among the primitive
ova, so that when these latter are ripe each ovum is surrounded
by a follicular membrane (Pl. 22, figs. 2 and 3, Fo.).

Generally speaking, all the ova contained in the ovary are
of about the same age (PI. 22, figs. 2-5).

Tor MATURATION OF THE Hac: THE YOLK NUCLEUS.

The chief point of interest in the process of the maturation
of the egg of Flustrella is the appearance of a “yolk
nucleus,” apparently homologous with that type described
by van Bambeke (1) as occurring in the egg of Pholcus
phalangioides. The existence of a yolk nucleus does not
appear to have been hitherto recorded in any of the Kcto-
proctous Bryoza, although a similar structure has been figured
by Braem (4 and 6) and by Kraepelin (17) as being present
in the egg of Plumatella among the Entoprocta.

The Yolk Nucleus.—The history of this body in the egg
of Flustrella hispida is briefly as follows :

In very young ovaries in which the ovarian cells are only
just recognisable, there are, in addition to the germinal
vesicle, certain darkly-staining granules surrounding the
nucleus and lying in close contact with the latter (Pl. 22, figs.
4-5; Pl. 28, fig. 29, Y.N.). These granules at a later period
coalesce to form the structure, which, following van Bambeke,
may best be termed the “yolk nucleus.” They originate, as
has been said, quite close to the germinal vesicle, and their
appearance is so very similar to that of certain intra-nuclear
elements as to suggest that they have originated from the
nucleus. In fact, in one case (PI. 22, fig. 5; Pl. 23, fig. 29)
these granules seemed to be in process of actually passing out
from the germinal vesicle. The granules at this period are
homogeneous in appearance, and their behaviour with any of

DEVELOPMENT OF FLUSTRELLA HISPIDA. 443

the staining reagents employed exactly resembles that of the
chromatin granules of the nucleus.

Shortly after this stage these extra-nuclear granules
coalesce, and eventually come to form a crescentic body—
the yolk nucleus—which (PI. 22, figs. 8, 11, 18) becomes sur-
rounded by what appears to be a clear space ; and at an earlier
stage a similar clear zone frequently also occurs around indi-
vidual granules, or groups of granules, prior to their complete
coalescence (PI. 22, figs. 6-7, a). This clear area may abut
directly on to the germinal vesicle, or it may be separated
from it by a thin layer of protoplasm (PI. 22, figs. 11, 13, 14).
At about this time also, vacuoles begin to appear in the body
of the yolk nucleus (Pl. 22, figs. 11 and 13).

The appearance of the yolk nucleus as seen in sections
depends, of course, upon the point through which the section
is taken (PI. 22, figs. 9-12). In figs. 9 and 10 the position of the
yolk nucleus is marked only by the clear area which usually
surrounds it; in fig. 11 the yolk nucleus shows a crescentic
cross section, and it is seen to contain numerous vacuoles ;
while in PI. 22, fig. 12, it has the appearance of a cap overlying
the nucleus.

The yolk nucleus gradually passes from a crescentic to a
hemispherical form (Pl. 22, fig. 14) ; and its growth being
proportionately more rapid than that of the egg as a whole,
this hemispherical form becomes still more marked in later
stages, so that in certain sections the yolk nucleus may even
appear as a complete ring encircling the nucleus (Pl. 22,
fig. 15). The vacuoles increase in number, and frequently con-
tain crystalloid bodies at this stage (Pl. 22, figs. 13, 14, cr.).

The yolk nucleus next loses its originally homogeneous
appearance and shows signs of degeneration. This is evidenced
by the appearance of a peculiar reticulate structure, the sub-
stance between the meshes staining less deeply than the net-
work (Pl. 22, figs. 15, 16). The yolk nucleus then loses its
regular outline (Pl. 22, figs. 16, 17), and it finally breaks up
into more or less finely-divided, darkly-staining fragments,
surrounded each by a clear zone. The process of fragmenta-

444, R. M. PACE.

tion of the yolk nucleus appears to vary somewhat in its
details in different eggs (Pl. 22, figs. 18-21). The fragments
of the yolk nucleus retreat towards the periphery of the egg,
and there for a time form a disconnected ring of darkly-stain-
ing substance, each segment of which is still surrounded by
a clear zone (PI. 22, figs. 22, 23). Soon after this the fragments
of the yolk nucleus lose their identity, and at about the same
time the first indications of the true yolk make their appear-
ance as small granules, which are scattered about in the cyto-
plasm (Pl. 23, fig. 24, Y.). At the same time vacuolisation of
the cytoplasm occurs; the yolk spherules come to lie within
these vacuoles and rapidly increase in size until they appear
to occupy the entire egg (PI. 23, fig. 25).

The whole process of the growth and disintegration of the
yolk nucleus and the formation of the yolk takes place some
time before the egg is released from the ovary.

That the yolk nucleus is a true cell organ, and not merely
an appearance due to the coagulation of proteid by the fixing
reagent used, is proved by the fact that it is visible in the
living egg as a dark mass surrounding or overlying the
nucleus.

The nature and origin of the clear zone or space which has
been described as surrounding the yolk nucleus are somewhat
doubtful. It may possibly contain a fluid, but all attempts to
prove by means of staining reagents that this is so have
failed; and it would seem perhaps more probable that this
apparent space is only an artificial one caused by the con-
traction of the substance of the yolk nucleus during fixation.
This latter view is supported by the fact that the clear zone
has not been detected in the living egg, nor in those eggs
(Pl. 23, figs. 26,27) which have been fixed by chromo-nitro-
osmic acid, and also by the fact that the shape of the sup-
posed space corresponds so exactly with that of the yolk
nucleus which it surrounds.

With the view to determine whether the formation of oil
globules, which has been described and figured by von Bam-
beke (1) as preceding yolk formation in the case of Pholeus

DEVELOPMENT OF FLUSTRELLA HISPIDA. 445

phalangioides, occurs also in the egg of Flustrella, a
number of eggs were preserved with Dr. Allen’s chromo-
nitro-osmic mixture; these were sectioned without previous
bleaching, so that any fatty matter might remain intact, but
the results obtained have so far proved somewhat difficult of
interpretation. Even in very young eggs in which the yolk
nucleus is still in quite the initial stages of development,
large drops of a fatty substance are found to be present in
the region of the developing yolk nucleus; and as the egg
enlarges, the number of the fatty drops also increases, but
the latter always remain in the immediate neighbourhood of
the yolk nucleus, while similar globules are also visible within
that body itself (Pl. 23, figs. 26, 27). The amount of fatty
material increases as degeneration proceeds. It collects
especially towards the periphery of the egg, and at the same
time the development of the yolk commences. The two sub-
stances increase in quantity side by side, so that the mature
egg has the appearance of a mass of yolk spherules inter-
spersed with fat globules (Pl. 23, fig. 28). This condition,
the presence of fat and yolk side by side, continues so long
as any yolk is discernible in the larva—that is until after the
degeneration which precedes metamorphosis has commenced.
At no period are the fat globules arranged in any definite
relation to the nucleus. It is hoped that it may be possible
to elucidate this question of the relation of the fat globules
and the food yolk to one another and to the yolk nucleus, by
investigating in greater detail the history of the yolk nucleus
in the eggs of such other Bryozoa in which it may prove to
be present.

Prolonged treatment with xylol will cause the fat globules
to disappear from the yolk.

The Germinal Vesicle.—To determine whether there is
any connection between the yolk nucleus and the germinal
vesicle, the latter body has also been carefully studied. Pl. 23,
fig. 29, shows a young germinal vesicle to which particles of
the yolk nucleus are in close approximation, and it will be
seen that one of these latter appears to be actually in process

4.46 R. M. PACE.

of passing out through the nuclear membrane. The chromatin
network is somewhat dense, and it has at its nodes deeply
staining granules, which are similar in their appearance to
the original elements of the yolk nucleus, while a faintly
staining substance occupies its interstices.

The germinal vesicle at first grows relatively more rapidly
than the egg as a whole, and the chromatin network becomes
more attenuated (Pl. 23, figs. 80-32) ; but this latter fact is
probably due rather to the increased size of the nucleus than
to any emission of chromatin from it. No further changes
take place in the germinal vesicle until after the formation of
yolk has been completed, although the irregular contour of
the nuclear membrane observed in certain sections (PI. 23,
figs. 24, 25) during the period of yolk formation may possibly
denote amoeboid movements in connection with the latter
process: such amoeboid movements have been described by
Bambeke (1) in the case of the egg of Pholcus phalan-
gioides.

After the completion of yolk formation the chromatin net-
work begins to thicken (Pl. 23, fig. 33), and the substance
between its meshes now stains more deeply (Pl. 23, fig. 34).
The nucleolus also becomes relatively very large, and at the
same time the nuclear membrane loses its regular outline
(P]. 23, figs. 34, 35). These processes continue until all trace
of the chromatin network has disappeared, and the nucleus
stains uniformly throughout (PI. 23, fig. 86). At this stage,
which is immediately prior to that of the formation of the polar
bodies, the nucleus begins to decrease in size relatively to
the rest of the egg and becomes amceboid: the nucleolus is
still present,

The Nature and Function of the Yolk Nucleus.—
The term “yolk nucleus” has been applied by various
authors to bodies which appear to be totally different in
their origin, development, and appearance, and which
would seem to have only this much in common, that all
have been regarded as being in some way connected with
the phenomenon of yolk formation. In the present instance,

DEVELOPMENT OF FLUSTRELLA HISPTDA. 44.7

the body described approximates very closely to that type
of yolk nucleus described by Bambeke (1), Crampton (9),
Wallace (26), and Calkins (7) for the eggs of Pholcus
phalangioides, Molgula manhattensis, Zoarces
viviparus, and Jiumbricus respectively.

It is not proposed in the present paper to enter into a
detailed consideration of the former work bearing on this
subject. A very complete bibliography of the yolk nucleus
is furnished by the papers of Jordan (15), Mertens (18),
Henneguy (12), Calkins (7), Wilson (27), Bambeke (1), and
Crampton (9).

As has already been stated, the yolk nucleus of the ege of
Flustrella hispida is closely comparable with that type
described by Bambeke, Calkins, Crampton, and Wallace, but
on comparison with these apparently closely related bodies
certain points of difference may be noted.

Bambeke (1) describes in the egg of Pholcus phalan-
gioides a type of yolk nucleus which corresponds closely
in its appearance and in its mode of growth and degeneration
with that occurring in Flustrella hispida. He recognises
four stages in the history of the yolk nucleus and nutritive
yolk of the egg of Pholeus: (a) the appearance of small,
darkly staining granules, which he believes to be of nuclear
origin, and which coalesce to form a crescentic structure
containing vacuoles, with included crystalloid bodies, and sur-
rounded by a clear zone; (b) the degeneration of the yolk
nucleus; (c) the appearance of oil drops; (d) the formation of
the true yolk. The most important difference between Bam-
beke’s account of the yolk nucleus and the present one is that
in the case of the ege of Pholcus the nucleus appears to
take an active part in the process of yolk formation, and that
the yolk nucleus itself gives rise to the vitellus by first under-
going a metamorphosis leading to the formation of oil
elobules, these latter becoming resorbed by the protoplasm
from which the true yolk is then elaborated. Now, as has
already been pointed out, the presence of oil globules in the
case of the egg of Flustrella hispida bears but little

VOL, 50, PART 3.—NEW SERIES. 32

448 R. M. PACE.

obvious relation to the process of yolk formation. In this
form oil globules are present in very young eggs, both in the
yolk nucleus itself and also scattered throughout the sur-
rounding protoplasm ; and, although they are certainly present
in increasing numbers as the yolk develops, they do not
disappear with the completion of yolk formation, but persist
throughout larval life. So far, also, in Flustrella it has not
been possible to confirm Bambeke’s views as to the importance
of the nucleus as a factor in yolk formation. The irregular
shape of the nuclear membrane, which is sometimes observed
in the early stages of yolk formation in the egg of Flustrella
(Pl. 23, fig. 24), may possibly indicate amceboid movements ;
but these have no apparent reference to the distribution of
the oil drops, which collect chiefly towards the periphery of
the egg rather than in the immediate neighbourhood of the
nucleus; and neither the oil globules nor the yolk spherules
show the radial arrangement which is described by Bambeke.

Crampton (9), in his paper on the early history of the
Ascidian egg, described a yolk nucleus of similar type in the
egy of Molgula manhattensis; and he attempted to
determine its chemical nature by differential staining. The
general account given by Crampton of the history of the yolk
nucleus in Molgula agrees with the above description of
this structure in Flustrella. Crampton, among other
stains, made use of Heidenhain’s iron-heematoxylin, and states
that this stain is taken up by the chromatin of the germinal
vesicle, but that it has no effect on the yolk nucleus ; he also
mentions the difficulty experienced in washing out this stain
from the-yolk after yolk formation, without at the same time
decolourising nuclear structures. In the case of the egg of
Flustrella, iron-hematoxylin stains both the yolk nucleus
and the chromatin elements of the nucleus with equal inten-
sity, and a similar difficulty with regard to the washing of
eges containing yolk has been experienced. PI. 23, fig. 25,
shows a section of an egg in which, even after prolonged
washing with iron-alum, the centres of many yolk spherules
remained darkly stained. Crampton’s researches led him to

DEVELOPMENT OF FLUSTRELLA HISPIDA. 4.4.9

the conclusion that the yolk nucleus, although of nuclear
origin, does not consist of chromatin, as has been maintained
by several writers, owing to its apparent origin from and simi-
larity of appearance to the chromatin granules of the germinal
vesicle; but that it is either purely albuminous or consists of
nucleo-albumin containing a large percentage of nucleic acid
constituents. According to this author, accounts of the
origin of the yolk from the cytoplasm at a point distant from
the nucleus, or from several centres, or of its formation all
over the egg, refer only to its later history, and do not take
into account an earlier stage, which is marked by the appear-
ance of the yolk nucleus. If Crampton’s view be the correct
one, a study of younger eges should in such cases lead to the
discovery of this supposed early stage.

Wallace (26) describes in the egg of Zoarces a yolk
nucleus of the type occurring in Flustrella; and by Mr.
Wallace’s courtesy, I have been enabled to examine many of
his preparations and drawings, which show that the yolk
nucleus agrees in almost all respects with that of Flustrella
hispida. Wallace, however, found that fixing reagents
containing nitric acid dissolved out the yolk nucleus, while,
as has already been stated, this is not the case with the egg
of Flustrella. Further, Wallace agrees with Bambeke
that the formation of oil-drops precedes true yolk formation.

The Yolk Nucleus and Yolk Formationin Bryozoan
EKggs.—As has been stated earlier in this paper, no yolk
nucleus has hitherto been noted in the eggs of any of the
Ketoprocta.| Among the Entoprocta, however, a structure,
which appears to be similar to the yolk nucleus occurring in the
ego of Flustrella hispida, has been figured by Kraepelin
and by Braem as present in the ovum of Plumatella.

Kraepelin (17) points out that shortly before the egg

1 It has lately been possible by the courtesy of Miss A. Heath to examine
some preparations of a species of Aleyonidium which contained young ova.
The material, which was collected at Millport in September, 1905, contained
abundance of young ova, and these were found to contain “yolk nuclei”
apparently similar to that which has been described in Flustrella.

450 R. M. PACE.

ripens, the germinal vesicle becomes more or less surrounded
by a differentiated mass of protoplasm, which persists until
after fertilisation, but subsequently vanishes.

Braem (4 and 6) mentions the presence of a light zone of
protoplasm surrounding the nucleus, which makes its appear-
ance in the ovum at an early stage. This zone is at first not
clearly defined from the outer layer of the protoplasm, but
later on it becomes sharply demarcated. The outer zone takes
no part in segmentation. Small, darkly-staining bodies next
arise in the outer zone of the protoplasm. ‘These bodies are
of varying sizes, and are each surrounded by a clear area.
Braem states that they are similar in appearance to the
nucleolus, and suggests that they resemble the latter in
chemical composition. He believes that they originate from
the outer zone of the cytoplasm, but similar bodies may also
occur in the inner zone. Braem suggests that these bodies
in the ovum of Plumatella may be homologous with the
yolk nucleus of certain other animal ova.

In the case of the eggs of Flustrella hispida, the facts
so far obtained all point to the conclusion that the yolk
nucleus is a true cell organ originating from the nucleus at
an early stage in the history of the ovum; that after under-
going a series of changes it finally disintegrates ; and that
the process of disintegration is in some manner intimately
connected with yolk formation. So far as can be seen, the
germinal vesicle plays no direct part in the formation of the
yolk. It has not yet been practicable sufficiently to investigate
the presence or otherwise of a yolk nucleus in the eggs of
other Bryozoa, and the results of this investigation must be
reserved for a future paper.

The Centrosphere.—No trace of a centrosphere has yet
been detected in the ovum of Flustrella hispida.

The Polar Bodies.—The formation of the polar bodies
had only been observed in one case. The egg (PI. 24, fig. 51)
had already passed into the tentacle sheath, and the first polar
body (P.B.) was lying on the surface of the ovum, while the
second was in process of formation.

~

DEVELOPMENT OF FLUSTRELLA HISPIDA. 45]

FERTILISATION, AND THE PassaGE OF THE EGG INTO THE
TENTACLE SHEATH.

Fertilisation.—The act of fertilisation has not yet been
observed, only such ova as were either preparing for, or had
already undergone, fertilisation having been obtained. It has
already been stated that the spermatozoa and ova ripen for
the most part at different times, and no ripe spermatozoa
have been observed in the zocecial cavity after March.

Passage of the Eggs into the Tentacle Sheath.—
The development of the ova takes place within the tentacle
sheath, and the eggs pass from the zocecial cavity into the
tentacle sheath in the interval between the degeneration of a
polypide and the formation of a new bud. ‘his bud attains
complete maturity, and not, as has been described by Joliet
(14) in the case of Valkeria cuscuta, only partial deve-
lopment. As has been mentioned, from four to five eggs
generally ripen at the same time, and these enter the tentacle
sheath together and develop side by side. As development
proceeds, the larvee, while still enclosed in the tentacle sheath,
increase in size, and gradually come to fill the entire cavity
at first occupied by the polypide, and the latter now ceases
to exist. Those zocecia occupied by advanced larvee contain
large quantities of a slimy, mucus-like substance, which
surrounds the developing embryos.

SEGMENTATION, AND THE FORMATION OF THE GERMINAL LAYERS.

The earlier stages of segmentation have been studied in
detail in the hope of elucidating the problem of the origin
and subsequent history of the mesoderm and of the endoderm
in this form. So far, however, the formation of mesoderm
has not been definitely traced, but it is hoped that it may
be possible to determine this point more satisfactorily at a
later date.

The Primitive Cleavages.—After fertilisation the egg
becomes separated from the vitellme membrane by a wide
space (Pl. 28, fig. 37). The first cleavage (Pl. 23, figs. 37-42)

452 Re VPAON:

divides the egg into two symmetrical halves, each containing
equal quantities of yolk: it is completed in about twenty
minutes. ‘This division is followed by another, in a plane at
right angles to the first, dividing the egg into four spheres,
which are to all appearance equal in all respects (PI. 25,
fig. 43; Pl. 24, fig. 52).

To simplify the following account of the cell-lineage of
Flustrella, these four first-formed cells have been dis-
tinguished in the figures by the letters A, B,C, and D. Cells
arising from these are denoted by the letter of the particular
cell from which they have been derived, with the addition of
a negative index to indicate the generation to which the cell
belongs, and of a positive index to denote the number of cells
in that generation at the moment of the formation of any
given cell, and the order of their formation: thus A# denotes
the fourth cell derived from A in the fifth generation.!

The 8-cell Stage.—The four cells A, B, C, and D, again
dividing in a plane at right angles to each of the former
divisions, give rise to eight cells: of these, the four lower
cells are larger, and contain more yolk than the four upper
ones (PI: 28, fig. 44).

The polar bodies are seen in the living egg to lie on the
surface of the smaller cells, and these smaller cells and their
derivatives always lie on the upper surface of the egg, which
later on becomes the dorsal or aboral surface of the larva.

The 12-cell Stage.—The four small upper cells next
divide (Pl. 23, figs. 45 a-b) each into two unequal cells by
a cleavage lying at an angle of 45° to the primitive cleavage
plane. ‘The ring of small cells which is thus formed becomes
shortly afterwards rearranged into two rows of four cells
each, so that the embryo now consists of two parallel series
of small cells overlying the four large yolk-laden ones.

Owing to this new arrangement of the embryonic cells, it is

1 In view of the use which has been made of a somewhat similar system of
notation by other authors, it should be stated that the symbols here used have
been adopted only for the sake of clearness, and that they have no reference
to any of the theories of cell genesis which have been put forward.

DEVELOPMENT OF FLUSTRELLA HISPIDA. 4538

now possible to distinguish between a longitudinal (long) and
a transverse (short) axis of the larva. At a later period,
however, cell division is found to occur more rapidly along
the transverse than along the longitudinal axis, and the larva
assumes a spherical shape; so that it is not therefore prac-
ticable to establish any direct correlation between the long
and short axes of these early stages and the long and short
axes of the mature larva.

The 16-cell Stage.—Each of the four large lower cells
next divides into two cells of unequal size. The plane of
cleavage is a vertical one lying more or less at right angles
to the long axis of the embryo, and the four central cells
formed are the larger; so that now the oral surface of the
larva also consists of two rows, each of four cells, which
immediately underlie those of the upper series. The larva
(Pl. 23, figs. 47a-c) is now, therefore, built up of sixteen cells
arranged in four parallel rows, which are disposed in two
tiers ; that is, it consists of an upper, aboral tier of two rows,
each of four cells, and of a lower, oral tier of larger cells,
also of two rows, each of four cells. All of these sixteen cells
belong to the fifth generation (the unsegmented ovum being
regarded as the first generation). The four central cells of
the oral series are much larger than the lateral ones of the
same series, and these again are larger than those of the
upper tier.

At this stage, the segmentation cavity (S.C.) becomes
visible, and it is noteworthy that it contains a substance
which stains feebly with eosin, safranin, borax-carmine, etc.

It may be mentioned that the arrangement of the cells in
two tiers of two parallel rows each of four cells is charac-
teristic of many Polyzoan larve at the 16-cell stage.

The 20-cell Stage.—The next division (Pl. 23, fig. 48)
also takes place in the four large central oral cells, which each
again divide into two unequal cells—A} and AZ, Bi
and B2, Ci and C2, and Di and D 2—cleavage taking
place in a vertical plane at right angles to the previous
division. The embryo (Pl. 238, figs. 49 a-c) now consists of

454 R. M. PACE.

twenty cells; that is to say, there is an aboral series of
eight small cells which are disposed in two parallel rows and
belong to the fifth generation ; below these a ring of eight
larger cells, four of which—A 3, B 3, C3, and D 3—are of
the fifth generation, and four others—A }, B}, C3}, and
D t—of the sixth generation; while the oral surface is
formed by four large central cells—A 2, B 2, C 2, and D 3—
which are also of the sixth generation, and which project
upwards into the segmentation cavity (Pl. 24, fig. 53). The
aboral series of cells contains, generally speaking, less yolk
than the lower series of cells, of which the before-mentioned
four large oral cells are particularly rich in yolk (Pl. 24,
figs. 53, etc.).

The 32-cell Stage——The beginning of the next stage
is shown in Pl. 24, figs.. 54a-b, and 55a-b.: ‘The
small cells composing the two rows of the upper tier first
divide horizontally, giving rise to four rows each of four cells,
and arranged in two tiers (Pl. 23, fig. 50b, and Pl. 24, figs.
54a-b, 55 a-b). This cleavage is followed almost imme-
diately by the division of the four cells—A ’, B#, C3, and
D 3—which were the first cells originally budded off by the
large oral cells. These four cells divide (Pl. 24, figs. 54},
and 55b) by a vertical cleavage lying at an angle of 45° to
the primitive cleavage of the segmenting ovum. ‘The embryo
(Pl. 23, figs. 50 a.-c., and Pl. 24, fig. 56) now, therefore, con-
sists of thirty-two cells, which are all of the sixth generation,
and which are arranged in the following manner. The aboral
surface of the larva is composed of sixteen small cells,
disposed in two tiers, each tier consisting of two parallel
rows of four cells each. Below these sixteen aboral cells is a
ring of twelve larger cells, overlying and partially surrounding
the four large oral cells, which still occupy the lower surface
of the larva. The upper halves of these four oral cells, being
surrounded by the ring of twelve intermediate cells, are thus
enclosed by them within the segmentation cavity, and the
latter is almost entirely obliterated at this stage (Pl. 24,
fig. 55 a).

DEVELOPMENT OF FLUSTRELLA HISPIDA. 455

The Formation of Endoderm.—P]. 24, fig. 57, illus-
trates a somewhat later stage than that just described. The
small aboral cells and the ring of larger cells underlying
these have divided, and the latter are now seen to enclose
about two thirds of the four large central oral cells within
the segmentation cavity. These four large oral cells have
also again divided, but this time the plane of cleavage
has been a horizontal one; and the four larger upper seg-
ments resulting from this division he within the segmentation
cavity, and represent the primitive endoderm. The four lower
segments retain their original oral position.

The EKectoderm.—From this time onwards, cell division
becomes less regular, and for a time at least, it takes place
more rapidly in the transverse than in the longitudinal direc-
tion, so that the larva tends to become spherical in form. ‘The
small aboral cells divide repeatedly, forming the aboral ecto-
derm ; while the ring of larger intermediate cells, which were
shown to have been initially derived from the large oral cells,
A ?, B%, C3, D 2, in like manner give rise to the oral ecto-
derm. A study of the living embryo and of sections (PI. 24,
fies. 53-60) shows that the cells of the aboral ectoderm tend to
remain, throughout embryonic development,smaller than those
of the oral ectoderm, the difference becoming more marked
as development proceeds, but it is not possible at this, or at
any later stage, to distinguish any definite ring of cells which
can be correlated with the ciliated ring of the mature larva.
The true origin of this structure will be dealt with later, but
it may not be out of place to say here a few words in order to
explain how it has come about that the existence of such a
ring of cells has been supposed by Barrois and other authors
to occur at this stage. It is true that observations made on
the entire egg at this period, especially when it 1s viewed
from the aboral surface, give somewhat the appearance of
there being an equatorial ring of cells, but this appearance
is a deceptive one. As stated above, an equatorial ring of
twelve cells actually did exist at the thirty-two-cell stage,
but the cells of this ring, as has already been described, have

456 R? MM. PAGE.

divided to form the oral ectoderm. The cells of the oral
ectoderm are all equal in size, though they are larger than
those of the aboral ectoderm. A study of sections shows that
the ring-like appearance seen at this stage, when viewing the
ege from the aboral surface, is due simply to this difference in
size between the cells of the oral and aboral surfaces causing
the former to project out beyond the latter (PI. 24, figs,
58-60). In later stages this appearance is enhanced by the
development of the aboral groove or mantle cavity just above
the line of junction of the two sets of ectoderm cells (Pl. 25,
fig. 61). Barrois (2) was misled by this deceptive appear-
ance, and, not having checked his observations on the living
ege by the examination of sections, published figures (2,
pl. xii, fig. 6), purporting to represent the larva at this stage,
showing a prominent equatorial ring of large cells, while the
remaining cells are represented as being of the same size both
above and below this supposed ring.

The Mesendoderm.—Owing to the rapid growth of the
oral ectoderm, the four central oral cells eventually become
surrounded and enclosed in the segmentation cavity, thus
forming, together with the four cells originally segmented
off, eight mesendoderm cells in all. These eight cells divide
repeatedly, and give rise to a solid mass of tissue, which in
time comes to fill the segmentation cavity. Pl. 24, figs. 57-60,
represent stages in this process. The asymmetrical position of
the mesendoderm cells in Pl. 24, fig. 58, seems somewhat
peculiar, and might well be considered to be abnormal, but
the arrangement shown in this figure has been met with in
all the sections examined, and must, therefore, be regarded
as quite normal.

The Blastopore and Gastrulation.—The oral ecto-
derm remains for some time open at the point of enclosure of
the mesendoderm cells, and this opening may be regarded as
the blastopore (Pl. 24, figs. 58, 59, Bl.). All trace of the
blastopore vanishes in later stages (Pl. 24, fig. 60). It is
impossible to say whether the enclosure of the second set of
mesendoderm cells be due to the pressure of the surrounding

DEVELOPMENT OF FLUSTRELLA HISPIDA. 457

cells—that is to say, whether it represents a process of in-
vagination, or whether the oral ectoderm simply grows over
and encloses the mesendoderm. In either case the process
may be regarded as a form of gastrulation, but no trace of an
archenteron has been observed.

Later Segmentation Stages.—At the stage represented
in Pl. 24, fig. 60, except that the oral surface is slightly
flattened, the embryo has become almost spherical in form.
Aborally, it is covered bya layer of small, flat, ectoderm cells,
while the oral surface is composed of large, high cells; and
from these latter a mass of endoderm or mesendoderm cells
projects into the segmentation cavity, which is still visible at
this stage. A shallow groove, bounded below by a slight
protuberance, is noticeable on the exterior of the larva rather
above the equatorial line. A study of sections shows that
this groove marks the junction of the oral and aboral
ectoderm, and that it is due to the difference in the size of
the cells in this region. It is to be again noted that there
are no specially enlarged cells, such as have been described
by Barrois, in this region.

Up to this stage in the development, the oral surface has
been relatively larger and more convex than the aboral, but
now cell division becomes more active on the latter surface,
so that it in turn becomes relatively larger and more convex
than the oral surface. Owing to the increased growth of the
aboral surface, the slight groove and the protuberance already
noticed now lie below the equatorial line, and in sections this
groove is now seen to be the result, partly of the difference
in size of the ectoderm cells at the line of junction of the oral
and aboral series, and partly of an actual bulging out of the
oral ectoderm (Pl. 25, fig. 61, M.C.). This encircling groove
represents the initial stage of the aboral groove or mantle
cavity.

The stage thus briefly described marks the close of the
segmentation period. The larva now passes gradually from
a spherical to an elongated form, which is typical of the
mature larva. When viewed in section (Pl. 25, fig. 61), the

458 R. M. PACE.

difference in size and character between the cells of the oral
and aboral ectoderm is now seen to be very marked, and the
mesendoderm forms a solid mass entirely obliterating the
segmentation cavity.

Summary.—To sum up the results obtained from the
foregoing study of the cell division :—

Four equal cells—A, B, C, D—result from the primitive
cleavages. These four cells divide and give rise to eight
cells, arranged in two series, four small upper cells, and four
large lower cells, each of which series is destined to play a
distinct part in the subsequent history of the cell layers.
The four small upper cells, Ai, B4, C1, D1, give rise only
to the aboral ectoderm. The four large lower cells, A3, B3,
C4, D%, give rise in part to the oral ectoderm, in part to a
tissue, which may for the present be best termed ‘‘ mesendo-
derm.” ‘he cells from which the oral ectoderm is derived
are cut off from the four large oral cells by two successive
vertical divisions at right angles to each other. The four
large oral cells then divide a third time horizontally, and
the four upper products of this cell division pass into the
segmentation cavity, and there give rise to the primitive
endoderm, or rather, mesendoderm. Owing to the rapid
growth of the surrounding ectoderm, the four remaining oral
cells also eventually become enclosed within the segmentation
cavity, giving in all eight mesendodermic cells, which, by
subsequent division, eventually form a solid mesendodermic
mass.

A true blastopore, which does not close until after the
formation of a considerable mass of mesendoderm, is present
during the stages referred to.

The formation of mesoderm has not been actually observed;
it seems probable, however, that the mesoderm is derived
from the primitive mesendodermic mass at a later stage than
those hitherto dealt with, and this point will be further dis-
cussed (p. 467) when dealing with the degenerating larva.

In view of the fact that the most essential of the larval
organs are developed from the oral ectoderm, it seems of

DEVELOPMENT OF FLUSTRELLA HISPIDA. 459

especial interest to note that this tissue is from the first
distinct from the aboral ectoderm; the difference in the
relative sizes of the oral and aboral ectoderm cells, which is
so marked in early stages, is noticeable throughout larval
life.

Comparison with other Bryozoa.—As yet, compara-
tively little has been written on the early development of the
Bryozoa; the most important papers on the subject being
those by Barrois, Joliet, Repiachoff, Vigelius, Harmer, Braem,
Prouho, and Calvet.

Barrois has published descriptions of the early stages of
several Bryozoan larve, and among others he dealt with (2)
that of Flustrella hispida. His observations, however,
were made solely upon entire eggs and larvee. His descrip-
tions and figures are in entire agreement up to the thirty-two-
cell stage with the general results described in the present
paper, but he does not make any special mention of the
lineage of these cells, and he was also unable to study the
formation of the endoderm. As has already been pointed
out, Barrois erroneously describes a later stage, in which
both dorsal and ventral ectoderm are said to be composed of
small equisized cells, the two series being separated by an
equatorial ring of large cells (2, pl. xu, fig. 6), and he figures
this ring as being present in all subsequent stages. Had
Barrois sectioned any of his material, instead of relying solely
on external appearances, he would have seen that no equatorial
ring of single cells, such as he described, is present at any of
the stages figured.

In another paper (8) Barrois describes the enclosure, in
Schizoporella unicornis, of four primitive endoderm cells
by epiboly. He also describes the formation of two bands of
mesoderm, which at a later period fuse with the endoderm
to form a single mesendodermic mass.

Repiachoff (23), in a paper on Tendra zostericola, also
describes the endoderm as originating by the enclosure and
division of four large oral cells. He, however, states that
the process is followed in Tendra by the formation of an

460 R. M. PACE.

archigastrula with an opening to the exterior: no such stage
occurs in Flustrella hispida.

Vigelius (25) studied the early stages in the development
of Bugula, and noted the presence of four large dorsal cells
within the segmentation cavity and the subsequent division
of these to form the endodermic mass.

Harmer (11), in his paper on Alcyonidium, mentions the
presence of a blastopore as occurring in that form.

Prouho, in a paper (20) on Flustrella hispida, makes
no mention of the early stages of this form. In a later paper
(21), however, he describes the formation of endoderm in
the Cyphonautes larva. He states that at the thirty-two-cell
stage the embryo is flattened along an axis perpendicular to
the plane of the first segmentation, the four oral cells being
larger than the other cells and especially rich in yolk; these
four large cells subsequently become enclosed by the rapid
growth of the ectoderm. In the case of Alcyonidium
albidum, Prouho records that the endodermic cells each
divide into two before becoming enclosed by the ectoderm,
and he defines the blastopore in this case as the point at
which the ectodermic cells close over the four large cells.

Braem, in his account (5) of the embryology of Paludi-
cella Ehrenbergi, points out that segmentation in the larva
of this species is total and almost equal. At the eight-cell
stage the segmentation cavity is visible, and the four upper
cells are somewhat larger than the lower cells. ‘The sixteen-
cell stage is similar to that which occurs in Flustrella
hispida. Atthe thirty-two-cell stage the embryo is spheri-
cal, and forms a typical blastula; the blastula has a large
segmentation cavity, and the cells of the vegetative pole are
larger than those of the animal pole and do not increase in
number as rapidly. The four central cells of the vegetative
pole then become surrounded by the ectoderm cells, and are
pressed into the segmentation cavity. After their enclosure,
these four cells se2ment to form other endoderm cells, which
multiply and give rise to a many-celled archenteron opening
to the exterior. A cell layer containing muscle fibres lies

DEVELOPMENT OF FLUSTRELLA HISPIDA. 461

between the ectoderm and endoderm, and Braem considers
that this tissue may represent the mesoderm, and that it is
possibly derived from the initial cells separated off from the
original four large endoderm cells.

Calvet (8), in his general account of the embryology of
Cheilostomes and Ctenostomes, makes the general statement
that segmentation is equal and regular up to the thirty-two-
cell stage ; but, from the above description of the process in
Flustrella hispida, this is obviously not invariably the
case. Calvet describes the formation of endoderm as taking
place in a manner similar to that in which it originates in
Flustrella, and he arrives at the conclusion that the
endoderm arises partly by endocytulation, partly by planu-
lation. .

The endoderm, therefore, appears to originate in a similar
manner in Flustrella and in the few other Bryozoa in
which its formation has so far been investigated.

THE DEVELOPMENT OF THE LARVAL ORGANS.

The formation of mesendoderm being completed, the first
traces of larval organs soon appear in the shape of a two-fold
invagination of the oral ectoderm, and this is followed soon
after by a third invagination at a point anterior to the
previous ones. Both ectoderm and mesendoderm cells have
by this time begun to lose their definite cell outlines.

Ectodermic Organs.—The internal sac is the first
organ to be formed. It arises as a median invagination
of the oral ectoderm ; and in PI. 25, fig. 62, it is seen as an
elliptical space (J.S.) communicating with the exterior by
means of anarrow opening. The cells lining this sac are short
and flat, and their nuclei lie close to the periphery. These
cells are seen to contain large globules of a substance closely
resembling the yolk spherules in their general appearance and
in their reactions to staining reagents; and drops of this sub-
stance are also found to be exuding from many cells (Pl. 25,
fig. 62 g.). This substance subsequently disappears entirely

462 R. M. PACE.

from the cells of the internal sac (Pl. 25, fig. 65). The signi-
ficance of its appearance will be discussed in a later section.

The second oral ectodermic invagination is destined to
form an organ representing the pharynx; it occurs in front
of that from which the internal sac arises, and appears at a
slightly later period. The cells bounding its opening are
large and high, their nuclei lie on the side nearest the open-
ing, and the cells themselves are much vacuolated. The
vacuoles (Pl. 25, figs. 62-64) in the walls of the pharynx are
filled with a substance similar in appearance to that which
has been described as occurring in the cells of the internal
sac (Pl. 25, fig. 62), and, as in the latter case, this substance
also entirely disappears at a later stage (Pl. 25, fig. 65).
From its mode of origin, the pharynx is to be regarded as a
true stomodeeum.

The pyriform organ arises, at a somewhat later period
than the internal sac and the pharynx, as an oral ecto-
dermic invagination anterior to the latter organs. In PI. 25,
fies. 62, 63a, the internal sac and the pharynx are shown
well developed, while at this stage the pyriform organ is
represented only by a slight invagination (Py.).

The aboral organ, the “calotte” of French authors,
is at the stage figured in Pl. 25, fig. 62, already visible as a
thickened mass (Ca.) of aboral ectoderm overlying the pharynx
and provided with numerous nuclei. From this organ a
delicate network of fibres and nuclei passes to the developing
pyriform organ (Pl. 25, figs. 62, 63 a).

Organs of Mesendodermal Origin.—The three-fold
ectodermic invagination leading to the formation of the
internal sac, pharynx, and pyriform organ has the effect of
compressing the mesendoderm into a solid mass, which lies
in the posterior part of the larva with its anterior end over-
lying the inner end of the pharynx (Pl. 25, fig. 62, Hd.).
This mesendodermal tract consists of a mass of yolk
spherules with scattered nuclei and it rapidly loses all trace
of definite cell structure. At a slightly later period, the
mesendodermic mass becomes hollowed out and forms what,

DEVELOPMENT OF FLUSTRELLA HISPIDA. 463

from its origin, appearance, and position, there can be no doubt
is a vestigial stomach (PI. 25, figs. 63 a-b, 64, 65 a-c, St.). This
supposed stomach is identical in its structure with that de-
scribed by Harmer (11, p. 446, Pl. XXVII, figs. 1, 2) as occur-
ring in the larva of Aleyonidium only, unlike the stomach in
that case, communication with the exterior by the pharynx is
never established. As is the case also in Alcyonidiun,
the stomach is lined simply by a protoplasmic mass in which
nuclei and numerous yolk spherules are embedded. The
lining epithelium shows no trace of any glandular character,
and the organ itself remains entirely vestigial and disappears
before free life commences.

In describing the development of the stomach in the
Cyphonautes larva, Prouho (21) says: “La depression orale
devient de plus en plus profonde, pendant que la région
aborale devient de plus en plus conique ... Les cellules
de la masse endodermique, qui se sont un peu multipliées
pendant que l’embryon subissait les modifications ci dessus,
se desposent autour dun axe et forment une masse, pleine,
allongée, oblique, dont une extrémité vient s’appuyer contre
le fond de Vinvagination orale; cette masse endodermique,
définitivement rejetée a l’arriére de la larve occupe d’ores et
déja la position de futur estomac.”’

The organ which has been regarded as the stomach in the
larva of Flustrella hispida has been shown by its origin
and position to correspond to the endodermic mass which
ultimately gives rise to the stomach in the Cyphonautes
larva, and it is also closely comparable to the rudimentary
stomach of the larva of Alcyonidium, except for the fact
that it never communicates with the pharynx. The exceed-
ingly sight development of the stomach in Flustrella is
easily explained when the short duration of free larval life,
and the correlated abundance of the supply of food yolk are
taken into consideration.

According to Prouho (20) the organ which has been here
termed the pharynxis to be regarded as a rudimentary intestine
much less differentiated than that of Aleyonidium. But, as

VOL 50, PART 3,—NEW SERIES. 39

464 R. M. PACE.

we have seen, this structure is a true stomodeum, and from
the above account it will be evident that the alimentary
apparatus is much better developed in Flustrella than
Prouho had supposed to be the case, in that it really com-
prises both stomach and pharynx, though both it is true are
of a rudimentary character.

It will be remarked, when viewed in relation to the
vestigial character of the stomach, that the considerable
development of the pharynx is somewhat surprising; but
may it not be that the pharynx has, in accordance with the
exigencies of larval life, assumed another function? Mention
has already been made of the globules which are present in
the vacuoles of the cells of the pharynx. At about the time
that these disappear, the larva becomes freed from the vitel-
line membrane, and at the same time the slimy, mucus-like
substance already noted becomes very abundant around the
embryos. Possibly, therefore, the pharynx has assumed a
elandular function, and it may well be that this mucus-like
substance has been derived from the globules previously con-
tained in the pharyngeal cells. Attempts to prove, by treat-
ment with Mayer’s mucicarmine, that the globules in the
pharynx and internal sac really consist of mucus, have so far
given negative results; and this point must, therefore, for the
present remain unsolved. The suggestion here put forward
may also afford an explanation of the existence of the drops
previously noticed as exuding from the cells of the internal sac.

In Pl. 25, figs. 63 b, 64, are shown two bands of tissue
marked * which appear to be budding off from the sides of
the mesendodermic mass. Their significance will be discussed
later.

Toe Mature Larva.

In Pl. 25, figs. 65 a-c, the larva is represented at a stage
shortly anterior to that at which degeneration commences.
It is now enclosed in a chitinous bivalve shell, and has
escaped from the vitelline membrane. From this point

DEVELOPMENT OF FLUSTRELLA HISPIDA. 465

onwards the development of the larval organs, with the
exception of the stomach, has been already so fully described
by Prouho (20) as to make detailed description unnecessary.

The internal sac has become much elongated, so that it now
occupies the greater part of the interior of the larva. Its
_lining epithelium has become much thickened, and has already
lost all trace of cell structure.

The pharynx has altered somewhat in appearance, owing
to the loss of the large globules already described.

The pyriform organ is now fully developed. On the
exterior two depressions are noted—an anterior depression
(Prouho’s ‘ fossette supérieure’) and a_ posterior one
(Prouho’s ‘“‘fente ciliée”), and between these lies a tuft
of cilia, the ‘ papille du plumet vibratile.’ Internally,
corresponding to the “ fossette supérieure,” is the “sys-
téme glandulaire supérieure,’ which consists of a single
mass of cells lying in the longitudinal axis of the larva, while,
similarly corresponding to the “ fente ciliée,” are two masses
of cells, which are placed in the transverse axis of the larva,
one on either side of the posterior part of the “ systéme
elandulaire supérieure,” and which represent the “ systéme
elandulaire inférieure.”

The aboral organ (Pl. 25, fig. 65a) at this stage is fully
developed. It consists of a tuft of long cilia arising from the
thickened patch of aboral ectoderm which has already been
mentioned. This organ is best seen in the living larva, in
which it is visible protruding between the two valves of the
shell; and connecting the aboral organ and the pyriform
organ is seen the neuromuscular cord. In the living larva
the jerking movements of the neuromuscular cord are dis-
tinctly visible, but it has not been possible to draw any
conclusions from these movements as to the functions of
either the pyriform or the aboral organs. The structure of
the aboral organ and neuromuscular cord have already been
fully described by Prouho (20). As stated by this author
the nerve-muscle tract, on reaching the pyriform organ,
breaks‘into three strands, one of which passes to that organ

4.66 R. M. PACE.

between the cells of the two glandular systems, the other
two passing to the cells of the ciliated crown.

The ciliated crown (the “couronne” of French
authors), first visible at this stage, does not consist of a
single ring of cells as it is has been described in other
Bryozoan larve, and as figured by Barrois (2) for Flustrella
hispida, but of two or three rows of cells, as is shown in
Pl. 25, figs. 65 a-c (C.). All trace of definite cell walls in the
ectoderm has vanished by this time, but certainly the ciliated
crown contains at least three series of nuclei, corresponding
presumably to originally three rows of cells.

The cells from which the cilia originate lie rather below
the aboral groove, and can at will be retracted within the
valves of the shell (Pl. 25, fig. 66). Later on the ciliated
crown certainly does have the appearance described by
Prouho (20), of a single series of flat discoidal cells with long
vibratile cilia, each imbedded ina cuticle, and prolonged below
this into a triangular mass of protoplasm (Pl. 25, fig. 66) ;
each of these cells contains a single large nucleus: this
appearance does not, however, arise until quite late in larval
life. No traces of the ciliated crown, or of any specially
enlarged cells, are visible before the larva escapes from the
vitelline membrane, despite Barrois’ assertion to the con-
trary. It is evident, therefore, that in Flustrella the
ciliated crown is formed by a series of cells, and that it
is only late in larval life that these unite to form a single
row of large cells.

The chitinous bivalve shell is developed as a secretion of
the aboral ectoderm, and is closely adherent to the latter.

The aboral groove, which is now strongly developed,
occupies its original position above the ciliated crown (PI. 25,
figs. 65 a-c, M.C.). It is best seen in transverse section.

The stomach, as a result of the growth of the larva, has
become much more elongated, and, owing partly to this,
partly to the absorption of food material, the yolk spherules
surrounding the stomach have become much reduced in
number.

DEVELOPMENT OF FLUSTRELLA HISPIDA. 467

THe DEGENERATING LARVA.

Shortly after the stage which has been briefly described
above, the degeneration of the larval organs commences.
The initiation of this process is shown in PI. 25, fig. 66.

The internal sac becomes enormously thickened, and its
lining ectoderm highly modified, especially near the opening
of the sac to the exterior, where it now assumes a granular
character. The pharynx gradually loses its cellular structure ;
the pyriform organ is at this stage still fully developed, as
are also the aboral organ and neuromuscular tract. The
ciliated crown, as has already been stated, consists at this
stage of a single ring of large cells. The stomach has prac-
tically vanished, its position being marked only by a number
of scattered yolk spherules and of nuclei lying between the
internal sac and the aboral ectoderm.

Origin of the Mesoderm.—Among the above-men-
tioned scattered elements of mesendodermic origin, and
apparently developed from them, occur fibres (Pl. 25,
fig. 66, MZ), which are presumably muscular in nature. Others
of these supposed muscular fibres occupy the former position
of, and are probably developed from, the lateral bands of
tissue previously noted as budding off from the stomach
It is this mesendodermic mass of yolk spherules, nuclei, and
fibres which Prouho (20) regards as representing the meso-
derm in Flustrella hispida. He maintains that the
mesoderm occurs as a distinct layer of cells lying beneath
the aboral ectoderm, generally thickened at the aboral pole;
that a similar membrane overlies the internal sac, and also
that all the muscular elements of the larva are of mesodermic
origin. But, as has already been shown, it is quite impos-
sible at any early larval stage to differentiate the mesoderm
from the general endodermic mass, and Prouho’s so-called
‘“‘mesoderm ”’ is, therefore, undoubtedly not simply mesoderm,
but endoderm, or perhaps rather mesendoderm, since it is, of
course, possible that in this endodermic mass he enclosed the
elements of the future mesoderm, from which the muscles are

468 R. M. PACE.

now formed. The two bands of tissue apparently budding off
from the main mesendodermic mass have already been
noticed ; possibly these may represent the true mesoderm,
which in that case would appear to develop only late in
larval life. Prouho, owing to the fact that he only studied
the later stages of larval life in which the stomach had already
begun to disappear, was led to regard the whole mesendo-
dermic mass as mesoderm, or as mesoderm containing some
endodermic cells, but he himself suggests the necessity of a
detailed study of the larva in its earlier stages in order to
decide the correctness of his view.

Tue ALIMENTARY CANAL IN LARVAL ECTOPROCTA.

The presence of an alimentary canal in Hctoproct larve
has been described by Barrois, Vigelius, Repiachoff, Prouho,
and Harmer.

Barrois (2 and 8) at one period regarded larval Hctoprocts
as having an alimentary canal, owing to his mistaken supposi-
tion that the internal sac represented the stomach. Later on
he saw reason to modify his views, but he pointed out and
depicted (2, pl. vil, fig. 15) an invagination between the
pyriform organ and the internal sac, which he regarded as
the rudiment of a pharynx. He therefore believed that many
Kctoprocts were originally provided with a digestive tube,
and that in cases where no such system is formed the endo-
derm arises as in other Bryozoa where the digestive system is
better developed, but that later on it degenerates to a mass
of yolk spherules filling the interior of the embryo.

Vigelius (25) stated that in the early larval stages of
Bugula, a slight split occurs in the mesendodermic mass
filling the interior of the larva; this split he regards as
representing a primitive stomach, which however, plays a
purely passive role, and does not open to the exterior. Later
on the supposed stomach vanishes, and the endoderm forms
a simple cell mass.

Repiachoff (28) describes for Tendra zostericola the

DEVELOPMENT OF FLUSTRELLA HISPIDA. 469

formation of an archigastrula with a definite opening to the
exterior : this communication becomes obliterated at a later
stage. The fully-developed larva has a stomach which com-
municates with the exterior, similar to that which has been
described by Harmer (11) as present in the developing larva
of Aleyonidium.

Prouho (21) describes the presence of a functional ali-
mentary canal in the “Cyphonautes” larva, the stomach
being developed from the internal endodermic mass in the
manner described above. In an earlier paper on Flustrella
hispida (20), he regards the pharynx as a rudimentary
digestive tube less differentiated than that of Aleyonidium,
and representing either an attempt to form a digestive organ
or a vestige of one which has vanished.

Harmer (11) found, in a species of Alcyonidium, a
definite alimentary canal closely resembling that occurring
in Flustrella hispida, but communicating by a narrow
opening with the exterior. In the same paper he points out
the probability that a similar structure occurs in Flustrella
hispida.

From its endodermic origin, its position, appearance, and
mode of development, and from the close agreement in
structure with the stomach described in Alcyonidium, there
can be no doubt that the larva of Flustrella hispida
possesses a vestigial stomach, and that this and the associated
pharynx must be regarded as vestiges of a digestive system
in which degeneration has proceeded a stage further than it
has in Alcyonidium, since no communication with the
exterior is ever established at any stage.

GENERAL SUMMARY.

The main points in the foregoing paper may be summarised
as follows :—

(1) A “yolk nucleus ” of the type described by Bambeke,
as occurring in the egg of Pholcus, is present in the de-
veloping egg of Flustrella hispida. .

4.70 R. M. PACE.

(2) Segmentation and cell-lineage have been followed out
in detail up to the 32-cell stage.

(3) The formation of the endoderm has been traced.

(4) The oral and aboral ectoderm are differentiated as
early as the 16-cell stage, and remain quite distinct from that
time onwards.

(5) The ciliated ring of the larva is formed by the coales-
cence of several originally distinct rows of cells, and not by
the hypertrophy of a single row.

(6) A stomach, comparable to that of Alcyonidium, is
present also in Flustrella.

REFERENCES.

1. BamBeke, C. v.—“ Contributions A l’histoire de Ja constitution de l’uf,”
iii, ‘‘ Recherches sur l’oocyte de Pholeus phalangioides (Fuessl.),”
‘ Arch. Biol.,’ vol. xv, pp. 511-598, pls. xxiii-xxviii, 1898.

2. Barrots, J.—‘ Recherches sur |’Embryologie des Bryozoaires,’ Lille, 1877.

3. Barrots, J.—* Mémoires sur les métamorphoses des Bryozoaires,” i,
**Métamorphose des Escharines,” ‘Ann. Sci. Nat.,’ ser. 6, Zool.,
vol. ix, No. 7, 67 pp., pls. xili-xvi, 1880.

4, Brae, Fritz.— Untersuchungen tiber die Bryozoen des stissen Wassers,”
‘ Bibliotheca Zoologica,’ No. vi, Cassel, 1890.

5. BrasM, Frirz.— Die geschlechtliche Entwickelung von Paludicella
Ehrenbergii,” ‘Zool, Anz.,’ vol. xix, pp. 54-57, February, 1896.

6. Brarm, Fritz.—‘‘ Die geschlechtliche Entwickelung von Plumatella
fungosa,” ‘ Zoologica,’ Heft 23, Stuttgart, 1897.

7. Catxins, G. N.—‘ Observations on the Yolk-nucleus in the Eggs of
Lumbricus,” ‘'l'rans. New York Acad.,’ vol. xiv, pp. 222-230, 1895.

8. Catvet, L.— Contributions a |’Histoire Naturelle des Bryozoaires
Ketoproctes Marins,” ‘Trav. Inst. Zool. Montpellier,’ ser. 2, No. 8,
1900.

9. Crampton, H. E.—‘‘ Studies upon the Early History of the Ascidian
Kgs.” Part I—“The Ovarian History of the Egg of Molgula
manhattensis,” ‘Journ. Morphol.,’ vol. xv, suppl., pp. 29-56, pl. iii,
December, 1899.

10. Foor, K.—“ Yolk-nucleus and Polar Rings,” ‘Journ, Morphol.,’ vol. xii,
pp. 1-16, pl. i, May, 1896.

11. Harmer, 8. F.—‘Sur l’embryogénie des Bryozoaires ectoproctes,”
‘Arch. Zool. Expt.,’ ser. 2, vol. v, pp. 443-458, pls. xxvii, xxviii, 1887.

12.

13.

14,

15.

16.

LF fe

18.

19.

20.

2.

22.

23.

24.

25.

DEVELOPMENT OF FLUSTRELLA HISPIDA. A71

Hennecuy, L. 'l.—‘ Le corps vitellin de Balbiani dans l’ceuf des verte-
brés,” ‘Journ. l’Anat. Physiol.,’ vol. xxix, pp. 1-39, 1893.

Hiycxs, THomas.— Notes on British Zoophytes, with Descriptions of
some New Species,” ‘Ann. Mag. Nat. Hist.,’ ser. 2, vol. vii, pp. 853-862,
pl. xiv, November, 1851.

JouiEt, L.— Contributions a l’histoire naturelle des Bryozoaires des
cotes de France,” ‘Arch. Zool. Expt.,’ ser. 1, vol. vi, pp. 198-304,
pls. vi-xiii, 1877.

Jorpan, E. O.— The Habits and Development of the Newt (Diemyc-
tilus viridescens),’ ‘Journ. Morphol.,’ vol. viii, pp. 269-366,
pls. xiv-xvilJ, May, 1893.

Korscn£ Lt, E., and Herper, K.—‘ Lehrbuch der vergleichenden Entwick-
lungsgeschichte der wirbellosen Thiere,” ‘‘ Aligemeiner Theil,” pt. i,
Jena, 1902.

KRAEPELIN, Karu.— Die Deutschen Siisswasser Bryozoen” II, ‘‘ Ent-
wicklungsgeschichtlicher ‘Teil.,” ‘Abhandl. Naturw. Ver., Hamburg,’
vol. xii, 1892.

Mertens, H.— Recherches sur la signification du corps vitellin de
Balbiani dans Povule des mammiféres et des oiseaux,” ‘ Arch. Biol.,
vol. xiii, pp. 389-422, pl. xiv, 1895.

Ostroumorr, A.—‘‘Zur Entwicklungsgeschichte der cyclostomen
Seebryozoen,” ‘ Mitth. Zool. Stat. Neapel,’ vol. vii, pp. 177-190,
pl. vi, April, 1887.

Provuno, H.— Recherches sur la larve de la Flustrella hispida
(Gray),” ‘Arch. Zool. Expt.,’ ser. 2, vol. viii, pp. 409-459, pls. xx1i-
xxiv, 1890.

Provuno, H.—* Contribution a histoire des Bryozoaires.,” ‘ Arch. Zool.
Expt.,’ ser. 2, vol. x, pp. 557-656, pls. xxili-xxx, 1892.

ReEpDFERN, Peter.—“‘F lustrella hispida and its Development,” * Quart.
Journ. Micr. Sci.,’ ser. 1, vol. vi, pp. 96-102, 1858.

Repracnorr, W.—“ Ueber die ersten embryonalen Entwicklungsvorgange
bei Tendra zostericola,” ‘Zeitschr. wiss. Zool.,’ vol. xxx, suppl.
pp. 411-423, pl. xix, May, 1878.

Van DER Srricut, O.—“ La structure de l’ceuf des Mammiféres. lre.
Partie. ’Oocyte au stade de l’accroissement.,”’ ‘ Arch. Biol.,’ vol. XX
pp. 1-101, pls. i-iii, December, 1904.

VieeLIus, W. J.—“ Zur Ontogenie der marinen Bryozoen,” ‘ Mitth.
Zool. Stat. Neapel,’ vol. vi, pp. 499-541, pls. xxvi-xxviii, June, 1886.

472 R. M.. PACE.

26. WaLLacr, W.—‘‘ Observations on Ovarian Ova and Follicles in Certain
Teleostean and Elasmobranch Fishes,’ ‘Quart. Journ. Micr. Sci.,’
n.s., vol. 47, pp. 161-213, pls. 15-17, 1903.

27. Witson, E. B.—‘‘ On Protoplasmic Structure in the Eggs of Echinoderms
and Other Animals,’ ‘Journ. Morphol.,’ vol. xv, suppl., pp. 1-25,
pls. i, i, December, 1899.

Postscript.—Since the above account of the yolk nucleus
in the egg of Flustrella hispida was written, a paper (24)
has appeared by O. van der Stricht on the yolk nucleus in the
eggs of mammals. The account of the yolk nucleus given by
this author appears in the main to confirm Bambeke’s observa-
tions and views.

EXPLANATION OF PLATES 22—25,

Illustrating Mrs. R. M. Pace’s paper on “The Harly Stages
in the Development of Flustrella hispida (Fabricius).”

REFERENCE LETTERS.:

A. Aboral surface of larva. 4.#e. Aboral ectoderm. Bl. Blastopore.
C. Ciliated crown, corona, or “ couronue.” Ca. Aboral organ or “ calotte.”
Ci. Fente ciliée.” er. Crystalloid bodies. €.7Z'. “ Papille de plumet vibratile,”’
ciliated tuft. Ze. Hctoderm. Hd. Endoderm, or mesendoderm. Jo. Follicle
cells. Fs. “‘Fossette supérieure.” Zw. Funicle. G. Chromatin granules.
g. Globules in cells of pharynx and internal sac. JZ. Intestine. J.8. Internal
sac. MM. Muscle fibres. M.C. Aboral groove or mantle cavity. Jc. Mesen-
chyme lining of zocecium. WV. Nucleus. 2. Nucleolus. 1.M. Neuro-
muscular cord. O. Oral surface of larva. Oe. Cisophagus. O.#c. Oral
ectoderm. O2. Oil globules. Ov. Ovary. P.B. Polar bodies. Ph. Pharynx.
Py. Pyriform organ, &. Rectum. S. Shell. S.C. Segmentation cavity.
S.G/ “Systéme glandulaire supérieure.” 8.4.” “Systeme glandulaire in.
férieure.” St. Stomach. 7. Testes. 7S. Tentacle sheath, vc. Vacuoles
in the yolk nucleus. V.J/. Vitelline membrane. x. Space or clear zone
surrounding the yolk nucleus. J. Yolk spherules, Y.N. Yolk nucleus.
Z. Zoccial cavity. Those tissue tracts which are possibly to be regarded as
being destined to give rise to the mesoderm are marked by a small asterisk.

DEVELOPMENT OF FLUSTRELLA HISPIDA. 473

In the figures of segmenting ova, the first four cells formed have been
lettered, merely for convenience of reference, “ A,” ‘* B,” “C,” “ D,” and the
daughter cells arising from these latter are distinguished by index numbers, a
negative index indicating the generation to which a particular cell belongs, and
a positive index its place in that generation; thus A* denotes the fourth cell
derived from A in the fifth generation (see also p. 452).

Allthe figures, both of sections and of entire larve, have been drawn by the
aid of the Zeiss camera lucida.

PLATE 22.

Fic. 1.—Transverse section of a young colony of Flustrella hispida,
collected in March. The section is taken close to the apex of the colony,
and shows the position of the testes on the lateral walls in the front of one
zocecium, and the ovary lying on the funicle at the back of a neighbouring
zocecium. x 6d.

Fig. 2.—Section of a young ovary lying on the funicle, showing the follicle
cells commencing to grow in among the young ova. X 225.

Fie. 3.—Section of an older ovary ; the follicle cells are now seen to have
increased in number. xX 225.

Fic. 4.—Section of a young ovary showing four young ova, the walls of
which are still unformed. The yolk nucleus is present in the form of small
dark granules. x 400.

Fig. 5.—Section of a somewhat older ovary showing four ova around which
the follicle cells have not yet developed. The dark granules which represent
the yolk nucleus have increased in number, and in many cases are seeii to lie
in close contact with the membrane of the germinal vesicle. x 400.

Fic. 6.—Section of a young ovum more advanced than those shown in
Fig. 5. The granules of the yolk nucleus are seen to lie in four groups, each
containing two granules, and three of these groups are surrounded by clear
spaces (7). xX 400.

Fic. 7.—Section of a young ovum of the same age as that shown in Fig. 6.
The granules of the yolk nucleus are grouped together and lie within a clear
space. X 400.

Fic. 8.—Section of a young ovum somewhat more advanced than that
shown in the preceding figure. The majority of the dark granules have
become fused to form a single large yolk nucleus lying within a clear space (7)
separated from the germinal vesicle. Two vacuoles have already appeared in
the yolk nucleus. x 400.

Fics. 9—12.—A series of four sections through a slightly older egg, to
show the difference in the appearance of the egg according to the point at

AI74, R. M. PAGE.

which the section is taken. Fig. 9.—A section of the egg showing the
commencement of the clear region in which the yolk-nucleus generally lies.
In Fig. 10 the section passes through the nucleus and nucleolus, and the
clear space has assumed a hemispherical shape. In Fig. 11 the yolk-nucleus
is shown as a well-developed crescentic body lying in the clear space visible
in the preceding figures, and in close contact with the germinal vesicle; the
section passes through only the upper part of the nucleus. Fig. 12.—The
yolk nucleus is seen lying in a clear space. ‘The germinal vesicle is no longer
visible, but lies underneath the cap formed by the yolk nucleus. x 400.

Fic. 13.—A section of an ovum of about the same age as that illustrated
in the preceding figures. The yolk nucleus has become rather more crescentic
in form, and three large vacuoles are present and contain crystalloid bodies
(er.). x 400.

Fic. 14.—A section of an egg of the same age as that shown in fig. 10.
The yolk nucleus is more markedly hemispherical in form, and the number of
vacuoles is larger than at previous stages. x 400.

Fie. 15.—Section of an older egg. The yolk nucleus has now assumed the
form of a ring surrounding, and in close contact with the germinal vesicle. An
indication of approaching degeneration is seen in the reticulate structure of
the yolk nucleus. x 400.

Fic. 16.—A_ scction of a still older egg in which the degeneration of the
yolk nucleus has commenced. The yolk nucleus has lost its regular outline,
and shows a markedly reticulate structure, the meshes of the network staining
more deeply than the interlying substance. ‘I'he outlines of the surrounding
space have also become somewhat irregular. x 400.

Fic. 17.—A section of an egg at a slightly later stage. The degeneration
of the yolk nucleus has advanced considerably and is very marked. The body
has completely lost its regular outline, and the boundary of the surrounding
space is very irregular. x 400.

Fic. 18.—Section of an older egg, showing the complete disintegration of
the yolk nucleus, which has now assumed the form of numerous minute,
darkly-staining granules which lie in a loose ring around the germinal vesicle
x 400.

Fic. 19.—Section of an egg of about the same age as that shown in
fig. 18, but which exhibits a somewhat different method of fragmentation of
the yolk nucleus. In this case the products of disintegration have assumed
the form of dark, irregular patches lying within clear spaces, and forming an
open ring around the germinal vesicle. x 400.

Fies. 20, 21.—Sections of eggs showing still other methods of fragmenta-
tion of the yolk nucleus. x 400.

DEVELOPMENT OF FLUSTRELLA HISPIDA. A475

Fies. 22, 23.—Sections of slightly older eggs in which the products of dis-
integration of the yolk nucleus have retreated towards the periphery of the
egg, there forming an open ring of deeply staining patches, each of which lies
within a clear space. x 400.

PLATE 23.

Fre. 24.—Section of an egg showing the first appearance of the yolk in the
form of minute globules scattered in the protoplasm. All trace of the yolk
nucleus has vanished. x 400.

Fre. 25.—Section of an egg in which the yolk is fully developed. The
section was stained with iron hematoxylin, and the centres of many of the
yolk spherules remained stained even after prolonged washing. x 400.

Fies. 26, 27.—Sections of eggs of about the same age as that illustrated in
figs. 14 and 15, and which have been treated with osmic acid. The yolk
nucleus is in immediate contact with the germinal vesicle, and the clear space
in which it usually lies is not present; oil globules (O/.) are present both in
the yolk nucleus and in the surrounding protoplasm. x 250.

Fie. 28.—Section of an egg stained with osmic acid, in which yolk-forma-
tion is occurring. All trace of the yolk nucleus has vanished; oil-drops are
scattered among the developing yolk, and are especially abundant towards the
periphery of the egg; they may be distinguished by their darker colour.
x 250.

Fie. 29.—Section through the germinal vesicle of a young egg. ‘The
chromatic network is well developed, and darkly staining nodules are present
at itsnodes. Lying at the edge of the germinal vesicle is a dark nodule (¥.W.)
similar to those which occur at the nodes of the chromatin reticulum, and
which is apparently in process of passing out through the membrane of the
germinal vesicle to become one of the granules which will coalesce and form
the yolk nucleus. x 650.

Fie. 30.—Section of the germinal vesicle showing the gradual attenuation
of the chromatin network as maturation proceeds. x 650.

Fies. 31, 32.—Sections of the germinal vesicle in somewhat older eggs,
showing the increased size of the germinal vesicle and the attenuation of the
chromatin network. %X 650.

Fic. 33.—Section of the germinal vesicle showing the thickening of the
chromatin network which occurs in later stages. xX 650.

Fic. 34.—Section of the germinal vesicle of an egg in which the formation
of the polar bodies will shortly take place. The chromatin network is much
thickened, and the substance between its meshes stains more darkly than at
earlier stages. he nuclear membrane has become irregular. Xx 650,

476 R. M. PACE.

Fic. 35.—Section of the germinal vesicle at a stage slightly later than that
shown in fig. 34. The nucleolus has become relatively very large, the
chromatin network is thicker, and the substance between the meshes stains
still more deeply than at previous stages. x 650.

Fic. 36.—Section of the germinal vesicle in a mature egg before the
formation of the polar bodies has taken place. The germinal vesicle has
decreased in relative size and has assumed an ameeboid form. All trace of the
chromatic network has vanished. x 650.

Fies. 37-42.—Series illustrating the primitive cleavage of the egg. x 65.

Fic. 43.—The segmenting egg at the four-cell stage. x 90.

Fic. 44.—The segmenting egg at the eight-cell stage. Lateral view.
x 90.

Fic, 45,—The segmenting egg at the twelve-cell stage. xX 90.

(a) Lateral view. (4) Aboral view.

Fic. 46.—Oral view in optical section of the larva at the twelve-cell stage,
showing the formation of spindles in the four oral cells prior to the division
which gives rise to eight oral cells. x 400.

Fic. 47.—Larva at the sixteen-cell stage. x 90.

(a) Oral view. (4) Lateral view. (c) Aboral view.

Fie. 48.—Oral view of the larva at the sixteen-cell stage, showing the
formation of spindles in the four central oral cells prior to their division to
form four new oral cells of the sixth generation. x 180.

Fic. 49.—Larva at the twenty-cell stage. x 90.

(a) Oral view. (4) Lateral view. (c) Aboral view. The four large
oral cells are of the sixth generation; of the eight surrounding cells, four
belong to the fifth and four to the sixth generation.

Fie. 50.—Larva at the thirty-two cell stage. x 90.

(a) Oral view. (4) Lateral view. (c) Aboral view.

PLATE 24.

Fic. 51.—Section of an egg in which the first polar body has been formed and
the second is in process of formation; the first polar body lies outside the
egg. x 250.

Fic. 52.—Section of a larva at the four-cell stage showing the formation of
the nuclear spindles prior to the division to form the eight-cell stage. x 225.
Fie. 53.—Longitudinal section of the larva at the twenty-cell stage. x 225.

Fie. 54.a.—Transverse section of the larva at the twenty-cell stage, showing
the formation of nuclear spindles in the aboral cells A?, C2, prior to division,
The four oral cells belong to the sixth generation. x 225.

DEVELOPMENT OF FLUSTRELLA HISPIDA. 477

Fie. 544.—Transverse section of the larva at the twenty-cell stage, show-
ing the formation of nuclear spindles in the aboral cells A1—C}, and in the
oral cell A, prior to division to form cells of the sixth generation. xX 225.

Fic 55 a, 6.—Transverse sections of the larva at the twenty-cell stage, but
ata slightly later period than that illustrated in Figs. 54a, 4, the division of
two of the aboral cells being now completed. x 400.

Fie. 56.—Longitudinal section of the larva at the thirty-two cell stage.
x 225.

Fie. 57.—Longitudinal section of the larva showing the formation of the
primitive mesendoderm cells by the division of the four large central oral cells.
x 225.

Fie. 58.—Section of the larva showing the enclosure of the four large
central oral cells within the primitive segmentation cavity, forming, together
with the four cells previously enclosed, in all eight mesendoderm cells; a
blastopore (B/.) is represented. x 225.

Fic. 59.—Section of the larva ata slightly later stage than that shown in
the previous figure. The difference in size between the cells of the aboral and
oral ectoderm is noticeable; the eight primitive mesendoderm cells have
divided, forming a mesendodermic mass which projects from the oral face into
the segmentation cavity ; the blastopore is still open. x 225.

Fic. 60.—Section of the larva at a stage subsequent to the closure of the
blastopore. J.C. marks the position of a slight groove which is to be seen
on the exterior of the larva, and the slight protuberance visible below this in
the external view is due to the difference in size of the cells of the oral and
aboral ectoderm, which meet along this line. The mesendoderm almost com-
pletely fills the segmentation cavity. x 225.

PLATE 25.

Fie. 61.—Section of an older larva in which the formation of the mesen-
doderm is completed, but in which the larval organs have not yet begun to
form. The difference in size between the cells of the oral and aboral ectoderm
is now very marked; the groove M/.C. has deepened, and the oral cells below
it are slightly protruded, giving rise to a raised ring on the external surface
of the larva; this ring is, however, quite unconnected with the ciliated ring
which develops later. The mesendodermic mass has by this time completely
obliterated the segmentation cavity. x 225.

Fic. 62.—Longitudinal section of a young larva at the period when the
larval organs are beginning to form. Three oral invaginations are observable;
these will later develop into the internal sac, pharynx, and pyriform organ,
respectively. Globules (v.) are seen to be exuding from the cells of the internal
sac, and similar globules are also present within the cells of the internal sac

478 R. M. PACE.

and pharynx. The neuro-muscular cord (V.I/.) is in process of development.
The mesendoderm forms a solid mass in the posterior part of the larva.
Viewed from the left side. x 225.

Fie. 63 a.—Longitudinal section through the larva at a slightly later stage
than that illustrated in Fig. 62, viewed from the left side. The mesendodermic
mass has now become hollowed out to form the stomach. Viewed from the
left side. x 225.

Fic. 63 6.—Transverse section through the region of the pharynx of a larva
of the same age as that shown in Fig. 63a. Two bands of tissue (*), which
may possibly represent the true mesoderm, are budding off from the mesendo-
dermic mass. X 225.

Fic. 64.—Transverse section of a larva older than that which is represented
in Fig. 63; the plane of section passing through the stomach, internal sac, and
pharynx. x 225.

Fic. 65 a-c.—Sections of mature larve.

a. Longitudinal section of a larva, viewed from the right side, in which
the larval organs are all fully developed. The pharynx and internal sae
have lost the globules which were present in earlier stages; the pyriform
organ now consists of the “systéme glandulaire supérieure” (8.G.') with
the “ fossette supérieure” (Fs.) and the “systéme glandulaire inférieure,”
(S.G.") only the right side of which is visible: the ‘“‘fente ciliée” is not
shown. The stomach has become very elongated. ‘The ciliated crown (C.)
is now visible. (From a. drawing by Dr. 8. F. Harmer.)

b. Transverse section of the larva in a plane passing through the stomach and
internal sac. The two bands of tissue which appear to be budding off from the
mesendoderm are again noticeable. x 225.

c. Transverse section passing through the stomach and pharynx of the larva.
x 225.

Fic. 66.—Transverse section of a larva in which degeneration has com-
menced. ‘The walls of the internal sac have become much thickened and the
protoplasm near the opening to the exterior has assumed a granular appear-
ance. The ciliated crown now consists of a single row of large cells. The
stomach has vanished, and the mesendoderm is represented only by scattered
nuclei and yolk spherules, among which lie numerous muscle-fibres. Xx 225.

ORIGIN OF THE TRABECUL® OF THE OPTIC NERVE. 479

Researches on the Origin and Development of
the Epiblastic Trabecule and the Pial Sheath
of the Optic Nerve of the Frog, with illus-
trations of Variations met with in other
Vertebrates, and some Observations on the
Lymphatics of the Optic Nerve.

By

J. EF. Gradon, WMi.A.,
St. John’s College, Oxford.

With Plates 26, 27.

INTRODUCTION.

Just over a year ago I began to feel dissatisfied with
Assheton’s (1) conclusion that the cells of the optic stalk do
nothing more than serve as a conductor for the fibres of the
optic nerve.

As I was aware that Assheton’s (1) opinion had been fully
endorsed by Professor Ryder (8) in the embryological sec-
tion of Norris and Oliver’s ‘System of Diseases of the Hye,’
I thought it unnecessary to go any further into the literature
of the subject before beginning the present researches, and
unfortunately I had finished them before I found that the
part of the epiblastic trabecule that I shall speak of as
transverse, had been dealt with by W. Miller (6), Kélliker
(5), Robinson (7), Studnicka (9), and Froriep (3a). But, as
all of these well-known investigators have dealt with the
transverse fibrils as though they were the whole epiblastic
trabeculee of the optic nerve, instead of being only a part of

VOL, 50, PART 3,—NEW SERIES. 34:

480 J. T. GRADON.

its complex framework, it will be my aim in the present
paper to describe its origin and development as a whole.
The subject will be dealt with under the following heads:

PAGE
I. The Relation of the Optic Stalk to the Nerve-fibres . . 480
II. Cellular Segmentation . : . 481
III. Obliteration of the Lumen of the Optic Stalk 5 . 482
IV. Period of slow growth, followed by one of great activity, conse-
quent on the formation of the Arachnoid Sheath and the
Enclosure of the Subarachnoidal Lymph Space ; . 483
V. Pigmentation of the Trabecule . : : . 485
VI. Supplementary Remarks 486
VII. Formation of the Pial Sheath, and its Relation to the Membrane
Limitans Externa . 487
VIII. Some Variations of Epiblastic Trabecule: oe with in ie Deve
loping Optic Nerve of the Mouse, the Trout, the Dog-fish,
and the Chick : : : ; . 487

I. Tae RELATION oF THE Optic STALK TO THE NERVE-FIBRES.

In discussing Assheton’s (1) contention that the first nerve-
fibres, though lying along the posterior border of the stalk,
are at first entirely outside it and separate from it, Robinson
(7), after showing that this conclusion is altogether at vari-
ance with the observations of W. Miiller (6) upon the lamprey,
of K6lliker (5) upon rabbits, pigs, and calves, of Keibel (4)
upon reptiles, and of Froriep (8) upon cartilaginous fishes,
makes the following very important statement :—‘“If the
condition which Assheton (1) found in the frog is present in
mammals also, then it follows that the sustentacular frame-
work of the optic nerve of man may consist, for the most part,
like the framework of an ordinary cerebro-spinal nerve, of
mesoblastic tissue surrounding and embedding the epiblastic
nerve-fibres, but if Miiller’s and Kolliker’s statements are
well founded, then the sustentacular tissue of the optic nerve
in man and mammals must consist chiefly of epiblastic tissue
derived from the primitive epithelial cells of the optic

ORIGIN OF THE TRABECULAE OF THE OPTIC NERVE. 481

stalk; ... this is a matter of some morphological, and cer-
tainly of pathological, importance.”

All my specimens undoubtedly confirm the observations of
the authorities quoted by Robinson (7), and his own state-
ments, which are based upon observations made on human
embryos, cats, ferrets, sheep, rabbits, rats, and mice, viz. that
the ingrowing nerve-fibres lie within the membrana limitans
externa, throughout the whole of their course in the optic
stalk, and that they enter the stalk along the ventral wall;
though Froriep (3a) has lately stated that, in his specimens
of rabbit embryos, the earliest bundles of nerve-fibres grow
in higher up on each side of the ventral wall, and that the
nuclei that lie above the ingrowing nerve-fibres are pushed
up towards the lumen of the stalk, whilst those that lie below
are pushed still further down, as the number of nerve-fibres
increases.

In tadpoles of 6 mm. in length I have invariably found the
earliest bundles of nerve-fibres, as they issue from the optic
cup, occupying a central position just within the membrana
limitans externa of the ventral wall of the stalk, and, as they
approach the brain, getting more and more towards the pos-
terior side of it, though in 8°5 mm. tadpoles they seem first
of all to travel a little anteriorly for a very short distance,
just after leaving the optic cup. These observations are in
agreement with Froriep’s (8a), figs. 237—239, taken from
tadpoles.

II. CeLLtuLaArR SEGMENTATION.

Robinson (7) has referred to the difficulty of obtaining in-
dications of definite cell-territories in the early stages of the
embryonic optic stalk of the rat.

In tadpoles of 4°5 mm. in length cell limits are certainly
recognisable (fig. 1), but in those of 6 mm. in length they
can rarely be distinguished from the pigmented fibrils of
the protoplasm that encircles the granules of food-yolk, or

482 J. T. GRADON.

takes up the position lately occupied by those that have been
assimilated.

An inspection of fig. 2 will show that the entrance of the
nerve-fibres along the ventral wall of the stalk produces a
confluence and stretching of these delicate protoplasmic
fibrils, and, at the same time, brings into prominence the con-
nections subsisting between nucleus and nucleus.

Further ingrowth of the nerve-fibres resolves these fibrils
into a complex framework of supporting elements which,
from transverse, longitudinal, and horizontal sections, may be
seen to radiate in every direction from the border of each
nucleus of the stalk.

This intermediate arrangement of the condensed proto-
plasmic fibrils finally becomes differentiated with the multi-
plication of the nuclei of the stalk into a transverse, oblique,
and longitudinal framework which, as we shall afterwards
see, also provides a complete system of lymph channels
throughout the interior of the optic nerve.

III. OBLITERATION OF THE LUMEN oF THE Optic STALK.

The obliteration of the lumen of the stalk has received con-
siderable attention from previous observers. Assheton (1)
ascribes it to pressure from the cartilaginous walls of the
cranium, whilst Robinson (7) considers that this cannot be
looked upon as an important agent, and concludes that the
obliteration ‘‘is brought about by developmental changes in
growth and relationship of the constituent parts of the stalk,”
and that ‘‘ with these is associated the invasion of the optic
nerve-fibres.”

In fig. 12 we can see that pressure is exerted on the stalk
by the cartilaginous walls of the cranium, and it is also cer-
tain that pressure produced by contact with the back of the
eye is the cause of the very decided oval shape of the stalk
at this point; in an 11 mm. tadpole (fig. 4), its shape, when
free, is almost round. But there are probably several causes at

ORIGIN OF THE TRABECUL& OF THE OPTIC NERVE. 483

work—both within the stalk itself and outside it—in bringing
about the obliteration of the lumen.

The pressure everywhere outside the stalk is evidently
greater than that within its lumen for, although the first
nerve-fibres lie just within the external membrane, the pres-
ence of the smallest bundle is enough to produce a certain
amount of bulging of the upper border of the ventral wall
into the lumen without in the slightest degree altering the
regularity of the outline of the external membrane underneath
it (fig. 2).

It is true that further ingrowth of nerve-fibres produces a
considerable change in the outline of the stalk, as shown in
fig. 3, but by this time, the lumen has almost been closed,
and still further ingrowth of nerve-fibres at the sides, com-
pletes its obliteration, and, at the same time, restores the
slightly longer axis of the stalk to the horizontal position
(figs. 4 and 5).

IV. Periop or Stow GRowrH, FOLLOWED BY ONE OF GREAT
ACTIVITY, CONSEQUENT ON THE FORMATION OF THE ARACH-
NOID SHEATH AND THE ENCLOSURE OF THE SUBARACHNOIDAL
LympH SPACE.

Between the stages shown in figs. 4—6, representing
tadpoles from 11 mm. to 21 mm. in length, the diameter
of the stalk increases only very slightly. This is due to the
fact that there is scarcely any karyokinesis going on within
the stalk, and the protoplasmic framework, which is now
binding the nerve-fibres together, seems unable to accommo-
date itself to further expansion.

Meanwhile, the stalk is being continually more and more
stretched between the eye and the brain, so that it is possible
to obtain transverse sections of a 21 mm. tadpole that do
not contain a single nucleus, only the protoplasmic fibrils
proceeding from nuclei that lie in the preceding and suc-
ceeding sections.

484 J. T. GRADON.

As there are no blood-vessels inside the optic nerve of the
frog, and very few capillaries on the pial sheath, it will be
evident that, up to this stage, the nutrition of the stalk is at
a very low ebb ;. there are, however, no indications of degene-
ration; in fact, it is possible to show a solitary instance of
cell-division now and again where the fibrils of the trabecule
are in contact with the delicate capillaries that have crept up
the pial sheath from the pia mater of the brain (fig. 9).

But when we turn to a 27 mm. tadpole (fig. 7) it is evident
that a remarkable change has taken place; mitosis is every-
where abundant, the number of cells has already greatly
increased, and the total diameter of the stalk is considerably
greater than it was in the preceding stage.

This sudden change coincides with the more complete
enclosure of the subarachnoidal lymph space, which has come
about through the formation of the arachnoid sheath. Before
the dural sheath has had time to form, the arachnoid itself is
directly connected with the ophthalmic artery by a band of
connective tissue, by means of which transudation of lymph
from the artery doubtless takes place.

The subarachnoidal space, which is evidently not suffi-
ciently enclosed until this stage has been reached, is now
filled with nutritive material, which passes through the pial
covering, and then reaches every nucleus of the stalk by
means of the elaborate system of tiny channels that follow
the course of each fibril of the epiblastic trabecule (figs. 8
and 13, and pe. fig. 17).

Moreover, we may now find, among the meshes of the |
connective-tissue cells that joi the ophthalmic artery and
the optic nerve, numbers of lymph-corpuscles, all in various
stages of cell-division, though I have only shown them in
outline in fig. 7.

Development now proceeds very rapidly, but I have not
thought it necessary to publish any drawings of stages
between that shown in fig. 8 and the adult stage shown in
fig. 10.

In the latter figure we can see the final arrangement of th

ORIGIN OF THE TRABECULE OF THE OPTIC NERVE. 485

epiblastic trabecule from the sector that I have filled in with
nuclei and fibrils; the other part of the drawing only shows
the distribution of the nuclei and some of the thicker fibrils.

It will also be evident from this drawing that the stellate
arrangement of the nuclei and the trabecule, found in a
32 mm. tadpole, does not persist; it 1s gradually lost in
succeeding stages.

V. PIGMENTATION.

In the earliest stages of the stalk the fibrils of the proto-
plasm surrounding each granule of food-yolk, are pigmented,
and can be traced perhaps separately.

But, when there has been a confluence of probably several
of these, through ingrowth of nerve-fibres, it makes the
condensed transverse fibrils, seen in sections transverse to
the stalk, stand out in bold relief, and, when one has learnt
what to look for by means of these sections, allows the con-
densed oblique and longitudinal fibrils to be seen, in longi-
tudinal sections, without having recourse to special methods
of staining.

But as soon as the increased flow of lymph, that we have
ascribed to the enclosure of the subarachnoidal lymph space,
takes place throughout the nerve, the trabecule, excepting
sometimes a very short and thickened piece close to the
nucleus, become completely depigmented.

This renders it afterwards very difficult to follow the
delicate, oblique and longitudinal fibrils among the nerve-
fibres.

But even in the adult state the amount of pigment which
the thickened end of the fibril sometimes contains near its
nucleus is sufficient to catch the eye when the rest of the
fibril would easily escape notice.

I have not thought it necessary to publish drawings of
longitudinal sections from stages later than that represented
in fig. 14, as there is nothing further to show than a con-

486 J. T. GRADON.

tinual increase of the number of nuclei and fibrils of the
trabecule, without any apparent increase in the thickness of
the latter.

It clearly follows, from what has been said, that the cells
of the optic stalk are spongioblasts, and that they, therefore,
take no part in the production of optic nerve-fibres, which
arise, according to the researches of Ramon y Cajal (2) and
other well-known investigators, from neuroblasts, chiefly in
the retina.

VI. SuprpLEMENTARY REMARKS.

Even when there is not sufficient protoplasm surrounding
the resting nuclei of the stalk to be represented in the
drawings, it will be understood that there is still an ex-
tremely delicate layer of it covering them, and that the fibrils
of the trabecule form the continuations of it. This thin sheet
of protoplasm may, however, be distinctly recognised around
the nuclei that are undergoing division (vide especially
fig. 9).

The fantastic outlines of the nuclear walls are accounted
for by the fact that each fibril of the trabecule is being
stretched by continual ingrowth of nerve-fibres, and is,
therefore, pulling its nucleus towards the point of its attach-
ment.

In this connection it will be interesting to compare the
nuclear outlines of the densely-packed optic nerve of the
frog (fig. 8), with those of the much less densely packed
optic nerve of the dog-fish (fig. 17).

The nerve-fibres contained in the optic nerve of the latter
are so comparatively few in number, and the lymph channels
so wide and numerous, that, when favourable transverse
sections of it are viewed with a very low power, the nuclei
themselves appear to form a well-defined framework, through
being pulled into mere threads between the nerve-fibres.

In the frog, on the other hand, the lymph channels, though

ORIGIN OF THE TRABECULE OF THE OPTIC NERVE. 487

numerous, are so comparatively narrow, and the nerve-fibres
so densely packed around the nuclei, that the pull exerted on
a nucleus by each fibril of the trabecule, can only result in
the production of a short cone,

VII. Formation oF THE PIAL SHEATH, AND ITs RELATION TO
THE MempBrana Limirans ExTrrna.

I have figured the formation of the pial sheath from the
earliest stages to show how the mesoblastic cells that enter
into the formation of its connective-tissue layer, gradually
unite with the external membrane of the stalk or its later
representatives—the ends of the epiblastic trabecule.

But the union is only apparent, for a regular system of
lymph spaces is formed between the ends of the trabeculz
and the layer of connective tissue, which may be separated
in sectionising (fig. 17).

I have referred to the scantiness of its vascularity in the
frog in a preceding section, p. 484.

In rat embryos of 8 mm. in length Robinson (7) found the
peripheral boundary of the stalk clearly defined, but was
unable to demonstrate a distinct external limiting membrane.

In the frog there never is any doubt about the external
limits of the stalk, though the boundary is naturally more
delicate in a 6 mm. tadpole than in those of succeeding
stages.

The optic stalk of the chick, containing a very great
number of cells, shows a well-defined external limiting
membrane, supported by numerous mesoblastic cells, when
the nerve-fibres begin to grow in, on the fourth day of
incubation.

VIII. Some Variations or Epiptastic TRABECULZ MET WITH IN
THE DeEvetopine Optic Nerve or tHE Mouse, tHe TRovt,
THE Doa-FIsH, AND THE CHICK.

Although I have selected the tadpole for tracing the com-
plete development of the epiblastic trabecule, still we can

488 J. T. GRADON.

find in other embryos some interesting variations which
assist us very materially in gaining a fuller comprehension of
the subject.

Fig. 15, which represents a longitudinal section of the
optic nerve of an embryo mouse of fourteen days, gives us a
better idea of the longitudinal fibrils than we have been able
to gain in considering the later stages of these fibrils in the
frog, as they lie in the same plane for a much greater distance,
and are, at the same time, rather thicker for a greater part
of their length than those we have seen in the frog.

On the other hand, in fig. 16, representing a longitudinal
section of the optic nerve of a developing trout (length of
optic stalk *5 mm.), we see fibrils that are extraordinarily
thick, and consequently very easily seen near their nuclei,
but undulating to such an extent that it is only possible
to follow them a very short distance away from their nuclei.

In fig. 17 I have shown a transverse section of the optic
nerve of a 33 mm. dog-fish, and on p. 486 I have compared
the nuclear outlines of the densely packed optic nerve of the
frog (fig. 8) with those of the much less densely packed optic
nerve of the dog-fish, shown in this figure, so that I need
only point to the difference in the arrangement of the nuclei
themselves, though, as I have stated on p. 485, the stellate
arrangement of the nuclei and the trabecule in the optic
nerve of the tadpole is lost before the adult stage is reached.

Another point of difference between the frog and the dog-
fish hes in the fact that the pial sheath of the latter is richly
supplied with blood-vessels, though not represented in the
drawing.

A transverse section of the optic nerve of an eight-day
chick (fig. 18) shows certain peculiarities of trabecular forma-
tion; the arrangement of the nuclei is free, like that of the
dog-fish, but the nuclear outlines more closely resemble those
of the frog, due, in my opinion, to the same causes as those
I have given to account for the peculiarities of these outlines
in the frog.

I have shown one of the numerous capillaries that supply

ORIGIN OF THE TRABECULH OF THE OPTIC NERVE. 489

the interior of the optic nerve of the chick with blood and
lymph descending into it from the pial sheath which is richly
supplied with blood-vessels, and the adjacent fibrils of the
epiblastic trabeculee may be seen in contact with it (fig. 18).

In conclusion, I deeply regret to say that since this article
was written, the sudden death of the Linacre Professor of
Comparative Anatomy has rendered it impossible for me to
publicly express my thanks to him for allowing me to carry
on my researches in ocular embryology in the Department of
Comparative Anatomy at Oxford, and more especially for the
kind interest that he always took in my work. But I grate-
fully avail myself of this opportunity of thanking Dr. J. W.
Jenkinson, Assistant to the Linacre Professor, for kindly
providing me with unlimited material and preparations for
the purpose of the present article.

SUMMARY.

We have seen that our trabecule are entirely epiblastic in
origin, for we have shown that the entrance of the nerve-
fibres along the ventral wall of the embryonic optic stalk
produces a confluence and stretching of the protoplasmic
fibrils of the epiblastic cells of the stalk, which result in a
complex framework of supporting elements radiating in every
direction from, the border of each nucleus of the stalk, and
that this complex framework afterwards becomes more or less
differentiated into a transverse, oblique, and longitudinal
trabeculee with the multiplication of the nuclei of the stalk
and without any admixture of mesoblastic cells, for we have
also shown that the nerve-fibres lie, throughout the whole of
their course, in the optic stalk, within the membrana limitans
externa, on the outside of which we have followed the gradual
formation of the connective-tissue layer of the pial sheath.

We have noticed the obliteration of the lumen of the stalk,
and have ascribed it to various causes operating within the
stalk itself and outside it, though chiefly to the ingrowth of
nerve-fibres.

4.90 J. T. GRADON.

We have seen that, in the development of the optic nerve
of the frog there is a period of slow growth, followed by one
of great activity, and we have felt justified in ascribing this
sudden outburst of activity to a greatly increased flow of
lymph into it by means of the elaborate system of minute
channels that follow the course of each fibril of the epiblastic
trabeculae, and consequent upon the formation of the arach-
noid sheath and the enclosure of the subarachnoidal lymph
space.

We have therefore shown that the cells of the optic stalk
perform the following three functions:

Ist. They conduct the nerve-fibres, which, in their turn,
resolve the constitution of the cells of the stalk, so that they—

2nd. Provide the nerve-fibres with asupporting framework
which—

8rd. Provides the whole interior of the optic nerve with an
elaborate system of minute lymph channels.

ALPHABETICAL List oF LITERATURE REFERRED TO.

1. Assurton, R.—‘On the Development of the Optic Nerve of Vertebrates,
and the Choroidal Fissure of Embryonic Life,” ‘Quart. Journ. Micr.
Sci.,’ vol. 34, 1892.

2. Casat, Ramon y.—‘ Sur la fine structure du lobe optique des oiseaux, et
sur l’origine réele des nerfs optiques,” ‘ Monthly International Journal
of Anatomy and Physiology,’ vol. viii, parts 9 and 10, 1891.

3. Frorier, A.—“ Ueber die Entwickelung des Sehnerven,” ‘Anat. Anz.,’
vi Jahrg., 1891.

* Wntwickelung des Auges,” ‘ Handbuch der vergleichenden und
experimentellen Hntwickelungshlehre der Wirbeltiere,’” O. Hertwig,
Bd. ii, Abt. 2, Jena, 1905.

4, KripeL.— Ueber die Entwickelung des Selnerven,” ‘ Deutsch. Medizin
Wochenschr.,’ xv Jahrg., 1889.

5. Kouiiker.— Zur Entwickelung des Auges und Geruchsorgans mensch-
licher Embryonen,” ‘ Festchrift der Universitat Ziirich dargebracht,’
Wiirzburg, 1883.

3a.

ORIGIN OF THE TRABECULAE OF THE OPTIC NERVE. 491

Ko.urker.—“ Entwickelungsgeschichte des Menschen und des hoheren
Thiere,”’ Leipzic, 1879.
6. Mier, W.—‘‘ Ueber die Stammesentwickelung des Sehorgans der
Wirbeltiere,”’ Leipzic, 1874.
7. Roprnson, A.—‘‘On the Formation and Structure of the Optic Nerve,
and its relation to the Optic Stalk,” ‘Journ. Anat. and Physiology,’
vol. xxx, April, 1896.

8. RypeEr, J. AA—‘‘ Development of the Eye,” ‘System of Diseases of the
Kye,’ Norris and Oliver, vol. i, London and Philadelphia, 1897.

9. StupnicKa, F. K.—‘‘ Untersuchungen iiber den Bau des Selhnerven,”’
‘Jenaische Zeitschr. Naturw.,’ Bd. xxxi, 1898.

EXPLANATION OF PLATES 26 anp 27,

Illustrating Mr. J. T. Gradon’s paper, “ Researches on the
Origin and Development of the Epiblastic Trabecule
and the Pial Sheath of the Optic Nerve of the Frog.”

ALPHABETICAL List oF REFERENCE LETTERS FOR ALL THE FIGURES.

Br. Brain. c.c. Cartilage of the cranium. c.c. Connective tissue cells
between the ophthalmic artery and the optic nerve. §. Eye. e. Outline of
pigmented epithelium of retina. e.m. Membrana limitans externa. 7.m. Mem-
brana limitans interna. /. Lymph-corpuscle. /.c. Lymph channel. /.s. Lumen
of optic stalk. m.a. Mesoblast of the arachnoid sheath. 2, Optic nerve
fibres. .s. Space occupied by nerve fibres, not represented. 0.z.c. Interior
capillary of optic nerve. p.c. Capillary of pial sheath. p.p. Pigmented
protoplasm. p.s, Pial sheath. sa.s, Subarachnoid space. ¢r./. longitudinal,
tr.o. oblique, ¢r.¢. transverse trabecula. y. Granules of food-yolk.

Fixing. agent: Aceto-corr. subl. Stain: Borax carmine + picro-indigo-
carm.

All the figures have been drawn with the Abbe camera.

The terms transverse and longitudinal apply to the optic stalk.

Figs. 1 to $ are transverse sections taken from tadpoles, the lengths of
which are given below.

They are all taken from the distal fourth of the stalk, except Fig. 2, which
is taken from the proximal fourth.

They show the gradual formation of the transverse trabecule and the pial
sheath.

The nerve fibres have only been represented in some of the figures, but they
will be understood to occupy the spaces between the trabeculae,

492 J. T. GRADON.

The resting nuclei of the stalk, except in Fig. 16, have been outlined only,
and all the nuclei of cells entering into the formation of the pial sheath have
been shaded. :

The top of the page represents “ dorsal.”

. PLATE 26.

Fic. 1.—4°5 mm. xX 800. (See previous page.)

Fic. 2.—6 mm. x 800.

Fig. 3.—8°5 mm. x 800.

Fic. 4.—ll mm. x 800.

Fie.5.—15 mm. x 800.

Fic. 6.—21 mm. x 800.

Fie. 7.—27 mm. x 800.

Fic. 8.—32 mm. The nerve fibres and lymph channels are shown in part
of the drawing only. x 500.

Fie. 9.—Oblique transverse section, from the same series as Fig. 6, taken
close to the brain. x 800.

Fic. 10.—From a transverse section of the optic nerve of an adult frog.
Taken 160 » from the eye. Only a sector has been filled in with the trans-
verse trabecule. x 500.

Fic. 11.—Horizontal section from an 8°5 mm. tadpole; showing fibrils of
the transverse, oblique, and longitudinal trabecule. x 800,

PLATE 27.

Fic. 12.—Longitudinal section, from a 15 mm. tadpole. Taken near the
brain. x 800.

Fig. 13.—Peripheral longitudinal section, from a 29 mm. tadpole. Taken
midway between the eye and the brain. x 800.

Fig. 14.—Longitudinal section, from a 32 mm. tadpole. x 800.

Fic. 15.—Longitudinal section, from a 14 day embryo mouse. xX 500.

Fic. 16.—Longitudinal section, from a developing trout. Body length not
known; length of stalk *5 mm. xX 500.

Fic. 17.—Transverse section from a 33 mm. dog-fish. The two large
spaces between the pial sheath and the membrana limitans externa, here
represented by the epiblastic trabecule, are due to sectionising, and they
show that the pial sheath forms a separate layer around the optic nerve.
x 500.

Fic. 18.—From a transverse section of the optic nerve of an 8-day chick.
x 800,

PIROPLASMA MURIS. 493

Piroplasma muris, Fant., from the Blood of
the White Rat, with Remarks on the
Genus Piroplasma.

By

H. B. Fantham, B.Sc.Lond., A.R.C.S.,

Derby Research Scholar, University College, London; and Demonstrator in
Biology, St. Mary’s Hospital Medical School.

With Plate 28.

ContTENTS.
PAGE
I. Introductory . ‘ : : . 493
II. Technique ‘ : : . 495
III. Occurrence of the Parasite in the White Rat . 496
IV. Morphology of the Parasite : : . 497
V. Note on Piroplasmosis in the White Rat. . 502
VI. Systematic; the genus Piroplasma ; . 503
VII. Summary of Results , : : . 508
VIII. Literature. : ‘ ‘ el
TX. Explanation of Plate. : : preoila

I. IntropuctTory.

Amone the Hemosporidia few genera are of greater interest
to-day than Piroplasma. Its complete life-cycle is still
unknown, yet “ piroplasmosis”! is a dreaded malady which
attacks many mammals, including man. In 1893 Smith and
Kilborne (46) published their epoch-marking monograph on
P. bigeminum, the parasite of Texas fever in cattle.
According to Koch, this ranks as one of the three great dis-

We owe this useful term to Ligniéres,

4.94, H. B. FANTHAM.

coveries in the etiology of protozoal diseases, the other two
being the discovery of the human malarial parasite by Laveran
in 1882, and the Trypanosome of ‘‘Nagana” by Bruce in
1895.

Since 1893 other species of Piroplasma have been notified
in horses, dogs, and sheep in various parts of the world.
Recently Wilson and Chowning (50) have described P.
hominis (Manson), the pathogenic agent of Rocky Mountain
or Spotted Fever, while the parasite of ‘ kala-azar ”’ has been
referred by Laveran and Mesnil to this genus.

Some months ago I had three white rats (Mus rattus,
albino variety) affected with ulcerations on the ears and tail,
and alopecia. On further examination all of these white rats
were found to be suffering from piroplasmosis; one died
almost immediately, another lived but a short time, while the
third and last one died towards the end of November, 1905.
Unfortunately pressure of work precluded my devoting much
time to the examination of these rodents when they first came
into my possession, and I was only able to give undivided
attention to the last one, having, perforce, to be content
with a partial examination of the others. Attempts at mocu-
lation of infected blood from diseased into healthy white rats
were unsuccessful, and the strain has thus been unfortunately
lost. Under these circumstances, and in view of my non-
success, up to the present, in procuring other white rats
suffering from piroplasmosis, I have thought it might be of
interest to publish my results on the morphology of the para-
site, hoping later on to continue my researches, if possible,
on fresh material.

Preparations of the blood of infected white rats were ex-
hibited by me before the Zoological Society of London on
December 12th, 1905 (12). I then proposed for the parasite .
the specific name of muris,! from its habitat. I would then

1 It might, perhaps, be urged that ‘‘muris” at once suggests “ mouse,”
whereas the parasite occurs in ‘‘rats.”” However, I have followed the well-
established custom of naming the species of the parasite after the genitive of
the generic name of the host.

PIROPLASMA MURIS. 495

designate the parasite described in this paper as Piro-
plasma muris.

II. TecHNIQueE.

Fresh blood-films, obtained from the tail or ear, were ex-
amined from time to time, and in some cases a few red blood-
corpuscles exhibited, in their interior, small, bright, usually
ovoid bodies with dark contour, which occasionally showed
slight motility. The change of position inside the corpuscle
was usually from near the periphery towards the centre and
back again to the periphery, and it sometimes set up slight
rotation of the corpuscle. No especial change of shape was
noticed. ‘I'he small size of the intra-corpuscular or endo-
globular bodies increased the difficulty of observation in
freshly-drawn blood.

Most of the observations hereafter recorded were, however,
made on fixed and stained preparations of thin smears or films
of blood from the peripheral circulation, and on smears made
from certain of the internal organs as soon after death as
possible. Many of these were stained for some time with
various modifications of the Romanowsky method, especially a
combination of the methods of Laveran and Plimmer, using
Bleu Borrel, erythrosin, and tannin orange. I also obtained
good results with Leishman’s stain, which possesses the added
advantage of simplicity ; and with an adaptation of Giemsa’s
stain, using a 1 per cent. aqueous solution of azur u1, together
with a 0'1 per cent. aqueous solution of erythrosin,! mixed on
the slide after fixation with pure methylalcohol. Leishman’s
stain is useful in that 1t imparts to the erythrocyte-cytoplasm
a pink colour, which affords—quite easily—a contrast with
the blue cytoplasm and red chromatin of the enclosed para-

1 The respective quantities were:—1 drop of azur ii to 2 drops of ery-
throsin, diluted with 5 to 8 drops of distilled water. The preparation may
afterwards be stained with a dilute solution of tannin orange. The azur ii
may be used first, followed for a short time by erythrosin, with good results.

VOL. 50, PART 3.—NEW SERIES. 35

496 H. B. FANTHAM.

site. Laveran’s method gives the best results when successful,
though it is a little difficult to manipulate and somewhat un-
certain. Methyl alcohol fixes more sharply and rapidly than
ethyl (absolute) alcohol.

Other blood-films and smears of organs were fixed with a
mixture of mercuric chloride (two parts) and absolute alcohol
(one part), or with osmic acid, and then stained with a dilute
acidulated solution of Delafield’s haematoxylin followed by
eosin. The staining is slow, at least twenty-four hours being
necessary, but the fixation is superior to that obtained with
alcohol alone.

I also used, on a few occasions, a slightly alkaline solution
of methylene blue, after fixation with absolute alcohol. Por-
tions of liver, kidney, and spleen were fixed in formalin,
embedded in paraffin and sectionised. Affixed to slides, sec-
tions of these organs were placed in a dilute aqueous solution
of Leishman’s stain for about twelve hours (vide Christophers
[4] and Graham Smith [14]), then treated with a dilute solu-
tion of acetic acid (1 volume of acid to 500 of water) for a
short time till pink, washed in distilled water, rapidly dried,
then immediately moistened with xylol and mounted in balsam.
Unfortunately, formalin is not a satisfactory fixative for these
tissues, causing shrinkage.

Most of the observations hereafter recorded were made on
material stained by the lLaveran-Plimmer or Leishman
methods.

III. OccurRENcE oF THE PARASITE IN THE Watte Rat.

The parasites were rarely met with in the peripheral circu-
lation, judging from observations on blood-smears from the
tail of infected rats or from scrapings of the ulcers on the
ears. On an average about 1 per cent., or rather less, of
these erythrocytes were infected.

In smears of the internal organs—as the liver, kidneys,
spleen, bone-marrow, lung, heart-muscle, and brain—the

PIROPLASMA MUORIS. 4.97

parasites were more numerous. ‘They were most plentiful in
red corpuscles occurring in the capillaries of these organs, as
seen in sections, especially in the kidneys, liver, and spleen
(fig. 23, dilated capillary of liver).

Extra-corpuscular stages of the parasite, free in the blood-
plasma, occurred in groups, probably resulting from the dis-
integration of the corpuscle host (fig. 21). Such groups were
noticed sometimes in the peripheral circulation, more fre-
quently in spleen blood.

Leucocytes were relatively rather more numerous than
usual in the smears above mentioned, especially in those
taken from the sores on the ears.

IV. MorpHotoay oF THE PARASITE.

In the red-blood corpuscles of the mammalian host ovoid
or pear-shaped organisms! were noticed, which, after ade-
quate staining by modifications of the Romanowsky method,
showed a definite contour, blue cytoplasm, and a red or
purple chromatin body, without any trace of pigment. Such
characters are diagnostic of the genus Piroplasma.

These endoglobular bodies may be centrally placed in the
blood-corpuscle, but more usually they are rather peripheral
in position. They represent the trophozoite stage of the
parasite, and may occur either singly (figs. 1 to 3) or in
pairs (figs. 6 to 9) within the erythrocytes. Double, and
even multiple, infection may be observed, as a dividing
trophozoite together with a single pyriform trophozoite, or
two trophozoites each in process of division (fig. 10), may
be seen simultaneously inside blood corpuscles. In the
smaller, and apparently younger forms, the internal chro-
matin body is somewhat flattened and peripheral in position
(fig. 1). The chromatin body (‘‘ nucleus” or ‘‘ karyosome”
of various authors) is, indeed, seldom quite central in posi-

' The chromation of the parasite stains purple with azur ii alone; this is

a test for a parasite (cf. Koch and Theiler), in contra-distinction from an
artefact.

498 H. B. FANTHAM.

tion, but usually polar (figs. 2 and 3), that is, nearer to the
rounded or blunt end of the pyriform trophozoite.

The smaller ovoid forms of the parasite measure 0°5 pu to
1‘5in diameter, while the pear-shaped forms are from 2 u
to 3m long, and from 1 pw to 1'5y broad.

Four pyriform bodies are sometimes seen in the red
corpuscles of the peripheral circulation (fig. 10), but rarely
more than four. In the spleen six and eight small pyriform
bodies may occur in an erythrocyte (fig. 13). In some cases
in the spleen they are rather more irregular in shape than
strictly pyriform, affording examples of the so-called “ame-
boid ” trophozoites (figs. 14 and 18) known to occur in other
species of Piroplasma, and first described by Piana and
Galli-Valerio in the case of P. canis.

In one case a vermiform trophozoite was noticed with a
chromatic appendage. This had been fixed and stained
towards the conclusion of the act of entering a red blood
corpuscle (fig. 19). Flagellate forms of P. canis have been
described by Bowhill and Le Doux (2), of P. equi by
Bowhill (1), and of the Piroplasmata of cattle by Lignieres
(30) and others. The suggestion that such flagellate forms
may possibly be microgametes seems to me premature and
doubtful in the present state of our knowledge, as the
“flagella”? described by Bowhill and Le Doux are beaded,
and may really be only pseudopodia.

Another vermiform or gregariniform, but entirely intra-
corpuscular, trophozoite, containing a chromatic dot attached
to an irregular rod-shaped portion of chromatin, is shown in
fio. 16.

The cytoplasm of P. canis is described by Nuttall (41) as
“vacuolated or trabecular.’ In the case of the smaller
organism, P. muris, the protoplasm is hyaline, and appa-
rently finely granular, though it is very difficult to observe
the finer structural details of so small an object through the
wall of the enclosing blood corpuscle.

A clear zone of protoplasm often occurs around the
chromatin body in the case of some of the large trophozoites

PIROPLASMA MURIS. 499

(fig. 17), while a distinct vacuole, more or less polar in
position, may occur in other forms (figs. 1, 2, and 5).

The outer border of some of the larger forms of the
parasite often takes up the stain more intensely than the
more central cytoplasm, and so appears of a distinctly blue
tint after Romanowsky staining.

Usually there is only one chromatic dot in each ovoid,
pyriform, or amceboid body (figs. 2,3,and 17). Occasionally
two chromatic dots are seen (figs. 4 and 15, note also fig. 18),
while in one case, as already mentioned, there was a chromatic
appendage, somewhat flagellum-lke, protruding from the
body of the parasite, and even outside the erythrocyte host
(fig. 19). The chromatin body averages 0°3 u to O°5 fe in
diameter, and may be irregular in outline.

A few remarks seem necessary respecting the chromatin
of the trophozoites of P. muris and other Piroplasmata. In
view of the recent researches of Schaudinn and others on
the “‘ vegetative ” and “ reproductive” differentiations of the
chromatin in parasitic Protozoa one may well hesitate
nowadays to use indiscriminately the terms “ nucleus” and
“karyosome.”! lLaveran (24), in 1901, used the term
“karyosome” to designate the chromatic body of the tro-
phozoite of P. equi. However, since all the species of
Piroplasma are comparatively small, it is difficult, in the
present state of our knowledge, to be quite precise in naming
the chromatin body or dot occurring in a trophozoite of
Piroplasma, and it would seem best simply to refer to such
structures as ‘chromatin bodies ” or “ chromatic dots.”

No bacillary (25) or rod-like forms were seen, types which
are characteristic of P. bigeminum in the blood of immune
Bovines (48), and are also common in the case of P. parvum
(48, 49).

The mode of multiplication? of the trophozoite is, as

1 Siedlecki has lately (Oct., 05) written on the ‘Significance of the
Karyosome,” putting forward somewhat different views (‘ Bull. Acad. Se.
Cracovie,’ 1905, No. 8, pp. 559-81).

? This mode of multiplication, which is endogenous, is in this genus simple,

500 H. B. FANTHAM.

in other Piroplasmata, by the primitive process of binary
fission (figs. 6 and 8). Hach “ merozoite” or daughter
trophozoite formed from a dividing trophozoite (‘‘ schizont”’)
may, in turn, similarly divide. I am sorry that, in view of
the smallness of this intra-corpuscular parasite, | am unable
to give cytological details of the process of binary fission
other than observing that the “‘chromatin body” of the
ellipsoidal trophozoite (schizont) divides into two parts,
arranged at the poles (fig. 5), and that the twin merozoites,
when longitudinally constricted apart (fig. 6), remain
attached for a time by their pointed ends (fig. 8). How-
ever, in the case of erythrocytes enclosing several parasites
the numbers so enclosed are usually multiples of two (vide
figs. 10 and 13, but exception fig. 12).

I have not yet seen examples of multiple fission forming
“rosette ” stages, as figured by Laveran (24), in the case of
P. equi, or “cross” stages of four, as mentioned by Koch
(21) in the case of P. parvum.

Large extra-corpuscular, sausage-shaped “ gamete” forms
have been figured by Nuttall and Graham-Smith (41, Pl. 9,
figs. 59—62) in the case of P. canis. Stephens and
Christophers (47, Pl. 3, fig. 10), too, have figured a pair of
large intra-corpuscular forms, containing much chromatin, as
possible ‘‘ gamete” forms of P. bovis. In each case it is
only tentatively suggested that such are “ gametes ” (perhaps
more strictly “ gametocytes ”), but the correctness of these
and similar interpretations has not yet been established.
(Cf. Minchin [87, pp. 269-70] on Hunt’s “crescents,”
Ligniére’s “‘ gametocytes” and Doflein’s views, where he
remarks :—‘‘ The relations of the various phases hitherto
observed, and their true role in the life-cycle is, at present,

so it might be considered unnecessary to use the term “schizogony” for
simple binary fission, or ‘“schizont’”’ for a trophozoite so dividing, as the
trophozoite is not here sporulating into many daughter forms, but usually
into two only, which themselves, perhaps, need not be specially termed
“ merozoites.” Another view, however, is that schizogony is here witnessed
in its simplest form, and this seems the better view to take.

PIROPLASMA MURIS. 501

. . . purely conjectural.”) Nor do these easily fit in with
Koch’s recent researches on P. bigeminum in the tick (21).

In sections of the internal organs, as the liver and kidneys,
infected corpuscles are seen to be numerous in the capillaries
(fig. 23). Endogenous reproduction of the parasite would
appear to be especially prevalent in these parts.

Regarding the sporogony of P. muris, I have, unfortu-
nately, no observations, nor, indeed, have any well-authenti-
cated details been published by any observer on the exogenous
reproduction of any species of Piroplasma, with the possible
exception of a preliminary note by Lingard and Jennings
(33) on Piroplasmata in Mammals and even birds and
lizards (!), wherein figures of what purport to be sporogony
(“sexual”) stages are given. ‘This account by Lingard and
Jennings, although avowedly preliminary, is, unfortunately,
somewhat condensed and disconnected, and so not very clear.
Information on the sporogony of many, indeed most, of the
Heemosporidia is still wanting.

With respect to the dissemination of piroplasmosis among
white rats little can at present be stated. I carefully searched
the infected rodents for ticks, but found none. Lice were
abundant on one rat, and fleas also occurred, but no clearly-
defined further stages of the Piroplasma were seen in the
internal organs of these insects. The ‘ intermediate” (in-
vertebrate) host is probably a tick, as has been shown in all
other cases of true piroplasmosis hitherto examined. The
three rats infected with P. muris and discussed in this
memoir were, I understand, obtained from two separate
sources in the EHast-end of London, but as to the manner
in which they became infected in the first instance I have no
information.

Very few examples of phagocytosis of the parasites were
observed. One or two instances of apparently free parasites
being engulfed by leucocytes were noticed (fig. 22), but no
cases of leucocytes actually destroying infected erythrocytes
could be detected. The paucity of examples of phagocytosis
has been emphasised by other observers of Heemosporidia.

502 H. B. FANTHAM.

V. Nore on PIROPLASMOSIS IN THE WuHiTeE Rat.

As Iam not a medical man I would crave indulgence for
deficiencies in the diagnosis or setting forth of symptoms in
the following outlines of piroplasmosis as exhibited by white
rats.

These rodents, in whose blood P. muris was found, at
once attracted attention by the presence of pronounced ulcers
on the ears. There were also smaller sores on the tail, and
sometimes slight ones on bald patches on the body, and in
one case on slight swellings near the anus and the snout.
The bald patches, from which the fur had quite disappeared,
were variously distributed in the different rats; there was a
marked patch devoid of fur on the necks of the rodents’
examined. The body temperature of the rats, determined
per rectum, was at times above the normal (for example,
readings of 102° F. and 101°6° F. were obtained), and
indicated an irregular fever. After death in one case a
yellowish discoloration of the skin and some tissues was
noticed, apparently due to biliary fever. There was also
slight anemia, and a relative increase in the number of
leucocytes, with enlarged spleen. Before death the rats
became emaciated, and showed gradually increasing loss of
appetite. Some solid bile was found in the bile-ducts, and in
one case the urine was dark coloured (hemoglobinuria).

From the comparatively long time two of the infected rats
lived while suffering from the disease (two to five months),
and the fewness of the parasites in the peripheral circula-
tion, the cases perhaps approached the chronic type.

A few nucleated red cells occurred in the blood, but none
of these were noticed to be infected, while erythrocytes
containing many parasites were sometimes slightly enlarged
(that is, greater than 7 in diameter), and when stained
were pale in colour, sometimes approaching a slight blue
tint after the use of Leishman’s stain.

It has already been mentioned that the parasites were

PIROPLASMA MURIS. 503

more numerous in the internal organs, such as the liver
spleen, kidneys, heart-muscle, lung, and bone-marrow, espe-
cially in the capillaries of these, which were enlarged. In
the liver, and to some extent in the kidneys, the outlines of
the cells were not easily apparent or were even broken down,
the cytoplasm was ill-defined, and the nuclei of the hepatic
cells were often hypertrophied (fig. 23).

Probably most, if not all, of the symptoms outlined above
are those of piroplasmosis, judging from published accounts
of cases of the disease in other mammals.

It would be interesting to determine if the disease is
periodic ; possibly it occurs in the spring or early summer.
Information is also required, as already remarked, regarding
the invertebrate host, probably a blood-sucking Arthropod,
which may be concerned in the spread of the disease.

Further, the disease may not be strictly limited to the
white rat, that is, the albino variety, but will perhaps be
found in black (Mus rattus) and brown (M. decumanus)
rats.

Cultures of infected blood, made by adding sodium citrate
and a little citric acid to freshly drawn blood, showed no
further stages or development of the parasite, even after
several days.

VI. SystTemMatic; THE GENUS PIROPLASMA.

Summarising briefly some of the more important charac-
teristics of the parasite! already described, we notice the
usually ovoid or pyriform shape of the trophozoite, generally
with a single well-marked chromatin dot, multiplication by
binary fission into two merozoites, the absence of melanin
pigment, and the cytozoic habitat within a red blood-cor-
puscle during the endogenous stages. From these features
it may be concluded that the parasite is a Hemosporidian,

' The sizes of the various forms of the parasite are given on pp. 498 and 508.

504. H. B. FANTHAM.

belonging to the order Acystosporea! on account of its
simple body form, and to the genus Piroplasma on account
of its ovoid or pyriform shape and simple fission in schizo-
gony. Since it occurs in the white rat and is apparently con-
fined to rats, I have proposed the new specific name muris.

The other well-authenticated species of Piroplasma’
(Patton, 1895), as mentioned by Laveran (28) and others,
are—

(1) P. bigeminum (Smith and Kilborne, 1893), the parasite
of Texas fever in the Bovide, which has since been observed
in most parts of the world. This species is sometimes called
P. bovis (as by Nuttall [40], and Stephens and Christophers
[47]). The correct name of the species is doubtful. Judging
from the illustrations of the relative sizes of the parasite
and its corpuscle host, as figured by Smith ana Kilborne
(46) in cases from Texas, and by Stephens and Christophers
(47) in cases from Madras, there would seem to be more
than one species. Ligniéres (1900) also thought there were
two species of P. bigeminum (P. bovis) in Argentina, and
Nuttall (40) has emphasised this point. Some authorities,
again, apply the name P. bovis to the parasite of bovine
hemoglobinuria in Kurope, spread by Ixodes reduvius,
separating it from P. bigeminum, which latter name is
restricted to the parasite of Texas Fever (Tristeza, Red-
water).

(2) P. parvum, separated by Theiler (48, 49) in 1904, as a
distinct species from the former, and found in Bovide suffer-
ing from East Coast Fever (Tropical Bovine Piroplasmosis,
“ Rhodesian Redwater ” [16, 18, 19, 20]). It also occurs in
Transcaucasia (11).

(3) P. canis (Piana and Galli-Valerio, 1895), occurring in

1 The distinction between the sub-orders Acystosporea and Hemosporea
is not now so sharp as formerly considered, since the discovery of intermediate
hosts in the case of several Hamogregarines, and the finding of Hemo-
gregarina gerbilli by Christophers in a mammal.

2 The synonymy of the generic name “ Piroplasma” is given by Minchin
(37, p. 269). Probably the strictly correct name, by priority, is “ Babesia,”
though the name “ Piroplasma” is almost universally used.

PIROPLASMA MURIS. 505

“malignant jaundice” in dogs in South Africa, India, Sene-
gambia, Italy, and France (13, 34, 39, 40, 41, 51).

(4) P. ovis (Starcovici, 1892) in sheep in Hungary, Rou-
mania, Italy, and Germany (“‘ Carceag’”’ [38]).

(5) P. equi (Laveran, 1899), the pathogenic agent of biliary
fever in horses in South Africa and Italy. The same or a
closely allied species occurs in donkeys in South Africa
(Dale [8]).

(6) A species, apparently unnamed, has been described
by P. H. Ross (44) in 1904 from monkeys (Cercopithecus)
in Africa.

(7) P. hominis (specific name due to Manson in 1903,
though the parasite was first described by Wilson and
Chowning) in cases of “Tick ” or “ Spotted Fever” in man
in the Rocky Mountains.

(8) P. donovani (Laveran and Mesnil [27, 28]), for the
Leishman-Donovan bodies (9, 10) found in cases of kala-azar
and Oriental sore in man in India, Arabia, China, Egypt,
and Tunisia. There is doubt as to the accuracy of placing
these bodies in the genus Piroplasma (see below).

(9) Lithe (33a, p. 201) mentions a little known, and appa-
rently unnamed, species of Piroplasma found by Ziemann
(53) in the Cameroons in the blood of sheep, goats, horses,
and donkeys (‘ Tier-Malaria ”’).

P. muris, as I have found and measured it, seems dis-
tinctly smaller than P. canis, and apparently slightly smaller
than the type species, P. bigeminum, though the sizes of
the latter, as given by different observers from various
localities and cases, vary somewhat. Indeed, this variation
in size seems to apply to many of the species of Piroplasma,
according to case, locality, and observer, perhaps due to the
smallness of the parasite and consequent difficulty in precise
measurement, as wellas to differences in fixation and staining.

The genus Piroplasma stands distinctly apart from the
other Hemosporidia. It may be that the Hemosporidia, as
at present understood, is really a heterogeneous group, which
will ultimately have to be broken up. Laveran (22), one of

506 H. B. FANTHAM.

the founders of this group of the Sporozoa, divided it in
1901 into three great genera, namely, Hemameba,
Hemogregarina, and Piroplasma. Some authorities,
although allowing the correctness of the basis of this arrange-
ment, would recognise more genera (vide Schaudinn’s
monograph on the ‘ Tertian Parasite” and Minchin [87,
p- 265]). However, the classification and nomenclature of
the Hemosporidia is still in a confused state, indeed few
groups of the animal kingdom are so involved from this
point of view. Laveran, in a recent essay (238), returns to
this matter, and reiterates his former classification, giving
also a list of recognised species to date (October, 1905).
The species which Laveran enumerates under the genus
Piroplasma have just been set forth above, and, in addition,
P. H. Ross’s species from Cercopithecus (44). The species
P. donovani, for the Leishman-Donovan bodies of kala-
azar, 1s open to discussion.

To consider this point (the systematic position of the
Leishman-Donovan bodies) at length is hardly within the
purview of this paper. Some of the more important debat-
able points may, however, be very briefly set forth, to show
the connection, or otherwise, of these bodies with the genus
Piroplasma.

The Leishman-Donovan bodies are endocellular in habitat,
occurring in spleen cells, endothelial cells, leucocytes, and
possibly in erythrocytes. Their occurrence in the latter
(erythrocytes) is not now generally held, and the first obser-
vations of them in this position have been variously inter-
preted. These bodies are piroplasmoid in shape, but are
bounded by a perfectly definite external layer, more marked
and consistent than in a Piroplasma, and possess two well-
marked chromatic bodies, differentiated in character, as well
as an internal “tail.” In view of these differences Ross (45)
has proposed for the parasites found in cases of kala-azar a
new and separate genus Leishmania,!

1 Containing two species, L. donovani (from Kala-azar) and LL.
tropica (from Delhi boil or Oriental sore).

PIROPLASMA MURIS. 507

Rogers (42, 43) and others (4, 5,6) have obtained flagellated
organisms from cultures of the Leishman-Donovan bodies.
These flagellates are obtained in an essentially artificial
medium, namely, by mixing infected spleen blood with sodium
citrate and slightly acidifying with citric acid. In nature
flagellate stages of these bodies, probably similar in character
to those obtained in .artificial media, might occur in the ali-
mentary canal of an Arthropod, but have not as yet been
observed. It would seem, then, a little premature to refer
these flagellates, developed in citrate cultures, to the genus
Herpetomonas, as the ‘‘Herpetomonas of kala-azar,”
Rogers (48).!

Apparently flagellates have not yet been obtained from
cultures of the similar Cunningham-Wright bodies of Oriental
sore (7, 35, 52).

A true Piroplasma possesses only one” chromatin body, and
no typically flagellated stages are yet known in its life-
history.

Koch (21) has recently published some short, but stimu-
lating observations, on stages of P. bigeminum in the gut
of ticks just gorged with infected bovine blood, and in tick
eges, observed in German Hast Africa. He states that the
Piroplasmata in blood-corpuscles taken into the alimentary
canal of ticks already, or very soon, show division of their
chromatin into two, and that radial processes are developed
from the parasite after it leaves the blood-corpuscle. Similar
radiate forms are mentioned in the case of P. parvum.
Later, copulation stages (probable zygotes) of the Piroplasmata
are seen in the alimentary tract of adult ticks. Large pear-
shaped forms of the parasite are described from tick eggs. I
have myself seen similar forms in eggs of ticks infected with
P. canis. There are no recorded observations of the para-
sites in larvee and nymphs of ticks.

1 Unfortunately Rogers, in his paper, writes of the “ group Hepatomonas,”
apparently in mistake for the genus Herpetomonas.

2 See Addendum for remarks on Lihe’s researches, and the presence of a
blepharoplast in P. canis.

508 H. B. FANTHAM.

Graham-Smith (15) has recently (October, 1905) recorded
an intra-corpuscular parasite from the erythrocytes of moles.
Although at first thought to be piroplasma-like, yet apparently
the parasites do not belong to the genus Piroplasma,
according to their discoverer, but are “longer or shorter rods
of irregular shape” occasionally even devoid of chromatin.
Graham-Smith does not appear to have-named them yet.

It is interesting to note that a rodent, Spermophilus
columbianus, is said to be concerned in the spread of human
tick fever in the Rocky Mountains. Wilson and Chowning
(50) give reasons for thinking that this Spermophilus is a
third host of Piroplasma hominis, and consider that it
is really the normal or true host of the parasite. In the
Columbian Spermophile the P. hominis is non-pathogenic,
and the human subject would seem to be not the true host
but one in which the parasite lives with, perhaps, some diffi-
culty, and wherein it consequently sets up pathogenic reac-
tions resulting in human “ spotted ” or “ tick fever.” With
this may be compared the action of Trypanosoma bruce},
which is non-pathogenic in the “ wild game ” of South Africa,
its true hosts, but is pathogenic or hurtful to the imported
horses not indigenous to the country ; similarly T. lewisi is
non-pathogenic in the rat, which is apparently its true or
natural host.

VII. Summary or Resutts.

The parasite described in this memoir occurs in the blood
and certain organs, as the liver, spleen, kidneys, lung, heart-
muscle, and bone-marrow of white rats, three of which came
under my observation, but only one of them lived long enough
to allow of continued study, and that only for a comparatively
short time, too short to allow of observation on the methods
of cross-infection.

The parasites are intra-corpuscular in habitat, occurring in
the erythrocytes or hematids of the host, and belong to the

PIROPLASMA MURIS. 509

order Heemosporidia, of the class Sporozoa. They were
not found to be numerous in the peripheral circulation, but
occurred in greater numbers in the internal organs above
mentioned. f

The trophozoites are ovoid (fig. 3) or pear-shaped (figs.
1,2, and 7), the former varying from 0°5y to 15min dia-
meter, the latter being from 2 to 3y long and 14 to lbp
broad, and devoid of melanin pigment (fig. 8). There is
usually only one chromatin body or dot which may be peri-
pherally or centrally placed, more usually near one end. A
clear zone of protoplasm often surrounds this chromatin
body (fig. 17), and a vacuole (fig. 5) may occur in the cyto-
plasm of the parasite. Pairs of trophozoites often occur in a
host-corpuscle, but single trophozoites are also not infrequent.

Some so-called “amoeboid” trophozoites (fig. 14) were
seen in the spleen.

Endogenous multiplication takes place inside the rat’s red
blood-corpuscle by simple fission. Double infection (fig. 10)
of a blood-corpuscle may occur, while free ovoid forms of the
parasite have also been seen in the plasma (figs. 20, 21).
Sometimes four parasites may be found in a corpuscle of the
peripheral circulation, and as many as six or eight in cor-
puscles in the spleen (figs. 13, 14).

Some of the pathological effects in the white rats, very
probably directly due to this parasite, were ulcers on the ears,
alopecia, emaciation, anzemia, biliary fever, enlarged spleen,
etc., and in each case death resulted.

From the foregoing characteristics the parasite may be
placed in the genus Piroplasma. A short account of my
exhibit of this parasite before the Zoological Society of
London appeared in the ‘ Proc. Zool. Soc.,’ 1905 (12), where
I proposed the new specific name of muris, from its occur-
rence in a member of the Muride. I would, then, call this
parasite Piroplasma muris.

The appended list of literature cannot be set forward as in
any sense complete. ‘lo compile a complete list would need

510 H. B. FANTHAM.

long searching of zoological, medical, veterinary, and even
general scientific journals, taking full advantage of the
several catalogues of scientific literature now published, and
even then allowing a margin for the rapid growth of the
literature on this and allied subjects. A full bibliography
of Piroplasma canis up to 1904 is given by Nuttall (40),
together with references to many papers on other Piroplas-
mata. A complete list of papers relating to P. donovani,
if it really be a Piroplasma, would also be difficult to
compile, and even more difficult to collect and read. I
only enumerate the more important papers relating to the
systematic position of the Leishman-Donovan bodies. Since
the intermediate host of P. muris has not yet been deter-
mined, I have not given many references to literature on
ticks. Nevertheless, I hope that in the following I have not
omitted any important papers on Piroplasma, although I
have only enumerated the papers more or less directly referred
to in the text.

In conclusion, I would take this opportunity of thanking
Professor Minchin for the pleasure and help I have derived
from attending his recent course of lectures on the ‘ Parasitic
Protozoa,’ which has been of use to me in writing the latter
part of this paper, and for general help at all times.

March, 1906.

ADDENDUM.

Since writing the foregoing, there have appeared impor-
tant works on Piroplasma by Liihe (83a, 33b), wherein
the generic name of Babesia is preferred (see my footnote,
p. 504). Having worked recently on P. canis, the largest
species of Piroplasma, under a magnification of 3000 dia-
meters, Liihe states that the pyriform trophozoites only are
endoglobular, and that, in addition to the “ principal
nucleus,” there is a small chromatic dot nearer the pointed
end comparable to the blepharoplast of a Trypanosome (ef.

PIROPLASMA MURIS. 511

my figs. 4 and 15, also figs. 11 and 18). I have only been
able somewhat hurriedly to look over again some of my
preparations of this smaller species, P. muris, but without
obtaining any new observations. Liihe enumerates the
various species of Piroplasma hitherto recorded, and dis-
cusses them in detail. From him I have inserted Ziemann’s
parasites (58) as No. 9 in my list on p. 505. Space does not
admit of further discussion of Liihe’s valuable treatise
(33a) on the Heematozoa, which should be consulted in the
original.

June, 1906.

VIII. Rererences To LITERATURE.

1. Bown11, T. 1905.—‘‘ Equine Piroplasmosis, or ‘ Biliary Fever,’ ’’ ‘ Journ.
Hygiene,’ v, pp. 7-17, pls. i—iil.

2. Bowuiuu, T., and Le Dovux, C. A. 1904.—‘‘A Contribution to the
Study of Piroplasma canis—Malignant Jaundice of the Dog,”
‘Journ. Hygiene.,’ iv, pp. 217-18, pl xi.

8. Brucr, D. 1905.—‘*Stock Diseases of South Africa,’’ Presidential
Address to Physiological Section, Brit. Assoc.—‘ Nature,’ Ixxii,
pp. 496-503.

4. CuristorHers, S. R. 1904.—‘ On a Parasite found in Persons suffering
from Enlargement of the Spleen in India,” Preliminary Report, ‘Sci.
Mem. Officers Med. and San. Depts., Govt. of India,’ new series, No. 8,
17 pp., 2 pls.

5. CuristorHers, 8. R. 1904.—[Same title.] Second report, op. cit.,
No. 11, 21 pp., 2 pls.

6. CuristorHEers, 8. R. 1905.—[Same title.]. Third Report, op. cit.,
No. 15, 14 pp., 1 pl.

7. CunnincHam, D. D. 1885.—‘* On the Presence of Peculiar Parasitic
Organisms iu the Tissue of a Specimen of Delhi Boil,” ‘Sc. Mem. by
Med. Off., Army of India,’ pt. 1.

8. Dats, T. H. 1903.—‘ Pyroplasmosis of the Donkey,’ ‘Journ. Comp.
Path. and Therap.,’ xvi, pp. 312-19.

9. Donovan, C. 1904.—‘ Human Piroplasmosis,” ‘ Lancet,’ 1904, ii,
pp. 744-50, 1 coloured plate.

VOL. 50, PART 3,—NEW SERIES. 36

512 H. B. FANTHAM.

10. Donovan, C. 1905.—‘‘ Human Piroplasmosis,” ‘ Lancet,’ 1905, i,
pp. 155-6.

11. Dscuunxowsky, E., and Luus, J. 1904.—‘‘ Die Piroplasmosen der
Rinder,” ‘ Centrbl. Bakter.’ (1), xxxv, orig. pp. 486-92, 3 pls.

12. Fanruam, H. B. 1905.—“ Piroplasma muris, n. sp.,” Note on a
microscopic exhibit, ‘ Proc. Zool. Soc.,’ 1905, ii, p. 491.

18. GaLui-VaLERI0, B. 1904.—‘‘ Die Piroplasmose des Hundes,” ‘ Centrbl.
Bakter.’ (1), xxxiv, ref. pp. 867-77.

14, Granam-Smiru, G. 8. 1905.—‘ Canine Piroplasmosis, III, Morbid
Anatomy,” ‘Journ. Hygiene,’ v, pp. 250-67, 2 charts and 2 pls.

15. Granam-SmituH, G.S. 1905.—‘ A New Form of Parasite found in the
Red Blood Corpuscles of Moles,” ‘Journ. Hygiene,’ v, pp. 453-9, .
pls. 18, 14.

16. Gray, C. E., and Ropertson, W. 1902.—‘‘ Red Water in Rhodesia,”
‘ Acric. Journ., Cape Good Hope,’ xxi, pp. 435-58.

17. Hunt, J. S.—‘ Progress Report on the Reproductive Forms of the
Micro-organism of Tick Fever, etc.,’’ ‘Queensland Agric. Journ.,’ ii, 3,
pp. 211-20.

18. Hutcueon, D. 1903.—‘ Virulent Red Water in the Transvaal,” ‘ Agric.
Journ., Cape Good Hope,’ xxiii, pp. 39-60.

19. Kocu, R. 1903.—*The Rhodesian Cattle Disease,” ‘Agric. Journ.,
Cape Good Hope,’ xxiii, pp. 33-39.

20. Kocu, R. 1903.—‘‘On Rhodesian Red Water or African Coast Fever”
(4 reports), ‘Journ. Comp. Path. and Therap.,’ xvi, pp. 273 and 390.

21. Kocu, R. 1905.—* Vorlaufige Mitteilungen tber die Ergebnisse einer
Forschungsreise nach Ostafirika,” ‘Deutsch. med. Wochenschr.,’
No. 47, 15 pp.

21a. Kosset, H., Scntrz, Weper, and MigssnER. 1903.—‘‘ Ueber die
Hamoglobinurie der Rinder in Deutschland,” ‘ Arbeit. a. d. Kaiserl.
Gesundheitsamte,’ Bd. xx, 1, 77 pp., 3 pls.

22. Laveran, A. 1901.—* Essai de classification des Hématozoaires endo-
clobulaires,” ‘C. R. Soc. Biol.,’ lili, p. 798.

23. Laveran, A. 1905.—Hemocytozoa, essai de classification,’ ‘Bull.
Inst. Pasteur,’ iii, No. 20.

24. Laveran, A. 1901.—‘Contribution a l’etude de Piroplasma equi,”
‘C. R. Soe. Biol.,’ lili, p. 385.

25. Laveran, A. 1903.—‘‘Sur la Piroplasmose bovine bacilliforme,”
‘OC. R. Ac. Sei.,’ exxxvi, pp. 648-53, 18 text-figs.

26. Laveran, A. 1903.—‘ Au sujet du réle des tiques dans la propagation
des piroplasmoses,” ‘C. R. Soc. Biol.,’ lv, pp. 61-63.

PIROPLASMA MURIS. bis

27. LAvERAN, A., and Mesnit, F. 1903.—‘*Sur un Protozoaire nouveau
(Piroplasma donovani, Lav. et Mesn.) parasite d’une fiévre de
Pinde,” ‘C. R. Ac. Sci.,’ cxxxvii, pp. 957-61, 17 text-figs.

28. Laveran, A., and Mersnit, F. 1904.—‘* Nouvelle observations sur
Piroplasma donovani, Lav. et Mesn.,” ‘C. R. Ac. Sci.,’ exxxviii,
pp. 187-9.

29. Laveran, A., and Nicotute, M. 1899.—*‘ Contribution 4 l'étude de
Pyrosoma bigeminum,” ‘C. R. Soc. Biol.,’ li, pp. 748-51, 15 figs.

30. Licnikres, J. 1901.— Sur la ‘Tristeza,’ ‘Ann. Inst. Pasteur,’ xv,
pp. 121-8, pl. 6.

31. LieniERzEs, J. 1903.—“‘ La piroplasmose bovine,” ‘ Arch.. Parasit.,’ vii,
pp. 398-407, pl. 4.

32. Linearp, A. 1904,.—‘Can the ‘Piroplasma bigeminum’ find a

habitat in the human subject?” ‘Centrbl. Bakter.’ (1), xxxvi, orig.,
pp. 214-16, 1 pl.

83. Linearp, A., and Jennines, KH. 1904.—‘‘A Preliminary Note on a
Pyroplasmosis found in Man and in some of the Lower Animals,”
‘Ind. Med. Gaz.,’ xxxix, pp. 161-5, 3 pls.

33a. Linz, M. 1906.—* Die in Blute schmarotzenden Protozoen und ihre
nachsten Verwandten,” in Mense’s ‘ Handbuch der Tropenkrankheiten,’
Leipzig, Bd. iii, ‘‘ Babesia,” pp. 193-202, Taf. 8.

83b. Line, M. 1906.—**Zur Kenntnis von Bau und Entwicklung der
Babesien,” ‘ Zool. Anzeiger,’ xxx, Nr. 1, pp. 45-52.

84. Marcuovx, E. 1900.—* Piroplasma canis (Lav.) chez les chiens du
Sénegal,” *C. R. Soc. Biol.,’ lii, pp. 97-8.

35. Mesniz, F., Nicouxs, F., and Remuinerr, P. 1904.—‘“Sur le Proto-
zoaire du bouton d’Alep,” ‘C. R. Soe. Biol.,’ lvii, pp. 167-9.

86. Merram, A. EH. 1905.—‘ A Note on Bovine Piroplasmosis,” ‘Journ.
Hygiene,’ v, pp. 271-3.

87. Mincnin, E. A. 1903.—‘“ Sporozoa,” in Lankester’s ‘Treatise on
Zoology,’ pt. i, fase. 2, pp. 150-360.

38. Moras. 1903.—‘Sur le role des Tiques dans le développement de la
piroplasmose ovine (Carceag),” ‘C. R. Soc. Biol.,’ lv, pp. 501-3.

38a. Moras. 1902.—‘‘La piroplasmose ovine, ‘Carceag,’’’ ‘C. R. Soc.
Biol.,’ liv, pp. 1522-4.

39. Nocarp and Moras. 1902.— Contribution a l’étude de la Piroplasmose
canine,’ ‘ Ann. Inst. Pasteur,’ xvi, pp. 256-90, pls. v, vi.

40. Norraut,G.H.F. 1904,—*‘ Canine Piroplasmosis, I,” ‘ Journ. Hygiene,’
iv, pp. 219-52, 2 pls.

514 H. B. FANTHAM.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

Norra, G. H. F., and Granam-Smitu, G.S. 1905.—‘ Canine Piro-
plasmosis, II,” ‘ Journ. Hygiene,’ v, pp. 237-49, with pl. 9.

Rocers, L. 1904.—“On the Development of Fiagellated Organisms
(Trypanosomes) from the Spleen Protozoic Parasites of Cachexial
Fevers and Kala-azar,”’ ‘Quart. Journ. Micr. Sci.,’? 48, pp. 367-77,
pl. 25,

Rocers, L. 1906.—‘‘ Further Work on the Development of the Hepa-
tomonas [Herpetomonas] of Kala-azar and Cachexial Fever from
Leishman-Donovan Bodies,” ‘Proc. Roy. Soc.,’ 77, B. 517, pp.
284-93, pl. 7.

Ross, P. H. 1905.—“A Note on the Natural Occurrence of Piro-
plasmosis in the Monkey (Cercopithecus), ” ‘Journ. Hygiene,’ v, pp.
18-23, 3 charts.

Ross, R. 1903.—‘‘ A New Parasite of Man (Leishmania),” ‘Thompson
Yates Lab. Rep.,’ 5, ii, pp. 79-82, 1 pl.

Smita, Tx., and Kingorne, F. 1893.— Investigations into the Nature,
Causation, and Prevention of Texas or Southern Cattle Fever,” ‘8th
and 9th Ann. Rep., Bureau of Anim. Industry,’ pp. 177-304, with
10 pls. Washington, U.S.A.

STEPHENS, J., and CHRIsTOPHERS, S. R. 1904.—‘* The Practical Study
of Malaria and other Blood Parasites,’ 2nd edit.; ‘‘ Piroplasma,” pp.
332-7, pl. 3; “ Ticks,” pp. 337-49.

TueEIter, A. 1904.—*The Piroplasma bigeminum of the Immune
Ox,” ‘Journ. Roy. Army Med. Corps,’ ili, pp. 469-88.

TuHEILER, A. 1904.—‘ Hast Coast Fever,’ ‘Journ. Roy. Army Med.
Corps,’ Dec., 1904, 22 pp.

Witson, L. B., and Cuownine, W. M. 1904.—‘‘Studies in Pyro-
plamosis hominis (‘Spotted Fever’ or ‘ Tick Fever’ of the Rocky
Mountains),”’ ‘Journ. Infect. Diseases,’ i, pp. 31-57, 2 pls.

Wricut, J. A. 1905.—‘‘ Canine Piroplasmosis: IV. On certain Changes
in the Blood,” ‘ Journ. Hygiene,’ v, pp. 268-70, 3 charts.

Wricut, J. H. 1903.—* Protozoa in a case of Tropical Ulcer (‘ Delhi
sore ’),” ‘Journ. Med. Research, Boston,’ x, pp. 472-82, 4 pls.

ZieMANN. 1903.—“ Vorlaufiger Bericht iiber das Vorkommen des
Texasfiebers der Rinder in Kamerun und weiteres iiber die Tsetse-

krankheit sowie tiber die ‘Tiermalaria,’” ‘Deutsch. med. Wochen-
schr.,’ Jahrg. 29, Nr. 16, pp. 289, 290.

PIROPLASMA MURIS. pili

IX. EXPLANATION OF PLATE 28,

Nlustrating Mr. H. B. Fantham’s paper on “ Piroplasma
muris, Fant., from the Blood of the White Rat, with
Remarks on the Genus Piroplasma.”

The figures were all carefully outlined with camera lucida, under Zeiss’
3 mm. homog. immersion lens, apert. 1:40, and compensating ocular 18
(except in the case of Fig. 23).

The scheme of colouring adopted is approximately that of Leishman’s
stain, within the limits of the two colours used, blue and pink. Other
modifications of the Romanowsky method used have, for the sake of uni-
formity and simplicity, been also thus represented, the tint of the cytoplasm
of the erythrocyte only needing to be sometimes modified in such cases.

The magnification is in all cases approximately 1950 diameters, except
where otherwise stated.

Fic. 1.—Pyriform trophozoite, young, with peripheral chromatin and
vacuole,

Fic. 2.—Typical pear-shaped trophozoite.

Fic. 3.—Ovoid trophozoite.

Fig. 4.—Pyriform parasite with two chromatic dots.

Fie. 5.—Trophozoite (‘‘ schizont ”) in process of longitudinal division, with
chromatin bodies at the poles and well-marked vacuole.

Fie. 6.—Typical longitudinal fission of parasite in red blood corpuscle.

Fic. 7.—Two daughter trophozoites (‘‘ merozoites ”’), bigeminate.

Fic. 8.—Two pear-shaped parasites, still connected by a thin strand of
protoplasm at their pointed ends.

Fic. 9.—Two ovoid forms, probably resulting from a simple binary fission
of the parent parasite.

Fic. 10.—Two pairs of parasites ina red blood corpuscle, the pairs lying
partly over each other. The members of the pairs are still connected, though
at different stages of separation. ‘This is probably a case of double infection
of the blood-corpuscle host.

Fic. 11.—Three intra-corpuscular parasites ; possibly a fourth behind the
heart-shaped smaller pair. Two chromatin bodies occur in each of the
members of the heart-shaped pair.

VOL. 50, PART 3,—NEW SERIES. 37

516 H. B. FANTHAM.

Fic, 12.—Tbree small parasites in a small blood corpuscle from the spleen.

Fic. 13.—Six intra-corpuscular parasites inside a corpuscle from the
spleen.

Fic. 14.—Six “ameeboid ” parasites from the spleen.

Fic. 15.—Pear-shaped trophozoite, with somewhat pointed apex at the
broader end, and two chromatin bodies, from tail blood.

Fic. 16.—Gregariniform trophozoite with rod-like, drawn out chromatin
body, perhaps preparing for division.

Fic. 17.—Rather large, somewhat spherical trophozoite, with chromatin
body lying in clear zone of protoplasm. From spleen blood.

Fic. 18.—‘‘ Ameeboid” trophozoite, with a single pseudopodium in which
lies a chromatin dot. From spleen blood.

Fic. 19.—Pyriform parasite with chromatic appendage still protruding
from the erythrocyte. Probably a “flagellate” form, but no bulb or bead
seen on the appendage. This may be a somewhat abnormal form of parasite,
as no other similar one was observed, though it was quite distinct.

Fie. 20.—Bigeminate pair of free parasites from tail blood.

Fic. 21.—Group of free parasites in the blood plasma.

Fie. 22.—Leucocyte probably containing the remains of degenerate, in-
fected blood-corpuscles, or remains of free parasites. Only the red chromatin
masses of the parasites are left.

Fie. 23.—Portion of section of liver of infected white rat, showing dilated
capillary containing many infected red blood-corpuscles and several (three)
leucocytes, with large nuclei. 1000 approx., somewhat diagrammatic.

With Ten Plates, Royal Ato, is.
CONTRIBUTIONS TO THE KNOWLEDGE OF RHABDOPLEURA
AND AMPHIOXUS.
By E. RAY LANKESTER, M.A., LL.D., F.R.S.
London: J. & A. CHurcuity, 7 Great Marlborough 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/-.

The issue of SupptumMent Numpers being found inconvenient,
and there being often in the Editor’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/-, 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 twenty-five copies of their
communication gratis.

All expenses of publication and illustration are paid by the
publishers.

As a rule lithographic plates, and not woodcuts, are used in
illustration.

Drawings for woodcuts 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. ApLtarD & Son, 224, Bartholomew Close, E.C., on
the following terms:

For every four pages or less—

25 copies : : : : 5/-
30 335 ‘ ; ‘ 6/-
LO0st: 7/-

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. CHURCHILL, 7 GREAT MARLBOROUGH STREET,
Lonpon, W.

THE MARINE BIOLOGICAL ASSOCIATION
UNITED KINGDOM.

Patron—HIS MAJESTY THE KING.

President—Prof. E. RAY LANKESTER, 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 TO 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.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 . . é . £1 1 Oper annun.
LIFE MEMBERS . : : ; . 15 15 0O Composition Fee.
FOUNDERS . ee OO S70 ee) ss »

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.

ala

New Series, No. 200 (Vol. 50, Part 4). Price 10s. net.

Subscription per volume (of 4 parts) 40s. net.

NOVEMBER, 1906.

THE

QUARTERLY JOURNAL

OF

MICROSCOPICAL SCIENCE.

EeaWAY LANKESTER, M.A. D.Se., LL.D), FBS.

NHONORAKY FELLOW OF EXETER COLLEGE, OX¥YORD;, CORRESPONDENT OF THE INSTITUTE OF FRANCE
AND OF THE IMPERIAL ACADEMY OF SCIENCES OF ST. PETERSBURG, AND OF THE ACADEMY
OF SCIENCES OF PHILADELPHIA, AND OF THE ROYAL ACADEMY OF SCIENCES
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 LINCEI OF ROME, AND OF THK AMERICAN
ACADEMY OF ARTS AND SCIENCES OF BOSTON, ASSOCIATE OF THE
ROYAT ACADEMY OP BELGIUM: HONORARY MEMBER OF THE
NEW YORK ACADEMY OF SCIENCES, AND OF THE
CAMBRIDGE PHILOSOPHICAL SOCIETY, AND OF
THE ROYAL PHYSICAL SOCIETY OF EDIN-
BURGH, AND OF THE
BIOLOGICAL SOCIETY OF PARIS, AND OF THK CALIFORNIA ACADEMY OF SCIENCES OF SAN FRANCISCO;
FOREIGN ASSOCIATE OF THK NATIONAL ACADEMY OF SCIENCES, U.S., AND MEMBER OF
THE AMERICAN PHILOSOPHICAL SOCIETY 5
DIRECTOR OF THE NATURAL HISTORY DEPARTMENTS OF THE BRITISH MUSKUM, LATE FULLERIAN
PROFESSOR OF PHYSIOLOGY IN THER ROYAT, INSTITUTION OF GREAT BRITAIN;
LATE LINACRE PROFESSOR OF COMPARATIVE ANATOMY AND FELLOW OF MERTON COLLEGE, OXFURD.

WITH THE CO-OPERATION OF

ADAM SEDGWICK, M.A., F.RS.,

FELLOW AND TUTOR OF TRINITY COLLEGE, CAMBRIDGE 4

AND

SYDNEY J. HICKSON, M.A., E.RS.,

BEYER PROFESSOR OF ZOOLOGY IN THE OWENS COLLEGE, MANCHESTER.

WITH LITHOGRAPHIC PLATES AND TEXT-FIGURES.

Ure oe
‘J. & A. CHURCHILL, 7 GREAT MARLBOROUGH STREET.

1906.

Adlard and Son, Impr., ] {London and Dorking.

CONTENTS OF No. 200.—New Series.

MEMOIRS :

PAGE
On the Structure of the Nephridia of Dinophilus. By CresswE.u
SHEARER, Trinity College, Cambridge. (With Plates 29 and 30) . 517

Contributions to our Knowledge of the Anatomy of Notoryctes
typhlops, Stirling,; Part III, The Eye. By Gzorcina Sweet,
D.Sc., Melbourne University. (With Plate 31) 4 . 547

Structure and Origin of Canker of the Apple Tree. By James E.
Bromrietp, M.A., M.D.Oxon. (With Plate 32) ; 67S

Review of Dr. Richard Goldschmidt’s Monograph of Amphioxides.
By A. Wittey, Hon.M.A.Cantab., D.Se.Lond., F.R.S. (With

seven text-figures) 581

The Modification of the Sexual Characters of the Hermit Crab caused
by the Parasite Peltogaster (castration parasitaire of Giard). By
F. A. Ports, B.A., Trinity Hall, Cambridge. (With Plates 33 and
34). : : : ; ; . 599

On the Medusa of Microhydra ryderi, and on the Known Forms
of Medusz inhabiting Fresh Water. By Epwaxp Ports, of Phil-
adelphia, U.S.A. (With Plates 35 and 36). : . 623

On the Freshwater Medusa liberated by Microhydra ryderi, Potts,
and a Comparison with Limnocodium. By Epwarp T. Brownz,
B.A., Zoological Research ee University College, London.
(With Plate 37). ‘ ; . 635

Wirtu Tire, Contents, anD InpEx To Vot, 50.

a a ee

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 517

On the Structure of the Nephridia of Dinophilus.

By
Cresswell Shearer,
Trinity College, Cambridge.

With Plates 29 and 80.

4. ScHMIDT (22) was the first to draw attention to the

~ nephridia of Dinophilus in 1848, which he simply men-
tions in D. vorticoides as two longitudinal vessels. They
seem to have escaped further attention till Korschelt (11) in
1882 again mentioned their presence in D. apatris. It is
to Harmer (8), however, that we owe the most complete
description of the nephridia. In 1889 he described five
pairs in D. teniatus, a species found abundantly in the
tidal pools of Plymouth Bay. Since then they have been
studied by Schimkewitsch (28) in D. vorticoides from the
White Sea. This last author investigated them by means of
the intra vitam method of staining with methylene blue,
but was able to add little to Harmer’s account of their struc-
ture.

I have long suspected from both Harmer’s and Schimke-
witsch’s descriptions that the nephridia of Dinophilus
would finally prove to be closed internally by flame-cells
similar to those Goodrich (5) has described under the name
of solenocytes, in certain Polychets, Amphioxus, and
the Actinotrocha larva of Phoronis. Only during the
present season, however, have I been able to verify this, and
to definitely determine that they are furnished with typical

- solenocytes.
This is, I think, a point of some morphological interest.
VOL, 50, PART 4,.—NEW SERIES. 38

518 CRESSWELL SHEARER.

Tt not only adds a very primitive Annelid form to the already
large class of animals possessing uephridia of this primitive
type, but the presence of solenocytes on all four or five
pairs of nephridia in Dinophilus shows, I think, that in
the primitive Annelids not only the head-kidney and the
immediately following segments, as in Poly gordius, were
furnished with solenocytes, but the nephridia of all the
segments ; and that while in some Annelids they have been
retained on the nephridia of most of the segments, others
have lost them or retained them only on the nephridia of the
anterior segments. Moreover, the presence of solenocytes in
Dinophilus is of interest on account of the many relation-
ships this group shows with that of the Platyhelminths.
Weldon (24) has pointed out the similarity of the muscular
cesophageal appendage of Dinophilus to the pharynx
of Planarians, and Harmer (8) has already called atten-
tion to the median position of the generative pore and
the method of fertilisation of the male D. teniatus as
showing a certain affinity to Platyhelminths. The crawling
swimming movements of Dinophilus and its manner of
feeding is also suggestive of some relationship to this group.
As will be seen from consulting figs. 12, 13, and 15 of the
present paper the ventral surface of Dinophilus is defi-
nitely thickened into a crawling pad as in the Turbellaria.
The resemblance of the nephridial canals of Dinophilus
to the excretory canal of the flame-cells of Thysanozoon
is unmistakable; with the exception that one is furnished
with solenocytes and the other is not, there is little difference
between the two. The granular walls of the organ gradu-
ating into a delicate tube-like canal is found in each, and
their whole appearance is remarkably the same in both cases.
From this resemblance I think it is not impossible to look
for the discovery of solenocytes in some of the higher Platy-
helminths. Their discovery in this class will then definitely
establish the homology of Platyhelminth protonephridia with
the solenocyte-bearing nephridia of Annelids and the kid-
neys of Amphioxus. It may then be possible to trace a

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 519

consecutive series of changes in the form of the nephridia
from the more primitive flame-cells of the Turbellaria to the
complicated organs of Polychets, and so establish a definite
connection between the nephridia of these two groups, which
at present would seem to be separated distinctly from one
another.

Although I made a careful search for Dinophilus at
Plymouth during the latter part of June, 1903, the season
was too late, and the adults had already disappeared from
their usual haunts. Until the present season I have had no
opportunity of revisiting Plymouth. ‘lhis year, however, no
difficulty was experienced in obtaining an abundant supply
of material, although limited to the male form. Towards
the latter part of April I examined a considerable quantity of
Dinophilus material daily without observing a single
female. This apparent absence of the female I am unable
to account for, except on the ground that the breeding
season may have been passed, as young worms of all sizes
were observed in numbers, and the males seemed to have
already discharged their spermatozoa.! Both sexes disap-
pear entirely at Plymouth, I believe, towards the end of May
or the first part of June. During July and August they are
not to be found at Plymouth.

This disappearance of Dinophilus during certain months
has already been noticed by Hallez (7), Weldon (24), and
Harmer (8). Weldon (24) observed the adults of D. gigas
in April undergoing degeneration after the discharge of the
sexual products, and concluded that the worms periodically
died off after the breeding season. That this can hardly be
the case has been shown by Schimkewitsch (23) in D. vorti-
coides, where the females continue to live long after they
have laid their eggs, and in fact pass through several breed-
ing periods during the course of the year. In an American
species Moore (17) has found that the worms have the power
of forming capsules and of encysting themselves. On being

1 Harmer (8) states that about April 18th the females were still carrying
their eggs.

520 CRESSWELL SHEARER.

placed in small watch-glasses of sea water the worms after a
short time throw out a considerable quantity of mucus, which
hardens about them forming a firm capsule, coiled up inside
of which the worm can be plainly seen. The capsule soon
loses its bright red colour, when it is quite indistinguishable
against any small piece of alga to which it may be attached.
This year towards the end of the season I noticed a number
of males settling on small pieces of sea-weed and coiling up ;
they remained in this position for several days, when they
seemed to have secreted a certain amount of mucus about
themselves. These worms may possibly have been under-
going encystment; I was unable, however, to keep them
long enough to determine their ultimate fate. As the worm
during the encapsulated condition easily escapes observation
as the result of its minute size and lack of colour, it can
readily be understood how this encystment might account
for their periodical disappearance. The mucus thrown out
by the worm is evidently derived from the clear large cells
that have been frequently noticed in the epidermis. These
cells answer to all the usual staining reactions for mucus.
They recall similar cells seen in a large number of Tur-
bellaria, and this power of throwing out mucus to form a
capsule is perhaps another point of resemblance between the
two groups. In Histriobdella,!a form closely related to
Dinophilus, and usually classed with it in the group of the
Archiannelida, I have observed that the female in laying its
eggs surrounds each egg with a mass of mucus which subse-
quently hardens about it, securely fastening it in a stalked
capsule to the eggs of the lobster on which this animal lives
parasitically. Inside this capsule the embryo undergoes its
entire development, only emerging in the adult. state.
Among the Turbellaria a number of forms also lay their
eggs in somewhat similar fashion, as, for instance, Lepto-
plana, Plagiostomum,? and numerous parasitic forms.

' I follow Haswell (9) in retaining the name Histriobdella instead of

Foettinger’s (4) modification Histriodrilus.
? See Bresslau (2), figs. 73 and 74.

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. aval

Another point of resemblance between these groups is that
in the majority of marine and fresh-water Turbellaria deve-
lopment is simple and direct, as in Dinophilus and His-
triobdella. While itis true other members of the Archian-
nelida, as Protodrilus, Polygordius, and Saccocirrus,
may possess larval stages, these forms are more nearly
related to the Polychetz than to either Dinophilus! or
Histriobdella. Saccocirrus is undoubtedly a Poly-
cheet, and can hardly be considered at all an Archiannelid,
as Goodrich (5) has shown. Protodrilus is also evidently
closely related to the same class, for young larve I obtained
at Naples in 1903 plainly showed the presence of a well-
marked ciliated ring, and in external appearance bear a
certain resemblance to young Nereis larve.? The larval
form of Polygordius is of course well known. Moreover,
Haswell (9), in a paper on Histriobdella, has advanced
strong reasons for including Dinophilus and Histriob-
della in one class separate from that of Polygordius and
Protodrilus. The great difference of metamerism in the
two cases, the head segments, the relation of the brain
commissure to the mouth, and the great difference of the
reproductive organs in the two groups, shows that their
affinity is remote. On the other hand, Dinophilus and
Histriobdella show more relationship with one another in
the possession of a distinct head, a nervous system consisting
of a metamerically arranged series of ventral ganglia, an
alimentary canal essentially similar in both forms, and a
close resemblance in the reproductive organs, especially in
the male. I agree therefore with Haswell in grouping
Histriobdella and Dinophilus in one class separate from
Polygordius, Protodrilus, and possibly Ctenodrilus.
Regarding this last it is hard to say anything until some-
thing has been determined of its life history, as so far it has

1 Nelson (19) claims that Dinophilus shows a remarkable resemblance
in its cleavage stages to Polychats, and thinks that in this respect it cannot
- be considered a primitive form.

2 See Pierantoni’s (20) figs. 1 @ and 1 4.

HDD CRESSWELL SHEARER.

only been observed reproducing (Zeppelin 27) by fission.
Possibly Ctenodrilus may be intermediate in position
between Polygordius and Dinophilus, but the posses-
sion of distinct setee sacs would seem to denote that it is a
degenerate Polychet. lHisig (8) has also expressed the
opinion that Histriobdella is a very degenerate form and
not an Archiannelid, and it is possible that all the members
of this class are after all very degenerate forms, in which
their simple structure is by no means primitive. This would
seem to be borne out by a number of features in their deve-
lopment. In Dinophilus and Histriobdella, certaimly
the most interesting forms of this group, the direct mode of
development affords unfortunately little evidence of their
affinity, and what can be obtained from this source seems to
point to their being degenerate forms, as Hisig holds to be
the case.

I. GENERAL DESCRIPTION OF THE NEPHRIDIA.

The nephridia of Dinophilus can be satisfactorily
examined by placing the living worm on a glass slide and
covering it with a cover-glass to holdit and prevent excessive
movement. The most convenient method of doing this is to
support the cover-glass by a little soft wax which prevents
the worms from being crushed too much, but at the same
time allowing of their being compressed as required. The
structure of the nephridia can then be readily investigated
under an oil-immersion lens. As the nephridia frequently
run through the testis, which forms the largest part of the
body tissue in the male (see fig. 14) it is sometimes necessary
to examine the preparation from both sides in order to follow
a particular nephridium throughout its entire course. Tor
this reason I soon discarded ordinary glass slides for making
my preparations, and adopted big cover-glasses such as those
used for making large embryological slides ; the thinness of
the glass allows of the preparation being focussed under high
power lenses from either side as required.

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 523

The normal orange pigment of the worm gives a yellow
tone to the light passing through the preparation when
placed under the microscope, which rendered the finer
details of the structure of the nephridia more visible; the
outlines of the transparent solenocytes being more clearly
defined in a yellow than in a pure white light.1_ For the study
of the head-kidney of transparent Trochophore larve I have
found the employment of a yellow glass screen most useful,
and in Dinophilus the yellow pigment serves much the
same purpose.

Although I placed the worms in various solutions of methy-
lene blue and indigo-carmine I never succeeded in getting
the nephridia to take up these colours, as they readily do in
the White Sea species. While the nervous system and the
sensory cells of the epidermis absorb the colour by the
methylene blue method of impregnation, the nephridia remain
entirely unstained.

Lang has remarked on the difficulty of obtaining good pre-
parations of the flame-cells of Platodes with living material
on account of the inhibiting action of the cover-glass on the
flagella of the flame-cells. Much the same difficulty is experi-
enced with Dinophilus; as soon as the worms are com-
pressed the action of the flagella in the nephridial canals is
at once stopped, being only resumed some time afterwards,
and then in only one or two of the nephridia at the most. It
is impossible to get a preparation in which all the nephridia
are to be seen at once active ; sometimes it will be the first,
sometimes it will be the third, but rarely two consecutive
nephridia are seen active, and it is thus difficult to determine
accurately the relations of one nephridium to another when
they have to be observed from different preparations. This
is especially so with regard to the terminal portion of the
nephridium, where it is a canal so delicate as to be almost
indistinguishable except under the most favourable conditions.

1 Although blue violet light has greater resolving power than yellow light,

it remains a fact nevertheless, that the solenocytes are more readily seen in a
yellow light.

524, CRESSWELL SHEARER.

As the following observations apply only to the male worm,
I have never observed the fifth nephridium which, according
to Harmer (8), is present in the female. Schimkewitsch (28)
has confirmed this observation on the White Sea species, so
there can be no doubt of the presence of this fifth pair in the
female. Nelson (19), in an American species obtained like
Korschelt’s (11) from an aquarium tank, has been unable to
find any trace of the nephridia; this is the more remarkable
on account of their almost constant presence in other species.
They can so readily be observed in D. taniatus that it seems
almost impossible they could have been overlooked if present
in this American species. In D. gyrociliatus,! D. vorti-
coides and D. tzniatus there are five pairs of nephridia in
the female, and this would seem to bear out Harmer’s sugges-
tion that possibly the body in the genus Dinophilus is com-
posed of fivemetameres. What we know of the disappearance
of nephridia and the apparent ease with which they can be
dropped from various segments, as, for instance, in the
anterior segments of the Arenicola larva, the presence of
more than one pair to a segment, as in Capitella, and their
very erratic behaviour in the development of Oligochets,
renders it doubtful whether they offer trustworthy evidence
as an index to the number of segments, and their total
absence may possibly exemplify this in the case of the Ameri-
can species of Dinophilus. Again, Ctenodrilus, a form
usually classed with the Archiannelids, possesses only one
pair of nephridia although the animal is plainly divided at
least into seven metameres.? On the other hand, in His-
triobdella, according to Foettinger (4), there are five pairs
in the female as in Dinophilus. Harmer has emphasised
this as indicating another point of relationship between them.
I have, however, re-examined the nephridia in Foettinger’s

1D. gyrociliatus according to Repiachoff (21) is identical with D.
apatris of Korschelt’s (11) paper.

? I have confirmed Kennel’s (10) and Zeppelin’s (27) observations on a
species of Ctenodrilus found at Naples, and there is only one pair of these
structures present.

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. a20

species, and I cannot find more than four pairs in the female.
It is certain their number is not five, as in Dinophilus.
Haswell (9), in Stratiodrilus, which is almost identical
with Histriobdella, finds four nephridia in the female. It
is plain in Dinophilus, the number of nephridia are remark-
ably constant, while the same cannot be said of the number
of metameres.

As already stated the nephridia in the male consist of four
pairs placed very much in the positions Harmer has indicated
—that is, the head of the first nephridium is very slightly
behind the second pre-oral ring of cilia on a level with the
anterior margin of the muscular pharynx organ, the second
at the anterior end of the stomach, the third about the middle
of the stomach, while the fourth lies at the posterior end of
the stomach close at its junction with the intestine. Their
ducts run outwards and round the segments to open on the
ventral surface not very far from the median line, not in the
segment to which they belong as stated by Harmer (8), but
in the subsequent following segment as in Annelids. This
point is difficult to exactly determine as the field of vision is
so limited under an oil immersion objective (with which it is
necessary to examine this terminal part of the canal), that it is
thus hard to tell where the segment septa lie in relation to the
end of the nephridial canals. A number of successful pre-
parations, however, seemed to clearly show the canals running
some distance backwards into the next segment before ending
under the epidermis, and I think I am quite right in saying
that the canals end in the segment following that te which
they belong, as in Annelids. ‘he closed internai ends of the
canals bearing the solenocytes project slightly into the irregu-
lar space surrounding the gut—the so-called body-cavity
(fig. 17). This space is of variable dimensions, being lined
by no distinguishable membrane, and is traversed in many

1 Asin Dinophilus they project into the so-called body-cavity where
they appear to end as a single flame-cell. In every respect they resemble
the structure of the head-kidney of the Trochophore larva, the whole organ
of which is quite comparable to a single very enlarged solenocyte.

526 CRESSWELL SHEARER.

directions by irregular strands and muscular fibres. About
the gut it forms a sort of sinus in which a few large orange-
brown granules are constantly seen (see fig. 3, e.gr.) ; these
move up and down it from one end of the stomach to the
other as the worm twists and bends. Roughly, the contour
of the space follows the general form of the external segmen-
tation, sending prolongations towards the surface at the end
of each segment, which augment the marked pseudo-metameric
appearance of the animal when seen under the microscope.
That this space cannot be strictly regarded as a coelomic
cavity is amply testified by the fact that the spermatozoa are
seen enclosed in an entirely different set of spaces lined with
a definite membrane, by no chance ever being seen in this
space or any of its numerous prolongations into the general
mass of the testis tissue (fig. 4, sp.m.). Also by the fact that
it is traversed in all directions as already mentioned by
numerous muscle-strands and cells which denote its primitive
blastoceelic nature. It seems to me this space is directly
comparable to the great blastoccelic cavity of the head-segment
of the Polygordius larva, into which the larval nephridia
project with their solenocytes, but with which their canals do
not communicate, and to the blastoccelic collar space of the
Actinotrocha larva, into which their larval organs project
under similar conditions ; the space that subsequently gives
rise to the circular blood-vessel ring of the adult in this
animal.

In Dinophilus this cavity sends two prolongations from
the corners of the stomach forward into the head region,
these run outward to terminate in two small triangular en-
largements beneath the skin a short distance behind the eyes
(fig. 14, blce.1). These spaces are shown partially on either
side of the cesophagus (fig. 15). Into these the heads of the
first pair of nephridia bearing the solenocytes project, the
solenocytes standing out on the end of the nephridial canal
into the lumen of the space like bristles from a brush (fig. 1).
The protoplasmic parts of the solenocytes are but imperfectly
seen on account of their transparent nature, and their presence

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 527

is only shown by the refractive tubes which make them look
not unlike a number of pins inserted on the ends of the
nephridial canals. With proper lighting, however, and
the use of a coloured glass screen their actual outline is
brought into view, when they are seen to be pear-shaped
bodies inserted on the end of the blind nephridial canal, being
attached by a delicate hyaline tube piercing the nephridial
wall, each tube having a long flagellum in its interior which
passes into the lumen of the nephridial canal and down almost
the whole length of this canal (fig. 2). Ican find no evidence
of the canal itself being ciliated as Harmer (8) has described.
On the first nephridium I have counted at least thirty soleno-
cytes, but their number is possibly double this as it is impos-
sible to accurately count them on account of the active
movements of the worm.

Collectively they form the “ triangular body ” or “ ciliated
appendage ” mentioned by Harmer (8) as being inserted into
the distal extremity of the nephridial canal. They by no means
form really a triangular body, as they can be frequently seen
to be spread out fan-lke, and to be composed of a large
number of separate solenocytes. As the preparation is com-
pressed, however, sometimes the solenocytes become separated
into two distinct masses, and this would seem to happen
more frequently in the case of the first nephridium than with
any of the others (fig. 1). This appearance has undoubtedly
given rise to Harmer’s (8) supposition that the ciliated
appendage is sometimes bifid, that this condition is merely
artificial being temporarily assumed, and that the solenocytes
form a single mass on the end of the canal can be easily
ascertained with a little carefulexamination. The space into
which these structures project, like the main portion sur-
rounding the gut, usually contains a few small brown granules.
These, during the movements of the animal under compression
are frequently forced in among the solenocytes, where they
lodge as if stuck amongst the hairs of a brush. Sometimes
they are seen to be forced directly against the end of the
nephridial canal of the point where the solenocytes are

528 CRESSWELL SHEARER.

attached, and from the fact that they are never seen to enter
the canal itself, being soon swept free again into the general
space of the cavity, demonstrates that the end of the canal
is not open. And this is also borne out by the appearance of
the end of the canal itself, which shows no trace of any such
opening. A drawing is shown in fig. | of the first nephridium
under compression, in which the solenocytes are seen divided
into these two groups as just mentioned.

Under normal conditions the organ is not so spread out as
shown in this figure. The head of the nephridium projects
into the body-cavity space so as to look outwards and back-
wards, lying remarkably close under the epidermis. If for
convenience of description we follow Harmer (8) in dividing
the nephridium into three portions; then the solenocyte
body being the first part, the large thick-walled part of the
canal marked nep. c. 2 in the figure will form the second part,
while the fine delicate duct into which this rapidly passes
composes the third part (nep. c. 3).

The second portion of the first nephridium is shorter and
less developed than the corresponding section in other
nephridia. It reaches its greatest development in the third
nephridium. Its walls quickly widen out, and are not so
granular or marked with orange pigment as in the case’ of
the following organs. These granules are so placed as often
to give it the appearance of being composed of numerous
cells: of being an inter-cellular instead of an intra-cellular
canal. ‘This condition is again more marked in the case of
the third nephridium. The third portion of the nephridium
is remarkable for its uniform diameter, which remains the
same till it almost reaches its point of termination, where it
then narrows down to a very fine duct. It forms by far the
longest portion of the nephridium, running backwards and
ventralwards close under the skin, to terminate in the base-
ment membrane of the epidermis close to a conspicuous
vacuole just over the line that marks the anterior border of
the following segment (fig. 9). In sections of fixed material
I have been unable to trace the course of the nephridial

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 529

canals. Schimkewitsch (23) shows these in some of his trans-
verse sections. I have, however, never been successful in
seeing them in sections. In sections like those shown in
fies. 12—15, taken from material treated with cocaine and
carefully fixed in Hermann’s fluid, and afterwards treated

with Zenker’s fluid, no trace of the nephridial canals could
be found.

Il. Tae First NeEpHrRipium.

The position and general course of the canal of the first
nephridium has been very correctly given by Harmer (8),
indicated in his figure (Pl. 10, fig. 15). ‘The main portion of
the nephridium lies in a plane slightly dorsal to the wall of
the pharynx when the animal is flattened slightly. The
testis extending into the head region is pierced by the canal
which terminates ventrally to it, but probably some distance
lateral to the median line. Throughout its course it is
seldom seen to undergo any variation in size during the
movements and contractions of the worm, which suggests
that its walls are composed of some fairly firm substance,
which does not allow of the canal being readily compressed
and the lumen of the canal obliterated. In none of the
nephridia can any trace of an actual opening of the canal on
the surface be seen, and in all cases they would appear to
terminate at the basement membrane of the epidermis. The
closest examination fails to reveal any traces of an external
pore. The point at which the canal terminates, and up to
which the flagella in its interior can be traced, is but a very
short distance from the surface, careful measurement show-
ing it to be less than :004 of a mm., yet the external surface
is perfectly intact. The terminal point of the nephridium is
well shown in fig. 7, which represents a portion of the margin
of a preparation under high magnification. The canal is
plainly seen to end at the limiting membrane of the epidermis
opposite a large vacuole into which sometimes the ends of the
flagella are seen beating, not, however, under normal condi-

530 CRESSWELL SHEARER.

tions, but only when the preparation has been pressed con-
siderably out of shape. In calling these clear spaces vacuoles
I do so through lack of a more suitable term; they are in all
probability filled with fluid, as in preserved material they
appear (as in fig. 13) as a series of narrow chinks, while
in the living state they are seen as large refractive spaces as
shown in fig. 7. Bohmig (1) in Triclada maricola shows
the excretory canals terminating in the basement membrane
beneath the epidermis, much as they doin Dinophilus. In
his fig. 19, Pl. 20, he shows this very clearly, and this figure
might very well do to illustrate the conditionin Dinophilus.
In this Triclad the epidermis also contains numerous clear
vacuoles, the walls of which lie close against the terminal
point of the excretory canals. In spite of much time devoted
to this point in Dinophilus I have been quite unable to de-
termine anything further regarding the relationships of the
nephridial canals to the vacuoles or whether these vacuoles in
turn open on the exterior. This point is of some interest in
connection with Korschelt’s (11) observation of the probable
presence of a fine system of anastomosing canals in the base-
ment membrane with which the nephridial canals may possibly
connect, although he could not succeed in D. apatris in
establishing any such connection. This is well shown in
Korschelt’s (11) fig. 29, but no similar system is to be found
in the case of D. teniatus. Korschelt points out the resem-
blance of this network to the system of canals into which the
ducts of the flame-cells of Polyclads open. ‘These in turn
open on the exterior in two dorsal pores. By certain im-
proper focussing, however, of the muscle-strands which run
irregularly throughout the body, an appearance is obtained
of a series of anastomosing canals looking not unlike the pic-
ture shown in Korschelt’s (11) fig. 29, and possibly it is this
Korschelt has taken for a system of canals ; that this network
is really made up of anastomosing muscle-fibres can readily
be determined by proper focussing. When it is considered
how notoriously difficult it is to see the external openings of
the ducts of the head-kidney in the Polygordius larva, the

EE

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 531

dorsal pores of Polyclads, and also the openings of the larval
excretory organs of the Actinotrocha larva, it may be claimed
justly that I have adduced, perhaps, no conclusive evidence
for their absence in Dinophilus, to this I can only say that
after prolonged investigation of this point I can find no evi-
dent traces of their presence. It is, perhaps, worthy of note
that Schimkewitsch (23) in the White Sea species also can
find no evidence of external openings to the nephridia by the
methylene-blue method of impregnation.

In several preparations the canals could be traced down to
these vacuoles which appeared closed, but careful examination
of the external surface over the vacuole seemed to show the
presence of a slight depression. In this depression a number
of granules were adherent to the external surface, these
moved backwards and forwards as if disturbed by the escape
of fluid from the interior of the vacuole. Frequently they
were noticed to have a peculiar dancing movement. I could
never see distinct evidence, however, of the escape of any
fluid, and the granules were themselves never seen to be de-
tached or actually washed from their places as one might
suppose would take place if any fluid was being discharged
from the vacuole. Close examination of the surface of the
vacuole when the movement of the granules was most evident
failed also to reveal traces of any external pore. In some
cases the granules seemed to dance round one particular point
in the middle of the surface of the vacuole, but close examina-
tion of this under a Zeiss 2 mm. oil-immersion objective com-
bined with 12 m. and 18 m. compensating ocular, with careful
lighting, failed to show the presence of any opening. I am
inclined to think the fluid escapes from the vacuole by osmo-
sis through its wall, and this escape of fluid accounts for the
movement of the granules which themselves may be of an
excretory nature. They are frequently orange coloured like
the granules of the blastoccelic space, though very much
smaller, and are probably deposited on the external surface
from the slowly excreted fluid of the vacuole as it reaches the
exterior.

532 CRESSWELL SHEARER.

The solenocytes of the first nephridium are quite like those
of the following nephridia in appearance, although a number
of observations have led me to believe that they are somewhat
finer and more pin-head-shaped in outline. Unlike the
solenocytes of Polycheets, it is difficult to see the lumen of
their solenocyte tubes, although their flagella can be traced a
little way up the narrow refractive stalk which attaches the
solenocyte to the end of the nephridial canal ; neither can the
end of this tube be seen projecting into the nephridial canal
as in some Polychets. In the head of each solenocyte is a
clear refractive dot; this is so regular in shape and size as
to preclude its being a nucleus. Hach solenocyte sends a
single flagellum down the canal of the nephridium, so that
the accumulated flagella of all the solenocytes form a mass
which almost fills the lumen of the canal. The waves of
ciliary motion starting in the solenocytes travel progressively
down to the ends of their flagella. As already mentioned, I
believe each flagellum extends the length of the nephridial
canal (fig. 6). In cases when the ciliary motion has almost
ceased and the movements of the flagella are consequently
slow, the wall of the canal is exposed from time to time as the
flagella move from side to side; it is then seen to be bare
without any trace of the insertion of cilia (fig. 5). At the
same time the individual lashes of the flagellum can be dis-
tinctly seen, and in some instances traced to their respective
solenocytes. As each lash beats backwards and forwards in
the canal it causes its solenocyte to vibrate with it in the body-
cavity space, and as the waves of motion pass down the
flagellum in a metachronous manner the solenocyte vibrates
backwards and forwards at each wave. As the solenocytes
are inserted on the ends of the nephridial canals in all possible
directions with regard to one another their vibrations are not
necessarily in the same plane. Sometimes it happens that
they are arranged so that a large number of them beat in the
same plane, when the whole mass moves together; usually,
however, this is not the case, each solenocyte has its own
separate motion, and they all appear to beat independently of

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 533
one another. As the preparation gradually dies they are the
first to stop, while their flagella in the nephridial canal still
continue to make a few sluggish movements. Under com-
pression they seldom remain active longer than fifteen minutes,
at the end of which time their action has become quite slow.
They never show any tendency to unite together for mutual
support in the body-cavity as in some Polychets, but always
remain separate. Their protoplasm is clear and remarkably
free from granules, and is, for this reason, highly refractive.
As already mentioned, the solenocytes of the first nephridium
seem somewhat finer than those of the following segments,
and in several cases their ends appear as if slightly flattened
and hook-shaped ; these are smooth, and never throw off pro-
toplasmic processes into the body-cavity as in Polygordius.
There is also considerable variation in size between those
attached to the margin and those attached to the centre of
the end of the nephridial canal, the latter being much longer
and more decidedly pear-shaped in form; their long stems
densely packed together afford considerable support to the
mass of solenocytes. Beyond the refractive granule men-
tioned (figs. 2 and 5) as sometimes distinguishable in their
heads, no obvious evidence of the presence of a nucleus is
visible in the living state, nor in the second portion of the
nephridium, the thick granular-walled part, could I ever dis-
tinguish the presence of nuclei. It will be seen from Meyer’s
(16) figure of one of the nephridial canals of D. gyrociliatus
(fig. 10) that a conspicuous nucleus is present on one side of
the canal. I have never observed any such nucleus in
D. teniatus.

I have never seen excretory matter passing down the
nephridial canals in the form of granules, and they would
seem to excrete clear watery fluid alone. Whatever function
the so-called body-cavity performs, it is certain that the hght
orange fluid filling it must at least play a considerable réle
in the areation and the removal of waste products from the
body tissues among which it ramifies. In the removal of
this fluid by osmosis the solenocytes, with their hyaline tubes,

VOL, 50, PART 4,—NEW SERIES. 39

534 CRESSWELL SHEARER.

undoubtedly take a great part. Their action is probably
selective. If we take the average length of a solenocyte tube
to be one hundredth of a millimetre, and allow sixty soleno-
cytes for each of the eight nephridia, a number very much
below their number in the case of the second, third, and
fourth pairs of nephridia, we get a total length of 4°8 mm.
of solenocytes for the whole animal, which represents a con-
siderable area for osmotic exchange in an animal whose total
length is under one millimetre. It is thus evident that a
considerable amount of fluid could be rapidly excreted from
the body-cavity by means of the solenocyte tubes. ‘The
nephridial canals, however, never show the presence of fluid
passing down their interior, and the process of excretion is
possibly a very slow one.

III. Tot Seconp NEPHRIDIUM.

The second pair of nephridia occupy the corners of the
body-cavity space opposite the anterior end of the stomach
(figs. 3 and 4). Into the triangular ends of these spaces
their solenocytes project from the ends of the canal portions
of the nephridia. The solenocytes are more numerous than
in the case of the first nephridium, and, in fact, the whole
nephridium is better developed. Numerous granules are
seen, as in the case of the first nephridium, frequently
lodging among the solenocytes. The end of the canal
towards the body-cavity in this case is distinctly closed, its
walls being marked out by dense masses of deep orange
pigment arranged in irregular masses along the first part
of its course. The granules in the wall are large and refrac-
tive, and render the nephridium readily visible even when
the flagella in its interior are not in motion. About the
middle of the testis this portion of the nephridium graduates
into the third part at the point at which the canal turns to
run through the testis in a ventral direction (see fig. 4). The
canal then continues its course ventral to the testis and
slightly backwards into the next segment, to terminate in

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 9535

the basement membrane of the epidermis near the median
line on the ventral surface of the worm. On the course of
this portion of the nephridium the canal sometimes widens
into lacuna-like spaces in the living worm under compression.
A similar space is shown by Meyer (16) on all of the
nephridia in D. gyrociliatus (fig. 9). In the enlarged figure
(fig. 10) this space is again seen very much as it looks in D.
teniatus (fig. 16) under compression. The walls of this
space are remarkably thin and uniform in thickness, and
have a transparent appearance, the masses of orange pigment
seen about the rest of the canal being usually wanting in the
walls of the spaces or lacune. Through the middle of these
lacune the flagella, which, as I have mentioned, can be
traced down from the solenocytes, beat passing on and
through and down the nephridial canal (fig. 6). Meyer
shows these spaces lined with cilia. I think, however, they
are not so lined in D. teniatus, there being considerable
grounds for believing that the flagella beating in the spaces
where they spread out somewhat produce a false appearance
of ciliation. It is difficult to describe this in some instances,
where these spaces, on the contrary, would seem to be
furnished with cilia.

A more important difference between the canals of D.
gyrociliatus and D. teniatus is that in D. teniatus
they are never folded, as shown in Meyer’s (16) fig. 10. The
lacunar appearance of these spaces is greatly increased by
the worms being under some compression when examined,
and they may be in great part produced by this compression.
While Meyer figures these spaces on all five pairs of
nephridia in D. gyrociliatus, I have observed them only
in one instance on the first nephridium and on the second,
but almost constantly on the third, in about five hundred
preparations examined altogether. Meyer (16) also shows a
well-marked external pore. It must be remembered that in
D. gyrociliatus, from Repiachoff’s (21) figures and de-
scriptions, as well as in the two figures given by Meyer, the
body-cavity space is much better developed than in D,

536 CORESSWELL SHEARER.

teeniatus, and the testis in the male is a much smaller
structure; this allows room for the nephridial canal becoming
coiled, and therefore more complex in structure. In D.
teeniatus the muscular system is also better developed, and
this probably plays some part in effecting the arrangement
of the nephridial canals.

In one instance the canal of the second nephridium could
be traced back into the next segment to about the level of
the head of the third nephridium, and so to the middle of
the segment. This relationship, however, may be somewhat
incorrect. It has to be remembered that the worm during
examination is sometimes able to crawl a little bit, in doing
which under the cover-glass the proper position of the dorsal
side of one segment above its ventral surface is displaced.
When the preparation is viewed vertically from above, what
is ventral does not correspond always to what is dorsal as
belonging to the same segment, and so relationships like
these established on living material may sometimes be in-
correct. In saying, therefore, that the nephridia of the first
and second pairs almost overlap, it must be kept in mind
that this is applied to living material observed under some-
what abnormal conditions. As the canal passes through the
testis it frequently passes close to a number of the large
vesicles in which the spermatozoa are seen actively moving ;
these, however, are always distinctly walled off from the
canal, and even under great compression never rupture or
discharge into the nephridial canal. In every case the spaces
containing the spermatozoa are always completely shut off
from the body-cavity ; while they are seen in almost every
other part of the body, they are never seen in the a cavity
or in any of its connecting spaces.

IV. Tue Turrp NEpuHRIDIUM.

The third nephridium is the most characteristic of all the
nephridia (figs. 5, 8, 10, and 16). It is situated in the wall
of the body-cavity space about opposite the middle point of

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 537

the stomach. Here its solenocytes project into the lumen of
the space, as in the case of the former organs. In this
nephridium all three parts of the structure reach their
fullest development, the solenocytes being most numerous,
the second portion being large, and the third portion remark-
able for its great length (fig. 10). The thick granular wall
of the second portion immediately catches the eye on the first,
examination of the preparation, its dark red pigment and its
hyaline granular walls render it striking. The greater part
of its canal lies dorsal to the main mass of the testis. It
pierces this to run ventral to it about its outer third, termi-
nates near the median line on the ventral surface as in the
previous organs, and like them in the segment following that
to which it properly belongs. The course of this nephridium
is shown infig. 10. It terminates in the basement membrane,
as shown in fig. 7.

V. Tue Fourta NeEpPHRIDIUM.

The head of the fourth nephridium is situated opposite the
posterior end of the stomach about its junction with the gut.
It lies for the most part ventral to the testis, running back-
wards and outwards to terminate on the ventral surface near
the median line. Its solenocytes project into the posterior
prolongation of the corners of the body-cavity space, and are
somewhat difficult to see on account of the stomach usually
folding over and hiding them when the animal is compressed.
It is more feebly developed than the third nephridium, and
from the denseness of the testis to which it lies ventral its
canal is difficult to follow. It terminates beneath the surface
of the epidermis in a manner similar to the other nephridia.
The second portion of the nephridium is not marked out by
the masses of red pigment as in the previous cases, its walls
are but slightly granular. The canal passes among several
large clear spaces crowded with spermatozoa. With these it
never shows any connection.

538 CRESSWELL SHEARER.

VI. Tse Firra NEpPHRIDIUM.

The fifth nephridium I have had no opportunity of exam-
ining personally through lack of material as already stated.
According to Harmer (8) this nephridium is situated in the
fifth segment of the female “on the ventral side of the
intestine (behind the cecal end of the stomach).” In general
structure it resembles the other nephridia just described.
In the male Harmer suggests it is represented by the vesicula
seminalis, and this view I wish to discuss under the present
section. This suggestion of Harmer’s concerning the vesiculee
has found considerable favour among subsequent investi-
gators, so that it is important to consider the grounds on
which it is based. Schimkewitsch (238) thinks it highly
probable that the seminal vesicles of the male represent the
modified fifth pair of nephridia of the female, while in the
female the nephridia of the sixth segment are represented by
the oviducts. Haswell (9) in the allied form of Histriob-
della states that “in the fourth segment the nephridia are
probably represented in the female by the oviducts, in the
male by the vasa deferentia.”

The grounds which Harmer (8) has advanced in support of
this contention are partly anatomical and partly embryo-
logical. The anatomical reasons are—that the seminal
vesicles in the male occupy the position of the fifth pair of
nephridia of the female; in the second place these vesicles
are furnished with ciliated funnels, opening into the cavity
of the testis, which resemble the ciliated appendage of the
nephridium (which Harmer considered probably opened in
the primary body-cavity by funnel-like apertures) ; thirdly,
cases occur among Annelids in which we know the nephridia
are transformed into genital ducts and functions as such in
the adult animal.

The embryological reasons are derived from the study of
the immature vesicles of young male worms. One of these
is shown in Harmer’s (8) figure 5, and concerning which

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 539

figure he states, p. 14, ‘‘ the vesiculee seminales were in their
definitive position in the fifth body segment, and their iden-
tification as vesicule was rendered sufficiently certain by
the fact that they contained ripe spermatozoa. The vesiculee
were arranged in an obliquely transverse position, their
outer portions ending blindly at the level between the two
ciliated rings of the fifth segment, their inner ends opening
into the cavity of the testis. A part of the vesicula imme-
diately succeeding the internal aperture was lined with long
cilia; the next part of the tube contained a small mass of
spermatozoa. The penis was well developed, and obscure
indications of a duct leading from the vesicula to the penis
was observed ; the existence of the duct was not, however,
completely proved. The resemblance of the young vesicula
seminalis to an ordinary nephridium was manifest not only in
its shape and position, but still more conspicuously by the
fact that its walls contained an orange pigment, exactly
resembling that so commonly found in the walls of the
excretory tubes.”

Numerous stages between this form and the mature con-
dition were observed. “The final form is acquired by the
gradual distension of the originally subcylindrical tube by
spermatozoa, this distension being accompanied by an altera-
tion in the direction of its axis, the result of which process
is that the end which, in the young vesicula, is external,
is situated in the adult condition in front, the whole organ
having now acquired an antero-posterior direction. The
funnel during the above changes will naturally come to be
situated near the posterior end of the organ.” ‘‘ It must be
especially noted that the funnel of the vesicula is in a posi-
tion corresponding with that of the ciliated appendage of an
ordinary nephridium, and that the original external aperture
of the modified nephridium was probably (in the phylo-
genetic history of the organ) at the opposite end of the tube,
which ultimately becomes the blind anterior end of the
vesicula. he relations of the outer ends of the young
vesicula to the ciliated rings of the fifth segment further

540 ORESSWELL SHEARER.

support this conclusion.” ‘Thus, while the nephridia project
into the space of the body-cavity, the modified nephridia of
this segment (the fifth pair of the female) open into the
cavity of the testis. However, it seems to me there are
many difficulties in the way of accepting this origin of the
seminal vesicles.

I have already called attention to the closed nature of the
spaces of the testis tissue and the fact that they are entirely
separated from the primary body-cavity about the gut
(fig. 4, s. p.m. and fig. 11, m. v.), and Dinophilus would
seem to be one of the few animals in which both the primary
and secondary body-cavity exist alongside of one another at
the same time. Repiachoff (21) at some length has con-
sidered the relationships of these two cavities in Dino-
philus, and has clearly pointed out how the secondary body-
cavity of Aunelids is probably represented in Dinophilus
by the spaces in the testis tissue. These towards the poste-
rior end of this structure fuse together to form the large and
roomy cavity of considerable size well shown in Schimke-
witsch’s fig. 43> (23). It is lined by a definite epithelium,
while the primary body-cavity about the gut possesses no
such lining membrane.! As the nephridia in Dinophilus
are related to the primary body-cavity alone it is necessary
to suppose on the basis of Harmer’s theory that those of the
fifth segment have lost their connection with this structure,
and acquired openings into the testis cavity. I have also
shown that the nephridia do not open into the primary body-
cavity, but are closed; therefore the funnel-like openings of
the vesicule seminales are new structures that have been
developed since this relationship has been established, and
cannot have been transferred, as Harmer thought, from the
primary body-cavity.

The resemblance of the male reproductive system of Dino-
philus to that of the Turbellaria is so similar in many
respects, although the reproductive system in these forms is

' Repiachoff (21) has described a peritoneal lining to the outer wall of the
gut in D. gyrociliatus, but no such lining is present in D, teniatus.

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 541

further complicated by their hermaphroditic condition, I
think it would be as just to assume that in them the vesiculi
and vasa differentia also represent modified portions of the
nephridial system. In the Turbellaria either the male or the
female sexual organs are readily reducible to the conditions
presented in Dinophilus. This, connected with the changes
necessary in the position of the relative parts of the nephri-
dium in order to make it agree with the observed growth of
the vesicles in young worms as described by Harmer (8)
renders it probable that they are not modified nephridia. It
is true, as Montgomery (18) has pointed out, no hard or fast
distinction can be drawn between the blastoccel and ccelom
as morphologically distinct spaces; since their presence as
separate cavities is dependent to a large extent on the form
of cleavage and gastrulation, which often differs so greatly
in closely allied forms. While realising no great import-
ance can be attached to the separation of these spaces,
nevertheless in Dinophilus this separation is so well
marked that it is hard to suppose that the nephridia of the
fifth segment could readily lose their connection with one
space, and acquire openings into that of the other in the
manner Harmer supposes without showing more obvious
evidence of this change. ‘This view will certainly need more
conclusive evidence in its favour than has so far been ad-
vanced for it, and it is possible ultimately it will prove to be
wrong.

VII. Summary.

In the present paper it has been shown that the nephridia
of Dinophilus are of the primitive solenocyte-bearing type
so frequently found in Annelids. In the male there are four
such pairs of nephridia whose solenocytes project into the
primary body-cavity or blastoccelic space about the gut.
The shape and general position of the nephridia is the same
as Harmer (8) has already described for this species. The
terminal portion of the nephridial ducts, however, probably

542 CRESSWELL SHEARER.

end beneath the skin, just over the division line of the seg-
ment to which they properly belong. The nephridia are
furnished with typical solenocytes, and their canals are defi-
nitely closed, and do not open into the primary body-cavity.
They are not ciliated, but the flagella of the solenocytes
beating down the length of the canals give them the appear-
ance of being ciliated. The presence of solenocytes in
Dinophilus is a point of considerable morphological import-
ance on account of the relationship this worm shows with
lower forms, especially the Turbellaria. Their discovery in
Dinophilus, on the other hand, in the absence of our know-
ledge of their presence in lower forms may be held to indi-
cate close affinity with the more highly developed Annelids,
and especially the Polycheets.

LITERATURE CITED IN TExr.

1. Boumic, Lupwie.—“ Tricladenstudien I, Triclada maricola,” ‘ Zeit.
f. Wiss. Zool.,’ vol. 1806, p. 344.

2. BressLav, E.— Beitrage zur Entwicklungsgeschichte der Turbellarien I,
Die Entwicklung der Rabdocélen und Alloiocdlen,” ‘ Zeit. f. Wissen.
Zool.,’ vol. Ixxvi, p. 213, 1904.

3. Hisic, H.—* Die Entwicklungsgeschichte der Capitelliden,” ‘ Mitth. Z.
Stat. Neapel,’ Bd. xiii, p. 1, 1898.

4. Fouttincer, A.—-“ Recherches sur lorganisation de Histriobdella
homari,” ‘Archives de Biologie,’ vol. v, p. 435, 1884.

5. Goopricu, E.8.— On the Nephridia of the Polychete,” Part 3, ‘ Quart.
Journ. Mier. Sci.,’ vol. 43, p. 715, 1900.

6. Goopricu, LH. 8.—**On the Structure and Affinities of Saccocirrus,”
‘Quart. Journ. Micr. Sci.,’ vol. 44, 1901.

7. Hauiez, P.—‘ Contributions a |’Histoire Naturelle des Turbellaries,’
Lille, 1879, p. 155.

8. Harmer, 8. F.—“ Notes on the Anatomy of Dinophilus,” ‘Journ. of
Marine Biol. Assoc.,’ N.S., vol. i, pp. 1—25, 1889.

9. HasweLt, Wituram A.—‘“‘On a New Histriobdellid,”’ ‘Quart. Journ.
Micr. Sci.,’ vol. 43, p. 299, 1900.

10. Kennex, J. V.—‘ Uber Ctenodrilus pardalis,” ‘Arbeit. a. den.
Zoologisch. Zootomische Inst. in Wiirzburg,’ vol. v, p. 373, 1882.

ile

12.

13.

14,

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 949

Korscuext, Evcen.—* Uber Bau und Entwicklung des Dinophilus
apatris,” ‘Zeit. Wiss. Zool.,’ vol. xxxvii, p. 315, 1882.

Layo, A.—“‘ Der Bau von Grunda segmentata und die Nerwandl-
schaft der Plathelminthen mit Coelenteraten und Hirudineen,” ‘ Nat.
Zool. Stat. Neapel,’ vol. iii, p. 187, 1881.

Lane, A.—“ Die Polycladen (Seeplanarien) des Golfes von Neapel und
der angreuzenden Meereabschnitte,” ‘Fauna Flora. Golf. Neapel,’
Monograph xi, p. 678, 1884.

Lane, A.— Beitrage zu einer Trophocdltheorie Jenaischen,” ‘Zeit. f.
Nat. Wiss.,’ vol. xxxviil, p. 1, 1903.

LanKkesteER, EB. Ray.—‘A Treatise on Zoology,’ part ii, chap. 2, “The
Enteroceela and the Ceelomoceela,” 1900, p. 1.

Meyer, H.—* Studien iiber den Korperbau der Anneliden,” ‘ Mitt. a. d.
Zool. Stat. zu Neapel,’ vol. vii, 1886-7 (pl. xxvii, figs. 9 and 10).

Moort, AnnE.—“Dinophilus gardineri (sp. nov.),’” ‘ Biological
Bulletin,’ No. 1, p. 15, 1899.

Monrcomery, T. H.—‘‘On the Modes of Development of the Mesoderm
and Mesenchyme with reference to the supposed Homologues of the
Body-cavities,” ‘Journ. of Morph.,’ vol. xii, p. 355, 1897.

Netson, J. AA—“* The Early Development of Dinophilus: a Study in
Cell-linage,” ‘Proc. Acad. Nat. Sci. Philad.,’ p. 687, 1904.

PrerantoniI, UmBerto.—‘‘Sullo Svilluppo del Protodrilus e del
Saccocirrus,” ‘Mitt. a.d. Zool. Station Neapel,’ vol. xvii, p. 575,
1906.

Rerracuorr, W.—“ On the Anatomy and History of the Development
of Dinophilus gyrociliatus,” O. Schmidt, Odessa, 1886. (The
paper is in Russian: a translation was kindly lent me by Dr. Harmer.)

Scumipt, Oscar.—‘ Neue Beitrage zur Naturgeschichte der Wurmer,’
Jena, 1848.

ScuimKewitscu, W.—‘ Zur Kenntuis des Baues und der Entwicklung
des Dinophilus vom Weissen Mere,” ‘Zeit. Wiss. Zool.,’ vol. xlix,
p. 46, 1895.

We poy, W. F. R.—“On Dinophilus gigas,” ‘Quart. Journ. Mier.
Sci.,’ 1886, p. 109.

WitHeEmI, J.—‘‘ Untersuchungen tber die Excretionsorgane der Suss-
wassertricladen,” * Zeit. Wiss. Zool.,’ vol. Ixxxi, p. 544, 1906.

Vespovsky, F.—‘ Zur Hamocdltheorie,” ‘Zeit. f. Wissen. Zool.,’ vol.
Ixxxii, p. 80, 1905.

Zevretin, Max Grar.—‘ Uber den Bau u. die Theilungsvorgiinge des

Ctenodrilus monastylos, Nov. Spec.,”’ ‘Inaugural Dissertation,’
Leipzig, 1883, p. 1.

544. GRESSWELL SHEARER.

DESCRIPTION OF PLATES 29 & 30,

Illustrating Mr. Cresswell Shearer’s paper on “ The Structure
of the Nephridia of Dinophilus.”

LETTERING.

ble.1. Blastoccelic space of the head segment. blc.2. Blastoccelic space of
the trunk region. das.mem. Basement membrane of the epidermis. e.gr.
Excretory granules of the blastocclic spaces. y/. Flageila of the nephridial
canal. jlg. Flagellum of a solenocyte. gv.s. Granule in the head of the
solenocyte. G¢. Gut. msc. Muscle fibres. m.v. Limiting membrane of the
vesicule. zep.c.2. Second portion of the nephridium. zep.¢e.3. Third por-
tion of the nephridium. pig.str. Pigment cells. pir. Pharynx. phr.o. Pha-
rynx organ. sep.d. Line of division between segments. so/. Solenocytes.
sp.m. Sperm reservoirs or spaces of the testis. Sm. Stomach. Tes. Testis.
vac. Vacuoles of the epidermis. v.c. The ventral thickened crawling surface
of the worm. vesc. Vesicula seminalis,

PLATE 29.

Figs. 1—12 were drawn under a 2 mm. oil-immersion obj. with oe.
6, except where otherwise indicated. They were drawn from living
preparations, although no attempt has been made to exactly portray the
appearance of the living protoplasm, which is simply roughly indicated as
granular. The sections are from material treated with cocaine and fixed in
Hermann’s fluid and Zenker’s fluid.

Fic. 1.—A portion of the first nephridium, somewhat unduly compressed
and flattened. Obj. 2 mm., oc. 12.

Fic. 2.—A solenocyte of the third nephridium. Obj. 2 mm., oc. 18.

Fic. 3.—The second nephridium. Several granules being seen in the
blastoccelic space about the gut.

Fic. 4.—The second nephridium.,
Fie. 5.—The first and second portion of the third nephridium.
Fic. 6.—A portion of the flagella of the third nephridium.

Fig, 7.—The termination of the canal of the third nephridium in the base-
ment membrane of the epidermis. Obj. 2 mm., oc. 18.

Fic. 8.—Head portion of the third nephridium.

a Se

STRUCTURE OF THE NEPHRIDIA OF DINOPHILUS. 545

Fie. 9.—Terminal portion of the third nephridium.
Fie. 10.—The third nephridium.
Fie. 11.—A transverse section through the region of the vesicule.

Fre. 12.—A transverse section through the region posterior to the stomach,
showing the testis surrounding the gut.

Fic. 13.—Transverse section through the region of the junction of the
gut with the stomach.

Fie. 14.—Longitudinal coronal section in a plane slightly below the middle
of the stomach.

Fie. 15.—Transverse section through the region of the pharynx.
Fie. 16.—The third nephridium.

PLATE 30.

Fre. 17.—A general diagrammatic figure showing all the nephridia,

EYE OF NOTORYCTES TYPHLOPS. 547

Contributions to our Knowledge of the Anatomy
of Notoryctes typhlops, Stirling.

Part III.—The Eye.

By

Georgina Sweet, D.Sc.,
Melbourne University.

With Plate 31.

— -———.

INTRODUCTION.

Tue work, of which this paper records a part, has been
done in the Biological Laboratory of the University of Mel-
bourne, for the use of which I am indebted to Professor
Baldwin Spencer. ‘To his kindness I owe all the material
used, together with much criticism and advice, and assistance
in obtaining literature, without reference to which this
research would have been incomplete. I wish also to thank
Mrs. H. R. Elvins, who had commenced this work on the
eye, but was prevented from proceeding with it, for the use
of her series of sections and sketches.

Part I of these Contributions, dealing with the Nose and
Jacobson’s Organ, and Part II dealing with the Blood
Vascular System, were published in the ‘Proceedings’ of
the Royal Society of Victoria in 1904.

The Structure of the Hye and associated parts.

The degenerate eye of Notoryctes has been previously
referred to by Dr. E. ©, Stirling, who has noted concerning

548 GEORGINA SWEET.

it [10, pp. 22, 23], that “it is not visible exteriorly,” has “no
bony orbit,” and “ consists of a small circular pigment spot ”
and three years later, a longer reference was made to it
[11, pp. 159, 180] giving its position in the head and its
general characteristics, concerning which Dr. Stirling writes :
—‘“T cannot be absolutely certain of its relations, or of the
morphological value of its constituent parts.” As the result
of careful study of some seven or eight heads, I am able not
only to corroborate Dr. Stirling’s observations so far as they
go, but also to record a quantity of additional information
which will, I think, make clear the homologies of this very
interesting relic, forming at the same time a striking study in
degeneration and convergence. Considerable variations are
met with in the different points of structure, as might be
expected in a degenerate organ, and the following description
takes note of the more important of these variations, while
at the same time, it is based on the most constant features.
It is a matter for regret that the embryo of this animal has
not yet been obtained, as no doubt considerable light would
be thrown on the degeneration of the eye by the study
of its early stages.

In seeking to trace the homologies of the eye of Noto-
ryctes typhlops, in the following account of its structure, I
have refrained from making comparisons or contrasts with
other forms, except in such cases as seem of assistance or
special interest.

Position.

The whole eye, with its gland, the nerve, blood-vessels,
muscles, etc., is situated at a distance of 13 mm. from the
anterior end of the snout, deeply seated beneath the tem-
poralis muscle, and often covered with a mass of connective
tissue (contrast T'yphlops, where the ordinary subcutaneous
tissue is thinner [6, p. 119], and compare Siphonops [6,
p- 105] and Typhlomolge [2, p. 51]). In addition to this
the dermis and epidermis always pass over this region un-
altered in general structure and thickness, in contrast to Pro-

EYE OF NOTORYCTES TYPHLOPS. 549

teus, in which it is thicker [6, p. 75], and Siphonops [6,
p. 105] and Typhlops [6, p. 117], in which it is thinner.
From the exterior, therefore, even after shaving off the hair
(contrast Scalops [8, p. 335] and Talpa [6!, p. 13]), there is
no sign whereby the presence of an eye may be detected
(compare Myxine [6, p.49], Troglichthys rose [1, p. 578],
Amblyopsis [1, p. 560], Typhlichthys [1, p. 570], and
Rhineura [4, p. 535] ; and contrast Zygonectes [1, p. 548],
Chologaster [1, p. 549], Siphonops [6, p. 102], Proteus [6,
p. 72], Typhlops [6, p. 116], and Typhlotriton [8, p. 34],
Scalops [8, p. 335], and Talpa [6', p. 13], in which latter
three it can be found by the eye cleft). Of those forms in
which the deeply-seated position of the eye is comparable to
that of Notoryctes, may be mentioned Myxine, Amblyopsis,
Rhineura, and especially Troglichthys, in which, as here, its
wall is in contact with the skull.

There are, however, to be found in Notoryctes in the head
region, in common with the region of the modified “ ischio-
tergal”’ patch, curious organs formed as modifications of the
epidermis. These are presumably tactile in function, and will
be described in detail in Part IV of these Contributions.

The eye, which is not much more than a hollow ball of pig-
ment, lies within the anterior wider end of a fibrous cone-
shaped capsule (fig. 1, fic.), 5°25 mm. to 5°7 mm. long, 1:14
mm. to 1°75 mm. in vertical diameter, and ‘7 mm, to ‘87 mm.
in transverse horizontal diameter. This capsule is attached
by its small posterior end to the bony wall of the skull, and
passes forwards close against the periosteum of the lachrymal
bone, being only partially open in front. This is not to be
confused with the fibrous capsule mentioned by Dr. Stirling
[11, p. 180], as will be seen later. Structures comparable to
this conical capsule are found only in Chologaster papil-
liferus [1, p. 550] and Amblyopsis speleus [1, p. 560],
in each of which cases orbital fat is found inside with the
eye and eye muscles; in Notoryctes, fat has only been seen
in one single eye, and that in small quantities. In Siphonops
(6, Taf. vii, figs. 66 and 67] a somewhat similar fibrous cone

VOL. 50, PART 4,—NEW SERIES. 40

550 GEORGINA SWEET.

covers the front of the eye, immediately beneath the skin,
instead of lying beneath the muscle and connective tissue as
in Notoryctes.

Muscles.

In the posterior small end of the fibrous capsule are to be
found the bands of muscle (fig. 1, m.) which represent the
degenerated eye-muscles, being attached posteriorly to the
wall of the skull. There are four constant muscles usually
lying beneath the eye mass. Often there is a fifth slp of
muscle, separated off from the upper and most external one
of these, which very early becomes attached to the pyriform
fibrous capsule around the eyeball. The other four continue
forwards for a varying distance (fig. 1), sometimes stopping
behind the level of the pigment ball, sometimes, but rarely,
extending to the level of its anterior border. Invariably,
however, they end in a more or less broken or irregular
fashion in the outer fibrous capsule, or rarely within it,
amongst the connective tissue, which is variable in quantity.

Variations are met with in the proportionate length of this
hinder muscular region. Most frequently it is merely one
sixth of the total length, in which case the gland mass is
larger. In several instances the muscles are one half to
two-thirds the length of the capsule, the gland being then
proportionally smaller. At the extreme posterior end of the
capsule there enters an arterial twig from the internal
carotid, which appears to supply these muscles. At the
anterior part of this muscular region there leaves the capsule
a large vein, which runs downwards and outwards to join the
facial vein.

These muscles are, in contrast to those of Mammalia gene-
rally, of the non-striated kind, resembling them in size,
shape, character of the nuclei, and apparent absence of
sarcolemma. As far as can be seen those which are external
and inferior in the present condition are, as a rule, larger and
run the farthest forward, in other words, degenerate last.

EYE OF NOTORYCTES TYPHLOPS. bol

It is probable, however, that the present positions of these
muscle slips are not indicative of their homology with the
muscles of a normal eye. ‘They are very variable both in
length, position and size, and having changed their insertion
from the true eye capsule to the abnormal fibrous cone, and
lost their striated character, it seems not unreasonable to
suppose that they may have lost their typical positions rela-
tive to each other. The condition of degeneration of these
muscles may be compared with that in Amblyopsis [1,
p- 561] and Troghchthys [1, p. 579], and contrasted with
that in such forms as Scalops, Proteus [6, p. 77], Typhlo-
triton [8, p. 35], Siphonops [6, p. 100], Typhlops [6, p. 1241],
Talpa [G, p. 30], where they are well developed, or with
such as T'yphlichthys [1, p. 571], Typhlomolge [2, p. 52],
and Rhineura [4, p. 537], where they are still more degene-
rate than in Notoryctes, and even sometimes absent alto-
eether.

In Proteus, although the six muscles are present, the indi-
vidual fibres are, according to Kohl, in an embryonic condi-
tion, not having yet developed into striated fibres. In
Typhlops and Talpa also he considers that the unstriated
fibres often found mingled with others are evidence of the
imperfect development of the muscles, while in Myxine
[6, p. 51; and 6}, p. 195] the masses of connective tissue in
the place of the recti are accepted as being the predecessors
of the true muscles, and not degenerate eye muscles. I
cannot, however, but think that in Notoryctes at all events
we have eye-muscles which have degenerated in company
with the remainder of the eye, unless, indeed, we are to
think that these fibrous bands are neither the relics nor the
forerunners of the normal eye-muscles, but are new develop-
ments in connection with the abnormal conical capsule.

Glands.

At about the level of the vertical plane of the front of the
muscle region there appears the enormous development of

552 GEORGINA SWEET.

elandular material (figs. 1—4, l.g.), which, as will appear
later, is undoubtedly the representative of the lachrymal or
Harderian glands, probably both; of normal forms. The
histological structure of this mass, which often occupies most
of the conical capsule, shows it to be a true serous gland,
comparable exactly to the parotid of the same and other
animals. In one instance, in which the whole head had
been cut in a series of transverse sections, on one side there
was seen an even greater development of glandular material
outside and posterior to the conical capsule, and extending
immediately beneath the dermis. It was, however, con-
nected anteriorly with the gland inside, with which its duct
also communicated. In two other eyes, which had been
removed for sectioning purposes, there was a similar large
development of gland outside the capsule. In these speci-
mens the conical capsule is more or less undefined and irreg-
ular in its posterior portion, but is normally definite ante-
riorly. In harmony with the broken character of the outer
capsule in the two latter cases, and the overflow of the gland
beyond the capsule to which it is usually confined, the
muscle bands appear to he ina special mass of connective
tissue which is here strongly developed, but anteriorly they
end, one or sometimes two, in the conical outer capsule, and
the others in the connective tissue outside the capsule in-
stead of insideit. These are really the most degenerate cases.

Embedded in the gland mass within the cone lies the pyri-
form remnant of the eye proper (figs. 1, 3, 4, pl. 51), generally
‘87 mm. to ‘96 mm. behind the front of the tube. It always
lies closely apposed to the bony wall, from which it is only
separated by the periosteum and the conical capsule. The
oland mass does not enter this part of the cone (figs. 1 and 4) ;
it appears to do so in fig. 3 to a small extent, but in the
succeeding sections it is quite absent here also. In front of
the eye a varying amount of gland is present, and also some-
times a small fibrous band, running straight, forwards and
outwards, for a very short distance from the pigment wall, or
a line of pigment in a similar position, The main feature of

EYE OF NOTORYCTES TYPHLOPS. 553

this part, however, is the large irregular sac (figs. 1 and
2, ¢.s.), the hinder wall of which is in close contact with the
median anterior and antero-lateral part of the pyriform
fibrous capsule (p.c.), which is here thinner though more
compact than elsewhere. The size of this sac varies as follows :
Length (antero-posterior) *7 mm. to ‘52 mm., vertical diameter
‘02 mm. to “42 mm., horizontal diameter 42 mm. Into its
cavity there open some four to six ducts (l.g.d.) from
the large lachrymal gland, while from it anteriorly there
pass off two ducts, the small and outer of which (fig. 2, e.d.)
runs obliquely outwards and forwards for a variable distance,
‘ol mm. to ‘79 mm. long, and ends blindly in the dermis near
the skin. In the case of the shorter of these outer ducts the
pigment in its wall is so thick as to obscure its minute struc-
ture. The larger and inner of the ducts, however, descends
obliquely forwards and inwards, and then passes through a
definite opening in the skull wall, which is visible in the
dried skull (contra Dr. Stirling [11, p. 162]), near the lachry-
mal notch, sometimes higher or lower.

The sac (c.s.) in front of the eye is lined by somewhat
columnar epithelium, continuous with that of the two ducts
leading from it, that of the exterior blind duct (e.d.) becoming,
however, more cubical and thick-walled in structure. ‘The
sac also is covered with a thick layer of circular muscle-
fibres, which becomes much thinner both where the sac wall
is in contact with the eyeball, and over that part of the
duct outside the skull wall, and is almost lost on that part
inside the bone. In both sac and duct there is often to be
found a quantity of secretion. The duct then descends
from its entrance, as a definite tube through the skull
wall, obliquely downwards and forwards close to the wall
of the skull in a distinct groove (rarely absent) formed by
the angle of the bony floor of the lateral cavity of the
nose and its external side wall. It is separated from the
olfactory mucous membrane by a more or less developed
glandular mass, of which more has been said when dealing
with Jacobson’s organ in Part II of these Contributions. In

554. GEORGINA SWEET.

this position, where it hes dorsal to the ophthalmic nerve, the
duct is definitely flattened, and is lined by columnar epithelium
with very darkly-staining nuclei, the inner border of the cells
of which appears slightly cuticular, while it is surrounded by
concentric fibres. The duct soon comes to lie nearer to the
middle line, and then sinks with a small artery, a veiulet, and
a small bundle of nerve-fibres from the ophthalmic nerve,
into a depression in the dorsal surface of the maxillary bone.
This becomes enclosed anteriorly forming a canal [12, fig. 5,
n.l.d.] by which the duct now more cylindrical in shape, and
its accompanying structures may enter the inferior meatus of
the nose. About the level of the posterior end of the organ
of Jacobson it runs in a groove in the bony jaw, which forms
the lateral boundary of this part, and it is here flattened and
still lined by columnar epithelium. Its greatest diameter
(vertical) is, in this region, °25 mm. It then descends again
into a canal [12, fig. 4, »./.d.] in the bone, which it leaves
opposite the Stenonian canals to lie near the bony floor of
the nasal cavity [12, fig. 3, n.l.d.] with its artery and veins
between the premaxillary bone and the cartilage which forms
the floor of the nose in this anterior region. Immediately
behind the connection of the ali-nasal cartilage with the ven-
tral cartilages it runs towards the middle line embedded in a
gland mass, becoming more and more cuticular inside and
quite flat (‘25 mm. in greatest vertical diameter), until it
opens into the nasal furrow by an oblique aperture, on the
under surface of the small primary lateral ridge or ‘‘ concha”’
found near the external opening of the nose.

With regard to the general course of this naso-lachrymal
duct, for such it undoubtedly is, we may see by reference to
Klein’s descriptions of this structure in the guinea-pig [7,
p- 224] and rabbit [7, p. 567], that it is here very similar to
what he has described in those forms, as also its accompanying
artery and veins. He, however, has described the epithelium
as a stratified columnar layer, the columnar cells being next
the lumen of the duct, and more cubical cells outside this. I
find an identical appearance of nuclei, but consider that there

EYER OF NOTORYCTES TYPHLOPS. 555

is only one layer of ordinary columnar cells which are cut
obliquely, so that there appear to be two or three layers of
nuclei, whereas there is but a single layer of cells.

It is interesting to compare the state of development of this
gland mass, conjunctival sac, and ducts in other forms. Thus
we find in Typhlomolge [2, p. 52], whose eye that of
Notoryctes resembles in many points of structure, that there
is no sign of a conjunctival sac, nor is there any glandular
structure connected with the eye. In Rhineura likewise [4,
p- 536] there is no conjunctival sac, but Harder’s gland is
well developed ; its secretion, on the other hand, is poured
directly into the tear-duct, and so into the nasal cavity. In
Siphonops [6, p. 101—105] we have a very well-developed
gland, and also a conjunctival sac, into which, in one case,
Dr. Kohl found the lachrymal glands emptying, and from
which the naso-lachrymal duct passed off through the sur-
rounding tissues and nasal wall, to open into the nasal cavity
far forwards, as in Notoryctes. More generally, however, in
Siphonops, the gland tubes open directly into the naso-
lachrymal duct, while in T'yphlops [6, p. 119—121] they open
into a duct which leads into the back of the mouth cavity.

The relations of the glands, which are apparently true
Harderian and lachrymal glands are similar in Typhlops to
those in Notoryctes, as also the presence of blind ducts
ending in the subcutaneous tissues.

In Scalops, Talpa, and Typhlotriton we find instructive
stages in the closing up of the conjunctival sac, intermediate
between that of Notoryctes and the normal eye. In both
Scalops [8, pl. xvii, figs. 4, 7, 8; and pl. xix, fig. 16]
and Talpa [6!, Taf. i, figs. 3, 4, 5] the eyelids have closed
over to such an extent that the only connection between this
pre-corneal space and the surface of the body is by an open
though small canal, which is of very little if any use for the
passage of light rays.

In Typhlotriton [8, p. 41, figs. 1 and 1 a], the lids are
merely overlapping slightly, a shallow groove indicating the
position of the eye cleft. The comparatively small amount

556 GEORGINA SWEET.

of degeneration in Talpa is also indicated by the fact that
the Meibomian glands connected with the eyelids are still
functional though small, indicating that the closing over of
the lids is but recent [6', p. 26]. Although Scalops is
undoubtedly “ much more degenerate in all its parts” than
Talpa [8, p. 361], Notoryctes leaves Scalops far behind in
this matter.

Bloodvessels.

The arteries supplying the eye region are very large in
proportion to the size of the eyeball, to cope no doubt with
the demands of the large lachrymal glands. They are
derived as usual from the external carotid and facial arteries,
with a twig from the internal carotid artery entering among
the muscles at the apex of the conical capsule. As pre-
viously stated the large vein empties itself into the facial
vein.

Nerves.

The second or optic nerve is discussed in connection with
the eyeball itself.

Of the third or motor oculi, and fourth or trochlear nerves,
I can find absolutely no trace in any specimen, either in
connection with the eye or the brain.

The sixth or abducens is equally wanting, though in one
animal there was to be seen extending through some two or
three transverse sections (of the whole head) a slight swell-
ing on the ventral surface of the medulla oblongata, near the
middle line, just where one might expect to find the abducens
leaving the brain, though I could find no sign in its structure
of its being the root of a nerve.

The gland mass receives its innervation from a branch of
the ophthalmic nerve, which it leaves in company with the
nasal branch of that nerve, again proving the relationship of
this gland with the lachrymal gland of the other forms.

The muscles, now no longer accessory optic structures,

EYE OF NOTORYCTES TYPHLOPS. 507

and so receiving no innervation from the normal source,
come to derive their nerve supply from a twig of the
lachrymal branch of the ophthalmic nerve, viz. that which
supplies the lachrymal gland.

The following data from one typical eye will help to give a
better idea of the relative sizes of the parts described :

Length of whole conical tube : . 5:25 mm.
Vertical diameter of conical tube (in front) . 1:14 ,,
e a : (greatest) . 1:75 ,,
a * PY (behind eye) 1:40 ,,
Horizontal __,, ig (iaeronG) enn oe ee.
Fe be ns (greatest) . ‘87 ,,
3 - if (behind eye) °75_,,
Distance of eye from front of tube = 9D 4,
Length of eye : : eo Onn:
Vertical diameter of eye . ; ee Oaee
Horizontal diameter of eye , sa) SOO. Gs
Length of conjunctival sac (antero-posterior) ‘7 ,,
Vertical diameter of sac. a ero (mes
Horizontal . 42 ,,

On comparison with Talpa and Scalops [8, p. 357] it will be
seen that the eye of Notoryctes is slightly bigger than that
of Scalops, though distinctly smaller than that of Talpa.

The Hyeball.

In size this varies from *96 to 1:1 mm. long, ‘61 to 1:05 mm.
in vertical diameter, and *52 to ‘87 mm. in horizontal dia-
meter.

Tt is enclosed completely in a tough pear-shaped capsule
(figs. 1, 3, 4, p.c.e.), consisting of closely layered fibres with
scattered nuclei, and occasionally with small elongated
patches of granular pigment in its inner part.

Rarely one could detect a small bloodvessel running in
this layer. It would seem that this must be regarded as a
sclero-choroid, the boundary between the two membranes

558 GEORGINA SWEET'.,

being indistinguishable. In two eyes a small oval nodule of
hyaline cartilage (fig. 4, ¢, 1), containing some six cells, is
to be found lying in the sclero-choroid, just on the inner
side of what ought to be the exit of the optic nerve. In
another case a similar though smaller nodule was found lying
in front of the eye, a little to its outer side. The structure
of this sclero-choroid is very much like that of Rhineura
[4, p. 537] among others. As in that also this layer is pro-
longed proximally, forming a sheath which must represent
the covering of the optic nerve, being connected posteriorly
to the skull along with the conical capsule. Small bars of
cartilage comparable to the nodule are also found in Rhineura

and Amblyopsis [1, p. 563}.

Pigment Epithelium.

Immediately within this sclero-choroid lies the conspicuous
hollow ball of pigment, dense and thick-walled anteriorly
and exteriorly, as well as above and below, and very thin on
its inner side and posterior end. At first, in the absence of
embryonic material, one is led to consider this as a much
thickened choroid, but, on comparison with other forms in
which developmental changes can be followed, it is seen to
be probably the greatly changed pigment epithelium of the
retina, though there is nothing in the irregular, broken, and
jumbled masses of granular pigment to suggest such a well-
defined cell-layer as the retinal pigment. No trace of cell-
structure can be seen enclosing the pigment, though, in
occasional gaps between the masses, small oval cells can be
seen without any granules within them. There are. never
any processes inwards or outwards. It will be noted that,
as in Troglichthys [1, p. 581], and Typhlomolge [2, p. 53,
and pl. 3, figs. 1, 6, and 7] the pigment is much denser in
front, where it might be expected to be absent, and very thin
or wanting posteriorly, where it ought to be thicker. In
this these three forms differ from all other known degene-
rate vertebrate eyes. ‘The irideal region is much reduced

EYE OF NOTORYCTES TYPHLOPS. 559

insomuch that, as a rule, no remnant of iris epithelium, pupil,
lens, vitreous humour, or hyaloid membrane can be seen, the
unbroken wall of pigment forming the front of the eye. In
this condition the eye of Notoryctes is almost exactly com-
parable to that of Troglichthys, which Eigenmann claims as
the most degenerate of all vertebrate eyes. In the less
degenerate specimens, however, structures are present which
appear to be the remnants of those parts. ‘Thus in one eye, a
little to the outer side of what may be taken as the optic axis,
is a Slight gap, in which the pigment epithelium is much less
dense than around it. This gap is tubular, sloping obliquely
outwards and forwards; in it posteriorly can be seen a double
layer of cells with oval and flattened nuclei, lying, not edge
to edge, but obliquely against one another. ‘hese are
continuous internally on either side, with a similar layer
sometimes two or three cells thick instead of one, and
extending over the anterior region of the eye. In one or
two other eyes these flattened cells are present in a less
definite manner over this anterior region, but more restricted
in extent, and not being connected with any gap. There
ean be no doubt, I think, that this imperfect gap represents
the last vestige of a pupil piercing the greatly-developed
pigment layer (pars iridis), the two walls of the gap being, in
that case, the edges of the iris. Similar vestiges have been
found by EKigenmann in Amblyopsis [1, Taf. xi, fig. 9, Taf.
xi, figs. 31 and 38, Taf. xiv, fig. 40], in Typhlichthys [1,
Taf. xiv, fig. 48], and in Troglichthys [1, Taf. xiv, figs. 54
and 56], and Typhlomolge [2, plate 3, figs. 6 and 7]. On
the other hand, the iris and lens are fairly well developed
in such forms as Petromyzon [6, Taf. ii, figs. 18 and
19], Chologaster [1, Taf. xi, fig. 4], Siphonops [6, Taf. vii,
fig. 67], Typhlops [6, Taf. viii, fig. 84], and Talpa [61, Taf.
iii, figs. 27 and 28]. In Proteus [6, Taf. vi, fig. 59] the lens
is absent in the adult, but the iris is readily distinguishable,
though reduced.

In Typhlomolge [2, pp. 53, 54] Higenmann considers the
pigment filling up the pupil as being of choroidal origin.

560 GEORGINA SWEET.

There is, however, nothing in Notoryctes to indicate that it
is any different from the rest of the pigment of the eye, i.e.
it is uveal in character. Apparently the edges of the pupil
have fused almost completely. Rhineura [4, p. 537], like
the more degenerate cases of Notoryctes, is even more
reduced than these forms since there, even this rudiment of
the irideal epithelium has gone. The lens was but rarely
seen in Rhineura, while in Typhlomolge it has gone alto-
gether. In respect of the iris Scalops [8, pl. xviii, fig. 7,
and pl. xix, fig. 9] apparently resembles Rhineura, but in the
mole the lens is well marked, though only consisting of cells.

Vitreous Humour, Hyaloid Membrane, Retina, and
Optic Nerve.

With regard to these, there seem to be three stages in
reduction in Notoryctes. I. The first and most highly de-
veloped state found in this “mole” is that of which one sec-
tion is shown in figs. 1 and 4. In it the retina shows a division
into outer nuclear and outer molecular layers, and less defined
inner nuclear and inner molecular layers. The section drawn
shows fibres coming from a group of cells at the anterior
end, evidently the remnant of the ganglion cell layer, and to
this extent is reminiscent of Amblyopsis [1, Taf. xiii, fig. 34],
Typhlichthys [1, Taf. xiv, fig. 46], Troglichthys [1, Taf. xiv,
figs. 4 and 56, and 6, Taf. vin, fig. 77], and Proteus [6,
Taf. v and vi, figs. 54, 55, 59]. In the next sections to that
shown in the figure the outer molecular layer is not visible,
the inner molecular Jayer being, on the other hand, more
sharply defined. In the sections ventral to that drawn there
appears a longitudinal split (the vitreous cavity) separating
the fibres into two layers, and bounded by a thin membrane
with elongated nuclei, evidently the hyaloid membrane.
Anteriorly the split spreads in a Y-shaped manner, its an-
terior face being formed by the flattened cells of the uveal
layer of the iris, before described (cf. Typhlichthys [1, Taf.
xiv, fig. 42], Typhlomolge [2, plate iii, figs. 2, 6, 7], and

EYE OF NOTORYCTES TYPHLOPS. 561

Proteus [6, Taf. vi,- fig. 59]). In this eye there was no
choroid split nor pupil visible. The fibres from the “ nerve-
fibre layer ”’ collect posteriorly, and leave the eyeball as shown
in fig. 4; passing individually through the pigment and
sclero-choroid layer which is not broken here, and then back-
wards, becoming lost among the connective tissue fibres,
which form a sheath for them, continuous with the sclero-
choroid itself, which “tails off’? here posteriorly. Un-
doubtedly these fibrils represent a degenerate optic nerve,
but it is impossible to actually prove their nervous character,
as I have no fresh material to work on. Indeed, they might
readily be accepted as fine connective tissue fibres with
elongate nuclei, lying lengthwise between the fibres, there
being no difference whatever in appearance, as indicated by
fig. 4. In one other section there was a distinct but narrow
cleft in the pigment wall near the same region, but nothing
could be found passing through. Nor in the material at
hand have I been able to detect any differentiation in the
cells of the retinal layers. One is led to believe that a thin
more or less definite layer of cells, exterior to the ‘‘nerve-
fibre” layer, represents the ganglion cells, but no difference
in structure can be seen from cells with slightly more deeply
staining nuclei, which are scattered all through the nuclear
layers.

In the other two stages of degeneration the cavity of the
ball is occupied by a mass of cells, generally much crowded
together, without any trace of special arrangement.

In Stage II (to which belongs the eye in which the indi-
cation of a pupil was seen, as described above) a choroid
fissure was present at the outer and ventral side, which
opened into a semicircular, very narrow slit-like cavity,
bounded posteriorly by a hyaloid membrane. It was almost
filled by connective tissue fibres, which have entered from
the sclero-choroid by the choroid split (cf. Myxine [6, Taf.
iv, fig. 40], Typhlichthys (1, Taf. xiv, fig. 49, and pp. 575-6],
and Rhineura [4, pl. xxxiv, fig. 4, and p. 538]. The anterior
boundary of this vitreous chamber, such as it is, is formed in

562 GEORGINA SWEET.

Notoryctes, as in Typhlichthys [{1, Taf. xiv, fig. 46], by the
rudimentary epithelium of the pars iridis retinalis. No nerve
fibres could be detected in this stage within the pigment
epithelium, though the “tailing off” of the sclero-choroid to
form a sheath was well marked, and the elongate nuclei
between the fibres outside the eye (? connective tissue or
nerve-fibres) were very regularly arranged in this specimen.
Just within the pigment wall, on the outer side of this one
eyeball, is a definite, very narrow layer of cubical cells, with
deeply staining nuclei, extending from the hinder edge of
the vitreous split, posteriorly to the proximal end of the eye.
I am unable to suggest any homology for them, unless they
be concerned with the pigment epithelium, which is specially
thick just here in this specimen. = *

In this eye also the walls of a small capillary blood-vessel
could be seen running straight inwards and across the eye
from the choroid split, and losing itself in the retina, some-
what similar to the eyes of Proteus [6, p. 86] and Typh-
lichthys [1, p.576]. No blood was present in it.

Stage III.—A choroid fissure filled with connective tissue
was present in some cases, but no vitreous chamber (cf.
Amblyopsis [1, laf. xiii]) nor differentiation of cells of any
kind within the eyeball, and no nerve fibrils. This repre-
sents the lowest stage of degeneration, to which the eyeball
has reached in any of these blind vertebrate eyes.

No Miiller’s fibres nor their nuclei, rods, nor cones have
been seen under any conditions in Notoryctes. There seems
to be a marked tendency in all the preparations of Noto-
ryctes eyes for the retinal cells under the influence of re-
agents to separate away from the enclosing pigment (figs. 1
and 4).

Of optic nerve-fibres the most careful examination has
failed to show a trace other than the fibrous appearance in
one eye noted above, and seen in fig. 4 (0. n.f.); though in
one specimen, in which the whole conical capsule was
mounted intact and unstained, there is an undoubted though
very short and faint double line of pigment, ‘41 mm. long,

EYE OF NOTORYCTES TYPHLOPS, 563

which must certainly indicate the position of the sheath for
the nerve., I have examined both brain and eye very
minutely, both by means of dissecting lens and microscopic
sections through the brain and eye and through the whole
head, and have found no further trace of it. In one speci-
men the dissecting lens showed a connective-tissue connec-
tion of the conical eye capsule with the brain, but this is
seen to be merely superficial on close examination, and is
probably the remnant of the sheath of the optic nerve still
connected with the membrane of the brain,

This may be compared with the pineal eye, in which
although clear remnants of the eye may still persist in many
cases, the nerve is completely wanting. In Amblyopsis
[1, p. 568], although the optic nerve can be traced to the
brain in the young, it is not so in the older form. In Typh-
lichthys [1, p. 574] the nerve is not so distinct in the eye,
but can always be traced to the brain (cf. Myxine also [6,
Taf. iv, fig. 40|). In Siphonops [6, p. 114] as in Notoryctes
it or its relic can only be seen for a short way from the eye,
no connection being found with the brain.

It will be seen that the eye of Notoryctes in all its present
stages is much more degenerate than is that of Talpa or
Scalops, its analogous forms in other parts of the world. In
each of these the vitreous humour, lens, retina, and optic
nerve are comparatively well developed, the retina in Scalops
being simply over-crowded, while the optic nerve is normal
in the adult,

SUMMARY OF STRUCTURE AND COMPARISONS.
A. Structure.

1. The eye has retired far beneath the skin which passes
over it unaltered but for the presence of sense organs (? tac-
tile).

2. A conjunctival sac is present, and the lachrymal glands

564 GEORGINA SWEET.

are extremely well developed—both being concerned with
some function not connected with the power of vision.

3. Eye muscles—abnormal in position, structure, and very
variable in development. Their usual nerve-supply is absent,
its place being taken by a branch of the ophthalmic nerve.

4. Sclerotic and cornea not distinguishable from one
another, nor from the choroid.

5. Lens absent always.

6. Vitreous body is practically absent, even in the least
degenerate forms.

7. Pigment layer of retina very thick distally, and thinner
proximally.

8. Pupil absent, iris being only represented by a few
elongate nuclei in some cases. These also are absent in
other eyes observed.

9. Rods and cones are absent, simplified nuclear and
molecular layers being present rarely—otherwise the retina
is represented by an undifferentiated mass of cells.

10. Optic nerve-fibres. Probable remnants are found only
in one case within the eye. They cannot be traced towards
the brain except rarely, and for a short distance by its
connective-tissue sheath.

B. Comparisons.

To summarise the most important of these.

1. Aquatic Forms.—Of these Troglichthys and Ambly-
opsis are the most nearly comparable in structure with that
of Notoryctes (higher stages). Typhlichthys and Typhlo-
molge have reached a similar condition in many points, but
the muscles in each are more degenerate, and there is no
conjunctival sac or lachrymal gland, and in Typhlichthys the
nerve can always be traced to the brain.

The eye of Proteus is, in all points, more highly developed
than that of Notoryctes, the retina only being similar to that
of the higher stages of this form.

9. Burrowing Forms.—In Siphonops we find most of
the parts such as the muscles, iris, and lens well developed,

EYE OF NOTORYCTES TYPHLOPS. 565

as also the gland and duct, which leads either into a con-
junctival sac, or direct into the nasal cavity. The optic
nerve, on the other hand, is about in the same stage of
degeneration as in Notoryctes, there being no connection
with the brain.

In Typhlops all parts are comparatively well developed,
the large glands, with their blind ducts towards the skin,
being, however, very like those of Notoryctes, though the
destinations of the internal ducts are different.

In Rhineura the most degenerate of previously-described
eyes in burrowing animals the muscles and conjunctival sac
are absent, though the gland is well developed. The lens is
found but rarely, and a single cartilaginous bar is present in
the sclero-choroid.

In the presence of a choroid split, and the general struc-
ture of the retina, this eye resembles the higher stage of
that of Notoryctes, and, in the absence of an iris epithelium,
the lowest stage.

Scalops and Talpa are so much less degenerate than
Notoryctes that a comparison is needless.

It is, therefore, not possible to compare the eye of
Notoryctes with that of any other animal in toto, but,
omitting the lachrymal glands and ducts, which are well
developed in all burrowing animals, and the muscles con-
nected with the fibrous capsule, its higher stages are almost
identical with that found in Troglichthys, which is regarded
by EKigenmann as the most degenerate of vertebrate eyes.
Thus it will be seen that the more degenerate condition
found at present in Notoryctes is without a parallel among
the Vertebrata, consisting simply of a fibrous sclero-choroid
containing a hollow pigment ball filled with a mass of cells,
devoid of all arrangement, and without any nerve or blood
supply.

The presence of such abnormally well-developed lachrymal
glands and of the ducts in all burrowing animals—except
Scalops and Talpa, where they are as yet unaffected—is most
interesting. The blind outer duct present in Typhlops, and

VOL. 50, PART 4,—NEW SERIES. 4A]

566 GEORGINA SWEET.

sometimes in Notoryctes, remains in the less degenerate
forms as an indication of the path along which the conjunc-
tival sac was drawn inwards by the retreating eye. This
duct doubtless, as now in T'alpa and Scalops, formed for
some time a last direct communication with the exterior.
This later on was lost, since evidently it was not of much
use to the animal, the sole remaining escape for the secretion
of the gland then being through the nasal duct into the nose.
The action of the muscular layer round the conjunctival sac,
as also that of the muscle bands which are attached to the
conical capsule, would be, by their contraction, to increase
the pressure on the conjunctival sac, directly and indirectly.
Their common innervation with the gland alveoli also
suggests that possibly efferent motor fibres may be associated
with the efferent secretory fibres of the latter. The gland
must have now a considerable functional value, since, with
increasing degeneration of the eye and closing of the direct
passage to the exterior, the gland has increased in size.

Presumably the present function of the secretion is—(1)
to keep the snout and nasal cavity moist, and (2) chiefly, to
hinder the entrance or accumulation of particles of sand in
the nasal cavity when burrowing, as this animal does so
rapidly in the fine sand in which it lives.

Indeed, this cavity is often so full of coagulated secretion
that at first, in sections, no cavity at all can be found.

The least degenerate of the eyes which I have been able to
examine is that in fig. 1, in which the gland is small, the
blind duct of the skin much longer, larger, and more definite,
while the appearance of the nerve-fibres is absent in every
other case—and there is no doubt whatever that, as the
eye becomes more degenerate, so the gland increases in size

and importance.

CoNnCLUSIONS.

Higenmann [1, p. 546] remarks :—“It must be apparent
that an experiment on a vast scale has been conducted by
nature, leaving us but to read the results. Moreover, the

EYE OF NOTORYCTES TYPHLOPS. 567

experiment is one of evolution without the assistance or
intervention of natural selection.” This latter statement
may or may not be true in the case of the blind fishes with
which he is dealing. To explain their degeneration he
invokes the aid solely of “disuse” of the eyes in “ animals
already predisposed to shun the light, or creep under rocks
or into crevices” [4, p. 555].

But there can be no question of use or want of use in the
eyes of Notoryctes, which “is, in reality, a more surface
animal than the Huropean mole,” as already pointed out by
Professor Spencer in the report of the Horn Expedition [9,
peo:

Here several factors have come into play.

First, as the result of natural selection, degeneration has
taken place because the presence on the surface of the head
of such a sensitive structure would be deleterious to the
animal. ‘The fine grains of sand through which it burrows
would have been a fruitful source of irritation, resulting
constantly in the production of inflammation,” and, as Pro-
fessor Spencer continues, “ more than counter-balancing the
advantage to be gained from the possession of an eye when
it did come to the surface”’ [9, p. 51].

Second, when the eye had receded beyond the reach of
irritation this factor would no longer operate. ‘Then the fact
of disuse may have operated to intensify the degeneration
already well advanced, at all events in the accessory parts,
such as the eyelids and character of the muscles.

Thirdly, the great degeneration of the eyeball itself with
its nervous structures has undoubtedly gone on side by side
with the great development of the gland structures both in
the eye region and in the nasal region [see 12] itself an
adaptation to the burrowing habit, and so probably controlled
by direct natural selection. Possibly the used organ has
developed at the expense of the food and room of the disused
organ by the law of compensation and economy of growth.
It cannot be due to merely disuse or diminished use from
want of light, since disused organs that are not concerned in

568 GEORGINA SWEET.

the struggle for room or food may maintain themselves for a
long time.

Higenmann [1, pp. 600, 601] doubts the efficacy of the
struggle for room and food in the degeneration of the eyes of
the Amblyopsidee, since the position and room formerly occu-
pied by the eye is now filled with fat. But in Notoryctes we
found very little fat, and that only in one specimen, while on
the other hand the more degenerate the eyeball itself the
more highly developed the gland arrangements were found
to be.

As to the object of the burrowing of the animal we must
conclude that it is mainly for food, as suggested in the Horn
Expedition Report, and as a means of escape from its
enemies.

The loss of the eye as a means of knowledge of the external
world is compensated for by the great sensibility of the animal
to sound, and probably from the presence of the supposed
tactile sense-organs to touch.

Its relation in point of degeneracy to Scalops and Talpa is
interesting, since we may regard it in two ways. Hither the
eye of the Marsupial has had a longer time since it took to
burrowing life in which to become reduced than has that of
the insectivorous EKuropean Talpa or American Scalops, or,
more probably, the sand in which Notoryctes lives is, and has
been, more deleterious to the eye than the earth in which
Scalops and Talpa burrow, and so degeneration has gone on
more quickly in Notoryctes.

Professor Spencer suggests that this view is strengthened
by the evidence as to the recent past of Central Australia.
There we find deep gorges and broad river valleys compara-
tively intact, though now dry except on rare occasions, show-
ing that this region was favoured with a more liberal rainfall
at no very great distance of time. Then when the conditions
became dryer the decomposition of the rocks and the wearing
of their débris produced a finer and finer sand. Probably,
up to this time, Notoryctes was a burrower with its eyes in a
condition comparable to those of Talpa, and as the sand be-

EYE OF NOTORYCTES TYPHLOPS. 569

came finer and the temperature higher, its eye would still
further degenerate very rapidly with the increased liability
to irritation. This also confirms the previously expressed
opinion with regard to the cause of the degeneration of the
eye of the Notoryctes.

In the eye of Notoryctes, then, we have an example of a
very specialised sense-organ, degenerating in virtue of its en-
vironment, losing its original sensory function, and assuming
an importance of quite a mechanical nature, also rendered
necessary by reason of the habitat, and, further, this trans-
ference of function is even now in an incomplete and transi-
tional stage, though it has in some animals reached a point
of reduction not known in any other Vertebrate eye.

I understand that the brain of the Notoryctes is being in-
vestigated, and it will be of great interest to note whether
the sensory nuclei in the optic centres have kept pace with
the degeneration of their peripheral end-organ.

BroLocicaL LaBporatory,
MELBOURNE UNIVERSITY ;
February 28th, 1906.

BIBLIOGRAPHY.

1. Eigenmann, C. H.—*The Eyes of the Blind Vertebrates of North
America,” I. Amblyopside, ‘Archiv fiir Entwickelungsmechanik der
Organismen,’ Band viii, Heft 4, 1899.

2. Kigenmann, C. H.—* The Eyes of Typhlomolge rathbuni, Stej-
neger,” ‘Contributions from the Zoological Laboratory of the Indiana
University,’ No. 29, ‘Trans. Amer. Mic. Soc.,’ vol. xxi.

3. E1genmann, C. H.—< Typhlotriton speleus,” ‘ Biological Bulletin,’
vol. ii, No. 1, 1900.

4. Eicgrnmann, C. H.—‘The Eyes of Rhineura floridana,” ‘Proc.
Wash. Acad. Sci.,’ vol. iv, Sept., 1902.

5. Ergenmann, C. H.—“ History of the Eye of the Blind Fish Ambly-
opsis,” ‘Mark Anniversary Volume,’ Art, ix, 1903.

6. Kout, C.—‘ Rudimentare Wirbelthieraugen,” ‘ Bibliotheca Zoologica,’
Bd. v, Heft 18, 1892.

570 GEORGINA SWEET.

6!. Koni, C.—“ Rudimentire Wirbelthieraugen,” ‘ Bibliotheca Zoologica,»
Bd. v, Heft 14, und W. Nachtrag, 1893, 1895.

7. Kue1n.— Quart. Journ. Mic. Soe.,’ vol. xxi.

8. StonakER, J. R.—* The Eye of the Common American Mole (Scalops

aquaticus machrinus),” ‘Journ. of Comparative Neurology,’ vol.
xii, 1902, p. 335.
9. Spencer, W. B.—*“ Report of Horn Scientific Expedition to Central
Australia,” Pt. il, ‘ Zoology.’
10. Strruine, E. C.—‘ Trans. Roy. Soc. S. Australia,’ 1888.
ll. Stinuine, E. C.—‘ Trans. Roy. Soc. 8. Australia,’ 1891.
12. Sweet, G.— Proc. Roy. Soc. Victoria,’ vol. xvii (N.S.), Pt. 1.

EXPLANATION OF PLATE 31,

Illustrating Miss Georgina Sweet’s ‘ Contributions to our
Knowledge of the Anatomy of Notoryctes typhlops,
Stirling. Part I[I.—The Hye.”

REFERENCE LETTERS.

c.n. Nodule of cartilage. ¢.s. Conjunctival sac. e.d. Blind external duct.
J.c. Conical fibrous capsule. 7g. Lachrymal glands. ¢.g.d. Lachrymal duet.
m. Muscle bands. m./. Internal molecular layer. o.x./. Optic nerve fibrils.
p. Pigment layer. p.c.e. Fibrous capsule of eye =sclero-choroid. 7.7’. Cells
representing retina. s.@. Small artery. v. Veinlet.

All figures were drawn with the aid of the camera lucida, and Zeiss lenses
were used.

Fic. 1.—Longitudinal section of eye of Notorycetes typhlops, showing
long tube (f/c.) containing two muscle bands (m.) at this level, lachrymal
glands (/.7.) and ducts (/.g.d.), with the precorneal space (c.s.); the eye itself
with pigmented wall (p.), and remains of retina (7.), and surrounded by its
own capsule (p.c.e.); two small arteries (s.a.) and veinlet (v.) ; also nodule of
cartilage (c.z.) posterior to the eye. Zeiss A*, oc. 2.

Fic. 2.—Longitudinal section of anterior portion of conical tube enlarged,
showing same parts as Fig, 1, and, in addition, the longest of the external

KYE OF NOTORYCTES TYPHLOPS. 571

ducts (e.d.) running out towards the skin and ending blindly beneath it.
Zeiss A, oc. 4.

Fie. 3.—Transverse section taken across the tube through the anterior part
of the eye, showing parts as above, and, in particular, the close relationship
of the precorneal space to the sclero-choroid capsule. Zeiss A, oc. 4.

Fic. 4.—Longitudinal section of small portion of outer tube, showing same
parts as in Fig. 1, but more enlarged. This figure is compiled from two con-
secutive thin sections, and shows more clearly than does Fig. 1 the fibres
(o.z.f.) running from front to back of eye and through the pigment wall.
Zeiss A, oc. 4.

= » =

CANKER OF THE APPLE TREE. ie

Structure and Origin of Canker of the Apple
Tree.

By
James E. Blomfield, .A., M.D.(Oxon.).

With Plate 32.

Tue object of the present communication is to give an
account of the structure and origin of the tumours produced
on the apple by the woolly aphis, Schizoneura lanigera.

These tumours are familiar objects in many orchards, and
are well known to gardeners who call the disease ‘“ canker.”

My reason for investigating these tumours and their mode
of origin arose from the circumstance that in examining a
transverse section stained differentially in carmine and methy|
green there appeared to be a transition of the wood cells into
the tumour cells suggestive of a malignant process. The
literature of vegetable pathology, as far as I could gain access
to it, gave me no help in discovering the meaning of this
appearance. Prillieux had described the structure of the
tumour in ‘ Bull. de la Soc. Bot. de France,’ i875. His re-
searches are quoted by Kiister in his ‘ Pathologische Pflanzen
Anatomie, but this authority notes that the subject of
wood-galls requires more exact investigation. This book,
published in 1903, gives exhaustive references to previous
observations in vegetable pathology, and I think that I am
justified in concluding that there is no satisfactory account of
how these tumours originate.

574. JAMES E. BLOMFIELD.

The insect which produces this disease is an aphis of a
purplish-red colour. Crushed between the fingers it leaves
a blood-red stain, whence it derives its German name of
Blood-louse, but its chief characteristic which makes its pres-
ence easily noted is a fluffy covering of a white or greyish
colour, sticky and resinous in consistence which exudes from
tubercles on the back of the insect, and which, from the fact
that it is insoluble in water, serves to protect the insect from
wet and damp, and to make difficult its eradication from an
orchard when once it is seriously invaded by the aphis. The
material is soluble in alcohol, and I imagine, though I can
speak from no practical experience, that the reason that in-
secticides are so useless is owing to the fact that this resinous
material is not removed before they are applied.

The general structure of the Schizoneura is the same as that
of other aphides except that the cornicles are atrophied. The
rostrum in the immature forms is as long as, or longer than,
the body. The rostrum or haustellum is as in other Hemi-
ptera an extension of the labium. It consists of three joints
which are grooved on their upper surface to receive the sete
or lancing organs which prick the juicy parts of plants to
cause a flow of sap on which the insects feed. The set are
three in number, and represent the mandibles and maxille of
other insects. During the act of sucking the rostrum is
closely applied to the plant surface and secured there by
coarse hairs at its tip, the sete are run along their groove
into the soft plant tissue, which they lance and stab to ensure
a flow of sap which can be sucked up partly by capillary
attraction and partly by a pumping action on the part of the
insect.

‘he question whether salivary glands are present in all
aphides appears to be yet unsolved. They were found by
Buckston (‘ Brit. Aphides,’ Ray Society) in some specimens,
the name of which he does not give, but in connection with
the present subject the matter would seem to have some im-
portance because it is to the secretion of some such eland
that we must look as a cause for the peculiar action of the

CANKER OF THE APPLE TREE. IFAS)

gall-producing Aphides, Lachnus, and Schizoneura. The
common aphides of the ivy and the rose prick and suck the
soft young stems, but there is no specific reaction on the part
of the plant such as we see in the case under consideration.

The natural history of the Schizoneura is as follows. Dur-
ing the winter months the mature insects find shelter in the
cavities and crannies of the nodosities, and in the early days
of spring their presence is noticeable from the patches of
white fluff. These increase in size owing to growth in number
of the insects, and as the summer advances masses of sticky
fluff envelope the branches in which are found crowds of
aphides. Some of these become “ nurses ” and produce living
young parthenogenetically. A new generation is said to be
produced every fourteen days, and as the young twigs of the
tree grow new colonies are founded. The place where the
colony is started on the twig of the year does not appear to
be a matter of chance but rather of selection. At the time
that the twigs are invaded the young leaves are well formed,
the distal end of the stalk is green, but towards the parent
tree the green passes into a reddish-brown colour indicating
the formation of a periderm, which, in the case of the apple,
is derived from the epidermis. The place of selection is not
the green portion of the twig at the extremity, but nearer to
the old wood at a point where the wood has definitely formed
and the formation of the periderm commenced. Some of the
“nurses”? descend to the roots as in the Phylloxera and
establish colonies there producing deformities similar to those
on the stem.

If a tumour is selected for study on which the aphides are
actively feeding, and after fixing, hardening, etc., sections
are prepared and stained in a manner to differentiate the
tissue, such as by iron hematoxylin method and fuchsin, the
appearance represented by fig. 3 will be seen.

On the outside is a layer of cells two or three in thickness,
which is the periderm, beneath this is the cortex with strands
of sclerenchyma. If we follow the cortex over the tumour
we shall see that it has undergone slight alteration only, a

576 JAMES E. BLOMFIELD.

stretching and thinning from the growth that is taking place
beneath it. The sclerenchyma bundles are, however, less
numerous and less defined. Beneath the cortex is a layer
consisting of bast and cambium, the separation of which is
marked in all parts except over the tumour, where no distinc-
tion can be made. In the centre of the section is the pith,
surrounded on all sides by the wood, the continuity of whose
ring is interrupted by the tumour which dips into the wood
ring in a wedge-like manner. Asa rule it does not reach the
pith unless the section has passed near a leaf-shoot, which
may happen, because a favourite place for the young colony
is just above a leaf-shoot which serves to shelter and protect it
from wind and rain, both of which are very disastrous for the
propagation of the species.

In such a section as I have described all lignified cells are
stained pink with the fuchsin, and it is easy to distinguish
the soft cellular cells of the tumour from the lignified cells of
the wood. If examination is now made with a higher power
it will be found that no hard and fast line separates the
tumour from the wood, but that the tumour cells seem to
arise by alteration of the wood parenchyma cells. Among
the tumour cells, however, will be discovered pink-stained,
large, irregular cells, which evidently are altered, pitted, and
scalariform wood-vessels, enlarged in size, and irregular in
shape. These must have arisen from altered cambium cells.
‘he general arrangement of the cells of the tumour is in a
radial direction. Large oblong nuclei can be seen in each
cell, and in some cases there are several in each cell, a fact
which was pointed out by Prillieux. ‘The soft walls of the
cells are composed of cellulose, as shown by Schultz reagent.
There is hardly any starch present, as revealed by iodine,
but there is sugar in relatively large quantities. If a section
is made of a fresh tumour, and tested with Fehling’s solution,
this fact comes out plainly by the reduction that takes place
in the tumour. A rough quantitative estimation showed
l per cent. of sugar in the tumour, while small fragments
from the same stem hardly yielded any reaction at all. A

CANKER OF THE APPLE TREE. 577

small quantity of coagulated plasma may be found in each
cell (fig. 4).

If the cambial region of the tumour is examined with a
high power it is evident that this tissue is in a state of great
activity. In appropriate sections the tracks left by the sete
of the insect may be traced through the periderm and cortex
till they terminate in the cambium, and it is around these
terminations that the greatest activity is taking place. The
distinction between bast and cambium is made out with
difficulty, as both kinds of cells are enlarged, and contain
large, well-defined nuclei. In places the division of the
nuclei exceeds in rapidity that of the cells, with the result
that a multinucleated mass is produced. This cell division
no doubt takes place by mitosis, evidence of which I obtained,
but the material is difficult to cut with sufficient accuracy for
a study of this process. After the cells are produced by the
cambium further division is undergone, or multiplication of
nuclei may take place without corresponding cell division.
The protoplasm exhibits vacuoles, which increase in size till
the whole cell consists of a wall of cellulose, a small quantity
of plasma, with a nucleus and a large quantity of cell sap,
consisting chiefly of sugar.

That these changes are produced by the aphides is shown
by the fact that, if they are swept away by wind or rain, the
cambium resumes its normal activity, and gives rise to cells,
which pursue their destiny of lignification in a normal
manner, enclosing a portion of the tumour, which itself
undergoes lignification, but, from the displacement, increase
in number (hyperplasia) and in size (hypertrophy), the
elements are abnormally arranged, and produce a condition
which is known as wound wood (fig. 6).

If the aphides linger on their tumour and the weather is
dry the soft parenchymatous tissue may split and allow a
genus of destructive fungi such as Nectria to enter, pro-
ducing necrosis and ulceration of the tissue. This the plant
tries to counteract by its powers of healing, and new cambial
tissue is produced, to be quickly utilised by the aphides for

578 JAMES E. BLOMFIELD.

their nourishment, till large gnarled nodosities are produced
consisting of dead, necrosed wood, hypertrophied tumours,
and wound wood, which may attain the size of a man’s fist.
It is not necessary that the tumour be split for Nectria to
gain entrance. In comparatively young tumours the fungus
and its necrosing action may be seen, though there is no
breach of the surface except the punctures made by the
insect, and it is, no doubt, by these that the germ gains
entrance.

To shortly resume the origin of these tumours, we have
seen that they are produced by the pricks of the aphides.
That during this process some influence is brought to bear
on the active cambial cells which leads to their enlargement
and increase. That the cells are arrested in their normal
development and destiny, and that as long as this influence
lasts they serve the purposes of their parasitic victors, that,
when these retire, they are able again to pursue their deve-
lopment and destiny, but in such a way that the traces of
their experiences are not obliterated.

I have sketched out this view in the diagrams (fig. 7).

What is the agent of this influence which the Schizoneura
is able to exercise on the cambium of the young twig? As I
noted above, mechanical irritation we must dismiss as a cause,
and we can only fall back on the hypothesis of a ferment,
such as Beyernick suggested under the term growth enzyme.
This may come from the salivary glands of the Aphis.

I have tried to test this question by acting on a suggestion
made by Prof. Farmer of inserting a glycerine and water
extract of the insects by means of capillary tubes as near the
cambium as possible in such a way that the liquid would
constantly bathe the cells. No success has followed these
attempts. The slight reaction visible at the point of insertion
did not amount to more than that produced by a fine wire
inserted in a similar way.

CANKER OF THE APPLE TREE. 579

EXPLANATION OF PLATE 382,

Illustrating Dr. James KH. Blomfield’s paper on “ Structure
and Origin of Canker of the Apple Tree.”

Fic. 1 is a photograph of branchlet of an apple representing the growth of
three years. Each year’s growth, numbered successively 1, 2, and 3, is
infected by Schizoneura, and shows the characteristic boil-like swelling.

Fic. 2 is a photograph of two sections, one through a normal stem, and
the other through a canker. At * there is a recent swelling caused by
Schizoneura,

Fie. 3 is adrawing of a slightly magnified section through a young tumour,
A is periderm commencing in the epidermis. B is the cortex. C is a
sclerenchyma strand. D is the bast. E is the cambium. F is the wood.
G is the tumour with a few large vessels cut across.

Fic, 4 is a portion of a longitudinal section of tumour, x 700, showing the
pointed cells (W) which would have been wood-cells, and the square-shaped
cells (M) which represent modified medullary ray cells.

Fie. 5 represents a portion of the tumour in the region of the cambium.
A are cortical cells. B are modified bast cells. CC, modified cambial cells.
D, modified wood cells. , the track of a seta with proliferating cells in its
neighbourhood, x 700.

Fic. 6 shows a tumour passing into a condition of wound wood. ‘The
enlarged wood cells are twisted, dislocated, and separated by collections of
cells, which represent the hypertrophied medullary rays.

Fic. 7 represents in a diagrammatic manner the changes undergone by a
cambium cell (A) in becoming wood, and by an affected cambial cell.

The series a to f represents the normal course; a! tof’ the course under
the influence of the Schizoneura, with its cessation after a time

Cay

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 581

Review of Dr. Richard Goldschmidt’s Mono-
graph of Amphioxides.!

By
A. Willey,
Hon.M.A.Cantab., D.Sc,Lond., F.R.S.

Amone the many acquisitions of the German Expedition to
the Deep Sea in the 8.8. ‘‘ Valdivia”? (1898—1899), which
was organised under the direction of Professor Carl Chun,
not the least valuable was the relatively large series (27
specimens) of pelagic Acraniata belonging to the genus
Amphioxides, Gill. This material was entrusted to the
skilled hands of Dr. Richard Goldschmidt, who may be con-
gratulated on his important monograph, which takes its
place in a section of zoological literature associated with the
honoured names of Johannes Miiller, A. von Kolliker, de
Quatrefages, Kowalevsky, Hatschek, Huxley, Ray Lankester,
van Wijhe.

Dr. Goldschmidt’s memoir will almost certainly tend to
enhance the morphological importance of Amphioxus (if it be
permitted still to employ this name in a cursory, non-italicized
sense), based as it is upon such careful observation and
logical deduction. It is, however, frequently necessary to
discount the force of logic when dealing with discussions of
the kind before us.

The type species of Amphioxides was originally described
from a single specimen taken in the Pacific Ocean, under the
name Branchiostoma pelagicum by Dr. Giinther (1889)
in his report on the Pelagic Fishes collected during the
“Challenger” Hxpedition. Further examples from the
Indian Ocean have been recorded by C. Forster Cooper

1 Richard Goldschmidt, ‘A mphioxides,” ‘ Wiss. Ergebn. der deutschen
Tiefsee-Expedition,’ Bd. xii, 1905, pp. 92, ten plates and ten text-figures.

VOL. 50, PART 4,—NEW SERIES. 42

582 A. WILLEY.

(1903),1 W. M. Tattersall (1903),? and G. H. Parker (1904).8
Two other species have been added by Dr. Goldschmidt, A.
valdiview and A.stenurus. All three species were taken
during the “ Valdivia” Expedition at considerable depths in
the vertical tow-net in the high seas, often several hundred
miles from the nearest coast. Specimens were also obtained
in the Atlantic Ocean ;* the distribution of the genus is there-
fore circumequatorial, but there is no correlation between
the specific forms and their geographical range. In the Bay
of Bengal, 300 miles east of Ceylon, twelve examples were
captured simultaneously at one station between a depth of
2500 metres and the surface, of which nine belonged to A.
valdiview, three to A. pelagicus. Off the west coast of
Africa A. valdiviz was taken south of Teneriffe and A.
pelagicus in the Gulf of Guinea. Lastly, all three species
have been taken in the neighbourhood of the Seychelles.

No sexually mature individual has yet been seen. Dr.
Goldschmidt has found the immature gonads developing only
on the right side, lying in the gonoccel which is shut off from
the ventral ends of the myotomes. No specimens in the
“ Valdivia ”’ collection exceed 10 mm. in length. At this size
the gonads were observed to be as far developed as in a
Branchiostoma lanceolatum of 28 mm.

1 ©. F. Cooper, ‘‘ Cephalochorda,” ‘Fauna and Geography of the Maldive
and Laccadive Archipelagoes’ (J. Stanley Gardiner), vol. i, part 4, 1903, p. 352.

2 W. M. Tattersall, “Report on the Cephalochorda collected by Professor
Herdman at Ceylon in 1902,” ‘Ceylon Pearl Oyster Fisheries,’ part 1, 1903,

. 214.
: 3 G. H. Parker, “Maldive Cephalochordates,”’ ‘Bull. Mus. Harvard,’ vol.
xlvi, 1904.

4 It is interesting to note that no examples were procured during the
‘* Plankton” Expedition. Hensen (Hinige Ergebnisse der Plankton Exped.,
1892, p. 24—25) says that they frequently obtained young Amphioxus
lanceolatus up to some centimetres in length, as many as two to ten indivi-
duals in one catch of the Plankton Net in the North Atlantic. He notes that
it is remarkable that they should remain so long at the surface over great
depths, because Amphioxus is a coastal and littoral form, only the larve
being pelagic in the coastal zone. These observations are important as indi-
cating that the prolongation of the pelagic life does not involve a persistence
of the larval asymmetry in A. lanceolatus.

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 583

Amphioxides is remarkable for the possession of many
characters which are proper to the larva of the European
species of Amphioxus, e.g. the absence of a closed atrial
chamber, a sinistral slit-hke mouth, an unpaired series of
gill-clefts which lie in the mid-ventral line, an anterior dextral
endostyle, a club-shaped gland, and a sinistral preoral pit.
On account of these and some other characters the genus is
made the type of a new family, Amphioxidide Goldschmidt,
in contrast with the first family, Branchiostomide Bona-
parte, 1846.

The characters mentioned above are regarded by Dr.
Goldschmidt as being essentially primitive, emphatically not
as indications of a persistent larval organisation. In other
words, in his opinion Amphioxides is the most primitive
Acraniate, and stands more or less in the direct line of
Vertebrate descent.

One of the finest additions to our knowledge of the anatomy
of Amphioxus which has been made in recent years is
Professor J. W. van Wijhe’s discovery of the sinistral inner-
vation of the mouth. Anyone who has handled Amphioxus
will probably subscribe to this statement. The conclusion
drawn by van Wijhe from this discovery and accepted by
Goldschmidt, namely, that the mouth of Amphioxus is from
the beginning to the end an organ of the left side, may seem
to be clearly indicated, and is held by Dr. Goldschmidt to be
a fact of primary phylogenetic significance in Amphioxides
where the sinistral position of the mouth is said to be
dependent upon the structure of the pharynx. It may be
mentioned here that my own theory is still what it was
fifteen years ago so far as its essential point is concerned,
that the mouth, or situs oris, of the lancelet has migrated
from a dorsal position such as it holds in the Ascidian larva.

The pharynx of Amphioxides is characterised by the
presence in its floor of a median series of gill-perforations to
the number of thirty-four, opening directly to the exterior
on the ventral side of the body between the metapleural
folds. Above the gill-arches the wall of the pharynx projects

584. A. WILLEY.

inwards as a ridge on each side, delimiting a dorsal pars
nutritoria from a ventral pars respiratoria. The gill-
openings are simple clefts destitute of tongue-bars, but the
gill-arches which bound them are considerably folded, and
exhibit a bilaterally symmetrical structure being incompletely
divided by a deep median groove into right and left halves.
The gill-arches are somital, the gill-slits intersomital, in their
topographical relation to the myotomes.

The primitive condition of the pharynx of Amphioxides

Fic, 1.—A. pelagicus. Section through the middle of the
branchial region. After Goldschmidt. ao. Aorta. 4é.a. Branchial
artery. d.m. Branchial muscle. p.”. Pars nutritoria of pharynx.
p.r. Pars respiratoria.

would therefore consist in the median ventral series of un-
paired gill-slits, the partial separation of the ventral pars
respiratoria from the dorsal pars nutritoria, and, con-
sequent thereupon, the perforation of a sinistral mouth into
the dorsal division, and the development of a ciliated
glandular organ, the endostyle, opposite to the mouth, also
leading up to the dorsal division. The club-shaped gland is

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 085d

regarded as an accessory organ to the endostyle ; it is well
developed in Amphioxides, opening into the pharynx
behind the posterior end of the endostyle into the dorsal
nutritive portion. Goldschmidt has not found an external
orifice of the gland such as was first observed by Hatschek
(1881), and subsequently confirmed by Lankester and Willey
(1890), and Willey (1891) in the larva of Branchiostoma
lanceolatum.

Now there is a reflection which must occur to the minds
of those who may be conversant with the living larve of
Amphioxus, and with the extreme contractility of their
tissues, which may raise a doubt concerning the fundamental
importance of the lateral ridges of the pharynx of Amphi-
oxides as interpreted by Dr. Goldschmidt. There is not a
shadow of doubt that the gill-slits and gill-arches of
Amphioxides would wear a very different appearance
from that which they present in Dr. Goldschmidt’s excellent
figures if they were seen fully expanded with the body ina
state of turgidity, and it seems not unlikely that under those
conditions the projecting ridges! of the pharynx would
vanish and the folds of the gill-arches straighten out.

Again, it follows, from the interpretation of facts which
has been outlined above, that the anterior dextral position of
the endustyle, with its unequal limbs, is also a primitive
feature ; but Goldschmidt has not observed in Amphioxides
those paired ciliated peripharyngeal bands which proceed
upwards and backwards from the anterior ends of the endo-
style in the larva of Amphioxus, and offer such a striking
analogy with the similar organs of the Tunicata. These
bands, while distinctly pointing to an affinity with the
Ascidians, also indicate that the endostyle, although asym-

1 These ridges, the so-called limiting folds or plice limitantes, are
described as being lined by cylindrical flagellate cells, the long flagellum of
which is connected through the cell-body with the nucleus by a clear and
deeply staining rod. It seems highly probable, after all, that they are homo-
logous with the peripharyngeal bands of Amphioxus. Forster Cooper also
figures them in larve of Amphioxides taken at the Maldive Islands.

WILLEY.

A.

eo)
te)

‘yid jerowid Jo ajosnut OJOIUYsUOD "wad *

"qjnouw puv ytd eroaid waaa4oq suvormmuimtos snojng ‘o's “41d peromag ‘ded
itd £I0,ORJTQ *@. “(S40 AOsnJov0) dit samo] Jo aposnpy ‘wu ‘yqnow jo diy

TOMOT "M "BAAS [BIIOSIA IO spfoy avunfiwag *f ‘ayAysopuy ‘a ‘pueys podeys-qnig *4s'9 -sayosnur yerpoursg *ag

"payt[duis 4ywyaAowos “ypruosplox Jaqyy

WA

‘OPIS 4JaT OY} Wor Apoq oy} Jo uoldar JoWajUy ‘“BIAIpyea "y—'Z 2) Wt |

Te ZU ATE SMES

wn

= oO
GOOO00 OO OBO00G6 O00 GO6L000000

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 587

metrical in the larva of Amphioxus, was originally sym-
metrical.

The gill-shts of Amphioxides are arranged eumeta-
merically, that is to say, there is, in general, an exact
correspondence between branchiomerism and myomerism in
the branchial region. An individual having twenty-nine
gill-slits, for example, will have thirty gill-arches, of which
the first arch corresponds with the first myotome, while the
thirtieth arch lies in the thirtieth segment. This applies to
two of the species, A. pelagicus and A. stenurus. In A.
valdiviz repeated counting showed that there were more
gill-shts than myotomes in the branchial region, so that in a
specimen with twenty-seven gill-arches the last lay in the
twenty-third segment, or, more correctly, under the twenty-
third myotome. Such a specimen is interpreted as having
four supernumerary gill-arches ; the highest number observed
was seven. ‘lhe supernumerary arches are considered to
indicate the occurrence of prosomital gill-shitsin A. valdivie.
This unexpected interpretation appears simpler in the writing
than in the illustration (fig. 2). Those who can appreciate
it will be able to make the best use of it. For my part, the
lack of coincidence between gill-slits and myotomes in A,
valdiviz appears as an indication of the independence of
branchiomerism and myomerism in the Acraniata, however
closely they may be correlated.

In A. valdivie the ventral halves of the anterio: myo-
meres are bent backwards at avery acute angle in correlation
with the great length of the mouth, 1—1‘4 mm. (fig. 2).
This again strikes me as being a slight myomeric disturbance
independent of the gill-slits.

Above the gill-arches and between every two gill-slits the
pars nutritoria of the pharynx is fused with the body-
wall, the lines of fusion coinciding with the dissepiments of
the myotomes. Between two lines of fusion the body-cavity
extends upwards as a pouch to the base of the notochord.
Seen from the side, the lines of fusion appear as semilunar
folds (fig. 2). This is the inter-segmental concrescence of the

588 A. WILLEY.

gut with the myosepta, in other words, a partial segmentation
of the ventral mesoderm.

The metapleural folds closely resemble the corresponding
structures in the larva of B. lanceolatum; the right fold is
larger than the left (especially in A. valdivie), and reaches
farther forwards. Goldschmidt finds, in general accord with
previous observations on Branchiostoma by MacBride
(1898) and van Wijhe (1902), that in Amphioxides the
pterygoccel or metapleural lymph-space communicates in front
by a fine opening with the general body-cavity, on the left
side behind the mouth, on the right side in front of the
mouth. Posteriorly the metapleural folds terminate freely,
directly behind the last gill-slit, and are thus independent of
the median ventral fin as in Branchiostoma. Goldschmidt
says that they cannot serve as equilibrating organs in
Amphioxides on account of their ventral position and
asymmetry; he thinks they are gill-covers and can become
turgid by fluid-pressure, thus approximating together and
closing under the gill-clefts, synchronously with the respi-
ratory movements. This view, however, is not confirmed
by observations on the living larve of Branchiostoma.
Another suggestion made by Dr. Goldschmidt may conve-
niently be mentioned in this place. He says (p. 32) that in
life the lateral folds of the pharynx (plice limitantes)
may be capable of being pressed together and so prevent
food from falling into the pars respiratoria. I think this
is highly improbable, and our author seems to overlook the
circumstance that the ingestion of food and the respiratory
current in Acraniata are alike effected by ciliary action.

From the arrangement of the branchial musculature which
is described in detail, Goldschmidt deduces a mechanism of
breathing by expansion and reduction of the body-cavity,
analogous to lung-breathing. As I have stated above, how-
ever, the respiration of Branchiostoma, like that of Asci-
dians, is primarily promoted by ciliary currents, not by
muscular contraction. The branchial muscles serve only for
the protection and regulation of the branchial apparatus.

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 589

Their contraction occurs under stimulus, but, far from in-
ducing respiratory currents, it temporarily inhibits them. In
point of fact rhythmic muscular respiratory movements have
not been observed in Acraniata.

The gill-arches of Amphioxides (fig. 3) possess the essen-
tial qualifications of a true vertebrate gill-arch, namely, the
endodermal pharyngeal epithelium; the ectodermal portion
of the body-wall; the branchial ccelom; the branchial mus-
cles, which are true visceral muscles derived from the wall of
the ceelom. ‘They appear, in the preserved material, to pro-

Fic. 3.—Diagram representing the structure of the branchial
apparatus of Amphioxides. The ventral half of the body with
the pars respiratoria has been exposed by a frontal incision. From
Goldschmidt. 4&sp. Gill-shit. 4725. Gill-arch. mit. Median furrow
of gill-arch. 4z.coe. Colom of gill-arch. Coe/. General celom.
b.w. Body-wall.

ject into the pharynx like hollow sacs between the gill-
slits, each arch being apparently divided incompletely into
right and left compartments by a median groove, and the
suggestion is made that this bilateral disposition may be a
stage towards the duplication of the slits. The respiratory
epithelium is described as a many-layered ciliated epithe-
lium ; a similar appearance may be noted in other species,
but it has been shown, in the first place by Langerhans, that

590 A. WILLEY.

the branchial epithelium consists of very high filiform
columnar cells in which the nuclei occur at different levels ;
it is probable that the appearance of stratification is still
further increased by the compression of the epithelium as a
consequence of contraction.

What may be described as a sensational conclusion is that
which proceeds from the author’s comparison of Amphi-
oxides with Branchiostoma, i.e. that the so-called
secondary gill-slits which suffer a retardation of develop-
ment in the larva of Branchiostoma are heterogeneous
formations, not homologous with the primary series. Of
course this conclusion is not rendered in an arbitrary manner,
but is led up to by a number of arguments based upon the
grand assumption that Amphioxides alone is primitive,
and that the larval development of Branchiostoma points
no farther back, but is a mere recapitulation of the charac-
ters of the original pelagic Acraniate as represented by the
former genus.

Dr. Goldschmidt adopts the method of assuming that his
form is primitive, and then explaining the facts on that
assumption ; and he claims to explain all the facts, whereas
other theories only explain some of them. But Amphi-
oxides may be as highly adapted to a pelagic life as
Branchiostoma to a benthonic life. Iam aware that this
is an easy objection, but it is none the less true. In most or
many other sharply defined orders, those forms in which an
entire organic system is functionally deranged or obsolete,
are not usually the most primitive. A simple example is the
eyeless condition of many cave animals, deep-sea animals,
and othe cryptozoic forms ; another is the limbless condition
of some T'eleostean Fishes, Batrachia, and Reptiles. They
may be primitive in other respects, but not in respect of
their lack of parts. Perhaps a closer analogy is afforded by
the pelagic Tunicata of the class Copelata which, like the
Acraniata, comprises two families, Kowalevskide and Appen-
dicularide, the former characterised by the absence of an
endostyle, but it is not suggested that the Kowalevskidx

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 591

are the more primitive on that account.' In the same way it
seems to me to be improbable that Amphioxides is primi-
tive in respect of the absence of an atrial chamber, of anti-
meres to the gill-slits, and of the hepatic czecum.

Another unexpected conclusion to which Dr. Goldschmidt
has been led by his researches is the virtual denial of the
homology of the hepatic caecum of the other Acraniates with
the liver of Craniota, but he does not refer to recent work on
this organ.”

Turning now to the cavities in the rostrum of Amphi-
oxides we find a remarkably clear account of their relations
to each other and of their connections with the cavities
(myoccel) of the first pair of myotomes. ‘The special character
of these myotomes, as compared with those which follow,
has been pointed out by Hatschek (1881), and more recently
by MacBride (1897), who compared them with the collar
cavities of Huteropneusta.

The diagram (fig. 4) and the transverse section (fig. 5)
illustrate the arrangement, which is sufficiently elucidated in
the explanation of the figures. Perhaps the most important
fact to note is that the ventral rostral cavity is that which
represents the right head-cavity of the embryo; it is aptly
described as the ontogenetic partner of the preoral pit, for
which it provides a splanchnoceel and a visceral musculature.

From the mode of branching of the cephalic nerves, their
relation to the ventral rostral cavity, and the alleged pro-
somital gill-clefts of A. valdiviz, Dr. Goldschmidt deduces
a scheme of the segmentation of the head, for the particulars
of which the reader should consult the original monograph.
I regret that I cannot follow it myself, chiefly because I
cannot believe that there is such a fundamental difference
between closely-allied species as would follow if it were
actually true that A. valdiviz is the possessor of pro-

1 Of. H. Lohmann, ‘Die Appendicularien der Plankton-Expedition,’ Kiel
and Leipzig, 1896.

2 Guido Schneider, “ Hiniges tiber Resorption und Excretion bei Amphi-
oxus lanceolatus, Yarrell,” ‘ Anat. Anz.,’ xvi, 1899, pp. 601— 605.

592 A. WILLEY.

a.drk

om

Al
(=)
inatett

=:
nat

SC

HH
iyi} ~
wh ~

Fic. 4.—A. valdivia. Diagram of anterior end from above to
show the relations of the rostral cavities. After Goldschmidt,
somewhat simplified. 2.2. Lateral rostral cavity proceeding from
the first myocel. 2d.v. Ventral rostral cavity. ork. Ventral
rostral canal proceeding below the notochord from the ventral
canalis communicans (gév) between the first myotomes. sbr&. Sub-
dorsal rostral canal proceeding above the notochord from the dorsal
canalis communicans (gkd). drk. Dorsal rostral canal continued

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 593

somital gill-clefts which are lacking in the other two
species.

In his description of Hatschek’s nephridium Dr. Goldschmidt
records the important discovery of the presence of solenocytes
in this small tube, which occurs sinistrally under the noto-
chord, and opens into the przoral portion of the gut. Ina
specimen of 8 mm. it attains the considerable length of half
a millimetre. It is closely applied to the anterior end of the
aorta, so that only a thin membrane separates the two struc-
tures. The part of the tube in the vicinity of the orifice into

Fic. 5.—A. valdivie. Section through rostrum. After Gold-
schmidt. /k. Lymph-canals which arise from the dorsal rostral
canal (dri). sbrk. Subdorsal rostral canal at its point of origin from
the dorsal canalis communicans between the lateral rostral cavities,
Rh,i, (Anterior ends of cavities of first myotomes.) vrk. Ventral
rostral canal. £/.v. Ventral rostral cavity.

the preoral gut consists of high cubical cells. Dorsally this
epithelium ceases, and gives place to isolated large round

forwards from the fin-chambers (/7.), and communicating by a canal
on the right side only (a.drk) with the cavity of the first myotome.
M.J and MII. First and second myotomes. sep¢. Septum between
the ventral rostral cavity and the splanchnocel. Coe.d. Anterior
end of the splanchnoceel on the right side communicating with the
first myoceel. coe.s.k. Anterior coelomic canal placing the splanch-
noceel in communication with the first myoccel on the left side.
prod. Preoral termination of the gut. y.p. Anterior and posterior
borders of the preoral pit. 5/4. Sclerotome diverticulum of the
second myoceel,

594. A. WILLEY.

cells which lie upon the membrana limitans of the aorta.
These latter cells are identified as solenocytes, since each of
them gives off peripherally a long and delicate tubule, which
passes straight across the lumen of the nephridium to the
opposite wall, where it penetrates between the cells of the
cubical epithelium. The solenocytes occur along the entire
dorsal wall of the organ, and all the tubules converge towards
the orifice, from which numerous fine long flagella depend
into the preoral gut; these are the flagella of the solenocytes.
Goldschmidt therefore defines Hatschek’s nephridium as a
portion of the ccelom constricted from the left head-cavity
(which gives rise to the preoral pit or preoral organ), and
effecting a communication with the preoral gut by a pronephric

Fic. 6.—Diagram of Hatschek’s nephridium, showing solenocytes
and the orifice into the preeoral gut. After Goldschmidt.

canal. It thus appears that Hatschek’s nephridium has a
structure analogous to that of Boveri’s tubules as corrected
by Goodrich.

Dr. Goldschmidt has found no other excretory tubes in
Amphioxides, but describes some structures which he calls
“ Schwammkorper ” occurring segmentally on the left side at
the dorsal recess of the body-cavity in the region of transition
between pharynx and intestine ; they appear as a feltwork or
mass of spongy tissue with nuclei and solenocyte tubules.!

Noteworthy features in the vascular system of Amphi-
oxides are the absence of a portal system, the presence of
an unpaired aorta, and especially the fact that the branchial
artery, which is a direct continuation of the sub-intestinal

* Possibly they have some relation to Lankester’s brown funnels (see also
Burchardt, ‘Jena Zeitschr.,’ 1900).

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 595

vein, is displaced to the right side. In front of the pharynx
the branchial artery bends up more dorsally and ends
blindly, co-extensive with the aorta and the preoral gut.

One word concerning the classification adopted by Dr.
Goldschmidt. He only recognises two genera of Branchio-
stomide, namely, Branchiostoma, Costa, 1834, and
Epigonichthys, Peters, 1876. Opinion may be reserved
regarding the necessity of abolishing certain other generic
terms which have been introduced, but the resuscitation of
Epigonichthys is clearly correct.

A great deal more information is contained in Dr.
Goldschmidt’s monograph than what I have outlined above.
In order to render the presentation of the portion of his
theoretical excursions which I have selected for criticism
more complete, it should be added that he traces the origin
of what he considers to be the primitive condition of the
pharynx as determined by the ventral series of gill-slits, to
the secondary extension of the segmental musculature towards
the ventral side ; this circumstance (and here he is in agree-
ment with Boveri) would also account for the existence and
peculiar method of development of the atrial chamber of
Amphioxus. It may be admitted that there is very likely a
good deal of truth in this correlation when regarded from the
point of view of the mechanical conditions of development,
without prejudicing supposed morphological or phylogenetic
relations one way or the other.

By the courtesy of the Cambridge University Press! I am
able to reproduce a set of diagrams which may serve a useful
purpose as indicating different points of view, and thus
helping to clear the issues.

In a recent article Professor van Wijhe? states that I have

1 * Zoological Results’ (A. Willey), part vi, 1902, p. 728. The matter is
introduced into that portion of my “Contribution to the Natural History of
the Pearly Nautilus,’ which is devoted to ‘‘ Personal Narrative.” I take this
opportunity of noting an unfortunate misprint on page 800 of that publication,
where the word “ Branchial” should have been “ Brachial.”

2 J. W. van Wijhe, ‘‘ Die Homologisirung des Mundes des Amphioxus und
die primitive Leibesgliederung der Wirbelthiere,”’ ‘Petrus Camper,’ April,
1906, p. 17 of reprint.

596

A. WILLEY.

confused topographical with morphological conceptions in
regard to the organ of fixation in the larva of Ciona intes-

ih,

Fic. 7.— Diagrams of an Enteropneust (A), an Ascidian larva (B),
and a Craniate embryo (C). After Willey (‘Zoological Results,’
part vi, 1902), by permission of the Cambridge University Press.

A.—p. Proboscis. p.p.c. Proboscis pore-canal opening externally
close to the anterior neuropore. ¢.v. Collar nerve-tube. ep. Epi-
physial roots. s. Stomochord. m. Mouth. ce. Collar region. g.
Gut. ¢. Trunk region. d.z.¢, Dorsal nerve tract. a. Anus. py,
Pygochord.

B.—fo. Organ of fixation. ¢. Endostyle. m. Mouth. a.n.e.
Anterior neurenteric canal. s.2.g. Subneural gland. ¢.v. Cerebral
vesicle. sp. Medullary tube. p.z.c. Posterior neurenteric canal.
n. Notochord.

C.—ep. Epiphysis cerebri or pineal organ. .0., m.b., and 4.0.
Fore-, mid-, and hind-brain. Ay. Hypophysis cerebri. ¢/. Thyroid
gland. Other letters as above.

a 7U.¢.

GOLDSCHMIDT’S MONOGRAPH OF AMPHIOXIDES. 597

tinalis which I have likened to a prxoral lobe. I may be
allowed to remark that whether or not there has been any
objective confusion, there has at least been no unconscious
mental confusion on my part on this particular point.

The preeoral lobe or proboscis shouid, I suggest, be regarded
as an axial organ, forming part of the normal body-length,
neither dorsal nor ventral. The functional situs oris is
determined by special factors (such as its relation to the
anterior neurenteric canal in the Tunicate larva) and should
be considered on a basis of its own. The mouth may be
dorsal, ventral or lateral in actual position. That the preeoral
lobe is essentially axial is indicated by the manner and order
of its development in the embryos of Acraniates and Entero-
pneusta, and also in the regeneration of the proboscis of the
latter (see Dawydoff, “ Ueber die Regeneration der Hichel
bei den Enteropneusten,” ‘ Zool. Anz.,’ xxv, 1902, pp. 551-6).

In conclusion, as to the relation of the Acrania to the
Ascidians, Dr. Goldschmidt is of the opinion that the deve-
lopmental tendency leads from Amphioxides to Amphioxus,
and beyond this in a straight line to the Ascidians, whose
organisation appears to him to have arisen by degeneration
from the Acrania. In opposition to this theory I submit
that the Ascidians have degenerated from an extinct ccelo-
mate perennichordate type, but not from a cephalochordate
type.

When, however, Dr. Goldschmidt asserts that the capacity
which resides in the pharynx of Acrania, as illustrated in the
particular instance of Amphioxides, of forming a gill-sht
between the segments over a great region of the body,
indicates the original existence of very numerous primitive
gill-slits, and supports the theory of the primary poly-
tremism of Vertebrates, | am glad to say that I agree with
him heartily.?

September, 1906.

1 Cf, A. Willey, ‘‘ Enteropneusta from the South Pacific,” ‘ Zoological
Results’ (Cambridge University Press), part iii, 1899; and R. C. Puimett,
“The Enteropneusta,” * Fauna and Geography of the Maldive and Laccadive
Archipelagoes,’ vol. ii, part 2, 1903, see p. 669.

VoL. 50, PART 4,—NEW SERIES. 43

HOLZOLTIS SA TO ASO Se aa

he : vou gd Delta | a | L fo

Le
1
=
<o
a =a
. ~ ered ad oe!

vi
f Fi ~ 4h | on)
‘ t
; ] nl
,
Aid oe Oe
= :
; —? res :
; ‘ sce
c is , TA 1 x é Ppt 7?
oA tf 3 i tucll tee
} 7 i A ;
} * ap’
: J ia
’ i» (eer
'
AS hy L) gs whe
7 l aru J 4 4
= -
wigMm
PAE feos
} "| stint } 4
‘ . ; :
Lan x the eee aa edfacev ii
ayitL! ri Pe if bt) Gee bray 2 a
rete A9 al it ah E .—
j 9 ‘+
ras : 7 . ' «cy

=
‘

HERMIT CRAB AND PELTOGASTER. 599

The Modification of the Sexual Characters of the
Hermit Crab caused by the Parasite Pelto-
gaster (castration parasitaire of Giard).!

By

F. A. Potts, B.A.,
Trinity Hall, Cambridge.

With Plates 33 and 34.

I. InTRODUCTION.

TE discovery of the fact that an alteration in the sexual
characters, both primary and secondary, of an organism may
be brought about by parasitic infection, and the demonstration
of the widespread occurrence of this phenomenon in the animal
kingdom, we owe to Professor Alfred Giard. Admirable
summaries of the work done in this department up till the
date of writing, are to be found in two papers of Giard’s cited
below (1 and 8), Most of the cases are drawn from the ob-
servations of Giard himself among the Crustacea, which group
affords particularly favourable opportunities for study, on

1 Mr. Sedgwick has pointed out to me that the term “castration para-
sitaire,” or at least its literal rendering in English, is somewhat misleading,
and should not be applied to the phenomena which form the subject of this
paper. Castration means destruction of the gonad, and though the phenome-
non here described may eventually culminate in this, yet the changes principally
here discussed occur before the complete destruction of the gonad, which is
consequent on the parasitism, and, so far from being the cause of the other
modifications, is probably merely an effect of the same order as the rest.

600 F. A. POTTS.

account of the well-marked secondary sexual differences and
the frequent ecdyses, allowing the modifying influence of the
parasite on the secondary sexual characters to be clearly
demonstrated. But that in cases of this phenomenon the
parasite and the host may belong to any phylum is shown by
such varied instances as the brittle-star Amphiura parasitised
by Orthonectids and a copepod crustacean, the Hymenopteran
insect Andrena infected by the parasitic strepsipteran Stylops,
the mollusc Lymnza by Trematodes, and the crab Inachus by
the gregarine Aggregata.!

In his conclusions Giard recognises a tendency of the
secondary sexual characters of either sex to become modified
toward the type of the other under the influence of parasitism,
while at the same time those originally possessed may tend to
be effaced. Thus inthe Oxyrhynchid crab Stenorhynchus
phalangium, attacked by Sacculina fraissei, the male
almost entirely loses its copulatory styles and assumes, in
greater or less degree, the abdominal ovigerous appendages
proper to the female; in the female these last-mentioned
characters become less accentuated, and the abdomen gradu-
ally approximates to the appendageless condition of the
male, Giard regarding this as a true modification toward the
male type. In other cases, as, for instance, Palamon infested
by Bopyrid Isopods, there is a different type of effect, the
parasite merely preventing the full development of the
secondary sexual characters of the male, causing them to re-
main in the less advanced female condition. In the primary
sexual characters Giard observes a reduction in the size of
the gonads and suppression of sexual activity, and the altered
condition of the gonads he regards as leading in turn to the
above-mentioned modification of the secondary sexual char-
acters. his sterility of an infected animal is considered to
be merely temporary, though satisfactory cases of the re-
generation of the gonad and the resumption of the sexual
functions, after the death of the parasite, are not quoted.

1 This case does not occur in Giard’s original list, but is the subject of a
recent paper by Smith. See Bibliography (7).

HERMIT CRAB AND PELTOGASTER. 601

Lately, also, the study of the effects of the parasitism of
various members of the Rhizocephala on crustacean hosts has
been continued by Mr. G. W. Smith, of New College, Oxford,
and I am permitted to make a preliminary reference to his
results, which will shortly appear in a work on the Rhizocephala
in the Naples series of monographs. Two most important
points are clearly indicated which place our knowledge of this
phenomenon on an entirely fresh footing. The capability of
assuming the secondary sexual characters of the other sex is
shown to be in the cases studied really confined to the male,
and, as proved by experiment and observation, modified
males (and, apparently, males only) may, on recovery from
parasitism by Sacculina, regenerate a completely hermaphro-
dite gonad. ‘lhis examination of parasitic effects in the
hermit crab was undertaken at the suggestion of Mr. Smith,
to whom I am greatly indebted for much kind help, with the
object of ascertaining whether evidence supporting and ex-
tending these remarkable results could be derived from the
study of other cases.

Before commencing the description of the effect which the
parasite Peltogaster produces upon its host it will be well to
mention a few facts with regard to its life history. It belongs
to the Rhizocephala, and after completing its free-swimming
existence as a Cypris larva passes through a completely in-
ternal stage in the body cavity of the hermit crab, like that
described by Delage and confirmed by Smith in Sacculina.
The Peltogaster internus, which is constituted of a rami-
fying root system and a small central tumour, which later
forms the external body of the adult, was first discovered by
Pekarsky (6), and has also been found on some few occasions
both by Smith and myself. With regard to the duration of
existence of the internus stage we are still ignorant, owing to
the absence of success attending attempts to raise Peltogaster
larvee and artificially infect hosts, and also to the comparative
rarity in nature of crabs containing it. When, however, the
Peltogaster is ready for emergence to the exterior it need not
wait for the assisting agency of a moult, but may, as we have

602 FL A. | POTTS.

actually witnessed, dissolve its way through the soft skin of
the abdomen. In the external stage the Peltogaster is a suffi-
ciently striking object with its sac-like body of a clear red
colour and the great mass of dark green roots shining through
the semi-transparent skin of the host. Practically always it
is found in the same position of fixation between the second
and third abdominal segments on the left-hand side of the
body, occupying in a noteworthy way the space where, in the
normal female, the embryos are borne. Once become external
the adult size is quickly attained, and this latter, then, is no
indication of the age of the parasite, but only of the food-
supplying capabilities of the host. ‘The same crab frequently
supports two Peltogasters, and less often examples of the
Cryptoniscid Liriope pygmea, are found fixed on or near
the Peltogaster itself. But neither the size of the Peltogaster
nor the presence of one, or a plurality of parasites appear to
be correlated with the degree of modification, due to the para-
site, which I go on to describe. It may, however, be men-
tioned here that the character of the root system of the para-
site seems to change with the duration of infection, being more
diffuse and extensive in the earlier stages, and more con-
densed, compact, and dark coloured as time goes on.

II. Generat Errect oF THE PARASITISM OF PELTOGASTER
on EHuPAGURUS.

Parasitic castration has been studied by Giard amongst his
numerous other cases, also in the genus Hupagurus (2). At
Wimereux the occasioning parasites are the Bopyrid Phryxus
and the Cirripede Peltogaster (P. paguri), and these are
found on Hupagurus bernhardi. At Naples one of the
most frequently occurring species of hermit crab is HK. meti-
culosus,! very common at a depth of 10 metres, and 20—30

1 A small percentage of the crabs examined, infected and uninfected,
appear referable to the species HE. angulatus, which is distinguished from
E. meticulosus by certain characters of the chela. Yet intermediate
specimens are found, and the effect of the parasite is the same as in E.

HERMIT CRAB AND PELTOGASTER. 603

per cent. are infested with a species of Peltogaster (P. curva-
tus, Kossmann). The same parasite is also found on HE,
prideauxi, associated with EH. meticulosus on the same
ground and in almost equal numbers, but infected specimens
are very rare, and in nearly a thousand crabs of this species
examined I could only find three with Peltogaster.

Some hundreds of infected crabs of the former species were
examined with respect to the internal and external sexual
characters. It is quite easy to keep the animal in captivity,
and to observe that this differs from other cases of parasitic
influence in Crustacea in that the parasite does not completely
check the general growth of the host after the moult, at which
it becomes external, and that under aquarium conditions
moults are fairly frequent. No exact comparison of the fre-
quency of moulting in infected and uninfected crabs are made,
but experience seemed to show that normal crabs did not
moult so frequently in captivity as infected. That castration
increases the rate of growth of animals is by no means a new
idea, and is put into practice in the rearing of stock and
poultry, and since H. meticulosus, at Naples, reproduces
the whole year round, it would appear most likely that a
relief from the reproductive functions would increase the
somatic growth.

It is also interesting with regard to the general effect of
the parasitism to notice that infected crabs seem fully as
active and healthy as normal specimens. ‘Though the internal
organs (for instance, the heart in its abdominal extension)
may be entirely enveloped by the enormous root system of
the Peltogaster, yet, with the exception of the gonads, they
do not appear to be affected by the parasitism. But even in
the period of the internus stage one finds the gonad reduced
in size, and the brown mature ova of the female giving place
to white ova with little yolk. In the male, however, sperm
production still occurs at this stage, and even for atime after
the commencement of the externus stage. The reduction in

meticulosus. Probably then KH. angulatus is only a variety of the last
named species.

604. F. A. POTTS.

size of the gonad is first effected at the expense of the
glandular portion—the testis of an infected specimen con-
sists typically of a coiled duct with reproductive cells only
toward the termination,—and it is noticeable that the secretory
function of the duct is not checked, but even accelerated up
till an advanced stage of infection. The duct is at first filled
with irregular spermatophores, which may contain a reduced
number of spermatozoa, but at later stages, where the testis
is still clearly recognisable, the ducts present a swollen
appearance, and are filled with irregular and unequal masses
of chitin derived from the cells producing the dense cap-
part of the spermatophore (pl. 34, fig. 5, ch.). It was often
observable that the duct on the same side as the Peltogaster
presents this phenomenon in a greater degree. Perhaps the
excessive production of chitin in the male ducts and the
increased capacity for moulting may be closely related.

With regard to the way in which the reduction of the gonad
is effected it should be said that though the Peltogaster roots
often completely enclose the gonads, in no stage of infection
was it shown by dissection or serial sections that they pene-
trated the bounding wall. That the parasite derives its
nourishment from the general body fluid rather than the
gonads, and hence that the effect on the latter is indirect and
effected through the general metabolism, seems conclusively
shown by the fact that though the parasite in the internal
stage proceeds to envelop first the gonad of the side where
it is destined to become external—and so must exercise a
more extended action on that gonad,—yet in the occasional
cases observed in which the aperture of one side had dis-
appeared, indicating entire and long-standing abortion of the
corresponding gland and duct, this had taken place on the
side opposite to the Peltogaster. This may be contrasted
with the case of Andrena parasitised by Stylops, worked out
by Perez and quoted by Giard, where the testis of the side on
which the parasite is situated, becomes non-functional, but
that of the other side continues to produce sperm.

HERMIT CRAB AND PELTOGASTER. 605

We will now go on to consider more specially the effect on
the sexual characters.

III. Errect on THE SExvAL CHARACTERS.

A. The Secondary Sexual Characters.

The external characters by which the sexes may be distin-
guished in EH. meticulosus are the following :

(a) The Genital Apertures.—In the female these are
situated on the internal corner of the basal joint of the third
pair of walking legs, and are noticeable oval apertures.

In the male, on the basal joint of the fifth thoracic leg, there
is a papilla directed somewhat posteriorly bearing the male
aperture on its apex, bordered with setaceous hairs.

(b) The Abdominal Appendages.—In both male and
female there are appendages present on segments 2, 3, 4, 5,
6, a character which distinguishes this species from HE. bern-
hardus, longicarpus, and others in which the appendage
of segment 2 is absent in the male, and from E. prideauxi,
in which the appendages of segments 2,3, 4, 5 are all absent in
the male. Those of the last segment are paired and form with
the uropod the organ by which the crab maintains its hold
on the columella of the shell it inhabits. The fifth segment
is also furnished similarly in both sexes, bearing a single
appendage consisting of a basal joint and two rami, of which
the external member is long, and flattened and furnished with
an edging of stiff hairs; the internal is pointed and rudi-
mentary. ‘The remaining three segments also bear a single
appendage, but while in the female these are of markedly bi-
ramous character (pl. 33, fig. 6), in the male they are practic-
ally uniramous (pl. 33, fig. 1), and similar to that of the 5th
segment.

The relative development of the rami varies slightly in the
female. In the second segment the external ramus is but
slightly the larger, while in the third and fourth respectively
the development of the internal ramus is progressively less,

606 F. A. POTTS.

Also in young (though even then sexually mature) hermit
crabs of 1:0—1°5 cm. of carapace length, the internal ramus
is less developed than in the adult, and everything points to
the conclusion that the biramous character of the abdominal
appendage has been acquired to provide a more secure anchor-
age for the developing embryos. For this purpose the basal
joint and the internal ramus, both in the middle and apical
part, are provided with long, radiating bunches of hairs.
Of the bunch on the basal joint only is there any representa-
tive in the male. But though the external ramus is well
provided with hairs it is noticeable that eggs are never
attached to them, so that this ramus is left quite free to sub-
serve its proper function of aérating the interior of the shell,
for which it is admirably adapted by its oar-like shape and
border of stiff hairs.

In the male the external ramus is developed as in the
female and is used in the same manner; but in correlation
with its non-participation in the nursing functions, the internal
ramus is rudimentary in the first four appendages.

Account of Modification in the Male.

The effect of the parasitism of Peltogaster on the secondary
sexual characters of the male H. meticulosus is to stimulate
the development of the rudimentary internal ramus of the
first three abdominal appendages. ‘The extreme members of
the modified series reach a stage indistinguishable from the
corresponding normal female appendage. The modification is
rendered complete by the assumption of the typical gibbous
form and the bunches of egg-bearing hairs of the internal
ramus of the female.

Various stages of the complete series which exists between
totally modified and unmodified crabs are here figured (pl. 35,
figs. 1—5). An examination of these shows that, as in the
development of the female, so in the modification of the male,
it is in the first appendage that the internal ramus earliest
responds to the stimulus (fig. 3). And though in some cases

HERMIT CRAB AND PELTOGASTER,. 607

the internal ramus attains the same degree of alteration in
the first and second appendages, in the third it remains for
the most part in a condition of suppressed development (figs.
2,3, 4), though in the most marked examples of modification
this appendage exhibits the effect in the same degree as the
two first (fig. 5). The retardation of modification proceeding
backwards in the series is apparently in correspondence with
what takes place in the normal development of the female,
which reaches its extreme in the fourth abdominal appendage,
where the internal ramus remains throughout life in a rudi-
mentary state, and where also no alteration occurs in connec-
tion with parasite castration. These facts are apparently to be
explained by the supposition that the abdominal appendages
have been successively pressed into the service of bearing the
embryos, and the modification entailed by this is more complete
in those longest so employed.

On account of the completeness of the modified series it is
impossible to do more than roughly classify the infected
males with respect to the degree of modification. It may be
said, however, that at least a quarter of those examined
were quite unmodified, that in considerably more than half
the modification range from a condition of bare percepti-
bility to one in which the internal ramus, though distinctly
small, showed the typical female form, and the remainder
alone could be spoken of as completely modified.

Amongst the infected crabs there is a considerable propor-
tion which it is impossible to assign to their sex by merely
external examination, for they are entirely without genital
apertures, which have to be taken as the external criterion
of sex. In these then the modification has been of so long
standing duration that the gonad or at least the duct has
become completely extirpated, and succeeding moults have
closed the genital apertures. As referred to above, a mid-
way stage to this condition is to be noticed in certain crabs
in which one only of the apertures is closed, this in almost
all the cases of both sexes examined being the one opposite
the Peltogaster.

608 BCA, POTTS.

By dissection of these crabs of problematical sex it was in
vare cases possible to discover a diminished gonad, and on
such occasions this was more often an ovary than a testis,
though instances of both were found. The appendages of
these animals vary from slightly modified to completely
female type, but the former part of the series is represented
by only a small number of individuals. Probably then the
slightly modified individuals and a small percentage of the
fully modified belong to the male sex, while the rest which
form a majority are really females.

In Carcinus menas infected by Sacculina, as described
by Giard, the old males are entirely unmodified, and the
varying amounts of modification in different individuals are
hence connected with the age at which they became infected.
Though the solution of the question is to a certain extent
interfered with in Eupagurus meticulosus by the con-
tinued growth after the external appearance of the parasite,
yet the observation that young males of 1°0—1:3 cms. cara-
pace length were perhaps more often than not unmodified,
while the largest infected males found of 1°6—2:0 cms. cara-
pace length were equally likely to be modified or unmodified
shows that in this case the male sexual characters are not
more susceptible to alteration of type at an earlier age. It
has also been pointed out above that the degree of modifica-
tion is independent of the presence of a larger or smaller
parasite, of one or a greater number.

Apparent modification in the female.

Although the majority of the infected females resemble
the young rather than the adult normal female with regard
to the development of the internal ramus, yet cases in which
any real approach to the uniramous condition is found are
extremely rare, and I have been only able to procure three
specimens which were, in Giard’s sense, clearly modified
toward the male type. All three were of small size; in no
case were any of the first three appendages of pure male

HERMIT CRAB AND PELTOGASTER. 609

type, but in the second and third appendages the develop-
ment of the internal ramus was quite slight, while in the first
it had attained about a half of its usual size.

The consideration of some facts dealing with the later
metamorphoses of the hermit crabs may help us to decide
whether these cases fall under the heading of true modifica-
tion or retardation of development.

The life-history of Hupagurus meticulosus is not
known in detail, but that of the American species, H. longi-
carpus, has been the subject of careful research by Thomp-
son (8). This species as noted above differs from H. meti-
culosus in the loss of the first abdominal appendage by the
male.

After escaping from maternal gestation the larva leads a
pelagic life, the last stage of which is known as the Glau-
cothoé. In this the animal has an appearance and anatomy
nearly similar to the adult, but the abdomen is symmetrical
in form, and it also differs in the possession of a pair of
appendages for each segment. These are of male type,
though the rami are but shghtly developed.

In the stage, after the moult from the Glaucothoé, during
which the animal seeks a shell, the abdomen becomes asym-
metrical, the appendages of the right side are totally lost,
and the first appendage of the left becomes rudimentary,
while throughout the remaining appendages the external
ramus increases in size. At this time there is no differentia-
tion of sex, both male and female possessing appendages
with a rudimentary internal ramus. And it is not till thirty
days after the glaucothoé stage that the rudiment on the
second segment even begins to develop into a typical female
appendage. ‘len days after that the alteration to the female
type begins in those of the third and fourth segments, but
the complete development requires several further moults.
Quoting Giard’s case of the parasitism of Peltogaster on
Kupagurus bernhardus, where the form of the appen-
dages of the female is said to approximate to the male type,
Thompson says, ‘The parasite presumably might attach

610 F. \A.' POTTS:

itself any time within the first fifty days of adolescent life,
and yet be able to check the complete development of the
female type of appendage.” But the parasite does not exert
its retarding stimulus directly after fixation, and we are
without data which would give the period after which it
commences to do so. ‘To support this explanation of
the apparent modification of the female characters, it is
probably necessary to postulate a larval infection. During
the last moults of larval life the Peltogaster would remain in
the internus stage and in the time of the first adolescent
moults, whether or no its stage of development was suffi-
ciently advanced for an external appearance, the checking
effect on the further development of the female appendage
could be brought about.

Assuming then that the course of development runs fairly
similarly in Hupagurus longicarpus and meticulosus,
it is perfectly open to consider the exceptional cases men-
tioned above as due to unusually early, probably larval,
infection and consequent midway arrest of the assumption of
the female characters. That this is the true explanation of
the phenomenon is indicated by the unusually small size of
the infected individuals and by the fact that the develop-
ment of the first appendage was more distinctly female than
the other two. It will be later seen from other evidence that
the parasitism does not cause alteration in the once assumed
female characters.

EFrEcTt ON THE SECONDARY SEXUAL CHARACTERS OF
EUPAGURUS PRIDEAUXI.

Before proceeding to describe the effect of parasitism on the
primary sexual characters of Hupagurus meticulosus it
will be well to mention the conclusions gleaned from the small
amount of infected material of EK. prideauxi I have been able
to collect. ‘This species differs from E, meticulosus in the
complete absence of the abdominal appendages of the second,

HERMIT CRAB AND PELTOGASTER. 611

third, fourth, and fifth segments in the male. In the female
appendages are found on the second, third, and fourth seg-
ments, of the usual biramous type. One of three individuals
with Peltogaster was a female; in it I was unable to detect
any alteration from the female type. Of the two males one
was unaltered; the other was provided with three pairs of
abdominal biramous appendages exactly as in the female.
But these appendages, though of an exact female form, did
not possess the long hairs which serve as egg attachments in
the female, and were only about one third the size of those
of a slightly smaller normal female with which it was com-
pared. It may be mentioned that the modified male was
much smaller than the unmodified, but the examination of a
more considerable number of cases is needed before the
connection of modification with infection at an early stage is
asserted.

EVIDENCE CONCERNING THE PERMANENCE OF THE MODIFICATION
CAUSED BY PARASI'TISM.

(1) From moulting.

(a) A large number of infected crabs, particularly un-
modified males, were kept in aquaria and fed generously.
In periods varying from a fortnight to two months the crabs
moulted, and in no cases, either of infected females or males
of any degree of modification, was there any alteration of the
relative proportions of the two rami in the new appendage.

(b) Primarily, for another purpose, to be treated of below,
operations were made on a large number of crabs, the
external part of the Peltogaster being removed by scissors.
They were then kept as far as possible in natural surround-
ings and well fed, and after a month moults became of
frequent occurrence. Oncomparing the old and new append-
ages no difference in the relative proportion of the rami,
concurrent with the removal of the occasion of modification,
could be distinguished. It is interesting to compare the

612 EF. LA; POTTS:

stability of the effect produced here with the case recorded
by Giard of Phryxus paguri and Hupagurus bernhardi.
The parasite is fixed on the abdomen of the hermit crab,
and it is comparatively easy to remove it from its host
without injury to the latter. Guard did this in the case of
a modified male crab, and when, after being kept for a
month, it moulted, he thought he could observe an incipient
return to the unmodified condition of the appendages, the
gonads being still of small size. It may, perhaps, be affirmed
that the two cases are fundamentally different, and that
Phryxus is purely an ectoparasite, its effect naturally passing
away after removal, while it is impossible to entirely nullify
the effect of the Peltogaster by operation, the roots con-
tinuing to live and play the same part as the complete
parasite, but the same results have been obtained with
Eupagurus meticulosus even in cases where the partial
recovery of the gonads showed that the parasitic stimulus
had finally ceased to act.

(2) From regeneration.

As in the American species on which T. H. Morgan experi-
mented, the abdominal appendages of Hupagurus meticu-
losus are easily capable of regeneration. Various abdominal
appendages were cut off from crabs, in a large number of
which the Peltogaster was at the saine time removed. In all
cases at the next moult the appendages were found to be
regenerating. When the appendage was cut off proximal to
the bifurcation the rami regenerated but feebly, but if the
cut was made at a level slightly distal to the bifurcation an
almost complete regeneration took place. Unfortunately the
number of operations of the latter class were but few, and in
those of the former it was difficult to see whether the degree
of modification in the old and new appendages exactly corre-
sponds. But the point which comes out clearly as a result
of these experiments is that females and modified males
(pl. 34, fig. 1), from which the Peltogaster has been removed,
regenerate appendages in which both rami are of approxi-
mately equal size, while unmodified and slightly modified

HERMIT CRAB AND. PELTOGASTER. 613

males (pl. 34, fig. 2) regenerate appendages in which they are
manifestly unequal, and in which at first, indeed, only the
external ramus may be developed.

In those on which the Peltogaster was allowed to remain
the interesting fact was observed that a female with an
abdominal appendage removed regenerates one of purely
female type, and this, together with. the evidence derived
from moulting, seems entirely to prove that the Peltogaster,
at least in the later stages of parasitism, exerts no influence
at all on the secondary sexual characters of the female, or on
the processes which regulate their reformation. So also with
regard to the male it appears that the modifying cause
operates at an early stage in the history of infection, and
that the degree of modification is not subsequently increased.
Iam inclined to attribute the great variability in degree of
modification in infected male crabs to constitutional differ-
ences in the crabs themselves, not to date of infection, for
the same range of variation is found in crabs of all sizes, nor
to the comparative vigour of the Peltogaster, for quite large
Peltogasters may be present on unmodified, and small Pelto-
gasters on modified, hosts.

The permanence of modification in recovering crabs may be
compared with a kindred phenomenon in Mammalia. In
deer the females may assume the horns characteristic of the
male, as a consequence of atrophy or disease of the genitalia.
But cases are known of female deer with horns which, on
dissection, showed normal ovaries. Herbst (4) refers the
growth of the horns to a period when the ovaries were in a
diseased condition, but while these later recovered their
normal state, the modification of the secondary sexual
characters remained.

VoL. 50, PART 4,—-NEW SERIES. 44

614 F, A. POTTS.

EFFECT OF THE PARASITE ON THE CELLULAR CHARACTER OF
THE GONAD.

A large number of testes of infected males have been ex
amined by sections, showing a very curious cytological change
of a widespread nature. This is the appearance of small but
distinct ova-like cells in the glandular part of the testes
(pl. 34, figs. 4, 5, 0v1, ovg). They may be followed in their
development from spermatocytes by the enlargement of the
nucleus, the appearance of a single large and distinct nucle-
olus, and the loosening cf the chromatic network, together
with the aggregation of cytoplasm round the nucleus. This
phenomenon may occur while sperm production, though in
diminished activity, is still proceeding in other parts of the
testes (pl. 34, fig. 5). A very large number of infected males
with external Peltogasters exhibit it, and it seems likely that
all of them, when a certain period of infection was reached,
would be capable of producing ova in their testes. As to the
period of appearance it is impossible to be sure; but in the
only male I found, with an internal stage of Peltogaster,
spermatogenesis was but slightly checked, and there was no
sign of the appearance of female cells. But though in the
majority of infected crabs with small just external Pelto-
gasters the testis was of normal character, in one or two I
found signs of incipient change on sectioning. Probably
there is much individual variation in this particular, though
the variability of the organism is much more clearly demon-
strated in the reaction of the secondary sexual characters to
the parasitic stimulus.

It must not be forgotten that the occurrence of ova is a
phenomenon of normal occurrence in the testes of various
animals, such as, for instance, Orchestia and Homarus among
Crustacea, in Phalangids, and in the toad. A number of
normal testes of Hupagurus were examined, but without
finding that it could be added to the above list.

With the object of ascertaining whether this apparently

HERMIT CRAB AND PELTOGASTER. 615

hermaphrodite state of the gonad of the infected male is
merely transient, or persists on the removal of the parasite
and the recovery of the host, from a large number of infected
crabs of both sexes, the external part of the Peltogaster was
removed, cut off with scissors as close to the crab’s abdomen
as possible. As mentioned above, the operated animals,
which were fed regularly, soon became quite healthy and
moulted very frequently. They were kept for various periods,
varying from two to five months, but up till the time of
writing no crabs had so far recovered as to completely
regenerate the gonad. It is difficult in most cases to prove
that the gonads increase in size at all, but in one of the
crabs of problematical sex (in which the gonad, if present, is
of very small size), when two months had elapsed after re-
moval of its Peltogaster, a fairly large sized ovary with eggs
approaching maturity was found. From this fact and others
it appears certain that a slow regeneration of the gonad does
take place. How slow this regeneration is may be judged
from the evidence of a crab once infected, but which was
found naturally freed from its Peltogaster. Though, in the
few other crabs of the kind which were found, the experience
appeared to have been recently undergone, this one, a slightly
modified male, preserved no trace of the Peltogaster ex-
ternally, indicating that several moults must have occurred,
enabling it to lose the scar of the Peltogaster. On dissection,
the roots, though degenerate, were found, but the gonad of
the host, in spite of the presumably extended period since
the loss of the parasite, was still very small, though showing
some signs of regeneration.

With regard to the question whether the egg cells, which
appear during the period of activity of the parasite, are
resorbed when the influence is removed, or whether, with
recovery of the gonad, they grow and become functional ova,
there is some little evidence to be cited from a histological
study of the testes of operated crabs. In nearly all those,
sectioned ova were still present, and in one case at least
seemed to have attained a greater size than had been ob-

616 F. A. POTTS.

served in infected testes (pl. 34, fig. 6, ov,). Such evidence,
as far as it goes, seems to show that, as in the secondary
sexual characters, so in the primary, the effect induced by
the parasite is a permanent one, at least for a considerable
time, and does not pass away immediately with removal of
the primary cause.

Eupagurus prideauxi.—In one (the unmodified) in-
fected specimen of this species testes of medium size were
found on dissection. The ducts exhibited the same pheno-
menon of excessive chitin secretion; but while spermato-
genesis had entirely ceased, there was no appearance of egg
cells in the glandular part. The study of more material
might well, however, establish the existence of the pheno-
menon described above also in Hupagurus prideauxi.

In concluding this account of the hermaphroditism of
EKupagurus, due to the action of the parasite Peltogaster, a
case of an apparently naturally occurring hermaphrodite
must be mentioned. This animal possessed typical female
appendages, but perfectly formed male apertures. On dis-
section, the gonad of the right side appeared completely and
normally male, though of somewhat reduced size; but on the
left side, though testis and duct were present, there were
large masses of mature brown ova. It is just possible that
this specimen was one which had recovered from parasitism
of Peltogaster, for we have seen above that the attachment
scar of the parasite may disappear with lapse of time, and if
complete absorption of the roots also supervenes no further
evidence of former infection remains. But this isolated case
cannot be cited as serious evidence of the capability of an
infected male to regenerate a complete hermaphrodite gonad.
I may, however, indicate the fact that the animal was pre-
dominantly male, possessing male ducts and apertures,
and gonads principally male; and since we have seen that
the effect of parasitism is to occasion the appearance of
ova in the testis, and alter the male secondary sexual
characters toward the female type, so we may imagine that
some other unknown cause, acting on a primarily male

HERMIT CRAB AND PELTOGASTER. 617

organism at some unknown possibly embryonic period of
development, may produce a hermaphrodite of the character
described above.

It will be interesting here to briefly compare the pheno-
menon of hermaphroditism in the gonad of Inachus scorpio
caused by Sacculina neglecta, as described by Smith.
Here the testis of infected males, before completely atrophy-
ing, Showed no appearance of ova. In those, however, from
which the Sacculina was removed, and which regenerated
their testes, the externally modified and slightly modified
males show a completely male gonad; the perfectly modified,
on the other hand, possess gonads with mature products of
both kinds. In the males of Inachus scorpio, then, the
capability of assuming the internally hermaphroditic state
would appear to be more restricted than in the crustacean
here considered. I am unable to say, however, whether the
male infected hermit. crabs are able to regenerate a com-
pletely hermaphrodite gland in place of the testis on removal
of the parasite, as happens in Inachus.

Many examples of infected and recovering ovaries were
likewise examined for evidence of hermaphroditism. But
though the cellular elements are so reduced in size from lack
of nourishment that many appear exactly similar to sperma-
tocytes, none of the stages of spermatogenesis were ever
seen, and in recovering ovaries the cells throughout increased
in size and recovered their ova-like appearance.

CORRELATION OF THE EFFECTS ON THE PRIMARY AND
SECONDARY SEXUAL CHARACTERS.

It was expected at first that some correspondence between
the degree of modification exhibited in the primary and
secondary sexual characters respectively might be traced,
but this does not appear to exist, and the development of
ege cells appears to take place in the same degree in
entirely unmodified animals as in those in which some

618 F. A. POTTS.

degree of modification has been developed. It has been
noticed that the effect on the primary characters as one
which very often supervenes, and may in fact be general for
the whole male sex; while on the other hand individual
variation plays a great part in the effect produced on the
male secondary sexual characters, which possibly in about
a quarter may be entirely null. The correlation which
exists between the primary and secondary characters must
be of a loose nature, and this fact and more particularly the
observation of cases in which modification of the append-
ages had begun before the corresponding changes in the
gonad, appears to conclusively show that the modifying
cause, acting on the secondary sexual characters, is not to
be sought specifically in the altered character of a secretion
of the gonad, but rather in a more general rearrangement of
the metabolism, occasioned by the parasite, which affects
both primary and secondary characters.

SUMMARY.

(1) The infection of the hermit crab Eupagurus meti-
culosus by the cirripede Peltogaster curvatus has the
effect of diminishing immediately the size of the gonads and
suppressing their functions. This is probably effected
through interference with the general nutrition of the host,
and not through direct action on the gonads.

(2) At an early stage of the external parasitism ova make
their appearance in the glandular part of the testis. Hxperi-
ments planned to discover the fate of these ova on recovery
of the host from the parasite were not very successful owing
to the slow regeneration of the gonad, but it seems probable
that they persist and grow. No corresponding changes
could be traced in the ovary.

(3) The male secondary sexual characters are stimulated
to development towards the female type under influence of
the parasitism. There is a complete series between un-

HERMIT CRAB AND PELTOGASTER. 619

modified crabs and crabs which have almost entirely assumed
the female characters.

(4) The apparent cases of modification in the female
toward the male type are probably due to unusually early
infection, a retarding influence being exerted by the parasite
on the incompletely developed secondary sexual characters
causing them to remain in the less advanced male stage.

(5) In Eupagurus prideauxi a similar development of
the female secondary sexual characters in the male was
observed.

(6) Observations on the moulting and regeneration of the
abdominal appendages show that the final degree of modifi-
cation is attained early after the commencement of the
external stage of Peltogaster, and is never afterwards altered
either by continued action of the parasite or by its extirpa-
tion. The great range of degree in modification is the
expression of individual variability in the hermit crabs.

(7) The looseness of the correlation between the changes
in the primary and secondary sexual characters shows that
the latter are not directly consequent upon the former, but
rather that both are attributable to some change in the
general metabolism.

Norr.—In “ Expéd. Sci. Travailleur et du Talisman, Crus-
tacea Decapoda, pt. 1, Brachyura et Anomura, 1900, p. 228,”
A. Milne-Edwards and KH. L. Bouvier assign the crabs here
called Kupagurus angulatus to Herbst’s earlier species
K. excavatus, and that known as E. meticulosus to EH.
excavatus var. meticulosus.

September, 1906. He Ase:

BIBLIOGRAPHY.

1. Grarp, A.—‘ La castration parasitaire,’ ‘Bull. Soc. pee: du Nord,’
ser. 2, 10 Année, 1887.

620 ¥,. A. POTTS.

2. Grarp, A.—‘‘Sur Ja castration parasitaire chez l’Eupagurus Bern-
hardus et chez la Gebia stellata,” ‘C. R. Acad. Paris,’ t. civ,
p. 113.

8. GiarD, A.— La castration parasitaire (nouvelles recherches),” ‘ Bull. Soc.
Dep. du Nord,’ (3), 1, 1888, p. 12.

4. Hergst, C.—‘ Formative Reize,’ 1901.

5. Prrez.— Des Effets du parasitisme des Stylops sur les Apiaires du Genre
Andrena”’ (Act. Linn. Soc. Bordeaux, t. 40).

6. ScurmKewitscu, W.—“ Ueber die Untersuchungen von J. O. Pekarsky ;
ueber die Entwicklung von Peltogaster paguri,” ‘Trav. Soc.
Natural Petersbourg,’ vol. xxvill, p. 218, 1898.

7. Situ, Grorrrey.— Note on a Gregarine which may cause the Parasitic
Castration of its Host,” ‘ Mitt. Zool. Stat. Neapel,’ 17 Band, 3 Heft,
1905.

8. Tuomrsoy.—“ Metamorphosis of Hermit Crab,” ‘Mem. Boston Soc. Nat.
Hist.,’ 1903-4.

EXPLANATION OF PLATES 388 anp 34,

Illustrating Mr. F. A. Potts paper on ‘“‘The Modification of
the Sexual Characters of the Hermit Crab caused by the
Parasite Peltogaster (castration parasitaire of Giard).”

In all cases the first three abdominal appendages only are figured ; they
are denoted respectively a, b, and ec.

PLATE 33.

Fic. 1.—Normal male: abdominal appendages.

Fic. 2.—Infected male: abdominal appendages, Stage 1.

Fie. 3.—Infected male: abdominal appendages, Stage 2.

Fic. 4.—Infected male: abdominal appendages, Stage 3.

Fie. 5.—Infected male: abdominal appendages, Stage 4 (appendages 1 and
3 only are figured),

Fic. 6.—Normal female: abdominal appendages (appendages 2 and 3 only
are figured). e. External ramus, 7. Internal ramus.

HERMIT CRAB AND PELTOGASTER. 621

PLATE 34.

Fic, 1.—Regenerated appendages (Ist and 2nd abdominal) of modified
male, The animal, which had only undergone one moult since operation,
regenerated appendages in a modified sense.

Vie, 2.—Regenerated appendage (lst abdominal) of unmodified male.
(One moult after operation.)

Fie. 3.—Transverse section through normal testis. Tubule with duct
stained with Delafield’s hematoxylin and eosin. spe. Spermatocyte. Duct
filled with spermatozoa (spz.). Objective 7*, ocular 2 of Koristka.

Fic. 4.—Preparation of infected testis with Schneider’s aceto-carmine.
Small and irregular spermatophores (sp.p2.) with much reduced numbers of
spermatozoa. sp.ph.c. Chitinous cap-part of sphermatophores. spe. Sperma-
tocytes, and 0v,, 0v2, two stages in the development of ova. Obj. 7*, oc. 5,
Koristka.

Fic. 5.—Transverse section through infected testis tubule. Stained with
borax carmine. cd. Irregular masses of chitin in the duct. Other lettering
as before. Obj. 7*, oc. 2, Koristka.

Fic. 6.—Transverse section through recovering testis tubule. Reduced
number of egg-cells (possibly some resorbed); but ov, here represents a more
advanced state of development than in infected testes. Cytoplasm shows
reticular structure. Higher magnification than last. Stained with Mayer’s
carmalum. Obj. 7*, oc. 6, comp. Koristka.

(The figures are drawn with Zeiss’s camera lucida.)

MICROHYDRA RYDERI. 623

On the Medusa of Microhydra ryderi and on
the Known Forms of Meduse inhabiting
Fresh Water.

By
Edward Potts,
Of Philadelphia, U.S.A.

With Plates 35 and 36.

Tue three species of freshwater jellyfish as yet discovered
were announced during the years 1880, 1893, and 1897 re-
spectively. As the first of these dates was twenty-six years
ago, and as the paper reporting the North American form
was unaccompanied by illustrative figures, it has been sug-
gested to the writer that a general review of the subject, so
far as known, would be useful at this time, and the kind
assurance was given thatif it included ‘ good drawings and a
full account of the polyp and medusa” of Microhydra
ryderi, it would be gladly accepted and published by the
‘Quarterly Journal of Microscopical Science.’

To make such a re-presentation of value I propose first to
offer copies of the published figures of the earlier forms, viz. :
Limnocodium sowerbii, Allman and Lankester;.found
in the Victoria Regia tanks of Regent’s Park Gardens,
London, England (PI. 35, figs. 5, 6, 7, 8), and Limnocnida
tanganyice, Dr. R. Bohm, Lake Tanganyika, Central
Africa (Pl. 35, fig. 12). Following these I give figures
showing the lateral and polar aspects of the medusa of
Microhydra ryderi, Potts (Pl. 36, figs. 13 and 14). Both

624. EDWARD POTTS.

of the figures of this medusa show, as described in the
‘American Naturalist, the eight tentacles, one at the ter-
mination of each radial canal, and the others alternating
with them; the latter figure also shows at the centre, as in
an optical section, the quadrate character of the mouth parts
of the manubrium. The polyp of this species was found in
two localities in the neighbourhood of Philadelphia, Penn-
sylvania, U.S.A., in January, 1885, and studied more or
less continuously for twelve years before the meduse were
detected in August, 1897. The figures of the medusa here
given (figs. 138 and 14) were drawn by Dr. J. Percy
Moore, of the Biological Department of the University of
Pennsylvania, from medusz placed by myself in a very dilute
aqueous solution of formalin some time in August or
September, 1897; and, although possibly somewhat shrunken,
they correspond very closely with my remembrance of them
when fresh, and with my description published in the
‘American Naturalist’ of December in that year. These
medusee have as yet only been seen as they were developed
from the hydroids in comparatively small culture-jars at my
home, where they were quickly lost sight of amongst water-
plants kept floating at the surface. Under these circumstances
none probably reached an age of more than two or three days,
and can only be safely compared with figs. 5 and 6, Pl. 35,
supposed to be mature, by experts familiar with similar types
of marine medusa. During several weeks of the year 1897,
and occasionally during the two following years, Dr. Daven-
port, Prof. Cheyney, with myself and some others, were
cognisant of the budding and liberation of several medusz.
I have felt an especial pleasure in finding, as represented
in figs. 15 and 16, Pl. 36, two rough drawings (the latter by
camera lucida) among my papers dated September 5th, 1897.
They represent sketches made, during the afternoon and
evening, of the same medusoid bud at an. early stage of its
development; and, under date of the 7th of the same month
in my letter press-book, I found copies of the following letter
addressed to Dr. C. B. Davenport, referring to the same

MIGROHYDRA RYDERI. 625

observations. “ Most important was my success in removing
a swelling bud to the stage of my microscope, where its con-
nection with the stem of one of a colony of two hydroids was
unquestionable. The pup, of course, grew faster than his
daddy, who curled his head to one side in evident disgust,
the pedicle of the medusa-bud assuming the position of the
main stem. I made a rude camera lucida sketch at the
time.” (The freehand drawing (fig. 15) had been made earlier
in the day.)

“ On the following morning I was pleased to see that deve-
lopment had gone on unchecked by its removal. When
examined that evening the tentacles had been projected to
nearly their full length, the expansion and opening of the
dise was nearly complete, and pulsations had evidently con-
tinued for some time. This morn (September 7th) the jelly-
fish was swimming about the little dish.

‘Another important point. Atthe time of my last evening
examination, the cellular structure of the nearest hydroid
stem, as well as that of the pedicle of the medusa, had entirely
disappeared, leaving the struggling creature attached only
by a diaphanous cuticle almost impossible to define, within
which were a very few spherical floating cells and numbers
of detached thread-cells. This disintegration of tissues
serves to account for my previous failures to discern the sup-
porting stems, and to explain the long detention of the medusa
when there seemed to be no attachment.

“J do not mean that the Microhydra had entirely dis-
appeared. The stem bearing the other head had, in fact,
elongated, and the nearer stem seems to be re-adjusting its
conditions and reforming a capitulum.”

It must not be at all understood that this was the only in-
stance of the liberation of a medusa observed ; we saw several
of them, but none that so entirely eliminated the personal
equation of a poor memory.

Before introducing a number of drawings illustrative of the
structure of the hydroid form of Microhydra ryderi, I beg

626 EDWARD POTTS.

leave to reproduce for comparison figs. 3 and 4, and 9,
10 and 11, Pl. 35, given by Alfred Gibbs Bourne and F. A.
Parsons in their descriptions of a certain hydroid found in
the Regent’s Park Gardens (but in other tanks than those in
which the Limnocodium had been discovered), and assumed
to be the hydroid condition of that medusa. I have not been
informed that such a derivation had been actually traced.’

The remaining figures of Plate 36 (one or two exceptions
noted) have been very accurately copied by F’. von Iterson,
artist, from excellent drawings made some years ago by my
friend, Dr. John A. Ryder, of the University of Pennsylvania,
who was enthusiastic in the study of the polyp with which
his name is associated, but who died before the medusa had
been seen, much to my sorrow and to the loss of science.
They are—

Fig. 17 (Pl. 36).—A_ bifid or bi-capitate representation of
M. ryderi from life, of which one branch (a) taken by
itself may be considered as showing the appearance of the
single form of the hydroid, and (b) as the compound form,
where two are found branching near the base: cpm., the
capitulum, more or less spherical, terminating the distal end
of each mature polyp, and charged upon the surface with
varying numbers of .—nematocytes or thread-cells; p.d.
is the pedal disc or foot, and J. a fully-formed asexual larva
about to be liberated from one of the polyps. x 64.

Fig. 18.—Another bifid form which Dr. Ryder describes as
‘a, freehand sketch of M. ryderi on the side of the culture-
jar with the capitulum of one (much elongated branch ring-
like mouth depressed, stem of glass-like transparence, with
the exception of cell-ends and boundaries shimmering in the
centre.” Incomplete ‘basal branch (b) very granular and
relatively very opaque.”’—J. A. R., October 12th, 1891.

Fig. 19.—Of this Ryder says, “Oral end of adult polyp,

1 This derivation, however, had been traced in 1890 by Fowler. The chief
figure of the Limnocodium polyp, showing a terminal “ medusa” in
section, is reproduced here in Pl. 35, fig, 1.—E. Ray LanKEsteEr,

MICROHYDRA RYDERI. 627

showing manner of eversion of hypostomal organ, in longi-
tudinal section.”—J. A. R., X 365.

Fig. 20.— Oblique section of pedal disc of adult, showing
cuticular sheath.’—J. A. R., X 360.

Fig. 21.—Original diagram by the writer to show the
locality of the larval bud (J.b.), its gradual growth, and the
manner of separation of the fully formed larva. A course of
24 hours or less.

Fig. 22.—Highly magnified section (xX 244), showing bud-
ding larva (J.0.).

Fig. 23.—Another bicapitate hydroid, showing larval bud
in situ, lasso threads highly stimulated. x 67.

Fig. 24.—Young larva forming capitulum; thread-cells
accumulating at distalend. x 112.

Fig. 25.—Single hydroid adherent to glass of jar; base
and adjacent parts showing adherent threads of Nostoc and
adventitious particles. x 30.

Fig. 26.—Drawing by J. P. Moore from cross section of
mature hydroid by Harold Wingate, showing structure, en-
teric folds, etc. Magnified 480 diameters.

In the latter part of January, 1885, while examining under
the microscope the surface of some small fragments of gneiss
and mica-schist, chipped from larger rocks in the bed of
Tacony Creek, a small mill-stream near Philadelphia, Penn., an
affluent of the Delaware River, I first observed the primitive
freshwater hydroid, since classified as M. ryderi. The
stones had been brought into my house during the previous
summer to enable me to study the winter condition of a
newly discovered bryozoan, Paludicella erecta, mihi
(since named Pottsiella erecta by Prof. Kraepelin), that
grew abundantly upon them.

The polyp, as seen, was about one half a millimetre in
length by one tenth of a millimetre in thickness, nearly
cylindrical, sometimes simple, in other instances divided near
the base of support into two nearly equal, divergent branches,

628 EDWARD POTTS.

Body soft (hydra-like), with but slight power of extension or
retraction; without tentacles or cilia, but terminated by
conspicuous symmetrical capitula varying from time to time
from a generally hemispherical shape, and bearing upon or
near their surfaces fifty or more nematocysts, each with
slightly protruding palpocils. At the extremity of each
capitulum an oval aperture, expansible and contractile, was
with difficulty determinable, but was pee abundantly
proven (fig. 19, Pl. 36).

The creature was fixed in position, and apparently without
power of locomotion when once attached—standing or sway-
ing upon its proximal extremity, which latter was specialised,
as in the common Hydra, into a pedal disc or foot (fig. 20).
Besides the nematocysts collected within the limits of the
capitulum, others were rather sparsely distributed throughout
all parts of the ectoderm.

The animal was often entirely quiescent; at other times
its sole movement consisted in a very deliberate writhing or
swaying of the whole body, or a curving of the capitulum
abruptly against the body as a bent finger curves upon
itself.

During the fifteen weeks or thereabouts that these crea-
tures were under close daily and nightly observation, at least
fifteen individuals, old and young, were seen; and as the
positions of the mature forms seemed unchangeable, and the
larvee had but slight—or even accidental—powers of loco-
motion, there was little difficulty in identifying them from
day to day. It must be observed, however, that on account
of the scarcity of specimens, and the apparent impossibility
of removing any from its place of attachment (on rough
stones) without fatal result, I was unable to place them in
good positions for examination by transmitted light, and
that, standing as they generally did, upon opaque surfaces,
and in positions directed more or less toward the observer,
their internal structure, and the changes occurring therein,
could not be seen with entire success in the living animal. For
instance, notwithstanding the most careful watching during

MICROHYDRA RYDERI. 629

this long period, I had entirely failed to see the act of feeding
until the last days of the last surviving mature individual.

I copy from my note-book, under date of May 11th, 1885 :—
“On this date, for the first time, I saw M. ryderi capture
and swallow a small rotifer. A hydroid standing upon the
clear surface of my glass dish had developed a lateral branch,
which had only recently completed its capitulum. While
watching the branch this evening, one of the turtle-back
rotifers, that have been so plentiful, came swimming by and,
fortunately for me, touched some of the palpocils, when it
immediately lost all power of motion, and adhered, or was
held, to the surface of the capitulum. ‘his extremity was
then deliberately curved around in a direction toward the
rotifer, pressing and holding it against the adjacent body of
the hydroid, when, by some method not clearly discovered,
its head was directed toward and absorbed into the ex-
panded oral aperture of the latter. Its downward progress
was then relatively quite rapid, so that it was little more than
five minutes from the moment of capture before the caudal
appendage of the victim had passed into the orifice, and the
whole creature could then be followed through the channel
of nearly transparent cells to its resting place about the
centre of the polyp. ‘The so-called gizzard-motion of the
rotifer still suggested life; but Microhydra’s digestion must
have been quite energetic, for about an hour later the empty
test had already been ejected.

“ About the same time I witnessed the act of ejection of
the test of a similar rotifer from the main extremity of this
hydroid, and the next evening I counted six of such ‘ relics
of mortality’ surrounding this form of voracious simplicity.”

Two modes of reproduction or development were noticed.'
The most important of them, perhaps unique in its character,
was seen in the formation of asexual larvee non-ciliated and
inert (represented in diagram fig. 21), which, after one or
two weeks, became fixed upon their pedal discs, developed
capitule, and appeared essentially mature (fig. 24).

' This was twelve years before tle medusa was discovered

VOL. 50, PART 4,—NEW SERIES. 45

630 EDWARD POTTS.

One of the first discovered polyps produced larvee February
3rd, March 2nd, 11th, and 15th. Another produced larvee
on February 5th, 16th, and 28th,and March 3rd. They were
all thrown off from the same neighbourhood in the parents,
and the visible process was accomplished in a few hours.
In every case the act was preceded by an increased opacity
of the lower half of the polyp, as though it became filled with
granules, an effect, perhaps, of the rapid segmentation of
nuclei in the endodermal cells (see fig. 22). The external
stages in their formation, often completed within a very few
hours, are rather suggested than illustrated in diagram fig.
21 mentioned above. It will be noticed that the product,
whatever it may be called, is not a self-sustaining duplicate
of the parent hydroid, but an immature individual, needing
entire development of its functional parts before it becomes
capable of ministering to its own growth or of reproducing
others.

When detached the larve rest wherever they may chance
to fall, changing place only asa result of the slight writhing
motion sometimes noticed, or of a protoplasmic, amoeboid
action within the surface cells. They are long, oval, cylin-
drical bodies, the central parts exhibiting infinite numbers of
yolk-like granules from which nucleated endodermal cells are
gradually developed, while the superfices, even before separa-
tion, show a continuous, clear, hyaline ectoderm. The
external surface seems generally smooth, but a high magnify-
ing power brings out, particularly near the extremities, a
delicate, variable cumulation, the outer convexities of minute
ectodermal cells. A few nematocysts can very early be
detected, indifferently scattered around the margin. After
a few days of quiescence, one extremity (Ryder says that last
attached to the parent, the nearest to its proximal end),
becomes attached to the supporting surface, and the other is
gradually elevated into a more or less erect position when
the thread cells are formed in numbers about the distal
extremity (fig. 24), and maturity is reached with the fully-
formed capitulum.

MICROHYDRA RYDERI. 631

The other mode of reproduction or of propagation is by
branching. ‘This was seen in two instances, one of them
from the same polyp that had been already known to have
produced four larve, and the branch originated upon the
same side and locality. The action was even more rapid in
that the branch appeared and completed its capitulum within
twenty-four hours. Its formation was not preceded by the
turgidity and granular appearance mentioned in the case of
the larval buds, but an uninterrupted sequence of clear endo-
dermal cells was traced from the main stem into the branches.
These branches were, possibly, later constricted off, as in the
case of the common Hydra, to form independent hydroids,
but I did not see it.

During the twelve years intervening between 1885 and
1897 collections had frequently been made from the
Schuylkill canal, on the banks of the Schuylkill River, far
above tide water, and from a water-shed quite separate from
that supplying the locality previously mentioned. The
special objects of search were here also stones bearing
localised forms of Bryozoa, and, as before, I found M.
ryderi amongst them. These stones were dredged up from a
depth of six or seven feet at a place where a continuous
rush of water kept them from being covered, and their
encrusting fauna from being smothered by the mud that in
other places embeds the canal.

Whether, under these circumstances, medusz are commonly
and normally developed, and their eggs hatch into hydroid
polyps, or whether nature here provides more abundantly the
asexual larval method of reproduction above described must,
for the present, remain a problem.

In concluding, it is hardly necessary to invite the attention
of scientists to the fact that we have, in these three fresh-
water forms, an equal number of species, no one of which has
been conclusively studied ; that their appearance in three of
the five grand divisions of the earth points very plainly to
the probability that closer methods of research may very soon
discover others in familiar but unsuspected places; that,

632 EDWARD POTTS.

whatever we may believe as to the origin of Limnocodium
sowerbii, it is hardly possible to doubt that Limnocnida
tanganyice and M. ryderi are native to the districts in
which they have been found, and that, unless or until the
polyps of Bourne and Parsons have been seen to produce
the Limnocodium, it is quite within the limit of possibility to
suppose that they have had their origin in the Thames, or
from any other source from which the tanks in which they
were found may have been filled.

EXPLANATION OF PLATES 35 anp 36,

Illustrating Mr. Edward Potts’ paper, “On the Medusa of
Microhydra ryderi, and on the Known Forms of
Medusz inhabiting Fresh Water.”

PLATE 35.

Fic. 1.—Diagrammatiec section of a terminal medusa-bud of the hydroid
colony of Limnocodium sowerbii. After Fowler, ‘Quart. Journ, Mier,
Sci.,’ vol. 30, 1890, Pl. 82, fig. 8.

Fic. 2.—Portion of hydroid colony of Limnocodium sowerbii. After
Fowler, loc. cit., Pl. 32, fig. 2.

Fic. 8.—Hydroid form of Limnocodium sowerbii upon roots of pond-
weed. After Bourne, ‘Proc. Roy. Soc.,’ xxxviii, 18$4, p. 10.

Fie. 4.—Portion of one of the hydroids of Fig. 3 seen in optical section.
After Bourne, loc. cit., p. 10.

Fie, 5.—Limnocodium sowerbii, as seen floating. After Lankester,
‘Quart. Journ. Mier. Sci.,’ vol. 20, 1880, p. 357.

Fic. 6.—Limnocodium. After Allman, ‘Journ. Linn. Soc.,’ ‘ Zool.,”
xv, 1880, p. 134. x 8.

Frias. 7 and 8.—Two views of an embryo medusa of Limnocodium
sowerbii. After Lankester, ‘Quart. Journ. Mier. Sci.,’ vol. 21, 1881,
Pl. 18, figs. 4 and 1.

Fics, 9, 10, and 11.—Hydroid figured by Parsons, ‘ Journ, Quek. Micro.
Club,’ ser. 2, vol. ii, Pl. 6, p. 1380.

MICROHYDRA RYDERI. 633

Fie, 12.—Medusa of Limnocnida tanganyice. After Moore, ‘The
Tanganyika Problem,’ 1903, p. 299.

PLATE 36.

Fies. 13 and 14.—Medusa of Microhydra ryderi, lateral and polar
aspects. Drawings made by Dr. J. P. Moore. x 116.

Fies. 15 and 16.—Medusoid bud developing upon a hydroid colony of
Microhydra ryderi. Two stages, drawn by the author, during the after-
noon and evening of the same day.

Figs. 17—20.—Drawings of Microhydra ryderi, made by Dr. J. A,
Ryder (see text, pp. 626-7).

Fies. 21—25.—Drawings of Microhydra ryderi, by the author (see
text, p. 627).

Fie, 26,—Section across a mature hydroid of Microhydra ryderi.
Drawn by Dr. J. P. Moore. x 480.

Nore on THE Forecoina sy BE. Ray LANKESTER.

Mr. Potts has been so very kind as to send to me, at my
request, a specimen of the minute medusa liberated by
Microhydra and preserved in formalin. It is one of two
which remained in his possession. Whilst further study of
Microhydra and its medusa is urgently called for, it is
clear that further research is needed in order to settle the
doubt entertained by Mr. Potts as to the actual genetic con-
nection of Bourne’s hydroid with the medusa Limno-
codium. It is within the bounds of possibility that that
hydroid is a native Huropean hydroid, similar to Microhydra,
and not connected with the Limnocodium life-cycle.

1
2

"

i

o>
ou)
Or

MEDUSA OF MICROHYDRA RYDERI.

On the Freshwater Medusa liberated by Micro-
hydra ryderi, Potts, and a Comparison
with Limnocodium.

By
Edward TT. Browne, B.A.,
Zoological Research Laboratory, University College, London,

With Plate 37.

TaroucH the kindness of Professor Ray Lankester, I
have had the pleasure of examining a specimen of the
Medusa liberated from the freshwater Hydroid Microhydra
ryderi.

The veteran American naturalist, Mr. Potts, of Philadelphia,
recently sent to Professor Lankester a manuscript on “ Known
Forms of Medusee Inhabiting Fresh Water,” for publication
in the ‘ Quarterly Journal,’ and this communication of mine
forms a kind of appendix to it; it should be regarded as such,
since I have before me an advanced proof of Mr. Potts’
paper. The title given by Mr. Potts to his communication
did not adequately convey the importance of its contents, and
has been modified accordingly by the editor. Mr. Potts has
at last given us a description with excellent figures of the
Hydroid phase of Microhydra and the first figure of the
Medusa.

When Professor Lankester showed me the original draw-
ings of Microhydra I noticed the remarkable resemblance
between the hydroid phase and that of Limnocodium, but
we were doubtful about the Medusa, as Mr. Potts had not
given a detailed description of it. Since we were not sure
about the presence of sense organs, Professor Lankester asked

636 EDWARD T. BROWNE.

Mr. Potts if he could spare a specimen for further examination.
A specimen was very kindly sent over from America by Mr.
Potts, and I sincerely thank Professor Ray Lankester for his
generosity in handing it to me for examination.

The figures of Medusa illustrating Mr. Potts’ paper were
not drawn from a living Medusa, but from a specimen which
had been in weak formalin for several years. To obtain a
really satisfactory drawing of a Medusa it must be made whilst
the animal is alive. In drawing a preserved specimen allow-
ances have frequently to be made for contraction and dis-
tortion, and therein lies a source for error. After studying
living Medusz for a few years the allowance for defects can
be fairly well estimated, though occasionally one may go
badly astray.

The specimen which I examined was in formalin and in good
condition, but the umbrella was badly crumpled on one side.
It was a difficult object to examine owing to its minuteness,
being less than half a millimetre in diameter, and not easy to
fix in a definite position. The drawings which I have made
were finished before I received the proofs of Mr. Potts’ plates.
I then noticed that my drawing of the Medusa did not quite
agree with that made by Dr. Moore, so I again examined the
specimen, but found that it was not necessary to make any
alterations.

Tne Descrrprion or THE Mupusa or Micronypra (PI. 37, fig. 1).

Umbrella.—The umbrella is campanulate, a little broader
than high (0°4 mm. in width and 0°3 mm. in height), with
thin walls. No nematocysts could be found on the ex-
umbrella, though the ectoderm cells were plainly visible.
The velum is broad, and here also the ectoderm cells with a
rounded nucleus could be easily seen.

Stomach.—The stomach is large for the size of the
Medusa, being about three quarters the length of the cavity
of the umbrella. It appears to be more cylindrical than
quadrangular in transverse section, and tapers slightly
towards the mouth. In this specimen the mouth is fairly

MEDUSA OF MICROHYDRA RYDERI. 637

well expanded, but there are indications of four small lips,
which are simply infolds of the margin.

Canal system.—There are four radial canals, which are
not at all conspicuous, and have the appearance of thin lines
running from the base of the stomach to the margin of the
umbrella. In fact the canals are rendered visible by the
brownish colouration of their endoderm cells. A circular
canal is probably present, but the thick layer of ectoderm
round the margin of the umbrella prevented me from finding
a definite canal. T’o demonstrate the existence of a circular
canal would have necessitated the cutting of sections.

Absence of gonads.—The specimen does not show the
slightest trace of gonads. The figures of the Medusa illus-
trating Potts’ paper (Pl. 36, figs. 13 and 14) would lead one
to believe that gonads were present along the whole length of
the radial canals and that they also extended round the
stomach. The Hydromeduse either have their gonads upon
the stomach and its lobes, or else upon the radial canals, but
none are known to possess gonads in both positions. If a
Medusa was found with gonads upon the stomach and radial
canals it would have to be placed in a special order. These
shaded thickenings are undoubtedly misleading, and I fail to
see what they are intended to represent.

Tentacles (Pl. 87, fig. 2).—There are eight tentacles (four
per-radial and four inter-radial) similar in size and shape.
They are probably in a semi-contracted condition, which gives
them a rather stunted appearance. ‘lhe base of the tentacles
is apparently attached for a very short distance, on its upper
side, to the margin of the umbrella. The manner in which
the tentacles curl at their base and hang down points to an
attachment, though I could not clearly see it. There is no
indication of a definite basal bulb at the base of the tentacles.
Nematocysts are very scarce in the tentacles, and it was only
after a long search that I found any, as they were not on the
surface but underneath the ectoderm cells.

Around the margin of the umbrella there is a thick layer
of ectoderm cells, similar in structure to those of the tentacles,

638 EDWARD 'l’. BROWNE.

and between these cells there are a few nematocysts, like
those found in the tentacles.

Absence of sense organs.—When I saw the original
drawing of Plate 36, fig. 14, before I received the specimen,
I thought that the little circles at the bases of the tentacles
represented marginal sensory vesicles situated just above the
root of the tentacle. They certainly have the appearance of
sense organs with a single otolith. In diagrams, and often
in good drawings, sense organs are drawn in a similar manner.
I thoroughly searched the margin of the umbrella for sense
organs, finally using an oil immersion lens, but failed to find
any indications of a sense organ either at the base of the
tentacles or in between the tentacles. The margin of the
umbrella has a slight brownish colour, as if the Medusa had
been killed with Flemming’s solution. All the nuclei are of
a faint brownish colour, easily seen with an oil immersion
lens, and the cell walls are also well defined, so that if any
sense organs had been present they ought to have been
visible. The little circles in the figure are intended for
optical sections of the bases of the tentacles. They represent
the attachment of the tentacle to the margin of the umbrella,
and have nothing whatever to do with sense organs.

A COMPARISON BETWEEN LimNocoDIuM AND MiIcROHYDRA.

When I first saw the drawings of Microhydra and com-
pared them with the figures of Limnocodium, it seemed
quite possible that Microhydra might turn out to be the
well-known Limnocodium, but after examining the Medusa
of Microhydra the idea of such a possibility soon vanished.

Hydroid.—In the Hydroid phase the resemblance between
Limnocodium and Microhydra is very close, as will be
seen oncomparing Pl. 35, fig. 10, with Pl. 36, fig. 17. In both
forms the hydranth has degenerated to its simplest condition—
i.e. to merely a body without tentacles, and when the native
place of Limnocodium has been discovered we may obtain
a clue to the cause of degeneration,

MEDUSA OF MICROHYDRA RYDERI. 639

In the case of Microhydra the Hydroid appears to attach
itself to rocks and stones in swift-running streams.

Under such conditions long flexible tentacles, like those
possessed by the common freshwater Hydra, would stream
out with the current, and be of little use for the catching of
food. One would rather expect to find in such a situation a
Hydroid with very short and fairly stiff tentacles, like those of
Coryne, which lives between tide-marks. But Microhydra
has, perhaps, conquered its new habitat at the expense of
its tentacles, as it may be reasonably assumed that this
Hydroid is descended from one which formerly inhabited
the sea.

There is, however, a difference between the Hydroid phase
of Limnocodiumand that of Microhydra. The Hydroid of
Limnocodium secretes from its body a glutinous mucus, to
which adhere particles of mud and other débris, so that a
protecting case is formed round the body, leaving only the
oral end free, and this end is capable of contracting within
the tube.

The Hydroid of Microhydra, so far as I can judge from
the figures and description, forms no protecting case to its
body. Potts’ figure 25 shows the Hydroid attached to the
glass of the aquarium with the “base and adjacent parts
showing adherent threads of Nostoc and adventitious
particles.” Any one who has kept a freshwater aquarium
knows that the glass becomes thickly coated with unicellular
Alge. Mr. Potts figures this coating surrounding the base of
the Hydroid, the body of the Hydroid being shown by him
quite naked.

Medusa.—The comparison between the Medusa of Limno-
codium and of Microhydra is not so simple as that of their
Hydroids. In the first place a stage exactly similar to that
of Microhydra has not been described and figured for
Limnocodium.

Fowler has described the Medusa-bud on the Hydroid of
Limnocodium ata very early stage, whilst still attached to the
polyp (Pl. 35, fig. 1), but as his supply of material failed he was

640 EDWARD 'T’. BROWNE.

not able to proceed any further. Next we have some very early
free-floating stages described by Lankester (Pl. 35, figs. 7
and 8), Although there is still no absolute proof that the
Medusa-buds found upon the Hydroid do develop into the
Medusa known as Limnocodium, still the circumstantial
evidence is very strong.

I have seen many species of marine Hydroids bnd off
Meduse, but have never seen Meduse liberated at such an
early stage as those of Limnocodium, which look as if they
were developing direct from eggs. On the other hand, the
Medusa of Microhydra looks like a Medusa which has just
been liberated from its Hydroid, and Mr. Potts states that
probably none of the specimens seen were more than two or
three days old. The Medusa of Microhydra on liberation
is at a far more advanced stage than the earliest floating
embryos of Limnocodium. (Compare PI. 35, fig. 8, with
Pl. 36, fig. 13).

The embryo of Limnocodium has already got one sense
organ developed, so one would expect to find sense organs in
the Medusa of Microhydra if it had any. If an adult
Medusa has sense organs one always finds (I cannot remember
an exception) a certain number (generally about four or eight)
of sense organs present in the young Medusa when ready for
liberation. To pin one’s faith on the absence of sense organs
in preserved specimens is not a safe proceeding, because
sensory vesicles have at times a wonderful way of becoming
invisible after preservation, especially when alcohol is used.
Their disappearance is generally due to excessive shrinkage
of the tissues when the specimens are too rapidly transferred
from sea water to strong alcohol. The great advantage of
dilute formalin is that it does not produce a shrinkage of the
jelly, and that the sense organs can be more easily found.

The great difference between the Medusa of Microhydra
and that of Limnocodium lies in the structure of the
tentacles. A few years ago I found out that the shape and
structure of the tentacles, and particularly the shape of the
basal bulb, were an exceedingly useful and reliable aid in

MEDUSA OF MICROHYDRA RYDERI. 641

the determination of species. The tentacles of Limno-
codium are quite unlike those of Microhydra. I have a
few specimens of Limnocodium which came from Regent’s
Park in my collection, and they show the tentacles in all
stages of development. The tentacles even at their earliest
stage, when as mere buds upon the margin of the umbrella,
show a character which is not found in the tentacles of
Microhydra, nor have I found or yet met with it in any other
Medusa. The nematocysts are definitely arranged at the ends
of little papillae. At first there are one or two nematocysts
in each papilla, but later on the number increases to about
five or more. For the purpose of comparison I selected a
very small tentacle of Limnocodium, a little over one
millimetre in length, and made a drawing (PI. 37, fig. 3)
from the central portion of the tentacle to the same scale as
the drawing of the tentacle of Microhydra (PI. 37, fig. 2).
It will at once be seen that there is a marked difference between
the tentacles of these two Meduse. The nematocysts have
also a different shape (Pl. 37, figs. 4 and 5).

THe Repropucrion oF MICROHYDRA.

The hydroid has two methods of reproduction; one is
asexual, the other is sexual. Mr. Potts considers the
budding of new hydranths, which are not set free, to be a
second asexual method. ‘The Hydroid is at first a single
polyp, later on from its base another polyp is developed,
but as the second polyp is not detached a colony of two
individuals is formed. The Hydroid phase of Limnocodium
in the same manner is also colonial, but has from two to four
polyps. This is not a case of reproduction, as there is no
increase in the number of independent individuals, but simply
one of branching to form a colony.

The asexual method of reproduction of Microhydra seems
to me, from the appearance of the figures given by Potts
(Pl. 36, figs. 17, 21, 24), to be reproduction by fission, which

occurs in certain marine Hydroids.

642 EDWARD T. BROWNE.

Allman, in 1871, gave an account, with figures, of repro-
duction by spontaneous fission in a Hydroid which he named
Schizocladium ramosum. Although Allman had worked
for many years upon British Hydroids, yet he had never
before met with a case of fission amongst them. MHincks, in
1872, found Campanularia neglecta reproducing by
fission in a similar manner, and refused to accept Allman’s
new genus Schizocladium, “ which seems to rest on a single
character, the development of fission-frustules in a certain
way—a character which, there is reason to believe, may
have a wide range amongst the Hydroida.” Hincks suggests
that Allman’s Schizocladiumis probablyan Obelia. Allman
found the colony in Loch Long (Firth of Clyde), and states that
it bore a considerable resemblance to that of Obelia dicho-
toma. It was without gonosomes, and it was the absence of
the gonosomes that led Allman to establish a new genus for a
Hydroid that was reproducing by fission. Allman states that
the frustule on liberation has a distinct endoderm and ecto-
derm, but no perisare. It has no means of locomotion, and
attaches itself by a mucous excretion from its surface to the
glass of an aquarium. Soon after attachment the mucous
excretion forms round the fission-frustule a very thin tube,
which is the perisare. Once attached a hydranth develops
from it, then later on other hydranths are formed and a little
colony arises by branching.

Fission is not merely the nipping off a small portion of the
coenosare, as the fission-frustule contains all the elements neces-
sary for the formation of a new colony. It does not usually
take place at the same time as sexual reproduction. In Micro-
hydra the method of fission is different from that in
marine Hydroids, as there are no branches. An outgrowth
takes place from the side of the hydranth (PI. 36, fig. 21),
and this is nipped off and develops into a hydranth.

It is very probable that the Hydroid of Limnocodium
also reproduces asexually by fission. Parsons, who kept the
Hydroids in an aquarium, states “ that the polyps made their
appearance on the side of a sponge which had been in contact

MEDUSA> OF MICROHYDRA RYDERI. 643

with a pipe (the hot-water pipe in the Victoria Regia tank).
This fact leads me to the inference that the polyps were
developed from germs contained in the water which I brought
away with me, for I do not see how they could have got there
while the sponge was alive; moreover, they were in different
stages of development, the earliest stage seen by me being
a little mound of fuscous coloured sarcode.” Fowler has
figured a section of a bud, which “ may either remain attached
to the parent, or may be nipped off and settle close by, its
tissues in either case gradually undergoing the differentiations
which characterise the adult.”

The sexual method of reproduction of Microhydra is no
doubt by means of Medusze. Up to the present time only the
earliest stage of the Medusa is known. The young Medusa
has the appearance of an Anthomedusa, but it is impossible
to assign it to a definite family until the later stages have
been seen. It would be most interesting to know what
became of the Medusa after leaving the Hydroid. The Medusa
is set free in a stream or river, so that it must be carried along
with the current in the direction of the sea.

The exact method of the sexual reproduction of Limno-
codium still remains a mystery. During the period it lived
in England only the male Medusa was found, and not the
slightest evidence on the presence of the female sex was
obtainable. ‘The Medusa has not been seen in England since
1893, so it is evident that our stock has completely died out.

In 1901 Limnocodium suddenly appeared in the Victoria
Regia tank at Lyons, and an account of it is given by Vaney
and Conte. The Hydroid phase was searched for, but was not
found, and the Meduse were all males.

In 1905 Boecker recorded the appearance of Limno-
codium in the Victoria Regia tank at Munich, and again
only males were seen. The author does not mention the
occurrence of the Hydroid phase.

644. EDWARD T. BROWNE.

REFERENCES.

On Microhydra.

1885. Ports, E.—* Microhydraryderi,” Science Bulletin, p.v. (a supple-
ment in ‘Science,’ vol. v, No. 123).

1897. Ports, E.—“ A North American Freshwater Jelly Fish (Microhydra
ryderi),” ‘American Naturalist,’ vol. xxxi, pp. 1032-1035.

On Limnocodium.

1890. Fowxer, G. H.—“ Notes on the Hydroid Phase of Limnocodium
sowerbii,” ‘Quart. Journ. Micr. Sci,’ vol. xxx, pp. 507-513,
pl. xxxii. (This paper contains the bibliography of Limnocodium
up to 1885.)

1893. LanKesTER, HK. R.—“‘ Reappearance of the Freshwater Medusa
(Limnocodium sowerbii),” ‘Nature,’ vol. xlix, p. 127.
(Records the occurrence of Limnocodium at Sheffield, and
gives an excellent summary of its life-history.)

1901. Vaney, C., and Contr, A.—“Sur le Limnocodium sowerbii,”
‘Zool. Anzeiger,’ Bd. xxiv, pp. 5338-534; two figures in text.
(Records the occurrence of the Medusa at Lyons.)

1905. Borcxer, E.—‘‘ Ueber das Vorkommen von Limnocodium im
Miinclner botanischen Garten,” ‘ Biol. Centralbl,’ Bd. xxv, pp. 605-
606.

On Reproduction in Hydroids by Fission.

1871. Auuman, G. J.—‘‘ On a Mode of Reproduction by Spontaneous Fission in
the Hydroida,” ‘Quart. Journ. Micr. Sci.,’ vol. xi, pp. 18-21, pl. ii.

1871. Attman, G. J. ‘Monograph of the Gymnoblastic Hydroids,’ p. 151,
fig. 61.

1872. Hincxs, T.—“ Contributions to the History of the Hydroida,” ‘ Ann.
Mag. Nat. Hist.,’ Series 4, vol. ix, p. 390.

1901. Bittarp, A.—‘*De la Scissiparité Chez les Hydroides,”’ ‘ Comptes-
rendus. Acad. Sci. Paris,’ tom, 133, pp. 441-443.

EXPLANATION OF PLATE 37.
Illustrating Mr. HE. T. Browne’s paper “On the Freshwater
Medusa liberated by Microhydra ryderi, Potts, and
a Comparison with Limnocodium.”
Vic. 1.—Lateral view of the medusa of Microhydraryderi. x 150.

Fic. 2.—A tentacle of the medusa of Microhydra. Outer side. x 500.

MEDUSA OF MICROHYDRA RYDERI. 645

Fra. 3.—The central portion of a very young tentacle of the medusa of
Limnocodium. Drawn for comparison with fig. 2. x 500.

Fie. 4.—A group of three nematocysts in the tentacle of the medusa of
Microhydra. x 1000.

Fie. 5.—Nematocysts from the tentacle of Limnocodium. x 1000.

Postscript.—The writing of this paper has led me to
commence investigations on the methods of asexual repro-
duction amongst Hydroids, and the work is now being carried
on in the Marine Laboratory at Plymouth. I have found
Allman’s “Schizocladium ramosum™” and have observed
the formation of fission-frustules, their liberation, and subse-
quent development. My observations completely confirm
those made by Allman.

The frustule when nipped off consists of a thin, transparent
layer of ectoderm with nematocysts, a thick layer of endoderm
loaded with granules, and a hollow central cavity. The bud
which is detached from the hydroid phase of Microhydra
(Potts, Pl. 36, figs. 17 and 24) is exactly like the fission-
frustule of Schizocladium, both in shape and structure.

Hincks’ suggestion that Schizocladium is probably an
Obelia has turned out to be correct. Some of the colonies
have liberated Medusze which belong to the genus Obelia.

I have also found a Clava-like hydroid detaching numerous
fission-frustules from its hydrorhiza. These buds drop to the
bottom of the aquarium and lightly attach themselves by one
end to the glass. They are now a week old and have not yet
begun to develop.

Nov. 4th, 1906.

VOL. 50, PART 4.—NEW SERIES. 46

ENDS PO VOR. 50;

NEW

Amphioxides,
schmidt’s monograph, by A. Willey,
581

Anatomy of Oncholaimus vul-
garis, by F. H. Stewart, 101

Aphis, woolly, J. E. Blomfield, on
canker of apple tree, 573

Apple tree, canker of, by J. E. Blom-
field, 573

Blomfield, J. E., canker of apple tree,
573

Browne, E. T., medusa of Micro-
hydra, 635

Canker of apple tree, by J. EK. Blom- |

field, 573

Castration parasitaire, in hermit crab,

by F. A. Potts, 599
Contributions to knowledge of
Notoryctes, Part 3, Hye, by
Georgina Sweet, 547
“Cystobia” irregularis, by
H. M. Woodcock, 1

Development, etc., of fins of fish, by |

E. 8. Goodrich, 333
Development of Flustrella, early
stages in, by R. M. Pace, 435

review of Gold- |

SERIES.

Development of Nebalia, by M.
Robinson, 383

Dinophilus, nephridia of, by Cress-
well Shearer, 517

Karly Stages of Development of
Flustrella, by R. M. Pace, 435
Epiblastic trabeculee of optic nerve of

frog, by J. T. Gradon, 479
Eye of Notoryctes typhlops, by
Georgina Sweet, 547

Fantham, H. B.,
muris, 493

Fins of fish, notes on development,
ete., by E. S. Goodrich, 333

Fish, notes on development of fins of,
by E. S. Goodrich, 333

Flustrella, early stages in develop-
ment of, by R. M. Pace, 435

Freshwater meduse, by EH. Potts,
623, and E. TI. Browne, 635

Piroplasma

_ Frog, development of optic nerve of,

by J. T. Gradon, 479

Goldschmidt on Amphioxides, re-
view, by A. Willey, 581

Goodrich, E. 8., notes on fins of fish,
333

VOL. 50, PART 4.—NEW SERIES. 47

648

Gradon, J. T., development of optic
nerve of frog, 479

Gregarines, Cystobia and other
neogamous, by H. M. Woodcock, ]

Harrison, R. M., new organ in Peri-
planeta, 377

Hemoflagellates, by H. M. Woodcock,
151 and 233

Hermit crab, modification of, caused |
by Peltogaster, by F. A. Potts, |

599

Life-cycle of “Cystobia” irregu-
laris, by H. M. Woodcock, 1

Limnocodium, compared with
Browne, 635, and E. Potts, 623

Lymphatics of optic nerve, by J. 'T.
Gradon, 479

Median and paired fins of fish, by
E. 8S. Goodrich, 333

Medusa of Microhydra ryderi, by
E. Potts, 623, and E. T. Browne,
635

Meduse of fresh water, by i. Potts,
623, and KE. T. Browne, 635

Microhydra ryderi, medusa of,

by E. Potts, 623, and FE. T.
Browne, 635
Modification of hermit crab by

Peltogaster, by F. A. Potts, 599

Nebalia, development of, by M.
Robinson, 383

Nematodes, Oncholaimus and two
parasitic, by F. H. Stewart, 101

INDEX.

New organ in Periplaneta, by
R. M. Harrison, 377

Notes on fins of fish, by HE. S.
Goodrich, 333

Notoryctes typhlops, eye of, by
Georgina Sweet, 547

Oncholaimus vulgaris, by F. H.
Stewart, 101

Optic nerve of frog, development of,
by J. I’. Gradon, 479

Origin and development of optic nerve
of frog, by J. T. Gradon, 479

Pace, R. M., development of Flus-
trella, 435

| Parasitic nematodes, notes on two, by
medusa of Microhydra, by EH. T. |

I. H. Stewart, 101

_ Peltogaster, modification of hermit

crab caused by, by F. A. Potts,
599

Periplaneta, new organ in, by
R. M. Harrison, 377

Pial sheath of optic nerve of frog, by

J. 'T. Gradon, 479

Piroplasma muris, by H. B.
Fantham, 493

Potts, H., medusa of Microhydra,
623

Potts, F. A., modification of hermit
crab by Peltogaster, 599

| Preliminary account of new organ in

Periplaneta, by R. M. Harrison,
377

| Remarks on genus Piroplasma, by

“* Neogamous ” gregarines, by H. M. |

Woodcock, 1
Nephridia. of Dinophilus,
Cresswell Shearer, 517

by |

H. B. Fantham, 493

Researches on optic nerve of frog,
by J. T. Gradon, 479

Review of Goldschmidt on Amphi-
oxides, by A. Willey, 581

Review of present knowledge of Try-
panosomes, by H. M. Woodcock,
151 and 233

INDEX.

Robinson, M.,
Nebalia, 383

development — of

Schizoneura, J. E. Blomfield, on |

canker of apple tree, 573
Sexual characters of hermit crab

modified by Peltogaster, by |

F. A. Potts, 599

Shearer, C., nephridia of Dino-
philus, 517

Stewart, F. H., Oncholaimus and
two parasitic nematodes, 101

Structure of canker of apple tree, by
J. Ei. Blomfield, 573

Structure of nephridia of Dino-
philus, by Cresswell Shearer, 517

Sweet, G., eye of Notoryctes
typhlops, 547

649

Trypanosomes, by H. M. Woodcock,
151 and 283

| Willey, A., review of Goldschmidt on

Amphioxides, 581
White rat, Piroplasma from blood
of, by H. B. Fantham, 493
Woodcock, H. M., hemoflagellates,
151 and 233
Woodcock, H. M., on Cystobia, 1
Woolly aphis, J. EK. Blomfield, on
eauker of apple tree, 573

“Yolk nucleus” in egg of Ilus-

trella, by R. M. Pace, 435

PRINTED BY ADLARD AND SON, LONDON AND DORKING.

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. CHuRcHILL, 7 Great Marlborough 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 SuppLement Numbers being found inconvenient,
and there being often in the Nditor’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 twenty-five copies of their
communication gratis.

All expenses of publication and illustration are paid by the
publishers.

As a rule lithographic plates, and not text-figures, are used
in illustration.

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 ewtra copies of their
communications at their own expense can have them by applying
to the Printers,

Messrs. ApLARD & Son, 224, Bartholomew Close, E.C., on
the following terms :

For every four pages or less—

25 copies : : : : 5/-
DOs : 3 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 Mussrs. J. & A. CuurcHint, 7 Great MarLporouca Street,
Lonpon, W.

THE MARINE BIOLOGICAL ASSOCIATION

OF THE

UNITED KINGDOM.
30;
Patron—HIS MAJESTY THE KING.
President—Prof. E, RAY LANKESTER, LL.D., F.R.S.

me CE

‘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 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, jractical 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.'l. 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 MEMBERS . ‘ : . £1 1 Oper annum.
Lire MemBers . ; ‘ ‘ . 15 15 O Composition Fee.
FOUNDERS . a 00.50.20 + -

Governors (Life Members of Council) 500 O O

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.

>

Dart, Sounn. Mice Sci il, 50 WS?

Iusart, SousrncMir Sei Vil ONE, 2

+ cI>
GILES GES Io,
8 ©
* .. =~

isa wer
Phases ie!
Spas eh

Huth, Lith’ Londo»

a
L
é
£
7 i
oS
a
1
od
ae
=.
= s
.
: ~

ba

ce

My

e
”
;
We.
;
:
7
-

Duar GournMiar scale bo

Huth, Lith? London

=

.

“

Quart. Sourn Mien Sei Ui, NSH.

ey ae

Huth, Lith? London.

Vv

Quart: Journ. Mié JAR OC L, Uhl. 30, NSB.

f :

s
)

London

Huth, Lith?

Sines eae
ate est a

Pe ai Pee = :

be ra <=, 2d) hoah x

; } Ps ES ferns ye coe SSSI eS ea aS
z Seer .

Rut hee
Tp ceergee rn PR Restuaebare!

ph. i.

afY)

rh
ie,

Pines
La)

Huth, Lithy London.

7 “
-
-
- or
A *
*
/
1
—
°
>
‘
‘
bd

‘

Huth. Lith Londox.

2
y i oe SS — mea :
' F ~ a EEE
{ : -
\ .
Fi 7 ‘
|
7
al
\ Hy
} oe =y

Fig. 40.

Tah Lith? London.

2

RY

E.3.Goodrich del, pre Huth, Lith? Landon.

Huth Tith® London.

E.S.Goodrich del

Huth, Lith’ London.

E.S. Goodrich, del

ES. Goodrich, del Huth, Lith” London

i | | -

wopueT syaty UTA
‘TOP Woupoop oy

a nel
rani NHS ry

| \ \ \ ae AN \\
Sea vangagnnneennt

eran pee My

¥

st?

‘

Pe a eS
fete -

a C2) 9.
COS oe ee)

Huth, Tith® London.

Huth, Lith® London

M.R.. del.

2 yaa
2 ee

oe
— x roe

Huth Lath? Londen i ae
MB..del. F -

>

9

e

>
>
aS
CM.

ey

Qe ae,
ces BIS Ss ae

3 “ee =f a “es Z
\ Ee oe (ST

iS Qe Sry
= MN RE
eee BOO

Huth, Lith® London

Huth. Lith® London

Wa
7

art 7, ~ oie

ary ‘ y

; 4

oy

oe : : é
Miss

ive

Si
.

“

23.
Huth, Lith? London

uO

Huth, Lith? London, — ig

RMP dei ;

_ don

Huth, Lith’ London
RHP. del

yes; S20)
2, aatenehe

-* A
a

Huth, Lith* London

; th, Lath? London.
J.T. Gradon del. : Hu

Lew

JT Gradon, del

tet.

Ze Nae ye

a eS = cA
= \ E

z pepe

20! oe
==

p08
=i = — mS

a J

Quant Souren Mien Sei UL SONSHL28.

Huth, Lith? London.

PIROPLASMA MURIS,

if} ii t
Bid

" Ps..
é iy

Huth, Lith? London

Quart IournMcr: sca VA. BONSG. 30

Mouth

1. Nephridium.. an

2. Nephridium. ......... '@ i : \ = (5) OF NSH. N aet------.-- bic. 2.
Sol.

| 3. Nephridiam.-.--- ( Col {if © IA CAL 1) A SET tal.

4.Nephridium-------¥

~| \ JOY Cb.
at (COIN | 5
Za Arta

SAN Ae

VESO, ===

Fig 17.

GS. del. Hoth, Lith? London

(Vol SONSGL FL

“
o
c

ne)

Quart, Burn Mic

Huth, Lith? London

G Sweet del

om al Rendall

S
5
$

0
g200 4

Huth, Iaith? London.

\
il
a

te
ore
ae
ae 1.
ie 7 :
2 ‘¥
_—— a
- os ee
. ~~. a wa
>
4 -
x © a ne
a
yf. 7
Als f i
Ap : // -
f 7 a
A (y 7
i 4,
if (
t\
—
oy i!
mh Al
Ch / A
Mee

T\
L
ry

|

NN
Vv,
I \

a
97.9;

Quant Journ Mor Sc VA, SONS HE, 3h:

spphc. >

SP. phe

Huth ,Lith® London.

11. (Parsons)

Figs.ltoll, Meduse of LIMNOCOOIUM and supposed hydroids of the same.
rig.12, LIMNOCNIDA from Lake Tanganyika.

LU. 3 1.

Nal 2

Direitst Sarre Soi VoL ON

Froth, ‘Lith® London.

MICRO Y DRA, RY DERI, Potts.

abs 1. By dell:

b peel

me bs

Date Due

—————

ON iNigyy

3 2044 110 319 936