sy"

pepe as
TRB AG:

zat ks fac ey
ot i

Woh ss
z

“

\¥ i]
=

ae, |
Fins

———

THE

TRANSACTIONS

OF

Pat LINNEAN SOCEETY

OF

LON DON.

SECOND SERIES.—VOLUME IX.
ZOOLOGY.

LON D OR

PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET :

SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON HOUSE ;

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

1903-1907.

2

(a,
me
fe ay
ne

bi ee:
t
ee yee
a
4
itn sea
1
Pv al
ta ear 1 ox 20Oy a
one aad
"wales shegw cal: Le
<3
ras OL Te
Lye FT i
ae (Mie h

CONTENTS.

PART I.—Juny, 1903.

I. A Contribution towards our Knowledge of the Morphology of the Owls.—Part II.
OstEoLocy. By W. P. Pycrart, 4.L.8., 7.27.8. (Plates 1&2.) . pages 1-46

PAR humvee toOs:
II. On some Points in the Visceral Anatomy of the Characinide, with an Enquiry into

the Relations of the Ductus Pneumaticus in the Physostomi generally. By
Water S. Rowntret, B.Sc. 0.8. (Plates3&4.). . . . . . . 4f-81

PART III.—Dercemper, 1903.

Til. On the Evolution of the Australian Marsupialia ; with Remarks on the Relationships
of the Marsupials in general. By B. Anruur Benstey, B.A. (Tor.), Ph.D.
(Col.), University of Toronto, Canada. (Communicated by Prof. G. B. Howes,
DSc, DED PHBA, Se6.0.S.) (Plates 5-7.) . 2. . . « 4 4. 88-217

PART IV.—Fesrvary, 1904.

IV. The Labial and Maxillary Palpi in Diptera. By Wautrr Wescuk, F.RAILS.
(Communicated by GrorcE Masser, F.Z.8.) (Plates 8-10.) . . . . 219-230

PART V.—Junz, 1904.

V. On the Anatomy and Development of Comys infelix, “mbleton, a Hymenopterous
Parasite of Lecanium hemisphericum. By Auice L. Empieton, B.Sc., 1851
Exhibition Science Research Scholar ; Associate of the University of Wales
(Cardiff College). (Communicated by Dr. Daviv SuHarp, F.R.S., FL.S.)
BRIeCSOInete a ieee wl es a wl Cs CR

PART VI.—Juty, 1904.

VI. Littoral Polycheta from the Cape of Good Hope. By Arruur Witey, D.Sc.,
F.R.S., Colombo Museum, Ceylon. (Communicated by Dr. W. G. Riprwoon,
Mulee(cister lod 14) . 2. 6g . OY AND. | OM 205-268

2198

Vili.

VIII.

IX.

al:

Lal:

Lay aa

PART VII.—NovempBer, 1904.
On the Evolution of Topographical Relations among the Docoglossa. By H. J.
Fievre, D.Sc., Fellow of the University of Wales. (Communicated by Professor
W. A. Herpman, F.2.S., F.L.S.) (Plates 15-17.) . . . . . pages 269-290

PART VIII.—Decemper, 1904.

On some Species of the Genus Palemon, Fabr., from Tahiti, Shanghai,
New Guinea, and West Africa. By De. J. G. DE May, of Ierseke (Holland).
(Communicated by the Rev. T. R. R. Sressrne, I A., FLR.S., Sec.LSa
(Plates 18-20.) . <=) (2 (eRe ae teeter ere le

PART IX.—Juty, 1905.
Observations on some undescribed or little-known Species of Hemiptera-Homoptera
of the Family Membracide. By G. Bowpier Bucxroy, F.RS, PLS.
(Plates @1' W229i wre) WR ea TR Sr

PART X.—Juvty, 1906.
The Genitalia of both the Sexes in Diptera, and their Relation to the Armature
of the Mouth. By Waurer Wescot, F.R.WS. (Communicated by Joun
Hopkinson, #.0.8.) (Plates 28-80.). . . - «. . +... 2» = Sud=888

PART XI.—Marcg, 1907.
On a Collection of Crustacea, Decapoda and Stomatopoda, chiefly from the Inland
Sea of Japan; with Descriptions of New Species. By Dr. J. G. DE Man, of
Terseke (Holland). (Communicated by the Rev. T. R. R. Stessine, W.A., FL. BS.,
F.D.8.) (Plates 81-33.) ... si. ae 0

PART XII.—Juty, 1907.

. On Cercococecus eremobius, gen. et sp. nov., an Aberrant Form of Coccide. By

Huen Scorr, B.A. (Cantab.). (Communicated by J. J. Laster, W.A., F.R.S.,
FDS) (Plate S4:)-.. . . . cil eee erinAR UNEnnS et ole Winer eet

PART XIII.—Aveust, 1907.

Observations on Australasian Polyclads. By Professor W. A. HAswEt1, J/.A.,
D.S¢., F.BS., Fa. (Plates 35-37.) 2. oe

PART XIV.—Octoser, 1907.

Titlepage, Contents,and Indet . . . . « « » « = ROE |. ) . AS /—eee

TRANSACTIONS

OF

I. A Contribution towards our Knowledge of the Morphology of the Owls. |
Part II.—Ostrotoey. By W. P. Pycrart, 4.L.8., F.Z.S.

(Plates 1 & 2.)

Read 19th June, 1902.

Con TENTs.

Page | , Page
Hpesltntroduchory . -\cyeiysmisl-yelere-t-verierisi> << »\< ae Wl The PelviciGirdle.:\ Goctoaceeeteeice a 30
er vhe Skull of the Adults. ss 0. ceric 2 ee we. he Pectoral impr? 2,2. .02 sacteceirereeie re 3l
TIT. The Skull of the Nestling.............. 15 | Exe Phe Pelvie Thimb)- 3 :< wkies i arcreeere terete 33

IV. The Vertebral Column, with remarks on X. Observations on Generic and Specific
the Excalation of Vertebre .......... 21 Characterss ) tics octeray ary ten daw a gaeee 35
Nee her Rabs 2 Ss. ai cteitlauterte fuses be ae 26 SREP MOINMALY, hx ia,t o'sc1s cranks ventana etme te 38
VI. The Pectoral Girdle and Sternum........ 27 XII. Key to the Families and Species ........ 39

I. Inrropucrory.

THE present paper is the result of a careful study of the skeletons of the SrrigEs in
the Collection of the British Museum (Natural History), Although probably one of the
largest collections extant, unfortunately there are many gaps yet to be filled, and
consequently I have been unable to make my work as complete as could be wished.
A large number of genera are still wanting, among the most important of which are :—-

Scotopelia. Gisella.
Sceloglaux. Micropallas.
Scotiapex. Heliodilus.

Besides, nestling skeletons of all ages are greatly needed. Many gaps in this direction

could be filled, 1 am sure, by members of this Society, who, residing in the country,

must have many opportunities of procuring nestlings of the Long- and Short-eared Owls,

Tawny and Barn-Owls. Knowing the persecution which besets these most useful and
SECOND SERIES.—ZOOLOGY, VOL. IX. 1

2 MR. W. P. PYCRAFT ON THE

harmless birds at the hands of ignorant and thoughtless people, it is with the greatest
reluctance that I make this request; yet in the interest of Science, I feel justified
in doing so. Those who may feel as much reluctance in helping to furnish these
desiderata as I feel in making the request therefor, will, I hope, find consolation in the
reflection that my demands entail upon the species mentioned a tax which must be
quite imperceptible. Say thirty young, all told, drawn from an area probably of
hundreds of miles, and spread over a period of several years perhaps, can certainly not
seriously injure the species called upon to furnish these victims! Of course, it would be
well, if possible, to procure nestlings of various ages of at least one species in every
genus, or at any rate of all the more important genera: The skulls of the nestling
Nyctala and Photodilus, for example, would be extremely valuable.

For the only skeleton of Photodilus which the Museum yet possesses we are indebted
to Dr. Charles Hose, resident magistrate of Borneo. It has only recently come to
hand, so recently that at the moment of writing this, the skull only has been
prepared ; the rest of the body awaits dissection before the skeleton of the trunk can be
set up.

The nestling skeletons of Syrniwm we owe to Mr. W. Storrs Fox, who spared himself
no pains to procure the young at the ages I requested. The nestling skeletons of Scops
and Speotyto have been prepared from downy young sent me by my late and ever
lamented friend, Daniel Meinertzhagen, who looked forward with keen delight to the
prospect of making drawings of the skeletons thereof for me when this should be written.
I regret, as we all must, that he was not spared to fulfil his hopes, for his loss to the
ranks of Ornithology is very great.

Il. THe SKULE oF THE ADULT.

The skull of the Striges is desmognathous and holorhinal, and bears a strong superficial
resemblance to that of the Falconiformes, as well as a general likeness to the skulls of
the Caprimulgi.

It may be distinguished, however, from both Falconiform and Caprimulgine skulls by
the extreme shortness of the parasphenoidal rostrum, the thick spongy lachrymal, which
lacks a supraorbital process, the large and functional basipterygoid processes (not
universal in the Falconiformes or Caprimulgi), and by the fact that the palatines are
separated one from another posteriorly by the vomer. Other distinctive characters will
be enumerated as they arise in the course of the following remarks.

The Occipital Region.—The occipital condyle is sessile, and the plane of the occipital
foramen looks directly downwards, forming only a very oblique angle with the basi-
cranial axis. The supra-foraminal border is extremely thin, and passes on either side into
a swollen ridge doing duty for the base of a paroccipital process. A barely perceptible
cerebellar prominence forms a broad occipital crest passing upwards into the lambdoidal
ridge which, in the Bubonine, may almost be said to terminate at the apex of the
temporal fossa. In some species of Budo (e. g. B. magellanica) the crest is folded back
upon itself into a narrow loop, the lower limb of which may be traced outwards and

MORPHOLOGY OF THE OWLS. 3

downwards along the free border of the tympanic wing of the exoccipital. The
lambdoidal ridge can best be traced in the Asionidee and Strigidx, wherein temporal fossze
are wanting.

The paroccipital processes, so conspicuous a feature in Rhea, for example, and slightly
developed in the Falconiformes, are practically wanting in the Striges, the exoccipital
tympanic wings from which these processes are developed being very thin, and curving
forwards to complete the recessus tympanicus anterior.

The Roof of the Cranium.—In the Asionidee (including Photodilus) the interorbital
region stands out in strong contrast to the portion forming the cerebral roof, by reason of
the narrowness of the former and the great width of the latter, which often nearly equals
the whole length of the skull. The free edge of this interorbital region is generally much
thickened over the anterior region of the orbit, and furthermore this thickened area is
produced caudad into a pair of supraorbital spinous processes, recalling the supraorbital
process of the lachrymal. That portion of the interorbital region lying between the
thickened lateral borders just referred to, is also much thickened by pneumatic tissue,
developed, apparently, by the inflation of anterior ends of the frontal and the immediately
overlapping portion of the nasals. The result, in the adult skull, is peculiar, in that the
base of the culmen of the beak—formed by the nasals and premaxillary—looks as
though it had been wedged into a spongy mound of bone. In this way an imperfect
nasal hinge has been formed. The cerebral portion of the roof, as we have already
remarked, is of great width, expanding behind the orbits into a broad tongue-shaped
postorbital process, which may extend downwards to within a short distance of the
quadrato-jugal bar. The outline of the cerebral region, traced from the supra-foraminal
border upwards and forwards into the interorbital region, may be fairly described as
>-shaped, so that the crown of the head is nearly flat. Very shallow tongue-shaped
temporal depressions are always traceable in the Bubonine, but are widely separated
one from another in the median line.

The roof of the skull in the Strigidz differs conspicuously from that of the Asionidee
by reason, partly, of the excessive development of spongy, pneumatic tissue; and
partly, because the width across the skull at the postorbital processes is much less.
The width across the interorbital region is relatively somewhat greater than in the
Asionidz; the relations between the beak and the skull also resemble those of the
Asionidve, the former having the appearance of being wedged into the latter when in a
plastic state, but the resulting nasal hinge is still more imperfect. The cerebral portion
of the dome rises up into a blunt cone culminating at the interorbital region, and
marked along the middle line by a deep furrow which extends forwards to the base of
the beak, and laterally by a shallow depression above the orbits. This depression is
crescentic in shape, and extends from the base of the postorbital process upwards and
forwards, terminating above the middle of the orbit. A row of pits along the are of this
depression marks the position of the larger disc-feathers. Partly on account of the
relatively smaller brain-case, which causes the occipital foramen to appear to lie further
backwards, and partly because of the great development of spongy tissue, the outline of
the cerebral portion of the skull differs markedly from that of the Asionidze. The post-

1*

4 MR. W. P. PYCRAFT ON THE

orbital processes stand out conspicuously from the skull, and extend downwards so as
nearly to touch the quadrato-jugal bar.

Seen in section, the cranial roof presents some interesting features. As already stated,
this region of the skull is considerably thickened by the development of spongy tissue
which attains its maximum over the interorbital region. In Asio and Budo, for example,
the tabula externa and tabula vitrea are of extreme thinness, and separated by a mass of
diploé to which has been added, in the interorbital region, a mass of pneumatic tissue.
Between the layer of diploé, and, anteriorly, between the diploé and the pneumatic tissue,
there runs, from the parietal region forwards, a distinct pneumatic canal, which,
however, is lost in the region above the olfactory fossa. The great development of the
pneumatic tissue has caused the anterior end of the frontal to fold over upon itself so as
to overlap the base of the nasals and the proximal end of the nasal process of the
premaxilla. The pneumatic tissue, it is to be noticed, does not extend downwards
between the interorbital septum, save only along its anterior border where it runs
downwards and backwards to terminate at the great air-sinus, the anterior tympanic
recess.

Syrnium differs conspicuously from Bubo in that the diploé and pneumatic tissue
alike present a regular cellular arrangement of superimposed layers—a series of
separate plates of bone supported by delicate bony pillars. This laminate arrangement,
however, terminates near the anterior end of the frontal, giving place to the normal
irregular spongy tissue.

Stria agrees with Asio and Bubo in that the diploé and pneumatic tissue are of the
ordinary spongy type, having no regular arrangement; but the skull differs both from
Asio and Syrnium in that the spongy mass descends downwards from the frontal to
inflate the whole of the interorbital septum. To such an extent has this development
proceeded, that the groove into which the olfactory nerve passes in its passage from the
skull along the roof of the orbit to the olfactory cavity has now become entirely
surrounded by the pneumatic tissue, the tabula externa having been driven outwards,
leaving the groove, now converted into a tube, completely invested. In Eurystopus,
one of the Caprimulgi, we have a stage halfway between the normal arrangement and
that in Strix, the nerve running in a deep trough of pneumatic tissue the edges of which
nearly meet to form a tube as in Strix.

The Base of the Skull.—The basitemporal platform is roughly triangular in form, having
a very broad base, and its apex lying in the middle line between the basipterygoid
processes. Above the apex lie the Eustachian apertures. The anterior lateral borders
of the plate, save at the apex, have fused partly with the base of the parasphenoidal
rostrum and partly, on either side of the rostrum, with the wall of the recessus
tympanicus anterior, so that the Eustachian grooves are converted into canals, which
open into the floor of the mouth of the recess. The paroccipital notch, which is deep in
many Falconiformes, in the Striges is wanting, having been obliterated by the great
development of the meatus externus and the recessus tympanicus anterior. It is to this
same great development of the tympanic cavity that the loss of distinct paroccipital
processes is due, since the bony tissue used in their formation appears to have been

MORPHOLOGY OF THE OWLS. 5

utilized to complete the mouth of the meatus by joing with the tympanic recess
and closing the notch in question.

The position of the obliterated paroccipital notch is indicated by a foramen, for the
passage of a branch of the sinus venosus, which can be traced upwards and forwards to
its mouth on the inner side of the articulation for the quadrate. Midway between this
foramen and the occipital condyle lies the vagus foramen, and mesiad of the vagus lie
the condyloid foramina.

The basipterygoid processes are represented by two very short stout pillars rising from
the base of the parasphenoidal rostrum, which is of considerable breadth. In the
Strigidze these processes are much reduced in length, so as to be little more than
prominences.

The pre-condylar fossa is well marked.

The parasphenoidal rostrum of the Striges presents some interesting modifications.
Apart from its great breadth at the base, which appears to be characteristic of the
Owls, the parasphenoid, in the majority of the Asionidee, has undergone a very con-
siderable shortening, terminating at the level of the palato-pterygoid articulation, instead
of running forward to terminate just above the hinder end of the maxillo-palatine
processes, as in the Falconiformes, for example. This abbreviation of the rostrum is also
noticeable in the Caprimulgi.

Among the Owls, however, Surnia has a relatively long rostrum, since it terminates
midway between the pterygo-palatine articulation and the hinder border of the maxillo-
palatine processes. The Strigidze appear also to have an abbreviated rostrum, but until
I have an opportunity of examining nestlings I cannot settle this point, owing to the
extraordinary inflation of the tissues of this part of the skull.

The Lateral Surface of the Cranium.—The tympanic cavity is unusually large in the
Striges, and presents some interesting modifications. Studied in its least specialized
condition, the increased size of this cavity is found to be due to the great development
of the tympanic wing of the exoccipital and the external border of the squamosal,
the outgrowths from these two bones forming a conspicuous outstanding wing of
varying shape, according to the genera examined. The squamosal portion of the wing it
is to be noted is borne by the region corresponding to the squamosal prominence of the
Falconiform skull, a prominence generally more conspicuous among the Owls, in the
nestling skull; its presence in the adult skull is marked by the great thickening of
the skull-walls by the development of pneumatic tissue, a thickening which causes the
quadrate in the adult Striges to appear to be set much nearer the middle line than in the
nestling.

Not only is the tympanic cavity proper actually increased in size by the development
of the lateral wings above described, but a more or less perfect chamber is added thereto
by the outward extension of that portion of the wing which arises from the region of the
squamosal prominence. This additional chamber is especially well developed in a few
genera to be described presently.

It may be studied in its simplest form perhaps in Syrniwm aluco or Surnia funerea.
Here it forms a vestibule, opening forwards into a deep trough formed by the large post-

6 MR. W. P. PYCRAFT ON THE

orbital process, a trough which is, however, non-existent in the living bird, inasmuch
as it is filled up by the temporal muscles. During life, then, the tympanic cavity forms
a spacious chamber, which, traced inwards, leads to other and smaller chambers. The
latter, three in number, are approached through a common and well-defined elliptical
mouth, whose long axis slopes obliquely backwards. The hinder border of this mouth is
formed by the free edge of the tympanic wing of the exoccipital, and the free edge of
the inferior border of the squamosal, but this can only be made out in the nestling: the
anterior border is formed by the processus articularis squamosi. This last, arising imme-
diately behind the squamosal head of the quadrate, is continued, not infrequently,
downwards and backwards to meet the inferior external angle of the exoccipital wing.
In the Falconiformes this process is represented by a downwardly projecting spine,
serving, as in the Owls, to hold the quadrate in position.

Within this mouth we distinguish two apertures, one above the other, and divided by
a bony bar formed by the otic head of the quadrate. The upper aperture is the mouth
of the recessus tympanicus superior, a large pneumatic cavity hollowed out between the
pro-otic and the squamosal and further extended by the absorption of the spongy tissue
of the parietal, leaving only the inner wall of the bone to protect the brain. The
position of the aperture of this recess is interesting, inasmuch as it varies considerably in
different groups. Thusin the Steganopodes, for example, it opens in front of the articular
surface for the quadrate, whilst in the Falconiformes, e. g. Circaétus, its aperture divides
the articular surface into two portions, the quadrate thus serving as a bar across the
aperture. Furthermore, the position of the aperture in the Steganopodes is, as in Oircaétus,
occupied by adeep fossa. In the Striges, asin the Falconiformes, the mouth of the recess
divides the articular surface for the quadrate into two portions, but having apparently
shifted further backwards, the aperture lies almost wholly behind the quadrate. The
Caprimulgi agree with the Owls in this respect.

The lower of the two apertures now under discussion is often smaller than the upper,
and lying well within its mouth will be found a vertical bony column dividing two
tunnel-like openings, one in front of the other. The anterior opening leads to the
recessus tympanicus anterior, which is very spacious, running forwards and inwards
to meet its fellow of the opposite side in the middle line. In the formation of the
anterior end of this space the whole of the spongy tissue of the basisphenoid and para-
sphenoidal rostrum has been absorbed, so that the pituitary pit, and the tunnels for
the internal carotids which open into it, are completely isolated, the pit appearing as a
delicate thin-walled cup, supplied from below by two converging tubes. The hinder of the
two openings of this lower tympanic aperture lodges the foramen ovale and the foramen
rotundum.

The great size of the aperture of the recessus tympanicus superior, as well as the
surprising size of the cavity itself, can be fully realized if the quadrate be removed.
The aperture is then seen to lie between the squamosal and otic articular surfaces for
the quadrate, which are on this account widely separated one from another.

The tympanic cavity in the typical, less specialized skulls of the Asionide is over-
shadowed by a conspicuous squamosal prominence, e. g. Bubo, from which arises, be

—— ee oe

MORPHOLOGY OF THE OWLS. 7

it noted, the squamosal wing of the tympanic cavity just described. Above this
prominence lies a broad, shallow, linguiform temporal fossa, which, in some forms, to be
presently described, may be absent.

Regarding the tympanic cavity of Syrniwm as representing, at least approximately,
the archecentric (¢f. P. C. Mitchell, Trans. Linn. Soc., ser. 2, Zool. viii. (1901)
p- 181) condition, let us turn now to a consideration of the more important apocen-
tricities which may be traced therefrom. These are of two kinds: (a#) symmetrical,
and (0) asymmetrical. Amongst the symmetrical forms the departure from the arche-
centric type begins by the formation of large bulle derived by an increase in the size of
the squamosal wing which forms the posterior tympanic wall, and which may, as in
Scops asio, S. rutilus, and 8. semitorques, fuse by its superior external angle immediately
behind the base of the postorbital process, thus bridging the temporal fossa. By this
inflation of the tympanic wing a very spacious cavern is formed, bounded posteriorly
and superiorly by the squamosal and exoccipital wings, anteriorly by the postorbital
process. It is divided into two more or less well-marked chambers, an outer and an
inner, one leading directly into the other, the division being bounded by the quadrate.
In Gymnasio and Speotyto, by the way, the squamosal wing joins the postorbital process
directly, and thus not only bridges the temporal fossa but also roofs in the tympanic
cavity. The size of this wing not being increased, however, the resulting outer chamber
is very small.

The form of the tympanic cavity in Asio is still further modified. It appears to be a
further specialization of the type seen in Photodilus and Strix (seep. 9). The squamosal
and exoccipital moieties of the tympanic wing are here combined to form an outstanding
wing of bone extending upwards and forwards to terminate above the base of the post-
orbital process, from which it is divided, however, by a low, saddle-shaped ridge. The
exoccipital moiety of the wing curves downwards and forwards so as to form a spacious
antechamber leading directly to the apertures of the recessus tympanicus superior and the
meatus externus. The tympanic cavity in front is bounded by the postorbital process, of
which more anon. Above, the cavity is bounded only by a low ridge connecting the
angles of the postorbital and squamosal wings of the tympanic cavity just described.
The deep cavity enclosed by these wings corresponds to the ‘posterior division of the
cavernum” of the first section of this memoir. For brevity’s sake it may be called the
post-cavernum. The post-cavernum of Asio differs from that of Speotyto, for example,
mainly in its greater upward extent, in its relatively greater width, and in that it is less
perfectly closed above. But whilst the tympanic wing of Speotyto forms a conspicuous
laterally expanded bulla, the wing in Asto forms rather a vertical flattened plate, the
inferior portion of which, however, is bullate. This change in the form of the tympanic
wing has resulted in an interesting modification—to wit, the suppression of the temporal
fossee—and fresh attachments for the first portion of the ¢emporalis muscle. The
temporal fossa is in consequence obliterated in Asio.

In Speotyto, Scops asio, Scops rutilus, and Scops semitorques, it must be remembered, the
tympanic wing has also extended upwards; but in these cases the temporal fossa is not
suppressed, the wing being perforated to allow of the passage of the muscle.

8 MR. W. P. PYCRAFT ON THE

We must turn now to a consideration of the asymmetrical tympanic cavity as exhibited
in the skull of Wyctala Tengmalmi. This appears to be a modification of the type seen in
Speotyto, the squamoso-occipital tympanic wings forming large bulle.

On the left side of the skull (Pl. 2. fig. 1 a) the tympanic wing has grown upwards and
forwards to join the postorbital process, bridging the temporal fossa as in Speotyto,
But with this difference, the fossa is reduced to a mere vestige in the shape of a narrow
groove deep enough perhaps to lay a horse-hair in; this groove passes between the
junction of the tympanic wing and the postorbital process through a minute hole into
the tympanic cavity. I have not been able to dissect out the muscles of the head, but
when this is done it will probably be found that not more than a vestige of the first
portion of the temporal muscle remains ; possibly even this has disappeared. ‘The cavity
enclosed by the squamoso-parietal wing and the postorbital process corresponds to the post-
cavernum of Asio. It is considerably larger than the corresponding cavity in Speotyto,
and curves downwards and forwards, so as to form at the same time a shield for, and a
spacious antechamber leading to, the apertures of the recessus tympanicus superior and
the meatus externus. In the nature and function of the exoccipital moiety of the
tympanic wing, Wyctala, it should be remarked, agrees also with Asio and Speotyto, for
example. But the tympanic wing of the left side in Myctala differs from that of any
other of the forms with which it has been compared in the development of a large
tongue-shaped plate of bone which projects from the middle of the tympanic wing,
forwards and downwards, to terminate just below and outside the articulation of the
quadrato-jugal bar with the quadrate.

The tympanic cavity of the right side of the skull in Nyctala differs conspicuously
from the left, in that the squamosal portion of the tympanic wing extends upwards to
within a short distance of the crown of the head, and lies far behind the base of the
postorbital process. Near the middle, the free border of this wing, as on the left side, is
drawn out into a tongue-shaped process, which curves inwards at its tip to join the
inferior end of the postorbital process. Thus, on this side the articulation of the
quadrato-jugal with the quadrate is fully exposed.

If the post-cavernum of the right and left sides be compared, it will be found that that
of the right side has a very large oval aperture, the inferior border of the rim of which
lies considerably above the level of the orbital process of the quadrate. The cavity itself
may be likened to a deep pocket, which runs directly backward. The aperture of the
left side is incomplete, a large gap being present in its lower border, owing to the
failure of the tongue-shaped process of the tympanic wing to reach the postorbital
process. Furthermore, the aperture is narrower, crescentic in shape, and terminates
superiorly at the level of a line passing forwards from the upper limb of the crescent to
the base of the beak. Thus its upward extent is very considerably less than on the right
side. A comparison of the figures (PI. 2. fig. 1, 1 @) will make this plain.

The mesial wall of the post-cavernum of the right side, it may be mentioned, is marked
by a shallow groove which, traced upwards, is found to lead to a minute aperture,
corresponding to the point of union of the tympanic wing with the base of the post-
orbital process of the left side. But here, the tympanic wing having shifted backwards,

MORPHOLOGY OF THE OWLS. 9

the aperture in question lies considerably behind this process. From the aperture there
runs, backwards and downwards, a vestigial temporal fossa, also corresponding to that of
the left side.

The tympanic cavity of Photodilus and the Strigidz is of the same type as that found
in Asio. Indeed this cavity in the latter appears to be a further modification of that
seen in the two subfamilies above mentioned.

That of Strix is the least specialized of all; that portion of the cavity which lies above
the head of the quadrate, and between the squamosal wing and postorbital process, is of
comparatively slight extent, consisting merely of a small depression. In Asio, it will be
remembered, it forms a large, oblong, and concave fossa. Photodilus is intermediate in
some respects between the two. In the skull of this form the fossa is nearly as long as
in Asio, but very narrow from side to side. In the great size of the postorbital process
Photodilus differs both from Asto and Strix. In the development of the squamosal
wing it is also peculiar. This is larger than in Strix, smaller than in Asio. Asio, Strix,
and Photodilus all agree in the absence of the horizontal temporal fossa so conspicuous,
for example, in Budo.

The orbits in the Asionide are large, and their capacity is still further increased
by the lateral expansion of the postorbital processes. They are overarched by the
frontals, and bounded in front by the lachrymals, whilst the antorbital processes
(prefrontals) and the orbital process of the quadrates form an imperfect floor. Feeble
supra-orbital processes are commonly, but not invariably, present. These, it is to be
noted, are furnished by the frontals themselves, and not, as in the Falconiformes, by the
lachrymals. The interorbital septum forming the partition wall between the two orbits
is imperforate, save in Stwnia, Speotyto, and some species of Scops. The orbito-
sphenoid is completely ossified.

The optic foramina are widely separated one from the other, and lie on a level with a
line drawn forward from the bottom of the rim of the recessus tympanicus superior.

The trigeminal foramina open above the roof of the vecessus tympanicus anterior on
a level with a line drawn downwards and forwards from the anterior zygomatic process
to the basipterygoids.

Immediately external and posterior to the optic foramen lie the foramina of the
third, fourth, and sixth nerves.

The orbit in the Strigidze differs conspicuously from that of the Asionidee and Photo-
dilide in that it is extremely small. Its horizontal axis scarcely, or not at all, exceeds
that of the horizontal axis of the lachrymal, though it must be remarked this last-named
bone is relatively very much larger than that of the two suborders with which it is
contrasted. Even, however, where the lachrymal of the two suborders is relatively as large
as that of the Strigid, the greater size of the orbit with regard to the lachrymal would
still remain the same.

The Strigide are furthermore peculiar in having the interorbital septum, which is very
short antero-posteriorly, enormously inflated by the development of spongy tissue.

The Hthmoidal Region.—The mesethmoid, which by its posterior extension forms the
interorbital septum, agrees more nearly with the mesethmoid of the Caprimulgi than

SECOND SERIES.—ZOOLOGY, VOL. IX. 2

10 MR. W. P. PYCRAFT ON THE

with that of the Falconiformes, being relatively very short antero-posteriorly, and in
some genera enormously thickened by the development of spongy tissue. It is relatively
largest in Swrnia and Speotyto: a fact which can best be realized by comparing the
mesethmoid of either of these genera with that of, say, Bubo, Asio, or Syrnium. In the
last-named genera the anterior free border is sharply truncated, sloping obliquely back-
wards to pass into a remarkably abbreviated parasphenoidal rostrum (p. 5) immediately
above the pterygoidal articulation. In Swrnia the parasphenoid extends considerably
further forwards, passing insensibly into the anterior border of the mesethmoid at a point
corresponding with the level of a line passing behind the lachrymals. The horizontal
plate of the supero-anterior angle of the mesethmoid is not much developed, in conse-
quence of the slight development of the olfactory cavity. The antorbital plate
(prefrontal) in Swrnia and Speotyto is short and strap-shaped and projects from the
inferior angle of the mesethmoid. In Budo it appears to arise rather below the middle
of the anterior border, an appearance which is due to the fact that the anterior border of
the mesethmoid by its greater obliquity projects further downwards. In Aso the
antorbital plates are very long, nearly touching the lachrymals and quadrato-jugal bar.

The Strigidze differ conspicuously from the Asionidze in the form of the mesethmoid.
One of its most conspicuous features in the adult skull is its enormous thickness, due to
the great development of spongy, pneumatic tissue. The interorbital region, in addition
to its great thickness, is relatively shorter than in the Asionidee, whilst. the olfactory
region is relatively longer, and is furthermore peculiar in that its inferior border lies
almost horizontally rather than obliquely. The antorbital plate is also much swollen,
so much so as to become club-shaped. The interorbital septa, however, of Aséo and
Photodilus are exceptional, being swollen by pneumatic tissue as in Strix, though to a
slighter extent.

The olfactory cavity is fairly spacious, but contains no turbinal ossifications. Bounded
posteriorly by the ali-ethmoidal antorbital plates, it passes forwards into the anterior narial
aperture, which is divided from its fellow of the opposite side by an imperforate nasal
septum. A certain amount of ossification has taken place in the ali-ethmoidal walls of
the anterior region of the olfactory chamber, the nature and extent of which appear to
be best displayed in Swrnia, where the cartilaginous portions of the wall, removed by
maceration, have left a free edge of bone lying immediately within the anterior narial
aperture. Traced inwards, this ossification is found to have formed a cup-shaped cavity
forming the floor of this region of the chamber, and rising upwards, passes into the
nasal septum. Posteriorly the wall of the cup is broken down, thus placing the anterior
and posterior moieties of the chamber in communication with one another. The external
alinasal wall is somewhat more extensively ossified in 4séo and Bubo, thus rendering the
contour of the nasal aperture formed by the premaxillary limb of the nasal somewhat
irregular.

The floor of the anterior nasal cavity undergoes certain limited changes. Thus in Ninox
it is arched instead of hollow, lying above the level of the rim of the narial aperture: in
Bubo it is flat; in Strix it is perforated. The septwm nasi is separated from the mes-
ethmoid by the cranio-facial fissure, which is deep. Furthermore, in some Owls, as in

_—_— ew oe
.

MORPHOLOGY OF THE OWLS. 141:

Budo, it is perforated near its superior border, so as to place the hinder division of the
olfactory cavity in communication with its fellow, at least in the dried skull. In Stria
the perforation, by its extension backwards, is converted into a deep notch. It is
interesting to note that owing to the ect-ethmoidal ossifications above described, the
nasal septum is not visible from the ventral aspect of the palate, as in the schizognathous
palates of certain Falconiformes. That is to say, in the middle of the anterior palatal
vacuity of the schizognathous Falconiform palate the nasal septum is plainly visible,
the roof of the cavity beimg formed by the premaxilla; whilst in the Strigine palate the
roof of this cavity is formed almost entirely by the ossified ali-ethmoids, only the extreme
anterior end of this vaulted chamber being formed by the premaxilla.

The lachrymal differs conspicuously from that of the Falconiformes, and agrees closely
with that of the Caprimulgi. As in the last-mentioned group, there is no supraorbital
process, and the descending limb, which extends downwards to the quadrato-jugal bar, is
greatly inflated by spongy tissue, and is attached to the under surface of the fronto-nasal
region of the skull, instead of the external border thereof as in the Falconiformes. It
varies somewhat in form in the different genera, but typically, in the Asionidee, it may
be described as columnar, and marked by a deep groove on its external face, near its upper
third, the groove passing forward into the lachrymo-nasal fossa. The latter, by the way,
is extremely small, having been practically obliterated by the great development of the
maxillo-palatine processes, which rise upwards to cover almost the entire anterior border
of the lachrymal. In Budo it is almost crescentic, the horns of the crescent being
directed outwards. In some species of Scops it is very small, tapering from above
downwards. In Syrniuwm, Surnia, and Asio it is perhaps most nearly columnar, and in
the two latter the lachrymo-nasal groove is especially deep.

The lachrymal is largest in Str7x, where it is subconical in form, with the apex pointing
forwards into the lachrymo-nasal fossa, and the base hollowed to accommodate the eye.
It is grooved across the middle of its outer surface, deeply in some, slightly in others,
according to the species. Again, in some species, the lachrymal extends so far
forwards as to leave only a small hole representing the lachrymo-nasal fossa, whilst in
others, e.g. S. poensis, a comparatively large fossa is left, which is open below down to
the quadrato-jugal bar. These differences in the form of the lachrymal, it should be
noted, are correlated with others in other parts of the skull.

The Cranial Cavity.—The metencephalic fossa in the Asionidze very closely resembles
that of the Falconiformes, being both moderately wide and fairly deep. The vagus
foramen lies very near the exoccipital border. The internal auditory meatus is deep,
and separated from the vagus foramen by a swollen ridge. About midway along the
anterior lateral border of the fossa, mesiad of the mouth of the trigeminal foramen, lies
the small abducent foramen. The oculo-motor lies bebind and below the optic foramen
on a level with the external angle of the sella turcica. ,

In the Strigidee the vagus foramen is reniform, and the internal auditory meatus is in
a very shallow pit into which the foramina for the facial and auditory nerves open. The
foramen for the oculo-motor is minute, and lies on the extreme external angle of
the sella turcica between the optic and trigeminal foramina: passing through the wall,

9*

12 MR. W. P. PYCRAFT ON THE

it emerges immediately beneath the foramen for the first branch of the trigeminal. The
foramina for the 1st and 2nd-3rd branches of the trigeminal in the Strigidz, by the way,
pass out from a horizontal slit-like foramen on the superior lateral border of the fossa.
The abducent foramen, like the oculo-motor, is minute, and lies mesiad and below the
inner angle of the fossa for the trigeminal branches 1 and 2-3. It opens into a fossa,
immediately behind the optic foramen, which contains besides the apertures of the first
portion of the trigeminal and the oculo-motor. Thus, as in the Asionide, all three open
into a common pit.

The cerebellar fossa presents no characters of any importance. It appears to be
slightly larger, relatively, in the Strigidze. The floccular fossa is well-marked.

The mesencephalic fossa appears to be relatively largest in Syrniwm, and smallest
in Striz. In Syrnium and Bubo the boundaries of the fossa are extremely well-
defined.

The pituitary fossa is relatively much wider in transverse diameter than in the
Falconiformes, and differs furthermore therefrom in that its anterior wall rises upwards
into a steep wall to pass into a wide but ill-developed optic platform ( prepitwitary ridge).
The preoptic ridge is rounded, swollen and prominent in the Asionidee, ledge-shaped in
the Strigidee, and less prominent. At its outer angle this ridge plunges sharply down-
wards to join the tentorial ridge.

The optic foramina are widely separated one from another, owing to the enormous size of
the anterior tympanic recess, the anterior end of this great pneumatic chamber lying
immediately under these foramina.

The cerebral fosse have encroached upon the cerebellar fossa, appreciably reducing
the size of its upper portion. The tentorial ridge, where it leaves the preoptic platform,
plunges sharply downwards to the level of or below the sella turcica, when it sweeps
downwards and upwards over the pro-otic, meeting its fellow of the opposite side
immediately over the cerebellar fossa. From the point of this union there runs forwards
in the middle line a prominent swollen ridge, the bony falx, which is continued forward
to form the roof of the olfactory fossa. In the Strigidz the middle region of this ridge is
produced into a sharp edge. A more or less prominent ridge corresponding to the
Sylvian fissure in the cerebrum is present in all the Striges.

The olfactory fosse appear to be largest in Bubo, and are never very large.

The Premaxilla.

The premaxilla of the Striges is almost indistinguishable from that of the Falconi-
formes. It appears, indeed, to differ therefrom only in the greater share which it takes
in the formation of the anterior narial aperture, in the nature of the anterior palatine
vacuity (which in the Owls is roofed by the ossified alinasal floor, thus concealing the
nasal septum), and in that the inner surface of the curved tip is not provided with a
median ridge. Seen from the ventral surface, this bone recalls that of the Cathartee.
But in the latter both nasal septum and ect-ethmoidal ossifications are wanting, thus
revealing the nasal process of the premaxilla, bounded by very long anterior narial
apertures—features which in the Striges are conspicuous by their absence.

MORPHOLOGY OF THE OWLS. 13

The Maxillo-jugal Arch.
The maxilla is indistinguishably fused with the premaxilla, and is indeed entirely
covered by that bone externally.
The mawillo-palatine processes in the Asionidee are of considerable size, highly pneu-
matic, being composed entirely of spongy tissue, and project backwards in the form of

large bulle, rising on each side above the quadrato-jugal bar so as partly to obliterate \.

the lachrymo-nasal fossa, which is reduced in some cases to the vanishing point, in
Gymnasio, for example, only a minute hole being left. This lateral and upward growth
of the process forms in Bubo, Gymnasio, and Ninox a quadrate mass at the infero-
posterior angle of the beak. In Asio, Scops, and Surnia this lateral mass is less
developed, and it is wanting altogether in other forms, such as Swrnia and Speotyto, for
example, where the upward growth of the bullate process may be seen filling up the
lower portion of the lachrymo-nasal cavity. The lachrymal, it should be remarked, fits
very closely on the hinder end of this lateral maxillo-palatine mass. Compared with the
maxillo-palatines of the Falconiformes, it is interesting to notice that the lachrymo-nasal
orifice of the antrum is incomplete above, inasmuch as the inner border of the dorsal
surface does not rise upwards to join the descending process of the nasal. The palatal
vacuity in the floor of the antrum, which lies in the triangle formed by the palatine,
premaxilla, and maxilla (PI. 1. fig. 8), is large, as in the Falconiformes, and this traced
forwards will be found to lead into the spacious anterior palatal vacuity.

The maxillo-palatine processes of Photodilus differ from those of the remaining
Asionidz merely in their somewhat smaller size, their lateral extension being less. That
is to say, they do not extend outwards to fuse with the descending maxillary process of
the nasal. They may best be described perhaps as intermediate in character between
those of the Asionidee and Strigide.

In the Strigide the maxillo-palatine processes are relatively much longer than in the
Bubonide, and differ from them in shape, forming delicate backwardly-directed and some-
what spindle-shaped processes. Viewed laterally, they are seen to lie near the middle
line and to leave a large cavity between themselves and the descending (maxillary)
process of the nasal. The lachrymo-nasal aperture of the antrum is wanting altogether,
but the palatal aperture is large.

The palate of the Owls appears at first sight to be schizognathous; in reality, however,
it is desmognathous, for although the bullate portions of the maxillo-palatines are widely
separated in the middle line, the palate is nevertheless bridged by the ect-ethmoid
ossifications extending from the nasal septum outwards and downwards on either side
to join, posteriorly, the maxilla.

The Vomer, Palatine, Pterygoid, and Quadrate.

The vomer is short and blade-shaped, but generally more or less inflated in appearance,
so much so, in Aséo and Stria, for example, as to become fusiform. In Aséo it is pneumatic.
In Surniaitis vestigial. In Speotyto, Ninox, and some other genera it appears to be wanting.
It is supported by a pair of small spurs projecting inwards from the dorsal border of the

14 MR. W. P. PYCRAFT ON THE

palatines, which effectually prevent the latter from meeting one another in the middle
line, the space between them being equal to the width of the parasphenoidal rostrum.
This peculiar feature appears to be found only among the Owls.

In Photodilus the vomer is extremely reduced, most nearly resembling that of Ketwpa.
In this respect Photodilus differs conspicuously from the Strigidee.

In Strix the vomer, seen from below, appears as a fusiform body, highly pneumatic
and of extreme delicacy, the surface being converted into a delicate filagree work. Its
pointed end runs forward for some considerable distance between the maxillo-palatine
processes. Seen from above, the vomer presents the appearance of the bow-end of a
canoe, the sides of which are kept apart by a most delicate lattice-work of bone.
Posteriorly, the vomer fuses with the quadrato-palatine plate described below.

The palatines take the form of flattened rods, in the Asionidze and Photodilide having
a strong outward curve. In the larger skulls the mesial borders of the posterior ends—
on either side of the vomer—send down a sharp keel. In Swrnia the palatines are
relatively very short and broad.

In the Strigide the palatines are straighter and relatively longer than in the Bubonide,
and, moreover, they are not separated one from another as in the Bubonide.

The separation of the palatines in the Asionidz one from another by the development
of lateral spurs is an extremely interesting feature, and becomes still more so when
contrasted with the conditions which obtain in the Strigidz. Here the palatines, viewed
from above, appear to have fused one with another, posteriorly, and with the vomer,
presenting, in front of the pterygoid articulation, a broad quadrate wall sloping upwards
and forwards, and having a doubly notched free dorsal border.

The pterygoids are long, and, near the middle, send inwards a thick spur to articulate
with the basipterygoid processes. In some forms they are nearly straight, e.g. Bubo,
Scops; in others they are sigmoidally curved, e.g. Surnia, Strix. In all they are blade-
like rather than rod-shaped. They are sharply truncated anteriorly, and articulated with
the extreme postero-external angle of the palatine. Proximally they articulate with the
shaft of the quadrate, the articular end in some species expanding to encroach upon
the orbital process, e.g. Ninox.

It is interesting to note that the angle formed by the pterygoids with the long axis of
the skull isa much wider one in the Striges than in the Falconiformes. This appears to
be due to the very decided antero-posterior shortening of the skull, brought about by the
reduction of the interorbital region. The shortening is even more marked in the para-
sphenoidal rostrum, which in Budo, for example, is extremely reduced. As a consequence
of this abbreviation, the posterior ends of the palatines lie relatively much further back
than in the Falconiformes. This same shortening, as we have elsewhere pointed out, has
also brought the antorbital plate almost directly over the pterygo-palatine articulation.

In the great length of the pterygoids the Striges resemble the Caprimulgi more nearly
than the Falconiformes, the orbital process and squamosal head being subequal in
length.

The guadrate is Falconiform in its general shape. It differs from that of the Falconi-
formes chiefly in the greater length and distinctness of the otic process, and in the more

MORPHOLOGY OF THE OWLS. 15

backward and downward position of this, and in its greater separation from the squamosal
head. Furthermore, the pneumatic apertures in the Striges do not extend below the
base of the otic head, whilst in the Falconiformes they extend downwards to the base of
the orbital process. In the Asionidze the pneumatic apertures are two in number, and
open, one on either side of the base of the otic head, into a deep groove. In the Strigidee
there is but a single aperture; this is large and opens beneath the otic head.

The form of the quadrate in Sériv is markedly different from that of any of the
Asionide, the orbital process being reduced to a short spine, seated rather below
the middle of the body of the quadrate, which is bent sharply backwards forming
a Y-shaped angle with the process in question. Furthermore, the internal mandibular
condyle in the Strigidze is comparatively feebly developed, whilst in the Asionide it is
nearly as large as the external condyle, and in Budo, for example, projects downwards
far below the articulation of the hinder end of the pterygoid.

The squamosal head of the quadrate is bounded in front by a zygomatic process,
which is longest in Asio, and especially so in A. madagascariensis, in which the
maximum length appears to be reached, and behind by the processus articularis
squamosi. ‘This last often extends downwards and inwards to join the pretemporal wing
of the basitemporal.

The Mandible.

The most conspicuous feature of the mandible in the Asionide and Protodilide is the
wide angle formed by the rami, and the great length of the internal angular process.
The angular is sharply truncated. A ramal vacuity is generally present. The only
suture that can be cut is the dentary, and this is often obliterated. The coronoid ends
in a free point partly closing the ramal vacuity.

In the Strigidz the angle of the rami is less open; the coronoid is less degenerate
than in the Asionidee, but has fused more completely with the jaw, reducing the size
of the lateral vacuity.

The Hyoid.
The hyoid resembles that of the Accipitres among the Falconiformes. The basihyal
is triangular in form, deeply notched in front, and has the posterior angle drawn out
into a small pair of cerato-hyals. The basibranchial is more or Jess rod-shaped and

produced backwards between the cerato-branchials. The cerato-branchials are long and
slender, and surmounted by a pair of slender, ossified epi-branchials.

III. Tur SKULL oF THE NESTLING.

The sutures of the skull appear to remain distinct, at least until the young is nearly
half-grown. The changes which take place between the skull of the nestling and that

-of the adult are at once striking and instructive.

It is to be regretted that the Museum Collection contains no nestling skulls of the
Strigidee. The following description is based on skulls of Syrniwm aluco and Speotyto
cunicularia.

16 MR. W. P. PYCRAFT ON THE

The Cartilage-bones.

The basioccipital, seen ventrally, is more or less linguiform in shape, and has a
straight anterior border overlapped by the basitemporal plate. Its lateral borders are
convex, and approach one another towards the middle line, reducing the posterior border
to a small segment forming the centre of the occipital condyle, and forcing its way
between the exoccipitals to bound the occipital foramen. Its lateral boundaries are
formed by the exoccipitals only. Seen dorsally, it is found that its lateral boundaries
are formed by the pro-otic and exoccipital; the former occupying the anterior and
the latter the posterior half of the border.

The exoccipital is large and irregular in shape. Ventrally its mesial border, which is
concave, bounds the basioccipital; its anterior border is convex and overlapped by the
wings of the basitemporal plate, whilst its external lateral border bounds the tympanic
cavity, the posterior half of the border being produced to form the floor of the trumpet-
shaped mouth of that cavity. Its posterior border is more or less triangular in shape,
the outer face of the triangle being bounded by the squamosal and the inner face by
the supraoccipital. Dorsally, the exoccipital is entirely concealed by the pro- and epi-
and opisthotic bones, only a small portion of its mesial border being visible.

The exoccipital seen ventrally is of considerable size, yet of smaller extent than in the
adult. Mesially, of course, it bounds the basioccipital and lateral portions of the
foramen magnum; the increase in size which takes place as growth proceeds is due to
additions to the external lateral border forming the tympanic wing. In the nestling
one-quarter grown this external border is divided into two regions, an anterior and a
posterior, by a deep notch. ‘The anterior segment is represented by an almost quadrate
plate bounding the basioccipital and forming the floor of the pro-otic ; the posterior segment
belongs to that portion of the plate which bounds the foramen magnum. The external
border of this plate extends outwards to form the lower portion of the trumpet-shaped
mouth of the tympanic cavity. This posterior segment is bounded superiorly by deep
grooves filled by cartilage, covering the pro-otic, the groove being bounded on the other side
by the squamosal. In the adult this groove is completely obliterated, the exoccipital
passing insensibly into the tympanic wing of the squamosal, to be described presently.
Within the cranial cavity the exoccipital is almost entirely concealed by the pro-otic
and opisthotic, only an extremely small portion being visible, mesiad of the opisthotic.

The swpraoccipital is a relatively small bone, deeply cleft in the median line from
the crest of its superior border downwards to the middle of the bone. On either
side of the median cleft lies a long crescentic groove, which later becomes closed by
the meeting of the edges of the groove and converted into the channel for the vena
cephalica posterior. Between the supraoccipital and the squamosal is a large space
filled by cartilage, which, in the dried skin, shrinks and reveals the pro-otic. In the
adult this space becomes filled in by the approximation of the edges of the squamosal and
supraoccipital.

The pro-otic appears externally as an oblong mass of cartilage, of small extent,
between the supraoccipital, squamosal, parietal, and exoccipital bones. Internally it is

-

MORPHOLOGY OF THE OWLS. ally

of relatively large size, and is bounded superiorly by the parietal, anteriorly by the
alisphenoid, and mesially by the basioccipital, inferiorly by the opisthotic, and
posteriorly by the epiotic. The floccular fossa, which is deep and pit-shaped, lies wholly
within the pro-otic, but is closed posteriorly by the epiotic. The meatus internus is very
shallow. The pro-otic is concealed from view externally entirely by the squamosal, but
is widely separated from contact with the latter by a great air-space—the recessus
tympanicus superior.

The epiotic in the skulls now under description is completely ossified, and anchylosed
with the supraoccipital. Its relations to the pro-otic are made manifest through fine
sutures passing through into the floccular fossa.

The opisthotic has completely fused with the pro-otic and basioccipital, only a trace
of a suture remaining visible below the floccular fossa to mark the separation of the two
otic bones. Fusion with the basioccipital is complete. A well-defined suture remains,
however, to divide the opisthotic from the epiotic and supraoccipital bones.

The basisphenoid, seen from the cranial cavity is still distinct; seen in section, the
suture between itself and the basioccipital is also distinct. Only the inner table of
the basisphenoid is, however, represented; all the rest of this bone has been absorbed
to contribute towards the formation of the huge recessus tympanicus anterior. The
pituitary fossa forms a basin-shaped pit, which is externally completely surrounded by the
air-sinus just described. There is no indication of the ossification of the basisphenoid
from separate centres, such as is seen in the Pygopodes, for example.

The presphenoid, if present, is still represented only by cartilage.

The alisphenoid proves to be an exceptionally interesting bone in the Strigide. In
the young skull of Syrnéwm, this bone is partly concealed externally by the squamosal,
which overlaps its infero-external border. So much as is visible is roughly oblong
in form, its long axis extending from the interorbital septum outwards.

Besides being overlapped by the squamosal, it is embraced above by the orbital process
of the frontal, and below by the basisphenoid. Its precise form can best be studied
from the inside of the skull. Viewed from this aspect, it will be found somewhat lingui-
form in outline, and to send inwards from its anterior border a long arm-like process
which arises from a swollen base. But little change takes place in the form of this
bone in later development ; the space which is left above this arm-like process is filled in
by the orbital process of the frontal and the orbito-sphenoid.

The relations which obtain between the alisphenoid, parietal, squamosal, and frontal
are discussed on p. 18. They are especially worthy of note, seeing that they differ
completely from those which obtain in Speotyto, for example.

The orbito- and pre-sphenoids have not yet begun to ossify.

The mesethmoid (Pl. 2. fig. 7a) is yet only partly ossified. In the skulls in the
Museum Collection it forms a linguiform plate projecting downwards from the skull-
roof to the parasphenoid. Its anterior border is straight, its posterior border convex.
The dorsal border is expanded into an oval horizontal plate visible, in the youngest
skulls, on the surface of the skull, where it appears wedged in between the frontals,
and overlapped on either side by the nasals and nasal processes of the premaxilla.

SECOND SERIES.—ZOOLOGY, VOL. IX. 3

18 MR. W. P. PYCRAFT ON THE

The quadrate has not yet assumed its fully adult form, the orbital process being
still cartilaginous.

The columella in the older skulls has completely ossified, but stapedial rays are
not traceable.

The articular is still separately distinguishable.

Membrane-bones.

The parietal presents two distinct forms. In Syrniwm it may be described as oblong
in form, but having the infero-lateral angles obliquely truncated by the overlapping
of the squamosal. It extends outwards and forwards to the alisphenoid, and divides the
frontal from the squamosal. In a half-grown skull of Syrniwm aluco the frontal sends
down, immediately behind the alisphenoid, a tongue-shaped process to overlap the
supero-external angle of the parietal, and thus appears to diminish the distance between
itself and the squamosal.

Bubo, Scops, Ketupa, Gymnoscops, Syrnium, and Strix have this parieto-alisphenoid
articulation. In Scops, however, the skull appears to be undergoing a change in the
matter of the relations of the bones of this region, inasmuch as the oblique external
lateral border has been cut back so as to cause it to fail to meet the alisphenoid,
and allow the squamosal and frontal to meet.

Speotyto differs markedly from the type seen in Syrniwm in the form of the parietal,
and in the relations of this to the neighbouring bones—alisphenoid, frontal, and squamosal.
In Speotyto the superior border of the parietal, instead of being straight, rises upwards
and forwards for a considerable distance and is sinuous in outline. Its external lateral
border is deep, but separated from the alisphenoid by the whole width of the squamosal.
Finally, by the considerable backward extension of the squamosal, the inferior parietal
border is restricted to the supraoccipital region, instead of extending outwards above
the exoccipitals also.

The frontal, save in the differences in the relations between itself and the neighbouring
bones in the two types just described, differs but little in form in either. Its general
conformation can be seen in PI. 2. figs. 7, 8.

The squamosal, like the parietal, differs considerably in form and its relation to the
alisphenoid, frontal, and parietal bones. In Syrniwm it undergoes considerable changes
in course of growth: in the quarter-grown skull it is pentagonal in form, and bent upon
itself so as to present two distinct faces, an external lateral and a posterior; in the
half-grown skull it has become oblong in form and quadrangular. Both stages,
however, differ from the squamosal of Speotyto; for whilst in Syrniwm the superior
border of the squamosal is gently arched and runs forward beneath the parietal to the
alisphenoid, in Speotyto this border suddenly rises, near its middle, to form a large
quadrate plate, articulating by a long horizontal suture with the frontal, and widely
separating the parietal from the alisphenoid (PI. 2. figs. 7, 8).

The mesial border of the squamosal, as may be seen when this bone is dissected from
the skull, turns sharply inwards, forwards, and downwards, the resulting flattened

MORPHOLOGY OF THE OWLS. 19

face articulating and ultimately fusing with a similar flattened area on the parietal
and alisphenoid. As a consequence of this inturning of the convex squamosal,
a spacious chamber is formed—the recessus tympanicus superior.

A comparison between the skull of the adult and nestling of Syrniwm and Speotyto
reveals yet other characters of considerable interest ; these concern the morphology
of the postorbital process and the tympanic wing, and the formation of the temporal
fossa. All three points may conveniently be discussed here.

Commencing with the postorbital process, we may remark that this, in Syrniwm
and other forms with a similar type of skull, is formed by the alisphenoid, whilst
in Speotyto it is formed by the antero-inferior angle of the squamosal.

The tympanic wing in both types of skull is formed by the free edges, exoccipital and
squamosal.

The temporal fossa in the young skulls is wanting. It is formed, in Syrnium, by
a depression which has for its centre the sutures of the alisphenoid, squamosal, and
parietal bones, and later becomes more sharply defined by the excessive development
of the postorbital processes and tympanic wing (compare PI. 2. figs. 4, 6).

From the cranial cavity the squamosal in Syrniwm is only visible as a small hour-
glass-shaped tract of bone lying between the alisphenoid and parietal and laterad of
the pro-otic; but in Speotyto it-forms a large quadrangular plate bounded above by
the frontal, below by the pro-otic, behind by the parietal, and in front by the alisphenoid,
thus contributing to a very considerable extent to the formation of the brain-case.

This upward growth of the squamosal externally, accompanied by the gradual
absorption of the underlying frontal and parietal elements and the usurpation of their
function in the protection of the brain, is an extremely interesting feature, and marks
_ an advance in the evolution of the skull. Proof of this advance we may derive from the
fact that the arrangement seen in Syrniwm is a primitive one, agreeing almost exactly
with that seen in Dromeus for example. In the latter, it is furthermore interesting to
note, the squamosal has not yet succeeded in absorbing the parietal wall, and so is
entirely excluded from participation in the formation of the cranial cavity.

I propose to deal with this question shortly in another communication, in which
T hope also to be able to show that, in the evolution of the Avian skull, a gradual
increase in the length of the frontal has taken place, accompanied by a shifting
backwards of the parietal and supraoccipital, the last two moving backwards through a
quarter of a circle.

The nasal is truncated posteriorly, and does not extend backwards quite so far as
the posterior border of the horizontal plate of the mesethmoid. Mesially the nasals
meet one another in Syrniwm, and almost conceal the mesethmoid ; but in Speotyto
they only overlap the mesethmoid, leaving the centre of that plate fully exposed. The
form of the nasal cleft is holorhinal. The maxillary process of the nasal is truncated
inferiorly ; the premaxillary process is long and slender.

The lachrymal is placed rather far forwards, and lies entirely underneath the nasal.
It differs but little from that of the adult stage in form.

The premaailla has the nasal processes strongly arched, and terminating at the

3*

20 MR. W. P. PYCRAFT ON THE

anterior end of the mesethmoid plate. The palatal processes are vestigial, the
maxillary processes relatively large.

The maxilla extends backwards, from the level of the maxillary process of the nasal
to the level of the tip of the orbital process of the quadrate, in the form of a long
slender rod constituting the outer sheath of the quadrato-jugal bar, and forwards as far
as the level of the anterior angle of the anterior nares. Its maxillo-palatine process
is very large and swollen, so much so as nearly to obliterate the lachrymo-nasal fossa.
The antrum of Highmore opens far forwards on its floor, and leads forwards by an
extremely short channel into a vault in the premaxilla roofed by ossified alinasals.

The jugal has the usual elongated and splint-like form, and overlaps the quadrato-
jugal posteriorly and the maxilla anteriorly.

The quadrato-jugal is long, extending forwards along the inside of the bar to beyond
the middle of it.

The parasphenoid appears to present the usual elements—a basitemporal plate, a pair
of pretemporals, or alisphenoidal wings, and a rostrum. It seems to me a point of
some significance that the basitemporal plate is quite separate from the rostrum in
the youngest skulls in the Museum Collection, a sharply defined suture-line being visible
when the skull is seen in section. It suggests, however improbable it may seem on
reflection, that what appears to be the middle region of the parasphenoidal rostrum may
really be the floor of the basisphenoid; the parasphenoid in this case would be
represented only by the basitemporal plate and the tip which projects beyond the
basisphenoid in such a way as to look as if it had been thrust through the antero-ventral
angle of that bone. So much of the bone as would underlie the basisphenoid is thus
supposed to have become absorbed. Elsewhere (p. 17) I have interpreted the features
observable in this region by supposing the basisphenoid floor to have become absorbed _
and the parasphenoidal rostrum to be persistent. The reverse may be the case, though
it must be admitted the weight of probability is in favour of the earlier interpretation.
Instances of the replacement of one bone by another by absorption, till the invading
bone more or less completely establishes itself, are not unknown in the skull. The
stages in the process may be studied in the casque of the Cassowary, and in the growth
of the squamosal for example: both these elements were originally quite external, but
are now slowly working their way inwards so as to take part in the formation of the
cranial cavity. The point in question can only be determined by a careful examination
of much younger skulls than we at present possess. The remarkable shortness of the
rostrum in the Owls is a feature already commented upon.

The vomer is small, and quite divorced from the hemipterygoid segment of the
pterygoid. Its support has now been transferred to the palatines, which send inwards,
for this purpose, from their mesial edges a pair of quadrate spurs.

The palatine may be described as scimitar-shaped, increasing in breadth gradually
from before backwards. It terminates anteriorly near the tip of the beak in a fine
point, but the posterior extremity presents some features of interest. Viewed from the
ventral surface, it will be seen that the movement towards the middle line has caused
the articulation with the pterygoid, which takes place at about this time, to form on the

MORPHOLOGY OF THE OWLS. 21

extreme postero-external, instead of on the extreme posterior end of the rod. Viewed
from above, it will be seen that a strap-shaped bar has been developed from the mesial
border of the rod, and that this runs forwards and inwards to form a support for the
vomer. Immediately behind this process will be found a degenerate hemipterygoid.

The pterygoid differs from that of the adult only in so far as the articulation with the
palatine is concerned. In the young Owl, as in other Neognathe, it is only very
imperfectly formed. Viewed from above in a half-grown nestling of Syrniwm (Brit. Mus.
n. 99.7.19.1), the main shaft may be clearly distinguished from a hemipterygoid rela-
tively large but which yet fails to reach the vomer. The hemipterygoid rests upon the
mesial border of the palatine, and extends backwards to be received into an indistinct
cleft in the main shaft. The palatine immediately behind the hemipterygoid extends
backwards as far as the lower lip of the cleft. Ultimately a synovial joint is formed
with the palatine and pterygoid shaft, whilst the hemipterygoid, by fusion with the
palatine, disappears. The method of articulation recalls that of the Sphenisci.

The dentary resembles that of the adult skull.

The splenial is still distinct: in the youngest skulls it is of considerable size, termi-
nating posteriorly on a level with the lateral vacuity, and anteriorly near the distal fifth
of the ramus.

The coronoid overlaps the proximal end of the lateral vacuity, and posteriorly takes
part in the formation of the internal angular process.

The angular forms the inferior border of the proximal end of the ramus, and extends
forwards nearly as far as the distal extremity of the splenial, turning inwards on its way,
so that whilst the proximal half of the ramus has the inferior border formed by the
angular, this border for the anterior half is formed by the dentary.

The supra-angulare passes forwards between the dentary on the one side and the
splenial on the other.

IV. THE VERTEBRAL COLUMN, WITH REMARKS ON THE EXCALATION
OF VERTEBRAE.

All the presynsacral vertebrie are free ; the thoracic are heteroccelous. In general
shape they resemble those of the Falconiformes rather than those of the Caprimulgi,
but they possess characters which distinguish them from both of those groups. These
distinctions, however, are so slight that they only hold good when the vertebral column
as a whole is compared, single vertebrze being frequently easily confounded either with
the Falconiform or Caprimulgine vertebrae, according to the particular peculiarities.
Generally speaking, the vertebra of the Striges are rather less pneumatic than those
of the two groups above mentioned. The pleurosteites, and the diapophyseal lamellz
with which they fuse, form in the Striges a narrow and well-defined outstanding
projection on either side of the anterior end of the centrum, sharply contrasting with the
centrum itself, which is, as in the Caprimulgi, relatively slightly longer than in the
Falconiformes. Hyperapophyses, so conspicuous in the cervicals of the larger Falconi-
formes, are wanting or only feebly developed in the Striges.

The neural arches of the atlas rise almost straight upwards above the centrum, and

22 MR. W. P. PYCRAFT ON THE

arching over the neural canal form but a narrow bar. The odontoid ligament is
unossified. The neural arches of the 2nd to 4th cervicals form, as in the Falconiformes
and some Caprimulgi, a broad quadrate plate having the postero-external angle
produced into more or less prominent hyperapophyses. From the 5th to the 9th
vertebree, the arch is deeply notched posteriorly, and the neural spine is either wanting
or very feebly developed. From the 6th to the 9th the vertebra bear prominent
catapophyses, which at the 10th give place to hypapophyses, which are continued
backwards to terminate with the 2nd or 3rd thoracic vertebra.

The thoracic vertebree are all free save the last, which is fused with the synsacrum.
The hypapophyses, which vary in length and slenderness, never extend beyond the 3rd
vertebra. The neural spines (e. g. Asio, Strix) interlock by means of a pair of spinous
processes, which projecting backwards from the postero-superior angle of the vertebra
embrace the antero-superior angle of that next behind. This interlocking is further
strengthened, in Asio for example, by long backwardly directed processes extending
from the hyperapophyses of the 1st and 2nd thoracic vertebrz, and by similar processes
formed by the production forwards and backwards of the external angles of the transverse
processes of the 4th to 6th vertebrae. Conspicuous pneumatic apertures open beneath
the transverse processes of the thoracic vertebre.

‘he synsacrum includes from 13 to 14 vertebrae. Of these, two are thoracic, four may
be lumbar, three lumbo-sacral, two sacral, and three caudal.

The parapophyseal processes of the lumbar vertebree are short, the last abutting
against the inferior border of the preacetabular ilium near its middle.

A planum coccygeum is generally present, and is especially distinct in certain forms,
e. g. Syrnium, where it is isolated by a deep fovea pudendalis; whilst in Bubo, for
example, its distinctness is masked by retention of the parapophyseal processes of the 1st
and 2nd caudal vertebrae.

The transverse processes of the free caudal vertebrae appear to be relatively longest
in Asio and Sériz. Free intercentra are present in the most proximal vertebra ;
posteriorly these elements fuse with the centra, forming blunt hypapophyses.

As in other groups of birds, there is evidence of a slow process of reduction in the
length of the vertebral column, apparently brought about both by excalation and
absorption of the vertebrze, the former method taking place in the presacral region,
the latter in the postsacral region.

A secondary process of reduction, affecting the number of vertebre in the different
series, has caused the actual reduction in the total number of the series to be generally
overlooked. The reduction in the numbers of the different series chiefly affects the thoracic
and cervical series, and is largely due (1) to the backward shifting of the sternum, and
(2) to the reduction of the sternal facet for the sternal segments of the ribs. The
backward shifting has resulted in the divorce of certain sternal segments of ribs from their
articulation with the sternum, and the consequent transference of the vertebra bearing
the rib to the cervico-thoracic series. The degeneration and final disappearance of the
sternal rib rapidly follows this divorce: later the vertebral segment of the rib likewise
becomes reduced, finally disappearing also. At least this appears to be generally the

MORPHOLOGY OF THE OWLS. 23

case: sometimes, however, the capitulum and tuberculum and a remnant of the shaft
remain, and fusing with the di- and par-apophyses of the vertebra form a bridge for the
vertebral artery, thus converting the cervico-thoracic into a “true” cervical vertebra.

The sharp distinction maintained between the cervical and cervico-thoracic vertebre,
by reason of the fusion of the rib-element of the former with the centrum, is a point of
some interest. This fusion may be ascribed, perhaps, to kinetogenesis; or, more vaguely,
to the adaptation to the peculiar and characteristic movements of the neck in birds.
Only in very young birds are the riblets of the cervical vertebre free, but in Archeo-
pteryx they appear to have been free throughout life. The number of these cervical
vertebrze is, as we have already pointed out, being constantly augmented by additions
from the cervico-thoracie series. The latter series is fed by the thoracic, as the
sternum shifts further and further backwards.

The posterior thoracic vertebrze become severed from the sternum by the reduction
of the articular surface for the sternal ribs, a reduction sometimes associated with a
reduction of the sternal plate itself. Severance from the sternum brings about, as in
the case of the anterior ribs, first a reduction of the sternal, and finally of the vertebral
segments; the amount of the reduction being proportionate to the period of their
isolation. The result of the final disappearance of the last vestiges of these ribs is to
leave the vertebra to which they belonged indistinguishable in appearance from the
vertebree of the lumbar series, with which it is generally reckoned. Occasionally, small
spicules of the sternal ribs remain after the complete disappearance of their vertebral
segments.

The fate of this hindmost vertebra brings us to the consideration of the question of
reduction by excalation.

Excalation appears to occur most frequently in the lumbar region, but also in the
cervical, lumbo-sacral, and caudal regions.

In the cervical region the number of vertebre is constantly 14. So far I have
found only one exception to the rule, when the number was reduced to 13 (Athene,
Brit. Mus. n. 1095e). Of these 14 vertebrze the number bearing fused riblets varies
between 11 and 13, the remainder being cervico-thoracics bearing free ribs or vestiges of
them. Since in the skeleton of Athene the maximum number of thoracic and lumbar
vertebree are present, there can be no doubt but that the reduction in the cervical region
is a real, not an apparent one.

In the lumbar region the maximum number of vertebree is four (fig. A, p. 28). But
for the correct determination of the vertebrze in this region it is necessary carefully to
determine the number and limits of the thoracic series ; inasmuch as the last thoracic,
by the loss of the vertebral segments of the ribs, become indistinguishable from the
lumbar series (*7, fig. B, p. 28).

The number of thoracic vertebrae appears invariably to be 7, six of which may be
attached to the sternum, though frequently only four succeed in effecting such an
attachment. Carlier, in more primitive ancestral forms, as we have already pointed
out, probably as many as ten vertebrae reached the sternum; severance from it has
taken place at both ends of the sternal plate.

24, MR. W. P. PYCRAFT ON THE

The seventh pair of thoracic ribs are frequently found only in minute traces of their
sternal seements (figs. B, C, s.7. 7, p. 28), at times even these are wanting. Sometimes
traces may be found of vertebral as well as sternal segments, and still more rarely the
whole rib may be found, though it fails to reach the sternum (fig. A, s.7. 7, p. 28). In
three specimens in the Museum Collection the seventh rib is well preserved—Bubo
magellanicus (s.r. 7, fig. A, p. 28), a Nyctala 98.22, and a Minox 1382 a.

The constancy in the number of the cervical and thoracic vertebre and the fixed point
afforded by the sacral vertebre enable one to determine, with some certainty, the
number of the lumbar and lumbo-sacral series.

Only five skeletons in the National Collection have as many as four lumbar vertebre.
At first sight, two skeletons of this five appear to have five lumbars, indistinguishable
one from another ; as a matter of fact, the first vertebra belongs to the thoracic series, as
is proved by the vestiges of the 7th pair of sternal ribs. Thus, it is absolutely necessary
to fix the last thoracic before attempting to count the lumbar vertebree; and this
fixation is often only possible through the presence of the sternal segments of the
thoracic ribs just referred to.

The maximum number then of lumbar vertebree is four, and this is now rarely attained ;
in all, save five, of the skeletons under my charge, only three are present. Since there is no
suspicion whatever of the fusion or reduction by crowding of two contiguous vertebree in
this series, I gather that the loss of the vertebree in the series is due to excalation. It is
possible, in some instances, that the 4th lumbar has, by loss of its par- and di-apophyseal
processes, become merged in the lumbo-sacral series; but in this case, it is still a fact
that a vertebra has been lost, only it is a lumbo-sacral and not a lumbar that is missing.

The lumbo-sacral vertebrze are three in number, but in five cases they are reduced to —
two, as in fig. C, 1-2, p. 28. Instances of this can be seen in the annexed table.
Excalation appears to have been at work here again, since there is no trace of the
reduction and fusion of contiguous centra.

The primitive sacrals are two in number (s.v., figs. A, B, p. 28), but the ribs, either of
the 1st or 2nd pair, are occasionally wanting.

The sacro-caudal are three or four in number (1-3, fig. C, p. 28). Occasionally the
vertebra corresponding to the first free caudal fuses with the last sacro-caudal, so that
the synsacrum actually projects beyond the limits of the mesial borders of the post-
acetabular ilia.

The post-synsacral vertebree or free caudal vertebree are either 7 or 8 in number.
In very young birds, e. g. Speotyto, there are 9 free caudals, the additional vertebra
lying in front of the pygostyle. It is small and wedge-shaped, and gradually dimin-
ishing, partly by the mutual pressure of vertebre on either side. Later in development
all trace of this vertebra disappears, owing to its fusion with the pygostyle.

Proof of a reduction in the number of presacral vertebrae may be equally well demon-
strated if the component numbers of the various series be disregarded and the total
number of vertebre be counted between two fixed points—the atlas and second sacral
vertebree. As is shown in the following table, the number counted in this manner varies
between 28 and 30, whilst the variation in the total number ranges between 39 and 42.

MORPHOLOGY OF THE OWLS.

25

Table showing Variation in the Number of the Vertebre, caused partly by Excalation and

partly by Absorption.

Buso MaGeLtantcus. 1334 a. Buso macutosus. 1335 a. Buso ontenratis. 45a.
OL Saas ite .) 13 ) 13 7}
Pi .. 241 ju | 1 igh 1 i cat
a emer 2 30 pone oP ae an et ti 4+30
Mies css {i
LSc.. 3 f 3 3 ;
a... 2 J 2 i 2 J
8.C. 4 : a | 4
CAL: ia 8 \ We fii J 11 Caudal series incomplete.
Total=42, Total =41—Lost: 1 free caudal.
Buso maximus. 98.57.23. Nryox pooxsoox. 1332 a. Buso capensis. 98.6.8.2.
Cee ees. 11 ) 12 i) 11 \
OTh. .. 241 a ea 14} 241 aad
Th. 5+1+1 5 6+1 ) War 6+1 9
rey. 3 } 10 ( e 3 (Oe 3 ae eit
L.Sc.. 3 i 3 | 3
Sc. 2 J 2 a) 2 J
S.C. 4 4 ‘ 3 |
=... 8 i i 8 \ bs 8 lige
Total 41—Lost: 1 lumbar. Total 41—Lost: 1 lumbar. Total 40—Lost: 1 lumbar.
1 caudal,
Nycrata. 98.5.7.22. SPEOTYTO CUNICULARIA. 98.6.8.1. SURNIA FUNEREA. 98.5.7.17.
eae... 12 my 1 Veen 11 we)
Oth .. 141 p14 | 21 fi Teepe Ses
| 5 a
a S+1+11 19 S29 Jy aaa Bpray jay LOH a
ae 3 3 | 3 |
Sey... 3 | 3 2
Sk geaeee 2 J 2 3 2 J
S.C. 3 3 | 4 lisic
— 8 1 8 (nestling 9) { 1 8 he
Total 40—Lost: 1 lumbar. Total 40—Lost: 1 lumbar. Total 40—Lost : 1 lumbar.
1 caudal. 1 lumbo- 1 lumbo-
sacral. sacral.
GLavUcIDIUM. Gymwnoscops INsuLARIsS. 14,11.10.1. PULSATRIX TORQUATA.
Ce sass. 11 ) 11 er i(d) } y
CTh. an pty Baers) tell 14141 p14)
’ | = | i=
Th. 5+14+1\ 19 y98 S+1+1 119 \o8 5+1+1) 19 y98
Le 3 3 3
SGs.'.,..- 2 | 2 2 {
‘hy OS Z D) 2 J 2 3
8.C. a F 4 lio 4 9
C.. 28 \ o 8 tn 8 } te
Total 40—Lost: 1 lumbar. Total 40—Lost: 1 thoracic. Total 40—Lost: 1 lumbar.
1 lumbo-sacral. 1 lumbar. 1 lumbo-
sacral,
ATHENE. 1095. Scors. 98.6.8.3.
O. 12 ] i >
oth ...... 1 } ae | o41 fit !
1h Soe 6+1 5+1+1 } 9
a, a1 52s Seen” 52
TR SGsete «<< 2 2 |
aoe... 2 J 2 B,
S: Orr ties: 3 4
Oy ee 8 } 7 } a

Total 39—Lost: 1 cervical.

1 lumbo-sacral.

1 caudal.
SECOND SERIES.—ZOOLOGY, VOL. IX.

Total =39—Lost: 1 lumbar.
1 lumbo-sacral.
1 caudal.

4

26 MR. W. P. PYCRAFT ON THE

The component elements of the typical vertebral column may be expressed as
follows :—

$.Se.
—- ee
Oyo (Cy. ths. elbe REL). Lb: 4. ib.Se. 3: Se. 2: Cd: ese =42.
es es
7 12

It should be stated that the evidence for the excalation of vertebrze, which, it is
contended has taken place, rests upon the fact that the presacral vertebrze show no trace
whatever of fusion of elements, and the nerve-apertures are all perfectly normal. This is
a point of some importance, since in the vertebral columns of Amphibia possessing fewer
than the normal number of vertebree, the reduction appears generally to have been
brought about by the fusion or confluence of vertebrae, rather than by excalation.
The reduction in the number of the caudal vertebree among the Aves, however, is, as we

have just indicated, undoubtedly brought about by the absorption of the vertebree lying -

immediately in front of the pygostyle. This is well seen in Brit. Mus. 28.5.7.39, showing
the remains of the 7th free caudal, which consists, now, of a small portion of the centrum
only, and this forms a wedge between the 6th vertebra and the pygostyle. Later in life
this wedge fuses with, and forms part of, the pygostyle itself.

It may be contended that the inconstancy in the number of presacral vertebrz is due
to what Mr. Bateson calls “ Meristic variation.”” It seems to me, however, that the facts

submitted rather favour the view that a general and orderly reduction in the Jength of

the vertebral column is taking place, rather than that which implies simply a series of
sporadic variations of no apparent meaning. Before this matter can be settled satis-
factorily, it will be necessary to examine a much larger series of individuals of the same
species, and, if possible, of nestiings from the same nest.

V. Tuer Rzrps.

The cervical ribs are styloid and short. The heads of the ribs have the form of flattened
band-like lamellz and fuse above with the diapophysis, below with the catapophysis of

each vertebra, thus forming a canal for the vertebral artery. The head and the short
shaft are sharply contrasted one with another; so that, in a lateral view, the pleurapo-
physeal lamella stands out buttress-fashion against the centrum. The hindmost cervicals.

or cervico-thoracics, however, must be excepted. These are two or three in number.
The 1st and 2nd are generally represented by vestiges, only the tuberculum being present :
the third is generally long and styliform. From this we may gather that from two to
three of the thoracic vertebree have been transferred to the cervical series by the shifting
backward of the sternum, the sternal ribs disappearing after their severance from the
sternum.

The thoracic ribs are long and slender, and increase in length from before backwards :
as a result the thoracic cavity is very spacious.

The number of thoracic ribs may reach a total of seven pairs, of Baeen six pairs may
articulate with the sternum, e.g. Nénox. But this is rare, the number in the majority

PE

MORPHOLOGY OF THE OWLS. Dy,

of forms varying between 5 and 6 pairs, the last of which fail to reach the sternum.
Some species of Strix (e.g. S. delicatula), Asio, Carine, Surnia, Speotyto, Syrnium, have
five pairs, which articulate with the sternum, the sternal segment of the 6th pair being
generally bound by connective tissue with the lower fourth of the fifth corresponding
segment. In one skeleton of S¢riv, however, in the Museum Collection (1223 a) the
sixth pair articulates with an imperfect facet projecting from the lower one-third of
the fifth segment. Frequently only four pairs of ribs articulate with the sternum, the
fifth, or even fifth and sixth pairs having sternal segments too short to reach the
sternum.

Various stages in the reduction of the number of the ribs are to be met with. Leaving
out of consideration the anterior thoracic (cervico-thoracic), we may have as many as
seven pairs of true thoracics. Of these not more than six pairs ever articulate with
the sternum, the seventh in some cases being quite long, the sternal segment almost
reaching the sternum, e. gy. Athene (1095 e). In others vestiges of the vertebral segment
only remain, e. g. Gymnoscops (94.11.10.1), or of the sternal segment only, e.g. Bubo
capensis (98.6.8.2). The reduction of the 6th and 5th pairs proceeds in a similar way.

There is evidence to show that at least nine pairs of ribs recently articulated with the
sternum. Of these the first and second pairs, by loss of their sternal segments,
have become transferred to the cervico-thoracie series; whilst the last pair, by loss of
their vertebral segments, have become transferred to the lumbar series. In many species
the number of ribs has become, as we have already remarked, reduced to five pairs, of
which four pairs only articulate with the sternum, the reduction taking place sometimes
from the anterior, sometimes from the posterior members of the series.

The uncinate processes are well developed, but except rarely not more than five pairs
are present. Occasionally the last cervico-thoracic rib bears an uncinate, but the last
(7th) thoracic appears to have lost the appendages completely. They are long and
slender in shape, sloping obliquely upwards and backwards. ‘They are relatively
longest in Strix, extending backwards on to the third rib from their base of attachment.

VI. THe PEcTORAL GIRDLE AND STERNUM.

The pectoral girdle of the Striges is extremely uniform in character throughout the
group, and in certain characters bears a very close resemblance to that of the Falconi-
formes.

As a whole the girdle of the Striges can be distinguished from that of the Buteonine
section of the Falconiformes by the presence of a prominent procoracoid process, and
from the Falconine section, in which a procoracoid process is present, by the smaller
relative size of the acrocoracoid, which in the Falconide is large, and by the form of the
furcula. This, in the Striges, is relatively long and slender, only slightly curved dorso-
ventrally, and is furthermore frequently incomplete, the limbs failing to meet in the
middle line.

The pectoral girdle of the Striges can readily be distinguished from that of the
Caprimulgi, in that in the latter group there is no procoracoid process, and the furcula
is strongly curved dorso-ventrally.

4.*

28 MR. W. P. PYCRAFT ON THE

The coracoid is long, being nearly or quite as long as the sternum; it has a prominent
acrocoracoid and a large procoracoid process, broad at the base, and terminating in a
downwardly directed hook-shaped process articulating with the furcula.

The coracoid of the Asionide may be distinguished from that of the Strigidee in that
in the former the acrocoracoid affords an articular surface for the furcula, which
develops a corresponding articular facet by sending outwards from the distal end of each
limb an oval plate. This articulation between these two bones is wanting in the
Strigidee.

The coracoid of the Striges bears a close resemblance to that of the Falconiformes,
but it may be distinguished from all, except the Falconide, by the presence of the
procoracoid process. With the Falconidze, however, the case is different ; and it becomes
a matter of nice discrimination to tell the coracoid of the Falcones or Polybori from that

Bubo magellanicus. Bubo capensis. Scops.

Synsacra showing the reduction of the vertebra by excalation ; cf. pp. 23, 24.

of one of the Striges. The Falcons may be distinguished by the absence of a supra-
coracoid foramen ; but in the Polybori and Striges the foramen is present, and almost
identical in size and position. In the Striges, however, it will be found to lie somewhat
nearer the scapula. In the Striges the base of the coracoid is comparatively deeply grooved
to fit the dorsal lip of the groove; whilst in the Polybori, what corresponds to the dorsal
lip of the groove in the Owls is represented only by a low and incomplete ridge.

The scapula has a swollen acromion, with that portion of its surface which forms the
roof of the foramen triossewm perforated by pneumatic foramina.

The furcula is long, slender, and but slightly arched dorso-ventrally, and wants a
hypocleideum.

MORPHOLOGY OF THE OWLS. 29

In the Asionidze, its distal extremities are laterally compressed, and furnished with
an oval facet for articulation with the coracoid. In some species, the two limbs fail to
meet in the middle line, e.g. Speotyto, Glaucidium, Surnia, whilst in others the
degeneration of this region has not proceeded quite so far, the limbs being still united,
but by a very slender thread of bone. In the nestling Speotyéo it is interesting to note
that the furcula is still entire. The backward extent of the furcula varies considerably.
In many of the Asionidee the furcula reaches the inferior and anterior angle of the
carina, and is held in position by strands of connective tissue, no articulation taking
place. In others the fureula falls considerably short of the carina, e. gy. some species of
Bubo, Syrnium, Asio.

In some eases, e.g. Bubo maximus, the furcula is closely bound to the anterior border
of the keel, the union taking place near the middle of the border. The furcula is
pneumatic in all the Asionide.

In the Strigidz the furcula articulates with the carina, which at the point of articu-
lation has developed a lateral expansion. The furcula differs from that of the Asionide,
not only in the fact of this articulation, but in that the distal ends lack the articular
facet for the coracoid. Furthermore it is non-pneumatic.

The sternum is nearly as broad as long, and bears a well-developed carina. The spina
externa is moderately large, but the spina interna is wanting.

The sternum of the Striges more nearly resembles that of the Falconiformes than any
other group. This resemblance, it should be stated, is most marked where comparisons
are made between Strigine sterna and those of Accipitres. The following characters will
be found useful in determining between sterna belonging to these two very different groups.
The posterior border of the sternum in the Striges is never entire, and never fenestrated,
but always notched. With the exception of the sterna belonging to the Strigide, and
the sternum of Huhua nipalensis (p. 37) of the Asionide, there are two pairs of notches.
The single pair of notches of Huhwa are of great size. They lie on either side of the
metasternum, and extending forwards to beyond the level of the middle of the posterior
lateral process, cause the sternum of this bird to resemble closely that of Microhierax
among the Accipitres. The great difference in size, however, renders any possibility of
confusion on account of this resemblance impossible.

The anterior lateral processes are small and form blunt-pointed projections from the
antero-lateral angles of the sternal plate, which are deeply grooved for the origin of the
sterno-coracoideus, the groove extending back as far as the last rib articulation.

The spina interna is wanting, a deep notch occupying its place. The spina externa is
present only in the Asionide, and here it projects downwards rather than forwards.

The single pair of notches in the sternum of the Strigide resemble those of some
Accipitres, e.g. Hlanoides, in that they are very shallow, so that the posterior lateral
processes pass almost insensibly into the metasternum, being divided therefrom only by
a sinuous line. But the processes are relatively much longer in the Strigidee than in the
Accipitres, and the sternum is narrower.

30 MR. W. P. PYCRAFT ON THE

VII. THe PrEtvic GIRDLE.

The pelvic girdle of the Striges bears a very close resemblance to that of the Accipitres,
among the Falconiformes. As in the Accipitres, the innominate bones are never free in
the adult, the preacetabular ilium is very long, and the pectineal process is wanting.

The Strigine may, however, be distinguished from the Accipitrine girdle by the
following characters :—(1) The preacetabular ilium has the middle of its inferior border
deeply emarginate, so that a line drawn from the cephalic end of the emargination
inwards and forwards to the point where the superior border intersects the vertebral
column, cuts off a large triangular segment of the innominate. (2) The dorsal plane of
the postacetabular ilium is continued forwards and outwards to form a conspicuous shelf
overhanging the acetabulum. (8) The ischium is continued backwards into a point along
the pubis. (4) The pubis is always complete.

The preacetabular ilia never meet one another directly above the neural spines of the
lumbar vertebrae. In Gymnoscops, e.g. G. insularis, the two preacetabular ilia rise to
the level of, and just succeed in touching, the crest of the neural spine of the 1st
lumbar vertebra. By this a pair of conspicuous canales ileo-lumbales are formed. In the
majority of the Striges the hinder openings of these canals are much restricted by the
greater backward extension of contact between the neural spines of the vertebrz and
the superior iliac crest. In many cases the ilia are really rather widely separated, the
canal being roofed by lateral expansions of the crests of the neural spines which extend
outwards to fuse with the ilia. In many genera, e.g. Asio, Bubo, Nyctala, Strix, the
canals are closed posteriorly, and thus become converted into cave ileo-lumbales dorsales.
The postacetabular ilium lodges a fairly large iliac pocket.

The pre- is about twice as long as the post-acetabular ilium, and the latter, it should be
noted, is not sharply deflected as in the Accipitres. The fovea lumbalis differs from that
of many Accipitres in that, owing, probably, to the relatively shorter neural spine, the
vertebral column lies within the fossa. In the Accipitres, the vertebral column projects
beyond the margin of the cavity, and is plainly visible when the skeleton is viewed from
the side. The fovea ischiadicus is of considerable size, lofty, and strongly defined. Clearly
defined limits are not so characteristic of the fovea pudendalis. This region can best be
studied in Syrniwm, where a lofty chamber divides the sacrum from a well-defined planwm
coccygeum. Generally this chamber is obscured by the close approximation of the sacral
and caudal vertebrze, the ribs of the one, and the paraphyseal bars of the other, cutting
the chamber up into a number of compartments. The iliac recess is well developed in
all the Owls.

The pelvic girdle of the nestling is instructive. At this stage the preacetabular is
more than twice the length of the postacetabular region, and the neural spines of the
lumbar vertebrz are low and project above the innominate.

A feature of especial interest is the part played by the postacetabular ilium in the
formation of the iliac recess. If the innominate be removed from the synsacrum, it will
be seen that the mesial border of the postacetabular region curves inwards and down-

MORPHOLOGY OF THE OWLS. dl

wards in the form of a strap-shaped plate, finally meeting the ischium; at this point it
turns abruptly upwards and outwards to meet the postero-external angle of the dorsal
plane, and at the same time closes the ilio-ischiadic foramen. This peculiar downgrowth,
and its subsequent upgrowth, forms the recess in question. Seen from above, the
downgrowth leaves an unfilled triangular space immediately behind the hinder border
of the ilium which later becomes filled up. The above description is based on the
innominate of a nestling Speotyto. In a slightly younger Syrniwm aluco the mesial
inturned border of Speotyto was here twisted so far outwards as to make it appear that
the recess was formed rather by an extension backwards of the inner half of the
hinder, and not the mesial border of the iium. Furthermore, the downward extension
of the plate, and its union with the ischium, was slightly different, inasmuch as, seen
from below, it formed a scroll-shaped mass lying at right angles between the ischium and
the long axis of the pelvis. In the adult, not the slightest clue is obtainable of the
origin of this recess.

The ischium has the posterior extremity rounded, not poimted as in the adult, its
anterior extremity presents a sharply truncated face to unite with the pubis, and sends
upwards a long columnar spur to unite with the ilium, and close the acetabulum
posteriorly. The pubis develops from the dorsal surface of its anterior extremity a
cylindrical spur, truncated at both ends, which are wedged in between the ischium on
the one side, and a descending bar from the preacetabular ilium on the other.

VIII. Tar PrctoraL Limes.

The pectoral limb bears a strong resemblance to that of the Falconiformes. It may,
however, be distinguished by the relatively smaller pectoral crest of the humerus,
the deeply excised postaxial border of Ph. 1, Me. II, and the fact that the proximal
end of Me. III fuses with Me. II distad of the extremity of Mc. I. In the Falconi-
formes, excepting in the Catharte, the fusion of Mc. III takes place at a point
corresponding to a line drawn across the shaft of Mc. II from the distal extremity of
Me. I. The Cathartze agree with the Striges in the last particular.

The hwmerus in the Asionidze has a well-developed pectoral crest, rounded in outline,
the palmar surface of which, for the insertion of the pectoralis, is generally sharply cut
off from the shaft by a well-defined linea aspera. The sulcus transversus (coraco-humeral
groove) is very shallow or wanting, and the ‘neisuwra capitis is only moderately deep.
The tuberculum internus is large, and the tuberculum earternus distinct. The pneumatic
foramen is large. The crista inferior is only moderately developed. The linea aspera
marking the insertion of the deltoideus major extends a considerable distance down the
shaft, terminating in a sharp point. The scar for the insertion of the brachialis inferior
is linguiform and of considerable length. The shaft is nearly cylindrical, sigmoidally
curved, and presents a conspicuously expanded palmar surface, owing to a large, rounded
entepicondylar process. Both ulnar and radial trochlez are well developed, and there is
a small ectepicondylar process.

32 MR. W. P. PYCRAFT ON THE

The humerus of the Strigide differs from that of the Asionidze mainly in its greater
slenderness, and in the feeble development of the various crests and tuberosities.

The humerus, in the adult, is larger than the manus, but shorter than the forearm.
The humerus of the Striges may be distinguished from that of the Caprimulgi by reason
of the extremely well-defined radial tuberosity and the deep incisura capitis seen in the
latter. In the Owls these are not conspicuously developed.

The forearm offers no characters of systematic importance.

The wna has a well-developed olecranon process, feebly developed tubercles for the
remiges and under tail-coverts, and affords a large articular surface for the radius. Two
distinct glenoid surfaces are developed on the palmar surface for the articulation of the
metacarpus.

The radius, which is much more slender than the ulna, is more or less sigmoidally
curved, the distal half running parallel with the ulna, leaving but a narrow space between.
The forwardly bowed portion of the shaft starts abruptly from the proximal moiety
involved in the articulation with the humerus, giving the cotylus for the radial tuberosity
of the humerus the appearance of being supported on a rather long neck. A large
sesamoid, the os prominens, is generally found attached to the distal end of the radius,
as in some Falconiformes, e. y. Hlanus among the Accipitres. In Ninox connivens this
os prominens is of relatively enormous size, and should be carefully examined in the
living bird.

In a considerable number of Owls the second fifth of the radius develops from its post-
axial surface a delicate arch of bone for the extensor metacarpi radialis brevis. Is this
character to be regarded as due to kinetogenesis ?

The carpal bones call for no special comment: the most interesting character which
they present is a deep indentation on the radiale for the reception of the tubercular meta-
carpal I when the wing is fully extended.

The manus is long and slender. The carpo-metacarpus may be distinguished from that
of the Accipitres by the fact that the Mc. III joins the shaft of Mc. II distad of the
Me. I. The trochlea for the ulnare lies on a level with the inferior border of Me. III,
and not in the middle line of this as in the Falconiformes. The intermetacarpal space
is wide.

Phalanx 1 of Me. II has a deeply emarginate postaxial border, the proximal end being
cut away so as to leave a cylindrical shaft, whilst the distal end is expanded to form a
broad plate for the support of the remiges. Ph. 1 of D. III presents a deeply emarginate
postaxial border, the proximal end being broad, the distal end tapering.

The proportions of the various segments of the wing appear to vary considerably
between the nestling and adult stages. The measurements afforded by a comparison of
two nestlings of Speotyto, of different ages, and an adult illustrate this.

In the youngest nestling in the Museum Collection (98.5.7.39) the humerus is
slightly longer than the forearm, whilst the latter and the manus are almost subequal.
In a second specimen (98.5.7.38) the humerus is conspicuously shorter than the fore-
arm, whilst the latter is now markedly longer than the manus. In the adult, the humerus
is one-quarter shorter than the forearm, and the disproportion between the latter and

i

MORPHOLOGY OF THE OWLS. 33

the manus has become still greater. The accompanying diagram illustrates this. Similar
variations in the rate of growth obtain also in the pelvic limb.

Humerus. ae Manus. Humerus. ~ Manus. Humerus. yen Manus.
NESTLINGs. ADULT.

The pneumatic foramen, so conspicuous in the humerus of the adult, is wanting in the
nestling.

IX. Tse Petvic Limes.

The pelvic limb of the Striges is not pneumatic, but otherwise bears a strong
resemblance to that of the Falconiformes. It may be distinguished therefrom, how-
ever, apart from this character, by the fact that the tibio-tarsus lacks an extensor
bridge.

The femur is relatively long and slender, with a cylindrical shaft. The linea aspera
dividing the surfaces for the crwreus and vastus externus runs the whole length of the
shaft from the ant-trochanter to the base of the internal tibial condyle in the Strigide ;
in the Asionidze it bifurcates near the distal third of the shaft. The popliteal fossa is
shallow. . The rotular channel in the Bubonidz is relatively shallow, but broad and deep
in the Strigide.

The tibio-tarsus may be at once distinguished from that of the Falconiformes in
that the extensor bridge is conspicuously absent. Ecto- and ento-cnemial crests are not
markedly developed. The shaft, in the Asionide, is long, markedly inflected, and
bowed slightly forwards. The internal projects downwards below the level of the
external tibio-tarsal condyle, whilst the lateral borders of the posterior trochlear
surface are produced backwards and upwards into a pair of prominent ridges, a feature
which is especially marked in the Strigide. The tibial shaft immediately above these
ridges is somewhat deeply hollowed, so much so in some genera, e. g. Bubo, Speotyto, as
to be saved only by a thin plate of bone from perforating the extensor groove on the
other side of the shaft. A small but prominent tubercle projects from the side of the

SECOND SERIES.—ZOOLOGY, VOL. IX. 5

34 MR. W. P. PYCRAFT ON THE

lower fifth or sixth of the inner side of the distal end of the shaft. The fibula in Ketupa,
and less distinctly in Bubo, Nyctala, and Strix, may be traced downwards as far as
the proximal tarsal mass. In Sypeotyto it fuses indistinguishably with the tibial shaft
rather below the middle of it ; in Scops it reaches as far as the distal sixth of the shaft;
in Carine it extends some distance beyond the middle of the shaft, but not so far as in
Scops. The fibular ridge borne by the tibia is fairly strongly developed, but varies
slightly in length. It is short and near the proximal end of the shaft in Megascops,
Nyctala, and Carine, long and low in Asio, Bubo, Ketupa, and the Strigide ; for example,
Scops and Speotyto have this ridge only feebly developed.

The ¢arso-metatarsus is remarkably Falconiform in its general conformation ; it may be
distinguished, however, by the presence of an extensor bridge—which is rarely absent—
and the disposition of the trochlee.

The length of the shaft varies much, being in some shorter than the femur, e. g.
Scops, Nyctala, Carine; and in others longer than the femur, e. g. Speotyto, Strix. In
Bubo, Ketupa, and Gymnoscops these two segments are subequal.

The hypotarsus is simple, being formed by a prominent and more or less quadrate
bony plate arising from the inner border of the proximal end of the shaft. A deep
groove divides this from a much smaller, laterally compressed plate arising from the
outer border of the shaft somewhat higher up, so that its superior border contributes
towards the formation of the glenoid surface for the ectocondyle of the tibio-tarsus. Not
seldom the inner calcaneal process takes part in the formation of the glenoid surface for
the entocondyle of the tibio-tarsus.

The shaft is grooved both on its anterior and posterior surfaces. Anteriorly the
groove is confined to the proximal end of the shaft, and is both wide and deep. This
feature is especially noticeable in the larger species of Bubo. In these the groove takes
the form of a deep and wide fossa lying at the proximal end of the inner border of the
shaft, and is crossed by an extensor bridge.

Mesially this fossa is bounded by a flat wall forming one side of a triangle, the other
side forming the outer border of the shaft. Within this fossa lies a long narrow scar,
the impression for the ¢ibialis anticus. The posterior surface of the shaftis deeply grooved
throughout, and perforated just below the hypotarsus by a small oblong foramen. Seen
in section the shaft, below the hypotarsus, is #-shaped, but lower down @-shaped. In
Ketupa and the smaller Owls, the anterior groove extends across the whole face of the
shaft. The posterior groove agrees with that of Bwbo just described.

The trochlee differ from those of the Falconiformes in that they are disposed in a
more strongly curved arch, and in that the third trochlea is much shorter than the first
and second—which are on the same level one with another—and directed backwards.
There is, it may be remarked, a passing resemblance in the form and disposition of the
trochlez between the Striges, and especially Ketwpa and Pandion among the Accipitres ;
but whereas in Pandion the 2nd trochlea has its mesial horder strongly raised above the
level of the shaft, and sloping inwards to over-arch trochlea 1, in Ketupa the middle
trochlea presents no strongly developed ridges and is widely separated from trochlea 1.

MORPHOLOGY OF THE OWLS. 35

Further, the tarso-metatarsus of the Striges, as a whole, may be readily distinguished
from that of Pandion, inasmuch as in the former the hypotarsus is simple, in the latter
compound.

The metacarpal I is relatively smaller than in the Falconiformes.

The phalanges of the Striges differ from those of the Buteonine section of Accipitres
among the Falconiformes only in the matter of their respective lengths in different digits.

Ph. 1 D. IT is very short, being only about half as long as Ph. 2; Ph. 1 of D. III is
also very short, so also is Ph. 2, though slightly longer than Ph. 1; Ph. 3 is long, about
as long as Ph. 1-2 combined; Ph. 1, 2,3 of D. IV are all extremely abbreviated, their
combined length being less than that of Ph. 4. The ungual phalanges are all very large,
and in the larger species have the base encircled by a broad raised collar.

The phalanges of the Strigidz differ from those of the Bubonide in that Ph. 1, D. IT
is long, so also is Ph. 2 of D. III.

X. OBSERVATIONS ON GENERIC AND SPECIFIC CHARACTERS.

I hoped, when I commenced the present section of the memoir, to be able to give
diagnostic characters, not only for every genus, but also for the bulk of the species in
each genus. Unfortunately, this Collection, though undoubtedly an exceptionally good
one as compared with that of other Museums, is still far from complete—so much so,
as to render it impossible to fulfil my anticipation.

Many genera are entirely wanting, and those we have are represented for the most
part only very imperfectly. Thus the genus Syrniwm is represented, according to
Dr. Sharpe, in his ‘ Hand-list of Birds’ (26), by 31 species. The Museum Collection
contains but 3. Ciccaba, with 8 species, is represented only by 1. Ninow, with 44 species,
by 2. Scops, with 80 species, by 3. ;

It is extremely unlikely that all the species in such large genera will prove specifically
distinct, according to the skeletal characters; but I think it certain that a very
considerable number will be found to be more or less easily distinguishable if the sum
total of all the osteological characters be taken into account. At least three examples of
each species of a genus are necessary before full reliance can be placed on the apparent
specific characters, but the following analysis is intended to show what may be done.

The genus Aséo is represented, according to the ‘ Hand-list of Birds,’ by 14 species; of
these, five are possessed by the Museum—A. otus, A. accipitrinus, A. madagascariensis,
A. nisuella, A. major.

The skulls of 4. accipitrinus and madagascariensis are readily picked out from the
skulls of 4. otus, nisuella, or major, by reason of the fact that the postorbital process on
its outer border bears a prominent projecting tubercle for the attachment of the mem-
branous valve dividing the cavernum from the diverticulum in the external asymmetrical
ear. This tubercle is represented by a mere vestige in the remaining species. Further,
the latter are characterized by the greater prominence of their supraorbital processes,
and greater width of the “ post-cavernum ”—the uppermost limit of the tympanic cavity.

A. accipitrinus and A. madagascariensis are also distinguishable by their skeletons:

5*

36 MR. W. P. PYCRAFT ON THE

inasmuch as the latter, apart from its larger size, has a relatively deeper post-cavernum
overshadowed by a prominent tubercular swelling, a relatively larger postorbital
process, crescentic instead of spatular in form, and supporting a relatively larger post-
orbital tubercle which lies nearer the middle of the process; besides the vomer and
posterior ends of the palatines are fused asin the Strigidze. The metasternum is broader
and the external pair of notches wider than in 4. accipitrinus; and in the pelvis, the
dorsal plane is wider and projects further forwards, relatively. In the foot Ph. 1, 2 of
D. II are indistinguishably fused, but free in 4. accipitrinus ; whilst the furcula has the
limbs nearer together and the median apophysis marked by a distinct ventralward
flexure, which is wanting in A. accipitrinus.

The skulls of Strix, Asio, and Photodilus, when compared with regard to the tympanic
cavity, prove extremely interesting, forming a series increasing in complexity, from Striz,
through Photodilus, to Asio. The nature of the modifications can be studied in the
Pl. 2. figs: 2,002

The skull of Photodilus is in many respects besides an interesting one. In the form
of its maxillo-palatine it is intermediate in type between the Strigide and Asionide.
It has the vestigial vomer of Ketupa; the swollen antorbital and interorbital septum
of Strix; the lachrymal of the Asionidee ; the skull-roof of Syrniwm, though less swollen
by pneumatic tissue; a quite peculiar quadrate, inasmuch as its orbital process is
vestigial and recalls that of the Caprimulgi.

The skeletons of the larger species of Athene and Glaucidium are very difficult to
distinguish, so much so that it is probable that a comparison of a large series would
make it necessary to include both genera under one head. The small Glaucidium
ridgwayi—the only small member of the genus represented in our Collection—differs far
more from the larger members of the genus than the latter does from species of Athene
of similar size.

The species included in the genus Syrniwm require very careful study, judging from
the few skeletons in our collection. It is possible that besides Pulsatriz one or two
other forms will have to be redistributed. With regard to Pulsatria it is interesting to
note that whilst in its pterylography it is distinctly Buwbonine, in its skull and trunk-
skeleton it partakes of the characters which obtain in Syrniwm, being indeed almost
halfway between S. wralense on the one hand, and Bubo on the other. 8S. wralense
undoubtedly belongs to the genus Syrniwm, though its skull differs at first sight from that
of S. aluco. These differences, however, when examined, are only due to the exaggeration
of the characters seen in S. aluco. Syrnium seloputo is another peculiar form, inasmuch
as in its skull Bubonine characters are unmistakable, nevertheless Syrniine features
predominate. The skull of Ciccaba is truly Syrniine, but I have not had an opportunity
of examining Scotiapex in this particular. From the variability which obtains in the
skulls of Syruiwm it would seem that the group was but recently derived from the
Bubonine, and that but few of the connecting-links have yet disappeared. ‘This sub-
family will evidently repay much further research.

The parieto-alisphenoid articulation which obtains in Ketupa, Bubo, Scops, Gymno-
scops, Syrnium, and apparently Séria also, may likewise be found in other genera when

ae cl eile

MORPHOLOGY OF THE OWLS. 37

nestling skulls are available. It will be interesting to see how many genera have the
types of squamosal seen in Speotyto. I imagine it will be found in Gymnasio, Nyctala,
Surnia, Carine, and Glaucidium, perhaps in others.

Genera and species are founded generally by ornithologists upon external characters
only, and not seldom upon apparently slender foundations. In doubtful cases of this
kind, an appeal to osteology will sometimes afford unexpected help. Thus Hodgson (14)
founded the genus Huhua to include certain Owls hitherto regarded as belonging to the
genus Bubo. Sharpe, in his Catalogue of Birds (25), suppressed Hodgson’s genus, but
has revived it in his Hand-list (26) and includes therein five species. Of these, but
one is represented in the Museum Collection skeletons, in the shape of a trunk of
H. nipalensis. The sternum of this is remarkable in that it differs not only from that of
all the genus Bubo, but from that of all the Asionidie, in possessing but a single pair
of notches. It will be interesting to see whether this character is common to all the
species included in the genus Huhua.

In the genus Strix specific differences are very small. Out of a total of 26 species

recognized by Dr. Sharpe, only six are represented in the Collection. Of these six,

strangely enough, that with perhaps the more distinctly marked skull, S. poensis, is
regarded by Sharpe, on the evidence of external characters, as indistinguishable from
S. flammeus.

Remembering the paucity of material at my command, it is well to be cautious in
attaching importance to the distinctions which can be made out in these skeletons, but
I give the following diagnoses as a foundation for further work.

A, Size larger, not less than 3 inches long.
a. Pterygoids relatively thick, shaft with strongly curved anterior extremity ; supra-
orbital process feeble ; frontal not constricted in front of supraorbital process ;
sternum longer than coracoid . .. . : : . 8S. pratincola.
b. Pterygoid relatively slender, not much aid antanioney. 5 fupidantieal processes
large, triangular, frontal constricted in front of them ; sternum and coracoid
Gavelly Solar Rens act ts) vo OR PACER ee Pe, woe he aelcatuluss
B. Size smaller, not évecoiling 2 8 inches.
c. Palatines with a slightly emarginate posterior lateral border; vomer large, filling
space between maxillo-palatine processes ; width of interorbital region behind
lachrymal equalling distance from nasal hinge to anterior extremity of external
nasal fossa; proc. lat. basalis of coracoid small, with emarginate lateral border
not extending forward as far as proc. lat. anterior of sternum; sternum shorter
HHATINCOLACOIG) 96° 2 se = Shoe . S. jlammeus.
d. Palatines with deeply emarginate ace aa harden! vomer Perna not ‘filling
space between maxillo-palatine processes; lachrymal relatively small ; interorbital
region with frontal not greatly inflated ; width across interorbital region behind
lachrymal falling far short of distance from nasal hinge to anterior end of
nostril; cerebral dome large and with a deep median furrow; sternum longer
than coracoid; antero-ventral angles of preacetabular ilium produced forwards
TRUER; ks ke tt we tw we 8 8 6S pons,

38 MR. W. P. PYCRAFT ON THE

e. Palatines with posterior lateral border convex ; vomer large, filling space between
maxillo-palatines ; interorbital region behind lachrymal less in breadth than the
distance from nasal hinge to anterior angle of anterior nares ; processus lateralis
basalis of coracoid straight, and extending beyond level of anterior lateral process
OfsterMum..‘.) «2, 5 945 Reger 3) 0; oc). en Serr tise
f. Palatine with posterior lateral border nearly straight; vomer large; pterygoid of
great breadth distad of basipterygoid facet; lachrymal large and with very
shallow lachrymal ¢rooveyseucmmcmmcmeCre te) es) se eet eS rere

XI. SuMMARY.

It is doubtful whether, on the evidence of the skeleton alone, the Striges would ever
have been separated from the Falconiformes, their resemblance osteologically to the
Accipitres being most striking. The anatomy of the soft parts, however, shows
conclusively that there is no real affinity between these two groups. The hind limbs of
the Accipitres and the Striges, as Dr. Gadow has pointed out (12), are almost indistin-
guishable, yet the former has an ambiens muscle, the latter has not. Again, he reminds ~
us, that though both Accipitres and Striges are carnivorous, yet the former have vestigial
czeca and the latter extremely large czeca. In this absence of an ambiens, the form and
size of the czeca, and in the convolutions of the intestines, according to Mitchell, we have
a combination of characters agreeing more nearly with those of the Caprimulgi than with
those of any other group. The pterylography also points to the same conclusion. Again,
as in the Caprimulgi, the skull is often extremely pneumatic, e. g. Strix, Asio, Photodilus ;
whilst the peculiar form of the lachrymal of the Striges is met with elsewhere only
among the Caprimulgi. The same may be said of the suppressed lachrymo-nasal fossa
and the basipterygoid processes. The general character of the sternum is also Capri-
mulgine, though, strangely enough, the pelvis is most remarkably Accipitrine. The
form of the anterior palatal fossa is peculiar to the Owls.

A comparison between the skull of avery young nestling Syrniwm or Bubo with a skull
of similar stage of development of Steatornis will reveal some striking similarities of
structure which will still further aid in establishing the Caprimulgine theory of origin of
the Striges.

It will be remembered that in the first part of this memoir, in which the pterylologica
characters were dealt with, this group seemed to fall naturally into two families, the
first containing the Barn-Owls only, the second all the remaining forms; these were
further divisible into two sub-families—the Asioninee and the Nyctaline.

Judging by osteological characters alone, the main division into families still holds good,
but it would appear to be necessary to recognize about six sub-families belonging to the
Asionide, instead of two. This subdivision, it should be remarked, would be founded on
the characters of the skull only, the axial and appendicular skeletons being remarkably
alike in all the members of the family.

The six sub-families would be :—

1. Asionine.
2. Photodiline.

MORPHOLOGY OF THE OWLS. 39

3. Bubonine.

4. Syrniine=Syrnium, Scotiapex, Ciccaba.
5. Nyctaline=Speotyto, Gymnasio, Nyctala.
6. Surniine=Surnia, Athene, Glaucidiwm.

I do not, however, propose to supersede the classification I originally suggested by the
above. It is set down here merely for the purpose of contrasting the results given by
the two systems. It is almost certain that the outcome of a study of the myology,
convolutions of the intestines, and other characters will suggest yet other combinations.
Obviously, therefore, it is best to wait until a survey can be made of all the factors, before
really sound results can be obtained.

XII. Ky to tHE FAMILIES AND SPECIES.
A. Sxunu. (Plates 1 and 2.)

Upper jaw markedly hooked ; nostrils holorhinal and impervious; with functional basipterygoid
processes ; lachrymal without supraorbital process and placed far forwards so as to encroach upon the
lachrymo-nasa! fossa ; lachrymo-nasal fossa greatly reduced, partly by the encroachment of the lachrymal,
and partly by the great upward development of the maxillo-palatine processes; palate with a large
anterior palatine vacuity; desmognathous, the bridge being formed by ossification of the alinasal
cartilages forming the floor of the anterior olfactory chamber ; maxillo-palatine processes not meeting in
the mid-ventral line ; parasphenoidal rostrum very short.

A. Orbit very small; interorbital septum of great thickness; vomer large, inflated,

fusiform, and fused with palatines ; maxillo-palatines relatively small ; lachrymal

of great size, subconical or subquadrate, thrust far into the lachrymo-nasal fossa,

and leaving only a small passage for the lachrymo-nasal duct; roof of skull dome-

shaped, with a deep median groove, and highly pneumatic ; palatines fused one

with another in the middle line behind the vomer; nasal septum deeply notched

along its postero-dorsal border, the notch leading into the cranio-facial fissure ;

floor of anterior nasal cavity perforated; palatines fused with one another in the
mnie slimesbehindstHesvOMery sf 5c. ss) o jfeul eu sue asl die sis ee) SIDRIGIDA,

B. Orbit moderately or very large; lachrymal columnar ; interorbital septum thin *;
antorbital plates thin +; nasal septum without a notch posteriorly . . . . . ASIONIDR.

a. Orbit relatively small; tympanic cavity large, well-defined, the ‘ post-

cavernum ” rising upwards to level of supraorbital process ; postorbital

processes small, projecting obliquely from orbit, and arising in front of and

below level of squamoso-parictal tympanic wing; conspicuous supraorbital

processes ; frontal between supraorbital process and base of squamoso-parietal

wing with strongly bevelled free edges forming a shelving supraorbital plane ;

nostrils very long, and with irregular inferior border ; vomer relatively large ;

antorbital plate very long, nearly reaching quadrato-jugal bar ; interorbital

syiien Vee! oo 3 oS |; rn.

* Asio and Photodilus only among the Asionidw have a thick interorbital septum, but the peculiar form of the
tympanic cavity and of the supraorbital region render the skulls of these two genera perfectly distinguishable from
that of Striw.

+ In Photodilus the antorbital plate is thick, and resembles that of Stria.

40 MR. W. P. PYCRAFT ON THE

b. Orbit very large ; interorbital septum thick ; antorbital plate spongy ; roof of
skull arched with supraorbital processes ; postorbital processes of great size,
projecting far outwards ; tympanic cavity extending upwards as a narrow gorge
between the postorbital and squamoso-tympanic process (= Bae

temporal fossee wanting ; vomer vestigial. . . . . 2 + es Photodelust

c. With a small, ill-defined post-cavernum ; postorbital processes es laterally
projecting, and with the free edge passing into the supraorbital rim of the
frontal, which is sharply defined.
a’. Remains of vomer blade-shaped; palatines with mesial borders of posterior
ends straight and meeting in middle line; pterygoids sharply truncated,
expanded distally and touching the parasphenoidal rostrum; maxillo-
palatines rising upwards so as nearly to obliterate the lachrymo-nasal fossa ;
lachrymal small, not reaching the quadrato-jugal bar . . . . . . . . Ketupa.
b’. Vomer vestigial ; lachrymal reaching the quadrato-jugal bar.
a, Palatines with mesial borders of posterior ends hollowed, enclosing a
small space caudad of the vomer shut in by the articulation with
pterygoids ; with large triangular supraorbital processes ; lachrymo-
nasal fossa extremely reduced. . . . . ye eh ot lo yh ltemeieel a eLS Onze
. Palatines with strap-shaped processes on Pe hence keeping articular
ends apart.
a". Fronto-nasal region of interorbital roof widest ; lachrymal very large;
temporal fossa never bridged . . . . . . Benet es ae. eo UHH.
b/”, Widest portion of interorbital region of roof eadad of vestigial supra-
orbital processes.
a‘. Temporal fossa frequently bridged by bar of bone from squamosal wing
CLO. GSS ch to Sh ely Bee ng Bei oo! SRDS:

6*. Quadrato-jugal with a triangular process below postorbital process. . { Ninoz.

Sceloglaux.

c’. With prominent, blunt, supraorbital processes overarching middle of orbit ;
frontal laterally constricted behind supraorbital processes so as to cause

postorbital processes to stand out prominently; palatines with pie
Syrnium.
Ciccaba.
d’. Tympanic cavity closed above by junction of squamoso-tympanic wing with

space caudad of yvomer; vomer vestigial

postorbital process.
a', Tympanic chamber symmetrical.
a. es fossa large.
. Supraorbital processes long and pointed ; nostrils nearly circular . . Speotyto.
a Supraorbital processes vestigial; nostrils oblong. . . . . - « Gymnasio.
6". Tympanic cavity markedly unger temporal fossz estiial =) a) Nycrala:
e’. Supraorbital processes long and slender; quadrato-jugal with a triangular
process below postorbital process; tympanic cavity open above; ramal
vacuity of mandible large.
a’, Squamosal wing of tympanic with broad, squarely truncated outer border,
not projecting beyond postorbital process ; pterygoid long, slender, and
straight ; tympanic cavity comparatively shallow, not rising to the level
of the floor of the temporal fossa; palatines wide apart caudad . . . Surnia.

MORPHOLOGY OF THE OWLS. 41

6’, Squamosal wing of tympanic with broad rounded lateral border, projecting
considerably beyond base of postorbital processes ; pterygoid sigmoidally
curved ; tympanic cavity very large and bullate, extending upwards
and backwards to underlie the floor of the temporal fossa ; palatines
nearly touching caudad. . . . . . POR 'd,| thoes Carine:
', Squamosal wing of tympanic projecting be slightly or not at all Heyord
baselotmpostorbinalyprocessi; . . «© . . » « =o sieummeemine = (Glaucwlium*.

B. VERTEBRA.

All the presynsacral vertebre are free and heteroccelous; the last thoracic is included in the synsacrum.
Innominates fused with synsacrum. The centra of the thoracic bear more or less conspicuous pneumatic
apertures opening beneath the transverse processes. Only the 2nd to 5th and the 14th bear neural
spines. The cervical ribs of all save the 13th and 14th vertebrie are very short or vestigial ; the 13th
and 14th ribs are free, and the 14th may bear uncinates ; the catapophyses never meet to form a canal.
The fovea lumbalis ischiadica and pudendalis are all well defined, and thereby the vertebral column
may be readily distinguished from that of the Caprimulgi; but there seems to be no character of
universal application by which the vertebral column of the Striges may be distinguished from that of
the Faleoniformes, when birds of a similar size are being compared. The distinctions between the
vertebre of the two families of the Striges and the various genera thereof are very slight.

a. Neural spines high, sloping forwards and interlocking.
a’, Transverse processes of thoracic vertebrae with anterior and posterior angles
produced into lateral spinous processes; hypapophyses long, not extending
beyond 2nd thoracic.

a. Preacetabular ilium with dorsal border nearly straight ; ventral border deeply
emarginate ; preacetabular width nearly equal to width across: anti-
BVOGHANKCT: (MMIC TS UG |. eI Res, wed ces cas ASO:

6”. Preacetabular ilium with dorsal border nearly straight; ventral border
gently hollowed ; preacetabular width much less than width at anti-
UAOCNRINSE eee coy ye) ee AO COR. oo a ODO Pin San Circe CMEC? 5171008

c', Preacetabular ilium with strongly arched dorsal crest; ventral border
deeply hollowed; preacetabular width equal to width across anti-

PO CHAT ict ners ts ees: «pte od tuts bela fo) ool on MCDA.
b'. Transverse processes without spinous lateral angles; hypapophyses long,
extending to 3rd thoracic ; lumbar vertebree 4; lumbar parapophyses 1-
vestigial ; planum coccygeum wanting . . . . .« . « + « «+ + « « Carine.
b. Neural spines high, nearly vertical and feebly interlocking.
Hypapophyses short, not extending beyond the 2nd thoracic; 4 lumbar vertebrae;
roof of fovea pudendalis not obstructed by parapophyseal processes ; a con-
spicuous planum coccygeum . . . 2 : of ee <a Syrnium.
c. Neural spines high, not interlocking ; Abhasonhyed shane not rieridinis heya
iimchonacio-moulumbarvertepre . . . « . « bee & os 5 ss « « Ninow:
d. Neural spines low, interlocking.
a’. Hypapophyses long, spine-like, extending back to 3rd thoracic ; planum coccy-
geum indistinct.

* For remarks on Glaucidium, see p, 36.

SECOND SERIES.—ZOOLOGY, VOL. IX. 6

MR. W. P. PYCRAFT ON THE

aevAvlumbar vertebra . « ... > cane aioe Clave
Megascops.
GH BY ie 5 Po cy SG |? Ao fone enone Nyctala.
b'. Hypapophyses cbsolete, or 1-2 only ; lumbar vertebree 4; fovea pudendalis with
intervertebral perforations . . . . ... é Rete tae sO | Strix.
e. Neural spines low, not interlocking ; hypapophyses ing: spine-like, extend
back to 2nd thoracic.
a’. 4 lumbar vertebre ; canales ileo-lumbales opening near the middle of the strongly
RIO Gg od 0 0 0 oO 0 o 0 6 Spiele, roe Surnia.
Canales ileo-lumbales opening near the anterior aa of the crest of the pre-
acetabular lium . . . . Cus sesanccn SAIL
b'. 38 lumbar vertebre. . . . . .« Scops.

C. Srernum anv Pecrorat GirDLe.

Corpus sterni large, with the hinder border always notched ; carina well developed, extending backwards
to the extreme end of the sternum; anterior lateral processes small ; antero-lateral border of sternum
marked by a deep scar for the sterno-coracoideus, which extends backwards to the level of the last sternal

rib; coracoid grooves but slightly overlapping ; coracoid with a large processus procoracoideus ;

furcula U-shaped, and lacking a hypocleideum.

A. Sternal plate with a single pair of notches posteriorly ; furcula non-pneumatic,
articulating with antero-ventral angle of carina by a broad facet; distal end not

articulating with acrocoracoid by a special facet . . . . . s+ s+ ws Srricip#£.
B. Sternal plate with a pair of notches posteriorly ; fureula pneumatic, not articulating
with the antero-ventral angle of the carina by a facet; distal end articulating
with the acrocoracoid by aspecial facet . . . . 6 6 Speers : ASIONIDE.
. Outer pair of notches relatively small, but eveeaiinn’ the pth of the keel . Asio.
; Outer pair of notches obsolete. . . . - . crass 38 - . « Huhua.
c. Length of coracoid equals length of sternum fron its anterior border to the
base of the inner notch.
Bubo.
a'. Coracoid with large processus lateralis basalis . . . . . . . J Ketan
Surnia.
b’. Coracoid with a small processus lateralis basalis . . . . . . Speotyto.
Scops.
Ninoz.
d, Length of coracoid equals length of carina. . . . . ... » is
Athene.
Glaucidium.
; Syrnium.
e. Spina externa of sternum wanting.
a’. Ventral liplofcoracoidygroove entire). i). i-n cin nnn nn nn Te Nyctala.
b!, Site of spina externa marked by notch . . . «2 ee + 6 + ee + Gymnasio.

MORPHOLOGY OF THE OWLS. 43

D. Petvic Grrpe *.

Innominate fused with synsacrum ; preacetabular ilium very long; pectineal process wanting; the
preacetabular ilium with a deeply emarginate inferior border; the dorsal plane of the postacetabular
ilium continued forwards and outwards to form a conspicuous shelf overhanging the acetabulum ;
ischium continued backwards into a point along the pubis; pubis always complete ; an iliac recess
always present.

KK. Pecrorat Lins.

Humerus with a moderately large, rounded, pectoral crest, a small ectepicondylar, and large, swollen
entepicondylar tuberosities, a deep linguiform impression for the brachialis inferior ; the sulcus transversus
obsolete, the impression for the pectoralis occupying nearly the whole of the pectoral crest, and the
impression of the deltoideus major extending for a considerable distance down the shaft.

Forearm and hand non-pneumatic. Radius bearing a slender bony arch or traces of it, on the second
fifth of the postaxial border, and a sesamoid at its distal end.

Manus with the proximal end of Me. III joining Me. II distad of the articulation of the pollex with
its metacarpal ; Ph. 1 of D. II with a deeply emarginate postaxial border.

F. Prenvic Lis.

All the bones non-pneumatic ; ecto- and ento-cnemial crests not feebly developed ; a tibio-tarsal
extensor bridge wanting ; hypotarsus simple; Mc. I very short.

A. Phalanx 1 of D. II short; Ph. 2 of D. III short; tarso-metatarsus with a strong

extensor bridge. . . 40 3G te 46 BAEC . . ASIONIDs.
B. Phalanx 1 of D. II lots Ph. 2 of D. Ur long dereepinetateindl extensor bridge
Rae 5 SS es we ee ee ee ww ef} SURIGID RE.

List OF THE MORE IMPORTANT WORKS AND PAPERS REFERRED TO
OR CONSULTED.

1. Avotrnrt, H.—‘‘ Ueber Variationen der Spinalnerven und der Wirbelsiule anurer Amphibien,”

Morphol. Jahrb. xxv. (1896).
. Bateson, W.—Materials for the Study of Variation. 1894.
Bepparp, F. E.—Siructure and Classification of Birds. 1898.
. Bepparp, F. E.—* On Photodilus badius,” Ibis, 1890.
. Bepparp, F, E.—“ On the Classification of the Striges,”’ Ibis, 1888.
. Cottett, C.—‘‘ On the Asymmetry of the Skull in Striz Tengmalmi,” P. Z. S., 1870.
. Coxe, F. J—“Some Variations in the Spinal Nerves of the Frog, with a Tae on an Abnormal
Vertebral Column,” Trans. Liverpool Biol. Soc., vol. xv. (1901).
8. Cours, E.—Key to N.-American Birds.
g. D’Atton.— De Strigum Musculis Commentatis. Halis, 1837.
10. Evans, A. H.—Birds. 1899.
11. Gavow, H.—‘On the Evolution of the Vertebral Column of Amphibia and Amniota,” Phil. Trans.
Roy. Soce., B. vol. elxxxvii. (1896).
12. Gavow, H.—Bronn’s Thierreich, Bd. vi. Végel, 1891. (Anatom. Theil.)
13. Gapow, H.—Bronn’s Thierreich, Syst. Theil. 1893.

SOM PW D

* The variation of the pelvis among the Striges is so slight that workable characters for distinguishing the
Families and Genera cannot be found,

MR. W. P. PYCRAFT ON THE

. Hopeson, B. H.—Journ. Asiat. Soc. Bengal, vol. vi. (1837) p. 362.
. Howes, G. B.—* On some Abnormalities of the Frog’s Vertebral Column,” Anat. Anzeig., Jahrg.

1886.

. Howes, G. B.—“ Notes on Variations and Development of the Vertebr and Limb Skeleton of the

Amphibia,” P. Z. S., 1893.

. LypexKker, R.—Cat. Foss. Birds Brit. Mus., 1891.
. Mitnr-Epwarps, A.—Recherches pour servir 4 Histoire des Oiseaux Fossiles de la France, vol. i.

(1867-68).

. Mitnn-Epwarps, A.—‘“ Observations sur les Affinités zoologiques du Genus Photodilus,” Nouv.

Arch. Mus., 2° sér. t. 1. (1878).
Nuwron, A.—Dictionary of Birds. 1896.

. Parker, G. H.—“ Variation in the Vertebral Columns of Necturus,’ Anatom. Auzeig., Bd. xi.

(1896).

. Pecx.— Variation of Spinal Nerves in Caudal Region of Pigeon,” Journ. Morphol., vol. ii. (1893).
. Ripewoon, W. G.—“ On the Development of the Vertebral Column in Pipa and Xenopus,’ Anatom.

Anzeig., Bd. xii. (1897).

. Suarre, R. B.—A Review of Recent Attempts to Classify Birds. 1891.
. SHarpz, R. B.—Cat. Striges Brit. Mus., vol. i. (1875).
. SHarpe, R. B.—Hand-list of Birds, vol. i. (1899).

EXPLANATION OF PLATES 1 anp 2.
Explanation of letters.

a.s.=alisphenoid. no.p.= (error for @.op.).
ao.p.=antorbital plate. n.s.=nasal septum.
a.p.v.=anterior palatal vacuity. p-= parietal.
b.pt.=basipterygoid process. pa. =palatine.
b.s.=basisphenoid. po.p.=postorbital process.
col.=columella. pme.=premaxilla.
c., oY cor.=coronoid. pro.=pro-otic.
d.=dentary. pt.=pterygoid.
e.g.=eustachian groove. q-= quadrate.
exo.=exoccipital. qJj-=quadrato-jugal.
f.=frontal. r.t.a.=recessus tympanicus anterior.
ff.=floccular fossa. Miss 3 posterior.
h.pt.=hemipterygoid. s.d.=supra-angular,
io,s.=interorbital septum. $.0.=supraoccipital.
1.=lachrymal. so.p.=supraorbital process.
l.pt.= (error for b.pt.) sg.=squamosal.
map.=maxillo-palatine. $q.0.wW.=squamoso- occipital wing.
m.int.=meatus internus. t,f.=temporal fossa.
mes. =mesethmoid. v.=vomer.
n.=nasal,

A

MORPHOLOGY OF THE OWLS. 45

PuatTE 1.

Figs. 1-7, dorsal aspect of adult skulls ; 8-14, palatal aspects of adult skulls ; 15 and 16, of young skulls.

Fig. 1.

2

Fig. 1.

Skull of Bubo capensis, showing the forward position of the vestigial supraorbital processes of
the frontal.

. Skull of Budo perspicillatum, to show the large size of the supraorbital processes.
3.

Skull of Nyctala Tengmalmi, showing the marked asymmetry of the skull in the position of the
postorbital processes.

. Skull of Speotyto cunicularia, showing large postorbital processes and the temporal fossa

bridged by the superior angle of the squamoso-occipital wing of the tympanic fossa.

. Skull of Photodilus badius, showing the great size of the postorbital processes and the backward

position of the supraorbital processes.

. Skull of Strix jlammea, showing the absence of supraorbital processes, and the peculiar ‘ dome-

shaped” roof.

. Skull of Asio accipitrinus, showing an incipient supraorbital groove and peculiar postorbital

processes, to see which properly a comparison must be made with Pl. 2. fig. 2.
The interest about the supraorbital processes lies in the fact that they are developed by the
frontal bone, and not formed by the horizontal process of the lachrymal as in birds generally.

. Palate of Bubo capensis, showing large maxillo-palatines and a small vomer. The palatines do

not quite meet in the middle line.

. Palate of Bubo perspicillatum, the vomer has become suppressed ; and the palatines meet in the

middle line.

. Palate of Nyctala Tengmalmi, showing reduced maxillo-palatines ; a vomer; and the great width

of the base of the parasphenoidal rostrum.

. Palate of Photodilus badius, showing great width of parasphenoidal rostrum ; vestigial vomer ;

and large maxillo-palatines, and lachrymal.

. Palate of Strix flammea, showing peculiar shape of vomer ; great width of parasphenoidal rostrum

at base ; and the peculiar shape of the maxillo-palatines.

. Palate of Asio accipitrinus, showing swollen parasphenoidal rostrum; the palatines separated

one from another posteriorly ; and large maxillo-palatines.

. Palate of Speotyto cunicularia, showing maxillo-palatines and absence of vomer.
. Dorsal aspect of skull of Syrnium aluco, showing sutures.
. Dorsal aspect of skull of Speotyto cunicularia, showing sutures.—Note the great difference in

the form of the squamosal in the two skulls.

PLATE 2.
Lateral aspect of the skull.

Left side of Nyctala Tengmaimi, to show form of tympanic cavity.—Note the form of the
squamoso-parietal wing and contrast with that of right side. For the sake of convenience in
comparison the left side has been reversed.

la. Right side view of same specimen.—Note the form of the squamoso-parietal wing and its

2.

3.

relation to postorbital process, and contrast with fig. 1.

Skull of Asio accipitrinus, showing the enlarged tympanic cavity or “ cavernum ” and the peculiar
postorbital process.

Skull of Strix flammea, showing the reduced size of the orbit, large lachrymal, and the hgh
dome-shaped cranial roof.

SECOND SERIES.—ZOOLOGY, VOL. IX. 7

A6 MR. W. P. PYCRAFT ON THE MORPHOLOGY OF THE OWLS.

Fig. 4. Skull of Bubo capensis, showing the large postorbital process and a well-marked temporal fossa ;

the only skull in this series, it is to be noted, in which this is normally developed.

5. Skull of Photodilus badius, for comparison with Asio and Sériz.

6. Skull of Speotyto cunicularia, showing the large postorbital process ; a large triangular plate on the
jugal; and the temporal fossa bridged by the squamoso-parietal tympanic wing.

7. Skull of nestling Syrniuwm aluco, showing sutures. Compare the form of the squamosal, and its
relations to the parietal and frontal, with that of Speotyto, fig. 8.

7 a. Inner view of the same.—Note the position of the squamosal.

8. Skull of nestling Speotyto cunicularia, showing sutures.—Note the form and size of the
squamosal, and its relation to the frontal, and compare with fig. 7.

8 a. Inner view of the same.—Note the great amount of the squamosal surface visible from the
inside of the skull.

en A”

Pyerazt

H Grony, del. et lith,

Trans. Linn. Soc, Ser.2.Zoon.Vou.IX.Pl. 1.

West, Newman imp

OSTHOLOGY ortTrHr OWLS.

Pycratt Trans. Liyn. Soc. SER.2.Z00n.Vou. IX. Pl. 2.

H.Gronv. del et lith. West, Newman imp

Spi hOhOGY or THE OWLS

ee a il

él a

ae)

II. On some Points in the Visceral Anatomy of the Characinide, with an Enquiry into the
Relations of the Ductus Pneumaticus in the Physostomi generally. By WALTER
8S. RowntreEz, B.Se., F.L.S.

(Plates 3 & 4.)

tead 5th March, 1903.

Synopsis.

Page
PME HCUTAG HOMNVatNR Nar siete clot ois; cs trai sucha £0 Gils gre! a Pa -elerapevel searatiel cualtene chen cubionale Steuer e! eealelalavetate te a 47
The Alimentary Canal and its Appendages (including the Air-Bladder) ................ 52
The Ductus Pneumaticus and its relations in the Physostomi generally ................ 63
etter Ovamesrandicertain.ouler Swructures® aeryfarde ae iio saci e+ as oe ise ce ees cess Te
Note on a Parasite found in one of the Characinide, .............2--.-c0esceseteeess 76
MIDI aaa AN] UR CONGIUGI ON SMMNter ey ler tate e-ways, Peek a Mainratert sy whele oe? lees dh Sejsin ore, w ele start ee vielore 77
List of the chief Works consulted in the course of the Investigation .................. 79

INTRODUCTORY.

"THE notes here embodied represent an investigation, begun two years ago, in the course
of which I have had under examination a considerable number of spirit-specimens and
skeletons, representing some 55 species and 33 genera of the Characinidee. Access to
these has been permitted me by G. A. Boulenger, Esq., F.R.S., of the British Museum
of Natural History, of whose sympathetic assistance I cannot speak too gratefully. For
purposes of comparison I have also had under examination certain Cyprinoids, Siluroids,
Gymnotids, and other Physostomes, together with Polypterus, Protopterus, Acipenser’,
Amia, and Lepidosteus, for the two latter of which I am indebted to the kindness of
Professor Bashford Dean, of Columbia University, New York. The investigation owes its
initiation to Professor G. B. Howes, F.R.S., whose encouragement and helpful criticism
have been of the greatest value to me throughout.

Further, I gladly avail myself of this opportunity of expressing my indebtedness to
other friends for assistance rendered: to C. E. Stansfield, Esq., M.A., and to Fratilein
Stéy, for facilitating at various times the labour involved in searching the literature of
the subject ; and to the Rey. T. R. R. Stebbing, F.R.S., for kindly identifying for me a
rare parasitical crustacean mentioned in the text. The starting-point of my work has
been Sagemehl’s masterly papers on the skull of the Characinide, and on the accessory
branchial organ of Citharinus, published in the ‘Morphologisches Jahrbuch’ for 1885
and for 1886-7. I have also made large use of Boulenger’s ‘ Les Poissons du Bassin du
Congo’ and ‘ Matériaux pour la Faune du Congo’; of Giinther’s ‘ Catalogue of the Fishes
in the British Museum,’ vol. v.; of Cuvier and Valenciennes’s ‘ Histoire Naturelle des
Poissons,’ vols. xix. and xxii.; of Stannius’s ‘ Anatomje der Wirbelthiere,’ and of several
other works.

SECOND SERIES.—ZOOLOGY, VOL, IX. 8

48 MR. W. S. ROWNTREE ON THE

The Characinidee constitute a well defined yet widely diversified family of Teleostean
fishes, falling under the division Physostomi. They are entirely restricted to the fresh
waters of Tropical Africa and America, a distribution in itself sufficiently curious. - The
greater number of the five hundred odd species known are denizens of the Neotropical
region. No genus is represented on both sides of the Atlantic.

Amongst the more obvious and generally recognized characteristic features of the
family are :—

The fusion and modification of the first four vertebree, in relation to the Weberian
mechanism; a physostomous air-bladder, transversely constricted, as in the Carps, into
an anterior and a posterior chamber; usually an adipose second dorsal fin; a non-
protrusible mouth, the upper margin of which is commonly formed by both premaxilla
and maxilla; usually teeth on the premaxilla and the dentary, sometimes also on the
maxilla and the palatine; pelvic fins abdominal; pectoral fins set low on the body and
folding like the pelvic fins; body covered with scales, cycloid or etenoid ; head without
scales; parietals separated by a longitudinal fissure, or united by a sagittal suture;
opercular bones complete; symplectic present; a supraoccipital spine, often long;
suborbitals usually large; pyloric appendages present; barbels absent.

By the earlier writers on the subject the majority of the then known forms comprised
in this family were classed as Salmonoids, others as Clupeids. Johannes Miller *
(1843-4) was the first to separate these and to include them in the Characinidee, basing
this change on the possession by these forms of the Weberian apparatus. It was,
however, reserved for Sagemehl + (1884) to attach full weight to this character and to
unite together the families possessing it—the Cyprinide, Siluridz, Characinidee, and
Gymnotide—into one group, the “‘ Ostariophyses.”

The general effect of Sagemehl’s paper is the establishment of three main pro-

positions :—

(1) That the parts of the Weberian mechanism in the families named are homologous,
implying community of origin, not simply analogous, as previously held. This
conclusion he supports by arguments deduced from comparison of the shoulder-
girdle, the interparietal fissure and other cranial characters, the opercular
apparatus, and other structures.

(2) That the Characinide fall naturally into three distinct subfamilies, no one of which
can be regarded as being derived from another, but which have all branched off
from a common ancestral line. These he designates the Erythrinoids, the
Herbivorous True Characinidee, and the Carnivorous True Characinidz.

(3) That the Characinidee, and more especially the Erythrinoids, in numerous cranial
and other features, recall the conditions found in Ama calva and, in a less

degree, in other Ganoids.

* Joh. Miiller, “ Uber den Bau und die Grenzen der Ganoiden,” Abhandl. der Berl, Akad. d. Wissensch., 1844,
+ M. Sagemehl, “ Das Cranium der Characiniden,” Morph. Jahrb., Bd. x. 1834.

VISCERAL ANATOMY OF THE CHARACINID. 49

Sagemehl’s conclusions appear to have been drawn in the main from the examination

of the following twelve species :—
Macrodon trahira.

UM aalbantora is) 2) 2) Se ae Erythrinus uniteniatus.
Lebiasina bimaculata.
Herbivorous True Characinids. . . . . . Citharinus Geoffroy.

( Alestes dentex.
Tetragonopterus fasciatus.
5 melanurus.
Carnivorous True Characinids . . . ... 4 FA ; ate!
Anacyrtus gibbosus.
Hydrocyon Forskalit.
5 brevis.

Sarcodaces odoé.

L

The Amioid characters insisted on so strongly in the paper in question, some only of
which I can claim to have verified, I will attempt to summarize. They apply with
especial force to the Erythrinoids, but some more generally. Most have reference to
the skull. They are as follows :-—

The angle of inclination between the posterior face of the skull and the skull-roof.

The whole exterior form of the skull, its solid roof, its bony surface, its naked and
sculptured bone, its lateral shielding by the extended suborbital bones.

The naked dentigerous palatines, and the almost identical formation of the border of
the mouth. ,

Certain relations of the labyrinth and the recess in which it lies.

The relations of the suspensorium.

The presence of a median cartilage situated between the hypopharyngeal bones,
partially ossified and dentigerous in Amica, rudimentary in Lrythrinus, farther
reduced and fused in Macrodon, absent in the other Characinidze and apparently
in all other Teleosts.

The relations of the nasal bones with the premaxille and the ethmoid.

The form of the parietals and of the squamosals.

The relations of the mucous canals of the cranial bones.

The relations of the intercalare.

Certain relations of the occipital region of the skull and the associated nerves.

The development of a canal for insertion of the eye-muscles.

|
|

In all these characters the conditions existing in at least some Characinidee, usually
the Erythrinoids, are such, according to Sagemehl, as suggest a connection with those
found in Amia. The differences are, indeed, often obviously such as have been con-
ditioned by the development in the former of the Weberian apparatus. Nevertheless,
apart from the skull region, there appears to be but little in the comparative anatomy of
the two types to suggest close relationship ; and in the skull itself there is at least the
supraoccipital to be reckoned with as having no counterpart in Amia.

In any case, the hypothesis of a direct genetic relationship between Amia and the

S*

50 MR. W. 8S. ROWNTREE ON THE

Characinidee seems to be rendered untenable by the fact that, whereas the resemblance is
closest in the case of the Erythrinoids, the actually most primitive conditions in certain
particulars are to be found in the highly modified herbivorous forms—conditions, indeed,
the homologues of which have to be sought not amongst the Ganoids, but amongst the
more lowly Selachians. This fact appears also to negative the hypothesis which suggests
itself, that the 4Amia-like Erythrinoids represent the ancestral stock from which the other
two groups have diverged.

Sagemehl points out that the Citharinoids (if I may for brevity so style the herbivorous
forms) present more primitive characters than the Erythrinoids in the following points :—

(1) The greater extension of the cranial cavity towards the nasal region, and the
associated possession of long olfactory tracts.

(2) The less advanced ossification of the primordial skull, especially in the ethmoid region.

(3) The simple undifferentiated socket for the hyomandibular.

(4) The possession (in Citharinus at least) of two “ submaxillaries,” ossifications which
the author homologizes with the upper labial cartilages of Selachians. These are
not found in the other Characinid groups, nor in Ama, though they exist in
certain other Teleosts (Gymnotus, Perca, some Cyprinoids).

(5) The persistence, in Cétharinus and other herbivorous forms, of the epibranchial of
the fifth arch, standing in connection with the fourth epibranchial, and serving
to support the accessory respiratory organ. A similar vestige has been described
in a Clupeid. In the other Characinids, and in Amia, it is unknown.

With regard to the Carnivorous Characinidze (other than the Erythrinoids), Sagemehl
considers that the cranial conditions are entirely such as might conceivably have been
derived from the Citharinoid skull, but that the jaw-apparatus and the relations of the
hyomandibular suggest rather a connection with the Erythrinoids.

On the whole evidence, therefore, so far as the skull region is concerned, the conclusion
to be drawn seems to be that the carnivorous (non-Erythrinoid) forms present the most
advanced conditions of development, whilst the Erythrinoids and Citharinoids are more
primitive, but that these two latter groups do not show any nearer approach to one
another; in fact, that the three groups must have branched out independently from a
common stock.

The African form Sarcodaces, however, classed with the non-Erythrinoid carnivorous
group, presents, as Sagemehl insists, most striking detailed resemblances to the Ery-
thrinoids in its cranial and facial characters, though separated from them by the
possession of an adipose second dorsal fin, and by the absence of teeth on the palatine.

Moreover, the assumption that all the herbivorous forms are to be classed together—
that is, that the herbivorous habit has only originated once within the limits of the
family—would appear to be scarcely warranted by the examination of the single genus
(Citharinus) described by Sagemehl, though, it is true, he refers at times to others.

The three groups or subfamilies marked out by that writer can, therefore, it seems to
me, be only accepted as, at most, a provisional basis of classification,

My own observations upon this most interesting and perplexing family of fishes have

VISCERAL ANATOMY OF THE CHARACINID. 51

been, in the first instance, mainly directed towards the comparative examination of the
viscera. No detailed description of these organs throughout any considerable section of
the family at present, I believe, exists ; and it is perhaps hardly to be expected that such
plastic and adaptive structures should materially contribute to the elucidation of the
affinities of the group. Nevertheless, it has seemed to me that such an investigation
might not be altogether valueless; and amongst the results which I now make known,
pending the conclusion of further work, are some hitherto unrecorded relations which
seem to be of some interest and importance.

The following is a list of the species of Characinidze examined by me. Except in a
few instances, in which the skeleton only was available, they were represented by spirit-
specimens, for the most part fairly well preserved. With certain specimens it has only
been permitted me to make an abdominal incision, but with many a more complete
dissection has been possible.

For convenience the list has been arranged on the basis of Sagemehl’s classification,
but it should be pointed out, in using the terms “ carnivora” and “ herbivora,” that some
forms under the former head (Alestes for example) are clearly omnivorous in habit.

EryTHRINoIDs: no adipose fin.
(Tropical America.)
. Erythrinus uniteniatus.

. Macrodon trahira.
. Lebiasina bimaculata.

ew woe

. Pyrrhulina unifasciata.

True CHARACINIDE: CARNIVORA.

(Tropical Africa.) (Tropical America.)
5. Sarcodaces odoé. 26. Tetragonopterus abramis.
6. Hydrocyon brevis. 27. - multiradiatus.
de fA Forskalit. 28. sD fasciatus.
8 L goliath. 29. Se argentatus.
9. Bryconaethiops microstoma. | 30. Brycon falcatus.
10. Alestes nurse. 31. Chalcinus brachypomus.
ll. 4, Kotschy. 32. Chalcinopsis dentex.
12. » longipinnis. 33. Myletes brachypomus.
13. », adentex. 34, Anostomus fasciatus.
14. », macrolepidotus. 35. Leporinus Frederici.
15. Micralestes acutidens. 36. Nannosiomus lateralis.
16. Ay altus. 37. Salminus mazillosus.
Wize # Stormsi. 38. Serrasalmo piraya.
18. Petersius Leopoldianus. 39. Bs humeralis.
19. Eugnathichthys Eetveldii. 40. Anacyrtus microlepis.
20. ie macroterolepis. 41. i guatemalensis,
21. Phago Boulengeri.
Om); loricatus.
23. Paraphago rostratus.
24. Neoborus ornatus.

25. Ichthyoborus niloticus (besse).

MR. W. S. ROWNTREE ON THE

or
bo

True CHaracinip&: Hersivora.

(Tropical Africa.) (Tropical America.)
42. Distichodus niloticus. 49. Prochilodus lineatus.
43. & Antonit. 50. Curimatus dobula.
44. Nannocharax niloticus. syle 39 albula.
45, Xenocharax spilurus. 52. 5 Gilberti.
46. Citharinus Geoffroyi. 53. x cyprinoides.
47, x latus.
48. 5 congicus.
48 a. ies macrolepis.
48h. Citharidium Ansorgit.

THe ALIMENTARY CANAL AND ITS APPENDAGES.

The Stomach.

In Characinids the short gullet passes often insensibly into the stomach, but in some
cases the transition is marked by a sudden change of calibre or of wall-thickness. ‘This
latter condition I have observed more particularly in Distichodus, Serrasalmo, Lebiasina,
Alestes nurse, and Leporinus.

The stomach has the form of a more or less expanded sae, of variable size, bent sharply
on itself at the pyloric region. This flexure is usually ventrally directed (Pl. 3. fig. 1).
But I have found numerous exceptions. In Sarcodaces the flexure was dorsal in every
one of the nine specimens examined (fig. 2). In IWacrodon and Erythrinus also the
flexure was dorsal or leevo-dorsal (fig. 3). In all the sixty-three specimens of Jchthyo-

borus, Which happened to be examined for a special purpose, the flexure was directed to

the left side. In JZydrecyon Forskalii (several specimens) the stomach was flexed
towards the right side. In Nannocharax, Eugnathichthys, and Phago the flexure was
dorsal. The regularity with which these conditions were found was somewhat striking,
although one can scarcely attribute any great significance to them.

The anterior or cardiac chamber of the stomach is usually much more capacious than
the ascending or pyloric limb. It is, moreover, at least in some forms, highly distensible;
the lining membrane, which is often deeply rugose when the stomach is empty and
contracted, becoming quite smooth when the organ is distended with food. At the
same time the stomach-wall, which may in the contracted state appear quite thick and
fleshy, thins out and becomes so transparent that the contents become visible even in
detail. Thus in one instance (Sarcodaces), where an eutire Anabas was found in the
stomach, the species of the ingested fish could be easily identified without opening the
organ. The distension takes place longitudinally as well as transversely. The organ
may indeed stretch to the full limit of the body-cavity, although in its contracted state
not exceeding half that length. 1n the instance above referred to the length of the
distended stomach, drawn tightly over the contained fish, was 11 cms., the Sarcodaces
itself being only 25 ems. long. It is obvious, in view of the narrow and acutely
reflected pyloric limb, that the stomach must possess great digestive activity. Fig. 2

———

VISCERAL ANATOMY OF THE CHARACINID”. 55

shows the stomach of Sarcodaces contracted ; fig. 3, that of Macrodon distended ; both
are represented as having been opened.

The cardiac chamber is often prolonged beyond the pyloric flexure into a more or
less pronounced blind sac. This prolongation is large in Hydrocyon (fig. 1), still larger
in Salminus, and yet further exaggerated in Servasalmo. It is large also in Macrodon
(fig. 3) and Lrythrinus, in Bryconaethiops, and in Tetragonopterus abramis. In
Sarcodaces (fig. 2), Petersius, and at least some species of Alestes it is smaller. In the
other forms examined by me this sac can hardly be said to exist. This is most decidedly
the case with the herbivorous forms, the stomach and entire alimentary canal of which,
however, are altogether peculiar. The cardiac chamber is perforated near its anterior
end by the opening of the ductus pneumaticus, to which I shall refer later. I merely
wish at this stage to point out that in at least some cases the orifice of the duct seems
to be actually in the wall of the stomach, rather than the cesophagus, if one may judge
from the extension in front of it of the rugose folds. Such is the condition, for example,
in Lrythrinus and Macrodon, in Sarcodaces, Hydrocyon, Salminus, Ichthyoborus, and
many of the forms I have examined.

In other cases, however, in which a sudden transition is apparent from cesophagus to
stomach, the opening of the duct appears to be just at or behind the junction of the
two. This I have observed in Distichodus, Serrasalmo, Micralestes, Eugnathichthys,
and Xenocharax. In certain other cases—Citharinus, Lebiasina, Alestes nurse, and
Leporinus—the orifice is in the narrower and thicker part of the tract, which one is
inclined to regard as the cesophagus.

The ascending pyloric limb of the stomach is usually firmer and thicker-walled than
the cardiac chamber. This condition becomes greatly exaggerated in Citharinus and
the other herbivorous Characinidée ; so much indeed that whilst the cardiac chamber is
extremely thin and delicate, the pyloric limb assumes the character of a veritable
gizzard. The contents of the canal—vegetable matter and mud—are reduced after
passing through this gizzard to the condition of an extremely soft pulp. COvtharinus,
Distichodus, Xenocharax, Prochilodus, and Curimatus agree in this particular, as also
in others to be mentioned later. enocharax, however, in this as in other points
presents a lower stage of differentiation than the other forms just named, the difference
between cardiac and pyloric regions being less marked.

In the newly discovered form Citharidium *, which may be described, as regards its
external characters, as a Citharinus with ctenoid instead of eycloid scales, I have not yet
had the opportunity of observing these relations.

The Pyloric Ceca.
The stomach is marked off from the intestine, sometimes by a slight constriction,
especially in Citharinoids, but also by the pyloric appendages, which are constantly
present in considerable number in the Characinide.

* Boulenger, G. A., “Description of a new Characinid Fish discovered by Dr. W. J. Ansorge in Southern Nigeria.”
Ann. & Mag. of Nat. Hist., Ser. 7, vol. ix. 1902, p. 144.

or
=

MR. W. S. ROWNTREE ON THE

The conditions in my specimens have rarely permitted of accurate counting, and there
is reason to think that there is considerable variation even in individuals; but,
supplementing my own observations with those of previous writers, I may make the
following statement regarding the number of appendages commonly found in different

forms -—
IHGA 9 6 9 + 30 Cuvier & Valenciennes.
WMiacr0don nnn 60 (+) m 2 rows 5
Tebiasing) (inne 5 or 6 Fy
Pri 6 (5 right, 1 left) ee
Other Characinids . . 10 to 25, rising in Hydro- Boulenger.

cyon and Citharinus
to 35 or 40.

The appendages vary not only in number, but also in length and thickness, and in
their arrangement. ‘Thus, according to my observations, those of Hydrocyon are large
and fall into three groups, two opposing one another at the pylorus, the third posterior
to these and forming a sort of fringe to the intestine for about one-sixth of its entire
length (2°5 ems. in 14 cms.). In Sarcodaces the appendages are short and numerous,
and forma fringe to the intestine for about the first ninth of its course (1°5 ems. in
14 ems.). In Ichthyoborus they are numerous and long, but fine, and fringe the
intestine for nearly one-sixth of its length (1°5 to 2 cms.in 12 ems.). In Dacrodon they
are continued for about one-fourth or one-fifth of the length of the intestine (8-4 cms.
in about 18 cms.). In Déstichodus the extension of the cecal fringe is at its maximum
amongst the forms I have examined. In this fish, the appendages at the pylorus are
long, numerous, and fine, and are thence continued in an abbreviated condition as
a fringe or sacculated border for about 9 ems., which, however, is only about
one-seventh of the length of the intestine—very long in this and other herbivorous
forms. In Salminus they are large and numerous, extending about 2 cms. along
the intestine. In <Anacyrtus, Tetragonopterus, and Alestes I have noted that the
appendages were large. In Alestes macrolepidotus 12 and 14 respectively were counted
in two specimens. They apparently formed a single row, fringing the intestine for 2 ems.
In Leporinus they are long and set in two rows of 7 each. In Bryconaethiops I made
out two rows of rather long appendages—about 5 in each row. In NXenocharax they
appear to be much reduced in number, size, and area, being in fact quite inconspicuous.
In Anostomus they are rather long but not numerous; in Citharinus numerous, long,
fine, and restricted. In the other forms I have examined the appendages were in such
an unsatisfactory state of preservation that nothing can be confidently stated about
them.

Looking at the foregoing observations as a whole, I find myself unable to do more
than simply state the facts which have come under my notice. I may here, however,
point out the curious circumstance that Amia on the one hand, and the Cyprinide and
Siluridze on the other—families to which we are especially led to look for affinities with
the Characinidee—are entirely devoid of pyloric appendages, whilst the Gymnotide, as I
have verified, possess them.

EE ————— CC

eS

VISCERAL ANATOMY OF THE CHARACINID#. 55

The Liver.

The liver in the Characinidz presents much variation in form and size. Usually
three lobes are more or less distinct, but in some the middle lobe is not developed except
as a mere band embracing the cesophagus ventrally and connecting the right and left
lobes. In others, again, one of the two lateral lobes may be comparatively little developed ;
and in one form at least it is the middle lobe which is developed at the expense of the
others. These conditions, though apparently specifically fairly constant, I regard as being
mere adaptations to the exigencies of space and pressure in the abdominal cavity, and
therefore of but slight significance. The gall-bladder, in every instance in which I have
been able to identify it, was situated at the extremity of the right lobe, being thus
sometimes in quite the posterior region of the body-cavity. Its duct could in some cases
be traced along the right lobe to the middle liver, and thence to the duodenum, where it
became lost amongst the pyloric appendages.

The following are the chief variations I have observed in the form of the liver :—
Macrodon.—No middle lobe. Lateral lobes subequal and slender, extending to behind pelvic fins.

Left lobe rather the longer, and somewhat expanded at its extremity.
Lebiasina.—All three lobes long and slender, and nearly of equal length.

* Sarcodaces.—Lateral lobes slender, subequal, and extending beyond stomach. Middle lobe short and

conical. Left lobe slightly expanded at its extremity.

Hydrocyon.—Lateral lobes wide, ending bluntly, and of equal length, terminating just short of the
pyloric flexure. Middle lobe rather large, triangular, and extending to near the middle of the
stomach. ,

Salminus.—Right lobe extends beyond stomach. Left lobe pointed and only half length of right lobe.
Middle lobe little more then a bridge between the two, but somewhat drawn out toa point at
its left border.

Leporinus.—Liver small. Right lobe slender and rather shorter than stomach. Left lobe rather
shorter, tapering to a filament. Middle lobe short, triangular, also tapering to a filament.

Anostomus.—Liver small. Left lobe two-thirds length of stomach. Right lobe slightly shorter. Middle
lobe short and triangular. All pointed.

Serrasalmo.—Three lobes of moderate length. Lateral lobes subequal. Middle lobe shorter.

Bryconaethiops.—Body of liver a mere narrow band, drawn out into three long and very slender lobes.
Middle lobe very long, extending to end of body-cavity. Right lobe almost thread-like, extending
beyond stomach. Both slightly expanded at extremity. Left lobe the stoutest and shorter than
stomach.

Ichthyoborus and Neoborus.—Liver very small, consisting mainly of the somewhat thick left lobe which
ends about the middle of the stomach. Right lobe pointed, but very short. Middle lobe a narrow
bridge connecting the two.

Eugnathichthys—Lobes short, especially the right, which is broad and obtuse. Others subequal.

Phago.—All three lobes short.

Tetragonopterus argentatus.—All three lobes subequal and slightly longer than stomach. Right lobe
the shortest.

Tetragonopterus fasciatus.—Lateral lobes very long, extending almost to end of body-cavity.

Alestes Kotschyi.—Lateral lobes subequal, about length of stomach and moderately stout. Middle lobe
about half that length and triangular.

Alestes nurse.—Three lobes slender, subequal, and extending slightly beyond stomach.

SECOND SERIES.—ZOOLOGY, VOL. IX. 9

56 MR. W. 8S. ROWNTREE ON THE

Alestes macrolepidotus.— All the lobes very long, and subequal. Middle lobe the longest. Left lobe
the shortest.

Citharinus.—Liver very small. ‘Three lobes short, triangular, and subequal.

Distichodus.—Lateral lobes very short, but thick. Middle lobe drawn out as a narrow attenuated
ribbon, ending in a small bulbous mass quite in the posterior part of the body-cavity.

Xenocharax.—Liver small. Left lobe long and thread-like, extending nearly the length of the body-
cavity. Right lobe short, broad, ending bluntly. Middle lobe short.

Nannocharax.—Left lobe broad, blunt, and longer than stomach. Right lobe short and triangular, and
occupying usual position of middle lobe. Middle lobe small, triangular, and displaced to the
left.

Prochilodus.—Right lobe moderately long. Left lobe shorter and thread-like. Middle lobe reduced to
a bridge so narrow as not to be definitely traceable across the intervening space.

Curimatus.—Liver apparently very small.

The Intestine.

The intestine of Characinids is, in the carnivorous forms, including Erythrinoids,
short, simple, and comparatively uniform. From the pylorus it first runs backwards, then
at a point more or less short of the end of the body-cavity turns and runs directly forward
to about the level of the pyloric ceca, usually to the right of the stomach, and finally
proceeds straight back, parallel to its former course, thus forming a single elongated
loop. Its walls are usually thin—in spirit-specimens often fragile—but in Hydrocyon
apparently stouter. The terminal portion sometimes appears to be wider than the rest
of the course, but in other cases no difference is perceptible. The appearance naturally
depends much on the distension due to contained material. In the specimens of
Bryconaethiops examined by me the terminal segment appeared distinctly narrower than
the preceding part, which was unusually wide. The length of the intestine, reckoned
from the first pyloric appendages, is in these carnivorous forms approximately from two-
thirds to four-fifths that of the entire fish, including the tail. In four forms only—
Lebiasina, Leporinus, Anostomus, and Alestes macrolepidotus—out of the twelve for which
I have recorded both measurements has the intestine equalled the fish in length. In
the last-named the intestine was longer than the fish in the proportion of about 7 to 5.
But this fish appears to be mainly vegetarian. The others were Macrodon, Sarcodaces,
Hydrocyon, Alestes Kotschyi, Alestes nurse, Ichthyoborus, Tetragonopterus abramis, and
Tetragonopterus fasciatus. Very precise measurement of the intestine was often difficult ;
and it seems therefore better not to give the exact figures, especially as those obtained
from different specimens of the same species were not always in close agreement. All,
however, fell within about the limits stated, and for a general comparison the
measurements were sufficiently accurate.

Turning to the herbivorous forms—Citharinoids—very different intestinal conditions
from those above described are found to obtain. The intestine is long, sometimes
enormously so, and is packed into the form of a flattened spiral, consisting, it may be, of
several coils. The following are the measurements which I have thus far been able to
make in this subfamily :—

VISCERAL ANATOMY OF THE CHARACINID#. 57

Citharinus latus—16 cms. Intestine 80 cms. Ratio of intestine to fish 5:1.

Distichodus niloticus —23 cms. Intestine 60 (+) cms. Ratio of intestine to fish 3: 1 approximately.

Prochilodus lineatus.—16 cms. Intestine 45 cms. Ratio of intestine to fish 3:1 nearly.

Xenocharawx spilurus.—16'7 cms. Intestine 28 cms. Ratio of intestine to fish 2:1 roughly ; or more
nearly, 5:3.

In Curimatus dobula the intestine forms a beautiful and evidently very long coil,
almost mathematical in the precision of its concentric circles, and constitutes a most
striking object when the body-cavity is opened, looking like a small coil of cord. But it
is so fine—at any rate in the small specimens (8 ems.) on which my observations were
chiefly made—that I have not attempted measurement.

As in the stomach, so in the length of the intestine, Xenocharax evidences a lower
stage of specialization for the herbivorous habit than either Cétharinus or Distichodus. In
this fish, moreover, the intestine does not exhibit the uniformity of calibre which
characterizes the Characinids generally. Thus, in the specimen on which my observation
was made, the first 16 ems. were narrow, the next 6 or 7 ems. rather wider, and the
terminal 5 or 6 ems. very wide, resembling a second stomach in capacity. This latter
condition was something more than mere distension due to enclosed material. In
Distichodus also I have noted similar but less marked variations: the intestine being
rather wide for 3 or 4. cms.; then narrow, and fringed as above mentioned, for about
9 cms.; then narrow and plain for some 10 or 12 cms.; then apparently wider for a
long segment, which is intricately coiled; the terminal 8 cms. the widest of all. For
Citharinus my notes are less precise, but here too I have recorded a variation of width in
different parts of the intestine, the terminal 5 cms. being the widest. In Prochilodus
the intestine was of uniform width, as also in Curimatus, so far as could be seen without
minute dissection.

In these forms the intestine was found to be full of very fine soft mud, said to consist
largely of diatoms, though this 1 cannot assert from observation. The intestine is
thus of a dark brown colour. In Citharinus and Distichodus, possibly as a result of the
action of the spirit, the whole lining membrane of the body-cavity was infiltrated with
brown material which had apparently oozed through the intestinal wall. Ina radiograph
of Citharinus, taken under my direction, a curious effect is observable, the intestine
having been rendered opaque to the rays by the mud contained in it. In the carnivorous
forms subjected to the same treatment nothing of the kind resulted.

The Air-Bladder.

The air-bladder in this family of fishes is well developed, and resembles externally that
of the Carps. It is constricted into anterior and posterior chambers, the cavities of which
remain in communication through a narrow and extremely short tube. The bladder
lies immediately beneath the vertebral column, and extends usually from the diaphragm
to the posterior end of the body-cavity, where it often tapers to a point. It may,
however, be considerably shorter than the body-cavity ; and, on the contrary, in one
instance, to be referred to later, it is continued between the caudal muscles to the very

Q*

58 MR, W. 8S. ROWNTREE ON THE

root of the tail-fin. The posterior chamber, usually much the larger, communicates
with the alimentary tract through the ductus pneumaticus, which is often of considerable
length and stoutness. These more obvious characters are to be seen in PI. 3, fig. 1.

The air-bladder consists of two coats: the outer stout, fibrous, and more or less opaque ;
the inner thin, delicate, glossy, and transparent. In the anterior chamber the two coats
are non-adherent and easily separable. In the posterior chamber they are not separately
distinguishable. The inner coat in the anterior chamber is thickened ventrally by what
has the appearance of a narrow sheet of muscular tissue running longitudinally. In
front the air-bladder abuts against a stout transverse membrane which passes backwards
ventrally to the bladder as far as the constriction, where it merges into the outer tunic
itself, and ends. This membrane is secured dorsally at two points, to the centra of two
successive vertebree, the first attachment taking the form of a stout peduncle; also
laterally to the body-wall at two points about one-third of the distance from the mid-
dorsal to the mid-ventral line. At the second median attachment this membrane is
confluent with the true outer coat of the bladder, which is also attached to the body of
the next vertebra. Thus there are, in all, three median attachments, the last two only
of which involve the true coat of the bladder. ‘The first of the two coincides with the
insertion of the hindermost of the series of ossicles constituting the Weberian mechanism.
The description above given refers to Macrodon, in which my examination of these
relations went into greatest detail. But I have not noted essential differences in these
particulars in other forms.

The air-bladder lies outside the peritoneum, and thus in reality outside the body-
cavity. In the herbivorous forms Citharinus, Distichodus, and Prochilodus, and in these
only, this membranous partition, shutting off the bladder and kidneys from the other
viscera, has become tough and fibrous to such a degree as to be quite resistant to the
knife. It is also opaque, causing the illusion that the air-bladder is entirely wanting.
In the other herbivorous forms examined, Xenocharax and Curimatus, the membranous
partition is much more delicate, but still an advance on the condition met with in the
other Characinide. Here, again, it is noticeable that Yenocharar presents a lower stage
of differentiation than its allies Citharinus and Distichodus. What significance this —
particular modification can have is not clear to me. It is suggestive, however, of an
adaptation in some way either to the herbivorous or mud-eating habit as such, or else to
the ground-habit. It may be, and I am inclined to take this view, that it serves to
protect the air-bladder from pressure arising from the distension of the large mass of
coiled intestine. Itis obvious that such variations of internal pressure on the air-bladder
would interfere with its bathymetric sensibility.

Possibly to be connected with this feature is Boulenger’s * discovery that in these
three genera, and these alone so far as is known, the dorsal and ventral lateral muscles
diverge opposite the anterior chamber of the bladder, so bringing the organ into close
contact at that pomt with the external skin. This condition is suggestive of the
“lateral cutaneous areas” found in many Siluroids, and stated by Bridge and Haddon to

* ¢Les Poissons du Bassin du Congo,’ p. 199.

we ee ee

VISCERAL ANATOMY OF THE CHARACINIDZ. 59

be restricted to that family *, though by Boulenger apparently recognized also in certain
Cyprinoids. Whether this character has any physiological significance as a means of
increasing the sensibility of the bladder to external influences, or has simply arisen as a
consequence of certain lines of growth, appears to be doubtful. But, in any case, it is
worthy of notice as a further suggestion of affinity between the three herbivorous genera
Citharinus, Xenocharax, and Distichodus.

The posterior chamber of the bladder usually exceeds the anterior chamber in length,
in a ratio varying from about 2 to 2 A notable exception occurs in the case of
Serrasalmo, as will be shown later. The following figures represent approximately the
ratios in the forms in which I have recorded measurements :—

MCG UHORM st Pee Oe Prochilodus 3:1
CUNT alle one is aa 2 Anostomus ool
Alestes nurse 3:2 Ichthyoborus 4:1
Nannostomus 3:2 Neoborus 4:1
Sarcodaces . Dell Eugnathichthys 4:1
Lebiasina ell Phago 4:1
Leporinus ell Xenocharax 4:1
Bryconaethiops eel Alestes Kotschyi A: 1
Alestes macrolepidotus Pe Distichodus 5:1
Hydrocyon onl Citharinus 5:1

These ratios represent length, not capacity. They do not appear to carry any great
significance. Herbivorous forms, for example, are found at both ends of the scale. Even
two species of the same genus, A/estes, are seen to be similarly separated,; but Alestes
kotschyi is altogether peculiar as regards its air-bladder, and should not perhaps have
been included in the list given above. Perhaps the most that can be said is that in the
Erythrinoids the ratio is small; that in the herbivorous forms Yenocharaa, Distichodus,
and Citharinus it is large; and that in the allied genera Ichthyoborus, Neoborus, Phago,
and Eugnathichthys the ratio is a high one. Also it may be noted that in this, as in
many other points, Sarcodaces stands near the Erythrinoids.

In Alestes kotschyi a peculiar condition exists: the air-bladder extends backwards far
beyond the boundary of the body-cavity—as far, indeed, as the root of the caudal fin. In
this region it is slender and tapering, and lies between the caudal muscles to the right of
the interhzemals of the anal fin. Such was the position in the two specimens I examined,
but it is said to lie sometimes on the left side. Other members of the genus are stated
to present the same formation in different degrees.

In the genus Serrasalmo, in which the body-cavity is short and deep, the air-bladder
presents some curious features. It is much shorter than the body-cavity, and is unique
among Characinids, so far as my observations extend, in that the anterior chamber is not
only longer, but altogether larger than the posterior chamber. In Serrasalmo piraya,
indeed, the latter chamber is reduced so far as to appear like a mere conical cap on the
end of the former, which is several times its capacity. In Serrasalmo humeralis the

* «The Air-bladder and Weberian Ossicles in the Siluroid Fishes,” Phil. Trans. B. vol. 184, 1893, pp. 295-6.
+ Boulenger, ‘ Les Poissons du Bassin du Congo,’ pp. 148-9; also Cuvier & Valenciennes.

60 MR. W. S. ROWNTREE ON THE

disparity in size is less marked, though still considerable. The tube connecting the two
is not axial, as in most Characinide, but at the ventral border of the bladder. The
ductus pneumaticus arises from this tube so obliquely that it appears to spring from the
anterior chamber ; in reality its communication is as usual with the posterior chamber.
The condition in respect of the relations of these tubes is somewhat suggestive of that
met with in the Gymnotide, where, however, the two chambers of the bladder are much
farther apart. The posterior chamber in Serrasalmo is concave in front, fitting over
the rounded oviform anterior chamber in the manner of a ball-and-socket joint. The
margin of its concave front face is raised on each side into some five or six small czecal
appendages, corresponding to internal pouches. A thick sheet of muscle on each side of
the anterior chamber connects the outer tunic with the dorsal body-wall. Some of these
features are shown in Pl. 3, fig. 4, together with internal characters to be described later.
This peculiar air-bladder has been briefly described by Valenciennes, but only as regards
its most obvious external characters ; its marginal appendages and its internal structure
are not mentioned by this writer. It is shown partially dissected in fig. 4. In
Anacyrtus also, in which the body-cavity is short and deep, the air-bladder is peculiarly
shortened and rounded, its entire length being only one and a half times its depth. The
posterior chamber is, however, as usual, much the larger. Internally, also, there are
points of interest (Pl. 3, fig. 8).

In the diminutive Nannocharax niloticus the bladder is much reduced. It does not
extend more than half the length of the body-cavity. There is the usual division into
two chambers, of which the posterior is considerably the longer, but the cavities appear
to be largely obliterated. The anterior chamber is globular, with hard tough walls; I
believe, indeed, that it is partially ossified, but its diminutive size renders its condition
doubtful. The posterior chamber is narrowed almost to a filament. The ductus
pneumaticus, though fine, has its usual relations. In no other Characinid have J met
with so rudimentary a condition.

Internally, the walls of the anterior chamber of the air-bladder are plain and smooth
in all the Characinidee. With the posterior chamber, however, it is otherwise: its walls
are often ridged by longitudinal bands of a ligamentous nature, some of which may be
of considerable depth, so as to appear like septa dividing the cavity peripherally into
longitudinal furrows or pouches. Most of the septa are restricted to the anterior part
of the chamber, becoming rapidly shallower backwards, and finally disappearing after a
comparatively short course. There are, however, usually two, one dorsal and one
ventral, which extend the entire length of the chamber so as to meet one another at its
posterior extremity; but these also become shallower as they proceed backwards. The
others are more or less symmetrically arranged laterally, but in varying numbers and
varying degrees of development. In a few cases such bands as are visible are but slightly
raised, the wall of the chamber appearing nearly smooth. In only about two cases,
however, have I observed the dorsal and ventral median bands to be absent or much
reduced in length: in the Ichthyoborine the two chief lateral septa extend the whole
length of tne chamber, exactly like the dorsal and ventral pair, excepting that they are
somewhat less prominent.

VISCERAL ANATOMY OF THE CHARACINIDA, 61

It may well be that the structure above described serves to maintain this chamber of
the bladder, which is comparatively thin-walled, in a state of due distension, and to
cuard against its collapse. Such, indeed, one cannot but suppose, must be its effect.
But that it may also possess a vestigial significance is suggested by a further very
curious modification, which is met with in the genera Lebiasina and Erythrinus. In
these two forms, and in these only amongst the Characinid, the longitudinal septa are
connected for a certain distance by transverse septa, in such a way as to constitute a
true cellular air-bladder—a suggestion of Ganoid affinities which harmonizes with
deductions from the cranial characters. This structure, which is found only in the
anterior third or two-fifths of the chamber, has been described by Valenciennes for
Erythrinus as follows * :—‘ La vessie postérieure est conique, pointue; sa tunique
fibreuse est plus confondue avec la tunique interne, et on y remarque quatre brides
longitudinales: une supérieure, une inférieure et deux latérales. Le tiers antérieur de
cette seconde vessie offre des parois celluleuses ; les cellules sont déterminées par des
brides transversales nombreuses, serrées, paralléles entre elles, et perpendiculaires aux
grandes brides tangentes 4 la surface du cone. Ces brides elles-mémes sont réunies par
dautres plus petites, excessivement plus nombreuses et perpendiculaires aux transversales
que nous venons d’indiquer. Entre ces mailles on apergoit de nombreuses lamelles
entrelacées, auxquelles est due la cellulosité des parois de la vessie.”

Of the bladder of Zediasina this writer simply says that it resembles that of
Erythrinus. My own observations, however, lead me to state that the cellular structure
is distinctly more pronounced in Lebiasina than in Erythrinus. Figs. 5 and 6 (PI. 3.)
show that in the former the cellular part of the wall is of some thickness, materially
narrowing the lumen of the chamber, whilst in the latter the bands or septa are com-
paratively slightly raised, and do not appear to encroach appreciably on the central cavity.
It is remarkable that in Wacrodon, which otherwise so strongly resembles Hrythrinus,
and in Pyrrhulina, which seems rather to be related to Lebiasina, there are none but
the longitudinal septa (fig. 7). But in MJacrodon, at least, these latter are strongly
suggestive of cellular structure.

Regarding the resemblance in detail presented by this type of air-bladder to that
of Amia, I can express no opinion, and observers are at variance. Sagemehl + considers
the similarity to be closer than in the case of any other Teleost ; whilst, according to
Valenciennes }, it is not so great as in some species of Hemirhamphus.

In all the other Characinids, so far as they have been examined, cellulation does not
exist, transverse septa being entirely absent. Nevertheless, the longitudinal septa,
constantly present but presenting varying stages of development or degeneration, are,
as it seems to me, suggestive of a former cellular condition of the bladder. In the short
statement given below, the order represents, according to my observations, the degree of
development of the longitudinal septa.

*

Cuvier & Valenciennes, ‘ Histoire Naturelle des Poissons,’ vol. xix. p. 485.
Sagemehl, Morphol. Jahrb., Bd. x. 1884-5, p. 108.
Cuvier & Valenciennes, ‘ Hist. Nat. des Poissons,’ vol. xix. p. 493,

++ +

62 MR. W. 8S. ROWNTREE ON THE

In Serrasalmo, although, curiously and quite exceptionally, the dorsal and ventral
septa appear to be absent, there are some six lateral septa on each side which are of
considerable depth, and extend backwards quite to the middle of the chamber, dividing
the side-walls into a corresponding number of deep channels, continued at the anterior
margin into well-defined pockets or pouches which are conspicuous externally. Inter-
calated between these main septa are some nine much smaller septa on each side, causing
a further subdivision of the lateral space. I have not found any previous record of this
peculiar condition shown in fig. 4, though the external features of the bladder have been
described by Valenciennes *.

In Macrodon (P1. 3, fig. 7) the dorsal and ventral septa are large, and there are two strong
but shorter lateral septa with at least 80 smaller bands of very varying size intercalated.
The existence of lateral septa is denied by Valenciennes *, who only recognizes the two
median bands. In addition to the definite bands, the whole internal surface of the
anterior part of the chamber presents the appearance of closely-set but interrupted
longitudinal strive, apparently of the same character as the bands themselves.

In Prochilodus the dorsal septum is deep, standing out ‘boldly ; the ventral less so.
Both extend to the extremity of the bladder. There are also some five lateral septa on
each side, which are short and of moderate depth. The whole bladder is very large.

In Sarcodaces there are the two long median septa, one shorter lateral band on each
side, all well developed, and some three subsidiary septa on each side.

In Ichthyoborus, Neoborus, Paraphago, and Eugnathichthys, very closely allied forms,
the two lateral bands extend to the extremity of the bladder, like the median bands,
than which, however, they are less prominent. There are traces of about two other
very faint and short lateral bands. In their near relative—Phago—the two lateral
bands present an appearance which suggests a degeneration from the condition just
described ; they are unusually long, and each is continued to the extremity of the
bladder as a coloured streak, which is not the actual band ‘itself, but which implies its
former extension. In Alestes nurse and A. macrolepidotus the median bands are strong,
especially the dorsal, and there are three or four short lateral septa on each side.

In Bryconaethiops and Salminus two evtire median, and two shorter lateral, septa are
present, all well developed and of fair depth. In Hydrocyon, Tetragonopterus, Micralestes,
Anostomus, Xenocharax, Distichodus, and Citharinus the septa are as in the two forms
just previously mentioned, but are less distinctly raised.

In Leporinus and Curimatus the dorsal band is strong, the ventral apparently absent ;
two short weak laterals are present in Leporinus; none in Curimatus.

In Anacyrtus (Pl. 8. fig. 8) the condition differs from that of any of the foregoing. The
dorsal and ventral septa are almost obliterated, extending for only a short distance.
Vestiges of about five lateral bands on each side are seen running parallel to one another
for about half the length of the chamber, and scarcely perceptibly raised. The whole
lateral wall also shows a delicate network of fine, closely-set lines at right angles to one
another, which, however, are not perceptible to the touch, and do not appear to be

* Cuvier & Valenciennes, ‘ Hist. Nat. des Poissons,’ vol. xxii. p. 280.

VISCERAL ANATOMY OF THE CHARACINID. 63

superficial. In its first-mentioned characters this air-bladder faintly suggests that of
Serrasalmo, which it also resembles in the somewhat unusual thickness of its walls.

Upon the question whether or no the internal features of the air-bladder summarized
in the preceding pages possess any real significance bearing on the affinities or classi-
fication of this puzzling family of fishes, it is difficult to speak with confidence. Certain
facts, however, seem to be in some degree suggestive. Thus, of the Erythrinoids, two
forms—Erythrinus and Lebiasina—present a similar cellular structure, but in different
stages of development or degeneration, whilst Macrodon appears to follow suit after a
much greater interval. Of the fourth form—Pyrrhulina—I have no satisfactory
observation. Again, the members of the group Ichthyoborine agree, so far as they have
been examined, in a unique extension of the two main lateral septa. Also the allied
African mud-eating fishes—Distichodus, Xenocharax, and Citharinus —present conditions
similar to one another.

The resemblances above pointed out are sufficient to inspire some confidence in the
persistence and constancy of the structural features of the air-bladder ; and it seems not
unlikely that the examination of that organ in a larger number of forms might here and
there yield evidence of value as an assistance in classification. I do not, however, feel
justified in proceeding further with any such deductions from the slender material
before me. The groupings found in my description may be in some degree accidental.
Perhaps the most noteworthy points are :—(1) the isolated and aberrant conditions in
Serrasalmo (Pl. 3. fig. 4), and to a less degree in Anacyrtus (fig. 8); (2) the fact that, of
the African forms, Sarcodaces seems to stand nearest to the archaic Erythrinoids ; and (3)
the agreement of the Jchthyoborine in possessing two more completely developed lateral
septa than are found in any other forms.

In conclusion, it is worthy of mention that in Cyprinoids I have observed two
longitudinal bands in the posterior air-bladder, which are visible as streaks, but not
appreciable to the touch, the surface being quite smooth. This condition suggests a
further stage of degeneration from a primitive cellular air-bladder. A still further step
seems to be recognizable in the Gymnotidee, in which these bands are absent.

THE Ductus PNEUMATICUS AND ITS RELATIONS IN THE PHYSOSTOMI GENERALLY.

The ductus pneumaticus has already received a brief mention. There remain,
however, certain points in connection therewith which merit consideration.

The duct takes its origin in all the Characinidee from the posterior chamber of the
air-bladder, which is thus placed in communication with the alimentary tract in the
transitional region between the cesophagus and the stomach. In some cases the
opening appears to be in the stomach, in others in the cesophagus, and in others on
the border line between the two regions. This border, it is true, is sometimes difficult
to define ; but, on the other hand, there may be a sudden change of calibre or consistency
which one cannot but interpret as marking the transition, and this may be variously
situated in relation to the opening of the duct. :

In Physostomes generally the duct opens dorsally, and I am not aware of anything in
SECOND SERIES.—ZOOLOGY, VOL. IX. 10

64: MR. W. 8S. ROWNTREE ON THE

the literature of the subject to suggest that the opening is not median. It has, however,
long been known that in the Erythrinoids, or at any rate in the genera Hrythrinus,
Macrodon (Pl. 3. fig. 3), and Lebiasina (fig. 5), the duct opens not dorsally but laterally—
in fact, in the middle of the left side. Valenciennes does not seem to have been aware of
this peculiarity, but it is referred to by Stannius *, and has been discussed at some length
by Sagemehl ¢. No observations recording anything unusual in the other Characinidze
seem to have been made. In fact it is stated that these forms do not present any
features of interest in this connection. This [I shall show to be incorrect; and it is
somewhat remarkable that Sagemehl] at least—who bases an argument of some importance
upon the conditions found in Erythrinoids—should have overlooked the true relations
existing in the other Characinids.

I have paid special attention to these relations in every form which has come under
my notice, for a priori it appeared unlikely that the Erythrinoids should stand absolutely
apart from their allies in this matter. The results are as follows :—In the Erythrinoids
Macrodon, Erythrinus, and Lebiasina the ductus pneumaticus opens into the middle of
the left side of the tract. This is seen in fig. 3, and equally distinctly in fig. 5. In the
fourth member of the Erythrinoid group, Pyrrhulina, the opening is dorsal, but well to
the left side. This condition I shall refer to as “ levo-dorsal.” Amongst the other
Characinids, out of 85 species examined by me with reference to this point, 25 were
found to have the opening of the duct very decidedly to the left side; in three others
the opening was only slightly on the left side of the mid-dorsal line; in the remaining
seven, owing to the condition of the specimens, I was not able to satisfy myself
absolutely, but im at least 5 of the number the opening appeared to be slightly to the
left. In no instance was the orifice on the right side, and in only two did it appear to
be median. Itshould be stated that in some instances only one specimen was available ; in
others, several.

In Sarcodaces the duct was found to open quite in the middle of the left side
(fig. 2), much as in the three first-mentioned Erythrinoids. This observation was made
on several specimens. ‘Thus this character, if it possess any significance, serves to
strengthen the suggestions yielded by the skull of affinity between these forms. In none of
the other Characinids examined does the opening lie so far to the left. Jchthyoborus (fig. 9),
of the internal anatomy of which no description exists, stands next to Sarcodaces in this
relation ; but the opening of the duct is more dorsal. Neoborus and Hugnathichthys are
closely similar. It is also well to the left in Hydrocyon (brevis and forskalit) ; in Alestes
(nurse, longipinnis, and macrolepidotus) (fig. 10); in Micralestes stormsi; in Tetragono-
pterus (fasciatus and argentatus); in Petersius, Bryconaethiops, Salminus (fig. 11),
Anacyrtus, Anostomus, and Leporinus ; in Prochilodus (fig. 12), Xenocharax, Distichodus
(untonii and niloticus) (fig. 13), and Citharinus macrolepis (fig. 138 a); and in Serrasalmo
(humeralis and piraya).

* Stannius, ‘ Anatomie der Wirbelthiere,’ 1854, p. 224.
+ Sagemehl, “ Das Cranium der Characiniden,” p. 108.

VISCERAL ANATOMY OF THE CHARACINIDA, 65

In the following forms the duct opens, according to my observations, only slightly to
the left of the median line :—Alestes Kotschyi, Citharinus (latus and Geoffroyi).

Finally, with regard to Zetragonopterus (abramis and multiradiatus), Curimatus,
Nannocharax, and Phago, I can only state, with some hesitation, that the opening
appeared to be slightly to the left; and with regard to Paraphago and Wicralestes
acutidens, that it appeared to be in the mid-dorsalline. In these latter instances, however,
either the size or the condition of the specimen before me was always an obstacle to an
accurate determination. Viewing these results as a whole, I can state with confidence
that, in spite of some individual variations, the position of the duct is fairly constant
specifically, and is normally on the left side throughout the Characinidie, so far as my
observations have extended. The meaning of this asymmetric structural condition has
been discussed at length by Sagemehl, who, however, as already stated, only recognized
its existence in the Erythrinoid division. According to the view set forth by that writer,
this condition supports the argument for a primitively ventral air-bladder, and is
inexplicable on the hypothesis advanced by Boas * of the dorsal origin of that organ.
The matter is of sufficient importance to warrant a summary here of the argument set
forth by Sagemehl, and the fact that my own observations, as will be seen, have an
important bearing upon the question must be my justification for venturing to
restate it.

Discussing the homology of the air-bladder with the ventrally-lying lungs of higher
vertebrates, Boas contended that the primitively dorsal and unpaired organ must have
undergone longitudinal fission, together with its duct ; and that the now separate ducts,
with their orifices, must have travelled round the cesophagus until they merged ina
common ventral opening, producing the condition existing in Polypterus and in air-
breathing vertebrates.

Sagemeh! objects to this hypothesis on the following grounds :—

(1) The absence of any imaginable cause for the occurrence of the changes involved ;
whereas a wandering in the reverse sense would naturally result from the
acquisition of a hydrostatic function by a ventrally situated bladder.

(2) The absence of any satisfactory explanation of the conditions found in
Erythrinoids and in the Dipneusti, in which the single duct passes to one side
of the cesophagus.

(3) The relations of the lung-arteries in Ceratodus +, which arise on each side from
the fourth branchial arch, that from the left arch passing under the cesophagus
and following the ductus pneumaticus, which in the Dipneusti winds round to
the right. If the dorsal situation of the bladder were the primitive one, the
arteries on each side would pass directly backwards to their distribution, as
those on the right side actually do.

This last objection appears to be conclusive as regards the Dipneusti, and inferentially
for all dorsal air-bladders, although in Teleosts the bladder receives a different blood-

* Boas, “ Uber den Conus arteriosus und die Arterienbogen der Amphibien,” Morph. Jahrb., Bd. vii. 1881, p. 566,
+ Boas, ‘“‘ Herz und Arterienbogen bei Ceratodus und Protopterus,”’ Morph. Jahrb., Bd. vi. 1880,
167

66 MR. W. S. ROWNTREE ON THE

supply. It follows then, if the air-bladder originated as a ventral structure, that in
order to arrive at the Teleostean condition the bladder and its duct must have travelled
round the alimentary canal, to one side or the other, until finally both attained a dorsal
situation. On this hypothesis, the position of the duct in Erythrinoids—and, as I have
shown, in Characinids generally—becomes endowed with a peculiar interest as repre-
senting an incomplete stage in the evolution of the Teleostean air-bladder ; as indicating
the path along w hich the bladder and duct have travelled ; and as bearing witness to the
primitive character of the family of fishes under consideration.

It would appear, then, according to Sagemehl, and so far as the discussion has yet
proceeded, that the dorsal position of the air-bladder in Dipneusti and Teleostei must have
been arrived at in the two cases along two different routes, lying to the right and to the left
respectively, each of which may have had as its starting-point the condition which is still
persistent in Polypterus. It would seem that the two cases are further distinguished,
firstly by the fact that in the Dipneusti the actual orifice of the pneumatic duct has not
travelled more than a short distance, and secondly by the fact that in Teleosts the
bladder has changed its blood-supply from the branchial artery to the dorsal aorta.

It may be further pointed out that this interpretation of the more or less laterally
situated opening of the ductus pneumaticus in the Characinide would involve the
derivation of the entire family from forms more primitive than Ama or any of the
existing Ganoids, in all the families of which I find the position of the duct to be mid-
dorsal. A point referred to by Sagemehl in support of his argument is the fact that in
Polypterus the left sac of the air-bladder is decidedly the larger. Thus it only becomes
necessary to assume the complete atrophy of the already dwindling right sac.

Against this argument, we have at the outset the facts cf ontogeny, which seem, so far
as at present known, to represent the air-bladder as a dorsal outgrowth from the
cesophagus *. But this, it must be admitted, may conceivably arise from the elision of
antecedent stages.

The discussion of this question involves the consideration of the relations of the ductus
pneumaticus in the other families of Ostariophyses, and even in the Physostomi
generally. I have accordingly extended the investigation in that direction, with the
following results.

Of the Cyprinidze I have examined specimens from the five diversified genera :—

Leuciscus (L. rutilus & L. leuciscus) (Pl. 4. fig. 14),
Carassius (C. auratus),

Barbus (B. Bynni),

Varicorhinus (V. beso),

Catostomus (C. macrolepidotus) (fig. 15).

In all these forms the opening of the duct is leevo-dorsal, and, in some cases at least,
even far to the left of the median line. In fact, excepting that its position is usually

* Miklucho-Maclay, “Uber ein Schwimmblasenrudiment bei Selachiern,” Jenaische Zeitschr. f. Naturwiss.,
Bd. iii. 1868.

ne

=e

VISCERAL ANATOMY OF THE CHARACINIDA. 67

further forward, the condition is precisely that which is commonest amongst the
Characinide.

Thus far, the facts are in accordance with Sagemehl’s hypothesis.

In turning, then, my attention to the Siluride, I was prepared to find conditions
similar to those above described. To my surprise, however, investigation revealed a
precisely opposite state of things, for in these fishes the opening of the ductus pneu-
maticus is actually and unquestionably on the right side of the mid-dorsal line of the
esophagus (Pl. 4. figs. 16 and 17)—as far, indeed, to the right as it is in most Characinids
to the left.

There may possibly be some significance in the fact that, in all the Siluroids examined,
the pyloric flexure of the stomach was found to be lateral and to the left side. But,
after careful observation and consideration, I am unable to see that this modifies the
relations of the ductus pneumaticus with the cesophagus as above stated.

This statement is based upon the examination of forms taken from the following
eight widely divergent genera, more than one specimen in most cases having been under
examination :-—

Amiurus (A. nebulosus),

Clarias (C. lazera) (fig. 16),
Synodontis (S. gambiensis),
Malapterurus (M. electricus) (fig. 17),
Chrysichthys (C. auratus),
Siluranodon (S. auritus),

Schilbe (S. mystus),

Auchenoglanis (A. biscutatus).

In seven out of the eight the opening of the duct was decidedly on the right side. In
the eighth case—S?/uranodon—it appeared to be median; but in tbis case, and in this
case alone, it has to be stated that the condition of the viscera was unsatisfactory and
precluded a positive verdict.

Bridge and Haddon *, in their detailed study of the air-bladder in the Siluroid fishes,
describe the median connection of the ductus pneumaticus with the bladder itself, but
make no reference to its precise relation with the esophagus.

Of the Gymnotidee, the fourth family of those fishes which possess the Weberian
apparatus, I have examined three species from two genera—

Carapus fasciatus (1 specimen),
Sternopygus carapus (2 specimens) (fig. 18),
Sternopygus virescens (3 specimens) ;

with the result that the pneumatic duct was found to open into the cesophagus in the
median dorsal line in Carapus and on the left side in both species of Stevnopygus (fig. 18).
The duct may thus fairly be described as having a tendency towards the left side.

* “Contributions to the Anatomy of Fishes: the Air-Bladder and Weberian Ossicles in the Siluroid Fishes,”
Phil. Trans., B. vol. 184, 1893.

68 MR. W. S. ROWNTREE ON THE

The following is a summary of my observations on the other Physostomi :—

1. Satmo. Salmo trutta (8 specimens) and Salmo fario (1 specimen).—Duct short ;
opens distinctly on the right side of the dorsal wall of the cesophagus (Pl. 4. fig. 19).

2. CornGonus. Coregonus albus (4 specimens).—Duct of moderate length; opens
nearly medially, doubtfully to the right side.

3. OsmeRuS. Osmerus eperlanus (3 specimens).—Duct rather long ; opening apparently
median.

4, ToyMaLuus. Zhymallus vulgaris (2 specimens).—Duct rather short ; opens slightly
to the right side.

5. Mormyripm. Petrocephalus bane (2 specimens) and Gymnarchus niloticus (4
specimens ).—In both these forms the short, wide duct opens into the cesophagus leevo-
dorsally. In Petrocephalus (fig. 20), the connection of the duct with the air-bladder is
also on the left side. In Gymnarchus (fig. 21), which, by the way, has been erroneously
included in the Ostariophyses by Bridge and Haddon in the paper previously referred to,

as also by Wiedersheim *, this is not the case, and the connection with the cesophagus is
more nearly median than in Petrocephalus.

6. Notoprerus. Notopterus kapirat (2 specimens).—'The duct opens into the
cesophagus far forward, and slightly but distinctly to the left side (fig. 22). It is
short and wide, and opens also into the left side of the air-bladder itself, which
is longitudinally divided in the abdominal region by a vertically complete median
septum. This septum has a free edge anteriorly, not quite reaching the transverse
membrane, which partially separates this part of the bladder from the unpaired
anterior chamber, and it is just on a level with this free edge that the ductus pneumaticus
opens. These relations, with others not now before us, agree with those described by
Bridge f as existing in Notopterus borneensis, with the exception that in that form the
pneumatic duct is described as being connected with the cesophagus in the mid-dorsal
line.

7. Exors. lops saurus (2 specimens).—The air-bladder opens by a wide orifice,
rather than by a definite duct, into the stomach at about the middle of its length, just
behind the pyloric flexure, and in the front part of the very long blind sac. In one
specimen the opening was in the mid-dorsal line; in the other, the stomach was
irregularly shaped, and it became a mere matter of opinion whether the connection
ought te be regarded as median or leevo-dorsal. Setting aside this doubtful case, we must
take the evidence of the other as pointing to a median position of the duct (fig. 23).

8. AtBuLA. <Albula conorhynchus (1 specimen).—In this fish, primitive in the
possession of a valved conus, the duct is longer than in Zlops, and opens into the
stomach much farther back, quite near, but not at, the extremity of the blind sac.
It is in the mid-dorsal line (fig. 24).

* «Elements of Comp. Anat.,’ translation by W. N. Parker, 2nd edition, 1897, p. 226,
+ “The Air-bladder and its connection with the Auditory Apparatus in Wotopterus borneensis,” Journal of the
Linnean Society, Zool. vol. xxvii, 1900, p. 507.

—?

~~

VISCERAL ANATOMY OF THE CHARACINID. 69

9. CLuPBA. Clupea sprattus and Clupea harengus.—As is well known, the duct is
fairly long and opens medially into the extreme posterior end of the blind sac (fig. 25).

10. Currocentrus. Chirocentrus dorab (1 specimen).—In this fish, marked out as
primitive by its intestinal spiral valve, the wide ductus pneumaticus opens medially into
the dorsal surface of the stomach rather behind the middle of the blind sac, which is of
enormous length (fig. 26).

11. Ganaxtas. Galaxias truttaceus (8 specimens).—The long duct opens far back on
the stomach, quite near the extremity of the short blind sac, much on the same level as
in Albula; but it is not median, being in fact very markedly displaced to the right
(fig. 27).

12. Esox. soa lucius (5 specimens).—The short duct opens far forward into the
dorsal wall of the cesophagus. It is nearly median, but in all the specimens examined
was just appreciably to the left side (fig. 28).

13. CYPRINODONTIDE. Cyprinodon calaritanus, Orestias Oweni, Fundulus robustus,
and Goodea atripinnis (2 or more specimens of each).—Diminutive size here presented
an obstacle, in most cases, to satisfactory observation, but in Ovestias the duct was clearly
made -out to be situated far to the right of the mid-dorsal line. Apparently this was
also the condition existing in Fundulus and Cyprinodon. The specimens of G‘oodea
were not in sufficiently good condition for a satisfactory observation.

14. Percopsipm. Percopsis guttatus (2 specimens) and Columbia transmontana
(1 specimen).—In both these forms the ductus pneumaticus opens into the cesophagus
far to the right side of the mid-dorsal line.

15. Hyopon. Hyodon alosoides (8 specimens).—The short, wide duct opens far
forward on the cesophagus, nearly mid-dorsaily, but, somewhat doubtfully, a little to the
right side.

16. ANGUILLIDZ. Conger conger (2 specimens).—In both cases I found the duct so
difficult to trace, that I cannot commit myself to any statement as regards its relations
with the alimentary canal.

17. Hatosavurus. Halosaurus macrochira (1 specimen).—The rudimentary air-bladder,
situated in the posterior part of the body-cavity, narrows in front into a band which is
continued as a thread-like ligament. This ligament, apparently solid, terminates in the
mid-dorsal wall of the stomach quite near to its posterior end, no/, as has been stated *,
in the esophagus.

Finally, in the three Ganoids Acipenser, Amia, and Lepidosteus, as I have verified,
the communication of the air-bladder with the cesophagus is in the mid-dorsal line.

In Acipenser the duct is of some length, but in Ama and Lepidosteus the bladder opens
directly by a slit-like orifice into the cesophagus.

* Giinther, in the Report on the Voyage of the ‘Challenger,’ “* Deep-sea Fishes,” vol. xxii. Zool. 1887, p, 232.

70 MR. W. S. ROWNTREE ON THE

The foregoing analysis may be thus tabulated :—

VENTRALLY. DORSALLY.
ino | nenina |

Creme itis to ane aoe Opening practically median.

| | Ge : pening on :

| | Bladder ven-| side. ened Opening on

tral: its left! Bladder oes! Inclined Inclined right side.
jsacthe larger. dorsal. to left. to right. i

| Communicating] Polypterus. | Dipneusti*.|| Mormyride. | Hsov. | Acipenser. | Hyodon. _| Salmo,

with esophagus || Notopteridee. Amia. Coregonus ? | Siluridee.

or the part | || Characinidee, | Lepidosteus. Thymallus. | Cyprinodontidee.
Fishes with | of the stomach | Gymnotide. Osmerus. Percopside.
air-bladder | immediately Cyprinide.
communi- | behind it.
(ehabaresxyaugay ff uv | ae ;
alimentary | Ce bene lie
canal by |Communicating Elops. | Galaxiidee.
a ductus |withthestomach Albula.
pneumati- jin its middle Clupea. |
cus. or posterior Chirocentrus.

region. | Halosaurus.

Now, what is the meaning of the facts shown in the above table ?

The constancy of the conditions in the families most extensively examined is strong
evidence that these relations are not without some significance—not a mere matter of
chance. In no single genus, and indeed in no single family, have I found the duet
varying between right and left. Within any one family it may be median or more or less
displaced to the left side; or, on the other hand, it may be median or more or less displaced
to the right side. Beyond these limits I have not found any variation. Throughout the
whole series of observations I have endeavoured to guard against fallacy arising from a
mere mechanical twisting of the stomach on its axis. Such a twisting, it is true, does
sometimes appear to exist either as a normal or an occasional condition; but the torsion
produced does not involve the cesophagus in such a way as to lead to erroneous observa-
tions, except perhaps in such a case as Galavias, where the duct communicates with the
stomach near its extremity, and on what appears in relation to the pylorus to be the
right side, but in which the pylorus is itself directed to the right.

Looking now carefully at the table before us, we may, I think, note the following
fairly obvious points :—

1. The position of the ductus pneumaticus in relation to the alimentary canal is a
characteristic feature in at least some groups of fishes, and as such may have a certain
diagnostic or taxonomic value.

* Giinther, A.: “ Description of Ceratodus,” Phil. Trans., vol. 161, 1871.
Parker, W. N.: ‘‘On the Anatomy and Physiology of Protopterus annectens,” Trans. Irish Acad., vol. xxx. pt. 3.
Spencer, B.: “Contributions to our Knowledge of Ceratodus—Pt. I. The Blood-vessels,” Macleay Memorial
Vol., Linn. Soc. N.S.W.
Spencer, B.: ‘* Der Bau der Lungen von Ceratodus und Protopterus,” Zoolog. Forschungsreisen in Australia
und dem Malayischen Archipel, Jena, 1898.

——

VISCERAL ANATOMY OF THE CHARACINIDZ. 71

2. Even eliminating all doubtful cases, there remains a substantial minority of families
in which the opening of the duct is situated on the right side, the majority, however,
having the duct situated medially or to the left side.

3. Many of the most primitive or generalized forms, including the three Ganoids, have
the duct in the mid-dorsal line.

4. Families regarded as being allied do not always present similar conditions in
relation to the situation of the duct. For example, the Siluridee and the other Ostario-
physes ; and, again, the Esocide and Galaxiide.

Whatever may be the interpretation of the facts, it seems clear that the conditions
found to cbtain in the Salmonide and Siluride, at least, cut away the ground from
Sagemehl’s argument, in so far as it was based upon the supposed transitional conditions
presented by the Erythrinoids. His contention for a primitively ventral air-bladder,
which in the Teleostei has travelled round the left side of the cesophagus, may indeed
still be the correct view, but it can no longer be supported by reference to the Erythrinoid
conditions, unless with the reservation that the evidence supplied by the Salmonidz and
Siluride points to a journey in precisely the opposite direction.

Tf the facts of organization before us stood alone, we might legitimately deduce two
divergent lines of evolution for the two groups into which the Physostomi are thus made
to fall, eliminating those families in which the duct is median and conceivably to be
derived from either condition. The idea is not without a certain fascination. But its
demands are surely such as cannot be conceded.

It demands, for example, for the highly-specialized Weberian apparatus either an
independent origin in the Siluroids and the other Ostariophyses, respectively, or else
an antiquity at least equal to that of the postulated ancestral Polypteroid air-bladder.
The former supposition is incredible: the latter highly improbable. The association of
the Salmonidze with the Siluridee is a less serious difficulty, kinship not being necessarily
implied, but possibly only parallel development.

If we set aside this hypothesis, as I think we must do, and admit that the duct may
in all cases have travelled in the same direction, we may perhaps conclude from the
weight of evidence that the rotation has been on the left side. This, however, necessi-
tates the assumption that in the Siluridee and Salmonidee the duct has crossed the median
line and become definitely associated with the right side of the cesophagus—a proceeding
for which I am able to suggest no conceivable cause. Moreover, the possibility is not
absolutely excluded that the rotation may have been in the reverse direction, and that it
may be in the Characinid group of families that the crossing of the median line has

taken place.

There still remains another view: that the immediate ancestors of the higher
Teleostomi had already evolved, the evidence does not show how, a dorsal air-bladder
connected medially with the oesophagus, and that the pneumatic duct afterwards shifted
to one side or the other as the several families branched out. This view, however,
attempts no solution of the problem of the origin of the air-bladder : it merely suggests
an explanation of the varying conditions observed within the limits of the Teleostomi.
Like the preceding hypothesis, it is unsatisfactory in that no cause is apparent which

SECOND SERIES.—ZOOLOGY, VOL. IX. 11

72, MR. W. 8S. ROWNTREE ON THE

could have determined a definitely-directed wandering of the duct away from the median
line; and it is to be remembered that in some forms—notably the Erythrinoids—the
deviation is very great.

Of the three hypotheses here advanced, the third is most in harmony with the con-
dition met with in Amza and the other Ganoids, and, indeed, as it seems to me, with the
facts generally. As regards, however, the testimony afforded by the Ganoids and other
so-called ‘“ primitive” forms, it is sometimes necessary to remind oneself that they are
themselves final terms, not mean terms, of series which have survived to the present
day, and that, whilst retaining certain primitive characters, they are not on that account
necessarily less likely to have acquired extreme modifications in other directions.

With reference to the possible causes which may have operated in determining the
position of the duct, it may be pointed out that in some forms—notably perhaps in
Amia and Lepidosteus—the ductus pneumaticus is so short that any asymmetry in its
position would almost inevitably be attended by an asymmetry in the position of the
air-bladder itself. One can thus see a determining cause for its median position.
In many Physostomi, on the other hand, such as the Characinid, for example, the
duct is of such length and laxity that its asymmetry would appear at the present day to
be independent of the position of the bladder, whatever may at one time have been
the case. Whilst, however, it is fairly obvious that in many cases the precise position
of the duct in relation to the median line may be physiologically immaterial, we do not
thus get any light upon the cause, either of its uniform asymmetry throughout entire
families of fishes, or of the two opposite phases of asymmetry which characterize
different families.

The one clear outcome of the investigation which I wish to emphasize, is the
demonstration that Sagemehl’s interpretation of the Erythrinoid condition of the ductus
pneumaticus is not justified by a more extended knowledge of the facts. ‘That in-
terpretation may, indeed, yet prove to be the true one—the asymmetric condition of the
air-bladder in Polypterus suffices to lend colour to it; but it cannot be held to be
established by Sagemehl’s line of argument. For it must be remembered that the
argument was to a large extent based on the improbability of a wandering of the duct
from the mid-dorsal line having occurred; and it is abundantly evident from the facts
set forth in this paper that exactly such a wandering must have taken place, to one side
or to the other, or perhaps to both.

The view that the air-bladder originated as a ventral structure remains, however,
untouched. The blood-supply of the bladder in the Dipneusti would appear to be
conclusive on that point. But as to when and how the air-bladder became a dorsal
structure in the evolution of the Teleostomi, evidence is as yet wanting.

In comparison with the considerations just discussed, observations on the relative
length of the ductus pneumaticus carry but little interest. I may, however, state,
as the result of careful measurements, that its lengti in Characinids is commonly about
one-tenth or one-twelfth of the length of the body of the fish. Considerable divergences
from these proportions occur in certain forms. Thus in Macrodon the length found was

VISCERAL ANATOMY OF THE CHARACINIDA. 73

& that of the body; in Sarcodaces 1+; in Citharinus and Bryconaethiops +3 in
Hydrocyon, Salminus, Alestes, Leporinus, and Xenocharax from py to 35; in Ano-
stomus and Lugnathichthys y'3; in Prochilodus +g; and in Ichthyoborus +5. In some
forms, for example IJacrodon, Lebiasina, Serrasalmo, Distichodus, Anacyrtus, and
Aenocharax, the duct dilates at its orifice; this is especially noticeable in the Ery-
thrinoids, in which it terminates on a large papilla. In Serrasalmo, according to
Valenciennes, the opening is furnished with a valve.

Differences are also observable in the width or stoutness of the duet, but these are not
readily susceptible of measurement or of expression in definite terms. In one case,
Macrodon, as stated by Sagemehl*, the ductus pneumaticus is beaded by alternating
widenings and narrowings, in the manner of a rosary.

I ought to state that, through an error of judgment in the early stages of this
investigation, all comparisons made with the length of the fish include the caudal fin.

This concludes the remarks | have at the present time to offer upon the alimentary
canal and its appendages. Of the other abdominal viscera, the ovaries alone have

yet claimed my special attention.

OTHER ORGANS.

The Ovaries.

On the question of the type of ovary represented amongst the Characinide, the
literature of the subject contains contradictory statements. Thus Joh. Miller + states
that the ovaries are closed sacs, shut off from the body-cavity, and opening together
on the exterior of the body. Valenciennes {, on the other hand, states in the most
explicit manner that the ovaries are of the Salmonoid type, and that the ova fall freely
into the body-cavity. This statement he makes for each of the following forms:
Erythrinus, Macrodon, Lebiasina, Curimatus, and Parodon. Of Myletes he states that
the ovaries fill two-thirds of the body-cavity, without, however, speaking of their
character. With reference to the two first-mentioned forms— Lrythrinus and
Macrodon—Valenciennes himself makes apparently contradictory statements: thus, in
one place (vol. xix. p. 493), speaking of Erythrinus he says: “ La nature a donné aux
poissons dont nous traitons ici une organisation semblable 4 celle des Saumons; en ce
qui concerne les organes génitaux, nous les voyons en effet constitués par deux rubans
portant dans les femelles les ceuls sur des replis transverses; ces ceufs tombent dans
VYabdomen pour s’échapper par deux trous percés de chaque edté de Vanus.” But
on page 484 of the same volume, speaking of the same fishes, he says, “ Les sacs
ovariens n’ont point de communication avec l’intérieur de la cavité abdominale, de
sorte que ces organes ne sont pas faits comme ceux de beaucoup de genres de la famille
des Saumons ou des Anguilles.”’

* Sagemehl, ‘“‘ Das Cranium der Characiniden,” p. 108.

+ Joh. Miiller, ‘‘ Untersuchungen iiber die Eingeweide der Fische,” Abhandl. der Berl. Akad. a. Wissenschaften,

1843, p. 189.
+ Cuvier et Valenciennes, ‘ Histoire Naturelle des Poissons,’ vols. xix, & xxii.

11?

7A MR. W. S. ROWNTREE ON THE

Of Macrodon, he says on page 513, “ J’ai trouvé a ouverture de corps, les deux sacs
ovariques enveloppés dans leurs replis peritonéaux, et adhérant chacun a la paroi
abdominale”; but on page 515, “Quant aux organes génitaux, ils ont, comme nous
Vavons déji dit, la disposition de ceux des Truites.’ The ovaries of the other
Characinids are not mentioned by Valenciennes.

Sagemehl * states that he has satisfied himself that the ovaries in these fishes are closed
sacs, thus confirming Miiller’s description. He does not say in what forms he has made
the observations.

I have up to the present time examined the ovaries in the following 14 forms:
Alestes nurse, Tetragonopterus abramis, Petersius Leopoldianus, Micralestes altus, Alestes
longipinnis, Micralestes Stormsi, Hydrocyon Forskalii, Sarcodaces odoé, Anacyrtus
microlepis, Curimatus Gilberti, Macrodon trahira, Erythrinus uniteniatus, Lebiasina
bimaculata, Pyrrhulina semifasciata. In some eases, it is true, the condition of the
specimen was not such as to justify any statement about the ovaries, owing to the
disappearance of the membranous parts. In other cases, however, the membranes were
intact, and I was able to satisfy myself, by the use of the seeker and of the blowpipe
under water, that no communication existed between the ovarian sacs and the general
hody-cavity ; the sacs, through the medium of their membranous continuations, opening
directly upon the exterior by a common post-anal orifice. In all, the ovaries of the two
sides were equally developed, arising at the level of the anterior chamber of the
air-bladder.

Nevertheless, it has seemed to me that two different ovarian conditions are dis-
tinguishable amongst the Characinids above mentioned: Swrcodaces (PI. 4. fig. 29) may be
yegarded as typical of the one, and Alestes nurse (fig. 30) of the other; but the
difference is, I think, only a question of degree in the backward extension of the
ovigerous tissue, and the reciprocally developed forward extension of the oviducal
membranes. In Sarcodaces the ripe ovaries extend back to just behind the level of the
anus, where they become closely approximated, but are not confluent. Their membranes
there continue as a single, very short common duct to the exterior. Each ovary has the
appearance of being, as it were, slung by its membranes, which invest it completely,
except on the dorsal side, where a space or canal runs without interruption to the
external opening.

Similar conditions appear to obtain in Macrodon. Of the other Erythrinoids I can
speak with less confidence, but I believe their ovaries to have the relations just
described.

In Alestes and Hydrocyon, on the other hand, according to my observations, the
ovaries fall distinctly short of the anus, so that the common oviducal space enclosed
by the backward continuation of their membranes is of considerable extent, instead
of being exceedingly small as in Sarcodaces.

The two conditions I regard as being modifications in different degrees of the same
ovarian type. Although differing considerably at first sight, they agree absolutely

* Sagemehl, ‘* Das Cranium der Characiniden,” p. 115.

VISCERAL ANATOMY OF THE CHARACINID. 75

in the absence of any communication between the ovarian tract and the general body-
cavity, and in the continuation of the ovarian membranes to form a common duct
to the exterior.

In some of the other Characinids in which ovaries were found, these organs appeared
to coalesce at a point anterior to the anus, and to continue to the exterior as a common
mass of ripe ova. Some such appearance I observed in Peéersius and Micralestes.
This condition I interpret as being referable to the Alestes-type; for obviously, if in
Alestes the common oviducal space were crammed with ripe ova in transit, the
appearance of the whole apparatus would be suggestive of an early coalescence of
the ovaries themselves. In the Erythrinoid form Lebiasina, also, there was the
appearance of coalescence, but beginning further back, about the level of the anus;
also, but still further restricted to the posterior portion, in Tetragonopterus abramis.
In all these instances of apparently coalesced ovaries the investing membranes them-
selves were not recognizably present, but the indications seemed to me to demand their
existence.

It will be observed that the ovarian relations above described present no resemblance
to those existing in Ama, in which the ova fall freely into the body-cavity, and thence
pass by wide orifices into the oviducal funnels.

The testes seem to be of the normal Teleostean type, presenting no features worthy
of remark. They are seen as narrow bands, arising at the sides of the anterior chamber
of the air-bladder, and passing back uniformly to their junction just before opening
to the exterior.

The Kidneys.

The kidneys in my specimens have not, generally speaking, been well preserved, and
I have no minute observations to record. It may, however, be stated that these organs
in the Characinide extend for the whole length of the body-cavity, from near the
diaphragm backwards immediately under the vertebral column. In some cases there
seemed to be an anterior enlargement (“ head-kidney ” ?) in which the anterior end of
the air-bladder was more or less imbedded. No enclosure in muscle was observed.

The Heart.

I carefully examined the heart in one of my largest specimens, Wucrodon, for signs
of a valvular conus. No valves, however, were found to exist, beyond the two small
pockets at the exit of the ventricle and that guarding the auriculo-ventricular aperture.

The Respiratory Organs.

Concerning the respiratory organs, I have as yet no new observation to record. The
gills themselves seem to present no special features. There is, however, in certain of the
Characinidee an accessory branchial organ, arising as a blind sac from the upper margin

76 MR. W. S. ROWNTREE ON THE

of the last gill-cleft. This organ, which has been described briefly by Kner * and in
detail by Sagemebl +, appears to be peculiar to the herbivorous forms, having been
recorded by these writers for Curimatus, Cenotropus, Hemiodus, Ctitharinus, and
Prochilodus; and more recently by Boulengert for Xenocharaz. Discussing the
morphology of this curious organ, Sagemebl expresses the view that we have here to do
with a structure arising from the gill of the rudimentary fifth branchial arch or
“inferior pharyngeal bone.” If this conclusion be correct, the presence of the organ in
the herbivorous Characinids, and in these only, becomes of great interest, in view of the
fact that traces ofa fifth gill have not been found in any living Ganoid, but only in fishes
of yet lower organization—certain Selachians and Dipnoids. That this character was at
one time widely diffused is shown, as Sagemeh] points out, by its presence in forms so
far separated as Notidanus and Protopterus. 'Two important conclusions seem to follow
from the identification of the organ under consideration with the gill of the fifth arch :
(1) that the herbivorous Characinids which possess it form a natural group; and (2) that
this division of the family cannot be derived from either the Erythrinoids or the other
carnivorous Characinids, but is at least as ancient as either of those groups. It should
be mentioned that an organ showing a somewhat similar minute structure has been
described by Hyrtl § and by Gegenbaur || in certain Clupeids.

Another structure—a rudimentary opercular pseudo-branch—is referred to by
Sagemehl | as being present in some Characinids, as also in Amia. The only forms in
which he mentions having found it are Citharinus and the following carnivora—Hydro-
cyon, Anacyrtus, Alestes, and Tetragonopterus. The Erythrinoids show no trace of it,
contrary to what might have been expected from the consideration of their markedly
Amioid cranial characters.

A PARASITICAL CRUSTACEAN.

An Tsopod crustacean, about 1°5 cm. in length, was found within the body-cavity of
a specimen of Anacyrtus microlepis from Asuncion, Rio Paraguay. Mr. Stebbing, as I
have already mentioned, kindly identified it for me as Artystone trysibia **, family
Cymothoide. I learn from him that the species is of some rarity, having been previously
only twice recorded, once from La Plata, and also from Brazil. My specimen, now in
the Biological Museum of the Royal College of Science, South Kensington, was enclosed
in a thick-walled capsule adherent to the ventral body-wall of the fish, and entirely
within the body-cavity. The rectum and adjacent parts were unfortunately cut before

* Kner, “Die Kiemenanhange der Characiniden,” Verhandl. d. zoolog.-botanischen Gesellsch, in Wien, Bd. xi,

+ Sagemehl, ‘‘On the Accessory Branchial Organ of Citharinus,” Morph. Jahrb., 1886-7.
+ Boulenger, ‘ Les Poissons du Bassin du Congo,’ p. 199.

§ Hyrtl, “Ueber die accessorischen Kiemenorgane der Clupeaceen,” Denkschr. d. k. Akad. zn Wien, Mathem.-
naturw. Klasse, Bd. x. 1855.

| Gegenbaur, ‘* Ueber das Kopfskelet v. Alepocephalus rostratus,” Morphol. Jahrb., Bd. iv. 1878, Suppl.

| Sagemehl, ‘ Das Cranium der Characiniden,’ p. 113.

* Schiddte & Meinert, ‘‘Symbole ad Monographiam Cymothoarum: Pt. IV. BGC, 1883. Stebbing,

T. R. R., “* History of the Crustacea,” Intern. Sci. Ser., vol. lxxiv. Cymothoide, 1893.

VISCERAL ANATOMY OF THE CHARACINID®. 17

the discovery was made, and the precise relations of this capsule were difficult to deter-
mine, but it had every appearance of being completely closed, and no communication
could be made out between it and the exterior, either directly or through the rectum or
“bladder.” Mr. Stebbing tells me that the situation was unusual for such parasites, but
that it is difficult to assign any limits to the liberties taken by them with their hosts.

SUMMARY AND CONCLUSIONS.

The main conclusions to which I have been led in the investigation detailed in this
paper are the following :—

(1) The ductus pneumaticus has a more or less asymmetric connection with the
alimentary canal, not only in the Erythrinoids, in which it has been long recognized, but
also in the Characinide generally, in the Cyprinide, the Gymnotide, the Siluridee, the
Salmonide, the Hsocide, the Mormyride, the Notopteride, the Galaxiide, the
Percopsidze, the Cyprinodontidie, and perhaps other families.

(2) This asymmetry is to the left side in certain families of fishes, and to the right side
in certain others.

(8) In many families, including some of the most primitive and generalized, no
asymmetry exists, the duct being median.

(4) The evidence furnished by the whole series of observations seems to point to the
derivation of the asymmetric condition from a pre-existing symmetric or mid-dorsal
position of the duct, this being contrary to the view advanced by Sagemehl, who

regarded the Erythrinoid condition as being archaic.

(5) In certain fishes—Notopterus and Petrocephalus (a Mamiend) the ductus
pueumaticus has also an asymmetric connection with the air-bladder.

(6) The posterior sac of the air-bladder presents in the Characinidse a characteristic
arrangement of longitudinal ligaments or septa. The similarity of these septa amongst
the members of some natural groups, such as the Ichthyoborine, suggests a certain
constancy in this character, which may be found to be of systematic value. On the
other hand, the part played by these septa in the formation of a cellular air-bladder in
Erythrinus, and especially in Lebiasina, together with their existence in varying degrees
of development, suggests that they possess a significance as a vestige of a formerly
cellular bladder in the ancestors of the family.

(7) The ovaries in the Characinidie are closed sacs, without communication with the
body-cavity. This statement, which is, however, based upon the examination of a limited
number of genera, is in accordance with the statements of Joh. Miiller and of Sagemehl,
but is at variance with some of the assertions made by Valenciennes.

(8) I have described two somewhat different ovarian conditions, of which Sarcoduces
and Alestes respectively may be taken as typical, the difference lying in the degree of
backward extension of the ovaries. Both cystoarian *.

* Huxley, T. H., “Contributions to Morphology—Ichthyopsida. No. 2: On the Oviduets of Osmerus, with
Remarks on the Relations of the Teleostean with the Ganoid Fishes,” Proc. Zool, Soc. Lond., 1883.—Howes, G. b.,

“On some Hermaphrodite Genitalia of the Codfish (Gudus morrhua), with Remarks upon the Morphology and
Phylogeny of the Vertebrate Reproductive System,” Journ. Linn. Soc., Zool., xxiii. 1891.

78 MR. W. 8S. ROWNTREE ON THE

(9) The Characinid stomach is with or without a blind sac. The cardiac and pyloric
regions are well differentiated. In the herbivorous forms the pyloric portion is highly
muscular and gizzard-like. The pyloric flexure is usually directed ventrally, but in
some forms is fairly constantly turned in some other direction.

(10) Pyloric czeca are always present in the Characinidee, but in very varying number
and development. They sometimes fringe the intestine for a considerable distance.

(11) The length of the intestine in the carnivorous Characinids does not exceed the
body-length, and is usually less. In the herbivorous and mud-eating forms it is very
much longer, varying from about twice to five times the length of the fish. It may vary
also in calibre in different regions. No trace of a spiral valve was observed.

(12) The liver in the Characinidee is usually tri-lobed, the proportionate development
of the three lobes being very variable. In some forms one or more of the lobes may be
greatly elongated, or may be much reduced. The gall-bladder is usually attached to the
right lobe.

(18) The African Characinid Sarcodaces strikingly resembles the Erythrinoid
Characinids, not only in cranial characters, as shown by Sagemehl, but also in certain
visceral characters, notably in the opening of tie ductus pneumaticus far to the left on
the alimentary canal, in the character of the ovaries, and in the features of the air-
bladder.

(14) The herbivorous Characinids, more especially Distichodus, Xenocharax, and
Citharinus, resemble one another in the marked thickening of the sheet of peritoneum
which shuts off the air-bladder from the body-cavity. Jn all the distinguishing visceral
characters of this group, Xenocharax seems to present a lower stage of specialization than
the other two associated forms.

(15) In the structural features of the air-bladder, Serrasalmo and Anacyrius seem to
be separately and somewhat widely divergent from the other Characinids.

(16) If the ridging of the posterior air-blaader be regarded as having a vestigial
significance, the Characinide, Cyprinide, and Gymnotidze would seem to represent
successive stages of air-bladder evolution.

(17) With the possible exception of the indications of a cellular air-bladder, there
appears to be nothing in the visceral anatomy of the Characinide which strengthens the
deductions made from the skull as to the Amioid affinities of the group. In opposition
to such deductions are especially the cystoarian ovaries, the asymmetric ductus
pneumaticus, the presence of pyloric appendages, and the absence of all trace of a
valvular conus and of an intestinal spiral valve.

VISCERAL ANATOMY OF THE CHARACINID. 79

List OF THE CHIEF WORKS CONSULTED IN 'THE COURSE OF THE INVESTIGATION.

Boas, J. E. Vi—Ueber Herz und Arterienbogen bei Ceratodus und Protopterus. Morph. Jahrb., Bd. vi.
1880, p. 321.

Boas, J. E. V.imUeber den Conus arteriosus und die Arterienbogen der Amphibien. Morph. Jahrb.,
Bd. vii. 1881.

Bovrencer, G. A.—Les Poissons du Bassin du Congo. Brussels, 1901.

Bouencer, G, A.—Matériaux pour la Faune du Congo. Ann. Mus. Congo, Zool. i. 1898-1900.

Bourencer, G. A.—Description of a new Characinid Fish discovered by Dr. W. J. Ansorge in Southern
Nigeria. Ann. & Mag. Nat. Hist., 7th series, vol. ix. 1902, p. 144.

Brroez, T. W.—The Air-bladder and its Connection with the Auditory Organ in Notopterus borneensis.
Journ. Linn. Soc., Zool. vol. xxvu. J900, p. 503.

Brivex, T. W., and Happon, A. C.—Contributions to the Anatomy of Fishes: II. The Air-bladder
and Weberian Ossicles in the Siluroid Fishes. Phil. Trans. Roy. Soc. Lond., vol. 184, 1893,
p. 65.

Bunveert, J. Si—On some Points in the Anatomy of Polypterus. Trans. Zool. Soc. Lond., vol. xv., 1901,

Cuvier et VaLencrennes.—Histoire Naturelle des Poissons, Paris. Tom. xix., 1846; Tom. xxil,
1849.

Dean, Basurorp.—Fishes, Living and Fossil. 1895.

Gecensaur, C.—Ueber das Koptskelet v. Alepocephalus rostratus. Morph. Jahrb., Bd, iv. 1878, Suppl.

GtntueEr, A.—Catalogue of the Physostomi in the Collection of the british Museum. Vol. V. London,
1864.

Ginraer, A.—An Introduction to the Study of Fishes. Edinburgh, 1880.

Guntuer, A.—Description of Ceratodus. Phil. Trans. Roy. Soc. Lond., vol. 161, 1871.

Gtnrurr, A —Report on the Voyage of the ‘Challenger,’ Vol. xxti. Zoology, 1887. _Deep-sea Fishes.

Howes, G. B.—On some Hermaphrodite Genitalia of the Codfish (Gadus morrhua), with Remarks upon
the Morphology and Phylogeny of the Vertebrate Reproductive System. Journ. Linn. Soc.,
Zool. xxiii. 1891, p. 539.

Houxtey, T. H.—Coutributions to Morphology. Ichthyopsida: No, 2. On the Oviducts of Osmerus,
with Remarks on the Relations of the Teleostean with the Ganoid Fishes. Proc. Zool. Soc. Lond.,
1883.

Hyntt, S.—Ueber die accessorischen Kiemenorgane der Clupeaceen. Denkschr. Akad. Wien, Bd. x.,
1855.

Kner, R.—Die Kiemenanhange der Characiniden. Verhandl. zoolog.-botanischen Gesellsch. Wien,
Bd. xi., 1861.

McLeop, J.—Rechercbes sur la Structure et le Développement de |’Appareil Reproducteur femelle des
Téléostéens. Archives de Biologie, tom. ii. 1881, p. 497.

Max Weper.—Die Abdominalporen der Salmoniden nebst Bemerkungen iiber die Geschlechtsorgane
der Fische. Morph. Jahrb., Bd. xii., 1886-7.

Miter, J.—Untersuchungen tiber die Hingeweide der Fische. Abhandl. Berl. Akad. Wissensch.,
1843.

Mtuirr, J.—Ueber den Bau und die Grenzen der Ganoiden. Abhandl. Berl. Akad. Wissensen.,
1844.

Miter, J., und Troscnet, F. H.—Hore Ichthyologice. Beschreibung und Abbildung neuer Fische.

Die Characiniden. Berlin, 1845.
SECOND SERIES.— ZOOLOGY, VOL. IX. 12

80 MR. W. S. ROWNTREE ON THE

Mixxucno-Mactay.—Ueber ein Schwimmblasenrudiment bei Selachiern, Jenaische Zeitschrift fiir
Medicin und Naturwissensch., Bd. iii. 1868, p. 448.

Parker, W. N.—On the Anatomy and Physiology of Protopterus annectens. Trans. Irish Acad., vol. xxx.
pt. ii.

RemNuarpt, J—Ueber die Schwimmblase in der Familie Gymnotini. Archiv fiir Naturgeschichte,
Berlin, 1854, p. 169.

Sacumrut, M.—Beitrage zur vergleichenden Anatomie der Fische: IJI. Das Cranium der Characiniden.
Morph. Jahrb., Bd. x., 1884.

Sacemrnt, M.—Die accessorischen Branchialorgane von Citharinus. Morph. Jahrb, Bd. xii., 1886-7.

Srencer, B.—Contributions to our Knowledge of Ceratodus: Pt. I. The Blood-vessels. Macleay
Memorial Volume, Linn. Soc. N.S.W., 1893, p. 1.

Spencer, B.—Der Bau der Lungen von Ceratodus und Protopterus. In Semon, Zoolog.
Forschungsreisen in Australia und dem Malayischen Archipel., Jena, 1898.

Scui6pre and Mrrnert.—Symbole ad Monographiam Cymothoarum : Pt. 1V. Cymothoide. Naturhist.
Tidsskr., Bd. xiii., 1883.

Srannrus, H.—Handbuch der Anatomie der Wirbelthiere: Die Fische. Berlin, 1854.

Sressine, T. R. R.—History of the Crustacea. Intern. Sci. Ser., vol. Ixxiv. (Cymothoide). 1893.

Wieversuerm, R, (W. N. Parxer).—Comparative Anatomy of Vertebrates, 2nd Ed. London, 1897.

DESCRIPTION OF THE PLATES.

PLATE 3.

Fig. 1. Hydrocyon brevis. General view of air-bladder and stomach. Liver removed. Stomach flexed
ventrally. Blind sac somewhat large.

2. Sarcodaces odoé. General view of air-bladder and stomach. Liver removed. Stomach flexed
dorsally. Ductus pneumaticus opens into left side of alimentary canal. Air-bladder
opened to show ridging.

3. Macrodon trahira, Stomach greatly distended and flexed dorsally. Large blind sac. Orifice
of ductus pneumaticus on the left side.

4. Serrasalmo humeralis. Air-bladder partly opened, showing ridging and pouching. Posterior
sac the smaller.

5. Lebiasina bimaculata. Cellular air-bladder. Ductus pueumaticus opening into left side of
alimentary canal.

6. Erythrinus uniteniatus. Posterior sac of air-bladder opened to show rudimentary cellular
structure.

7. Macrodon trahira. Posterior sac of air-bladder opened to show ridging of its walls.

8. Anacyrtus microlepis. Air-bladder partly opened to show faint ridging of posterior sac.

Figs. 9-28 represent the dorsal aspect of the stomach in a series of forms, with the relative position

of the opening of the ductus pneumaticus.

Vig. 9. Ichthyoborus niloticus.

=

10. Alestes nurse.
11. Salminus mavxillosus.

VISCERAL ANATOMY OF THE CHARACINID.¥, 81

PLATE 4.

Vig. 12. Prochilodus lineatus.
13. Distichodus Antoni.
13 a. Citharinus macrolepis.
14. Leuciscus rutilus.
15. Catostomus macrolepidotus.
16. Clarias lazera.
17. Malapterurus electricus.
18. Sternopygus carapus.
19. Salmo trutta.
20. Petrocephalus bane.
21. Gymnarchus niloticus.
22. Notopterus kapirat.
23. Elops saurus.
Albula conorhynchus.
Clupea sprattus.

S
2 Ole

. Chirocentrus dorab.
Galazias truttaceus.
Esox lucius.

=

ww wo iwi
o:

we

29. Sarcodaces odoé. Ovaries form closed sacs, with investing membranes which continue as
short ducts uniting just before their opening on the exterior. No communication with
general body-cavity. Ova large.

30. Alestes nurse. Ovaries differ from the preceding in having a shorter backward extension, with
a proportionally larger common oviducal space. No communication with general

body-cavity.

REFERENCE LETTERS.

bl., air-bladder. @., esophagus.

b.s., blind sac. ov., ovaries.

ce., cellular structure. p.c., pyloric ceca.
c.0.8., common oviducal space. st.c., cardiac stomach,
d.p., ductus pneumaticus. st.py., pylorie stomach.
Lr., longitudinal ridges. i.r., transverse ridges.

m.s., tembranous sac.

(Note by Author.)—The Plates are the work of Mr. James Green, to whom my thanks are due. ‘The
figures are from my own dissections and have been drawn under my direction. All are about natugal
size or slightly enlarged.—W. S. R.

Pee a ATC! ov RL ne eS ee ee

ie Jr ,
o
Car
'
‘
’
i. .
\
%
y
? »
tl 'rd ’
Rh,

Trans. Linn.Soc.Ser.2.Zoon.Vou IX. Pxu.3.

W.S.Rowntree

St_C

(Ge

pets
—

Lr.
ai
te ALi

‘
fj mi

Wire
PTF

F* co “ae

Mantern Bros inp

J.Green del.ct ith, ANATOMY OF THE CHARACINID&

Trans. Linn. Soc.Ser.2.Zoon .Vou. 1X. Pi .4

W.S.Rowntree

Mintern Bros amp

ANATOMY OF THE CHARACINIDA.

J.Green del.et lth.

ra

III. On the Evolution of the Australian Marsupialia ; with Remarks on the Relation-
ships of the Marsupials in general. By B. Arraur Benstey, B.A. (Tor.),
Ph.D. (Columbia), University of Toronto, Canada, (Communicated by Prof. G. B.
Howes, D.Sc., LL.D., F.R.S., Sec.L.8.)

(Plates 5-7.)

Read 7th May, 1903.

Conrents.

Page
Timi hCHON eobSe bGoooneaages ns Uu.coo dhs Boon eaves UooHdd sesh Orn bbAsaoonane 83
The Adaptive Modifications of the Dentition in the Australian Marsupials ............ 88
The Adaptive Modifications of the Foot-structure in the Australian Marsupials ........ 162
The Identification of the Stem-Form of the Australian Marsupials .................. sey)
The Phylogenetic Arrangement of the Australian Marsupials ..................005. 192
The Question of the Time and Mode of Origin of the Australian Radiation............ 204
The Major Classification of the Marsupials in general ...........- ee cece cece cee emes 207
The Systematic Arrangement of the Australian Families .............-.0 0000 eee ee 210
SMO arena o50 cog dock POeoUDEO Dd bn oe GbOD Conor Hon bord bOUUCdOOU SCO OMDDOnOOnC 211
Bgdennmnoaei Une IARI 3.6 doomacoon Cobo ood cobb cTe Dooo ooUdo oOo UmodcoCoaraS 214

INTRODUCTION.

On reviewing the general progress of opinion in mammalian phylogeny, we note
the fact that the discussions bearing on the affinities of the Marsupials have largely
centred around the broader question of the relationships of these animals with the
Placentals, and that in respect to the latter there has been a gradual change from an
older view, by which the Marsupials were assigned an ancestral position, on the
assumption that they represent a lower grade of organization, to a newer one, by which
they are regarded as a collateral or even a derived group.

The three groups into which the existing Mammalia are divisible have been commonly
recognized as representing, to a certain extent, three grades of organization—the
Placentalia, mainly distinguished by their vascular nutritive allantois, standing highest
in the scale; the Monotremata, in virtue of their sauropsidan characters, especially their
oviparity, lowest; while the Marsupialia occupy a somewhat intermediate position,
especially in respect to their combination of a non-vascular allantois, which they possess
in common with lower oviparous forms, with the actual viviparity of the Placentalia.
Following the evolutionary principle, the tendency has been to connect these groups

SECOND SERIES.—ZOOLOGY, VOL. IX. 13

84 DR. B. A. BENSLEY ON THE EVOLUTION

into a line of descent, the lowest being associated with the Reptilia and indirectly with
the Amphibia, or directly with the latter group. It has been found, however, that the
Monotremata do not occupy a wholly prototypal position with reference to the
Marsupialia, and that the latter are by no means wholly prototypal to the Placentalia ;
and the necessity has thus arisen for the recognition of hypothetical or ideal groups
through which the evolutionary sequence might be assumed to have been established,
and from which the existing groups might be supposed to have arisen by independent
specialization. The clearest exposition of this principle is to be found in Huxley’s essay
of 1880, in which three type groups are characterized and designated as Prototheria,
Metatheria, and Eutheria, these being assumed to have arisen in the order named, and
to have given rise successively to the Monotremata, Marsupialia, and Placentalia.

Monotremata. Marsupialia. Placentalia.

| Eutheria,
i

Metatheria.

Prototheria.

A phylogenetic plan, showing the primary relationships of the Mammalia.

Naturally, the Eutheria, being a terminal group, are practically equivalent to Placentalia.
Huxley remarks :—“ It is a fact, curiously in accordance with what might be expected
on evolutionary principles, that while the existing members of the Prototheria and the
Metatheria are all extremely modified, there are certain forms of living Eutheria which
depart but little from the general type. . . . There is no known Monotreme which is not
vastly more different from the Prototherian type, and no Marsupial which has not far
more widely departed from the Metatherian type, than Gymnura or, indeed, Lrinaceus
have from the Eutherian type” *.

* Careful analysis shows that it is not absolutely necessary to recognize more than two primary mammalian
groups, one including the stock-forms of the Monotremata, the other those of the Marsupialia and Placentalia. he
question arises, How are these groups to be designated? and this leads to the further question, In fixing the
designatious, are we to rely on strict priority or on common usage? As early as 1872 Gill divided the Mammalia
into two primary groups of Prototheria and Eutheria; and this division has been used by Cope (1889) and recently
by Osborn (1899). Gill’s Eutheria were not assigned a placental or aplacental character. Common usage has it that
the Eutheria are placental, as specified by Huxley. It accordingly follows that if we revert to Gill's classification,
we must characterize the Eutheria as fundamentally equivalent to Huxley’s Metatheria and as aplacental like the
majority of the Marsupialia, unless it is proved that the latter animals are of placental origin, as supposed by Wilson
and Hill, and Dollo. Even in the latter event, it will be apparent that allowance must be made for a Metatherian
stage in development, although not for a definite Metatherian group.

OF THE AUSTRALIAN MARSUPIALIA. 85

The precise distinction of Huxley between the ancestral or metatherian characters of
the Marsupialia and their special characters definitely removes the latter animals from
an ancestral to a collateral position with reference to the Placentals. The principle is
essentially one of reduction in number of the characters in which the Marsupials appear
to occupy a prototypal position. For example, the single-tooth change of these animals
was regarded by Flower (1867) and afterwards by Thomas (1887) as prophetic of the
more complete tooth-change of Placentals, but was pointed out by Huxley as a special
character of the Marsupials and as derivative of a former diphyodont condition common
to the Metatherian ancestors of both Marsupial and Placental groups, but more com-
pletely retained in one of them than in the other. The correctness of this position has
been amply confirmed through the embryological investigations of several zoologists,
including Rése, Leche, Kiikenthal, Woodward, Wilson and Hill, and Dependorf.

Recent research has, however, gone still further in the reduction of the recognized
prototypal characters of Marsupials, and certain observers, namely, Wilson and Hill,
and Dollo, are in favour of removing the most essential one, namely, the possession of a
non-placental allantois, which they regard as a secondary feature of the Marsupials and

_as derivative of a former placental condition. This view is based on Hill’s discovery of a

placental connection in Perameles ; and the reduction of the allantois in other Marsupials
is supposed by Dollo to have been associated with premature birth and fundamentally with
arboreal habit. In a former paper (1901 0), however, the present writer has pointed out
the probability that the aplacental condition of most Marsupials is actually primitive,
and that the placental connection in Perameles, like a multitude of other characters in
which Marsupials resemble Placentals, has been independently acquired ; in other words,
that it represents a convergent or homoplastic * development.

However this may be, we have the more definite fact that the Marsupials and
Placentals are collateral and, in a certain sense, equivalent groups, of common parentage ;
and this conception may be welcomed as clearing the way for a better perception of the
d tails of their secondary evolution or adaptive radiation. Especially is this true of
the former: any attempt to explain their secondary differentiation on the basis of a
Marsupial ancestry of the Placentals must naturally result in confusion, because of the
lack of distinction between those of their adaptive characters which are, generally
speaking, common to all of the members of the group and those distinctive of minor
divisions, leading to the doubly erroneous conclusion that the evolution of the Marsupial
group is not finite, and that characters of both kinds have been carried over from
Marsupial to Placental stages. At the present time, while the evidence at our disposal
may not be of sufficient extent to furnish a complete plan of the whole Marsupia}
radiation, we at least have the advantage of being able to form a clear conception as to
what the problem involves. Referring to the plan above presented, the whole situation
in Marsupial phylogeny may be summed up in three questions. In the first place,
What are the characters of the ancestral Marsupio-placental or Metatherian forms ?
Some of these characters, as, for example, the narrow cranium and projecting zygomata,

* Of. recent paper on Homoplasy by Osborn (1902).

13*

86 DR. B. A. BENSLEY ON THE EVOLUTION

or the narrow iliac bones, we already know to have been carried over to the primitive
forms (Didelphyidee, Creodonta) of both of the derived groups, and reduced in the more
specialized ones. Others, such as the double dentition and the normal pentadactyl pes,
were carried over to the Placentals and lost in the Marsupials; while still others, such
as the marsupial bones and the non-placental allantois, were carried over to the Marsupials
and lost in the Placentals. Secondly we have to ask, What were the features of a
divergent adaptation which, acting on the ancestral forms, brought about their division
into Marsupial and Placental series, and gave to the former the characters which now
distinguish them, namely, the reduced milk-dentition, the prehensile pes, and premature
birth? The assumption by the ancestors of the Marsupial division of arboreal habit may
be noted as a possible answer to this question. Given the primary differentiation of the
Marsupials as a group, we have finally to ask, What is the sequence of events in their
secondary adaptive radiation ? This question involves the history of several faunas com-
prising the Jurassic series of the northern hemisphere (providing its members were
actually related to the Marsupial division), the Oligocene Didelphyide of the same region,
the Miocene series of South America, the existing didelphyid fauna of that country, and,
finally, the Marsupial fauna of the continent of Australia and the adjacent islands.

Generally speaking, the history of a fauna can only be satisfactorily determined by
reference to palzeontology. In the Marsupial fauna of Australia we have the result of a
radiation which we may be sure has proceeded along much the same general lines as
that which has given rise to the existing Placentalia. Many members of each series
present characters which have more or less perfect counterparts or parallels in the other.
One has only to compare the dental characters of the Dasyuride, the Phalangeridee, or
the Phascolomyide with those of the Creodonta, the Primates, or the Rodentia to
appreciate the intimate structural correspondences between the members of the two
groups. In attempting to define the details of the Australian radiation, however,
we are at once confronted with the difficulty that the fossil deposits of that country have
not as yet furnished ancestral or collective types. The reason of this is rather obscure.
There is still the possibility that older deposits may be found containing such types;
and when we consider the progress of paleontological discovery in other countries,
it seems difficult to believe that the finding of such forms in Australia is more than a
matter of industrious exploration. There is a further possibility, however, that the
ancestors of the Australian fauna passed the incipient phases of their evolution in another
country, such as Asia or even South America. Again, the researches of Huxley and
Dollo have shown that the ancestors of the existing Marsupials must have been arboreal
animals ; and this raises a suspicion that as such they may have been protected to a much
greater extent than terrestrial animals living in caves, or upon the open plains, or
frequenting the water-side, from those conditions which would result in the preservation
of their remains by fossilization. However this may be, apart from a comparatively
small number of arboreal forms (Petawrus, Dromicia) and some small terrestrial forms
(atpyprymnus, Perameles), remains of which have been brought to light in cave-

OF THE AUSTRALIAN MARSUPIALIA. 87

deposits *, the extinct fauna is composed of large specialized terrestrial types showing no
more approximation to theoretically prototypal forms than, if we may cite an analogous
case, the members of the giant fauna of the mid-Tertiary portion of the placental
radiation do in comparison with the actually ancestral Creodonta. It will be apparent,
therefore, that beyond the limited although valuable evidence afforded by geographical
distribution, or the occurrence of allied faunas living at different periods in other
countries, we are dependent for the history of the Australian series on the sequence of
adaptive characters as presented by living forms.

The results recorded in the present paper represent an attempt to construct a plan of
the Australian radiation by tracing the sequence of adaptive modifications in the dentition
and foot-structure of the various genera, in some cases of the species. The possibility
of making such an arrangement with advantage was suggested to the writer some time
ago by the perusal of the opinions expressed by Huxley and Dollo with reference to
an arboreal ancestry of the Marsupials, as determined by the characters of the pes. The
extensive observations of the latter writer in particular, although primarily designed to
show the presence throughout the group of a prehensile type of pes, appeared to indicate
further that most, if not all, of the progressive changes in foot-structure through which
the Australian Marsupials have passed from the time of their origin are actually repre-
sented either in the existing Australian forms or in the American Didelphyide, and that
it would accordingly be possible to explain in a measure the relationships of the various
genera by simply arranging them on a basis of the sequence of adaptive modifications
of the pes. The fact, however, that phylogenetic conceptions based on the adaptive
characters of single structures have frequently been found to be misleading when applied
to others, partly on account of the occurrence of convergent developments, and the
presence of a conspicuous example of convergence in foot-structure between the
Peramelidze and Macropodidie, indicated clearly the advisability of checking such an
arrangement by reference to the adaptive changes of other organs. For this purpose
the characters of the molar teeth were selected, the latter offering the advantage that
the main features of their evolution have already been demonstrated in the parallel case
of the Placentalia.

The results of this study were published in a preliminary paper (1901 @), but on
account of many limitations both of material and literature they included only a general
or family arrangement. The opportunity of examining the case in detail was afterwards
afforded by a visit to London, where, through the kindness of the officials of the
Geological and Zoological Departments of the British Museum, the writer was able
to study the splendid collections both of fossil and existing Marsupials there preserved.
The latter form the basis of the present paper.

Reference has been made above to the necessity of depending in the present case on

* Of. Broom (1896). The presence of these forms in cave-deposits will be noted as in all probability due to no
efforts of their own, and therefore as throwing no light on the conditious affecting their ancestors before the

differentiation of definite carnivorous forms,

88 DR. B. A. BENSLEY ON THE EVOLUTION

the adaptive characters as presented by existing animals. Objection may be justly taken
on general principles to basing phylogenetic conceptions on the characters of living forms,
not only on account of the necessity it involves of selecting ancestral and derived
characters from contemporaneous types, and thereby incurring a risk of mistaking the
direction in which the evolution may be proceeding, but also on account of the necessity
of recognizing hypothetical collective forms. Many instances might be cited in which
the identification of such forms has not got beyond their original creation. While
preferring to claim too little than to assert too much in favour of the stability of
the present arrangement, the writer believes that such difficulties as those referred to
are, in the ease of the Australian radiation, more than compensated for by other
advantages. It may be observed, in the first place, that in utilizing the characters
of the dentition and foot-structure we are dealing with organs whose adaptive modi-
fications are, to a much greater extent than those of others relating to the internal
organization, of a definite and irreversible stamp. The Australian Marsupials represent,
furthermore, a peculiarly homogeneous group. During their development they have
been protected from competition with other animals. They have apparently been free
from such disturbing conditions as result from the development of a large carnivorous
element. Their range of lateral radiation has been restricted by their geographical
limitations; so that their progressive development has not been greatly disturbed by
divergence. Finally, their evolution has not proceeded to the stage of over-population
and the obliteration of less specialized intermediate types. Under these conditions we
may expect to find—as, except in a few cases, we actually do—that the adaptive modi-
fications follow oae anuther with a precision which leaves no doubt as to the manner in
which the evolution began and the direction in which it is proceeding.

It will be borne in mind, however, that the present arrangement is morphogenetic
rather than truly phylogenetic, because, while doubtless supported in different ways by
the modifications of other organs *, it is based solely on the dentition and foot-structure,
whereas true phylogeny must account for the modification of all parts of the organization.
Hence, while genera or species indicated as ancestral or derived may be actually so in a
phylogenetic sense, they are only advanced as such so far as the morphological differen-
tiation of their dentition and foot-structure is concerned.

Tur ADAPTIVE MODIFICATIONS OF THE DENTITION IN THE AUSTRALIAN
MARSUPIALS.

The following details of nomenclature may be noted :—

With reference to the cheek-teeth, it is a familiar fact that in the Marsupials the
numbers of premolariform and molariform teeth are the reverse of those in the Placentals,
being respectively 3:4 in the former and 4:3 in the latter. The question of their serial
homologies is one which has given rise to considerable controversy ; and several plans
for the nomenclature of marsupial teeth have been proposed. A consideration of the

* A paper on the general correlation of changes in the mammary apparatus is in course of preparation.

OF THE AUSTRALIAN MARSUPIALIA. 89

opinions expressed by Thomas (1887), Leche (1892, 1893, 1895), Woodward (1896),
Wilson and Hill (1897), Dependorf (1898), and Lydekker (1899) leads to no definite
conclusion ; and it therefore appears inadvisable at the present time to adopt any system
of nomenclature which will imply, intentionally or otherwise, definite homologies with
the Placentals. Thomas (1895) has recently adopted the plan of designating the three
premolariform teeth of Marsupials, in the order of their position from before backwards,
as anterior, middle, and posterior; and this plan has been adopted throughout the
present paper (with the substitution of the term median for middle), as presenting the
advantage over numerical systems that it conveys the idea of homologies only within
the limits of the Marsupial group. It may be observed that even these terms have
some disadvantages, since, in order to express homologies, it has been necessary in
several cases to disregard their descriptive applicability. In the advanced members of
the Dasyurine, for example, where the posterior premolars have disappeared, the
remaining teeth are still distinguished by the terms anterior and median.

Fig. 1.
pr pl.
' Hi ee La Pe.
> me. me,
HL i

Cusp-nomenclature of upper and lower molars.

A, B, crown and profile of upper, C, D, of lower molar of Peratherium ; E, F, upper and lower molars of Trichosurus ;
G, lower molar of Perameles Bougainvillei ; H, upper molar of Pseudochirus ; I, J, upper and lower molars of
Macropus. Abbreviations: pr., protocone; pa., paracone; me., metacone ; hy., hypocone ; pl., protoconule ;
ml., metaconule; a, b, hb, ¢,, ¢, ¢,, external styles; pr¢., protoconid ; pat., paraconid ; me?., metaconid ;
hy*., hypoconid ; en%., entoconid ; hl¢., hypoconulid; a.e.s., antero-external shelf.

The molariform teeth of the adult animals have been designated throughout as first,
second, third, and fourth molars, the first tooth being regarded as a true molar rather
than as a formerly deciduous premolar whose successor has been aborted (Lydekker,
1899).

The nomenclature used in describing the patterns of the molars is that proposed by
Osborn (1891) in connection with the tritubercular theory of molar evolution. An
exception has, however, been made in the case of the external cusps or “ styles” in the
upper teeth of polyprotodont forms. Certain of these elements, as they occur in the

90 DR. B. A. BENSLEY ON THE EVOLUTION

Cretaceous Mammalia and some of the Ungulata, have received special designations ;
and three of them, as they occur in Marsupials, have been shown by Winge (1882) to
have definite homologies in the Placentals. Careful study, however, shows that in the
Marsupials no less than six of them are represented, and all of these may occur together
even in primitive forms (certain Didelphyide). It has therefore seemed preferable to
express their homologies in the Marsupials by designating them alphabetically, rather
than to extend the nomenclature without a reconsideration of their number and
arrangement in the Placental series. The main points of cusp and style nomenclature
are illustrated in the accompanying diagram (fig. 1, p. 89).

DASYURIDA.

Viewing the dental characters of the Australian Marsupials from an adaptive rather
than a taxonomic standpoint, the Dasyuride may be broadly distinguished from the
remaining members of the series by the fact that they present a trend of evolution
involving a progressive change from primitive insectivorous to specialized carnivorous
modifications. The position of the family in the Australian radiation is approximately
equivalent to that of the main insectivorous-carnivorous line represented by the
Creodonta, the existing Talpidee, and the Carnivora in the general placental radiation.

All of the Australian Marsupials—and the same appears to be true of mammals
generally,—which are not at the present time in an insectivorous phase of dental
development, have passed through that phase in the course of their evolution. We
accordingly find in the smaller insectivorous members of the present family dental
modifications which are prototypal not only to those of the more advanced carnivorous
forms, but also, to a considerable extent, to those of the omnivorous Peramelide and
Phalangeridze, forms which are in turn prototypal in many respects to the members of
the herbivorous section. The smaller Dasyuridze would, no doubt, be wholly prototypal
in dentition, were it not for the fact that the Peramelidze present a more primitive
arrangement of the external cusps of the upper molars and a more primitive condition
of the upper incisor formula. In other words, the prototypal dental condition is nearly,
but not quite, realized among the Australian Marsupials in the smaller forms of the
Dasyuridee. It is only in Peratherium and the existing Didelphyide of South America
that all of the prototypal characters relating to the dental evolution of the Australian
Marsupials may be found associated in a single form.

The dental evolution of the Dasyuridz is not entirely homogeneous; two of the
constituent genera—namely, Thylacinus and Myrmecobius—may easily be shown to have
undergone independent development. Zhylacinus has undergone a carnivorous evolution
resembling so closely that of certain Neotropical forms (Sparassodonta of Ameghino),
while differing from that of the typical Australian members of the Dasyuride, as to throw
- doubt on the propriety of its inclusion in the latter family. Myrmecobius has undergone a
special development, characterized by incipient retrogression of the dentition, as a result
of the adoption of the ant-eating habit.

OF THE AUSTRALIAN MARSUPIALIA, on

The various genera may be arranged on a basis of their molar and premolar characters
as follows :—

A. Protocone of upper molars well developed ; external styles obsolete. Lower
molars with broad shelt-like talonid; no metaconid. Posterior premolars

well developed Pete es oe ee hy taGininse,

Genus: Thylacinus.
B. Protocone of upper molars variable; external styles well developed. Lower
molars with basin-shaped talonid, which is well developed or reduced ;

mietaconid present, variable. Posterior premolars variable or absent.
a 4 ;
a. Dentition normal; molar formula ei suede) Oth sy el ah Pore ee a aa. Dasy urine.
Genera: Sminthopsis, Antechinomys, Phascoyale, Chetocercus,

Dasyuroides, Dasyurus, Surcophilus.

6. Dentition retrogressive ; molar formula = Sele nn eye coum
Genus: Myrmecobius.

DasyURIN».

From the standpoint of their dental sequence, the Dasyurinze may be regarded as
forming a continuous progressive series. Every gradation is presented between a
primitive insectivorous condition and one indicative of high carnivorous specialization ;
there are no divergent developments of any significance, and the transformation takes place
without reference to generic distinction. ‘The evolution appears to be closely connected
with increase in size of the body, and carnivorous characters are for the most part
preceded by carnivorous habit *.

The general relations of the various forms appear to be as follows:—The two genera
Sminthopsis and Antechinomys in general represent the smallest and most primitive
members of the series. Although differing in no essential characters of the dentition from
the smaller species of Phascogale, they lack the range of modification which is characteristic
of the latter genus, and which prevents it from being prototypal to the same extent.
The larger species of Phascogale present specialized characters, such as the reduction of
the posterior premolars, which make them transitional between the smaller species
of the same genus and those of Dasywrus, to which they bear an ancestral relation.
Of the species of Dasywus, the form which approximates most closely to Phascogale is
D. hallucatus, this animal almost repeating the dental characters of ?. Wallacei or
P. Thorbeckiana. Of the remaining species, D. viverrinus and D. Geoffroyi are inter-
mediate, both in size and dental characters, between J. hallucatus and PD. maculatus.
The dental characters of the last-named form are exactly prophetic of those of Sarcophilus
ursinus. 'The two genera Chelocercus and Dasyuroides are simply terrestrial modifications
of Phascogale, and approximate closely in their dental characters to the larger species of
that genus and to Dasyurus hallucatus.

* The familiar accounts given by Gould (1863) and Lydekker (1894) of the habits of these and other Australian
Marsupials have been extensively supplemented by Semon (1896).

SECOND SERIES.—ZOOLOGY, VON. IX. 14

92 DR. B. A. BENSLEY ON THE EVOLUTION

Sequence of Molar Patterns.—The molar characters are remarkably constant in
the smaller species of the Dasyurinze; so that, so far as these teeth are concerned,
any one of them might be selected as representing the starting-point for the insectivorous-
carnivorous evolution. The following description is based on Sminthopsis leucopus, the
teeth of which are represented in Pls. 5 & 6 by figs. 2 a, 6 of the second upper and third
lower molars respectively.

The second upper molar presents the following characters:—The biting-surface is
triangular in outline, with the apex of the triangle directed internally. The latter is
separated from those of the neighbouring teeth in front and behind by triangular
spaces, into which are inserted, when the jaws are closed, the anterior triangular
pillars of the lower teeth. The longest axis of the crown is transverse, so that the
tooth appears to be slightly compressed in an antero-posterior direction. The crown
presents the trituberculate * pattern in general common to the insectivorous Dasyuridee
and Didelphyidew, and also to the placental Talpidee, the primitive Creodonta, and
the Cretaceous Mammalia. It bears three main cusps and an outer row of subsidiary
styles. Of the three main cusps, that placed internally, the protocone (pr., Pl. 5,
fig. 2a), is well developed and is supported on a separate root; its crown surface is
placed slightly below that of the two remaining cusps of the triangle. Of the latter
the anteriorly situate paracone (pa.) is comparatively small, while the posteriorly placed
metacone (me.) is greatly enlarged and forms the most important element of the crown.
Its posterior border is modified so as to form a trenchant spur, which shears against the
antero-external border of the triangular pillar of the succeeding lower tooth. The
shearing-edge is relatively short, and its direction is for the most part transverse. All
three cusps are of a piercing or insectiverous type, their tips being triangular in section
and sharply pointed.

The structures referred to as external styles are practically two in number, and are
separated by a considerable spac: from the outermost of the main cusps. One of
them (ab), placed opposite the paracone, shows indications of a composite structure.
‘The remaining style (¢) is much better developed; its tip projects almost te the same
extent as that of the adjacent metacone.

* Objection to this designation has been taken by Dr. Forsyth Major (1893, p. 199) on the grounds that so-called
trituberculate teeth are often in reality polybunous. ‘The justice of this view 1s obvious, such a tooth as that of
Sminthopsis leucopus being no more actually trituberculate than (if one may be allowed to cite an analogous case
of descriptive inapplicability) the Edentata are edentulous. The term is, however, of too great descriptive
conyenience to be easily dispensed with when employed to designate such teeth as those presenting three main cusps
arranged in the form of a triangle, with or without a complement of intermediate conules or external styles,

Descriptive terms such as the present indicate definite states rather than progressions, and the evolutionist
is hampered by the necessity either of extending their meaning to more than they literally signify, or of adopting a
cumbrous nomenclature to designate minor modifications. As to how far a tooth may be modified away from its
original type and still be designated as trituberculate is a question of individual opinion. The term can obviously
not be applied to those teeth which have passed the tuberculate stage in the carnivorous evolution, and should not
be applied to those which possess an incipient hypocone (cf. Pl. 5. fig. 8), indicative of an omnivorous evolution,
in view of the fact that the latter element, in subsequent stages, comes to rank in importance with the three
original cusps.

OF THE AUSTRALIAN MARSUPIALIA. 95

Of the remaining upper molars, the first differs from the second in being less com-
pressed antero-posteriorly, while the reverse is the case with the third. In the latter,
also, the styles are not so well developed, this being especially the case with style ce.
The fourth tooth is greatly reduced through the lack of development of the metacone,
there being no lower teeth situate behind it, against which it might shear. The cusps
represented are the protocone, paracone, and style ab. The paracone is trenchant
anteriorly.

The third lower molar (PI. 6. fig. 2) shows the following characters:—The crown is
relatively short and broad, and presents the tuberculo-sectorial pattern, being composed
of an anterior triangular pillar (trigonid) and a posterior heel (talonid). The sectional
area of the trigonid is slightly greater than that of the talonid. The former bears three
cusps, of which the externally placed protoconid (pr*.) is best developed, while of the
two internally placed cusps the posterior metaconid (me’.) is smaller and the anterior
paraconid (pa*.) still more so. The antero-external side of the trigonid, formed by
the combined edges of the paraconid and protoconid, is trenchant and shears against the
metacone-spur of the preceding upper molar. The trigonid also bears a small antero-
external shelf (q.e.s.).

The talonid is basin-shaped, and its edge bears one outer and two inner cusps, repre-
senting respectively the hypoconid (hy’.), hypoconulid (4/".), and entoconid (en*.). The
last-named cusp is, however, vestigial.

Of the remaining lower molars, the second repeats the characters of the third. The
first tooth (cf. Pl. 6. fig. 6, Sméinthopsis crassicaudata) presents the appearance of
transverse compression, the paraconid and metaconid being partially reduced. As
pointed out below, this tooth shows an interesting evolution connected with a gradual
change to a premolariform condition, in which the condition in the present species
represents the initial stage. The fourth molar shows a reduction of the talonid by
lateral compression, the whole structure having the form of a spur attached to the
posterior border of the trigonid.

Among the smaller Dasyurinze the departures from the type just described are few and
unimportant. There are certain minor variations in the characters of the external styles
to which reference will be made in a subsequent section. ‘he relative proportions of
the sectional areas of trigonid and talonid differ sliglitly in different cases. The vestigial
character of the entoconid seen in Siminthopsis leucopus is apparently common to all of
the species of that genus except S. erassicaudata, in which this cusp is well developed.
Jt is also found in Phascogale minutissima and in Antechinomys laniger. In the ease
of Sminthopsis and Antechinomys the apparently vestigial condition of the eutoconid
may in reality be due to mechanical wear, the animals being ground-feeding types.
The cause of the reduction in the case of Phascogale minutissima is seen in the reduced
character of the talonid, the latter structure being possibly in process of reduction
in this form. The small size of the entoconid is, in any case, a secondary character,
this element being well developed in all of the larger forms, as in S. crassicaudata, as it
is in the primitive Peramelide and all of the Didelphyide.

On examining the molar modifications of the larger species of Phascogale, we find

14*

94. DR. B. A. BENSLEY ON THE EVOLUTION

that the only important change from the condition in Sminthopsis leucopus relates to
inerease in size*. That this character is an essential feature of the carnivorous
evolution is seen from the fact that it is only after passing through the successively larger
species of Phascogale into the genus Dasyurus that we meet with animals of sufficient
size to present carnivorous characters. Chetocercus cristicauda, Dasyuroides Byrnei,
and Dasyurus hallucatus show no departure from the conditions in Phascogale; but,
beginning with the still larger forms D. viverrinus and D. Geoffroyi, and passing
through D. maculatus to Sarcophilus, we find the molar characters changing as
follows :—Taking the second upper tooth as a basis of comparison of the upper molars
(cf. Pl. 5. figs. 3 & 4, Dasyurus maculatus and Sarcophilus ursinus), there is (@)
reduction of the protocone; (4) enlargement of the metacone, with lateral compression
of its tip to form a trenchant blade, and encroachment of its base on the protocone-
root; (¢) elongation of the trenchant metacone-spur and rotation inwards of its distal
extremity, so that it comes to shear inwards rather than backwards, as in Sminthopsis ;
(d) lateral compression of the whole tooth, so that the external styles become
approximated to their respective cusps.

Throughout the series the remaining upper molars differ from the second in exactly
the same way as those of Sménthopsis leucopus. The first tooth always presents a
condition of greater lateral compression. In Sarcophilus the external styles, which
even in the second tooth are so closely approximated to their respective cusps that the
tooth practically possesses a double cutting-edge, are in the first only separated from the
latter by a narrow groove, style ab being almost indistinguishable from the paracone.
The third tooth always presents a condition of less transverse compression, and style ¢,
which was seen to be reduced in Sminthopsis, becomes quickly obliterated. The fourth
molar becomes further reduced by the reduction of the protocone.

The extent of rotation of the metacone-spur, which forms such a conspicuous feature
of the change, will be more apparent from the following data:—A line passing through
the tip of the metacone and the distal extremity of its spur, or, in other words, along
the shearing-edge, will pass through a tooth of the opposite side of the Jaw—a posterior
one in the case of the smaller primitive forms, and an anterior one in the case of the
larger carnivorous ones. Such a line, drawn through the metacone-spur of the third
molar of one side, will, in the case of Sminthopsis leucopus, Phascogale Wallacet, or
Dasyurus hallucatus, pass through the second molar of the opposite side. In D. viver-
vinus it will pass between the first and second molars; in D. maculatus through the
anterior part of the first molar, or between this tooth and the last (morphologically
median) premolar. In Sarcophilus it will pass through the posterior portion of the canine,

Taking the third tooth as a basis of comparison, we find the following changes in
the lower molars (cf. Pl. 6, figs. 38, 4, & 5, Dasyurus viverrinus, D. maculatus, and
Sarcophilus wrsinus):—There is (a) enlargement and lateral compression of the trigonid ;
(b) conversion of the tips of the paraconid and protoconid into trenchant blades ;
(c) rotation outwards of the shearing-edge formed by the adjacent edges of these cusps;
(d) reduction of the metaconid; (e) reduction of the talonid,

* (Of. general sequence in length of the tooth-rows given on p. 98.

OF THE AUSTRALIAN MARSUPIALIA, 95

In the first molar (cf Pl. 6. figs. 7 & 8, Dasyurus Geoffroyi and Sarcophilus
ursinus) there is a further reduction and final obliteration of the paraconid and
metaconid, the tooth assuming a premolar condition. In the fourth molar the talonid
becomes reduced to a vestige.

That the above-mentioned modifications, with the partial exception of those of the
first lower molar, represent the successive phases of an insectivorous-carnivorous
evolution, can be readily shown by reference to their functions. The minute teeth of
Sminthopsis leucopus serve three distinct purposes, adapting them for the comminution
of insect-prey. A piercing action is performed by all of the cusps of the relatively
broad crowns, a shearing one by the working of the metacone-spur of the upper molars
against the paraconid and protoconid of the lower, and finally a crushing action is
performed by the working of the protocone of the upper molars into the talonid of the
lower. In the extreme members of the series both the piercing and crushing actions
are abandoned, and the shearing action perfected. The lapse of the crushing function
is indicated by the reduction of the protocone and talonid, while the lapse of the
piercing action and the perfection of the shearing one is proclaimed by the lateral com-
pression of the teeth and the conversion of the originally conical cusps into trenchant
blades. The reduction of the metaconid is also attributable to the lapse of its piercing
function, there being in the advanced stages no cusps against which it might shear.
The rotation of the metacone-spur is no less a character of carnivorous adaptation than
the trenchant modifications of the cusps: by means of it the teeth are made to shear
together in a single longitudinal line instead of individually in parallel transverse lines,
as is the case with Sméinthopsis and its allies.

The changes which take place in the first lower molar, and result in the conversion of
a typical tuberculo-sectorial tooth into one of premolariform character, are associated
with the reduction of the posterior premolars shortly to be described, and are therefore
only indirectly connected with the carnivorous evolution. In the smaller species of
Phascogale, and in Sminthopsis and Antechinomys, where the upper posterior premolars
are well developed, or at most only slightly reduced, the latter teeth shear against the
protoconid and paraconid of the first lower molars in much the same manner that the
metacone-spur does in the succeeding teeth. With the reduction of the posterior
premolars in the larger forms this function is obliterated, and the protoconid loses its
shearing character, while the paraconid disappears. The metaconid becomes reduced
for the same reason as in the remaining teeth. In Dasywrus and Sarcophilus, where the
posterior premolars are entirely absent, the first lower molars bite loosely against the
median upper ones.

Incisors.—These teeth present a carnivorous evolution almost as marked as that just
described for the molars. The change affects chiefly the relations of the median upper
teeth and the length of the tooth-rows. Throughout the species of Phascogale the
median upper incisors are sharply differentiated from the lateral teeth, being elongated,
procumbent, and subcaniniform. They are separated at their bases and approximated
at their tips. The lateral incisors are short and of the usual spatulate type, except that

96 DR. B. A. BENSLEY ON THE EVOLUTION

their tips are slightly pointed. In the species of Sminthopsis there is usually a more
or less marked tendency for the median upper teeth to be separated at their tips as
well as at their bases, and to project downwards rather than forwards. This
condition is to be regarded as special rather than as primitive. Dasyurus hallucatus
shows much the same condition as Phascogale, but following the remaining species of
Dasyurus we find the median upper incisors undergoing a gradual reduction by
which they come to resemble the lateral teeth. In Sarcophilus they are indistin-
euishable from the latter, except for their slightly more rounded section and their basal
separation.

The differentiation of the median upper incisors in the smaller Dasyurine represents
an insectivorous adaptation developed for the purpose of making the terminal teeth
more serviccable in the capture of insect-prey, and their subsequent return to a normal
spatulate condition, in the course of the carnivorous evolution, is due to the lapse of
their original function. The reduction of these teeth is closely associated with a change
in the relative length of the incisor rows. In the smaller Dasyurine the latter are
relatively long, and meet in the middle line at an acute angle, so that the modified
median teeth are placed in the most advantageous position. Passing through the
species of Dasyurus in the order above mentioned we find a successive shortening of the
incisor rows, until in Sarcophilus they form an almost straight line across the front of
the muzzle. As mentioned below, there is also to be seen, in proceeding through the
same series, an increase in the functional importance of the canines, indicating a
gradual transference of the piercing and grasping functions from the median incisors
to the latter teeth. The shortening of the incisor rows is in part connected with
the thickening of the muzzle for the support of the enlarged canines, but is more
especially developed for the purpose of placing the incisors in a position where they
will least interfere with the action of the canines, namely as close as possible to the
canine roots.

Many of the placental Insectivora present a close resemblance to the Dasyurine in
respect to the differentiation of the median upper incisors. For example, in the
Soricidee, and in Centetes and Chrysochloris, the median upper incisors are much like
those of Phascogale, and the more specialized modifications of Smdénthopsis are repeated
in the genera Echinops, Ericulus, and Limnogale (Centetidve), and also in Macroscelides
(Macroscelidide). It is interesting to note that, with the partial exception of such
forms as Phascogale penicillata and its prototypal relative P. calura, the insectivorous
differentiation of the incisors in the Dasyurine is confined to the median upper teeth,
but that in many of the Insectivora it extends to other of the incisors as well. More
especially is this noticeable in the Soricidxe, where the median lower teeth are elongated
in such a way that they work against the median upper teeth after the manner of the
tips of a pair of forceps. This modification, while it is not indicated in the Dasyurine,
is highly characteristic of the Phalangeridee, and, as pointed out below, represents the
starting-point for the whole diprotodont modification characteristic of the last-named
family and its herbivorous derivatives. These facts taken together indicate that the
differentiated condition of the median upper incisors represents the first stage in the
insectivorous specialization of the incisors, and also that it represents a prototypal and

a

ia

OF THE AUSTRALIAN MARSUPIALIA. 97

not a special character in the Dasyurinee, since it must have been present in the
ancestral forms of the Phalangeridie.

Canines.—These teeth are well developed in all of the Dasyurine, but are relatively
weakest in Sntinthopsis and Antechinomys. In the carnivorous evolution, as exemplified
by Phascogale, Dasyurus, and Sarcophilus, they present a successive increase both in
actual and relative size. The increase in relative size is not at first sight apparent, but
is easily demonstrable by sketching to scale the profile aspect of the tooth-series in such
representative forms as Sininthopsis murina, Phascogale flavipes, P. Wallacei, Dasyurus
hallucatus, D. maculatus, and Sarcophilus ursinus.

Premolars.—In their general characters these teeth are of the simple piercing type
common to the Didelphyidze, Peramelidz, and primitive Phalangeridee (Acrobates and
Distaechurus), as well as to the placental Insectivora and Carnivora. In their special
characters they present a series of important modifications, which serve to support and
extend the plan of dental evolution as determined by the molars, at the same time
throwing light on the relations of the smaller forms with the remaining Marsupials,
especially the existing Didelphyide and Peratherium.

The chief modifications are as follows :—(a) successive reduction of the posterior
premolars, proceeding from a stage in which these teeth are the predominant elements
of the series (Sminthopsis and Antechinomys) to one in which they are altogether absent
(Dasyurus, Sarcophilus) ; (b) compensatory enlargement of the median premolars until
the reduction of the posterior teeth is completed; (¢) subsequent slight reduction of the
anterior and median premolars in the final carnivorous stages.

Some of the features of the premolar evolution of this group have already been
described by Thomas (1887), this writer having shown, by a carefully selected series of
diagrams (pl. 27. figs. 1-5), that the two-premolared condition in Dasyurus is only the
culminating stage of a process of reduction of the posterior premolars which is
exemplified by the species of Phascogale. Thomas has further pointed out that the
extent of reduction is always greater in the lower than in the upper teeth, so that in
certain cases, such as P. Thorbeckiana and P. apicalis, where the upper teeth are
vestigial, the lower may be entirely absent.

The modifications of the posterior premolars are of great interest, as showing much
more clearly than those of the remaining teeth the primitive position of Smdénthopsis and
Antechinomys as regards dentition. In both of these genera the posterior upper
premolars are always larger than the median teeth, and in some cases (S. leucopus,
A. laniger) disproportionately so, while in Phascogale, with the exception of P. minutis-
sima, in which they are again disproportionately larger, they are variable, being slightly
larger than, or equal in size to, the median teeth, or in some cases wholly vestigial. As will
be pointed out in greater detail in a subsequent section, Smdénthopsis and Antechinomys
make the closest approach to the prototypal condition found in Peratherium, where the
posterior premolars are apparently always predominant, and, furthermore, they take up
a position with reference to the latter genus approximately equivalent to that taken by
certain of the existing Didelphyide (Peramys).

98 DR. B. A. BENSLEY ON THE EVOLUTION

Relative size of upper Species | Length of upper |No. of specimens
posterior premolar. " tooth-row. measured.
|
| mm,
; : Phascogale minutissima .....+ 9°9 | 2
i, SRosterior) premolamyd1s prO-1\l|wemmcns ace eee | fe
: Sminthopsis hirtipes .......- 13 1
portionately larger than | < Teegees | eae 9
MCGLAN: veces sieneniereretens Bae ‘ee Z =
| | Antechinomys laniger........ We 2 R55 i
| L J
Sminthopsis crassicaudata.. .. 12 2
! .
e LON 5 oon Oe 12-13 4
: a3 3 MOCTOURGIN ta... « 13 1
II. Posterior premolar slightly 5 | 2 ae
2 Phascogale flavipes.........- 14:°5-15 3
larger than or equal to |< : a Bae 5)
eee Pe OURO, 5005000 15°5 2
ee eee: ws a Swaimsont.....-.- 16°5-17°5 3
es COVUrAe erent ee 16-17°5 2
IU pemeillata ...... 22-25 3
Phascogale Wallacet ........ 25 | il
: :

III. Posterior premolar smaller | | - Govsdlis, weteweitey ee 20 2
than median and be- |< AA Thorbechiana .... 26 1
coming vestigial ...... | Cheetocercus cristicauda ...... | ;

| | Dasyuroides Byrnei .......- 5
| ( Dasyurus hallucatus ........ 32-33 3
| as CANA Boonae oo 36-38 5
TV. Posterior premolar absent. . < 35 ViVErTINUS ........ 36-42 7
| Pe SNUCULALUS Wael tener ners 46-48 2
log ; 5 2 :
| \ Sarcophilus ursinus .1. 0.00 72 2

Like the carnivorous evolution of the molars, the reduction of the posterior premolars
is closely connected with increase in size of the animals, and the above table has
been arranged to show this by a comparison of the relative sizes of the upper posterior
premolars with reference to the median teeth, with the length of the upper tooth-rows.
The various species are here arranged in four groups in order to avoid unnecessary
deseription of the sizes of the premolars in each case. Within the different groups,
however, the forms have been placed where possible in their natural order. The lengths
of the tooth-rows have been measured in each case in a straight line from the bases of
the median incisors to the ends of the fourth molars.

Two notable exceptions to this plan are found in Phascogale apicalis and in
Sininthopsis macdonnellensis. In the former, two specimens measured showed the
length of the tooth-1ows to be 18 mm. ‘The posterior premolar is in an advanced stage
of reduction, so that the species presents a dental length characteristic of Group IT. and
the premolar characters of Group III. <A similar condition is found in the latter
species *, where, in animals of general proportions like the larger members of Group ILI.,
the posterior premolar is greatly reduced.

With regard to the premolar transformations in the later carnivorous stages, it has
already been mentioned that while during the reduction of the posterior premolar there
is a compensatory enlargement of the median tooth, there is subsequently a slight
reduction both of the latter tooth and of the anterior premolar. Neither of these teeth

* Only spivit-specimens of this species and of Chetocercus and Dasywroides were available, so that the tooth-rows
could not be exactly measured.

a ee

—

OF THE AUSTRALIAN MARSUPIALIA. 99

is as functional in the carnivorous forms as in the insectivorous prototypes. The
explanation of this apparently anomalous condition is doubtless to be sought in the
progressive increase in the functional value of the canines. In the insectivorous forms
the latter teeth are at most only moderately developed, so that the premolars possess
their full functional value as grasping and piercing organs. In the carnivorous forms
these functions become largely usurped by the enlarging canines, and the incisors
and premolars become reduced. In Sarcophilus and Dasyurus the premolars present
an appearance as if the material formerly used for their development had been utilized
in the development of the canines.

MyYRMECOBIIN®.

The dentition of the single representative of this subfamily is of great interest on
account of its marked departure from the usual conditions in the recent Mammalia, in
respect to the peculiar characters and extraordinary number of the molar teeth, and the
alliance thereby suggested between MWyrmecobius and the Mesozoic Mammalia.

It is a noteworthy coincidence that the original description of MJyrmecobius by
Waterhouse (1836) was published only shortly before a vigorous discussion, marked by
the opinions of such eminent zoologists as de Blainville, Owen, and Agassiz, took place
as to the nature of the first of the famous jaws from the Stonesfield Slate. The
discovery of Myrmecobius, with its peculiar dentition, furnished Owen (1836) with a
strong argument in favour of the view that their affinities were not only mammalian, but
also marsupial. Owen repeated his opinion in his subsequent publications (1846, 1871),
and even went so far in his estimation of the primitiveness of Myrmecobius as to suggest
a community of dental characters between it and the theriodont reptile Galesaurus
(1887). Following Owen, Thomas (1888, p. 312) has characterized Myrmecobius as an
“unmodified survivor from Mesozoic times,” and its possible affinities have been
commented upon by other writers in almost every case in which the Mesozoic mammals
have been described.

New interest has lately been added to the question through the discovery by Poulton
of the true teeth of Ornithorhynchus, the patterns of which were compared by that
writer (1888, p. 20) with those of Wyrmecobius, and by Cope (1888, p. 259) with those of
the Multituberculata. Following Poulton’s comparison, Leche (1891, p. 152) suggested
a possible community of type between the molars of the Multituberculata and those of
the Dasyuride, through Wyrmecobius, although he subsequently (1893, p. 114, footnote)
withdrew his opinion as to the resemblances between Ornithorhynchus and Myrmecobius
on the publication of Stewart’s description (1891) of a specimen of the former animal
in the collection of the Royal College of Surgeons with more complicated teeth
than those in the specimens previously described by Poulton (op. cit.) and Thomas
(1889).

Winge (1882, 1893) has expressed the opinion that the relations of IZyrmecobius are with
the Dasyuride, the extraordinary number of the molariform teeth being in bis estimation
the result of a retention of the normally deciduous premolars. Leche (1891), however,

SECOND SERIES.— ZOOLOGY, VOL. 1X. 15

100 DR. B. A. BENSLEY ON THE EVOLUTION

dissents from this proposition, and believes the increased molar formula to represent a
primitive character inherited from Mesozoic forms *.

In a former paper (1901 @) the writer expressed the opinion that the dental characters of
Myrmecobius have been derived by retrogression from those of the normal Dasyuride,
and this position has been amply confirmed by the examination of the extensive series
of specimens in the British Museum collection. Many of the characters of the incisors,
canines, and premolars, which appear at first sight to be primitive, are repeated in the
Peramelidee, where they are undoubtedly the result of retrogression. The patterns of
the lower molars are directly derivable from those of the smaller Dasyurine. In fact,
the only real difficulty in the derivation of the dentition relates to the patterns of the
upper molars, these being so extremely variable that the cusp-homologies are difficult
of determination. Nevertheless, even in them it is possible to recognize a ground-type
running through the variations, and this ground-type represents approximately the
pattern characteristic of the normal Dasyurine.

With reference to the origin of the excessive molar formula in Myrmecobius, it will be
seen that there are three possible explanations :—(a) That it is due to the retention of the
normally deciduous teeth (Winge); (0) that it represents a primitive condition carried
over from the Mesozoic Mammalia (Owen, Leche, and others); (¢) that it is due to
a simple reduplication of teeth from the posterior portion of the dental lamina, or to a
reappearance of formerly vestigial teeth in the same region.

As to the possibility of a retention of the deciduous teeth, a description will be found,
on pp. 106-107, of a young specimen in the collection in which the posterior premolars
have not yet begun to develop; and of three molariform teeth already formed, the first
are comparatively minute, while the second are approximately equal in size to the unworn
first molars of the adolescent animal. This appears to indicate that the minute first
molariform teeth represent deciduous premolars, which are afterwards replaced. It may
be observed that, even assuming a retention of the deciduous teeth, we would still have
to account for the occasional presence of an additional lower molar.

With reference to Leche’s suggestion that the increased formula is a primitive
character, the writer believes it to be counterindicated by the following facts :—(a) The
extreme variability of the dentition and the abundant evidence of retrogression from a
normal Dasyurine condition, added to the fact that there is an intimate correspondence
between Myrmecobius and si Dasyuridze in other respects, and more especially in the

reduced incisor formula ote 3 indicate that the relations of the animal are in every way

with the latter family. That this relation is, however, not an ancestral one, as the
increased molar formula would seem to indicate, is apparent from the fact that the
ancestral modifications of the Dasyuride are to be found in Didelphyide, which forms
present a more primitive condition of the incisor formula than is found in Myrmecobius.
(b) The data collected by Bateson (1894) show that reduplication of teeth may occur

* «Bs scheint mir somit festzustehen dass das Myrmecobius-Gebiss, was die Form der Backen-Zihne betvrifft,
theilweise reducirt ist, und demselben Typus wie dasjenige der Dasyuride und Ornithorhynchus angehart, dass
aber die gréssere Anzahl etwas Primitives, von Mesozoischen Siiugethieren Ererbtes ashen

OF THE AUSTRALIAN MARSUPIALIA. 101

without reference to homology, while the observations of Thomas (1888) and Allen (1901)
show, further, that an excessive number of molars may be found as a variation in normal
Marsupials (Bettongia and Didelphys). (c) The minute size of the molars in Myrmecobius
and the elongation of the palate and lower jaw offer just those conditions favourable for
the intercalation of new teeth, or, as indicated above, their origin by reduplication from
the posterior part of the dental lamina.

Unlike Leche, the writer sees nothing incompatible in the recognition of a retro-
gressive development of the already existing teeth and the synchronous addition of new
ones. The redaction of the function of an organ is not immediately followed to its
obliteration, the latter only taking place after a longer or shorter period of attempted
development. While, therefore, the molar teeth of Myrmecobius may be found in
a semi-reduced condition, the mere fact of their presence is sufficient evidence of their
tendency to develop in a normal way. It is accordingly reasonable to suppose that
under the favourable conditions of increased space in the molar region, the same
tendency as is seen under less favourable conditions in Bettongia and Didelphys to
produce new teeth should not only be present, but also be more strongly marked. In
fact, the existence of such a tendency is in evidence in the case of the antemolar teeth,
although the latter are admittedly less reduced than the others. Two of the British
Museum specimens already mentioned by Thomas (1888) show four instead of three lower
incisors, and Leche (1891) has described a specimen in which there is an additional
premolar in the left ramus of the lower jaw behind the posterior premolar. In the
Peramelidz we find several instances in which reduction of the canines is accompanied
by the appearance of new basal cusps, and an analogous case is seen in the Didelphyide,
where in Caluromys a retrogressive development of the external styles of the upper
molars (¢f. Pl. 5. fig. 27) is proclaimed by a reduction of the larger elements normally
present, and the development of a large number of smaller ones.

No surprise need be expressed that such new teeth should resemble in their characters
the already existing ones. The results of Bateson (1894) show that in the intercalation of
new teeth without definite homologies the same principle determines their patterns as
those of the normal teeth of the region in which they occur.

With reference to the comparisons which have been made between the molars of
Myrmecobius and those of Ornithorhynchus and certain of the Multituberculata, there
is no doubt that certain resemblances exist, but they are of much too general a kind to
be interpreted as indicating affinities. A careful study of the patterns presented by
Plagiaulax, Microlestes, Ornithorhynchus, and Myrmecobius fails to reveal any evidence
of homologous cusps. The fact, already referred to above, that the relations of
Myrmecobius are in most respects with the Dasyuridee, whose prototypal characters are
those of the Didelphyide, strengthens the view that such resemblances as do exist are
merely the result of convergent development. The molars of Myrmecobius differ from
those of the normal Dasyuride chiefly in lacking the angularity both of the general
contour and the constituent cusps, which, in the latter, is due to the mechanical precision
with which the teeth are fitted together. It is exactly in these obviously secondary

15*

102 DR. B. A. BENSLEY ON THE EVOLUTION

features that Myrmecobius resembles the plagiaulacid Multituberculata. As to what
kind of adaptations the teeth of the latter animals represent is wholly obscure. In the
writer’s opinion the tabulate teeth of Ornithorhynchus ave highly specialized, and
represent an adaptation for crushing the shells of small molluses. They are comparable,
in a general way, with those of the Trichechidee, or, better, the Sea-Otters. The excessive
crenulation of the molar margins in the specimen described by Stewart (1891, pl. 8),
which enhances their multituberculate character, is undoubtedly secondary, and repre-
sents a development of much the same order as that seen in the teeth of the Suide and
Ursidee. Without expressing an opinion as to the Multituberculata, the writer considers
the superficial resemblances existing between Myrmecobius and Ornithorhynchus to be
referable to different adaptations, and therefore not only secondary, but also much less
worthy of being designated as convergent developments than, for example, those seen in
the molars of the Phalangeridee and Primates, where similar quadrituberculate teeth
have been evolved independently for similar purposes.

Before considering the dental characters of MWyrmecobius in detail, it may be observed
that the whole dentition is extremely variable, and that no adequate conclusions can be
formed from a study of single specimens. The variability is much more pronounced in
the upper molars than in any of the remaining teeth, there being in them not only no
correspondence between the homologous teeth of different individuals, but also none
between those of opposite sides of the jaw in the same animal. The cause of the
variability is partly to be sought in the uneven mechanical wearing to which the teeth
are subjected, through the presence of particles of earth in the food. This is especially
noticeable in the case of the molars, where, as Leche (1891, p. 151) has shown, the outer
parts of the lower teeth and the inner parts of the upper ones tend to be obliterated
with age, while the remaining parts are, in each case, only slightly affected. Beyond
this, however, there is a variability of an inherent kind which is not the result of wear,
but of the retrogression which is taking place throughout the dentition. It is interesting
to note that a similar tendency towards variability is seen under similar conditions in the
Peramelidee.

Molar Patterns.—The lower molars are much more constant in their characters than
the upper, and also depart to a less extent from the normal dasyurine type, so that they
may be more conveniently described first.

As already mentioned, apart from their minute size, the molars of Myrmecobdius differ
from those of the Dasyurine chiefly in lacking the angularity both of the general
contour and of the constituent cusps. The lower teeth (Pl. 6. fig. 9) are roughly
oval in section, and their cusps are either conical or slightly curved. ‘The crown of an
unworn posterior tooth shows exactly the same number and arrangement of the cusps
as is seen in the teeth of normal dasyurine forms (cf. fig. 3), there being in all six
cusps, of which the three anterior together represent a trigonid, while the remaining
three represent a talonid. ‘These two portions of the tooth are not differentiated as in
normal forms, the reason being partly that the crown-surface of the talonid is placed

OF THE AUSTRALIAN MARSUPIALIA. 103

almost at the level of that of the trigonid, and partly that the cusps of both portions
have much the same characters. In fig. 90 is represented a tooth which had not yet
appeared above the bone, and in this the trigonid will be seen to show indications of the
angularity characteristic of normal forms. This specimen is of great interest, not only
as indicating the former presence of this condition in the lower molars, but also as
implying a similar condition in the more highly modified upper teeth. In some cases it
is possible to recognize in the lower molars an anterior tubercle representing the antero-
external shelf characteristic of normal forms.

The most peculiar feature of the lower teeth is, however, the marked differentiation
between the outer and inner cusps, the protoconid and hypoconid being reduced, while
the paraconid, metaconid, entoconid, and sometimes the hypoconulid are large and
elongated. These relations are exactly the reverse of those in the Dasyurinee. As
Leche has pointed out, the condition is partly attributable to mechanical wear. In the
anterior molars, which, being older, have been in use for a longer time, the protoconid
and hypoconid are found in a much more advanced stage of reduction than in the newer
posterior teeth. In extreme cases, where the outer cusps have been worn down to their
bases, the teeth present a curious appearance ; when viewed internally in profile they
appear triconodont, or, in cases where the hypoconulid is well developed, quadri-
conodont, the cusps being arranged in a linear series from before backwards. ‘The basal
ledge, which formerly supported the two outer cusps, presents the appearance of a wide
external cingulum. The reduced condition of the protoconid and hypoconid is not,
however, wholly attributable to mechanical wear, since the same tendency is seen in the
unworn teeth, although naturally in a lesser degree. It appears as if the repeated
obliteration of the outer cusps by wear had produced a development in that direction.

The first lower molar, believed by Winge to represent a formerly deciduous tooth, is
usually found in a much more reduced condition than the remaining teeth. Leche
regards this also as the result of use. In all of the older specimens which I have
examined, and also in the young specimen represented in Pl. 6, fig. 9c, the tooth
shows three or four cusps arranged in a linear series. ‘These apparently represent the
inner cusps of the posterior teeth, although apart from this their homologies are usually
obscure. Inthe young specimen represented in text-figure 3 (p. 106) of the second pair of
molariform teeth, that of the left side shows one outer and two inner cusps of doubtful
homologies, while that of the right side shows all of the cusps characteristic of the
posterior teeth. Itis probable that the same remark made for the posterior teeth is
true to a greater extent of the first, namely, that the reduction is partly a natural one
and partly the result of mechanical wear.

Examples of the patterns of the upper molars are given in Pl. 5, fig. 5. The cusp
designations, and also the descriptions here given, are based on a study of seven
specimens. As already intimated, these teeth present an infinite variety of minor
characters, but it is possible to recognize a ground-type apparently representing a
modification of the condition in normal forms.

Taking the intermediate members of the molar series as more typical than the exten

104 DR. B. A. BENSLEY ON THE EVOLUTION

anterior and posterior ones, we first notice the fact that as regards contour a roughly
triangular shape predominates, and that, further, the apex of the triangle is internal and
is formed of a large cusp (pr.), identifiable with the protocone in normal forms. In
some cases the protocone-like cusp is well developed, but the triangular shape is
obliterated by the great development of the more posteriorly placed cusps. In one case
the protocone was found to be considerably elongated and its edge crenulated. The
anterior portion of the tooth is usually occupied by one internal ( pa.) and two external
(ab and ¢) cusps. The anterior position of these elements is partially disturbed by the
occasionally anterior position of the protocone. The internal cusp, from its small size
and position, probably represents a paracone, while of the two external cusps the
anterior one, which stands in close relation with the paracone, is probably equivalent to
style ab of the Dasyurinz, and the more posterior one, judging from its large size and
position, is probably equivalent to style c. The paracone is often worn down and some-
times wholly absent. The posterior part of the tooth is usually occupied by two cusps,
one of which (me.), situate internally, is usually conspicuous for its large size. ‘These
cusps probably represent together the metacone of normal forms. ‘The reasons for this
view are: first, that their posterior position indicates a connection with the metacone ;
secondly, that they are sometimes connected by a trenchant edge (¢f. fig. 5a); and,
thirdly, that all of the well-developed cusps of the molars of Myrmecobius are elongated,
and that an elongation of such a cusp as the metacone of normal forms would very
likely result in the formation of an extra cusp from the distal extremity of its spur. It
must be admitted, however, that it is sometimes difficult to explain the relations of
these elements on this supposition, on account of a reversal of their relative sizes,
although it is probable that the latter condition is due in some cases to the wearing to
which the internal cusps are subjected.

As in the case of the lower teeth, the first molars are much simpler than the others.
The best-developed specimen examined showed four cusps arranged in a linear series,
the two intermediate members being much larger than the others ; and, in addition, two
minute cusps placed between the larger median elements, the one internally, the other
externally. The greatest departure from the condition in this specimen was found in
another case, in which the tooth formed a flat plate, with no indication of cusps beyond
that arising from a slight concavity of the edge. The homologies of the cusps, even in
the better-developed teeth, are extremely doubtful, except that they show a general
resemblance, when seen in profile, to the external styles in normal forms.

The general conclusion with regard to the derivation of the upper molars is that they
represent modifications of a trituberculate type. The examination of a larger series of
specimens is, in the writer’s opinion, necessary before this proposition can be accepted
as certain.

Incisors.—All of the incisor characters of J/yrmecobius are directly derivable from
those of the Dasyurinz or Didelphyide. The median upper teeth (text-fig. 2, a,b) are more
rounded in section than the lateral ones, and are also slightly procumbent, so that, while
they are not differentiated to the same extent as in normal insectivorous forms, they

OF THE AUSTRALIAN MARSUPIALIA. 105

show signs of having been so at an earlier stage. It is interesting to note that a similar
lack of differentiation characterizes the homologous teeth in the Peramelide (fig. 2, e).
The lateral incisors present a curious appearance, due to a subcaniniform modification
of their tips. This condition is only an extreme development of that seen in the
Dasyurinze and Didelpiyidee (fig. 2,¢), where the tips of the lateral teeth are already
acute and directed slightly backwards. A very similar appearance to that seen in
Myrmecobius has been figured by Flower and Lydekker (1891, p. 539) in the degenerate

The antemolar teeth of Myrmecobius fasciatus compared with those of other Marsupials.

a-c. Myrmecobius, upper and lower dentition of normal specimens; d. Lower dentition of abnormal specimen with
four lower incisors; e. Perameles obesula, upper incisors; f. Cheropus castanotis, upper incisors and canine;
g. Metachirus opossum, upper incisors and canine. Abbreviations: 7., incisors ; cn., canine : p-, premolars ;

m., molars.

carnivore Hwpleres, and a similar tendency towards a posterior extension of the tips of
the lateral teeth is found as a secondary character in certain of the Peramelide (fig. 2, e).
All of the lower incisors tend to be caniniform. Their contour is much less angular
than in the upper teeth, and a similar difference is observable in normal forms. The
median lower incisors are slightly enlarged, as in Phascogale penicillata *.

Canines.—These teeth show a distinct departure from the usual type found in the
Dasyuride, this being particularly the case with the upper ones. ‘The latter (fig. 2, b)
tend to be laterally compressed, and in some cases. present accessory anterior and
posterior cusps. In addition, they sometimes show a grooving of the root. ‘These
characters give them an almost premolariform appearance, and this is further enhanced
by the fact that there is a gradual reduction in size of the cheek-teeth proceeding from
the canine to the posterior premolar.

The fact that morphologically the canines are modified premolars might at first sight

* The incisors of AMyrmecobius, like all of the remaining teeth, tend to be separated by diastemata, and thus
present an appearance not unlike that seen in certain of the Mesozoic Mammalia. he condition is, however, a
purely secondary one, and is due to the elongation of the muzzle common to this and other ant-eating mammals,
It is interesting to note that even the comparatively advanced young show no indications of the elongated muzzle
characteristic of the adult, the facial region presenting a curious abbreviated appearance, which Leche has aptly
referred to as *‘ cine wirkliche Mopsform.”

106 DR. B. A. BENSLEY ON THE EVOLUTION

lead one to regard the condition in Wyrmecobius as primitive. That it is not so, however,
may be seen from a comparison of the Peramelide, certain species of which present
canine modifications of the same kind and even of a more extreme degree. An
example is given in text-fig. 2 of the teeth of Charopus castanotis. In all cases among
the Peramelidee the modification is not only secondary but also purely local.

Premolars.—Apart from a marked tendency to develop accessory cusps, which is
doubtless due to their separation by diastemata, these teeth present the same characters
as those of the Dasyurinz. As already mentioned above, they decrease in size from
before backwards. The exact explanation of this condition is doubtful; the small size
of the posterior tooth may be due simply to the circumstance that it forms an inter-
mediate member between the larger anterior teeth and the greatly reduced molars, or
that it has been formerly reduced as in the normal Dasyurinee. As regards the presence
of accessory cusps, it may be observed that exactly the same tendency is found in the
Peramelidze under similar conditions of increased space in the premolar region.

Milk premolars.—In describing the dentition of the Dasyurinz no reference has been
made to the modifications of the deciduous teeth for the reason that their characters do
not pertain to the category of secondary adaptations here considered, but to the primary
differentiation of the Marsupials as a group. Some reference to them is, however,
necessary in the case of Myrmecobius, on account of their possible connection with the
origin of the increased molar formula.

mu. pp. mp. ap.

Doubtful case of dental replacement in Myrmecobius fasciatus.

a. Internal view entire left ramus of jaw of young specimen (nat. length of jaw=23 mm.); b. External view of part
of right ramus of same specimen; c. Right ramus of older specimen; d. External view of right posterior
premolar and anterior molars of adult animal. Abbreviations: ap., mp., pp., anterior, median, and posterior
premolars ; d., doubtfully deciduous premolar ; m. 1, first true molar. All figures drawn to same amplification.

The milk-premolars have been recognized by Woodward (1896) in the young, but their
subsequent fate has not been elucidated. According to the view of Thomas (1887) and
Leche (1891) they are replaced by the posterior premolars in the ordinary way, while in
Winge’s (1882) opinion they remain in place, the posterior premolars being unable to
dislodge them on account of the elongation of the jaw.

The accompanying text-fig. 3 represents the anterior molariform teeth of the lower
jaw in three specimens, all of which have been drawn to the same amplification. Of
these a and b represent the teeth of a young specimen, in which the posterior premolars
are not yet formed. In a the first member of the molariform series is comparatively

OF THE AUSTRALIAN MARSUPIALIA. 107

minute, and in b the corresponding tooth of the opposite jaw is only slightly larger.
In d, which is the specimen described and figured by Thomas (op. cit., pl. 27. fig. 5),
the posterior premolar is almost fully formed, but is still beneath the bone. The
first members of the molariform series are considerably larger than in the younger
specimen, approximating more nearly in size to the second members of the latter. A
somewhat similar size relation is seen in c, in which the posterior premolars are fully in
place. The minute size of the first molariform teeth in specimen a appears to point out
these teeth as milk-premolars, which are afterwards replaced as in normal forms. The
dental variability is, however, so great in Myrmecobius that the evidence of a single
specimen must be regarded as suggestive rather than conclusive.

TAHYLACININA.

In a former paper (1901 a) the molar patterns of the Tasmanian Wolf (Zhylacinus
cynocephalus) were described as representing the final stage in the carnivorous evolution
of the Dasyuridze. To this conclusion, which turns out to be erroneous, the writer was
led partly by a comparison of the molars of the animal with those of Dasyurus, the
only other member of the family at that time available, and partly by the view expressed
by Thomas (1888), that the relations of Sarcophilus are with Thylacinus rather than with
Dasyurus. In the present paper Thylacinus is assigned to a separate division, and even
suggested as a foreign or unrelated element in the Australian family.

Reference has already been made to the fact that the various genera of the Dasyurinz
present successive phases of an extremely homogeneous dental evolution, the chief
features of the transformation being : (a) reduction of the protocone in the upper molars
and of the talonid in the lower; (2) approximation of the external styles of the upper
molars to their respective cusps; and (¢) reduction with final obliteration of the posterior
premolars and of the paraconid in the first lower molars. In all of these characters
Thylacinus stands apart from the typical Dasyuride, so that while it presents a degree
of carnivorous dental specialization only slightly inferior to that presented by Sarco-
philus, its evolution must have proceeded independently of that of the Dasyurine, at
any rate in so far as the carnivorous stages of the latter are concerned.

The resemblances between Thylacinus and the Sparassodonta of the South-American
Miocene have been noted by different writers and regarded as indicating affinity between
‘the latter group and the Dasyuridee. Lydekker (1899) has suggested the Sparassodonta
as the ancestral forms of the Australian family. The characters in which Thylacinus
resembles the Sparassodonta, however, prove to be exactly those in which it differs from
the advanced Dasyurinze, whose evolution can be directly traced to minute insectivorous
forms such as Sminthopsis and Phascogale. If the resemblances between Zhylacinus
and the Sparassodonta represent aflinity rather than parallel development we can
assume no closer relation between both of them and the Dasyaridz than is implied by
a possibly common origin of the South-American and Australian faunas.

It must not be supposed that Thylacinus presents any dental characters which would
prevent it from being theoretically derived from one of the smaller dasyurine forms.

SECOND SERIES.— ZOOLOGY, VOL. IX. 16

108 DR. B. A. BENSLEY ON THE EVOLUTION

It is only the fact that all of the remaining members of the Dasyuride as at present
defined, with the exception of IZyrmecobius, present a uniform dental evolution which is
thus the predominant and characteristic one for the Australian radiation, which accen-
tuates the resemblances of such an isolated form as Zhylacinus to the Sparassodonta,
whose type of evolution is just as characteristically South American, and lends probability
to the view that its origin is to be sought in the latter group.

Length of Tooth-rows.—On comparing the tooth-rows of Thylacinus with those of
Sarcophilus or the more advanced species of Dasyurus, we notice a conspicuous difference
in their relative lengths. In the latter, notwithstanding the lateral compression of the
molars, there has been a shortening of the whole dental series, while in the former an
elongated condition has been retained and possibly increased. It is interesting to note
that similar differences characterize the dental evolution of the placental Creodonta and
Carnivora, although in the latter Orders they are closely associated with the elaboration
of posterior or more anterior cheek-teeth as sectorials. The elongated condition in
Thylacinus is repeated in the South-American group.

Molar Patterns.—The characters of the upper molars are represented in PI. 5, figs. 6
& 7, of Thylacinus cynocephalus (m. 3) and T. speleus (m. 2). The protocone is well
developed, and is supported on a separate root. In this character Thylacinus differs
from the carnivorous members of the Dasyurinee, and points back to the insectivorous
forms. The crown-surface of the protocone in the adult presents a curious spur-like
appearance. Its posterior border tends to be slightly trenchant. The metacone is much
enlarged and trenchant. As in Sarcophilus, its tip is lanceolate from a filling out of the
concavity originally present on its outer side. The metacone-spur is well developed and
trenchant, but its distal extremity has not been rotated inwards to the same extent as
in Sarcophilus. The paracone is small and comparatively undifferentiated. Unlike
those of the Dasyurinze, the external styles are vestigial or wholly absent. Style Pad
is apparently always present in the first, second, and fourth molars. Style ¢ is not
represented, but a small element probably equivalent to style c, (cf. Pl. 5. fig. 1 ,
Peratherium) is apparently always present in the first molar, variable in the second, and
scarcely distinguishable in the third. The vestigial nature of the external styles gives
the teeth a very different appearance from that seen in Sarcophilus, where their great
development and approximation to their respective cusps result in the production of a
double cutting-edge. The fourth molar is small, and of much the same character as
that of the Dasyurine. The cusps represented are the protocone, paracone, and style ad,
the metacone being barely indicated by a small posterior protuberance.

The patterns of the lower molars are represented in Pl. 6, figs. 10 & 11, of the third
and first teeth of 7. cynocephalus. As in Sarcophilus, the paraconid and protoconid are
modified to form trenchant blades. The metaconid is wholly absent, as in Prothylacinus
and Amphiproviverra. 'The anterior portion of the base of the trigonid in the first and
second molars bears a small protuberance representing an antero-external shelf. Except
in the fourth tooth the talonid is well developed, and its crown-surface is flat and shelf-

OF THE AUSTRALIAN MARSUPIALIA. 109

like. In this character Thylacinus differs markedly from the Dasyurinze, more especially
the carnivorous forms, but resembles Prothylacinus and Amphiproviverra. 'The talonid
shows indications of the three original cusps. Of the latter the hypoconid is best
developed and tends to be slightly trenchant. In the fourth molar the talonid is
reduced to a small spur, as in Amphiproviverra. An interesting character is seen in the
first molar, where the paraconid is comparatively well developed, as again in the last-
named form. As already pointed out, this structure is absent in the carnivorous
Dasyurine.

Antemolar Teeth—The incisors are very similar to those of Sarcophilus. The median
upper teeth show signs of having been formerly procumbent, as in the insectivorous
Dasyurine, in being more rounded in section, or, in other words, less completely
spatulate than the lateral teeth, and also in being separated at their bases and approxi-
mated at their tips. The incisor-rows, unlike the premolar- and molar-rows, show the
same shortening exhibited by Sarcophilus. In this case the lower canines are almost
apical in position.

The canines show no essential difference from those of Sarcophilus, except that they
are rather more slender and more evenly curved.

The chief feature of the premolars is that they increase in size from before backwards,
the posterior tooth, as in the Prothylacinide, showing no indications of the reduction
which is so marked a feature of the Dasyurine.

PERAMELID.

Reference has already been made to the fact that the dental characters found in the
smaller insectivorous members of the Dasyuridze are prototypal not only to those of
the larger carnivorous forms of that family but also to a considerable extent to those
of the omnivorous Peramelidee and Phalangeridze. The close sequence of modifications
which is observable in the dental evolution of the Dasyurinze, and which is of so perfect
a kind that it is impossible to distinguish where the insectivorous evolution ceases and
the carnivorous evolution begins, shows at once that one is a direct natural continuation
of the other. Nevertheless, on the omnivorous side we find, in certain of the Peramelidze
and Phalangeridze, modifications which are almost as inseparable from those of the
insectivorous Dasyuridze as are those of their carnivorous relatives, but which at the
same time usher in advanced mcdifications of a widely divergent kind. While, therefore,
it may be shown that the differentiation of the Phalangeride in some respects represents
an insectivorous specialization, there is abundant evidence that as regards dentition the
primary division of the Australian radiation has been the differentiation of a carnivorous
and an omnivorous line from insectivorous prototypes.

It is interesting to note that the Peramelidie, or including with them the Phalangeride,
occupy much the same position as intermediate omnivorous types with reference to the
Dasyuridee on the one hand, and the members of the herbivorous section on the other,

16*

110 DR. B. A. BENSLEY ON THE EVOUUTION

that, among the Placentals, the Condylarthra do with reference to the Creodonta and
the Ungulata respectively. It is also an interesting fact that among the existing
Insectivora, a group showing a fairly wide range of dental differentiation, we find in the
Soricidse molar patterns of an incipient omnivorous type which parallel almost exactly
those of the Peramelidze, just as in the Talpidee and Chrysochloridze we find insectivorous
modifications paralleling those of the Dasyuridee and Notoryctide respectively.

Unlike that of the Dasyurine, the dental evolution of the present family does not
represent a uniform progression in one direction. In the members of the dominating
eenus Perameles the molars will be seen to present a progressive omnivorous develop-
ment, associated with an increasing hypsodontism, which is quite as gradual as the
carnivorous development of the Dasyurinze; while the antemolar teeth, on the other
hand, will be seen to present a varied assortment of local and more or less retrogressive
characters. The cause of this condition may be found in the peculiar manner in which
the family has been derived. The ancestors of the Peramelidee were arboreal animals,
and were probably either purely insectivorous or at most only slightly omnivorous. On
becoming terrestrial they have gradually become more completely omnivorous, as shown
by the successive modifications of the molars. But they have also become fossorial to a
varying extent, and this development appears to represent the prime cause of the
antemolar variation. The hypsodontism of the molar crowns represents a compensatory
development, the object of which is to counteract the effects of the mechanical wearing
down of the teeth caused by the presence of particles of earth in the food. It owes its
uniformity to its association with the omnivorous development of the molars. The local
and retrogressive characters of the antemolar teeth may be due to a slight extent to the
reduction of their function in the omnivorous evolution, but are more closely connected
with the displacement of these teeth from their original relations by a varying elonga-
tion of the muzzle. ‘The latter is here, as in many other cases, the result of fossorial or
semifossorial habit.

The family may be divided on a basis of molar characters as follows :—

A. Upper molars normal ; triangular, becoming quadrate by the development of a
hypocone. External styles not conspicuously enlarged . . . . . . . Perameline.
Genera: Perameles, Cheropus.
B. Upper molars aberrant ; quadrate through displacement inwards of the meta-
cone ; external styles greatly enlarged and formimg with the protocone and
metacone the main cuspsof the crown . . . +... . =... +. + » Thylacomyine.
Genus: Thylacomys.

Some doubt may be expressed as to the advisability of making a major division in the
Peramelidee on account of the fact that the family is a small one and the three genera
closely related. The type of molar modification found in Thylacomys is, however,
unique in the Marsupials, and rare in the Mammals generally, so that this form seems
fairly entitled to distinction.

OF THE AUSTRALIAN MARSUPIALIA. ant

PERAMELIN2.

Of sixteen species composing the whole family of the Peramelide twelve belong
to the genus Perameles of the present division. The latter, therefore, presents nearly
all of the dental modifications of evolutionary interest found in the family. It
also represents the parent form from which the less inclusive genera Charopus and
Thylacomys have been derived.

Although the modifications of the molars are comparatively uniform in this series,
the specific variability of the antemolar teeth is so great that a general synopsis of the
dental sequence is scarcely permissible.

Sequence of Molar Patterns.—TVhe order of specialization is as follows :—Perameles
Doreyana and P. Raffrayana, P. Bougainvillei, P. nasuta and P. Gunni, P. Cockerelli,
P. moresbyensis, P. macrura and P. obesula. Cheeropus castanotis is in some respects
as specialized as the extreme form P. obesula, but appears to be a derivative of
P. Bougainvillet or a closely allied form.

As indicated in the above plan, the most primitive modifications relate to P. Doreyuna
and P. Raffrayana. This remark, however, applies only to the species studied. The
collection contains no examples of the presumably primitive forms P. Broadbenti and
P. longicaudata (cf. Thomas, 1888, pp. 240-242).

The following description of P. Doreyana will apply almost equally well to P. Raffrayana.
The third upper molar of P. Doreyana (Pl. 5. fig. 8) is of an insectivorous type approxi-
mating closely to that found in the Dasyurine (cf. fig. 2, Sminthopsis leucopus). The
only important points of departure are seen in the arrangement of the external styles,
and in the addition of a new cusp, the hypocone (y.). As regards the arrangement of
the external styles, the Dasyurine have already been described as presenting a specialized
condition, style ¢ being greatly developed, while two anterior styles have been fused
together to form a single element ad. In P. Doreyana, as, in fact, in all of the Perame-
line, style ¢ is only moderately developed *, while the anterior styles @ and } are much
more completely separate. In both of these characters the Peramelinz approximate
much more closely to the Didelphyide (cf. Pl. 5, figs. 1 & 26, Peratherium, Metachirus
opossum) than do the Dasyurinze.

The appearance of the new cusp, the hypocone, is of great significance, since it marks
the first stage in the transformation of a piercing, cutting, and crushing tooth into a
grinding one, or, in other words, the passage from an insectivorous to an omnivorous
condition.

The remaining upper molars of P. Voreyana differ from the third in exactly the same
way as in the Dasyurine. ‘The first and second are more laterally compressed, and the
fourth is reduced through the loss of the metacone.

The lower molars of P. Doreyana (PI. 6. fig. 12) also approximate closely to those of

* In the Perameline, and also in the Didelphyide, as in the Dasyurine, style c is better developed in the first

and second molars than in the third. Allowance should be made for this fact when comparing the figure here
given of the third molar of P. Doreyana with those of the second molars of other forms.

112 DR. B. A. BENSLEY ON THE EVOLUTION

the Dasyurine (cf. figs. 2 & 3, Sminthopsis leucopus, Dasyurus viverrinus). Three
points of departure are, however, indicated. There is, first, a slight reduction of the
paraconid; secondly, a partial elevation of the crown-surface of the talonid; and, lastly,
a slight reduction of the hypoconulid. These remarks apply to all three anterior
molars, but not wholly to the fourth, the latter having the talonid reduced, as in the
Dasyurinee (cf. fig. 16, P. Bougainvillei).

Proceeding from P. Doreyana and P. Raffrayana we find the following changes in the
molar patterns. In the upper teeth (PI. 5. figs.9 & 10, P. Bougainvillet, P. obesula) the
general shape from being triangular becomes quadrate, the spaces originally existing
between the internal apices of the teeth being gradually filled partly by the growth of
the hypocone and partly by a very slight rotation of the body of the metacone. The
metacone-spur becomes partially reduced, and the former disproportion in size between
the paracone and metacone becomes less obvious. In the lower teeth (Pl. 6. fig. 13,
P. obesula) there is a further reduction of the paraconid and an elevation of the talonid,
the result of which is the production of a practically quadritubercular pattern. It is
interesting to compare the tooth of P. obesula here figured with that of Thylacomys
leucura (fig. 14), which shows a more advanced stage of the same evolution. In this
form the paraconid has completely disappeared, and the remaining cusps of the trigonid
have now the same characters as those of the talonid.

Viewing the molar changes from an adaptive standpoint it is seen that the teeth
of P. Doreyana and P. Raffrayana perform the same piercing, crushing, and cutting
functions as those of the smaller Dasyurine. In the insectivorous stage the general
movement of the teeth is vertical. In the omnivorous stage the teeth assume a trans-
verse grinding action, and the cusps become arranged in such a way as to present a
uniform crown-surface, and at the same time to obliterate all of the interspaces. The
crown-surface of the talonid, originally low, is raised to the level of that of the trigonid.
The protocone and hypocone are similarly raised. The change in the latter cusps is, how-
ever, not so essential as that in the talonid, since the movement of the teeth is necessarily
vertical and internal as well as transverse, and is accordingly slower. It is, in fact,
only completed in the more advanced Phalangeride (cf. Pl. 5. figs. 13 & 16, Diste@churus
pennatus, Trichosurus culpecula). The triangular spaces between the internal apices of
the upper molars become obliterated by the growth of the hypocone, and the somewhat
similar spaces between the external apices of the trigonids and talonids in adjacent
lower teeth become obliterated by the reduction of the paraconid and hypoconulid.

Hypsodontism of the Molar Crowns.—Yhe molars of the various members of the Pera-
meline furnish an interesting example of the perfect correlation of two adaptive
changes, which are due to different causes, in the association of a gradually increasing
hypsodontism of the molar crowns with the omnivorous elaboration of the patterns.
The hypsodont development is scarcely perceptible in the teeth of the primitive forms
P. Doreyana and P. Raffrayana, but becomes more and more obvious as we pass to the
final forms P. obesula and P. macrura. Unlike the hypsodontism which is found in the
Macropodidee, that of the Peramelide affects the bodies of the teeth rather than the

OF THE AUSTRALIAN MARSUPIALIA. 115

cusps. The appearance presented in an extreme stage is well represented in Pl. 6.
fig. 15 of the lower molar of Cheropus castanotis. The same modification of otherwise
comparatively simple teeth is found in the Notoryctide (PI. 6. fig. 17 0) and in many
of the placental Insectivora.

Incisors.—The most important feature of these teeth is the retention in all forms,
except P. Doreyana and P. Cockerelli, of a fifth upper pair. In this feature, which is
also presented by Zhylacomys, the members of the family are more primitive than any
of the remaining Marsupials of Australia, but exactly resemble the Didelphyide.

With the partial exception of the fifth upper pair, the incisors present a very
characteristic appearance on account of the great broadening of their tips. The median
upper teeth are not differentiated from the lateral ones as they are in the Dasyurine,
the Didelphyide, and the primitive Phalangeridee. This condition is apparently
secondary, as in Myrmecobius and Notoryctes. The anterior four pairs of upper incisors
are usually seen to have their cutting-edges squarely truncated, but this condition is
probably the result of wear. In young specimens of Chwropus castanotis and Perameles
nasuta the upper incisors show the same curious triangular shape as found in the
Didelphyidz and Dasyuridee (éf. text-fig. 2, f & g, p.105). As already noted above, the
unworn teeth of P. obesula show a backward prolongation of the tips somewhat like
that found in Myrmecobius. The fifth upper incisors tend to be more or less caniniform,
like all of the upper lateral teeth of the latter form.

‘The lower incisors are fairly constant in their characters. Their cutting-edges are
rounded and point to the type found in the Dasyuridee and Didelphyidee. The third
tooth shows a prominent posterior accessory cusp. This element probably represents a
secondary development. It is not present in the Didelphyide, which have four lower
incisors, but is frequently present in the Dasyuride, where, as in the Peramelid, the
incisors have been reduced to three.

The relative positions of the fifth upper incisors and their occasional absence furnish
the first instance of the variability of the antemolar teeth already referred to. In the
otherwise primitive form P. Raffrayana the muzzle is moderately elongated and the
fifth incisors are separated from the fourth by short diastemata; in the intermediate
forms P. nasuta and P. Gunni the muzzle is excessively elongated and the diastemata
are correspondingly increased; while in the final forms P. obesula and P. macrura, as
also in Cheropus castanotis, the muzzle is again shorter and the diastemata are scarcely
in evidence. Again, of the two primitive forms P. Doreyana and P. Raffrayana, the
fifth upper incisors are present in one and absent in the other, while they are present in
all of the specialized forms with the exception of P. Cockerelli. These characters may be
taken, in connection with certain others of the canines and premolars, as showing that
the sequence indicated by the modifications of the molars cannot be relied upon as
indicating the true relationships of the various species.

Canines.—These teeth are even more variable in their characters than are the fifth
upper incisors in their position. In P. Raffrayana the upper tooth is short, moderately

114 DR. B. A. BENSLEY ON THE EVOLUTION

stout, and curved. The lower tooth is so short as to be scarcely caniniform, and bears a
small posterior accessory cusp. In P. Doreyana we find both upper and lower teeth
greatly reduced, the former showing usually small anterior and posterior cusps like
those of the premolars, and also a slight grooving of the root. The lower tooth shows
indications of a posterior cusp, and its root is also slightly grooved. A somewhat
similar condition is seen in Charopus castanotis, except that in this form there is an
anterior cusp on the lower tooth. Perameles Bougainvillei also shows a similar con-
dition, except that the upper tooth is not grooved. In P. xasuta both upper and lower
teeth are of a normal caniniform type; they are, in fact, better developed in this form
than in other members of the family, excepting the species of Thylacomys. In P. obesula,
P. macrura, P. moresbyensis, and P. Gunni the upper canines are short and curved, as
in P. Raffrayana. The lower teeth of P. moresbyensis and P. macrura are not so
caniniform as those of P. obesula, They show a posterior cusp as in P. Raffrayana, but
in addition an anterior one. Apart from their stouter build, they thus resemble those
of P. Doreyana, P. Bougainvillei, and Chaeropus. In P. obesula and P. Gunni the
lower teeth are rather short, but are otherwise normal.

Apart from special phylogenetic considerations, the characters of the canines in this
series are of interest as illustrating the effects of retrogressive development. The
reduction in length, the addition of anterior and posterior basal cusps, and the grooving
of the root mark the return of a caniniform tooth to a premolariform one, from which
it originally evolved. ‘The retrogressive characters of the Peramelidee find a direct
parallel in Wyrmecobius. It is interesting to note that retrogressive characters of a
somewhat similar kind are seen in Ca@nolestes (cf. Pl. 5. fig. 38) and in the Phalangeride,
and further that in certain members (Dactylopsila, Phalanger) of the latter family the
opposite development is illustrated, namely the conversion of an anterior premolar into
a caniniform tooth.

Premolars.—Except for a slight tendency towards the reduction of the posterior teeth
in Cheropus castanotis, the premolars are well developed throughout the series. They
resemble in their general characters those of the Dasyurinz and Didelphyidze. Except
in Cheropus they increase in size backwards, and this character places the Peramelidze
in a more primitive condition than the Dasyurinze or even the existing Didelphyide.
In their relative positions the premolars show considerable variation. In cases where
the diastemata separating the teeth are of moderate length, the latter show a pronounced
tendency towards the enlargement of the basal cusps. ‘The separation of the teeth being
due to the elongation of the muzzle, the characters show much the same relations in
the different species as those of the fifth upper incisors.

From their association with the molars, the posterior premolars show indications of a
progressive evolution. In P. Doreyana and P. Raffrayana they are laterally compressed
and of the usual primitive trenchant type common to the Dasyuride and Didelphyide ;
in the more specialized forms they become more massive and rounded in section, as in
the more advanced Phalangeridee.

OF THE AUSTRALIAN MARSUPIALIA. 115

THYLACOMYILN 2.

The antemolar teeth of Thylacomys call for no description, since they resemble in their
characters those of the Peramelinze and throw no further light on the evolution of the
family.

The most important features relate to the characters of the upper molar patterns.
As will be seen from a comparison of Pl. 5. fig. 11 with figs. 8, 9, 10, the upper
molars show a quadrate contour very like that seen in the more advanced forms of the
Peramelinze, but that while in the latter the quadrate shape is due to the development
of a hypocone, in Zhylacomys it is due to the displacement inwards of the metacone,
or, what amounts to the same thing, a shortening of the protocone. The paracone
retains its original characters and position. Of the external styles, only % and ¢ are
represented. ‘They are greatly enlarged and form with the protocone and metacone the
main cusps of a functionally quadrituberculate crown. ‘The second tooth here figured
will be seen to show a small hypocone wedged in between the protocone and the meta-
cone. This element is absent in all of the other teeth of this specimen, including the
corresponding one of the opposite side. As to the extent of its occurrence in Thylacomys,
no further data have been available, except that it has not been figured or described by
Spencer (1896) in the new species 7. minor. The specimen of 7. leucura here described
is unfortunately the only example of this genus in the collection which shows the
unworn molar patterns.

The first molar of 7. leuwcwra resembles the second and third here figured. The fourth
is not visible in this specimen, but in adults of 7. /agotis it appears to be of the same
reduced type as in the Peramelinze and Dasyuridee.

The upper molar patterns of Thylacomys have apparently been derived from such a
primitive type as is presented by P. Doreyana or P. Bougainvillec, the rudimentary
hypocone having been in all probability obliterated by encroachment of the metacone.

The lower molars of J. leucura (Pl. 6, fig. 14) are not modified away from the
perameline type, but present a much more advanced stage of development than is
found in any of the Peramelinz. ‘The anterior lobe of each tooth no longer bears any
resemblance to a trigonid, the paraconid having completely disappeared. The broad
ledge occupying the anterior border of the tooth represents the antero-external shelf,
a structure which is better developed in all of the Peramelidee than in the Dasyurine.
In the talonid the hypoconulid is quite vestigial, while the entoconid and hypoconid are
greatly enlarged and raised to the level of the anterior cusps.

The lower molars of 7’. lagotis resemble those of 7’. lewcura so far as can be judged
from worn specimens. In both species the hypsodont development is present to a
marked degree.

The Cusp-homologies of Quadrate and Quadrituberculate Molars.

The occurrence of two distinct lines of molar development in the Peraimelidee raises
an interesting question as to the homologies of the cusps in more specialized forms.
SECOND SERIES.—ZOOLOGY, VOL. IX. gy

116 DR. B. A. BENSLEY ON THE EVOLUTION

It will be seen that both in Perameles obesula and in Thylacomys leucura the upper
molars are functionally quadrituberculate, the main cusps being in the former the
protocone, hypocone, paracone, and metacone, but in the latter the protocone, metacone,
and styles and c. The upper molars of the Phalangeride present in most cases an
actually quadrituberculate pattern, which is ancestral to the advanced patterns of the
Macropodidee, Phascolomyide, and Diprotodontide. The question therefore arises as
to whether the four cusps of the Phalangeridze are homologous with those of Perameles
or those of Zhylacomys.

In the more primitive forms of the Phalangeridee (cf. Pl. 5. figs. 14 & 15, Dromicia
concinna, Petaurus sciureus) the upper molars show a vestigial external cingulum, which
appears to be equivalent to that bearing the external styles in polyprotodont forms, and
in one section of the family (Phascolarctinze) vestigial styles are actually present. This
indicates that, although resembling them in many respects, the two outermost of the
main cusps in the Phalangeridze are not equivalent to the modified external styles in

mu. me.

(Ge
: Normal and aberrant production of quadrate upper molars in Marsupials and Placentals.
Placentals:— A. Protogonodon; B. Euprotogonia; C. Phenacodus (Condylarthra); D. Coryphodon (Amblypoda).
Marsupials :—H. Perameles Doreyana; F. P. Bougainville’; G. P, obesula; H. Thylacomys. Abbreviations :
pr., protocone; pi., paracone ; me., metacone ; hy., hypocone ; ps., parastyle ; st.b., style b; st.c., style c.

Thylacomys. Furthermore, in the primitive Phalangeridee (¢f. Pl. 5. fig. 18, Diéiste-
churus pennatus) the postero-internal cusp is relatively small, but increases in importance
in the more advanced stages (fig. 16, Zrichosurus vulpecula) in precisely the same way
that the hypocone does in the Perameline, and is therefore more probably homologous
with that cusp than with the displaced metacone of Thylacomys. Finally, in the
Placentalia the formation of a quadrituberculate tooth by the addition of a hypocone
to an originally triangular crown is the normal procedure, since it is indicated in various
members of the Condylarthra, Primates, Insectivora (Soricidee, Macroscelidee, Erina-
ceide). The formation of a quadrate tooth by displacement of the metacone is rare,
being only found, so far as the writer is aware, in the Amblypoda. The significance
of this will be more apparent from the accompanying diagrams (text-fig. 4), in which

OF THE AUSTRALIAN MARSUPIALIA. ny,

the typical method of the formation of a quadrate tooth is illustrated by the teeth of
the Condylarthra and Perameline, on the one hand, and the atypical method by those
of the Amblypoda and Thylacomys. All of these facts indicate that the main cusps
represented in the typically quadrituberculate Phalangeridze and their herbivorous
derivatives are directly homologous with those of the Perameline.

The lower teeth of the advanced quadrituberculate forms call for no consideration,
since in the Thylacomyine, as also in the Amblypoda, they show no departure from the
condition in other forms.

NOTORYCTIDA.

During the twelve years which have elapsed since the original description of Notoryctes
by Stirling (1891), the dental characters of the animal have been commented upon by
several writers, including Ogilby (1891), Cope (1892), Gadow (1892), Winge (1895), Forsyth
Major (1897), Spencer (1896), and Tomes (1899). The following remarks are therefore
largely directed towards an arrangement and discussion of opinions already expressed.

It may be observed at the outset that no adequate conclusions concerning the affinities
of Notoryctes can be formed froma study of the dentition alone. While the median
lower incisors present an undifferentiated condition, which shows the animal to be related
to the members of the polyprotodont section, namely the Dasyuridee and Peramelidie, the
dentition is otherwise so completely modified away from the usual polyprotodont type
that it does not present any of those special characters which distinguish the latter
families from one another. As will be pointed out in detail in a subsequent section, in
the absence of definite characters of dental affinity in Notoryctes the primary characters
which separate the Dasyuridz from the Peramelidee do not relate to dentition, and in
addition do not represent merely a family distinction, but, on the other hand, serve to
separate the Dasyuridze from all of the remaining Australian forms. These characters
may be mentioned as relating to the non-syndactylous and syndactylous modifications
of the second and third digits of the pes, respectively characteristic of the two series.
Notoryctes, in its foot-structure, shows affinities with the syndactylous section, indicating
that of the two polyprotodont families its relations are rather with the Peramelidie than
with the Dasyuridz. These relations should be borne in mind in attempting to derive
the characters of its dentition.

Dental Formula.—A conspicuous feature of the dentition of Notoryctes, and one which
is in all probability responsible for most of the remaining characters presented, relates to
the foreshortened condition of the tooth-rows. One of the effects of this development
is seen in the variability of the dental formula, implying a reduction, which is still in
progress, of certain of the less functional teeth. The formula has been variously
determined by Stirling, Ogilby, Gadow, and Spencer, the most trustworthy figures being
those given by the last-named writer. Spencer states that the full formula, “as far as at
present known,” is: @. 3 C. s p: 5 m. : Concerning this he remarks as follows (p. 47) :-—
“Tt was only after examining twenty-nine specimens that one was found in which the

Lz*

118 DR. B. A. BENSLEY ON THE EVOLUTION

full number of incisors was present in the upper jaw, whilst in only a comparatively
few specimens an additional premolar was present in the lower jaw, always small and
pushed out to the side of the jaw in front of the first molar.” The formula given by
Spencer differs only from that determined by Gadow from Stirling’s specimens in the
addition of one upper incisor and one lower premolar.

Gadow has expressed the opinion that the affinities of Motoryctes are with the
Dasyuridee, and the subsequent discovery of a fuller incisor formula of appears at
first sight to support this view. It must be borne in mind, however, that the upper
incisor formula of 4 only represents one limit, so far as known at present, of the varia-
bility which is normal for this genus; it cannot be regarded as indicating the presumably
stable formula of the ancestral form from which Notoryctes has been derived. Further-
more, no greater importance can be attached to the occurrence of a fourth upper incisor
in one of twenty-nine specimens than, for example, to the occasional presence of a
fourth lower incisor in Myrmecobius or a fifth in Didelphys, even though the latter be
regarded as reversional, a proposition which, as Bateson (1894) has shown, is rather
doubtful.

With reference to the premolar formula, Gadow considered the number : determined
by him to represent the anterior and median premolars of other Marsupials, the posterior
teeth having disappeared as in the Dasyuride. This, again, cannot be regarded as
indicating a special affinity with the latter family, since the same process of reduction
of the posterior premolars is found in other families. In the Phalangeride the genus
Acrobates shows a reduced condition of the posterior premolars, and in Déisteechurus
(cf. Pl. 5. fig. 89) the upper teeth are much reduced and the lower are absent, as in the
more advanced members of the Dasyuridw. The same tendency towards reduction is
seen in the Didelphyide, and in two species, Thylacomys lagotis and Cheropus castanotis,
of the Peramelide.

The dental reduction of Voforyctes appears to be confined to the antemolar teeth, the
molars always presenting the formula 4 characteristic of all of the remaining Marsupials
with the exception of Myrmecobius, Acrobates, Distechurus, and two species of
Dromicia. ‘the cause of the stability of the molar formula is to be sought in the
greater functional importance of these teeth. Like the antemolar teeth they have been
affected by the shortening of the jaws, but only to the extent of an antero-posterior
compression.

Derivation of the Molar Patterns —The characters of the upper teeth are represented
in Pl. 5, fig. 12, and those of the lower in Pl. 6, fig. 17.

The appearance of the molar crowns has been figured by Stirling (1891) in connection
with his original description. Tomes, however, remarks that, “in Dr. Stirling’s figures
of the grinding-surfaces of the molar teeth, it is shown that the middles are worn into
concavities, and that the retention of the cuspidate form is not due to the retention of
sharp enamelled cusps, but that it is due to the upstanding of the edges.” While this is
probably true of the lower molars figured by Stirling, the prominences shown in his
illustrations of both upper and lower teeth will be found to correspond with those here

OF THE AUSTRALIAN MARSUPIALIA. 119

given, which have been taken from a specimen in which no signs of wearing are
apparent.

Both the upper and lower molars are extremely simple in structure. Each of the
upper teeth is composed of a high triangular pillar, to which is attached two small
external cusps and a large internal one. The apex of the triangular pillar forms a
large cusp, which in the pointed character and the triangular section of its tip resembles
the paracone or metacone of a normal insectivorous tooth. Of the external cusps, one
is attached anteriorly to the tip of the triangular pillar, while. the other occupies a
similar position posteriorly. The former shows indications of a composite structure.
The large internal cusp is crescentic in shape when viewed from the crown, and its tip
is placed at a much lower level than that of the median pillar. Each of the lower teeth
shows a triangular pillar similar to that of the upper teeth, but with the section-apex
reversed. The crown-surface bears three cusps, showing the same arrangement and
proportions as those of the trigonid in the Dasyuridz and primitive Peramelidee. The
talonid, which in the latter forms is well developed, is, however, only represented by a
minute tubercle attached to the postero-internal angle of the trigonid in the first and
second molars.

It will be observed that the molars of No/oryctes present a less complicated condition
than is found in any of the other polyprotodont families; and this raises the interesting
question, involving the origin not only of Notoryctes but that of all recent Marsupials,
as to whether the modification represents a more primitive phase or is the result of
special development proceeding from the type represented by the teeth of the Dasyuridie
and Peramelide.

Cope has called attention to the fact that there is an intimate resemblance between
the molars of Notoryctes and those of Chrysochloris, and has further stated that “the
tritubercular molars . .. . show Notoryctes to be a primitive type”; while Winge, on the
other hand, has expressed the opinion that the teeth of Notoryctes have lost some of
their cusps, and has further compared them with those of the Centetide. A similar
comparison with the Centetidz has been made by Forsyth Major.

Whether the teeth of Notcryctes are primitive or specialized, one thing is apparent,
namely that there are no intermediate stages connecting them with those of any other
marsupial forms. ‘heir relations can therefore only be surmised from the parallel case
of the Centetide. The evolution of the teeth in the latter family has been carefully
studied by Forsyth Major (1897), and the following statement of the case is based to a
considerable extent upon this writer’s views:—In certain of the Insectivora (Talpide )
(of. text-fig. 5, p. 120) the upper molars show an arrangement of the cusps whicb is very
similar to that seen in the Didelphyidee, there being in each case a triangle of three
main cusps and an outer row of styles. In the Potamogalide the two outer cusps of
the triangle, the paracone and metacone, are partially fused together, while the
protocone and the external styles tend to retain their original characters. In all of
the Centetide, and also in the Chrysochloridze and Solenodontide, the paracone and
metacone are completely fused together to form a median pillar, in which it is not

120 DR. B. A. BENSLEY ON THE EVOLUTION

possible to recognize the constituent elements. The protocone is either reduced
(Oryzoryctes) or in the final stages (Hriculus, Echinops, Centetes) wholly vestigial or

absent.
Referring to the consolidation of the paracone and metacone in the upper molars,
Forsyth Major adds :—“ It is only in Notoryctes, which in other respects also is highly

specialized and forms a parallel to the Centetidze and still more to the Chrysochloride,
that this fusion likewise occurs in all molars; the protocone in Notoryctes is still
developed to a considerably greater extent than in the Centetidee.”

With reference to the lower molars of the Centetide, this writer points out that the
chief feature is the reduced condition of the talonid, a character which he believes to
be secondary, from the fact that this structure is well developed both in the lowest
‘Tertiary and Cretaceous Mammalia.

C<
. pad me.
D. E.

Homoplastic consolidation of central cusps of upper molars in Marsupials and Placentals.
A. Talpa; B. Potamogale ; C. Oryzoryctes ; D. Peramys; E. Notoryctes. (Abbreviations as in text-fig. 4, p. 116.)

According to this view, the molars of Notoryctes bear the same relation to those of
the Dasyuridie, primitive Peramelide, and Didelphyide as those of the Centetidee and
especially Oryzoryctes bear to those of the Talpidee. The triangular pillar of the upper
molars represents the conjoined paracone and metacone, the large internal cusp a
protocone. The external styles, which are virtually three in number, the anterior one
being composite, represent a, b, and ¢ of the normal polyprotodonts. The triangular pillar
of the lower teeth represents an elongated trigonid, the talonid being vestigial.

It is apparent that Cope’s designation of the molars of Wotoryctes as trituberculate, —

although quite applicable to the lower teeth, is misleading as applied to the upper, if by
the term “ trituberculate” we mean to designate such teeth as those of the Didelphyidee,
'Talpidxe, and primitive Creodonta, which present a triangle of three main cusps with or

without intermediate conules or external styles. It is further apparent that his view —

that the minute proportions of the talonid in the lower molars of Notoryctes, Chryso-
chloris, and the Centetide represent a primitive condition, which is prophetic of one
in which this structure is well developed, is also erroneous. While the fact pointed out

OF THE AUSTRALIAN MARSUPIALIA. 121

by Forsyth Major that in the early Tertiary and the Cretaceous Mammalia the talonid
is found in a well-developed condition, does not in itself invalidate the view that the
reduced condition of this structure in the Centetidze represents a primitive character,
the reverse is indicated by the fact that the modifications of the talonid in the lower
molars of the Insectivora follow those of the paracone and metacone and also the
protocone in the upper. In the Talpidee, where the protocone is well developed, and
the paracone and metacone entirely separate from one another, the talonid is found in a
well-differentiated condition. In the Potamogalide, where the paracone and metacone
are partially fused and the protocone slightly reduced, the talonid is also reduced ; while,
finally, in the Centetidee, where the fusion of the paracone and metacone is complete and
the protocone vestigial, the talonid is also vestigial. Such an assumption, therefore, as
that the changes of the lower molars in the above-mentioned Insectivora represent
successive stages in the development rather than the reduction of the talonid, implies
the further assumption of a similar sequence in the upper molars, namely the gradual
development of a protocone and the formation of a paracone and metacone by the
splitting apart of an originally single cusp, a proposition which could not be seriously
entertained. This conclusion, however, in nowise affects the general thesis that the
tuberculo-sectorial type of lower molar, as exemplified by the Didelphyidz, Dasyuride,
Talpidze, and the Creodonta, has arisen by the addition of a talonid to a previous existing
trigonid, nor that the trituberculate condition of the upper molars of the same forms
represents a primitive type from which many other modifications have proceeded. The
conditions in Notoryctes, the Chrysochloride, Centetidee, and Potamogalidee simply
represent another line, similar to those leading to the carnivorous Dasyuridze and the
placental Carnivora, or to the diprotodont Marsupials, the Primates and Ungulata, along
which the transformation of trituberculate, tuberculo-sectorial teeth may proceed.

The dominating principle in the molar evolution of Notoryctes appears to be the
economy of space in the molar region, involving an antero-posterior compression of the
teeth, but the development has probably been assisted by the hypsodont modifications
of the molar crowns. In the upper molars of the Dasyuridee and Didelphyide, the
paracone and metacone are not placed on exactly the same level as the protocone, but
arise together from a short median platform, to the base of which the latter cusp is
attached. The hypsodont development of such a crown would tend to produce just such
a condition as is seen in Noloryctes, providing it were accompanied by antero-posterior
compression. In the Peramelidee, where there is a tendency towards an elongation
rather than compression of the tooth-rows, the hypsodont development is found un-
accompanied by a fusion of the paracone and metacone, although even in these forms
the latter cusps, instead of being individually elongated, are borne on an elongated
median pillar very similar to that seen in Voltoryctes. The reduction of the longitudinal
diameter of the lower molars by obliteration of the talonid calls for no explanation.

As already pointed out, the dentition of Noforyctes gives no indications of special
affinity either with that of the Dasyuridie or that of the Peramelide. ‘The only molar
characters on which reliance may be placed in the distinction of the latter families from
one another relate to the presence of a hypocone in the Peramelide and its absence in

1223 DR. B. A. BENSLEY ON THE EVOLUTION

the Dasyuride. The hypocone, however, represents at most an outgrowth from the
protocone, the crown-surfaces of the two cusps being practically continuous. Even
supposing the upper molars of Votoryctes to have originally possessed a hypocone, we
could scarcely expect to find this element differentiated in teeth which have undergone a
consolidation of two cusps, namely the paracone and metacone, which are originally much
more distinct from one another than are the protocone and hypocone in the Peramelide.

As shown in Pl. 6, fig. 16, there is considerable resemblance between the fourth
lower molar of Perameles Bougainvillei and the normal molars of Wotoryctes. This
resemblance is largely due to the hypsodont condition of both, and since the original
condition in the Peramelid is a brachyodont one, it cannot be strictly regarded as
indicative of affinity. A certain resemblance also exists between the fourth upper
molars of the Peramelidze with the normal molars of Notoryctes, but this resemblance
also extends to the Dasyuridee and Didelphyide.

Antemolar Teeth.—Apart from their variability in numbers, the chief feature of these
teeth is their lack of differentiation, the incisors, canines, and posterior premolars being
all of the same short peg-like character. The median upper incisors are separated at
their bases and approximated at their tips, but, like those of the Peramelidee, show
otherwise no indications of having been formerly differentiated as in the Dasyuride,
Didelphyidee, and primitive Phalangeridze. The last (morphologically median) premolar
is better developed than any of the remaining antemolar teeth, probably on account of its
association with the better developed molars.

PHALANGERID&.

The present family has already been referred to as forming with the Peramelide an
intermediate group, with dental characters connecting those of the insectivorous
Dasyurine with those of the herbivorous section. The association of the two families
will be seen to depend fundamentally on the recognition of the Peramelidze as Omnivora
rather than as Polyprotodontia. The polyprotodont characters may now be considered
together with certain progressive omnivorous ones of a lower rank than those seen in the
Phalangeridze, as defining the position of the latter family with reference to the
Peramelidze as in many respects a derived one. ‘These characters are as follows :—
(a) Retention in the Peramelidz of an undifferentiated condition of the median lower
incisors, which is replaced in the Phalangeride by a differentiated condition involving
extensive reductions of the antemolar teeth. (2) Retention in the Peramelidee of the
piercing character and pyramidal shape of the molar cusps ; replaced in the Phalangeridze
by a bunoid or selenoid modification of these structures. (c) Retention of a full
complement of external styles; in the Phalangeridz these elements are either wholly
vestigial (Phascolarctine) or absent (Phalangerinze). (d) Retention of the constructive
stages in the formation of either functionally or actually quadrituberculate molars —
from others of insectivorous type, only indicated in the Phalangeride, and to a very
limited extent, in the development of the hypocone in the upper molars.

ete
Ai ea

OF THE AUSTRALIAN MARSUPIALIA. 123

Considered alone, the above characters might be regarded as indicating that the dental
evolution of the Phalangeride represents a continuation of that of the Peramelidee. That
this is not the case, however, will be seen from the following additional characters :—-
(a) The primitive members of the Phalangeride retain a differentiated (insectivorous)
condition of the median upper incisors. This is not indicated in any of the Peramelide,
although present in all other primitive polyprotodonts (Dasyuridze, Didelphyidee). (%) The
same forms show signs of dasyurid affinity in the retention of normal insectivorous
characters in the upper canines and premolars, in which respect they are quite as
primitive, and, in view of their smaller size and the tendency towards retrogression of
these teeth, especially the canines, in the Peramelidze, even more so than the latter family.
(c) The primitive members of both families present an incomplete condition of the
hypocone in the upper molars, while the advanced members in each case show this
element in a perfect condition. (d) Finally, the Phalangeridze show none of those
hypsodont developments of the molars which characterize the other family.

The explanation of these two series of characters becomes more apparent on comparison
of the foot-structure in the two families. The present facts may be taken as indicating
that the common ancestors of the Peramelide and Phalangeridz were in all probability
animals combining the incipient omnivorous molar characters of the primitive members
of the former family, and also their incisor formula, with the normal antemolar characters
of the smaller Dasyurine, and it may be mentioned in anticipation that the two families
have undergone a divergent evolution in the foot-structure—the Peramelidee having
become terrestrial, while the Phalangeride have remained arboreal, so that their
omnivorous dental evolution must have proceeded independently as far as the later
stages are concerned.

The Origin of the Diprotodont Modification.

Closely connected with the question of the dental relations of the Phalangeride and
Peramelidze is that of the origin of the diprotodont modification, which appears for the
first time in the former family, although characteristic to a marked degree of the whole
herbivorous section.

It is a familiar fact that of the recent Marsupials the Didelphyide, Dasyuridie, and
Peramelidée on the one hand, and the Phalangeridee, Macropodidie, Phascclomyidi, and
Diprotodontidee on the other, form two broad divisions, differing markedly in the
characters of the antemolar teeth, which distinction led Owen to propose a primary
division of the whole group into Polyprotodontia and Diprotodontia. ‘The relative
merits of this division will be discussed in a subsequent section, the object of the
following remarks heing to show that the diprotodont modification, although charae-
teristic of the herbivorous section of the Marsupials, is the result of an insectivorous
adaptation which must have been developed in the minute ancestors of the Phalangeridw
during the incipient stages of the omnivorous evolution, but after the separation of the
peramelid stem.

The fundamental characters of the diprotodonts may be enumerated as_ follows :—
(a) Reduction of the upper incisor formula to 3 or even 1. (%) Reduction of the lower

SECOND SERIES.— ZOOLOGY, VOL. IX. 15

124 DR. B. A. BENSLEY ON THE EVOLUTION

functional inciscr formula to 1, the median teeth becoming greatly enlarged, while
the lateral ones, together with the canines and anteriorly placed premolars, become
reduced or disappear.

The fact has already been noticed that the insectivorous members of the Dasyurine,
like the Didelphyidee and many of the placental Insectivora, show a procumbent
development of the median upper incisors, which adds considerably to their functional
value as grasping and piercing organs in the capture of insect prey, but that, with the
partial exception of certain forms such as Phascogale penicillata and its prototype
P. calura, they show no corresponding differentiation of the median lower incisors ;
further that in the placental Soricide the latter teeth are enlarged and elongated in
such a way as to act in concert with the median upper ones after the manner of the tips
of a pair of forceps. ‘The primitive members of the Phalangeridee (Acrobates, Diste-
churus, Dromicia) show exactly the same differentiation of the median upper teeth as is
seen in the Dasyurinve, combined with the same differentiation of the median lower ones
as is found in the Soricidee. This circumstance furnishes a complete clue to the origin
of the diprotodont modification, since it can be shown that all of the changes which take
place in its formation are directly dependent on the differentiation of the median lower
incisors as grasping and piercing organs.

It may be observed that nore of the actually constructive stages in the development
of the diprotodont modification are illustrated in the existing Phalangeridee, the most
primitive condition found in that family being of the kind represented in Pl. 5, fig. 39,
of the dentition of Distechurus pennatus. Outside of the Australian group, however,
we find in the Epanorthide and their allies much more primitive relations, and a com-
parison and figure of the dentition of Cenolestes obscurus, the existing representative of
that family, has accordingly been introduced.

If we suppose such a polyprotodont animal as Phascogule penicillata to undergo a
further enlargement of the median lower incisors, the result would be twofold. In the
first place, the elongation of these teeth would be accommodated by a shortening of the
anterior portion of the lower jaw, and this would disturb the relations of all of the more
anteriorly placed lower teeth with the corresponding upper teeth. On comparison of
tine figure of Canolestes (Pl. 5, fig. 38) *, it will be seen that, while the lower molars, as
well as the median and posterior premolars, occupy their normal positions with reference
to the corresponding upper teeth, all of the remaining lower teeth situate behind the
enlarged median incisors have been displaced backwards, and are now in a vestigial
condition. A continuation of the same development will be seen in Déistechurus
(fig. 39), where, with a further shortening, at least two teeth, probably representing the
canine and anterior premolar, have wholly disappeared. In the second place, there
would be a reduction of the posterior upper incisors, since only those teeth with which
the tips of the median lower incisors come into contact could persist. ‘hus in

* This specimen of Cenolestes, kindly lent to me by Mr. Thomas, differs from the type specimen described by
him (1895 6) in possessing an additional “ intermediate” tooth on each side of the lower jaw. ‘The full antemolar
formula, which is thus 8 in this specimen, is greater than in any of the existing Marsupials excepting the
Didelphyidee.

OF THE AUSTRALIAN MARSUPIALIA. 125

Cenolestes the fourth upper incisor shows a tendency towards reduction, while in
Distechurus, as, in fact, in all of the Phalangeridse, the number of upper incisors is
already reduced to three, the lateral two together serving as a stop for the enlarged
lower teeth.

The presence of the diprotodont modification in herbivorous Marsupials is accordingly
not to be explained on the supposition that it represents an advantageous condition, but
rather that these animals are the descendants of minute forms, which, like Acrobates,
Dislechurus, and Dromicia, had undergone a special insectivorous development of the
median lower incisors, and that their ancestors were able to remodel the original
insectivorous condition to suit the necessities of an herbivorous evolution. In the
Macropodide both the upper and lower incisors present trenchant modifications
adapting them more or less perfectly for a grazing habit; while in the Phascolomyidie
and Diprotodontidee (Pl. 5. figs. 40, 41) the median upper and lower teeth present an
open-rooted condition, and a formation of anterior enamel bands fitting them for a
rodent habit.

Apart from its bearing on the evolution of the Phalangeride, the origin of the
diprotodont modification as an insectivorous modification, primarily concerning the
median upper and lower incisors, and its conversion into a rodent modification in
the Phascolomyidze, is of interest as furnishing a possible clue to the origin of the
placental Rodentia.

With regard to the special dental evolution of the Phalangeridxe, we may distinguish
three lines of development, characterized by the modifications of the molar patterns as
follows :—

A. Molars quadrituberculate, with selenoid cusps. Upper teeth with reduced
external styles ; with or without intermediate conules. . . . . . . . Phascolarctine.
Genera: Pseudochirus, Petauroides, Phascolarctus.
B. Molars quadrituberculate, with bunoid cusps. Upper teeth without external
Beesorintermediate conules . =. =... =... »« «. . . . » Phalangerine.
Genera: Acrobates, Distechurus, Dromicia, Gymnobelideus,
Petaurus, Dactylopsila, Phalanger, Trichosurus.
C. Dentition degenerate; cheek-teeth haplodont, wholly vestigial . . . . . . Tarsipedine.
Genus: Tarsipes.

PHALANGERIN&E.

As compared with the Phascolarctine and Tarsipedine the present group will be seen
to represent the dominating division of the family, the bunodont modifications of the
molars by which it is distinguished being found in eight of a total number of twelve
genera. It also represents the ancestral group from which the herbivorous families
have been derived.

The various genera of the Phalangerinze form a progressive series of much the
Same order as that seen in the case of the Dasyurinze, the dominating principle appearing
to be, as before, the increase in size of the body. The more primitive forms, such as
Acrobates, Distechurus, and Dromicia, in their comparatively minute proportions and,

18*

126 DR. B. A. BENSLEY ON THE EVOLUTION

apart from the diprotodont modification, in their dental characters, make an interesting
approximation to the insectivorous Dasyurine, while the most specialized forms, such
as Phalanger and Trichosurus, in their large size and incipient herbivorous modifi-
cations, present an opposite extreme, bordering on the conditions in the smaller of the
Macropodide.

It may be observed, however, that the sequence of dental modifications in the present
division does not indicate an evolution as completely homogeneous as that of the
Dasyurinze. Each genus will be seen to present one or more specialized modifications
which, while not disturbing the general sequence, prevent it from being wholly
prototypal to a succeeding one. The cause of this condition will be better discussed
after a review of the dentition of the group, but it may be mentioned at this point that
the omnivorous-herbivorous evolution, preceded as it is by the diprotodont modification,
appears to represent a critical stage in the general dental progression. There appears
to have been a double conflict between the continuation of the reduction of the
posterior premolars and their elaboration as sectorials, and between the use of the
incisors as piercing and cutting organs and the transference of their functions to
the posterior premolars, The molars will be seen to show an evolution as gradual
as that of the Dasyurinze and Peramelinze, the only important exception being that in
Acrobates, Distachurus, and two species of Dromicia there is a reduction of the formula
by one tooth above and below.

Sequence of Molar Patterns.—Apart from certain characters of specialization, to which
reference will be made below, the molars of Distechurus pennatus, the form selected for
illustration, may be taken as representing the starting-point for the evolution of the
bunodont section, as far as the stages of this are indicated in recent forms.

The upper molars of Déistachurus decrease in size from before backwards, the first
tooth being more than three times as large as the third. ‘his character and also
that represented by the redueed condition of the molar formula are secondary and
not prototypal. Each of the upper teeth (Pl. 5, fig. 13) is roughly triangular in
shape, and shows only four cusps, representing the protocone, paracone, metacone, and
hypocone (cf. p. 89, text-fig. 1). The pretocone and hypocone are placed below the level
of the remaining cusps, as in the Peramelinze and, in the case of the protocone, the
Dasyurinee and Didelphyidxe. Together they appear to form an internal ledge attached
to the bases of the outer cusps. The hypocone is not fully developed, and this accounts
for the somewhat triangular contour of the crown. Both the paracone and metacone
are well developed, and present a very different appearance from that seen in the
Dasyurine and Peramelidi, there being no disproportion in their relative sizes, and no
tendency towards any excavation of their external faces, so that, apart from a fine
trenchant line extending over their tips, they are wholly bunoid. External styles are
entirely unrepresented. There are also no traces of a metacone-spur.

The lower molars of Distechurus do not show the disproportion in size characteristic
of the upper teeth. The first is only slightly, if at all, larger than the second, and the
latter is slightly larger than the third. The second (PI. 6. fig. 18) and third are oblong

OF THE AUSTRALIAN MARSUPIALIA. 127

in shape and quadrituberculate, only the protoconid, metaconid, hypoconid, and entoconid
being present, as in Zhylacomys. ‘The protoconid and hypoconid are not yet completely
bunoid, but tend to retain the triangular section characteristic of more primitive forms.
As in the latter also the two cusps are separated externally by an angular notch. At
the anterior and posterior margins of each tooth a minute ledge is seen to be connected
with a fine ridge running over the tips of the outer cusps. ‘This structure is not very pro-
nounced in the present form, but becomes very conspicuous in the more advanced species.

The first lower molar of Déstachurus (P1. 6. fig. 20) shows an insectivorous specialization,
the protoconid being elongated and curved after the manner of a canine, while the
metaconid is reduced so as to form a small tubercle attached to the inner side. This
curious development also appears in a reduced state in more advanced forms.

The molars of Acrobates are very like those of Distechurus. The upper teeth are,
however, more definitely quadrate, in which character they are more specialized, but they
also decrease only moderately in size backwards, in which respect they are more primitive.
In the first lower tooth the protoconid has not been elevated to the same extent,
so that the disproportion in size between it and the metaconid is not so marked.

The anterior upper molars of Dromicia resemble those of Acrobates in their pro-
portions and the completeness of the hypocone, but this is more nearly true of D. lepida
and D. nana than of D. concinna, which tends to resemble Distachurus. In the third
upper molar the hypocone is absent, so that the tooth is triangular and trituberculate.
In the first lower molar the anterior cusp is elongated only in D. concinna. The
metaconid is fairly separate in D. lepida, as in Acrobates, but is absent in D. concinna
and D. nana.

Passing from <Acrobates, Distechurus, and Dromicia, it is probable that the next
higher member of the series is Gymnolelideus. This genus is, however, not represented
in the collection. Its volant form, Petawrus, which is said by Thomas to be identical
with it in dentition, presents the following features. The molar formula is unreduced,
and the teeth decrease gradually in size from before backwards. The upper molar
patterns almost repeat those of the preceding species, this being especially true of
P. breviceps and P. sciureus. It is interesting to note in this connection that
P. breviceps makes the closest approach in size to Dromicia, P. sciureus being inter-
mediate in size between P. breviceps and P. australis. ‘The upper molars of the last-
named species begin to show advanced characters. In those of P. sciureus (Pl. 5, fig. 15)
a thin ridge is seen to pass over the protocone and hypocone, and to terminate anteriorly
and posteriorly in narrow ledges. This structure is very faintly indicated in P. breviceps
as in Distechurus, but is more definite in P. australis. The internal sides of the
paracone and metacone show slight indications of transverse ridging. The lower teeth
differ from those of Distachurus in having the protoconid and hypoconid more completely
bunoid, and in having the external band slightly more pronounced. The anterior lobe
of the first molar shows the same modification as in Distechurus, except that the
protoconid is not so conspicuously elongated, and the metaconid is absent. It now
becomes apparent that this tooth has been formerly modified as a piercing-organ and is
undergoing reduction.

128 DR. B. A. BENSLEY ON THE EVOLUTION

Dactylopsila represents the next higher member of the series both in size and dental
characters. It differs from Petaurus in having the protocone and hypocone fully
differentiated in the upper molars and placed on a level with the outer cusps, except in
the case of the first tooth, where these elements are low as in the smaller species. The
external cusps do not, however, show quite so marked a tendency towards transverse
ridging as those of Petaurus. Dactylopsila appears to be very slightly aberrant in this
character and in its insectivorous habit as compared with its size, judging from the
conditions presented by Pefaurus and more advanced forms. In the first molar the
protoconid is low, as in Petaurus. The remaining teeth have their cusps completely
bunoid.

With Petaurus and Dactylopsila the omnivorous development of the molars may
be regarded as complete. In the next member of the series, Phalanger, and also in the
final form, Zrichosurus, herbivorous modifications begin to be apparent. The upper
teeth of the latter genus (Pl. 5, fig. 16) are oblong in shape; the hypocone is fully
developed, and the cusps are all of the same height. The internal band described for
Petaurus is now extremely conspicuous and involves completely the inner cusps. The
internal sides of the outer cusps are sharply ridged, each of the ridges extending from
the tip of an outer cusp towards the base of an inner one. In Phalanger this ridging is
slightly less prominent than in Zrichosurus.

In the lower molars of Trichosurus (PI. 6. fig. 19) the same lophodont developments
are again in evidence, the opposite (internal) cusps being affected. The external band,
like the corresponding internal band of the upper teeth, is very pronounced and involves
completely the outer cusps. The first molar, both in Zrichosurus and Phalanger
(Pl. 6. fig. 21), shows the protoconid enlarged but low, as in Petawrus and Dactylopsila ;
the metaconid is practically absent.

The development of transverse crests in the final members of the Phalangerine
represents the beginning of the herbivorous evolution. As will be seen from a
consideration of the Macropodidz and Diprotodontidie, it is on these structures that the
herbivorous evolution of the molars is mainly developed. It is interesting to observe
the transverse ridges already appearing so far back in the evolution as Petaurus.

Origin of Bunodont Molars.—On following the modifications of the molar patterns in
the Phalangerinze, it will be seen that the conditions obtaining im the final members
Phalanger and Trichosurus represent a considerable advance on those of the initial
forms Acrobates, Distechurus, and Dromicia. Nevertheless, the molar patterns even of
the latter animals do not represent an actually incipient, but a fairly advanced phase
of the bunodont evolution, the first constructive stages having already been obliterated.

The available evidence goes to show that the upper molar patterns of the Phalangerinze
have been derived from a trituberculate type like that seen in the Dasyurinze, the main
facts being as follows :—In the first place, we find the earlier stage represented in the
group characterized by an imperfectly developed condition of the hypocone. This
appears to indicate that the hypocone is now in process of development as a supplement
to an originally trituberculate crown. The same differences in the degree of development

OF THE AUSTRALIAN MARSUPIALIA. 129

of the hypocone are seen in the Peramelidz, whose molars are undoubtedly of trituber-
culate origin (cf. PL. 5, figs. 8,9, & 10). Secondly, while in the majority of the existing
Didelphyide we meet with molar patterns of the same insectivorous type as those seen
in the Dasyurine, in one form, Caluromys, we find indications of a bunodont omnivorous
development, which is now in its very incipient stages, but which if continued would
probably give rise to exactly the same conditions as are seen in Diéistechurus and its
allies. Comparing the upper molars of Caluromys (Pl. 5. fig. 27) with those of a normal
dideiphyid form (tig. 26), we notice the following modifications:—The cusps are lower
and their triangular section not so sharply indicated. The protocone is conspicuously
broadened. The disproportion in size between the paracone and metacone is not so
obvious, the metacone-spur being reduced. The external styles are reduced, and the
ridge bearing them is thin and usually crenulated. In all of these characters the pattern
of Oaluromys is prophetic of that of Distechurus (Pl. 5, fig. 13), and if the protocone
were still further broadened to a stage of differentiation of a hypocone, and tie
triangular section of the outer cusps slightly more reduced, tie result would be much as
in the latter genus.

Confirmatory evidence may be obtained from a consideration of Cenolestes. In this
form (¢f. Thomas, 1895 4, pl. 1. fig. 6) the first and second upper molars show a bunodont
quadrituberculate condition, which is very similar to that seen in Petaurus, while the
third shows a trituberculate condition, which is intermediate between that in Calwromys
and that in Distechurus, the outer cusps being bunoid and the external styles absent, as
in the latter. A somewhat similar relation may, in fact, be seen between the second and
third molars of the three-molared Phalangerinze themselves.

The derivation of the lower molars of the Phalangerinze presents no great difficulties.
The quadrituberculate pattern is the same as that seen in Zhylacomys, the latter being
shown by analogy with the Perameline to be of tuberculo-sectorial origin. A prototypal
condition is again seen in Caluromys, where there is a tendency towards the reduction
of the paraconid and hypoconulid, and a rounding off of the remaining cusps. As
already mentioned, the molars of Déistechurus show a tendency to retain the original
angular character of the protoconid and hypoconid.

Incisors.—The general characters of these teeth have already been referred to in
connection with the origin of the diprotodont modification, so that only the sequence of
their special modifications calls for consideration.

As in the case of the molars, the most primitive conditions are to be found in the three
genera Acrobates, Distechurus, and Dromicia. In all of these the median incisors are
of much the same procumbent piercing type as those of the Dasyurinie, the only note-
worthy difference being seen in a slight tendency towards the lateral compression of the
tips. The upper lateral teeth (cf. Pl. 5. fig. 39) are relatively small, this being especially
true of the anterior one (i. 2). In the unworn condition they are spatulate, as in the
Dasyurinz, but their edges are now slightly turned inwards so that they stop the tips of
the lower teeth instead of letting them pass inside. he latter are recurved and sharply
pointed.

130 DR. B. A. BENSLEY ON THE EVOLUTION

The larger forms Petaurus and Dactylopsita show a more advanced stage of the same
modification. In the former the median upper incisors are decidedly procumbent.
They tend to be triangular in section from a thickening of their antero-internal edges.
In Dactylopsila they are still more completely triangular in section, and the procumbent
character is very conspicuous (¢f. Thomas, 1888, p. 127, profile figure). The upper lateral
teeth of Petaurus are like those of the smaller forms, and the same is true of Daciylopsila,
except that here the edge of the posterior tooth (¢. 8) is much broader, and its anterior
portion slightly more inwardly rotated than in Pefawrus. In both genera the lower
teeth are elongated and recurved, as in the smaller species, and this character is again
more pronounced in Dactylopsila than in Pelaurus.

Phalanger and Trichosurus show a more specialized modification of the incisors, but
it is of a different type from that seen in the preceding forms. The procumbent develop-
ment of the median upper teeth, which is so pronounced in the next lower member of
the series, Dactylopsila, is here scarcely in evidence. In Phalanger these teeth are
more rounded in section, and their length is reduced, so that their tips project only
slightly beyond those of the lateral teeth. In Zrichosurus the length is still further
reduced, and, except for their greater curvature, they begin to show the characters of
the lateral teeth. In Phalanger ursinus and P. melanotis the upper lateral teeth are
much like those of Petaurus and Dactylopsila, the second serving chiefly as a stop for
the lower ones, while the third, which are again spatulate at their tips, have their
anterior portions rotated inwards. The remaining species of —Phalanger show a
specialized condition, the third upper teeth being reduced through the encroachment of
the canines. In Zrichosurus the upper lateral teeth are like those of P. uwrsinus
and P. melanotis. In both genera the lower teeth are flattened, lanceolate, and only
slightly curved. In neither case do they show any tendency to continue the progressive
elongation and piercing developments characteristic of the lower members, especially
Dactylopsila.

The incisor modifications just described furnish the first instance of the conflict of
developments above referred to as characterizing the omnivorous evolution of the
Phalangerine. In the minute forms Acrobates, Distechurus, and Dromicia the
insectivorous purpose of the diprotodont modification is fully realized, the procumbent
character of the median upper teeth, and the elongated, recurved, and sharply-pointed
character of the lower ones, combining to make these structures of obvious value in the
prehension of food. Proceeding from these more or less initial forms we meet with a
series of genera showing a gradual increase in the size of the body, which we might
reasonably suppose, both on general principles and from the analogous case of the
Dasyurine, to demand the abandonment of insectivorous habit and a gradual reduction
of insectivorous characters. What we do find in the present case, however, is a tendency
to continue the insectivorous developments. This is not surprising in the case of
Petaurus, especially the smaller species P. breviceps and P. sciureus, but it certainly is
so in the case of Dactylopsila, both on account of its larger size and its actually
insectivorous habits. Passing to the still larger forms Phalanger and Trichosurus we
find the conditions abruptly changed, the insectivorous characters being replaced by

OF THE AUSTRALIAN MARSUPIALIA. 131

others which can be easily shown from their constant occurrence in the Macropodide to
be of a herbivorous type. As will be seen from a further consideration of the antemolar
teeth of this series, the changes noticeable in the incisors of the latter genera are closely
associated with others in the premolars which are almost as abrupt in their appearance
and are also characteristic of the Macropodidee.

Functional Canines.—These teeth show a fairly homogeneous evolution. As before,
the most primitive conditions are found in the three genera Acrobates, Distechurus,
and Dromicia. In the two former the teeth are long, curved, and sharply pointed, and
are apparently quite as functional as in the Dasyurine. As will be seen from the
figure of Distechurus (Pl. 5, fig. 89) they work over the sides of the lower incisors.
In both forms they have undergone an anterior progression, so that the space originally
occupied by the notch for the reception of the lower canines, and presumably also by
additional posterior incisors, has been obliterated, the third incisors being in close contact
with the canines. In Dromicia the canines are not so well developed, being shorter, less
curved, and slightly compressed; they are also separated from the third incisors by
short diastemata. Their shortened condition in this genus represents the beginning of
a process of reduction which characterizes the canine evoluuon in the present series. In
Petaurus the canines are so short as to project only slightly beyond the third incisors ;
their tips are rounded and laterally compressed. In some cases the compression is
accompanied by a grooving of the root as in P. Doreyana, Charopus, and Myrmecobius.
In Dactylopsila the canines are reduced as in Petaurus, and are also slightly compressed.
In Phalanger we find two different conditions. In P. wrsinus and P. melanotis the
canines are short and rounded as in Dactylopsila, and they project to about the same
extent as the third incisors, while in the remaining species they are enlarged, in all
probability secondarily, and present the characters of normal canines. Z'richosurus
shows much the same condition as P. wrsinus and P. melanotis, the only noteworthy
difference being a slightly greater reduction. In this form the third incisors frequently
project beyond the canines.

Functional Premolars.—The most primitive conditions are found in Acrobates. In
this form the upper premolars are of a piercing insectivorous type, exactly as in the
Dasyurine. The resemblance to the latter is increased by the fact that the posterior
teeth are slightly reduced as compared with the median, while the latter are slightly
larger than the anterior teeth. The lower functional premolars are the median and
posterior teeth. They also resemble those of the Dasyurinze in being of a primitive
piercing type; the posterior teeth are again slightly reduced. Distwehurus (Pl. 5, fig. 39)
shows a slightly more advanced stage of the same modification. The upper anterior
and median premolars are simply enlarged, while the posterior teeth are still further
reduced. In the lower jaw the posterior teeth have entirely disappeared, recalling the
condition in advanced forms of the Dasyurine. The median lower teeth show an
elongated and subcaniniform condition.

The three species of Dromicia present considerable differences in their premolar
SECOND SERIES.—ZOOLOGY, VOL. IX. 19

132 DR. B. A. BENSLEY ON THE EVOLUTION

characters. Those of D. lepida are of interest as being prototypal to those of Petaurus.
In this species the upper anterior and median teeth are small as compared with the
posterior ones, the latter, unlike those of Acrobates and Distechurus, being enlarged
instead of reduced. They present a rather different appearance from those of Acrobates,
being relatively short and provided with accessory cusps. The tips of the upper
posterior premolars are slightly flattened and bifid. Of the lower premolars the only
functional members are the posterior teeth, which are comparatively well developed
and also bifid. In D. concinna the disproportion in size between the upper anterior
and median and the posterior premolars is almost more marked than in D. lepida,
not, however, on account of the enlargement of the latter, these being in reality less
enlarged, but because of the styliform and wholly vestigial character of the former.
In the lower jaw of this species none of the premolars are functional. In D. nana,
a larger and more specialized form than the others, the anterior and median upper
premolars are vestigial, as in D. concinna, while the posterior teeth are greatly
enlarged, and much more completely bifid than those of D. /epida. As in the latter,
only the posterior premolars are functional in the lower jaw; they are enlarged but
not bifid.

As already mentioned, Petaurus approximates to D. lepida. This is especially true of
P. breviceps, the smallest of the three species. The upper anterior and median premolars
are low, with prominent accessory cusps. The median teeth, which in D. lepida are
reduced as compared with those of Acrobates and Distechurus, are here further
reduced. The upper posterior premolars are moderately developed ; they are not bifid
at the tips, although they show a tendency towards such a development in a grooving of
their outer sides in the region of the edge. ‘The lower posterior premolars are reduced ;
they show a faint tendency towards a grooving of the edge. It appears probable that
both upper and lower teeth have been formerly modified as in D. lepida.

Dactylopsila shows an advance on Petaurus, except as regards the upper anterior
premolars. The latter are seen to be undergoing a new development, becoming less
compressed, single-rooted, and subcaniniform. Apart from its striking character, this
development is of great interest as being carried over from the present form to the
succeeding genus Phalanger. The median premolars are still more reduced than in
Petaurus, being wholly vestigial or even absent. The posterior teeth are moderately
developed, and their edges are slightly compressed in an oblique direction, the anterior
portion of the edges being placed internally. his is the reverse of the condition in
Phalanger, where the edges are compressed in such a way that the anterior portions are
external. None of the lower premolars are functional in Dactylopsila; the posterior
teeth are still more reduced than in Petaurus.

In the succeeding genus Phalanger the anterior and median upper premolars show
the same trend of evolution as those of Dactylopsila, while the posterior teeth above and
below show a pronounced departure. Of the former teeth the anterior elements are
single-rooted and subcaniniform, except in Phalanger ursinus, while the median ones
are vestigial or wholly absent. The posterior premolars are enlarged as sectorials, as in
Dromicia nana. 'The various steps in the elaboration of the sectorials are illustrated to
a certain extent in the different species. In the upper teeth of P. celebensis, P. breviceps,

OF THE AUSTRALIAN MARSUPIALIA. 185

P. melanolis, P. orientalis, and P. maculatus we find that the original tip (protocone)
is anterior and more or less external in position, and that the posterior slope is com-
pressed and slightly trenchant. The cutting-blade thus formed bears one or two short
superficial grooves ending in notches at the edge. In P. melanotis, where only one
groove is present, the teeth resemble those of Dromicia nana and D. lepida. In all
cases the posterior portion of the trenchant edge tends to be elevated, but this tendency
is much more apparent in certain forms (P. orientalis) than in others (P. breviceps,
P. celebensis). In P. lullule, P. leucippus, and P. Rothschildi the trenchant character
of the edge is much more pronounced, the posterior portion of the edge being still more
elevated and bearing three grooves.

An interesting feature of the development of the trenchant edge is its axial rotation.
In P. melanotis, P. uwrsinus, and P. maculatus we find a practically unrotated condition,
so that a line drawn through the cutting-edge, if it passes through the middle line of the
skull at all, will pass posteriorly in the region of the condyles. In the case of P. orientalis,
P. lullule, and P. leucippus it will pass through the middle line forwards in the region
of the palatal vacuities. Increase in relative size seems also to be a part of the sectorial
elaboration. In P. celebensis and P. breviceps the teeth are small, while in such extreme
forms as P. leucippus and P. Rothschildi they are considerably enlarged, now exceeding
the first molars.

It is interesting to note at this point that in the Macropodidee the sectorial evolution
has proceeded along two lines, one of which, represented by Hypsiprymnodon and
Bettongia, appears to present in an extreme degree the grooved and rotated condition
found in Phalanger, while the other seems to represent a modification of the simpler,
practically ungrooved and unrotated type found in Dromicia and Petavrus.

In Trichosurus the upper anterior premolars are single-rooted, as in the two preceding
genera, Phalanger and Dactylopsila; they are on the whole, however, more reduced.
The median teeth have now disappeared. ‘he posterior premolars above and below
show much the same conditions of sectorial elaboration as are seen in the more
specialized forms of Phalanger. Their trenchant edges are elevated, as in P. leucippus,
and are rotated to about the same extent as in the latter form or P. orientalis. They
show indications of three notches. A significant feature is that the sectorial teeth
show, as a rule, more decided indications of wear in Zvichosurus than in Phalanger,
this character supplementing the evidence of the molars, canines, and incisors as to the
greater herbivority and more advanced position of the former genus.

Apart from special phylogenetic considerations, the modifications of the premolars in
Phalanger and Trichosurus possess a general significance. While these genera may be
shown by their lack of development of the median premolars to be not very nearly
related to the ancestors of the Macropodidze, they show the incipient phases of a sectorial
development of the premolars, which is continued and perfected in the latter family. The
fact that the formation of sectorials may have taken place not only within the limits of
the family but within the limits of two genera, indicates the futility of such general
comparisons as have been made in suggesting aflinities between the specialized sectorials
of the Macropodide and those of certain of the South-American Miocene forms, or of the
Plagiaulacide.

Lo?

134 DR. B. A. BENSLEY ON THE EVOLUTION

The modifications of the posterior premolars in the Phalangerinz will be seen to
furnish a further example of the conflict of developments already described for the
incisors. In Acrobates and Distechurus these teeth present two stages of a reduction
similar to that which is so characteristic of the Dasyurinze. Within the limits of the three
species of Dromicia, we find the same tendency towards reduction indicated in D. concinna,
but an opposite tendency towards enlargement as sectorials in D. lepida, and still more
in D. nana. In Petawrus and Dactylopsila the posterior premolars are comparatively
poorly developed, this being especially the case with the lower teeth, while in the
succeeding forms Phalanger and Trichosurus they are greatly developed as sectorials.
The marked difference in this respect between Phalanger and Dactylopsila is doubtless
due in part to the retention in the latter of insectivorous modifications of the incisors,
the latter teeth being thus the more functional elements; but, in general, the modifica-
tions of the whole series represent a conflict between a reduction of the posterior teeth,
as in the Dasyurine, and their elaboration as sectorials, as in the Macropodide.

Lower intermediate Teeth—Yhe functional or vestigial character of the various
teeth situate between the median lower lower incisors and the first molars in this series
probably depends on several circumstances, the elongation of the median incisors being,
however, the most important factor in the reduction of the elements of this region.
In all forms, except Trichosurus and Phalanger, there are normally two vestigial teeth —
situate immediately behind the median functional ones. They are sometimes separated
from the median premolars by diastemata which, in all probability, were formerly
occupied by the canines and anterior premolars. The vestigial nature of these teeth
can only be ascribed to the enlargement of the median incisors. ‘The latter would
doubtless have brought about the reduction of the median and posterior premolars in
all. forms, as in Petaurus and Dactylopsila, were it not for the fact that it has been
prevented from so doing by other developments. Thus, in Acrobates and Distaechurus
the median premolars have been prevented from becoming vestigial by being developed
as piercing organs. In Phalanger, Trichosurus, and Dromicia nana the posterior teeth
have been prevented from becoming vestigial by being developed as sectorials.

The only feature of phylogenetic interest presented by the intermediate teeth concerns
their relations in Phalanger and Trichosurus. In the former three vestigial teeth are
usually, although by no means always, present *, while in Trichoswrus only one is as a
rule indicated, although two may occur as a variation. This fact supports the evidence
already given as to the more advanced position of the latter genus.

TARSIPEDINE.

The single representative of this division, Zursipes rostratus, owes its distinction from
the remaining members of the family to the wholly degenerate character of its dentition,
the only teeth represented being the upper incisors, the median lower incisors, the upper

* Bateson (1894, p. 253) has given a detailed account of the variability of these disappearing elements in
Phalanger and other members of the present family.

OF THE AUSTRALIAN MARSUPIALIA. 135

canines, and a varying number of undifferentiated cheek-teeth, all of which are vestigial *,
The upper incisors are usually two in number, judging from four specimens in the
collection, and are always extremely minute and styliform. Apart from their minute
size, the upper canines bear a certain resemblance to those of Dromicia in being only
slightly curved and rounded at the tips. The median lower incisors are almost straight
and extremely slender, but are otherwise of the usual diprotodont type. The cheek-
teeth (Pl. 6. fig. 24), like the upper incisors, are minute and either simply styliform or
slightly curved. Their number appears to be usually three, and their homologies are
indeterminable.

The exact dental relations of Tars¢pes with the remaining Phalangeridze are doubtful,
and its mode of origin can only be conjectured from the association of its reduced
dentition with its mellivorous habit. It is extremely probable that the omnivorous
evolution of the Phalangeridee began with diminutive animals which, like Acrobates,
Distechurus, and Dromicia, were able to live among the smaller branches of trees, and
to supplement their insectivorous fare with blossoms and honey. Certain of these must
have continued the omnivorous evolution in a normal way, giving rise to the Phalan-
gerinze and Phascolarctine, while others resorted largely to a mellivorous habit, giving
rise to Zarsipest. The reduction of the dentition is due to the same cause as in
ant-eating forms, the food requiring no mastication, and its collection being perfectly
provided for by the prehensile development of the tongue.

PHASCOLARCTIN®.

The three genera forming the present division are of general interest as presenting a
fresh instance of the remarkable parallelisms between marsupial and placental types.
In the Macropodide, and in their prototypes the Phalangerinze, we have two groups,
roughly distinguishable by the hypsilophodont and brachybunodont modifications of the
molars. In respect to these modifications the two groups parallel the perissodactyl
Ungulata, especially the tapiroid representatives of the latter, and the Condylarthra.
The Phascolarctine now present selenodont modifications similar to those of the
artiodactyl Ungulata, especially the more primitive forms which retain the brachyodont
condition.

In the case of the Marsupials the evolution of the selenodont section is of very
limited extent, the reason being that, unlike the case of the bunodont Phalangerine,
none of the members of the original arboreal stock have given rise to terrestrial
successors. The evolution of the bunodont section would be scarcely less limited were it
not for the existence of the terrestrial Macropodidee, Phascolomyidx, and Diprotodontidee.

With regard to their special relationships, the dental characters indicate that the
genus Pseuwdochirus represents the ancestral type from which the remaining genera
Petauroides and Phascolarctus have been derived. Like Phalanger, Pseudochirus is

* Of. Waterhouse (1846).
+ The interesting observation of Gould (1863) with reference to the fly-catching propensities of a captive Tursipes
may be noted in this connection.

136 DR. B. A. BENSLEY ON THE EVOLUTION

conspicuous for its wealth of species, a condition which is doubtless due in a large
measure, as in the former genus, to the advent of a new and successful adaptation— in
the present case the formation of quadrate selenodont molars, in Phalanger of sectorials.
Petauroides, as pointed out by Thomas (1888) is indistinguishable in dentition from
Pseudochirus, and otherwise represents simply a volant form of the latter. Phascol-
arctus, although seemingly isolated by its great specialization, shows in all of its dental
characters a direct advance on Pseudochirus.

Sequence of Molar Patterns.—Notwithstanding their wealth of numbers, the species
of Pseudochirus are remarkably constant in their molar characters, and the following
description of P. peregrinus will apply almost equally to all. In P. peregrinus
(Pl. 5. fig. 17) first, second, and third upper teeth are fully quadrate in shape, like
those of the advanced Phalangerinze and Perameline, so that they are not separated
internally by angular spaces. The functional cusps are arranged in three longitudinal
rows, a condition which at first sight suggests that seen in the Peramelinze. On closer
examination, however, the teeth are seen to possess a fourth (external) row of vestigial
cusps representing the external styles of the third row in the Peramelinze and the poly-
protodont families in general. The predominant cusps of the crown are those of the first
and third rows, these representing, as in the Phalangerinz, the protocone, paracone,
hypocone, and metacone. The protocone is slightly predominant. The hypocone is
complete. There is no disproportion in size between the paracone and metacone, and
no trace of a metacone-spur. ‘The main cusps are comparatively low and are crescentic
in section, each being concave in its outer face. The protocone and metacone are placed
slightly below the level of the outer cusps. The cusps of the second row are com-
paratively small, but are of the same crescentic character as the larger ones. They are
usually smaller than in the specimen here figured. They represent the subsidiary proto-
and metaconules, which are fairly common in the Placentals, but rare in Marsupials. The
vestigial external styles are two in number; they appear to represent J (or ad) and ¢ of
polyprotodont forms. Their presence in the Phascolarctinz is of considerable interest, since
in all the remaining members of the diprotodont section they have completely disappeared.
Style c shows a slight tendency to be bifid, a condition which is partly natural and partly
the result of wear. In mastication the crowns of the lower teeth sweep transversely
across those of the upper in such a way that the triangular tips of the hypoconid and
entoconid pass through the triangular spaces separating the paracone and metacone.
Style ¢ is placed directly in the line of passage, so that its tip becomes notched by those
of the hypoconid and entoconid. In unworn teeth of P. peregrinus and P. Cook there
is seen a slight indication of a cleavage of this element, showing that the mechanical
notching of its tip has produced a natural development in that direction. In other
species of Pseudochirus there is a slight variation in the external styles: they are
obsolete in P. Dahli, and almost so in P. Forbesi and in Petauroides volans.

Like those of the Phalangerine, the upper molars decrease in size backwards. The
fourth tooth is reduced, although not to the same extent as in polyprotodont forms.
This element is in fact now undergoing a progressive evolution. In P. peregrinus it is

OF THE AUSTRALIAN MARSUPIALIA. 137

triangular in shape as in the Phalangerinz, and it shows three cusps, probably representing
the protocone, paracone, and metacone. In some species the metacone-like element
shows a tendency towards division, and this condition appears to be prophetic of that in
Phascolarctus, where the posterior lobe of the fourth molar bears two distinct cusps.
The same process of development, resulting in the formation of new cusps, with
apparent, but without actual, homologies, is also observable in the same teeth of the
Macropodidee. and also in the first lower molars of that family. Apart from its special
significance, this condition throws an interesting light on the general problem of the fate
of vestigial members. In the most primitive of the polyprotodont forms we find the
fourth upper molars in an advanced condition of reduction, while passing through a
series of intermediate forms in the Phalangeride and lower Macropodide we find these
teeth becoming finally the largest and most perfect elements of the upper molar series.
Both in the Macropodidze and the Phascolarctinze the cause of this peculiar development
is to be sought in the greater possibilities of service involved in the substitution of a
transverse grinding action of the teeth for a vertical cutting and piercing one. In
primitive polyprotodonts, as already indicated, the action of the metacone of the anterior
upper molars is directed backwards against the trigonid of the succeeding lower teeth.
In the fourth upper molar the metacone is reduced, the reason being that there is no
trigonid working behind it, that of the lower tooth being in relation with the third upper
one. In the Phascolarctine, as in the Macropodide, on the other hand, the action of
the metacone is directed forwards against the hypoconid and entoconid. The fourth
tooth may therefore become capable of service by the development of its vestigial
metacone into a functional cusp, and may increase its utility by the addition of a new
cusp simulating the hypocone in normal teeth.

The lower molars of P. peregrinus (Pl. 6. fig. 22) are brachyodont, and, like those of
Thylacomys and the Phalangerinze, completely quadrituberculate. The selenoid modi-
fications are confined for the most part to the outer cusps, the inner cusps being
rather blade-like than crescentic. It will be seen that the concavities of the outer cusps
occupy their internal faces, this condition being the reverse of that in the upper molars.
In the first lower molar the protoconid is greatly reduced and flattened against the side
of the metaconid. In P. Albertisi, however, it tends to be set apart, a condition which is
found to recur in Phascolarctus. From the presence of two distinct cusps in the latter
form, however, it is possible that the element here designated as a metaconid may be
in reality a protoconid, the former having disappeared as a result of an insectivorous
specialization, as it undoubtedly has in the case of the Phalangerinze. According to this
view, the outer element represents a new accessory cusp. Such a development of a
new protoconid-like element can be directly traced in the Macropodide.

Passing from Pseuwdochirus to Phascolarctus we find a more specialized stage of the
same modification found in the former genus. In the upper teeth (Pl. 5. fig. 18) only
the main cusps are well developed, the intermediate conules being very minute. All of
the main cusps are more completely selenoid than in Psewdochirus. ‘The protocone is

138 DR. B. A. BENSLEY ON THE EVOLUTION

usually more prominent than the hypocone from an inward extension of its base. This
condition appears to be a remnant of that seen in more primitive trituberculate forms.
The spaces between the cusps present a curious appearance, the surface being thrown
into small ridges. The posterior shearing edge of the paracone is frequently supple-
mented by a larger ridge passing down its postero-internal side. The external styles are
almost as conspicuous as in Pseudochirus, not, however, on account of their better
development, but from the greater concavity of the outer cusps; they are in reality
more vestigial than in the latter genus. Style & does not project beyond the level of the
paracone ridge which joins it ; style ¢ is very minute and bifid.

In the fourth upper molar the posterior lobe bears two perfectly developed selenoid
cusps, so that the tooth is almost completely quadrate like the others. As already
mentioned above, this represents the final stage of a process of cleavage of an originally
single cusp.

The lower molars of Phascolarctus (P1. 6. fig. 23) resemble those of Pseudochirus even
more closely than do the upper. ‘The selenoid character is more pronounced in the outer
than in the inner cusps. The latter are, however, more selenoid than in Pseudochirus,
a slight concavity having now appeared on their inner faces. These teeth show a
tendency towards the addition of an external cingular ridge, obliterating the notch
originally present between the protoconid and hypoconid.

Origin of Selenodont Molars.—Like those of the Phalangerine, the molars of the
present division are comparatively well specialized. Apart from the more or less
completely selenoid modifications of the cusps, the hypocone will be seen to be well
developed in the upper teeth, while the paraconid and hypoconulid are wholly absent in
the lower. They are thus considerably removed from the secodont type characteristic of
the polyprotodont section. The question here arises as to the intermediate stages which
have been passed in the development of the selenodont molars up to the Pseuwdochirus
stage.

It may be admitted at the outset that the evidence on this subject is of a rather
fragmentary kind. There is, in the first place, no marsupial group outside of the
Phascolarctinze which shows incipient stages of the same evolution, as Caluromys does
with reference to the Phalangerinz, and, in the second place, the evidence as to the
origin of the selenodont molars in the parallel case of the artiodactyl Ungulata is not
wholly satisfactory *.

Winge has expressed the opinion that the molar patterns of Phascolarctus are
ancestral to those of the Phalangerinz, intermediate conditions being found in Pseudo-
chirus. The main reason advanced for this view is that Phascolarctus shows the
same angular character of the molar cusps as is found in the polyprotodont forms,
and further that it possesses vestigial external styles, which have disappeared in the
Phalangerine.

Three propositions may be considered, namely: (a) that the molar patterns of the
Phascolarctinz are ancestral to those of the bunodont Phalangerinz ; (0) that the reverse

* The available evidence fayours bunodont origin of selenodont (artiodacty]) types.

OF THE AUSTRALIAN MARSUPIALIA. 139

is the case; (c) that the molar patterns of each group have been independently derived
from a secodont type, like that seen in Perameles Doreyana.

On comparing the mclars of Phascolarctus with those of the poiyprotodonts, it will be
seen that there is ample evidence in favour of Winge’s view that certain characters
of the former are directly derivable from those of the latter. J°or example, the selenoid
modifications of the protocone, paracone, metacone, hypoconid, and protoconid are
already present in the polyprotodont forms, and are only more fully elaborated in the
Phascolarctinz. The hypocone is crescentic in the Peramelide. It may also be shown
that the selenodont molars could not have originated from a bunodont type: in the first
place, on account of the fact that the external styles are absent in the latter; in the
second place, that the outer and inner cusps in the lower molars of the Phascolarctin
present a differentiated condition which would in all probability not be indicated if all of
the cusps had been derived from a common bunoid type. Jn the polyprotodonts the
outer cusps are quite selenoid, with their concavities external as in the Phascolarctine.
The inner cusps, and especially the metaconid, also tend to be selenoid, but their con-
cavities are on their external sides, or, in other words, oa the wrong side, with reference
to the Phascolarctine. The blade-like character of the inner cusps in the latter seem
therefore to indicate an intermediate stage in the conversion of semiselenoid cusps with
external concavities into selenoid cusps with internal concavities. The condition found
in Pseudochirus is just on the line between the two phases, and that of Phascolaretus
just beyond.

There are, however, no reasons for concluding from this that the selenodont inolars of
the Phascolarctinze are ancestral to the bunodont patterns of the Phalangerinee, while to
regard the dental characters of Phascolarctus as ancestral to those of Pseudochirus is
simply to reverse the order of evolution. As already mentioned above, the various
members of the Phalangerine show a close general sequence of dental modifications, the
most primitive conditions being found in Acrobates, Dromicia, and Distaechurus. The
ancestors of the Phalangerinze were minute insectivorous forms, which, apart from the
diprotodont modification of the antemolar teeth, possessed a full antemolar formula.
Their secodont molars began to be modified in the same way as those of Caluromys are
being changed at the present time. They could not lave possessed any of the special
characters which now distinguish the Phascolarctine, with the exception of the
crescentic character of certain of the cusps and the possession of external styles. The
Phascolarctinz are comparatively large specialized forms, which are well advanced in an
herbivorous evolution, the latter being here, as in the Macropodidee, Phascolomyide,
Diprotodontide, and specialized Phalangerinz, secondary, and succeeding, not preceding,
the insectivorous evolution as exemplified by the primitive Phalangerinze. The fact that
the Phascolarctinze show resemblance to the polyprotodonts in the character of certain
cusps means nothing more than that in the formation of selenodont molars from others
of secodont type these cusps have been able to undergo a complete change of function
with little alteration of their original character. The molars of the Phascolarctinze
and Phalangerine have been derived by a divergent evolution from a common
insectivorous secodont type. It is possible that, as in the primitive Peramelinz, this

SECOND SERTES.—ZOOLOGY, VOL. IX. 20

140 DR. B. A. BENSLEY ON THE EVOLUTION

type involved an incipient hypocone and slightly reduced paraconid and hypoconulid.
The omnivorous stage in the evolution of the Phascolarctinzee must have been of
extremely short duration.

Incisors.—Within the limits of the genus Psewdochirus we find several variations in
the characters of the incisors, especially those of the upper jaw. There is, however,
evidence of a ground type corresponding in a general way to that represented by the
smaller Phalangerinee.

In P. Forbesi, a small form showing, for the most part, primitive conditions, the
median upper incisors are of an almost insectivorous type. ‘Their tips are rounded in
section and pointed, and they project a considerable distance beyond those of the lateral
teeth. They are directed towards one another in the middle line. The second and third
upper teeth are subequal in size, the second shows a very slight thickening of the tips.
The enlarged lower teeth are lanceolate and only slightly curved; their tips, which are
rounded, work against the posterior sides of the median upper teeth, while their sides
cut against the lateral ones, especially the second.

In P. Albertisi the median upper incisors are much as in P. Forbesi, but are stouter.
The second upper teeth have bulb-like tips. The third upper teeth are relatively small
‘And unmodified. The lower teeth are flattened like those of P. Forbest, but, unlike the
latter, their external cutting-edge is concave, so as to fit the convexity of the upper lateral
teeth. Somewhat similar conditions are found in P. Corinne, P.cupreus, and P. Dahli.
In P. Corinne, however, the median upper incisors are very much elongated, projecting
downwards, but only slightly forwards, and are also separated from one another in the
middle line, this development producing an appearance similar to that seen in certain of
the Insectivora and Hyracoidea. Much the same modification is found in P. Dahli,
although the teeth are not elongated to the same extent as in P. Corinne. With
reference to the second upper teeth, it is difficult to decide whether or not their tips are
bulbous in P. Corinne and P. Dahli, these teeth being much worn in all of the specimens
examined. In P. Dahli the external cutting-edges of the lower incisors are not concave
as in the remaining species except P. Forbes.

In P. peregrinus, P. occidentalis, P. Cooki, and in Petauroides volans we find modifi-
eations of a slightly different kind. The median upper incisors are reduced, so that they
project only slightly beyond the lateral teeth. They present a compressed appearance
which is due to the excavation of their postero-internal surfaces where they come in
contact with the lower teeth. They are separated at their bases and approximated at
their tips. Apart from their more reduced characters, they resemble somewhat those of
Petaurus. The second and third teeth are much as in P. Forbesi, being subequal,
slender, and on the whole unmodified. The lower teeth are only slightly curved and are
fully lanceolate. The internal edges are not concave.

The above facts, although not exhaustive, are sufficient to indicate that the main
feature on which the variations develop in this genus is one in which the median upper
incisors are pointed, elongated, and procumbent, the lateral incisors subequal and inturned

OF THE AUSTRALIAN MARSUPIALIA. 141

to act as stops for the lower teeth. The lower incisors appear to be already modified
throughout as cutting rather than simply piercing organs.

The incisor modifications of Phascolarctus indicate a much more specialized condition
than those of Pseudochirus. The chief difference relates to the tendency towards the
development of rodent characters. The lower teeth meet the upper at a wide angle
(about 90°), so that instead of cutting by their edges against the tips of the upper lateral
teeth they cut entirely by their tips against those of the median upper ones. The upper
lateral teeth therefore act as stops. ‘The median upper teeth are rounded in section and
flattened from before backwards at the tips. The lateral teeth are slender, and at most
only slightly thickened distally as in most species of Psewdochirus. The lower teeth are
lanceolate when unworn, and scalpriform when worn. In the peculiar manner in which
the median upper and lower teeth are fitted together, as also in the slenderness of the
lateral teeth, Phascolarctus shows an interesting approximation to Diprotodon, although
there is no tendency in the former towards the differentiation of enamel bands.

As in the case of the molars, there is nothing primitive in the characters presented by
Phascolarctus as compared with those of the Pseudochirus type. The unworn lower
teeth of Phascolarctus, on the other hand, furnish direct evidence that a cutting
modification like that of the latter type is the original condition from which the
scalpriform one has been derived.

Upper Canines.—These teeth show few points of phylogenetic interest. In all
members of the series they are reduced and apparently functionless, this condition being
a specialized one, as in the advanced Phalangerinz, and contrasting with the primitive
one in the smaller members of the latter group. They are short, often laterally
compressed, and in general bear little resemblance to normal canines.

Functional Premolars.—Pseudochirus and Petauroides are primitive in respect to the
possession of a full upper premolar formula. ‘The anterior teeth are reduced, the
median moderately developed, and the posterior slightly enlarged, the condition being
also fairly primitive. The median teeth show slight indications of compression of the
edges, and this character is still more pronounced in the posterior teeth. The latter,
unlike the compressed teeth of the advanced Phalangerinz, have their original tips in
the centre instead of at the anterior end. These teeth show signs of the formation of
an internal ledge and of a postero-internal cusp. In the lower jaw the conditions
are. more advanced, the only functional teeth being the posterior. Like the upper
teeth their edges are compressed, and they also show indications of an anterior accessory
cusp.

The condition in Phascolarctus appears to be a direct advance on this type. The
anterior and median upper premolars have disappeared, while the posterior have become
enlarged. The latter are of much the same character as in Psewdochirus, but the
edge is more uniformly trenchant and the internal ledge better developed. ‘he lower

20*

142 DR. B. A. BENSLEY ON THE EVOLUTION

teeth show much the same relations. The cutting-edge tends to be duplicated behind |
and a trace of this formation is also observable in Pseudochirus.

Lower intermediate Teeth.—The only interesting feature of these teeth is the fact
that they are usually indicated in Pseudochirus and Petauroides, but are absent in
Phascolarctus, thus supporting the evidence of incisors, canines, and molars as to the
more advanced position of the latter genus. Even in Psewdochirus there is considerable
variation in the extent of development of these vestigial members as in the Phalangerine,
the number varying from none to four.

MACROPODID.

From what has been said with reference to the preceding families, it will be apparent
that on passing successively through insectivorous and omnivorous modifications in the
Dasyuride, Peramelidie, and the primitive Phalangeridz, we finally meet with incipient
herbivorous characters in the advanced members of the latter family. As will be pointed
out in greater detail in a subsequent section, if we follow the modifications of the foot-
structure in the same forms we find that these successive phases of dental evolution are
closely associated with successive phases of arboreal adaptation. The development of
each new phase of dental evolution in arboreal animals appears to have opened a field
for the continuation of that evolution in terrestrial ones. At any rate, we find that at
the insectivorous stage the arboreal Dasyuride have given rise to a number of terrestrial
forms which now form a prominent section of the family, while at the omnivorous stage
another series of arboreal forms, now without typical representatives, have given rise to
the terrestrial Peramelide. inthe same manner the appearance of incipient herbivorous
characters in the Phalangeridz appears to have been marked by the origin of a series
of terrestrial herbivores differentiating into the Macropodide, Phascolomyidze, and
Diprotodontide.

All of the last-named families are of bunodont origin—that is, if we may judge the
derivation of the lophodont molars of the Diprotodontidz by analogy with those of the
Macropodidee. Of the two principal divisions of the Phalangeride their relations are
accordingly with the Phalangerine.

On comparing the dental characters of the herbivorous families it becomes apparent
that the phalangerine dentition is capable of giving rise to more than one line of
herbivorous evolution. ‘Thus the Phascolomyide present rodent modifications throughout
the dentition, while the Diprotodontidee present rodent modifications of the incisors
combined with lophodont characters of the molars. The Macropodidee alone show a
uniform continuation of that type of herbivorous evolution which is found in its incipient
stages in the advanced members Phalanger and Trichosurus of the Phalangerine, although
it may be shown from the retention of the median premolars that they are the
descendants of forms slightly more primitive than the latter genera.

OF THE AUSTRALIAN MARSUPIALIA. 145

Generally speaking, the dental evolution of the Macropodide has been connected
with the perfection of the dentition for a grazing habit, and this, in turn, appears to
have been closely associated with increasing capabilities of the animals for terrestrial
progression. Judging from the sequence now presented, the ancestors of the family
possessed a type of dentition fully as primitive as that of Petaurus, and their sphere
of action must have been greatly limited by their small size and the shortness of their
limbs. ‘Taking these phalangerine characters into consideration, it is probable that the
original forms lived among the low scrub, subsisting on leaves and shoots, and possibly
supplementing this fare with roots and grasses. The most specialized members of the
family, namely, the Kangaroos, judging from the present characters and habits of the
smaller forms, are the descendants of such phalangerine forms, which, on becoming larger
and longer limbed, and from browsing on the herbage of the scrub, have taken to grazing
in its open spaces, then to the open forest-glades, and finally to the open plains.

In ascribing to the Macropodide the perfection of grazing adaptations as the chief
feature of their dental evolution, a few reservations must be made. The most primitive
members of the existing family (Bettongiinze, Potoroinze) are forms which have not taken
part in the grazing evolution, the direction of their special evolution being, with one
exception, away from rather than towards the typical forms. Furthermore, it seems
probable that certain of the extinct forms, such as Procoptodon, judging from the
characters of the incisors, possessed feeding-habits of a different kind from those of their
grazing contemporaries. Finally, certain of the more advanced existing members of the
family, embracing the Tree- and Dorca-Kangaroos, from being terrestrial, have become
temporarily or permanently arboreal, and while retaining many primitive characters have
undergone a divergent evolution of sectorial premolars as a result of shoot-eating habits.

The various existing genera of the family may be arranged, on a basis of their premolar
and molar characters, as follows :—

A. Molars lophodont, their anterior and posterior pairs of cusps completely connected
by transverse crests . Es cs 6. 0 jets TemeEOtS, Sct . . Macropodine.
(2) Molars brachyodont : sectorial premolars excessively developed.
Genera: Dendrolagus, Dorcopsis, Setonyx.
(6) Molars hypsodont ; sectorial premolars moderately developed.
Genera: Lagorchestes, Lagostrophus, Onychogale, Petrogale, Macropus.
B. Molars quadrituberculate, their cusps only incipiently lophoid.
(a) Sectorial premolars with from two to four short superficial grooves . . , Potoroine.
Genera: Potorous, Caloprymnus.
(6) Sectorial premolars with six or more long, narrow, and prominent grooves . Bettongiine.
Genera: Hypsiprymnodon, Bettongia, Apyprymnus.

BETTONGIIN®.

The three genera Hypsiprymnodon, Bettongia, and Apyprymnus, which are included
in this division, owe their distinction to the possession of well-differentiated sectorial
premolars associated with primitive characters in the remaining teeth. They present a
line of dental evolution entirely distinct from that beginning in the Potoroine and

144 DR. B. A. BENSLEY ON THE EVOLUTION

continuing in the Macropodine, and have in all probability originated independently
from phalangerine forms possessing the same peculiarities. The extinct form Burramys,
described by Broom (1896), is amember of the same general series, although, as mentioned
by that writer, its position cannot be exactly determined at present beyond that it
presents phalangerine as well as macropodine resemblances, and that it is not directly
ancestral to Hypsiprymnodon.

The special evolution of the group is chieily marked by the progressive modifications
of the molars and sectorial premolars, the canines and incisors retaining for the most part
primitive relations throughout. The various forms, with the exception of Bettongia
cuniculus, appear to be related in the following order :—Hypsiprymnodon moschatus,
Bettongia penicillata, B. Gaimardi, B. Lesueuri, and Atpyprymnus rufescens. B. cuni-
culus appears to represent a side development of B. penicillata, or possibly B. Gaimardi.

Sequence of Molar Patterns.—The progressive characters noticeable in the molars are
as follows:—(a) tendency towards reversal of the relative proportions of the three
anterior teeth above and below ; (2) rotation inwards of the posterior ends of the tooth-
rows; (¢) slight hypsodontism of the crowns.

In the most primitive form, Hypsiprymnodon moschatus, the first and second upper
and lower teeth are approximately equal in size, while the third are slightly smaller than
the second and the fourth smaller than the third. The tooth-rows are straight, as in
some of the Phalangerinee. The teeth are quadrituberculate and brachybunodont, as in
the latter group. In the upper teeth the protocone and hypocone are placed on a level
with the outer cusps, and the internal sides of the latter elements tend to form low
transverse ridges, both as in the Phalangerine. There are no traces of external styles.
The last upper molar is triangular, as in the Phalangerinze, but its posterior lobe shows
indications of division into two cusps. In the lower molars the inner cusps, like the
outer cusps of the upper teeth, show indications of the development of low ridges. As
in the advanced Phalangerinze, the first lower molar has but one cusp on its anterior
lobe.

Passing through the remaining members of the series we find these characters
changing as follows :—The second teeth from being equal in size to the first in Hypsi-
prymnodon and in some specimens of Bettongia penicillata become slightly larger than
the latter in other specimens of B. penicillata, in B. Gaimardi, and in B. Lesueuri, and
considerably larger in Zpyprymnus. The third teeth from being smaller than the
second in Hypsiprymnodon and Bettongia become larger in A/pyprymnus. The changes
in size of the fourth molars are not serial: in B. penicillata and B. Lesweuri they are
of about the same size as in Hypsiprymnodon, and as the former are much larger
animals they appear out of proportion to the remaining teeth. In B. Gaimardi and
in Apyprynmnus they are of about the same relative size as in Hypsiprymnodon,

The changes just described, namely, the reversal of the relative proportions of the
molars, are of considerable significance, in that they represent a necessary feature of the
herbivorous evolution, since they are indicated not only in the present group but also

OF THE AUSTRALIAN MARSUPIALIA. 145

in the Potoroinze and Macropodine. As will be more apparent from a consideration of
the last-named division, the advent of herbivorous habits involves extensive wearing of
the molar crowns, which in the course of evolution is counteracted in part by a hypsodont
development and in part by a functional replacement of the teeth from behind forwards.
The anterior molars, since they come into service before the posterior ones, are naturally
worn down in the young animal, and if the teeth decreased in size backwards, as in the
Phalangerine, the adult would be provided with a molar dentition totally inadequate for
its needs. Instead of this, the posterior molars increase at a greater rate than the
anterior ones, and also replace the latter when worn down, so that the animals are
provided with an efficient dentition up to an advanced age. In the Phalangerinze and
in those forms of the Macropodidee which have not yet advanced to a definite herbivorous
phase, the anterior molars may conveniently be larger than the others, because they
come into service at an earlier period and are functional throughout life.

In Hypsiprymnodon the molars are wholly brachyodont, as in the Phalangerine, in
Bettongia their crowns are slightly elongated, and in Zpyprymnus considerably so. The
molar patterns of Gettongia present the same characters as those of Hypsiprymnodon.
In Lpyprymnus the increasing hypsodontism of the crowns causes the transverse ridges
to be more prominent than in the remaining forms. The posterior lobe in the fourth
upper molars shows indications of two cusps in all of the forms. The anterior lobe of the
first lower molars, both in Bettongia and pyprymnus, bears an accessory cusp on its
outer slope. The new cusp takes up the position of a protoconid, but is obviously
not homologous with that element, the true protoconid being the inner cusp, the true
metaconid having been lost in the phalangerine stage.

Incisors.
gated. They are scarcely procumbent, and in this respect resemble those of the
Potoroinze and some species of Pseudochirus. The tips of the upper lateral teeth are
spatulate when unworn, and turned inwards so as to stop the passage of the lower teeth.
The latter are lanceolate. The general condition is little removed from that of the
advanced Phalangerine. The special developments are few. Settongia penicillata
practically repeats the characters of Hypsiprymnodon. B. Lesweuri shows a slight curving
of the median upper teeth with antero-posterior compression of their tips, and also a
turning inwards of the anterior parts of the tips of the third teeth. All of these
characters are indicated in Zpyprymnus, and in a more marked degree,

Throughout the series the median upper incisors are enlarged and elon-

Canines.—These elements are fairly constant throughout, being greatly reduced but
not wholly vestigial. The canines are the last of the reduced elements to disappear in
the antemolar region. Their reduction is not, as in the case of the lower intermediate
teeth, attributable to the presence of the diprotodont modification, since they are seen
to persist long after its introduction, but, as in ungulate Placentals, to the effects of
herbivorous evolution.

14.6 DR. B. A. BENSLEY ON THE EVOLUTION

Median and posterior Premolars.—The modifications of these teeth are of interest as
showing much more clearly than those of the remaining teeth the sequence of species in
the present series and their general relations with the Potoroine. It appears to be on
them that the dental change in this series has been mainly developed. The evolution is
marked by the following characters :—(@) increase in size and functional importance,
with increase in the number of the grooves; (4) gradual rotation of the cutting axis
inwards anteriorly. In Hypsiprymnodon (PI. 5, figs. 31, 34) the posterior upper pre-
molars are comparatively short and their cutting-edges rounded. They are marked by
six prominent curved grooves ending in notches. In Bettongia penicillata (Pl. 5.
fig. 832 0) these teeth are more elongated, and the number of grooves is increased to
seven. The most interesting feature, however, is that their anterior portion has the
same characters as the whole tooth in Hypsiprymnodon, showing approximately the
same number of grooves and the same rounding of the edge. They have apparently
been derived from such a type as is seen in Hypsiprymnodon, simply by the lengthening
of the posterior portion. In B. Gaimardi, B. Lesueuri, and Aipyprymnus the anterior
portion of the cutting-edge is successively levelled, so that all parts of it come to project
to about the same extent. In B. Gaimardi the number of grooves is increased to eight, in
B. Lesueuri to ten, and in _Apyprymnus to eight. Correlated with these modifications is a
rotation of the cutting-edge, by which the latter from being directed outwards anteriorly
comes to be directed inwards. 'The main features of this peculiar change are represented
in Pl. 5, figs. 34, 35, 36, where the upper teeth of the left side in Hypsiprymnodon,
B. penicillata, and B. Gaimardi have been drawn so that their cutting axes form the
same angle with reference to the border of the Plate that they do with reference to the
middle line of the palate. In Hypsiprymnodon the tooth is rotated outwards anteriorly,
so that the edge forms a considerable angle with reference to the middle line. In
B. penicillata what has been referred to as the original or anterior portion of the tooth
is rotated outwards, as in Hypsiprymnodon, while the newer posterior portion is approxi-
mately parallel to the middle line. In B. Gaimardi the whole tooth is straight, while
in B. Lesueuri it is slightly rotated inwards, and in _Zpyprymnus considerably so.

It appears at first sight extraordinary that such a change as that just described,
involving a torsion of a sectorial tooth through an angle of nearly 90°, should represent
a natural progression. The explanation becomes apparent, however, on comparing the
modifications of the muzzle in the different species. In Hypsiprymnodon the latter is
relatively long, but in the succeeding members it becomes gradually shorter and broader.
At the same time the sectorial premolars become elongated antero-posteriorly. If under
both of these changes the premolars should retain their original outwardly rotated con-
dition, either their anterior and presumably most essential portion would have to be
supported on a weak projecting pillar or their inner portion would have to encroach on
the palate. As it is, the rotation of the teeth in the opposite direction enables them to
be constantly supported directly in the line of the outer border of the jaw.

The above descriptions of the upper posterior premolars apply almost equally to the
lower teeth. The median premolars, which perform the same functions in the young as
the posterior teeth, replacing them in the adult, also show a progressive development

OF THE AUSTRALIAN MARSUPIALIA. 147

of much the same kind. It is an interesting and suggestive fact, however, that the
modifications of the median premolars always represent a more primitive stage of the
same modifications seen in the posterior premolars of the same forms. Thus the median
upper premolar of B. penicillata (Pl. 5. fig. 32) is much simpler than the posterior
tooth, being shorter and bearing a smaller number of grooves. It will be apparent that
if the present series is a natural one then the posterior premolars of the more primitive
forms should resemble the median premolars of more advanced forms. The species of
Bettongia and Aipyprymnus are scarcely removed to a sufficient extent from one another
to present such relations ; but between B. penicillata and Hypsiprymnodon, two forms
which, except for the dental resemblances already described, we would scarcely suspect
to be related as successor and ancestor, from a comparison of other characters, the
correspondence is exact and unmistakable, the median premolar of the former (fig. 32 a)
almost repeating the characters of the posterior premolar of the latter (fig. 31).

A further application of this principle to the median premolars of HHypsiprymnodon,
the existence of which is, however, only presumed at present, might throw some light
on the characters of the posterior teeth in its phalangerine ancestors, and also facilitate
the comparison of the animal with the peculiar form Burramys (Broom, 1895), in which
the posterior premolars are modified, as in the present series, while the median teeth,
although placed far back in the jaw, are not modified and not replaced in the tooth
change.

POTOROIN&.

The chief interest of the two genera Potorouws and Caloprymnus, composing the
present division, lies in their general prototypal relations with the Macropodine. It will
be seen on comparison of the present series with the Bettongiinze that while both groups
present primitive phalangerine characters in the general dentition, the Potoroinz alone
present premolar characters which correspond to those found in the more primitive
forms of the Macropodine ; none of the members of the two latter groups show the finely
grooved and rotated condition of the sectorials found in the Bettongiinie.

The species of Potorous form a natural series, of which the incipient or most primitive
member is P. platyops, a form which shows an interesting approximation in many of its
dental and also of its cranial characters to Petaurus, suggesting an aflinity with Gymno-
belideus, the non-volant form of the latter. The remaining members of the series are, in
order, P. Gilberti and P. tridactylus, the typical Tasmanian form of the latter (P. apicalis)
being more advanced than the New South Wales form. It is interesting to notice that
the evolution cf this series is associated with an elongation of the muzzle, a modification
which is exactly the reverse of that characterizing the Bettongiinze. Caloprymnus does
not continue the evolution of Potorous, but probably represents a side development from
a primitive species of the latter. It combines nearly all of the ancestral phalangerine
characters seen in Potorous with certain special modifications pointing towards the
Macropodinee. Unlike Potorous, it shows no tendency towards an elongation of the
muzzle.

SECOND SERIES.— ZOOLOGY, VOL. IX. 21

148 DR. B. A. BENSLEY ON THE EVOLUTION

The evolution of Potorous presents a further instance of the principle of increase in
size of the body already noticed in the case of the Dasyuridz and Phalangeride. An
interesting exception is, however, found in the final forms, where of the two Tasmanian
derivatives of the New South Wales species, P. t7idactylus, one is a more progressive and
larger form, while the other (P. rufus) is a dwarf (¢f. Thomas, 1888, p. 120).

Sequence of Molar Patterns——In Potorous platyops the molars are comparatively
small and lightly built, as in Pefawrus. The tooth-rows are very slightly bent inwards
posteriorly. In the upper jaw the first molars are equal in size to the second, from
which point they decrease gently in size backwards. The same conditions obtain in the
lower jaw, except that the first teeth are slightly smaller than the second. It will be
observed that we have here the same primitive size-relations common to the incipient
members of the Bettongiine and to the Phalangerine. Both upper and lower teeth are
of the usual quadrituberculate bunodont type. ‘The protocone and hypocone in the
upper molars are placed on a level with the outer cusps, and the latter, as well as the
inner cusps of the lower teeth, show a tendency towards the formation of transverse
ridges. The fourth upper teeth are almost quadrate; they have evidently been
triangular at an earlier stage, but now show a division of the posterior lobe into two
cusps. The patterns of the first lower molars are doubtful, these teeth being much
worn in the specimens examined. All of the molars are longer in the crown than those
of Petaurus, but would nevertheless be described as brachyodont.

Passing through the remaining species of Potorous we find the molars showing but
little modification. They become slightly more robust and more hypsodont. Their
proportions change, so that the three anterior teeth above and below become equal in size.
In P. Gilberti and P. apicalis the anterior lobe of the first lower molar bears an accessory
outer cusp. As in the Bettongiinie, this tooth is probably changing from the insecti-
vorous modification characteristic of the Phalangerinz to the quadrituberculate one seen
in normal molars.

In Caloprymnus the general characters of the molars are as in Potorous. The upper
teeth decrease gently in size backwards, while the lower increase in size to the third.
The tendency towards hypsodontism is almost more pronounced than in the final form
P. tridactylus (apicalis) of Potorous. The patterns of the molars are as in Potorous,
except that the ridging of the cusps is, if anything, more pronounced. On the whole,
this form points towards the Macropodine.

The molar patterns of Caloprymnus are represented in Pl. 5, fig. 19, and Pl. 6. fig. 25,
for comparison with those of the phalangerine genus Zvichosurus (Pl. 5. fig. 16, Pl. 6.
fig. 19), on the one hand, and with those of the Macropodinee (PI. 5, figs. 20-23, Pl. 6.
figs. 26-29), of which they may be taken as representing the ancestral type. The chief
progressive changes relate to the development of the transverse ridges. As already
described, in the case of the Phalangerine, Bettongine, and Potoroinz, with the
exception of Caloprymnus, to a certain extent also of Mpyprymnus, what little ridging
is indicated is confined to the outer cusps of the upper molars and the inner cusps of
the lower ones. In Caloprymnus the ridges in both upper and lower teeth will be seen

OF THE AUSTRALIAN MARSUPIALIA, 149

to show a tendency towards extension to the remaining cusps. This development is
continued in the Macropodine with the rapid. obliteration of the originally cuspidate
character. It is associated with a vertical lengthening of the newly formed crests, and
the partial conversion, in most cases, of the internal longitudinal bands of the
upper teeth, and the corresponding external bands of the lower teeth, into longitudinal
crests,

Incisors—In Potorous platyops these teeth are much as in Petaurus. The median
upper incisors are prolonged beyond the others, and are also stout, gently curved, and
slightly flattened. The second upper teeth are small, and act as a stop for the lower
teeth, while the third are flattened and have their cutting-edges elongated. The lower
incisors are not greatly elongated as in Petawrus, but have their tips gently curved
upwards as in that form.

Passing through the remaining species of Potorous, we find much the same conditions
as in P. platyops, except that the median upper incisors become slightly straighter,
stronger, and more angular, the second teeth larger, and the lower teeth straighter and
more elongated, so as to project forwards but only slightly upwards.

The incisors of Caloprymnus resemble those of P. platyops, but show some characters
pointing towards the Macropodine. These relate chiefly to the tendency of the median
upper teeth to become flattened so that their edges shear against the tips of the lower
teeth instead of simply piercing against them. Both the second and third upper teeth
serve as a stop for the lower ones, but show a slight tendency to shear against their
sides. The latter show much the same proportions as those of P. platyops.

Canines.—These teeth are reduced, but not wholly vestigial, in all of the species of
Potorous, as in the Bettongiine. In Caloprymnus they are quite vestigial, more so, in
fact, than in many of the Macropodine.

Median and posterior Premolars.—As in the case of the Bettongiin, these teeth
furnish the most valuable indications as to the direction of the dental evolution. Unlike
those of the latter division, their axes are never rotated, but are always placed in a line
with the molars, as in Macropodine. In Potorous platyops the posterior premolars are
extremely simple. The upper teeth are small; their edges are emarginate and bear two
short superficial grooves. In P. Gilberti these teeth are more elongated, and the number
of grooves is increased to three. In P. tridactylus (N.S.W.) the teeth bear three grooves
and indications of a fourth, while in the Tasmanian form, P. apicalis, there are four
fully developed grooves. In both of the latter species the teeth are still more elongated.
In the lower teeth a similar sequence is observable. In P. platyops and P. Gilberti
there are two grooves, in P. ¢ridactylus and P. apicalis there are three with indications
of a fourth. he dwarfed Tasmanian form (P. 7u/us) shows three grooves in both upper
and lower teeth.

In Caloprymnus (Pl. 5. figs. 33, 37) the posterior upper premolars are slightly

elongated, and the edge bears two superficial grooves, with indications of a third.
21*

150 DR. B. A. BENSLEY ON THE EVOLUTION.

These grooves are, however, so poorly marked that they do not produce the conspicuous
notching of the edge seen in Potorous. A noteworthy feature is the presence of a
postero-internal cusp. While indicated in Hypsiprymnodon and Bettongia, this element
is highly characteristic of the Macropodinw. The lower teeth are similar to the upper,
but the accessory cusp is external.

As in the Bettongiinze, the modifications of the median premolars follow those of the
posterior teeth, but represent a backward stage of development. ‘These teeth are
unfortunately not indicated in any of the specimens of P. platyops and P. tridactylus.
In P. Gilberti they are decidedly more primitive than the posterior teeth, the upper ones
being comparatively small and broad, and bearing two notches like the posterior teeth
of P. platyops, while the lower ones show only one. In both of the Tasmanian species
the upper teeth show two notches, while the lower show two. In Caloprymnus they
show indications of two notches, above and below.

A comparison of the most primitive member Potorous platyops of the present series
with the most primitive form of the Bettongiine, Hypsiprymnodon moschatus, in respect
to the characters of the posterior premolars, will show sufficiently the necessity of
regarding the existing Macropodide as of diphyletic origin. In the Bettongiine the
premolars are well developed, conspicuously grooved, and their axes rotated outwards in
the most primitive of the known forms, and the occurrence of similar modifications in
the allied form Burramys, which shows indications of phalangerine affinity in its median
premolar, as in its mandibular characters, furnishes some evidence of the former
existence of the same peculiarity in a section of the Phalangerine now without repre-
sentatives. The posterior premolars of Hypsiprymnodon resemble in a very general way
those of the more specialized forms of Phalanger. In the Potoroinew the posterior
premolars, in the most primitive condition known, are small and scarcely grooved, and
their axes are in the same line with the remaining teeth. We can only assume, there-
fore, that their characters have developed from somewhat similar ones in phalangerine
ancestors. The single-notched lower premolar of P. platyops is not far removed from
the type found in the upper teeth of D. lepida or the upper and lower in D. nana, in both
of which the tips are bifid. During the development of the Phalangeride there has
been a development of sectorial premolars along at least four different lines, one of
these being indicated in the two genera Phalanger and Trichosurus, a second in the
Phascolarctine, a third in the ancestors of the Bettongiine, and a fourth in Dromicia
and the ancestors of the Macropodine and Potoroinz. It is interesting to note that if
one may judge from the somewhat similar bifid condition in the less specialized forms
of Phalanger and in Dromtcia, on the one hand, and the somewhat similar rotated and
grooved condition in the more specialized forms of Phalanger and in Hypsiprymnodon,
the first, third, and fourth lines are more closely related to one another than any one of
them is to the second.

MACROPODINZA.

In the case of the Potoroinz and Bettongiinz the dental evolution is of an extremely
limited range, and for the most part of an indecisive kind. In the present series, on the

OF THE AUSTRALIAN MARSUPIALIA. 151

other hand, it is much more comprehensive, the herbivorous adaptations being taken up
at a stage not far in advance of that represented by Caloprymnus, and rapidly brought to
a high state of perfection.

The extent of modification may be briefly summarized as follows:—(a) In the molars,
complete conversion of the anterior and posterior pairs of cusps into transverse crests,
the latter becoming successively elongated; development of longitudinal crests from
already existing longitudinal bands, or as new structures ; reversal of the relative pro-
portions of the teeth, by which they come to increase in size backwards; retardation of
the posterior teeth, with functional replacement from behind forwards. (0) In the upper
incisors, reduction of the length of the median upper teeth, by which they cease to project
beyond the lateral ones; compression of their edges; elongation of the cutting-edges of
the lateral teeth, frequently accompanied by reduplication. (¢) In the lower incisors,
perfection of the spatulate development, by which they come to cut against all of the
upper incisors, and in some cases against one another. (d) In the canines, rapid
obliteration proceeding from an already reduced condition. (e) In the median and
posterior premolars, successive elaboration of the teeth as sectorials.

As mentioned above, the dental evolution of the existing Macropodinz appears to have
proceeded along two different lines. One of them is represented as a series of typical
forms, whose evolution has been a terrestrial one throughout and has culminated in the
true Kangaroos; while the other is represented by an aberrant series, whose evolution,
at least in the case of two genera, has been effected by a return to arboreal life. In the
following descriptions these groups are respectively referred to as the hypsodont and
brachyodont series.

The Brachyodont Series. (Dendrolagus, Dorcopsis, Setonyx.)

The derivation of Dendrolagus and Dorcopsis is an interesting question, on account of
the fact that while both forms are more primitive in their general dental characters and
more specialized in their premolar characters than the true Kangaroos, Dorcopsis appears,
and has been commonly supposed from its terrestrial habit and its limb proportions, to
represent a connecting-link between the latter and the tree-living Dendrolagus. On
tracing the dental sequence, it is found that neither one of the two genera is nearly
related to the true Kangaroos, the resemblances shown by Dorcopsis to the latter being
the result of convergent development. The evolution appears to represent a Papuan
migration, the sequence of species being as follows:—Dendrolagus Lumholtzi and
D. Bennettianus (Queensland forms); D. inustus, D. Dorianus (both Papuan forms) ;
Dorcopsis Macleayi, D. luctuosa, D. Muelleri (all Papuan forms). The species of
Dorcopsis are derivatives of Dendrolagus Dorianus, or a closely allied form, which have
become terrestrial and have begun to assume characters similar to those of the true
Kangaroos.

The relations of Setonyx (= Macropus brachyurus, Quoy & Gaim.) are rather obscure.
It resembles Dorcopsis and Dendrolagus in the characters of its incisors and molars, and
also in the general characters of its sectorial premolars. In the relative size of its

152 DR. B. A. BENSLEY ON THE EVOLUTION

premolars it is, however, more primitive than Dendrolagus Lumholtzi, while in the
extent of reduction of its canines it is more advanced than the final members of
Dorcopsis. There is no evidence that the dental characters of this form are the result
of arboreal habit. Thomas (1888) estimates its systematic position to be with the members
of the small Wallaby section of the genus Macropus, and it probably represents an
aberrant form of this group which has assumed feeding-habits similar to those of the
arboreal forms.

Molars.—TVhe characters of these teeth are almost constant throughout the series. In
all forms the tooth-rows are either straight or but slightly bowed outwards in the middle.
The upper teeth increase gently in size from before backwards from the first to the third,
the fourth being slightly smaller on account of the meagre development of their posterior
lobes. The lower teeth increase in size from the first to the fourth. These proportions
are more primitive than those seen in the true Kangaroos, but represent an advance on
those of the Potoroine and Bettongiinz. All the teeth are functional at the same
time, the first teeth being only partially worn when the fourth are fully in place. Their
tips accordingly project to the same extent, and thus present a very different appearance
from that seen in the Wallabies and Kangaroos, where the functional rotation of the
teeth produces an arcuate arrangement of the tips. Both in the upper and lower teeth
(ef. Pl. 5. fig. 20; Pl. 6, fig. 26) the originally anterior and posterior pairs of cusps are
completely connected by transverse crests. In this character and in the relative
shortness of the crests they represent a transitional stage between the cuspidate
condition of the Potoroine and Bettongiine and the hypsilophodont condition of
the Wallabies and Kangaroos. Longitudinal crests formed by the modification of the
original longitudinal bands are indicated, and only to a moderate extent, in the median
valleys of the upper molars.

Incisors.— Although essentially macropodine rather than phalangerine in their general
characters, these teeth are noticeable for their comparative primitiveness. The median
upper incisors are rounded in section, but slightly flattened at their tips. They tend to
project slightly beyond the lateral teeth. The latter are comparatively small, and appear
to serve as much for grasping as for cutting organs. Their crown-surfaces are basin-like,
the rims of the basins being, however, incomplete externally. This modification recalls
that seen in the incisors of the Equide, the pseudo-invagination having, however, taken
place internally in the latter. The lower incisors are distinctly lanceolate and sharp-
edged, as in the members of the hypsodont series. In Se/onyx the incisors are more as
in the hypsodont series, both median and lateral upper teeth being distinctly flattened,
the lower lanceolate.

Canines—These teeth are present in a reduced condition in all the species of
Dendrolagus and Dorcopsis. They are in general more primitive as regards the extent
of reduction than in the hypsodont series; in the latter they are, as a rule, either wholly
vestigial orabsent. In Setonya, notwithstanding its small size and general primitiveness,
the canines have wholly disappeared.

OF THE AUSTRALIAN MARSUPIALIA. 153

Posterior Premolars.—As in the Bettongiinze and Potoroinz, these teeth furnish the
most decisive evidence concerning the sequence of the different species. Leaving out of
consideration for the present the genus Sefonyx, the most primitive conditions are found
in the two species Dendrolagus Lumholtzi and D. Bennettianus. In both forms the upper
teeth are scarcely as long as the first and second molars. Their length is about equal to
that of the upper incisor rows. They are comparatively broad, but the cutting-edge is
thin and emarginate. The edge ends anteriorly in a pyramidal column, probably repre-
senting the protocone. There is a prominent internal ledge, ending posteriorly in a well-
developed cusp like that seen in Caloprymnus. There is also a postero-external cusp.
In the Papuan species, D. Dorianus, we find a more advanced stage of sectorial evolution.
The teeth are of about the same size as in the preceding forms, and show the same
anterior pyramidal cusp, internal ledge, and postero-internal cusp. The postero-external
cusp is, however, absent ; and the internal ledge is present in the form of two rounded
protuberances. ‘The emarginate portion of the edge is furthermore thickened, and now
bears two distinct vertical ridges with corresponding grooves. In the remaining Papuan
form of Dendrolagus, D. inustus, we find a transitional stage between D. Dorianus and
D. Iumholtzi, the cutting-edge of the premolars being thin, the internal ledge not in the
form of protuberances, the ridges scarcely distinguishable, and the postero-external cusp
absent.

Passing to Dorcopsis, we find a more advanced stage of the same evolution. In
D. Macleay? the teeth are slightly longer than the first and second molars, and decidedly
longer than the upper incisor rows. The edge is uniformly thickened and not
emarginate, so that the differentiation of the anterior pyramidal cusp is not so marked.
The sides of the teeth are marked by four prominent ridges, with indications of a fifth,
the most anterior ridge corresponding to that formed by the anterior pyramidal cusp in
Dendrolagus. As before, there is a postero-internal cusp, but no postero-external one.
There is probably an internal ledge: this part of the teeth was found to be worn in the
single specimen examined. In JD. luctuosa the teeth have the same characters as in the
preceding species, except that they are still more elongated, being now equal in length to
the first and second molars together with the tip of the anterior crest of the third. There
are again four ridges, with indications of a fifth. The internal ledge is prominent and
tuberculate, as in D. Dorianus. In D. Muelleri we find the last stage of the evolution.
The teeth are exactly similar to those of D. ductwosa, except that they are again more
elongated, their length being now equal to that of the first and second molars together
with the whole anterior lobe of the third. The character of the internal ledge is
doubtful.

In Setonyx the outer premolars show the same stout proportions as those of Dendro-
lagus. Their length is about equal to that of the first and second molars combined. In
this character they approximate to Dendrolagus Lumholtzi and D. Bennettianus. There
are no traces of a postero-external cusp found in the latter forms. The cutting-edge is
not emarginate ; it bears three stout ridges, with indications of a fourth. There is a
well-developed postero-internal cusp and an internal ledge which does not appear to be
tuberculate.

154 DR. B. A. BENSLEY ON THE EVOLUTION

The Hypsodont Series. (Lagorchestes, Lagostrophus, Onychogale, Petrogale,
Macropus.)

The present group scarcely calls for generic consideration, since the characters by
which the genera are distinguished are for the most part others than those of the
dentition. Generic differentiation appears in this case to have resulted from the
assumption by the animals concerned of different kinds of environment while still
pursuing a grazing habit.

The main stages of progressive dental evolution concern the species of the predominant
genus Macropus. As recently pointed out by de Vis (1895), the three sections into which
the latter have been classified by Thomas (1888)—namely, the Small Wallabies, Large
Wallabies, and Kangaroos—represent successive phases of herbivorous evolution. It is
interesting to notice that, as in the Dasyuride, Phalangeride, and Potoroine already
referred to, the successive phases indicate increase in size of the animals as well as dental
specialization.

Progressive Characters of Molars.—Reference has already been made to the fact that
the assumption of an herbivorous habit involves an extensive wearing-down of the molar
crowns, rendering necessary a change from such a condition as is seen in the Phalangerine,
where the teeth are short-crowned, and being functional throughout life tend to decrease
in size backwards, because those first formed, namely the anterior ones, are longest in
use and may advantageously be largest. This change is now seen to include in its
entirety (a) reversal of the relative proportions of the teeth, by which they come
to increase in size backwards ; (0) hypsodont development of the crowns; (c¢) functional
replacement of the teeth from behind forwards, the anterior teeth being worn off and
shed while the posterior are coming into service.

As already pointed out, the tendency to reverse the proportions of the molars is
indicated, although to a very limited extent, in the Bettongiinz and Potoroine. In the
brachyodont division of the Macropodinew, where we meet with dental modifications
intermediate in extent of specialization between those of the foregoing smaller divisions
and the present hypsodont series, the upper teeth are seen to increase gently in size from
the first to the third, the fourth being smaller on account of the imperfect develop-
ment of its posterior lobe, while the lower teeth increase in size backwards throughout.
In the present series we find both the upper and lower teeth increasing markedly in size
from the first to the fourth. In respect of this character there is little progressive
variation, the modification being only carried to such a stage as to provide the adults
with a dentition as functional as that of the young. The forms which are highest
in other respects, namely those of the Kangaroo section of J/acropus, are at most only
slightly more advanced in this character than the smaller more primitive forms of the
same genus.

With regard to the functional replacement of the molars, it is seen that in the
brachyodont series this principle is scarcely indicated, the first molars being only slightly
worn and still very obviously functional when the fourth molars are fully in place; so

OF THE AUSTRALIAN MARSUPIALIA. 155

that, as in more primitive forms, all of the teeth are functional at the same time. In
the present series the first molars may be worn to the bases of their transverse ridges
before the fourth teeth come into place. This functional rotation is much more
conspicuous in the Kangaroos than in other forms, and has there been aptly compared
by Thomas (1888) to that found in the Proboscidea and some Sirenia. Correlated with
the functional rotation is a throwing-off of the worn-out anterior teeth. On examining
a series of individuals of different ages, it is seen that the worn median premolars and
milk-premolars are first thrown off by the eruption of the posterior premolars, and that
afterwards the latter, and finally the first molars, are thrown off; so that, of seven
cheek-teeth, only the last three to appear remain. In the Kangaroos that, portion of the
margin of the jaw which bears the cheek-teeth is raised above the anterior edentulous
portion ; and De Vis, in his interesting paper on the fossil Kangaroos, has pointed out
that the anterior teeth are in a way pushed over the edge of this platform, the posterior
premolars not being large enough in these forms to retain their own position and that of
the succeeding first molars. However this may be, the condition in the Kangaroos
represents a distinct adaptive advance on that in the Wallabies, where even in small
premolared forms the anterior teeth are retained.

Hypsodontism of the molar crowns is more pronounced in the Kangaroos (PI. 5. fig. 25
& Pl. 6, fig. 29, Macropus rufus) than in the smaller forms (Pl. 5. fig. 21 & PI. 6, fig. 27,
Lagorchestes), and is better indicated in all of the group than in Dendrolagus or Dorcopsis.
M. Eugenii furnishes a partial exception, the length of the molar crests in this form being
scarcely greater than in the last-named genera.

In respect to the molar patterns, the present group shows an advance on the brachyo‘ont
series in the formation of longitudinal crests supplementing the transverse ones. Mention
has already been made of the presence in the Phalangerine, the Bettongiine, and
Potoroinz of longitudinal bands connecting the protocone and hypocone in the upper
molars and the protoconid and hypoconid in the lower, and ending anteriorly and
posteriorly in both cases in thin ledges. In the brachyodont series these elements are
scarcely indicated; but in the present group they appear as longitudinal crests. One of
these crests connects the anterior and posterior transverse crests. It is placed on the
internal side in the upper mclars and on the external side in the lower. The upper
longitudinal element opposes the posterior transverse crest of a lower tooth, while the
lower longitudinal crest opposes the anterior transverse crest of an upper tooth. In the
lower teeth the longitudinal band is represented in front by a ridge extending from
the tip of the protoconid to the anterior ledge. Passing from such a type as Calo-
prymnus (P1. 6. fig. 25) to the Kangaroos (Pl. 6. fig. 28), we find this ridge assuming a
more internal position, its outer wall becoming pouched in such a way that a structure
resembling the antero-external shelf of polyprotodonts is formed. In the upper moiars
the anterior ledge has lost its connection with the protocone, but tends to be connected
externally with the paracone. This connection is obvious in all of the forms, with the
exception of the Large Wallabies and Kangaroos: in the former it is barely indicated,
while in the latter the anterior ledge is practically free externally. In the Kangaroos
an apparently new ridge is developed internally. It is well developed in all forms

SECOND SERIES.— ZOOLOGY, VOL. IX. 22

156 DR. B. A. BENSLEY ON THE EVOLUTION

excepting WZ. rufus and I. magnus, in the latter of which the anterior ledge has been
reduced. In certain fossil Kangaroos, as shown by the descriptions of Owen (1877) and
De Vis (1895), a greater degree of molar complication is found than is indicated in existing
forms. In Procoptodon accessory ridges are very numerous and conspicuous.

Incisors.—In the two lower subfamilies of the Macropodid the median upper incisors
tend to retain the original predominating size-relations and the grasping character found
in the smaller Phalangerinz, Caloprymnus alone (like the higher Phalangerinze, Phalanger
and Trichosurus) showing the opposite tendency to reduce the vertical length of these
teeth while flattening their tips in order to adapt them for a cutting function. The
members of the brachyodont series also tend to preserve the original relations, although
in a lesser degree. In the present group the modifications begun in Caloprymnus are
followed throughout, these teeth being always of a distinctly cutting nature. The
median lower teeth present the lanceolate cutting modification throughout. ‘The upper
lateral teeth show successive stages of sectorial elaboration ; but there is not a perfect
progression in this respect from the smaller towards the larger forms. Primitive
conditions are found in Lagorchestes, Lagostrophus, Petrogale, Onychogale, certain of the
Small Wallabies (JZ. Hugenii), and also in certain of the Large Wallabies (JZ. tr7ma and
M. Greyi) which appear to be derivatives of J. Hugenii. In these forms, especially
Onychogale, the cutting-edges of the upper lateral teeth are comparatively narrow. In
the Small Wallabies, with the above exception, the edges of the third upper teeth tend
to be elongated. A more primitive condition is found in JZ. Covent and MW. Billardieri
than in WZ. Thetidis, M. Wilcoxi, and M. stigmaticus. The same expansion is noticeable
in the Large Wallabies, except in the species referred to, and also in the Kangaroos,
Even the latter present differences of a somewhat similar kind to those in the lower
forms—J/. robustus and MM. giganteus, for example, showing a condition of greater
expansion than is found in I. rufus and MW. magnus. It is an interesting fact that
in this elaboration of cutting-edges, which is so important a feature of the grazing
evolution, the second upper incisors play a much less important 7/e than the third, their
edges being at most only slightly expanded. The explanation is probably to be found in
the greater freedom of the third teeth, there being no teeth situate behind them to oppose
their posterior elongation.

Reduplication of the edges of the upper lateral teeth, already referred to in the case of
the brachyodont group, is also indicated in the present series, although the tendency
is here in the direction of reduction. In such forms as Dendrolagus and Dorcopsis,
where no expansion of the edges is indicated, the tips of the teeth are basin-shaped ;
while in more specialized forms, such as those of the present series, where the upper
internal teeth cease to act as stops for the lower and become definite cutting-organs,
yeduplication is less apparent on account of the transverse compression. It is still
indicated, however, by a notching of the posterior or lateral sides of the teeth. In the
Kangaroos the notching is absent in the second incisors and barely indicated in
the third, while scarcely more primitive conditions are found in the Large Wallabies.
It is apparent that in the development of these forms there has been an increase
of muscular perfection admitting of a finer adjustment of the incisor teeth.

OF THE AUSTRALIAN MARSUPIALIA. 157

Sectorial Premolars.—The posterior upper teeth may be taken as typical of these
elements, which include the upper and lower median and lower posterior teeth. Their
modifications are of interest as throwing a side-light on the evolution of the series, since
they are not connected with the development of the grazing habit, but with the passage
of the shoot- or twig-eating habit. In their general character they preserve the orthal
position in the jaw, as in the Potoroinee. They are never so well developed as in the
typical shoot-eating forms of the brachyodont series. In Onychogale the teeth are
extremely small as compared with the first molars. They are pear-shaped in section, the
larger end being posterior. There are three cusps—two outer, and one inner and
posterior: the former appear to represent the fore and aft portions of the cutting-edge
of Caloprymnus, and the latter the postero-internal cusp. In Lagostrophus, Lagorchestes,
and Petrogale the posterior premolars are fairly well developed. In the first-named
genus and in Lagorchestes hirsutus the teeth are equal in length to the first molars, and
their cutting-edges bear three well-developed grooves; in Lagorchestes conspicillatus,
LL. leporoides, and Petrogale they are slightly longer than the first molars and bear four
grooves on their cutting-edges. Among the Small Wallabies, WZ. Wilcox, M. Thetidis,
M. stigmaticus, and M. Billardieri have the posterior premolars equalling in length the
first molars, with the tip of the anterior crest of the second, with two grooves on
the cutting-edges. In JZ. Hugenii and MW. Bedfordi the length of the posterior premolars
is almost equal to that of the first molars; the postero-external cusp tends to be
grooved. The whole condition is not far removed from that seen in Onychogale.
Among the Large Wallabies we find JZ. ruficollis, M. Parryi, and WM. Greyi showing a
similar approximation to Onychogale ; the posterior premolars are here shorter than the
first molars. In JZ. agilis, on the other hand, the teeth are well developed, their length
being equal to that of the first molars and the anterior crests of the second. In
the Kangaroos the posterior premolars are shorter than the first molars, and resemble
those of the small premolared lower forms. The explanation of these facts seems to be
that in the lower genera, including the Small Wallabies, there has been local elaboration
of sectorials in connection with a shoot-eating habit, grazing adaptations being incomplete
at that stage. Small-premolared and large-premolared Small Wallabies have apparently
given rise to similarly conditioned Large Wallabies. The relatively small size of the
posterior premolars in the Kangaroos appears at first sight to indicate that elaborated
sectorials have formerly been present in some of the lower forms and have been sub-
sequently reduced; but the more probable explanation is that these animals are the
descendants of successively small-premolared Small and Large Wallabies, which have
become more and more dependent on grazing habit and the development of grazing
adaptations.

Canines.—With the partial exception of Lagorchestes conspicillatus, none of the
members of this group show such well-developed canines as those of the brachyodont series
(excepting Setonyx), the Potoroine and Bettongiine, although even in the latter they
are greatly reduced. These teeth are present in a vestigial condition in Lagorchestles
hirsutus and L. leporoides, vestigial or absent in Onychogale, and typically absent in the

remaining forms (except ZL. conspicillatus).
9DO*

158 DR. B. A. BENSLEY ON THE EVOLUIION

PHASCOLOMYID.E.

The dental characters of Phascolomys, the single living representative of the family,
are of interest as indicating another line along which the herbivorous evolution begun
in the Phalangeridee has proceeded, and also as furnishing still another example of
convergent developments between Marsupials and Placentals, the general condition being
here very similar to that in the Rodentia.

The exact relations of Phascolomys with the phalangerid genera are not wholly
demonstrable as regards the dentition. The main evidence is (a) that the unworn molar
patterns present resemblance to those of the advanced bunodont Phalangerine ; (b) the
incisor modifications represent a more advanced stage than is found in any of the latter
forms or in the Diprotodontide ; (c) the moderately elaborated posterior premolars bear
a general resemblance to those of Dromicia on the one hand and, at least in the case of
the lower, to those of Nototherium and Diprotodon. The animal appears to represent a
specialized offshoot of the same line leading from the more primitive Phalangerine to
the Dipretcdontide.

Melars.—In the upper molars of the adult Phascolomys the sectional area is seen to
decrease in proceeding from before backwards, while in the lower it is more uniform.
As in the more specialized members of the Rodentia, the teeth are elongated and open-
rooted. Their bodies are curved, outwards in the case of the upper teeth and inwards in
that of the lower; so that the grinding-stress does not fall on the unsupported bases,
but on the side of an arch. Each tooth shows two triangular pillars; the sectional
apices are directed outwards in the lower teeth and inwards in the upper. In the
adult animal, the crown-surfaces being entirely worn down, the tips present the appear-
ance of two triangles joined together by their contiguous basal angles. In all of the
above characters, with the exception of the first, Phascolomys is not only highly
specialized, but also unique among existing Marsupials.

The only safe evidence concerning the molar derivation is afforded by the unworn
patterns. The latter are represented in Pl. 5. fig. 24 & Pl. 6. fig. 30. Both the figures
and accompanying descriptions are based on a young specimen in which only three molars .
above and below are formed, the first and second having barely pierced the gum, while
the third is still concealed in the alveolus. The upper molars bear a close resemblance
to those of Trichosurus (P1. 5, fig. 16). The two triangular pillars represent the anterior
and posterior lobes of the phalangerine tooth. Each of the pillars bears two large cusps ;
those of the anterior one correspond almost exactly in their characters and position to
the protocone and paracone of Trichosurus, while those of the posterior pillar similarly
correspond to the hypocone and metacone of that genus. As in Trichosurus, the cusps
are wholly bunoid, and the protocone and lypocone take part in the formation of a con-
spicuous internal band enclosing the internal anterior and posterior portions of the tooth.
‘The paracone and metacone are massive, and their internal sides show indications of the
development of transverse ridges. A peculiar feature is the presence of a large number

OF THE AUSTRALIAN MARSUPIALIA. 159

of small nodules arising from the sides of the main cusps. Apart from the obvious
quadrituberculate ground-pattern, the presence of these structures gives the teeth an almost
multituberculate appearance. This development appears to be unique among the
Marsupials.

The lower teeth also present a quadrituberculate pattern similar to that in Trichosurus
(Pl. 6. fig. 19). As in the latter, the protoconid and hypoconid take part in the formation
of an external band enclosing the external anterior and posterior portions of the teeth.
The entoconid and metaconid are massive. The development of accessory nodules is
again in evidence. The appearance of these elements is well illustrated in the internal
profile view of the second lower molar given in PI. 6. fig. 30 8.

A conspicuous feature of the unworn molars is the small sectional area of the crowns.
The crown-surface of the second upper molar of an adult Phascolomys may be twice as
long as in the young and more than twice as broad; and similar proportions are
observable in the remaining molars. In the young the upper teeth decrease in size
backwards from the second ; the lower teeth increase in size from the first to the third.

Incisors.—As in the Rodentia, with the exception of the Duplicidentata, the incisors
are limited to one on either side of the jaw above and below. They are greatly enlarged,
arched, and open-rooted. In the unworn condition (PI. 5. fig. 41) they are pointed at the
tips, and thus resemble those of the Phalangerinz ; while in the worn condition they
are scalpriform, as in the larger Rodentia. The scalpriform modification is, however,
incomplete, the reason being that the teeth meet at such a narrow angle (40°) that the
wear falls obliquely on their tips rather than on their posterior sides. The degree
of differentiation of the enamel bands appears to vary in different individuals; while
always thicker on the anterior sides of the teeth than anywhere else, the enamel sometimes
extends to their sides or even all the way round to their posterior surfaces,

Posterior Premolars.—With the exception of the incisors, these are the only members
of the antemolar series present in the genus. Like the molars they are elongated, curved,
and open-rooted, but unlike them are not definitely bilobed. In the unworn lower teeth
there are indications of three cusps, including a somewhat larger anterior element and
two smaller posterior ones; the latter are connected by a small transverse ridge. The
lower teeth bear a certain resemblance to those of Nototheriwm and Diprotodon. The
upper teeth are bifid at the tip, and resemble somewhat those of Dromicia.

DIPROTODONTID&.

The available evidence concerning the dental derivation of this family points to a
connection with the Phalangerine, in part through the Phascolomyidee., Nototherium
appears to represent one of the ancestral types leading towards Diprotodon.

Molars.—Apart from their comparatively huge size, the molars of Diprotodon resemble
closely those of the Macropodine in having the opposite cusps completely connected by

160 DR. B. A. BENSLEY ON THE EVOLUTION

transverse crests (cf. Pl. 5. fig. 25 & Pl. 6. fig. 31). If one may judge by analogy with
the latter family, these teeth are of quadrituberculate bunodont origin, as are also those
of the Phascolomyide. In both upper and lower teeth the transverse crests are greatly
elongated vertically, as in the Macropodine, but unlike the latter they are also curved.
The longitudinal crests characteristic of most of the Macropodine are not represented.
The upper molars show the anterior basal cingular ridge which is commonly developed
in the latter group. The lower molars show a posterior cingular ridge which is not
represented in the Macropodine.

The molar characters of the smaller form, Nototherium, are much as in Diprotodon..
There is a trace of a longitudinal band in the median valley. In this character
Nototherium is slightly more primitive, because, while longitudinal bands are highly
characteristic of the more specialized of the Macropodide, their actual presence is a
primitive phalangerine character, and their absence is secondary.

Incisors —The incisors of Diprotodon are in general intermediate in character between
those of the Phalangeridze and those of the Phascolomyidx. The median upper teeth
are open-rooted, arched, and scalpriform, as in the latter family. ‘They are even more
specialized as regards the differentiation of the enamel bands. The second and third are
retained, and serve as stops for the lower teeth, as in the Phalangeride ; but, unlike
the latter, they are also open-rooted.

Nototherium approaches Diprotodon in that the second and third upper teeth are
present and those of opposite sides are in contact in the middle line. The median upper
and lower teeth are of a more primitive type, neither being scalpriform. ‘The lower
especially are rounded in section and definitely pointed. It is interesting to note that
among the fossil forms referred to the Macropodidz we find two types of lower incisors,
the latter being usually of a spatulate cutting type, as in the existing forms of grazing
Macropodinz, but sometimes (Procoptodon) of a pointed piercing type, as in Woto-
therium. Considered alone, the lower incisors of Procoptodon would be readily taken as
belonging to a lower and smaller member of the Nototheriwm-Diprotodon line.

Premolars.—As in the Phascolomyid, only the posterior teeth are represented.
The following comparisons include only the lower ones. In Nototherium these teeth are
very much as in Phascolomys, except that they are not open-rooted. They are com-
paratively poorly developed; they show a main anterior cusp, and a posterior crest
formed by the connection of two cusps. In Diprotodon, also, these teeth are poorly
developed. When slightly worn they present anterior and posterior crests, the former
appearing imperfect. The posterior crest obviously corresponds to that in Mototherium,
although better developed than in that form. ‘The anterior crest cannot be regarded as
having arisen from the connection of two opposite cusps, as in the molars, because no
case occurs in the herbivorous Marsupials where these are found in posterior premolars,
the only conclusion being, therefore, that it is the result of compression of the anterior
cusp found in the smaller forms, Nototherium and Phascolomys.

It will be seen that, although the dental characters of both families are highly modified

OF THE AUSTRALIAN MARSUPIALIA. 161

‘away from what would be considered a primitive type in comparison with the Pha-
langeridze, and also from one another, the Phascolomyide and Diprotodontidze appear to
be allied in several features, pamely : (a) the (presumably) quadrituberculate bunodont
origin of their molars ; (%) the rodent modifications of the incisors; (¢) the similarity in
size and arrangement of the cusps of the lower posterior premolars; (d) the obliteration
of the anterior and median premolars. To these may be added: (e) reduction of the
deciduous premolars.

It is a question not wholly answerable, however, whether or not some of these
characters may be the result of convergent development. Phascolarctus presents points
of general resemblance with Phascolomys, and as regards its dentition shows three of the
characters above referred to, namely, 0, d, and e. There is good reason for believing all
of the latter to be the result of convergent development. Phascolarctus is isolated by
its advanced selenoid molar patterns; and while there is overwhelming evidence, both in
Marsupials and Placentals, of the conversion of the bunodont type of molar into a
lophodont one, no case is known of the conversion of the selenodont type into a similar
lophodont one. As already pointed out, although the molars of Phascolarctus and the
other Phascolarctine have retained, with but little modifications, certain of the cusp-
characters of the polyprotodont forms, which are alike ancestral to them and to the
Phalangerinze, they are much too advanced in a special selenodont direction to have
given rise to those of the last-named division. Furthermore, Phascolarctus occupies a
derived relation to Pseudochirus, a genus lacking characters 6 and d. Finally, the
smaller members of the Phalangerine, such as Acrobates, are far more primitive in their
general dental characters than the Phascolarctinee. We have, however, no such proofs
of convergence between the Diprotodontidee and Phascolomyidz as we have between both
families and Phascolarctus; so that if it is permissible to generalize on the evidence
available, we may assume that their common features are characters of affinity—on the
whole, that they have diverged from common ancestors possessing a reduced premolar
formula, a reduced tocth-change, bunodont quadrituberculate molar patterns, and
incipiently rodent incisors. As mentioned below, an analogous case of a combination of
fundamental characters of resemblance with special characters of divergence is observable
in the foot-structure of the two families.

THYLACOLEONTIDA.

The recent discussion by Broom (1898) of the controversial question of the habits of
the extinct form Thylacoleo appears to allow of no further doubt as to its predaceous
carnivorous character. ‘The question now arises as to how the relationships of the animal
with the Phalangeridze are to be explained. Throughout the present paper the effort has
been made to show that the Australian radiation began with insectivorous prototypes and
proceeded along two primary lines, one of them carnivorous, the other omnivorous and
finally herbivorous. In the second line all the advanced forms are diprotodont, and all
the typical terminal forms are highly specialized herbivora. Thylacoleo is clearly a

162 DR. B. A. BENSLEY ON THE EVOLUTION

member of the second line, which, in all probability after the cmnivorous stage, has
become carnivorous instead of herbivorous. In view of the general sequence in the
Australian group, this development is to be looked upon as an aberrant one. Analogous
cases of perverted habit and dental adaptation are to be found in the Placentalia, a con-
spicuous case being afforded by the fish-eating rodent Ichthyomys (Thomas, 1893) of South
America, in which the herbivorous principle of the Rodentia has not been adhered to.

The presence of enlarged posterior premolars in Thylacoleo points to a former condition
of at least incipient herbivority. The excessive enlargement and ungrooved character of
the teeth represent more special carnivorous adaptations, repeated in the placental
Carnivora, but without parallel in marsupial carnivores.

The reduced character of the upper canines possesses a similar significance. In the
smaller members of the Dasyuridee the canines are moderately developed, and following
the carnivorous evolution they show a successive increase in functional importance. In the
Phalangeride we find the smaller members presenting well-developed upper canines, but
following the omnivorous-herbivorous evolution there is a successive decrease in functional
importance, and in derived families total obliteration. Thylacoleo is quite as advanced in
the reduction of the canines as Phalanger or Trichosurus.

In Thylacoleo the piercing elements of the dentition are the median upper and lower
incisors. It is interesting to note that in the Dasyuride a piercing modification of the
upper incisors appears as a prototypal insectivorous character, and that in the carnivorous
evolution the piercing function is taken over by the canines. In the more primitive
Phalangerid we find both upper and lower median incisors modified as piercing-organs ;
while in more advanced forms the lower teeth assume a cutting character, and the upper
become slightly reduced. It is difficult to say whether the incisors of Thylacoleo were
formerly modified as cutting elements; but it is apparent that the animal must have
developed along normal phalangerine lines to such an extent that only the median
incisors were left as possible piercing-organs.

The reduction of the molars, although repeated in the most specialized of the placental
Carnivora, contrasts strongly with the sectorial elaboration met with in the typically
carnivorous Dasyuridze, and has probably resulted from the presence in the ancestral
form of inadaptive quadrituberculate bunodoxt patterns as much as from the sectorial
elaboration of the posterior premolars.

THE ADAPTIVE MODIFICATIONS OF THE FOOT-STRUCTURE IN THE AUSTRALIAN
MARSUPIALS.

On comparing the general types of foot-structure in the Placentalia it is seen that
certain forms, more especially some of the Creodonta, present an extremely close
approximation to the ideal pentadactyl type, which we assume to represent the starting-
point for the foot-evolution of the Mammalia generally, and, further, that the great
majority of the remaining forms present advancec modifications of this type bearing a
decidedly terrestrial stamp. The predominance of terrestrial modifications and the

OF THE AUSTRALIAN MARSUPIALIA. 163

presence of prototypal conditions in such a central group as the Creodonta indicate
that the evolution of the Placentalia is typically a terrestrial one.

In the Marsupials the reverse appears to be the case. As already pointed out by
Huxley in 1880, none of the members of this group present an ideal pentadactyl pes,
the hallux, whenever present, showing more or less extensive indications of opposability.
And, as recently further shown by Dollo (1899), all of the primary modifications of the
pes found in the group are in the direction of prehensilism, while all other modifications,
whether indicative of terrestrial or aquatic (Chironectes) adaptation, are derivatives
Unlike that of the Placentals, therefore, the evolution of the
Marsupials is primarily arboreal, and secondarily terrestrial.

The following summary, representing to a considerable extent an adaptation of Dollo’s
views to the present case, is designed to show the general sequence of arboreal and
terrestrial phases in the various marsupial families :—

of arboreal phases.

slow terrestrial
with
podials and more or less radiating digits.

a, Primary terrestrial phase .
(Unidentified Marsupio- Placeual acy )

Pes adapted for progression :

plantigrade, pentadactyl, short meta-
}. Primary arboreal phases.
Hallux

bo’. First arboreal phase .
(Didelphyide in part.)
6’. Second arboreal phase . a N
(Some Didelphyide, pHnatsenae except
Tarsipes.)
6°. Third arboreal phase
(Tarsipedinz.)
c. Secondary terrestrial phases.
c’. First terrestrial phase
(Dasyuride in part.)
. Second terrestrial phase

(Notoryctidz, Phascolomyide, mitetedon.

tide, Peramelide, Macropodide.)

Pes prehensile for arboreal progression.
opposable.

Characters of 6'. 4th digit elongated ; 2nd and
8rd digits reduced and syndactylous.

Characters of 4?. Claws of 4th and 5th digits con-
verted into nails,

Derivative of b'. Pes elongated; hallux reduced

or absent. Plantigrade or subdigitigrade.

Derivative of 6°, Pes variable. Plantigrade or
digitigrade ; pentadactyl, tetradactyl, or func-

tionally monodactyl.

DASYURID.

On comparing the distribution of arboreal and terrestrial phases in the Australian
Marsupials, we find that they are, as a rule, so completely differentiated as to be

distinctive of family divisions.

The Dasyuridze, however, present a notable exception,

there being evidence of a main line of arboreal or semi-arboreal evolution, with several

subsidiary lines of terrestrial adaptation.
is here illustrated in a single family.

The general principle of the whole radiation

The main line is represented by the numerous species of Phascogale, together with

the two species D. hallucatus and D. maculatus of Dasyurus.

Its identification does

not depend on the presence of progressive changes, but on the retention of primitive

characters in successively larger animals.

SECOND SERIES.— ZOOLOGY, VOL. IX.

The subsidiary lines are constituted as
23

164 DR. B. A. BENSLEY ON THE EVOLUTION

follows :—The first by the species of Sminthopsis and by Antechinomys laniger ; a second
by Chetocercus cristicauda and Dasyuroides Byrnei; a third by Dasyurus Geoffroyt
and D. viverrinus; and a fourth by Sarcophilus ursinus. The species described by
Spencer (1896) as Phascogale macdonnellensis may represent another terrestrial line, but
the animal is doubtfully separable from Sminthopsis. The genera Myrmecobius and
Thylacinus represent two more terrestrial lines whose exact connections are doubtful.

The above reference to the main line of the evolution as arboreal or semi-arboreal
calls for some explanation. It will be seen from a general survey of the family that
even in the most primitive species the hallux, although of an opposable type, is much
smaller than in typical arboreal forms, such as those of the Phalangeridze and
Didelphyidee, and the digits present a parallel rather than a primitive radiate arrange-
ment. Thomas (1888) remarks that the species of Phascogale are strictly arboreal, while
Winge (1893) states that the foot of Phascogale is more modified for terrestrial service
than in the Didelphyide.

The fact that the forms here referred to as constituting successive lines of terrestrial
evolution present definite terrestrial characters, shows that modifications presented by
Phascogale must be explained on some other assumption than that they represent
terrestrial adaptations.

It is a well-known fact that the Didelphyide, which approximate very closely to
the present family, present typical prehensile developments both in the pes and in
the tail. In the Dasyuride these prehensile developments are absent, if we except the
presence in some of striated plantar pads and the opposability of the reduced hallux,
The conditions in the Dasyuridz thus appear at first sight to indicate that the family
has been derived originally from forms which had already embarked on a terrestrial
evolution. This proposition may, however, be dismissed with the simple statement, since —
in another section of the Australian fauna, namely the Phalangeride, we meet with
hallucal characters of the same primitive arboreal stamp as those of the Didelphyide.
The alternative explanation is that the Dasyuridze have abandoned modifications of
prehensilism for others better suited for either arboreal or terrestrial progression in
connection with their insectivorous habits. Comparison with the Placentals shows that
arboreal habit does not necessarily demand the adoption of prehensilism, although it is
indicated to a marked degree in some forms (Primates), as it is in the Marsupials. The
modifications of the pes in the primitive Dasyuride indicate a cursorial development
involving elongation of the pes and parallel arrangement of the digits as well as
recession of the hallux. The ancestral forms of the family appear to have abandoned
their prehensile modifications for semi-cursorial ones of undoubted advantage in
insectivorous life, the pes having thus been made serviceable for rapid progression
either in the trees or on the ground. At the same time they have lost nothing in the
abandonment of modifications of prehensilism, these being unnecessary in animals
which are capable of balancing their bodies in arboreal progression by means of
rapid locomotion. Those of their successors which have become definitely terrestrial
are not necessarily more completely cursorial, although their special characters are
largely the result of a substitution of digitigrade for plantigrade progression.

OF THE AUSTRALIAN MARSUPIALIA. 165

For descriptive purposes the foot of Phascogale flavipes may be taken as representing
the type from which the modifications of the remaining forms have been derived. In
this animal the foot is relatively short and broad. In the specimen here figured (PI. 7.
fiz. 2) the ratio of the breadth, measured directly behind the hallux, to the length,
measured from the tip of the hallux to that of the third digit, excluding the claw, was
found to be in the proportion of 1:3°5. The hallux is greatly reduced as compared
with that of the Didelphyide and Phalangeridze. As in the latter, however, it is set at
a considerable angle to the remaining digits, and is, further, clawless and provided with
a slightly swollen terminal pad. The remaining digits are approximately equal in size
and provided with strong curved claws. Their most conspicuous feature is their parallel
arrangement and consequent restriction of lateral motion. The plantar surface is naked
and granulated, and bears five prominent pads, all of which are elongated and _ trans-
versely striated. One of these is placed at the base of the hallux, and extends from the
notch separating that digit from the second backwards towards the heel; its posterior
two-thirds is separated from its anterior third by a sharp constriction. ‘Three other pads
are situated at the bases of the outer digits; they are completely separated from one
another, and the middle one is placed somewhat in advance of the others. A fifth pad
is placed on the outer margin of the sole opposite the middle of the hallucal pad. It is
probable that even in arboreal progression these structures are not so functional as in
those arboreal forms in which the hallux is well developed and the outer digits are
radially arranged. The original function of the hallucal pad is to oppose the outer
digital pads anteriorly and the marginal pad posteriorly, just as the hallux opposes the
outer digits.

In the remaining species of Phascogale and in Dasyurus hallucatus we find much the
same condition as in P. flavipes, such differences as do exist, apart from the progressive
increase in size, being of a minor character. In P. Swainsoni and its Tasmanian repre-
sentative P. minima the pads are shorter and more oval than in P. flavipes. The
proportion of breadth to length in P. minima is approximately 1:3. In P. dorsalis
these characters are repeated, except that the constriction seen in the hallucal pad is
here replaced by a distinct gap. The proportion of breadth to length is 1: 3:1. In
P. Wallacei the proportion is.1: 3:8, and in P. Thorbeckiana 1:4. In the latter the
hallucal pad is undivided. PP. apicalis, which was seen to be aberrant in its premolar
relations as compared with its size, is also aberrant in its foot-structure, the heel being
elongated, so that the ratio of breadth to length becomes 1: 4-2. The hallucal pad is
undivided, and the marginal pad is very thin posteriorly. In P. calura and its larger
representative P. penicillata the pads are elongated as in P. flavipes; the proportions
are respectively 1: 5 and 1: 3°75.

The form described by Spencer as Phascogale macdonnellensis shows a_ partial
departure in foot-structure from the normal members of that genus. The foot is
relatively short, the proportion of breadth to length being 1 : 3°26. It presents a peculiar
swollen appearance. The hallux is reduced to scarcely more than a tubercle. All of the
pads are comparatively well developed, and their surfaces are transversely striated as in

Phascogale. The hallucal pad is here subdivided. According to Spencer’s account, the
25°

166 DR. B. A. BENSLEY ON THE EVOLUTION

habits of the animal are terrestrial. Its foot-characters appear to represent an inter-
mediate stage between the condition in Phascogale and that in the most primitive form
(S. leucopus) of Sminthopsis, with, however, a substitution of a plantigrade for the
saltatorial modification which distinguishes the latter genus.

Passing from Phascogale to Sminthopsis and Antechinomys, we meet with a com-
paratively well-graded series of minute forms which show a terrestrial and saltatorial
evolution. The most primitive condition is found in Siminthopsis leucopus (Pl. 7. fig. 3),
the foot-structure of which approximates very closely to the type presented by
Phascogale flavipes. The foot is relatively longer than in P. flavipes, the proportion
being 1 : 5:5. The hallux is slightly more reduced and its terminal pad scarcely
swollen. ‘The plantar surface is not wholly naked as before, the marginal hairs of both
sides of the foot showing a tendency to encroach on its posterior portion. ‘The heel
is almost completely hairy. The hallucal pad is scarcely distinguishable from the
surrounding granular surface; its striated area is minute and oval. The digital pads
are as well marked as in P. flavipes, but are much more rounded and less distinetly
separated. The striated surface does not extend over the whole pad, but is confined to a
small area at the apex, the granulated surface having encroached on the base. The
marginal pad is practically absent, its position being only indicated by certain larger
granules.

Sminthopsis larapinta while agreeing with S. /ewcopus in the proportions of the foot
(1: 5:1) presents a more advanced stage in respect to the characters of the pads. The
hallucal pad is practically absent, being only represented by a row of three large smooth
granules. The digital pads are closely set together and their bases are fused. ‘he
basal granules of each have now encroached to such an extent that the formerly well-
developed and striated surfaces have become much reduced. ‘The marginal pad is
absent ; its former presence is indicated by one or two larger granules.

Sminthopsis murina shows a variation or side development of the condition seen in
S. leucopus. The proportion of breadth to length is 1: 5. There is a small elevation at
the base of the hallux, with a larger granule anteriorly. The digital pads are well
developed and separate, except at their bases; but their surfaces are completely
eranular. There is in each case a larger granule representing a formerly striated area,
‘The marginal pad is absent.

In Sminthopsis crassicaudata (Pl. 7. fig. 4) we find a more advanced stage than that
in S. leucopus and 8. larapinta. The ratio of breadth to length is 1:75. Except fora
slightly enlarged granule at the base of the hallux, the hallucal pad is unrepresented.
The digital pads are still more completely fused together basally. ‘Their surfaces are
entirely granular, but the apex of each is occupied by a large differentiated granule,
which obviously represents the remains of a formerly striated surface, as in S. murina.
The marginal pad is wholly absent.

OF THE AUSTRALIAN MARSUPIALIA. 167

In Sminthopsis hirtipes (Pl. 7. fig. 5) we find what is apparently another modification
of the condition in S. larapinta. The proportion of breadth to length is 1 : 6:3, the foot
being relatively shorter than in the preceding species. The hallux is more reduced than
in S. larapinta. 'The hallucal pad is represented by a small protuberance. The digital
pads are more specialized than those of the preceding species, being so completely fused
together that their former distinctness might be questioned were it not for the fact that
the compound structure shows a trilobate contour.

In Antechinomys laniger (PI. 7. fig. 6), which is the final member of the series, we
find a further advance on S. crassicaudata. The proportion of breadth to length is
1:10. The hallux is absent, us are also the hallucal and marginal pads. At the bases
of the outer digits there is a single large pad representing the now completely fused
pads of the more primitive forms. The whole surface is covered with granules, which
increase in size towards the apex. The posterior portion of the sole is completely hairy
as far forwards as the base of the digital pad.

A second line of terrestrial evolution, with relations with the larger instead of the
smaller species of Phascogale, is represented by Chetocercus cristicauda and Dasyuroides
Byrnei. These forms repeat, on a larger scale, the special modifications found in
Sminthopsis. In Chetocercus (Pl. 7. fig. 8) the proportion of breadth to length is
1 : 4°53, much the same as in Phascogale. The hallux is more minute than in the latter
genus. ‘The hallucal pad is represented only by a small protuberance, bearing granules
which are slightly larger than those of the surrounding surface. The digital pads are
well developed, but are laterally compressed and fused at their bases: their surfaces
are entirely granular; but, as in S. crassicaudata, there is an enlarged apical granule
representing a formerly striated area. ‘The middle line of the plantar surface is hairy
for about 10 mm. posteriorly.

Dasyuroides Byrnei (Pl. 7. fig. 9) shows an advance on the preceding species. ‘The
proportion is 1:75. The hallux is absent, as are also the marginal and hallucal pads.
The digital pads are well developed and fused basally. As in Cheéocercus, their surfaces
are granular, and a larger apical granule is present ; the latter shows signs of transverse
striation. The sole is hairy for 10 mm. posteriorly, as in the latter species, and the
marginal hairs of both sides of the foot, as far forwards as the pads, tend to encroach
on its granular surface.

Within the genus Dasyuwrus we find both arboreal and terrestrial phases. In
Dasyurus hallucatus (P\. 7. fig. 10), the smallest and most primitive of four species *,
we find an arboreal foot type which is scarcely distinguishable from that of Phascogale,
The proportion of breadth to length is 1:37. All of the pads are well developed and

* The collection contains no examples of the Papuan form, D. albopunctatus,

168 DR. B. A. BENSLEY ON THE EVOLUTION

transversely striated. The sole is completely naked. The hallucal pad is subdivided
and its posterior portion moderately elongated.

Dasyurus maculatus is represented in the collection only by dried skins. So far as
can be judged from these, the pes practically repeats the arboreal type of D. hallucatus.
The plantar pads are well developed and transversely striated. The proportion of
breadth to length is 1:4. The hallux appears to be relatively more reduced than in
D. hallucatus.

The terrestrial forms Dasyurus Geoffroyi and viverrinus, on the other hand, show
resemblances to Chetocercus and Dasyuroides. In the former the proportion of breadth
to length is 1:4°7. The hallux is relatively smaller than in D. hallucatus. The digital
pads are partially fused together, and their surfaces are completely granular. The
hallucal and marginal pads are entirely absent. The posterior portion of the sole is
hairy, as in Chetocercus. In Dasyurus viverrinus (Pl. 7. fig. 11) we find a direct
advance on the terrestrial type of D. Geoffroyi. The pes is more elongated, the
proportion being 1:6. The marginal fur of the foct tends to encroach to a greater
extent on the sole. The digital pads are still more definitely fused, and the hallux has
disappeared.

In all of the terrestrial forms above described, with the exception of Phascogale
macdonnellensis, the tendency is towards digitigrade modifications. In Sarcophilus
(Pl. 7. fig. 12) plantigrade conditions have been retained, the foot being relatively short,
as in Phascogale, the proportion of breadth to length being in the neighbourhood
of 1:35. The hallux is absent, as in other advanced terrestrial forms. ‘There
are no plantar pads, the sole being uniformly covered with a papillated skin. The
whole condition is unique in the family, but is derivable from an arboreal Dasyurus
type.

In three of the four terrestrial lines here recognized we have animals of respectively
small, medium, and large size showing the same type of foot-modification. This
appears, at first sight, to indicate genetic sequence, but such a view is rendered
improbable by the fact that the evolution of the family is primarily an arboreal one, and
that in the arboreal series we have animals of successively larger size. The terrestrial
lines have arisen in succession with the development of larger and larger species in the
arboreal series. It, furthermore, seems impossible to regard the generic characters of
Dasyurus as the result of convergent development in two sets of species, one of
arboreal, the other of terrestrial derivation, as we should be obliged to assume in
adopting the former view.

The foot-structure of the terrestrial form Myrmecobius is, in many respects, unique,
and there is little in the way of direct evidence concerning its relations with the
Phascogale type. The proportion of length to breadth is 1 : 5:5, the foot being thus
slightly elongated. The sole is naked anteriorly and provided with a smooth tough
skin; its posterior portion is hairy for about 12 mm. forwards. ‘There is no external
hallux. An oval elevation in the region of the tip of the reduced hallucal metatarsal bone

OF THE AUSTRALIAN MARSUPIALIA. 169

appears to represent the hallucal pad of Phascogale. Digital pads are present as in the
latter genus; they are well developed and separate. Each pad has a punctate area at
its tip, probably representing a formerly striated surface. The first digital pad tends to
be approximated to the second, as in some syndactylous forms.

The derivation of the type met with in TVhylacinus is quite as obscure. The
modification is a digitigrade one, as in the smaller terrestrial forms. The foot is not
greatly elongated, the proportion being in the neighbourhood of 1:4. The hallux is
absent, as in other advanced terrestrial forms. A single large pad, placed at the bases
of the digits, shows signs of trilobate structure, and may have originated from the fusion
of three separate pads, as in smaller forms. It is difficult either to demonstrate a
connection of this type with those represented by other members of the family, or even
to prove that it is of arboreal derivation.

PHALANGERID.

In this family the same general principle is exemplified as in the arboreal line of the
Dasyuridee. We have a series of animals showing increase in size of body and _pro-
gressive characters in other respects, but with foot-patterns of much the same type
throughout. The modification is an arboreal one, and in some respects more primitive,
in others more specialized than in the primitive Dasyuridz. Just as in the case of the
dentition, the exact prototypal characters are not found associated in any single form,
but are distributed over two families—in this case, the present one and the Dasyuride.

The pes of Dromicia (Pl. 7. fig. 18, D. nana) may be taken as representing the
prototypal condition not only for the Phalangeride, but for all of the Australian families
with the exception of the Dasyuride. The general characters are those of the second
arboreal phase (¢f. p. 163). The primitive characters, as compared with the Dasyurid:e,
are: (a) the short broad proportions of the foot and the radiating character of the digits ;
(b) the unreduced character and wide opposability of the hallux; (¢) the more typical!
development of the five striated plantar pads and of the terminal pads of the digits.
The more advanced characters are: (a) the great development of the fourth toe *; (0) the
reduction and syndactylism of the second and third toes.

In the remaining members of the family, excepting Tarsipes, the modifications of this
type are few and insignificant. Acrobates pygmeus presents the same conditions as
Dromicia nana, except that there is an accessory pad on the postero-external side of the
hallucal pad and another smaller one on the outer side of the third digital pad. arsipez,

* With reference to the recognition of the enlarged nature of the fourth toe as an arboreal adaptation (Winge,
Dollo), it is interesting to note that a different explanation was given by Owen (1879). While acknowledging the
fact that there has been a successive enlargement of this member in the Macropodide, Owen considered its original
predominance to be a reptilian character. A somewhat similar opinion has been expressed by Leche (1891). This
writer remarks as follows :—‘ Bis auf Weiteres neige ich zu der Ansicht hin, dass die Prevalenz der +. Zehe bei
Marsupialia ebenso wie diejenige der 3. bei Ungulata unabhangig von der besondern Function durch Vererbung
erworben ist.”

170 DR. B. A. BENSLEY ON THE EVOLUTION

as pointed out by Dollo (1899), presents a more extreme type (third arboreal phase, p. 163),
the second and third digits being greatly reduced, and almost completly enclosed in a
common integument; while the fourth digit is greatly enlarged, and like the fifth is
provided with a nail instead of a claw, as in the placental Primates. In addition to these
obviously progressive characters, Tarsipes presents certain other peculiarities. The
terminal pad of the hallux is not swollen to the same extent as is usual in arboreal forms.
The hallucal plantar pad is very short. The second digital plantar pad is subdivided and
its outer portion has migrated on to the base of the fourth digit.

The larger forms Petawrus (breviceps) and Dactylopsila almost repeat the characters
of Dromicia. In the former there is an accessory hallucal pad as in Acrohates.
Dactylopsila differs only in that the claws are very powerful and the digits appear to be
permanently bent at the joints of the first and second phalanges.

Among the species of Phalanger we find in P. orientalis and P. lullule a slight
deviation from the general type. Except for a small area on the inner side of the
foot between the hallucal and the first digital pad, and another between the marginal
and the outer digital pad, the whole plantar surface is occupied by striations. This
condition is obviously the result of a fusion of the originally separate pads. ‘The
hallucal pad is very broad, and its striations, which run in an oblique direction, join
posteriorly with those of the marginal pad. The striations of the digital pads cross from
one to the other, and, in the case of the second and third, they also extend backwards
on the sole to meet those of the hallucal and marginal pads. The hallux is set rather
further back than in Dromicia, so that its opposability is more perfect.

In Zrichosurus the extension of the plantar striations is not indicated. The hallucal
pad is well developed, but is not sharply differentiated from the sole on the outer side.
Its striations are very short. The digital pads are oval and well separated from one
another. The marginal pad shows much the same characters as the hallucal one. At
first sight the non-extension of the plantar pads appears to represent a distinction
between this form and Phalanger; but in P.celebensis we find much the same condition
as in Trichosurus, showing that the condition in the above-described species of Phalanger
is a special one.

In the structure of the pes there is no such difference between the phascolarctine and
phalangerine groups as is observable in dentition. The foot of Psewdochirus (Forbesi)
conforms to the type in Dromicia. The first and second digital pads are closely set
together and have their striations concentrically arranged. The conditions in Phas-
colarctus give no indication of special relationship with Pseudochirus. The general
conformation of the digits is the same, but the plantar pads have all been reduced, the
sole being uniformly covered with a soft granulated skin somewhat as in Sarcophilus
and Phascolomys. The hallux is, however, set back to a much greater extent than
in Pseudochirus, so that it is more perfectly opposable. With only a slightly greater
displacement, its axis would be in the same line as that of the fourth digit which it
opposes.

OF THE AUSTRALIAN MARSUPIALIA. 171

NOTORYOTID.

Apart from the original description by Stirling (1891), the chief observations on the
foot-structure of Notoryctes are those of Gadow (1892), Winge (1893), and Dollo (1899).
Of the last-named writers, Gadow and Dollo have made comparisons with other
Marsupials with conflicting results, the former finding resemblances to the Didelphyidee
and Dasyuridee, the latter to members of the syndactylous series, more especially the
Peramelide. The following remarks are designed to show that the balance of evidence
is in favour of the latter view.

It may be observed at the outset that the pes of Notoryctes is so highly specialized
that the exact manner of derivation cannot be affirmed with certainty.

The plantar surface shows no indications of arboreal pads, being covered by a leathery
wrinkled skin. The hallux is well developed and possesses the full number of phalanges
as in the Phalangeridee and Didelphyide. Its terminal phalanx is provided with a
flattened claw. In this respect Notoryctes is unique among the Marsupialia. Winge,
who follows Gadow’s view as to the dasyurid affinity of the animal, suggests the
possibility that the hallucal claw is of secondary origin; and this explanation is in all
probability the correct one—Notoryctes being highly specialized in other respects. The
entocuneiform bone is comparatively long. As pointed out by Owen, and by Dollo for
the present form, this character represents an arboreal adaptation. Gadow mentions
the presence of a prehallucal element attached to the entocuneiform as indicative of
affinity with Didelphys, in which it is also present. An examination of other genera
of Didelphyidze shows that the element is of fairly general occurrence in this family.
It is absent in the adults of all the Australian forms, but Emery has described it in
embryos of certain species.

The hallucal articular facet of the entocuneiform is figured by Stirling as terminal
(cf. text-fig. 6, B, p.172). In an exhibition skeleton in the B.M. Collection (text-fig. 6, C)
the terminal portion of the entocunciform is wedge-shaped, the one side of the wedge being
applied to the proximal portion of the second metatarsal, while the other bears the
articular facet. The hallux thus occupies a position of partial opposability. This
condition is more apparent in one foot than in the other. A somewhat similar tendency
towards variation is found under the same conditions in Phascolomys. In Notoryctes
a formerly opposable hallux is apparently returning to an orthal position while remaining
functional, while in Phascolomys a formerly opposable hallux retains a more or less
opposable position, but is becoming vestigial.

Gadow regards the second and third digits of Notoryctes as free as in “ Didelphyide,
Dasyurus, Thylacinus, Phascogale, Myrmecobius, Phascolomys. ... . There is at the utmost
a very slight indication of syndactylism of the second and third toes, far less obvious
than it is even in Phascolomys.” Dollo, on the other hand, recognizes a slight reduction
and syndactylism of these digits, as in the arboreal Phalangeride. A spirit-specimen
(¢f. Pl. 7, fig. 19) examined by the writer shows very definite indications of syn-
dactylism. It is true that the condition is not so marked as in normal forms, because

SECOND SERIES.—ZOOLOGY, VOL. IX. 24

172 DR. B. A. BENSLEY ON THE EVOLUTION

in Notoryctes all the digits are much less distinctly separated from one another. The
same condition is observable in Phascolomys, also a terrestrial plantigrade type. As
shown below, there is not the slightest doubt as to the derivation of the latter from a
normal syndactylous type such as is represented by the arboreal Phalangeride. As
regards the size of the second and third digits in the skeleton of Notoryctes, they are not
distinctly more slender than the others, especially the fourth, as in phalangerine forms.
The same condition is again observable in some specimens of Phascolomys, although in
others they are of the slender phalangerine type. Notwithstanding their uniformity in
size, however, the relations between second and third digits in Notoryctes are plainly
with one another rather than with those on either side of them. ‘They show a very

Dorsal views of phalangerine pes and terrestrial plantigrade modifications.

A. Trichosurus (after Flower) ; B. Notoryctes (after Stirling) ; C. Votoryctes (B.M. specimen); D. Phascolomys (after
Owen); E. Diprotodon (from Dollo, after Stirling and Zietz). Abbreviations: a., astragalus ; c., caleaneum ;
m., nayicular ; en., entocuneiform ; m., mesocuneiform ; ¢c., ectocuneiform ; cb., cuboid.

close correspondence in the general size and length of the metatarsals and phalanges,
which is strongly suggestive of former syndactylism.

Dollo points out a predominance of the fourth digit as a further indication of affinity
with the syndactylous series. In Stirling’s figure (text-fig. 6, B), here reproduced, the
fourth digit is seen to project a considerable distance beyond the third. In the spirit-
specimen above referred to (Pl. 7. fig. 19) the fourth digit does project, in a sense, but

OF THE AUSTRALIAN MARSUPIALIA. 173

the projection is due entirely to the elongation of the claw. In the skeleton examined
(text-fig. 6, C) the fourth digit projects to about the same extent as the third. It is
nevertheless to be regarded as predominant. The metatarsal of the fourth digit is
displaced backwards, so that the head of that of the third articulates with it as much
as with the conjoined trapezoid and cuboid. Its head is prolonged backwards beyond this
articulation, so that its tip extends to the margin of tle foot, overlapping that of the
enlarged fifth metatarsal. The fourth digit is thus still predominant, but has simply
been displaced backwards for better support. It appears to be one of the most functional
digits of the pes, and the fact that the sole is turned outwards and the hallux downwards
in action, the longest digit thus becoming subjected to great lateral stress, shows
sufficiently the reason of this adaptation.

The available evidence is decidedly in favour of the derivation of the Notoryctes type
of pes from one of ordinary phalangerine type. The resemblance pointed out by Gadow
between Notoryctes and Didelphys is probably indicative of affinity, since, considering
the polyprotodont character of its dentition, Notoryctes must have been one of the first
derivatives of the arboreal syndactylous line. The latter is directly traceable to the
Didelphyide.

PHASCOLOMYID#.

The pes of Phascolomys (Pl. 7. fig. 20) represents a derivative of the second
arboreal phase. The original plantigrade condition has been retained, as in Notoryctes
and Diprotodon. The hallux is reduced to a tubercie externally. In the skeleton
(cf. text-fig. 6, D)it occupies a position of opposability. The entocuneiform to which it
is attached is elongated, and its articular facet is external, so that the axis of the hallux
is placed at right angles to that of the second digit. The terminal phalanx of the
hallux has disappeared, and the proximal phalanx may also be absent. Externally there
is little indication of syndactylism of the second and third digits, the reason being that
-all of the digits are more or less connected by integument. The condition is more
apparent in the immature specimen here illustrated than usually in the adult. In the
skeleton there is some variation in these digits. In some specimens they are very
distinctly slenderer than the fourth and fifth, while in others this disproportion is not
indicated. The fourth digit is always well developed, as in the Phalangeridee. The
plantar surface shows no indications of the arboreal pads of the Phalangeridie, the whole
of it being occupied by a tuberculate skin, as in Phascolarctus and Sarcophilus.

DIPROTODONTIDZ.

In their original description of the foot-structure of Diprotodon, Stirling and Zietz
(1899) show that it presents resemblances with the Phalangeridse and the Phasco-
lomyide. Dollo (1899) has pointed out that the pes is of arboreal derivation, as
shown by the opposability of the hallux, the reduction of the second and third digits and
predominance of the fourth. The following summary, which is based partly on the
original description of Stirling and Zietz and partly on the plaster-casts recently
acquired by the British Museum, is designed to show that, as in the case of the

24,*

174 DR. B. A. BENSLEY ON THE EVOLUTION

dentition, there are general characters of resemblance and special characters of diver-
gence between the pes of Diprotodon and that of Phascolomys, pointing to a common
origin of the two genera or the families they represent.

Both in Diprotodon (ef. text-fig. 6, E) and Phascolomys the plantigrade condition is
retained. ‘The hallux is reduced ; in the former only the metatarsal element is retained,
and its distal portion is swollen to form a knob-like structure. Although reduced in
respect to the loss of phalanges, this digit was apparently functional in Diprotodon,
serving as an antero-internal support, the remaining antero-internal elements, especially
the second and third digits, being poorly developed. In both forms the entocuneiform is
elongated and its articular facet (as opposed to terminal) is external, as in the Phalan-
geride ; so that the hallux is set away from the remaining digits. ‘The two forms differ in
the relations of the mesocuneiform element. This is free in Phascolomys and joined to
the entocuneiform in Diprotodon. In both forms the second and third digits are dis-
proportionately smaller than the fourth. Dollo has correctly pointed out that the larger
size of the fourth digit in Diprotodon is partly inherited and partly adaptive. In Phasco-
lomys the second and third digits have regained their functional importance, while in
Diprotodon they have not. In the former the walking-stress is unifermly distributed
to all of the digits ; in Diprotodon it must have fallen on the outer side of the foot, only
the hallucal metatarsal, beyond the more posterior navicular and calcaneal protuberances,
serving for support internally. This is apparent not only in the general disproportion
between the fourth and fifth digits on the one hand, and the second and third on the
other, but also in the massiveness of the astragalus, with its internally directed facet,
caleaneum, cuboideum, and fifth metatarsal. The shifting outwards of the walking-axis
is attributable in the first place to the reduction of the opposable hallux common to all
terrestrial forms, and in the second to the reduction of the second and third digits as a
phalangerine arboreal adaptation, the three internal digits being thus largely thrown out
of service. The divergent characters presented by Phascolomys and Diprotodon are
explainable on the assumption that in the former the second and third digits have
regained their functional importance, and that in the latter the hallux has remained
functional to a slight extent by being modified as a prop or internal balancer. There is
little doubt that the two forms are divergent members of a single terrestrial plantigrade
line leading from the Phalaugeridee.

PERAMELID&.

The general type of pes in the Peramelide agrees with that of the three preceding
families in that it represents a derivative of the second arboreal phase. It differs,
however, in the substitution of digitigrade developments for plantigrade ones, with a
tendency towards functional monodactylism.

The Peramelidee are without existing arboreal representatives. The only way in which
it is possible to arrange their foot-patterns in proper sequence is, therefore, according
to the extent to which they depart from the arboreal type as presented by the
Phalangeride. In such an arrangement the most primitive conditions are seen to be
presented by the two forms Perameles Raffrayana and P. Cockerelli. It is an interesting

ai

OF THE AUSTRALIAN MARSUPIALIA., 175

fact that P. Cockerelli is very closely related to P. Doreyana, and that the latter, together
with P. Raffrayana, are the most primitive forms in dentition ; also that all three are
Papuan in their distribution. The foot-patterns of the Australian or mainland forms
are traceable to that of P. Raffrayana. These species present a division into short- and
long-footed series: the former including P. macrura, P. obesula, and P. barrowensis; the
latter P. Bougainvillei, P. nasuta, P. Gunni, Thylacomys, and Cheropus. In the long-
footed series development has probably proceeded along three subsidiary lines.

In P. Cockerelli (Pl. 7. fig. 14) we find a very close approximation to the typical
condition in the Phalangeride. The foot is comparatively short; the proportion of
breadth measured immediately behind the hallux to the length measured from the
heel to the tip of the fourth digit, excluding the claw, is approximately 1: 4°3. It is not
possible to make a similar measurement of a phalangerine foot, but it may be observed
that if the foot of P. Cockerelli were provided with a well-developed instead of a reduced
opposable hallux, its proportions would be almost exactly as in the Phalangeridze. In
other words, the plantigrade and non-elongated condition has been perfectly retained.
P. Cockerelli is specialized after the manner of all other secondarily terrestrial forms in
the reduction of the opposable hallux. It is primitive, as compared with other Pera-
melide, in the more radiate arrangement and greater freedom of the outer digits, this being
a phalangerine character which is lost in the more advanced forms. Another primitive
character is observable in the position of the notch separating the conjoined second and
third digits from the fourth on a level with that separating the latter digit from the
fifth. Of the plantar pads characteristic of the arboreal type of pes, only the digital
elements are represented; they are well separated, but have no striated areas, their
surface being granular like the rest of the sole. The claws of the fourth and fifth digits
are slightly curved: this is also a primitive character—in the more specialized sub-
digitigrade forms the claws are stouter and also straighter.

The above description of P. Cockerelli applies almost equally to P. Doreyana. In
P. Raffrayana (PI. 7. fig. 15) the pes is more specialized in its proportions, the ratio of
breadth to length in a specimen measured being 1:6°3. The hallux, however, is almost
better developed, its proximal phalanx and metatarsal being quite large. The plantar
pads are also more primitive. At the base of the hallux there is a small elevation
representing a hallucal pad, the apex of which bears a smooth area, probably representing
a formerly striated surface. The digital pads are well developed and show the same
characters as the hallucal one.

In P. moresbyensis we find much the same characters as in P. Raffrayana. The
proportion of breadth to length in a specimen measured was found to be 1:61. The
hallux is more reduced. ‘There are no signs of a hallucal pad and the first digital pad is
obsolete. The second and third digital pads are well developed, but their apical smooth
areas are not so well marked. Practically the same characters are found in P. obesula.
The proportion of breadth to length in a specimen measured was found to be 1: 6°3.
A dried preparation of P. macrwra showed a proportion of 1: 7°5, but this measurement
cannot well be compared with those of spirit-specimens.

The species just referred to are essentially short-footed forms. In the remaining ones,
with the partial exception of the annectant form P. nasuta, the pes shows definite signs

176 DR. B. A. BENSLEY ON THE EVOLUTION

of elongation and digitigrade development. In P. Bougainvillei (Pl. 7, fig. 16) the
proportion in a specimen measured was found to be 1:11. The hallux is now
reduced almost to a tubercle. The second and third digits are more completely bound
together and also more reduced; they are set backwards in such a way that the notch
separating them from the fourth digit is placed considerably behind that separating the
fourth digit from the fifth. There is no hallucal or first digital pad. The second and
third digitals are well developed and show faint signs of striation. The sole is hairy as
far forwards as the base of the hallux.

In P. nasuta and P. Gunni the conditions are more primitive, although the animals
are larger and approximate more closely to those of the short-footed series. A dried
specimen of P. nasuta showed a proportion of 1: 6:5. A spirit-specimen of P. Gunni
showed a proportion of 1:8. In the latter species the hallux is not so reduced as in
P. Bougainvillei, and the second and third digits are not displaced backwards. 'The
characters are otherwise much as in that species.

Thylacomys leucura (Pl. 7. fig. 17) presents an advance on P. Bougainvillei. The
proportion of breadth to length is about 1:15. The two digital pads (morphologically
second and third) are fused together basally, but the compound structure thus formed
shows two small, smooth, apical areas representing the formerly striated surfaces. The
sole is completely hairy. Judging from a dried specimen, the pes of Z. lagotis
corresponds closely with that of 7. leucwra just described.

Cheropus (Pl. 7, fig. 18) presents a more advanced stage of the digitigrade development
than Thylacomys. In a dried specimen the proportion was found to be 1: 21-4. The
sole is extremely narrow and completely hairy. The haliux is absent and the second
and third digits wholly vestigial. There is a large digital pad which is probably the
result of fusion of the second and third pads, as in P. Bougainvillet. It is an interesting
fact that while in P. Bougainvillei the second and third digits are placed behind the fifth,
in Charopus they are placed in front. Thus, in the reduction of the second, third, and fifth
digits, in connection with the monodactylous elaboration of the fourth, the recession,
as shown by P. Bougainvillei, first influences the second and third, but ultimately, as
shown by Chwropus, these elements are passed by the fifth*.

* It is interesting to notice, in connection with this form, the peculiar conditions which must have attended the
origin of its foot-modifications. In both the front and hind feet Chwropus presents digitigrade cursorial develop-
ments analogous to those of the placental Ungulata. The hind feet have become functionally monodactyl like
those of the Equidee in the perissodactyl series ; but whereas in the latter the third digit is the predominant member,
in Cheropus it is the fourth, Dollo has already shown that the reason of this is that while in the Equide the mono-
dactyl condition has proceeded from a normal pentadactyl one, in Chwropus it has proceeded from a specialized
arboreal type like that seen in the Phalangeridie, in which the third digit was already reduced and the fourth digit
predominant. ‘The front feet present an analogous case. They are functionally didactyl, as in the artiodactyl
Ungulata ; but whereas in the latter division the predominant digits are the third and fourth, in Cheropus they are
the second and third. The reason is that before becoming cursorial the foot of Charopus was already specialized
in a fossorial direction, In the species of Perameles the first and fifth digits, and to a lesser extent the fourth, are
reduced, and the second and third enlarged for digging purposes. It is interesting to note that in Charopus the
two functional toes are not exactly of the same size, the second being slightly but perceptibly smaller.

In developing into a cursorial animal, Charopus has been, in a sense, hampered by the possession of arboreal
characters in its hind feet and of fossorial characters in its front ones.

OF THE AUSTRALIAN MARSUPIALIA. IAT)

MACROPODIDH.

The general type of pes in the Macropodidz represents a terrestrial derivative of the
second arboreal phase. ‘The tendency is towards digitigrade modification and functional
monodactylism. There is an almost perfect parallelism between this family and the
Peramelidee *, although the two are easily proved to be of totally independent derivation,

The pes of Hypsiprymnodon (PI. 7. fig. 21) is familiar as presenting an intermediate
condition between the general type represented by the arboreal Phalangeride and that
characteristic of the Macropodide. The whole foot is comparatively short. The
hallux is fairly well developed and possesses the full number of phalanges. It is
opposable, although not to the same extent as in the Phalangeride. This condition
contrasts strongly with that in the remaining Macropodidze, where the hallux is in-
variably absent. The plantar surface presents the full number of transversely striated
pads. ‘The first and second digital ones tend to fuse together in connection with the
incipient narrowing of the foot.

Contrary to what might be expected from their dental relations, Hypsiprymnodon
shows no points of special affinity in its foot-structure with Lettongia and Afpyprymnus.
In both of the latter the pes is already well specialized; it is very narrow and greatly
elongated, much more so, in fact, than in many otherwise more specialized forms. There
are no indications of plantar pads, with the exception of a large elevation at the base of
the functional digits. The various species show comparatively little variation.

In Potorous (Pl. 7. fig. 23) we find a fairly close approximation to Hypsiprymnodon.
The pes is much shorter than in the preceding genera, and the digits show the lesser
degree of elaboration and also the freedom of lateral movement characteristic of Hypsi-
prymnodon and the Phalangeridee. The full number of plantar pads has been retained,
but their transverse striations are broken by irregular longitudinal markings. The
hallucal and marginal pads are in process of reduction, as in the Peramelidee. ‘The first
and second digital pads are partially joined together, as in Hypsiprymnodon. So far as
can be judged from dried specimens, there is little variation in the different species.
Caloprymnus shows a much closer approximation in its foot-structure to Bettongia and
Ajpyprymnus than to Potorous, as seen in the greater elongation, the greater elaboration
of the fourth digit, and the absence of plantar pads, with the exception of the large
basal structure described for the former genera. The length of the pes as compared
with the breadth is greater in Caloprymnus than in Bettongia, with the exception of
B. Gaimardi, and almost twice as great as in Potorous. These relations appear at first
sight to oppose the serial arrangement arrived at from a study of the dentition; but
there is no doubt that the dental relations are of a more fundamental character. ‘The
fact that the foot-structure of the Phalangerine is practically homogeneous throughout
shows that the greater resemblance of Hypsiprymnodon to Potorous means only

* The arrangement of the tarsals and metatarsals furnishes a point of distinction, In the Peramelidwe the
enlarged fourth metatarsal is supported in part by the ectocuneiform and in part by the cuboid, while in the
Macropodidee it is practically supported by the cuboid alone.

178 DR. B. A. BENSLEY ON THE EVOLUTION

derivation from the same type. Potorous has retained primitive characters by assuming
a more or less fossorial habit, while the remaining genera Bettongia, Aipyprymnus,
and Caloprymnus, of both subfamilies, have, like the Macropodinze, assumed a freer
ground-living habit, with the development of a more elongated saltatorial modification in
the pes.

Passing to the Macropodine, we find two chief types, respectively characteristic of
the arboreal forms (species of Dendrolagus) and the terrestrial ones. The pes of Dendro-
lagus (Pl. 7. fig. 22) is comparatively short. ‘The proportion of breadth to length in a
spirit-specimen of D. wrsinus and also in a skeleton of D. énustus was found to be approxi-
mately 1:3. This appears to indicate a more primitive condition than even in Hypsiprym-
nodon, where, disregarding the hallux, the proportion is about 1:6; but it is very
probable that the shortened condition of the pes has been reacquired. The plantar surface
shows no indications of pads. These, like the hallux, have been lost during the antecedent
terrestrial phase. The second, third, fourth, and fifth digits resemble more closely those
of the Phalangeridze than those of the terrestrial Macropodinze, and similar conditions
characterize the metapodials and phalanges in the skeleton. In the fourth and fifth
digits the claws are definitely curved. It is difficult to state exactly how far these
characters have been secondarily developed or have been carried over from the phalan-
eerine ancestors through a primitive terrestrial form. As to whether there is a special
resemblance between Dorcopsis and Dendrolagus in foot-structure as there is in dentition
has been difficult to ascertain from a comparison of the available specimens, all ef those
representing the former genus being dried preparations. The general conformation of
the pes in Dorcopsis is as in other terrestrial macropodine forms. The ratio of breadth
to length is approximately 1: 7-9, the foot being relatively short as in Potorous. It is
an interesting fact that the pes of D. luctuosa presents a distinctly swollen condition as
compared with that of D. Muelleri or D. Macleayi, and the heel, and in fact the whole
pes, is less elongated than in the latter species. These characters point to those of
Dendrolagus, and afford some evidence for the conclusion arrived at from a study of the
dentition, namely, that Dorcopsis is a secondary terrestrial derivative of the secondarily
arboreal Dendrolagus.

The members of the macropodine series, Petrogale, Onychogale, Lagorchestes, Lago-
strophus, and Macropus, show very great uniformity in the general pattern of the pes,
and the specimen of Macropus dorsalis here figured (Pl. 7. fig. 34) may be taken as
representative of the group. The ratio of breadth to Jength in this specimen is 1: 13°5.
The fifth and the intimately joined second and third digits are now more closely associated
with the fourth, and thus show a great reduction of the free movement characteristic
of more primitive plantigrade forms. The fourth digit is perfectly axial in position, and
corresponds very closely in stoutness with the middle portion of the foot. The plantar
surface is covered with a tuberculated skin, and of the plantar pads the marginal as well
as the hallucal one has completely vanished ; while the three digital pads are completely
fused together, forming a prominent protuberance which shows not the slightest traces
of the striated areas formerly present.

Comparison of the proportions of the feet in the various genera reveals a slight

OF THE AUSTRALIAN MARSUPIALIA. ie

variation, inasmuch as those of Zagorchestes, Lagostrophus, and Onychogale are re-
latively more elongated than those of the remaining species with which they are most
closely associated in size. The following data will suffice to illustrate this difference,
although the figures given must be taken as roughly approximate, being based on dried
preparations :—Three specimens representing the three species of Onychogale, O. frenata,
O. lunata, and O. unguifcra, show a ratio respectively of 1: 14°5-14°6-17:1; a specimen
of Lagostrophus fasciatus, 1:15°1; three specimens representing the species of Lagov-
chestes, L. leporoides, L. hirsutus, and L. conspicillatus, respectively 1: 15-15:7-18.
On the other hand, specimens representing the species of Petrogale, P. penicillata,

P. brachyotis, P. concinna, and P. inornata, show a ratio respectively of 1: 10°7-10°9-:
y i J

11°8-14; while the Small Wallabies of the genus J/acropus show a range from 1:7
(1. Coxeni) to 1:13°5 (IL. Lugenii). In Lagorchestes and Lagostrophus, moreover, the
sole of the foot tends to become hairy.

As regards the three sections of the genus Macropus, namely the Small Wallabies,
Large Wallabies, and Kangaroos, there is a marked increase in the actual size of the
foot as in the size of the body proceeding from one group to the next, and this difference
has been made use of by Thomas (1888) in defining the three sections. There appears,
however, to be no broad differences in the proportions of the feet separating these sections,
although it is possible to demonstrate by measurement that the members of the Small
Wallaby group show a greater tendency towards the retention of the original shortened
form than do those of the other two. This is indeed what we should expect, because
while the initiation of a saltatory method of progression demands the development of
an elongated pes, yet its perfection is more a matter of the actual size of the animals
and the proportion of the hind limb as a whole.

The species referred to as Setonyx brachyurus corresponds in the proportion of the:

foot with with the Small Wallabies, the ratio in a specimen measured being 1: 7°9.

THr IDENTIFICATION OF THE STEM-FoRM OF THE AUSTRALIAN MARSUPIALS.

So far as deducible from the modifications of dentition and foot-structure, the data
already given in the two preceding sections are suflicient for the construction of a
phylogenetic or morphogenetic plan, except in one important particular, namely, that
they fail to express the relationships of the primary families Dasyuridee, Peramelide,
and Phalangeride. None of the latter present the characters of a prototype, such
characters being distributed over all three. Thus the Dasyuridie in dentition occupy
a prototypal position, except in respect to the upper incisor formula and the characters
of the upper molar styles, in which they give place to the Peramelidie; while the
Phalangeridxe occupy a prototypal position in foot-structure, except in the syndactylous
condition of the second and third digits, in which they give place to the Dasyuridie.
The Australian fauna considered by itself represents a radiation without a recognizable
source, and in order to complete it it is necessary to add a hypothetical form combining
the prototypal characters of the three primary families.

SECOND SERIES.—ZOOLOGY, VOL. IX. 25

180 DR. B. A. BENSLEY ON THE EVOLUTION

Such a generalized type while not represented in the Australian group is, except for a
more primitive condition of the lower incisor formula, exactly represented by the
American Didelphyidze, and the present section is devoted to a consideration of the
various members of this family both in their relations to the Australian fauna and to
one another.

Reviewing the opinions already expressed with respect to the relationships of the
Australian fauna, we find them to be of the most diverse kind. Writing in 1871 Owen
remarked :—‘‘ Among these initial forms of Marsupialia [referring to the Mesozoic
Mammalia] we may see in Amphitherium the prototype of Myrmecobius ; Peralestes has
culminated in Sarcophilus; Triconodon in Thylacinus; Plagiaulax is to Thylacoleo
what the Weasel is to the Lion. But derivative change has not advanced to the
long-limbed saltatory type of Marsupial; nor has any evidence yet been had of a
Mesozoic predecessor of the climbing Koala, the volant Petaurist, or the burrowing
Wombat.

“The Marsupial type... has in America progressed to and been succeeded by the
more specialized form of Didelphys.

‘If Australia possessed Marsupials as far back in time as did America and Europe,
analogy would lead us to suppose that the primitive diminutive multimolar insectivorous
type prevailed. It has not there yet become extinct ; but it seems to have been reduced
to the solitary exceptional form of Myrmecobius.”

Wallace, in his ‘Geographical Distribution of Animals’ (1876, v. p. 2), has expressed a
similar view of a connection of the Australian fauna with northern Mesozoic forms :—“ As,
however, no other form but that of the Didelphyidse occurs there [in Europe] during
the Tertiary period, we must suppose that it was at a far more remote epoch that the
ancestral forms of all the other marsupials entered Australia; and the curious little
mammals of the Oolite and Trias offer valuable indications as to the time when this
really took place. .. . It was probably far back in the Secondary period that some portion
of the Australian region was in actual connection with the northern continent, and
became stocked with the ancestral forms of marsupials.”

To the same order belong the views of Falconer and others who have sought to
establish a relationship between the Australian diprotodont forms and the Plagiaulacide,
and that cf Cope (1882, 1884), who actually referred the form Thylacoleo to the family
Plagiaulacide, at the same time connecting the family with the Macropodide through a
hypothetical ancestor Tritomodon.

The ‘ Systematische Phylogenie’ of Haeckel (1895) contains much more definite views
of a connection of the Mesozoic Mammalia with existing Marsupials. This writer
recognizes a Jurassic group of Prodidelphia giving rise to the Marsupials along poly-
protodont and diprotodont lines.

Apparently the only writers who have anticipated what is probably the true relationship
of the Australian fauna are Winge (1893) and Lydekker (1896), both of whom favour
a. more direct connection with the existing American Didelphyide than with Mesozoie
forms. In a general scheme of the marsupial families Winge indicates three lines of
development—one of them represented by a group composed of Diprotodon, Thylacoleo,

OF THE AUSTRALIAN MARSUPIALIA. 181

Phascolomys, and Phascolarctus, a second by the Peramelidee, and a third by the Dasyuridee,
from didelphyid prototypes. Lydekker, in his ‘ Geographical History of Mammals,’
remarks as follows :—‘* Recent researches have tended to show that the alliance between
the Dasyuridx and the Didelphyide is much more intimate than was formerly supposed
to be the case. This being so, it is a fairly safe assumption that both families are
descended from a single common ancestral stock... . Not improbably polyprotodont
Marsupials survived in south-eastern Asia till the early portion of the Eocene division of
the Tertiary epoch, and in this region both Dasyuride and Didephyid were differentiated.
Representatives of the former family soon afterwards found their way into Australia,
while the Opossums would appear to have dispersed in one direction into Europe, and in
the other into North America.”

Certain opinions have, however, been expressed in favour of a South-American
origin of the Australian fauna. Ameghino (1891) regards the forms described by him as
Microbiotheriidze as ancestral to both polyprotodont and diprotodont Marsupials, as well
as to the Insectivora and Chiroptera. Spencer (1896) has pointed out the possibility
that the Australian Marsupials and the Didelphyide may have originated at some time
during the Cretaceous period from South-American ancestors. In a recent paper
Iydekker (1899) has suggested the origin of the Dasyuride from South-American
Prothylacinide (Sparassodonta), this view being based for the most part on the resemblances
of Thylacinus to that family.

The main evidence as to the general relationships of the different groups of Marsupials
may be summarized as follows:—During the Oligocene period the Didelphyide,
represented by Peratherium, were widely distributed in the northern hemisphere, and if,
as seems very probable, the Microbiotheriidee of Ameghino, or at least some of them, are
in reality members of this family, they were present at a slightly later period in South
America as well. Of fossil forms at present known, the Oligocene Didelphyide are the
earliest which may definitely be referred to the Marsupialia. The idea that the more
ancient fossil Mammal]s must be Marsupials is untenable, and the only evidence on which
the identification of the Jurassic forms as Marsupials now depends is the presumed
existence of a single tooth-change in 7’riconodon, concerning which, as Lydekker (1899)
has pointed out, there is room for doubt.

But whether the Oligocene Didelphyide were the first true Marsupials to be
differentiated, or whether they were the descendants of earlier Marsupials as yet
unrecognized, they are the ancestors of all of the later-appearing forms. In the develop-
ment of the latter there is evidence of at least three different radiations. The first aud
most extensive one is that represented by the Australian fauna. A second is, in all
probability, represented by the Miocene fauna of South America. The existing
Didelphyidee of South America, which might at first sight be regarded as surviving
remnants of the original didelphyid radiation, may be shown to represent a third radiation
which is at the present time in its very incipient stages. Of these three radiatious the
Australian and the existing South-American ones are directly traceable to minute
primitive didelphyid forms like the existing genera Marmosa and Peramys, or the
extinet Peratherium.

25*

182 -DR. B, A. BENSLEY ON THE EVOLUTION

Winge (1893), who has made a careful study of the existing Didelphyide, divides the
family as follows * :—

I. Hindmost of the three outer cusps of the upper molars well developed.
A. p. 1 and m. 3 comparatively well developed. Molar teeth with high sharp cusps. Body of the
lower jaw shallow. Proc. angularis sharply inflected. Ribs narrow. No marsupium.
Grymeomys {| = Marmosa}.
B. p. 1 and m. 3 reduced. Molar teeth with low rounded cusps. Body of the lower jaw deep.
Proce. angularis only slightly inflected. Ribs broad. Marsupium in the form of folds.
Philander | = Caluromys}.
II. Hindmost of the three outer cusps of the upper molars reduced.
A. Tail long. Terminal phalanx of hallux broad.
(a) No swimming-membrane between the toes. Skin of palms and soles normal.
Didelphys | = Didelphys and Metachirus}.
(6) Swimming-membrane between toes. Skin of palms and soles provided with peculiar
papille.
Chironectes.
B. Tail short. Terminal phalanx of hallux compressed.
Hemiurus [= Peramys].

He considers the relationships of the above genera to be as follows :—

Chironectes. Peramys.
\ f
\ ;
Philander. Didelphys.
/
\ hs

Grymcomys.

The genus Peramys is regarded by Winge as derivative of Didelphys, but in many |
respects it is prototypal not only to that genus but also to Warmosa. It will be observed
that Winge’s main division associating Didelphys and Peramys is based on the characters
of those elements of the upper molars here referred to as external styles. These structures
are regarded by him as typically three in number and as the original cusps of the crown
(triconodont type). Winge remarks with reference to Phascogale, Dasyurus, and their
allies in the Dasyuride that they present a more primitive condition of the upper molars
than is found in any other of the existing Mammalia, the three outer cusps being
persistent, and the median one of them, like the median one of the three internal cusps
of the lower molars, being larger than the others. He also states that in Grymeomys
(Marmosa) the three outer cusps are present, but the median one has lost its pre-
dominance, while again in Hemiurus (Peramys) and in Didelphys the hindmost cusp
has become reduced. He illustrates these cases by the teeth of Phascogale penicillata,
Marmosa cinerea, and Peramys domestica, designating the three cusps numerically in
the order of their occurrence from before backwards.

* On account of these results being written in the Danish language the writer has taken the liberty to substitute
a translation for the original text.

OF THE AUSTRALIAN MARSUPIALTA. 183

That the Dasyuride should present more primitive characters than the Didelphyide, a
family which is otherwise so completely prototypal, appears at the outset improbable. In
addition it may be shown that the three cusps figured by Winge in Marmosa cinerea
and Phascogale are not homologous, the third cusp of the latter animal not being present
in Marmosa, whose third cusp is equivalent to the second of Phascogale, and whose
second cusp represents a new development.

In Pl. 5. fig. 14 will be found a carefully drawn external profile view of the inner
molar of Peratherium, in which the outer cinguler ridge will be seen to bear in all six
elevations. The latter may be designated in the order of their occurrence from before
backwards as a, 0, b,, ¢,, ¢, ¢)*, the letters indicating those elements which are well
developed and have, for the most part, definite homologues in other polyprotodont forms,
while the letters to which numerals are appended indicate smaller and more subsidiary
elements. The cusps 0 ( or ab)—see account of Dasyuridze,—e, and Cy represent the three
outer cusps figured by Winge in Phascogale penicillata. On examining the arrange-
ment of these structures in the various species of Warmosa and Peramys, with reference
to the type presented by Peratherium, there is found to be not a single character of
them which will serve to separate one genus from the other. The stylar formula varies
to a certain extent in different individuals of a species, and even in different teeth of
the same animals, the range being from a formula of a, 0, ¢ to one of a, b, b,, ¢), 6, ey.
In Peramys the more subsidiary cusps are on the whole poorly represented; ¢,
(Winge’s 3) may be present or absent. In Marmosa c, is nearly always present and
comparatively well developed, the only notable exception being in M. elegans, in which
only two of some twelve or fourteen specimens presented signs of intermediate cusps.
In Marmosa murina, WM. cinerea, and M. rapposa, ec, is better developed, as compared
with the other cusps, than in the remaining species of the genus, and a similar condition
is found in some cases in Caluromys. This cusp is that designated as 2 in Winge’s
figure of Marmosa cinerea. Its developmental stages, as here interpreted, are repre-
sented in Pl. 5. figs. 1, 28, 29, 380 of Peratherium, Peramys Iheringi, Peramys
americana, and Marmosa cinerea.

The question here arises—How can it be shown that the cusp c, is a subsidiary
element which is becoming enlarged in such forms as Jf. cinerea rather than an original
cusp which is becoming reduced? According to Winge’s view, the molars of IW. murina
are more primitive than those of the smaller forms I. microtarsus and WM. pusilla,
because in the former the median outer cusp is better developed and the teeth are less
compressed in an antero-posterior direction. But in reality they are much more specialized,
because not only J. cinerea but also I. murina and M. rapposa are closely allied to
Caluromys, of which they represent an ancestral type. Their teeth are transitional in
pattern between those of Calwromys and those of the smaller forms, and this is true in
respect to the peculiarly elongated canines and the reduced anterior premolars as to the
molars. The molars of Caluromys are noticeable for their lateral compression and the
obsolete character of the external cingulum (¢/. Pl. 5. fig. 27). The apparently well-

* Cf. text-fig. 1, p. 89.

184 DR. B. A. BENSLEY ON THE EVOLUTION

developed condition of style c, in Marmosa murina, M. cinerea, M. rapposa, and in some
specimens of Caluromys in reality represents a general levelling down of the whole stylar
series prior to its obliteration. In some specimens of Calwromys, as in that here figured,
ihe cingular ridge is seen to be extremely small and thin, and to bear, instead of definite
styles, a number of minute crenulations. All of the molar characters of this genus
are indicative of an incipient omnivorous development, pointing towards that seen among
the Australian Marsupials in the Phalangerinz, where, even in the most primitive forms
(of. Pl. 5. fig. 18, Distachurus), the external styles have already disappeared.

In the writer’s opinion the relationships of the Didelphyidse are as indicated in the
following modified plan :—

Chironectes. Didelphys.

\
Culuromys. ‘

\

Metachirus.

Dromiciops.
x
x \

—Peramys.

Marmosa.

This represents an incipient radiation, since the diverging lines are directly traceable
to Marmosa. It is an interesting fact that the representatives of these lines are still
scarcely separable systematically from that genus. As in the Australian families,
increase in size of the body appears to have been an essential feature of the evolution.
With the exception of Dromiciops, the two genera J/urmosa and Peramys include the
smallest and most primitive forms of the family. Peramys bears much the same relation
to Marmosa as Sminthopsis does to Phascogale, being in general more primitive in
dentition, but specialized in a terrestrial direction in foot-structure. The most important
dental character demonstrating the prototypal position of Peramys relates to the posterior
premolars, which are, as a rule, larger than the median teeth. In Peratherium they are
apparently always larger, but in Marmosa, as well as in the more specialized forms, they
show definite signs of reduction. The foot-characters in which Peramys is specialized
relate to the reduction of the terminal phalanx of the hallux, reduction of the striated
plantar pads, and shortening of the fifth digit, all of which modifications are connected
with a secondarily terrestrial habit. Peramys is also more specialized than Marmosa
in the shortened, hairy, and, apparently, non-prehensile character of the tail. It is
interesting to note that in the Didelphyide the tail ranks with the teeth and feet as an
organ of adaptive change. As indicated in Winge’s plan, Marmosa represents the
parent form of all the remaining genera. The existing species have, however,
undergone a specialization of the posterior premolars, a reduction similar to that found
ona larger scale in Phascogale. Some of the species have further undergone a special
development of the molars, canines, and anterior premolars pointing towards the genus

OF THE AUSTRALIAN MARSUPIALIA. 185

Caluromys. Marmosa is completely prototypal in foot-structure, apart from a slight
tendency in some species towards syndactylism of the second and third digits, the pes
being of a primitive arboreal type, with widely opposable hallux and well-developed,
striated, plantar pads. The tail is long and prehensile. The minute form 1. velutissima
forms a connecting-link between Marmosa and Peramys.

Caluromys is an advanced form of one of the larger species of Marmosa, such as
M. rapposa, M. murina, ov M. cinerea. In its foot-structure, as also in the character
of its tail, it continues the arboreal developments of Marmosa. In dentition the main
progressive character is the omnivorous type of molar pattern in an early stage of
differentiation. The anterior premolars are reduced and the canines are elongate and
compressed. This character, at first sight appearing to indicate carnivorous habit, is in
all probability of secondary sexual significance.

The peculiar form Dromiciops (Thomas, 1894) is a derivative of Marmosa, approxi-
mating to Caluromys in the characters of its molar teeth, and to Peramys in the
shortened and hairy condition of the tail.

The two genera VWetachirus and Didelphys are so closely allied that it is not possible to
separate them on the characters of the dentition or foot-structure. Their dentition
repeats on a larger scale the characters of Marmosa. In the upper molars style ¢, is
usually indicated on account of the greater space available on the metacone-spur.
Intermediate styles are poorly represented. Both genera show a slight tendency
towards terrestrial specialization in foot-structure. In other respects Metachirus is
more primitive than Didelphys, approaching Marmosa in its smaller size and the
incomplete development of the marsupium and its uniform coloration, the homogeneous
character of the fur, and the conformation of the ears. It is possibly related to Marmosa
through I. elegans.

Chironectes has been pointed out by Thomas (1888) as very closely allied to Metachirus,
and in its cranial characters indistinguishable from that genus. Apart from its obvious
metachirine derivation, the special features of Chéronectes are remarkable and unique
among Marsupials, as indicating an aquatic adaptation, the tail being flattened and
the feet, which have a modified opposable hallux, being provided with swimming-
membranes connecting the digits.

In identifying the characters in which the Didelphyide occupy a prototypal position
to the Australian Marsupials, it is not possible to rely entirely on fossil forms, the
information available concerning them not being sufliciently full. In utilizing the
characters of the existing Didelphyidze to supplement the evidence of Peratherium, the
facts pointed out above must be borne in mind, namely, that they represent a con-
temporaneous, although less extensive, radiation. ‘Three different classes of characters
may thus be expected, namely : (a) those which indicate a more primitive condition than
is found in any of the Australian forms, as, for example, the increased lower incisor
formula or the arboreal type of pes with opposable hallux and non-syndactylous second
and third digits; (2) those representing conditions exactly equivalent to those already

136 DR. B. A. BENSLEY ON THE EVOLUTION

mentioned as prototypal for the Australian series, as, for example, the tritubercular type
of upper molar; (¢) those which indicate a similar potential of evolution, such as the
reduction of the posterior premolars and the development of syndactylous modifications
in the pes. It will be observed that all the characters of the dentition and foot-
structure described as prototypal, and distributed over three families in the Australian
series, may occur together in the Didelphyide in a single species.

Dentition,

Molar Patterns.—Yhe upper molars of the smaller Dasyurinee have already been
mentioned as completely prototypal, except in the arrangement of the external styles.
We may now notice the fact that their patterns are practically identical with those of
Peratherium and the existing Didelphyidee (cf. PL. 5, figs. 1 a-0, 2 a-b, 26, of Peratherium,
Sminthopsis leucopus, and Metachirus opossum). The only important exceptions to
this general statement are as follows :—In the first place, the molars of the more advanced
Didelphyidze (fg. 26), while differing from those of the Australian forms (advanced
Dasyurinze and Peramelide) in tending to retain the original insectivorous character,
show less antero-posterior compression than those of the smaller forms, and also
Peratheriumand the smaller Dasyurinze. Secondly, one member of the family, Caluromys,
shows indications of an omnivorous development, which, if continued further, would
parallel that of the Phalangerine. ‘Thirdly, there are minor variations in the
arrangement of the external styles; all of them, however, both in the Didelphyide and
the Australian forms, are easily traceable to the Peratherium-type.

Comparing the secoud molar of Sminthopsis leucopus here figured with that of
Peratherium, it is seen that in the former styles a and 6 are fused together. Intermediate
styles are not represented. Style ¢ is greatly enlarged. In other members of the
Dasyurinze style ¢, is occasionally developed. In the British Museum Collection it is
present in m. 2 of eight specimens of Phascogale flavipes, in m. 1 and m. 2 of one
specimen of P. Swainsoni, in m. 1 of three (of five) specimens of P. penicillata, and in
m. 2 and m.3 of two. One specimen of P. minutissima shows both style c, and 6, in
m.1. Style ec, is generally absent in the smaller forms. Even in Peratherium it is so
small as to be scarcely recognizable. It is altogether absent in all of the British Museum
specimens of Sminthopsis. It is present in m.2 in one specimen P. Swainsoni; it is
absent in all of eleven specimens of P. flavipes. In Phascogale penicillata, the type
figured by Winge in connection with this element, it is present in m.1 and m. 2 in three
of five specimens. It is absent in m.3 of three specimens, and in each of the two others
it is represented by two minute tubercles. It is present as a trace in m.1 and m.2 of
one specimen of P. dorsalis and one of P. Thorbeckiana, and is barely indicated in most
specimens of Dasyurus (cf. Pl. 5. fig. 3).

The arrangement of the external styles in the existing Didelphyidz has already been
referred to. It needs only to be added that not only are the variations traceable to
a Peratherium-type, but also that there is less departure from that type than is seen in
the Dasyurine. The same tendency towards the fusion of styles a and 6 is noticeable
(cf. Pl. 5, figs. 28, 29), but is not so apparent as in the Dasyurine. ‘The condition is

OF THE AUSTRALIAN MARSUPIALIA. 187

nearer that shown by the Peramelidee (Pl. 5. fig. 8). Style ¢ does not become greatly
enlarged as in the Dasyurin, but retains more the proportions seen in Peratheriwm
and the Peramelide. Finally, there is a greater tendency towards the retention of the
intermediate styles 0, and ¢,.

The following additional characters of resemblance between the Didelphyidee, including
Peratherium, and the Dasyurinz and Peramelidee may be noticed. The first molar is
always more laterally compressed than the second, while the reverse is true of the
third. Tn the latter also the external styles are poorly developed. The fourth molar
is in every case reduced.

Passing to the lower molar patterns, we note the fact that the smaller Dasyurinz
present prototypal relations which are exactly duplicated in the Didelphyide. The
pattern figured for Sménthopsis leucopus (Pl. 6. fig. 2) will be seen to be practically
identical with that of Peratherium (fig. 1) and Metachirus (fig. 32). An exception will
be noticed in the absence of the entoconid in Sminthopsis, but, as already mentioned in
connection with the Dasyurine, the condition in this form and other species of
Sminthopsis is not typical of the subfamily. It is interesting to note that, notwith-
standing the increase in size of the body in the advanced members of the Didelphyide,
the primitive molar type is for the most part retained, while in the similarly advanced
members of the Dasyurinz carnivorous modifications begin to appear.

As noticed above, the first lower molar in the Dasyurinz shows a series of reductions
of the paraconid and metaconid, beginning with a type like that figured for Sminthopsis
crassicaudata (Pl. 6. fig. 6), in which these cusps are fairly well developed. In the
primitive Peramelidee, in Peratherium, and all of the existing Didelphyide (cf Pl. 6.
fig. 33) we find prototypal relations, these elements being well developed throughout.

In the fourth lower molar of the Dasyurinie there is a marked tendency towards the
reduction of the talonid. The same tendency is seen in a somewhat lesser degree in
the existing Didelphyidxe. Five specimens of Peratheriwm in the’ British Museum
Collection show the characters of these teeth, and in three of them the talonid is fairly
developed, and all of the three cusps are recognizable, while in two others the talonid is

reduced *.

Incisors.—On account of the lack of information concerning the incisors of Pera-
therium the following comparisons include only the existing Didelphyidee. In the
Australian series, and apart from the incisor formula, the smaller Dasyurine are the
only forms showing prototypal characters. In these we find no less than five points of
resemblance with the Didelpbyidee. The median upper teeth are procumbent and sub-
caniniform. The upper lateral teeth, when unworn, have triangular tips. The median
lower incisors are unmodified. Like the lower lateral teeth their tips are rounded.
The insertion of the three anterior lower teeth is arranged so that the root of the second

* These specimens are P. arvernensis (Puy-de-Dome, M. 27700), an undetermined species from Auvergne
(M. 27811), an undetermined species from Hordwell, Hants (M. 30350), P. Lamandini (Caylux, M. 2888 d), and
P. afjinis (Tarn-et-Garonne, M. 2388 a). Cf. Lydekker (1887).

SECOND SERIES.— ZOOLOGY, VOL. IX. 26

188 DR. B. A. BENSLEY ON THE EVOLUTION

is placed behind that of the first, while that of the third is placed externally to that
of the second.

In the incisor formula the Peramelidze are the most primitive of all the Australian
forms, the number being The Dasyuride come next with the vniform formula
of e. The Didelphyidze present the prototypal condition of 2 . The mutual homologies
‘of the teeth in the three families are not at all clear. Thomas (1888), Winge (1895), and

Dependorf (1898) have estimated the original marsupial formula as > but there is no

5)

agreement as to how reduction from this type has been effected in the different families.
The various views may be tabulated as follows :—

Wilson and
Thomas (1888). Winge (1895). Woodward(1895). Hill(1897). Dependorf (1898),

9° 9345
Didelphyide ...... rete a

12345 12345 12345% 12345
Peramelidce ....-- 12300 OSs Boe Tuae 0234? 10340

: 12340 12340 12045 12340
Dasyuridee ......... 12300 02340 oma 5s) +e) aa 02340

The evidence on this point is admittedly unsatisfactory, but it is probable that the
plan of reduction suggested by Winge is the correct one. In the first place, there is
evidence to show that the anterior lower incisor has disappeared throughout. Winge
has pointed out that the median lower incisors are in relation with the second upper
teeth, with which they are accordingly homologous. There are no lower teeth corre-
sponding to the median upper ones, and it may thus be supposed that the development
of a procumbent caniniform condition in them may have disturbed the relations of their
original homologues in the lower jaw, leading to their reduction. Furthermore, both
Woodward and Dependorf have described vestigial germs situate in front of what are
later on the first functional lower teeth in embryos of Dasyuwrus. Wilson and Hill
mention a possibly equivalent germ in Perameles, although this is not mentioned by
Dependorf for that genus.

In the second place, there are reasons for believing the three lower incisors of the
Dasyuridee and Peramelidze to be homologous with one another and with the first three
lower incisors of the Didelphyidee. Wilson and Hill have called attention to the fact
that in embryos of Perameles the germs of the lower functional teeth occupy a peculiar
relation to one another, the second being situate in a line posterior to the first, while
the third, in consequence of a sharp bending outwards of the dental lamina, is situate
externally to the second, so as to appear in the same transverse sections. Rdése (1898)
has figured the same condition in a model of the tooth-germs of Didelphys aureus,
although he has mistaken the homologies in designating the external germ as second
instead of the third, and vice versd. Exactly the same conditions are seen in the adults
of all three families, Dasyuridee, Didelphyidz, and Peramelidze, where the second tooth
is inserted behind the first and the third tooth externally to the second, the tips at the
same time forming a uniform row. According to Dependorf’s idea, the lower teeth of

* As determined from their description of the tooth-development of Perameles (1897).

OF THE AUSTRALIAN MARSUPIALIA. 189

Perameles and Dasyurus are not homologous throughout, since in the former he
recognizes a vestigial germ of a morphological second incisor (dé. 2). This element,
however, is probably not a member of the same series as the functional teeth, but is more
likely equivalent to the calcified germ described as di. 2 by Wilson and Hill, or, in other
words, the predecessor of the functional second (morphological third incisor), and also
homologous with the germ described as di. 3 in the upper jaw, which, according to
Wilson and Hill, is large, but degenerates early in Perameles, and is present and calcified
in Dasyurus.

Finally, there is evidence of the presence of vestigial germs in the posterior incisor
region in the Dasyuride and Peramelidee. Dependorf describes, in his fourth and fifth
stages in Perameles, a vestigial germ (di.5), presumably corresponding to the fourth
lower functional tooth of the Didelphyidz. This element is not mentioned by Wilson
and Hill for Perameles, but is described by Woodward for Thylacomys (Peragale), and
by both the latter writer and by Dependorf for Dasyurus. A fifth upper vestigial germ
is also mentioned by Dependorf for Dasywrus, although it is not mentioned by Woodward,
this writer having apparently concluded that the reduction took place in the middle
of the series. Dependorf states his inability to recognize either the upper or lower
intermediate teeth described by Woodward for Dasyuwrus. Further study of the whole
question from the standpoint of comparative embryology is greatly to be desired.

Canines.—These teeth call for little consideration. They present exactly the same
insectivorous characters in the smaller Dasyurine, and, in the case of the upper, in the
more primitive of the Phalangerinze (Acrobates and Distechurus), as in the Didelphyide,
with the partial exception of Caluromys.

Premolars.—In the general characters of the premolars the smaller Dasyurine and
the Didelphyidze are inseparable. The chief special features relate to the proportions
of the posterior teeth. As already mentioned above, the Dasyurine present successive
stages of reduction of the posterior premolars, beginning with a condition in which they
are slightly or disproportionately larger than the median teeth, and ending in their
total obliteration. This reduction has also been mentioned for certain of the Phalan-
gerinze (Acrobates and Distachurus), and for two members, Thylacomys lagotis and
Cheropus castanotis, of the Peramelide. Peratherium presents a prototypal condition,
the posterior premolars being apparently always larger than the median teeth*. In
the existing Didelphyidz we find indications of the same process of reduction seen in
the Dasyurinz, with the difference that it does not proceed to the obliteration of the
teeth or even to the loss of their functional value. Of all the existing Didelphyide
Peramys makes the closest approach to Peratheriwm, the upper posterior premolar
being always larger, and the lower posterior tooth slightly larger than, or equal to, the
median.

* This statement is based on five specimens in the British Museum collection, including P. Aymardi (2388),
P. arvernensis (27700 and 27810), an undetermined species (27811), and P. eailis (27806, type), and also on the
description given by Gervais (1859), by whom two of the specimens have been figured. Cf. Lydekker (1887),

26*

190 DR. B. A. BENSLEY ON THE EVOLUTION

Deciduous Premolars.—Except in the case of IMyrmecobius, these teeth have not been
considered in dealing with the dentition of the Australian forms, the reason being that
their characters dc not relate to the secondary evolution of the group *, but with the
primary differentiation of the Marsupials as a whole. They furnish such conspicuous
proof of the prototypal position of the Didelphyide, however, that their characters may
well be used to supplement the evidence already given.

It is a well-known fact that in the differentiation of the Marsupials one of the
dentitions has been obliterated, although as to which dentition is indicated is still a
matter of controversy. Wilson and Hill (1897), in expressing the opinion that it is the
milk-dentition which has been reduced, have pointed out the fact that in the Didel-
phyidze the deciduous premolars are well developed, while in the Dasyuride and
Peramelid they are reduced to vestiges. This they regard as indicating a final stage
in the obliteration of the milk-dentition, the posterior deciduous premolar being at the
present time in process of reduction. The following remarks, although specially
designed to show the prototypal position of the Didelphyidz, may be taken as con-
firmatory of that view.

The deciduous premolars are well represented in the British Museum specimens of
the Didelphyide, partly on account of the completeness of the collection, and partly
from the fact that the teeth are not replaced until well on in life, and thus appear in all
of the younger specimens. They show very little variation throughout the family. The
lower teeth resemble almost exactly the true molars, and the same remark applies to
the upper teeth, except that here there is a fusion of the paracone with style d. The most
important feature of the teeth is, however, their large size and obvious functional
condition. They are at most only slightly smaller than the first true molars.

The deciduous premolars are to be seen in only a few of the British Museum specimens
of Dasyuridee, as Sminthopsis murina, Phascogale flavipes, P. minima, P. penicillata, and
Thylacinus cynocephalus. In all of these the deciduous teeth are found in an advaneed
state of reduction. They are, in fact, quite vestigial. In all except Thylacinus they tend
to retain a molariform shape. ‘There is an internal cusp, a large central cusp, a metacone,
and a third anterior smaller cusp, respectively representing protocone, metacone, and
paracone. ‘The teeth are relatively more reduced in P. penicillata than in the smaller
forms. In Thylacinus we find an extreme stage of reduction, the teeth being represented
by minute triangular plates which are displaced before birth (cf Flower, 1869). In
Dasyurus, where the posterior premolars are absent, the deciduous premolars only occur
as calcified vestiges in the embryo (Woodward, 1895; Dependorf, 1898).

Thomas (1887) has expressed the opinion that the reduction of the deciduous premolars
in the Dasyuride is associated with the reduction of their successors, the posterior
premolars. While this is in a certain sense true, it must be remembered that the
association is only incidental. In Thylacinus, where the deciduous teeth are greatly
reduced, the permanent teeth are well developed and predominant. The same condition
is seen in the Peramelidze and in Phascolarctus,

* A partial exception is afforded by the Macropodide and advanced Phalangerine (vide infra).

OF THE AUSTRALIAN MARSUPIALIA. 191

Of the Peramelide the deciduous premolars are represented in specimens of three
species of Perameles (P. obesula, P. nasuta, and P. Cockerelli), as well as in Charopus
and Thylacomys. In all of these, as in the Dasyuridee, they are greatly reduced as com-
pared with those of the Didelphyidee. The specimen of P. xasuta presents the same
appearance as one figured by Wilson and Hill (1897, pl. 32. fig. 78), the minute deciduous
tooth being supported on the posterior slope of its enlarged successor. A reduced
condition of the deciduous teeth has been figured by Flower (1869) for P. Bougainvillei
and by Rose (1892) for P. Doreyana.

None of the specimens representing the smaller species of the Phalangeridze show the
deciduous premolars. In Phascolarctus, as pointed out by Thomas (1887 0) and
Dependorf (1898), they are wholly vestigial. In the final members of the Phalangerine,
Phalanger and Trichosurus, as well as in the Macropodidee, we find a modification of
quite the reverse kind, the deciduous teeth being well developed and functional until
well on in life. In the phalangerine forms they perform the same sectorial functions
for the young as the posterior premolars do for the adult, while in the Macropodide,
where the median premolars are retained as sectorials in the young, the deciduous teeth
take on, or perhaps retain, a molariform shape. In the Phascolomyide the deciduous
teeth have apparently been reduced in the ordinary way. In view of the concise
evidence in the Dasyuridze and Peramelidée of retrogression from a didelphyid type, it
is surprising to find the deciduous teeth so well developed in the most advanced forms
of Marsupials, such as the advanced Phalangerinz and Macropodidee, where one might
expect to find them wholly absent. Whatever may have been the original cause of the
reduction of one dentitien—Leche has ascribed it to the peculiar suckling conditions
attendant on premature birth—it appears probable that in these forms the reduction of
the last element, namely, the deciduous premolar, has been checked by its becoming
serviceable in the secondary herbivorous evolution.

Foot-structure.—With regard to the foot-structure of the Australian forms it will be
seen that the prototypal characters are distributed over two families. The Dasyuridee
are primitive in the non-syndactylous and unreduced condition of the second and third
digits. The Phalangeride are primitive in the possession of a full complement of well-
developed and transversely striated plantar pads, and of a fully developed and completely
opposable hallux associated with a non-elongated condition of the whole pes. It is
apparent that the two types respectively characteristic of these two families are deri-
vatives of a common form combining the prototypal characters of both. This common
form is exactly represented by Marmosa (P1.7. tig. 7) among the Didelphyid, with the
partial exception that there is here, in some species, an indication of the syndactylous
condition of the Phalangeridze. Marmosa is completely arboreal, just as the ancestor of
the Australian forms must have been. The conditions in this genus are repeated in
Caluromys. Peramys, the genus which is most primitive in dentition, is slightly
_ Specialized in a terrestrial direction in foot-structure (PI. 7. fig. 1), but, except for the
partial reduction of the plantar pads and a tendency to shorten the fifth digit, retains
the character of a prototypal form. It is an interesting fact that we have here in

192 DR. B. A. BENSLEY ON THE EVOLUTION

Marmosa and Peramys a beginning of separation of persistent arboreal from terrestrial
or semiterrestrial types, just as must have taken place in the original separation of the
Dasyuride from the common ancestors of the Phalangeride, Peramelide, and Noto-
ryctidee. It is also an interesting fact that the larger members of the Didelphyide are
conservative in foot-structure exactly as they are in dentition. Both Metachirus and
Didelphys, the larger terminal forms of this incipient radiation, like Peramys, tend to
depart from arboreal conditions, but in doing so do not become modified to an appreciably
ereater extent from the original arboreal type.

ae PHYLOGENETIC ARRANGEMENT OF THE AUSTRALIAN MARSUPIALS.

In the three preceding sections the attempt has been made to bring together and
classify the data at present available as to the origin and sequence of forms in the
Australian radiation, by reference to the characters of the dentition and foot-structure,
the latter being considered for the most part separately. In the present section a
veneral statement is given of the phylogenetic * conclusions which may be drawn from
their combination.

The accompanying plan, showing the main lines of the radiation, may be consulted as
preliminary to the more detailed statement of generic and specific relationships given below.

Dasyuride. Peramelide. Phalangeride. Macropodide.
\ =
\ ; Bee
Thylacoleontide. A
a
\ s ape in
\ ve Se S
ee
“——~Diprotodontide.
Omn.
Terr.
Tns.-carn.
Arb.-terr.
Notoryctidee Inst
eg alien
S Properamelide (hyp.).
O/}F
é
[a]
Didelphyide: { 1%:
aes Arb.

* Cf. closing paragraph of introductory section.

OF THE AUSTRALIAN MARSUPIALIA. 195

DASYURIDA.

Beginning with the hypothetical stem-form, to which we assign the characters of the
Didelphyidie, we first note the fact that the radiation has proceeded along two primary
lines. One of these, distinguished by its insectivorous-carnivorous dental evolution,
and by its arboreal-terrestrial foot-evolution, in which the primitive eleutherodactylous
condition of the digits has been retained throughout, leads to and culminates in the
present family.

Thylacinus. Sarcophilus.
. Dasyurus,
terr. spec.
a
SS pe:
Dasyurus,
arb. spec.
Dasyuroides.
4
Cheetocercus. ;

\
ve
\ ae.
\ $
Antechinomys. ve Myrmecobius.
NS es

~ . .
Sminthopsis.

9
Phascogale.

As shown by the dentition, the stem-form for the Dasyuridze must have been a small
species of Phascogale. Considering size, dentition, and foot-structure together it is
probable that P. flavipes makes the closest approach to this prototypal form. The
dental evolution of the predominant section of the family, or Dasyurine, has been
referred to as homogeneous and progressive throughout, but the general sequence is
disturbed by the occurrence of successive lines of terrestrial modification in foot-
structure. We are thus obliged to recognize a main insectivorous-carnivorous arboreal
line extending through the species of Phascogale and ending in Dasyurus. Sminthopsis,
which has been referred to as, on the whole, more primitive than Phascogale in dentition,
this remark referring chiefly to the less reduced or wholly unreduced character of the
posterior premolars, in reality represents a terrestrial derivative of that genus, the
various species, beginning with S. Jewcopus, showing successive stages in the development
of a digitigrade type of pes. The small size of the species of Sminthopsis and their
primitive dental characters, together with the transitional type of pes presented by
S. leucopus, distinguish them from the larger terrestrial forms, and establish the fact of

194 DR. B. A. BENSLEY ON THE EVOLUTION

their origin from a diminutive form of Phascogale. Antechinomys is a derivative of
Sminthopsis, since it shows the same dental characters as the latter genus, and a more
advanced stage of the same digitigrade modification of the pes characterizing that genus.
Chetocercus cristicauda represents a terrestrial modification of Phascogale, which
parallels Sminthopsis, but is distinguished by its dental characters and size as a
derivative of a larger species. Dasyuroides represents a more advanced member of the
same terrestrial line leading to Chetocercus. Within the genus Dasywrus we find
D. hallucatus directly continuing the evolution of Phascogale in size, dental characters,
and foot-structure. The advanced carnivorous form J. maculatus represents a further
continuation of the same line and also its culminating point as far as the arboreal
evolution is concerned. D. viverrinus and D. Geoffroyi, which are intermediate between
the two last-named species in dentition, represent terrestrial derivatives in respect to
foot-structure. Of the two species D. viverrinus is the more advanced, the hallux having
entirely disappeared. Sarcophilus, which, in dentition, represents the culminating
stage of the carnivorous evolution of the Dasyurinie, is a terrestrial plantigrade
derivative of Dasyurus. It has no relation with the remaining specialized carnivorous
form Trylacinus.

The foot-structure of Myrmecobius cannot be relied on for evidence of the special
affinity of the animal. A close study of the dentition shows a balance of evidence in
favour of derivation by retrogression from Phascogale rather than any direct relation
with the Mesozoic Mammalia.

Thylacinus, so far as one may judge from the dentition and foot-structure, may be a
derivative of Phascogale; if this is the case, it must be derived from a very primitive
form of the latter genus, since it retains the original proportions of the posterior pre-
molars. The evidence of the dentition, however, favours a relation with the Sparasso-
donta of the South-American Miocene, the most important considerations being, first,
that the dentition is well advanced in the carnivorous evolution, but its characters
are in many respects different from those gradually developed in the Dasyuringe and
perfected in Sarcophilus, so that Thylacinus comes to occupy an isolated position in the
Australian family, and, secondly, that the dentition resembles that of the carnivorous
Sparassodonta in exactly those characters in which it differs from that of Sarcophilus.
This may not be regarded as indicating that the Dasyuride are the descendants of
Sparassodonta through forms like Thylacinus, but that Thylacinus is a foreign element
in the Australian series.

PERAMELIDS.

The line of evolution leading to the Dasyuridie has already been referred to as one of
two primary lines along which the Australian radiation has proceeded. 'The fundamental
characters of the second line include an insectivorous-omnivorous progression in denti-
tion, coupled with a continuation of the arboreal evolution of the stem-form in foot-
structure, up to what has been designated as the second arboreal phase or that in which
syndactylism appears. The arboreal evolution of the pes culminates in the Phalangeridee,

OF THE AUSTRALIAN MARSUPIALIA. 195

but the dental evolution changes from omnivorous to herbivorous, and only culminates
in terrestrial derivatives of that family. The latter include the Macropodide, Diproto-
dontidze, and Phascolomyidie. The present family and the Notoryctide are early
terrestrial derivatives of the same general line, in which the dental evolution has
been arrested at the omnivorous or insectivorous stage, thence proceeding in special
directions.

The relations of the Peramelidee are shown by the following characters :—(a) the
retention of the polyprotodont modification of the antemolar teeth; (0) the development
of omnivorous modifications of the molars; (¢) the development of terrestrial, and more
or less digitigrade, modifications of the pes, the latter being otherwise of an arboreal

Thylacomys. Cheeropus. P. obesula.
P. Gunni.
P. macrura.
ae
Ze
P. nasuta. ey
P. Bougainviller. P. moresbyensis.
ed ee
=~
P, Cockerelli. a
\ we
ws
He #
-_____P, Raffrayana.

Properamelidx
(Arb. prototypes).

phalangerine type. The first two characters indicate a general ancestral relation of the
Peramelidze to the remaining members of the main line, with the partial exception of
the Notoryctidse, while the third is a primary character of divergence. The presumed
common ancestors of the Peramelide and Phalangeridz are designated in the general
plan (p. 192) as Properamelidie (hyp.).

The present family is notable for the paucity of characters of generic rank. Its
special evolution largely concerns the distribution of the species of Perameles. The
two forms Thylacomys and Cheropus are independent derivatives of a form not far
removed from Perameles Bougainvillet.

- No member of the family combines the characters of a stem-form. P. Doreyana and
P. Raffrayana are prototypal in their molar patterns, which are only slightly specialized
in an omnivorous direction from the insectivorous type presented by the Didelphyidse
and smaller Dasyurine. P. Cockerelli and P. Doreyana are prototypal in the proportion
SECOND SERIES.—ZOOLOGY, VOL. IX. 27

196 DR. B. A. BENSLEY ON THE EVOLUTION

of the pes, which shows on the whole little departure from the phalangerine type. Both
species are specialized in the reduction of the upper incisor formula and in the reduction
of the plantar pads. P. Cockerelli is more specialized than P. Doreyana in its molar
patterns. P. Raffrayana is not far removed from the stem-form of all the Australian
mainland species, as shown by its retention of a full upper incisor formula, in common
with all of the latter, retention of prototypal characters in the molars, retention of more
primitive proportions of the pes and of plantar pads.

The evidence of the dentition and foot-structure will be seen to favour the view of a
Papuan origin of the family, the three forms P. Doreyana, P. Raffrayana, and P. Cockerelli,
possessing the most primitive characters, being confined to New Guinea, while all the
advanced forms are Australian.

Thomas (1888) has remarked that the Australian species fall into two series typified by
P. obesula and P. Gunni respectively. These divisions indicate two lines of development,
giving rise respectively to short-and long-footed types. The former includes P. obesula,
with its dwarfed relative P. barrowensis, and P. macrura, the latter the remaining species
of Perameles, together with Thylacomys and Cheropus. The short-footed forms are,
speaking generally, also distinguished by the retention of the original shortened condition
of the muzzle, in which character they are even more primitive than P. Raffrayana, and
by the possession of highly specialized molar patterns, the latter being of the most
advanced type found in the family.

Assuming a Papuan origin of the Payne the question arises as to whether the
Australian mainland species are derivatives of a single stem-form or are diphyletic.
This question is not easy to decide because of the lack of adherence to type which is so
conspicuous a feature of the whole family. The Papuan species P. moresbyensis is, as
pointed out by Thomas (1888), closely allied to the North- Australian P. macrura. It
represents a very slight advance on P. Raffrayana in foot-structure, and a very decided
advance in its molar patterns, the latter being almost as specialized as those of P. obeswla.
P. macrura is not, as might be expected, prototypal to the southern and Tasmanian
P. obesula, but is more specialized both in dentition and foot-structure. The long-footed
series must be traced to a form combining the dental characters of P. Bougainvillet with
the pedal characters of P. nasuta. ‘The presence in New Guinea of P. moresbyensis with
affinities with P. macrura and P. obesula appears at first sight to indicate that there has
heen a differentiation in New Guinea of two stem-forms, one combining a primitive type
of dentition with a slightly elongated type of pes, the other combining a specialized type
of dentition with retention of ae original shortened type of pes. It is much more
probabie, however, that P. moresbyensis has arisen independently of the Australian
forms P. macrura and P. obesula, and has become specialized in dentition. If this is
the case, it bears the same relation to P. Raffrayana as P. Cockerelli does to P. Doreyana.
P. Cockerelli, it will be observed, is also independently specialized in its molar patterns.
There is a remote possibility that the presence of P. moresbyensis in New Guinea
is secondary.

Of the mainland species P. macrura may be safely regarded as a more specialized
offshoot of the same form which gave rise to P. obesula. The distribution of the two

OF THE AUSTRALIAN MARSUPIALIA. 197

series does not indicate that P. odesula is an original or ancestral form. P. nasuta,
although recognized as a member of the long-footed series, is not far removed in the
proportions of the pes from P. obesula. It is more primitive in foot-structure than its
ally P. Gunni, with which it agrees in dentition. Both forms are more primitive than
P. Bougainvillet in foot-structure, but more specialized in dentition. P. Gunni is
undoubtedly a derivative of P. nasuta, and in this relation it is interesting to note the
typically Tasmanian distribution of the latter and the South-east Australian distribution
of the former.

P. Bougainvillet is probably the ancestral form of both Thylacomys and Cheropus.
The chief derivative features of Thylacomys are the further digitigrade development of
the pes and the peculiar elaboration of the molars, in which an incipient hypocone, like
that of P. Bougainvillet, has-been obliterated by displacement inwards of the metacone.
It is interesting to compare the distribution of the two forms: P. Bougainvillei is
West, South, and South-east Australian, and Thylacomys West, South, and Central
Australian.

Cheropus is the most specialized member of the family in foot-structure, and presents
a great advance in this respect on Thylacomys, which comes second. Cheropus is,
however, to be regarded as an independent derivative of P. Bougainviilei, since it does
not show the peculiar type of molar found in 7hylacomys. It is interesting to note
again the distribution, which includes, in this case, West and South Australia, with
Western New South Weles and Victoria.

NOTORYCTID.

The different opinions as to the affinities of Motoryctes have already been mentioned.
The available evidence is decidedly in favour of Dollo’s * suggestion of a relationship with
the syndactylous Peramelidee. In the general plan (p. 192) Notoryctes is placed as a
terrestrial derivative of the Properamelidze (hyp.), the presumed arboreal ancestors of
the Peramelidee and Phalangeridz. The main facts determining this position are as
follows :—(a) Excluding Noloryctes, the primary lines of the Australian radiation may
be defined as insectivorous-carnivorous and insectivorous-omnivorous in their composition,
and may be distinguished by the respective absence and presence of a hypocone in the
upper molars. Votoryctes is too far advanced in a special insectivorous reduction of the
molars to show definite resemblances with either series. (%) The two main lines may also
he distinguished by the fact that they are made up respectively of eleutherodactylous and
syndactylous types, Noloryctes showing definite affinities with the latter. (¢) Unlike all
the remaining members of the syndactylous line, with the exception of the Paramelide,
Jotoryctes has retained the polyprotodont modification of the antemolar teeth, indicating

* “Caracteres communes aux Peramelide et aux Notoryctide, polyprotodontie, prédominance du quatriéme
orteil, réduction du deuxiéme et du troisiéme, syndactylie de ceux-ci, entocun¢iforme allongé, caisse tympanique
volumineuse, rotule osseuse, poche s’ouvrant en arri¢re. Ces deux familles proviendraient-elles d'un ancétre
commun?” (Dollo, 1899.)

27*

-~

198 DR. B. A. BENSLEY ON THE EVOLUTION

that its divergence as a terrestrial type must have taken place, as in the case of the latter
family, at a comparatively early stage. (d) Its Central-Australian range and the
probability of a Papuan origin of the Peramelidee do not favour the view of a very
intimate relationship with the latter family.

PHALANGERID.

The phylogenetic position of this family may be estimated as follows:—-(a) By the
retention in foot-structure of the second arboreal stage through which all the remaining
members of the second or syndactylous line have passed ; (0) by the appearance of the
diprotodont modification of the antemolar teeth; (c) by the loss of the insectivorous
characters of the molar cusps; (d) by the retention of a number of dental characters
ancestral to those of the Macropodidze, Diprotodontidse, and Phascolomyide. In respect
to character @ the family is prototypal. In characters } and ¢ it is more advanced
than the Peramelidze and Notoryctide.

Phascolarctus. : Trichosurus.

Phalanger.
Dactylopsila.
Petauroides.

Petaurus.
Pseudochirus.

Nt Gymnobelideus.

x 7

< Disteechurus. Dromicia,

\ Aerobates.
—

Le Tarsipes.

|
Properamelide (hyp.).

The Phalangeridee are the derivatives of minute insectivorous or incipient omnivorous
prototypes combining the dental characters of the more primitive of the Peramelide with
the type of pes at present characteristic of the group. The diprotodont modification was
developed in these forms as an insectivorous adaptation primarily affecting the median
lower incisors.

OF THE AUSTRALIAN MARSUPIALIA. 199

The foot-structure is so homogeneous throughout the family that it may practically be
neglected in making a phylogenetic arrangement of the genera. Evolution has proceeded
along two main lines leading respectively to the Phalangerinze and Phascolarctine, the
chief distinctive features of these being the development of selenodont and bunodont
modifications of the molars. A third subsidiary line, represented by Tarsipes, is
characterized by dental retrogression and the development of a more advanced stage of
arboreal elaboration in the pes than is found in the remaining members of the family.

The facts of the dentition show that the phascolarctine line cannot be derived from
the phalangerine one, or vice versd. In the former the most primitive form is
Pseudochirus. Petauroides, which agrees with Pseudochirus in dentition, must be
placed as a derivative of that genus on account of its volant character. Phascolarctus,
the terminal form, represents, in all its dental characters, a direct advance on
Pseudochirus *.

In the Phalangerine the most primitive members are Acrobates, Distachurus, and
Dromicia, these forms showing in some respects an interesting approximation to an
insectivorous prototype. Acrobates and Distechurus are more closely related to one
another than either of them is to Yromicia. They are derivatives of a common stem-
form, from which they have become specialized by the reduction of the posterior
premolars; they have diverged through the development of Acrobates into volant
animals. In the extent of reduction of the posterior premolars, and in the development
of special insectivorous characters in the lower median premolars and first lower molars,
Distechurus represents a more advanced form. Dromicia shows a more marked tendency
towards the omnivorous modifications characteristic of the larger specialized genera.
The volant form Petaurus occupies a derived position with reference to Dromicia, so
that its non-volant relative Gymmnobelideus, which is said by Thomas to be identical with
it in dentition, is probably a derivative of that genus. Dactylopsila represents a direct
advance in dentition on Petawrus, and is to be regarded as another derivative of
Gymnobelideus. 'The derivation of the genus Phalanger is a somewhat difficult matter.
As noted above, it represents a direct advance on Dactylopsila in the reduction of the
vestigial teeth, especially the median upper premolars, and also in the development of
incipient herbivorous characters in the molars. On the other hand, it departs from both
Dactylopsila and Petaurus in the abrupt substitution of herbivorous for insectivorous
characters in the incisors, and of well-developed sectorial premolars, somewhat like those
of Dromicia nana, for vestigial ones. Zrichosurus is plainly a slightly more specialized
form of Phalanger. ‘The fossil form Burramys (Broom, 1896) should probably be
included in the Phalangerinie, on account of the lack of development of the median pre-
molars, which removes it from the Bettongiinse. The characters of the sectorial premolars

* In connection with the present interpretation of the affinities of Phascolarctus it is interesting to note that in
the structure of the manus the animal agrees with Psewdochirus, and differs from all the remaining members of
the family, the first and second digits being arranged so as to oppose the third, fourth, and fifth. The peculiar
straightening out of the normally inflected angle of the jaw, which is so noteworthy a feature of Phascolarctus, is also
indicated in Pseudochirus Cooki, and to a certain extent in some other species. Other species of Pseudochirus,
such as P, Albvertist and P. Corinne, have the angle fully inflected as in normal phalangerine forms.

200 DR. B. A. BENSLEY ON THE EVOLUTION

offer a slight suggestion of special relationship between Burramys, the Bettongiinee, and
Phalanger.

Spencer has recently described under the designation Wyniardia bassiana a fossil
Marsupial from the Table Cape Beds of Tasmania, in which he identifies characters of
resemblance to both polyprotodont and diprotodont sections, and which he states “may be
regarded as intermediate between the former and the latter, and as indicative of a stage
in the development of the Australian Marsupials when the ancestors of the recent
Diprotodontia were beginning to diverge from the original polyprotodont stock, from
which they have been developed within the. limits of the Australian region.” This view
is based on a detailed study of the parts preserved, which unfortunately include neither
the dentition nor the front or hind feet. While it would be difficuit to add to the
excellent comparisons presented by Spencer, it is probable that the reference to the
animal as an intermediate form must be excluded, if for no other reason, on account of
its large size. Throughout the present paper the effort has been made to show that the
actual prototypal or central forms must have been comparatively minute insectivorous
animals. In the Didelphyidee, which occupy an ancestral position to all of the Australian
Marsupials, we find that the most primitive forms are the minute animals belonging
to the genera Marmosa and Peramys. The species of the Oligocene Peratherium, which
is closely related to the preceding genera, and probably the actual ancestor of the
Australian series, were animals of small size. The most primitive members of the
Dasyuridee, and even of the Phalangerida, are of like proportions. Thus all the
forms which approach the hypothetical intermediate type are of small size. Not only
this, but the diprotodont modification itself may, as already explained, be shown to
represent an insectivorous adaptation which could only have taken place in comparatively
small animals. The relations of Wyniardia are more probably with one of the advanced
genera, such as Pseudochirus, Phascolarctus, or Phalanger.

MACROPODID Ai.

With regard to the relationships of the Potoroine and Bettongiine it has been
mentioned above that the modifications of the pes appear to contradict the general
subfamily division and generic sequence as determined by reference to the dentition,
Hypsiprymnodon showing a closer approach to Potorous than to Betlongia, while
Caloprymnus shows a closer approach to Betéongia than to Potorous. There is no doubt,
however, that the correct plan of division is according to dentition, because tlie divergent
characters of the sectorial premolars, on which the dental division mostly depends, are
of a very fundamental kind. It is impossible to derive the straight and comparatively
unelaborated sectorials of the Potoroinee and Macropodine from the multigrooved
rotated sectorials of Hypsiprymnodon. On the other hand, while Beétongia is widely
separated from Hypsiprymnodon in foot-structure, it must have passed through a
HHypsiprymnodon-stage in arriving at its present condition. The Potoroinz have also
passed through a Hypsiprymnodon-stage in foot-structure, but this does not prove affinity

OF THE AUSTRALIAN MARSUPIALIA. 201

with that genus, because there was probably, in the phalangerine ancestors of both
forms, as in the existing Phalangeride, several types of dentition, but only one general
type of pes. Potorous has retained a fairly primitive type of pes on account of its more
fossorial and less saltatorial habit.

The connections of the Bettongiinee with the Phalangeride are doubtful, but the
characters of the sectorials afford a slight suggestion of a common origin of three lines
ending in the present division, in Burramys, and in Phalanger. Hypsiprymnodon is
closely allied to, if not itself the actual ancestor of Bettongia. The annectant species is
B. penicillata. Bettongia is further the prototype both in dentition and foot-structure
of Apyprymnus.

Macropus
(Kangaroos).

Macropus
(Large Wallabies).

|

Onychogale. |
Dorcopsis.

Petrogale. Macropus

(Small Wallabies).
Lagorchestes. ae

Dendrolagus.

Lagostrophus.

Epyprymnus. Tay

Bettongia. Caloprymn us.
Potorous.

|

Setonyw.

Tlypsiprynmodon.
SK

.

Primitive Phalangerine.

In the Potoroine, a common stem-form, not far removed from Petaurus or, better,
Gymnobelideus, has given rise to Potorous platyops as a short-limbed fossorial form and
to Caloprymnus as a free plains living cursorial form. ‘The former appears to have
spread from West Australia eastward to Tasmania, giving rise to the remaining species.
Caloprymnus shows the closest approach of all the Potoroinze and Bettongiine to the
Moacropodine.

In the brachyodont section of the Macropodine we have first the genus Dendrolagus,
representing an arboreal derivative of some primitive form rather than a specialized one
such as Macropus. Dorcopsis represents a secondarily terrestrial derivative of Dendro-
lagus, which begins to parallel Macropus. The sequence of modifications of the sectorial
premolars and the retention of certain primitive characters in the general dentition
preclude the possibility of a derivation of Dendrolagus from Macropus through

202 DR. B. A. BENSLEY ON THE EVOLUTION

Dorcopsis. The evolution of the series indicates a migration from Australia to New
Guinea.

The derivation of Dorcopsis is interesting from a general biological standpoint,
inasmuch as there is direct evidence that in the whole course of its evolution it has been
three times terrestrial and twice arboreal. The alternation of arboreal and terrestria
phases may be expressed as follows :—

Arborealifonms)sots ieee Didelphyide —+ Phalangeride ........ Dendrolagus.

wa ANE wee Dorcopsis.

Terrestrial forms.... Marsupial ancestors ...... Primitive Macropodinee ——+ Macropus.

Setonyx does not appear to be nearly related to Dendrolagus and Dorcopsis, although
it agrees with those forms in the peculiar characters of the sectorials and the small
degree of hypsodontism of the molar crown. It is removed partly by its small size and
complete canine reduction, and partly by the purely terrestrial character of the pes; all
the secondarily terrestrial species of the brachyodont line, namely those of Dorcopsis,
are Papuan, while Setonyx is Australian. It is probable that Setonyx is a member of
the Small Wallaby section of the genus Jacropus which has assumed feeding-habits
somewhat similar to those of the tree-living Dendrolagus. The fact that Dorcopsis,
although terrestrial, continues the sectorial evolution begun in Dendrolagus, shows that
such developments do not necessarily begin and end with arboreal life.

The members of the hypsodont series are best arranged as representing a minor
radiation, the divergent characters of the various genera being, for the most part,
associated with differences of environment, and not affecting to any appreciable extent
the dentition or foot-structure. Onychogale, however, appears to be more nearly related
to Macropus than to the remaining genera, and Lagorchestes, of which Lagostrophus is a
modified form, appears to be similarly related to Petrogale. The typical and predominant
forms of this radiation are the species of Jacropus. It is an interesting fact that this
genus, in which the most important stages of the grazing evolution are to be found in
point of numbers of species, stands out as the successful type of the family, just as the
genus Phalanger does in the Phalangerinz, and Pseudochirus in the Phascolarctine.
Within the genus the Small Wallaby, Large Wallaby, and Kangaroo sections represent
successive stages of the same graminivorous and saltatorial evolution. The predecessors
of the Small Wallabies appear to have undergone a distributional differentiation with
the production of the various species comprising the section. The latter appear to have
given rise locally to Large Wallabies, and these in turn to Kangaroos. This sequence is
indicated, first, by the obvious affinities and wide distribution of the Small Wallabies;
secondly, by the fact that it is possible to differentiate between small and large premolared
species, especially in the two lower sections ; and, thirdly, that there are indications of
relationships between certain members of the Small and Large Wallaby sections having
similar geographical ranges. Thus the West-Australian Small Wallaby (Macropus
Lugenit) is prototypal to the Large Wallaby (J/. ima) of the same region, and also to

_
cial re ss Fn

OF THE AUSTRALIAN MARSUPIALIA. 203

the more southern J. Greyi, and a similar relation is observable between JZ. Bedfordi
and MW. dorsalis. The existing Kangaroos are apparently the descendants of successively
small-premolared forms.

THY LACOLEONTID.

The evidence available concerning the phylogenetic position of Thylacoleo is very
incomplete, partly on account of the absence of annectant types, and partly because
of the lack of information as to the foot-structure. ‘The dentition furnishes indications
of two lines of development—one of normal phalaugerid type, represented by the
diprotodont modification, involving enlargement of the median incisors and reduction of
the posterior antemolar teeth, and by the enlargement of the posterior premolars as
sectorials; the other of a special carnivorous type, represented by the piercing develop-
ment of the median incisors, the reduction of the molars, and the excessive enlargement
and smoothing of the cutting-edges of the sectorials. ‘This evidence does not warrant a
closer estimation of the relations of the animal than that it is a derivative of the
Phalangeridze, which, unlike all the remaining derivatives of that family, has, after
the passage of the omnivorous stage, or even during the herbivorous stage, become
varnivorous instead of continuing the herbivorous evolution in the normal way.

PHASCOLOMYIDA anp DIPROTODONTID &.

Thsyailable evidence concerning the phylogenetic position of these families may be
briefly's, mmed up as follows:—It is possible to recognize, both in the dentition and
foot-strt\nre, three kinds of characters—primary ones of phalangerine affinity, secondary
ones of \yity with one another, and characters of divergence. The first refer to the
ee of the dentition and the bunodont origin of the molars, and, in
the foot-str ure, the reduction of the second and third digits, enlargement of the
fourth, and osability of the hallux. ‘The second refer to the rodent modifications of
the incisors, th. duction of the premolars to the posterior pair, which present somewhat
similar patterns q the terrestrial plantigrade modification of the pes. 'The divergent
characters refer
presence of lopho
side of the pes, wit
mesocuneiform with
internal balancer. In
and the molars have
bunodont. There has b
shifting of the walking a3
normal or less differentiate

e retention in the Diprotodontide of the upper lateral incisors, the
rooted molars, and the shifting of the walking axis to the outer
argement of the astragalus, caleaneum, and cuboid, fusion of the
entocuneiform, and functional retention of the hallux as an
Phascolomyidee the upper lateral incisors have been obliterated,
e rootless and rodent in their character while remaining
re-development of the second and third digits by which the
the pes has been avoided. The latter is thus of a more
e. The entocuneiform has remained distinct from the

i _ The halla
Saeed : Bas not been modified as a balancer, but has been reduced
i yay as in othe é
in the same way Asal dorms:

SECOND SERIES.—ZOOLOGY, \,

204 DR. B. A. BENSLEY ON THE EVOLUTION

These relations point to the origin of the Phascolomyide and Diprotodontidee from a
common stem-form, which is in turn traceable to the Phalangerinz. The absence of
annectant types makes it impossible to fix the affinities of the two families with any

existing phalangerine genera.

THE QUESTION OF THE TIME AND MODE OF ORIGIN OF THE AUSTRALIAN RADIATION.

Partial reference has already been made to the views of different writers as to the
time and conditions under which the Australian fauna originated in dealing with the
identification of the stem-form.

With regard to the direction from which the ancestral forms entered Australia there
appear to be only two possibilities—that they came either from the North, as supposed
by Wallace (1875) and Lydekker (1896), or from South America, as thought by Ameghino
(1891) and Spencer (1896).

According to Lydekker’s view the Dasyuridze and Didelphyide may have originated
from common ancestors inhabiting South-eastern Asia, from which region the former
family migrated into Australia, while the latter one dispersed in two directions into
Europe and North America. In advocating a South-American origin of the family
Spencer takes exception to this view on the general grounds of the paucity of pol
protodont types in New Guinea, through which the ancestral forms would have past
the difficulty of explaining the absence of Didelphyide in Australia, and the la
evidence concerning the former existence of Marsupials in Asia.

With reference to the first of these objections, it may be pointed out that the
does not concern so much the relative numbers of polyprotodont forms in Net none
and Australia, since, in view of the greater area and greater possibilities of dis So, hapa
differentiation of the latter region, we may expect to find, as we actuall“
contains a greatly predominating portion of the polyprotodont fauna. T” question

: .- pay be faul
rather whether or not New Guinea contains prototypal forms. ‘bef _ —_—

iestion
Guinea

answered in the affirmative. On referring to the Dasyuride we find tha, = eae
the specialized forms Phascogale Dorie, P. dorsalis, P. Wallacei, and ; occurs shee
which are confined to the Papuan region, the prototypal form P. flavi, Peramelidsel
well as in the eastern and south-eastern parts of Australia. Of three genora.of Ale
the species presenting prototypal characters are Papuan. Of tly, puan represen neon
Phalangerinz which present pr imitive characters, Dromicia has hon fined to that region.
(D. caudata), Acrobates one (A. pulchellus), while Distechur us , New Guineal anti
Tt is true that migrations may have taken place from Australig
factor should be kept in mind as bearing on the question, be
such migration in the distribution of Phalanger, Dorcopsis, an
Small Wallabies. /

Spencer’s further objection as to the lack of evidence éo
of Marsupials in Asia is, of course, unanswerable. We have
community between Asia and North America, poirtg

'/

use there is evidence of
o a certain extent of the

rning the former presence
y the indications of faunal
intermigration over what is

OF THE AUSTRALIAN MARSUPIALIA. 205

now Behring Straits, and the occurrence of fossil opossums in the Oligocene rocks of
Europe and North America.

The difficulty as to the absence of Didelphyidze in Australia presents itself whether we
assume a northern or a South-American origin. Spencer explains the condition by
assuming that the land-connection between South America and Australia was broken at a
time soon after the ancestral forms of the Australian Marsupials had passed across, and
while the Didelphyidze were being developed in the more northerly portion of South
America. If this explanation be the correct one, we may ask why the form Thylacinus,
whose affinities are decidedly with the South-American Sparassodonta, gained access
to the Australian region, while the Microbiotheriidze, some of which at least present
the characters of the Didelphyidee and are found fossil in the same formations, were
excluded. The same question might be asked of the Edentate fauna of those formations.
There is also the difficulty of recognizing a definite period of rupture of the antarctic
connection and that of accounting for the origin of the Didelphyid of the northern
hemisphere.

In a former paper the writer referred to the difficulty presented by the absence of the
Didelphyidz in Australia as an apparent one, due to the recognition of the family as a
modern derived group. Weare accustomed to look upon the Didelphyide not only as
contemporaries of the Australian Marsupials, but also as forms possessing fixed family
characters. When we consider, first, that the modern family represents an exceedingly
plastic group, the members of which are at the present time undergoing an incipient
radiation ; secondly, that the structural differences separating the family from the
Dasyuridz and the presumed common ancestors of the Peramelidze and Phalangeridee are
extremely slight; and, thirdly, that the ancestors of the Australian fauna on entering
that region must have found themselves under conditions most favourable for differential
development in a place offering diverse and unoccupied sources of food-supply—it is not
difficult to conceive that the Didelphyidz in establishing the foundation of an extensive
radiation may have thrown aside their original didelphyid characters. It seems
preferable to believe that the Didelphyidee were formerly present in Australia, as well as
Europe and North and South America.

The evidence in favour of a South-American origin of the Australian fauna is bused
partly on the general faunal evidence of an antarctic land-connection, concerning which
there seems to be little doubt, and partly on the presumed special resemblances between
the Australian fauna and that of the Patagonian Miocene. Fuller reference to these
resemblances will be made in the next section, but it may be mentioned at this point that
those occurring between the diprotodont forms of both series are of too general a kind to
be interpreted as indicating more than a parallel development from common ancestral

_ types, and that those between Thylacinus and the Sparassodonta do not indicate a South-
American origin of the Australian fauna, since there are no definite reasons for believing
Thylacinus to have a special affinity with the Dasyuride, and it has not the slightest
resemblance to a prototypal form. The evidence of a connection of the Patagonian with
the Australian forms at present limits itself to the possibility that Zhylacinus may have
migrated from South America, and a further possibility of a former distribution of
28*

206 DR. B. A. BENSLEY ON THE EVOLUTION

Didelphyide across the Antarctic continent with centres of radiation in Australia and
South America.

It may be mentioned in passing that if an antarctic connection existed there is
evidence from the Marsupials that it did not include Tasmania, because the fauna of that
region is made up for the most part of specialized forms having identical or prototypal
representatives on the mainland. In the Dasyuride we have Phascogale minima, a
typically Tasmanian form, and in P. Swainsoni a closely related form inhabiting
Tasmania and the adjacent portion of the mainland. While both of these forms are
fairly primitive, they are specialized as compared with the Papuan and Australian form
P. flavipes. Of the species of Dasyurus, two are common to East Australia and Tasmania.
One of them, D. maculatus, is the most specialized member of the Dasyuridze, with the
exception of Sarcophilus and Thylacinus. The remaining form, P. viverrinus, is only
slightly less specialized. The actually primitive species, D. hallucatus and D. albo-
punctatus, are North Australian and Papuan respectively. Nothing can be inferred from
the present Tasmanian distribution of Sarcophilus and Thylacinus, since both are found
fossil on the mainland.

Passing to the Peramelidee we find two Tasmanian forms, P. Guwani and P. obesula.
The former species is easily shown to be a specialized derivative of the East-Australian
P. nasuta, while the latter, itself one of the most specialized members of the family,
enjoys a wide distribution on the mainland.

In the Phalangeridze only one of the three primitive genera, namely Dromiécia,
possesses Tasmanian representatives. These are D. lepida, a fairly primitive type,
possibly ancestral, in some respects at least, to Gymnobelideus, and D. nana, which is
not only highly specialized but is also not typically Tasmanian, since remains of it have
been found in cave-deposits of New South Wales.

Passing to the Potoroine division of the Macropodidee, we find in the distribution
of the species of Potorous a clear case of migration from Australia to Tasmania.
Proceeding from the West-Australian P. platyops we find a progressive specialization
passing through the South-Australian P. Gilberti, the New South Wales form P. ¢ridac-
tylus, to the Tasmanian P. “ apicalis.”

These facts are suggestive of a comparatively recent derivation of the Tasmanian fauna
from forms inhabiting the adjacent portions of the mainland.

As to the time at which the ancestral forms of Marsupials gained access to Australia,
the most diverse opinions have been expressed. Owen supposed the group to have been
present there in Mesozoic times, while Wallace estimated the time of entry as Jurassic.
According to Spencer’s view it was Cretaceous, while in Lydekker’s opinion it was early
Eocene. While it is interesting to notice in these opinions an increasing appreciation of
the newness of the Australian fauna, it is probable that even the lowest estimate of the
duration of their evolution is still much too large. In the first place, on comparing the
Australian radiation with the general radiation of Placentals, we note the fact that the
former is in a backward stage of development. For example, the differentiation has not
proceeded beyond the production of families, although, as recognized by several writers,
these families have the potential value of placental orders. In the existing didelphyid

—

OF THE AUSTRALIAN MARSUPIALIA. 207

radiation, which is of still more recent origin, differentiation has not proceeded beyond the
production of genera. Furthermore, in some cases, even the family differentiation has
proceeded to such a meagre extent that family divisions are based on comparatively
trivial characters. In fact, the Marsupials present much the same composition as must
have existed at a very early stage of the placental radiation. The Australian radiation,
confined as it has been, has not proceeded to the stage of very great specialization or to
over-population, with the obliteration of less specialized intermediate types. We have
such cases as that of I/yrmecobius, in which, while dental reduction has taken place as a
result of ant-eating habits, it has not proceeded to the stage of total obliteration of the
teeth, as in placental ant-eating forms. All these facts indicate that the Australian
radiation is of comparatively recent origin, and if, as appears probable from the central
position of the Creodonta, the placental radiation began in the early Eocene or late
Cretaceous, it seems unlikely that the marsupial radiation could have begun until well
on into the middle of the Tertiary period. To this structural evidence we may add the
final fact that the Didelphyidee are the ancestral forms of the Australian radiation, and
that they are typically Oligocene forms.

THE Masor CLASSIFICATION OF THE MARSUPIALS IN GENERAL.

The recognition of the Dasyuridze as a primary division in the Australian radiation
raises a doubt as to the applicability of Owen’s classification of the Marsupials into
Polyprotodontia and Diprotodontia. The difficulty presented by the Peramelidee, which
while essentially polyprotodont possess a syndactylous type of pes otherwise characteristie
of the diprotodont series, has already been commented upon by Flower, but the relations
of this family have been explained by Thomas on the assumption that their syndactylism
has been independently acquired. |

The diprotodont modification has obviously been derived from a polyprotodont one ; it
is traceable to an insectivorous specialization of the median lower incisors. The members
of the diprotodont section may be shown to have passed through an omnivorous stage, and
to have possessed omnivorous molar characters similar to those of the primitive Pera-
melide. The primitive Phalangeride are to be connected with the Peramelidze not only
as syndactylous, but also as omnivorous forms, rather than separated as Diprotodontia,
while the Dasyuridze are to be separated as eleutherodactylous and _ progressively
carnivorous forms,

In selecting a differential character on which a classification may be based we are
obliged to turn to other structures than the dentition. The absence of a definite hypocone
in the upper molars of the Dasyuridee would furnish exactly the kind of character
required, since the addition of this element is highly characteristic of the omnivorous
evolution, were it not for the aberrant developments presented by Notoryctes and
Thylacomys. Turning to the foot-structure, we find in the respectively eleutherodactylous
and syndactylous condition of the pes a differential character of suflicient importance
and applicability. If by a major classification we intend to designate primary lines of

208 DR. B. A. BENSLEY ON THE EVOLUTION

descent, the division should not be into Diprotodontia and Polyprotodontia, but rather
into Syndactyla and Diadactyla.

Owen’s divisions of the Polyprotodontia naturally included the Didelphyide as well as
the Australian Dasyuride and Peramelide. Such a group as the Didelphyide could
have no place in the above classification on account of the fact that its component genera
present both syndactylous and eleutherodactylous types of foot-structure. The Epanor-
thidee present similar difficulties both with reference to this and to Owen’s classification.
In commenting on the systematic position of Canolestes, the living representative of the

Jpanorthidee, Thomas remarks: “ As to the general classification of the Marsupials, a
subject already difficult in view of the puzzling possession by the Peramelide of
polyprotodonty combined with syndactyly, Cenolestes apparently only adds to the difficulty,
being non-syndactylous like most polyprotodonts, while it has by dentition nothing
whatever to do with them. If anything, however, this fact tends to confirm the tentative
opinion expressed in the ‘Catalogue of Marsupials,’ that the primary division of the
order should be by dentition and that syndactyly isa secondary character. Were syndactyly
the primary character, the Epanorthidee would be thrown with the Didelphyid with
which they have clearly nothing to do, and separated from what appear to be their
nearest allies, the Phalangeridee.”

Difficulties such as these are only apparent, arising from the attempt to apply the
same principles of classification to isolated and independently evolved groups. In the
first place, how have the various marsupial faunas arisen? As to the origin of the
Didelphiyide, the evidence is wholly unsatisfactory. They may have come from a spur of
an earlier Jurassic radiation, or they may have been themselves the original stock of the
marsupial division. As to the subsequent history of the family, the evidence is more
complete. They enjoyed a wide distribution in the northern hemisphere during the
Oligocene period, and were probably present in South America as well. The Australian
fauna shows indications of didelphyid origin. The existing Didelphyidze of South America
represent a minor radiation, proceeding from forms approximating closely, so far as may
be judged from the dentition, to Peratheriwm. It is extremely probable that at least
some of the Miocene Microbiotheriide of South America were in reality Didelphyide,
and were the ancestors of the Epanorthidze and their allies and of the Sparassodonta. In
other words, the derivation of the different marsupial faunas is the history of successive
radiations of Didelphyide. Now it will be apparent that it is impossible to apply the
same principles of classification to two geographically isolated faunas unless their
resemblances are of such a nature as to lead us to suppose that two or more differentiated
portions of one of them represent migrated portions of the other. In the case of the
marsupial faunas we have to ask whether the resemblances between them can be traced
to such migration or whether they are the result of convergent development.

There appears to be no possibility of a connection of the existing didelphyid radiation
with the Australian one, the former being much too modern. We may take advantage
of this to point out certain resemblances as due to convergent development. We notice
especially that the Didelphyide parallel the Dasyuride in the tendency towards reduction
of the posterior premolars. Certain of them parallel the Phalangeride in the develop-

7e

OF THE AUSTRALIAN MARSUPIALIA. 209

ment of syndactylous modifications of the pes, and Calwromys shows signs of the
parallel development of omnivorous characters in the molars as in the latter family.
These resemblances are not indicative of affinity except in the broader sense that
they possibly imply a similar potential of evolution carried over from common
ancestors.

The case of the South-American Miocene and Australian groups is admittedly difficult
of discussion on account of the lack of definite information as to the primary differential
characters of the former. The dental resemblances of Thylacinus to the Sparassodonta
are sufficiently close to warrant a belief in their common origin. Even so, however,
there is no evidence that the evolution of Thylacinus has been connected in any way
with that of the Dasyuridee, the reverse being indicated by the fact that the various
forms of the Dasyuridve present sunecessive stages of a dental evolution, which, while
carnivorous in its character as in Thylacinus, is otherwise of a totally different facies.

The fact that the diprotodont modification of the dentition occurs in the South-
American as well as in the Australian forms has been regarded as a mark of affinity.
As indicated above, even the most primitive forms of the existing Phalangeridz are
well removed as regards dentition from the condition which must have obtained at the
time of introduction of the diprotodont modification. Furthermore, the origin of the
peculiarly grooved sectorials of the Bettongiine is not illustrated in the Australian series.
In view of the presence of such gaps in the series, we could not wish to deny the possi-
bility that the prototypes of these forms may yet be found in the South-American group.
The mere fact of the presence of diprotodont modifications in each series, however,
means nothing. ‘The development of a diprotodont modification has taken place in the
Soricidze, in the Redentia, and in the Multitubereulata. The bunodont molars of the
Epanorthidze resemble those of the Phalangeridze, but such teeth have been developed
in the Condylarthra and the Primates. If convergent development may occur between
groups which have little in common beyond their mammalian character, how much
greater is the chance of convergent development in smaller groups of common parentage ?
If resemblances between the Australian and the South-American groups are to be
recognized as indicative of affinity they must not be of a broad general kind such as those
already pointed out between the Epanorthidz and their allies and the Australian dipro-
todonts, but must be of much more special application. The forms presenting them
should not differ in a greater degree from one another than do the successively specialized
genera of the Australian families. For all we are able to say at present, the South-
American radiation may have proceeded on general lines of polyprotodonty and
diprotodonty or on some other unrelated character.

If the above interpretations are correct, how are the Marsupials to be classified ?
The farther the identification of intermediate types proceeds, the more difficult becomes
the systematic classification to which we are accustomed. We are now placed in the
somewhat paradoxical position of attempting to recognize and characterize concrete
groups which our knowledge of evolutionary sequence tells us could have had no
separate existence. If natural classification recognizes lines of descent, major classifica-
tion must recognize radiations, and distinctions of time and geographical distribution

210

DR. B. A. BENSLEY ON THE EVOLUTION

are as valuable as those of anatomical structure. From this point of view the Marsupials

may be arranged as follows :—

Te

©

ws

Primary marsupial radiation—Arboreal derivatives of Metatheria. Jurassic ? Arctogzeic.
Jurassic Mammalia in part. (‘Lhe identification of this group is wholly problematical.)

Arctogeic and Neogaic didelphyid radiation.—Derivative of 71. Oligocene—Lower Miocene.
Didelphyidz, Microbiotheriide ? in part.

Notogeic radiation.— Derivative of 2. ? Mid-Tertiary—recent. Dasyuride, Peramelidie,
Phalangeride, &ce.

First Neogeic radiation.—Derivative of 2. Lower Miocene—recent (Cenolestes). Prothylacinidie
(Sparassodonta), Epanorthide, Decastid, Abderitidie, &c.

Second Neogeic radiation.—Derivative of 2. ? Pliocene—recent. Existing Didelphyidee.

Tuer Systematic ARRANGEMENT OF THE AUSTRALIAN FAMILIES.

In the case of the Australian radiation the systematic arrangement which most
nearly expresses the lines of family derivation is as follows :—

A. Diadactyla (second and third digits of the pes separate).

Incisor formula = upper molars with no postero-internal cusp (hypocone).

Elalluxsreducedior absent: 3.4...) 2) Geteussm Selon Ce ee D ASMOR EDIE

3. Syndactyla (second and third digits of the pes conjoined).

a,

db.

Dentition polyprotodont.

Tncisor formula = posterior premolars well developed. Upper molars

with paracone and metacone separate, witn hypocone (except Thylacomys).
Lower molars with well-developed talonid. Pes terrestrial, subdigitigrade ;
hallux vestigial or absents., «9. ../94GRiieat © 078 NE ae Pe: nt PEA

Incisor formula = posterior premolars absent. Upper molars with con-

joined paracone and metacone; no hypocone. Lower vestigial talonid.
Pes terrestrial, plantigrade ; hallux well developed, not opposable, with
secondary claw). .«. «* |G.) fplet Urea ellie Sil evies oe 1g HNomG@R yuma

Dentition diprotodont.

Pes arboreal; hallux well developed, fully opposable. Antemolar dentition
primitive. Anterior upper premolars always, and vestigial intermediate
teeth usually present.

Functional incisor formula e Median lower incisors piercing or
trenchant. Molars rooted, brachyodont, quadrituberculate (except in
Tarsipes), with bunoid or selenoid cusps 3? eee PHALANGERIDA.

Pes terrestrial ; hallux not definitely opposable, reduced or absent. Ante-
molar dentition specialized. Anterior upper premolars and vestigial inter-
mediate teeth absent.

Pes plantigrade. Incisors scalpriform. Median premolars absent.
Upper lateral incisors present. Molars hypsilophodont, rooted. Pes

with walkimg-axis external; axial elements of tarsus enlarged.
Hallux retained as an internal balancer. Mesocuneiform fused
with entocuneiform <9. so 6 6 + 5 + wl ee eee DD ROODONTID Rae

ee

OF THE AUSTRALIAN MARSUPIALIA. AL

Upper lateral incisors absent. Molars quadrituberculate, bunodont,
hypsodont, rootless. Pes of normal proportions. Mesocuneiform
free ; hallux vestigial . ai ts? th eet DS ee PHASCOLOMYID®.

Pes digitigrade; incisors trenchant. Median premolars developed as
sectorials in young.
Incisor formula 2 Molars rooted, bunodont, brachybunodont and
quadrituberculate, or lophodont. Sectorial premolars grooved . Macroropip®.
Incerte sedis. Pes? Functional incisor formula = Iucisors piercing.

Sectorial premolars excessively enlarged and smooth-edged. Molars

; PeRsteICWOraOSEN ty i) tg ue ss et st tw ls se ys oC ORBONTEDA:

‘ : M f
The writer takes this opportunity of expressing his indebtedness to the officers and

others of the Geological and Zoological Departments of the British Museum who

rendered him assistance during the preparation of the present paper, more especially to

Dr. Henry Woodward, Dr. Smith Woodward, Mr. Oldfield Thomas, and Mr. Richard

Lydekker.

At an early stage of the work it was the intention to attempt to trace the sequence
of dental and particularly incisor reduction in the Marsupials by reference to embryonic
forms, and while this had ultimately to be abandoned for lack of time, the writer is
indebted to Professor G. B. Howes, of the Royal College of Science, for affording him
access to the laboratory and research material of that institution. He also wishes to
record the kindness and liberality with which the late Mr. Martin F. Woodward placed
his valuable private collection of foetal Marsupials at his disposal.

Final acknowledgment is due to Professor Henry F. Osborn, of Columbia University,

_ New York, for friendly advice and supervision.

BIBLIOGRAPHY.

‘ Igor. Atuen, J. A. A preliminary Study of the North-American Opossums of the Genus Didelphis.

Bull. Am. Mus. Nat. Hist. vol. xiv. art. 11, pp. 149-188.

1891. Amucurino, F. Nuevos restos de mamiferos fésiles decubiertos por Carlos Ameghino en el
eoceno inferior de Ja Patagonia austral. Revista Argentina de Historia Natural, tom. i., Oct.
1891, pp. 289-328.

1894. Barzson, W. Materials for the Study of Variation. 8vo. London.

19014. Brnstry, B. A. A Theory of the Origin and Evolution of the Australian Marsupialia. Am.

a Naturalist, vol. xxxv. no. 412, pp. 245-269.

On the Question of an Arboreal Ancestry of the Marsupialia, and the Interrelationships
of the Mammalian Subclasses. Am. Naturalist, vol. xxxv. no. 410, pp. 117-1388.

1896. Broom, R. Report on a Bone Breccia Deposit near the Wombeyan Caves, N.S.W.; with
} Descriptions of some new Species of Marsupials. Proc. Linn. Soc. N.S.W., part 1,
q April 28th, pls. 6-8.

“1898. —— On the Affinities and Habits of Thylacoleo. Proc. Linn. Soc. N.S.W., part 1,
a April 27th, pp. 57-74.

SECOND SERIES.— ZOOLOGY, VOL. IX. 29

1gord,

DR. B. A. BENSLEY ON THE EVOLUTION

Corr, E. D. The Ancestry and Habits of Thylacoleo. Am. Naturalist, vol. xvi. pp. 520-522.
Tertiary Vertebrata. Report of the U.S. Geol. Surv. of the Territories, vol. iii

Washington.

The Mualtituberculate Monotremes. Am. Naturalist, vol. xxii. p. 12.

Synopsis of the Families of the Vertebrata. Am. Naturalist, vol. xxiii. pp. 849-877.
On the Habits and Affinities of the new Australian Mammal Notoryctes typhlops.
Am. Naturalist, vol. xxvi. pp. 121-128.

Derrenvorr, T. Zur Entwickelungsgeschichte des Zahnsystems der Marsupialier. Zool.

Forschungsreisen in Australien (Semon), Bd. ii. Lief. 2, pp. 245-402. Jena.

De Vis, C. W. On the Phalangistide of the Post-Tertiary Period in Queensland. Proc. Roy.

Soe. Queensl. vol. vi. pp. 105-114.

A Review of the Fossil Jaws of the Macropodide in the Queensland Museum. Proce.
Linn. Soc. N.S.W., vol. x. pp. 75-133.

Dotto, L. Les ancétres des Marsupiaux, étaient-ils arboricoles? Miscellanées Biologiques,
pp. 188-203. Paris.

Le pied du Diprotodon et l’origine arboricole des Marsupiaux. Bull. Scientifique de
la France et de la Belgique, pp. 275-280.

Frowrr, Sir W. H. On the Development and Succession of the Teetn in the Marsupialia.
Phil. Trans. Roy. Soe. vol. 157 (1868), pp. 631-641.

Article “Mammalia.” Encyclopzedia Britannica, 9th ed. vol. xv.

Frower, Sir W. H., and R. Lyprxxer. Mammals Living and Extinct. 8vo. London.

Gapvow, H. On the Systematic Position of Notoryctes typhlops. Proc. Zool. Soc. London,
pp- 361-370.

Gervais, P. Zoologie et Paléontologie frangaises. 2nded. 4to. Paris.

Git, T. Arrangement of the Families of Mammals. 8vo. Washington.

Goutp, J. The Mammals of Australia. 3 vols. London.

Haxrcxer, E. Systematische Phylogenie. 3 Theil. 8vo. Berlin.

Huxtry, T. H. On the Application of the Laws of Evolution to the Arrangement of the
Vertebrata, and more particularly of the Mammalia. Proc. Zool. Soc. London, pp. 649-
662.

Lecuz, W. Beitrige zur Anatomie des Myrmecobius fasciatus. Biol. Foren. (Stockholm) ;
Verh. d. Biol. Vereins in Stockholm, 11. Bd. Nr. 8, pp. 136-154.

Studien tiber die Entwickelung des Zahnsystems bei den Saugethiere. Morph. Jahrb.

Rd. xix. pp. 502-547.

Nachtriige zu Studien tiber die Entwickelung des Zahnsystems bei den Saugethiere.

Morph. Jahrb. xx. Bd. pp. 118-142.

Zur Entwickelungsgeschichte des Zahnsystems der Saugethiere. Bib. Zool. xvii. Heft
Stuttgart.

Lypukker, R. British Museum Catalogue of Fossil Mammalia. Part 5. 8vo. London.

A Handbook to the Marsupialia and Monotremata. (Allen’s Naturalist’s Library.)
12mo. London.

— A Geographical History of Mammals. Cambridge.

— The Dental Formula of the Marsupial and Placental Carnivora. Proc. Zool. Soe.
London, pp. 922-928.

Masor, C. I. Forsyra. On some Miocene Squirrels, with Remarks on the Dentition and
Classification of the Sciurine. Proc. Zool. Soc. London, pp. 179-215.

On the Change and Form of the Teeth in the Centetide, a Malagasy Family of

Insectivora. Ann. & Mag. Nat. Hist. vol. xx. pp. 525-544,

ae

OF THE AUSTRALIAN MARSUPIALIA. ONS

1891. Ocmsy, J.D. Hand-list of Australian Mammals. Australian Museum.
1891. Ossorn, H. F. Fossil Mammals of the Wahsatch and Wind River Beds. 1. Homologies and
Nomenclature of the Mammalian Molar Cusps. Bull. Amer. Mus. Nat. Hist. vol. v.

pp- 84-91.
1899. —— The Origin of Mammals. Amer. Journ, of Sci. vol. vii. pp. 92-96.
| 1902. —— Homoplasy asa Law of Latent or Potential Homology. Am. Naturalist, vol. xxxvi.
; pp. 259-271.

1871. Owen, Sir R. Monograph of the Fossil Mammalia of the Mesozoic Formations. Palaonto-
graphical Soc. London.

1877. ——~ Researches on the Fossil Remains of the Extinct Mammals of Australia, with a Notice
of the Extinct Marsupials of England. (Collected Papers from Phil. Trans. Roy. Soc.)
2 vols. 4to. London.

1879 On Hypsiprymnodon, Ramsay, a Genus indicative of a distinct Family (Pleopodide) in
the Diprotodent Section of the Marsupialia. Trans. Linn. Soc. London, 2nd series, Zool.
vol. 1. pp. 573-582.

1887. On the Skull and Dentition of a Triassic Saurian (Galesaurus planiceps, Ow.). Quart.

Journ. Geol. Soe. vol. xliii. pp. 1-6.

1888. Povtton, E. B. The True Teeth and Horny Plates of Ornithorhynchus. Quart. Journ. Micr.
Sci. vol. xxviii. pp. 9-48.

1892. Rosz, C. Ueber die Zahnentwickelung der Beutelthiere. Anat. Anzeiger, vil. Jahrg. Nr. 19, 20,
pp. 639-650, 693-707.

1896. Szmon, E. Im Australischen Busch. S8vo. Leipzig.

1894. Spencer, B. Preliminary Notice of two new Species of Marsupials from Central Australia.

; Proc. Roy. Soc. Victoria, art. 22, pp. 222-224.

Preliminary Descriptions of certain new Marsupials from Central Australia, together

with Remarks upon the Occurrence and Identity of Phascologale cristicauda. Proc. Roy. Soe.

Vict. art. 2, pp. 5-13.

Report on the Horn Expedition to Central Australia. 8vo. Melbourne.

A Description of Wyniardia bassiana, a Fossil Marsupial from the Tertiary Beds of
Table Cape, Tasmania. Proc. Zool. Soc. London, pp. 776-795.

Srewart, C. On a Specimen of the True Teeth of Ornithorhynchus. Quart. Journ. Micr.
Sci. vol. xxxiii. pp. 229-231.

Stiruine, E.C. Description of a new Genus and Species of Marsupialia, Notoryctes typhlops.
Trans. Roy. Soc. South Australia, vol. xiv. part 1, pp. 154-187.

Sriruine, E.C.,and A.H. C. Zierz. Fossil Remains of Lake Callabonna. Part I. Description
of the Manus and Pes of Diprotodon australis, Owen. Mem. Roy. Soc. South Australia, vol. 1.
part 1. Adelaide.

. Taomas, M. R. O. On the Homologies and Succession of the Teeth in the Dasyuridie, Ke.

Phil. Trans. Roy. Soc. London, vol. 128, pp. 443-462.

. — On the Milk-dentition of the Koala. Proc. Zool. Soc. London, pp. 338, 339.

— British Museum Catalogue of Marsupialia and Monotremata. 8vo. London,

— On the Dentition of Ornithorhynchus. Proc. Roy. Soc. London, vol. xlvi, pp. 126-131.

— On some Mammals from Central Peru. I. Ichthyomys. Proc. Zool. Soc. London,
pp. 333-341.

—— On Micoureus griseus, Desm., with the Description of a new Genus and Species of
-Didelphyide. Ann. & Mag. Nat. Hist. (6) xiv. pp. 184-188.

On some Mammals collected by Mr. Albert Meek on Fergusson Island, D’Entre-

casteaux Group. Novitates Zoologicz, vol. ii. (footnote, p. 165).

29*

214 DR. B. A. BENSLEY ON THE EVOLUTION

1895. Tuomas, M. R. O. On Ceenolestes, a still-existing Survivor of the Epanorthide of Ameghino,
and the Representative of a new Family of Recent Marsupials. Proc. Zool. Soc. London,
pp. 870-878.

1897. Tomes, C. 8S. Note on the Minute Structure of the Teeth of Notoryctes. Proc. Zool. Soe.
London, pp. 409-412.

1876. Watxtace, A. R. The Geographical Distribution of Animals. 2 vols. 8vo. New York.

1836. Warernouse, G.R. Description of a new Genus of Mammiferous Animals from Australia,
belonging probably to the Order Marsupialia. Trans. Zool. Soc. London, vol. ii. (1841)
pp. 149-154.

1846. A Natural History of the Mammalia. Vol. I. Marsupiata. 8vo. London.

1897. Wrtson, J. T., and J. P. Hizi. Observations on the Development and Succession of the Teeth

in Perameles, &c. Quart. Journ. Micr. Sci. vol. xxxix. n. s. pp. 427-588.

1882. Wiyer, H. Om Pattedyrenes Tandskifte, isaer med Hensyn til Taendernes Former. Vid.
Medd. f. d. Naturh. Foren. (Copenhagen), pp. 15-67.

1888.

Jordfunde og nulevende Pungdyr fra Lagoa Santa, Minas Geraes, Brasilien, med Udsigt
over Pungdyrenes Slaegtskab. Aftryk af “ E Museo Lundi.”” 4to. (Copenhagen.)

1896. Woopnwarp, M. F. On the Teeth of the Marsupialia, with especial Reference to the Premilk-
Dentition. Anat. Anzeiger, xii. Bd. Nr. 12 u. 13, pp. 281-291.

EXPLANATION OF THE PLATES.

PLATE 5.
Illustratiug the patterns of the upper tecth.

Nore.—In the preparation of Plates 5 and 6 the figures have been arranged, wherever convenient,
in lines indicative of adaptive sequence, and numbered accordingly. No attempt has been made
to represent the relative sizes of the teeth, those of the smaller forms having been enlarged to a
much greater extent than those of the larger ones for the purpose of illustrating, to the best advantage,
the peculiarities of their patterns. In all of the crown-views the uppermost margin of a figure
represents the anterior side, and the left-hand margin the internal side. In the profile views the
left-hand margin represents the anterior side. The external profile is represented throughout in the
upper molars and premolars, and the internal profile in the lower molars (except Pl. 6, figs. 14 & 15).

The numbers given in brackets are those by which the figured specimens are designated in the
registers of the British Museum.

Vig. 1. Peratherium sp. Right m. 2 reversed. a, crown; 4, profile. (Lower Miocene, Antoign ;

No. 27807.)

2. Sminthopsis leucopus, Gr. Left m.2. a, crown; 8, profile. (No. 52. 1. 13. 3.)

3. Dasyurus maculatus, Kerr. Left m.2 and m.3. (No. 983 0.)

4. Sarcophilus ursinus, Harr. Left m.2 and m.3. (No. 42 6.)

5. Myrmecobius fasciatus, Waterh. Left m.2 andm. 3. a, crown; 6, m.3, profile (No. 76.11. 27.3);
c, left m.2, crown (No. 44. 1. 22. 21).

6. Thylacinus cynocephalus, Harr. Left m.3. (No. 65. 10. 9. 17.)

7. Thylacinus speleus, Owen. Left m.2. (No. 43887.)

wo 735
a Oo oo

5 © W W
“I

OF THE AUSTRALIAN MARSUPIALIA.

Perameles Doreyana, Quoy & Gaim. Left m.3. a, crown; 4, profile. (No. 95. 5. 8. 1:
9. Perameles Bougainvillei, Quoy & Gaim. Left m.2. (No. 41. 1178.)

. Perameles obesula, Shaw. Left m.2. (No. 87.5. 18. 8.)

. Thylacomys leucura, Thomas. Right m. 2 and m.3 reversed. (No. 83. 10. 19. 17.)

2. Notoryctes typhlops, Stirl. Left m.2. (No. 97. 11. 3. 13.)

3. Distechurus pennatus, Pet. Left m.1. (No. 94. 2. 14. 1.)

. Dromicia concinna, Gould. Left m.1. (No. 60. 1. 5. 27.)

. Petaurus sciureus, Shaw. Left m.1. a, crown; 4, profile. (No. 245 d.)

. Trichosurus vulpecula, Kerr. Left m.1. (No. 43. 8. 12. 40.)

. Pseudochirus peregrinus, Bodd. Left m.2. a, crown; 4, profile. (No. 55. 12. 24. 69.)
. Phascolarctus cinereus, Goldf. Left m.2. (No. 253 c.)

. Caloprymnus campestris, Gould. Left m.2. (No. 46. 4. 4. 45.)

. Dorcopsis luctuosa, D’Alb. Left m.2. (No. 76. 10. 28. 3.)

. Lagorchestes hirsutus, Gould. Left m.2. (No. 289 e.)

2. Macropus giganteus, Zimm. Left m.3. (No. 40. 2. 28. 1.)

. Macropus rufus, Desm. Jueft m.3 (worn pattern). (No. 56. 4. 7. 2.)

. Phascolomys Mitchelli, Owen. Left m.2 and m.3. (No. 87. 3. 1. 3.)

. Diprotodon australis, Owen. Right m.3 reversed. (No. 38609.)

. Metachirus opossum, Linn. Left m.3. (No. 98. 9. 5. 9.)

. Caluromys philander, Linn. Left m.2. (No. 51. 8. 30. 5.)

. Peramys Iheringi, Thomas. Left m.2. (No. 61. 12. 2. 9.)

. Peramys americana, Mill. Left m.2. (No. 55. 12. 24. 74.)
. Marmosa cinerea, Temm. Left m.1. (No. 99. 4. 3. 23.)

. Hypsiprymnodon moschatus, Rams. Posterior upper premolar. (No. 1694 a.)

46. 4. 4, 41.)

46. 4. 4. 45.)

. Hypsiprymnodon moschatus, Rams. Crown view of fig. 31.

5. Bettongia penicillata, Gray. Crown view of 32 0.

. Bettongia Gaimardi, Desm. Posterior upper Jeft premolar. (No. 55. 12. 29. 199.)
. Caloprymnus campestris, Gould. Crown view of fig. 33.

. Cenolestes obscurus, Thomas. Profile view of dental series.

. Distechurus pennatus, Pet. Profile view of dental series. (No. 94. 2. 14. 1.)

. Diprotodon australis, Owen. Profile view of incisors, greatly reduced. After Owen (1877,

vol. ii. pl. xix. fig. 1).

view of lower incisors. (No. 87. 3. 1. 3.)

(For cusp-abbreviations vide text-fig. 1, p. 89.)

PLATE 6.

Illustrating the patterns of the lower molar teeth.

Caylux, Tarn-et-Garonne. )

. Sminthopsis leucopus, Gray. Right m.3. a, crown; 2, profile. (No. 52. 1. 13. 3.)
. Dasyurus viverrinus, Shaw. Right m.3. a, crown; 0, profile. (No. 44, 2. 29. 1.)
. Dasyurus maculatus, Kerr. Right m.3. a, crown; 4, profile. (No. 34a.)

. Bettongia penicillata, Gray. a, median; 6, posterior upper premolar. (Nos. 42. 6. 29. 16 and

. Caloprymnus campestris, Gould. Posterior upper premolar. (Nos. 46. 4. 4. 44 and

. Phascolomys Mitchelli, Owen. a, profile view of incisors in young animal, enlarged; 4, ventral

. Peratherium Aymardi, Filh. Right m.3. a, crown; b, profile. (No. M. 2388; Upper Eocene,

216

Fig. 5.
‘ . Sminthopsis crassicaudata, Gould. Right m.1. (No. 46. 8. 3. 37.)

. Dasyurus Geoffroyi, Gould. Right m.1. (No. 86. 1. 26. 10.)

. Sarcophilus ursinus, Harr. Right m.1. (No. 42.)

. Myrmecobius fasciatus, Waterh, a, crown views of right m. 2, m. 3, m. 4; b, internal profile of

N“N

Noe )

eel
oon

wWwwnvnnnnndn wn
DPmMONawpawresd

(Je)
o

co CO
wm

DR. B. A. BENSLEY ON THE EVOLUTION

Sarcophilus ursinus, Harr. Right m.3. a, crown; b, profile. (No. 42 6.)

m.2, m.3; ¢, external profile view of median and posterior premolars and first three molars.
(No. 44. 1. 22. 21.)

. Thylacinus cynocephalus, Harr. Right m.3. a, crown; 3, profile. (No. 93. 4. 13. 1.)
. Thylacinus cynocephalus, Harr. Right m.1. (Drawn from No. 77. 2.6.12 and a foetal

specimen in the Index Collection.)

. Perameles Doreyana, Quoy & Gaim. Right m.3. a, crown; 4, profile. (No. 95. 5. 8. 13.)
. Perameles obesula, Shaw. Right m.2. a, crown; 4, profile. (No. 87. 5. 18. 8.)
. Thylacomys leucura, Thomas. a, right m. 2, crown; J, left m.2, external profile. (No.

83. 10. 19. 17.)

. Cheropus castanotis, Gray. Left m. 3, external profile. (No. 48. 1. 27. 41.)
. Perameles Bougainvillei, Quoy & Gaim. Right m.4 (for comparison with 176). (No.

Al. 1178.)

. Notoryctes typhlops, Stivl. Right m.2. a, crown; 4, profile. (No. 97. 11. 3. 13.)

. Disteechurus pennatus, Pet. m.2. (No. 94. 2. 14. 1.)

. Trichosurus vulpecula, Kerr. Right m.2. (No. 43. 8. 12. 40.)

. Distachurus pennatus, Pet. Right m.1. (No. 94. 2. 14. 1.)

. Phalanger orientalis, Pall. Right m.1. (No. 67. 4. 12. 90.)

. Pseudochirus peregrinus, Bodd. Right m.2. a, crown; 6, profile. (No, 55. 12. 24. 69.)
. Phascolarctus cinereus, Goldf. Right m.2. (No. 253 ¢.)

Tarsipes rostratus, Gerv. & Verr. Three of the vestigial cheek-teeth. (No. 13810.)

. Caloprymnus campestris, Gould. Right m.2. (No. 46. 4. 4. 45.)

. Dorcopsis luctuosa, D’Alb. Right m.2. (No. 76. 10. 28. 3.)

. Lagorchestes leporoides, Gould. Right m.2. (No. 41. 1127.)

. Macropus giganteus, Zimm. Right m.3. a, crown; 0, profile. (No. 40. 2. 28. 1.)

. Macropus rufus, Desm. Right m.3. (No. 56. 4. 7. 2.)

. Phascolomys Mitchelli, Owen. Right m.2 and m.3. a, crown; 2b, profile (m.3). (No.

87..8:1.3,)

. Diprotodon australis, Owen. Left m.3 reversed. (No. M. 474.)
. Metachirus opossum, Linn. Right m. 3. a, crown; 4, profile. (No. 98. 9. 5. 9.)
33.

Caluromys sp. Right m.1. (No. 95. 8. 17. 19.)

[Norr.—In figs. 6, 7, 8, 11, pa. should read pa*.; figs. 2a, 9a, 10a, 13a, he’. should read hi’. ;

fig. 9c, pm. 3 and pm. 2 should read pp. and mp. |

PLATE 7.

Illustrating the modifications of the pes in the Australian Marsupialia.

Nore.—All the figures on this Plate, with the exception of fig. 19, refer to the plantar surface of the
right foot. All, with the exception of figs. 11 and 18, are drawn from spirit-specimens.

Fig. 1. Peramys brevicaudata, Erxl. (No. 89. 10. 26. 14.)
2. Phascogale flavipes, Waterh. (No. 86. 5. 15. 8.)
3. Sminthopsis leucopus, Gray. (No. 86. 5. 15. 9.)

OF THE AUSTRALIAN MARSUPIALIA.

Sminthopsis crassicaudata, Gould. (No. 97. 11. 3. 9.)

. Sminthopsis hirtipes, Thos. (No. 97. 12. 17. 1.)

. Antechinomys laniger, Gould. (No. 97. 11. 3. 12.)

. Marmosa pusilla, Desm. (No. 82. 9. 30. 42.)

. Chetocercus cristicauda, Krefft. (No. 97. 11. 3. 2.)

. Dasyuroides Byrnei, Stirl. (No. 97. 11. 3. 7.)

. Dasyurus hallucatus, Gould. (No. 84. 9.11.3.) (Immature.)
. Dasyurus viverrinus, Shaw.

. Sarcophilus ursinus, Harr. (No. 52.1. 15.2.) (Immature.)
. Dromicia nana, Desm. (No. 60. 11. 29. 34.)

. Perameles Cockerelli, Rams. (No. 82. 10. 27. 10.)

. Perameles Raffrayana, M.-Edw. (No. 83. 3. 29. 2.)

. Perameles Bougainvillet, Quoy & Gaim. (No. 70. 8. 30. 1.)

. Thylacomys leucura, Thos. (No. 83. 10. 19. 17.)

. Cheropus castanotis, Gray. (No. 44. 7. 9. 22.)

. Notoryctes typhlops, Stirl. Dorsal view of left foot.

. Phascolomys Mitchel, Owen. (No. 87. 3. 1. 3.)

Hypsiprymnodon moschatus, Rams.

. Dendrolagus ursinus, Schl. & Mull. (No. 84. 4. 22. 7.)
. Potorous tridactylus, Kerr. (No. 90. 5. 19. 1.)
. Macropus dorsalis, Gray. (No. 85. 1. 30. 1.)

21

*

Bensvry. TRANS. LINN. SOC. SER. 2, ZOOL.. VOL. IX. Pl. 5.

i

H. Griinvold, de. Collings, photose

TEETH OF MARSUPIALS.

’
‘
‘
,
‘
4
re
)
.
‘

BENSLEY. TRANS. LINN. SOC. SHR. 2, ZOOL. VOL. XI. PI. 6.

2A.

H. Grinyold del, Collings, photose

TEETH OF MARSUPIALS.

Beys.ey. TRANS. LINN. SOC. SER. 3, ZOOL.

H. Grinvold, del, Collings, photose

FEET OF MARSUPIALS.

IV. The Labial and Maxillary Palpi in Diptera. By Water Wescuk, FR U.S.
(Communicated by G. MassEr, F.L.S.)

(Plates 8-10.)

Read 16th April, 1903.

Hiruerto only one pair of palpi have been known in Diptera, and these have
generally been regarded as homologous throughout the order. They are described as
maxillary by Savigny, and this nomenclature has been accepted and adopted by the large
majority of morphologists.

The mouth-parts in the various families differ much in shape and armature: the
(stride have only a few tubercles, while the Tabanide have a nearly complete organ.
The trophi of the Muscidz are considered typical of the order, and Calliphora, on
account of its large size and great abundance, has usually been selected for study ;
consequently, the rudiments of palpi, absent in Calliphora, but present in a very minute
form in Musca domestica and many species related to it, have escaped notice. The
second pair of palpi have been searched for, as Kirby and Spence * mention that Savigny
thought he had seen the rudiments of the labial palpi in Tabanus +. Westwood gives as
a character of the order, ‘always destitute of labial palpi” ¢. Later writers who have
studied the proboscis in the Muscide hold similar views.

Dr. Benjamin T. Lowne worked at Calliphora erythrocephala in the larval, nymph,
and imago stages. The subject has been most exhaustively treated, sections of the
various parts having been largely used in his studies. He says§ that the disc on the
end of the proboscis (presumably the labella) is derived from the first pair of maxilla,
_and the palpi present are, “ without the slightest doubt, maxillary palpi.”

Kraepelin ||, Chatin YJ, and Macloskie** are of opinion that the extremity of the
proboscis is derived from a fusion of the labial palpi in the median line. Macloskie thus
translates Kraepelin :—“ The labella which Burmeister and Erichson have shown to be

* “Tntroduction to Entomology,’ Letter xxxy.

+ Savigny, in the first part of his ‘ Mémoires sur les Animaux sans Vertébres,’ entitled “‘ Théorie de la bouche,”
gives a figure of the labium of Tabanus italicus (plate 4, pp. 51-53). A ventral view is given, and the rudiments
consist of two minute tufts of hair, symmetrically placed on slight projections of the labium (‘ tige ”), immediately
posterior to the labella. These he calls “vestiges des palpes?,” adding a note of interrogation. I have not
seen any preparation of 7’. italicus, but I cannot find any similar structure in Hematopata pluvialis, or in any
of the numerous drawings of Zabanus given in H. J. Hansen’s ‘ Fabrica oris Dipterorum.? In Vabanus bromius
and in 1’, sudeticus are tufts of hair on the ventral side of the labium, but no structures suggesting the rudiments
of palpi.

t= ‘ Modern Classification of Insects,’ ii. p. 496. § ‘Anatomy and Physiology of the Blowfly,’ p. 131.

|| “Zur Anatomie und Physiologie des Riissels von Musca,” in Zeitschr. wiss, Zool. xxxix. (1883) pp. 683-
719, tt. 40, 41.

' §[ ‘Machoires des Insectes’ (Paris, 1897, 8yo, pp. 202). ** Kraepelin’s ‘ Proboscis of Musca,’
SECOND SERIES.— ZOOLOGY, VOL. IX. 30

220 MR. W. WESCHE ON THE LABIAL AND

labial palps.”” Meinert *, of Copenhagen, has done much work on the trophi throughout
the order. He has drawn, in a most admirable manner, the mouth-parts of the various
families, but has passed over the large group of the Anthomyiidze, considering it typified
by the Muscidze. According to Dr. Sharp f, Meinert (who writes in Danish) considers
that not only the appendages to a head-segment, but also part of the body of the
segment, may be used in the construction of the mouth-organs.

H. J. Hansen {, also of Copenhagen, has published beautiful drawings of the mouth-
parts. His work is in Danish, with a Latin explanation of the plates. His studies were
mostly confined to the Orthorrhapha, Zabanus being elaborately treated. A novel and
unsatisfactory terminology is to be found both in this and in Meinert’s work.

Trophi of Blatta.
The lettering as in Plates 8 and 9, as explained on page 229.

In the order Hemiptera the rostrum is considered homologous with the proboscis in
Diptera. Those morphologists who have derived the labella in Diptera from the labial
palpi have included the rostrum in their theory. In 1901, Dr. N. Leon §, a professor
at the University of Jassy, in Roumania, found rudimentary labial palpi on the rostra of
several aquatic Hemiptera.

I shall now endeavour to homologize the proboscis of Musca with the typical trophi of
Blatta, as that has an intimate connection with the labial and maxillary palpi. The

* ©Trophi Dipterorum ’ (Kjébenhayn, 1881, 4to). + ‘Cambridge Natural History,’ Insects, part 1. p. 444,
¢ ‘Fabrica oris Dipterorum,’ 1883. § ‘ Recherches morphologiques sur les piéces labiales des Hydrocores.’

MAXILLARY PALPI IN DIPTERA. 221

observations I have made mostly confirm the ideas of Savigny. Speaking of the mouth-
parts in Insecta he says: ‘the organ is the same, the use alone is modified or
changed” *, His nomenclature of the palpi in the Muscidee must be excepted from this
statement. Of all the morphologists known to me, Robineau-Desvoidy alone calls these
labial. Dr. Lowne ¢ thus quotes him:—The base of the labrum “is enveloped by the
base of the labium, of which the palpi are always present.”

Before proceeding further, it may be as well to say that I use the terminology, much
the same as that of Kirby and Spence, given in C. O. Waterhouse’s { essay on the
* Submentum,” to which I wish to express my obligations.

In this investigation I have been guided by the following facts :—

1. Mandibles.—In the Tabanidz, which have all the mouth-parts present except the
labial palpi, the proboscis or labium has tracheated labella, and is so like in appearance
and position that there can be no doubt of its homology with the proboscis in the
Muscide. If a labium of Hematopota pluvialis has the soft parts dissolved away and
is mounted with pressure (and I may remark, for the sake of brevity, that all the parts
referred to later have been treated in this manner, unless expressly stated to the contrary),
and is examined with the higher powers of the microscope, it will be seen that the dorsal
surface has no chitinous structure, the necessary stiffening being afforded by a chitinous
plate on the ventral side, the mentum (PI. 8. fig.1). I have examined Tabanus sudeticus
and 7. bromius, and have found the labium identical in this respect with H. pluvialis. It
is, however, as well to mention that Chrysops cecutiens, which has a very elongated labium,
has chitinous structure on the dorsal side, while still retainimg the mandibles. But this
species, judging from the length of the labium, is a later and more specialized type
than Tabanus, and the chitinous plate followed as a consequence of increased length
requiring more “stiffening.” It may generally be assumed, though not on conclusive
evidence, that the archaic types in Diptera were furnished with a short labium and
that extreme length is a case of specialization. A fairly obvious example is found
in Siphona.

We may then say that Tabanus, having the mandibles present, has no chitinous
structure on the dorsal side of the labium,

When we compare the corresponding parts in the Syrphidee and the Empide, families
in which the mandibles have disappeared, though the maxille are present, we find well-
marked chitinous rods on the dorsal side. We may reasonably presume that the
mandibles are soldered into the dorsal surface of the labium. This view is held by
Professor Packard, and has general acceptance with little or no opposition (Pl. 8.
figs. 2, 3).

2. If we examine Calliphora we find a similar chitinous structure on the dorsal side
of the labium. Dissection shows that the parts readily separate into two lateral rods
with a broad central plate (Pl. 8. fig. 7).

In Fristalis tenax, E. arbustorum, and Syrphus balteatus these rods are different in

* ¢ Mémoires sur les Animaux sans Vert¢bres,’ Bouche, p. 11.

+ ‘ Anatomy and Physiology of the Blowfly,’ p. 130.

t ‘The Labium and Submentum in certain Mandibulate Insects’ (London, 1895, Svo, pp- 12, 4 col. pls.).
30)*

222 MR. W. WESCHE ON THE LABIAL AND

shape and are continued along the sides of the main trachese. The central plate seems
to be detached, but that the parts are homologous with those in Calliphora (Pl. 8. fig. 7)
is evident.

3. Maxille—Again examining Calliphora we find that, in addition to the mandibles,
the maxillee, which are present in the Syrphide and Empide, have disappeared. But if
the base of the labium is dissected, the cardines of the maxillze will be found embedded,
and working the labrum as levers (Pl. 8. fig. 4). That these are cardines of the maxille
may be proved by comparing the parts with those in Hristalis and Helophilus (Pl. 8.
figs. 3, 5, & 6). A comparison with Culex pipiens only strengthens this proposition
(Pl. 9. fig. 24) *.

We also find the palpi in a different position. In Calliphora they are above and
posterior to the base of the labrum; in Hmpis and Hristalis they are in a more anterior
position and firmly attached to the cardines of the maxille (Pl. 8. figs. 2, 3, & 4)f.
“A dissection of the maxille in Helophilus pendulus and Eristalis tenax shows the palpi
attached to the maxille by a membrane, which has at its extremity a chitinous plate,
which is probably the palpifer. The bases of the palpi connect with the maxille just
before they emerge from the enclosing membrane of the labium. There are well-marked
ridges halfway down the maxillz that indicate the boundary of the cardo. The laciniz
continue, the gale having aborted. The palpifer shows only on the palpus, being pulled
apart from the maxilla (Pl. 8. figs. 5 & 6). Further, I have dissected the maxille in
Culex pipiens, 2 (Pl. 9. fig. 24), in Tabanus bromius, T. sudeticus, Chrysops cecutiens,
Asilus crabroniformis, and a species of Simuliwm, all species with a nearly complete
armature, and I find that invariably the palpus is firmly attached to the stipes and cardo
of the maxilla.

4, Labium.—tThe palpi in the Muscide are in quite a different position, attached to
the membrane of the labium; at their bases in Calliphora are plates of chitin and rows
of two or three bristles. The palpi have no connection with the levers of the labrum,
which, as we have seen, are the cardines of the maxille in Hristalis. They are there-
fore labial and not maxillary palpi, and have worked round from a lower part of the
labium to the upper. The chitinous plates and rows of bristles in Calliphora have their
analogies in a great number of related species. In some the chitin is evident, in others
it has almost disappeared, but the row of bristles is very constant, and can be seen in all
the species possessing labial palpi that I have examined. It is fairly obvious that these
parts represent the palpiger.

5. A dissection of the trophi of Culex pipiens throws some light on the subject under
discussion. In both sexes the labium is simply a chitinous sheath for the lancets. In
its paired and segmented flaps may be seen the homologies of the labella.

The cardines and stipites of the maxille are present, having much the same appearance
as in the Muscide, and a palpifer can be traced supporting the three-jointed palpus
(Pl. 9. figs. 23 & 24).

* See note on page 229.
+ The stipes and cardo are so soldered together that it is not possible, in the majority of cases, to define them.
The palpifers are attached to the stipites and not the cardines.

MAXILLARY PALPI IN DIPTERA. 223

In the female the maxilla, mandibles, hypopharynx, and labrum are very beautiful
weapons. In the male the lancets are obviously atrophying, and they can only with
difficulty be dissected out from the enclosing labium. Here is seen the process, which
has been completed in the Syrphidz and Empidee, actually proceeding, or suddenly halted
in a stage of the process.

In the males the cardines of the maxille hinge on to the maxillary palpi, which, in
this sex, are very long and highly developed (Pl. 9. figs. 25, 26, and text-fig. 2).

Summary.—tiIn the Muscidee the mandibles are embedded in the dorsal side of the
labium. The maxillary palpi, galeze, and lacinize are aborted, but the cardines remain.
The palpi present are labial, and the palpigers are represented by chitinous plates with
two or three sete springing from or near them.

In the Syrphidz and Empide the mandibles are similarly placed, but the maxillee are
represented by the laciniz, the palpi, cardines, stipites, and palpifers. The labial palpi
are aborted.

Rule.—A rule can he formulated :—The maxillary palpi when present in Diptera are
always in contact with the upper part of the cardines, the stipites.

Extremities of maxilla and mandible of Culea pipiens, 2.

Rudimentary Maxillary Palpi.—l was fortunate in discovering a number of species,
in the Anthomyiidz, Muscide, and other families of the Cyclorrhapha, with four
palpi which afford useful clues to the surrounding parts. They appear in different
stages of atrophy, from a fairly developed organ to a rudiment of a few hairs. Some
examples were described in a paper read at the meeting of the Royal Microscopical
Society on June 18th, 1902, and since published in the Journal of that Society *.
These palpi enable us to indicate the position of the maxillee with certainty, and,
by proving the other pair to be labial, establish the proposition that the whole organ is
the labium.

We have now the mandibles, the maxille and palpi, the labium and palpi, and the
mentum clearly defined.

Paraglosse, Ligule——We now come to the parts of the labium, the paraglossz and

* Joum. Roy. Micr. Soc. 1902, p. 412. ‘ Undescribed Palpi on the Proboscis of some Dipterous Flies, with
Remarks on the Mouth-parts in several Families.”

224. MR. W. WESCHE ON THE LABIAL AND

ligule. In the Muscide, Tabanide, Syrphidze, and other families we find, at the
extremity of the labium, two paired organs, the labella and the transverse chitinous
levers that support them. They are in the situation where they might be expected to
be found, and I have little doubt represent—the labella the paraglosse, and the
transverse levers the ligule (PI. 8. figs. 4,8; Pl. 9. fig. 17, &c.). In the paraglossze of
Culex (Pl. 9. fig. 23) I have seen two strengthening rods, which confirm me in the above
homology. They only appear on dissection and are not shown in the figure. In species
of Empis, such as HL. chioptera (Pl. 8. fig. 2), and in Siphona geniculata, the ligule
have aborted.

Teeth.—The large majority of species in the Muscidee, from Tachina down to Scato-
phaga, have an armature of teeth at the bases of the false trachea. These may be traced.
through such specialized species as Stomoxys and G'lossina, and may even be found in a
most minute size in some of the species of the parasitic Hippobosca. They are arranged
in symmetrical crescents, and are so close to both the mandibles and the ligule that
they might possibly be parts of those organs; but the simpler hypothesis that they are
outgrowths from the paraglossz is probably the correct one. Some light is thrown on
this point by the very exceptional dentition shown in Ephydra (Pl. 9. fig. 21). Here
each of the pseudotrachez is furnished along its whole course with many short chitinous
teeth disposed at regular intervals.

Labrum, Lingua, and Mentwm.—All morphologists seem agreed that the lancet-case
corresponds with the labrum, and the hypopharynx with the lingua. The homology of
the ventral plate on the labium with the mentum is fairly clear (Pl. 8. fig. 4).

Submentum.—The submentum is possibly membranous, but it may be the “ fulcrum.”
This part is described as the chitinized walls of the pharynx, but a long tracheated
tube is easily dissected out from the trophi of all the species I know, and seems to be
the true pharynx. There is so little to guide that this point must remain doubtful ;
allowing this, we thus have the parts complete.

Summary.—The proboscis in Diptera may be regarded as homologous with the
trophi of the typical insect mouth. In the Muscide it has been modified into a rod-like
organ, enclosed in a thin membrane, and capable of extension and retraction. It is
mainly formed of the labium, and has the labial palpi well developed and placed near
chitinous ridges, which are the palpigers. The palpi have worked round from the under
to the upper side—or, taking the usual position of the labium into account, from a
posterior to an anterior position, where it is obvious that they would be of greater service
to the insect.

The proboscis has two symmetrical tracheated flaps at the extremity, which represent
the paraglosse. The teeth, which are often present, may be considered outgrowths of
the paraglossze. The transverse chitinous levers which expand the labella are the ligula.
Mr. Waterhouse shows (‘ Labium and Submentum ’) that the paraglossze have a tendency
to enlarge at the expense of the ligule, and this has occurred in the Muscide.

The labium has absorbed into its structure the mandibles and maxille on the dorsal
side, and the mentum, and possibly the submentum, on the ventral side.

MAXILLARY PALPI IN DIPTERA. 225

The mandibles act as supports and assist in working the labella.
The upper parts of the maxille, which are present in Syrphus, have aborted, leaving

the cardines and stipites, which find use as levers, acting on the labrum and hypo-|
pharynx. The maxillary palpi have mostly aborted, but in many species they are ~.

present in a more or less rudimentary state. When the maxillary palpi are present,
they are based on structures homologous with the palpifers. These are attached to the
stipites. The palpi are situated at proportionally the same distance from the bases of
the cardines as the palpi in the Syrphidee and the Empidee (PI. 8. figs. 3, 5,6; & Pl. 9.
fig. 17). The proboscis has, on its dorsal surface, a lancet-case, the labrum, which acts
as a sheath for the hypopharynx, which homologizes with the liagua.

At the base of the hypopharynx, and continuing it, is a tube, the pharynx; this is
sometimes tracheated, and is well marked in this state in Calliphora and Stomoays.
This tube curves upwards under the base of the labrum.

The proboscis in the Syrphidze and the Empidee corresponds with that in Musca,
except that, in their case, the labial palpi have aborted, and the laciniz of the maxille
and maxillary palpi are present.

Rudimentary Maxillary Palpi.—l shall now give a short description of the rudi-
mentary maxillary palpi in a few species, figuring various forms.

Family AntHomy1ip#, subfamily M/ydeine.—Polietes lardaria, Fabr. Palpi pyra-
midal in form, strongly haired, short pointed projection at extremity of stipes, palpifer
marked ; base 3+; in., length 3$; in. (PI. 8. fig. 13.)

Same family and subfamily.—Hyetodesia lworum, Zett. Very elongated, strongly
haired, long blunt projection at extremity of stipes, palpifer marked; base 77:5 in.,
length 337 in. (PI. 8. fig. 16.)

Same family and subfamily.— Hyetodesia perdita, Meig. Acutely conical, well haired,
no projection, palpifer indistinct ; base $5 in., lenggh gho in. (PI. 8. fig. 10.)

Same family and subfamily— Wydea impuncta, Fallen. Broad at the base; base not
haired ; tapers off with a curve, where it is thickly haired ; stipes broadens very much
at end; palpifer indistinct; base ¢}5 in., length gj in. (PI. 8. fig. 9.)

Same family and subfamily —Hydrotea dentipes, Fabr. Short and thick; thickly
haired, blunt projection of stipes; base g}5 in., length g§> in. (PI. 8. fig. 15.)

Same family and subfamily.—Ophyra leucostoma, Wied. Broad at the base, short and
stout, thickly haired, palpifer distinct ; base 735 in., length g}> in. The cardines are
very strong in this species. (Pl. 8. fig. 14.)

Same family, subfamily Homalomyiine.—Homalomyia canicularis, Linn. Very small,
conical in shape ; base y4/5p in., length s$5 in. (PI. 9. fig. 28.)

Family Muscipm.— Cyrtoneura stabulans, Fallen. Tubular and short; stipes projects,
well haired ; palpifer fairly distinct; base 55 in., length g3> in. (PI. 8. fig. 11.)

In a very rudimentary form I have found the palpi in a number of species. They
have all very much the same appearance, so that a separate description would be
superfluous. I have figured four species. In most cases only a few hairs remain, but
in Lasiops I have found quite a tuft.

a -
&

\
\e,

296 MR. W. WESCHE ON THE LABIAL AND

Family ANTHOMYIID2. Subfamily Homalomyiine.
Subfamily Anthumyiine. Azelia Macquarti, Stig.
Hylemyia cardui, Meig. (Pl. 9. fig. 27).
H. pullula, Zett. (very small). Family Sarcornacip2.
Lasiops ctenoctema, Kow. Myiocera carinifrons, Fallen.
Anthomyia radicum (Linn.).
A. pluvialis (Linn.). Family Muscipa.
A. sulciventris, Zett. Graphomyia maculata, Scop. (Pl. 9. fig. 31).
Phorbia floccosa ?, Macq. Musca domestica, Linn. (Pl. 8. fig. 12).
Pegomyia bicolor, Wied. (Pl. 9. fig. 29). M. corvina, Fabr.

Acalyptrate Muscide.—In a number of species in the Acalyptrate Muscide the palpi
are quite fully developed, but, on account of the small size of the flies, are very difficult
to make out. The maxille, when present in Diptera as hitherto observed, have only
the lacinie and palpi present, the gale having aborted. Such a state of things is seen in
the Culicide, Simulidee, Tabanide, and Asilide. In a small and common species of
Ephydra (PI. 9. figs. 21 & 22) I found a maxilla which has all the parts represented.
This species has also the remarkable dentition previously alluded to.

In Balioptera combinata, Linn. (Pl. 9. fig. 18), an arrangement is seen differing from
those already mentioned, but there are several species that follow this type. It is
difficult to say whether the palpi or the lacinize are aborted. Comparing this species
with the Syrphidz one is inclined to think that the palpi are absent.

In Nemopoda cylindrica, Fabr. (Pl. 9. fig. 19), maxillary palpi can be made out, but
the cardines are so faintly chitinized that their presence can only be suspected. In
the nearly related Sepsis cynipsea the cardines and stipites are exceedingly short.

In Spherocera subsultans, Fabr., the four palpi are found in a more developed state
than any hitherto met with (Pl. 9. fig. 20.)

In the figure a dorsal view is gfven of the parts, showing the fulcrum, the mentum
through the transparent membrane, and the labrum above it. I have examined several
species of this family, but have only found similar palpi in Borborus suillorwm, Hal.
In Limosina lugubris, Hal., L. sylvatica, Meig., L. fuscipennis, Hal., and Spherocera
denticulata, Meig., the palpi were of much the same type as in B. combinata
(Pl. 9. fig. 18).

Archaic Type.—The varying armature and shape of the trophi in Diptera suggest
a speculation as to archaic types.

I. The labium was probably short as in 7%pula and has gradually elongated. The
labial palpi have remained at or near their original distance from the head, but have
worked round to the upper part so as to be of service. Where this has not occurred
(1) they have aborted as in the Syrphidee, Empidz, Bombylide, Asilidz, Tabanidze, and
Culicide ; (2) or they have followed the labella down, as in Dilophus (P1. 8. fig. 8).

Siphona geniculata is an example of the extreme development caused by the advantages
gained by elongation. In this species many parts usually present on the labella, teeth,
and transverse expanding levers (ligula) have aborted, and no traces of the maxillary
palpi are to be seen, From this we may infer that this type is more specialized,

ae

oe

MAXILLARY PALPI IN DIPTERA, 227

and consequently less archaic, than Ophyra leucostoma (Pl. 9. fig. 17), which has an
abnormally short labium, the maxillary palpi well marked, and the parts on the labella
very evident.

II. Pubescent Eyes.—All the flies in the Cyclorrhapha with pubescent eyes that I
have examined, either have the maxillary palpi only, or possess rudiments of them,
e. g., Syrphidee, Polietes, Hyetodesia, Hydrotea occulta. Some species of Phoridee
are apparently an exception; it is, however, very doubtful whether this family is in its
proper division.

Ill. Venation of Wings (text-figs. 3 & 4).—The rudimentary palpi seem to follow the
venation of the wings in a curious manner. ‘The cell formed by the 3rd and 4th
longitudinal veins (subapical or 1st posterior cell) is open in the Anthomyiide, begins to
close in Cyrtoneura stabulans, is nearly closed in Musca domestica, and is quite closed
in Phito melanocephala. The palpi gradually diminish through these species till no sign
is left in P. melanocephala.

Wing of Polietes lardaria. Wing of Phito melanocephala.

IV. Tegule.—The tegule also seem correlated :—

(1) When all the palpi are well developed, as in Spherocera subsultans, the subapical
cell of the wing is open, the tegulz are absent.

(2) When the palpi are more or less rudimentary, as in the Anthomyiide, the cell
still remains open, but the tegulze are well marked.

(3) When the palpi are rudimentary, as in Cyrtoneura stabulans, the cell begins to
close and the tegulze increase in size.

(4) When the palpi are very rudimentary, as in Musca domestica, the cell is still more
closed, the tegulz remain large.

(5) Wher no trace of the palpi can be seen, as in Calliphora, the cell is still more
closed, the tegulz remain large.

(6) In Phito and Melanophora the tegule are still more developed, no trace of palpi
can be seen; the 4th longitudinal or median vein has closed the 1st posterior cell
so much that it has left the margin.

(7) In Gstrus the mouth-parts are rudimentary, the venation has left the lower part
of the wing, and the tegul are at their largest development.

Working on the above data, a fly may be imagined with hairy eyes, a plumose arista,

a venation approximating to that in the Anthomyiide, mandibles fused into the labium,
which would be short and stout; maxille also absorbed, but with both palpi well

SECOND SERIES.—ZOOLOGY, VOL. IX. 31

228 MR. W. WESCHE ON THE LABIAL AND

developed, and the tegulze small or absent. This might be an ancestral or primitive type
of the Muscidee.

Polietes and Hyetodesia would only differ from this insect by less developed maxillary
palpi, a longer labium, and larger tegule. Ophyra leucostoma, while retaining the short
labium, has lost the plumose arista, the hairy eyes, and retains fairly large rudiments
of the palpi. Cyrtoneura has also lost the hairy eyes, developed a slightly longer labium,
but retains the plumose arista and fairly large rudiments of the palpi.

In Hydrotea the palpi are present, but the plumose arista has gone. Some species
retain the short labium, and others have lost the pubescence on the eyes. Finally, we
come to the little Lasiops, with hairy eyes and only rudiments of the maxillary palpi.

My diagram is drawn up on this system—that is, as the species lose primitive
characters they become more recent ; and it would be tedious to trace the matter further,
as a glance at the scheme will make my meaning clearer than any written explanation.
(Plate 10.)

Classification —It will be seen that this arrangement shows what excellent natural
groups the species of the Anthomyiidee and the Muscidze have been divided into, and
how one character more, the rudimentary palpi, follows the order in which the genera
have been placed. It is well marked in the subfamily Mydeinz, dwindles in the
Anthomyiine and Homalomyiine, disappears in the Ccenosiine, and no trace is found
in the nearly related Scatophaga.

In this paper I have endeavoured to avoid a controversial attitude, but it must be
obvious that if the conclusions I have arrived at are correct, the contrary must be the
case with the works of several theorists, and that the generally accepted dogma that the
palpi in Diptera are homologous and maxillary can no longer stand. ‘Therefore it will
probably be urged by some (and, looking at the subject from their point of view, I admit
quite fairly) that my methods are out of date, and that comparative anatomy must give
way to a minute sectional study of the insect from the ovum, through the metamorphoses,
to the imago state. To this I answer that this method, so apparently promising and
conclusive, when applied to the trophi of Diptera, is discounted by its results. Hither
it is a tool of such complexity and nicety that no observer has hitherto used it correctly,
or the facts observed have not been properly weighed and understood.

Since the preceding pages were written, Professor V. L. Kellogg has published a paper
which lends valuable aid to my contention as to comparative anatomy and ontogenetic
study, when applied to the homologization of the mouth-parts of Diptera *.

If the presence of labial and maxillary palpi in Diptera is admitted, and I cannot see
how, unless my facts are traversed, this can be denied, such speculations as the derivation
of the labella from the fusion of the labial palpi or from the jist maxille cannot
be entertained, though my investigations quite agree with the theory that the labium is
a modified double maxilla, derived from the second pair of jaws.

* «The Development and Homologies of the Mouth-parts of Insects,” The American Naturalist, yol. xxxyi.
(Sept. 1902), pp. 683-706.

|
|
|

a

al

MAXILLARY PALPI IN DIPTERA,

Note.—The Homology of the Cardines.

It has been objected that the homology of the levers of the labrum in the Muscidee
(the “apodemes” of Dr. Benjamin Lowne) with the cardines of the maxille is not quite
clear. As this is of vital importance to my theory, I have enlarged my argument and
summarized it as follows :—

t.

J)

a

In Vespa and Apis the cardines are on the same plan as in Musca. They are
generally admitted to be the “hinges” of the “ posts’? supporting the palpifers
and other parts. If these are cardines here, so must they be in Diptera.

It may be urged that the fact that the levers are enclosed is against the theory
that they are cardines. This objection has no weight. The cardines being the
lowest part, they would (if there were any tendency in that direction) be the
first part to be embedded.

In Culex, 2, the levers are undoubtedly cardines, the four-jointed palpi are affixed
to them, and though the palpifer is differentiated, I have failed to see the point
where the stipes and cardines fuse.

In Eristalis and Helophilus parts of the maxillee are present with palpi. It is

impossible to deny the homology of the posterior ends with the apodemes.
Therefore as these parts bear the palpi, they must be in some part the stipites ;
and as similar parts in Vespa and Apis are admitted to be the cardines, in
Eristalis and Helophilus they must be the fusion of the stipites and cardines,
which applies equally to the Muscide. ,

It may be suggested that in the Muscide the levers are so obviously necessary to
work the labrum and hypopharynx that they would specially evolve. To this it
may be answered that the levers are present in Culex. In the female they binge
on to the blades of the maxille; in the male they work the long palpi charac-
teristic of thatsex. So in Diptera we know of three uses of the cardines, showing
there exists a tendency to utilize this part.

In the small Hphydra figured (which, judging from a determination of Mr. Piffard,
is Hydrellia griseola, Fallen), in the same places as the apodemes are also found
complicated organs, which are evidently the complete maxillee.

The negative proof—if they are not cardines, what are they ?

EXPLANATION OF THE PLATES.
The lettering applies to all the figures.

a, labium. k, stipes.

b, paraglossa, | m, cardo.

¢, ligula. | n, maxillary palpus.

d, labial palpus. | o, palpifer.

e, palpiger. , | p, mentum.

f, mandible. | r, submentum.

g, maxilla, s, labrum.

h, galea. | t, hypopharynx or lingua.
i, lacinia.

Ve 4

230 ON THE LABIAL AND MAXILLARY PALPI IN DIPTERA.

PuLaTE 8.
Fig. 1. Labium of Tabanus (Hematopota pluvialis), lateral view. The remaining parts (the mandibles,
maxillz, max. palpi, labrum, and hypopharynx) have been removed.
2. Head of Empis chioptera.
3. Proboscis of Evristalis tenaz, lateral view.

3 Calliphora erythrocephala, lateral view.
5. Dissection of the maxilla and palpus of Helophilus pendulus.
6. = 9) 5) Eristalis tenax.

7. Dissection of the dorsal plate on the labium of Calliphora erythrocephala, showing the embedded
mandibles. The lower end shows a portion of the pharynx, dorsal view.
8. Labium of Dilophus albipennis, ventral view.
9. Maxillary palpus of Mydea impuncta. Only the anterior end of the stipes is shown, with a
lateral view. This applies to all the figures of palpi on these Plates.
10. Maxillary palpus of Hyetodesia perdita.
ils 6 se Cyrtoneura stabulans.
12. Rudiment of maxillary palpus, Musca domestica.
13. Maxillary palpus of Polietes lardaria.

14. 39 44 Ophyra leucostoma.
be 3 A Hydrotea dentipes.
16. a3 a Hyetodesia lucorum.
PLATE 9.

Fig. 17. Proboscis of Ophyra leucostoma, dorsal view.

18. > Balioptera combinata, lateral view.

19. sf Nemopoda cylindrica, lateral view.

20. a Spherocera subsultans, dorsal view.

Pls Dp a species of Ephydra (Hydrellia griseola, Fallen ?).

22. Maxilla of same, more highly magnified.

23. Part of the labium of Culex pipiens, 2, showing the labella, dorsal view.

24, Dissection of the cardo, stipes, palpifer, and palpus of the maxilla of Culex pipiens, ? .
9

ol

. Extremity of the labrum of Culex pipiens, 8. This has affinities with the labrum in some
species of the Syrphidee.

26. Extremity of the labrum of C. pipiens, 9, showing the higher organization of the female. _

27. Rudiment of maxillary palpus, Hylemyia cardut.

28. 3 S Ay Homalomyia canicularis.

29. ns x 5 Pegomyia bicolor.

30. End of stipes, Phito melanocephala.

31. Rudiment of maxillary palpus, Graphomyia maculata.

Puate 10.

Diagram showing a speculative arrangement of genera and species, placed
as they recede from a hypothetical primitive form.

Wesche. Traws.Linn.Soc.SER.2. Zoot Vou IX Pl. 8.
¥ “ %

West,Newman imp.

PALPI IN DIPTERA. ae

=. "ie —
Wesche.

Trans.Liny, Soc. SER.2.Z00L Vou IX P12 ‘

A

& Fy Se ya
Band

RTE Na tere =

ee

a

ee
PALPI IN DIPTERA. a

Wescué. TRANS. LINN. SOC. SER. 2, ZOOL. VOL. IX. Pl. 10.

.

Primitive form: Hairy eyes, four palpi, mandibles
and maxille in labium, arista plumose,
short proboscis.
; Cyrtoneura Ophyra

q Polietes.
- Explanation.

s fe) Hairy eyes.

HyetodesiaZ

Hydrotea:

[= arista.. Hyle in((() Lasiopa {||}
(Som arista.

Musc

ma((|)
Azcia([)

Ceenosie )

Stomoxys.

Hippobosca.

W. Wesché, del, Collings, photose.

HYPOTHETIC ARRANGEMENT OF GENERA OF MUSCIDA,

V. On the Anatomy and Development of Comys infelix, Embleton, a Hymenopterous
Parasite of Lecanium hemisphericum. By Atice L. Empieron, B.Se., 1851
Exhibition Science Research Scholar; Associate of the University of Wales
(Cardiff College). (Commamicated by Dr. Davip Suarp, F.R.S., PLS.)

| (Plates 11 & 12.)

Read 4th June, 1908.

ConTENTs.
: Page Page ~
Ble ENOUUICHON ois ieee eee vente e Waele ees 231 IIL. Development, and Structure in Karly Stages. 237 .
Characters of Comys infelia .......... 232 TV Amatomyvof Imago Qe aces ace se 244
Il. Natural History: Classification ........ 232 Ver Anatomy of lmago- gy scree leler-lls 250
Wisdesor Occurrence, Ke, 2. .4..5-+--- 233 WAG Lohllboemliny £655 cc0o00ouc o7aeeocoo vor 252
Keonomic Aspects ...... areca ern Gis: 235 VIL. Explanation of the Plates.............. 254

I. INTRODUCTION.

ee subject treated of in this paper is of interest both to the biologist and to the
economist. Comys infeliv is a very small Hymenopterous fly belonging to the family
-Chalcididee, the members of which have long been recognized as beneficial from an
economic point of view, inasmuch as they are great destroyers of the Coccid pests
that are so inimical to cultivation in all parts of the world. Biologically this species of
_ Hymenoptera claims attention by virtue of its remarkable life-history, the early stages of
which are passed inside a scale-insect, while the imagines, male and female, lead a free
life after emergence from the Coccid. In spite of their two-fold importance, little is
_ known about these insects, and their life-histories remain obscure. Recently Bugnion (9)
_ has published the results of his researches on the development, anatomy, and habits of
Eneyrtus fuscicollis, a form allied to Comys infeliv, though with very different habits.
He mentioned the fact that many embryos of this species were to be found enclosed ina
capsule, or tube, in the interior of the host—a caterpillar, but he did not thoroughly
elucidate the matter. Since Bugnion’s paper was published, Marchal (39, 40, 41) has
discovered that his species (now called Ageniaspis fuscicollis, Dalm.) offers in its develop-
ment an example of the phenomenon so rare in zoology, and of extreme intecest from a
philosophical point of view, of dissociation of the embryo. In an early stage of the
development, the embryo breaks up and forms from fifty to a hundred embryos. The
development of this insect is being studied by this accomplished French savant, and we
anticipate most interesting results. He suggests that the species of Ayeniaspis afford
SECOND SERIES.—ZOOLOGY, VOL. IX. 32

232 MISS A. L. EMBLETON ON THE ANATOMY

examples of chronic segregation, the individuals being separated in the times of their
appearance in conformity with the habits of the species of which they are parasites.
No morphological differences have yet been detected between the “seasonal races ” in
this case.

Apart from the investigations of these two naturalists, there has been, within the last
few years, work done on the Excyrtine by the American school of entomology, especially
by Dr. L. O. Howard and Dr. Ashmead, but their studies have not been directed to
investigation of the ontogeny.

I propose, therefore, in this paper to record the results obtained after work for eighteen
months on the Chaleid, Comys infelix; from the nature of the case there are necessarily
some points which are left obscure.

As the species Lam dealing with has at present only been characterized bya few diagnostic
words, it may perhaps be well here to give very briefly the distinguishing features serving
to identify it among other species of the genus—especially to distinguish it from Comys
dicolor, which it resembles to some extent. At first I took it to be that species, but
further investigation pointed to the conclusion that I was dealing with a new species.
I submitted it to Dr. L. O. Howard, who confirmed my opinion, as did also Mr. P.
Cameron and Mr. R. Newstead.

CoMYs INFELIX, Embleton (Trans. Entom. Soc. London, 1902, i. pp. 219-229).

A small fly, 2 mm. in length, 3°75 mm. across the outstretched wings. Colour
extensively black, but variegated, the head and thorax dark brown, abdomen black; the
coxie of the jirsé legs are silvery white, the other parts fuscous and the tarsi black ; tibial
spur normal; in the second legs the coxze are black, the femora white shading to black,
the tibie being expanded at the distal ends, and bearing a long, powerful spur; the
third legs have white coxee, brown femora, and dark brown tibiz with a normal spur,
the tarsi begin white and then shade to black at the tip; the tarsi are all five-jointed.
The antenne are black, club-like at the tips, compressed in the female, but subeylindrical
in the male; there are cleven segments, all more or less clothed in fine hair; the funicle
is six-jointed, while the pedicel is shorter than the next segments. Between and rather
behind the two large dark eyes are three ocelli. The scutum is raised and triangular,
hearing an apical tuft of long dark setee, directed backwards. The fore wing is mottled
in blackish fuscous patches. The aldomen is short, and united to the thorax by an in-
conspicuous petiole; laterally the abdominal segments become ‘looped up” over a small
oval hinged plate on which are four long sete. The ovipositor, almost entirely hidden,
is composed of two large expanded chitinous plates, and a central sting made up of two
pointed rods.

II, Narurau History.

(a) Classification.—The creature belongs to the subfamily Lncyrtine, according to
tloward in Comstock’s report (10). ‘The distinguishing characters of the Chalcidide

as given there are as follows:—Tarsi five-jointed ; middle tibize with a very stout spur at —

i

AND DEVELOPMENT OF COMYS INFELIX. 233

=

the tip. The Eneyrtine come under this family and are distinguished from the Aphelinine
by the fact that the antenne are more than eight-jointed. It is perhaps best to quote
the more detailed diagnosis of the distinguishing characters of the Encyrtine as given in
Comstock’s report :

“Subfamily Hncyrtine.—Tarsi five-jointed ; middle tibizee somewhat dilated towards
the tip, and furnished with a long stout spur; antennze more than 8- usually 11- or 10-
jointed. Parapsides of the mesoscutum not separated by furrows; mesothorax prominent,
broad in the middle ; vertex with an acute occipital margin ; abdomen usually short and
sessile. The members of this tribe are small active Chalcids, which, while by no means
confined to Coccids as hosts, still are much more often parasitic on insects of this family
than upon those of any other. Dr. Mayr (42), in his paper upon the Eurepean Encyrtine,
tabulates the species according to their hosts, and we inay briefly condense by saying
that one species is parasitic upon an Hymenopterous insect, two upon Coleoptera, four
upon Lepidopterous eggs, sixteen upon Lepidopterous larve, four upon Diptera, while
forty species are parasitic upon Hemiptera, of which thirty-nine infest bark-lice, the
remaining one being found upon two species of Aphides. Ratzeburg (46) mentioned
two species of Lncyrtine parasitic upon Hymenoptera, four on Coleoptera, four on
Diptera, twelve upon Lepidoptera, and no less than twenty-five upon Hemiptera. Even
these facts, however, cannot be taken as fairly indicating the proportion of these insects
which are parasitic upon the Coccidie, since the latter family has heretofore beea so little
studied in comparison with other groups, that doubtless many of its parasites have never
been reared. When as much biological work shall have been done upon it as, for
instance, upon any one of the families of Lepidoptera, we may expect to find that the
proportion of Exeyrtine parasitic upon insects of other families will become dwarfed by
comparison.”

* Genus Comys, Forster (19, 20, 21).—Antenne rather long, eleven-jointed ; funicle
six-jointed ; pedicel slightly shorter than the succeeding joints, from joint three the
joints of the flagellum gradually decrease in length, with the female they become more
and more compressed towards the tip of the club, with the male remaining subcylindrical.
Head and face coarsely punctured. Mesoscutum without silvery white hair. Scutellum
three-cornered, with a somewhat rounded tip, near which is a tuft of erect, long, stiff
dark hairs. Ovipositor entirely, or almost entirely, hidden. Fore wings brownish on
the distal half, the nearly clear basal half having a brownish cross streak ; marginal
vein shorter than stigmal; post-margival and stigmal long, Males very similar to
females, antennal characters giving the only absolute distinction; wings sometimes clear,
and sometimes brownish as with females.”

(b) Mode of Occurrence, §c.—Comys infelix was first noticed in August, 1901, on an
Asplenium fern parasitized by Lecanium hemisphericum * vay. filicum, and by Chionaspis
aspidistre, Signoret. I observed that Comys infeliv emerged from the Lecanium, with

* I notice that some of the most recent writers (in 1902) have called this Coccid Saissetia hemispherica, 1 have
not been able to discover why this alteration has been effected, if it is so, for the name Suissetia was used in Mollusca
in 1900, and is therefore preoccupied.

32*

~

234 MISS A. L. EMBLETON ON THE ANATOMY

which the fronds of the fern were covered; a small round hole was left in the dorsal
shield of the scale as the only indication of its destruction by the fly; neither then nor
since have I found more than one fly emerge from one host or victim. From the small
white Chionaspis a very minute fly hatched out in abundance; it is apparently
Aspidiotiphagus citrinus, Howard (35 mm. in length), but I have not at present been
able to continue the work on this species.

I soon discovered that the same kind of fly existed in the Lecanium when this
occurred on other ferns—viz., Aspidium falcatum, Asplenium spp., Pteris spp. (?):
usually the Ptevis was most abundantly stocked with the scale; it also occurs on
Beaumontia.

It is an interesting fact that the flies hatched out all through the year, though in
fewer numbers in the winter, but appearing in great profusion in spring and early
summer; similarly the earliest larval stages were most common in late summer and
early autumn, but it was always possible to obtain the different stages all the year round.
This continuity of production may merely be the result of the artificial conditions under
which the insects live, for the palms and ferns which their hosts inhabit are all hot-
house varieties experiencing fairly constant conditions as regards food-supply, warmth,
and moisture, whereas in a state of nature they might be subject to seasonal changes.
I have never found the parasite on ferns growing out of doors, and the Lecanium itself
is cousidered to be exclusively a greenhouse species in Europe and America.

The females are much more common than the males, which appear comparatively
rarely, and then orly in small numbers; so far, I have found them in the spring and
early summer. I shall subsequently mention that this disproportion of the sexes in
Comys infeliz is very great, perhaps a thousand females to one male: this is the only
circumstance I have noticed that suggests the natural time of emergence, which, if this
may be relied on, is the first warm weather of spring. It was at this period, too, that
the females were produced in greatest numbers; the sporadic appearance of other
examples is therefore quite probably due to the unnatural conditions of existence as
stated above. The male is so rare that I have had but little opportunity of observing it,
and owing to this rarity my remarks may be considered to apply only to the female
except when the other sex is specially mentioned. During the time when the flies
were so abundant, | kept them under cages with flowering Genista plants and some
parasitized ferns. They were always most active in the direct sunshine, and in the first
week of March, when I had a cage of them in the sun, I observed pairing to take place
for the first time ; subsequently I observed it repeatedly.

IT was not fortunate enough to observe oviposition, but I will quote an account
given by Mr. R. Newstead (43 @) (which he has kindly placed at my disposal) of ‘the
occurrence in an allied form—Blastothrix sericea, Dal. ‘ On the 17th of October, 1901,
after long and careful watching, I observed for the first time one of the chalcidid
parasites in the act of laying eggs in the body of a coccid. When first seen, the parasite
was running swiftly from place to place, evidently searching for a suitable host; its
antennze were bent downwards almost at right angles to the long scape forming their

basal half, and were moved up and down rapidly and alternately, the tips each time

|
|

AND DEVELOPMENT OF COMYS INFELIX. 235

touching the path of the insect as it progressed. Many coecids were examined, and,
when a suitable one was found, the parasite turned its head towards the anterior
extremity of the coccid, and, resting with all its feet upon the body of the latter,
inserted its ovipositor into the centre of the thoracic area; it then slowly moved its
abdomen up and down, and apparently laid its eggs in the puncture; the parasite then
withdrew its ovipositor, and, turning round abruptly, feeling its way again with its
antennee, seized with its jaws the lips of the wound made by its ovipositor, and distinctly
closed them upon it and apparently pressed the edges together; finally it passed the
palpi over the wound, and then left the coccid to its fate.”

These insects do not—as is the case with so many species—seek the light; on the
contrary they prefer the shadow. I tested this many times by having them in a glass
vessel, of which one half was illuminated and the other half shaded; they always left
the light for the darkened end. This behaviour is all the more unexpected seeing that in
the sunlight their activity was always so marked, but it may be that they were resting
and therefore preferred to be in the dark. For the most part, they sit inactive on the
plant on which they are bred; they seldom offer to leave it, flying but rarely and then
only under the stimulus of strong sunshine; such flights are short, never exceeding the
distance between one fern-frond and the next. ‘Their most usual mode of locomotion is
walking at a relatively rapid rate, supplemented by sudden jumps, effected doubtless by
the powerful tibial spurs of the second legs; by this leap they often cover a distance of
one or two feet. This rapid running movement gives them a certain likeness to ants,
which they resemble curiously in colour and size, for their wings are carried folded flat
upon the dorsal surface and are inconspicuous. The antennz are constantly feeling
and moving, as is also the case with ants, and at first sight this motion is sufficient to
suggest the resemblance.

(ce) Heonomic Aspects.—From an economic point of view Comys infelix is of signal
importance inasmuch as it destroys one of our most injurious scale-insects. Coccide
attack a great many plants of commercial value and do immense damage; usually the
horticulturist combats their ravages by means of spraying with insecticides, but the
question of parasites is of the greatest importance. In a preliminary paper on
the “ Economic Importance of the Parasites of Coccide ” (17), referring to C. infelix, I
have said that “as far as I can judge from the facts that have come under my observation,
TI am led to rate very highly the value of these parasitic Hymenoptera as destroyers of
Coccid pests. In the case of L. hemisphericum, King’s statement that it is one of the
commonest pests in greenhouses applies to the district round Cambridge as well as to

the United States, and the pest is satisfactorily controlled by the parasite. If the

parasite is not found in other districts where the scale is injurious, it should be introduced
there. Considerable difficulty has been experienced in the attempts to distribute the
predaceous enemies, but in the case of the internal parasites, the task is much simpler,
and success will be easier to attain, for it is only necessary to transmit a small plant

hearing a few parasitized Coccide. From my work on this species Iam led to believe

that the Hnxcyrtine are remarkably tenacious of life in their early stages.” As regards

ie the quarantine regulations that are so strictly enforced in many parts of the world, I

236 MISS A. L. EMBLETON ON THE ANATOMY

remarked that ‘‘ the creatures may be imported on merchandise or by insects as well as
on plants. Neither should it be forgotten, that if.a pernicious scale be once introduced,
then the parasites that may be contained in the scales are excluded. ‘To avoid this, a
knowledge of the marks by which parasitized scales can be distinguished from others, is
really essential in the carrying out of the quarantine regulations” ... “it is of utmost
importance that, previous to any attempt to destroy the Coccide, it should be
ascertained whether internal parasites are present or not. If they are found in large
proportion, then time should be allowed before any insecticide be used, so that the ~
parasites may emerge from all those individuals containing them, so as to allow the
beneficial creature to be perpetuated and increased. After the emergence of the flies,
then the Coccidie, if any such there be, may be killed by using those insecticides which
are especially adapted to the particular case.” There seems, however, to be some
difference of opinion on this subject of the importance of Hymenopterous parasites, but
I find myself more and more convinced of the views expressed in the preliminary
paper, because of the simple fact that having had this scale-insect under observation for
a long time I can testify to the well-nigh total destruction of it by Comys infelix. I
can readily understand the reluctance of those who have not had this advantage to
admit the fact, when I recall that there may be a thousand scales on a plant, and that
though nearly every one may be realiy parasitized and consequently utterly destroyed,
there is nothing to reveal this. One must first understand the signs of parasitism in
the scale before being able to appreciate the work done by the fly. In economic
work of this description it is therefore of the utmost importance that the marks of
parasitism should be recognized. JI have shown a plant covered with parasitized
Lecanium to people used to making biological observations, and they could not believe
that the scales were parasitized and destroyed until they saw the fly actually emerge,
and that although the parasitism was nearly complete.

It is necessary, in dealing with all insect pests, to use much vigilance and caution, but
at the same time to combat them whenever they appear; but this can only effectually
be through a proper knowledge of their habits and life-histeries, for it is possible, as
already stated, to do harm by destroying some beneficial creatures which live upon the
pest. In countries which have been under cultivation for centuries a balance between
the destructive and beneficial forces has been produced by adaptation, and in this way
the destructive species are kept in check; it is therefore essential that care be taken not to
remove any of the factors which help to maintain this necessary and delicate equilibrium.
It is, at the same time, of utmost consequence that as soon as an injurious creature ©
makes its appearance, it should be watched carefully; in the event of a species coming
into the country, in the very early stages it may be dealt with successfully by resorting
to hand picking, giving it, if possible, no chance to become established. An added
danger to the introduction of an injurious species from another country, is the fact that
in its native land it is accompanied by some enemy which keeps it in check; but when ©
it sets up in the new home, it is free from this restraint and spreads enormously in —
consequence. In the case of parasites of injurious creatures it often happens that they —
can also be introduced with ease, if, for example, they happen to be beetles, flies, birds,

AND DEVELOPMENT OF COMYS INFELIX. 237

lizards, &e. It is a recognized method to reduce a pest by introducing some enemy
from another country. This has been done in America against injurious scales, by
introducing little Chalcid flies from Europe.

Thus the basis of economic entomology is life-history work, which has wider bearings
than merely following and observing the different stages in the life-cycle of an insect,
for all these separate pieces of work on life-histories contribute towards the solution of
the great biological problems of distribution, adaptation to environment, parasitism, and
even add to the data on the questions of heredity and variation*. In order to arrive at any
knowledge of how to control injurious insects, it is obvious we must first learn the facts
of their life-histories, together with a knowledge of the distribution, food, enemies,
climatic needs, &c., until its limitations can be outlined with approximate correctness—
then any steps taken to deal with it will probably be successful. This subject is at
present warmly debated in various parts of the world, and some contend that in the long
run it is better to trust to Nature than to extensive artificial operations. My contention
is, that we may only trust to Nature when we have obtained a knowledge that will
warrant us in so doing, and that will probably enable us to restore natural conditions
when they have been abruptly infringed.

Since my paper, above referred to, was published, Professor Antonio Berlese, of the
R. Scuola di Agricoltura, at Portici, has taken up this question and published two
important papers on the method of Economic Zoology; and as he takes the same
view of the subject as I have done, I hope he may be successful in his effort to attract
greater attention to the economic importance of the parasites of Insects.

III. DEVELOPMENT, AND STRUCTURE IN EARLY SrTaGEs.

(7) Eee.

I have examined the ovaries of a large number of the flies immediately after their
emergence from the scales: though I cannot f say how long the specimens had been in
the imago condition inside the host, I found the eggs were present in the tubes in an
already advanced stage of development (PI. 11. figs. 2-5). It usually happened that the
ovaries became crushed so that the eggs were most of them free. I have not been able
to decide definitely whether the different stages of growth of the eggs are connected with
different positions in the egg-tubes; but probably it is so, the egg nearest the oviduct in
each tube being the most advanced. Except those in the very earliest stages, the eggs
had the appearance of two masses of yolk connected by an isthmus; it is clear that this
isthmus ultimately becomes the appendage of the perfected egg, which, as will be seen.

* A German economic entomologist, Dr. L. Reh, of Hamburg, is at present engaged in enlightening public
opinion as to the value of knowledge of this sort.

t The reason why I am unable to say when the imago condition is reached, is because the creature is inside the

Coccid, and to expose it is to destroy it, and so no individual case can be traced. ‘This difficulty will always be met
____with, and is insurmountable. From external signs it is possible to tell roughly when the creature pupates, because

Pthen it turns black, and this can be detected through the shell of the Coccid. But there is no such clue to the time
when the imago condition is assumed,

>

238 MISS A. L. EMBLETON ON THE ANATOMY

from Plate 11. figs. 6, 7, 8, is of a very remarkable form. It was difficult to decide in
certain cases whether this neck was or was not equally connected with both masses ;
if so, it certainly soon becomes the exclusive property of the mass that will become
the egg. There is a considerable difference in the two masses, more especially as regards
the contents. The future egg-mass is almost oblong in shape, while the other portion is
spherical and smaller; the granules in the former are larger and more concentrated than
in the latter. It is a fact of interest and of some significance that the two masses or
parts of the egg always respond in strikingly different ways to various staining reagents ;
for instance, when treated with methyl green in acetic acid and mounted in dilute
glycerine, the true egg-mass and isthmus show a clear green-blue colour, while the other
mass is purple. I observed that this stain behaved with similar distinction in other cells
belonging to different parts of the body, the nuclei presenting the same colour as the
egg-mass, and the cytoplasm of the cells the same as the yolk-mass. I could find no
nucleus in the egg.

In the earliest stages the egg is an oval body containing coarse granules ; this stage is
followed by one in which a slight constriction is observable in the middle; this gets
more and more marked until the condition of two masses connected by a fine isthmus is
assumed. All these stages are found within a membranous covering (Pl. 11, figs. 2-4).
In the latest stages a curious structure is present on the neck part of this dumbbell-
shaped egg; it is a valve-lke projection, pointing towards the yolk-body like a lip.
There is still a connection between the two masses (PI. 11. figs. 6-9), but where this lip
is situated there is an appearance of fine papillation or striation on the wall of the tube.
At the distal end of the yolk-mass there is evidence of some thickening of the wall, but
T cannot find any definite structural peculiarity beyond this. This dumbbell-shaped
condition measures *85 mm.

T also examined flies which I had kept alive for a week or two on flowering plants,
and in them I found eggs as stalked bodies ; that is, the yolk-mass had disappeared, and
left only the true egg-mass with the isthmus appearing as a stalk. ‘This appendage ends
as a curious forked apparatus, representing the valve or lip present on the connecting
tube in an earlier stage. The papillated appearance is now more conspicuous; in the
neck of the stalk, at the point where it forks, there is a plug of protoplasmic matter
(Pl. 11. fig. 7). From this stage, though I could not observe the act of oviposition, I
traced the egg to its destination in the body of the scale, where it is always situated
on the right side of the anus (if one looks at it dorsally, towards the head). The form
of the egg now is identical with that when last seen in the fly; it measures -25 mm.,
and always has the plug of matter in the stalk. In older Coccids, where the young
larva is present, then the tail of the larva is seen to be capped with the remains of the
stalked egg-case, the body of the egg having split into two when the larva emerged.
This stalk measures ‘05 mm. .

To summarize the processes I have described, I may say that the mass of matter from
which an egg is developed becomes constricted in the middle until it finally assumes a
dumbbell form; 7. e., two masses are connected by an isthmus. In this condition,
changes between the two masses take place by means of the isthmus, as the result of

AND DEVELOPMENT OF COMYS INFELIX. 239

which, one mass becomes predominant and forms the egg; a portion of the isthmus
remains attached to the egg as its stalk or appendage.

It is desirable to call attention to this, as I am not aware that any form of oogenesis
quite similar to this has been described. I may add that, when I had made only a few
observations, I thought it probable that the original mass of formative material divided
into two, and that we might find ourselves in the presence of a condition suggesting an
initial stage of dissociation of the embryo, but I was soon convinced that this was not
the case. The fact that chemical stains and reagents acted differently on the two masses
is of importance, indicating an essential dissimilarity in the nature of the two halves.

(6) Larva.

In tracing the early life-history of the fly a serious practical difficulty is met with,
inasmuch as it is impossible to follow the same individual through the different stages,
for to be able to observe it in the scale it must be exposed, and thus killed. In this
way there is no certainty regarding the exact continuity of the observations; I have
endeavoured to obviate this as much as possible by examining a great number of the
parasitized scales in all stages of growth and at all times of the year; even so, lam
sorry to say that there are some links missing from the chain.

General points.—In preparing these larvee for the microscope, the best results were
obtained with those which had been removed from the Coccid, and then suspended on
a watch-glass in the vapour of osmic acid over the mouth of a bottle containing a
‘5 per cent. solution. In a minute or two the larve look brown; they are then washed
carefully in water and mounted in glycerine. Others were mounted in Canada balsam,
but in these preparations some shrinkage always occurred. Some I stained with borax-
or alum-carmine, but the osmic acid preparations possessed many advantages. This
applies also to the later stages in the metamorphosis before the pupal stage is reached,
when the creature becomes black and chitinous.

i. First observed Stage. (Plate 11. figs. 10, 11, 12, 13.)

External form.—t refer to this as the first “ observed” stage, because it is probable
that the larva immediately on hatching may be different from the youngest specimens
I have succeeded in finding; the hypermetamorphosis may be greater even than that
which I shall describe.

The larva of the fly is found in the younger Coccide as a soft white tapering maggot
‘75 mm. in length; the head is not differentiated in any conspicuous way from the rest
of the body, the anterior end being merely bluntly rounded. Behind what may be called
the “oral segment” are thirteen segments, including the bifurcated tail-segment ; this
segment is a most unusual structure, for the body terminates in two long tapering tails,
each with a tracheal trunk continuing to its tip, suggesting that this is an adaptation
subserving respiration. These delicate prolongations are always embedded near the anus
of the host, and very frequently are capped with the old egg-case. The question of the

_ respiratory significance of this apparatus will be discussed later.

SECOND SERIES.— ZOOLOGY, VOL. Ix. 33

240 MISS A. L. EMBLETON ON THE ANATOMY

Tracheal system —The main trachez run in two lines parallel with the sides of the
body; being full of air, they can be seen plainly through the semitransparent body-
walls ; anteriorly they are connected by a transverse commissure in the second segment
behind the head ; there is also a posterior connection.

Bugnion (9) states that there are nine pairs of stigmata in the larva of Hncyrtus fusci-
collis, occurring in the segments two to ten. Reinhard (47) has counted the same
number in Pteromaline larvee (Decatoma, Callimone, Eupelmus, Pteromalus). Laboul-
béne (35) describes also nine in the larva of Pimpla Fairmairei; but Ratzeburg (46), in
the larva of Anomalon circumflexum, a parasite of Bombyx, found no stigmata at all,
which he says is due to the fact that the larva lives inside the caterpillar. In the larva
of Comys infelix at this stage I have been unable to detect any stigmata, though I have
employed many methods in preparing the larve for the microscope, but always with
negative results. Possibly the tail-apparatus takes the place of stigmata, for it is a most
remarkable arrangement, and suggests a parallel with the metapneustic tracheal systems
of some Dipterous larvee in which there exist only a terminal pair of spiracles. In some
metapneustic Dipterous larvee a pair of anterior spiracles appear later; this is then
called an amphipneustic larva. The arrangement in which the normal stigmata are
present is called peripneustic.

Alimentary system.—In the bluntly rounded anterior end there is a circular mouth
with a soft rim or lip (Pl. 11. fig. 11); inside there are two chitinous mandibles, each
being a simple tooth or claw, one of which overlaps the other; these mandibles possess
a powerful muscular apparatus. From the mouth, the alimentary canal runs to the tail ;
immediately behind the mouth there is a funnel-shaped pharynx, leading through an_
cesophagus to the stomach, in about the third segment. On either side of the stomach
is a gland, probably possessing a salivary function (Pl. 11. fig. 15); these glands open
into the pharynx. The alimentary tract is straight, apart from the stomach, which is
sac-like and fills the larger part of the body-cavity, and contains fat-globules and other _
dark-coloured granules; it looks darker than the rest of the body, which is white, and,
under a high power of the microscope, is seen to be granular. There is a short intestine
behind the stomach, but at this stage it is not connected with the anus.

ii. Second Stage.

External form—The body now measures 1:75 mm. As the two-tailed larva grows its —
contents get more and more aggregated into ball-like masses. The body becomes rounder
and thicker, while the tails atrophy (Pl. 11. fig. 13), till the stage is reached (fig. 14) in
which the posterior end is rounded, but the two tracheal trunks are still visible projecting
out of the body as withered ends not yet quite cast off.

Tracheal system—Apart from the difference in the tail-apparatus, the two main
trachere remain unaltered except for the fact that in each segment they give off a
group of secondary branches. In this stage I have found the anterior and posterior
commissures still present ; in front of the anterior connective the main trunks continue,
ultimately ramifying in the head; behind the posterior commissure the trunks are.

AND DEVELOPMENT OF COMYS INFELIX. 241

continued, finally projecting out of the body as withered atrophied remains of the tail-
apparatus (Pl. 11. fig. 14). Between these two transverse connectives are eight pairs of
side branches, more marked than the finer anastomosing branches of tracheze that, as
stated, are given off in each segment. Of these eight pairs of conspicuous side branches,
the first and last have each a definite spiracle, but the six pairs between have no spiracles,
or else they are so small as to be unfunctional. This is an interesting condition between
the peripneustic and the amphipneustic arrangements.

Alimentary system.—The mouth and mandibles show no alteration, and the alimentary
canal is unchanged except that its contents are more marked in colour and quantity,
and that the communication with the anus is now probably established. The coloured
refringent granules in the stomach are most noticeable.

ii. Third Stage.

: External form.—The larva now has become more swollen and measures just over
38mm. in length (PI. 11. fig. 15), and it is only very occasionally one finds traces of the
tail-trachez outside the body.
Imaginal discs.—It is now possible to discern in outline the early rudiments of the
future organs of the imago: these rudiments Weismann called “ imaginal discs ” ;
Kiinckel d'Herculais (34) called them “ histoblasts,” and Miall refers to them as
“imaginal folds.’ Gradually they assume a form (PI. 11. fig. 15) in which the antenne
can be seen as a pair of plate-like bodies near the mouth. Behind and almost above
_ these is a pair of circular discs from which the eyes will arise. The buds ‘which will
7 ultimately form the buccal folds are also discernible. ‘The leg-rudiments (Pl. 11.
fig. 15, /.‘-*) come behind the eyes, and on the outer margin the wing-rudiments are
plainly traceable. Nearer the posterior end of the body are two oval masses of cells,
representing the future sexual glands (Pl. 11. fig. 15, s.g.)—ovaries or testes; and
_ following closely on these are three pairs of bud-like bodies, which are the rudiments of
the sting (Pl. 11. fig. 15, st.’~).

Tracheal system.—The tracheal trunks still have a lateral course, connected by an anterior
_ and posterior transverse commissure. The stigmata are as in the preceding stage, but in
- connection with each of the four spiracles described as functional in that instar, there
_ is an apparatus by which intercommunication is apparently set up between the respiratory
systems of the host and parasite (Pl. 11. fig. 15). From each of these four spiracles
there is a double tracheal tube running out into the host’s body; these two branches
become subdivided and ramify in the host-tissue, and by this means, it seems, the
parasite obtains its air-supply. Whether these ramifying trachez originate from the
tracheal system of the Coccid host, or from that of the parasite for whose benefit we may
_ presume they are functioning, is quite obscure. If the former be the correct inter-
pretation, we must conclude that the host develops a respiratory system for the benefit
of the parasite that is destroying it. If, on the other hand, the structures originate
ith the parasite itself, we have to deal with the remarkable fact that they constitute a
tracheal system entirely outside of the creature to which they belong. Buguion says that in
33*

242, MISS A. L. EMBLETON ON THE ANATOMY

Encyrtus fuscicollis the trachez of the host ramify on the tube containing the parasite,
and serve for respiration of the young Lncyrtus. The only other case I can find bearing
on this point, is that described in 1837 by Dufour, of Andrena aterrima and a parasitic
Dipterous larva*. Dufour claims that two of the host’s trachee grow into the body of
the larva, which is thus supplied with air; but there seems some doubt about the inter-
relationship in this case.

As this instar passes into the next, the posterior pair of external trachez become
marked by the appearance of three red-brown plate-like structures on each side of the
body of the larva (Pl. 11. figs. 16, 17, 18, pl.) ; they are situated at the point where the
radiating external tracheze are united with the body-wall. These curious plates are
present in the pupal condition, but are left behind when the pupal skin is cast.

Thus we see, in summarizing the facts recorded above, that in the larval condition
respiration is at first carried on by means of a bifurcated tail-apparatus in which the
two lateral tracheal trunks are continued; in this condition there are no spiracles, but
later on the two tapering tails atrophy, and spiracles are developed. This condition is
superseded by an arrangement of trachee ramifying outside the parasite, in the body of
the host; there are now 8 pairs of stigmata, and the posterior pair of radiating tracheze
arise in the 9th segment behind the head.

Before the next (or prepupal) instar is reached, a process of histolysis takes place by
which the internal organs are completely broken down, to be built up into the pupal
tissue by means of a process of histogenesis, or regeneration.

Jn this instar, or at any rate during the later period of its existence, the larva appears
to possess a membranous coat, or sheath. It is toa membrane of some kind or other
that Bugnion refers in the passage quoted above as to the respiration of the larva of
Encyrtus fuscicollis: he states that the embryos are all in one long common capsule,
and he is of opinion that this is a product of the egg; but the case differs so widely from
that of Comys infelix, that his facts cannot be of much service in making a parallel
argument. As to the nature and origin of this sheath in Comys infelix, I am quite in
doubt. Subsequently, as we shall see in the next instar, there exists a well-defined and
ample membranous sheath, but whether the membrane observed in the third instar is
really the same as that found later, or not, I cannot say with any certainty; no actual
ecdysis has, however, been observed, and I think it is probable that the sheath is a
larval skin retained on account of the respiratory apparatus attached to it externally.
The question is complicated by the appearance in the next (or prepupal) instar, at the
base of this respiratory apparatus, of additional structures in the form of certain oval
red plates (Pl. 11. figs. 16, 17, 18, p/.), which appear to be applied to the sides of the
body at the spot where the tracheze radiate out. if these were developed outside an old
skin, it would be a very remarkable fact: I have alluded to the possibility that these
respiratory structures are really formed by the host, and in that case there would be no

difficulty in understanding that they undergo further development even on the separated
skin of the parasite.

* Of. Cambridge Natural History, Insects, ii. (D. Sharp), p. 26, fig. 13.

a

=

AND DEVELOPMENT OF COMYS INFELIX. 248

I have been unable to settle this question, for in order to do so with any finality it
would be necessary to find the creature at the exact time when the change is taking
place, and this would entail the examination of a vast number of individuals, and as
each examination involves a careful dissection of the small host first, the time required
would be considerable.

It is at this point in its life-history that the creature turns completely round inside
its host.

iv. Prepupal Stage. (Plate 11. figs. 16, 17.)

External form.—A great change has taken place between this and the preceding
condition, transforming the larva into the prepupa, or pseudonymph. In this instar
the creature is still creamy-white, like the larval forms already described; it is, however,
enclosed in a membrane (PI. 11. figs. 17, 18, m.), which is probably the remains of the
original larval cuticle, and the parasite is now head to tail in the Coccid, instead of
lead to head as was the case in the larval condition; its dorsal surface is now closely
pressed to the inside of the dorsal surface of its host, whose body it now almost
completely fills. ‘The appendages, &c., of the imago body are well defined, though still
imperfectly developed in this instar (Pl. 11. fig.17). The head is differentiated from the
rest of the body, the antenne and eyes being clearly outlined, while the thoracic region
has become more rounded. The rudiments of the legs and wings are folded across the
ventral surface, and the segmentation of the abdomen is distinct, so that the creature
now exhibits very definitely the form of the perfect insect. Internally, the organs have
undergone less change as a result of the histolysis and subsequent regeneration.
Packard (44, p. 67) has called this form the semipupa, and Siebold’s term pseudonymph
applies to the same instar; the nymphal form is attained after an ecdysis.

Tracheal system.—There are in this instar four large stigmata—three situated in the

_ thoracic region and one in the abdominal segments, with six smaller abdominal stigmata

which may not be functional, for only traces of the first five can be seen in the imago,
and the seventh has entirely disappeared, or has fused with the sixth to form the large
functional spiracle which is so conspicuous.

Situated on each side of the abdomen there are three oval red plates (Pl. 11. figs. 16,
17, 18, pl.), referred to above. These appear to be connected in some way with the
respiratory function of the creature, for they are placed at the base of the large forked
tracheal tube, which at that point effects its union with the body, or, in other words,
branches out of the body at that point, to ramify over the membrane which encloses the
insect. This inembrane has been discussed when the preceding instar was being
considered. Each of these curious plates is an oblong body applied closely to the outside
wall of the abdomen; under a high power of the microscope, one of these plates is
seen to contain red globules, of different sizes. It is not easy to determine what the
function of these structures can be, but I suggest that they have some connection with

respiration, seeing the radiating trachew arise at the same place as these anomalous

plates. Nothing of this nature has been recorded by Bugnion as occurring in Hneyrtus
ruficollis, nor can I find elsewhere any statement bearing on these curious structures.

244, MISS A. L. EMBLETON ON THE ANATOMY

v. Pupa. (Plate 11. fig. 18.)

External form.—The pupa measures 2°25 mm. in length. The earliest pupal form is
not of a uniform black colour, the dorsal surface alone being black. The fully formed
pupa is, however, completely black and chitinous, though still enveloped by the cuticular
membrane whose origin has been already discussed. The first organs to assume their
final colour are the eyes and ocelli, which early become charged with dark red-brown
pigment, at a period when the rest of the body is creamy-white; gradually the colour
deepens all over the thorax and abdomen, though the head, apart from the eyes and
ocelli, remains almost transparent for a considerable time. When the body has become
uniformly black, the creature is in an inert, passive condition, unlike the preceding
larval stages, in which slow movements from side to side take place as a result of
external stimulus. But after a time this black, inert pupa becomes the perfect insect,
and breaks through the enveloping membrane (Pl. 11. fig. 18, m.) and the old pupal
cast skin, and with its mandibles cuts a small round hole in the dorsal shell of the
Coccid, and escapes, to lead its free existence.

Tracheal system.—The abdominal plates (fig. 18, pl.) and the radiating trachez are still
present, but when the imago emerges they are left behind with the old sheathing
membrane. The pair of tracheze which radiate out from the head region never develop
these oval red plates at their bases; it may be that these plates are more connected with
excretion than respiration. The nymphal respiratory arrangements are apparently
very transitory.

In order to obtain, if possible, more detailed and exact knowledge of the larva and
pupa while still within the host, I cut the undissected host into a series of sections with
the microtome. To do this I tried various methods for softening and cleaning the
chitin, but the most successful results were obtained when I employed eau de Javelle
or eau de Labarraque, as suggested by Looss. The chitin in this way was rendered
transparent and permeable to reagents. I first hardened the specimens, and then
left them in 25 per cent. eau de Javelle for 18 to 24 hours, afterwards washing out
thoroughly with water. They were then dehydrated, embedded in paraffin, and good
sections were obtained ; sometimes I stained, before embedding, with alum-carmine or
picro-carmine (five to six days), but usually I found it best to stain the sections when on
the slide.

But though the sections were good, there was little or nothing to be‘made out by this
method of investigation, which was not seen better in macroscopic dissections ; except,
of course, histology, but that I have been content to omit.

IV. Anatomy oF Imago, °.
(a) Measurements.—Length (without antenn) 2°5 mm.; length of thorax ‘9 mm. ;
width of head ‘75 mm.; width of thorax *74 mm.; width of abdomen ‘74 mm.; length
of abdomen ‘85 mm.; length of antenne 1:25 mm.; extent of fore wing 15 mm. ;
extent of hind wing 1:25 mm.; width of fore wing “62 mm.
(b) Antenne (Pl. 11. figs. 19, 20; Pl. 12. fig. 21).—The antenne are relatively long -

AND DEVELOPMENT OF COMYS INFELIX. 245

and club-like, composed of eleven segments, the funicle being six-jointed, while the
pedicel is shorter than the succeeding segments. From the third segment, the joints
which go to make up the flagellum decrease in length, but increase in width and become,
in the female, compressed towards the tip; in the male they remain almost cylindrical.
Except for the pedicel the whole organ is clothed with fine black hair. The antenna is
light brown at its base, shading to black in the flagellum.

(c) Head (Pl. 12. figs. 26, 27)—The surface of the head and face is curiously
sculptured and punctured ; it is golden-brown, almost black beneath the eyes, which are
very dark and composed of fine facets devoid of sete. Between, and slightly behind, the
eyes are three ocelli, the two basal being nearer the true eyes than they are to
one another.

(d) Mouth-parts.—The maxillary palpi are the most conspicuous parts, being black and
projecting laterally. The mazil/a (Pl. 12. fig. 24) consists of a stout tooth made up of
three joints, the second bearing the palp on its outer surface and a small club-like
organ on its inner surface; there is next a broad flattened part of two halves; the palp
comprises four pieces, and ends in a curious club-like structure with long sete on its
inner face and tip. ‘The lower lip can be seen as a median structure below the
mandibles (Pl. 12. figs. 22, 23); laterally it has a foliaceous setose apparatus united
basally by a triangular membrane, to which is attached a labial palp on either side; this
is made up of three segments. The mandibles are very simple, curved, plate-like organs,
with sete ; they are approximately triangular in outline, the apices curving inwards, the
bases being just above the maxillary palpi (figs. 22, 25). Overhanging the whole mouth-
apparatus is an upper lip, with a setose edge (fig. 22, w./.).

(e) Thoraz.—Except for the shining black collar (PI. 12. fig. 29, ¢.) the thorax is
yellow with conspicuous darker patches of pigment and black hairs. The pronotum
(Pl. 12. fig. 29, pr.) is almost circular; laterally it carries the small dark tegule.
Behind the pronotum is the raised scutellum (fig. 29, sc.), a triangular area with the apex
pointing backwards, and terminating in a tuft of long stiff black hairs, arising apparently
in two longitudinal rows. The mesonotum and metanotum are black and glossy, the
former serving as attachment for the fore wings, and the latter for the hind wings and
third legs (Pi. 12. fig. 28). The thoracic spiracles are arranged in three pairs (PI. 12.
fig. 29, st.~*); the mesothoracic are in a suture, only seen by looking from the side; the
metathoracic are conspicuous (figs. 28, 29).

(f) Fore wing (Pl. 12. figs. 19, 20)—The wings are mottled with black fuscous areas,
and are covered with fine hairs, forming a fringe round the distal margin. At the base
of the short incurving vein is a group of dark sete, with a second group just behind.
From the description of the genus as given in Comstock’s report, quoted above, and
from my figures, it will be easy to see the disposition of the wing-veins.

(g) Hind wing (Pl. 12. figs. 19, 28).—These wings are smaller and more delicate than
the fore wings, and are not mottled, being uniformly transparent and clothed with fine

; short hairs.

(h) First legs (Pl. 12. fig. 31).—The core are silvery white, while the femora are

dark, separated from the coxz by two trochanters; the tibie are slender and fuscous,

246 MISS A. L, EMBLETON ON THE ANATOMY

the spur being no longer than the first tarsal joint; the tarsi comprise five segments,
the last. being the claw, made up of two hooked down-curving setee. The whole leg is
covered with short dark hairs.

(i) Second legs (Pl. 12. fig. 30).—These differ from the first only in colour and in the
length of the tibial spur. The cove are black, the two trochanters rather paler, while
the femora are almost white; the ¢édie@ are yellow, and dilated distally; the spur
is long and powerful, and by its means the creature performs the characteristic hopping
movements; in preserved specimens the second legs project beyond the third (Pl. 12.
figs. 19, 20, 30).

(j) Third legs (Pl. 12. fig. 28)—The small rounded cove are silvery white, the femora
are brown and the ¢ibi@ black, while the ¢arsi are almost white, the tip of the claw being
black. The spur is of normal size, as in the first leg.

(k) Abdomen (PI. 12. figs. 32, 33, 34, 35, 36, 37).—The abdomen is of a deep glossy
black, with numerous long dark hairs, and ventrally a soft down of short hairs. There
are seven segments succeeding the petiole. The structure of the exoskeleton of the
abdomen in the perfect insect is not easily ascertained, for there are many obscure points,
and it was only after dissecting a great number of specimens that I was able to arrive at
even an approximately correct knowledge in this matter. After trying many methods,
I found that the most satisfactory specimens were those that had been macerating for a
day or two in water, until they were quite soft and greatly distended ; in such prepara-
tions the black abdominal plates stand clear of each other with the swollen white
internal tissue showing between: this was most useful, for in such a condition one could
clearly and unmistakably make out the relative positions of the parts before dissection.
Another advantage is that the pieces of the exoskeleton can be removed one by one if
ereat care be used, and the whole series put together again on a slide and mounted in
Canada balsam (PI. 12. figs. 34, 35, 36,37). An alternative method, which often proved
very helpful, was to leave the flies in 10 per cent. (or 5 per cent.) KOH for 24 hours,
when all the soft parts are destroyed, the chitinous parts remaining uninjured and in a
condition in which they can be easily separated from one another; when properly
washed, dehydrated, cleaned, and mounted in Canada balsam, the structure is seen to
even better advantage than in the former method, for the chitin has become, to some
extent, transparent. Nevertheless, the difficulty always remains that curved and _
rounded pieces of the armature must be flattened when mounted, rendering some
distortion unavoidable. This difficulty is met with chiefly in the matter of obtaining a
correct interpretation of the structures which go to make up the petiole (figs. 34, 35, 36,
37). This has a dorsal and a ventral face, each composed of two small chitinous plates,
of which the dorsal pair is much the thicker and stronger, being always conspicuous in
mounted preparations, while the ventral pieces need careful search before they are
discovered. In these flattened mounted specimens, the dorsal factors of the petiole
appear as a dark mass situated between the side wing-like parts of the abdominal plate
immediately behind the petiole (figs. 34-37). In life these two lateral flaps of the —
abdominal plate curve round the abdomen to the antero-ventral face, leaving the petiole —
standing out in front; its dorsal elements also curve round, forming a sort of short tube,

oo

n5 jew;

AND DEVELOPMENT OF COMYS INFELIX. 247

the remaining (ventral) face of which is completed by the ventral pieces of the peticle.
This dark basal portion of the petiole-ring is made up of two separate elements, meeting
in the middle line, where they are very narrow; at the sides they are much deeper
(Pl. 12. fig. 36), so that, in outline, the two together are cup-like in mounted specimens,
with an almost straight base resting on the segment behind the petiole; the two sides
are prolonged to form blunt points, which articulate with the propodium and make a
concavity between, filled by a pallid membrane, like a semicircle between two horns.
The (anterior) rim of this semicircle is a clear strong chitinous band. The chitin of
these parts is smooth, and devoid ot surface reticulations such as mark the abdominal
plates. Ventrally (Pl. 12. figs. 36, 37) the structure of the petiole is very similar
to that seen on the dorsal side, but the chitin of which these parts are composed
is thin and reticulated, making them obscure and difficult to detect. Another difficulty
too in investigating the ventral structures arises from the fact that these almost
transparent parts often remain attached to the metathorax, and so get lost. But with
careful manipulation it is seen that they are borne between two lateral horn-like projec-
tions of the segment abutting on the petiole, and that their anterior edge is not concave
as in the dorsal region, but straight or slightly convex. The two strong outstanding points
from the ventral abdominal segment apparently form the articulation, and in the natural
position would point towards the dorsal surface on the face that meets the thorax; they
would not point up and forward, as is necessarily the case when the parts are flattened-
out under a cover-glass. This petiole, therefore, differs greatly from what prevails in
either the Aculeate Hymenoptera or the Ichneumonidze. Dorsally, behind the petiole
there are seven separate plates (figs. 34, 36). Each plate, except the last, is made up of
a median area connecting two lateral flaps. In the first of these (and what must be
looked upon as the true second segment), the lateral parts are broad and wing-like, their
function being to fold round the anterior face of the abdomen, curving towards the
ventral surface; this segment bears no trace of a spiracle.

The 2nd, 3rd, 4th, and 5th plates form a progressive series, each with an atrophied
spiracle. The flaps are bent back more in each segment (Pl. 12. figs. 34, 36) until in
the 6th this curvature reaches its maximum, for the side parts are very slender until
near the tips, where they expand. At this point there is the only functional spiracle in
the abdomen: it is a large circular orifice (Pl. 12. fig. 40) surrounded by a chitinous
ring; in the clear centre can be seen a tongue-like structure; the large trachea can be
traced running from this spiracle into the body. The median part of the seventh plate

is slender and deeply concave in front. The last dorsal plate has no lateral expansions ;

at its points on either side is situated a remarkable organ, which, for want of a better
term, I refer to as the “ tactile plate” ; it is over this that the preceding plates have
heen looped, as it were (figs. 33, 34, 36, 39). This organ consists of a pear-shaped plate
(figs. 42, 43) placed transversely to the length of the body, the pointed end being on the
inner side; the plane of the plate is at right angles to the dorsal surface. The plate

itself consists of a membrane stretched over a loop-like ring of chitin and supplied

‘strongly with nerves; atits pointed inner end there is a solid chitinous rod, which, in the

natural position of parts, points forwards at right angles to the length of the oval plate ;

SECOND SERTES.—ZOOLOGY, VOL. IX. od

:
ie

248 MISS A. L. EMBLETON ON THE ANATOMY

this rod is almost the same length as the plate, to which it is attached by a hinged joint,
as may easily be proved by moving the rod to and fro on the plate-rim. The plate itself
carries four long setze, the two inner being the biggest, and the outer the smallest. It
is hard to say what is the function of this curious apparatus. I have watched the living
insect walking about for hours on a fern-frond, hoping to get some clue as to the use of
this “‘ tactile plate”; [have examined its structure microscopically, but cannot be certain
of its function. In life, the long hairs were always standing out conspicuously from the
side of the abdomen, and occasionally it seemed they were brought into a more erect
and bristling attitude, but beyond that I observed nothing.

Turning now to the ventral elements of the abdominal exoskeleton, it is found that
there are six large segments (PI. 12. figs. 32, 35). Behind these there are two small sub-
circular pieces, one on each side of the base of the sting, and two large pieces still
further behind these, nearer the middle line close to the apex of the ovipositor; probably
these are the vestiges of the seventh and eighth segments. The ventral plates are
almost all alike, except that the one adjoining the petiole has the two horns pointing
forwards (cf. petiole). The others all have the lateral parts rounded at the ends, and
the surfaces are reticulated and bear fine hairs, becoming more numerous on the
posterior segments. The last of the veutral series of plates covers the ovipositor, which
is almost entirely hidden in life, though sometimes in a lateral view the tip may be
seen protruding slightly (fig. 33).

I find but little information in books as to the structure of the abdomen of Chalcidide.
More attention has been given to the external anatomy of the Aculeata, but the structure
of the abdomen in Comys infelix is so different from that of the Aculeates that the
homologies of some of the parts are still an enigma to me. Bugnion, who dealt with
the anatomy of a form allied to Comys infelix, left the difficult points of the morphology
untouched; and he is also silent as to some of the important points of simple anatomy
of this part of the body.

Ovipositor (Pl. 12. fig. 41).—The ovipositor or sting has been largely studied in the
Aculeata, but I can find little information about it in the Chalvidide. Certain portions
of this organ in Comys infelix are so different anatomically from the sting of the
Aculeates that it appears at first sight almost hopeless to homologize the two by mere
comparison, and it is necessary to thoroughly understand the development to arrive
at really permanent results. As regards this a preliminary difficulty exists: it is now
generally admitted that the sting of the Aculeates is developed from appendages situate
in the pupa on the eleventh and twelfth body-segments, the segments themselves being —
subsequently very highly modified to form parts of the complex apparatus. Bugnion —
has figured the ovipositor of Encyrtus fuscicollis, and it agrees largely with what I find —
to exist in Comys infelix; he also figures the buds of the appendages in the larva, and
shows that they arise in three different segments, viz., the three in front of the anal
segment. If this is correct, great caution is necessary in comparing a sting formed by |
appendages of three segments with a sting that is developed from appendages of two
segments only. I have no information to give on this point in Comys, and shall therefore —
briefly describe the structures I have figured.

AND DEVELOPMENT OF COMYS INFELIX. 249

The ovipositor is furnished on either side with two large expanded chitinous plates ;
these are held together anteriorly by a three-cornered nodule of chitin (Pl. 12. fig. 41).
Of these two pairs of plates, the inner is the larger; each half carries a fringe of short
bristles round the distal border. The outer edge of this plate is strengthened by a
thickening of the chitin; this thickening leaves the margin and turns into the middle
of the plate before it reaches the extremity. Where these two inner plates articulate
with the triangular nodule, there arises a curved rod which runs round and back, meeting
its fellow of the opposite side in the middle line; these two curved bars end near the
extremities of the inner plates, and they form the pointed stinging-apparatus of the
ovipositor. A little below the level of the articulating nodules, these two curved rods
converge together and are embraced by a thin membranous sheath; at this point the
enveloping sheath expands into a pair of oval claspers or clamps, through which the rods
of the sting run. The membrane is continued on as a covering or director to the sting-
points. At first I was uncertain as to whether these rods were separate from the
membranous sheath or not, but later on I found it was possible to dissect them out as
perfectly free structures, grooved to the tip from the point where they become swathed
in the sheath.

As regards the origin of the parts of the ovipositor—or sting—in Hymenoptera, there
seems to exist a considerable amount of confusion in the literature on the subject; the
point on which I find authors disagree is as to whether the three pairs of buds (or
imaginal discs) present in the larval condition belong to two segments or to three.
Bugnion says they arise in three different segments representing three pairs of
appendages, and he figures them accordingly; he says: “la ¢ariére (gorgeret, stylets et
valves), représentant également trois paires d’appendices, dérive de six petits disques qui
se montrent dans la seconde moitié de la période larvaire, de chaque cété de la ligne
médiane, a la face inférieure des trois derniers segments.” . . . “ J’ai observé moiméme la
formation des six disques de l’armure génitale chez les larves d’ Zucyrtus, la transformation
de la partie centrale de ces organes en petits bourgeons digitiformes, puis la division des
deux bourgeons intermédiaires en quatre (?), mais il ne m’a pas été possible de suivre leur
développement ultérieur.”” He seems to be in some doubt as to the division of the
median pair into two pairs of buds. Kraepelin (33), in Apis mellifica, states that in the
earliest instar ‘an der Bauchseite der drei vorletzten Segmente findet man um diese Zeit
je ein Paar langlich runder Wiilste, welche, von Tracheen umsponnen und augenschein-
lich der Hypodermis entstammt, man nach Weissmann’s Definition als Imaginalscheiben
zu bezeichnen das Recht hat.” . . . “* Bald zeigen diese Wiilste weitere Differenzirungen,
namentlich die des dreizehnten Segmentes. An letzteren gewahrt man nach kurzer Zeit
eine Langstheilung, derart dass jeder Wulst nunmehr aus zwei nebeneinander liegenden
cylindrischen Zapfen besteht, deren basale Theile unter sich wie mit dem correspond-
irenden Wulst der andern Seite verbunden sind. Die Wucherungen des zwolften
Segmentes sind zu linglichen, gekriimmten Zapfen geworden, wihrend die des elften

Ringes zwei rundiche mit je einem langen Faden in Verbindung stehende Blasen

repiasentiren.”
3d*

250 MISS A. L. EMBLETON ON THE ANATOMY

But Dewitz (16) also worked at the development of the sting in Apis mellifica,
and he asserts that the three pairs of sting-rudiments arise from two segments, and his
figures show this very unmistakably. He says: “auf der Bauchseite der beiden
vorletzten Segmente, dem 11. und 12., sicht man die erste Anlage des Stachels als 6 kleine
Wiirzchen, von deren 4 dem vorletzten, 2 dem drittletzten Leibesringe angehoren.
Die beiden letzteren kriimmen sich bei ihrem spiiteren Wachsthum mit den Spitzen nach
den Seiten und sie sowohl, als auch die 4 tibrigen liegen unmittelbar unter der Oberhaut
in kleinen Hohlungen.” He finds the development is similar in Locusta viridissima,
and also in Vespa vulgaris, of which he remarks: “die beiden vorletzten Segmente, das
11. und 12. hinter dem Kopfe zeigen wieder die 6 Stachelwarzchen in kleinen Vertiefungen
unter der Oberhaut liegend, welche ebenfalls aus Imaginalscheiben entstanden sind.”

Janet (31), too, supports the view that the parts of the sting arise in two, and not in
three segments: his remarks apply to Iyrmica rubra and are as follows: “sur le 11°
anneau il ya deux appendices qui sont les rudiments des stylets. Sur le 12° anneau
nous voyons le rudiment de la glande a venin, deux appendices qui se souderont plus
tard en une piéce unique impaire et donneront le gorgeret et, enfin, sur les cétés du
gorgeret, deux appendices qui deviendront les valves protectrices de l’aiguillon.” His
figure shows this very clearly.

Packard (44) quotes these writers, but seems to be unaware of any discrepancy; he
reproduces the figures of the development of the sting in Bombus, as given by Dewitz,
showing that only two segments are involved, but he says, “as shown, then, by our
observations and those of Dewitz, the rudiments of the ovipositor consist of three pairs
of tubercles, arising, as Kraepelin and also Bugnion have shown, from three pairs of
imaginal discs, situated respectively on the seventh, eighth, and ninth uromeres, or at
least on the three penultimate segments of the abdomen.” So far as my observations
go, they show that there are three pairs of tubercles, but that they arise in two, and not
in three, segments, there being two pairs in the posterior segment.

Regarding the ultimate fate of the sting-buds, Kraepelin remarks that the parts
arising from the first of the three pairs that he admits to be present go to form the
oviduct and passages ; but it is very difficult to suppose that the external buds figured
by Bugnion on the corresponding segment could become transformed in such a manner,
for it would involve a complicated process of invagination.

T have been unable to find anything further relating to the development of the sting
in Hymenoptera ; Zander’s (66, 67) papers are on the morphology of the adult structures,
and do not consider the larval condition. One of the most recent papers (Anglas, 1) on
the metamorphosis of the wasp does not deal with the sting origins at all.

V. Anatomy oF Imaco, ¢. (Plate 12, figs. 44, 45.)

The male of this species differs in several respects from the female. It is smaller than.
the largest females, being about the same size as the smaller specimens; it looks some-
what longer than it is in reality on account of the fact that the wings, when folded on

AND DEVELOPMENT OF COMYS INFELIX. 251

the back, extend a short distance beyond the end of the body; this is true for the hind
wings as well as for the fore wings.

In colour, the male fly differs considerably from the female. The body is entirely
black, except several white joints of the legs. The thorax and head are dull black, and
not golden brown as in the female. The antennz are completely black, and do not end
in a club-like dilatation, for the component segments are almost equal in size. ‘The
antennee of the male fly can also be distinguished from those of the female by the fact
that they possess a conspicuous covering of fine black hairs, relatively long; the elbow,
too, is not so strongly marked as in the female antenna. When the male is looked at
from the side, it is seen that the antenne are arched upwards and forwards, the tips
curving down again.

The wings are, proportionately, much larger than in the female; they differ also in
not possessing the shaded fuscous patches so noticeable in the other sex; they are, on
the contrary, of a shining iridescent colour, having an almost metallic appearance. The
group of hairs so marked on the surface of the female fore wings, is absent in the male,
in which the entire surface of the wing is covered with a soft down of fine hairs.

The legs are similar to those of the female, the second pair being furnished with the
large tibial spur so characteristic of the species.

The abdomen is relatively smaller than that of the female, but, as regards the arrange-
ment of the segments and of the curious lateral plates, the form is identical. ‘The
reproductive apparatus differs entirely from that seen in the female. In the natural
condition, there is a small pointed process projecting from the tip of the body, on either
side of which can be seen a little wing-like organ. When dissected out, the male organs
are found to be composed of a central hollow piece, or penis (PI. 12. fig. 45, p., h.), in
which run the ducts communicating with the glands. On each side there is a jointed,
hooked rod, ending in three sharp teeth. Outside these parts, latcrally, there is another
rod-like organ, which ends in a long spike, or seta. The tip of the penis has curious
little papilize, which are the openings of the ducts leading from the gonads.

Note—-While performing the work recorded above, I have been fortunate in having
the valuable help of Dr. D. Sharp, who has, with constant generosity, given me the
benefit of his knowledge and experience; I am, therefore, only too happy to take this
opportunity of acknowledging, with gratitude, the essential service he has rendered.

Balfour Laboratory, Cambridge,
May 14th, 1903.

22.

33:

Sy aA RY PL

. Janet, C. 1898. Etudes sur les Fourmis, les Guépes et les Abeilles. 18. Aiguillon de la Myrmica

MISS A. L. EMBLETON ON THE ANATOMY

VI. BrpLioGRAPHy.

Alphabetical List of Authors consulted.

Referred to in the text by the Author’s name and the corresponding number given in this list.

. Aneras, J. 1901. On the Metamorphosis of the Wasp. Bull. Sci. France et Belgique, xxxiy.

pp. 363-473.

. Asumeap, W. H. 1886. Tr. Amer. Ent. Soc. xii.

1887. 14th Bull. U.S. Dep. Agric.

—— 1888. Ent. Amer. iv.

— 1893. Bull. Ohio Exp. Stat. 1.

—— 1896. Tr. Amer. Ent. Soc. xxiii.

— Classification of the old Family Chalcidide. P. Ent. Soc. Washington, iv. pp. 242-249.

— 1900. On the Genera of the Chalcid-flies belonging to the Subfamily Hncyrtinw. P. U.S.
Nat. Mus, xxii. pp. 323-412.

. Buenion, E. 1891. Recherches sur le développement postembryonnaire, |’anatomie et les mceurs

de V Encyrtus fuscicollis. Rec. Zool. Suisse, v. pp. 435-534, pls. xx.—xxv.

. Comsrocx, J. H. 1881. Report of the Commissioner of Agriculture for 1880. Rep. Entomolo-

gist, Dep. Agric. Washington.

. Curtis, J. 1829. Guide to British Insects.

1832. Brit. Entom. ix. p. 395.

. Datta Torre, C.G. 1885. Jahresber. nat. Ges. Graubiindens, xxviii. p. 61.

1898. Catalogus Hymenopterorum, v. [ Chalcidide, ete. }.

. Datman, J. W. 1820, 1821. Svenska Ak. Handi. xli. p. 371, ete.
. Dewitz, H. 1875. Ueber Bau und Entwickelung des Stachels und der Legescheide einige

Hymeuopteren und der griinen Heuschrecke. Zeit. wiss. Zool. xxv. pp. 174-200, pls. xu. & xii.

. Emsurron, A. L. 1902. On the Economic importance of the Parasites of Coccide, Tr. Kut. Soe.

London, ii. p. 224, ete.

. Fonscotomser, Boyer pr. 1840. Ann. Sci. Nat. sér. 2. Zool. xiii. pp. 186-192.
. Forster, A. 1841. Beitr. Monogr. Pteromal.

1856. Hymen. Stud. i. pp. 32-144.
Progr. Realsch. Aachen, p. xxxili.

. Howarp, L. O. 1882. Ist Ann. Rep. Insects New York.

1885. Descr. N. Amer. Chalcidide.
1888. Insect Life, i.

1888. P. Ent. Soc. Washington, i.
1890. Insect Life, 11.

1894. Journ. Linn. Soc. London, Zool. xxv. p. 95. .
1895. P. U.S. Nat. Mus. xvii. p. 611.

1896. P. U.S. Nat. Mus. xvii.

1896. Journ. Linn. Soc. London, Zool. xxvi. pp. 129-178.

Boar

rubra, appareil de fermeture de la glande a venin. Inst. France, Ac. Sci. 1896. Paris, 1898.
27 pp. :
Koxrscuett, E. 1887. III. Die Bildung des Chorions und seiner Auhiinge bei Nepa cinerea (eine
abweichende Entstehungsweise des Chitins). Acta Ac. Germ. li. pp. 224-252, 5 pls.
Krarretin, C. 1873. Untersuchungen tiber den Bau, Mechanismus und die Entwickelungs-
geschichte des Stachels der bienartigen Thiere. Zeit. wiss. Zool. xxiii. pp. 289-230, pls. xv. & xvi.

ae os

re = £2 per

aa

AND DEVELOPMENT OF COMYS INFELIX.

bo

53

. Ktyexet p’Hercurais. 1876-1882. Rech. sur le dév. et Vorg. des Volucelles. Paris, (Ouvr.

conronné par |’ Acad. des Sciences.)

. Lasourskne, A. 1858. Histoire d’un Ichneumon parasite des Araignées. Aun. Soc. Ent.

France, i. p. 808.

. Larreitir, P. A. 1809. Gen. Crust. Insect. iv. p. 31. [Genus Encyrtus erected.]
. LEPELETIER DE Sarnt-Farceau, A. 1825. Encycl. méthod., Insect. x.
. Leuckart, R. 1855. Ueber die Micropyle und den feinern Bau der Schalenhaut bei den

Insekteneiern. Arch. Anat. Physiol. pp. 90-264, pls. vii.-xi. [Hymenoptera, pp. 235-244]

. Marcnat, P. 1898. La dissociation de l’ceuf en un grand nombre d’individus distincts et le cycle

évolutif chez l’Eneyrtus fuscicollis, C. R. Ac. Se., Feb. 28, cxxvi. pp. 662-664. [In English,
Ann. Nat. Hist. (7) i. pp. 28-30.]

= ——— 1898. C.K. Soc. Biol. 1898) p. 238.

1899. Comparaison entre les Hyménopteéres parasites 4 développement polyembryonnaire et
ceux 4 développement monoembryonnaire. C. R. Soe. Biol., July 22, p. 711.

. Mayr, G.L. 1875. Die europiischen Encyrtiden. Verh. Ges. Wien, xxv. pp. 675-778.
43°

1876. Ditto, p. 691.

43a. Newsreap, R. Coccide of the British Isles. Ray Soc. 1903, vol. ii.

44.

45.

46.
47.

48.
49.
50.
51.
. Warxker, F. 1837. Ent. Mag. iv. p. 48.

Pacxarp, A.S. 1898. Text-book of Entomology. [Pp. 168-173. Development of the sting. ]

Raspait, X. 1900. [On the results of interference with natural conditions.] Bull. Soc. Acclimat.,
summarized in Feuill. Natural. xxxi. pp. 119 & 143.

Rarzesurc, F.T.C. 1844. Ichneum. Forstinsect. i.-iii.

Reinnarp, H. 1865. Zur Entwickelungsgeschichte des Tracheensystems der Hymenopteren.
Berlin. ent. Zeitschr. ix. p. 187, etc.

Suarp, D. 1895. Insects, i—ii. Cambridge Natural History.

Sito, J. B. 1900. Rep. Ent. Dep. Agric. New Jersey ; Agric. Coll. Exper. Station.

Tuomson, C.G. 1857. [Ageniaspis, Dahlbom.] Ofv. Ak. Forh. xiv. p. 292.

1875. [Ageniaspis fuscicollis.] Hymen. Scandin. iv. pt. i, p. 182.

1837. Ent. Mag. v.

1839. Monogr. Chalcid.

1841. Entomologist.

1843. Ann. Nat. Hist. xii. pp. 46-49, 103-104.
1844. Aun. Nat, Hist. xiv. pp. 14-17, 18-22.
1816. Ann. Nat. Hist. xvii. p. 181.

1846. List Hymen. Brit. Mus., Chalcid. i.
1847. Ann. Nat. Hist. xix. p. 229.

1847. Ann. Nat. Hist. xx. pp. 19-29.

1850. Ann. Nat. Hist. (ser. 2) v. pp. 125-133.
1851. Ann, Nat. Hist. (ser. 2) vii. pp. 210-216.
1860. Ann. Nat. Hist. (ser. 3) vi. pp. 8357-359.

Peer |

. Westwoop, J.O. 1837. Phil. Mag. (ser. 3) x. p. 441.
. Zanver, E. 1899. Zeit. wiss. Zool. xlvi. p. 289.

- — 1900. Beitrage zur Morphologie der miannlichen Geschlechtsanhange der Hymenopteren.

Zeit. wiss. Zool, Ixvii. pp. 461-488, pl. 27. (Summary in Zool. Centralbl. viii. p. 174.)

7D

254

Fig. 1

.

ON THE ANATOMY AND DEVELOPMENT OF COMYS INFELIX.

VII. EXPLANATION OF THE PLATES.
RrATE ala

. Portion of fern-frond attacked by Lecanium hemisphericum var. filicum, the Coccid being
parasitized by Comys infelix.

Tigs. 2, 3, 4,5. Early stages in the development of the egg within the egg-tubes of Comys infeliz.
Vig. 6. Fully formed egg previous to the disunion of the two parts, showing the valve-like apparatus on

~
4

Tig. 21. Ventral view of head.
22. Mouth-parts in situ. w.d.=upper lip ; md.=mandible ; p.=maxillary palpi.
23. Labial palpi. 24. Maxilla and palp (p.).
25. Mandible, drawn on scale of 4, mm. Length ="18 mm.
26. Surface of face, with two ocelli and part of an eye. 27. Surface of the eye.
28. Hind wings, 8rd legs, and 8rd thoracic spiracles (sp.’).
29. Dorsal view of thorax. c.=collar; sp.!-*=spiracles; pr.=pronotum; ¢.=tegule; sc.=

30. Second leg with tibial spur.

32. Ventral view of abdomen of specimen distended with water, showing relative positions of parts.
33. Lateral view of abdomen as seen in lite, showing sting.

34. Dorsal plates of abdomen and petiole.

35. Ventral plates of abdomen flattened, showing petiole and sting.
36. Dorsal plates and petiole, flattened.

38. End view of the tip of the abdomen, showing the arrangement of the plates round the anus.
39. Tactile plate, in situ.
40, Large functional spiracle on the 7th dorsal plate.
41. Sting-apparatus.

Figs. 42, 43. Tactile plate and hinge.

Fig. 44. Comys infelix, 3, dorsal view.
AD) 1,5 » 9 Skeletal parts of the reproductive organs. p.=penis; 4.=hooks.

. First observed larva with the bifurcated tail. Length =:75 mm.

|. Second instar in which the tails remain only as withered ends of the trachez. sp.1~?=spiracles.
. Third instar showing imaginal dises. /.1~*=leg-buds; st.1~*=sting-buds; s.g.=sexual glands.
. Prepupa in situ in the host (4.). (The insect is represented in an inverted position.)

. Prepupa removed from host but still in its sheathing membrane (m.). p/.=oval red plates.

. Pupa in membrane. Length =2:25 mm.

. Comys infelix, ? , dorsal view. Length =2°5 mm. 20. The same, ventral view.

the connecting neck. A=true egg-mass ; B=yolk-mass. |

. Egg after losing the yolk-mass; the valve-apparatus is retained as a bifurcated foot to the staik
which is plugged with a quantity of protoplasmic matter. Egg-cases like this are found in the
host’s body, and are often seen capping the tail of the larva,

. Valve-apparatus more highly magnified.
. Egg before the separation of the two masses, drawn on} mm. scale. Length ='35 mm.; length
of stalk after separation =-05 mm.

. Mouth and mandibles of first larva, with buccal glands. 12. Posterior segments of same.
. Larva just before the second instar, in which the tails are atrophying.

PLATE 12:

scutellum.

1. First legs with thoracic segment of attachment.

7. First ventral plate and petiole.

#

Trans. Linn Soc. Smr.2.Zoon. Von. IX P11

al

Zs

ee

+a
<
West,Newman imp

Embleton.

COMYS INFELIX.

er
Embleton. Trans. Linn Soc.SeR.2.Zoou. Vou. IX.P1.12.

4

West, Newman imp.

COMYS INFELIX.

|
bo
or
(ori

fat

VI. Littoral Polycheta from the Cape of Good Hope. By Arraur Wiey, D.Sc., \2}
P.RS., Colombo Museum, Ceylon. (Communicated by Dr. W. G. RipEwoop, \¥
ES.) .

(Plates 13 & 14.)

Read 3rd December, 1903.

THE Annelids here described were collected by Mr. W. F. Purcell in the years 1896
and 1900, with the co-operation of Messrs. G. H. Glasson and R. M. Lightfoot. The

collection was sent, by arrangement, from the South African Museum to the British

(Natural History) Museum, and intrusted to me for examination. Most of the specimens
were preserved in an alcoholic solution of corrosive sublimate, and, in many cases, care
had been taken to procure the extrusion of the proboscides, which is a matter of

importance in the systematic study of errant Annelids.

There is a pronounced Mediterranean and Northern element in the Annelid fauna of
the Cape, a feature which has already been noted by Dr. von Marenzeller *, and, indeed,
it would appear that the geographical distribution of marine Anmnelids is primarily

determined by thermal considerations. Many species are eurythermal, and hence

cosmopolitan or pan-oceanic; where this is not the case, we frequently meet with
instances of discontinuous distribution, the areas of distribution being»separated by

thermal barriers. The only terrestrial barriers of first importance are the Isthmus of Suez

exchange of types is clear from the fact that the Annelid fauna of the Indo-Pacific
- region may be said to be composed of an assemblage of endemic, Caribbean, and
4 Mediterranean constituents.

The following are the species dealt with in this paper :—

1. Euphrosyne capensis, Kinberg. 15. Eriphyle capensis, Kinberg.
2. Lepidonotus clava semitectus, Stimpson. 16. Marphysa sanguinea hemasoma (Montagu).
3. Polynoe scolopendrina, Savigny. 1 capensis (Schmarda).
4. Hemilepidia erythrotenia, Schmarda. 18. Purcellana, sp. 0.
5. Parmenis capensis, sp. n. | 19. Lysidice capensis, Grube.
6. Sthenelais fuliginosa capensis, Claparéde. 20. Maclovia iricolor capensis (Montagu).
7. Eulalia capensis, Schmarda. 21. Lumbriconereis coccinea, Renier,
8. Phyllodoce sp. ? 22. nardonis, Grube.
9. Glycera convoluta africana, Keferstein. 23. capensis, Grube.
10. Neanthes latipalpa, Schmarda. 24. Cirratulus atrocollaris, Grube.
ak capensis, sp. 1. 25. tentaculatus meridionalis (Montagu).
12. Mastigonereis operta (Stimpson). 26. capensis, Schmarda.,
18. Perinereis mendax (Stimpson). 27. Flabelligera luctator, Stimpson.
14. Platynereis striata (Schmarda). 28. Lipobranchus capensis, sp. u.

* Marenzeller, E. yon, “ Polychiiten der Angra Pequena-Bucht,” Zool. Jahrb. Syst. Bd. iii. pp. 1-24 (1888),
SECOND SERIES.—ZOOLOGY, VOL. IX. 35

256 DR. A. WILLEY ON LITTORAL POLYCHATA

I am greatly indebted to Prof. F. J effrey Bell for his kindness in looking over the
proofs of this paper.

1. EvpHROSYNE CAPENSIS, Kinberg. (Plate 13. figs. 1-3.)

Euphrosyne capensis, Kinberg, 1857, Ofv. Ak. Férh. Stockholm, 1858, p. 14; Grube, 1867, ‘ Novara’
Exped, Anneliden, p. 6; McIntosh, 1885, ‘ Challenger’ Polychzta, Reports, vol. xii. part 34,
p- 1; Marenzeller, 1888, Polychiiten der Angra Pequena-Bucht, p. 1.

Euphrosyne polybranchia, Schmarda, 1861, Neue wirbellose Thiere, ii. p. 136.

The identity of Schmarda’s species with Kinberg’s 2. capensis was first established by
Prof. McIntosh.

The collection contains seven examples taken ‘‘ between tide-marks among rocks,
St. James, False Bay,” and “‘among roots of sea-bamboo off Woodstock and Salt River
beaches.” The colour of the living worms is stated to have been brick-red.

The largest specimen has 64 segments, a length of 64 mm., and width of 15 mm.
Others with 54 segments measured 20-22 mm. in length, and one with 52 segments
measured 46°5 mm. in length, showing that there is no fixed correlation between the
total length and the number of segments.

The anterior pair of eyes is placed upon the ventral side of the head (fig. 1), and on
either side of them there isa minute parophthalmic tentacle which has not hitherto
been described. With strong reflected light they are easily seen under a low power,
more clearly in some specimens than in others (fig. 2).

The cephalic caruncle occupies the median dorsal area of the first seven segments
(fig. 3).

2. LEPIDONOTUS CLAVA SEMITECTUS, Stimpson. (Plate 18. fig. 4.)
Lepidonotus semitectus, Stimpson, “ New Marine Invertebrates,” Proc. Acad. Philad. vil. 1855, p. 393 ;
Marenzeller, 1888, Polychiiten der Angra Pequena-Bucht, p. 3.
The very numerous examples of this species contained in the collection present a
varying aspect, differing in colour from mottled dark brown or black to mottled scarlet
on the elytra. ‘They represent the Cape community of the Mediterranean and British

species, Lepidonotus clava (Montagu). The remaining synonymy is given by Maren- ~

zeller. The above trinomial designation of the species requires some explanation. It
seems there is no real specific distinction between the Cape and the northern forms.
The worms in this collection are topotypes of Stimpson’s L. semitectus, and the word
semitectus is merely employed here to denote this fact. Their colour-mean, average
dimensions, and periodicity probably diverge more or less from those of the northern
members of the species. If Stimpson had employed a geographical epithet for the
trivial name of his specimens, it would have better suited our purpose, because the
typical L. clava is also “ semitectus”’ in respect of the elytra; but as he did not, I retam
his term with the view of recognizing his rights and avoiding a controversy concerning
priority.

The opposed scales sometimes touch in the middle line, leaving uncovered diamond-

i te

FROM THE CAPE OF GOOD HOPE. 257

shaped patches (presenting a white nucleus surrounded by reddish-brown pigment)
along the middle of the back; successive scales also meet each other sometimes, but not
always. The scales are orbicular, margin unfringed, often reddish in colour, with pale
outer border.

The length of the ¢tentaculum impar varies, and this may be due either to normal
variation or to regeneration after injury. In one specimen the median cephalic tentacle
was thick and white, and only half the length of the paired antennie, though there was
no sign of abnormality beyond the absence of pigment. In another the tentaculum
was barely longer than the antennz and rather stouter. In a third the tentaculum
was half as long again as the antennz, as long as the palps, and, like the latter,
terminating in a flagelliform appendix (flagellum).

A moderately large specimen showed 26 segments, and measured about 22 mm. in
length, with width of 10 mm. over the setze, 8 mm. without the sete.

One tube contained forty-one specimens taken between tide-marks amongst rocks,
St. James, False Bay. In another there were six examples, taken amongst roots of
sea-bamboo off Woodstock beach, Table Bay, in 8-10 feet of water.

The elytra are tuberculate on the surface, with plain margins. The papille of the
palps are disposed in six longitudinal rows.

3. POLYNOE SCOLOPENDRINA, Savigny. (Plate 13. fig. 5 and fig. 25.)

Hemilepidia tuberculata, Schmarda, Neue wirbellose Thiere, Bd. i. 2, 1861, p. 149.

Polynoe attenuata, McIntosh, 1885, ‘Challenger’ Polycheta, Reports, vol. xii. part 34, p. 120;

cf. Marenzeller, 1888, Polychiten der Angra Pequena-Bucht, p. 5.

Schmarda’s Hemilepidia tuberculata cannot, in my experience, be specifically dis-
tinguished from Polynoe scolopendrina (Savigny) *.

Segments about 110; length 70 mm.; width without sete 6°5 mm., with sete 9 mm.
The dorsal tubercles, of which there are a median row and a lateral row on each side of
the dorsum, commence about the 20th segment. ‘The ventral (nephridial) papillz are
large, visible without the use of a lens. Dorsal cirri alternate with the elytra in
anterior region of body, becoming consecutive behind the last elytron. Cirri anales 2,
stout, subulate, with filiform tip.

The elytral formula is the same as for Hemilepidia erythrotenia, namely, 2, 4, 5, 7, 9,
11, 18, 15, 17, 19, 21, 23, 26, 29, 32, always counting the segment which carries the
tentacular cirri, 7. e. the buccal segment, as the first segment of the trunk.

Locality. Amongst roots of sea-bamboo off Woodstock beach, Table Bay, ten specimens.

The elytra of the first pair are round and larger than the succeeding oval scales.
The anterior eyes occupy the frontal peaks.

A specimen in another tube, from St. James, False Bay, presented a pale flaccid
appearance.

* Cf. Baron de Saint-Joseph, “ Les Annélides . .. de Dinard,” Ann. Sci. Nat. (7) vy. 1888, p. 183; and McIntosh,
W. C., ‘British Annelids,’ Ray Society Mon. 1900, p. 390.
35*

258 DR. A. WILLEY ON LITTORAL POLYCH ETA

4. HEMILEPIDIA ERYTHROTHNIA, Schmarda. (Plate 13. fig. 6 and fig. 26.)
Hemilepidia erythrotenia, Schmarda, 1861, Neue wirbellose Thiere, ii. p. 150; Marenzeller, 1888,
Polychaten der Angra Pequena-Bucht, p. 4.

This polymeric Polynoid has, so far as I am aware, only been recorded from the Cape
region, and constitutes one of the features of the South African Annelid fauna.
Luphrosyne capensis is another characteristic component of this fauna.

The pigmentation of the elytra consists of a broad sharply defined black area at the :
mesial borders, which just meet in the middle line. The rest of the surface of the
elytra is colourless, except for a small dark spot over the scars.

The principal character by which it differs from the type of Polynoe scolopendrina is
in the curved tip of the ventral setee, which is smooth in H. erythrotenia and bidentate
in P. scolopendrina.

Locality. Amongst roots of sea-bamboo off Woodstock beach, Table Bay.

The tentaculum and antenne, especially the latter, are beset with small squamiform
papilla. The dorsal surface of the body is devoid of tubercles.

5. PARMENIS CAPENSIS, sp. n. (Plate 13. figs. 7 & 8 and figs. 27-29.)

In the diagnosis of his genus Parmenis, Malmgren * includes the following charac-
ters :— Elytra, paria 15, totum dorsum imbricatum tegentia. Setze rami superioris
seriatim transverse spinulose, breviores et crassiores quam sete rami inferioris. He
infra apicem glabrum bifidum vel profunde bidentatum, dente superiore apice curyato.”

In the definition of the species P. Ljungmani an error has crept into the text, the
dorsal setee being described as ‘* paullum ¢fenuiores quam sete rami inferioris” instead
of * paullum crassiores.”

The Cape specimens which I refer to this group have 15 pairs of elytra and as many
as 39 segments, the last 6 segments uncovered, as happens also in Lagisca. The elytra
of the first pair are circular and very much smaller than the succeeding elytra, which
have an ovate form with long diameter placed obliquely with reference to the longi-
tudinal axis of the body. The outer and posterior borders of the elytra are fimbriated,
apparently differing in this respect from the northern type, which is described as having
elytra “ margine glabro.”’

The anal cirri resemble the dorsal cirri in length and form, and, like these, are densely
fringed + with elongate papillee.

The pigment of the elytra is sparse, with an interrupted submarginal tract of neutral
tint and a scar-patch.

The dorsal fascicle of sete is cespitose; the sete are numerous and much shorter,
though only a little thicker, than the ventral.

The ventral setze are strongly bidentate.

The posterior elytra, more especially the penultimate, are noticeable on account of
their larger size.

Locality. Amongst roots of sea-bamboo off Woodstock beach, Table Bay.

* Malmgren, A. J., ‘ Annulata Polycheta,’ 1867, p. 11.
+ The term “ciliated ” is commonly employed in a special sense to describe this condition.

FROM THE CAPE OF GOOD HOPE. 259

6. STHENELAIS FULIGINOSA CAPENSIS.

T can find no distinctive character in what I take to be the Cape representative
of the Mediterranean Sthenelais fuliginosa, Clapareéde *.

The length is 28 mm., width (including setze) 4 mm., width of ventral surface without
parapodia 1:5 mm. Segments between 70 and 80 in number.

Locality. One specimen found amongst roots of sea-bamboo (arborescent Fucus) oft
Woodstock beach, Table Bay, in 8-10 feet of water.

7. EULALIA CAPENSIS, Schmarda.

Eulalia capensis, Schmarda, 1861, Neue wirbellose Thiere, ii. p. 86; McIntosh, 1885, ‘ Challenger ’
Polycheeta, Reports, vol. xii. part 34, p. 168; Marenzeller, op. cit. 1888, p. 5.

Tentaculum impar between, and slightly in advance of, the large eyes, longer than the
frontal antenne. Cirri tentaculares 8, arranged in the following manner on the first
three segments :—I - Il = Til coop Proboscis crowned with 17-19 marginal
papillee and densely beset with papillee over the surface, except in its posterior portion.
Length 8 mm., width 3°5 mm. over all; length of papillose portion of proboscis
15 mm.

As indicated in the formula for the tentacular cirri, the third segment carries
the fourth tentacular cirrus above and a cirrus ventralis foliaceus below, on each
side.

Locality. Ten specimens from a depth of 8-10 feet off Woodstock beach, Table Bay.

Colour during life, green.

8. PHYLLODOCE sp.

In the absence of information concerning the structure of the proboscis, I refrain
from giving a definite name to three specimens of Phyllodoce dredged at a depth of
10 feet in Table Bay on a mud bottom. Like Lulalia capensis, the colour in the
fresh condition was green. There are upwards of 172 segments; length 40°5 mm.;
width without sete 1:5 mm., with sete 2 mm. ‘The specimens had all lost the
proboscis. :

The head is rotund, not longer than broad, sometimes narrower in front ; its posterior
margin, near which the eyes are placed, is entire.

The tentacular cirri are disposed as in Carobiat; they are short and stout, their

length not exceeding the width of the body.

The character of the proboscis is absolutely essential to the definition of species of
Phyllodoce and its subgenera Anaitis and Carobia.

* Of. Marenzeller, “ Zur Kenntniss der adriatischen Anneliden,” SB. Ak. Wien, i. Abth. Bd. Ixix. 1874,
p. 421.

+ Cf. Marenzeller, op. cit. (Adriat. Annel.) 1874, p. 426; and same author, 1879, “ Siidjapanische Annel.,”
~ Denkschr. Ak. Wien, xli. (2nd Abth.) p. 127.

260 DR. A. WILLEY ON LITTORAL POLYCHATA

9, GLYCERA CONVOLUTA AFRICANA.

Glycera convoluta, Keferstein, 1862, Zeitschr. wiss. Zool. xii. p. 106 ; Grube, 1869, Jahresber. Schles.
Ges. Breslau, 1870, pp. 59 & 63; Grube, 1877, Monatsber. Akad. Berlin, p. 510 (Table Bay,
50 fathoms); de Saint-Joseph, Ann. Sci. Nat. (sér. 7) xvii. 1894, p. 27.

Glycera africana, Arwidsson, 1898, Bergens Mus. Aarbog, no. xi. p. 21 (no locality).

From the description which Dr. Arwidsson gives of G. africana in his recent studies
on the Glyceridee and Goniadide, I am unable to recognize its distinctness from
Keferstein’s G. convoluta, and the author makes no mention of the fact that the latter
species was recorded by Grube from Table Bay among the Annelids obtained during the
cruise of 8.M.S. ‘Gazelle.’ Keferstein pointed out that the species of the genus Glycera
fall into two sections, according to the presence or absence of gills. The present
species belongs to the gill-bearing section, and is distinguished by its biannulate
body-segments and by the rounded truncated character of the ventral portion of the
hifid posterior lip of the parapodium. The simple unbranched branchiz, absent from
about a score of segments in the anterior region, attain their greatest dimensions in the
mid-region of the body. There are upwards of 140 segments; length 32 mm.

Locality. Two specimens found in the mud on the mud-banks in the lagoon at the
mouth of the Knysna River; one example dredged on mud-bottom in Table Bay at
a depth of 10 feet.

10. NEANTHES LATIPALPA TYPICA. (Plate 13. fig. 9 and Plate 14. figs. 1-2 a, 0.)

Nereis latipalpa, Schmarda, 1861, Neue wirbellose Thiere, ii. p. 104.

Neanthes latipalpa, Kinberg, 1865, Ofv. Ak. Férh. p. 171; Marenzeller, 1888, Polychaten der Angra

Pequena-Bucht, p. 6.

Schmarda committed an undoubted indiscretion in ‘applying the same specific name
to two different Nereids from the Cape, WV. latipalpa and Mastigonereis latipalpa, and
introduced a further element of confusion by making one and the same figure (Taf. xxxi.
fig. 244) do duty for the two species.

The principal character of the species is afforded by the paragnaths of the order VI,
which constitute a monostich of large triangular teeth about 23 in number, con-
fluent across the middle line, so that the group V cannot easily be separated, and is
therefore to be regarded as quasi-existent. In Kinberg’s original specimen, which
T have had the opportunity of examining, there were 28 teeth in the row—11 on each
side and 1 in the centre.

The first specimen in the collection of the South African Museum which I looked at had
the same number of teeth in the groups V+ VI as in Kinberg’s type. But the number
is subject to considerable variation, both in different individuals and on the two sides of
the same individual, ranging on either side from 8 to 15. The paragnaths of order I may
be represented by a single tooth or by two or three, placed, as usual in this group, one
behind the other. The teeth of group VI may be flattened and linear instead of erect
and conical.

The feet are approximately equal throughout the length of the trunk (pedes equales),
and the dorsal cirrus is equal to or rather less than the length of the dorsal ligule.

> Na ate

FROM THE CAPE OF GOOD HOPE, 261

The head is long, and the eyes are placed at the corners of a wide trapezium.
The paragnaths of the distal or maxillary division of the proboscis are much smaller
and feebler than those of the proximal or oral division.
Length 95 mm.; width in front (gradually tapering backwards) is 4 mm. without
the feet, 6 mm. inclusive measurement.
Locality. Forty-one atokous specimens taken among rocks at Green Point, Table
Bay, in November 1896 ; twelve epitokous examples taken in the mud on the mud-
banks of the Knysna lagoon,
In the structure of the parapodia and their armature this species apparently does not
differ from Grube’s Nereis brevicirris * from St. Paul, but with regard to the arrange-
_ ment of paragnaths, there is, in the latter species, a group of three teeth of the order V
placed in a triangle behind the confluent monostich of VI. In spite of this apparent
_ difference, which is not great in view of the frequency of meristic variations and of the
common occurrence of supernumerary teeth, I think Grube’s species would be more
suitably entitled Neanthes latipalpa brevicirris.
Another representative of the same specific group was collected in Ceylon by
Mr. L. A. Borradaile, who has added the specimen to the material of the British Museum.

11. NEANTHEs CAPENSIS, sp. n. (Plate 18. fig. 10 and Plate 14. figs. 9 & 10.)

A number of small Nereids taken in company with Platynereis striata off Woodstock
beach have all groups of paragraths represented in the proboscis by conical sclerites,
and therefore belong to Kinberg’s genus Neanthes.

The third pair of tentacular cirri stretch over 4-8 segments. There is a faint
moniliform pattern along the centre of the back over the dorsal vessel, which serves to
distinguish them, when the proboscis is not exserted, from Platynereis striata. The
ligules of the feet are rounded in front and become conical behind.

___ An incomplete specimen had 63 segments ; length 31 mm., width over all 3°5 mm.
q The length of the antennz seems to vary somewhat from about half the length of the
_ prostomium to more than half this length.

The paragnaths of group VI are disposed in an acervus, and the species therefore

falls into line with V. acuminata, Ehlers, and N. erucifera, Grube.

12. MasTIGONEREIS OPERTA. (Plate 13. figs. 11 & 12 and Plate 14. figs. 7-8 a, .)

Nereis operta, Stimpson, 1855, Proc. Acad. Philad. vii. p. 392.

Mastigonereis latipalpa, Schmarda, 1861, Neue wirbellose Thiere, ii.

Mastiyonereis retrodentata (Quatrefages, 1865, Hist. Nat. Annel. i. p. 557) ; ¢f. Marenzeller, 1858,

Polychaten der Angra Pequena-Bucht, p. 7.

The collection comprises atokous and epitokous forms, with transitions from one
condition to the other. The occurrence of epitoky is a fact of great bionomic interest,
but from a strictly systematic standpoint an epitokous Annelid, although it has
achieved its highest development, is of no more practical use for diagnosis than the
same worm in the atokous condition. That is to say, the substitution of reniform sete

* Grube, A. E., “ Anneliden,” Novara-Reise, Zool. Bd. ii. (Vienna, 1867), p. 19, Taf. ii, fig. 2.

262 DR. A. WILLEY ON LITTORAL POLCHATA

for normal setze and the development of the natatory membranes (ligular lobes) on the
parapodia add no useful character to the definition of a particular species. This is
a rather curious fact. In an epitokous worm the anterior region of the body retains on
the whole its normal specific character, but in the modified posterior region the specific
features are concealed below the profusion of secondary natatory appendages of the feet,
although the characteristic form of cirri and ligules is retained and can be recognized
by careful examination below the mask of epitoky.

An atokous example measured 93 mm. in length, 8 mm. in width (including the feet),
and possessed 114 setigerous segments. An epitokous female was nearly 140 mm. long.

Locality. Six atokous specimens taken amongst seaweed off Woodstock beach in
August 1896, and seven epitokous specimens from the same locality collected in
December 1900.

13. PERINEREIS MENDAX (Stimpson). (Plate 13. fig. 18 and Plate 14. figs. 3-6.)

Nereis mendax, Stimpson, Proc. Acad. Philad. vil. p. 392 (1855).

Mastigonereis podocirra, Schmarda, 1861, Neue wirbellose Thiere, ii. p. 108; Marenzeller, 1888,

Polychiten der Angra Pequena-Bucht, p. 7.

Nereis Stimpsonis, Grube, 1867, ‘ Novara ’ Exped., Anneliden, p. 18, Taf. i. fig. 8.

Kinberg’s genus Perinereis is determined by the character of the paragnaths of
order VI, which occur as one or two linear or broadly conical or arcuate chitinous
sclerites on each side of the median group V. I have examined the types of Kinberg’s
venera of Nereidze by special arrangement between the authorities of the British
Museum (Natural History) and the Royal Museum at Stockholm, and am bound to say
that I am not clear as to the distinct generic properties of Perinereis, Paranereis,
and Pseudonereis *.

Very numerous examples of this species were taken between tide-marks at St. James,
False Bay, and two specimens from Woodstock, Table Bay. They are described as
living “in holes between the barnacles &c. on the upper sides of rocks exposed at low
tide.”

14. PLATYNEREIS STRIATA (Schmarda). (Plate 13. fig. 14 and Plate 14. figs. 11 & 12.)
Platynereis striata (Schmarda), see Kinberg, 1865, ‘“ Annulata nova, Nereidum dispositio nova,”
Ofv. K, Vet.-Akad. Férh. 1865, Stockholm, 1866, p. 177. .
The length of the antenne is equal to that of the prostomium. The third pair of
tentacular cirri stretch over 10-14 segments. There are about 80 segments in all;
length about 52 mm.; width without feet 3 mm., with feet 5 mm.
‘The dorsal cirrus is about twice the length of the dorsal ligule, and this proportion
does not appreciably alter through the length of the body.
‘The paragnaths have the form of minute granulations characteristic of the genus
Platynereis, which has priority over Malmgren’s genus Leontis. The groups VII+ VIII

* The Pseudonereis anomala of Gravier (“* Contribution 4 l’étude des Annélides . . . de la Mer rouge,” Arch. Mus.
Paris, (8) xi. 1900, pl. xii. figs. 50-52) is not a Pseudonereis in Kinberg’s sense. I have seen specimens of it from
Karachi, and think it is worthy of subgeneric rank at least, unless it be regarded as a Wereis s. str.

FROM THE CAPE OF GOOD HOPE. 263

are represented on the ventral side of the oral division of the proboscis by five distichous
or tristichous acervuli; the sides of the proboscis are unarmed, so that there is a long
interval between these ventral acervuli and the group VI. Group V is unrepresented.

Numerous examples taken among roots of sea-bamboo off Woodstock beach, Table
Bay, at a depth of 8-10 feet.

15. ERIpHyLe CAPENstS, Kinberg.

Eriphyle capensis, Kinberg, 1864, “ Annulata nova,” Ofv. Ak. Férh. Stockholm, p. 561.

See Marenzeller, 1888, Polychiiten der Angra Pequena-Bucht, p. 7, ubi syn.

According to Dr. von Marenzeller, this species is distinct from E. aphroditois by the
structure of the falciform and scalprate sete. I have not succeeded in convincing
myself on this point, and incline towards the trinomial designation EZ. aphroditois
capensis.

In the collection of the South African Museum there is a specimen, 304-8 mm. lone,
with diameter of 13 mm., taken “ between tide-marks amongst rocks, St. James, False
Bay.” The colour when the worm was alive is stated to have been brick-red.

The branchie occur as simple filaments on the 8th, 9th, 10th, and 11th setigers,
12-pinnate on the 12th setiger, rising in succeeding segments to a maximum of about
14 pinnee.

16. MARPHYSA SANGUINEA H#MASOMA. (Plate 13. fig. 15.)
Marphysa sanguinea (Montagu), cf. Marenzeller, 1888, op. cit. p. 11.

I cannot detect any essential difference between Marphysa hemasoma, Quatrefages
(Hist. Nat. Annel. i. 1865, p. 334), and the European IZ. sanguinea, except a difference
of size.

One small specimen was taken between tide-marks at St. James, False Bay.

The compound sete are spinigerous; the branchiz commence on the 17th segment,
and occur as simple filaments through 7 segments.

17. Marpuysa CAPENSIs (Schmarda). (Plate 13. fig. 16.)
Marphysa capensis (Schmarda), 1861, Neue wirbellose Thiere, ii. p. 126.

Numerous specimens taken among roots of sea-bamboo off Woodstock beach, Table Bay.
In one specimen, 125 mm. in length, the branchiz commence on the 22nd segment,
the first half-dozen being simple filaments. There are about 35 posterior segments
‘without branchie. Width of body measured over the ventral cirri 8 mm., tapering
gradually behind. Compound setie falcigerous. Jaws:—lII r. 4, 1. 8, large teeth only
_ at anterior end of the long jaw-piece; III 1.4; IV 1.3, r.6; V 1—1.
In another specimen the branchiz commence simple on the 15th foot, becoming
biramous on the 24th and triramous on the 2Sth foot.

18. Marpuysa PURCELLANA, sp. n. (Plate 13. fig. 17.)

This interesting species is closely reluted to WZ. adenensis, Gravier (‘ Contribution A
Yétude des Annélides Polychétes de la Mer rouge,” Arch. Mus. Paris, (4) ii. fase. 2,

SECOND SERIES.—ZOOLOGY, VOL, IX. 36

264 DR. A. WILLEY ON LITTORAL POLYCHATA

1900, p. 270, pl. xi. figs. 91-92), and my own inclination is to adhere to the trinomial
system by the designation Marphysa adenensis Purcellana.

Tt differs from JZ. adenensis in that the prostomium is broader than long and its frontal
border is emarginate ; the median antenna is the shortest, shorter than the prostomium,

The branchiz are pinnate and the compound setz falcigerous, these being the
principal characters which relate it to Jf. adenensis.

The pinnate branchiz occur on segments 10-30. The feet are low. There are
136 segments (in the specimens examined), followed by an apparently regenerated
tail-end of about 10 segments, terminated by two slender anal cirri, at the base of which
are two quite short cirri. The total length is about 95 mm., and the width in the
branchial region 5 mm.

This species, in common with JZ. adenensis, differs from JZ. Belli in the form of the
setee and branchiz, but resembles it in a striking manner in the localization of the
branchise *.

I have much pleasure in dedicating this species to Mr. W. F. Purcell, by whom
it was collected.

19. LysrpIck cAPENSIS, Grube. (Plate 13, fig. 18 )
Lysidice capensis, Grube, 1867, ‘ Novara’ Exped., Annel. p. 12, Taf. 1. fig. 4.

Five examples of this species were taken between tide-marks at St. James, False Bay.

20. MACLOVIA IRICOLOR CAPENSIS. (Plate 13. figs. 19 & 20.)

One specimen taken among roots of sea-bamboo off Woodstock beach, Table Bay.
It consists of 215 segments, incomplete behind, 102 mm. long, 3 mm. wide. The dorsal
cirree are evanescent, but the setze which enter them are present. It can hardly be
separated specifically from Jf. tricolor (Montagu) f, differing only in size, so far as I can
ascertain from the alcoholic material. Another specimen comes from St. James,
False Bay.

21. LUMBRICONEREIS CCcCINEA, Renier. (Plate 18. fig. 21 and Plate 14. fig. 18.)
See Ehlers, Borstenwiirmer, 1868, p. 389.

A tube contained 28 Lumbriconereids from St. James, False Bay. Most of them
appeared to belong to this species, which is characterized by the breadth of the anterior
end, and especially by the subglobular prostomium. Compound falciform setee (in the,
specimen examined) occurred in the first 13 setigerous segments, simple hamate
sete thereafter. Simple limbate capillary setze occurred in the first 28 setigers, and on
one side I found them again cropping up in segments 41, 42, and 438. This shows (what
I have often observed before) that the distribution of the various forms of sete in
the Lumbriconereide is subject to considerable variation.

A specimen of 70 mm. had 100 setigerous segments.

* Compare also Eunice stragulum, Grube (Philippine Annelida, 1878, p. 163).
t Cf. Willey, A., ‘On Maclovia iricolor (Montagu),” J. Mar, Biol. Assoc. (n. s) vi. pp. 98-100.,

Oy

FROM THE CAPE OF GOOD HOPE. 265

22. LUMBRICONEREIS NARDONIS, Grube. (Plate 18, fig. 22.)
See Ehlers, Borstenwiirmer, 1868, p. 381.

It is with some hesitation that I assign a specimen found in the same tube with the
preceding to this species. It is difficult to distinguish Lumbriconereidz, as a rule,
from one another. Almost the only difference (the only one which I can recognize)
between LZ. coccinea and L. nardonis relates to the form of the prostomium, which
is subglobular in the former and subconical in the latter.

Capillary setze occur in the first 46 segments, up to 7 in a fascicle.

The two species L. coccinea and L. nardonis are associated together in the Adriatic,
and it would seem that this is also the case in Table Bay, though further observations
are required in confirmation of this statement.

23. LUMBRICONEREIS CAPENSIS, Grube.

Lumbriconereis capensis, Grube, “ Fortsetzung . . . iber Eunicea: II. Lumbriconereide,” Jahresber.
Schles. Ges. 1878 (Jhrg. 56), Breslau, 1879, p. 95.

Probably synonymous with L. cavifrons, Grube (* Novara’ Exped., Annel. 1867, p. 18) ;
it cannot be distinguished satisfactorily from L. Diibeni, Kinberg, 1864.

Six specimens from St. James, False Bay. One was much slenderer than the rest,
having 300 segments, head conical, capillary setee in about 60 segments. Another
shorter specimen had capillary sete (frequently 3 in a foot) in about 36 segments.
No compound sete.

Another specimen (125 segments, incomplete behind) had simple curved limbate setze
in the first 50 setigerous segments, thereafter the hamate limbate sete. Prostomium
rather longer than three succeeding segments, ovate.

24. CIRRATULUS ATROCOLLARIS, Grube.
Cirratulus atrocollaris, Grube, “ Annel. Gazelle,” Monatsber. Akad. Berlin, 1877, p. 536.

Body round, smooth, and short; segments over 200, crowded and short; there is a
half-collar of black pigment on the third segment below; length nearly 40 mm.;
ventral aciculz no stronger than the dorsal, slightly curved; aciculee absent from about
30 anterior segments; most of the curved ends of the acicule are broken off; the sets
are excessively brittle; branchiz in paired acervi.

Twelve specimens “in mud on mud-banks in the Knysna Lagoon,” a large salt-water
lagoon formed by the sea entering the mouth of the Knysna River.

25. CIRRATULUS TENTACULATUS MERIDIONALIS.

Cf. Marenzeller, Polychiten der Angra Pequena-Bucht, 1888, p. 16.

Ventral acicule commence at the 46th segment (52nd, Marenz.), the dorsal after
the 100th (184th, Marenz.); anterior branchial filaments numerous, forming a con-
tinuous transverse acervus; length 70 mm., width 2°5-3 mm.

Twenty-one examples between tide-marks amongst rocks at Sea Point, Table Bay.

266 DR. A. WILLEY ON LITTORAL POLYCHXTA

96. CIRRATULA CAPENSIS, Schmarda.

Cirratulus capensis, Schmarda, 1861, Neue wirbellose Thiere, ii. p.56 ; McIntosh, 1885, ‘ Challenger’
Polycheta, vol. xii. p. 383 ; Marenzeller, 1888, Polychiten der Angra Pequena-Bucht.

Segments 1st-8rd achzetous, 4th-28th with capillary sete only, 29th ventral acicule
commence, 41st dorsal aciculzee commence; on the dorsum of segments 6 and 7 on each
side an acervus of about 20 branchial filaments with slender insertions and thickened
extremities; then for about 20-25 segments branchiz occur in each segment,
afterwards becoming more and more irregular, reappearing in greater numbers and with
more dorsad insertion towards the posterior end; ventral acicule 5, sometimes 3, in a
fascicle ; a few capillary setee occur throughout the length of the body in the ventral
fascicles; the ventral aciculz are stout and strongly curved; dorsal acicule slender
and nearly straight; eye-spots on sides of head; length 114 mm., width 5-8 mm.;
colour in life, orange.

Numerous examples off Woodstock beach, Table Bay.

27. FLABELLIGERA LUCTATOR, Stimpson.
Cf. Marenzeller, 1888, op. cit. p. 15.

Large pro-eminent brown hooks (festuce), one to each segment, with one in reserve ;
setze of flabellum numerous; dorsal surface convex, smooth, about 46 segments without
the flabellum, attenuate behind; length 33 mm., maximum width nearly 5 mm,

IT cannot properly distinguish this species from the northern F. affinis, M. Sars.
It is evidently the Cape form of the species, and I think the name should read F. affinis
luctator. Of course, F. affinis capensis would be the more appropriate designation, but
it would probably introduce confusion, as the name Jwctator has been applied to the
Cape members of this race of Flabelligeridz.

Seven specimens among roots of sea-bamboo off Woodstock beach; five specimens
from St. James, False Bay.

28. LiPOBRANCHUS CAPENSIS, sp. n. (Plate 13. figs. 23 & 24 and Plate 14, fig. 14.)

A single specimen of a small black Scalibregmid with white transverse head was
collected between tide-marks amongst rocks at St. James, False Bay, by Mr. W. F.
Purcell.

I submitted this worm to Dr. J. H. Ashworth *, who pronounced it to be unlike any
of the Sealibregmidze known to him.

There are 4 fascicles of sete in all segments commencing immediately behind the
head; setee of two kinds—simple, smooth, capillary settee and furcate setz ; a fringe of
papillz surrounds the terminal anus ; branchiz absent ; more than 60 segments.

The collection also contains a Capitellid (tube No. 87) from roots of sea-bamboo off
Woodstock beach, which I was unable to identify.

* Ashworth, J. H., “ The Anatomy of Scalibregma inflatum, Rathke,” Quart. Journ. Micr. Sci. vol. xly. pp. 237-
309 ; see p. 297, on the family Scalibregmide.

Fig.

wore

FROM THE CAPE OF GOOD HOPE. 267

EXPLANATION OF THE PLATES.

PLATE 13,

. Euphrosyne capensis. Anterior end from below, showing the fori buccales. x 8.
. Same. Ventral surface of prostomium, enlarged to show the anterior pair of eyes, with the

minute parophthalmic tentacles.

. Same. Anterior end from above, showing the caruncle. x 8.
. Lepidonotus clava semitectus. Anterior end from above; the first two pairs of elytra have

been removed. x 8.

. Polynoe scolopendrina. Anterior end from above. x 8.
. Hemilepidia erythrotenia. Anterior end from above. x 8.

Parmensis capensis. Anterior end from above. x 8.

. Same. Head from the right side.

Neanthes latipalpa. Anterior end from above. x 4.

. Neanthes capensis. Anterior end from above. x 8.
. Mastigonereis operta. Anterior end from above; in the extruded pharynx the paragnaths of

group V are irregular; the tentacular cirri are rather abnormal. x 4.

. Same. Epitokous phase; eyes enlarged and contiguous. x 4.
. Perinereis mendaz. Auterior end from above. x 4.

Platynereis striata. Anterior end from above. x 8.

. Marphysa sanguinea hemasoma. Anterior end from above. x 3.
. Marphysa capensis. Anterior end from above. x 2.

Marphysa Purcellana. Anterior end from above. x 3.

. Lysidice capensis. Anterior end from above. x 2.

. Maclovia iricolor capensis. Anterior end from aboye. x 4.

. Same. Ina state of protraction, x 4.

. Lumbriconereis coccinea. Anterior end from above. xX 4.

. Lumbriconereis nardonis. Anterior end from above. xX 4.

. Lipobranchus capensis. Anterior end from above. x 15.

. Same. Anterior end from below. x 15.

. Polynoe scolopendrina. Ventral sete: A, superior; C, inferior. x 130.
. Hemilepidia erythrotenia. Corresponding sete. x 130.

. Parmenis capensis. Foot from an elytra-bearing segment. x 20.

Same. ‘Tip of a ventral seta.
Same. Portion of an elytron.

Pruavre 14,

1. Neanthes latipalpa. Wighth foot of right side. x 26.
2. Same. Seventy-third foot of right side. x 26.
s. 2a & b. Details of ventral sete.

N.B.—The few sete represented in the feet are designed to illustrate the distribution of
the homogomph and heterogomph varieties.

3. Perinereis mendax. Thirty-ninth foot of right side during commencing epitoky. x 24.
COND SERIES.—ZOOLOGY, VOL. IX. 37

268 ON LITTORAL POLYCHATA FROM THE CAPE OF GOOD HOPE,

Figs. 4, 5, & 6. Perinereis mendax. Eleventh, sixty-second, and seventy-eighth parapodia parently
of the right side of an atokous individual. x 24.
7 & 8. Mastigonereis operta. Thirty-eighth and eighty-fourth feet - -respectively of the right
x 24,
8a& b. Details of sete. ‘
9 & 10. Neanthes capensis. Twelfth and fifty-seventh feet of right side, with details of s se
x 55.
11 & 12. Platynereis striata. Twelfth and seventieth feet of right side. x 40.
lla & 12a. Details of setz.
Fig. 13. Lumbriconereis coccinea, Compound ventral sete from the sixth foot of right side.

14, Lipobranchus capensis. Furcate seta. x 350.

Willey. . Trans.Linn. Soc. Ser. 2.Zoon Vou.lX.P1. 13.

=)
3
’
shi

Z

Ji

Ae

y a

A
P.Highley del. B West,Newman lith. London

CAPE POLYCHATA.

Willey. Trans.Linn. Soc. SeR.2.Zo0nVou!X Pl. 14.

= . = Je
SS ——— rs _—
: West, Newman lith London

P.Highley del.
CAPH POLYCHATA.

VII. On the Evolution of Topographical Relations among the Docoglossa. By H. J.
Frievre, D.Sc., Fellow of the University of Wales. (Communicated by Professor
W. A. Herpman, F.2.S., F.L.S.)

(Plates 15-17.)

Read 17th December, 1903.

THE researches, the results of which are here described, were pursued in the labora-
tories of the University College of Wales, Aberystwyth, and owe a great deal to the
kind encouragement and stimulating advice of my friend, Professor Ainsworth Davis.
Prof. Yves Delage was so good as to allow me to work in the Lacaze-Duthiers Laboratory
at Roscoff during the summer of 1902, and I have to thank him for affording me
the possibility of collecting material and observing the habits of the animals studied.
I must also thank my friend Mr. H. N. Adair for very valuable help in improving and
finishing the sketches which accompany this paper.

Part I.—THE comMON ANCESTOR OF THE PROSOBRANCH GASTROPODS.

The Docoglossa form a well-marked group, having many characteristic features
possessed by all the members. Such features include the oval foot, the horseshoe-
shaped shell-muscle, the general characters of the visceral hump, the position and form
of pericardium and heart, the characters of the kidney, and the disposition of the gonad.
These features, therefore, seem to have been acquired before the members of the group
diverged amongst themselves, and it is thus of special interest to trace their history
from an origin somewhat farther back than their latest common ancestor. The
group retains several very primitive Gastropod characters, such as the close approach to
external symmetry, the symmetry of the shell-muscle, the strong labial commissure, and
the two kidneys each possessing excretory tissue and each communicating with the
pericardium. ‘This allows the conclusion that Docoglossa branched off from near the
‘base of the Gastropod stem, and it is therefore best to take as starting-point the latest
common ancestor of the Prosobranch Gastropods.

The present account of this hypothetical form is based upon inferences drawn from
a direct study of the detailed anatomy of various Docoglossa—Emarginula, Fissurella,
Haliotis, Scissurella, and Trochus. I have also utilized, as far as possible, the results
of the work of Pelseneer, Haller, Boutan, Thiele, and Woodward on various archaic
Gastropods. The form described is for convenience referred to as the Prostreptoneure.
This name is selected because the form is supposed to have already undergone that torsion
of the branchial region and visceral hump which characterizes the Prosobranchis, and
therefore to have possessed the twisted visceral loop of the nervous system.

It is, however, possible that the Gastropods had begun to diverge among themselves
‘SECOND SERIES.—ZOOLOGY, VOL. IX. 38

270 DR. H. J. FLEURE ON THE EVOLUTION OF

before that process was complete, and the Prostreptoneure is to be regarded as a
general type of what were probably a few closely related forms.

The Prostreptoneure was probably far more symmetrical externally than many of its
descendants. This view is supported by the following considerations :—

1. The Docoglossa remain symmetrical throughout development, and at an early
stage Haliotis has a symmetrical pair of shell-muscles (Bowtan). The earliest
post-torsional condition of the shell-muscle is almost certainly a paired one, and
that of Scissurella, which shows this condition, may be primitive.

2. Among the earliest Gastropod fossils we find many feebly spiral shells which
are almost or quite symmetrical. Among the oldest Gastropod fossils also is the
large order of the Bellerophontidz, which usually possess symmetrical shells.

3. The remains of symmetry, both external and internal, are far more marked among
the more primitive than among the more specialized Prosobranchs.

The torsion, however, had so profoundly disturbed internal symmetry that the retention

of complete external symmetry is improbable.

It is therefore supposed that the shell of the Prostreptoneure was nearly, but. not
quite, symmetrical and possessed a moderately-developed spiral, coiling in or near the
sagittal plane. In the anterior edge of the shell there was a sinus or slit, which was
situated in the median plane or somewhat to the right.

The Foot was fairly long and primarily useful for creeping over rocks and on seaweeds.
It probably possessed a broad, and possibly bilobed, front edge, an adaptation to the
habit of creeping upon the surface of the water. This breadth may have also
permitted swimming-movements to a moderate extent. There was a moderately
developed operculum on the postero-dorsal surface of the foot.

The approach to external symmetry indicates, perhaps, a certain ability to move
through water, probably by swimming and surface-creeping. It is also adapted to the
circumstances of a life on floating seaweeds, but the Prostreptoneure was more probably
a shore-living form, perhaps favouring among other places the rock-pools. The marked
asymmetry of Pleurotomaria, the Trochidz, and typical Monotocards seems to have been
developed as a further specialization to a creeping habit.

The Alimentary Canal was in the form of a U with some amount of extra coiling on the
distal limb of the u. This can be inferred with some certainty, asa gut of that type
occurs in Pleurotomaria, Haliotis, many species of Trochus, and various Tzenioglossa,
and it also characterizes the Cephalopoda. The general disposition of the gut will
be further discussed in dealing with the consolidation of the Docoglossan visceral hump
(see Pl. 15. fig. 6).

Within the lips the lining epithelium of the mouth was cuticularized, and there must
have been a pair of lateral projections bearing extra strong cuticle which formed the jaw-
plates. These jaw-plates were probably united by a median dorsal piece, but the arch
with the strong front edge seen, for example, in the Patellide is perhaps a specialization,
though it certainly resembles the upper jaw of Cephalopods, as Thiele has remarked. _

It is certain that the Docoglossan odontophore has been greatly specialized during
the evolution of the group, and one can consider the relations in Plewrotomaria as

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA. 271

corresponding more closely to those of the Prostreptoneure. There were certainly one
pair of buccal glands and probably a pair of dorso-lateral pouches in the buccal cavity.
Farther back occurred valves, probably dorsal and ventral, and behind them a pair of
lateral cesophageal pouches opened into the gut-cavity. In this region the appearance
of the gut in transverse section would be that of a very wide cavity with a pair of dorsal
longitudinal infoldings and a median ventral infolding forking into two towards its free
edge (Pl. 17. fig. 22a). The cesophagus reached back (PI. 15. fig. 6) as far as the bend
of the U, and was followed by the stomach, the region in which the digestive secretions
were intimately mixed with the food.

With the Stomach communicated the spiral caecum, the great digestive gland, and
probably a crystalline style-sac.

The Digestive Gland in Cephalopods, 7. e. pre-torsional forms, consists of a pair of
compact lobes; but the torsion must have considerably altered the disposition of such a
massive organ, which could not change its position without affecting the external form
of the animal. It was probably squeezed out during the torsion and development of the
visceral hump, so that it afterwards filled the interstices between the gut and the body-
wall. It is doubtful whether any trace of the original paired condition remained in the
Prostreptoneure. The ducts of the digestive gland opened into a groove in the stomach-
wall, and this groove was, as in Haliotis, continued into the cavity of the spinal cecum.
This, and the fact that the ducts of the digestive gland in Cephalopoda open into
the spiral czcum direct, suggests that one function of that outgrowth is to provide
a temporary store when necessary for the copious secretion of the digestive gland. In
that case the loss of the cecum by the Docoglossa may be correlated with the huge
increase in length of the stomach (PI. 15. fig. 8), and the consequent improbability of
wastage of the secretion due to its flowing on too far in the intestine.

A crystalline style-sac is well known among Lamellibranchs ; it also occurs in Nawétlus,
and Moore and Randles have found it in various Architzenioglossa. It may therefore
have occurred in the Prostreptoneure. This structure must not be confounded with the
spiral ceecum, as they occur together in various forms, as Moore and Randles have shown.

A rectal gland may have been present.

The Nervous System has been studied in such detail by Lacaze-Duthiers, Bouvier,
Haller, Woodward, Pelseneer, Thiele, and others that its probable primitive Gastropod
condition is in the main a matter of general knowledge and agreement. It is not
therefore necessary to describe it in detail here.

Controversy has arisen more especially with regard to the innervation of the epipodium
and the evolution of pleural centres. The first question seems to be practically settled
in favour of the view, advanced by Huxley, Pelseneer, and Haller, that the epipodium is
a pedal organ, and that the ventral ganglionic cords of Pleurotomaria, Haliotis, &c. are
to be called “ Pedal Cords.” On this point, however, Thiele maintains another opinion ;,
but the matter does not directly concern the Docoglossa, and it can therefore be passed
over without further comment.

The pleural centres were certainly imperfectly developed in the Prostreptoneure, and
the lateral portions of the cireumesophageal ring were divided into two connectives on

38*

272 DR. H. J. FLEURE ON THE EVOLUTION OF

either side, as is the case in the Docoglossa— Pleurotomaria, Haliotis, &e. From the
outer one of these two connectives on either side arose the nerves of the mantle-region
and the connectives which form the visceral loop. In Haléotis the points of origin are
very near that end of the connective which abuts upon the posterior part of the ring,
and Thiele holds this to be the primitive condition. In Pleuwrotomaria the points of
origin are farther forwards, and Woodward believes this type to be primitive, especially
as among the Trochide and Haliotide there is a tendency for the pleural centre to
mount up to the visceral nerve, a specialization along a peculiar line. Among the
Docoglossa the imperfectly developed pleural centres are not in such close connection
with the pedal cords as in Haliotis, but they also show the tendency for the pleural
centre to mount up the visceral nerve, and perhaps therefore on the whole support
Woodward’s view, as they are beyond dispute very early Gastropod forms. The pallial
nerves were most probably not concentrated into a single trunk (Woodward). WScissurella
resembles the Docoglossa, and therefore Pleurotomaria rather than Haliotis, and so
seems to support Woodward’s view. ‘The nervous system in general thus seems to
have resembled that of Pleurotomaria and the Docoglossa, and in many details,
such as the nerve-supply of the slit-region, that of Haliotis. The visceral loop was
certainly smaller than in Haliotis and nearer the median plane than in that form or
the Docoglossa.

The Circulatory System (PI. 16, fig. 16a) no doubt resembled in general features that
of the Fissurellidee, and to a less exient that of Scissurella, Haliotis, and Pleurotomaria.
The ventricle was developed around the rectum, and gave rise to an aorta at its posterior
end. This aorta sent a blood-channel (visceral artery) to the upper part of the visceral
hump, and then went forward to the head to open into a sinus surrounding the cushion
of the odontophore. Thence the blood streamed into the lower part of the visceral
hump through a sinus surrounding the radular sac, and another connection led the flow
of blood into the paired longitudinal sinuses surrounding the pedal nerve-cords. The
course of the aorta in Haliotis, where, on its way towards the head, it surrounds the
radular sac, is certainly a specialization. .

The Respiratory System included a pair of subequal ctenidia in a median or nearly
median anterior branchial cavity, whose roof contained a pair of mucous glands. In
Pleurotomaria and Haliotis we find a secondary elongation of the ctenidia and hyper-
trophy of the mucous glands. The right ctenidium and corresponding mucous gland
and osphradium were smaller than the corresponding organs of the other side.

There were certainly two kidneys, right and left of the rectum and pericardium.
The right kidney was functionally by far the more important of the two and possessed
several intervisceral lobes, showing a structure resembling that in Haliotis and Trochus.
A right reno-pericardial pore has been found in practically all the archaic Gastropods,
and therefore must have been present in the Prostreptoneure.

The condition of the left kidney is not so easy to ascertain. This organ is very small
among the Docoglossa, but possesses some amount of excretory tissue and retains the
reno-pericardial canal (Pl. 1%. fig. 24). Among the Fissurellide it is reduced almost to
vanishing-point. In Pleurotomaria, Haliotis, and the Trochidx, on the other hand, it is

= ©

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA. 273

large, and forms the papillary sac, lying practically in the mantle-roof in front of the
pericardium ; it seems to possess a pericardial canal in these forms also. Lankester and
Pelseneer have therefore homologized the left kidney of these Rhipidoglossa with the single
kidney of Monotocards, and their view has been supported by the ontogenetical work of
Erlanger, which was, however, confined to the rather specialized Paludina. The difficulties
in the way of this view have been well urged by Perrier, who found that the papillary
sac is no longer an organ for the excretion of nitrogenous waste, but, quite on the other
hand, a builder of reserve material. Another of his arguments was that the left kidney
of these forms is represented by the nephridial gland of the Monotocards. Woodward

has supported Perrier’s view in a tentative fashion, but the most recent work inclines in ~

favour of the older opinion. In particular, Miss Drummond has confirmed Erlanger’s ©

ontogenetic work, and Thiele has described a structure in Trochus cinerarius which
resembles the nephridial gland and is here an outgrowth of the left kidney, as that of
Monotocards is an outgrowth of their single kidney. The difficulty remains that it seems
almost impossible to derive the Monotocard kidney from an organ specialized in another
direction, such as is the case with the left kidney of Zrochus and even of Pleurotomaria.
We should therefore, apparently, have to derive the Monotocards from forms more
primitive than Tvrochus or Pleurotomaria, in which the left kidney had not yet
acquired the special epithelium and the highly peculiar circulatory arrangements of a
papillary sac.

However this may be, it would seem to correspond best with the facts were we to
endow tle Prostreptoneure with a left kidney, more compact and much smaller than
the right, but still retaining ordinary excretory epithelium. The Prostreptoneure
possessed a single gonad whose duct led either into the pericardium or, more probably,
into the reno-pericardial canal of the large right kidney, as is the case, at any rate, in
Haliotis and Trochus.

At the end of this account of the hypothetical ancestor of the Prosobranchs it
seems important to urge that no attempt is made to derive that group from the

Docoglossa, which are certainly highly specialized on lines of their own. It is only

supposed that the Docoglossa diverged from a very primitive Prosobranch stock, and
thus, though the divergence may be considerable, the group is an important one in
connection with the problems of Gastropod phylogeny.

The difference between the view here developed and that brought forward by Thiele
touches the extent of the external asymmetry of the ancestral form. Thiele, influenced
by the paired ctenidia and the many other admittedly primitive characters of Pleuro-

tomaria aud Haliotts, would imagine an ancestor somewhere intermediate between the

two, not so high as the former, but not so distinctly flattened as the latter. It possessed

a paired arrangement of the shell-muscle, but the spiral was already very distinctly

asymmetric.

Such a view appears to me to connect too closely the development of the torsion and
that of the conical spiral hump. In a previous essay (22) I have endeavoured to clear
up one or two difficulties in connection with the former process, and to show more
especially why the torsion was essentially a forward and upward movement of the

274 DR. H. J. FLEURE ON THE EVOLUTION OF

branchial cavity along the right side. The main argument there employed is that
the pretorsional right ctenidium and left kidney and gonad, the gonad probably earliest
of all, acquired predominance in their respective functions. This gonad is the important
one already in the Cephalopods, and its predominance would probably ensure a pre-
dominance of the kidney which it used as an excurrent channel. The gill of that side
would therefore be more liable to get soiled than its fellow, which latter would acquire
greater importance. In such a condition, movement (that is, selection of variations of
position) of the branchial cavity up the right side would enhance the advantage of this
ctenidium, and therefore increase the efficiency of respiration. This had previously been
diminished through the development of the visceral hump above the posteriorly placed
branchial cavity.

It seems to me that the paired shell-muscle is a post-torsional, or almost post-torsional,
development which may have helped to set the branchial cavity in a completely
forward position. Such a shell-muscle corresponds to a fairly symmetrical shell, and
the embryological and paleontological facts, though too scanty to be of the first
importance, seem to support the view here advanced, that the Prostreptoneure had not
advanced far in the direction of external asymmetry. There is also nothing in connection
with the torsion-process as just sketched out which in any adequate fashion accounts for
or involves the development of the typical asymmetric spiral. This last development
can be understood on other grounds, which will now be discussed, though it may
first be said that the difference between Thiele’s opinion of the Prostreptoneure and that
advanced here is by no means fundamental, as a Haliotis not yet flattened would be very
much like Pl. 15. fig. 3, in which, it is freely admitted, the amount of asymmetry may be
somewhat under-estimated.

As the Prostreptoneure seems to have possessed an operculum, it must have been
able to protect itself from unfavourable circumstances by retracting the delicate parts
more or less into the shell. It is probable that variations of reduction of the left shell-
muscle accumulated very early among the Gastropods, in consequence of the increased
freedom this would give to the incoming stream bathing the more important ctenidium.,
Disappearance of this muscle and inward motion of the area of insertion of its fellow
would facilitate complete retraction, there being now only one fixed point some distance
in, instead of two comparatively near the edge of the shell. With the loss of the
left shell-muscle, the left side of the shell and visceral mass lost their support, and we
may suppose them to have sagged, especially as a median spiral meant a centre of
gravity high above the foot and consequent unstable equilibrium, as the shell had by
this time grown considerably in length to increase the efficiency of retraction. Finally,
the shell has wound around its point of support, and so lodged a maximum length
of cavity in a minimum space and retained the centre of gravity as low as possible,
Such a course of evolution seems to correspond to the facts so far as they are known,
and obviates the necessity of connecting the extreme external asymmetry of Plewro-
tomaria and Trochus with the torsion, and of ascribing it to the Prostreptoneure. The
conical spiral is therefore correlated with the creeping habit, and contrasts with the
symmetry of the Docoglossa adapted to an adhesive habit.

el rl

—

— —

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA. 275

Part II.—Tuxr Dococtossa. Foot anp EnGE oF SHELL.

The present attempt to sketch the evolution of topographical relations among the
Docoglossa is based upon a detailed study of Acmea virginea, A. testudinalis, A. corti-
cata, and A. fragilis, Patina pellucida, a species of Nacella, and Patella vulgata,
P. cerulea, and P. ornata. I have also had the advantage of Haller’s notes on Acmea
galathea (he uses the name Scutellina galathea), Lottia viridula, Scurria, Nacella,
Ancistromesus, and Patella magellanica. The chief types of structure within the
group are therefore taken into consideration, so there seems reason to hope that the
course of evolution here sketched out offers, at any rate, a basis for constructive
criticism.

The Docoglossa exemplify the adaptation of the Prostreptoneure to a life more
especially in exposed parts of the tidal zone, coupled with the adoption of the method of
adhesion for protection. In the specialization to the adhesive habit is involved the loss
of the imperfectly developed method of protection by retraction, 7. e. the reduction of the
operculum, and the cessation of evolution on the lines above sketched out for Plewro-
tomaria and the Trochide. For efficient adhesion there should be an adhesive surface
broad in proportion to the height of the adherent body, a shell capable of being pulled
down and held down symmetrically, and a shape which would give the least possible
purchase to the waves.

The Foot, already broad in front, was shortened by loss of the tail-process adapted for
a creeping habit, and was further broadened into an oval—a change which hastened the
disappearance of the habit of retraction. The animal probably possessed symmetrically
disposed shell-muscles, but if that symmetry had previously been disturbed, it was now
re-established.

The Edge of the Shell now grew downwards and outwards, ever enlarging the rim so
as to form a complete cap for the animal. Stages of such a development have been
seen and figured by Boutan (5). It then became necessary to be able to pull down the
shell into contact with the rock or other surface by a force operating as symmetrically
as possible and as directly as possible at every point. Backward marginal extension of
the paired shell-muscles secured this to a large extent, and had the advantage that it
gave the greatest possible moment to the downward force. Ultimately, the extensions
met behind and gave the horseshoe form to the shell-muscle, forming at the same time
a wall around the soft viscera within (Pl. 15. figs. 3-7).

The fibres of the shell-muscle went down into the foot, the outer ones almost vertically,
so as to hold the shell-edge down, the inner ones obliquely, so that they exerted a
downward and inward pull on the shell-edge. The inner oblique fibres, which very
greatly strengthen adhesion, seem to characterize more especially the Cyclobranchs,

The horseshoe (marginal) shell-muscle of the Docoglossa contrasts with the central
muscle of the Haliotidz, which pulls the roof-like shell down into close contact with
the contracted animal, without, however, bringing the shell-edge down against the rock.
A cap-shaped shell, which it was possible to draw down against the rock at every point,
must have imprisoned the epipodium within a narrow space and led to selection of

276 DR. H. J. FLEURE ON THE EVOLUTION OF

variations of reduction of the latter. The mantle-edge, always near the shell-edge,
would also, under the circumstances, perform most efficiently that function of giving the
animal an impression of its immediate surroundings which had been the raison détre
of the epipodium. We therefore understand the reduction of the epipodium and the
development of sensory structures on the mantle-skirt. The edge of the latter was
already occupied by shell-secreting glands, so the sensory zone was differentiated ventral
to this.

No trace of the epipodium exists among Monobranchs, but a lateral glandular streak
occurs along the sides of the foot in Patina, among the Nacellide, and in very young
specimens of Patella. This structure has been homologized with the epipodium by
Pelseneer, but the homology is denied by Haller because of the non-development of the
streak among the Monobranchs, which he considers to be the ancestral group. Thiele,
too, denies the homology, though he looks upon the Cyclobranchs as the more primitive
group; his opinion rests on the view, which he alone adopts, that the epipodium is
no part of the foot, but rather the equivalent of the ‘“‘ Notaum”’ of Chitonidze, the organ
which surrounds and secretes the shell-plates, and is by other specialists homologized
with the mantle of Gastropods. In any case, there is no doubt that the epipodium
has been very much reduced among the Docoglossa, and this is all that concerns us at
this stage.

As the adaptation of the shell-edge to the form of the underlying surface became
important, and the available space was limited by competition of barnacles, alge, &c.,
the possession of a “home” became a valuable consideration. Since the animals
still crept about to some extent, residence on one spot necessitated the evolution of the
* Homing-Power”’ which is such a well-known and remarkable feature of Patella.

be)

The sensory arrangements which make homing possible have been the subject of
discussion, and Lloyd Morgan has credited the cephalic tentacles with this function.
Ainsworth Davis has, however, found that Patella homes after excision of the cephalic
tentacles, and he ascribes the homing faculty to the pallial tentacles. In his favour is
to be said that the mantle-tentacles become differentiated among the Docoglossa (Pl. 17%.
fig. 21 a—c), they are undeveloped in Acmea virginea, small and fairly equal in A. testu-
dinalis, and very numerous in Patella and most Cyclobranchs, where they are divided
into large and small tentacles. The concomitant development suggests correlation,
which is also otherwise probable ; too much stress must, however, not be laid ‘on this, as
the pallial gills have also developed concomitantly with them, and the mantle-tentacles
may be correlated with protection of these latter from unfavourable influences.

Cyclobranchs often possess a depressed scar, in regard to which I agree with those
observers who think it due to wear and tear, shuffling of the foot, &c. The edge is the
deepest part of the scar, no doubt as a result of friction with the shell-edge. This
depressed edge of the scar possesses a special biological value, for it places the shell-
edge, the animal’s vulnerable point, in a groove from which it is all the more difficult to
dislodge it.

As Haller and others have stated, the eyes have degenerated, because the head always
remains under the shadow of the protecting shell.

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA. 277

Part IJ].—Tue Viscerat Hump or tax Docoetossa. INTRODUCTORY.

With the shortening of the foot the mass of the visceral hump would come to lie quite
behind it—that is, with its weight operating against the force of adhesion. In such a
condition of the spiral, too, the waves would have great power. Variations of growth
would therefore be selected, in which the extension of the shell-edge was especially
marked in the posterior region. In this way the spiral or its remains would acquire
a position over the posterior part of the shell-mouth, and a continuance of the process
gives the condition in very young Acmee (Pl. 15. fig. 2). Later, even this vestige of
the spiral disappears, and the well-known Docoglossan cone-like shell remains; the
disappearance has undoubtedly increased efficiency of adhesion. The cone-shaped shell
is highly advantageous under the conditions of life, for a large part of the pressure of
the waves glances off without appreciable effect, and another part by striking down-
wards only strengthens the resistance to removal.

As the shell-rim grew backwards, a space remained above the broadening foot which
became occupied by the posterior part of the contents of the spiral. The contents of the
apical region of the spiral, now fairly far forwards, settled down in the upper part
of the cone (figs. 3-5)—that is, above the contents of the base of the spiral. This may be
called the first stage of the process of consolidation of the viscera. Its occurrence is
inferred partly from the topographical relations of organs to be considered later, and
partly from a consideration of Boutan’s figures of embryos of Aemea.

It is of interest to note that enlargement of the shell-rim, so that the spiral no longer
extends beyond it, is also found among the Bellerophontidee (fig. 1).

Part I1V.—BRANCHIAL CAVITY AND HEART.

The slit in the anterior edge of the shell weakened the force of attachment, and
natural selection operated to bring about its disappearance, though this reduced the
efficiency of the branchial cavity. The settling of the viscera as just described also
reduced its efficiency by pressing on the cavity from behind. The apical viscera
accommodated themselves on the right side (fig. 7), as the left ctenidium was by far the
more important of the two, and free space for it was therefore a necessity. This
rendered the right ctenidium still more inefficient, and it disappeared, the anus and
excretory openings meanwhile shifting to the place it left vacant (Pl. 17. figs. 19 & 20).

The respiratory arrangements have been discussed by Dall, and are the bases of the
modern classification of the Docoglossa. These are, therefore, relatively well known,
and need not be described at length on this occasion.

Pressure of the viscera reduced the branchial cavity, and led to selection of variations
whereby the position of the remaining ctenidium became oblique, and a greater length
could be sheltered in the cavity, while a larger surface was thus turned to the incoming
stream. Considerations of space meanwhile led to the reduction of the mucous glands,
and these two changes greatly increased the chances of damage to the ctenidium due to
excrement. ‘These disadvantages, together with the pressure on the branchial cavity, its
imperfect currents, and the frequent prolonged exposure to air, brought about the

SECOND SERIES.— ZOOLOGY, VOL. Ix. 39

978 DR. H. J. FLEURE ON THE EVOLUTION OF

ultimate disappearance of the left ctenidium. In Aecmea fragilis, Miss Willeox found
a muscular strengthening of the excurrent channel of the large right kidney and
described it as a penis, a description which raises special difficulties and is highly
improbable. In the light of the above considerations, it is, however, easy to explain the
development of muscularity as an attempt to force excreta beyond the limits of the gill.

Haller has shown that, even before the loss of the left ctenidium, the development of
respiratory outgrowths occurs along the ventral surface of the mantle-skirt. With the
loss of ctenidial respiration aniong the Cyclobranchs, this process is carried much further
and a circle of gills appears on the mantle-skirt which—in Ancistromesus, for example—
are even branched.

The roof of the branchiai cavity in Cyclobranchs is thin and honeycombed with blood-
spaces, so that it probably serves for the breathing of damp air among those forms which
live high up the shore.

With the loss of the right ctenidium, the right auricle of the heart was reduced,
and variations in the position of the heart were selected which did away with the
now purposeless curves in the blood-stream. This involved the shifting of the heart
towards the left, so that the auricle lay directly behind the remaining ctenidium.
The ventricle would then turn to the left, so as to lie behind the auricle, and its
previously antero-posterior axis would run obliquely backwards from right to left
(Pl. 16. fig. 16 a—c). The communication between ventricle and aorta shifted at the same
time to a position directly behind the auriculo-ventricular aperture. The pressure of the
consolidating viscera meanwhile reduced the size of the pericardium, and the presence of
the rectum became rather an obstruction than a support, so that we find complete
separation of the two, the rectum running parallel to the ventricle, but just outside the
pericardium.

A feature of the Docoglossan heart is that the ventricle is united to the dorsal wall of
the pericardium along its morphological longitudinal axis, 7. e. along a line going
obliquely backwards from right to left (figs. 17 & 18, Meso.). It is accepted that the
pericardium arises as a pair of sacs, one on each side of the rectum, and that the
ventricle of the heart develops in the musculature of the partition-wall between them,
either above, around, or below the rectum. ‘The remains of the partition are “ meso-
cardia,’ but they usually disappear in Gastropods. The connection between ventricle
and pericardial wall in Patella may be such a imesocardium, possibly a secondarily
persistent embryonic character: figs. 18 a and 0 show this condition as seen in section of
Acmea and of Patella. A mesocardium occurs in Chiton, and it is quite possible that
it was also present in the Prostreptoneure.

The Docoglossa possess a special intrapericardial “‘ Bulbus Aortze ” separated from the
ventricle by a valvular aperture. It is a development of the base of the aorta, and
is relatively strongly muscular, especially on the side against the ventricle. Its special-
ization is probably due to the fact that the two arterial streams diverge in opposite
directions, both of which are at 90° to the direction of the blood-stream through the
ventricle. This curve could not be straightened out as the others had been, and, as an
alternative, extra muscular tissue has been developed at the basis of the aorta which is
included within the pericardial cavity.

a

ee EE eee

>
a

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA. 279

The change from ctenidial respiration to respiration by the pallial gills and the roof of
the branchial cavity has led to dwindling of the efferent ctenidial vein and strengthening
of the pallial vein which joins it (Pl. 16. figs. 16 & 17). Small vessels from the roof
of the branchial cavity have become connected with the auricle direct (fig. 17), instead
of with the efferent ctenidial or the pallial vein.

Part V.—CoNSOLIDATION OF THE VISCERAL Mass. First And Seconp STAGEs.

With regard to the consolidation of the viscera, it has been argued in preceding
paragraphs :—

1. That the contents of the ancestral spiral visceral hump settled in the space left free
between the broadening foot and the conical shell, and that the contents of the
upper part of the spiral were laid down with apex forwards over those of the
basal part (PI. 15. fig. 4).

2. That the’ apical viscera settled to the right, as the left side was occupied by the
ctenidium, which was still important.

3. That the pericardium moved to the left side after the disappearance of the right
ctenidium, and pulled the rectum with it for some distance.

4, That the anus and excretory openings moved to the right to the space left free by
the disappearance of the right ctenidium.

In this way the rectum became connected with the upper portion of the viscera and

connected them with the pericardial region.

The steps in the process of consolidation of the visceral mass are inferred largely from
a study of the topography of the included organs. For the sake of brevity and clearness,
however, the stages of consolidation will first be described, and then will follow an
account of their effects on the mutual relations of the different organs.

The reduction of height was an important consideration for the Docoglossa, and this
led to the settling of the high visceral mass with its “ upper” and “ lower” parts side
by side. The apical viscera Jay to the right, and the remainder of the upper region was
affected by the shift of the pericardium to the left, so that it followed and came to lie
along a line going obliquely backward from right to left (Pl. 15. fig. 7). The lower
portion therefore settled on the right side in the posterior region, and led forwards to
the left beneath the pericardium and so on to the median anterior head. The lower
viscera thus crossed beneath the upper just about beneath the last remains of the spiral,
and it is possible that the coincidence of position is evidence of correlation between these
changes. This stage is represented in fig. 7, and may be called the second stage of the
consolidation of the viscera.

The general concentration made the apical viscera squeeze under the branchial cavity,
and this pushed the lower viscera in this plane towards the left (curve Q). «A special
pressure was also exerted from behind (Z, fig. 7) as the antero-posterior axis shortened,
and this pressure helped to reduce the branchial and pericardial cavities. It is interesting
to note that Haller found a more spacious pericardium in Acme galathea, which seems
to be very primitive, than in other forms. Another effect was that the pericardium was
pushed forwards, and we notice that the posterior boundary of the pericardium is more

Z 39*

280 DR. H. J. FLEURE ON THE EVOLUTION OF

oblique in the, on the whole, more primitive 4emea than in the more specialized Patella
(Pl. 17. figs. 19 & 20). These pressures from behind, and from the right as the apical
viscera settled beneath the branchial cavity, may be referred to as the third stage of the
process of consolidaticn.

Owing to the special circumstances of dearth of free space for the lengthening gut,
the topographical relations of the intestinal coils are of value in tracing out the history
of the consolidation.

The gut of the Prostreptoneure has already been mentioned, and its probable disposition
was as follows :—

1. It was essentially a U-tube with extra coiling on the intestinal limb of the U. The

rectum ran through the pericardium (figs. 6 & 16).
2. The parts of the gut were :

(A) The Fore-gut making the proximal limb of the u.

(B) The Stomach and the beginning of the intestine. This ran forwards, perhaps
above A,

(C) The middle regions of the intestine which probably lay near B proximally and
near D distally.

(D) The terminal part of the intestine which came forwards through the peri-
cardium.

One of the first changes among the Docoglossa was the lengthening of the wide
stomach-region, 7. e. the first part of B, and it is permissible to suppose that this extension
would be dorsalwards where space was freer. It is permissible to correlate this
lengthening with the special difficulty of digesting the food, which consists mainly
of small tough Algze &c. growing on exposed rocks, or, in the case of many species of
Acmea, of calcareous coralline Alge.

After the first stage of consolidation of the viscera, it is therefore probable that B lay
more or less in the upper part of the visceral mass, A, C,and D as before. The loops
B-C and C-D were certainly lengthening and the lengthening parts of B and C and of
Cand D would probably keep together.

In the second stage of consolidation of the visceral mass, region A would go with the
lower part of the viscera, and would thereafter lie along a line going back from the
median anterior position towards the right. 4, going with the upper viscera, would
make a dorsal loop from its junction with A round the back and left side and behind
the pericardium towards the right anterior corner. C, as before, would lie at first near
B and further on near D. D would be pulled out towards the left as the pericardium
shifted in that direction; it would therefore run back from the anus, behind the peri-
cardium to the left side, and would meet C at the back (see Pl. 15, fig. 8).

A gut practically of this type has been figured by Haller for Aeme@a galathea (fig. 9)
and for Lottia viridula, and the preceding paragraphs on the disposition of the gut aim
at tracing its previous evolution. They are admittedly hypothetical, but are inserted for
the sake of completeness. Their correctness hardly affects the value of subsequent
deductions, as it would have been easy to take Acmea galathea as our starting-point,
and derive the other types of gut from its actual arrangements.

————eE————— Ls, CC CU eh t—~—“—~; CC

oe

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA. 281

It would be possible to describe the second stage of the process of consolidation
by saying that the posterior portion of the high visceral hump laid itself down on its left
side (Pl. 15. fig. 7). In such a process the fore-gut, whose front end was fixed, must have
undergone a counter-clockwise torsion approaching 90°. The torsion of the visceral
hump through 180°, to which all Gastropods have been subjected, had already had its
effect on the fore-gut. The total twist on this organ in the Docoglossa therefore should
approach 270°. As Amaudrut, Willcox, Robert, Woodward, and others have shown,
the cross-section of the gullet allows us to infer the amount of torsion to which it
has been subjected. This is possible because of the presence of a primitively dorsal
pair of longitudinal folds and a ventral subdivided fold opposite them. In the
Docoglossa these folds stretch very far back, and a series of sections allows us to trace
them round (PI. 17. figs. 22 & 23) till the dorsal pair, having passed the mid-ventral
line, mount far up the right wall. The maximum torsion traceable in Acmea (fig. 23)
seems to be about 250°, but it is as much as 300° in Patella. This correspondence with
theory is good evidence for the second stage of the consolidation-process above mentioned,
and further evidence is obtainable from the nervous system.

In the Prostreptoneure, as in primitive Gastropods generally, the visceral loop of the
nervous system is in intimate relation with the fore-gut. As the latter moved to the
right, it must have pressed the visceral loop before it, and, since we find that loop on
the right side in all Docoglossa, this may be quoted as evidence of the second stage of
the consolidation.

The extra 30° of torsion in Patella and the separation of gut and visceral loop must
be dealt with later.

Part VI.—ConsoLIDATION OF THE ViIscERAL Mass. Laver Sraces.

In Acmea virginea, testudinalis, corticata, and fragilis the gut is arranged practically
as in Haller’s types, but with two differences of detail (Pl. 16. fig. 11). The limb A is
placed as if it had been pushed back from the right anterior corner, and the arrange-
ment of the junction of Band C points to the same inference. This is what is here
and above described as the third stage of the process of consolidation, a squeezing-in of
the viscera of the apical region in the right anterior corner—it is continued farther
in pursuance of the process of general concentration. At the same time, the viscera of this
region exerted a pressure on the fore-gut in the ventral region to the left, and thus gave
it a curve to the left (Q, fig. 11). The junction of regions B and C now formed a
loop Z, which characterizes most of the remaining Docoglossa and which may be double
(fig. 11). Changes in the head-region carried this process of inpushing from the right
anterior corner still further, and may be referred to as the fourth stage of consolidation.
The food of the Docoglossa must be raked up from hard rocky or tough algal surfaces,
and in correlation with this the odontophore has grown relatively to the other parts and

differentiated in various ways so as to secure firmness and exact symmetrical adjustment.

A further fact is that the buccal glands increased markedly in mass, many forms
possessing two pairs. The acinous mass of the gland also shifted back from the wall
of the buccal cavity, thus increasing the free space above the radula, and accumulated

282 DR. H. J. FLEURE ON THE EVOLUTION OF

on the anterior surface of the visceral hump. These two causes further increased the
backward pressure (1).

As region dA of the gut moved backwards and inwards, the visceral loop of the
nervous system, moored probably by its osphradial and rectal nerves, remained on the
right and was in this way separated from the gut. Its very reduced size among so-
many of the Docoglossa is probably a consequence of this separation. The gut of the
Cyclobranch Ancistromesus, according to Haller’s figure (Pl. 15. fig. 10), shows the effect
of the first two stages of consolidation and of the others to some extent, but particularly of
a fifth stage. A fairly proximal part of region C, apparently situated previously on the
right side, has been pushed in over the dorsal surface of the visceral hump, or perhaps
rather has lengthened in this particular direction because of the direction of the pressure
on it. The pressure in question was due to the concentration of the shell-muscle on either
side, butt took effect mainly on the right (Y), as the pericardium on the left was hardly
susceptible of further compression. The part of region C which has spread on the dorsal
surface of the visceral hump is referred to in later paragraphs as loop I/; it characterizes the
group of the Cyclobranchs. The position of the loop C—D dorsal to B in Ancistromesus,
instead of at its left side in other types, is not a difference of theoretical importance.

In Patella vulgata (Pl. 16. figs. 12 & 13) are seen the effects of all the processes thus
far enumreated.

Regions dA and B and loop Z are fundamentally as in Acmea testudinalis, save
that 4 is pushed still farther back from the right anterior region, to which it is now
distinctly concave, forming the curve Q,. This is to be understood as a further effect of
pressure in processes 8-5. It is probable that this pressure has also increased the
torsion of the fore-gut, and so accounts for the extra 30° of torsion beyond the 270°
whose origin has already been traced.

Processes 3 and 4 seem to have had the additional effect of pushing back a loop WV on the
ventral surface of the visceral mass, as usual from the right anterior corner. Process 5
accounts for the loop J on the dorsal surface of the mass, as in Ancistromesus.

Patella cerulea (Pl. 16. figs. 14 & 15) is particularly interesting in that of itself it
makes us infer process 5. The junction of 4 and & and that of Cand D are pushed in
from the right over the dorsal surface, and this process is also evidenced in the same
way by P. radians and P. ornata. In the two latter forms loop WV is not apparently
developed. In Patina pellucida the gut resembles that of Patella cerulea, save that
the junctions mentioned are even more strongly pushed in over the dorsal surface, and
the loop J/ is for this reason pressed to the left ; loop WV is, however, not developed.

In the Nacellidz we find very marked elongation of the gut, but it is arranged as in
Patina, save that loop JZ has its two hind limbs close to one another, and they run round
the left side to the back in close contact with B. The inpushing from the right is very
marked, and may account for this change, which, however, may be due to a close
connection between J/and 6 or JZ and the loops C-D at an early stage. At all events,
there is no fundamental difference between Nacella and Patina in this respect.

In Scurria the Docoglossan gut attains its greatest length, but even here it only
shows local lengthening of what is fundamentally a Nacellid gut.

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA. 283

We therefore see that the Docoglossan gut is arranged on what is fundamentally
a single type complicated by progressive lengthening in a confined space.

From the disposition of the parts of the gut and of the various organs of the hump it
has thus been possible to infer the course of the process of consolidation of the visceral
hump. This has been described as consisting of five stages, which, though necessarily
separately discussed, must be understood to have acted in part simultaneously.

Part VII.—Stummary oF DocoGtossaN EvoLurion.

Other processes enumerated contributed to the development of the Docoglossan

characters, so that the evolution of the group may be summarized as follows :—

I. The foot shortened and broadened into its characteristic oval shape, and the tail-
process disappeared. This caused a forward pressure on the viscera, which is
discussed in dealing with the third stage of the consolidation-process (IV. 3, below).

II. The shell-edge grew downwards and outwards, giving the cone-like form. Growth
was particularly marked along the posterior edge, and in this way the apex came
to lie far forwards. Correlated with this are the reduction of the epipodium, the
sensory differentiation of the mantle-edge, the degeneration of the eyes, and the
evolution of the horseshoe shell-muscle.

III. The changes in II. reduced the efficiency of the branchial cavity, which was
already impaired by the loss of the slit. As a result, we find alteration of the
position of the Monobranch ctenidium, and complete disappearance of the
ctenidia in Cyclobranchs. In both groups the mantle-skirt becomes an important
respiratory organ, and in the latter group a circle of pallial gills is developed.
Concomitant changes occurred in the heart, which moved to the left side.

IV. Consolidation of the viscera. This has been described in five stages :—

1. The contents of the spiral settled down in the space above the broadened foot.
The viscera of the base of the spire went from front to back ventrally, and
those of the apical region from the back to a right anterior position dorsally.

2. The viscera further consclidated so as to reduce height. The upper parts of
the mass, together with the rectum, drew to the left us the pericardium (III.)
moved thither, and thus came to lie along a line going obliquely backwards
from the right anterior corner to the left side. The lower viscera, on the
other hand, arranged themselves along a line going forwards from the right
side to the head. They crossed under the upper viscera about beneath the
apex of the cone.

3. The shortening of the antero-posterior axis caused pressure forwards on the
viscera and pericardium. The apical viscera were at the same time squeezed
in under the branchial cavity, and exercised a pressure backwards from the
right anterior corner.

4, The growth of the buccal mass and the increase of the salivary glands increased
the backward pressure just mentioned.

5. The concentration of the shell-muscle led to inpushing of the viscera, and the
effects are visible particularly on the right side.

284 DR. H. J. FLEURE ON THE EVOLUTION OF

Part VIII. APPENDIX TO EVOLUTIONARY SKETCH.

It now remains to deal with various details of the group's evolution that would have
appeared to be digressions had they been treated in the discussion of the main scheme.

The Buccal Mass has been specialized in important points, whose bionomical correlatives
must be briefly mentioned, though a detailed discussion would not be profitable without
comparison throughout with other primitive Gastropods. The specialization is on different
Jines in Aemea and Patella, thus supporting the view that Cyclobranchs and Mono-
branchs have diverged from a common ancestor rather than evolved one from the
other.

The odontophore of the Prostreptoneure showed in transverse section a V-form, which
was modified among the Docoglossa as an adaptation to efficiency of raking. The
cartilage-pieces drew close together and altered so as to fill the groove, the remains of
which are more conspicuous in dcmea than in Patella. The teeth of the radula have
become strong and specialized hooks or claws, and the differentiated median tooth has
been reduced or has become similar to the others, as a specialized median piece would
scatter rather than rake in.

In species of Aemea, which in several cases feed on calcareous Alge, the cartilages of
either side have fused, giving a much enhanced firmness.

Among the Patellidz, on the other hand, we find general growth, increase of the
number of muscles, and differentiation of extra cartilages to which muscles are attached,
and which thus increase the possibilities of adjustment.

Reference has already been made to the jaws, the buccal glands, and the lost spiral
excum of the stomach. It is noticeable that the cesophageal pouches of the Docoglossa
do not show the marked glandular development which characterizes these organs in
Haliotis, &c., and, perhaps in correlation with this, the valves are reduced which in the
latter prevent food from returning to the buccal cavity when it is in these pouches.
The cesophagus or crop has become very much more complex within the group. Within
the limits of the Docoglossa the nervous system has not greatly altered. The changes
in the visceral loop have already been mentioned. The anterior ends of the pedal
commissures are far wider apart in Acmeidze and Nacellide than in Patellide. The
latter show therefore a concentration perhaps correlated with the increased efficiency of
adhesion which undoubtedly characterizes them. The figures of the Docoglossan
nervous systems due to Bouvier, Pelseneer, Haller, and Miss Willcox are so widely distri-
buted in text-books that I have thought their reproduction in connection with this
paper unnecessary.

Thiele has observed various special patches of sensory epithelium whose phylogenetic
importance is not yet determined, so I pass them by and give merely a reference to
pages 326 and 327 of his recent paper (46).

The necessity for compactness of the viscera explains the development by the
Docoglossan right kidney of superficial lobes around the visceral hump superseding the
intervisceral lobes of the kidney of the ancestral form. The only intervisceral lobe
which remains is the small subrectal one into which the reno-pericardial canal opens.

SS ae ee ee” CU

TOPOGRAPHICAL RELATIONS AMONG TIE DOCOGLOSSA. 285

As the pericardium turned to the left and forwards, the left kidney was pushed before it
into the roof of the branchial cavity; it was also pulled with its excretory aperture to the
right and its Docoglossan position (Pl. 17. figs. 19, 20, & 24) is thus accounted for. The
reno-pericardial canals have already been correctly described by Goodrich and Pelseneer,
and it only remains to be said that their elongation, as evidenced by the position of their
renal apertures far from the pericardium (fig. 24), is a consequence of the pushing of the
excurrent apertures to the right and of the pericardium to the left. If the space-
relations in my sections (fig. 24) do not exactly correspond with those shown in the
drawings of other workers, I would ascribe this to individual variation and variation
with age, which undoubtedly occurs.

The moliuscan gonad shows an extreme adaptability of position, and, considering the
special pressure on the right, it is natural that it should usually occur on the left among
the Docoglossa.

Part [X.—AFFINITIES OF THE DocoGuossa.

The affinities of the Docoglossa amongst themselves have been discussed by Pelseneer,
Dall, Haller, and Thiele. It is now generally admitted that the Lepetide are specialized
forms adapted to a deep-sea life, though, in the retention of the median tooth of the
radula and of vestiges of the spiral, they show primitive features. They may be a basal
offshoot, but, as it is at any rate certain that the other groups are not descended from
them, I have ventured to omit them from the foregoing evolutionary sketch.

Pelseneer, Dall, and Haller all think that the Cyclobranchs have descended from
Monobranch-like forms, but Thiele disputes this because :—

(a) The ctenidium of Acemea has, according to him, no homology with that of other

Gastropods—it is a new development.

(2) The Cyclobranchs show equal remains of both ctenidia (in the osphradia and
neighbouring tissue).

(c) Dall and Haller have described rudiments of mantle-gills among the Monobranchs,
and believe these to be very early stages in the evolution of the Cyclobranch
gill-wreath. Thiele, however, sees in them disappearing vestiges.

The conclusion (@) above seems to me to be quite unjustifiable on the grounds given—
that the ctenidium is oblique and free except at the basis in Acmea, attached and free
only at the apex in other Gastropods; that ctenidium and osphradium are not so nearly
connected as in Haliotis, &c.; that the structure of the Acmeean gill is simpler than
that of others ; and that the efferent ctenidial vein joins the great mantle-vein.

This last is to be expected, as the great mantle-vein is undoubtedly a development of
the veins from the mantle which join the efferent branchial vein even in Zuliotis. The
simplicity of structure is easily understood when we believe that the Acmzean gill is
a structure arrested in an early stage of degeneration, and the other two characters do
not seem to me to be of morphological importance. I therefore accept the general
opinion that the ctenidium of Acmea is the left ctenidium of the Prostreptoneure,
probably in a somewhat degenerate condition. Perhaps Haller has made the evolution
of Cyclobranchs from Monobranchs appear too direct, but if we rather suppose, as

SECOND SERIES.— ZOOLOGY, VOL. IX. 40

286 DR. H. J. FLEURE ON THE EVOLUTION OF

I think he means, that both groups have descended from a common ancestor nearer the
Monobranchs, Thiele’s difficulty about the ctenidial vestiges disappears. It is easy to
understand that along one line of development both ctenidia were early reduced, while
along the other the reduction of one was arrested, so that it persisted in a probably
slightly degenerate state. The respiratory importance of the branchial roof in Patellidze
explains the maintenance of the osphradial sensory structures.

This view of the affinities of the subgroups, which is, I think, in full accord with
Pelseneer’s opinion, is also supported by the presence of probable vestiges of the
epipodium among the Cyclobranchs; the common ancestor would have retained this
ancestral organ to some extent. The odontophore also shows that the two orders have
diverged along somewhat different lines. The account given here of the consolidation of
the viscera shows, I think, that the Monobranchs are the more primitive group, but
that the ancestor of both may be supposed to have been intermediate between Acmea
galathea, for example, and Ancistromesus.

The affinities of the Docoglossa with other forms are difficult to trace on account
of the antiquity of the group. The Haliotide, Plewrotomaria, and the Trochide have
evolved far along other lines, especially as regards the shell-muscle, the branchial cavity,
the visceral loop of the nervous system, the disposition of the visceral hump and
its contents, the heart and pericardium, and the kidneys. Scdssurella is certainly some-
what more like the Docoglossa in external features in some species and in the condition
of the branchial cavity, but it is in other respects nearer the Haliotide &c. The
Fissurellidee have the same form of the shell-muscle and the condition of the kidneys is
similar to that among the Docoglossa, but it is probable that among them the complete
external symmetry is secondary. They have, besides, evolved on lines of reduction of
the shell and unique specialization of the branchial cavity, while the process of consoli-
dation of the viscera was also probably very different. It is possible that the Docoglossa
may be connected with the Bellerophontacea, which possessed symmetric shells that in
some cases show a tendency towards expansion of the rim, analogous to that which has
had such far-reaching effects among the Docoglossa. This is, however, quite problematic,
owing to our very limited knowledge of the palzozoic Gastropods. We are only
justified in hinting that the Docoglossa and Bellerophontacea are two of the earliest
offshoots from the Gastropod stem.

A List or Papers DEALING WITH THE DOCOGLOSSA.

1. Apams, L. E.—Deep Limpet “ Scars” [of Patella vulgata]. ‘ Naturalist,’ 1890, p. 335.

2. AMauprut, A.—La partie antérieure du tube digestif et la torsion chez les mollusques gastéropodes.
Ann. Sci. Nat., Zool. sér. 8, t. vii., 1898.

3. Bernarp, F.—Recherches sur les organes palléaux des Gastéropodes prosobranches. Ann. Sci. Nat.,
Zool. sér. 7, t. ix., 1890.

4. Bouran, L.—L’organe glandulaire périphérique de |’Helcion pellucidum. Arch. de Zool. exp. sér. 3,
t. v., 1897.

5. Bouran, L.—La cause principale de l’asymétrie des Gastéropodes prosobranches. Arch. de Zool.
exp. sér. 3, t. vil., 1899. (Includes an account of development of Acmea.)

2,

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA. 287

. Bouran, L.—La Patelle commune. Zoologie Descriptive par Apostolides ete. Paris, 1899.
. Bouvier, E. L.—Systéme nerveux des Gastéropodes prosobranches. Ann. Sci. Nat., Zool. sér. 7,

t. i1., 1887.

. Cooxe, A. H.—% Mollusca.” Cambridge Natural History, 1895.

. Cunnineasam, J. T.—The Renal Organs of Patella. Q.J. M.S. vol. xxiii., 1883.

. Cuvier, G.—Mémoires pour servir 4 Vhistoire naturelle des Mollusques. Paris, 1817.

. Dati, W. H.—Preliminary Sketch of a natural Arrangement of the Docoglossa. Proc. Boston Soc.

Nat. Hist. vol. xiv. p. 51.
Dati, W. H.—On the Extrusion of Seminal Products in Limpets. Sc. Res. Expl. Alaska, vol. i.
1876.

. Dati, W. H.—Review of the Docoglossa. Proc. U.S. Nat. Mus. iv. p. 400 (1882).
. Dart, W. H.—Ospkradia of Acmea. Reports Results Dredging ‘ Blake,’ vol. xxix. On Mollusca,

ii. Also Bull. Mus. Comp. Zool. xviii. p. 342 (1889).

. Dart, W. H.—Docoglossan Phylogeny. Proc. Ac. Philad. 1893, p. 285.

. Davis, J. R. Arnswortu.—The Habits of Limpets. ‘ Nature,’ xxxi. 1885, pp. 200-201.

. Davis, J. R. Atnsworto.—The Habits of Limpets. ‘ Nature,’ li. 1895, pp. 511-512.

. Davis, J. R. A., and H. J. Frevrse. “The Common Limpet.” Liverpool M. B. C. Memoirs,

1903.

. Ertaneer, R. von.—On the paired Nephridia of Prosobranchs. Q. J. M.S. vol. xxxiii., 1892.
. Fiscuer, H.—Recherches sur la morphologie du foie des Gastéropodes. Bull. scient. I'r. et Belg.

t. xxiv., 1892.

. Fiscuer, H.—Quelques remarques sur les mceurs de Patella. J. Couch. vol. xlvi. p. 214.
. Frevre, H. J.—On the Relations of the Kidneys in Hadliotis tuberculata, &c. Q.J.M.S., Jane

1902.

. Forszs, E., and 8S. Hantey.—History of British Mollusca. London, 1853.
. Greppzs, P.—On the Mechanism of the Odontophore in certain Mollusca. Trans. Zool. Soe. vol. x.,

1879.

- Gemnitt, J. F.—On some Cases of Hermaphroditism in Limpets. Anat. Anz. Bd. xii., 1896.

. Grsson, R. J. H.—Anatomy of Patella vulgata. Trans. Roy. Soc. Edin. vol. xxxu., 1885.

. Goopricu, E. S.—On the Reno-pericardial Canals in Palella. Q J. M.S. vol. xli., 1898.

. Hatier, Bera.—Studien iiber docoglosse und rhipidoglosse Prosobranchier. Leipzig, 1894.

. Hanitry S. See Forsus, E., and S. Haney.

. Lane, A., and K. Hescneter.—Lehrbuch der vergleichenden Anatomie der wirbellosen Tiere.—

Mollusea, Jena, 1900.

. Lanxester, HE. Ray.—Article “ Mollusca.” Encye. Brit. 9th ed.
. Lanxestrr, E. Ray.—On some undescribed Points in the Anatomy of the Limpet. Ann. & Mag.

Nat. Hist. ser. 3, vol. xx., 1867.

. Laygester, E. Ray.—On the originally bilateral Character of the Renal Organs of Prosobranchs.

Ann. & Mag. Nat. Hist. ser. 5, vol. viii., 1881.

. Lanxzsrer, E. Ray.—On the Ccelome and Vascular System of Mollusca and Arthropoda.

Q. J. M.S. vol. xxxiv., 1893.

. Kew, H. W.—The Faculty of Homing in Gastropods. ‘ Naturalist,’ 1890, pp. 307-318.
. Morean, C. Luoyp.—Animal Behaviour. London, 1900.
. Newsiain, (Miss) M.—On certain green Chlorophylloid Pigments in Invertebrates. Q. J. M.S.

vol. xli., 1898.

. Parren, W.—Embryology of Patella. Art. Zool. Inst. Wien, Bd. vi., 1885.
. Petsenrer, P.—Traité de Zoologie par R. Buancnarp. Fasc. xvi. Mollusques, par P. Petsenerr.

Paris, 1897.

288 DR. H. J. FLEURE ON THE EVOLUTION OF

40. Petsensrr, P.—Recherches morphologiques et phylogénétiques sur les Mollusques archiaiques.
Mém, cour. etc. pub. par Acad. Roy. de Belgique, t. lvii., 1899. Also references in other papers.

41. Perrinr, Remy.—L’anatomie et Vhistologie du rein des Gastéropodes prosobranches. Aun. Sci.
Nat., Zool. sér. 7, t. viil., 1890.

42. Pussry, W. H.—Tryon’s Manual of Conchology. Continued by Piispry. Marine Univalves,
vol. xill.

43. Scuorreip (and Uxrica). Geological Survey of Minnesota, vol. ii. pt. 2. “ Lower Silurian
Gastropoda.”

44. Srmroru.—Mollusca. Bronn’s Klassen und Ordnungen des Tierreichs.

45. SPRENGEL, J. W.—Die Geruchsorgane und das Nervensystem der Mollusken. Zeit. f. wiss. Zool.
Bd. xxxv., 1881.

46. Tuinie, J.—Die systematische Stellung der Solenogastren und die phylogenie der Mollusken.
Zeitschr. fiir wiss. Zool. xxii. (In this paper will be found a complete list of the author’s
previous publications dealing with Molluscan morphology.)

47. Tryon, G. W.—Manual of Conchology. Continued by W. H. Pirssry. See Piussry, W. H.

48, Uxnricn and Scuorie.p.— Lower Silurian Gastropods.””? Geol. Survey Minnesota, vol. iii. pt. 2.

49. Weemann, H.—Notes sur Vorganisation de Patella vu'gata. Rec. Zool. Suisse, t. iv., 1887.

50. Winicox, (Miss) M. A.—Zur Anatomie von dAemea fragilis. Jenaische Zeitsch. f, Naturw.
Bd. xxxii., 1898.

51. Wuiticox, (Miss) M. A.—Notes on the Anatomy of Aemea testudinalis. ‘ Science,’ vol. xi.

52. Wittzy, A.—Tidal Migrations of Patella. ‘ Nature,’ liv. 1896, p. 125.

53. Witx1ams, T.—On the Mechanism of Aquatic Respiration in Invertebrates. Ann. & Mag. Nat.
Hist. ser. 2, vol. xvii. 1856, pp. 28-42, 142-154, 247-258.

54. Woopwaxp, S. P.—A Manual of the Mollusca. London, 1851.

55. Zivrer, K. A. von.—Handbuch der Paliontologie. 1 Abth. Paliozoologie, Bd. u., Mollusken,
1885.

EXPLANATION OF THE PLATES.
(Hyp.= Hypothetical.)
PuatTe 15.

Fig. 1. Shell of Bellerophon expansus. Wenlock Limestone, Dudley.
2. Shell of embryo Acmea showing vestige of spiral.
3. The hypothetical Prostreptoneure, seen from the right side, the shell being supposed to be
removed.

Figs. 4 & 5. Stages illustrating the transition to a Docoglossan type.

Fig. 6. A supposed dissection of the Prostreptoneure, from the left side. The gut and pericardium
are shown as embedded in the general mass of the remaining organs of the body ; details of
the latter are omitted for the sake of simplicity.

7. Schematic diagram showing the settling-down of the viscera in the pro-Docoglossa in the
second stage of consolidation of the visceral mass.
8. The gut at the hypothetical stage represented in fig. 7. The rectum is only dotted in.
(The letter Q in figs. 8-10 should be deleted.)
9. The gut of Acmea galathea. After Haller.
10. The gut of Ancistromesus. After Haller.

TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA.

289
PLAte 16,

Fig. 11. The gut of Acmea virginea, A. testudinalis, or A. corticata.

Figs. 12 & 13. The proximal and distal portions of the gut in Patella vulgata.
Figs. 14 & 15. The proximal and distal portions of the gut in Patella cerulea.
Fig. 16a. The heart and pericardium of the hypothetical Prostreptoneure.

24. A

L.Au.
L.Ct.V.
ENG

Docoglossa.

17a. The heart of Lottia viridula.
17 6. The heart of Patella vulgata.
18 a. Oblique section through the pericardium of Acmea,

166,&c. Hypothetical diagrams illustrating the transition to the condition in the primitive

After Haller.

PuatTE 17,

Fig. 184. Oblique section through the pericardium of Patedla.
19. Dorsal view of Acmea testudinalis after the removal of the shell, the pigmented epithelium, and

the roof of the branchial cavity.

respectively.

20. Corresponding view of Patella vulgata.
2la,b,&c. A portion of the mantle-edge in Acmea testudinalis, A. virginea, and P. vulgata

22 a-f. A series of transverse sections through region A of the gut of Patella vulgata, showing

merely the longitudinal folds.

23. A corresponding diagram, showing the maximum torsion observed in Acmea virginea.

transverse section showing the renal communications of the reno-pericardial canals in

Patella.

. = Anus.

. = Anterior aorta.
. = Intrapericardial ‘ Bulbus Aorte” in fig. 18.

Wide region of absorption and admixture of secretions in gut.

. = Buccal glands.

Reference Letters used in the Plates.

. = Fore-gut or special secretory region.

. = Branchial cavity.

= Intestinal region of gut.

. = Ctenidium.

= Foot.

. = Terminal region of intestine.

. = Folds which are dorsal in the gullet-wall.
; = bive:

. = Epipodium.

Glands of mantle-edge.

= Greater mantle-tentacle.

= Great mantle-vein.

Patella, &e.
Left auricle.

. = Great digestive gland.
. = Boundary between the two kidneys in fig. 24.
. = Special wide loop of intestine (region C) which immediately follows region B in Acmea,

= Left efferent ctenidial vein.

Left kidney.

SECOND SERIES.—ZOOLOGY, VOL. IX. 41

290 ON TOPOGRAPHICAL RELATIONS AMONG THE DOCOGLOSSA.

L.K.Ap. = Aperture of post-torsional left kidney,
L.-M. = Junction of loops L and M of the intestine.
L.M.T. = Lesser mantle-tentacle.
L.R.P. = Reno-pericardial canal of the left kidney (renal pore).
M. = Dorsal intestinal loops characteristic for Cyclobranchs.
M.-L. = Junction of loops L and M of the intestine.
Meso. = Connection between ventricle and dorsal wall of pericardium.
Mt. = Mantle-skirt.
M.Tent. = Mantle-tentacle.
N. = Ventrally-placed loop of intestine following loop M in Patella vulgata.
N.R.V. = Veins from nuchal roof.
Oes.P. = Lateral esophageal pouches.
Op. = Operculum.
Osph. = Osphradium.
P. = Pericardium.
Post.A. = Posterior aorta,
Pr. and arrows indicate direction of pressure of viscera on pericardium.
P.Z. = Pigment-zone.
Q (curve) indicates the part of region A of the gut which is pushed to the left by pressure from
the right anterior corner.
(Q, indicates the part of region A of the gut which is pushed inwards and backwards by pressure from
the right anterior corner.
R. = Rectum.
R.Au. = Right auricle.
R.Ct.V. = Right efferent ctenidial vein.
RK. = Right kidney.
R.K.Ap. = Papilla and aperture of post-torsional right kidney.
R.R.P. = Reno-pericardial pore of the right kidney.
S.M. = Shell-muscle.
Sp.C. = Spiral cecum.
St. = Slit in mantle.
T. = Cephalic tentacle.
V. (upper & lower) = Upper and lower parts of the consolidating visceral mass.
V.F.G. = Folds which are ventral in the gullet-wall.
Vn. = Ventricle.
X and 8 arrows indicate pressure obliquely backwards from the right anterior corner of the visceral
mass causing curve ).
Y and 2 arrows indicate pressure inwards from the right side of the visceral hump,
Z and 1 arrow indicate pressure forwards at the back of the visceral hump.

i

Trans .Linn.Soc.Snr.2.Zoor .Vou.IX.Pn.15.

ANATOMY OF THE DOCOGLOSSA. ’
MP Parker lith.

Mintern Bros.imp.

Blears ; Trans .Linn.Soc.Ser.2.Zoor.Vou.IX.PxuI16.

a, ee)

; AO : SS _ er . 16 b. (Hyp)

-

s i.

t

aa

ANATOMY OF THE DOCOGLOSSA.
M.P.Parker lth.
Mintern Bros imp.

-
.
.
*
a4
.
.
wh
' i ‘4 .
. -
i
‘
7 P . id
‘ ae
(eee as
a
ae!
Pern ‘ a
4% * ie

Fleure.

Trans.Linn .Soc.Ser.2.Zoou. Vou. IX. Px.17.

ANATOMY OF THE DOCOGLOSSA.

4
]
J

VIII. On some Species of the Genus Paleemon, Fabr., from Tahiti, Shanghai, New
Guinea, and West Africa. By Dr. J. G. pe May, of Jerseke (Holland).
(Communicated by the Rev. T. R. R. Sressine, I.A., F.R.S., Sec.L.8.)

(Plates 18-20.)

Read 3rd March, 1904.

THE following Report contains the description of some species of the genus Palemon,
Fabr., several of which were sent to me for examination by Prof. F. Jeffrey Bell, of the
British Museum (Natural History), whereas the others, belonging to my private collec-
tion, were gathered in fresh water at Catumbella, near Benguella, by Mr. P. Kamerman,
the same gentleman to whom I am also indebted for the interesting Brachyura from
Angola described four years ago in Mémoires Soc. Zool. de France, 1900, pp. 31-65,
pls.i., ii. One species from Catumbella has not previously been observed in West Africa,
and may even eventually prove to be new to science; another, from Cameroon, unfortu-
nately represented by only one specimen, seems to be the rare P. Foai, Cout., the exact
habitat of which is still unknown ; and finally aremarkable hitherto unknown character
of P. asperulus, v. Martens, was observed, namely, that some segments of the abdomen
are carinated.
The following species are described :—

Palemon (Eupalemon) lar, Fabr. Palemon (Macrobrachium) jamaicensis (Herbst),
Palemon (Parapalemon ?) asperulus, v. Martens. var. Vollenhovenii, Herklots.

Palemon (Macrobrachium) latimanus, vy. Martens. Palemon (Macrobrachium) Olfersii, Wiegm.
Palemon (Eupalemon) macrobrachion, Herklots. Palemon (Macrobrachium ?) sp.

Palemon (Eupalemon) Foai, Cout.

A.—INDO-PACIFIC SPECIES.

PaLZMON (EUPALZMON) LAR, Fabr. (Plate 18. fig. 1.)

Confer: Spence Bate, Report on the ‘Challenger’ Macrura, 1888, p. 789, pl. 129. fig. 1; de Man,
Notes from the Leyden Museum, i. 1879, p. 168 (sub nomine Pal. ornati, Oliv.), and in Max
Weber, Zoolog. Ergebn. Reise Niederl. Ost-Indien, ii. 1892, p. 445, and Abhandlungen der
Senckenbergischen Naturf. Gesellschaft, Bd. xxv. Heft iii. 1902, p. 774.

Four adult males from fresh water at Tahiti. British Museum (Natural History).
The tooth-formulee and the measurements of the legs of the second pair are indicated
in the Table. These specimens closely resemble the figure quoted from the ‘ Challenger ’
Report. The rather slender rostrum is as long as or somewhat longer than the peduncles
SECOND SERIES.—ZOOLOGY, VOL. IX. 42

292 DR. J. G. DE MAN ON SPECIES OF PALZMON

of the upper antenne, though not reaching to the end of the scaphocerites ; the upper
margin is slightly convex above the eyes, and its distal half is more or less turned
upward. The third tooth is situated just above the orbital margin; in three specimens
this tooth is as long as the second, and both are longer than the others, but in the male
No. 4 these teeth are not longer than the rest.

The telson of No. 1 and No. 2 ends in a short median acute tooth (fig. 1); the inner
of the two spines on each side is three times as long as the outer, and overreaches the
median tooth by its distal half. In the two other specimens the median tooth and the
spines are more or less worn off. In these four specimens the second legs are of about
equal size ; in all the slender fingers are gaping, just as in the figure quoted, in consequence
of the strong development of the teeth with which they are armed, and they are also
somewhat curved inward. The dactylus is regularly curved towards the tip and usually
a little shorter than the immobile finger; the latter is provided with a conical tooth at a
fourth of its length from the articulation and with a smaller one between this tooth and
the proximal end of the finger. The dactylus bears also a conical and compressed tooth
at a third of its length from the articulation, which is not smaller than the tooth of the
other finger, and between this tooth and the articulation is seen four or five much
smaller obtuse teeth ; sometimes, however, the latter are more or less rudimentary.

The second legs are of a fine dark purple colour, darkest on the fingers, which
appear almost black ; the articulation of the fingers is beautiful orange-red, as also
the articulation between carpus and palm, and the fingers are marked with pale spots,
as in the figure in the ‘Challenger’ Report. The dorsal surface of cephalothorax and
abdomen and the posterior margin of the upper teeth of the rostrum have also a purple
colour, though paler than that of the second legs, being more cherry-red. As regards
their colour, our specimens apparently agree with those from the River Papeuriri in
Tahiti, described in the Report on the ‘ Challenger’ Macrura.

An adult male and an adult female from Patani, on the island of Halmahera, are now
before me (vide de Man, in Abhandl. Senckenb. Naturf. Gesellschaft, xxv. Heft 3, 1902,
p- 777). In these specimens the cephalothorax and the abdomen are wot adorned with
the beautiful purple colour observed on the specimens from Tahiti, and the three posterior
pairs of legs are longer and more slender. So, e.g.,in the adult male from the River
Tobelo, that is 115 mm. long, the meropodites of the fifth pair are 16 mm. long, and,
measured on their outer side, in the middle 1-6 mm. broad; the carpopodites are 10°5 mm.
long, and 1:6 mm. broad at the distal end ; the propodites havea length of 17 mm. and are
0-96 mm. broad in the middle, the terminal joints finally are 4-5 mm.long. In our male
(No.4), however, of exactly the same size, the meropodites of the fifth pair are 14 mm. long
and 1°85 mm. broad in the middle; the carpopodites are 8 mm. long and 1°6 mm. broad
at the distal end; the propodites are 15 mm. long and 1:12 mm. broad in the middle, the
terminal joints finally are 3°5 mm. long.

The Tahiti species may therefore be regarded as a local variety, for which I propose
the name spectabilis, because it is certainly identical with P. spectabilis, Heller, from
the same island.

FROM THE INDO-PACIFIC AND WEST AFRICA. 293

Measurements in millimetres.

= = = - — —__——_____—— _ _-- = = |

| 1. 2 3. 4

| é. d 3
Tenpth of the animal ................ 125 118 116 110
Length of the whole leg .............. 155 «145 153 147 122 116 138 185

ieneth of the merus ...............505 30 29 28:5 28 24 23 27 26-5

| Width of the merus at the distalend ....) 54 52 | 5 5 45 45 | 475 4:75
Weneth) of the carpus .....5..2...5..+.- 27 26 27 26 21 21 25 25
Width of the carpus at the distalend ....| 5:5 5 55 5:25 475 4°75 5 5
onetbvor the palms \).)j.). 25.2.0. os | 42 37 43 40 33 30 37 36
Breadth of the palm in the middle ...... 5°75 4°75 5°33 (525 475 45 5:25. 5
Thickness of the palm in the middle .... 5 4°25 45 4:5 4:25 4 45 45
Menoinvor therfingers, (2 «lis cfa. «ste es 305) e2ii 29 28 23°5 21:5 | 2675 25
Length of the whole hand ............ 725 64 72 68 565 515 63°5 61
Formula of the rostrum .............. ; = 5 f

PALAZMON (PARAPALZMON ?) ASPERULUS, v. Martens. (Plate 18. figs. 2-8.)

Palemon asperulus, v. Martens, Archiv Naturg. xxxiv. Jahrg. 1868, p. 43, Taf. 1. fig. 5.
Palemon asperulus, Ortmann, Zool. Jahrb., Syst. v. p. 708.

One young female from South Hu-peh, China. British Museum (Natural History).

Though P. asperulus was described a third of a century ago, it has apparently never
been observed since that time, and, so far as I am aware, no mention of this species
has been made since 1868, except by Dr. Ortmann, who, however, did not have any
specimens at his disposal. We must therefore regret that only one young female has
been collected, for, according to von Martens, P. asperulus isa common species in the
fish-market of Shanghai.

Our female is 47 mm. long from tip of rostrum to the extremity of the telson, quite
young therefore, this species attaining, indeed, a length of 83 mm. The rostrum (fig. 2)
is short and reaches to the distal end of the antennulary peduncles ; it is lanceolate, and
appears very slightly arcuate above the eyes. The upper margin bears 10 equal teeth,
the first of which is placed just twice as far from the posterior as from the anterior
margin of the cephalothorax. This first tooth stands a little farther from the second
than the following teeth, that are equidistant and reach to the tip of the rostrum. The
under edge bears 8 teeth, which are smaller than those of the upper margin; the first
is situated just below the seventh tooth of the latter, and these teeth have also about
the same length and extend to near the tip. The rostrum has a stout shape and that

part which is situated above the lateral crest appears, in the middle of the rostrum,
. 42*

294 DR. J. GQ. DE MAN ON SPECIES OF PALAMON

as high as that below it. Three teeth are on the cephalothorax, the fourth is situated
just before the orbital margin. :

Von Martens describes the rostrum as being as long as the scaphocerites (J. ¢. p. 34),
but in his figure it does not reach so far; this figure, however, is inaccurately drawn, for
the lateral spines of the cephalothorax have been forgotten altogether.

The hepatic spine is a little smaller than the antennal, and is situated posterior to and
somewhat below the latter.

The cephalothorax is covered, though not closely, with very small spinules, which are
only 0:04 mm. long. The abdomen is finely punctate. The first segment (Pl. 18. fig. 3)
presents a remarkable form, which has not hitherto been observed in any species of
this genus, the dorsal part of it being distinctly tricarinate by three, not very sharp,
longitudinal carine, that reach from the anterior to the posterior margin of the segment.
The two lateral carinze converge very slightly backwards, and the upper surface is concave
between the median crest and each lateral one; even on the outer side of each lateral
carina the surface appears a little concave, though a second lateral crest does not exist.
Though the second segment is not at all carinate, its surface appears, however, on each
side of the median line, near the posterior margin, very slightly concave. The third and
fourth segments are rounded above, but the fifth is distinctly carinate in the median
line of the dorsal surface, though only along the posterior half; on each side of the crest
the surface appears a little concave. The telson tapers rather strongly towards the
posterior extremity (fig. 4), that ends ina sharp acute median tooth, which is shorter than
the internal of the two spines on each side of it. The upper surface of the telson is
roughened by minute spinules, which are still smaller than those of the cephalothorax.
The anterior pair of spines on the upper surface is a little farther from the anterior end
of the telson than from the posterior extremity; it is, of course, very seldom that a
third spine occurs on the left side, close to the anterior one.

The short filament of the upper antennee is united for a very short distance with the
outer one, this distance being only one-third of the length of the third joint of the
peduncle. The external footjaws project half their terminal joint beyond the peduncles
of the outer antennz. The legs of the first pair overreach the scaphocerites by the
length of their hands: the fingers are just as long as the palm. The carpus, which is
somewhat thickened as usual at its distal end, is 5:5 mm. long, the hands are 3°6 mm.
long; so that the carpus is only once and a half as long as the hand (fig. 5).

The legs of the second pair (fig. 6) are equal and rather feeble; they project a
fourth of their wrist beyond the scaphocerites. Measured along its upper margin,
the merus appears 5:2 mm. long; this joint is almost cylindrical and thickens but very
slightly anteriorly, so that it is 16 mm. thick at the distal end. The carpus, which
is 6-4 mm. long, is distinctly somewhat longer than the merus: von Martens says that
both joints are equally long, but in his figure the merus appears distinctly shorter.
The carpus also thickens gradually towards the distal end, and is, moreover, slightly
compressed ; in consequence of this, the distal extremity appears 1-9 mm. broad looked
at from above, but only 1°65 mm. when it is measured at the outer side (fig. 7).

a

es

FROM THE INDO-PACIFIC AND WEST AFRICA. 295

The carpus bears on its outer side a longitudinal ridge along its whole length, and this
ridge is continued for a short distance, about 15 mm., along the outer margin of the
palm. Both merus and carpus bear, moreover, on their upper surface an impressed
longitudinal line, which runs in the same direction from the distal end of the ischium to
that of the carpus (fig. 6). The hand is 13 mm. long, the palm measures 7°5 mm., the
fingers 5°5 mm.; the palm is thus somewhat longer than the carpus and the fingers, and
the proportion between palm and fingers also conforms to the original description,
according to which their proportion should be as 3:2. The lateral margins of the palm,
which is 2 mm. broad in the middle, are parallel with each other, so that its upper surface
presents the same breadth almost along its whole length; the palm is 15 mm. thick in
the middle, so that it appears a little compressed in the proportion of 4:3. The fingers
shut close together and are somewhat curved inward, so that the inner margin of the
hand runs very slightly concave ; the fingers are also a little broader than thick
and they have the same breadth to near their acute tips. Hach finger bears a sharp
cutting-edge, at the proximal end of which is seen a small conical tooth; the dactylus
bears, moreover, a second tooth between it and the articulation, so that the tooth of the
immobile finger is situated just between the two of the dactylus. One observes on each
side of the cutting-edge the usual short hairs as in other species, and on the upper and
lower surface of the fingers small tufts of short hairs, which are somewhat longer near
the tips, but otherwise the fingers are quite smooth. The straight inner border of the
palm is beset with very small spinules, 0-12 mm. long, visible only by means of a
magnifying-glass; still smaller ones are seen on the outer margin; the lower surface of
the palm bears also a few microscopical spinules, but the upper is nearly smooth. The
palm is also a little hairy; the hairs are short, fine, and widely separate. The other
joints of these legs are everywhere covered with similar small spinules and similar short
fine hairs. The second legs are 32 mm. long, measuring just ¢wo-thirds the length of
the animal; they have a pale flesh-colour.

The three posterior pairs of legs are short and stout. Those of the third pair reach

nearly to the end of the antennal scales, the two others are but little shorter. The

meropodites of the third pair (fig. 8) are 5°33 mm. long, measured along their upper
margin, and 1 mm. thick on their outer side; the propodites are 5 mm. long and 0°72 mm.
broad in the middle, the terminal joints finally are 1:9 mm. long. ‘The meropodites of
these legs are thus five, the strongly compressed propodites seven times as long as broad
on their outer side, and the terminal joints are little longer thana third of the propodites.
The propodites are armed with nine mobile spines along the posterior margin of their
upper and lower surfaces; these spines measure 0°32-0°38 mm., and are about half as
long as these joints are broad. The propodites are a little hairy on their upper and on
their lower surface, as also along their anterior margin, but are otherwise smooth; the
other joints are also a little hairy and almost smooth, though a few microscopical spinules
may be distinguished under the microscope.

Palemon asperulus should very likely be referred to the subgenus Parapalemon.

Those species with which P. asperulus is most closely allied are P. (Parapalemon)

296 DR. J. G. DE MAN ON SPECIES OF PALAMON

javanicus from Java and Sumatra, Palemon (Parapalemon) Horstii, de Man, from
Celebes, and P. (Hupalemon) elegans, de Man, from Java. <A female with eggs of
P. javanicus, Heller, described by myself in the ‘ Notes from the Leyden Museum,’ i.
1879, p. 181, is now before me, and also type specimens of P. elegans from Buitenzorg.
The female of P. Horstii is still unknown, so that the type specimens of this species
were not required.

The female of P. elegans differs at first sight from our female of P. asperulus by
the legs being much more slender and by the slender carpus of the second legs being
distinctly longer than the palm; the abdomen is not carinate, and there are other
differences.

The abdomen of P. javanicus is also rounded, without a trace of the longitudinal
carinz that are characteristic of the Chinese species. In the female of P. javanicus
all the legs are, moreover, more slender. The carpus of the first pair is twice as long as
the hands. In the second legs, which are also considerably more slender than those of
P. asperulus, the palm appears in the right leg just as long as, but in the left distinetly
shorter than, the carpus.

The legs of P. (Parapalemon) Horstii are short and stout, and evidently bear a close
resemblance to those of P. asperulus; but if my figures of the second legs of the male
(in Max Weber’s ‘ Decapoden des Indischen Archipels,’ 1892, Taf. 27. fig. 39) are
compared with the figure of P. asperulus in von Martens’s paper, there can be little
doubt that the species are different. The rostrum closely agrees in both species, but in
that of P. Horstii there are four teeth on the cephalothorax.

P. (Parapalemon) asperulus, v. Martens, has hitherto only been observed at
Shanghai.

PaLm=MON (MACROBRACHIUM) LATIMANUS, v. Martens. (Plate 18. figs. 9-12.)

Palemon latimanus, von Martens, Archiv Naturg. Bd. xxxiv. 1868, p. 44.

Palemon latimanus, de Man, Archiv Naturg. Jahrg. liii. 1888, p. 557; and in Max Weber’s Zoolog.
Ergebn. Reise Niederl. Ost-Indien, ii. 1892, p. 477, Taf. 28. fig. 45; and in Abhandl. Sencken-
bergischen Naturf. Gesells. Bd. xxv. Heft 3, 1902, p. 780.

Palemon (Macrobrachium) latimanus, Nobili, Annali Museo Civico Genova, Ser. 2, vol. xx. (xl.) 1900,
p- 485, figs. 3 a-c & 4.

Palemon latimanus, Schenkel, in Verhandl. Naturf. Gesells. Basel, Bd. xi. Heft 3, p. 512 (1902).

Five adult males and one young female from Dinawa, Owen Stanley Range, 120 miles
inland from Yule Island, at an altitude of 4000 feet.

In the third of my papers quoted above I supposed that the two male specimens from
Celebes, 107 mm. and 103 mm. long, were adult, because they were even a little longer
than the type specimen described by von Martens, which measured 97 mm. The
five adult males from Dinawa, which are all about the same size, are, however, still
considerably larger, for they measure from tip of rostrum to the end of the telson 130-
140 mm. Specimens of this size have never been examined, so far as I am aware.
Unfortunately, however, they have all lost the legs of the second pair, but one second
leg is lying loose in the bottle.

FROM THE INDO-PACIFIC AND WEST AFRICA. 297

In the large males the rostrum extends to the middle of the terminal joint of the
antennulary peduncles, or to the end of it ; it closely resembles the figure published by
Nobili of the type specimen of von Martens, but in some specimens the upper margin is
more strongly inclined downward. In these five individuals the upper margin bears five,
the lower two teeth, and in all the first ¢wvo teeth stand on the cephalothorax. In
two males the second tooth of the upper margin is just as long as the following two
together (fig. 10), but in the others it is but little longer than the third tooth (fig. 9).
In Nobili’s figure of the type the second tooth of the lower margin is as far from
the first tooth as from the tip, but in the Dinawa males the two teeth stand close
together and the second is considerably farther from the tip than from the first tooth.
In the young female which is 75 mm. long the rostrum is -5 mm. longer than the
peduncles, and it is armed above with 6, below with 3 teeth ; two stand again on the
cephalothorax, the second is but little larger than the following, and the foremost tooth
is still a little farther from the tip than the length of the fifth and of the sixth tooth
taken together. The three teeth of the under margin are smaller than those of the
upper; the third reaches to the distal end of the penultimate joint of the upper peduncles.

Only in one single male (fig. 11) the extremity of the telson ends in a short acute
median tooth ; the inner of the two spines on each side is twice as long as the outer
and extends beyond the median tooth a third of its length. In the other specimens
the extremity of the telson is more or less worn off, and in the female this segment is
wanting.

In these adult males the outer footjaws project the distal fourth of their penultimate
joints beyond the peduncles of the outer antenne. The legs of the first’ pair extend a
third or a fourth of their wrist beyond the scaphocerites. The fingers are just as long
as or a little longer than the palm, and the carpus is once anda half as long as the
hand.

Only in one male, which has a length of 130 mm., the first four joints of the right
leg of the second pair are still present; the merus, which is 23 mm. long, extends
2 mm. beyond the scaphocerites.

The large left leg of the second pair, which lies loose in the bottle (fig. 12), is
152 mm. long, and is consequently somewhat longer than the animal, for it no doubt
belonged to one of the five males. It closely agrees with Nobili’s figure of the type,
but the palm is comparatively longer and the dactylus appears a little longer than
the immobile finger. Measured along the outer margin, the merus appears to be
33 mm. long; looked at from above the proximal extremity that articulates with the
ischium, it appears to be 5 mm., the distal end, however, 8 mm. broad. ‘The carpus is
22 mm. long and measures, in conformity with the original description, two-thirds of the
merus; viewed from above its proximal extremity, it appears 3°75 mm_ broad, and its
greatest width, at a fourth of the whole length from the distal end, measures 10 mm.
The hand, measured along the outer margin, appears to be 71 mm. long, the palm
42°5 mm., the tingers 28°5; the palm is twice as long as the carpus, and the fingers are a
little shorter than the merus. ‘The palm is, in the middle of its length, 11 mm. broad
and 7-75 mm. thick, being compressed in the proportion of 1:1°4; in the type specimen

298 DR. J. G. DE MAN ON SPECIES OF PALZMON

figured by Nobili this proportion is as 1; 1°22 or as 1: 1:23 (vide de Man, in Max Weber’s
‘ Decapoden des Indischen Archipels,’ p. 479). This difference depends, no doubt, upon
the difference of age, or is perhaps individual. In a male specimen from Minahassa the
palm of the left leg was 25 mm. long, 10°5 mm. broad, and 85 mm. thick; though
the palm was a little less broad than in the leg that I am describing, it was, however,
somewhat thicker, so that the proportion was as 1:1:23 (de Man, in Kikenthal’s
‘ Decapoda,’ 1902, p. 783). Zhe palm is thus four times as long as broad; in the type
figured in the ‘ Annali del Museo Civico di Genova’ the palm appears three times as
long as broad, but the type specimen is much younger, being only 97 mm. long, and the
leg that is figured measures only 85 mm. The fingers of our specimen shut close
together, and the dactylus is almost 15 mm. longer than the immobile finger. At the
end of a short black-coloured cutting-edge, which extends almost along a third of the
length of the finger from the tip, is seen a small conical tooth, and another of the same
size stands in the middle between this tooth and the articulation; between this second
tooth and the articulation the finger presents two smaller teeth, and between this second
tooth and that at the end of the cutting-edge are seven small teeth that gradually
decrease in size distally. The black-coloured cutting-edge of the immobile finger is
somewhat longer, extending along two-fifths of its length, and it presents a conical
tooth at its proximal end; between the latter and close to the articulation this finger
carries fifteen small teeth, of which two or three in the middle are a little larger than the
others. The dactylus, however, carries only ten teeth between the cutting-edge and the
articulation. The number of these teeth apparently increases with the age; in a young
male from the island of Rotti, 50 mm. long, each finger bore only nine teeth between
the tooth at the end of the cutting-edge and the articulation (de Man, in Max Weber's
‘Decapoden des Indischen Archipels,’ 1892, p. 480). The whole leg is covered with
small spinules. On the outer margin of the merus they are less numerous than
elsewhere; but on the outer, and in a less degree also on the inner, side of the carpus
these spinules are more crowded than on the upper and lower surface of this joint. The
outer margins of the palm and of the dactylus are closely beset with these spinules, but
their number gradually decreases towards the inner margin of the palm, and on the upper
as also on the lower surface of both fingers they are few in number. The spinules on the
inner margin of the immobile finger are very slightly larger than those on the inner
margin of the palm and than the crowded spinules on the outer margin of the dactylus.
This leg has a reddish colour, but on the inner side of merus and carpus it is marbled
with black, as are also the distal extremities of the red-coloured fingers.

The three posterior legs are stout. Those of the third pair project a third or a fourth
of the length of their carpopodites beyond the scaphocerites ; the fourth pair is somewhat
shorter; and the fifth legs extend scarcely beyond the middle of the antennal scales.
The meropodite of the fifth legs of one of the adult males from Dinawa is 17 mm. long
and 2°75 mm. thick in the middle; the carpus, measured to the articulation with the
propodite, appears 10°5 mm. long and 2°66 mm. thick on its outer side; the propodite
is 18 mm. long and 1°75 mm. broad in the middle, and the dactylopodite measures 5 mm.

In all the specimens of P. latimanus that I have examined anteriorly, the upper

i ae

a ives. *

FROM THE INDO-PACIFIC AND WEST AFRICA. 299

margin of the rostrum was armed with 10, 9, rarely 8 teeth, which reached to the tip, and
Nobili describes also 8 or 9 teeth for specimens from the Mentawei Islands; in seven
individuals from Celebes, recently described by Schenkel, three were armed with 9, the
four others with 7, 8,11, and 12 teeth. The type specimen from the island of Samar
finally had a rostrum armed with 6 teeth on the upper margin. We may consequently
conclude that the number of these teeth is subject to much variation, and there are
perhaps local varieties in which it is constantly larger. The male specimen from North
Halmahera, 74 mm. long, that I have described in the third of the papers quoted
(p. 782), is lying before me. The meropodites of the fifth pair are 9 mm. long and
1:33 mm. broad on their outer side; the carpopodites are 5 mm. long and 1:2 mm. thick at
the distal end; the propodites are 10 mm. long and 0°85 mm. broad in the middle; the
dactylopodites finally are 2°75 mm. long. The three posterior legs are thus comparatively
a little less stout in proportion to their length than those of the adult specimens from

‘Dinawa.

Palemon (Macrobrachium) latimanus has hitherto been recorded from the following
localities :—Mentawei Islands (Nod¢li); Flores (de I); Rotti (de I); Timor (de JZ.) ;

_ Amboina (de M.); Halmahera (de JL); Celebes (de IL, Schenkel); Samar (von Martens) ;
Fiji Islands (Ortmann). Fresh water.

B.—SPECIES FROM WEST AFRICA.

PaLaMon (EUPAL#MON) MACROBRACHION, Herklots. (Plates 18. and 19. figs. 13-29.)

Palemon macrobrachion, Herklots, Additamenta ad Faunam Carcinologicam Afric Occidentalis, L. B.,

1851, p. 15.

Palemon macrobrachion, de Man, in Notes from the Leyden Museum, i. 1879, p. 177.

Confer: Carl W.S. Aurivillius, ‘ Krustaceen aus dem Kamerun-Gebiete,’ in Bihang till K. Svenska Vet.-
Akad. Handlingar, Bd. xxiv. Afd. iv. No. 1, 1898, p. 19.

The following collection is lying before me :—

One male of medium size, obtained by Messrs. Biittikofer and Sala in 1881 in Liberia.
(Leyden Museum.)

One adult male and one somewhat younger female without eggs, from the River Prah,
South of Ashantee, West Africa. (British Museum, Natural History.)

Nine specimens of medium size (1 ¢, 8 ¢) from the Congo Coast, probably from
Ambriz.

About 300 young and very young specimens, half of which have unfortunately lost
their legs of the second pair, from fresh water at Catumbella, near Benguella ; presented
to me, together with the Congo specimens, by my cousin, Mr. P. Kamerman.

Dr. Aurivillius, who has also studied a large number of specimens of this species
collected in Cameroon, was led to the conclusion that P. macrobrachion ought to
be identified with P. acanthurus, Wiegm.; and he is followed in this opinion by
Miss Rathbun in Proc. U.S. National Museum, xxii. 1900, p. 315. The identity of
both species appears very probable also to me: nevertheless I prefer to describe the

SECOND SERIES.—ZOOLOGY, VOL. IX. 43

300 DR. J. G. DE MAN ON SPECIES OF PALZMON

African specimens still provisionally under the name given to them by Herklots, because
neither Aurivillius nor myself was enabled to study also American individuals of
Palemon acanthurus. It may be possible, indeed, that slight differences exist between
the two species, differences that have been overlooked in the published descriptions,
and we must consider that von Martens, who examined American specimens of
P. acanthurus together with one from Sierra Leone, was unable to decide whether
the latter specimen was to be referred or not to P. acanthurus (vide Archiv f. Naturg.
xxxv. Jahrg. 1869, p. 30).

The rostrum of the male specimen* from Liberia, the cephalothorax of which is
closely covered with minute spinules, differs rather much from all the other specimens in
the large collection lying before me. In the first place, ¢¢ is but little longer than
the peduncles of the internal antenne, much shorter than the scaphocerites, and it
extends straight forward, being not at all upturned at the extremity (Pl. 18. fig. 13). The
11 teeth of the upper margin extend to the tip, there being no smooth interspace
separating one or two apical teeth from the preceding. 'The first two stand on the
carapace and are somewhat more distant from one another than the others. One
observes 5 teeth along the distal half of the lower margin, the first of which is situated
just below the seventh tooth of the upper border.

The right leg of the second pair is almost once and a half as long as the animal,
the left is but 4 mm. shorter. As regards the length and the shape of the joints,
these legs fully agree with that of the adult male from the River Prah described below,
and the same conformity is observed as regards the number, the form, and the arrange-
ment of the spinules with which the joints are beset. Hach finger is armed with a small
conical tooth, but between this tooth and the articulation there are only two or three
smaller teeth; both fingers of each leg are thickly covered with a woolly felt. The
three posterior legs are comparatively just as long as in the adult male from the River
Prah and agree also in the other characters; the meropodites of the legs of the third
pair are 11°5 mm. long and, in the middle, 1:66 mm. broad.

The rostrum of the adult male (Table, No. 2) from the River Prah is broken off,
only the three posterior teeth of the upper margin are still present; two of them are
on the carapace, the third just above the orbital margin. The cephalothorax is somewhat
roughened anteriorly, especially on the sides, by microscopical spinules, only visible by
means of a magnifying-glass. They exist also on the telson. In my first description of
P. macrobrachion (1. c. 1879) the carapace is said to be smooth. The anterior pair
of spinules on the upper surface of the telson is situated exactly in the middle, the
posterior pair just half-way between the anterior and the extremity. ‘The latter
ends in an acute median tooth in the middle, on either side of which are inserted the
two usual spines; the internal spines extend considerably beyond the median tooth.

The inner or shorter of the two outer flagella of the upper antennze are 20 mm. long
from their extremity to the end of the peduncle, being slightly longer than the latter;

* For the measurements of the body and of the second legs, as also for the toothing of the rostrum, I refer to the

Table on p. 821, where the dimensions of 88 specimens are given. The joints of the second legs haye been measured
along their outer margin.

FROM THE INDO-PACIFIC AND WEST AFRICA. 301

they are united for a fifth of their length with the outer flagella and are slightly
serrated.

The external maxillipedes extend a third of their terminal joint beyond the peduncle
of the lower antenne. ‘The legs of the first pair, which are 42 mm. long, project with
_ little more than their hands beyond the tip of the antennal scales. The slender carpus
is 18 mm. long, and 1°33 mm. thick at the distal end; it is almost three times as long
as the chela, which is 6:66 mm. long and the fingers of which are slightly longer than the
palm. Only one leg (fig. 14) of the second pair is still present ; this leg is a little longer
than the animal. The merus extends for about a third of its length beyond the antennal
scales; it-is straight and thickens slightly to a little behind its distal end, its greatest
width being 6-5 mm.; this joint appears consequently five times as long as thick. The
carpus is one-fourth longer than the preceding joint; it is slender and thickens slightly
and rather regularly towards the distal end, its diameter being here about one-eighth
of the whole length. The carpus is, however, not quite straight, but very slightly
curved, so that the inner margin appears a little concave, the outer very slightly convex.
The palm has about the same length as the merus and is as much shorter than the
carpus as the latter. The almost quite straight palm is cylindrical, presenting over
its whole length a diameter of 4 mm., so that this joint is 8 times as long as thick. The
fingers, which are thickly covered with «@ woolly felt, quite close together ; the immobile
finger is straight, the dactylus, however, very slightly arcuate. The immobile finger is
armed with a small conical tooth, about 0°75 mm. high, at one-third of its length from the
articulation, and between this tooth and the articulation seven or eight smailer teeth are
observed, placed in a longitudinal row. ‘The dactylus carries likewise a small conical
tooth, 1:5 mm. beyond that of the immobile finger, and between it and the articulation
five or six smaller teeth. The inner and the inferior sides of merus, carpus, and palm are
covered with slender spinules, those on each side are somewhat larger and arranged here
more or less distinctly in two longitudinal rows, four in all therefore, but the upper and
outer sides are closely beset with much smaller spinules.

The following legs are stout, a little hairy, and thickly covered, except on the lateral
parts, with minute spinules; the legs are, however, partly broken and incomplete. The
propodites of the third pair reach almost to the end of the scaphocerites; the carpus of
the following pair reaches to the cornea of the eyes, that of the fifth pair finally extends
to the anterior margin of the cephalothorax. The meropodites of the legs of the third
pair are 17 mm. long and 2°5 mm. thick in the middle.

The cephalothorax of the female from the River Prah is smooth. The upper margin
of the rostrum (fig. 15) is straight till a little beyond the middle, then it curves very
slightly upwards, reaching to the end of the scales. It is armed above with 9 basal
teeth, the first two are on the carapace, the third just above its anterior margin, and the
ninth tooth reaches almost to the distal end of the peduncles of the inner antenne;
finally with two small apical teeth close to the tip, the posterior of which is separated by
a smooth interspace, which is 5 mm. long, from the foremost tooth of the basal portion.
The first and the second teeth are twice as distant from each other as the second is from
the third, and the three or four foremost teeth of the basal portion increase very slightly

43*

302 DR. J. G. DE MAN ON SPECIES OF PALZMON

inlength. The convex inferior margin carries 5 teeth, the first of which is situated below
the seventh of the upper margin, the fifth just opposite the middle of the smooth
interspace. The two portions into which the rostrum is divided by the lateral crest
are equally high.

The outer footjaws project beyond the peduncles of the lower antenne by a third of
their terminal joint. The legs of the first pair extend by half their hand beyond the
antennal scales ; the slender carpus, 11°5 mm. long, is two and a half times as long as the
chela, which measures 4°5 mm., and the fingers of which are somewhat longer than the
palm. Only the right leg of the second pair is still present (fig. 16). It is considerably
feebler than that of the male and reaches with the distal third of the carpus beyond the
antennal scales. The carpus, 2 mm. thick at the distal end, is 1°25 mm. thick at the
extremity. The fingers shut together and are thickly covered with felt; the palm is
1:9 mm broad when looked at from above, but only 1'7 mm. in the middle when the
lateral side is measured, so that it is not exactly cylindrical. The spinules with which
the joints are covered are comparatively much smaller than in the adult male, but
their arrangement is the same. The hand is a little longer than the carpus, and the
carpus is much longer than the merus. The following three legs are somewhat less
stout than those of the adult male and their joints are almost smooth ; those of the fifth
pair reach to the end of the scaphocerites ; the two other pairs, which have about the
same length, are a little shorter. The meropodites of the third pair are 10°5 mm. long
and, in the middle, 1°33 mm. thick.

The situation of Boutry, where the type specimens of Herklots were obtained, is in
the immediate vicinity of the River Prah, for Boutry is situated close to Dixcove, and
that place is not far distant from the mouth of the river.

The cephalothorax of the only male from the Congo Coast is almost smooth. The
slender rostrum (Pl. 19. fig. 17), which projects about 1 mm. beyond the antennal seales,
is not straight, for the distal half curves somewhat upward and the upper margin is very
slightly arcuate above the eyes. The basal portion carries 9 teeth, three of which are on
the cephalothorax, and the ninth tooth reaches to the distal end of the penultimate joint
of the antennular peduncles; then follows a smooth interspace, 5 mm. long, finally two
smaller apical teeth close to the extremity. The first and the second teeth are a little
more distant from each other than the following. The inferior margin is provided with
G6 teeth, the first immediately before the eyes, the following increase in length, and the
sixth is situated below the penultimate apical tooth of the upper margin. The rostrum
resembles that of the female from the River Prah, but is more strongly turned upward
and has a more slender form. The first legs overreach the scales only by the length of
their fingers; the carpus is 7°5 mm., the hand 3°66 mm. long.

Though this specimen is but little smaller than the male from Liberia, the legs of the
second pair are considerably shorter, and have the same form and length as in the
females. This fact, however, is often observed in the males of this genus. As regards
the measurements of the right leg, I refer to the Table; they nearly agree with those of
the female from the River Prah. The whole leg is somewhat shorter than the animal
and fully agrees in its other characters with that female. I will only add that the merus

ee ee eee ee

FROM THE INDO-PACIFIC AND WEST AFRICA. 303

reaches as far forward as the peduncle of the upper antenne, and that the palm is 2°1 mm.
broad looked at from above and 1:°9 mm. in the middle measured on the outer side.
The fingers are already covered with the characteristic woolly pubescence. The left leg
is only 30 mm. long and almost quite smooth. The following legs are also smooth;
those of the fifth pair reach to the end of the antennal scales.

Five of the eight females from the Congo Coast are provided with eggs; the largest of
these five is 80 mm., the shortest 58 mm. long from the tip of the rostrum to the end of
the telson. The cephalothorax is smooth. The rostrum (fig. 18) agrees, in its general
characters and outer appearance, with that of the female from the River Prah and in all
it curves more or less upward at the distal end; the teeth-formule are given in the
Table. In all eight females 8, 9, or 10 teeth of the basal portion are separated by a
smooth interspace from one or two, rarely three apical teeth, and these apical teeth are
either closely situated near each other or a little distant; two or three teeth are on the
carapace and the lower margin is armed with 4, 5, or 6 teeth.

In the larger specimens the fingers of the first pair project beyond the antennal scales,
in the smallest one these legs are just as long as the scaphocerites. The legs of the
second pair agree with those of the female from the River Prah. They are generally
equal, half as long or little more than half as long as the animal. In the largest female,
which is 80 mm. long, these legs project a fourth of their carpus beyond the antennal
scales ; inanother (No. 7), which has almost the same size, the second legs (PI. 18. fig. 19)
are just as long as the scaphocerites. In the former specimen the carpus is 12 mm.
long, and 15 mm. thick at the distal end, in the other these numbers are 10 mm. and
1:6 mm.; so that the carpus of No. 5 appears somewhat slenderer than in the other.
In two females the carpus is a little shorter than the hands, but in the others these
joints have about the same length; the fingers, which in all are covered with a felt-
like pubescence, though the tips are generally devoid of it, are usually a little shorter

‘than the palm. The fingers carry about the same teeth as in the male, but they are

fewer in number and smaller. So, e. g., in the female (No. 6) the dactylus (figs. 20-22)
carries an obtuse tooth at one-third of its length from the articulation and behind it four
other teeth, which are also obtuse, a little smaller and of unequal size. The immobile
finger is also armed with an obtuse tooth at the end of its cutting-edge, and this edge is
a little longer than on the other finger; between this tooth and the articulation are
seen six smaller teeth, also of unequal size. Looked at from above the palm presents
the same breadth along its whole length, but viewed from the lateral side, palm and
fingers slightly decrease in height or thickness and the fingers taper gradually towards
their tip. In the largest female (No. 5) the upper surface of the palm is 15 mm. broad,
but looked at from the outer side the palm appears 1:4 mm. thick at its proximal end,
1:3 in the middle, and 11 at the base of the fingers. Fine short hairs are distributed
on the joints, and as regards the distribution and size of the small spinules on their inner
and lower surface, these Congo specimens fully agree with the female from the River
Prah. So, e. g., the spinules on the inner margin of the carpus of the leg figured
(fig. 20) are rather slender and 0°25-0°27 mm. long (fig. 21); similar spinules occur
on the inner margin of the palm; those on the inner margin of the merus are, however,

304 DR. J. G. DE MAN ON SPECIES OF PALZAMON

a little shorter and thicker. The spinules on the outer margin of the joints are much
thinner and feebler. The three posterior legs agree also with those of that female. In
the largest female (No. 5) the legs of the fifth pair exceed the antennal scales by the
length of their dactylopodites; the two preceding pairs are a little shorter, extending
only to the end of the scales. Their joints have a smooth, shining surface.

The numerous specimens from Catumbella, 28 of which have been measured, are
almost all very young, and the few of somewhat larger size are, however, younger than
the Congo individuals described above. The largest specimen is a male, measuring
60 mm. (No. 11 of the Table). The cephalothorax is apparently still smooth. The
rostrum (Pl. 19. fig. 23), which extends to the end of the scales and curves very slightly
upwards at the extremity, makes a rare exception to nearly all the other individuals: there
are, for instance, no apical teeth on the upper margin; the 10 teeth of the upper margin,
the third of which is situated just in front of the anterior margin of the cepbalothorax,
reach to the distal end of the antennulary peduncles; and the space between the tenth
tooth and the apex, 2°75 mm. long, is slightly convex, without apical teeth. The first
legs overreach the antennal scales by half the palm; the carpus is 7°5 mm., the hand
35mm.long. The left leg of the second pair (the right is wanting) measures two-thirds
the length of the animal, and projects almost with two-thirds of the carpus beyond the
antennal seales. Though this specimen is but little shorter than the Congo male
described above, this leg, however, is much shorter, less stout, and the surface of the
joints is almost smooth, the spinules being less distinctly developed ; the measurements,
however, present the same proportions, and the fingers are already covered with the
woolly pubescence characteristic of this species. The other legs agree with those of that
male, the fifth pair reaching to the end of the scales.

In the other specimens the rostrum, which is as long as the scales or projects very little
(0°5-1°5 mm.) beyond them, is yenerally more or less turned upwards at the distal end
(fig. 24); it is usually armed above with 8, 9, or 10 teeth on the basal portion and with
one or two apical teeth, separated by a smooth interspace from the preceding; and when
two apical teeth are present, then they are situated close together near the tip of the
rostrum or at some distance from one another. The lower margin is armed with 4, 5, or
6 teeth. So, e. g., in a male specimen that is 837 mm. long, the rostrum, the formula of
which is soe and the slender distal end of which curves a little upward, projects slightly
beyond the scales for half a millimetre, as in many other individuals. ‘The first legs
reach to the end of the antennal scales, and those of the second pair, that are just half
as long as the animal, overreach the antennal scales by the whole length of their hand.
In another male of exactly the same length the first legs are just as long as in the
preceding specimen, but those of the second pair project a fourth of their carpus
beyond the scaphocerites ; the legs of the third pair reach as far forward as the terminal
spine of the lateral margin of the scales.

In the largest female (No. 22), which is 58 mm. long, the first legs project only by
half the length of their fingers beyond the distal end of the antennal scales; the chele,
in which the fingers are distinctly longer than the palm, are just half as long as the
slender carpus. The legs of the second pair (Pl. 18. fig. 25 and PI. 19. fig. 26) are little

FROM THE INDO-PACIFIC AND WEST AFRICA. 305

more than half as long as the animal, and project beyond the distal end of the scales by a
third part of their carpus. ‘The chelz are also slender and somewhat shorter than the
preceding joint; the fingers, already slightly pubescent, area little shorter than the palm.
The palm is 1 mm. broad above, and its breadth remains the same along its whole length,
but it narrows slightly towards the base of the fingers when the lateral side is looked at:
the palm is 0°75 mm. high in the middle, so that it is but slightly compressed, namely in
the proportion of 4:3. The second legs are nearly smooth and the fingers are marked
with small red spots. The third and fourth legs reach almost to the end of the scales.

In younger females, which are 35 mm. long, the first legs reach as far forward as the
lateral terminal spine of the scales; the second legs, half as long as the animal, project
with only their chele beyond the distal end of the scales. This is also the case in still
younger individuals, though sometimes the wrist projects a little beyond the scales.
Now it should be remarked that the carpus of the second legs has generally in these
younger specimens from Catumbella a somewhat slenderer form than in the older
individuals from the Congo Coast, as will be found by comparing in the Table the
proportion between the length of this joint and its thickness at the distal end in different
individuals.

As regards the teeth of the fingers in these Catumbella individuals, the following
may be remarked :—Each finger is provided with a cutting-edge, at the end of which
stands an obtuse tooth, and between this tooth and the articulation are seen a few other
teeth. So, e.g.,in a male specimen that is 835 mm. long, the dactylus carries two rounded
teeth between the tooth at the end of the cutting-edge and the articulation, but ‘on the
immobile finger there are here no teeth at all.

Of another male (No. 20), which is 29 mm. long, one of the equal second legs is
represented in fig. 27, and the teeth of its fingers in fig. 28 of Pl. 19. The cutting-edge
of the dactylus extends along two-thirds of its length, that of the immobile finger
is a little longer. The obtuse tooth at the end of the cutting-edge of the dactylus is a
little smaller than that of the other finger. Between that tooth and the articulation
on the dactylus two obtuse teeth of equal size are seen, as large as the distal tooth; on
the immobile finger, however, three much lower teeth, also of equal size, but considerably
smaller than the tooth at the end of the cutting-edge.

In the largest female (No. 22), which has a length of 58 mm., the dactylus is armed
with four, the immobile finger with two teeth between the obtuse tooth at the end of the
cutting-edge and the articulation ; all are smaller than the distal teeth, and those of the
dactylus are of somewhat unequal size (Pl. 19. fig. 26). In another female, which is 33 mm.
long, there are but two small teeth on each finger between the tooth at the end of the
cutting-edge and the articulation ; in another, 30 mm. long, three, and in this individual
the middle of the three teeth on the dactylus appears double. In both females the
cutting-edge of the dactylus is a little shorter than that of the immobile finger. Ina
fourth young female, 25 mm. long (No. 38), the teeth of the fingers are shown in PI. 19.
fig. 29. The obtuse tooth at the end of the cutting-edge is situated on the dactylus at
one-third of its length from the articulation, and behind it there exists still only one,
situated about twice as far from the articulation as from the distal tooth and of about

306 DR. J. G. DE MAN ON SPECIES OF PALZMON

the same size. Just in the middle between both teeth one sees the distal tooth of the
immobile finger, which is as large as that of the dactylus; posterior to it this finger carries
still two low teeth that are a little smaller.

Palemon (Eupalemon) sundaicus, Heller, from the Java Sea, is perhaps the most
closely allied form. One of the four female specimens which I described a few years
ago is lying before me (vide Zool. Jahrb., Syst. ix. 1897, p. 779, Taf. 37. fig. 71).
It is 75 mm. long, and its size is thus about the same as that of the macrobrachion
specimens from the Congo. The rostrum has about the same form and characters, but
it arises almost in the middle of the cephalothorax; the distance of the first tooth from
the posterior margin of the cephalothorax is indeed ¢wice as great as its distance from
the anterior margin; in P. macrobrachion, however, five to eight times as far, so that
in the African species this tooth is situated much nearer to the orbital margin. In
P. sundaicus the two spines on the sides of the cephalothorax are situated about in the
same horizontal line, but the hepatic spine of P. macrobrachion is situated much lower.
The legs of the second pair much resemble each other in both species; but the fingers
are smooth, not covered with the woolly pubescence characteristic of the West-African
form, and their toothing is different, the dactylus being armed with only two teeth, the
immobile finger with only one. The second legs are, moreover, otherwise coloured. The
other legs present also a great conformity in both species. According to Max Weber,
P. sundaicus inhabits the coast of Natal (Max Weber, Zool. Jahrb., Syst. x. 1897,
p. 165).

Palemon (Eupalemon) macrobrachion, Herklots, inhabits the rivers of West Africa
from Liberia to Benguella.

PaL#Mon (EvpALZMON) Foat, Coutiére. (Plate 19. figs. 30-37.)
Palemon (Eupalemon) Foai, nov. sp., Coutitre, Bull. Muséum d’ Hist. nat. Paris, 1902, No. 7, p. 517.

One single male, collected in the River Kribi, 25 miles from the coast of Cameroon,
West Africa. This specimen belongs to the British Museum.

This species, certainly different from P. (Hupalemon) macrobrachion, Herklots, is
apparently related to Palemon paucidens, Hilgd., a species discovered at Adeli, near
Bismarckburg, Togo Country, described by Hilgendorf in Sitzungsber. Gesellschaft
Naturf. Freunde, Berlin, 1893, No. 5, p. 155. As Coutiére does not compare his species
with P. paucidens, it will be done in the course of this description.

Our specimen is 8 cm. long from the tip of the rostrum to the end of the telson;
Coutiére’s largest specimen was 70°5 mm. long. The cephalothorax is smooth. The
rostrum is stout and projects straight forward exactly to the end of the antennal scales ;
it is armed above with six, below with two teeth. ‘The rostrum arises on the anterior
half of the cephalothorax ; the distance of the first tooth from the anterior margin of the
cephalothorax is just one-fourth the whole length of the upper surface from the
posterior to the anterior margin. The upper margin is very slightly convex above the
eyes, but in P. paucidens the rostrum, which is armed above with 7-8, below with
1-2 teeth, and scarcely extends beyond the distal end of the antennulary peduncles, has
been described as high, foliaceous (‘‘ blattférmig’’), and short. Just as in the species

FROM THE INDO-PACIFIC AND WEST AFRICA. 307

from Togo Country, however, only one tooth is situated posterior to the orbital margin,
the second standing already before it. The second tooth is the longest of all, for the
distance of its tip from that of the first, 5-5 mm. long, measures just one-third the
distance of the tip of the first tooth from the posterior margin of the cephalothorax.
The two following teeth are shorter than the second, 3 mm. long, and the fifth and the sixth,
each a little more than 2°5 mm. long, are again a little shorter than the two preceding ;
the sixth is 2 mm. distant from the tip of the rostrum. The teeth of the upper margin
areall rather small. In the middle of the free portion of the rostrum, immediately posterior
to the point of the fourth tooth, that part which is situated above the lateral crest is but
very little higher than that below it; the rostrum is not at all high. In the middle of
the rostrum the lower margin appears slightly convex; the two teeth are still smaller
than those of the upper margin and are situated, about as in P. paucidens, on the
anterior half of the inferior margin. The first is situated a little beyond the middle of
the fifth tooth of the upper margin, the second just behind the middle of the sixth; the
distance of the second tooth from the tip of the rostrum is twice as long as the length of
that tooth and twice as long as the distance of the sixth tooth of the upper margin from
the tip. The hepatic spine is situated below and posterior to the antennal one, exactly
below the first tooth of the rostrum. Our specimen also agrees with Hilgendorf’s
description in this character, that the two pairs of spinules on the upper surface of the
telson are situated more backward than in P. macrobrachion, as is represented in my figures.
The telson (figs. 31 & 32), which is 11 mm. long and 4 mm. broad at base, ends in an acute
median tooth, with two spines on each side as usual, the inner of which extends beyond
the median tooth, about as in P. macrobrachion. The anterior pair of spinules on the
upper surface is situated 6 mm. from the anterior end of the telson, émmediately behind
the middle ; it is 2°25 mm. distant from the posterior pair, and the latter 2°75 mm. from
the tip of the median tooth at the end of the segment. In P. macrobrachion, however,
the first pair of spinules is situated just defore the middle of the telson. The eye-
peduncles are a little longer than broad. The short flagellum of the upper antenne,
which is distinctly serrated and 12°5 mm. long, is united only for a sixth of its length
with the outer one, and is scarcely longer than their peduncle.

The outer footjaws reach two-thirds of their terminal joint beyond the distal end
of the peduncles of the outer antenne. ‘Ihe legs of the first pair (fig. 33) project
a fourth of their wrist beyond the distal end of the antennal scales. The slender
carpus is 11°5 mm. long and 0°9 mm. thick at the distal end; the hands are half as long
as the wrist and the fingers are just as long as the palm.

Unfortunately, one of the second legs is lost; the other, 75 mm. long, has almost the
same length as the body. It is a rather feeble leg in proportion to the size of this
specimen; but I have shown in the description of P. macrobrachion that the length
and the size of the second legs differ sometimes very much in different specimens :
compare, e. g., in the Table on p. 321 the male (No. 1) from Liberia and that from the
Congo (No. 4). The second leg (fig. 34) of our specimen from the River Kribi
much resembles that of P. macrobrachion as regards the proportion of the length of
the joints, but it differs at first sight in the absence of the woolly felt with which the

SECOND SERIES.—ZOOLOGY, VOL. IX. 4s

308 DR. J. G. DE MAN ON SPECIES OF PALA MON

fingers are covered in the last-named species. The merus is 14°5 mm. long, the carpus
18 mm.,the hand 25°25 mm., viz. the palm 16 mm., the fingers 9°25 mm. Just as in
Palemon macrobrachion, the carpus is distinctly longer than the merus and shorter than the
whole hand; it is also a little longer than the palm and almost twice as long as the fingers.
The latter measure little more than one-third the length of the whole hand. The slender
merus thickens gradually and regularly towards the distal end, and, being here 2'4: mm.
broad, this jot appears just six times as long as the thickness at the distal end. The
merus of P. macrobrachion thickens not so regularly towards the distal extremity. The
slender carpus widens likewise regularly towards its distal extremity, and, being here
2:5 mm. broad, this joint appears just seven times as long as thick at the distal end. The
hand closely resembles that of the male of P. macrobrachion from Liberia described
p. 308. Viewed from above, the palm appears 2°2 mm. broad, and presents the same
breadth along its whole length; in a lateral view, just as in P. macrobrachion,
the breadth slightly decreases towards the base of the fingers, and in the middle of its
length the palm is 1:9 mm. thick. The palm is not quite cylindrical, but very nearly
so, and it is a little less broad than the distal end of the carpus. Just as in the male
from Liberia, the fingers are a little curved inwards, so that the inner margin of the hand
is slightly concave. The fingers shut close together. The dactylus appears at its base a
little broader or thicker than the other finger, and tapers regularly towards the tip; it
is armed with a small conical tooth just at a third of its length from the articulation,
and between it and the latter there are still two other similar teeth. The immobile
finger likewise carries a tooth, a little nearer to the articulation than in the other finger,
and also two other teeth between the articulation and the distal tooth. Between the
distal tooth and the tip a sharp cutting-edge is seen on each finger. The joints are
covered with small spinules which on the upper outer side are smaller and less numerous
than elsewhere; on the outer border of the palm they are arranged in a longitudinal
row, but they are here a little smaller than on the inner margin. These spinules
occur also on the fingers. The joints of this leg are very slightly hairy; the hairs,
however, are very short and fine, and the leg has a reddish colour. In the larger leg of
the second pair of P. pawcidens from Togo Country, the merus was 11:3 mm. long, the
carpus 11°5 mm., the palm 15°5 mm., and the fingers 9°6 mm.; merus and carpus were
thus shorter than in our specimen, especially the carpus, which was scarcely longer than
the preceding joint.

In the species described by Hilgendorf the fingers were each armed, beside with the
basal teeth, with seven spines (“‘ Dornen”), standing on both sides of the cutting-edge ;
these spines are wholly wanting in our species from the River Kribi.

The three posterior legs of our specimen resemble those of P. macrobrachion;
they are, however, a little stouter. Those of the third pair project, by little more than
their terminal joints, beyond the scaphocerites, those of the two other pairs only by
half their dactylopodites. The meropodites of the fifth pair are 11°5 mm. long and
1:04 thick, measured on the outer side ; the carpopodites are 6°5 mm. long and 1:5 mm.
thick at the distal end; the propodites are exactly as long as the meropodites and
0°7 mm. broad in the middle; the terminal joints are 3'1 mm. long.

FROM THE INDO-PACIFIC AND WEST AFRICA. 309

In a female specimen of Palemon macrobrachion from Congo of about the same size,
being 78 mm. long, the legs of the fifth pair have the following measurements :—The
meropodites are 9°5 mm. long and 0°86 mm. thick in the middle; the carpus is 6 mm.
long and 0°82 mm. thick at the distal end ; the propodites are 11°5 mm. long and 0°5 mm.
broad in the middle, the terminal joints finally are 3°2 mm. long. In P. macrobrachion
the propodites of these legs are thus distinctly longer than the meropodites and, like the
carpopodites, a little more slender than in the male from the River Kribi. The three
posterior legs of the male from the River Kribi are covered with scattered microscopical

-spinules, especially on their upper margin. WHilgendorf has not described the three
posterior legs of P. paucidens.

[For note received since this paper was in type, see p. 327.—Szc.L.8. ]

Pat#Mon (MACROBRACHIUM) JAMAICENSIS (Herbst), var. VOLLENHOVENII, Herklots.
(Plates 19. and 20, figs. 38-53.)

Palemon Vollenhovenii, Herklots, in Tijdschrift voor Entomologie, i. 1858, p. 96; de Man, in Notes
from the Leyden Museum, i. 1879, p. 178.

Palemon jamaicensis (? Herbst), Benedict, in Proc. U.S. National Museum, vol. xvi. 1893, p. 540.

Palemon jamaicensis, Herbst, var. Vollenhovenii, Aurivillius, in Bihang till K. Sv. Vet.-Akad. Handl.,
Ba. xxiv. Afd. iv. No. 1, 1898, p. 16, Taf. 2. figs. 1-5.

The following collection is lying before me * :—

One adult male from Liberia. (Leyden Museum.)

Three nearly adult males and two young females, from the River Prah, Ashantee.
(British Museum.)

190 specimens of different size, collected in the river at Catumbella, near Benguella.
(Private collection.)

Some time ago Dr. Ortmann, who was enabled to compare adult males of this species
from Cameroon with American specimens of P. jamaicensis (Herbst), finally concluded
‘that the African form was quite identical with the American type (Ortmann, ‘Os
Camardes da agua doce da America do Sul, 8. Paulo, 1897, p. 209). Aurivillius,
however, who compared ten specimens from Cameroon with one adult individual from
Central America, is inclined to regard the African form as a variety of P. jamaicensis ;
and I like to follow him provisionally in this opinion, because no American specimens
are at my disposal. Nevertheless, I suppose that fresh descriptions of the African form
will still be welcome.

The largest specimen of all lying before me is the male from Liberia. The rostrum
(fig. 38), closely resembling that which has been figured by Aurivillius (J. ¢. fig. 1), reaches
to the end of the peduncles of the upper antenne, and is armed above with 16 teeth, a
number already observed by that Swedish naturalist ; they are equidistant, but the first

* The measurements of the body and of the second legs are indicated in the Table (p. 322), also the formule of
the rostrum. he joints of the second legs are measured on their upper surface, the merus, ¢. g., up to the distal end
of the upper margin of the ischium. Nos. 4 and 5 are the Congo specimens described in my paper of 1879.

44*

310 DR. J. G. DE MAN ON SPECIES OF PALAMON

is slightly more distant from the second than the others, and the distance of the foremost
tooth from the tip is a little longer than the interspace between the preceding teeth.
Five teeth are on the cephalothorax. The rostrum is slightly convex above the eyes and
its distal half is directed downward.

In young individuals of this species (fig. 39) the posterior extremity of the telson
ends in an acute triangular point, and of the two spines on each side the inner larger
one reaches beyond the acute extremity; the anterior pair of spinules is situated imme-
diately behind the middle. In older individuals, however, the telson extremity is
usually more or less worn off, and it is therefore quite erroneous to describe it as
largely rounded, as H. Milne-Edwards and Ortmann have done (Ortmann, in Zool.
Jahrb., Syst. v. 1891, p. 729, Taf. 47. fig. 7). In the male from Liberia the
extremity of the telson appears still triangular, but the acute point is already
worn off.

The outer footjaws of this specimen are just as long as the peduncles of the upper
antenne, overreaching those of the lower antennze by their terminal joint.

The legs of the first pair extend for half their carpus beyond the scaphocerites ; and
the hands, which are 11°5 mm. long, and in which the fingers appear a little shorter
than the palm, are slightly more than half as long as the wrist, which measures 20°5 mm.

The larger leg (fig. 40) of the second pair is situated on the left side. The merus
reaches almost to the end of the antennal scales. The carpus, which is 10 mm. broad at
its distal extremity, is not shorter than the merus when the upper surface is measured ;
and an individual of about the same size, in which both joints were also equally long,
has been described by Aurivillius from Cameroon (J. c. p. 17). The hand, a little more
than half as long as the body, is somewhat more than three times as long as the carpus,
and the palm appears almost twice as long as this joint. The palm, somewhat more than
three times as long as broad, is a little broader than the carpus and a little less thick
than broad, viz. 11:5 mm. thick and 12°5 mm. broad. The fingers, a sixth shorter than
the palm, are armed each with a conical tooth, that on the immobile finger is some-
what larger and situated somewhat closer to the articulation than on the dactylus ;
between this tooth and the articulation the dactylus carries in addition three small obtuse
teeth, the immobile finger only one. The fingers are slightly curved inward, and leave
therefore a narrow interspace between them when closed. The whole leg is closely beset
with small acute denticles, which are larger and less numerous on the inner side. The
right leg fully agrees with the other, but the carpus is a little shorter than the merus,
the fingers are almost as long as the palm, and the teeth with which they are armed
are considerably smaller.

The legs of the third pair reach as far as the scaphocerites, those of the fourth as far
as the external mayillipedes, and the legs of the last pair extend to the end of the
peduncles of the outer antenne. The meropodites of the fifth pair are 16 mm. long and
2-25 mm. thick, measured on their lateral side ; the carpopodites 9 mm. long and 2°25 mm.
thick at their distal end; the prepodites 15 mm. long and 14 mm. thick in the middle,
measured on their outer side.

The three males from the River Prah are about the same size as that from Liberia ;

FROM THE INDO-PACIFIC AND WEST AFRICA. 311

in all the rostrum is broken off, and this is also the case with the legs! A larger leg of
the second pair, lying loose in the bottle, belongs very likely to the male No. 2 that still
bears its shorter left leg. This loose leg (No. 2, ”) closely resembles the larger leg of the
male from Liberia, but the fingers are a little longer in proportion to the palm, and the
large tooth of each finger is somewhat more distant from the articulation. The palm,
11 mm. broad, is 9°5 mm. thick in the middle. The other, shorter leg fully agrees with
the shorter leg of the male from Liberia, which in this specimen is borne on the
right side.

In the female No. 3, from the River Prah, the rostrum, slightly arcuate above the
eyes, reaches almost to the end of the antennal scales, and resembles that of the male
from Liberia. The last two joints of the outer footjaws, the right leg of the second
pair, and the three posterior legs are wanting. The merus of the left leg reaches to the
middle of the antennal scales ; the carpus, 9 mm. long, is 3°25 mm. broad at the distal end.
The inner margin of the chela is concave, the palm is 3°75 mm. broad and 3 mm. thick
in the middle, and the fingers, that shut close together, are only armed with minute teeth
near the articulation, and are as long as the palm. ‘The other female has lost its legs.

Such a large and fine collection as that of the 190 specimens from the river
at Catumbella has most probably never been at the disposal of any naturalist.
Unfortunately, in half of them the legs of the second pair are wanting.

The teeth-formule of the rostrum of 22 males and as many females (that are all
provided with the legs of the second pair, so that they certainly belong to this species)
are the following :-—

Males.—One specimen 2. five specimens e3 six specimens 2 ; one specimen = ; one
specimen - ; one specimen 3; one specimen o ; one specimen 7: two specimens = :
one specimen =F one specimen 2; one specimen = :

In one specimen the upper margin bears 11 teeth, in twelve 12, in three 13, in four 14,
in one 15, and in one 18. ‘The lower margin is armed in seven individuals with 3 teeth,
in eleven with 4, in two with 5, and in two with 6.

Females.—Six specimens = ; Six specimens 2; one specimen =; six specimens 2 ‘
three specimens *.

In twelve specimens 12 teeth on the upper margin, in seven 13, in three 14. The lower
margin is armed in seven females with 3, and in fifteen with 4 teeth.

In the males the number of teeth on the upper margin varies from 11 to 18, the most
Frequent number is 12; the teeth of the lower margin vary from 3 to 6, the usual number
is 4, In the females the upper teeth vary from 12 to 14, the lower from 3 to 4; the
usual number of the former is 12, of the latter 4, exactly as in the males.

In twenty-three specimens 4 teeth are on the cephalothorax, in four 5, and in two 3;
in ten specimens the fourth tooth stands just above the orbital margin, in five the
fifth.

In not a single one of the 190 specimens from Catumbella is the rostrum limited to the
length of the antennulary peduncles ; in the larger individuals (fig. 41) it generally reaches

312 DR. J. G. DE MAN ON SPECIES OF PALAMON

more or less beyond these peduncles, though still not extending to the tip of the scales ;
but in the numerous specimens of smaller size, the quite young ones, the rostrum usually
extends to the extremity of the scales, and even rarely a little beyond them. It has the
same form as in the specimens from Liberia and from the River Prah that have been
described above, and it agrees also with the figures in the paper of Aurivillius (J. ¢.
figs. 1 & 2). The upper margin is slightly convex above the eyes, and the distal half is
directed downward, but the acute tip is generally somewhat turned upward ; the rostrum
is slender, tapers gradually to a point, and the part situated above the lateral crest is
scarcely higher than that below it.

In the largest male but one (No. 7) the outer footjaws project with their terminal
joint beyond the peduncles of the lower antennee, just as in the adult male from Liberia;
in quite young individuals they overreach these peduncles about three-fourths of the
terminal joint. The legs of the first pair project beyond the antennal scales for half
their wrist; this joint is 14 mm. long; the hands are 7°34 mm. long, and the fingers
are very slightly shorter than the palm. Only the left leg (fig. 42) of the second pair
is present, and this leg is apparently the larger one. The merus reaches as far forward
as the antennal scales. When this leg is compared with the larger leg of the male from
Liberia, or with the described larger leg of the specimen from the River Prah, the joints
appear less stout, being somewhat less broad or thick in proportion to their length. The
merus and the carpus are equally long; the former is 5-25 mm. thick, and the diameter
of the carpus at its distal extremity, measured above, is 6 mm. broad. The palm is nearly
as long as that of the larger leg from the River Prah, viz. 35 mm., but only 8 mm.
broad instead of 11 mm.; the palm is 6 mm. thick in the middle. The spinules on the
inner side of the palm are somewhat larger than in the individuals from Liberia and the
River Prah. The fingers are a little shorter in proportion to the palm, and their tips are
less abruptly curved inward than in the specimens from Upper Guinea; the teeth are
typically developed, so that the fingers leave an interspace between them. The legs of
the third pair project for half their terminal joints beyond the end of the scaphocerites,
those of the fourth reach to the end of the peduncles of the internal antenne, and the
last pair is but very little shorter. These legs are thus a little shorter than in the adult
male from Liberia, but this may depend upon age; for the rest, they agree in other
characters.

In the male No. 8 the rostrum is slightly upturned at its distal end and reaches to
the end of the antennal scales. The right leg is the larger (fig. 43), but it is consider-
ably smaller than in the specimen just described ; the fingers are but little shorter than
the palm and shut close together, as the teeth are still quite small. The dactylus bears
two small teeth between the articulation and the distal tooth; likewise the immobile
finger. The second legs appear quite as slender as in the preceding individual. The
merus of the right leg is 35 mm. broad, the carpus 4°25 at its distal end, and the palm
is 45 mm. thick in the middle.

In the largest male (No. 6) the legsof the second pair are rather feeble and short,
when compared with the two males already described. In both legs (fig. 45) the carpus
reaches to the distal end of the scales, and the right leg is but 3 mm. longer than the

FROM THE INDO-PACIFIC AND WEST AFRICA. 313

left. The palm of the right leg is 3-5 mm. broad and 8 mm. thick in the middle ; the carpus,
which has a length of 10°5 mm., is 3°6 mm. thick at the distal end, so that in this specimen
the palm appears no¢ broader than the carpus. The dactylus carries five small teeth of
equal size between the tooth at the end of the cutting-edge and the articulation ; on the
index, however, only one obtuse tooth is seen behind that at the extremity of the cutting-
edge. All these teeth are very small and of equal size, so that the fingers shut close
together. This specimen is evidently an individual variety, in the same way as
Palemon vagus, Heller, is a variety of the well-known Indian P. lar.

In the male No. 12 the rostrum fully resembles that of the male No. 8 described
above ; it reaches to the end of the scales, but there are only 11 teeth on the upper
and 3 on the lower margin; three teeth stand on the cephalothorax. The outer foot-
jaws project for four-fifths of their terminal joint beyond the peduncles of the lower
antenne. The carpus of the first legs is 7 mm. long, and projects for two-fifths of its
length beyond the distal end of the scales; the hand, 3°75 mm. long, is little more than
half as long as the carpus, and fingers and palm have the same length. The left leg of
the second pair is the larger, projecting for a third of the carpus beyond the scapho-
cerites. The merus is distinctly longer than the carpus and 1:4 mm. thick at the distal
end; the carpus is 1:8 mm. thick at its distal extremity. The palm is 2:25 mm. broad
and 1:8 mm. thick in the middle.

In the quite young male (No. 16) the rostrum (Pl. 20. fig. 46) reaches, as in most youne
individuals, to the end of the scales, and four teeth stand on the cephalothorax. The
outer footjaws project for three-fifths of their terminal jot beyond the peduncles of
the external antennz. The carpus of the legs of the first pair, 5°3 mm, long, extends
for one-fifth of its length beyond the scaphocerites ; and the hand, 3 mm. long, is again
a little more than half as long as the wrist. The larger leg of the second pair is the left
one (Pl. 19. fig. 47); the carpus extends for a fourth of its length beyond the scales ;
the merus is ‘8 mm., the carpus 1:16 mm. thick, at their distal extremities. The palm is
1:42 mm. broad and 1°31 mm. thick in the middle. The immobile finger (Pl. 20. fig. 48)
bears one, the dactylus two obtuse teeth between the conical tooth at the end of the
cutting-edge and the articulation. The legs of the third pair reach to the end of the
scales, the two following are somewhat shorter.

In the largest female but one from Catumbella, 84mm. long, which bears no eggs, the
rostrum (fig. 49) reaches midway between the distal end of the antennulary peduncles
and that of the scales. It fully resembles that of the preceding specimens from the same
locality : the fifth tooth stands just above the orbital margin, the upper margin is slightly
convex above the eyes and the tip is somewhat turned upwards; the third tooth of the
lower margin reaches to the middle of the terminal joint of the antennulary peduncles.
The outer footjaws extend almost for the whole terminal joint beyond the extremity
of the peduncles of the outer antenne. The carpus of the first legs is 13 mm. long, and
projects for half its length beyond the antennal scales; the hand is 7 mm. long. The
left leg (fig. 50) of the second pair is considerably larger than the right (fig. 51). The
merus is 3 mm., the carpus 3:4 mm. thick, at their distal ends, and the latter projects
for half its length beyond the end of the scales; the palm, slightly longer than the

314 DR. J.G. DE MAN ON SPECIES OF PALEMON

fingers, is 455 mm. broad and 3°5 mm. thick in the middle. The fingers shut close
together; the dactylus is armed with two small teeth, the immobile finger with one
besides the tooth at the extremity of the cutting-edge.

In the young female No. 28 finally the rostrum reaches to the end of the scapho-
cerites and agrees with the other specimens. The outer footjaws project for two-thirds
of their terminal joint beyond the lower peduncles, and the legs of the first pair for
a fourth of the carpus beyond the distal end of the scales. The larger leg of the
second pair is the left (fig. 52), measuring just two-thirds of the body. The merus is
1 mm., the carpus 1°33 mm. broad, at their distal ends, and the latter projects for a
third of its length beyond the antennal scales. The palm is 152mm. broad and 1:27 mm.
thick in the middle ; as usual, the dactylus (fig. 58) is provided with two, the index with
one small tooth behind the distal tooth at the end of the cutting-edge. The legs of the
third pair reach to the extremity of the scaphocerites.

The larger leg of the second pair of the large male (No. 7) from Catumbella has a pale
yellowish-green colour; the carpus appears on the inner, the palm on its outer side dark
green, and the fingers are also dark green to their tips, but they show a yellowish-red
tinge at their proximal ends and at their articulation. In the other specimens these legs
are of a pale flesh-colour, reaching to the base of the fingers; the latter are dark bluish
coloured, often with pale tips.

It is, indeed, a pity that in Liberia and in the River Prah no younger specimens have
been collected or at Catumbella individuals of larger size, for then it would have been
possible to decide whether the more slender appearance of the second legs of the Angola
specimens is caused by their younger age or not ; in the latter case the Angola form
would be a distinct, new variety of Paleemon jamaicensis. Benedict has already observed
(J. c.) that specimens from the Quanza River at Cunga are ‘“a little more slender” than
others from Old Providence, West Indies. If further researches should prove this to be
really the case, I propose for this variety the name angolensis.

PALa@MON (MACROBRACHIUM) OLFERSI, Wiegm. (Plate 20. figs. 54-74.)
Palemon Olfersii, Wiegmann, in Archiv fiir Naturg. Jahrg. 2, vol. i. 1836, p. 150; Greeff, in
Sitzungsber. Gesells. zur Beférderung ges. Naturw. Marburg, 1882, No. 2, p. 30; Ortmann, in
Revista do Museu Paulista, No. ii. 1897, p. 212, Est. i. figs. 10 & 11; Aurivillius, 2. ¢. p. 23.
Palemon spinimanus, H. Milne-Edwards, Hist. Nat. Crust. ii. p. 399; von Martens, in Archiv fir
Naturg. 1869, p. 26, Taf. ii. fig. 3.
Bithynis Olfersit, Rathbun, The Brachyura and Macrura of Porto Rico, Washington, 1901, p. 124.

One adult male from the River Prah, Ashantee. (British Museum.)

60 young specimens, a third of which are males, from the river at Catumbella, near
Benguella. (Private collection.)

The adult male from the River Prah is 58 mm. long from the tip of the rostrum to
the end of the telson. The rostrum (fig. 54) reaches almost to the end of the antennulary
peduncles, is slightly directed downward, and armed above with 16, below with 5 teeth;
the upper teeth are equal, small, equidistant, and five are on the cephalothorax. The
telson tapers rather much towards the triangular pointed extremity and the anterior pair
of spinules is situated just before the middle. The outer footjaws are as long as the

FROM THE INDO-PACIFIC AND WEST AFRICA. 315

rostrum, reaching almost for their whole terminal joint beyond the peduncles of the
external antenne. A third part of the carpus of the first pair of legs, that is 85 mm.
long, extends beyond the antennal scales, and the hands are 4°5 mm. long.

The right leg of the second pair is wanting, the left has been figured (fig. 55).
Measured along its upper or outer margin, the merus appears 12 mm, long and 6 mm.
thick in the middle, the carpus 13°5 mm. long and 5:25 mm. thick in the middle; the
palm is 17 mm. long, just half as broad (85 mm.). and 58 mm. tliick in the middle, the
fingers finally are 12°5 mm. long. The upper and inner sides of the palm are covered with
felt and with scattered soft longer hairs; along the inner margin of the palm long,
slightly curved, acute spines are seen, which gradually grow smaller on the index towards
the tip. A second row of spines occurs on the upper surface of the palm near and
parallel with that on the inner margin, and a few spines stand also on the upper and on
the lower surface of the immobile finger at its base; a row of much shorter and more
numerous spines is found along the arcuate outer margin of the palm, and both upper
and lower surface of palm and fingers are beset with similar short spinules. The merus,
strongly swollen in the middle, bears acute, slightly curved spines on its inner surface,
as also the carpus, and these spines are arranged more or less distinctly ¢” three or four
longitudinal rows. Similar, though shorter, spines stand all round the carpus; on the
outer side, however, of the merus they are very short and small, so that the surface
appears here almost smooth, even under a feeble magnifying-glass. The fingers are
armed with small rounded teeth along the whole length of their inner margins, 15 or 16
on the immobile and 12 or 13 on the other finger, with short hairs on each side of them.
The meropodites of the fifth pair are 7°5 mm. long and 1:2 mm. broad in the middle on
their outer side.

The numerous specimens that have been gathered in the river at Catumbella by
Mr. P. Kamerman, and presented by him to my collection, are all young and many
are of very small size. The rostrum is rather variable as regards its form and the
number of teeth with which its margins are armed, as is proved by the following
_ descriptions and a glance at the Table of measurements. A short apical part of the rostrum

armed with one or two apical tecth, which are somewhat more distant from the preceding
than these are from each other, is constantly distinguishable; and this apical part is
generally more or less turned upward. As regards the legs of the second pair, I wish
to remark that the palm appears Jess enlarged, when compared with the adulé individual
from the River Prah, and also that the preceding joints present a more slender form than
in the adult. There can, however, be no doubt that these specimens are really the young
of P. Olfersii.

The large male (No. 3 of the Table) is 42 mm. long from the tip of the rostrum to the
end of the telson. The rostrum, slightly convex above the eyes (fig. 56), reaches almost
to the distal end of the scaphocerites and is armed above with 14 teeth, of which 5 stand
on the cephalothorax; the foremost tooth is smaller than the penultimate and situated
nearer to the apex than to that tooth. The lower margin of the rostrum is armed with
5 teeth on its distal half. The telson tapers strongly towards the posterior extremity,
which is long-pointed (fig.57); the inner of the two spines on each side reaches considerably

SECOND SERIES.—ZOOLOGY. VOL. IX. 45

316 DR. J. G. DE MAN ON SPECIES OF PALZMON

beyond the acute tip. The anterior pair of spinules on the upper surface is situated just
before the middle of the telson. The external maxillipedes extend for their terminal
joint beyond the distal end of the lower peduncles. The carpus (fig. 58) of the first legs,
5°33 mm. long, projects for little more than the hands beyond the antennal scales, and
the hands are 3mm.long. The left leg (fig. 59) of the second pair is considerably stronger
and longer than the right, and extends beyond the antennal scales for three-fourths of
the carpus. ‘This leg, 36°5 mm. long, is little shorter than the body, merus and carpus
are subequal in length, and the chela is about twice as long as the carpus. Merus and
carpus are swollen, the former in the middle, the latter nearer the distal extremity ;
both joints resemble those of the adult male from the River Prah. The hand also agrees
with it, but the palm is less enlarged in proportion to its thickness and the fingers are
still closer together. Along the inner margin of the hand is seen a@ longitudinal row
of rather strong spines which gradually decrease in length towards the tip of the fingers,
and on the upper surface of palm and fingers spinules occur similar to those on the adult
leg from the River Prah, with the soft flexible hairs and the felt also similar. The inner
margins of the fingers (fig. 60) are armed along their whole length with small obtuse teeth,
fourteen or fifteen in number; on each finger, however, the third or fourth is conical,
acute, and a little larger than the others.

The other leg (fig. 61) is much shorter and projects for nearly the whole hand beyond
the antennal scales. The chela is twice as long as the carpus. The palm is very little
broader than the carpus and distinctly shorter than the fingers, that shut close together.
Both fingers (fig. 62) bear a sharp cutting-edge, at the end of which one observes an obtuse
conical tooth. On the dactylus this edge extends along three-fourths of the finger; on
the immobile finger it is a little longer; posterior to the tooth at the end of the cutting-
edge the mobile finger is armed with three somewhat smaller, obtuse, equidistant teeth,
the index with four that are still smaller.

In another male (No. 5) of about the same size the rostrum reaches to the end of the

scaphocerites ; it is very slightly arcuate above, the distal end somewhat turned upward

16+1

and ~;- dentate. The first ten or eleven teeth stand much nearer to one another than

the following, and the sixth tooth is placed just above the orbital margin. The right leg
of the second pair is little shorter than the animal and almost once and a half as long as
the left. At a third of its length from the articulation the immobile finger of this larger
leg is armed with a conical tooth and posterior to it with four very small ones; between
the conical tooth and the tip of the finger one observes ten or eleven rounded teeth which
gradually decrease in size towards the tip, and the first of which is but little larger than
the four teeth near the articulation. The dactylus bears a similar conical tooth, larger
than the others and a little farther from the articulation than on the immobile finger ;
posterior to it there are only three, which are, however, slightly larger than the four of the
index, and between this tooth and the tip of the finger ten or eleven small rounded teeth
are seen, which again decrease in size. The fingers of the smaller chela are provided
with a cutting-edge, at the end of which stands a tooth and between it and the articulation
three or four smaller teeth.

In another young male (No. 7), which has a length of 36 mm., and that as regards its

* aa

FROM THE INDO-PACIFIC AND WEST AFRICA. 317

other characters agrees with the preceding, the right leg is lost, but the left (fig. 63) differs
from them by the palm being considerably shorter than the fingers and by the fingers being
much more curved and leaving a wide interspace between them, filled up with hairs. This
leg is the shorter one, because the still present coxopodite of the other leg is much larger ;
and in this case it is no doubt that leg which in de Saussure’s figure of Palemon Faustinus
is seen on the right side (de Saussure, ‘ Mémoires pour servir A l’Histoire Naturelle du
Mexique, ete.’ 1° livr. Crustacés, 1858, pl. iv. fig. 30). Each finger bears a cutting-edge
with a tooth at the end of it; and between this tooth and the articulation the dactylus is
armed with two somewhat smaller teeth, the immobile finger with four that are still
smaller than those of the dactylus (fig. 64).

This phenomenon, that the fingers of the smaller leg gape in some individuals, the inter-
space being then filled up with hair, but closed together in others, has already been described
for other species, e.g. for Palemon (Eupalemon) endehensis, de M., which occurs in
fresh water of the island of Flores (de Man, ‘ Decapoden des Indischen Archipels,’ 1892,
p. 465, Taf. xxvii. fig. 42).

The rostrum of the male (No. 8) that is 34 mm. long is a little (0°75 mm.) longer than
the scaphocerites, and agrees with the other specimens. The outer footjaws project for
half their terminal joint beyond the distal end of the lower peduncles, and the legs of the
first pair extend beyond the scaphocerites by the length of their fingers. The left leg
(fig. 65) of the second pair projects with the whole hand, the right leg with the fingers,
beyond the distal end of the antennal scales. The inner margin of the palm of the larger
lee is armed with ten large spines that are 0:16 mm. long, ¢. e. one-seventh of the breadth
of the palm; five or six spines occur along the inner margin of the immobile finger.
Each finger bears a sharp cutting-edge that extends along two-thirds of the dactylus
and almost along three-fourths of the immobile finger; at the end of each cutting-edge
stands a small obtuse tooth. Between this tooth and the articulation (fig. 66) there are
on each finger still four other teeth of about the same size. The fingers of the other leg
agree with them, but they are armed with only three low teeth between the tooth at the
end of the cutting-edge and the articulation.

In the youngest male (No. 9) that has been measured, which is 27 mm. long, the
rostrum is still a little longer than the dorsal surface of the cephalothorax, and projects
14 mm. beyond the distal end of the scaphocerites. The slightly convex basal part reaches
to the end of the antennulary peduncles, and the apical part is slightly turned upward ;
there are three teeth on the cephalothorax, the fourth standing immediately before the
orbital margin. The legs of the first pair reach to the end of the antennal scales. The
second legs are equal and project for almost their whole hand beyond the secaphocerites.
The carpus is just a little longer than the merus and shorter than the hand. The tooth
at the end of the cutting-edge stands on the dactylus at about one-third of its length from
the articulation, and there are three similar teeth between the latter and that tooth; on
the immobile finger the cutting-edge reaches a little farther, and here also three low teeth
are found posterior to it.

In the largest female (No. 10), which is 41 mm. long, the apical part of the rostrum
(fig. 67) is rather much turned upward and reaches slightly beyond the antennal scales ;

45*

318 DR. J. G. DE MAN ON SPECIES OF PALZMON

the basal part is, as usual, very slightly arcuate. Two-thirds of the terminal joint of
the external maxillipedes extend beyond the peduncles of the external antennez. The
legs of the first pair extend beyond the scaphocerites for the distal fifth of their wrist.
The carpus is 5 mm. long, and 052 mm. thick at the distal end; the hand is 2°72 mm.
long, the palm measuring 1°56 mm., the fingers 1:16 mm., and the palm is 0°56 mm.
broad. The larger leg of the second pair (fig. 68) extends almost with half the carpus
beyond the scales, the shorter leg with only the hand. The cutting-edge (fig. 69) of the
immobile finger extends along three-fourths of its length, but it is somewhat shorter
on the dactylus; the dactylus bears three low and obtuse teeth of equal size posterior to
the somewhat larger conical tooth at the extremity of the cutting-edge ; on the immobile
finger, however, the inner margin runs somewhat irregularly between the distal tooth and
the articulation, but distinct teeth are not distinguishable. The fingers of the shorter leg
(fig. 70) present the same characters, but the teeth at the extremity of the cutting-edges
arerudimentary. The legs of the third pair overreach the scaphocerites with half their
dactylopodites.

Though the Catumbella specimens are all young or even very young, still there is one
ova-bearing female amongst them. This specimen is 40 mm. long; the eggs are very
numerous, about 0°5 mm. long and a little less broad. The rostrum, which reaches to the
end of the scales, is very slightly arcuate above the eyes, and the acute tip is somewhat
turned upward. The upper margin is armed with 16 acute teeth besides the two apical
ones; the fifth tooth stands just above the orbital margin, and the two apical teeth, of
which the anterior is much the smallest, are situated on that distal part which is slightly
turned upward. The posterior apical tooth is a little more distant from the preceding
than the latter are from one another. The inferior margin bears five teeth.

The chelze of the first pair of legs extend beyond the distal end of the scaphocerites ;
they are 2°66 mm. long, the carpus 45 mm. Only the right leg of the second pair is
present, the carpus reaches almost to the end of the antennal scales. The carpus
is, as usual, little longer than the merus and than the palm; the fingers, that shut
close together, are provided with the usual cutting-edge and a small tooth at the end
of it. The dactylus bears, moreover, three small teeth between that tooth and the
articulation, the immobile finger also has traces of teeth. This leg is somewhat hairy,
and there are a few spinules along the inner margin of the joints, but for the rest it is
smooth ; it resembles the smaller leg of the described young males, but it appears to be
more slender.

The legs of the third pair reach to the end of the antennal scales, the following are
hardly shorter. The meropodites of the fifth pair are 48 mm. long, the propodites
5 mm.; the former are 0°64 mm., the latter 0°42 mm. broad in the middle.

In the younger female (No. 13), which is 31 mm. long, the legs of the first pair project
with their fingers beyond the antennal scales. Those of the second pair are equal and
overreach the antennal scales with one-fifth of their carpus. The tooth at the end
of the cutting-edge stands, on the immobile finger of the right hand, at a little more
than one-fourth of its length from the articulation, and between both one observes three
smaller teeth of equal size; the cutting-edge of the dactylus is somewhat shorter, and

FROM THE INDO-PACIFIC AND WEST AFRICA. 319

posterior to it only two teeth are recognizable, one of which is conical, the other low
and rounded. The other hand agrees with that described.

In the youngest female (No. 17) finally, which is 21 mm. long, the apical part of the
rostrum (fig. 71) is slightly turned upward and extends just a little beyond the antennal
scales. The legs of the first pair reach to the distal end of the scaphocerites. The
carpus is 2 mm. long, and 0:22 mm. thick at the distal end; the hand is 1:16 mm. long,
the palm measuring 0°62 mm., the fingers 0°54 mm., and the palm is 0°25 mm. broad.
The second legs (fig. 72) are almost equal and their wrists reach almost to the end of
the scales, so that the fingers and two-thirds of the palm project beyond the scales.
In both chelz the cutting-edge of the immobile finger extends along about two-thirds
of the finger, that of the dactylus is somewhat shorter; at the end of each cutting-
edge there is a well-developed conical tooth and, between it and the articulation, on each
finger two other teeth that are a little smaller, especially those of the immobile finger
(fig. 73).

PALZMON (MacroBRacHium ?) sp. (Plate 20. figs. 75-80.)

One male and one egg-bearing female from Catumbella, near Benguella.

I describe first the female, because it is still provided with both legs of the second
pair. The specimen is 46 mm. long from tip of rostrum to the end of the telson. The
eggs are very numerous and small. The rostrum (fig. 75) is short, a little arcuate above
the eyes, then inclined downward, but the apex is very slightly turned upward again ;
the rostrum extends a little beyond the distal end of the penultimate joint of the
upper antenne. It is armed above with 11 rather strong teeth, that reach the distal
extremity, four of them standing on the cepbalothorax, the fifth just above the orbital
margin ; the first tooth, a little more distant from the second than the following, is
situated twice as far from the posterior margin of the cephalothorax as from the
anterior. The lower margin bears two quite small teeth immediately in front of the eyes.
The cephalothorax is smooth. The hepatic spine is smaller than the antennal one and
situated behind and rather far below the latter. Both in the female and in the male the
extremity of the telson (fig. 76) is triangular, but the acute tip itself is apparently worn
off, so that the extremity appears to be truncate (fig. 77); on each side one observes
the usual two spines, of which the inner reaches beyond the truncate tip. The anterior
pair of spinules is situated just behind the middle. The outer footjaws project with
their terminal joint beyond the peduncles of the external antenne.

The legs of the first pair (fig. 78) extend with their chelee beyond the aiaen end of
the antennal scales; the carpus is 4°8 mm. long; the hands, the fingers of which are
distinctly somewhat shorter than the palm, measure 3°5 mm. ‘The carpus is 0°74 mm.
thick at its distal end, this being almost one-sixth of its length, so that it has a rather
stout shape.

The legs of the second pair are equal (fig. 79) and 28 mm. lung, a little more than
half the length of the body; they project for a small part of their carpus beyond the
antennal scales. ‘The merus of the left leg is 5 mm. long and 1°75 mm. thick anteriorly;
the carpus is 4°75 mm. long and 2 mm. thick at the distal end; the palm is 55 mm. long,

320 DR. J. G. DE MAN ON SPECIES OF PALZMON

2-25 mm. broad, and 1°66 mm. thick in the middle; the fingers finally are 5 mm. long.
The palm is thus very little broader than the carpus. The fingers, about as long as the
palm and as the merus, but a little longer than the carpus, shut close together; their
inner margins (fig. 80) are provided with a sharp cutting-edge, and between it and the
articulation each finger is beset with three or four very small teeth. ‘The whole leg is
covered with acute spinules, which on the inner margins of the joints are a little longer
than elsewhere ; the joints seem, moreover, to be covered with felt and are somewhat
hairy.

Similar to the legs of the first pair, the three posterior pairs are stouler and less
slender than those of the young males of P. Olfersii that have been described on p. 315.
In the young male of this species, which has a length of 41 mm. (No. 5 of the Table), the
meropodites * of the third pair of legs (fig. 74) are 5 mm. long and 0°96 mm. broad in the
middle, measured on their outer side, 7. e., five times as long as broad; the propodites
are 4°4: mm. long and 0°54 mm. broad in the middle, a little more than eight times as
long as broad. In our female (fig. 81), however, the meropodites of the third pair are
55 mm. long, but 1-4 mm. thick, so that they are only four times as long as broad,
and the propodites, 46 mm. long and 0°7 mm. broad, are hardly seven times as long as
broad. The legs of the third pair reach in the female to the end of the scaphocerites,
those of the fourth pair are as long as the outer footjaws, and the last pair are still some-
what shorter. The three posterior legs are hairy along their upper and lower margins,
but for the rest quite smooth.

The male, that unfortunately has lost its second legs, has exactly the same size as the
female. The rostrum fully agrees with that of the female, bearing also 11 teeth above,
but it reaches to the middle of the terminal joint of the upper antenne, and there are
three small teeth on the lower margin. The outer footjaws reach almost to the end of
the scales, projecting a fifth part of the penultimate joint beyond the lower peduncles.
The first legs overreach the antennal scales with half their carpus; this joint is 5°5 mm.
long and 0°73 mm. thick at the distal end, so that it appears a little more elongate than
in the female ; the chel are 3°8 mm. long. The two following pairs of legs are wanting,
those of the fourth pair reach almost to the end of the scales, the last pair is incomplete.
I at first thought that the female was that of P. Olfersti, and that the legs of this
species were much thicker in the female than in the male. This would, however, be
quite an exceptional phenomenon, and this opinion was fully refuted by the examination
of the male, in which the legs are just as stout and thick as in the female. It is therefore
to be regretted that the second legs of the male are wanting. I have not succeeded in
identifying this species with any yet known, but it bears apparently a great resem-
blance to P. (Macrobrachium) Iheringi, Ortm., from Brazil.

* The joints are measured on their outer side, along their upper margin.

“eTTOQUINALD ‘ge-TT ‘SON {OSu0D ‘O[-F ‘SON fYIg IOANT ‘Eg pur Z ‘son { ulsoqry ‘T ‘oN

re
NX
oD FET | LL | $8 | SLL |FFL |98-T | 29-1 |9FT | F9-1| G9-T | 4T |68T (86 | 8-6 €& | GE |osr | ce-T | OFT |*' smesug “
9T | FT | 99-T | OFT |9LT | F9-T 6-1 | 131 LZ @ | 1E3 |98-3 IL-8 | 8-8 F | eer G | ¢9-T Gi |e ae OL [eC es
#83 | $3 6 | 9:2 | Zs € | GS |16-8 | 9-2] S9-¢ F |SLF |¢-9 | ¢9-9 L |$8-4 193-6 & | OF |''** puey ony jo yyWa00T
.0 | ae. : ape : : : ; : 0 |*9- ; 4 : ; * + pue Te }sIp ety 98
96:0 | GE-0 | 18:0 | LE-0 |GF-0 |S88-0 | 9F-0 |9F-0 | SFO | 4F-0 | 69-0 |*9-0 |8-0 | 8-0 | 8-0 |98-0 I F-0 | FF-0 | sai oy Jo sojau10IGg
6S | F¢E | LE | OFS | ee-F LF | 6o-F F LE g ¢ G \@-9 L g | ¢g |¢3-0 |\zee Gp | osndivo “ -
<j GE | FS | 936 | Fo | 6S | 8-3 | ELE | 6S | 18-€] OF | OFE |OFE ILS | GG |99-9 9 |S49 || 83 @ |** snzour 049 Jo Y}800T
(@) |
ra Se ae Ae a oe : : aed puoses 013 Jo
“A ie fe a. a” Y .
fs GIT eu) Gil FE SI G91 Nit 1G | 6G #8 rat ie 943 Jo 433097 9[O4 A
= poh. Cee 9 9 ¢ Gg ¢ Gu CG 9 ha ¢ ¥ e g ¢ ¢ | c ¢
A Er6 [+1+8] [+OL | GTOL |[+I +8 BA Olay eae ere B+6 | oté tit | 1+6 |I+6 +6 Stor I+01 Ce GO eT | eo Uy ters ee
[<a] G Go . Go Go co o
E cS cB 9% rs! Bee |) elke og| Te Ccleece eg| 98] 0s Tg¢| 36S Fc 8¢ | 1z 63 |** [eutUT oy} Jo yySU0T
A ——<—|__ |_| __-, a SS ee SS eS aS SS }
a ro) re) *} é 8 g *} } *} } ae) “Se $ § re) $ | “p "P
O ate "le ‘9S “G¢ HAS "6s "CS 1g ‘0S 63 83 ‘LG “9G cB GE "$3 ‘GG mid ‘0G
fe _ = ee —__—___— | 2
‘)
a g9-1 SL | 13 | 3] cor | Ge lee | LE 9 |¢L8 | oP G | ¢¢ L | Gg | GZ Z\¢-8i |) Gor | srs0y |e
' |
& id 6 | #3 | Fe] & | Gor G | oF 8 || oF 9 | ¢G |G49 | ¢8 | $9 OL || 6| #8 i mia ji aa igs fei
Zz |
i= €9-8 8 | Gr | 69) 66-6 | G22 |G48 | 3&8 FL 1Sa-8 | S-0L | G-0L | 4-61 | ¢-ST GI | GLI || of Gee | ¢-eg |° °° puBYy og} Jo TASUOT
Fi 9 | eg 2 fi a s "+ puo [RySIp ey9 4B
ES G0 | 9F-0 | 9-0 |T8-0 T | 60 T | 60 |99-0 66-0 |€en | FL | 91 |e2-n | Gt j S GE ELE Nendavo ayy Jo sjownrg
a ¢ |cLr | FE |9:0 | GL 4 | 98 | 96 | GIT] 8 |se0r | ¢-0T OL |S2-8T GI | SFL |¢-e1| GF 96.1] Sa rd os
EI 9F-E | 9FE F 18h | 39 G 9 |c%9 |4a8 ||GL¢ | Gel] GL | G8 | G6 | o8 TI |¢-01] 3&8 ZS |°* sniew oy} Jo qysueT
os 20 cho 50 oi = = , de) f ‘+ ared ptoaas ey} Jo
9 g ¢ G ii ¢ Gg F ¢ | 2G c ig 9 9 te 9 TG ae g say
I+1[+01| stor} stor |i+6\s+i+s) stor |jo+i+s) ste | otor | stor |iti+e\[+1+6| 1+01 | ctor |I+1+8 1+1+6 |o+6 Tq SRE BOT Oty JO eee
G @ ie & & {J é € & é € € & € $s ie
rats eg) 68} etl &F cr]; 0s ae 09 6g +9 89 9L SL 08 e9 | cg eer! oz |** emrue omg Jo yy8u07
P O19) ‘Pp ‘p “Pp =a) me) one) i) | oes) ‘4 Oras 3S ey or “p we} “Pp ‘Pp
‘OL ‘SL AT ‘OL "GT FI ‘SI “GL sie |

DR. J. GQ. DE MAN ON SPECIES OF PALAMON

322

‘ep[aquingey ‘FZ—-9 ‘SON ‘odu0p ‘g pus F ‘son ! yeig t9any ‘g pu g ‘son Se1taqry ‘[ ‘ON

ee losg | 9-9 log @¢] 6 Gel 6 O11¢It oeticeot| + Glar PH| 9 98 ley: L |i 7+ + sroSuy oq} jo yySuary
o1.0 |S@T | PI lest GI) 2 Gaeloe Pelee Fel FE (LT F1 TT oLl| Fr wt Gon hr i aed’ outer
t}or | ort feos et|re oaloe eloe srl ar lar stir etfot sot] te 66|\ ona caput ie cape
ge | 9-9 |e¢@ | 8 g¢) 66 Ol] OL TI Gtr G-OL| “oT lea gee )] op 99% \egg0) eo) imO;yiu mmm) me ser cya teen
L |GL-01 |GL-0T jee9L TT] GST 61] GI 12| 83 GE G6a |oo8 Go-01) 6 68 |SITE SIGSEl FI" pug “«
eerlep | ep | 9 giccs Leh Ges.ieze = eillcemdiee Gr le oe \egpi mo emcee gmimmmas sudagny = tae cay
ge | SLF G¢ |¢9 99/€2 GL/G8 G48 |916 BI) GIT] F GLb| ar F| G eg leg 99 fo sns0m 043 Jo yySue'T
Gp aOke | 96a 98 Z\- SF SF Or 0S) 29 TL “Oo ce 9c) ac (92a) = Yoeeacneae sf) | ei te Ave ON
| 2 I oi I oe 1 rs 1 Fr 1 a | ae | on ‘| JO So] OTOYA 944 Jo YASuET
dle aE ek ca $ & cs Aa % ca fora ae een |
€1 rd ra gI éI eI al zi al al iI eI RED AOTOUt i> BI Rua oa
P F P P , v g g y F i? hd |
1¢| or| 2 ao 29 el #8 ool; se OF IF Gh t+ Spoq oy Jo yySueT
———— | | | ef | | SS |
6) re) 5 5 78) ae) *} (wao) P ‘p “p “p
FG 8G CG ‘la ‘0G ‘6l ‘eI “LI 91 SI FI ‘SE
=
¢9 G&L! 6 98] 91 11] 9-2. gz-21] 61. stI| 972 |G-rL ¢-ElllFe | GF | G31 | €@ 96] 0 ge |''**** sxe8uy 013 Jo WySu07 |
FI Bl /ES ess] F S995 SLs jPH coe eso |re ef || € | 96 ¢€9 jet gor lee aitadions Segeaonee
| | RoR ON re
G91 93 [943 Gos |er elle ore|e¢ F| 8 ioe ee | ‘a GLE II 8 | 6 ¢8~ arate one ee
sz: Gelliga Gee lou eileen oer] so £1] Ss | of Sepplee jc: || geet og . -Leulide: sacyal Sei capaceteee = ae!
GEL 91| G-8I S8IlSE8 GFs| oa oho] Tr Gel G-69 |¢-62 alse |LOT] 92 69. - $@aalIRZOm neue creaming
GS 989/99 89 | IT $68 |SL8 GL8|9-€L GTI] 8 |S0L or|os |ce | 6 |Se6t €91 |¢-6t €2|-7*7**sndmo “
9 o9/ 2 L| 11 6866 6 /E€T G11] S21 | I SOr}es jee | 26 |eeer AT | 06 eg |-7*** enaom oma Jo qy3ueT
ee ogg} si OF) -99 ee! GG cl es Ox! “ern 99) — coierr ser g SEL 901] LaT calls ia nage et
GS | € Jy 1 lig 9 Ed ne 3 | ax | 1 in g 6 | Fe i T react AE jo 39 a[OyAr olf} Jo qysue'T
+ & 9 2 % & 5 a) sel 2 : OS well ee
I tI ra cl al Fa at ai | SI el Bed or TBO Se Us Oe SEO
£ P P P F g g y v ¢
0g F¢ OL TL 08 06 96 OSL| OFT!) 06 GFT ynoqa || 9€T "stts* Kpoq eyj Jo yySuaT
PO: "Pp Die) “Pp ‘Pp ms) “Pp ae) mo) | re} “0 "p
‘EL TI ‘OL 6 2 m7 | ‘9 9 | > g % ‘T

"S20]YLOTT “UUPAOYUSTIO A “42 ‘(squazZ) sisuoorvetuel (UINIYOVIqoIOVY) UoMIAeg fO saujawajjiU We SpUsMaINSDaTT

3

FROM THE INDO-PACIFIC AND WEST AFRICA.

“OTPPIUL OY} UL “TOAIMOY ‘saTeUT LoyYO oY} UT f AjIMOI}XO [eYSTp OT} 4B pornsvom st sndxeo puB sNIOUL Jo IOJOWIVIP 04} ‘G—J “SON SoTvUT oY} UT puL ‘sayeUIOZ OY} UT

PI 86-1 |98-T 6-1 j 6 | & FS |9TS BS | OG GS | FE Po a A fet See " sroSuy oy] Jo qy3u0T
vE-0 €60 | 70 0 | 90 +0 | $0 $90 | G0 GO | £0 %9-0 | 48.0 | 20 &T as Oi wed oxy jo ssouyoryy,
FF-0 F-0 9:0 69-0 9:0 $€9:0 | 89-0 60 |110 F210 T 98-0 | FLL 196-0 PRL Zak Masi PN sare wyed a3 Jo Taper

PI FEL |FLS OL | 13 G | &G € |886 FFG | FE § | Fé Soh SOPoe | maaan as iaentg wed

8: 8G. b 90-7 | LP b | &b FES |'9F 99% | 69 Gg 86-9 8-9 Ce ees sore sss* pueq ayy Jo yysue'T
86-0 860 |€¢0 0 290 64-0 |€9-:0 4:0 |€9:0 9:0 |28:0 F140 | LT 88-0 &T ** sndivo oj Jo cojouer(y

GSB GILG € 906 |96€ GEE | GE 996 |6¢E LE | Lr 68 | F CLF ¢9-¢ "++ *sndueo oy Jo yysu07T

60 16-0 F-0 GF-0 |GF0 GF-0 G0 8¢0 |8F0 G¢-0 | 99-0 9:0 | c8:0 8L-0 T ** snioul oY} jo Jojomery
FLE 90-6 LG LG L6G FL:6 & GE |90-6 FO GS FE | 9S b 99-7 Tere eres sss: snieur oqy Jo yysueT

G6 66 | GEL 9-81 tL SPI GI G.9T GL GT 6L G41 9-61 16 96 sad PUod9s a4} Jo ssa OY} Jo YASUg] [20],

2 ; Ae T a r :e t a . wu 4 a T
g eth __ t x] ig ao _g
Cie G+EL G I+el G+SE1 StI tal I+ ST Stor +91 ceeeeeeeess © TUNIz80L O44 JO BULLET
3 ¥ v |
1G TG 86 6G 1S 98 | OF If ctereeccesees TomITE 04 Jo UISuET
°S) 26) fe) "6) °) ae) Gi S
| DT ‘OL ‘aT PI ‘eI ra “yah ‘OL
8 G3 9Le | 9 G FF |SL SEG | 8E Gh | 6F 4 | Gr Gh | SBr| "++ sxosag oy} jo Waa07
| ¥:0 ¢9-0 8:0 | GL GL-T IE GG G1 8:0 Gl GL G9-G L Cole imo ** wed 043 Jo ssouyoryy,
86:0 68-0) OLE G So Sh | eS. Sil Gal = eS: en |SGa SOS Glee rote oat Tress ss wiped ayy Jo qapvarg
G EG FEE 19-6 |GL-9 GE |GL-8 ¢6-F 6-6 Fb |GL-6 G8 GS Gigs a= ir ieee ck toa earn a
8-6 G e-9 6 GLIL G-L 9T &-6 T-4 G8 | 99-8 GGT 8 Gé-ILT || S-66 |° "Tos 5" puwy 049 Jo q98u0T
| GO FF-0 |64:0 G6-0 8-1 G EL GG Wl LT FL GI 9-6 G1 9:1 OGG) * | tog er le eal snd.ivo oy} Jo Loy ouUrcy
/F8-G 8-6 € L& | oP GG F L BP GF COF | Fr GEL i G-¢ GEL ott ts sndaro oxy Jo yysu0T
| FO 98:0 9:0 F1:0 | GL G9-T LL FG GFL |66-0 GL FL GG TT 9-1 g | UCC c tte ss" snieur ey} Jo JoZoWBICT
9G 99-6 & 9S v g FP |S6-9 oP 6 Gor P G9 | GLE S GT || Sf eG tee ea! “snout OY} JO YysuO'T
G9-6T GGL 61 | G&G 66 16 G& CG 16 6 66 G-9E GE OG-8 | 99 | ared puoodas aug Jo sso] 049 Jo YySuoy [ejOY,
ig 1 a Hr 7 te : a 4 ee As es Ae 1 |
=) G85 Bt id f aes 29's 9
o+FI GtGL peread ¥1 I+91 I+$1 I+81 1+61 or |’ Terese ss ss Uantzsor ogy Jo BTNUIOT
€ g g ¢ v 3
16 tS 96 G-8E IP GF cP TP BG prot rsseescss 2 TeuoE oug JO ussooN,
“2 "p ‘2 “p "p ‘p “2 “2 'P
6 8 L 9 g * g % aT

‘mbar j{ USIOJ[O (ummpovaqoror yy) wouaed Jo savjameyyy Up Spuauaunsvary

No)
=

SECOND SERIES.—ZOOLOGY, VOL. IX.

324 DR. J. G. DE MAN ON SPECIES OF PALAMON

TABLE OF THE WEST-AFRICAN SPECIES OF THE GENUS P4Lzmoy, FABR.,
CONTAINED IN THIS PAPER.

Miss Rathbun, in her instructive paper “The Decapod Crustaceans of West
Africa”? (Proc. U.S. National Museum, vol. xxii. 1900, p. 315), mentions four species
of the genus Palemon as inhabiting West Africa. Three species ought now to be added
to them, viz., P. (Hupalemon) Foai, Cout., another undetermined species of the same
subgenus from the Upper Congo (Coutiére, op. cit. p. 519), and finally the species from
Catumbella belonging to the subgenus Macrobrachium, related to P. Ihering, Ortm.
Some principal characters of the five species described in this paper are the following :—

A. Large chelipede with palm cylindrical. (Subgenus Eupalemon.)
(8, 9, or 10) + (1 or 2
4, 5, or 6

Apical teeth of the upper margin usually present.
Anterior pair of spinules situated before the middle of the telson.

a. Rostrum ) dentate.

Carpus of second legs longer than merus, always longer than palm, fingers covered with felt.
: ; P. macrobrachion, Herklots.
8. Rostrum 5 dentate, no apical teeth.
Anterior pair of spinules situated immediately behind the middle of the telson.
Carpus of second legs longer than merus and longer than palm, fingers without felt.
P. Foai, Cout. (River Kribi.)
B. Large chelipede with palm more or less compressed. (Subgenus Macrobrachium.)

4or5
y. Rostrum = dentate, no apical teeth. :
Carpus of second legs about as long as merus or a little shorter, always shorter than palm.
Three posterior legs slender.

P. jamaicensis, Herbst, var. Vollenhovenii, Herklots, and var. ? angolensis, de M.

6. Rostrum apical teeth present.

3-5
(13-16) + (1 or 2) Alenéate
4-7 2.
Carpus of second legs about as long or slightly longer than merus, either a little longer or a little
shorter than palm.

Posterior legs slender . . - - - + ++... =. + « P. Olfersit, Wiegm.

e. Rostrum 2 dentate, no apical teeth.
Carpus of second legs about as long as merus, but not longer than it, slightly shorter than palm.
Posterior legs, also the others, thicker than those of the two preceding species.
Palemon sp, (Catumbella.)

Fig. 1.

Fig. 2.

Fig. 13.

. Left leg of the second pair of the largest female, long 58 mm., from Catumbella, x
. Right leg of the second pair of the male (No. 20), long 29 mm., from Catumbella,

Fig. 17.

24.
26.

FROM THE INDO-PACIFIC AND WEST AFRICA. 325

EXPLANATION OF THE PLATES.

PLATE 18,

Palemon (Eupalemon) lar, Fabr.
x 6.

Palemon (Parapalemon?) asperulus, vy. Martens.
the young female, x 3.

Extremity of the telson of the largest male, long 125 mm.,

Lateral view of cephalothorax and rostrum of

. Abdomen of the same, x 3.

. Extremity of the telson, x 25.

. Right leg of the first pair, x 6.

. Right leg of the second pair, looked at from above, x 6.

. Lateral view of the carpus of the same leg, viewed from the outer side, the upper surface is

situated on the left hand, x 6.

. Right leg of the third pair, x 6.
Fig. 9.

Palemon (Macrobrachium) latimanus, vy. Martens. Lateral view of cephalothorax and rostrum

of an adult male from Dinawa, x 2.

. The same of another adult male, x 2.
. Extremity of the telson of an adult male, x 10.
. Left leg of the second pair of an adult male, viewed from above, x 14.

Palemon (Eupalemon) macrobrachion, Herklots.
of a male, long 70 mm., from Liberia, x 2.

Lateral view of cephalothorax and rostrum

. Right leg of the second pair of the adult male from the River Prah, viewed from above, x 13.
. Lateral view of cephalothorax and rostrum of the female from the River Prah, x 2*.

. Right leg of the second pair of same female, looked at from above, x 2.

. Left leg of the second pair of the female (No. 27), long 76 mm., from the Congo Coast, x 2

~e

. Right leg of the second pair of the female (No. 6), long 78 mm., from the Congo Coast, x 2.
. Two spines of the inner margin of the carpus of this leg, x 25.
. Teeth of the fingers of this leg, x 25.

The hairy felt has been omitted.

2.
x 5.
PLATE 19,

Palemon (Eupalemon) macrobrachion, Herklots.
of the male (No. 4) from the Congo Coast, x 2.

Lateral view of cephalothorax and rostrum

. The same of the female (No. 7) from the same locality, x 2.
. Lateral view of cephalothorax and rostrum of the largest male, long 60 mm., from Catumbella.

x 2.
The same of the largest female, long 58 mm., from the same locality, x 2.
Teeth of the fingers of the left leg of the second pair of this female, x 25.
dactylus are on the left hand.

The teeth of the

* Though this figure, as also fig. 18 of Plate 19, are quite accurate, their appearance is not natural, as they
have been turned downward by the draughtsman, so that the rostrum runs horizontally forward and somewhat

downward, instead of being turned slightly upward.

—.
hy
f e.. é ; r)
j C P CG
Loi LIBt ')2
= 2? / -
\ - \ ; ‘ Pe sy
\" 8 9. LY
NH ae! WS

Fig.

we

DR. J. G. DE MAN ON SPECIES OF PALAMON

Teeth of the fingers of the right leg of the second pair of the male (No. 20), long 29 mm., from
Catumbella, x 50.

. Teeth of the fingers of one of the legs of the second pair of the young female (No. 38), long

25 mm., from Catumbella, x 50.

. Palemon (Eupalemon) Foai, Cout., male, from the River Kribi. Lateral view of cephalothorax

and rostrum, xX 2.

. Telson, x 3.

. Extremity of the telson, x 12.
. Leg of the first pair, x 2.

34.
5. Two spinules of the inner margin of the palm of the same leg, x 25.

Left leg of the second pair, x 2.

. Two spinules of the longitudinal row on the outer margin of the palm, and six thinner ones

near that row on the upper surface of the palm of the same leg, x 25.

. Teeth of the fingers of the same leg, x 25. The teeth of the dactylus are on the left side.
. Palemon (Macrobrachium) jamaicensis (Herbst), var. Vollenhovenii, Herklots. Juateral view of

cephalothorax and rostrum of the adult male from Liberia, x 13.

. Extremity of the telson of a young male (No. 14), long 41 mm., from Catumbelia, x 25.
. Larger (left) leg of the second pair of the same male, x 14.
. Lateral view of cephalothorax and rostrum of the male (No.7), long 90 mm., from Catumbella,

ee

. Left leg of the second pair of the same male, x 1t.

. Right leg of the second pair of the male (No. 8), long 80 mm., from Catumbella, x 14.

. Left leg of second pair of same male, x 1}.

. Right leg of the second pair of the largest male (No. 6), long 96 mm., from Catumbella, x 14.
. Larger (left) leg of the second pair of the young male (No. 16), long 38 mm., from Catum-

bella, x 3.

PLATE 20.

. Palemon (Macrobrachium) jamaicensis (Herbst), var. Vollenhovenii, Herklots. Lateral.view of

cephalothorax and rostrum of the young male (No. 16), long 38 mm., from Catumbella, x 3.

. Teeth of the fingers of the larger (left) leg of the second pair of the same male, x 13.
. Lateral view of cephalothorax and rostrum of the largest female but one from Catumbella

(No. 18), long 84 mm., x 2.

. Left leg of the second pair of the same female, x 14.

. Right leg of the second pair of the same female, x 14.

. Left leg of the second pair of the young female (No. 23), long 40 mm., from Catumbella, x 3.
. Teeth of the fingers of the same leg, x 13.

. Palemon (Macrobrachium) Olfersii, Wiegm. Lateral view of cephalothorax and rostrum of

the adult male from the River Prah, x 3.

. Left leg of the second pair of the same male, x 14.
. Lateral view of cephalothorax and rostrum of the male (No.3), long 42 mm., from Catumbella,

xioe

. Extremity of the telson of the same male, x 25.

. Left leg of the first pair of the same male, x 6.

. Larger (left) leg of the second pair of the same male, x 2.
. Teeth of the fingers of the same leg, x 5.

. Right leg of the second pair of the same specimen, x 2.

Teeth of the fingers of the same leg, x 10.

OE —

FROM THE INDO-PACIFIC AND WEST AFRICA. 327

Fig. 63. Left leg of the second pair of the young male (No. 7), long 36 mm., from Catumbella, x 2.

64. Teeth of the fingers of the same leg, enlarged.

65. Left leg of the second pair of the young male (No. 8), long 34 mm., x 3.

66. Teeth of the fingers of the same leg, x 25.

67. Lateral view of cephalothorax and rostrum of the largest female (No. 10), long 41 mm., from
Catumbella, x 38.

68. Left leg of the second pair of the same female, x 3.

69. Teeth of the fingers of the same leg, x 25.

70. Right leg of the second pair of the same female, x 3.

71. Lateral view of cephalothorax and rostrum of the youngest female (No. 17), long 21 mm., from
Catumbella, x 6.

72. Right leg of the second pair of the same female, x 6.

73. Teeth of the fingers of the same leg, x 25.

74, Right leg of the third pair of the male (No. 5), long 41 mm., from Catumbella, x 5.

Fig. 75. Palemon (Macrobrachium?) sp. Lateral view of cephalothorax and rostrum of the female from

Catumbella, x 3.

76. Telson of the male, x 5.

77. Extremity of the telson, x 25.

78. Right leg of the first pair of the female, x 5.

79. Left leg of the second pair of the female, x 3.

80. Teeth of the fingers of the same leg, x 25.

81. Right leg of the third pair of the female, x 5.

[Nore received since the foregoing pages were in type :—
The preceding description (pp. 806-309) agrees quite well with that of Coutiére. Of

the three specimens described by him, a male long 70°5 mm. and two somewhat younger

1 1 1
8

? the two females respectively ; and

In the male the carpus of both legs of the second pair was a little longer than in the
specimen from the River Kribi, being once and a half as long as the merus, and the fingers
of the larger left leg measured not quite one-third the length of the whole hand. The
joints of these legs are described as smooth, except some sharp spinules on the lower
border of tle palm; when touched, the legs appeared, however, to be slightly scabrous.
Coutiére makes no mention of the characteristic position of the spines of the telson, and
it is remarkable that he does not compare his species with P. paucidens, Hilgd.—
17th November, 1904. J. G. pE M.]

8
3

females, the male head for its rostrum the formula

SECOND SERIES.—ZOOLOGY, VOL. IX. AT

Trans. Linn. Soc. Srr. 2. Zool. Vol 1X. Pl. XVIIL

DE MAN DEL. JT RENNIE REID, LITH. EDIN®

SPECIES OF PALAEMON.

Trans. Linn, Soc Szr.2. Zool. Vol. IX. Pl. XIX

Dr Man

Sea ee le eee ee oe

JT RENNIE REID. LITH EDIN®

DE MAN DEL.

SPECIES OF PALAEMON

°
‘
‘ ©
:
‘
'
‘
‘

.
*
A . .

fi *

Trans. Linn. Soc. Ser.2 Zool. Vol. IX. Pl] XX

DE Man.

JT RENNIE REID, LITH.EDIN®

DE MAN DEL.

SPECIES OF PALAEMON.

a

pvaze

IX. Observations on some undescribed or little-known Species of Hemiplera-Homoptera
of the Family Membracidee. By G. Bowpter Bucxroy, F.R.S., LS.

(Plates 21 & 22.)

Read 17th November, 1904.

THE efforts made during the last few years to systematize, in some natural manner, the
Homopterous family of Membracidee have met with varying success. The somewhat
numerous species have hitherto engaged less attention from the entomologist than they
deserve, though they constitute a remarkable group, in which diversity of form suggests
problems and theories on the effects of environment, and mimicry also, which will
exercise the patience of the experimentalist and the acumen of the biologist. The little
interest shown may partly be ascribed to the comparatively small size of some species,
which necessitates an appeal to the microscope so as to reveal their chief characteristics ;
yet many entomological collections contain undescribed examples of these Hemiptera
which will repay the investigation into their distribution and life-history, &c.

An examination of a few examples not hitherto described —or, at least, not adequately so
for identification—constitutes an excuse for offering some notes to the Linnean Society.

Another cause of the entomologist’s indifference may be the fact that very few repre-
sentatives of the family are known as denizens of Europe. Species are mostly exotic to
England, and are at present most numerously represented on the continents of North
and South America. The European species typified by Linnzeus in Centrotus cornutus
and the American in Membracis foliacea hy Fabricius, are now expanded into several
subfamilies and numerous genera.

The significance of their remarkable forms and their dependence on environment,
also on their protective and aggressive mimicry, have been discussed by Prof. Poulton.
To him and to the Rev. Canon Fowler, both Fellows of our Society, Iam indebted for
information as to the economics of the species and their persistence throughout the
struggle for life.

Where the man of science can frame a tenable hypothesis he often produces more
valuable fruit than the compiler of facts, however unanswerable these may be; but the
work of the resolver of what appears to be a confusion into a consistent order has its value.
On this footing I offer to the Linnean Society the present contribution. It is advanced
as tentative, and must be so until our knowledge of the life-history of Membracid shall
add to the bare facts now alone at our disposal.

The recently published memoirs by W. W. Fowler, in the ‘ Biologia Centrali-Americana’
of Godman and Salvin *, and my ‘ Monograph of the Membracide ’ +, may be consulted

* Biol. Centr.-Amer., Rhynch.-Homop. (W. W. Fowler), vol. ii. (1894).
t+ Mon. Membracide (Buckton) (1903).
SECCND SERIES.— ZOOLOGY, VOL. 1X. 48

330 MR. G. BOWDLER BUCKTON ON UNDESCRIBED OR

as to related genera. In the last work an attempt has been made to classify the family
as it is at present known.

The ‘ Biologia’ above alluded to of course only deals with the American insects.
Prof. Ign. Bolivar has obligingly forwarded to the writer specimens from the Musée
d’Histoire Naturelle de Madrid, and the writer has also received examples of unnamed
species from the Musée Belgique. These form the subjects of the present memoir.

RHYNCHOTA-HOMOPTERA.

MEMBRACIS MICANS, sp. n. (Plate 21. fig. 1.)

Pronotum foliate and flattened laterally ; metopidium rising perpendicularly above the
eyes. Colour pale stramineous, shining like mica, with a pale brown fascia reaching
from each shoulder to the thin upper edge, succeeded by two other interrupted fasciz, the
latter carried to the posterior end. Legs pale ochreous; fore legs slightly spatulate with
brown claws. ‘Tegmina hyaline with yellow neuration. Twelve specimens at my
disposal were pretty constant in their colouring.

From the Belgian Museum. Collected by M. Van Voixem. Size 9X6 millimetres.

MEMBRACIS VERGENS, sp. n. (Plate 21. fig. 2.)

Rather large. Colour a dull coal-black. Seen in profile, with a bright ochreous band
which extends backwards from above the eyes on the metopidium to the dorsal edge;
this forms a waving stripe on a black ground. Metopidium rather overhangs the head.
Legs black and slightly spatulate.

Hab. Sta. Caterina.

Coll. Camilie, Van Voixem. Size 126 millimetres.

Hama NoposuM * (genus et species nov.). (Plate 21. fig. 3.)

Pronotum turgid and prolonged into a knot-like sinuous process, furnished with
numerous small spines. This serpentine appendage is continued free from the dorsum,
and is nearly as long as the tegmina. ‘These last are short, each furnished with a broad
corrugated limbus and with a brown coarse neuration. Metopidium high and crested.

H. nodosa is of a concolorous shining coal-black. The tegmina are ochreous and
diaphanous at the tips. Legs black, except the hind pair, which are rufous. The
posterior process is contorted into segmental knots.

Hab. The Kameroons, W. Africa. Size 4x2 millimetres.

Allied to Sphongophorus.

MIcROSCHEMA MUCRONATA, sp. n. (Plate 21, fig. 4.)

Larger. Pronotum rising perpendicularly from the frons into a pointed dorsal process,
obtuse in the outline and continued as a straight line to the sharp posterior apex.
Suprahumerals rather short and divergent. Colour bright red, with punctured dots.

* "Apa, a knot.

LITTLE-KNOWN SPECIES OF MEMBRACID, 331

The upper edge of pronotum broadly black at the summit, which shade is continued
as a black line nearly to the apex of the tegmen. The tegmina ample and pointed at
the tips, with a broad limbal edge and of a fine purple-brown colour, too dense to
show the neuration. rons pale sordid brown. Legs rather spatulate. Size 15 x6
millimetres.

Musée de Madrid.

ACONOPHORA OBFUSCATA, sp. n. (Plate 21. fig. 5.)

Pronotum porrect, or projected forwards as a flat sharp horn, laminated at the edges.
Colour dark fuscous and mottled. Pronotum carried to the posterior end, which
terminates in a point nearly reaching to the tips of the tegmina. Legs rather long, with
yellow tibize; 'Tegmina pale ochreous with a fuscous newration.

This insect is somewhat like A. flavipes, but it is not so large and the given locality
differs also.

Hab. Mexico. Size 11x 4 millimetres.

TRAGOPA TRIANGULATA. (Plate 21. fig. 6.)

Small. General aspect scutiform. Pronotum, viewed from the dorsal aspect,
trapezoidal or four-sided. Tegmina short, much corrugated, and difficult to separate
from the abdomen. Suprahumerals hardly visible, but by the frontal aspect rather
auriculate, suggesting some affinity to Chelyoida*. Eyes large and prominent. Legs
short and robust. Colour sordid ochreous, with dark fuscous on the thorax and on the
abdomen. Size 4x4 millimetres.

Musée de Madrid.

PoPPEA SUCCINEA, sp.n. (Plate 21. fig. 7.)

Pale amber-yellow, rather transparent. Pronotum raised into bulbous tubercles, the

posterior bulb of which forms two stout processes, somewhat similar to the suprahumeral

horns. Eyes prominent. Tegmina hyaline, glistening, but corrugated, and with a broad
limbus. The suprahumerals are divergent and united between the shoulders to a button-
like scutellum, which joins the tuberculous dorsum.

This insect has a considerable resemblance to Poppea concinna 7, but is larger and has
more robust suprahumerals.

Hab. Mexico. Size 9X3 millimetres.

Musée de Madrid. :

ELECTROPHINA PACIFICATA (genus et species noy.). (Plate 21. fig. 8.)

Has some of the characters of a Cevesa, particularly in the neuration of the tegmina,
which are remarkable for their length, viz. about twice that of the body, and also by
the occurrence of conspicuous suprahumeral horns. These are barely visible in Ceresa.

. | * Mon. Membracide, pl. 33. fig. 2, p. 156.
+ See Mon. Membracide, pl. 34+. fig. 5.
48*

332 MR. G. BOWDLER BUCKTON ON UNDESCRIBED OR

Electrophina pacificata is a relatively large insect, almost concolorous yellow, with the
pronotum slightly inflated and punctured. It is not laminated, or flat, as in Ceresa.
A dull fuscous patch over the metopidium, a transverse stain on the pronotum,
and a fuscous tint at the posterior horn, are the sole variegations of the yellow colour
of the insect. The horn is free above the large abdomen. ‘The tegmina show ovoid
membranous cells which are bounded by pale fuscous nervures. Legs moderate in
length and brown.

Hab. Coll. de Pacifico. Size 135 millimetres.

Musée de Madrid.

CERESA NITENS, sp. n. (Plate 21. fig. 9.)

Pronotum arcuate and flat, and ending in a sharp point. Abdomen large and
ringed. Metopidium, when seen from the front, high and furnished with short supra-
humerals. Legs short. Surface very shining, like corrugated tale; colour sienna-
yellow or of an amber hue. Pronotum marked by a conspicuous brown or black
transverse fascia. Tegmina slightly ferruginous, but with clear membranes. Legs
ferruginous brown.

Hab. Chiriqui. Size 9X5 millimetres.

Musée de Madrid.

ENTYLIA M@sTA, sp. n. (Plate 21. fig. 11.)

Small. Metopidium, as scen from the front, rising into a pointed and punctured
prominence, which appears broad and truncated by the profile view. Pronotum rises
behind into a hump, which falls off to the posterior apex. Colour dingy ferruginous,
with a pale carina on the procephalon and two other streaks down the hump. Tegmina
short, with grey patches on the shoulders and corrugated grey on the tips. Legs
stout.

Hab. Mexico. Size 5X3 millimetres.

Musée de Madrid.

ENTYLIA FUSCODORSA, sp. n. (Plate 21. fig. 10.)

Larger than the last insect. The procephalon smaller and less truncated. Colour
pale greenish yellow. Tegmina with deep punctures and bro wnish blotches between the
venations, and with still larger blotches below the pronotal horn. The lower margin of
the pronotum notched where it joins the metopidium. The dorsal hump is often, but
not invariably, ferruginous brown. Tegmina olive-grey. Legs ferruginous. Size 6X 4
millimetres.

HYPSAUCHENIA JUGULATA, sp. n. (Plate 21. fig. 12.)

Dorsum. with a yellow patch between the procephalic horn and the dorsal prominence.
The long curved cephalic process has a yellow line on each side, which runs from the eye

LITTLE-KNOWN SPECIES OF MEMBRACIDA. 330

to the summit. The fore legs obscurely spatulate, the other legs yellow. The general
colour of the insect brownish black, more or less covered with a fine corrugated
punctuation.

The species hitherto described are distributed over several islands of the Indian Ocean
and the Philippines, but this is the first record of their occurrence in Sumatra.

Hab. Sumatra. Size 8x9 milimetres.

Musée de Madrid.

OURANORTHUS PALUS * (genus et species nov.). (Plate 22. fic. 1.
(5 i >

Although this somewhat singular insect does not strictly conform to the diagnosis
given by Fairmaire for his genus Lamproptera, I think provisionally it may be placed
under that classification. The erect horn in the dorsum is single, not seen as “ cornubus
duobus,” and is inserted at a right angle just above apex of the abdomen. Lanceolate in
form, it is neither carinated nor glabrous. The metopidium rises above the eyes into
a tumid hump, and then it proceeds nearly straight to the apex. Legs stout and
slightly spatulate. Frons furnished with two short recurved suprahumerals. The colour
fine yellow and the surface devoid of hairs. Tegmina yellow, with orange-coloured
nervures.

Hab. Bangalore, India. Size 8x5 millimetres.

Musée de Madrid.

KLEIDOS PALMATUS, Sp. 0.

Tegmina sombre brown, but inclining to red at the tips. Posterior horn vomerate or
like a plough-share, with fine serrations on the lower edge. A slight tubercle occurs
above the geniculate angle of the horn. In other respects it resembles Kleidos vomeris
(figured in Mon. of the Membracide, pl. xlviii. fig. 2) and is the second example of
that genus.

Hab. Lanzibar. Size 9X6 millimetres.

Kleidos vomeris occurs in Ceylon.

ANCHON STRIGATUM, sp.n. (Plate 22. fig. 3.)

Procephalon conical, with the summit divaricate or split into two leaves, which the
insect appears to be able to close and open at will. The base of the cone has a white
line which runs to the top of the same. Posterior horn ulnate and tapers to the end,
without any dilation. There are no suprahumerals, or they may be represented only
by obtuse points. Tegmina bright and of a shining yellow, but corrugated and stained
with fuscous on the limbus, a spot on the costa, and a patch on the inferior edge. Tibiw
yellow. Size 9x6 millimetres.

This insect recalls Anchon albolineatum, but it is distinct.

Hab. Cameroons, W. Africa.

*-oipa, tail; avopOdw, I erect.

334 MR. G. BOWDLER BUCKTON ON UNDESCRIBED OR

ANCHON FUSCUM, sp. n. (Plate 22. fig. 2.)

Concolorous light brown, except at the tips of the tegmina, which are darker, and fuli-
ginous near to the costa, and the legs obscurely ferruginous. The procephalon is without
suprahumerals and the summit is divaricate, as in the last species. Posterior horn ulnate
and sinuous. Size 7x5 millimetres.

Hab. Cameroons, W. Africa.

TALOIPA TINCTORIA* (genus et species nov.). (Plate 22. fig. 4.)

Small, robust. Concolorous black, except the tegmina. Suprahumerals short and
square to the frons and to the metopidium. The posterior horn of the pronotum very
short, and not equal to half the tegmen. Frons and face hirsute. 'Tegmina ochreous,
but diaphanous, corrugated with a brown neuration. The base stained with a red
suffused fascia, giving the wings a mottled tint.

Hab. Manila, Philippines; Bangalore. Size 7 x 4 millimetres.

This insect has mixed characters of Centrotus and Otinotus, &c.

LEUCOTHORAX VILLOSA (genus et species nov.). (Plate 22. fig. 5.)

Large, robust. Posterior horn simple, but rather curved and shorter than the
tegmina. Suprahumerals acute by the profile view, but truncated by the dorsal aspect.
Colour dark shining brown, furnished with a broad white villous space on the thorax
and at the wing-insertion. Two white spots on the dorsum. Legs strong, black, with
yellow at the tips of the tibiz and the tarsi. Membranes of the tegmina corrugated
and shining, but the neuration is obscure.

This is a striking species, partly from its diverging horns and tomentose coating.

Hab. Cameroons. Size 126 millimetres.

Musée de Madrid.

LEPTOCENTRUS IMPUNCTUS, sp. n. (Plate 22. fig. 6.)

Suprahumerals stout and recurved. Procephalic horn rather short, cylindrical, and
distant from the abdomen. General colour dark brown, shining, with a tendency to show
« white pilose spot on the pronotum. ‘Tegmina long, with warm ferruginous and brown
corrugations and nervures.

In the Madrid Museum there are several specimens of this species, which show
slightly different sizes and also colouring, but they may be considered as identical.

Hab. Padautsin (?). Size 105 millimetres.

IBICEPS RUFIPENNIS, sp. n. (Plate 22. fig. 8.)

Colour dark brown, nearly black. Metopidium rather high, with erect suprahumerals
and a free cylindrical posterior horn which is longer than the abdomen. The tegmina
brown, with a broad rufous or yellow spot occupying the apical area. This spot is more

* ra Nora, the residue.

LITTLE-KNOWN SPECIES OF MEMBRACIDA, 399

obvious in some examples than in others. There is also a greyish sheen spread over the
basal portions of the wings.
Hab. Cameroons. Size 8X 4 millimetres.

OPHICENTRUS SERPENTARIUS, sp. n. (Plate 22. fig. 7.)

This species is characterized in great part by the sinuous form of the posterior process
or horn. Although the examples given by Canon Fowler in the Biol. Centr.-Amer. are all
American, this species from Africa and from Tasmania has its significance.

Colour dark brown, showing a slight grey pubescence. Metopidium high. Posterior
horn stout, much gnarled and bent into a tapering curve shorter than the tegmina, which
last are warm reddish fuscous with dark neuration and a corrugated limbus. Legs
stout. Abdomen and the rest of the body dark brown.

Hab. Cameroons. Size 8x4 millimetres.

Musée de Madrid.

There are other smaller specimens in the same Museum which have broad fuscous
bands on the tegmina, and these as varieties may be designated Ophicentrus minor var.,
from the Cameroons.

POLOCENTRUS LABATUS, sp. n. (Plate 22. fig. 10.)

This genus is characterized by the clavate apex of the posterior horn, which is serrated
below the clubbed extremity. The suprahumerals are short and obtuse by the profile
aspect. Colour ochreous-orange, mottled with fuscous. TFrons square and brown.
Legs stout, flat, with yellow tibie. Tegmina with yellow cellules and with broad brown
nervures.

The usual habitat of the genus appears to be Southern India, but this species is from
Abyssinia.

Musée de Madrid. Size 8x4 millimetres.

POLOCENTRUS CAUDATUS, sp.n. (Plate 22. fig. 9.)

Suprahumerals shorter and more erect than in P. lobatus, and the tegmina not
brocaded with brown but diaphanous. Colour uniformly bright ochreous yellow. Legs
flattened and almost spatulate. The clavate apex of the posterior horn is large and
serrated on the lower margin.

Hab. Natal. Size 85 miilimetres.

TRAPEZOIDA HIRSUTA (genus et species nov.). (Plate 22. fig. 11.)

The pronotum quite covers the scutellum and is domed in outline when seen from
the side, but it has a somewhat four-sided outline from the dorsal aspect. The tegmina
ample, broad, and longer than the posterior horn. Metopidium square, with short
suprahumerals.

Frons covered with hairs. Eyes prominent. Legs with black femora and yellow-
fringed tibia. Colour uniformly dark fuscous, but with a yellow carina on each of the

336 MR. G. BOWDLER BUCKTON ON UNDESCRIBED OR

suprahumerals and two wide yellow fascize across the dorsum, leaving the apex black.
The tegmina are dense and do not readily show the neuration.

Hab. Central America. Size 7x4 millimetres.

Musée de Madrid.

Perhaps this insect might be included in the original genus Centrotus; yet it differs
from Linnzeus’s typical Centrotus cornutus, which has been retained for reasons set
forth in my ‘ Monograph of the Membracide,’ p. 245.

There is no waste in the products of animal life, and it is a fact familiar to all
observers that the effete excretion of one animal is often the food for another lower in the
biological scale. The sweet secretions from many Homoptera are much sought for by
ants and by the members of some insect families separate from them both in habit and
classification. Thus we have Aphidze, Cercopidee, Fulgoridee, and Membracidze all laid
under contribution for the pleasure or nourishment of different orders of insects. Whilst
in the Aphidee we find at least two discharging orifices or nectaries for such excreta,
Mr. E. Green has shown that in Centrotus nectavis of Ceylon the lurve have but one
duct, which is capable of extension like the tube of some telescopes.

The larve of another species are common at the Cape of Good Hope, probably
belonging to the genus Oxyrhachis, which carry their single nectary erect from the apex
of the abdomen. They also are visited by ants. Although the winged insect has not yet
been ascertained, a figure of this curious larva or pupa may be added to the singular
forms which represent the pupal and immature stages of the Membracide.

It may be remarked that these pup are incapable of flight, yet they have the
rudiments of the tegmina much developed, and that they are very active in their move-
ments. ‘They possess eyes and antenne.

Future observation will show, indeed, whether they are pup or only arrested imagoes.

Pupa. (Plate 22. fig. 12.)

Robust. Colour wholly black, except the eyes, which are large, prominent, and
sordid ochreous. Metopidium continued into a single sharp and erect horn which
slopes nearly straightly to the apex of the abdomen, where it rises into an erect
coriaceous and conical nectary, wide at its base and tapering to its summit. ‘This is
perforated and forms the nectary or duct for ejaculation, just as in Aphis.

The wing-cases or rudimentary tegmina are short, pointed, and black, with traces
of an obscure neuration. Legs very stout, rather flat, with coarse tarsi. Size
5x38 millimetres.

These pupe are probably the immature forms of an Oxyrhachis which develops
simultaneously in the month of January, at Wynberg, a suburb of Cape Town, Africa.

‘The fact that these pup have only a single horn above the metopidium, instead of the
double horn of Oxyrhachis, need present no difficulty when we consider the extra-
ordinary shapes often assumed by certain insect organs which are not really homologues,
though they may appear to be such. The legs of a larva need not he necessarily those
of the corresponding imago which emerges from it.

LITTLE-KNOWN SPECIES OF MEMBRACIDA, 337
EXPLANATION OF THE PLATES. : \
) =>\
Lad 3
PLATE 21, \gZ\ 0 By
Fig. 1. Membracis micans. The imago has a glistening and tale-like surface. Size 9x6 millimetres. ~ 4 P ¢ A SSS
Ne &y

wo

Fig. 2. Membracis veryens. <A large species remarkable for its dark hue and its broad ochreous streak.
Size 12 x 6 millimetres.
Fig. 3. Hamma nodosa. Small, pronotum rugose and contorted into knots.
3a. The frons and metopidium. The sides are furnished with small spines. Size 4x2
millimetres.
Fig. 4. Microschema mucronata, Large, remarkable for its bright colour, and sharp apex to the
dorsal edge of the pronotum.
4a. Frontal view of the frons and stemmata. Size 12 x 15 millimetres.
Fig. 5. Aconophora obfuscata. The imago shows a broad foliated summit of the porrect procephalon.
5a. Front aspect of the procephalon as seen on edge. Size 11 x 4 millimetres.
Fig. 6. Tragopa triangulata. Small and robust in figure. The pronotum does not reach beyond
one-half of the tegmina.
6a. Dorsal view. The pronotum forms an irregular triangle. The head is below the metopidium,
which last has two lateral ear-like processes. Size 4x 4 millimetres.
Fig. 7. Poppea succinea. This semitransparent insect has the pronotum more or less inflated into
coriaceous bubbles which have acute points. 'Tegmina hyaline and blistered on the surface.
7 a. The dorsal aspect of the insect. Size 9x3 millimetres.
Fig. 8. Electrophina pacificata. The long wings and the extended suprahumeral horns are distinctive.
8a. Frons and metopidium with horns. Size 13 x5 millimetres.
Fig. 9. Ceresa nitens. Very glistening, amber-coloured, robust.
9a. Front view of same insect, with its high metopidium and short suprahumerals. Size 9x5
millimetres.
Fig. 10. Entylia fuscodorsa. Imago with truncated summit of the pronotum, punctured with fine dots
within the sculptured carinz. Size 6 x 4 millimetres.
Fig. 11. Entylia mesta. Imago small, with greyish pubescence.
lla. Front aspect of the same showing the thin edge of the procephalon. Size 5 x3 millimetres.
Fig. 12. Hypsauchemia jugulata. This specimen from Sumatra has lost the summit of its procephalon
which, probably like the Indian species, was curved over the back. Size 8 x 9 millimetres.

PLATE 22.

Fig. 1. Ouranorthus palus. The winged insect is remarkable for the erect process proceeding from
the caudal apex of the pronotum. Though allied, it certainly is not Lamproptera capreolus
of Fairmaire.

la. The head, pronotum, and recurved suprahumeral horns of the same insect. Size
8x5 millimetres.

Fig. 2. Anchon fuscum. Allied to A. albolineatum, but it wants the white streak on the pronotum.

2a. The front aspect clearly shows that the procephalon is cleft into broad plates or foliations.
The insect is bright amber-coloured. Size 7 x 5 millimetres.
SECOND SERIES.—ZOOLOGY, VOL. IX. 49

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

ON UNDESCRIBED SPECIES OF MEMBRACID#.

3. Anchon strigatum. The dorsal process is here free and ulnate. The procephalic point probably
is cleft like that of the last-named species. Size 9 x 6 millimetres.

4, Taloipa tinctoria. The pronotum here is obtuse, and it ends posteriorly in a blunt process
shorter than the abdomen. The bright orange colour on the base of the tegmina may be
noticed.

4a. The hirsute frons with the square metopidium. Size 7 x 4 millimetres.

5. Leucothoraz villosa. Large, with truncated but short dorsal processes. The chest has a white
villous coat.

5 a. The dorsal view of the pronotum.

56. The frontal aspect of the insect. Size 12 x6 millimetres.

6. Leptocentrus impunctus. Remarkable for its long tegmina.

6a. The head and frons.

66. The venation of the tegmen of the same. Size 10 x5 millimetres.

7. Ophicentrus serpentarius. The undulating form of the dorsal process has been used generically
by Canon Fowler, and I do so tentatively, whilst the immediate cause of variation in secon-
dary organs is sub judice.

7 a. Frontal view of the insect. Size 8 x 4: millimetres.

8. Ibiceps rufipennis. The Imago. Bright shining, warm brown. ‘The grey on the tegmina as
shown by this figure is only meant to represent the light glancing on the corrugations of the
wing, and it is not due to any patches of grey colour. Its long pronotal horn is smooth,
instead of rough as represented on the Plate.

8a. The head and front view of the pronotum. Size 8x 4 millimetres.

9. Polocentrus caudatus. The winged insect is somewhat remarkable from its short posterior
horn, which, like the rest of this genus, is serrated below. Legs spatulate.

9a. Front view with the short suprahumerals. Size 8 x5 millimetres.

10. Polocentrus labatus. This insect is from Abyssinia, and has the characteristic serrated posterior
horn.

10a. The front view with the dark frons and pointed suprahumerals.

10%. The tegmen with its dark fuscous venation which encloses the chief ochreous cellules. Size
8x4 millimetres.

11. Trapezoida hirsuta. This insect from Central America has a square metopidium which is
strongly hirsute. The dorsal view of the pronotum shows something of a lozenge shape.

lla. Head and frons of the same.

11. Dorsal view of the insect. Size 7 x 4 millimetres.

12. Larva of a Membracid, not uncommon in the neighbourhood of Cape Town, and at Wynberg,
S. Africa, but the winged insect has not yet been determined. The erect caudal nectary
discharges a liquid, probably of a saccharine nature, and gives the insect a grotesque appear-
ance. Probably it is visited by Ants, as is known to be the case with the larva of Centrotus
nectaris of Ceylon. Size 5 x38 millimetres.

Buckton Trans. Linn. Soc.Ser.2.Zoor Vou.I[X PLZ

G.B.Buckton del West, Newman chror

SPECIES OF MEMBBACIDA:

m.c. Zoon. Von. IX P1l.22

Buckton

G.B. Buckton del West,!

a .
+
,
‘
.
“ .
7
> ry
7
«

ir *330u]

X. The Genitalia of both the Sexes in Diptera, and their Relation to the Armature
of the Mouth. By Waurer Wescut, F.R.IS. (Communicated by JOHN
Hopkinson, Esq., F.L.S.)

(Plates 23-30.)

Read 21st June, 1906.

WHEN, five years ago, I made some dissections of the genitalia of the males in
Scatophaga lutaria, Fabr.*, and S. stercoraria, Linn., I was much hampered by the
absence of any systematic nomenclature. Huxley, in his ‘Anatomy of Invertebrated
Animals,’ passed by the genitalia of the male cockroach with a mere allusion to the
complexity of the parts Tf.

L. Dufour ¢ in his, at all events, comprehensive review of the genitalia in Diptera,
never attempts to grapple with the separate parts, contenting himself by saying:
« T/armature copulatrice, receptacle de la verge, est une machine des plus compliqués,
destinée & se porter hors un corps lors de union des sexes. Les nombreuses piéces
plus ou moins symétriques, cornées ou coriacées qui la composent, combinent leur action,
soit entre elles, soit avec les organes externes de la femelle pour consommer V’acte de la
fécondation.”

Packard is no help, either in his larger works, or in his small paper on the homologies
of the ovipositor and the homologous parts in the male insect §.

Kirby and Spence classify the ovipositors, giving them some very unwieldy names,
but make no analysis of the parts of the male ||; an omission to be regretted, as they
are, at the least, of equal interest, and in their way of equal symmetry and beauty, with
the armature of the mouth-parts, though more difficult to dissect out.

But in a paper on the genus Phronia of the Mycetophilide by A. Dziedzicki] I
found a scheme of nomenclature for the external valves of the genitalia of both sexes,
but not for the complicated parts that are attached to the penis of the male, that
combination being called the “appendix interna” or “ adminiculum.” The only careful
and complete study of this part exists in Dr. B. Thompson Lowne’s monograph on the
Blow-fly ** ; he has named all the parts as they exist in this insect, and given admirable
figures of the male armature, drawn on a large scale, so as to be easy of comprehension.
So with the material drawn from Dziedzicki, from Lowne, and a small contribution
from Miall and Denny’s ‘ Cockroach,’ and my own resources, I am able to formulate a

* Journal of Quekett Micr. Club, April 1903. + p. 350 (1877 edition).
+ “Recherches anatomiques et physiologiques sur les Diptéres,” 1851, Mém. Prés. Ac. Sci. Paris, tome xi.
p. 198.

§ Proc. Boston Soc. Nat. Hist., xi. 1868.
|| Description of plates 15 & 16, 1828 edition.
{ Hor. Soc. Entom. Ross., tome xxiii, 1889.
** «The Anatomy, Physiology, Morphology, and Development of the Blow-fly,’ 1895.

SECOND SERIES.—ZOOLOGY, VOL, IX. 50

340 MR. WALTER WESCHE ON THE GENITALIA OF

complete nomenclature for the genitalia of Diptera. Lately, F. du Rosille* has
demonstrated the value of a study of these organs, by distinguishing a number of species
in the difficult Savcophaga genus that have previously passed as S. carnaria (Linn.),
by a comparison of the male armature.

There are two papers on the male genitalia of Lepidoptera by F. Buchanan White,
M.D. +, and P. H. Gosse{; but these works are mainly engaged with the characters of
the outer and visible parts, with the view of determining species. There is no attempt
to dissect out the interior or to classify the armature; but, nevertheless, the papers have
value, as they show, particularly in the beautiful drawings of Gosse, the endless variety
of shape and contrivance that may characterize any portion of the male armature.

The male genitalia in Diptera always consist of a central tube, surrounded by more or
less complicated appendages. All these parts are extremely variable in shape and
in their relations to each other, but it will be seen that they homologize, though they
undergo very remarkable transformations. For convenience they may be divided into
four divisions :—(1) the posterior external valves; (2) the anterior external valves ;
(3) the penis and its appendages; (4) the interior organs.

(1) Posterior external valves.—The posterior external valves consist of the (a) Forcipes
infericres and the () Forcipes superiores.

(2) Anterior external valves —The anterior external valves are the (¢) Laminez
superiores.

(3) Penis and appendages.—In the third section are a number of organs beginning
with (d) the part of the penis, usually a hyaline membrane, which is the orifice and
continuation of the ejaculatory duct; (e) the theca or sheath, which has ramifications
called by Lowne (f) the Paraphallus and (g) the Hypophallus; (4) the Spinus titil-
latorius, (7) the Forcipes interiores, (4) the Palpi genitalium, and (m) the Apodemes of
the penis, often called, for the sake of distinction, the double apodeme.

(4) Interior organs—The interior organs consist of (7) the Ductus ejaculatorius,
(0) the Sacculus ejaculatorius, (p) the Ejaculatory apodeme, (7) the Vas deferens, (s) the
Paragonia, (¢) the Vasa efferentia, and (a) the Testes.

(a) FORCIPES INFERIORES.

Structure.—These are a pair of more or less complicated hooks. They are on the
dorsal side of the abdomen in the Muscidee, but are laterally placed on the sides of
paired processes, the forcipes superiores, which are even more dorsal still, being on the
sides of the median line. Occasionally they have another hook articulated on to them
and are sometimes hairy, but generally they are structurally strongly chitinized, with
setze and spines disposed in contact-areas.

Nomencluture.—Vhey correspond with the valvulee externze of Lowne; they are also
known as zygapophyses or claspers.

* Mém. Soc. Linnéenne du Nord de la France, Amiens, 1905.

+ “On the Male Genital Armature in the European Rhopalocera,” Trans. Linn. Soc., ser. II. Zool. vol. i. (1876).

+ “On the Clasping-Organs ancillary to Generation in certain Groups of the Lepidoptera,” Trans. Linn. Soc.,
ser. II. Zool. vol. ii. (1882).

BOTH THE SEXES IN DIPTERA. 341

Cecidomyiide and Mycetophilide—They are clearly present in the Cecidomyiidee and
the Mycetophilide. In the genus Sciara they are quite valvular in appearance, and are
on the ventral side of the abdomen in their true inferior position (Pl. 23. fig. 1).

Bibio.—They are simple in Bibio and Scatopse (PI. 23. figs. 5, 7).

Tipulide.—In the Tipulidee they are remarkably developed. In Tipula oleracea,
Linn., the part consists of two hooks and a plate, forming an extraordinarily complicated
process (Pl, 24, fig. 25); the plate has a sense-organ, consisting of a number of socket-
like depressions and enclosed in a chitinous ring. In each socket is a membranous bulb
(Pl. 24. figs. 27, 28).

Rhyphus.—The part is complicated and strongly chitinized in R. fenestralis, Scop.,
hairy and smaller than the f. superioresin R. punctatus, Fabr.

Stratiomytide.—In the Stratiomyiide, where the genitalia of the two sexes are
exteriorly much alike, these parts are simple and quite on the ventral side; this is
markedly so in Chloromyia formosa, Scop., and Michrochrysa polita (Linn.).

Tabanus.—In Tabanus bromius, Linn., two pairs of organs are found, and those on
the ventral side I homologize with this part; they are normally bent at a right angle to
the median line, an unusual position (Pl. 29. fig. 109).

Asilide—tIn the Asilide they are smaller than the dorsal processes. In an un-
determined Australian Asilid the part has a hook articulated to it (Pl. 24. fig. 36).

Empide.—tIn the Empid these parts are absent in F. stercorea, Linn., but are
represented as relatively small in most of the genera.

Doliochopodide.—In the Dolichopodidee, where the “ so-called fan ” is well-developed,
this part appears to consist of two hyaline plates under the hypopygium (PI. 25, fig. 43).

Syrphide.—They vary much in form in the Syrphidze, but are not complicated ; they
are usually simpler and much less developed than those in Syritta pipiens (Linn.)
(Pl. 25. fig. 54).

Muscide.—In the whole of the Muscide they are prominent on the dorsal side, and
form useful specific characters ; in Glossina at least four species can be differentiated
from these organs alone (Pl. 26. figs. 61, 65; Pl. 27. figs. 76, 77, 79, 80).

Muscide Acalyptrate.—They are also present in those families where, the female
having developed a telescopic horny ovipositor, the penis has been modified into a long
ribbon-like structure (Pl. 28. fig. 96).

Homology with the ovipositor.—In the majority of flies these organs are on the eighth
segment of the abdomen, and correspond with the ventral egg-guides or valves of the
ovipositor, also on the eighth segment of the abdomen. This is very clear in Chloromyia
formosa (Scop.). From their position, as Lowne points out *, they “ correspond with the
inferior blades of the ovipositor in Locusta,” or with the ventral valves in Zipula
oleracea +.

* Note——In my two earlier papers on Scatophaga and Glossina I have called these parts the Forcipes superiores,
transposing the adjectives. As my studies had been confined to the Muscidw, where these hooks are always more
prominent, important, and articulated on the dorsal side of the hypopygium, this seemed the more reasonable
nomenclature. Study of the other families and of the ovipositor has enabled me definitely to settle the homology
of the part with that formulated by Dziedzicki.

+ ‘Blow-fly,’ p. 732.
50*

342 MR. WALTER WESCHE ON THE GENITALIA OF

(6) FoRCIPES SUPERIORES or VALVULE INTERN.

Structure.—These so-called hooks are mostly (exceptions will be found in some
Culicide and Tipulidee) valves, which are generally covered with hair, or bulbous and
palp-like in appearance, suggesting a sensory function.

Nomenclature.—Lowne is responsible for calling them valves, Dziedzicki for the first
name.

Mycetophilide.—In Phronia and Sciara they are hairy bulbs (PI. 28. fig. 1).

Bibionide.—in Scatopse notata (Linn.) they are large plates, quite overshadowing the
usually more prominent forcipes inferiores (Pl. 28. fig. 7).

Culicide.—In the Culicidee they are very elaborate; in Culex pipiens they are large
and are fitted with modifications of setee, several knife-like and one leaf-like, and have
besides single large hooks articulated on to their extremities (Pl. 23. fig. 16). Anopheles
cinereus has also a hook articulated to the forceps, but isin other respects simpler (PI. 23.
fig. 14).

Tipulide.—In the Tipulidee they consist of large hairy plates as in Tipula oleracea
(Pl. 24. fig. 22), of plates with an articulated hook as in Gynoplista bella, Westwood
(Pl. 23. fig. 19), or a more foliaceous form as in Ptychoptera scutellaris, Meigen (Pl. 28.

=r)
ge

7):
Stratiomyiide.—In the Stratiomyiide they are hairy and thinner in proportion to
their length than in Sciara.

Tabanide.—They are represented by two hairy bulbs in Tabanus bromius, Linn.
(Pl. 29. fig. 109).

Asilide.—They are relatively very large in the Asilidze (Pl. 24. fig. 37).

Empide.—A very marked form is seen in Empis stercorea, Linn. As arule it is much
less developed in this family (Pl. 25. fig. 40).

Dolichopodide.—Of all the families in Diptera they are in most extraordinary
development in the Dolichopodidee. They form the fans that have earned this tribe
the popular name of the “ fan-tailed flies.” They are in their greatest relative size in
the genera Dolichopus and Pecilobothrus (Pl. 25. fig. 43).

Syrphide.—They are variable, but never very prominent, in the Syrphide. In
Eristalis tenax (Linn.) they are represented by two small hairy processes between the
larger forcipes inferiores, which in this case are much like the f. superiores in shape and
structure; they are more developed in Syrphus pipiens, but still smaller than the
f. inferiores (Pl. 25. fig. 55).

Muscide.—They are hairy valves in the Muscidx, which have the anus at their base
and close over and protect the penis. Sometimes they are fused as in Rutilia splendida
and Olivieria lateralis, Fabr. (Pl. 30. fig. 120), or absent as in Glossina *, where they
are represented by two rows of hair on either side of the median line.

Cordyluride.—In Norellia spinimana, Fallén, there is rather a characteristic armature
(Pl. 28. fig. 91).

Use and Homology.—This part seems to act either as a protection to the genitalia or

* « Genitalia of G. palpalis,” W. Wesché in Journ, Quekett Mier. Club, no. 57 (Nov. 1905).

—

BOTH THE SEXES IN DIPTERA. 345

as a sensory organ. It homologizes with the superior valves of the ovipositor in Tipula
or with the paired valves in Musca, possibly with Huxley’s “podical plate” in
Periplaneta orientalis, Linn.

(c) LAMINA® SUPERIORES,

Structure.—This part consists of paired processes, often of a separate plate on the
anterior edge of the cavity containing the genitalia. Structurally it is strongly chitinized
and often with hooked bristles, sometimes with bunches of hair; so far as my observa-
tions have gone, the plate is always without spiracles.

Nomenclature.—It is so called by Dziedzicki, who also names the segment on which
the part rests the “‘ Lamina basalis.”

Mycetophilide.—It is well-marked in the genus Phronia, but seems absent in other |~

genera of the Mycetophilidx, nor can it be differentiated in the Culicide.

Tipulide.—It is obvious in Tipula oleracea, where the ventral edge of the abdomen
opposite the forcipes is studded with a number of tubercles; but it does not seem
developed in the Stratiomyiide, Tabanidie, Asilidee, Bombylidee, Empidie, Dolichopodidee,
or the Syrphide (Pl. 24. fig. 24).

Muscide.—In most of the Muscidz it is very evident and varies with species. In
Echinomyia fera (Linn.) and Phyto melanocephala, Meig., small plates with anterior clefts
are found. In Sarcophaga it is very marked, often having long hairy lateral processes
on the anterior edge, though in the nearly related Sarcophila latifrons, Fall., it is hardly
perceptible. In Musca domestica, Linu., two marked lateral processes project, but in
Stomozys calcitrans (Linn.) and Pollenia rudis, Fabr., only cleft hairy plates represent
the part. It can be differentiated, but is not prominent, in Calliphora erythro-
cephala, Meig.

In Glossina, where the whole of the last segment is turned in under the abdomen,
the forcipes inferiores rest against an arched horny band which represents this part ;
but in G. morsitans, Westw., where the whole ventral surface consists of a coriaceous
membrane without any chitinous plates, there is in the usual situation of this part a
lunule of chitin with strong short bristles thickly spread over it; this is absent in
G. palpalis, Des.

It has soft-haired, bulbous tubercles, small in size, in Morellia hortorum, Fall.,
and Hyetodesia obscurata, Meig. In Hydrotea, where the legs have many contrivances
for holding tie female, it is but little developed. ‘lhe most striking elaboration of the
part that I am acquainted with is on the abdomen of a small undetermined fly of the
genus Anthomyia that was common inside the houses in Maryborough, Queensland; on
this insect are two pectinated arms, articulated on to the segment opposite the genitalia
(Pl. 26. fig. 66).

Sepside.—In an undetermined species of Nemopoda from Jersey this part is much
developed ; on it are two pairs of hairy prominences, a pair of membranous tubercles,
somewhat like palpi in structure, and a pair of barbs (Pl. 29. fig. 103).

Cordyluride.—tIn Scatophaga litorea, Fall., and Norellia spinimana are characteristic
developments of this part (Pl. 27. fig. 87; Pl. 28. fig. 97).

344 MR. WALTER WESCHE ON THE GENITALIA OF

Use—From the situation of the part, the number of spines and hooks usually
found on the organ, it is obviously used in holding the female.

Homologies with ovipositor.—It appears to homologize with the ventral portion of the
fourth segment of the abdomen of the female, counting the last segment of the ovipositor,
that bearing the valves or egg-guides, as the first. Properly speaking, the first segment
of the abdomen is that next the thorax, but, as the number of the segments is variable,
the only way of comparing the genitalia of the sexes with exactness is to count that
bearing the valves as the first segment.

(d) Tar EXTREMITY OF THE PENIS.

Structure.—This part has to be differentiated from the theca or cover, therefore for
the present purpose I shall regard it as the orifice of the ejaculatory duct. It is
mostly a delicate flexible hyaline membrane with characteristic triangular, more or less
chitinous plates often forming part of the structure, or it may be a free, transparent
chitinous tube as in the Tipulide.

Mycetophilide —In Sciara thome (Linn.) it is difficult to make out, but appears to
be a structureless hyaline stile (PI. 23. fig. 1).

Bibionide.—In Bibio hortulanus (Linn.) and B. marci (Linn.) it appears as a plate,
with the edges bent over, which do not meet but are covered by a delicate membrane.
It has the appearance of the organ as seen in some Coleoptera (Vermestes). In Seatopse
notata it is membranous, and with the character of the whole organ approximating
to that in the Muscidee in the external, and to that of the Tipulide in the internal
organs (Pl. 23. fig. 7).

Culicide and Tipulide.—I have failed in trying to dissect out this part in the Culiemt e
as well as in the Tipulid Gynoplistia bella, but a comparison of these parts with
those of Tipula oleracea and Pachyrrhina maculata, Meig., easily supplies the lacune.
In these insects a long tube or flagellum proceeds, bending in a circle in the process,
from the ejaculatory sac, of which it forms a part and continuation, to an organ which is
evidently the theca, passing longitudinally through a passage in it, and working quite
freely in the passage; it is capable of extrusion and retraction. It is contained in
a membranous envelope, noticed by Dufour, which has been thought to be the containing
membrane of the spermatozoa, but, as I shall show later, this does not agree with my
explanation of the working of the apparatus (Pl. 28. fig. 15; Pl. 24. figs. 23, 24, 29).

Rihyphide.—\n Rhyphus fenestralis this part in the interior presents the appearance
of a long tangled thread, and it is similar in 2. punctatus.

Stratiomyiide.—In Beris vallata (Forster) it is a hyaline tube, less stiliform in shape ;
springing from the same base, it has on either side two knife-like blades with serrated
edges at the extremities (Pl. 24. figs. 32, 33).

Tabanide.—In Tabanus bromius two slender stiles seem capable of extrusion through
the theca; they are enveloped in a delicate membrane, much covered with fine blunt
setee (Pl. 29. fig. 109).

Asilide.—In the Asilidze an arrangement is found which is somewhat similar to that

BOTH THE SEXES IN DIPTERA. 845

of Beris, but the stile, obvious in that genus, is quite absorbed by the chitinous theca
(Pl. 24. fig. 38).

LEmpide.—Nearly the same form is found in the Empidee, though here the tube is
again quite stiliform. Hmpis stercorea, Linn., is the simplest combination that I am
acquainted with; the part is readily seen on the pinned insect as a long stile. On a
prepared specimen the junction with the theca is seen even with the lower powers of the
microscope (PI. 25. fig. 39). The part is practically the same in Hilara cilipes, Meig.,
but with more elaborate surroundings (PI. 25. fig. 41). In the ovipositor of the female
is a curious notched process which may possibly act as a guide to the delicate flagellum
(PI. 25. fig. 42).

Dolichopodide.—In spite of the greater development of the holding organs, the
structure of the penis in the Dolichopodid shows a close relationship to the Stratio-
myiide, the Asilidee, and the HEmpide. ‘The hyaline stiliform tube has developed
processes, presumably of use in coitus, serrations as in Dolichopus griseipennis, Stannius,
and a leaf-shaped excrescence in D. nobdilitatus (Linn.) (Pl. 25. figs. 43, 44, 49, 50).

Syrphide.—In the Syrphidee the part is often a flexible membrane, as in Catabomba
pyrastri, Linn. (Pl. 25. fig. 53). In Hristalis tenax (Linn.) and Syritta pipiens (Linn.)
the ejaculatory duct opens into a horny box situated at or near the extremity of the
theca (Pl. 26. figs. 56, 57, 64). In Spherophoria scripta (Linn.) it consists of a
membrane studded with the usual triangular plates; this is capable of inflation, and
when in that condition takes a tricuspid shape.

Muscide.—In Sarcophaga carnaria (Linn.) and the majority of the Muscidee, calyptrate
and acalyptrate, it is a hyaline flexible membrane. In Glossina tachinoides, Westw.,
this part has membranous wings, studded with minute plates as in S. seripta, and
capable of inflation or erection, as I have demonstrated by alternately raising and
lowering the cover-glass.

Acalyptrate.—In the Ortalidee, Trypetidee, and Lonchzeidee it is a long membranous
tube, normally rolled on itself as a watch-spring is coiled, and when extruded not unlike
an ovipositor in appearance (Pl. 28. fig. 96).

Ephydride.—A very peculiar shape is found in Notiphila nigricornis, Stenhammar,
where a membrane, studded with aculeations of various size, rises from a chitinous
theca (Pl. 28. fig. 98). The part in Parydra coarctata, Fall., is much like that in
Bibio hortulanus. This small fly has several peculiarities about the structure of the
mouth-parts that suggest that it is of very archaic type (Pl. 29. figs. 110, 111).

. Homology with ovipositor.—This part appears to be homologous with the membranous
structure of the ovipositor. This hypothesis is strengthened by the fact that in the
Ortalidze the penis has no chitinous parts except at the extremity.

(e) Tae TuEca or PENIS-SHEATH.

Structure.—This is the chitinous case that protects the ejaculatory duct; it is always
highly chitinized, often with an anterior suture. It may be quite simple or consist of a
number of plates. When appendages are present it invariably supports them, whether
articulated or not. No socketed hairs or bristles have been noticed, unless they

346 MR. WALTER WESCHE ON THE GENITALIA OF

represent an aborted organ, as in Oliviera lateralis; but it often develops hooks, or
processes like the paraphalli, without any articulation or break in the structure.

Bibionide.—In Bibio hortulanus and B. marci it is a ribbon-like process supporting
the appendages and surrounding the curious plate which forms the penis, and which
itself is a part of the theca (Pl. 23. fig. 4). In Scatopse notata it supports the membrane
of the penis (Pl. 23. fig. 7).

Culicide.—In Culex pipiens, Linn., and an undetermined species, it is a membrane
supported by lateral chitinous processes, an arrangement similar to that found in
certain Tipulide (Pl. 23. figs. 18, 15; Pl. 24. fig. 29).

Tipulide.—In Tipula oleracea it is a very clearly differentiated part, supporting a pair
of appendages and fused at the base with the double apodeme. The penis works freely
through it, and there is also a central rod, which I regard as a part of the theca, which
forms a third apodeme. These organs have muscular attachments which practically
anchor the theca, as that part is not extruded, the long flagellum being pushed through
it (Pl. 24. fig. 23).

In Pachyrrhina maculosa, Meig., are two chitinous pieces that represent the atrophying
apodemes and support the theca, which is well developed and carries the appendages,
but has not the ceutral process of Tipula oleracea (Pl. 24. fig.29). In Gynoplistia bella,
Westw., an Australian insect, as well as in an undetermined British Tipulid, this part is
very extraordinary, developing numerous hooks and supported by lateral columns as in
the Culicidee (Pl. 238. fig. 15).

In Ptychoptera scutellaris, Meig., the levers appear absent, but the theca is much
developed. I have not succeeded in finding the orifice of the ejaculatory duct, but
I have had but little material for examination (Pl. 24. fig. 20).

Stratiomyiide—In Beris vallata it is attached to the last segment of the abdomen ;
it does not enclose, and is not soldered to, the penis, but appears to act as a grooved
channel to guide it (Pl. 24. fig. 33).

Tabanide.—In Tabanus bromius the theca is quite similar to that of Pachyrrhina
maculosa, but is articulated to the double apodemes, which are well-developed (Pl. 29.
fig. 109). In Pangonia longirostris the part has much the same form.

Asilide.—In the Asilide is strongly chitinized, and broadening out at its bases
contains the ejaculatory sac (Pl. 24. fig. 38).

Empide.—tn Empis stercorea the theca has developed paired barbs, which are probably
homologous with the paraphallus and the hypophallus (Pl. 25. fig. 39). In Hilara cilipes
a number of barbed serrations cover a very pronounced organ; the penis is held in
a neat appliance, through which it is extruded or retracted (Pl. 25. fig. 411).

Dolichopodide.—In Dolichopus plumipes, Scop., and other species of this genus, the
whole segment is fused with the theca, though a suture can be detected (Pl. 25. fig. 43).
In D. festivus, Haliday, paired barbs are found; in D. griseipennis only a single
pair appears (Pl. 25. figs. 44, 50).

Syrphide.—tIn Catabomba pyrastri an arrangement is found not unlike that in
Bibio hortulanus; the theca surrounds the penis with an enclosing wall, which supports
the appendages (Pl, 25. fig. 53). In Eristalis tenax and Syritta pipens this part is

BOTH THE SEXES IN DIPTERA. 347

much enlarged and has developed levers at the base, by means of which it can be rotated
(Pl. 25. figs. 56, 57, 58).

Muscide.—In many of the Muscidz the theca not only forms a guard to the tube,
but is also a platform which is fused with and supports the part called by Lowne the
bulb * (the lower part of the chitinous process which sustains the ejaculatory orifice),
the appendages (the spine, palpi, and hooks), as well as the paraphallus and hypophallus
(Pl. 26. figs. 62, 68). In Glossina the theca forms a wall round the double apodeme
(Pl. 27. figs. 78, 84).

Loncheide.—A very beautiful adaptation of this part is found in Toxoneura muliebris,
Harris. This insect has a long ciliated penis with considerably more chitinous structure
than Ulidia nigripennis, Meig. This penis at its base passes through a plate which bears
the usual appendages, and is obviously the theca (PI. 30. fig. 122). The same structure can
be made out in Lonchea nigrimana, Meig. In Palloptera ustulata, Fall., a remarkable
chitinous process is found at the extremity of the ciliated ribbon which forms the male
organ; a series of folds enclose the tube; from this rises a small column carrying the
ejaculatory duct; attached to the folds is a single relatively large plate formed into a
hook; this appears to be the upper part of the theca, much separated from its base,
which is, as in 7. muliebris, a chitinous ring bearing some appendages.

Ephydride—tin Parydra coarctata two rings of chitin support the part carrying the
hyaline membrane of the duct; the larger ring bears two appendages (PI. 29. fig. 111).

Homology with ovipositor—The homology of this part with any distinct part of the
ovipositor is not obvious, but possibly it may be represented in Musca domestica by the
rods on the first two segments, counting, as before, the segment bearing the egg-guides
as the first.

(f) THE PARAPHALLI.

Structure ——There are two lateral rods springing from the back part of the theca and
coming forward. They are always strongly chitinized, and characterized in the genus
Musca by minute transverse serrations on the extremities.

Nomenclature.—They were first pointed out, and are named, by Lowne.

Empide and Dolichopodide.—They are only markedly present in the Muscide; and
we seem but to get a hint of their existence in the Empide and Dolichopodide,
where the barbs on the theca already alluded to are probably homologous (Pl. 25.
figs. 39 & 44).

Syrphide.—In Syritta pipiens the theca is hollowed out at the point, and in this half
segment of a circle are one or two curious processes which may, or may not, represent
these parts (Pl. 26. fig. 56).

In Eristalis tenax the theca is continued up to two serrated blades, which appear to
be the paraphalli (Pl. 26. fig. 58).

Muscide—Iin Sarcophaga these parts are represented by two plates springing
laterally trom the theca above the bulb, quite unlike the appearance of the parts in

* ¢Blow-fly,’ p. 738.
SECOND SERIES.—ZOOLOGY, VOL. IX. 51

348 MR. WALTER WESCHE ON THE GENITALIA OF

Musca (PI). 26. fig. 60). They are well developed and characteristic in Pollenia rudis,
Fabr., and some species of Lucilia (Pl. 26. fig. 62).

They were described from Calliphora erythrocephala, Meig., where the serrations on
the extremities are very marked (Pl. 26. figs. 68, 71). In Glossina the paraphalli,
together with part of the theca, form a framework which supports an elaborate sensory
apparatus and protects the orifice of the ejaculatory duct. The serrated structures at
the ends of these parts are still to be found at the extremities of the parts in Glossina
palpalis, Des. (Pl. 26. figs. 63, 69).

They are present in Anthomyia radicum (Linn.), with curious processes at the
extremities (Pl. 27. fig. 85).

Sepside.—In the Sepside there is an aculeated membrane which springs out of the
lower part of the theca, which may be homologous with these parts (Pl. 28. fig. 99;
Pl. 29. fig. 101).

Use.—The use of these various modifications will obviously vary with the type. In
Oalliphora and Pollenia the shape and the broadened serrated ends, to prevent the
extremities slipping, suggest some application where their elasticity would come into play.
As to homology with the ovipositor, they may be treated as outgrowths of the theca.

(g) Tue HyPoPHALLUs.

Structure and Nomenclature.—Lowne has so-named the excrescences of the theca
which project from the front of that part, the long paraphalli starting from the back.
It is a paired organ, open in front, but this is only to be seen by a very careful and
difficult preparation of the part. It is strongly chitinized, but sometimes shading to a
quite hyaline appearance, and usually has a wrinkled surface with more or less minute
barbs.

Empide and Dolichopodide.—What I have said with regard to the presence of
homologies of the paraphalli in the Empide and Dolichopodide will also apply to this
part.

Muscide.—In the Muscide it is generally present; it is evident in Sarcophaga
carnaria as the upper of the two anterior processes, cup-shaped in Pollenia rudis.
Seen from the front of the organ in Calliphora, it has the appearance of lateral plates
(Pl. 26. figs. 62, 68). In Anthomyia radicum it is well-marked, also in most of the
Cordyluridee (Pl. 27. fig. 85; Pl. 28. fig. 88).

Sepside.—It is to be seen in Sepsis cynipsea (Pl. 28. fig. 99), but it is not so
obvious in the Nemopoda that is figured (Tl. 29. fig. 101).

(h) SPINUS TITILLATORIUS.

Structure—This part is a single unpaired organ, which lends it distinction, that
otherwise it might not merit, as it is often absent, even in those families where it may
be found in many species. It is situated immediately behind the penis. It is
occasionally articulated, at other times it springs from the base of the theca without any
suture. It is highly chitinized, but has a tendency to become membranous at the
extremity.

BOTH THE SEXES IN DIPTERA. 349

Nomenclature —Lowne calls it “the spine”; I observed it in Scatophaga lutaria,
Fabr., as an articulated organ, and thought it might be homologous with the “ titillator ”
in Periplaneta orientalis, the hook which Lowne suggests is used in the transference of
spermatophores * from the male to the female organisms.

Dolichopodide.—I cannot find any trace of this part in the families of Diptera till I
come to the Dolichopodide ; there it is large and well developed, and in company with
the other appendages to the theca (Pl. 25. fig. 45).

My identification of this part as the unpaired lancet under the hypopygium of
Dolichopus is confirmed by the observations of Mr. Robert E. Snodgrass in his paper on
the genitalia of that family f.

A number of species show the spinus in its usual position as shown in Pl. 25. fig. 438,
though the part varies much in shape. But when the appendages are absent from this
situation, as in Psélopa sipho, the penis is surrounded by the theca, which supports two
leaf-like lateral processes, and between them a single median stile which I recognize as
the spinus.

Muscide.—It is absent in Sarcophaga carnaria and Oliviera lateralis (Pl. 26. fig. 60;
Pl. 29. fig. 112), but it is generally present in the Muscide (PI. 26. fig. 68; Pl. 27. fig. 85).

Loncheide.—One of the most curious migrations of this part is in the Loncheide.
In the Dolichopodidze we have found it near the base and at the extremity of the penis,
and also in Toxoneura muliebris, Harr., it appears to be at the base of the long ciliated
ribbon which forms that organ (Pl. 30. fig. 122), while in Palloptera ustulata, Fall., it
appears as a relatively large hook, symmetrical in shape and articulated to the rather
complicated chitinous part which forms the extremity of the penis, which in this
species also consists of a long ciliated ribbon.

Use.—In use it fits into a fold of the ovipositor, as can be seen in Berlesi’s remarkable
section of Musca domestica “in copula”? ¢.

Homology.—\t is obvious that its homology in the ovipositor must be looked for after
the determination of the homology of the theca.

(¢) FORCIPES INTERIORES.

Structure——These are two small blade-shaped hooks on both sides of the theca,
generally articulated; they are more or less highly chitinized, with very short sharp
spines springing from sockets ; the surface is usually smooth.

Nomenclature—Lowne calls them “ posterior gonapophysis,” but this is obviously
indefinite, and I prefer my name of “forceps interior,” as not only more definite, but
descriptive of the part.

Mycetophilide.—In Sciara thome they are probably the paired blades springing from
the root of the adminiculum (PI. 23. fig. 1).

Bibionide.—In Bibio hortulanus they are difficult to differentiate, but I think are

* « Blow-fly,’ p. 664.
+ “ Hypopygium of the Dolichopodide,” Sept. 28, 1904. Proc. Cal. Acad. Science, ser. 3, vol. iii. Zool. no. 11.
+ Riv. Patol. Vegetale, ix. (1902),

Gl*

350 MR. WALTER WESCHE ON THE GENITALIA OF

represented by the free, pointed, posterior processes (Pl. 23. fig. 4). They are quite
evident in Scatopse notata, which has every part complete with the exception of the
“spinus tittilatorius” (Pl. 23. fig. 7).

Culicide and Tipulide.—In the Culicidee and the Tipulidee they are represented, but it
is difficult to say which are forcipes and which are palpi (PI. 28. figs. 18, 15 ; Pl. 24. fig. 23).

Stratiomyiide.—In Beris vallata they appear to be the paired blades springing from
the root of the penis (Pl. 24. figs. 32, 33).

Dolichopodide.—In the Dolichopodide they can be differentiated from the palpi by
their structure (Pl. 25. fig. 46).

Syrphide.—In Catabomba pyrastri they take the form of the curious band which
encircles the penis (Pl. 25. fig. 53). In Syritta pipiens they are also much modified and
changed into the two strongly chitinized pectinated organs that are on the head of the
theca (Pl. 26. figs. 56, 57). In Eristalis tenax these organs are represented by two
curious hooks placed laterally on each side of a chitinous receptacle, which is the orifice
of the ejaculatory duct (Pl. 26. figs. 58, 59).

Muscide.—In the Muscidee they are generally present, but not in so well-marked a
type as in Calliphora, as may be seen by comparing that with Sarcophaga (Pl. 26.
figs. 60, 68).

In Glossina palpalis and G. tachinoides they are fused with the “ palpi genitalium”
(Pl. 27. fig. 86).

Ortalide and Loncheide.—They seem absent in Ulidia nigripennis, but are repre-
sented in Lonchea nigrimana, Meig., which has a similar form of penis.

Ephydride.—They are represented in Parydra coarctata and Notiphila nigricornis
(PL 28. fie. 985 Pl. 29. fig. 111).

Homology with ovipositor.—Vhere are in the ovipositor of Musca domestica eight single
rods and one double rod, fused at the base—besides this, seven or eight transverse setose
plates; and the homology of this part must be sought amongst this rather confusing
array *.

(k) PALPI GENITALIUM.

Structure.—These are two small, palpiform organs, like the forcipes interiores situated
on both sides of the theca but placed anteriorly to them. Sensory hairs and sete are
usually present.

Nomenclature.—Lowne calls the part the “anterior gonapophysis ” +, but I prefer the

* Note——In the Culicidee I have encountered a special difficulty in homologizing this part, as there are four
interior hooks ; these are placed symmetrically in front of the membrane supported by the lateral processes (Pl. 23.
fig. 13). These lateral supports are probably the palpi genitalium, and, if that is so, there must be a double pair
of forcipes interiores. Those figured as 10 and 12 represent a single pair.

In Dolichopus festivus there are apparently two pairs of palpi next each other; they can be differentiated by a
curious venation on the head of one pair (Pl. 25. figs. 47 & 48). This is a similar difficulty to the previous one, and
one which is not easy of explanation. There are many instances of failure of parts, but these two families are the
only ones which show a redundance. Mr. Snodgrass, in his paper already alluded to, shows some complicated genital
palpi; D. erenatus has several processes at its extremity, and this may be owing to a fusion of the double organ.

+ ‘Blow-fly,’ p. 740.

BOTH THE SEXES IN DIPTERA. 351

above name for the same reasons as in the case of the forcipes interiores. They are
only anterior so far as regards the penis itself—as regards the insect, they are often, as in
Dolichopus in posterior positions.

Bibionide.—In Bibio hortulanus I think these organs are represented by the bands
supporting the penis (Pl. 23. fig. 4). In Scalopse notata their extremities are columnar,
springing from a broader base; the column is capped by a bunch of sensory hairs, and
there is a seta lower down (PI. 23. fig. 6).

Tipulide.—\t is impossible with the facts at my command to say definitely that the
lateral arms of the theca in Tipula oleracea are not these organs, but at all events
their function seems supplied by the bunches of setze on the dorsal sides of the abdomen
(Pl. 24. fig. 24). They seem absent in Ptychoptera scutellaris.

Dolichopodide.—\l cannot identify these organs among the related families till we
reach the Dolichopodide. In the genera Dolichopus and Pecilobothrus, in the
remarkable group of weapons situated between the forcipes superiores and the forcipes
inferiores, are generally a pair of smooth large-headed organs with sensory setze (Pl. 25.
fig. 47) ; also a pair with a leaf-like venation, whose presence is not easy to account for
(Pl. 25. fig. 48).

Syrphide.—In Catabomba pyrastri they are two hairy plates articulated to the theca
(Pl. 25. fig. 53). In Syritta pipiens they seem atrophying, being represented by two
weak lateral bands, which, however, carry a sense-organ (Pl. 26. figs. 56, 57). In
Fristalis tenax they seem quite atrophied, two minute tubercles carrying scattered
setze only remaining.

Muscide.—In the Muscide they are at their greatest and most characteristic
stage, but are occasionally absent, as in Olivieria lateralis, where small tubercles and
bristles mark their site (Pl. 29. fig. 112). In Sareophaga carnaria they are developed,
but more of a hook shape, and though they are studded with hair-sockets, the hairs
appear to be absent, but the inner side of the part appears to be soft and membranous
(Pl. 26. fig. 60). In Pollenia rudis, Fabr., they are very characteristic both in shape
and pubescence (PI. 26. fig. 62). They are approximately of the same type in Calliphora
erythrocephala, in Lucilia cesar, Linn., and in Anthomyia radicum. In Glossina they
are not articulated, and are fused with the forcipes interiores in G. palpalis and
G. tachinoides (Pl. 27. fig. 86); in the former species they carry most remarkable
hairs*. In G. pallidipes, Austen, they are two broad plates with very long fine hair
arranged symmetrically on the anterior side of the central organ (Pl. 27. fig. 78).

In the Cordyluridz the parts often carry long setz as in Scatophaga litorea, Fall.
In Norellia spinimana, Fall., they seem to have exchanged places and functions as well
as shape with the forcipes (Pl. 28. fig. 89). In Fucellia fucorum, Fall., they are absent,
their places, exactly as in Olivieria, being marked by single bristles.

Loncheide.—They can be traced in L. nigrimana and Toxoneura muliebris at the base
of the penis, probably in an atrophying state (Pl. 30. fig. 122).

Sepside.—In Sepsis and Nemopoda they seem absent, though I have an ‘imperfect

* “Genitalia of the Tsetse-fly, Glossina palpalis,” Journ. Quek. Mier. Club, Noy. 1908, p, 236.

352 MR. WALTER WESCHE ON THE GENITALIA OF

preparation of S. eynipsea, Linn., which appears to show them, but unfortunately the
bases cannot be seen. They are certainly absent in many species, and the laminz
superiores have developed bunches of hair, and in at least one case (Pl. 29. fig. 108)
tubercles which appear to be sensory, to compensate.

Ephydride—tThough the forcipes interiores are present in Parydra coarctata, there
are no signs or remains of the palpi (Pl. 29. fig. 111).

Use.—These organs fulfil an important office, as when they are absent we find
compensating sense-organs. In T%pula oleracea lateral bunches of hair, in Pandora
scutellaris processes on either side of the theca, six bunches of fine black hair, and in
Nemopoda elaborated laminze superiores.

(m) THrt APODEMES OF THE PENIS, THE DOUBLE APODEME,
OR THE GREAT APODEMES.

Structure.—These are the organs (or organ) that rotate the penis, often considerably
elongating it in the process. They are found in three conditions: (1) as symmetrical
paired crgans, as in Bibio hortulanus, Gynoplistia bella, or Tabanus bromius; (2) as
partially fused together, as in Glossina or Sepsis; (8) or fused or “united in the
median line” as Lowne puts it, speaking of the part in Calliphora*.

The part has a markedly laminated structure; when it is fused, a highly chitinized
process runs longitudinally through the middle; it is fitted for the attachment of
muscles.

Nomenclature.—Lowne calls these parts the “ great apodemes”’: it is necessary to
have an adjective to distinguish them from the apodeme of the ejaculatory sac, which is
often very large and prominent; as I have found this part separate in several families I
have suggested the name of the “double apodeme” to distinguish it from the other
apedeme, which is always single.

Mycetophilide.—There are indications of the presence of such an organ in my
preparations, but not sufficiently definite to quote.

Bibionide.—In Bibio the theca is attached to two strongly chitinized levers, not
shown in the figure. In Scatopse notata the apodeme is fused. ‘These extraordinary
anomalies are quite in keeping with what I found while working on the homologies of
the mouth-parts. Bibio had a type of trophi approximating to that in the Muscide,
differing in this from nearly all the Nematocera, and in the Empide the genus Hybos
differed in arrangement from the other genera in the family f.

Culicide.—I have a preparation of Dinocerites cancer, Theobald, which shows two
powerful apodemes placed laterally, and articulating on to the two processes which
support the membrane of the penis, also one of Culex (?) which has the apodemes of
exactly the same type as Gynoplistia.

Tipulide.—In Tipula oleracea, as I understand the apparatus, the long flagellum is
pushed forward through the aperture in the theca by the rotation of the ejaculatory sac,

* «Blow-fly,’ p. 743.
+ “The Mouth-parts of the Nemocera,” Journ. Roy. Micr. Soe. 1904.

BOTH THE SEXES IN DIPTERA. 353

actuated by the apodeme of that part, and the organs under discussion, “ fallen from
their high estate,” are reduced to act as the lateral levers of the theca, to which they are
fused (Pl. 24. fig. 23). In Gynoplistia bella they are, on the contrary, of great
importance and development, attached by muscles to the complicated mechanism of the
theca, apparently to separately rotate the two central hooked processes (Pl. 23. fig. 15).
In Pachyrrhina maculosa they are represented by two rather weak chitinous plates,
laterally supporting the theca.

Strationytide, Asilide, Empide, and Dolichopodide—tIn Beris vallata, the Asilidee,
the Empide, and the Dolichopodidee these levers seem absent, or are fused to the theca,
the single ejaculatory apodeme doing all the work. In the undetermined Asilid, already
alluded to and figured, the anterior portion of the theca sends out processes for the
attachment of muscles, and possibly represents the fused apodemes (Pl. 24. fig. 38).

Syrphide.—This part is very difficult to make out in the Syrphide. It appears to be
present in Spheriophoria scripta, but seems in that case to be doing the work of the
ejaculatory apodeme, which I cannot find. There is a suture down the organ and large
lateral processes to support the theca. I cannot find it in Catabomba pyrastri, but am
not sure of its absence, as in this family, as I shall show, the part migrates in a surprising
manner. In Syritta pipiens, in the upper part of the theca, is an arched opening; this,
seen from the front, has some curious and minute structure on its floor, the orifice of
the ejaculatory duct. Examining the organ from the side, I am able to make out that
this floor is the upper part of a chitinous box; and welded to this box, and working in
the centre of the theca, is the great apodeme, quite single and extending downwards to
the top of the ejaculatory apodeme, which is very large and evident. From this central
position it may rotate the whole organ, but does not seem sufliciently powerful for that
purpose (Pl. 26. figs. 56, 57).

In Eristalis tenax there is no domed aperture, but the orifice of the duct is also
placed in a chitinous chest or box; to the front of this box is hinged the apodeme, which
rotates it and the two hooks (forcipes interiores) which are attached to it. The whole
organ (the theca) is rotated by its edges, which are rounded and thickened, and project
at the base, forming levers (Pl. 26. fig. 58).

Muscide.—In the Muscide the organ is generally at the base of the theca and is
fused for its whole length: it takes this characteristic form in Sarcophaga carnaria
(Pl. 26. fig. 60), in Olivieria lateralis (Pl. 29. fig. 112), in Pollenia rudis (Pl. 26. fig. 62),
in Calliphora erythrocephala (Pl. 26. fig. 68), in Anthomyia radicum (Pl. 27. fig. 85), in
Norellia spinimana (Pl. 28. fig. 89) and Scatophaga litorea (Pl. 28. fig. 88), besides the
great majority of the other species.

In Glossina there is another remarkable metamorphosis. It has become forked at its
upper part as in G. palpalis (Pl. 27. fig. 72) and G. tachinoides (Pl. 27. fig. 75), or has
spread out into a plate with an aperture to contain the ejaculatory sac, asin G. morsitans
(Pl. 27. fig. 74) and G. pallidipes (Pl. 27. fig. 73). It works through the theca, which
has become a wall surrounding it; and in G. palpalis, G. tachinoides, and G. pallidipes
the upper ends are articulated on to the paraphalli, which, in their turn, are articulated
on to the theca.

3504 MR. WALTER WESCHE ON THE GENITALIA OF

Loncheid@a.—In Toxoneura muliebris it can be traced in a similar forked form at the
base of the penis (PI. 80. fig. 122), and in a fused form in L. nigrimana.

Sepside.—tn the Sepsidze the part is very clearly seen to be a paired organ, though it
becomes fused at its lower end (Pl. 28. fig. 99; Pl. 29. fig. 101).

Use.—These organs in all their conditions are attached to powerful muscles, which
rotate the penis from its position of rest as in the Blow-fly, or extrude it as in Tabanus
or Bibio.

Homology with ovipositor.—tIn the base of the ovipositor of Musca domestica are two
rods which fuse at their lower end for no apparent reason; it has occurred to me that
they represent these apodemes. Firstly, because they are at the base of all the other
rods save one, as the apodemes are normally at the base of the theca; and secondly,
because they are paired organs, agreeing in this respect with the apodemes in two
important particulars (Pl. 29. fig. 107). .

(x) Ducrus EJACULATORIUS.

Structure.—This part is usually a delicate hyaline membrane. In Syritta pipiens it
differs, having the appearance of a striated muscular tube with double walls, which
spread out to form the ejaculatory sac, this peculiar structure ending here.

Homology with ovipositor.—It is obviously homologous with the vagina of the female
insect.

(0) “ SACCULUS EJACULATORIUS.”

Structure.—This most important organ of the genitalia undergoes bewildering
changes of sbape and situation; it is usually a hyaline pouch with the apodeme
adhering to it, proceeding downwards from the penis; or it may be a chitinous sae, as
in the Tipulide, or a part of the theca, as in the Asilidee and Dolichopodide.

It will be convenient to describe the apodeme with the sac, so before proceeding to
more details I shall give the structure of the

(p) ExsacuLatory APODEME,

Structure—This is a minute rod in many species, not unlike the microscopic lever
that is found attached to each spiracle, but is more often spread out at the base into a
spatulate or fan shape, to afford a larger space for the attachment of muscles. This
part, when so modified, has the curious laminated structure noted in the double
apodeme.

In what may be termed the head, that is to say the part in contact with membrane,
or which is used to close the sac, a number of lighter circles, not unlike bristle-sockets,
may be seen in some species. They are characteristic, but they require high powers and
skilful manipulation to reveal. I have seen them in Asélus crabroniformis, Linn., the
Asilid figured, and in the very distantly related Glossina palpalis and G. pallidipes.

Nomenclature—Lowne noticed the part in Calliphora, calling it a sclerite* ; but this

* ¢ Blow-fly,’ p. 665.

BOTH THE SEXES IN DIPTERA. 355

not being definite enough in view of its importance in other families, I propose to call it
the “ ejaculatory apodeme.”

Bibionide.—My preparations of Sciara and Bibio are not successful in showing these
organs, but in the minute Scatopse notata I have been able to make them out quite
clearly. The sac is an elastic globe, and the elaborate transverse levers, actuated by the
apodeme, compress it, and force the seminal fluid through the duct (PI. 28. figs. 8, 9).

Tipulide.—In Tipula oleracea the sac, as I said before, is hard and horny, with two
processes, fitted for the attachment of muscles, fused on to it. The duct opens widely
but soon narrows, forming the hyaline flagellum. The base of the apodeme is blunt,
fitting into an opening in the sac, and is capable of closing the opening of the duct.
The apodeme has two arms, and it is obvious, on looking at the parts, that when the
right-hand or dorsal arm is drawn in that direction by its muscular attachments, the sac
is rotated, the duct is closed (owing to the changed position of the apodeme in its
containing cavity), and the flagellum is withdrawn into the theca. On the other hand,
when the left or ventral lever is pulled, and the sac rotated in the contrary direction, it
not only opens the duct, but protrudes the flagellum (Pl. 24. figs. 24, 26).

In Pachyrrhina maculosa the same arrangement is found, but the apodeme is of the
more usual form, with a fan or disc instead of opposing arms (PI. 24. figs. 30, 31).

In Gynoplistia bella the apodeme is seen in the centre, between the apodemes of the
penis, but my preparations do not clearly show the structure of the sac (Pl. 23. fig. 15).

In Ptychoptera scutellaris an even more curious organ is found. Three flat plates,
with muscular attachments, are fixed on three sides of a hollow chitinized receptacle
which is situated immediately below the curious penis (Pl. 23. figs. 17,18; Pl. 24. fig. 20).

Stratiomyiide.—In Beris vallata the penis takes a curve, and is then bent on itself,
broadening out into a cavity which forms the sac. In this cavity the apodeme is
articulated in such a manner that when it is drawn towards the thorax it opens the sac
and extrudes the penis, and when towards the posterior, contrary motions take place, a
variation on the device in 7’. oleracea (Pl. 24. fig. 33).

Tabanide.—A very complicated mechanism exists in Tabanus bromius, and I only offer
the following as a hypothetical explanation of the action of the parts.

The apodeme is placed in the centre of the theca, and reaches well up into that part;
at the broadest’ part of the theca an articulation with another piece is apparent. From
this point two very fine rods curve outwards, and then approach each other till they
overlap and form a fine point, capable of easily passing through the opening in the
point of the theca. These rods are contained in a membrane, which is the sac. On the
apodeme being drawn forwards it would push the rods through the orifice of the theca.
On being still further drawn, or being drawn backward and forward, it would separate
the points and, at the same time, compress the sac against the containing sides of the
theca, thus opening the orifice and expelling the spermatic fluid (Pl. 29. fig. 109).

Asilide.—In the Asilidz a similar mechanism to that of Bevis is seen, but with the
difference that the part is much thickened and chitinized (Pl. 24, fig. 38).

Empide—tIn the Empide we have another variation on the mechanism in Zipula.
The flagellum, after turning back on itself, widens out into a sac, into which an apodeme

SECOND SERIES.—ZOOLOGY, VOL. IX. 52

356 MR. WALTER WESCHE ON THE GENITALIA OF

is inserted and acts in the same way, the muscles pulling it one way opening the
passage and protruding the flagellum, the other way, closing the sac and withdrawing
the flagellum (Pl. 25. fig. 41),

Dolichopodide.—In Dolichopus griseipennis a beautiful mechanism is present; the
base of the apodeme is pierced by a neat fitting for the vas deferens, which, when the
apodeme is in its normal position, fits against a pad and closes the duct. When it is
drawn back, as is seen in the drawings (Pl. 25. figs. 50, 51), an elastic membrane is
stretched across the space between the theca and the apodeme, leaving the ejaculatory
duct free to the orifice. The changes in mechanism from the type of the Tipulide are
brought about by the fixture of the flagellum to the theca; consequently the apodeme
only does its proper work of containing or releasing the spermatic fluid, and does not
rotate or extrude the penis.

Syrphide.—In Catabomba pyrastri the sac is a flexible hyaline membrane with a
minute apodeme (Pl. 25. fig. 53). In Syritta pipiens an unusually large apodeme is
found at the base of an enlargement of the ejaculatory duct which represents the sac;

the head of the apodeme has a transverse plate and a swelling which appears to be —

modified to fit this space, and closes or opens the duct by backward and forward
muscular action (Pl. 26, figs. 56,57). In Lristalis tenax the sac is represented by a
hyaline tube with a small apodeme, as in Catabomba, but all contained in the theca and
fairly close to the orifice. The chitinous box at the extremity of the theca is fitted with
a ciliated valve, allowing for a free passage outwards (PI. 26, fig. 64).

Muscide —In the greater proportion of the flies the sac and apodeme are of the same
type and situation as in Catabomba pyrastri. The action of this arrangement is by no
means obvious ; either the apodeme is pressed against some other part by the attached
muscles, so as to close the sac, or is drawn backwards and forwards, acting as a pump
and driving the fluid through the ejaculatory duct (Pl. 26. figs. 60, 62, 68).

In Calliphora the sac is at a much greater distance from the theca than in Pollenia
or Scatophaga. This part is found in a curious situation in Glossina, where the
apodeme rests in the fork of the double apodeme (Pl. 27. figs. 72, 73, 74, 75), practically
at the orifice of the duct, and it is a knowledge of the mechanism in this genus that has
suggested my explanation of the part in Tabanus.

Loncheide —In Toxoneuru muliebris the sac and small apodeme are at the base of
the long ciliated penis, below the forked apodeme (PI. 30. fig. 122).

Trypetide.—In Acidia heraclei (Linn.) the sac is in a precisely similar position, but
the base of the apodeme is spatulate.

Sepside.—In the Sepsidze the apodeme has a remarkably long stalk (Pl. 28. fig. 94) ;
this is also the case in the nearly related Diopside.

Use.—This apparatus is arranged to control the flow of the seminal fluid, being so
situated and held by muscular attachments that when the penis is in its normal position
the sac is closed; a different process seems to be the rule in those Muscidze where the
ovipositor is long, as in Musca domestica ; in such flies the valve would probably not be
opened till after the elaborate interlocking of the parts shown in Berlesi’s section had
taken place.

—— ="

|

a

BOTH THE SEXES IN DIPTERA. 357

Homology with ovipositor.—tThe sac at the base of the oviduct in the Blow-fly appears
to homologize with this part. It is in the passage leading from the sac into the uterus
that the ducts from the receptacula seminis discharge, impregnating the ova in their
passage. But Lowne regards it as homologous with the uterovaginal tube *, which is
after all exceedingly close to the part I suggest. In Musca domestica there is a cavity
at the end of the passage into which the ductus ejaculatorius discharges, which is possibly
also homologous. This is well shown in Berlesi’s section already alluded to, but any
longitudinal section of the ovipositor in its normal position in the abdomen which shows
the receptacula seminis, or spermathecee, will probably show the part [ mean.

(r) THE VAS DEFERENS.

The vas deferens is the tube which leads from the paragonia and testes to the sacculus
ejaculatorius.

Structure.—It is usually a subhyaline membrane with many longitudinal wrinkles.
It is of various lengths, short or long in different families. In Dolichopus it has a very
marked muscular structure, and broadens out at its junction with the secretory organs.
Besides its length, it has little to distinguish it in the various families.

Homology —I\t homologizes with the oviduct of the female.

(s) THE PARAGONIA or VESICUL® SEMINALES.

These are a pair of sacs which open into the vas deferens. Their function is obscure,
but Lowne 7} rejects the idea that they contain spermatozoa, and considers that in the
Blow-fly atleast, “the secretion coagulates with great rapidity in the ejaculatory
duct or in the vagina of the female insect, and is apparently concerned in the formation
of spermatophores.”

Homology.—They are easily homologized with the glands which Lowne calls parovaria
in Calliphora, and have been confused with the “ glue-glands ” ¢.

(¢) VASA EFFERENTIA.

These are ducts which lead from the testes to the vas deferens. They are so named
by Lowne §, but are the “vasa deferentia” of other writers. These organs are always
present, usually of a hyaline structure and of varying length. They homologize with
the “tubze” of Lowne, which lead from the ovaries to the oviduct.

(wv) THE TESTES.

These are paired sacs, which secrete the spermatozoa; are mostly separated, and of
an orange, white, or brown colour. Dufour says, however, that in Laphria fulva, Egger,
both testes are enveloped in one receptacle, ‘un véritable scrotum ” |}.

Structure —In Musca domestica they are of a deep brown, and have the appearance

* © Blow-fly,’ p. 675. tT ‘ Blow-fly,’ p. 663.
~ Lbid. p. 673. § Lbid. p. 662.
|| Recherches anatomiques, etc. p. 198.

cr
bo
*

358 MR. WALTER WESCHE ON THE GENITALIA OF

of being chitinous, but Lowne says that this colour is due to the “ pigmented epithelium
which forms the walls’ *. In Scatophaga they are a reddish brown; in the Tipulide,
according to Dufour, white.

Homology.—They homologize with the ovaries of the female.

Phylogeny.—tThe foregoing examination of the male armature shows that the greatest
variations are in the central organ and in the ejaculatory duct; also that the families
exhibit their relationship to the Tipulid or the Muscid type by the character of this
duct, whether it is of the nature of a stiff flagellum or of a flexible membane. The
Bibionidee seem anomalous ; the rest of the families appear to show that the structures
of this part are characters of the two great divisions of Orthorrapha and Cyclorrapha.

SEGMENTS OF ABDOMEN.

Before proceeding to describe the ovipositor, I propose making some remarks on the
number of segments in the abdomen of both the sexes. This number is known to
be variable, as the following table, taken at random from the mounted Diptera in
my collection, will show ; but I think the normal is probably eight. In consequence of
this variability, when making comparisons between the sexes it will be advisable to
count from the last segment, that bearing the valves.

| Number of | | Number of |
| | segments. | segments.
Family. a sir bot | Family. eee
rs) 2 3. 2

| Mycetophilide ........ Wee moi ceen yi eS so | 78
\Rikodidecis tote | 8-9 9 | Dolichopodidee ........ 7-8 8
Gaede tan eek I Aig eal geet | Syrpbides! 1) Jee Seg Gen ae
| AshiGESS Sogoume on aa 7-8-9 8 | Muscide .2-.:--: «e- 5-6-7-8| 6-7-8
| lganiienleey ¢Sgoe ome 49 | ass | 8 | Heliomyiide .......... 6-7 7-8
| Stratiomyiide ........ | 8-9 5-6" MGirealisten ie An dene eel Gi
| Tabanide ............ 8-9 9 Sepsids: iJ) tea tanientes 5 | 6
Wisthidis Fecaet ek. ei Tl Ephydride ........-. 5-6 Alf

Bombylide 32-5... | 8 Borboridse ie ererrers-t iste 5-6 7
| | |

THE OVIPOSITOR.

The ovipositor in Diptera varies in length and structure. In what are usually
considered (on the evidence of fossil remains) the older families, it consists of valves or
egg-guides, and is not telescopic or capable of extension (Pl. 24. fig. 21). It approxi-
mates to this form in Bzbie, Culex, Chironomus, Tabanus, and Tipula, and also in some

* © Blow-fly,’ p. 660.

BOTH THE SEXES IN DIPTERA. 399

of the Muscidee, calyptrate and acalyptrate. In Stratiomys, Empis, Dolichopus, and
Syrphus it is found to be longer and more capable of extension. In some of the
Muscidee it is at its greatest proportionate length in its membranous form. In the
acalyptrate Ortalide, Trypetids, and Loncheide it is equally long, but hard and horny
and capable of depositing the eggs under the cuticle of leaves. In the minute Phito-
myzidee it is shorter, but still a horny, minutely-aculeated organ.

“ Receptacula.”—An examination of a prepared slide of a female shows from one to
three, rarely four, chitinous sacs, mostly of an oval or pyriform shape, with more or less
short ducts leading from one end; these are the receptacula seminis or spermathece.
In Musca domestica they are situated at the end of the long fold of the ovipositor into
which the ductus ejaculatorius of the male discharges; they absorb the spermatozoa,
and again eject them when the ova in their passage down the oviduct compress the
receptacula.

In the preparations they are found either in the abdomen or in the extruded
ovipositor, and possibly, on a cursory examination, might be mistaken for eggs.

* Glue-glands.”’—Excretory glands will also be found in some ovipositors, sometimes
called “ glue-glands.”” The eggs adhere to each other or to their resting-places by means
of the fluid from these glands, with which they are anointed in their passage through
the ovipositor.

Types in the Muscide.—In the Muscide there are several types of this organ :—

(1) That in Musca domestica is long and telescopic, with three joints, and has on the
anal segment two feeling-organs or valves, a semi-ovoid and two triangular plates (these
latter probably representing the eighth segment of the abdomen) and two other segments
supported by ten chitinous rods (a pair of these rods being fused at the base), anda
number of transverse setose plates (Pl. 29. fig. 107).

(2) In Lucilia it is also long and telescopic, but has plates instead of rods, and on
the dorsal plate of the third segment, the anal segment counting as “one,” are two
pairs of spiracles close to each other.

(3) In Calliphora it is much shorter, but has plates and paired spiracles on the same
segment as in Lwcilia.

(4) In Polietes lardaria (Fabr.) it is again long and telescopic as in JZ. domestica,
has the usual valves on the anal segment, but neither rods nor plates, the segments
being marked by small transverse bristly bands of chitin, and without spiracles on the
third segment.

(5) In Hydrotea dentipes, Fabr., it is long and telescopic, has plates which seem to
be thickening into rods, and two pairs of spiracles on the third segment, as in Calliphora
and Lucilia.

(6) The very short ovipositor that is found in Scatophaga and Anthomyza (Pl. 28.
figs. 90, 100; Pl. 30, fig. 181).

(7) The horny, extrusive ovipositor that is found in those families already alluded to
(the Ortalidee, Trypetidee, Lonchzeide, and Phitomyzidie) (Pl. 28. fig. 95).

The Spiracies—In comparing the ovipositor in different families, and homologizing
it with the male armature, the presence or absence of spiracles on the segments might

360 MR. WALTER WESCHE ON THE GENITALIA OF

be thought to be a guide to relationships, but so far I have not been able to draw any
conclusion from my observations on this point, so I shall content myself with stating
them.

Lowne points out that on the sixth somite of the abdomen of the female Blow-fly, or
the third from the end, are two pairs of spiracles close to each other. I find a similar
arrangement on Calliphora or Protocalliphora grenlandica, Zett., and Lucilia sericata,
Meig., but no traces of spiracles on Musca domestica and the Anthomyid Polietes
lardaria, though in Hydrotea dentipes and Ophyra leucostoma, Wied., the double
spiracles are again evident.

In Hilara cilipes the number of segments differs, but counting from the anal
segment, in the same place, are a single pair of spiracles (Pl. 25. fig. 42).

Arrangement.—For convenience of discussion and description the ovipositor may be
divided up into four parts :—

(1) The egg-guides or sensory organs and the other plates at the extremity.

(2) The ultimate or anal segment.

(3) The penultimate segment and the elue-glands.

(4) The spiracle-bearing segments and the Receptacula seminis.

The Egg-Guides and the Appendages of the Ultimate Segment of the Ovipositor.

Egg-guides or valves—All the ovipositors that I have examined have some appendages
on this segment. In a complete state, five parts are present—two dorsal valves or blades,
two ventral, and a single plate between; but it is seldom that the organ is found in this
state.

Nomenclature.—Nothing definite has been settled on the point of nomenclature.
Dziedzicki has, in the paper before referred to, named the parts in the ovipositor, but
these do not, in my opinion, fit the organ so well as his male nomenclature ; so I propose
to call the parts the valvulie superiores, the valvulz inferiores, and the lamella anterior.
The valvulz superiores will correspond with the dorsal plates of Lowne*, while the
valvule inferiores are represented ia the Blow-fly by the anal scales.

Bibionide.—In Bibio hortulanus there are two large hairy valvule superiores, a
single plate (the lamella anterior), and two ventral subtriangular pieces (the valvule
inferiores) (Pl. 23. fig. 2).

Culicide—In this family these parts are not so prominent, and their bases are
usually hidden by the last plate. In Culex pipiens the upper valves are fairly
pronounced, but the lower are only represented by hairy palpiform processes, well up in
the dorsal region of the cavity made by the last plate. Dénocerites has also in the
female striking genitalia, horny upper valves and fairly large lower valves, placed in
the same situation as in Culex.

Tipulide.—In Tipula oleracea all the parts are present, the upper and lower valves
very hard and blade-like in form (PI. 24. fig. 21).

Trichocera hiemalis, Degeer, and Ptychoptera albimana (Fabr.) have only one pair of
large blade-like valves on the end of the abdomen, the dorsal and ventral sides meeting

* ¢ Blow-fly,’ pp. 745-46.

23-0 So) aa eS

eee a

BOTH THE SEXES IN DIPTERA. 361

at their base—these are the valvulz superiores; the valvule inferiores seem atrophied,
but there are signs in the interior of P. albimana of the presence of the lamella
anterior.

Stratiomytide.—In Beris vallata the valvulee superiores are two-jointed, which is very
unusual, the valvule inferiores very minute, a similar arrangement existing in B. nigra,
Meig. (Pl. 24. figs. 34, 35). In Stratiomys chameleon all the parts are present but
relutively small. ‘The valvulze superiores have a curious sense-organ on the apex.

Tabanide.—In Tabanus bovinus, Linn., all the valves are hairy, and, as in S. chameleon,
relatively small; the valvulz inferiores appear to have fused, but the plate is notched
on the median line (Pl. 30, fig. 126).

Limpide.—\n Hilara cilipes only a pair of hairy processes are present (PI. 25. fig. 42).
In Empis chioptera, Fall., these processes are adhering to a membrane, the segment or
apparent segment consisting of a dorsal and a ventral plate. These may represent the
other valves and plate.

Dolichopodide.—in Dolichopus a very remarkable development of this part is found; |
two horny valves (v. superiores) are laterally placed, and below them are two ciliated \—

lunules (v. inferiores). The valvulze superiores are articulated on to a plate, cleft in the
middle; on this is a remarkable series of blunt spines, presumably of use in coitus
(Pl. 25. fig. 52). They are a feature of the ovipositor in this family, and are often
double and treble the length of those in D. griseipennis. These contrivances for
holding the partner in coitus, common enough in the male, are quite rare in the other
sex. In the Bombylid Comptosia ocellata, New., the valvule inferiores have a number
of long blunt hooks, Pegomyia bicolor has pads of hair, and Norellia .spinimana (to
which I shall again refer) has rows of short blunt spines on the abdomen (PI. 28. fig. 90).

Syrphide.—tin the Syrphide the valves are hairy and rather insignificant, but the
plate is difficult to differentiate.

Muscide—tIn the calyptrate Muscide the dorsal valves are hairy ; they are opposite
to a plate which is probably a fusion of the ventral valves, and they are attached to
a strip of chitin which has, and sometimes has not, a suture in the middle (J/uscu
domestica and Calliphora erythrocephala) (P1. 29. fig. 106).

When the ovipositor is of the Ortalid type, generally a pair of hard short valves are
present, carrying a few sensory sete, the remaining parts not being distinguishable
(Pl. 28. fig. 95).

Cordyluride.—tIn Seatophaga stercoraria the valvule superiores are very hairy and
each carries a long seta; the valvule inferiores are horny plates, genuine egg-guides,
and the lamella is well-marked (PI. 30. fig. 131).

In Norellia spinimana are only two setose valves and two plates (PI. 28. fig. 90).

Geomyzide—In Anthomyza pallida, Zett., the valvule are peculiar; they are
chitinous plates, studded with blunt spines; opposite to them are two valves; the plate
is difficult to differentiate in my preparations (PI. 28. fig. 100). In an unnamed slide
in my collection (an Anthomyid) the dorsal valves have two short, strong, socketed
setze on each, besides many hairs; the other parts are of the MWusca domestica type,
but with very fine rods.

362 MR. WALTER WESCHE ON THE GENITALIA OF

The Ultimate Segment of the Ovipositor.

In families where the ovipositor is not extruded this part does not call for remark.

Stratiomyiide.—In Beris and Stratomys chameleon it is a chitious plate with no
suggestions of other structure, though the edges in B. valata are convoluted and
thickened transversely.

Empide.—tIn the Empidee it consists of a dorsal and ventral plate.

Dolichopodide.—In Dolichopus griseipennis there are indications of the formation
of two chitinous rods. On the ventral side, at the extremity, is a sensory membrane,
covered with minute triangular plates (Pl. 25. fig. 52). In Pacilobothrus nobilitatus
four rods show quite plainly.

Syrphide.—In Syritta pipiens are two chitinous levers which appear to be the
continuation of the valvulz, and there is a darkening of the chitin suggestive of the
formation of a rod.

In Eristalis intricarius there are three transverse ridges of chitin with setze. In the
widdle is the opening of a gland which is probably a glue-gland.

Muscide.—In Musca domestica this part contains four chitinous rods which are
shorter and stouter than the rods in the following sections; there are also several
transverse setose ridges. The membrane between this and the penultimate segment
has minute triangular sensory plates (Pl. 29. fig. 107).

In Calliphora erythrocephala there are no rods, the part consisting of a dorsal and
ventral plate; there are apodemes from the valves.

In Polietes lardaria ave two transverse setose ridges, dorsal and ventral; the rest of
the part is transparent membrane.

In Ulidia nigripennis this segment consists of apparently three plates forming a
rod-like piercing-organ.

In Norellia spinimana this segment is part of the abdomen, and appears to consist of
a chitinous plate, and bears the remarkable process of short spines already alluded to
(Pl. 28. fig. 90).

The Penultimate Segment.

Stratiomyiide.—In Beris vallata and B. nigra plates, which do not quite encircle the
abdomen, are found in this place. In S. chameleon what seems to be the opening of a
gland and lateral processes are found on this segment.

Empide.—In the Empidz two plates usually form this part. In Hilara cilipes is
the curious notched process previously alluded to, and in the interior wall is the opening
of a glue-gland (Pl. 25. fig. 42).

Dolichopodide.—In Dolichopus griseipennis this part appears to be quite membranous ;
there is also a glue-gland with a funnel-shaped opening in the interior wall (Pl. 25.
tig. 52). In D. festivus three plates are present, as is also the case in Pecilobothrus
nobilitatus, where there are also suggestions of the formation of rods.

Syrphide —In Syritta pipiens the part presents no noticeable structure.

Muscide.—In Musca domestica, at the posterior part of the segment, are four small,

BOTH THE SEXES IN DIPTERA. 363

transverse, setose plates, and three long narrow rods longitudinally support the
membrane (PI. 29. fig. 107).
In Calliphora erythrocephala there is a dorsal plate opposed by a shorter ventral plate.
Polietes lardaria has three transverse setose plates.
Ulidia nigripennis has four half-formed chitinous rods on this segment (PI. 98. fig. 95).
Before discussing the next section, “the spiracle-bearing segment,’ I propose making
a few remarks upon a part that is occasionally met with and whose presence accounts
for traces of chitinous structure found without any apparent reason in the ovipositor.

The Apodeme in the Ovipositor.

Simuliide.—In Simulium reptans (Linn.) * there is an apodeme at the extremity of the
abdomen ; it is a thin chitinous process and forks at its posterior part, the forks partially
enclosing an opening (which appears to be that of the vagina) and having at their
extremitics some connexion with the external valves. The function of this part is to
rotate the opening, as I have preparations which show the shaft of the apodeme in both
anterior and posterior relation to the opening.

Asilide.—In an undetermined Asilid from Queensland there is, in a similar position,
a plate with tie posterior extremity furcate ; near this part the orifices of the ducts of
the three receptacula seminis open. This apparatus appears to be homologous with
the apodeme in Simuliwm, and explains the frequent appearance of chitinous rods in
Syritta (Syrphide), Stratiomys (Stratiomyiidze), and other insects.

Homology.—In homologizing this part with the ovipositor of Musca domestica, its
furcation suggests that it is represented by the partially fused rods in the third segment,
but its position points to other structures in a more anterior position. It appears to me
that these levers are a development in the direction of the telescopic ovipositor, it beg
of advantage to insects to extrude the ovipositor even in a small degree.

Chironomide.—Chitinous structures are also found close to the valves of Chironomus
plumosus (Linn.) and C. riparius, Meig.

The Spiracle-bearing Segment.

Bibionide.—In Bibio hortulanus, Dilophus febrilis (Linn.), D. albipennis, Meig., and
Scatopse notata the abdomen has, on the segment next the penultimate segment, single
paired spiracles, as on this part in those families where the ovipositor is telescopic.

Stratiomyiide.—In Beris vallata the spiracles are on the membranes on the sides of a
central, ventral, smaller plate. In B. nigra the plate is large, consequently the spiracles
are farther apart.

Tabanide.—In this family, unlike the Bibionids, the spiracles appear to be on the
penultimate segment; but I think, from the evidence of a small triangular piece in
Hematopota pluvialis (Linn.) and H. crassicornis, Wahlb., that the last segment has
fused with the next and represents two.

Limpide.—tIn the Empidze the spiracles are about the middle of the segment.

* There is some uncertainty as to this insect, but it is of the same size and colour, and if it is not the genuine
S. reptans, Linn., it is almost impossible without special study of this family to separate the two species.
SECOND SERIES.— ZOOLOGY, VOL. Ix. 53

364 MR. WALTER WESCHE ON THE GENITALIA OF

Dolichopodide.—In Dolichopus griseipennis and D. festivus the spiracles are in the
anterior part of the segment.

Syrphide.—In Syritta this segment has no structure to notice except a single pair of
spiracles at the anterior part.

Muscide.—In Musca domestica three transverse setose plates at the posterior part of
the segment, anterior to them three fine long rods, two of which are fused together at
their anterior ends. I cannot find any spiracles on this segment (Pl. 29. fig. 107).

In Calliphora erythrocephala there is a dorsal plate opposed by a longer ventral plate ;
the posterior corners of the dorsal plate bear two pairs of spiracles close to each other.
This is a striking character, which I find also in Protocalliphora graenlandica, Zett., and
Lucilia sericata, Meig., as well as in the Anthomyids, Hydrotea dentipes (Fabr.),
Ophyra leucostoma, Wied., and Anthomyia pluvialis, Linn.

Polietes lardaria has no spiracles on this segment, the only structures being three
transverse setose plates.

In Ulidia nigripennis, on this part are three plates, and a pair of spiracles is laterally
arranged (Pl. 28. fig. 95).

Aberrant forms.—t shall now describe two ovipositors which are very unlike the
usual plan of the part.

In the Tachinid Phorocera serriventris, Rond., or concinnata, Meig., a remarkable form
is found. The valvule superiores are fused and formed into a single highly chitinized
hook, which is bent in under the abdomen and appears much like the penis in many
flies. Another process is found on one side with sete, and on the other side another
asymmetric part. Close to the hook are paired spiracles. The insect was in the vivi-
parous stage, the abdomen being full of fully-formed larvee (PI. 30. figs. 118, 119).

Lauxania enea, Fall., is also very divergent from the general type. The dorsal edge
is furnished with the usual paired organs (valvule superiores) ; opposite these the ventral
plate is drawn out into a single strong lanceolate process, having on either side two
plates which may be the homologues of the valvulze inferiores (Pl. 30. fig. 121).

The Receptacula seminis.

Number.—The number of these organs varies from one to four; but in the Muscidee
it is nearly always three. When the organ is single it is usually of large size.

Nomenclature—They are sometimes known as spermathece, but Dufour calls them
“glandes s¢bifiques.” I am under the impression that “ receptacula seminis ” is Lowne’s
name for the parts.

Libionide—In Bibio hortulanus and Dilophus febrilis (Linn.) they are three in
number and globular in shape. In Scatopse notata oval, relatively large, and single
(Pl. 23. figs. 3, 11).

Chironomide.—tIn a few females of Chironomus that I have examined I cannot find
any; but in the genus Ceratopogon three are found.

Culicide and Tipulide.—In the Culicidee and Tipulide these organs are globular and
three in number,

i

BOTH THE SEXES IN DIPTERA. 365

Stratiomyvide.—tIn the Stratiomyiidz they are more oval in shape and have charac-
teristic long chitinized tubes, which continue for a similar distance in a hyaline
condition, and are three in number.

Tabanide.—In Tabanus bovinus, Hematopota pluvialis, and H. crassicornis the recep-
tacula are peculiar, relatively small, quite pyriform in shape, attached to exceedingly
long tubes whose circumference is only a little less than that of the receptacula. They
end in muscular outgrowths, which have the appearance of columns, the capitals being
formed of a transverse plate which the tube pierces; these outgrowths penetrate the
walls of, and open into, the oviduct. They are three in number (PI. 30. figs. 128, 128).

Asilide.—Three globular receptacula with long muscular ducts are found in the
Asilidee.

Bombylide.—A. similar arrangement and number of receptacula to that of the
Tabanidee are found in Comptosia ocellata.

Empide.—I have only found one receptaculum in the females of the Empide that I
have examined.

Dolichopodide.—In the Dolichopodidee, as well as several other families, the recep-
tacula are not chitinous and do not show in my preparations. But the rectal papilla—
four glands which surround a portion of the anal passage and whose exact function is
not clearly demonstrated—are very prominent. They are shown on the ultimate segment
of the ovipositor of Dolichopus griseipennis (P|. 25. fig. 52), but this gives no real idea
of their true form; they are cone-shaped, the base is a chitinous hoop which has some
very fine aculeations scattered on its surface. This hoop rests on a membrane with
trachez in its structure, and supports a transparent or subtransparent cone which is
studded with numerous curious lamine, which have one of the edges pectinated,
resembling a scale of Lepidoptera (Pl. 30, figs. 124, 125).

They are present in equal development in both sexes, and are very evident in several
species of Dolichopus, as well as in Pecilobothrus.

This part is, strictly speaking, not in the subject-matter of this paper; but as the
receptacula in this family are difficult to find, and appear to be of a different structure
from that which is usually found associated with the part in Diptera, I think it as well
to draw attention to them, especially as they appear in Pl. 25. fig. 52, and as they might
easily be mistaken for receptacula, unless the observer was familiar with Lowne’s *
description of those organs in Calliphora.

Syrphide—tIn Syritta pipiens and Lristalis tenax the receptacula are three in
number, are of a flattened oval shape, and carry long tubes.

Muscide.—In the Muscide three receptacula are usually found; but my prepara-
tions of Stomoxys calcitrans (Linn.), Hematobia stimulans, Meig., Glossina morsitans,
and G. palpalis only show two receptacula. In the Anthomyzide the receptacula are
three in number, and often have marks on the cuticle quite distinctive of the species.
Anthomyia radicum has short dark papillee on the surface (Pl. 30. fig. 1383). Homalomyia
manicata, Meig., shows short spines on the inside cuticle. Hylemyia cinerosa, Zett. (?),

* « Blow-fly,’ pp. 417-418.
53*

366 MR. WALTER WESCHE ON THE GENITALIA OF

thick knobs with broad bases; Pegomyia bicolor, Wied., fine blunt spines inside (Pl. 30.
fiz. 134). In the Cordyluride it is also variable, quite long and vermiform in Scato-
phaga stercoraria (Pl. 29. fig. 115; Pl. 30. fig. 116). In Helomyza similis the receptacula
are four in number. In Ulidia nigripennis they are three in number and of very curious
form (Pl. 28. fig. 95). In Setoptera vibrans (Linn.) they are four in number, but double,
the pairs being joined by the tubes, each pair having only one duct (Pl. 30. fig. 132).
Several preparations of Lonchea and Toxoneura muliebris, two; Balioptera tripunctata,
Fall., and B. combinata (Linn.), two; Sepsis cynipsea, two.

Ephydride.—All the Ephydridz I have examined have a single receptaculum, but
this is of a peculiar design. In the small Hydrellia griseola, Fall., it is very large and
shaped like a thimble; in Parydra coarctata somewhat of the same shape, but longer,
and from the flat end proceeds a thick stalk (Pl. 29. fig. 114).

Borboride.—Borborus has two, and Limnosina three receptacula.

Remarks on those Families whose Genitalia have not been analyzed.

I shall now make a few remarks on the families that have been omitted in the fore-
going review.

Pulicide—tThe Pulicidee are a doubtful group, as some morphologists place them in
a separate order (the Aphaniptera or the Siphonaptera). Mr. Verrall, however,
includes them in his list of British Diptera. In Pulex irritans the penis is double and
consists of two chitinous stiles, to each of which a hyaline membrane adheres. These
stiles coil on themselves, at the interior extremities appearing to connect with a large
gland (the testes?). In an interior position is a strong apodeme ; between it and the two
stiles is a membranous tube, longitudinally wrinkled, which in the Muscide I should
recognize as the sacculus ejaculatorius but for the fact that no apodeme appears to be
present. The upper part, with which the great apodeme connects, is furnished with two
hairy processes (the forcipes superiores) and some complicated clutinous ones which my
preparation does not clearly define. Behind the forcipes superiores is the pygidium, the
curious sense-organ found in both sexes of the Pulicidee.

In the female the valves are represented by several hairy processes below the pygidium,
and one large receptaculum seminis of peculiar shape is clearly seen (Pl. 29. fig. 113),
In Ceratopsyllus jubatus, Wag., the flea parasitic on the bat, one of my preparations
shows organs at the end of the abdomen which may represent the appendages of the
theca, but these insects are so minute that I cannot definitely say so. The other parts
resemble those in P. irritans. The double stile, so far as my knowledge goes, is not in
favour of the idea that the Pulicide are degraded Diptera.

Cecidomyiide.—I\n the Cecidomyiidze one preparation shows an armature approximating
to that of the Chironomidee. In another the forcipes superiores and inferiores, as well
as the forcipes interiores, are evident and the penis is membranous; but it is difficult or
impossible to determine specimens till the group has been more systematized ; a study of
the genitalia of prepared insects will probably be the easiest means, as otherwise they
keep so badly that, after a few years, only specimens mounted in balsam are of much

use for comparison.

BOTH THE SEXES IN DIPTERA. 367

Stmutiide.—In Simulium ornatum, Meig., the forcipes superiores are small and hairy ;
the forcipes inferiores large, and each have a small spine placed symmetrically opposite
each other near the ends. On the theca appear to be barbed appendages, but my
preparation is not satisfactory, and I cannot trace the apodeme. In S. reptans (Linn.) (?)
I can make out an arrangement of the theca with some affinity to that found in the
Culicids ; there is, however, a central apodeme which is forked at its junction with the
theca. The females have the usual valves, a single receptaculum, and a peculiarity in
the shape of a bifurcate apodeme to the aperture of the vagina. This appears to answer
to the double apodeme or to the fused rods in the ovipositor of Musca domestica.

Chironomide.—The Chironomide are a large group, but, like the Cecidomyiide, of
fragile structure. I can trace all the valves and appendages in my preparations, but
not beyond that, the apodemes and ductus ejaculatorius eluding me. In the females
I cannot see the receptacula, these also not being of robust structure, except in
a single preparation of Ceratopogon obscurus, Winn., where I find three, and those
three relatively very large. I hope at some future time to make a special study of
these insects.

Orphnephilide.—The Orphnephilidze have only two European species, neither of
which is in my collection.

Psychodide.—The Psychodide are minute and difficult of preparation. The females
of some species have egg-guides somewhat similar to those on Ptychoptera albimana and
Trichocera hiemalis. A male has four relatively very large forcipes, and I think I can
trace the apodeme.

Leptide and Therevide.—The Leptide and Therevidee are small groups of the
Tabanid type at which I have hitherto had very little opportunity of working.

Scenopinide and Cyrtide—The Scenopinide and Cyrtide are even smaller, both
families only numbering five species in the British list. Dufour has some remarks on
Scenopinus fenestralis in his paper already quoted.

Lonchopteride.—tThe genitalia of the Lonchopteride, like the venation of the wing,
approximate to those of the Muscide; but the receptacula must be of a different
structure, as they do not show in the preparations cleared in potash, whereas they show
admirably in those of the Muscidze treated in this manner.

Platypezide.—I have not had an opportunity of examining any species of the small
group of the Platypezide.

Pipunculide.—In the Pipunculide the males have generally a very prominent hypo-
pygium, with an armature something between that of the Empide and the Syrphide.
In Chalurus spurius, Fall., the flagellum is fureate a short distance from the point ; the
theca also is furcate, and the flagellum enters it at this point. The rest of the theca
is much like that in the Empidew, but at the interior end, though it widens out, it does
not form a sac, nor is it articulated with the ejaculatory apodeme. Instead of this,
a hyaline membrane proceeds from it with the usual small apodeme, as it is found in the
Muscidee (Pl. 30. fig. 117).

Conopide.—In the Conopide the male has often a bulbous hypopygium, but the
forcipes are not of very definite shape, more valvular than hamate, but covered with

368 MR. WALTER WESCHE ON THE GENITALIA OF

many short spines and blunt bristles; both apodemes are easily made out in Sicus
Serrugineus (Linn.) and Myopa buccata (Linn.).

G@stride.—tThe male of Gastrophilus equi, Fabr., has large horny forcipes, like those
of the Conopidee, more valvular than hamate, but without spinous processes; the
appendages are large, but I have only very poor preparations of both sexes, and cannot
make out either of the apodemes. The female has hairy valves which enclose a curious
chitinous chamber in the vagina, and I can only make out two receptacula; but
obviously this insect requires more study from fresh preparations before anything more
definite can be said.

Phycodromidea.—The genitalia of the Phycodromidee are distinctly Muscid in type,
but have a long ribbon-like process at the extremity. A species of Cwlopa has on the
end of the penis five triangular membranous appendages; from the central triangle,
which is flanked on either side by a pair of the others, the ribbon is attached; a long
central apodeme is easily made out.. The ovipositor is much longer than that found in the
Cordyluridze, but not so long as in Musca domestica, and has the usual three receptacula.

Heteroneuride.—I have no preparations of the small Heteroneura family in my
collection.

Sciomyzide.—In the Sciomyzide the genitalia are of the Muscid type in most species.
Sciomyza cinerella, Fall., has both levers very much developed, and the female appears
to have only two receptacula, but these are remarkably horny, covered with short
barbs and with strongly chitinized stalks. The ovipositor is short, with hairy valves
and two apodemes at the base of the penultimate segment.

Psilide.—The males of Loxocera albiseta (Schrank) have small genitalia of a rather
indefinite character, and possibly other species of this group may give better results.

Micropezide and Piophilide.—In the Micropezide, also, I have had insufficient
material, and even less in the minute Piophilide.

Drosophilide.—The male of Drosophila funebris (Fabr.) has most elaborately hamate
genitalia of a decided Muscid type; the female has a short ovipositor not capable of
extrusion and only two receptacula (so far as a single specimen is to be relied on).

Chloropide.—The males in the Chloropide have apodemes of the Muscid type, and
the females long membranous ovipositors, which consist of three segments, and have
a characteristic longitudinal striation. The receptacula are not chitinous, as none
appear in my preparations that have been cleared in caustic potash.

Milichiide, Agromyzide, Astetide—I1 have no preparations of these families in my
collection.

Phoride.—In some males of the Phoride there are suggestions of a petiolated hypo-
pygium like that in the Dolichopodide, but the genitalia are peculiar and not easy to
classify. The females of some species have an ovipositor which is remarkably like that
of the Chloropidee, and which also does not show any receptacula.

Hippoboscide.—In the Hippoboscidee the penis appears to consist of several stiles,
as in the Stratiomyiide. In Melophagus ovinus (Linn.) the double levers are evident,
suggesting a Tabanid descent; the female has, however, the two pairs of paired spiracles
which we find in some Muscide.

BOTH THE SEXES IN DIPTERA. 369

Braulide.—In the male of Braula ceca, Nitz., the minute parasite of Apis mellifica,
small as it is, I can trace the appendages, and they appear of more Muscid type than
the genitalia of Melophagus.

Nycteribiide.—The Nycteribiidee are a difficult group to study, and I have only had
access to some preparations in the British Museum, and have had no opportunity of
making dissections, the only satisfactory way of studying genitalia.

The male of Nycteribia Dufourti, Westw., has a large pair of forceps, quite ventrally
placed, articulated at their bases ; their points are highly chitinized, and are not unlike
those of Glossina palpalis. The penis is small and the apodemes cannot be differentiated.
There are many spines in the region of the laminz superiores, but no actual plate ;
laterally there are two bulbous processes very thickly spined.

The ovipositor cannot be said to exist in the Pupipara, but there are relatively large
processes at the extremity of the abdomen of the female.

In Cyclopodia Hopei, parasitic on one of the flying-foxes, the male has a rather pointed
abdomen ; on the ventral side of this are articulated a neat pair of forceps which meet
at their points and quite cover the cavity of the hypopygium. Between their bases
two small chitinous knobs can be seen. There are no lobular spined processes as in
N. Dufourii, but the segment opposite the forcipes has a short row of blunt spines on
its edge, which represents the laminze superiores. The female cannot be said to have
valves, but has two large tubercles situated dorsally and ventrally on the extremity of
the abdomen ; the larger is on the ventral side and is tufted with bristly hairs, and that
side of the abdomen has many long strong spines.

In ancther species of this genus, also parasitic on a “ flying-fox,” labelled Nycteribia
Westwoodii, in the Cabinet of the Quekett Microscopical Club, the forceps are more like
those of NV. Dufourii; it has the row of blunt spines representing the laminz superiores,
but no lateral processes. This specimen, which has been cleared and mounted under
pressure, shows in the interior a very large aculeated membrane, with long chitinous spines
and astrong apodeme, which appears to be the penis, but no appendages can be made out.

Similarity of Appearance of Genitalia and Mouth-parts in Diptera.

A person acquainted with the mouth-parts of Flies, particularly with the armature of
Simulium and Tabanus, must be struck, when examining the genitalia of the Muscide,
by a similarity of appearance and arrangement. In both he finds a central organ
surrounded by aculeate and setose appendages, and in both the central organ pierced by
a tube or duct. It occurred to me that these parts, so widely separated in situation,
were intimately connected, and that they were both influenced by an édentical law of
growth and development.

By “identical law ” I mean such a law as governs the growth of the number of joints
on the limbs; a law not absolutely inelastic or immutable, as can be seen in the tarsi,
where, though five joints are generally found, a lesser number is sometimes met with.
It regulates the growth of the appendages of both extremities.

Quite different from this law is that which governs the secondary sexual characters

370 MR. WALTER WESCHE ON THE GENITALIA OF

that develop on all parts of insects: under the first law, though admitting infinite
variety of form, those forms are rigidly confined within the limits of a fixed scheme
and can be homologized throughout the Insecta; under the second law, though
admitting an even greater variety of form and contrivance (all, it is true, with a very
obvious objective), the forms are not capable of homologization; we can recognize that
an identical cause has broadened and shortened the tarsi of Platychirus and Dytiscus,
but the extraordinary suckers on the tarsi of the Beetle can only be recognized as
homologous in its family; they have no counterpart in the fundamental scheme of
Insecta.

A similar law, governed by a different necessity, has developed the “ ptilinum” in
both sexes of the Cyclorrhapha in Diptera.

If we could adopt the old idea of design in Nature, we might think of the genitalia
and mouth-parts as formed on the same plan, or as the working-out of the same idea.
Failing that, we assume that they might be derived from organs which, placed at both
extremities of the organism, shaped themselves by unknown but similar laws of growth.
So we find the extremities furnished with hamate appendages such as maxille or
forcipes interiores, and sensory appendages such as maxiilary or genital palpi, and these
appendages surrounding and attached to a centfal organ which itself has special sense-
organs attached to it, “ taste-hairs” on the labium, or the aculeate membrane on the
penis.

IT submit that if it were a matter of development along the “lines of least resistance,”
or of growth similar to the second law, we might expect frequent departures from the
unity of plan that we can trace in these organs; that possibly appendages fulfilling
the functions of mouth-parts would be found attached to the fore legs, if not in most, at
least in some species, as in that position they could be used with equal, if not greater,
convenience. This is contrary to all experience, but when we leave the true insects,
and examine such a group as the Arachnida, we notice the absence of the unity of
design both in the mouth-parts and the genitalia.

Genitalia in the Arachnida and Hydrachnide.—In the Spiders the males have the
genitalia on two palpiform processes which spring from the roots of the fore legs, as in
Meta segmentata and Stemonyphantes lineatus, and these vary from extreme simplicity
to extreme complexity; while in the Water-Mites a spermatophore is picked out of the
genital pouch by the claws of one pair of legs and transferred to the genital pouch of
the female, both organs having a similar situation on the centre of the abdomen.

In Odonata.—It may be argued that even in the true insects there is a departure
from the general type in the Odonata, where the copulatory apparatus is in a segment
far removed from the extremity of the abdomen, though the forcipes and the genital
pore remain in their usual position. I have made only one or two dissections in this
group, and these are not complete, but my preparations show a theca, genital palpi,
forcipes interiores, a penis, and a spine, which can be homologized (putting aside the
segmentation) with the corresponding parts in Diptera; this is in_#Wschna cyanea, Mill.

In Orthoptera.—Some of the Orthoptera also present a difficulty, as in many species
a spermatophore is transferred by a hook to the cloaca of the female; this is, according

BOTH THE SEXES IN DIPTERA. 371

to Lowne *, most probably the method by which the common Cockroach, Periplaneta
orientalis, Linn., is fertilized.

Homologies of the Genitalia in Periplaneta orientalis and Diptera.—With the view of
homologizing the genitalia of the two orders, I have made a number of dissections of
this P. orientalis, and have come to the following conclusions. The greatest difficulty
of the investigation was the unsymmetrical character of the parts.

1. The whole combination, which in all the other insects that I have examined is
always in the longitudinal plane of the abdomen, in Periplaneta is placed in a transverse
position. What Lowne calls the left gonapophyses (the spinus titillatorius) is usually
the most anterior part of the genitalia.

2. Though there is no duct opening in a penis, there are parts which represent the
theca, the hypophallus, and the paraphalli.

3. The part which I homologize with the hypophallus is highly chitinized, has a
structure of short triangular serrations similar to that on the same part in the Blow-fly—
a structure which is constantly recurring in the different families of Diptera (PI. 25.
fig. 41; Pl. 26. fig. 68; Pl. 29. fig. 101).

4. These are grouped round the “conglobate gland” of Professor Miall+, the
paraphalli forming the “saddle-shaped piece.”

5. Opposite and distinct from these organs, on the right side of the insect, is a
mechanism of a number of pieces forming a receptacle and fitted with a plate which
acts as a lid. Below this, and supporting the base, is an apodeme which, from its
structure and situation, I homologize with the ejaculatory apodeme in Diptera, more
particularly as it is found in the Tipulice.

6. There are also pieces which, though they are not symmetrical, represent the
forcipes interiores, and a hairy process which shows the site of an atrophied palpus.

7. The important great apodemes are difficult to find and equally difficult to display
in a preparation. On either side of the left-hand part of the genitalia (the penis and
its appendages), and forming part of the membrane at the bases, are two insignificant
islands of chitin quite surrounded by a sea of wrinkled, transparent membrane. They
appear to be functionless, and are nearly suboval plates, tapering to a point. Their
microscopic structure is consistent with my idea that they are the atrophied remains
of the apodemes.

These complete the list of pieces in the left-hand part, with the exception of a small
subtriangular piece at the base of the titillator, which appears to be some part of its
articulation to the theca.

A species of Brachytrupes has genitalia very different from Periplaneta. They are
symmetrical ; there is a central stiliform simple penis, which appears to have the opening
of the ductus ejaculatorius at its extremity. This is surrounded by a theca, bearing
forcipes interiores and palpi genitalium, fused to it. A pair of lateral apodemes are

* « Blow-fly,’ p. 664.
t+ ‘The Cockroach,’ Miall and Denny, p. 174.
SECOND SERIES.—ZOOLOGY, VOL. IX. o4

372 MR. WALTER WESCHE ON THE GENITALIA OF

articulated to the theca. At the base of the penis is a sac surrounded by muscles,
which may be some form of ejaculatory apparatus.

I have also made a few dissections and preparations in the Coleoptera, Lepidoptera,
and Hymenoptera, which show that the male genitalia, as might be expected, since
mouth and all other parts agree, can be homologized on the scheme originally con-
structed from the organs in the Diptera.

Coleoptera.—Lucanus cervus, Linn., has a long flagellum similar to that in the Ortalidze
(Pl. 28, fig. 96). This starts from a bulb articulated to the penis, which is supported by
two apodemes. The ejaculatory duct can be traced to the bulb of the flagellum, and
has a structure similar to that on the part in Syritta pipiens. The whole organ is
swathed in several chitinous envelopes, probably representing segments of the abdomen,
and the theca, which immediately surrounds the penis, supports the forcipes interiores.

Acilius sulcatus (Linn.), an aquatic insect, has the “ titillator’’ much developed and
occupying its usual position. This part is also to be seen in a preparation of a land
beetle which I cannot satisfactorily identify.

In Geotrupes stercorarius (Linn.) the theca is long and cylindrical; it bears at its
extremity two articulated plates, the forcipes interiores; through these protrudes a
sensory membrane, which envelops a chitinous structure ; this structure is extruded by
two slender apodemes.

Necrophorus interruptus, Steph., is of the same type, but the forcipes interiores are
longer, and the apodemes cannot be recognized.

Lepidoptera.—In the Lepidoptera the genitalia are mostly simple, often consisting of
only the forcipes superiores and a horny central penis; but in Arctia caja (Linn.),
which has a wide distribution, I find the forcipes superiores seemingly fused into a
single piece, the theca supporting a pair of forcipes interiores and surrounding the base
of a cylindrical penis, with a membranous process running through it; the membrane
having the triangular aculeations so common in Diptera and also to be found on the
genitalia of the drone in Apis mellifica. In Neuronia popularis a pair of apodemes can
be differentiated.

Hymenoptera —A Saw-fly, Cimbex ariana, Kirby, has the hypopygium turned in
under the abdomen exactly as in Glossina; the plate carries cerci and hairy organs,
representing the forcipes superiores and inferiores. It will be remembered that
G. palpalis has forcipes superiores and the remains of atrophied forcipes inferiores in an
identical situation. The theca surrounds the penis and bears the forcipes interiores and
the palpi genitalium ; but the penis in the Tenthredinide consists of paired convoluted
valves, each supported by an apodeme, and the mechanism of the ejaculatory apparatus
(if one exists?) is not obvious. I have found a similar penis in several species of
Nematus.

I think it probable that an examination of a number of families in Insecta will yield
types closer to Diptera than those considered, but there appears nothing to guide the
enquirer to a particular genus or family. I now recur to my main argument, the
relationship between the mouth-parts and the genitalia.

m= or 7 =

————————

BOTH THE SEXES IN DIPTERA. 3738

Further Remarks on the Relationship between Genitalia
and Mouth-parts.

In support of the idea that similar laws govern the growth of the genitalia and the
mouth-parts, I have, in a tentative fashion, prepared a table of relationships, which gives
the parts of the male and female which have already been homologized in the previous
discussion, as well as a list of portions of the mouth-armature which, from structure or
function, appear to be the counterparts of the genitalia. In justification of this selection
I have some evidence which, taken by itself, might be treated as mere coincidence, but
gains weight from the other parts fitting into their places.

Fulcrum or- Submentum.—tIn the mouth-parts of the Muscidze, embedded in the base
of the proboscis, is a chitinous plate which, from its median suture and lateral continua-
tions, is obviously a fusion of two organs. It is used to extrude and control the labium,
having attachments for many muscles. Lowne calls it the “fulcrum,” and says*:
“Gerstfeldt + is, I believe, the only author who has anticipated me in the statement
that the maxille enter into the composition of the fulcrum, but he merely observes,
‘The anterior lancet (labrum) shows distinctly, by the presence of a median raphe, that
it is formed of two halves, which must be the blades of the maxillz (Kieferladen).
They rest upon a piece extending backwards (the fulcrum) which appears to be the
united stipites, from which two slender, nail-shaped parts diverge downwards and
backwards.’ ”

I have called { the part the “ submentum,” which has been thought to be the cardines
and stipites of the second pair of maxille (usually called the labium), differing in this
from Lowne, who considers it to be the first pair of maxillee, because I have shown that,
together with this part (the fulcrum), there exist in Mydrellia griseola, Fall., nearly
complete first maxille, and in identical situations in other Diptera these are nearly
always present, the stipites and cardines often carrying the maxillary palpi.

Relation of the Great Apodeme to the Submentum.—The double apodemes, from their
situation below the theca, from their paired character, and from their function, I
consider to homologize with this part, the submentum or fulcrum.

I have shown the double apodemes to be separate paired organs in some species, in
others partially fused, the upper part being forked, and in other species totally fused. It
is remarkable how analogous this part seems to be with the submentum.

* ¢Blow-fly,’ p. 138.
+ ‘Ueber die Mundtheile der saugenden Insecten. 8vo. Dorpat, 1853.
~ “The Labial and Maxillary Palpi in Diptera,” Trans. Linn. Soc. London, ser, IL. Zool. vol. ix. (1903).

O74 MR. WALTER WESCHE ON THE GENITALIA OF

Analogies.

Great Apodeme.—(1) It is present in its paired form in several (judging from the
geological record) of the oldest families (Tipulidz and Tabanidz).

(2) It is present in its fused condition in the Cyclorrapha, only to be found in the
Oolite, whereas the Nematocera and the Brachycera date from the Lias, or, in simpler
language, it is found in the fused condition in the younger families, but still retaining
traces of its former state.

(3) It performs the functions of an apodeme, governing the movements of the
central organ.

(4) It is situated at the base of the central organ.

Submentum (fig. 187).—(1) This, when part of the second maxillee, consists of the
stipites and cardines of paired organs, the posts or levers, and hinges of a more or less
complicated mechanism.

(2) Itis fused in insects like Per¢planeta, or in the Cyclorrapha, but still retains traces
of its paired state.

(3) It performs the functions of an apodeme, governing the movements of the central
organ.

(4) It is situated at the base of the central organ.

A plate from the ovipositor of an undetermined Asilid from Queensland, that has
already been alluded to, is remarkably like in general outline and some points of
structure to the submentum (fulcrum) in Diptera. This I consider is homologous with
the apodemes, and consequently represents the submentum.

Labrum and Forcipes superiores——KEvidence in favour of another part is also forth-
coming—the identification of the labrum as the representative of the fused forcipes
superiores. I have repeatedly mentioned the hairy nature of these last-named parts,
which will now be seen to have a connexion with the matter under discussion. I will
state my reasons at length for my selection.

Hypopharynz (1).—There is only one wnpaired chitinous organ in the mouth-parts, the
hypopharynx, situated immediately behind the central organ, the labium.

Spinus.—There is only one unpaired chitinous organ in the genitalia, the spinus
titillatorius, situated immediately behind the central organ, the penis; therefore one
represents the other.

Composition of Labrum (2).—Covering and next to the hypopharynx is the labrum, an
obvious fusion of paired organs, with often a cleft at the extremity, and always a
median suture and symmetrical rows of bristle-cavities (Pl. 30. figs. 129,180); probably
a fusion of the laciniz of the maxilla, the true labrum being the mouth-edge, This
exists as a separate part in the Ephydride; Parydra coarctata, Fall.*, shows it
particularly well.

* See Journ. Roy. Micr. Soc., 1904, pl. 8. fig. 7, where, however, I mark the part as the Clypeus, following
the usual practice.

Ee ee ee eee ee eee ee ee ee ee ey

—— =

BOTH THE SEXES IN DIPTERA. 375

Forcipes superiores—Covering and next to the spinus titillatorius are the forcipes
superiores, nearly always hairy and setose, and occasionally fused as in Oliviera lateralis
(Pl. 80. fig. 120) or Rutilia splendida. Therefore the labrum, or fusion of the lacinice
of the maxillee, represents the forcipes superiores.

But it may be pointed out that both the labrum and mavxille are present in the
Empide and Syrphide, as well as in all the flies with complete mouth-parts ; to which
it may be answered that the parts of the maxille present are the galee, which are the
parts aborted in the Muscidze—the often setose character of the laciniz quite supporting
this view.

Fig. 137.

DiaGRAM OF MOUTH-PARTS IN DipTeRa.

pr. Paraglossa. )

m. Mandible. lg. Ligula.

g- Galea. Ip. Labial palpus.
mp. oa palpus. \ Masala. pg. Palpiger. Crain
pf. Palpifer. { mn. Mentum.

s. Stipes. | sm. Submentum.

c. Cardo. J ir, Labrum.

h. Hypopharynx. J

I now venture to submit a table of relationships of the genitalia and the mouth-parts
of Diptera, as, if what I have already advanced can be accepted, the other parts appear
to fall into their places. A counterpart for the very important labial palpi appears to
be wanting, but that may be explained in the same way as their absence in the large
Empid and Syrphid groups is explained.

7

376 MR. WALTER WESCHE ON THE GENITALIA OF

Table of Relationship of Genitalia and Mouth-armature in Diptera.

Mater Geniratia. Ovrposrror. Movru-parts.
lS atom Forcipes inferiores. Ventral piates or egg-guides. Mandibles.
imi eictdo | », superiores. Dorsal bs # Labrum or fusion of laciniz.
Cheer Lamine superiores. | Ventral plate of abdomen, 4th | Mentum.
from end.
Gi stejetaueee Orifice of ejaculatory duct. Aperture of vagina. Orifice of pharynx.
6 Teyeasisvess Theea. Rods and chitinous structures: | Labium.
lamina interior.
ipcacaco | Paraphalli. ” % » Ligule.
Goaanes | Hypophalli. » ” ” Paraglosse.
Hh sale covelhts | Spinus titillatorius. ” ” ” Hypopharynx.
eageon | Forcipes interiores. 3 + os Maxille (galez).
Hesrateusteiehe Palpi genitalium. » » 73 Maxillary palpi.
Mss Beene as Double apodemes. Double rod, Musca domestica. Submentum ; or fused stipites
| Apodeme, Simulium. _ and cardines of 2nd pair of

maxille.
Dk ores | Ductus ejaculatorius. Vagina. Pharynx.
Olevicatery: Sacculus ejaculatorius. | Utero-vaginal tube. —
jPeaosos Ejaculatory apodeme. | ——s? pee
PoAsons Vas deferens. | Oviduct. (Esophagus.
Bae Paragonia. Parovaria or receptacula seminis. —
Pe so Vasa efferentia. Tubsee. ==
Ub cra teuetessi| Testes, Ovaries. —

I have often sought for the reason why the armature of the male mouth is less than
that of the females in the blood-sucking Culicide and Tabanide.

Compensatiom of development of Trophi and Genitalia—The idea of an identical law
of growth affords some explanation, because if that law exists it would include com-
pensation ; that is to say, if the mouth-parts were much developed, the genitalia might
be expected to be simpler or vice versa. Let us apply this test, and see how it
works out.

Differences in mouth of males and females.—The armature of the female Tabanid
mouth consists of mandibles, maxillze (the stipites, cardines, and galez), the maxillary
palpi (the labial are aborted), the bypopharynx and the part usually called the labrum
(the fusion of the laciniz of the maxillz), and the labium bearing the usual tracheated
paraglossze (labella).

The males have no mandibles, and the maxille are rudimentary or atrophying, but
have the other parts that the female possesses.

1. Now as regards the genitalia, the females have a very simple type of ovipositor,

BOTH THE SEXES IN DIPTERA. 377

and one not capable of extrusion; while the male, as I have shown in Zabanus bromius,
has complex genitalia.

Nearly identical conditions are found in the Culicide.

Cases where the genitalia are complex in both sexes :—

2. The male has very complex, though not complete, genitalia in Tipula oleracea,
and the female a striking ovipositor. The mouth-parts are comparatively simple, the
mandibles being fused into the labium ; the galeze of the maxillz and the labial palpi
are absent; the labrum and hypopharynx are very imperfect; the trachez of the
paraglossee are not well-developed, and the whole organ seems to have undergone great
changes. Jn the Muscidee the genitalia are quite as complex, and we find nearly the
same armature on the mouth, except that the palpi are labial and not maxillary.

Cases where the mouth-parts are nearly aborted :—

3. In Gastropilus equi the mouth-parts consist only of an aperture and two hairy
processes. ‘The compensation in this case seems to be in size, as the ovipositor is a very
prominent part of the insect, and the genitalia of the male are large and chitinous.

Cases where the genitalia incline towards simplicity :—

4. Parydra coarctata has comparatively simple genitalia, the spinus and_ palpi
genitalium being absent. ‘The mouth-parts are proportionately very large; a distinct
clypeus (which I have submitted is the true labrum) is present, and the structure of the
tracheze is abnormal.

Case where the genitalia is simple in mechanism but well-developed as to size :—

5. In Hmpis stercorea, whose genitalia I have already commented on, the mouth-
parts are large, the labium, maxilla, hypopharynx, and labrum well-developed and
strong, the only weak parts being the maxillary palpi.

So-called teeth on Paraglosse :—

1. In all cases where the ovipositor is of the Musca domestica type, the teeth on the
labella are thin and transparent, or absent.

2. In all cases where the ovipositor has hardened, as in the Ortalide and Loncheidz,
the teeth are quite absent.

3. In most cases where the teeth are strongly developed, as in the Cordyluride, the
ovipositor is short as in Scalophaga stercoraria (Linn.) (Pl. 30. fig. 181); the exceptions
are in the Cenosia group of the Anthomyzidee, where very chitinized teeth are found,
together with a long, membranous ovipositor. The males are of the Muscid type, which
is complicated as a whole, but tends towards simplicity of the individual organs.

Summary as to the failure of parts in the male mouth.—From these cases it may be
assumed that when the male mouth-armature is failing or variable, it is brought about
by reason of a relationship between mouth and genitalia, the over-development of one
part causing, by compensation, the failure or simplification of the other. That being so,
the nature of the food in the blood-sucking species has no connexion with the failure
of the mouth of the male, though it may possibly have acted indirectly on the genitalia
by reason of its stimulating character.

The fact that in Stomorys, Hematobia, or Glossina both sexes are blocd-sucking and
identical in mouth-armature, supports this view, as the mouth-armature is, though

378 MR. WALTER WESCHE ON THE GENITALIA OF

highly modified, also much simplified. The labrum or upper lancet is only equalled in
simplicity by that part in the Culicidee, while the hypopharynx is in an atrophying
condition in Glossina. In Stomoxys the male genitalia are relatively small, though in
Glossina they are highly developed and modified, yet both are simplifications of the
Muscid type, owing to the abortion and fusing of the appendages of the theca, the
spinus, genital palpi, and interior hooks. There is no extrusile ovipositor in Glossina,
but in Stomoxys there are indications that at some period it resembled that of
Musca domestica ; the joints appear to be there, but in a fused condition. I spent a
considerable time unsuccessfully endeavouring to extend the organ. There is also
simplification in the number of receptacula, which is always two in this group.

Analogy in the Mammalia.—In addition to what I have advanced in favour of the
theory of the intimate connexion between the growth of the genitalia and of the
mouth-parts, it may be pointed out that there are analogies in the Mammalia, where
the effects of excision of the genitalia on the throat and voice, as well as beard, of
the male are common knowledge.

Summary of the main argument.—I have now shown: (1) That the male genitalia
and the mouth-armature are on the same general plan, of a central perforated organ
surrounded by aculeate and sensory or possibly sensory appendages. (2) That they
coincide in many details of structure and arrangement. (3) That the male and female
genitalia are homologous. (4) That the male armature, the ovipositor, and the mouth-
parts have a central mechanism (double apodemes, apodeme of ovipositor, fulerum or
submentum), which guides or governs the whole, or traces cr remains of such a part.
(5) That the application of an hypothesis founded on the above data accounts for the
hitherto unexplained failure of the male mouth-parts in the Culicidz and the Tabanide.
(6) That an intimate connexion is known to exist between the male genitalia and
the throat, voice, and the hairy appendages of the mouth in the Mammalia.

Ancestral form of the Arthropoda.—Further, | am informed by Mr. H. M. Bernard,
F.LS., &e., that the idea that genitalia and mouth-parts are homologous is not new ;
the hypothesis has been advanced that, in a primitive ancestral form of the Arthropoda,
these mechanisms were formed by adaptations of the legs of the end segments.

My opinions are arrived at by a close study of the mouth and genitalia of the
principal families in Diptera, and almost wholly by methods of comparative anatomy.
My arrival at what is practically the same conclusion, by an enquiry comparatively very
narrow in its scope, being confined to a single Order, is evidence in favour of this
hypothesis.

Weighing these facts, I submit that there is much evidence in favour of my main
thesis, which may be formulated thus :—

Formula.—The external appendages of both extremities in Insecta are derived from
two organs of the character of maxille, and all the variations of the parts are
adaptations of these organs.

BOTH THE SEXES IN DIPTERA. 379

Metuops oF Work.

For the convenience of those who wish to examine genitalia or to test my statements,
and are unacquainted with the usual procedure, I shall now explain my methods of
work on these very minute parts. I will presume that the enquirer has the usual
outfit of the worker in microscopy. Fresh or newly-killed insects are undoubtedly the
best subjects; but specimens preserved in alcohol are nearly as good, and old pinned dry
ones are often extremely useful. The genitalia, protected in a cavity, are as a rule
intact, though the insect may have lost every other appendage. If a species is to be
thoroughly studied, at least three specimens of each sex ought to be procured. I assume
that they are newly-killed insects. (1) Immerse one of each sex in 15 per cent. solution
of caustic potash; 10 per cent. solution will do but takes longer. ‘This is to clear the
insects, ‘and the length of the process will depend on the density of the chitinous
structure; a day may be enough, or it may take over a fortnight. In this process all
the muscles and nerves are dissolved and destroyed, leaving the membranes, the exo-
skeleton, and all the chitinous structures. (2) After this is accomplished the prepara-
tions must be thoroughly washed in water; and (3) then placed in glacial acetic acid for
24 hours to get rid of the potash crystals and any fat that may have escaped the action
of the chemical. (4) They must again be washed in water. (5) After this they are in
a flaccid state, and can be readily arranged on a 3-inch slip in any desired position.
The wings will be found to flatten better if the specimen is floated on to the slip. In
the case of the male a lateral arrangement is best, as it more clearly shows the genitalia
in their natural positions; but where possible it is well to have a preparation showing
the ventral surface as well. A pair of strong clips, such as are used to hold papers, is
useful in the next process. (6) After the insect has been arranged on the slide, another
slip is placed above it; it is squeezed flat by applying a clip at one end, and the process
is completed when the second clip is applied at the other end. There is now an oppor-
tunity for roughly examining the preparation with low powers to see that the organs
are properly displayed, and that (for instance) a claw does not cover the hypopygium, or
the trophi of the head. are not hidden. If unsatisfactory, undo the clips, separate the
slides (care is required in this), drop some water on the insect, and rearrange it. If,
however, the initial effort is successful, tie the slips tightly together with fine twine,
remove the clips, and (7) immerse in methylated spirit for at least 24 hours. In cases
where the preparations have been over cleared, a stain is useful. Aniline blue is best in
my experience ; fuchsin is also good, but care must be taken not to overstain. A drop
can be placed on the insect before the upper slide is applied. In the spirit the prepara-
tions will be dehydrated, and on the result of this process the success of the preparation
as a microscopic object depends. After that they will be ready for transference to spirit
of turpentine, which should be of the best quality procurable. (8) To do this, the slips
must be withdrawn from the spirit with a pair of forceps, the string untied, the slips
slowly and carefully separated, and the preparations taken from the slide on a section-
lifter. (9) After being at least 24 hours in turpentine they are ready for mounting in
Canada balsam, though many microscopists recommend a further clearance in cloye-oil.

SECOND SERIES.—ZOOLOGY, VOL. IX. 55

380 MR. WALTER WESCHE ON THE GENITALIA OF

(10) A preparation is then lifted out with a section-lifter and placed in the centre of a
slip, and examined on a microscope to see that no hairs, cotton-wool, or dust disfigure it,
and that it presents the ventral side (if it is not a lateral preparation). (11) The moiety
of the turpentine is then removed by means of blotting-paper, taking care that not too
much is taken, as otherwise air will enter into the cavities. (12) Canada balsam
dissolved in xylol is then applied, and the whole sealed with a cover-glass, which may be
pressed down by the weak clip sold for this purpose. These preparations will be dry
enough for use in about two days, and must be carefully studied, so that a good idea of
the form and of the relative situation of the parts is arrived at, before proceeding to the
separation and dissections which must be undertaken later. The microscopist will
require a good high power. I recommend a 7th of Mr. Pillischer, of New Bond Street,
as the best I know of for the purpose, as it is comparatively inexpensive, works at a
greater distance than any similar magnification that I am acquainted with (which is
its peculiar advantage over lenses of higher angular aperture), and gives good definition.
With this must be used a powerful substage condenser to enable the light to pierce
through the more chitinous parts.

To study accurately, drawings must be made, and where it is necessary to compare
shape and size, those drawings ought to be made on squared paper corresponding to
squares in the eyepiece of the microscope.

This is the comparatively easy part of the work; in what follows the student must
not be discouraged by repeated failure, as he has to acquire the steadiness of nerve and
nicety of movement necessary for dissecting under the microscope.

(1) The end of the abdomen must be removed and placed in a drop of water on a live-
box or compressorium (the cover being removed), and the organs “‘ teased”’ apart with
the finest needles procurable. The forcipes superiores and inferiores must be separated
from their articulated bases, and the penis and its appendages brought out free from its
adhering duct and muscles. Place the cover-glass on, and the separated parts can be
roughly examined with low and higher powers, and drawn, if desired. The muscles will
often be found very hampering and obscuring; a drop of potash solution will facilitate
their removal. If this is used, before proceeding further, drop water on the dissections
and endeavour to remove the potash as thoroughly as possible ; a little acetic acid will
help, but this also must be washed away.

(2) Take a 2" cover-glass (thickest size), place a small drop of spirit on it, and place
on it the parts, by means of a needle ora bristle. This is very difficult, and the cover-
glass must be examined with a lens to judge of success. Care must be taken to avoid
the loss of parts by dropping off the needle, &c. Place some more spirit on the glass
and place it on a piece of white paper, and both on the hot-plate, which must then be
gradually heated. (8) As the spirit evaporates it must be replaced by dropping fresh ;
a glass rod drawn to a point answers best. This must be continued till all the water
has been driven out of the dissections; it will sometimes be seen as a milky fluid on the
edge of the glass, and can be removed with blotting-paper.

(4) Turpentine must now be dropped on the (by now) hot slide, and allowed partially
to evaporate, always dropping fresh before the specimens are dry. Now put out the

_ a ae ! a

BOTH THE SEXES IN DIPTERA. 381

lamp and let the preparations gradually cool, never letting them get dry, or air will
spoil them. (5) When cool, remove the superfluous turpentine with blotting-paper ; and
as quickly as possibly (6) place Canada balsam and a smaller and thinner cover-glass on
the dissections.

(7) This $-glass, when dry, can be mounted between three slips of cardboard, the upper
and lower punched with a circular hole (fig. 139), and the middle cut to the shape of the
larger cover-glass (fig. 138). They can then be gummed together and placed in a press.

(8) For permanently sealing both these preparations I recommend spirit-varnish.

This last preparation mounted thus can be examined on both sides with high powers,
which is necessary for a proper understanding of the mechanism and of the real shape
of the parts.

Fig. 138. Fig. 139.

Middle piece. Lower and upper piece.

In many cases it will now be necessary to take a fresh insect, dissect out the penis,
and then undertake the still more difficult task of separating that into component parts,
proceeding to mount it exactly as in the foregoing case.

The dissections can be done with an inch objective on a strong microscope with a flat
Stage; a telescopic tube is a convenience, as it permits of a quick amplification of
magnitude.

The insects which have been preserved in spirit only need to be thoroughly soaked
in water before dissection, but they will be found more brittle, and consequently more
difficult to handle, though, on the other hand, muscles will detach more readily.

Dry specimens.—In the case of dried specimens, they may be placed in hot water for
half an hour before dissection, or soaked for a few hours in cold.

Air-bubbles—Air-bubbles can be driven to the side of the cover-glass by means of
heated needles (not red-hot, or they will crack the cover-glass), or a nail is better, as it
retains the heat longer.

55*

382 MR. WALTER WESCHE ON THE GENITALIA OF

Chilled balsam.—Sometimes, in spite of care, the specimen has not been sufficiently
dehydrated, and the balsam is “chilled” and the whole slide a failure, as nothing can be
clearly seen. ‘The remedy requires careful nicety :—(1) Place a piece of white paper on
the hot-plate; on the paper place the preparation. The paper prevents the glass
becoming too hot, and also shows the situation of the dissections, which often cannot be
seen without the white background. The balsam will soon melt, and the air will run
in; (2) then lift off the upper glass with a fine pointed forceps, and place it balsam-side
uppermost on the paper.

(N.B.—Do not use the forceps again till it has been dipped in spirit and the points
wiped. The neglect of this precaution has caused many disasters, and the flow of many
tears of the Recording Angel necessary for the obliteration of words *.)

(8) Drop turpentine on the preparation; this will drive all the balsam to the sides,
leaving the dissections in the middle, where the preparation can have a thorough
washing in evaporating turpentine repeatedly dropped on it. (4) Then put out the lamp
and proceed as before, dropping balsam, and sealing with the old or a fresh cover-glass.

Sections.—If abundance of material is available, sections of the hypopygium can
be cut.

Mounting without pressure—There yet remains another method, but it cannot be
applied to insects larger than the house-fly, Musca domestica. This is the mounting of
the whole insect without pressure ; the initiatory processes are identical till it arrives at
the pressing stage. (5) Instead of being placed on a slip, it is placed in a small saucer
on its back and a little water poured in. The wings, legs, and proboscis are arranged in
their desired position. (6) The water is drawn off and replaced with spirit; when the
specimen is quite stiff it may be lifted out with a section-lifter and placed in a closed
receptacle also filled with spirit, where it should remain for 48 hours or as much longer
as is convenient, to set thoroughly and dehydrate. (7) It is desirable that this receptacle
should have a flat bottom, as otherwise the arrangement of the insect will be altered
and possibly spoilt. (8) It is then lifted out very carefully, with the section-lifter as
before, and transferred to turpentine, but this must not be done till the operator is quite
satisfied that it is very thoroughly dehydrated. Failures in this part of the process are
more likely in this type of mounting than in the pressure type. (9) After 48 hours in
the turpentine the insect can be lifted out and placed on a slide. (10) Cover it with
plenty of balsam and place round it (11) three or four glass beads of a suitable size (I
prefer transparent ones, zine rings are equally good); these must have been prepared
(12) by being washed in alcohol and afterwards in turpentine. Transfer them straight
from the turpentine to the balsam, and their preparation will ensure that no air hangs
to them and that they sink readily into the thick medium, (18) Place a cover-glass on
the insect, which will not be injured by the pressure as the cover-glass will rest on the
beads. (14) Put a weight (a small bullet answers well) on the cover-glass and then run
in balsam under it till it is quite full; this will shrink in a day owing to the evaporization
of the xylol; (15) and fresh must then be run in till it is quite set and firm, when it
may be ringed in the usual manner,

* Tristram Shandy: ‘* And the Recording Angel, as he wrote it down, blotted it out with a tear.”—Sreryn,

BOTH THE SEXES IN DIPTERA. 385

This preparation is best suited for examination with a binocular microscope and low-
power objectives, and the position of the interior organs can be well seen in a successful
slide. It remains to say that when, only the chitinous and membranous structures are
wanted, dissection is greatly facilitated by 24 hours’ immersion in the potash solution.

EXPLANATION OF THE PLATES.

The following letters are used in the female genitalia :—

a. Valvula inferior, or ventral egg-guide. t. Tuba.

6. Valvula superior, or dorsal egg-guide. u. Ovarium.

d. Aperture of the vagina. w. Lamella anterior.

m. The apodeme or the double rod. z. Receptaculum.

n. Vagina. y. Egg (ovum).

o. Utero-vaginal tube. z. Glue-gland. :

7. Oviduct. x’, Rectal papilla (Dolichopus only). A

s. Parovarium. >
The following letters are used in the male genitalia :— ||

a. Forceps inferior. k, Palpus genitalium.

6. Forceps superior. m. Double apodemes. Ye

c. Lamine superiores. n. Ductus ejaculatorius.

d. Orifice of ejaculatory duct. o. Sacculus ejaculatorius.

e. Theca. p. Hjaculatory apodeme.

f. Paraphallus. r. Vas deferens.

g. Hypophallus. s, Paragonium.

A. Spinus titillatorius, t. Vas efferens.

i, Forceps interior. u. Testis.

PLATE 23.

Fig. 1. Genitalia of Sciara thome (Linn.), 3.
2. Genitalia of Bibio hortulanus, Linn., 9.
3. Receptaculum seminis of B. hortulanus, Linn., 2? .
4, Penis and appendages of B. hortulanus dissected out of abdomen.
5. Forceps inferior of B. hortulanus 3.
6. Palpus genitalium of Scatopse notata (Linn.), removed from its place on fig. 7, where it is

hidden by the forceps interior.

7. The genitalia of S. notata, drawn diagrammatically to show the arrangement of the parts;

seen from the ventral aspect.

8. Ejaculatory apodeme of S. notata, enlarged.
9. Diagram of the sacculus ejaculatorius and its apodeme and duct (S. notata).
10, One of the interior hooks in Culew (?).

11. Extremity of abdomen of S. notata, showing the single receptaculum.

12. Second interior hook in Culex (?).

This and that figured as 10 are opposed by a similar pair of

hooks on the front of the central organ.

13. Central organ in Culex (?).

14, Forcipes superiores of Anopheles cinereus.

— =

18.
19.

i=

Fig.

Ww WwW Ww Ww
woe

ri)
= ©

36.

MR. WALTER WESCHE ON THE GENITALIA OF

5. Central organ of Gynoplistia bella, Westw., showing the three apodemes and the changes in the

theca.

}. Forceps superior of Culew (?). This part in C. pipiens only differs by the articulated bristle at

the extremity of the hook being of a different shape.
. Ventral view of hypopygium of Ptychoptera scutellaris, showing the situation of the ejaculatory
apodeme ; the penis has been removed, and is shown in fig. 20.
Ejaculatory apodeme of P. scutellaris.
Forceps superior of Gynoplistia bella.

PLATE 24,

. Central organ (penis) in P. seutellaris.

. Ovipositor of Tipula oleracea, Linn., lateral view.

. Forceps superior of 7’. oleracea.

. Theca and flagellum of 7°. oleracea.

. Diagram of the last segments of the male abdomen of T. oleracea, to show the arrangement of
the ejaculatory apparatus. The forcipes superiores and inferiores have been removed.

. Forceps inferior of 7. oleracea.

. Ejaculatory sac and apodeme of 7. oleracea.

. One of the processes which constitute the sense-organ on the plate of the forceps inferior of
T. oleracea, seen from above.

. The sume part in section.

. Theca and flagellum of Pachyrrhina maculosa, Meig.

. Ejaculatory sac and apodeme of P. maculosa.

. The same apodeme enlarged.

. Extremity of the ductus ejaculatorius and forcipes interiores of Beris vallata (Forst.).

. The last segments of the abdomen of the male of B. vallata, drawn diagrammatically to show
the arrangement of the parts.

|. End of abdomen of B. vallata 2.

. End of abdomen of B. nigra, Meig., ? «

Forceps inferior of an undetermined Asilid from Mt. Gambier, S. Australia.

. Forceps superior of same insect.

. Ejaculatory duct of same insect, showing the arrangement of the ejaculatory sac and apodeme
in the base of the theca.

PLATE 25.

. Penis of Hmpis stercorea, Linn.

. Forceps superior of H. stercorea.

. Penis of Hilara cilipes, Meig.

. Ovipositor of H. cilipes.

. Hypopygium of Dolichopus plumipes (Scop.).

. Orifice of the ejaculatory duct and the end of the theca of D. festivus, Hal.

. Spinus titillatorius of D. festivus, Hal.

. Forceps interior of D. festivus.

. Palpus genitalium of D. festivus.

. Palpus genitalium ? of D. festivus (doubled part).

. End of penis of Pecilobothrus nobilitatus (Linn.).

. Penis, ejaculatory apodeme, and vas deferens of Dolichopus griseipennis, Stann.

. Part of same, enlarged and drawn diagrammatically to show the working of the ejaculatory
apparatus.

i

_—

Fig. 72.
. The same of G. pallidipes, Aust.
. The same of G. morsitans, Westw.

Fig. 88.
. Penis and appendages of Norellia spinimana, Fall.
. Ovipositor of N. spinimana.

BOTH THE SEXES IN DIPTERA. 385

. Ovipositor of Dolichopus griseipennis, Stann.

. Penis and appendages of Catabomba pyrastri (Linn.).
. Forceps inferior of Syritta pipiens (Liun.).

. Foreeps superior of S. pipiens.

PLATE 26,

. Penis of Syritta pipiens, drawn diagrammatically to show the apodeme and ductus ejaculatorius

in the interior; seen from the front.

. Side-yiew of the same part, and drawn in a similar manner to fig. 56.

. Penis of Hristalis tenax (Linn.).

. Forceps interior of E. tenaz.

. Penis of Sarcophaga carnaria (Linn.).

. Forceps inferior of S. carnaria.

. Penis of Pollenia rudis, Fabr.

. Extremities of the paraphalli in Glossina palpalis, Des.

. Diagram of the opening of the ejaculatory duct, and showing the sac and apodeme, of Eristalis

tenax.

. Forceps inferior of Anthomyia radicum (Linn.).

. Lamine superiores from an undetermined Anthomyid from Maryborough, Queensland,

. Ejaculatory apodeme of Glossina palpalis.

. Penis and appendages of Calliphora erythrocephala, Meig.

. Diagram of the arrangement of apodemes, theca, and paraphalli of Glossina palpalis.

. Forceps superior of Anthomyia radicum.

. Extremity of paraphallus of C. erythrocephala, to show the serrated edges characteristic of the

structure.

PLATE 27.

Double apodeme and ejaculatory apodeme of Glossina palpalis.

. The same of G. tachinoides, Westw.

. Forceps inferior of G. palpalis.

. Forceps inferior of G. tachinoides.

. Penis of G. pallidipes.

. Forceps inferior of G. morsitans.

. Forceps inferior of G. pallidipes.

. Forceps inferior of Scatophaga litorea, Fall.
. Extremity of penis of G. morsitans.

. Forceps superior of S. litorea.

. Penis and appendages of G. tachinoides.

. Penis and appendages of Anthomyia radicum,
. Fused palpi and forceps of G. tachinoides.

. Lamine superiores of S. litorea.

PLATE 28.

Penis and appendages of S. litorea.

. Forcipes inferiores and superiores of N. spinimana.
. Forcipes inferiores of Sepsis cynipsea (Linn.).
. Forceps interior (?) of S. cynipsea.

386 ON THE GENITALIA OF DIPTERA.

Fig. 94, Ejaculatory apodeme of Sepsis cynipsea.
95. Ovipositor of Ulidia nigripennis, Loew.
96. Penis of U. nigripennis.
97. Laminz superiores of Norellia spinimana.
98. Penis of Notiphila nigricornis, Stenh.
99. Penis and apodeme of Sepsis cynipsea.
100. Ovipositor of Anthomyia pallida, Zett.

PLATE 29.
Fig. 101. Penis and appendages from an undetermined species of Nemopoda from Jersey.
102. Forceps inferior of same insect.
103. Lamine superiores of same insect.
104. Diagram of genitalia of the male, showing arrangement of parts as usually found in Diptera,
seen from ventral aspect.
105. Ejaculatory apodeme of an undetermined species of Nemopoda from Jersey.
106. The plates at the extremity of the ovipositor in Musca domestica, Linn.
107. Ovipositor of M. domestica.
108. Diagram of the genitalia of the male, seen in profile.
109. Genitalia of Tabanus bromius, Linn,
110. Penis of Parydra coarctata, Fall.
11]. The same seen from the front.
112. Penis and appendages of Oliviera lateralis (Fabr.).
113. Receptaculum of Pulew irritans, Linn.
114, Receptaculum of Parydra coarctata, Fall.
115. Receptaculum of Scatophaga lutaria (Fabr.).

PLATE 380.

Fig. 116. Receptacula of Scatophaga HD (Linn.).

117. Penis of Chalurus spurius, Fall.

118. Ventral side of abdomen of Phorocera serriventris, Rob.-Des.

119. The hook of the ovipositor, as seen laterally, of the same insect.

120. Fused forcipes superiores of Oliviera lateralis.

121. Ovipositor of Lauxania enea, Fall.

122. The base of the penis in Towxoneura muliebris, Harr.

123. Receptaculum of Tabanus bovinus, Linn.

124. Rectal papilla of Pacilobothrus nobilitatus (Linn.).

125. A scale enlarged from the same part.

126. Ovipositor of Tabanus bovinus.

127. Extremity of penis of 7. muliebris.

128. Extremity of duct of receptaculnm of 7. dovinus.

129. Labrum of Parydra coarctata.

130. Labrum of Empis tessellata, Fabr.

131. Ovipositor of Scatophaga stercoraria.

132. Receptacula of Seioptera vibrans (Linn.).

133. Receptaculum of Anthomyia radicum (Linn.).

134. Receptaculum of Pegomyia bicolor, Walk.

135. Receptaculum of Homalomyia manicata, Meig.

136. Receptaculum of Schenomyza cinerella, Fall.

| ~S

Wescheé

|
WWesche del

GENITALIA

OF THE

Wescheé Trans Linn Soc. SeER.2. Zoor Vor. IX. Pu.24

} WWesche del London Stereoscopic Co. imp

GENITALIA OF THE DIPTERA

: ae 4
‘ . 7
: a
nw =>
ies :
Co 7 \
. _
: es
aa ;

y ei - ®

nel a ies

Wesche

is aan

WWesche del

Trans. Linn Soc. Ser.2. Zoor.Vor. IX. Pi.2d

=X

pee

'
, as

\
—,

imal
Pox
a bial

”

Landon Stereoscopic ©

Wesche Trans. Linn.Soc Ser 2. Zoonr. Vou 1X. Pu.26.

>" gg coe an

; I qua | “l/

2
V

OL eet
‘a v \
1

ets.
:

x

7
inerahy

7)

or,.Vor.1X

Z. 16

9
fa

Linw.Soc. SER

TRANS

TAT
Wesche

WWeeche del

z=
*
‘ = >.
- = 7
; ‘ “ha
x s . iar
= ™ >
P =
= aj
é 7
: - }
es re
y
‘ ra cy
=
bed =
>
:
; :
Pe
, b ~
.
= .
"
,/
-
——

W Weechs

: i
‘
'
¢ ‘
.
.
*
.
\

-

!
. ad
\ P 7
.
s
‘ ?
‘ \ 7. <
. 1 .
.
Py ‘
4
‘
¥
ra
, 7 es.
/ a n
| - 7 a
' ‘
My -
i fh
‘ q a and

WWesche del

ie : F : oni

« =
ol tithe
- “oh
¥
=i . _ ro
ao ~

Wesche

Trans lam Soc. sen.2.Zoor.Vor.IX, Prod.

WWesche del

London Stereoscopic Co im

F THE DIPTERA

[3sv 4

XI. On a Collection of Crustacea, Decapoda and Slomatopoda, chiefly from the Inland
Sea of Japan ; with Descriptions of New Species. By Dr. J.G.pE Man, of Terseke
(Holland). (Communicated by the Rev. T. R. R. Stersine, W.4., F.R.S., FL.S.)

(Plates 31-33.)

Read Ist November, 1906.

THE present collection, which was entrusted to me by Prof. F. Jeffrey Bell, of the
British Museum, London, consists, firstly, of 30 species of Decapod and 2 of Stomatopod
Crustacea, collected in the Inland Sea of Japan, mostly in deep water; secondly, of
7 Decapod species from four other different localities. The last named are interesting
not only on account of two novelties, a new Parathelphusa and a new Paleimon,
discovered respectively in the Chinese province of Yunnan and at Darjeeling, but also
by the Mediterranean Sicyonia sculpta having been captured off Bahia; the most western
limit of geographical distribution of this species was, indeed, hitherto the Cape
Verde Islands, so far as Iam aware. For Potamon spinescens, Calm., a new subgenus,
Parapotamon, is created.

The Crustacea from the Inland Sea of Japan proved also to be of much interest.
Five species are new to science, viz., a remarkable small Lambrus, for which a new
subgenus, Oncodolambrus, is created, two new species of Crangon, and two of the genus
Spirontocaris. Most of the other species are also remarkable. Thus a small species
of Pinnotheride, viz., the rare Asthenognathus inequipes, Stimps., was captured, a form
described in 1858 and not found again since that year. I wish also to draw attention
to the rare Arcania globata, Stimps., Galathea acanthomera, Stimps., and Leander
longipes, Ortm. The discovery of the male of Spirontocuris rectirostris (Stimps.) is
interesting; it shows considerable sexual differences from the hitherto only known
female. Spirontocaris pandaloides (Stimps.), of which several specimens were caught,
is also one of the numerous rare forms described, almost half a century ago,
_ by that eminent American naturalist, which have not occurred in literature since
that time.

Preliminary diagnoses of six new species have already been published in the ‘ Annals
and Magazine of Natural History,’ ser. 7, vol. xvii, 1906, pp. 400-406, and of the new
Parathelphusa in the ‘ Zoologischer Anzeiger’ of March 20, 1906.

SECOND SERIES.— ZOOLOGY, VOL. IX. 56

388 DR. J. GQ. DE MAN ON CRUSTACEA CHIEFLY

List of Species.

A.—InLAND SEA oF JAPAN.

Lambrus (Oncodolambrus) predator, de Man. Spirontocaris propugnatriz, de Man.

Lupa (Hellenus) hastatoides (Fabr.), de Haan. S. alcimede, de Man.

Platygrapsus depressus (de Haan). S. pandaloides, Stimpson.

Asthenognathus inequipes, Stimpson. Latreutes planirostris (de Haan).
Trigonoplaz unguiformis (de Haan). L. acicularis, Ortmann.

Leucosia rhomboidalis, de Haan. L. laminirostris, Ortmann.

Myra fugax (Fabr.). Hippolysmata vittata, Stimpson.

Arcania heptacantha (de Haan). Alpheus brevirostris (Olivier).

A. globata, Stimpson. A, japonicus, Miers.

Galathea acanthomera, Stimpson. Peneus (Metapeneus) lamellatus, de Haan.
Crangon consobrinus, de Man. P. (Metapeneus) akayebi, Rathbun.

C. cassiope, de Man. P. (M.) acelivis, Rathbun.

Sclerocrangon angusticauda (de Haan). P. (Parapeneopsis) tenellus, Sp. Bate.
Leander longipes, Ortmann. P. (Trachypeneus) curvirostris, Stimpson.
L. paucidens, de Haan. Chloridella affinis (Berthold).
Spirontocaris rectirostris (Stimpson). C. fasciata (de Haan).

B.— axe at Yunnan-Fu, Cura.

Potamon (Parapotamon) spinescens, Calman. | Potamon (Parathelphusa) endymion, u. sp.

C.—DarseeLine, BENGAL.

Palemon (Parapalemon?) hendersoni, de Man.

D.—Taurspay Istanp, Torres Srraits.

Penaeus (Peneus) latisulcatus, Kishinouye, var.?

E.—Coast orr Banta.
Peneus (Peneus) brasiliensis, Latr. Sicyonia carinata (Olivier).
Sicyonia sculpta, H. M.-Edw., var. ?

A.—INLAND SEA OF JAPAN.

LAMBRUS, Leach.
OncoDOLAMBRUS, de Man.
Oncodolambrus, de Man, in Ann. & Mag. Nat. Hist. ser. 7, vol. xvii. 1906, p. 400.

Carapace broadly triangular, much broader than long. Rostrum acute, projecting
and strongly deflexed. No postocular constriction. Branchial regions extraordinarily
swollen, globulate, rounded, much higher and broader than the narrow cardiac region
and devoid of tubercles and spines. Pterygostomian regions traversed by a ridge that
runs parallel with the antero-lateral border. Chelipeds of moderate length, their
margins dentate, their surfaces smooth. Ambulatory legs short.

Related to Platylambrus, Stimps., but distinguished by the considerably inflated and
swollen branchial regions that are not tuberculate. The subgenus Parthenopoides,
Miers, differs by the postero-lateral margins of the carapace running nearly in a straight
line with the posterior margin.

FROM THE INLAND SEA OF JAPAN. 389

LAMBRUS (ONCODOLAMBRUS) PREDATOR*, de Man. (PI. 81. figs. 1-3.)
Lambrus (Oncodolambrus) predator, de Man, in Ann. & Mag. Nat. Hist. ser. 7, vol. xvii. 1906, p. 400.

One male from Japan, the locality not defined.

Probably a species of small size. Measured in the middle line the carapace appears to
be 7 mm. long, the front included, and the greatest breadth at the angles between the
antero- and postero-lateral borders measures 10 mm.: the broadly triangular carapace is
thus nearly once and a half as broad as long. The triangular, subacute front is
prominent, but strongly, obliquely, deflexed. The lateral margins are parallel, though
slightly concave, between the eyes and then curve inward ; they are smooth and entire,
but, on each side, the subfrontal process is visible as a small tooth or prominence, when
the front is looked at from above, the subfrontal process being situated almost as
far distant from the tip of the rostrum as from a transverse line that runs along the
posterior border of the orbits. The breadth (1'8 mm.) of the front at its base is almost
one-fifth of the greatest breadth of the carapace. The smooth upper surface of the front
is concave between the eyes; the groove, here rather broad and deep, becomes gradually
more shallow anteriorly; the groove gradually narrows backward on the upper surface
of the somewhat elevated gastric region until its posterior end, when one observes a low
rounded tubercle in the middle line.

The gastric region is slightly inclined from behind forwards. A little in front of the
round tubercle the gastric region carries, on either side of the middle line, another obtuse
tubercle that is much smaller and much less prominent. The cardiac region carries, in
the middle line, two obtuse tubercles one behind the other, which are as large as the
tubercle at the posterior end of the gastric region; the anterior cardiac tubercle is once
and a half as far distant from the gastric tubercle as are the two cardiac from one another,
Behind these prominences, which are, however, not so high as the swollen, branchial
regions, one observes, on the posterior slope of the cardiac region, two other smaller
tubercles, the anterior of which is probably double. The slightly convex and granulated
posterior margin of the carapace carries five tubercles, namely, in the middle three
smaller ones, of which the median one is a little larger than the two others and are
contiguous to one another, and a larger tubercle on each side more laterally. The
tubercles of the gastric and cardiac regions as also those of the posterior border of
the carapace appear granulated under a very strong lens. The intestinal region carries,
on each side, in the angle between the cardiac and branchial regions, two very low
prominences, separated by a shallow groove, the anterior being somewhat larger than
the other.

The distance between the external orbital angles, which are not at all prominent,
measures almost one-third of the greatest breadth of the carapace. The hepatic ares
situated between the orbits and the swollen, branchial regions are deeply concave ; they
are smooth, like the gastric and branchial regions, but finely punctate, the purctuation
being more crowded on the gastric region. The considerably swollen ad inflated

* Predator, robber: because, when looked at from in front, the crab seems to be burdened on each side with
its prey.
56*

390 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

branchial regions are nearly globular and very large, being twice as broad as the gastric
region. There is no postocular constriction of the carapace. The antero-lateral margins
which run at first outward, then curve backward and upward, terminating on the
outer surface of the branchial globes in a triangular, compressed tooth which is directed
outward; at the level of the subacute tips of these teeth the carapace shows its greatest
breadth. The cristiform, antero-lateral margins are thus curved S-like; the described
largest tooth is followed anteriorly by three or four others, that gradually become
smaller, which, like the former, are granulated or denticulate on their margins, The
anterior part of the antero-lateral border, defining the hepatic region laterally, is entire,
not granulate. From the largest tooth the likewise cristiform and finely denticulate
postero-lateral margin runs at first backward and upward, then it turns suddenly
downward and inward at a right angle until near the base of the branchial regions;
at this angle the postero-lateral margin carries another, rather obtuse tooth, which is
smaller than that at the posterior end of the antero-lateral border and which is directed
backward and outward, At the base of the branchial regions, finally, the postero-lateral
margins curve for a short distance forward, not uniting therefore with the posterior
border of the carapace ; just at this curve they carry a rounded, obtuse tooth or promi-
nence. From the angle where the postero-lateral margin turns suddenly downward a
finely granulated ridge runs upon the upper surface of the branchial region forward
and inward; just outside of this ridge the upper surface is a little concave, but more
outward and forward it is regularly convex and also on the inner side of that ridge.
For the rest the branchial regions are smooth, very finely punctate, the puncta being
not crowded, except just near the granulated ridge above.

The orbital margins are smooth. The posterior wall of the orbits is marked with a
narrow, linear fissure, the lower wall has a large triangular notch, and the obtuse, internal
angle is little prominent. From the inner infraorbital tooth a ridge extends backward
that makes a right angle with the acute tooth at the antero-lateral angle of the buccal
frame. From the last-mentioned tooth a prominent granulated ridge runs obliquely
backward on the pterygostomian regions, parallel with the antero-lateral border of the
carapace; between the latter and the granulated ridge the subhepatic region is, just
below the orbits, deeply concave. At the level of the middle of the buccal frame the
pterygo-stomian ridge has a triangular notch.

The external maxillipeds are granular; on the inner half of the merus-joints the
granules are larger. ‘The sternum is granulated. The abdominal somites carry each a
compressed, transverse tooth in the middle and another smaller one at the lateral angles ;
the teeth are granulated and there are granules between them; the second somite is
visible when the carapace is looked at from above.

The chelipeds are subequal, the right a little longer than the left. The right cheliped,
16 mm. long, is little more than twice as long as the carapace, it is thus of moderate
length. The arm, 7 mm. long, is quadrilateral and projects only one-third of its length
beyond the carapace; its surfaces are smooth. The anterior surface makes a right
angle with the lower; the edge between both is beset with small, subacute teeth, Both
the anterior and the posterior borders carry small, compressed, triangular teeth, which

FROM THE INLAND SEA OF JAPAN. 391

are unequal ; the upper border is also somewhat denticulate. The posterior margin of the
carpus, Which is smooth above and below, is sharp; it carries one tooth just beyond the
middle and one at the distal end. The anterior border of the upper surface is granulate.
The three sides of the trigonal palm are also smooth; the anterior edge is crenulate, the
upper denticulate; the teeth are small, little prominent, but one, just beyond the middle,
is somewhat larger than the rest. The sharp, cristiform, posterior margin carries four
triangular teeth, one at each extremity and two in the middle; teeth and margin are,
moreover, finely denticulate : that of the left cheliped carries six or seven teeth. The
sharply-pointed fingers are much turned inward, the dactylus being at a right angle with
the upper surface of the palm. The upper border of the dactylus is granulated, the first
granule or tubercle near the articulation is much larger than the following, which become
eradually smaller; the cutting-edge of the dactylus of the larger cheliped carries five
low, obtuse teeth, of which the fifth, near the tip, is a little larger than the preceding.
The immobile finger carries two much larger, obtuse teeth in the middle, the second of
which is larger than the preceding one. The fingers of the left leg are less denticulate.

The ambulatory legs are also of moderate length, those of the first pair extend with
only half their dactylopodite beyond the distal end of the arm of the chelipeds; their
joints are laterally compressed. The upper margin of the merus is sharp, lamellar, and,
in the legs of the last pair, faintly denticulate ; the lower edge of the outer surface
is, in the last pair of legs, beset with prominent, rather acute and unequal granules; on
the meri of the two preceding pairs they are smaller, and on the legs of the first pair it
is not the lower edge of the outer, but that of the inner surface which is granular. The
upper margin of the two following joints is also lamellar and sharp, and the lower margin
of the propodites is finely granulated. The terminal joints, slightly longer than the propo-
dites, are tomentose, except at their tips. The coxe of the fifth pair carry two acute
teeth posteriorly, the outer larger than the inner.

The upper surface of the carapace is cream-coloured, the sides of the median regions
are marked with wine-red spots; the chelipeds are red, and the fingers dark brown on
their distal half, the tips being paler.

Lambrus (Purthenopoides) pleromerus, Ortm., from Japan, of which the type was
examined by me, is a quite different, much larger species.

LUPA, Leach.

Lupa (HELLENUS) HASTATOIDES (Fabr.), de Haan.
Portunus (Amphitrite) hastatodes, de Haan, Fauna Japon., Crust. 1835, p. 39, Taf. 1. fig. 3.
Neptunus (Amphitrite) hastatoides, de Man, in Spengel, Zool. Jahrb., Syst. viii. 1894-95, p. 557.
Neptunus (Hellenus) hastatoides, Alcock, in Journ. Asiat. Soc. Bengal, vol. xviii. pt. 11. 1899, p. 38.

One male from the Inland Sea of Japan, deep sea.

The two median teeth of the front are distinctly less prominent than the others, just as
in de Haun’s type specimen, mentioned by me (/. c.) ; in Indian specimens the two median
teeth are usually as prominent as or even more prominent than the others, as was stated
by me, and later also by Alcock, by whom a large number of individuals were examined.

392 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

The cephalothorax is 13°5 mm. long, measured in the middle line, the abdomen
excluded ; the external orbital angles are 10°75 mm. distant, and the tips of the large
lateral spines 30°5 mm. The lateral angles of the posterior margin are spiniform.
Penultimate joint of the abdomen 3 mm. long, its posterior margin straight, 2 mm.
broad.

In both chelipeds the anterior border of the arm carries 4 spines; the right cheliped
is a little larger than the left, the arms project nearly their whole length beyond the
carapace.

The tip of the dactylus of the last pair of legs shows no trace at all of a dark fleck.

PLATYGRAPSUS, Stimpson.

PLATYGRAPSUS DEPRESSUS (de Haan).
Grapsus (Platynotus) depressus, de Haan, Fauna Japonica, Crust. 1835, p. 63, tab. 8. fig. 2.
Platygrapsus depressus, Ortmann, in Spengel, Zool. Jahrb., Syst. vii. 1894, p. 716.

One male of medium size from the Inland Sea of Japan, caught in deep water.

This specimen, which has been compared with an adult typical male from the Leyden
Museum, is 1425 mm. broad and 12 mm. long; breadth of the anterior border of the
front 66mm. The right cheliped is much larger than the left, in both the inner angle
of the carpus is subacute; the fingers of the right cheliped, which is just as long as the
carapace, viz. 12 mm., are gaping and meet only at the tips; the arcuate and tapering
dactylus carries a denticulate prominence in the middle, and between it and the tip six
or seven small rounded teeth; the inner border of the lower finger carries also seven or
eight small, somewhat unequal teeth. The smooth outer surface of the chela is finely
punctate. The fingers of the other chela, which is 9°5 mm. long, are just as long as
the palm, straight and shut almost close together; the cutting-edge of the immobile
finger shows a dozen somewhat unequal conical teeth; as many teeth occur on the
dactylus, but here they are very small, those near the tip being a little larger than
the rest.

The legs are of a beautiful scarlet colour; the upper surface of the carapace is greenish,
but the front and the antero-lateral margins are also red.

ASTHENOGNATHUS, Stimpson.

ASTHENOGNATHUS INZQUIPES, Stimpson. (PI. 31, figs. 4-6.)
Asthenognathus inequipes, Stimpson, in Proc. Acad. Nat. Sciences Philadelphia, 1858, p. 107.

One egg-laden female from the Inland Sea of Japan, caught in deep water.

So far as [am aware, this species has not been found again since its first discovery almost
half acentury ago. It isa little smaller than Stimpson’s type, also a female, the carapace
of which was 6°8 mm. long and 9°5 mm. broad. The carapace of our specimen from the
Inland Sea is 48 mm. long, measured in the middle line; the well-defined and granular
antero 1ateral borders are slightly arched, diverging backward, and they meet with the

FROM THE INLAND SEA OF JAPAN. 393

somewhat shorter postero-lateral nearly at the level of the median part of the cervical
groove, ¢. e. a little behind the middle of the carapace. The upper surface shows here
its greatest breadth of 6°5 mm.; the proportion between this breadth and the length
fully agrees with that of the dimensions indicated by Stimpson. The likewise granulated
postero-lateral borders are also slightly arched, converging backward, but, different
from Tritodynamia (confer Nobili, in ‘ Annales Mus. Nat. Hungarici,’ iii. 1905, tab. 10.
figs. 1 & 2), their concave side is turned inward, in Tritodynamia, however, outward.
The postero-lateral borders almost reach to the posterior margin of the carapace. From
the point where the antero- and postero-lateral borders meet, a granulated line proceeds
backward on the side wall of the carapace, terminating above the antepenultimate legs;
it is here that the carapace has its greatest width of 7-2 mm.

The upper surface, which is one-third broader than long, is slightly convex longi-
tudinally ; the median transverse groove, which is situated a little behind the middle
and occupies about one-third of the breadth of the carapace, is broad and shallow ;
but the gastric region, which regularly curves into the strongly deflexed front, is barely
demarcated from the hepatic regions. Whereas the larger anterior half of the cardiac
region is slightly convex longitudinally, a shorter posterior part is somewhat depressed.
Just in front of the posterior margin of the carapace, parallel with it, a straight ridge
runs between the bases of the last pair of legs; at its lateral extremities this ridge
curves forward and, running above the last pair of legs, appears here granular. The
carapace is also slightly arched from side to side. Its upper surface is finely punctate,
for the rest smooth; examined under the microscope it appears very finely granulate
(“subtilissime granulata,” Stimpson).

The distance, 3°7 mm., between the external orbital angles, which are not at all promi-
nent, measures almost three-fifths of the greatest breadth of the upper surface and three-
fourths the length. The upper orbital margins regularly curve into the lateral margins
of the front, which converge forward, so that the much deflexed front appears somewhat
broader at its base than at its anterior border; the anterior border is 1:28 mm. broad,
about as broad as the orbits and one-fifth of the breadth of the upper surface; at its
base, however, the breadth of the front is almost one-third the greatest width of the
carapace. When the latter is looked at from above the anterior margin of the front
appears very slightly arcuate, but when the front itself is looked at from above the
anterior margin appears broadly triangular, because it projects a little forward in the
middle; the lateral margins of the front make distinct, somewhat obtuse angles with
the anterior border. The frontal and supraorbital margins are smooth; frontal median
furrow short and quite shallow.

Interantennular septum very narrow, if complete; antennular fossee barely broader
than long, well developed, like the antennulze, which fold transversely (Pl. 31. fig. 4).
The basal joint of the outer antennze, situated between the basal antennulary joint and
the small obtuse tooth at the inner lower angle of the orbits, is about as long as broad ;
the second joint, which is just as long, but only half as broad, reaches to the level of the
front, and the much smaller third joint extends beyond it; the flagellum is 1:65 mm.
long, longer than the breadth of the anterior border of the front, and reaching beyond

394 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

the orbits. The orbits are well developed but incomplete below; the movable eye-
peduncles are of a stout shape, being a little more than half as thick as long; the red-
brown cornes are distinctly facetted. The eye-peduncles are a little pubescent, and
these minute hairs are, like those of the carapace, a little setose or ramified. Infraorbital
ridge smooth, prominent, running a little below the orbits.

Epistome extremely short. The buccal cavern broadens backward, as the slightly
arcuate lateral margins distinctly diverge; posteriorly it is 25 mm. broad, little more
than one-third the greatest width of the carapace, and the buccal cavern appears once and
a half as broad as it is long. The greater median part of the anterior border of the
buccal frame is straight, whereas the smaller lateral parts are slightly convex. Palate
quite smooth, without any trace of a median or lateral ridges.

The external maxillipeds are widely distant; the greatest width of the gap, at the
boundary between merus and ischium, is more than once and a half as large as the
breadth of each footjaw at that place; these maxillipeds are rather feeble (hence the
name of Asthenognathus), for they do not quite reach to the anterior border of the buccal
frame, leaving a small gap between them and this border. The ischium is 0°'72 mm.
long, quadrangular, its slightly concave anterior border is 0°35 mm. broad; it becomes
somewhat broader backwards, so that it is 0‘6 mm. broad in the middle; the ischium
appears thus a little longer than broad in the middle. The slightly arcuate inner border
is granulate, and one observes on its outer surface a shallow groove that runs nearer to
the inner than to the outer border, with which it is parallel. The merus-joint is also
quadrangular, but smaller than the ischium, for it is only 0°62 mm. long between the
antero-external angle and the posterior border ; this joint, 0°54 mm. broad in the middle,
is but liltle longer than broad; the outer margin runs at first parallel with the somewhat
arched inner (PI. 81. fig. 4), but then it runs inward towards the ischium, so that it shows
an obtuse angle in the middle. The anterior border of the merus is barely broader than
the posterior, viz. 0-4 mm., makes right angles both with the inner and outer margins, and
the antero-internal angle is rounded; a longitudinal groove runs, on the outer surface,
near and parallel with the inner border along the whole length of the merus-joint. The
palp is of moderate size and articulates near the antero-external angle of the merus; it
consists of three joints that articulate at their distal ends. The carpus, 0°55 mm. long,
measured along its outer border, is a little shorter than the merus and nearly twice as long
as the following joint; the terminal joint is 0°38 mm. long, almost three times as long as
thick at base, conical or rather sugarloaf-shaped ; it is furnished with long setze, and the
inner borders of ischium and merus are also setose. The exognath, which is not concealed,
reaches almost to the distal third or fourth of the outer border of the merus; near the
middle of the ischium it is 0°23 mm. broad, about one-third of the breadth of this joint,
but it distinctly narrows anteriorly. The gap between the outer footjaws is a little
broader between the antero-internal angles of the merus-joints than at the base of the
ischium-joints. The anterior border of the sternum is coarsely granulate.

The abdomen is 7-jointed and 5°6 mm. broad, a little less than the upper surface
of the carapace; the penultimate joint of the abdomen, which is finely punctate and
pubescent, is 0°S8 mm. long, measured in the middle line, whereas the antepenultimate

Ee

FROM THE INLAND SEA OF JAPAN. 395

joint is 084mm. long. The terminal joint is triangular, 0°8 mm. long and 14 mm.
broad, almost twice as broad as long and barely shorter than the preceding ; its posterior
margin is arcuate, convex, the tip rounded.

The chelipeds (Pl. 31. fig. 5) are equal, rather feeble and small; they are 6 mm. long,
almost as long as the upper surface of the carapace is broad. The arm is triangular,
2 mm. long, unarmed; its upper border is strongly curved and carries about in the middle
a tuft of long sete that are half as long as the merus. Stimpson describes the merus as
“‘superne prominentia mediana setigera instructus,” but I see no prominence at all.
Carpus rounded internally. The chela, which is somewhat compressed and the fingers of
which are slightly curved inward, is 3 mm. long, once and a half as long as the merus.
The fingers, which are distinctly longer than the upper border of the palm, barely exceed
the length of the lower border; the palm is 1-1 mm. high, so that the chele are nearly
three times as longas broad. The fingers regularly taper to the pointed acute tips; they
are of equal size, equally broad at their base, and they leave a small interspace between
them that gradually narrows towards the tips; the cutting-edges are rather sharp, that
of the immobile finger carries 6 or 7 very low obtuse teeth, nearly of equal size and
extending along the two proximal thirds; the dactylus carries near the base two truncate,
somewhat larger teeth, the first of which is little larger than the other, and beyond them
the cutting-edge runs somewhat uneven, the distal third excepted. The upper border of
the palm is a little hairy and seems to be slightly granular, but it cannot be described as
sharp, as was done by Stimpson; his words ‘ superne acuta” are apparently applicable
to the dactylus. The outer surface of the palm and of the fingers is smooth, but a ridge
proceeds along the lower border from the carpal articulation to the tip of the index, and
the palmar portion of this ridge is granulated.

As regards their shape and their relative length, the ambulatory legs much agree with
those of Tritodynamia japonica, Ortm. Those of the antepenultimate or third pair (fig. 6)
are the longest of all, measuring 11-5 mm., i. e. once and a half the greatest width of the
carapace ; the legs of the fourth pair are 11 mm. long, barely shorter than the preceding ;
then follow those of the second pair, that are much shorter, measuring 85 mm. ;
whereas the legs of the fifth pair, 6 mm. long, are the shortest and smallest of all,
reaching but little beyond the merus of the penultimate pair. The meri of the third
legs are moderately enlarged, as they are almost three times as long as broad; the two
following joints are nearly equally long, and the dactyli are barely shorter than the
propodites. The straight dactyli are depressed and taper, about from the middle, to the
pointed extremity; their outer surface is longitudinally grooved in the middle, the
lateral margins are ridged, and one observes on either side of the ridges a fringe of stiff
outstanding sete. The lower margin of the outer surface of the merus is coarsely
granulated, the arcuate upper border more finely and the borders of the two following
joints are also partly granular. The legs of the penultimate pair much resemble those
of the third, but the carpo- and propodites are a little broader in proportion to their
length ; the dactyli are as long as the propodi, but those of the second pair are a little
longer than the propodi, measured in the middle. The dactyli of the small legs of the
fifth pair, which are also a little longer than the propodites, are slightly recurved, and they

SECOND SERIES.—ZOOLOGY, VOL. IX. 57

396 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

are furnished on their lower surface with a row of sete, of which the first are little
shorter than these joints, whereas the following regularly decrease in length towards the
tips ; the outer surface of the meri of these legs is distinctly granulated near the upper
and lower borders. The upper surface of the ambulatory legs is covered, the dactyli
excepted, with the same dark brown tomentum that one observes also on the side-walls of
the carapace and near the lateral margins of the upper surface; the lateral margins of
these legs are furnished with somewhat longer sete.

Eggs numerous, globular, small.

The upper surface of the carapace has a very pale ochraceous colour.

On the legs of the fifth pair several pedunculated Infusoriz were attached.

The genus 7’ritodynamia, Ortm., is apparently most closely related ; its chief difference
is probably presented by the external maxillipeds, the merus-joint of which is longer
than broad and not shorter than the ischium, and furthermore by the insertion of the
terminal joint on the inner border of the penultimate. The latter character was observed
by Nobili in a new species referred by him to Zritodynamia; but it is still unknown
whether this character occurs also in the typical species, Zit. japonica, because the
outer footjaws of Ortmann’s single specimen were much damaged. Probably, therefore,
the genus Tritodynamia ought to be referred to the subfamily Asthenognathine (confer
Alcock, in Journ. Asiat. Soc. Bengal, vol. lxix. part 11. 1900, p. 294).

Geographical Distribution —Hast coast of Nippon, 38° N. lat., on a sandy bottom
(Stimpson).

TRIGONOPLAX, H. M.-Edw.

TRIGONOPLAX UNGUIFORMIS (de Haan).
Ocypode (Elamene) unguiformis, de Haan, Fauna Japonica, Crust. 1839, p. 75, tab. 29. fig. 1, g 2, and
tab. H.
Elamena (Trigonoplax) unguiformis, Alcock, in Journ. Asiat. Soc, Bengal, vol. lxix. pt. mu. 1900,
p. 387.

One male, collected in deep water, Inland Sea of Japan.

H. Milne-Edwards, in Annales Sciences Nat. 3° série, Zool. t. xx. 1853, p. 224,
describes this species as having the carapace “arrondie en arriére et trés-déprimée.”
These words are not quite exact, but liable to be misunderstood. The carapace,
measured from the rather obtuse tip of the triangular front to the middle of the
concave posterior margin of the carapace appears to be 7-4 mm. long; the greatest
breadth above the insertion of the third pair of legs measures 9'4 mm. ‘The undivided,
smooth, and glabrous upper surface cannot be said to be “ trés-déprimée”; in a lateral
view of the carapace the middle part of the upper surface corresponding to the cardiac
area appears, indeed, to slope slightly downwards towards the front and more rapidly
towards the posterior and the slightly carinate antero-lateral margins. The postero-
lateral margins and the concave posterior border are distinctly lamellar.

The legs of the second and third pairs are five times as long as the length of the
carapace without the front; the upper border of the meri terminates in a small tooth.

ee eee eee

FROM THE INLAND SEA OF JAPAN, 397

The legs are yellowish, the carapace orange-coloured.
Geographical Distribution —Japan (de Haan); Bay of Tokyo, Kadsiyama, Kagoshima,
Japan (Ortmann); Gulf of Martaban (Henderson); Andamans (Alcock).

LEUCOSIA, Fabr.

LEUCOSIA RHOMBOIDALIS, de Haan. (PI. 31. fig. 7.)
Leucosia rhomboidalis, de Haan, Fauna Japonica, Crust. 1841, p. 134, pl. 33. fig. 5; Alcock, in Journ.
Asiat. Soc. Bengal, vol. lxv. pt. 1. 1896, p. 234.

One adult male from the Inland Sea of Japan, caught in deep water.

The carapace is 16 mm. long and 13°75 mm. broad; its upper surface is lead- or slate-
coloured, without the dark red spots described by Alcock. The abdomen (PI. 81. fig. 7)
does not exactly agree with the figure in the ‘ Fauna Japonica’; the penultimate segment
narrows more distinctly distally and its lateral margins are very slightly arched, not at
all concave ; the antepenultimate joint is distinctly constricted not far from its posterior
margin, as in Leuc. maculata, Stimps., which is regarded by Alcock as identical with
this species. The edge of the pterygostomian region that forms the anterior boundary
of the thoracic sinus is quite straight.

MYRA, Leach.

Myra FuGax (Fabr.).
Myra fugax (Fabr.), Alcock, in Journ. Asiat. Soc. Bengal, vol. lxv. pt. 1. 1896, p. 202.

One young male from the Inland Sea of Japan, deep water.

The carapace of this specimen, which agrees with the form described by Miers in 1879
under the name of Myra dubia, and apparently also with that described by Hilgendorf
as Wyra coalita, is 13°5 mm. long exclusive of the median spine, and 15 mm. when it is
included ; the carapace is 12-4 mm. broad. The upper surface, which is strongly convex
transversely, agrees in its general shape with the figure of Myra carinata in Bell's
Monograph, but the acute median spine is much shorter, being only once and a half as
long as the lateral ones. The median granulated ridge is quite distinct, as also the raised
cluster of granules on the well-defined intestinal region; punctiform granules are
scattered on the upper surface, except quite anteriorly. The front and the adjacent
parts of the flattened subhepatic regions are pubescent. Immediately behind the notch
three or four beads of the lateral border are dentiform, and one observes another just
above the last pair of legs. The chelipeds are 25 or 26 mm. long, not quite twice as long
as the cephalothorax.

There is still a very young male specimen, without definite locality, that no doubt
belongs also to this species ; it is, in my opinion, that form which has been described by
Dr. Alcock as a distinct species, Myra pentacantha (Alcock, l. ¢. p. 204). Measured in
the middle line, the carapace appears to be 6-4 mm. long the median spine included, and
55 mm. without it; it is 5-1 mm. broad. The carapace is less strongly convex; not
only is the intestinal region distinctly defined, but the branchio-cardiac grooves are also

57*

398 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

discernible. The median acute spine on the intestinal region is twice as long as the
lateral ones. The carapace is marbled with red on either side of the median ridge, on
each side of the front, and on the hepatic regions.

ARCANTA, Leach.

ARCANIA HEPTACANTHA (de Haan). (Pl. 31. figs. 8-10.)
Iphis heptacantha, de Haan, in Herklots, ‘Symbole carcinologice : Htudes sur la Classe des Crustacés,’
Leyde, 1861, p. 27.

Two males and two sterile females of somewhat larger size, from the Inland Sea of
Japan, deep water.

Through the kindness of the Direction of the Leyden Museum I was enabled to
compare these specimens with the single type of [phis heptacantha, de Haan, a descrip-
tion of which seems not to have appeared. This type specimen, the locality of which is
unknown, is an adult female of larger size than the Japanese specimens ; legs and foot-
jaws are unfortunately wanting. The Japanese specimens no doubt belong to this
species.

The cephalothorax of de Haan’s type specimen is a little broader in proportion to its
length; but this may be explained by its larger size (compare the measurements).
Alcock’s description of Are. septemspinosa (Fabr.), Leach (in Journ. Asiat. Soc. Bengal,
vol. Ixv. pt. 11. no. 2, 1896, p. 265), is applicable to de Haan’s type of Arc. heptacantha
except as regards the length of the spines and perhaps also the following. The cardiac
and intestinal regions are also separated on each side by a moderately deep groove
from the branchial regions, which, according to Alcock’s description, does not seem to
be the case in Arc. septemspinosa. In Arc. heptacantha the surface of the carapace is
finely granular; on the gastric, cardiac, and intestinal regions the granules are a little
larger than on the branchial; from each of the two spines with which the latter are
armed a somewhat irregular row of granules runs forward and inward on their surface ;
these granules have the same size as those of the gastric region, but between these rows
the granulation is finer than on the median regions. The concavity just behind the
crease or pucker that separates the hepatic from the branchial regions and the upper
surface of the front are smooth. The lateral spines, that are somewhat directed backward
and slightly curved upward, measure in de Haan’s type specimen 6 mm., 7. e. almost
one-third the breadth of the carapace without the spines; the median spine on the
transversely and longitudinally convex intestinal region, which is a little directed upward,
is the shortest of the seven spines, measuring 1°75 mm., not quite one-third the length of
the lateral spines. The four other spines, which have nearly the same size, are 2°4 mm.
long, so that they appear a little larger than the spine on the intestinal region, measuring
a little more than one-third the lateral spines. Exclusive of the spines, the cephalo-
thorax, which is strongly convex transversely and as much longitudinally, appears in the
Leyden type a little broader than long, in the largest of the Japanese specimens (a sterile
female) nearly as long as broad. In the latter specimen the lateral spines are 3°25 mm.
long, about one-fifth the width of the carapace without the spines, so that they are

FROM THE INLAND SEA OF JAPAN. 399

comparatively shorter than in de Haan’s adult female; the median spine on the intestinal
region is just as long as the two spines on the posterior border, viz. 13 mm., measuring
a little more than one-third the lateral spines, but the posterior two on the branchial
region are, in this specimen, the shortest of all, measuring 0°9 mm. The posterior
branchial spines are a little farther, viz. 7°3 mm., distant from the tips of the lateral
_ spines than from that (5°75 mm.) of the median spine. The front and the depression
between it and the gastric region are tomentose; the spines are also granular. In the
Leyden type the front is a little less prominent and its lateral margins run somewhat
more obliquely than in this female; but in the other specimens the obliquity is nearly
the same. The chelipeds are equal, 37 mmm. long, more than twice the length of the
carapace (posterior spine included); they agree with the quoted description of Arc.
septemspinosa and with the fig. 4, pl. 25, in Cuvier’s ‘Atlas du Régne Animal.’ The
slender fingers are one-fourth longer than the tapering hand, but carpus and palm
appear also finely granulated under a lens. The meropodites of the ambulatory legs are
finely granular, but the following joints seem to be smooth,

In the other female the five posterior spines are of equal length, but the two males
agree with the larger female. In these specimens the whole upper surface of the
carapace is slightly pubescent.

The abdomen of the male (PI. 81. fig. 9) consists of five pieces; the penultimate seg-
ment is once and a half as long as broad and onceand a half as long as the terminal piece.

Measurements in millimetres.

ile 2. 3 4, 5

= 2. 2 fo} 3.
Breadth of the carapace, the lateral spines included ...... 33 21 184 165 13°75
Length of the carapace, the posterior spine included...... com LOvone Lo 13 11:25
Breadth of the carapace, exclusive of the lateral spines... 21°5 = 15°5 134 115 10
Length of the carapace, exclusive of the posterior spine. 20°5 175 145 12 10:

No. 1. Leyden type of Iphis heptacantha, de Haan; Nos. 2-5. Inland Sea of Japan.

Whether this species differs from Arc. septemspinosa (Fabr.), Leach, by other characters
than the shorter spines, is difficult to say, because I was unable to compare it with
specimens of the latter. I will, however, observe that at the end of his quoted descrip-
tion of Arc. septemspinosa, Dr. Alcock adds :—‘ Of ninety-two specimens in the Indian
Museum the lateral spines are found to vary a good deal in length: they are usually, in
adults, about as long as the arm, and sometimes a good deal longer; but in the young
they are usually much shorter than the arm.”

Perhaps Arc. heptacantha is related to Arc. septemspinosa (Fabr.), var. gracilis, Hend.,
from the Gulf of Martaban, but it is difficult to decide, because his description is too
short (Henderson, in Trans. Linn. Soc., ser. 2, Zool. vol. v. 18938, p. 403).

Arc. quinquespinosa, W.-Mason (Ill. Zool. ‘Investigator,’ Crust. pl. 24. fig. 6), is
certainly a different species.

400 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

ARCANIA GLOBATA, Stimpson. (PI. 31. figs. 11-13.)
Arcania globata, Stimpson, in Proc. Acad. Nat. Sciences Philadelphia, 1858, p. 160.
Arcania globata, Miers, in Proc. Zool. Soc. 1879, p. 44.
Arcania globata, Ortmann, in Zool. Jahrb., Syst. vi. 1892, p. 577.

One young male, collected at a depth of 8 fathcms in the Inland Sea of Japan;
bottom sandy. It is said to be here very rare.

Measured in the middle line, this specimen appears to be 8 mm. ivione, whereas the
carapace is 7°25 mm. broad without the spines, and 9 mm. when they are included.
Without the front the cephalothorax appears semiglobular, for the outline is circular
and it is strongly convex transversely and also much longitudinally. The front, which
is characteristic, is prominent, extending a little beyond the eye-peduncles. Its upper
surface, which is a little convex longitudinally, is slightly furrowed in the median line,
makes a very obtuse angle with the upper surface of the carapace, and appears to be
situated at a much lower level than the latter, in a lateral view or when the carapace is
looked at from in front. The breadth of the front is nearly one-fourth that of the carapace
(the spines included) and it is a little broader than long; the slightly arcuate, lateral
margins terminate each in a small subacute tooth, and the anterior border of the front
between these two teeth is nearly straight, very slightly concave. The upper surface is
rather thickly beset with slender, subacute spines ; these spines, which are smooth and
almost of equal length, appear to be very slightly curved forward in a lateral view of the
carapace. A few similar spines, though much smaller, stand at the base of the front.
The spines on the upper surface of the carapace are about 80 to 90 in number.
Under a strong magnifying-glass the upper surface appears to be covered between the
spines with small pointed spinules, especially anteriorly. A much stouter though barely
longer spine stands on the middle of the well-defined intestinal region; this spine is
granular and also slightly curved forward. Round the margin of the carapace are ten
conical larger and acute spines that are all granulated and more or less curved upward ;
they are as stout as the already described stouter spine on the intestinal region. Of the
five spines on each side, the third or middle one is placed just in the middle of the lateral
margin and somewhat curved forward; the fourth spine has the same size as the third,
and it is as far distant from the third as from the spine on the intestinal region. A fifth
spine, a little shorter than the third and the fourth, is placed on the outer angle of the
narrow posterior border of the carapace ; this spine, which is directed backward and slightly
outward, is also as far distant from the fourth spine as the fourth from the third. The
second spine, which is a little smaller than the third, stands somewhat nearer to the latter
than to the first ; the distance between the first spine and the second is just two-thirds
of that between the third and the fourth. The first to fourth spines and also the spine
on the intestinal region are all placed at some, nearly equal, distance from the lower
border of the carapace, 7. e. from the base of the legs, but the fifth spine stands just near
the base of the last leg. One observes, moreover, two smaller spinules on the posterior
margin between the two spines of the fifth pair.

Eye-peduncles a little shorter than the front; the cornea, which is shining and dark

FROM THE INLAND SEA OF JAPAN. 401

brown, carries anteriorly a small conical tooth or tubercle. Both the outer and the inner
angle of the lower margin of the orbits are produced into an acute slender spine that
reaches not as far forward as the eye-peduncles ; the outer wall of the orbits carries on
its free border a smali spine, which is preceded on its outer surface by a somewhat larger
one. The outer wall is separated on each side by a furrow from the front and from the
pointed spine at the outer orbital angle; the latter spine carries a small acute tooth on
its outer margin (Pl. 31. figs. 10 & 11).

The lateral margins of the buccal frame are considerably thickened anteriorly at the
level of the merus-joint of the outer footjaws and terminate in a forwardly-directed
spine that reaches as far forward as the spine at the internal angle of the orbits. The
outer footjaws are granulated, like the lower surface of the carapace. The merus-joint,
measured along the inner border, appears to be 1 mm. long, the ischium-joint 1-6 mm. ;
the former is thus more than half as long as the latter.

The 5-jointed, strongly granulated abdomen resembles that of Arc. 11-spinosa, de Haan ;
the same rather coarse granulation exists on the sternum. The chelipeds, 13 mm. long,
are little more than once and a half as long as the carapace. The merus-joint, which is
a little stouter than that of Arc. 11-spinosa, de Haan, is covered above with rounded,
circular granules, mostly large, though with some smaller observable among them on
the distal half ; on the anterior border they are of a more conical shape and the posterior
border carries fowr strong, nearly equidistant, and subequal, subacute spines, which
are not described by the quoted authors, unless by Stimpson with the words “ granulis
plerumque subspiniformibus.” Similar circular bead-like granules as on the upper
surface also occur on the lower. Carpus and hand are closely beset with granules, which
are, however, much smaller than those of the arm; the slender fingers, which shut close
together and are almost once and a half as long as the upper border of the palm, are
deeply furrowed longitudinally ; they show a fine granulation under a strong magnifying-
glass, they are a little hairy distally, and their prehensile edges are beset with numerous
small teeth, a few of which are distinctly larger on the distal half of the fingers.

The ambulatory legs, smooth to the naked eye, are indeed covered with a close minute
granulation, visible only by means of a strong magnifying-glass; the anterior border of
the meropodites is spinwlose, being beset with 5-9 small, spiniform, acute teeth, and the
slender, slightly arcuate terminal joints are about as long as the propodites.

This pretty little crab has the front and a median band on the upper surface of the
carapace white, the median band being half as broad as the front; adjacent to the band
the upper surface is orange, but this colour gradually becomes paler laterally. The
spines are also of a pale orange-colour, but those that stand on the band are white.
The lower surface is uncoloured, but the sternum is marked anteriorly, on each side of
the abdominal groove, with a triangular orange-coloured fleck, between that groove and
the base of the chelipeds. The latter are pale reddish above ; the proximal extremity of
the merus is white, like the tips of the fingers. The ambulatory legs are uncoloured,
but carpus and merus are partly reddish.

Arcania 11-spinosa is at once distinguished by the different shape of the front, by the
carpus and chel being apparently smooth, and, no doubt, by other characters as well.

402 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

Geographical Distribution—Chinese Sea, lat. 23°, depth 16-25 fathoms, bottom sandy
or muddy (Stimpson); Corea Channel, lat. 34° 8’ N., long. 126° 24! E., at a depth of
24 fathoms (Miers); Maizuru, Japan (Ortmann).

GALATHEA, Fabr.
GALATHEA ACANTHOMERA, Stimpson. (PI. 31. figs. 14, 15.)

Galathea acanthomera, Stimpson, in Proc. Acad. Nat. Sci. Philadelphia, 1860, p. 252.
Galathea orientalis, Ortmann, in Spengel, Zool. Jahrb., Syst. vi. 1892, p. 252, tab. 11. fig. 10 (mee
Stimpson).

One male from the Inland Sea of Japan.

Through the kindness of Prof. Déderlein, of Strassburg, some specimens of Ortmann’s
Galathea orientalis from Kadsiyama, Japan, are lying before me, and though they show
a few slight differences, especially as regards the rostral teeth, they belong no doubt to
the same species as our specimen from the Inland Sea. As will appear from the
following description, this species ought to be referred to Gal. acanthomera, Stimpson, and
not to Gal. orientalis of the same author.

The carapace of our male is 7-2 mm. long and 5 mm. broad. The rostrum measured
from the tip to a transverse line uniting the bases of the first, ¢. e. the posterior teeth,
appears to be 2°7 mm. long and 1:5 mm. broad at its base; in a male from Kadsiyama
of the same size the rostrum is 2°55 mm. long, but just as broad as the other. The
length of the rostrum and its relative breadth are thus somewhat variable. The lateral
teeth of the rostrum are all acuminate and pointed. ‘The first or basal tooth is, in the
male from Kadsiyama, directed straight forward and measures just one-third of the second,
which is slightly turned inward, its outer margin being a little curved; the third tooth,
one-third longer than the second and therefore the longest of all, and also the feurth, which
is just as long as the second, are directed straight forward; the terminal spine, finally,
measures two-fifths the whole length of the rostrum, is once and a half as long as the
fourth lateral tooth, and its lateral margins carry a few, six or seven, microscopical teeth
and some setze. In another specimen the third and the fourth lateral teeth are nearly of
equal length and the fourth is slightly turned outward. In an adult female from
Kadsiyama the second tooth is also directed straight forward and its outer margin
straight, not curved inward.

In the male from the Inland Sea the first tooth measures a little more than one-third
the length of the second and is turned slightly outward ; the second tooth projects straight
forward and its outer margin is straight ; the third is once and a half as long as the second,
which is almost as long as the fourth, the third and the fourth being both directed straight
forward. ‘Length and shape of the rostral teeth are thus somewhat variable. Imme-
diately posterior to a transverse line uniting the base of the incisions between the first
and second lateral teeth, one observes, in the middle, two spines near together; these
spines are, in the male from the Inland Sea, a little shorter than the basal teeth of the
rostrum, they are twice as far distant from one another as they are long and a little
farther distant from the basal teeth than from each other.

FROM THE INLAND SEA OF JAPAN. 4.03

The upper surface of the rostrum is somewhat hairy in the middle, short sete being
arranged in curved, parallel rows on each side and near the middle line; a longer seta is
inserted at the base of the fourth lateral tooth, in the middle, another nearly in the middle
of the rostrum on either side of the median line.

The lateral borders of the carapace are armed with nine teeth, or rather spines. The
first spine is, in the male from the Inland Sea, a little larger than the first lateral tooth of
the rostrum and directed obliquely outward; it stands at the outer angle of the orbits.
The second spine, a little less turned outward, is placed somewhat nearer to the first than to
the cervical groove; one observes, between the second and this groove, the two following
spines, viz. the third, somewhat smaller than the second, placed on the upper surface quite
near the cervical groove and a little remote from the lateral margin, and the fourth, which
is as large as the second, just below the lateral margin. Behind the cervical groove
the lateral margin carries five other spines, which are equidistant and of equal size, as
long as the second, except the last which is somewhat shorter. From each of the two
spines, on the boundary between rostrum and gastric region, a ciliated ridge runs
laterally towards the base of the second spine of the lateral margin of the carapace ;
posterior to the two spines one observes seven ciliated ridges, all reaching the lateral
margins, except the second, which terminates at the cervical groove. Between the third
and the fourth runs a short transverse ridge immediately in front of the cervical groove ;
between the fourth and the fifth a ridge proceeds, parallel with them, from the lateral
border until at some distance from the middle line ; between the fifth and sixth two similar
shorter ridges run from the lateral border inward, of which the posterior, which terminates
at the ninth spine of the lateral margin, is almost twice as long as the other. » Between the
sixth ridge and the seventh a similar stria proceeds from the lateral border; this stria is a
little shorter than the posterior of the two between the fifth ridge and the sixth. A
ciliated stria runs from the fifth lateral spine inward along the cervical groove, another
shorter one from the sixth lateral spine. All these ridges are ciliated; the cilia are long,
viz.0°3-0°35 mm. The upper surface of the carapace, of the rostrum, and of the abdomen
is thickly and coarsely punctate; the anterior borders of the segments of the abdomen
are ciliate and carry, moreover, a few rather long setze, which occur also in very small
number on the lateral parts of the upper surface of the carapace.

The antepenultimate joint of the antennal peduncle is bispinose, carrying a strong
spine above and a similar one on the lower border; the penultimate joint is armed above
with a single, somewhat smaller spine.

The external maxillipeds (Pl. 81. fig. 14), partly described already by Dr. Ortmann, show
the following characters :—Measured along its outer margin, the ischium appears a little
longer than the merus; in the male from the Inland Sea the ischium is 1°5 mm. long,
the merus, however, 1:2 mm. The outer margin of the ischium terminates distally in a
sharp tooth, which is slightly turned inward; the inner margin ends in a conical, stouter
though shorter tooth. The two acuminate teeth on the inner margin of the merus are
larger than those of the outer border; the anterior spine on the outer border is somewhat
curved inward and stands at the distal end, the other nearly in the middle of the border.
The outer margin of the carpus is armed, in the male from the Inland Sea, with three sharp

SECOND SERIES,—ZOOLOGY, VOL, IX. 58

4.04. DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

spines, preceded by a very small, acute tooth; these spines are a little smaller than those
of the outer border of the merus, and decrease a little in length from the posterior to the
anterior. In the male from Kadsiyama the outer border of the carpus carries two spines,
which conform to Ortmann’s description, and they are also preceded by a very small
acute tooth. The slender peduncle of the exopodite reaches a little beyond the merus.

In the male from the Inland Sea the chelipeds are a little unequal, one being 20 mm.
long, the other 18 mm.: they agree with Ortmann’s fig. 10. The dactylus of the larger
chela carries a moderately strong, subacute tooth at one-third of its length from the
articulation, and between this tooth and the tip are seen 25 small obtuse or subacute
teeth; the immobile finger has only small teeth, no stronger ones, as also the fingers of
the smaller leg.

The three following legs are also characteristic. The meropodites of the second pair
(Pl. 81. fig. 15) are five times as long as broad, and their upper margin is armed, along its
whole length, with 11-12 strong sharp teeth nearly of the same size; the lower margin is
also a little denticulate. and terminates, at the distal end, in a sharp spine which slightly
projects beyond the rounded extremity of this joint. The outer margin of the carpus is
armed with 5-6 sharp spines, nearly of the same size as those of the merus; the spine at
the far end is a little larger than the preceding. The propodites, one-fourth shorter than
the meropodites and about six times as long as broad, carry, on the proximal half of their
upper border, three or four spines, which are a little smaller than those of the merus, and
their lower margin is beset with six movable spines, which have nearly the same size as
those of the upper margin. The terminal joints, little more than half as long as the pro-
podites, end in a curved claw, while their lower border carries six movable spines, which
gradually increase in length from the first to the sixth. The upper border of the mero-
podites is furnished with sete, which are partly plumose or ciliate; at the base of
each spine, on the posterior surface, is a long hair and one or two shorter hairs near
it. ‘The posterior surface of the meropodites shows transverse rows of short sete and
near the lower margin longer hairs. The following joints are also setose. The legs of
the third and fourth pairs agree with those described, but the spines on the upper border
of the meri are, in the legs of the fourth pair, comparatively smaller. -

The male from the Inland Sea is of a pale yellowish red, the rostrum is whitish, like
the lateral teeth of the carapace; the ciliated ridges on the carapace and the segments of
the abdomen are marked with small red spots. The mero- and propodites of the second
to fourth legs are adorned each with two wine-red rings.

Gal. orientalis, Stimps., from Hong Kong is, no doubt, a different species. The
carapace is described as ‘ brevissime pubescens,” whereas in Gal. acanthomera the cilia
are long. ‘The lateral margins of the carapace carry s¢x teeth, in Gal. acanthomera
nine; the first lateral tooth of the rostrum of Gal. orientalis is minute, the chelipeds
“ crassiusculi,” the chela depressed, the dactylus bidentate, all characters not observed in
Gal. acanthomera. The upper border of the meropodites is described as “ confertim
spinulata, spinulis minutis zequalibus,” that of Gal. acanthomera, however, as “ spinulis
robustis ad 11 armato.”

Geographical Distribution—Bonin Islands (Stimpson); Japan, Kadsiyama, Sagami
Bay, Maizuru, Tanagava, Kagoshima (Ortmann).

es

FROM THE INLAND SEA OF JAPAN. 4.05

CRANGON, Fabr.

CRANGON cCoNsoBRINUS, de Man. (PI. 31. figs. 16-19.)
Crangon consobrinus, de Man, in Ann. & Mag. Nat. Hist. ser. 7, vol. xvii. 1906, p. 401.
Crangon affinis, Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 531.

One adult egg-laden female from the Inland Sea of Japan, caught in deep water.

This species is closely allied to Crangon alaskensis, Lockington, from Mutiny Bay,
Alaska *, but as it is perhaps different, I think it well to publish a somewhat detailed
description.

Measured in the middle line, this specimen appears to be 48 mm. long, from tip of
rostrum to the end of the telson ; the carapace, inclusive of the rostrum, measures
113 mm., i. e. one-fourth of the whole length, without the rostrum it is 10 mm. long.

The rostrum is distinctly shorter than the eye-peduncles when they are directed
straight forward, and reaches only to the cornee; it is rather narrow, spatulate, the
sides nearly parallel for a portion of their length, though the rostrum is very slightly
narrowed behind the middle; the edges are somewhat upturned and the sides curve
anteriorly to the rather acute tip. The carapace is pubescent on each side, but glabrous
posteriorly and in the middle of the dorsal surface, but the short hairs are here perhaps
partially worn off. The single median gastric spine, which is of usual size and slightly
directed upward, is situated at one-fourth the length of the carapace from the tip of the
rostrum, the distance between both tips being 3 mm. On each side is the hepatic
spine, which has the same size as the gastric, and the three spines are situated in a
transverse line. In its general shape the abdomen resembles that of Crangon vulgaris :
it is three times as long as the carapace (rostrum included). The first, the second, and the
third segments are rounded above; the third, however, shows a slight depression on each
side of the median line just behind the middle. The fourth segment presents a trace of
carination along a very short space on the posterior half; the faint and obtuse carina does
not, however, reach either to the middle of the segment or to its posterior margin. The
Jifth segment is distinctly carinate ; the rather obtuse carina arises about at one-sixth the
length of this segment from its anterior extremi'y and terminates quite near the posterior
margin. The sixth segment, which is 7 mm. long, resembles that of Crangon vulgaris, but
its upper border has a shallow median groove; asin Crangon vulgaris, the sixth segment is
suleate beneath, the furrow is rather shallow, and, as in that species, there is a sharp tooth
at the posterior end between the bases of the uropods. The telson is 10 mm, long, almost
once and a half as long as the sixth segment and just as long as the carapace (rostrum
excluded); the slender and gradually tapering telson, which is Saintly grooved above,
terminates in a sharp tooth, on each side of which three movable spinules are inserted ;
the second is the longest of all, twice as long as the others, and extends, like the third,
a little beyond the extremity of the telson, The inner caudal swimmerets are just as
long as the telson, the outer are very little shorter.

The eye-peduncles (Pl. 31. fig. 16) resemble those of Crangon vulgaris. The antennular
peduncles reach just beyond the middle of the distance between the orbital margin of the

* Rathbun, ‘ Decapod Crustaceans of the North-west Coast of North America,’ 1904, p. 114.
58*

4.06 DR. J.G. DE MAN ON CRUSTACEA CHIEFLY

carapace and the tip of the antennal scales ; the process on the outer side of the base
is rather narrow and does not quite reach to the distal end of the first joint, hardly
exceeding the eye-peduncles when they are turned straight forward. The gradually
tapering, inner flagellum, which surpasses somewhat the antennal scales, is a little longer
than the peduncle, measured from the orbital margin of the carapace ; the outer flagellum
reaches to the end of the blade.

The external antennze are just as long as the body. The scale (Pl. 31. fig. 18), measured
along its straight outer margin, appears to be four-fifths the length of the carapace,
exclusive of the rostrum; it resembles that of Crangon alaskensis, but it is only three
times as long as broad; the end of the blade is rounded, not produced at the antero-
internal angle, and much broader than the spine at this level; the spine extends almost
as much beyond the blade as the end of the latter is broad. The antennal peduncle
extends as far forward as the penultimate joint of the external maxillipeds, which just
reach to the end of the blade.

The first pair of feet (fig. 19) are somewhat shorter than the antennal scales, reaching
a little beyond the antennal peduncles. ‘The chelzx, which are 5°6 mm. long and 1°5 mm.
broad at the base of the spinous pollex, are a little slenderer than those of Cr. alaskensis,
for they are almost four times as long as broad; the obliquity of the terminal margin is
in both species the same.

The legs of the fifth pair reach as far forward as those of the first.

There is a slender spine on the sternum between the third pair of legs.

The single typical specimen of Ortmann’s Crangon affinis from Maizuru, Japan, which is
lying before me (Ortmann, /. ¢.), seems to belong to this species ; the rostrum is, however,
a little longer and the process on the outer side of the base of the antennular peduncle
reaches almost to the end of the first joint. Ortmann’s specimen carries a Bopyrid on
the left side of the carapace.

Crangon affinis, de Haan, is certainly different. In this species, indeed, the external
maxillipeds are longer than the antennal scales, and the latter are just as long as the
carapace, the rostrum excluded. Nothing is said about the carination of the fifth
abdominal segment. Unfortunately de Haan’s types do not now exist in the Leyden
Museum.

Crangon propinquus, Stimpson, differs by the third and the fourth segments being
carinate, not the fifth. According to Miss Rathbun *, the rostrum of this species should
exceed the eyes.

CRANGON CASSIOPE, de Man. (PI. 32. figs. 20-25.)
Crangon cassiope, de Man, in Ann. & Mag. Nat. Hist. ser. 7, vol. xvii. 1906, p. 402.

Two egg-laden females from the Inland Sea of Japan, captured in deep water, common
on mud.

In its outer appearance Crangon cassiope much resembles the typical species of this
genus, viz. Crangon vulgaris, but it is at once distinguished by the sixth segment of
the abdomen, which is conver, not sulcate, beneath. This species appears therefore also

* Rathbun, in Proc. U.S. Nat. Mus, xxvi. 1902, p. 42.

FROM THE INLAND SEA OF JAPAN. 407

related to Orangon alba, Holmes, and Crangon holmesi, Rathb., from the North-west
coast of North America. From the former it differs, however, at first sight by the blade
of the antennal scale which agrees with that of Crangon vulgaris; from the latter also
by the antennal scale, which measures only two-thirds the length of the carapace,
exclusive of the rostrum, while the blade appears, moreover, broader at the extremity
than that of Crangon holmesi.

The two specimens are nearly of the same size: they are 46°5 mm. and 44 mm. long
from tip of rostrum to the end of the telson. In the larger specimen the carapace
is 12°5 mm. long, rostrum included, and 11 mm. without it; in the other it is 11°75 mm.
long, rostrum included, and 10°5 mm. without it, so that the carapace, rostrum included,
is a little longer than one-third the abdomen.

Viewed from above this species closely resembles Crangon vulgaris, but the numerous
small, dark spots with which carapace and abdomen of the common shrimp are mottled
are almost wanting in Crangon cassiope. Small violet spots are, however, seen on
the peduncle and inner flagellum of the inner antenne, on the antennal scales, on the
hepatic region of the carapace between the pterygostomian and hepatic spines, near the
posterior margin of the carapace, on the telson and on the uropods.

Even on close inspection the carapace shows no differences from that of Crangou
vulgaris. The narrow, triangular rostrum is as short in proportion to the eye-peduncles
as in that species, the gastric and the two hepatic spines agree also in both. The
abdomen, viewed from above, also closely agrees with that of Crangon vulgaris; all
the seven segments are rounded above, but neither the sixth nor the seventh shows any
tendency to become flattened or grooved, as is sometimes the case in Crangon vulgaris.
In the common shrimp the ventral surface of the sixth segment is marked by a
moderately deep groove, which usually begins near the anterior margin and more or less
gradually widens posteriorly ; on the posterior end is a sharp spine, which is directed back-
ward. In Crangon cassiope, however, the ventral surface appears in the middle of the
segment rounded and convex, but the posterior fourth is slightly concave, and there is also
a short, transverse, though quite shaliow pit or depression at one-third of the segment
from its anterior margin; instead of a sharp spine one sees in Crangon cassiope, at
the posterior end, a subacute conical tubercle. On each side of the middle line the
ventral surface is punctate; one observes numerous large puncta and between them
many others that are quite minute. The two pairs of antennie closely resemble
those of Crangon vulgaris. The antennal scales (PI. 32. fig. 20) measure along their
outer margin two-thirds the length of the carapace without the rostrum, and they are
two and half times as long as broad; the end of the blade (fig. 21) is slightly rounded,
makes a distinct angle with the inner margin, and is four times as broad as the adjacent
part of the spine, which reaches considerably beyond it. The antennal scales closely
resemble those of Crangon vulgaris and the outer antennze are just as long as the body.
As regards the inner antenn, I wish only to observe that the stylocerite is a little shorter
than the first joint of the peduncle, and that these antennz otherwise fully agree with
those of the common European shrimp.

The external maxillipeds, which reach to the end of the antennal scales, do not fully

4.08 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

agree with those of a female of Crangon vulgaris from this country. Thus the joints of
the endopodite are broader in proportion to their length. The terminal joint appears in
an egg-bearing female of Crangon vulgaris of the same size six times, but in the female
of Crangon cassiope five times as long as broad; the penultimate segment of Crangon
vulgaris is a little more than four times, that of Crangon cassiope a little more than
three times as long as broad; the antepenultimate joint, finally, is, in the common
shrimp, about four times, but in Crangon cassiope three times as long as broad.

The first pair of feet (Pl. 32. fig. 23), which reach nearly to the end of the antennal
scales, are stouter than those of vulgaris; the length of the chelze is only two and one-
third times the width measured from the inner base of the immovable spine, in Crangon
vulgaris, however, three times. The obliquity of the anterior margin is in both species
the same.

The second legs are also a little less slender than those of the common shrimp. The
legs of the fourth pair reach with their dactyli beyond the tip of the antennal peduncles,
those of the fifth (fig. 24) are but little shorter ; these legs differ especially from those of
Crangon vulgaris by comparatively shorter dactyli (fig. 25) and somewhat slenderer
propodites. For example, the propodites of the fifth pair in an egg-laden female of
Orangon vulgaris of the same size as the specimens of Crangon cassiope are seven
times, but in cassiope eight times as long as broad; the dactyli are in Crangon cassiope
half as long as the propodites, but in Crangon vulgaris they measure three-fourths the
length of these joints, appearing thus comparatively once and a half as long as in our
new species.

The globular eggs are small, diameter 0°45 mm.

SCLEROCRANGON, G. O. Sars.

SCLEROCRANGON ANGUSTICAUDA (de Haan).

Crangon angusticauda, de Haan, Fauna Japonica, Crust. 1849, p. 183, tab. 45. fig. 15; Stimpson, in
Proc. Acad. Nat. Sci. Philadelphia, 1860, p. 25.

Sclerocrangon angusticauda, Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 533, and in Proc. Acad.
Nat. Sci. Philad. 1895, p. 179.

One egg-laden female from the Inland Sea of Japan.

Length 32 mm. from tip of rostrum to the end of the telson; the carapace, 8°75 mm.

long, the rostrum included, measures little more than one-fourth the whole length.
Viewed from above, the rostrum, which is as long as broad at its base, appears a little
shorter than the eyes ; its slightly upturned lateral margins, which in a lateral view of the
rostrum appear a little arcuate, curving at first upward and then very slightly downward,

converge forward, so that the rostrum appears triangular, with rather obtuse tip.

De Haan, however, describes the rostrum as “ apice acutum.”

The obtuse, flattened, median carinz of the third to fifth abdominal somites are
bounded on each side by a hairy, longitudinal furrow, into which issues the transverse
furrow described by de Haan. The sixth segment carries above two obtuse carinz, which
converge backward and are even united fora short distance posteriorly ; between the two

FROM THE INLAND SEA OF JAPAN. 4.09

carinze it appears faintly furrowed, and the two carine are also bounded externally by a
hairy groove.
The external maxillipeds reach with half their terminal joint beyond the antennal scales.
Geographical Distribution—Japan (de Haan); Simoda and Hakodadi (Stimpson) ;
Kadsiyama (Ortmann).

LEANDER (Desm.), Stimpson.

LEANDER PAUCIDENS (de Haan).

Palemon paucidens, de Haan, Fauna Japon., Crust. 1849, p. 170, tab. 45. fig. 11.
Leander paucidens, Stimpson, in Proc. Acad. Nat. Sci. Philadelphia, 1860, p. 40.
Palemon paucidens, Rathbun, in Proc. U.S. National Museum, xxvi. 1902, p. 51.
Leander paucidens, Doflein, Ostasiatische Dekapoden, 1902, p. 640.

Ten specimens, among which are two adult, egg-laden females, from Hakone Lake,
Japan, caught in July 1896, 2400 feet above sea.

The two egg-bearing females are respectively 54 and 55 mm. long, the other specimens
are all smaller but one. The eggs are not very numerous, large, 18-2 mm. long and
1:4-1'5 mm. broad.

In five specimens the tip of the rostrum is injured, in the rest it is bifid at extremity ;
in one the rostrum is broken, two carry five teeth on the upper margin besides the
apical tooth, the rest only four; usually the second tooth stands immediately before the
frontal border of the carapace, rarely just above it. In two specimens the lower border
is armed with three teeth, in six with two, and in one specimen there is only one
tooth. In some specimens the rostrum is just as long as the scales, in others it
overreaches them a little; in the larger specimens it is slightly upturned at extremity,
in the rest it is straight.

In the larger, ova-bearing female, which is 55 mm. long, the external maxillipeds
reach a little beyond the antennal peduncle; the legs of the first pair extend to the end
of the scales and those of the second reach with their chele beyond them, the carpus
extending to the end of the scales. The carpus of the second pair is once and a half as
long as the chela.

Geographical Distribution—Japan (de Haan): near the town of Simoda, in fresh
water of a river, not far from the sea (Stimpson): Aomori, Rikuoku; Matsushima,
_ Rikuzen; Misaki, Sagami; Lake Biwa, Matsubara, Omi (abundant); Kawatana;
Kurume ; Nagasaki, Hizen (Rathbun): Korea, Fusan; Gensan, brackish streams flowing
into the sea (Rathbun): Nemuro, Yesso (Doflein): Iterup, Kurilen, August (Do/flein).

LEANDER LONGIPES, Ortmann. (Pl. 32. figs. 26-30.)

Leander longipes, Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 519, Taf. 37. fig. 18.

Palemon ortmanni, Rathbun, in Proc. U.S. National Museum, xxvi. 1902, p. 53 footnote.

Leander longirostris, de Man, in Notes from the Leyden Museum, iii. 1881, p. 141 (nec Palemon
longirostris, H. M.-Edw., Hist. Nat. Crust. ii. p. 394=styliferus, H. M.-Edw., ibidem, Errata,
vol. iii. p. 638, 1840),

One adult egg-laden female from the Inland Sea of Japan. are.

410 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

As usual, Dr. Ortmann has not indicated the length that this species attains: the
present female is 58 mm. long from tip of rostrum to the end of the telson. The
carapace is just half as long, viz. 29 mm., the rostrum included, and 11 mm. without it.
The slender, elongate rostrum, which reaches with somewhat less than half its length
beyond the antennal scales, is strongly recurved and about once and a half as long as the
carapace. Different from Leander pacificus, Stimps., L. serratus, Penn., L. treillianus,
Risso, and other species, the rostrum is hardly broadened at the level of the first tooth
of the lower margin, so that it appears stiliform and not emarginate at base. As
in Ortmann’s typical male specimen from the Sagami Bay, the third tooth is placed
immediately before the anterior margin ef the carapace, the first two teeth standing
upon it; the third tooth is followed by four other teeth, the first six are equidistant,
but the seventh, which is placed just on the middle of the upturned part of the rostrum,
is a little farther from the sixth than the sixth from the fifth. The seventh tooth
is a little smaller than the preceding. As in Ortmann’s specimen, there are three apical
teeth, which are smaller than the preceding ; the first apical tooth, z. e. the eighth of
the whole series, is as far distant from the second as the third apical tooth from the
tip. The first apical tooth is a little farther from the seventh tooth than the seventh
from the sixth. The first tooth stands immediately before the middle of the carapace,
and the fifth is situated above the distal end of the basal joint of the antennular
peduncle. As in the typical male, the lower margin carries eight nearly equidistant
teeth, of the same size as those of the upper border; the first is situated just below the
fifth, the eighth just below the eighth of the upper margin, #. e. the first of the three
apical teeti.

As was rightly observed by me in 1881 (J. c.), the branchiostegal spine, which is a little
remote from the margin of the carapace, is distinctly smaller than the antennal. The
abdominal segments are rounded. ‘The telson (Pl. 82. fig. 26), once and a half as long as
the sixth segment, gradually tapers backward and ends in a sharp tooth; of the lateral
spinules the outer, 0°34 mm. long, are a little shorter than the median tooth, but the
elongate slender inner spinules are four times as long and reach far beyond the latter
(fig. 27). There are two pairs of spinules on the upper surface as usual.

The short flagellum, as long as the antennular peduncle, is united for one-fourth its
length with the outer ; 10 or 11 joints are grown together.

The external maxillipeds reach with their terminal joint beyond the antennal
peduncle.

The legs of the first pair are as long as the scales ; the carpus is almost twice as long
as the chela, and the fingers are a little longer than the palm.

The legs of the second pair are unequal, the much longer right leg (fig. 28) reaches as
far forward as the rostrum, the other only to the end of theantennal scales. The carpus
of the right leg, 5°7 mm. long, is almost as long as the merus (6 mm.) ; the carpus,
0-5 mm. thick at the proximal end, is 0°92 mm. broad at the distal extremity, here thus
twice as thick. The chela, 8:35 mm. long, is almost once and a half as long as the
carpus, and also longer than the merus. The palm, a little shorter than the fingers, is
distinctly Broader than the distal end of the carpus, and its upper surface is about

FROM THE INLAND SEA OF JAPAN. 411

three times as long as broad, the palm being 4:1 mm. long and 1:3 mm. broad in the
middle. ‘The slender fingers, which shut close together, are 4°25 mm. long ; the dactylus
carries two, small, equal, obtuse teeth near one another (PI. 32. figs. 29, 30), the anterior
of which is situated at one-fifth the length of the finger from the articulation; just
opposite the middle between both teeth the immobile finger carries one single, somewhat
smaller, subacute tooth ; the finger-tips are strongly curved inward.

The three following legs are very slender: those of the third pair reach with their
dactyli beyond the scales, those of the fifth even with one-third of their propodites.
The propodites of the fifth pair, e. g., are 84 mm. long and 0:32 mm. broad in the
middle, twenty-five times as long as broad; they thicken somewhat at the distal end and
are beset with a few spines on their distal half. The slender, tapering dactyli of the
fifth pair measure little more than one-fourth of the propodites, viz. 2°36 mm. °

The very numerous eggs are small, 0:6-0'7 mm. long and 0:45-0°5 mm. broad.

Two specimens, collected near Amoy, China, were (/. ¢.) wrongly referred by me in
1881 to Leander longirostris (Milne-Edwards, Hist. Nat. Crust. ii. p. 394): as is proved
by the examination of one of them, now lying before me through the kindness of the
Direction of the Leyden Museum, they belong in fact to Z. longipes, Ortm. In 1881
the words of Milne-Edwards’s description, ‘‘ surmonté a sa base d’une créte sexdentée,”
were misunderstood or overlooked by me. As has been shown by Miss Rathbun,
1. c. pp. 50 & 51, the species described by Milne-Edwards, /. c. p. 394, under the name
of longirostris should henceforth bear that of styliferus, M.-Edw.

In my opinion, however, Miss Rathbun was wrong when creating for L. longipes,
Ortm., the name or/manni, because this species belongs to the genus Leander; de Haan’s
longipes, however, to the genus Palemon. In that case the species mentioned by the
learned carcinologist of Washington under the name of Pal. japonicus (Ortm.,) should
also be changed, because a “ Bithynis” japonica has been described by de Haan.

Geographical Distribution—Japan, Sagami Bay ( Ortmann).

SPIRONTOCARIS, Sp. Bate.

SPIRONTOCARIS RECTIROSTRIS (Stimpson). (Pl. 82. figs. 31-34.)
Hippolyte rectirostris, Stimpson, in Proc. Acad. Nat. Sciences Philadelphia, 1860, p. 33.
Spirontocaris rectirostris, de Man, in Ann. & Mag. Nat. Hist. ser. 7, vol. xvii. 1906, p. 403.

One male and one egg-laden female from the Inland Sea of Japan.

The female, which agrees pretty well with Stimpson’s diagnosis, was captured in deep
water; as the above were the only specimens caught, it is probably a rare species. Alive,
the female was of a Prussian blue, the eggs were orange. The female is 35°5 mm. long
from tip of rostrum to end of telson and has a stout shape; the carapace, rostrum
included, measures nearly a third the whole length. The rostrum, which arises with an
obtuse crest at one-third the length of the carapace from its posterior border, reaches to
the end of the antennular peduncle ; the free portion, which measures a little more than
half the length of the carapace, projects straight forward. The upper margin carries
six teeth, three of which are on the carapace, and the first of these stands just before its

SECOND SERIES.—ZOOLOGY, VOL IX. 59

412 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

middle; the first four are equidistant, the fifth is almost once and a half as long
as the fourth, and the sixth is as iong as the fifth; the tip of the sixth tooth
is three times as far from that of the fifth as from the extremity of the rostrum.
The rostrum is a little dilated distally, just below the sixth tooth, and carries here two
teeth, which are smaller than the first teeth of the upper border; the first is situated
just below the middle of the sixth tooth, the second just below its tip, and the tip of the
second is a little farther from the extremity of the rostrum than from the tip of the first
tooth. According to Stimpson, the lower margin should be armed with four teeth.
Posterior to the first tooth the lower margin appears slightly concave. Antennal spine
short ; pterygostomian spinule very small, but distinct.

Abdomen rounded above, geniculate at the third segment, which is slightly produced
posteriorly into an obtuse lobe; on either side of the middle the tergum of the third
segment carries a faintly impressed, longitudinal line, which runs from the posterior
border to a little beyond the middle. The pleura of the fourth segment, which is
a little longer than the fifth, ends in a small sharp spinule; the sixth segment, once and
a half as long as the fifth and almost twice as long as broad in the middle, terminates, as-
also the fifth, in a sharp tooth at the postero-lateral angles. The tapering telson, which
is one-third longer than the sixth segment, is armed on its flattened, upper surface with
four pairs of spinules and terminates in a small, sharp tooth ; of the two spines on each
side the inner are twice as long as the outer, which slightly reach beyond the median
tooth. The basal joint of the uropods, which are a little longer than the telson,
terminates in a sharp tooth at its postero-external angle.

The eye-peduncles, which carry a distinct ocellus close to the corner, reach with the
latter beyond the lateral margin of the carapace.

The internal antenne are little longer than the carapace and extend with half the
inner flagellum beyond the antennal scales. Their peduncle is as long as the rostrum
and reaches the middle of the antennal scales; the first joint is somewhat longer
than the eye-peduncles when they are directed forward, and carries one or two
spinules at the distal border of its upper surface; the large and broad stylocerite is
acuminate and reaches beyond the middle of the second joint ; the second joint, not quite
half as long as the first and as broad as long, is armed, at the antero-external angle,
with a strong spine, which is directed forward and outward ; the terminal joint, finally, is
half as long as the second and has a sharp tooth or spine at the distal end of its upper
border. The upper flagellum is considerably thickened along three-fourths its length
and the filiform terminal part extends beyond the scales. The basal joint of the
antennal peduncle carries a slender spine on the distal border of its lower surface;
the straight outer margin of the scale (Pl. 32. fig. 32), which measures one-seventh
the whole length of the body, and is two and a half times as long as broad, terminates
in a sharp spine, which reaches a little beyond the rounded or truncate extremity of the
laminar portion. The antennal peduncle is as long as that of the inner antennz, and the
flagellum is somewhat shorter than the body.

The external maxillipeds, which are devoid of an exopodite, project with one-third their
terminal joint beyond the antennal scales.

FROM THE INLAND SEA OF JAPAN. 413

The legs of the first pair, which barely reach to the end of the scales, can hardly
be described as “graciles,’ as they were by Stimpson. The merus, 3°1 mm. long,
is nearly three times as long as broad and carries on its lower border proximally six or
seven small, movable spinules and some plain sete; the latter are also observed on the
lower border of the ischium. The carpus, half as long as the chela, is somewhat
excavated distally. The chela is little longer than the merus, and the fingers, which
shut close together, measure one-third its whole length. Unfortunately, the right
leg of the second pair is wanting, the left reaches to the end of the scales. The carpus,
which was not described by Stimpson, is 8 mm. long, and seems to be composed of
six joints; the first is a little longer than each of the following and the fifth is the
shortest. The chela measures little more than one-third the carpus, and the fingers
are half as long as the palm.

The legs of the third pair reach to the end of the scales, the following are a little
shorter. The meropodites of the third and fourth pairs (the fifth pair are wanting) carry
on their outer surface four or five movable spines, whereas their lower margin, like
that of the ischium, is furnished with tufts of sete.

The three posterior legs are marked with blue rings.

The oblong eggs are very numerous and small,«0'6-0°65 mm. long, 04-045 mm. broad.

If the other specimen be really the male of Spiront. rectirostris, the sexual differences
are considerable: Stimpson apparently observed only the female, though he does
not mention it. This specimen is 345 mm. long from tip of rostrum to the end of
the telson, presenting the same size as the female, but the abdomen is less deep and
appears therefore slenderer. The carapace is 12°5 mm. long, a little longer in proportion
to the whole length than in the female. The rostrum, which just reaches beyond the
distal end of the antennular peduncles for about 0°75 mm., projecting straight forward,
arises more anteriorly than in the female, viz., at one-third the length of the carapace
from its frontal border; the upper margin is armed with six teeth, which are of equal
size and equidistant, and only two of them are placed upon the carapace. ‘The upper
margin appears between the most anterior tooth and the extremity of the rostrum
somewhat convex, different from the female. As in the latter, the rostrum is dilated and
just below the foremost tooth and the lower edge carries here also two teeth, which are
much smaller than those of the upper margin; the second of these is as far distant
from the tip of the rostrum as from the first. The antennal spine and the pterygo-
stomian spinule agree with those of the female.

The third segment of the abdomen resembles that of the female, but the two
impressed lines on the tergum are wanting. The fifth segment appears a little shorter
in proportion to the fourth than in the female, the fourth being once and a half as
long as the fifth; the sixth segment appears therefore twice as long as the fifth, but it
is only once and a half as long as broad. The postero-lateral angles of the fourth, fifth,
and sixth segments terminate in a sharp spinule. The telson, almost once and a half
as long as the sixth segment, agrees with that of the female, but the four pairs of
spinules reach farther backward, so that the most posterior pair is farther from the

59*

f(yuQor

414 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

penultimate than from the posterior border, whereas in the female the contrary is the
case.

The internal antennze agree with those of the female, and the distal border of the
first joint of the peduncle carries a sharp spine near the outer angle, which has the same
size as the spine with which the second joint is armed, and the thickened portion of
the outer flagellum reaches to the end of the antennal scales. The antennal scales
(Pl. 31. fig. 84) are more elongate than those of the female; they are 6 mm. long, nearly
one-sixth the whole length and a little more than three times as long as broad; the
flagellum is as long as the body.

The most prominent difference from the female is exhibited by the external maxilli-
peds, which are much longer. These appendages, 25:5 mm. long, are twice as long as the
carapace, rostrum included, and their last two joints extend beyond the antennal scales ;
the terminal joint, just as long as the other joints together, viz. 12°75 mm., is much
slenderer than in the female and terminates in one single, brown-coloured, sharp
point.

The legs of the first pair resemble those of the female, but they are much larger, half
their chele extending beyond the antennal scales. The lower margin of the merus
carries proximally eight small spines, similar to those of the female.

The legs of the second pair extend one-third of their carpus beyond the antennal
scales. The merus is not articulate, the carpus is 7-articulate; the third joint is nearly
as long as the first and the second taken together, and longer than the others; the sixth
is the shortest, the fourth little longer than the fifth; the first, the second, the fifth,
and the seventh are nearly the same length. The chela is a little shorter than the
last three joints taken together, and the fingers are somewhat shorter than the palm.

The other legs agree with those of the female: the third extend a little beyond the
scales; the merus of the third pair carries only three movable spinules on the distal
half of its outer surface, that of the fourth only two, that of the fifth only one, near the
carpal articulation.

The difference in colour is quite remarkable, for, when caught, the male is scarlet.

Geographical Distribution —Hakodadi, in deep water (Stimpson).

SPIRONTOCARIS PROPUGNATRIX, de Man. (Pl. 82, figs. 35-41.)
Spirontocaris propugnatrix, de Man, in Ann. & Mag. Nat. Hist. ser. 7, vol. xvii. 1906, p. 404.

One specimen from the Inland Sea of Japan, caught at a depth of 6 fathoms; bottom
sand, weeds, here and there stones. It was captured together with the specimens of
Spirontocaris pandaloides, Stimpson.

Apparently a new species, distinguished by the elongate rostrum and its characteristic
toothing. Measured from the tip of the rostrum to the end of the telson, this specimen,
which has a slender shape, appears to be 33°2 mm. long; the carapace, rostrum included,
is 15-2 mm. long, little shorter than the abdomen; without the rostrum, the carapace
measures one-seventh the whole length. The rostrum, which is @ little more than twice
(namely 24 times) as long as the remainder of the carapace, arises at one-third the

FROM THE INLAND SEA OF JAPAN, A15

length of the carapace from its anterior border ; it is s¢/liform, very little dilated at the
level of the first tooth of the lower margin (PI. 32, fig. 37), and tapers gradually to the
acuminate tip. The upper margin, which is somewhat arched above the eyes, whereas
the anterior half is gently ascending, is armed with 7 rather low teeth, two of which are
on the carapace ; these teeth, which reach to the middie of the free portion, grow gradually
somewhat longer, so that the two anterior, which are of equal length, are longer than the
preceding. ‘The lower margin is armed with 10 teeth, of which the first is small and
situated below the fifth of the upper margin; these ten teeth increase also in length
from the first to the last, and reach ¢o the tip of the rostrum (fig. 36). Two-fifths of the
rostrum extend beyond the antennal scales.

_ Antennal spine small ; supraorbital and pterygostomian spines wanting.

The abdomen is moderately geniculated, the upper border of the deflexed part making
an angle of 45° with the remainder. The third segment is slightly produced into an
obtuse lobe posteriorly. The fourth and fifth segments are of swbegual length; the
postero-lateral angle of the fourth is obtuse, but that of the fifth terminates in a sharp
tooth. The sixth segment (fig. 39), almost twice as long as the fifth, is ¢wice as long as
broad, and its postero-lateral angleissharp. The slender telson, almost one-fourth longer
than the preceding segment, tapers gradually, so that the posterior margin measures but
one-fourth its breadth proximally; the posterior margin (fig. 40) ends in the middle in
a sharp tooth, and of the two spinules on either side the outer are twice as long as the
inner. The upper surface carries four pairs of spinules; the anterior pair are as far
from the base of the telson as the posterior pair from the posterior border. The uropods
are barely longer than the telson.

There is a distinct ocellus near the cornea, and the rather slender eye-peduncles
project their whole terminal joint beyond the carapace when they are directed transversely
outward.

The peduncles of the internal antennze, measuring little more than one-fourth the
length of the rostrum, reach not quite to the middle of the antennal scales ; the acuminate
stylocerite reaches to the distal end of the first joint. The second and third joints
are together half as long as the first; the second, which is once and a half as long as
thick and twice as long as the third, is armed at its antero-external angle with a strong
spine; the thickened outer flagellum reaches to the distal third of the scales, whereas
the thin inner flagellum reaches slightly beyond them.

As in other species, there is a spine on the distal border of the lower surface of the
basal joint of the outer antennze. The scales are narrow, elongate, their outer margins
straight ; the membranous portion (fig. 41), which extends considerably beyond the strong
spine, is obliquely truncate. ‘The antennal peduncle reaches as far forward as that of
the inner antennz, the flagellum measures two-thirds the length of the body.

The external maxillipeds are very short, barely reaching to the end of the antennal
peduncles ; they seem to be devoid of an exopodite and an epipodite.

The legs of the first pair, still shorter, project with their fingers, which are half as long
as the palm, beyond the basal joint of the antennal peduncle. The legs of the second
pair extend with their chelze beyoud the antennal peduncle. The carpus, once and a half

416 DR. J. G. DE MAN ON CRUSTACHA CHIEFLY

as long as the merus, is 7-articulate ; the joints measure 0°32 mm., 0°22 mm., 0:54 mm.,
0°36 mm., 0°3 mm., 0:22 mm., and 0°46 mm.; the second and sixth the shortest, the
third the longest. The chela, 0°88 mm. long, is almost as long as the last three carpal
joints together, and the fingers measure two-fifths the whole length of the chela.

The following legs are slender. Those of the third pair project nearly with half the
propodites beyond the antennal peduncles; the meropodites, eleven times as long as
broad, carry on their outer surface a longitudinal row of 8 stout, movable spines, of
which the last is inserted near the carpal articulation; the lower margin of the
propodites is furnished with 9 movable spinules, which are smaller and thinner than
those of the merus ; the dactylus carries 6 spines on its lower margin, the last is stouter
than the terminal claw, so that the dactylus appears to terminate in two claws. ‘The
following legs are gradually shorter; the meropodites of the fourth carry 7, those of
the fifth 3 spines.

The nearest allies of Spiront. propugnatrix are Spiront. stylus (Stimpson), Spéront.
gracilis (Stimpson), ard Spiront. amabilis, Lenz (confer Rathbun, ‘ Decapod Crustaceans
of the North-west Coast of North America,’ 1904).

SPIRONTOCARIS ALCIMEDE, de Man. (PI. 82. figs. 42-46.)
Spirontocaris alcimede, de Man, iu Ann, & Mag. Nat. Hist. ser. 7, vol. xvii. 1906, p. 404.

Twelve specimens from the Inland Sea of Japan.

Closely related to Spiront. gracilis (Stimpson), and Spironé. flexa, Rathbun, from the
North-west coast of North America, but apparently different.

The largest specimen is 34 mm. long from the tip of the rostrum to the end of the
telson; the abdomen, which is strongly geniculated ata right angle at the third segment,
is almost once and a half as long as the carapace (rostrum included). The slender
rostrum, the free part of which is oxce and a half as long as the remainder of the
carapace, arises with an obtuse crest at one-third of the length of the cepbalothorax
from its anterior border ; it projects at first horizontally forward, but is gently ascending
from the anterior tooth of the upper margin and the acuminate extremity just reaches
beyond the antennal scales. The upper margin, which is slightly arched above the
eyes, is armed with 5, rarely 4, pointed teeth, of which two always stand on the carapace ;
these teeth grow usually a little longer from the first to the anterior, so that they
cannot be said to be equidistant. The anterior tooth is situated in the middle of the
free part or immediately behind it, so that the terminal half of the upper margin
or somewhat more appears devoid of teeth; rarely the foremost tooth is situated in
front of the middle, in which case the terminal part, devoid of teeth, appears
somewhat shorter than the remainder. In front of the foremost tooth the upper limb is
very narrow and cannot be followed to the tip. The lower limb (PI. 82, fig. 43) is shallow,
convex, as in Spiront. unalaskensis, Rathbun, and Spiront. tridens, Rathbun, the width of
the rostrum at the base of the lower margin being only 3+ of its whole length. The
lower limb, which graduaily diminishes anteriorly, is armed with 6, 7, or 8, rarely 9, teeth,
which are smaller than those of the upper border, grow usually longer distally and reach

FROM THE INLAND SEA OF JAPAN. 417

“T) to

tothe tip. The toothing-formulz of these specimens are the following :—- two specimens ;
2 2 2 2
; four specimens ; ; two specimens ; 4 three specimens ; 3 one specimen.

According to the woodcuts in Miss Rathbun’s excellent work, the lower limb of the
rostrum appears in Spéront. gracilis, Stimpson, and Spiront. fleca, Rathbun, narrow along
its whole length, hardly broader at its base than distally.

There is xo supraorbital spine, the outer angle of the orbital margin terminates in a
rounded tooth or lobe, and the antennal spine is of moderate length. In most specimens
the antero-lateral angle of the carapace is rownded ; in two specimens only (PI. 82. fig. 4.2)
a minute pterygostomian spinule occurs on one side of the body, whereas on the other
side the carapace is rounded. ‘The third segment of the strongly geniculated abdomen
is produced posteriorly to @ somewhat compressed hump or hunch, which is bent at a right,
though rounded angle. The lateral sides of the third segment are somewhat punctate,
near the posterior border, like the others, but also below the upper margin. The
fourth segment, distinctly longer than the fifth, is rounded at the postero-lateral angle,
but the fifth ends in a sharp tooth; the sixth segment, which is twice as long or almost
twice as long as the fifth, is twice or barely twice as long as broad; its postero-lateral
angle terminates in a sharp tooth. The telson, which is but little longer than the sixth
segment and somewhat shorter than the uropods, terminates in a sharp tooth, and of the
two spinules on either side of it the outer is halfas long as the inner. The upper surface
(fig. 44.) carries 4, more rarely 5, pairs of spinules; in seven specimens there are 4 pairs,
in two 5, in two 4 spinules are observed on one side, 5 on the other, and in the last
individual the telson has 3 spinules on one side and 4 on the other.

The eye-peduncles, which carry a distinct ocellus close to the cornea, measure a little
more than one-fourth the length of the carapace (rostrum excluded). The antennular
peduncle (fig. 45) attains to one-third of the antennal scale; the acuminate stylocerite
reaches ¢o the distal end of the first joint, but never beyond it; the second joint, much
shorter than the first, has a spine at its antero-external angle, and the third, half as long
as the second, carries also a spine at the distal end of its upper border; the thickened
portion of the outer flagellum reaches, in all the specimens, somewhat beyond the middle
of the antennal scale. Antennal scale a little longer than the carapace (rostrum
excluded), slender, six times as long as broad, hardly narrowing distally; the outer
margin is a little concave, and the distal spine is not nearly so advanced as the mem-
branous portion; there is a slender spine at the distal end of the basal joint at the lower
side; the peduncle reaches to the middle of the second joint of the antennular peduncle,
and the flagellum is little longer than the body.

The external maxillipeds, though produced a little beyond the antennal peduncle,
attain only to one-third of the antennal scale; they are devoid of an exopodite, the
upper margin of the antepenultimate joint terminates in a small acute tooth, and the
terminal joint carries 7 or 8 brown-coloured teeth at the distal end.

The legs of the first pair extend their fingers beyond the basal joint of the outer
antennie, those of the second reach to the middle of the antennal scales. ‘The joints of

418 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

the carpus present in a specimen 32 mm. long the following dimensions, from the
first to the last :—0°-42 mm., 0:29 mm., 0°73 mm., 0°42 mm., 0°3 mm., 0°26 mm., and
0°52 mm.; the cheia is 0°9 mm. long, and the palm is nearly once and a half as long as
the fingers. The sixth joint of the carpus, just half as long as the last, is the shortest,
the third, which is as long as the first and the second together, the longest, as in
Spiront. propugnatrix, and the chela is almost as long as the last three joints taken
together.

The third pair reach to the middle of the antennal scale, the following are a little
shorter. The meropodites of the third legs, which are ten times as long as broad, carry a
row of six spines on their outer surface, of which the last is inserted near the distal end
of the lower margin; the lower margin of the propodites is armed with 14 or 15 pairs of
spinules, those of the distal half increasing somewhat in length. The dactyli, which
measure one-third the length of the propodites, terminate in two claws, of which the
posterior is stouter than the other, and, between the former and tie articulation, the
posterior margin is armed with six movable spines, which slightly increase in length
distally. The following legs are a little shorter; the meri of the fourth pair are armed
with four, those of the fifth with three spines.

External maxillipeds and thoracic legs seem to be devoid of epipods.

Spiront. amabilis, Lenz, of Bare Island (Spengel’s Zool. Jahrb., Syst. xiv. 1901,
p- 432, pl. 32. figs. 2a@ & 3), a typical specimen of which was kindly sent me by
the Direction of the “Stadtisches Museum” at Bremen, differs as follows:—As in
Spiront. alcimede, the outer angle of the orbital margin ends in a rounded tooth or lobe;
in the figure in Lenz's paper it appears erroneously as sharp, and in both species there
is, just below this angle, a sharp antennal spine, at the level of the upper border of the —
basal joint of the outer antenne. Spiront. amabilis carries, however, below this spine,
another also sharp tooth, apparently the pterygostomian spine, but this is altogether
wanting in Spzront. alcimede. The third segment of the abdomen of Spiront. amabilis
is less strongly curved, the posterior deflexed part is much shorter in proportion to the
anterior than in our new species, and not compressed; the sixth has a slenderer shape,
being almost three times as long (5 mm.) as broad anteriorly (1‘°9 mm.), and, according
to the figure, the telson should carry siz pairs of spinules (in the type specimen the
telson is wanting). The carpus of the second legs is 7-jointed, in the figure it appears
erroneously 6-jointed; it agrees with that of Spiront. aleimede, but the chela is as long
as the last four joints taken together. There are, however, still more differences in the
toothing of the rostrum, &c. (Concerning this locality, see Note A on page 454.)

SPIRONTOCARIS PANDALOIDES (Stimpson). (PI. 82. figs. 47, 48.)
Hippolyte pandaloides, Stimpson, im Proc. Acad. Nat. Sciences Philadelphia, 1860, p. 34.

Seven specimens from the Inland Sea of Japan, captured at a depth of six fathoms,
bottom sand and weeds, here and there stones. These prawns are, when alive, brilliant
emerald-green, and conform to Stimpson’s “ color viridis.”

The slender fusiform body is moderately geniculated at the third segment of the
abdomen, the upper border of the posterior deflexed part making an angle of nearly 45°

FROM THE INLAND SEA OF JAPAN. 419

with the anterior. Six specimens are of subequal size, their length from the tip of the
rostrum to the end of the telson varying between 50 and 56 mm.; the seventh is younger,
38°5 mm. long: the first six specimens are thus a little longer than was indicated by
Stimpson, viz. 44 mm. The slender, stiliform rostrum is horizontal and straight, or the
distal half is slightly turned upward; the free portion of the rostrum is once and a half
as long as the upper border of the carapace, rarely a little shorter, and one-fourth or one-
jifth of it extends beyond the antennal scales. The upper margin, which arises with an
obtuse carina a little before the middle of the carapace, carries in four adult specimens,
in which the rostrum is normally developed, 7, 8, or 9 teeth. These teeth, which are
rather small and ¢wo of which are always (also in the other specimens) situated on the
carapace, reach either almost to the middle of the free part ora little beyond it, so that in
one specimen the terminal part, which is devoid of teeth, appears a little longer than the
rest, whereas in the others the terminal third or a little more appears unarmed. ‘Two or
more distal teeth are longer than the preceding ; in one specimen they gradually increase
in length, but in the others this is not the case, and the longer distal teeth are of equal
or unequal length. This species is apparently variable as regards the number and the
shape of these teeth. The lower margin is armed, in these four specimens, with 8,
10, or 12 teeth, that reach to the tip; they are partly larger than the teeth of the upper
border, and grow also, more or less regularly, longer towards the tip. The basal part
of the lower margin, posterior to the first tooth, is nearly straight. In these four

Be a2) 9 2

: ; ears Ges: : :
specimens the toothing-formulee are therefore: jo, {5, 5, and ;5; in two others, in which

the rostrum is apparently not well developed, reaching not or barely beyond the

2 2 2
antennal scales, the formule are: ; and a in the young specimen, finally, it is
Stimpson mentions ae as the toothing-formula, but in not one of our specimens do ten
or twelve teeth occur on the upper border. Neither the upper nor the lower margin of
the rostrum is ciliated.

_Carapace and abdomen are smooth, though finely punctate. Antennal tooth slender,
reaching to the middle of the basal joint of the outer antenne ; antero-lateral angle
rounded, devoid of a pterygostomian spinule. Abdomen rounded, the third segment
moderately produced backward in an obtuse lobe. The pleurze of the third segment are
rounded posteriorly, those of the fourth are obtuse, but the fifth are produced, at the
postero-lateral angle, in a sharp spine. Measured along its upper border, the sixth
segment appears almost twice as long as the fifth, resembling that of Spcrontocaris stylus
(Stimps.) (cf. Rathbun, ‘ Decapod Crustaceans of the North-west Coast of North
America,’ 1904, p. 84). The sixth segment, the lower surface of which is rounded, and
the postero-lateral angles of which terminate in a sharp tooth, is just twice as long as
broad in the middle. The telson, which is a little longer than the sixth segment and a
little more than four times as long as broad at its base, tapers rather strongly ; its
rounded upper surface, armed, according to Stimpson, with 6 pairs of spinules, carries in
the six adult specimens 5 pairs, though in one of them there are 6 spinules on the left
and 5 on the right side; the telson of the young individual has but 4 pairs. ‘Che telson

SECOND SERIES.—ZOOLOGY, VOL. IX. 60

4.20 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

terminates in a sharp tooth and carries two spinules on either side, of which the inner
are twice as long as the outer.

The eye-peduncles, which present a distinct ocellus near the cornea, project almost
entirely beyond the antero-latera] angle of the carapace, when directed transversely
outward, and reach to the distal fifth of the first joint of the antennular peduncle.

The inner antenne barely reach with their thin filiform, inner flagellum beyond the
end of the antennal scales. The first joint of the peduncle, which is not quite half as
long as the antennal scales, is three times as long as the second ; the pointed stylocerite,
the outer margin of which is straight, extends barely beyond the distal end of the first
joint. The second joint, the outer border of which terminates distally in a small sharp
tooth or spine, appears once and a half as long as broad when viewed from above, and
twice as long as the third joint, which has also a small sharp tooth at the distal end.
The shorter outer flagellum is considerably thickened along two-thirds of its length and
beset with olfactory setz.

The basal joint of the peduncle of the outer antenne carries a small spine at the
distal border of its lower surface. The blade of the scale, which is five times as long as
broad, and the outer margin of which is straight, exceeds the small spine considerably
by its rounded antero-internal angle. The peduncle reaches nearly to the distal end of
the second joint of that of the inner antennze, and the flagellum is about as long as the
abdomen.

The mandibles are typical, and consist of a strong molar-process, an incisor-process, and
apalp. The molar-process carries at the distal end a subacute conical tooth and another
that is more obtuse ; the distal end is yellow-coloured and thickly covered with short setulee
or bristles. The incisor-process, almost as long as the molar-process, but much narrower,
tooth-like, narrows somewhat towards the distal extremity, which is divided into four
acute teeth, the outer one of which is a little larger than the three others, which are of
equal size ; both processes are not connected and make a right angle with one another.
The palp that originates at the base of the incisor-process is two-jointed ; the terminal
joint, as long as the other, is 0°45 mm. long, spathulate; its margins are fringed with
pubescent sete, a few of which occur also on the basal joint.

The external maxillipeds are short, reaching only to the distal end of the antennal
peduncle, and are devoid of an exopodite ; the antepenultimate joint is deeply hollowed
out along the proximal half of its lower surface, and the penultimate is half as long
as the terminal joint, which is armed with six sharp teeth at the distal end.

The legs of the first pair are very short, reaching only to the distal end of the basal
joint of the antennal peduncle. The carpus, which slightly thickens distally, is a little
shorter than the merus and than the chela; the fingers are about half as long as the
palm; the dactylus terminates in two dark-brown claws, the fixed finger in one. The
second legs (PI. 82. fig. 48) barely reach beyond the antennal peduncle. The seven
joints of the carpus, which is 4°24 mm. long, not yet twice as long as the merus, are,
from the proximal to the distal end, respectively 0°65 mm., 0°38 mm., 1°02 mm., 0'7 mm.,
0-5 mm., 0°34 mm., and 0°65 mm. long; the chela is 1°22 mm. long, the fingers 0°52 mm.
These numbers show that the first and the seventh joints are equally long, that the

FROM THE INLAND SEA OF JAPAN. 421

second and sixth are subequal, that the third is the longest and the sixth the shortest.
The chela is almost twice as long as the last joint of the carpus, and the fingers are a
little shorter than the palm.

The legs of the third pair reach to the end of the antennal peduncle, those of the fifth
little beyond the anterior border of the carapace. The merus of the third legs is armed
on its outer surface near the lower margin with 7 spines, that of the fourth with 6, that
of the last pair with 3.

Geographical Distribution.—Hakodadi, Japan (Stimpson).

LATREUTES, Stimpson.

LATREUTES PLANIROSTRIS (de Haan).
Cyclorhynchus planirostris, de Haan, Fauna Japonica, Crust. 1849, p. 175, tab. 45. fig. 7.
Riynchocyclus planirostris, Miers, in Proc. Zool. Soc. 1879, p. 55.
Latreutes planirostris, Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 505, Taf. 37. figs. 4 d-/, 4n.
Platybema planirostre, Rathbun, in Proc. U.S. Nat. Mus. vol. xxvi. 1902, p. 46.

One egg-bearing female from the Inland Sea of Japan; rare.

This specimen fully agrees with the ova-bearing females described by Miss Rathbun
(l.¢.). Itis 28 mm. long from the tip of the rostrum to the end of the telson. The sixth
abdominal segment is not quite twice as long as the fifth, and the telson is about once
and a half as long as the latter. The carpus of the legs of the first pair is not carinate
above. The legs of the second pair extend with their chela beyond the distal end of the
antennal peduncles; the first joint is slightly longer than the third and both together
are almost as long as the second; the chela, which is a little less broad than the distal
extremity of the carpus, is little shorter than the second joint, and the fingers are
distinctly shorter than the palm.

Geographical Distribution —Japan (de Haan); Hakodate and North Coast of Nippon,
10-20 fathoms (Stimpson); Cape Sima, Nippon, 18 fathoms (J/ers) ; Bay of Tokyo and
Kagoshima, Japan (Ortmann); Hakodate, Hokkaido (Rathbun).

LATREUTES ACICULARIS, Ortmann.
Latreutes acicularis, Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 506, Taf. 37. figs. 6, 6 d-k, 6 n.
Latreutes acicularis, Doflein, Ostasiatische Dekapoden, 1902, p. 638.

One ova-bearing female from the Inland Sea of Japan ; deep water.

This specimen is 31 mm. long from the acute tip of the rostrum to the end of the
telson, the carapace (rostrum included) being just half as long. The rostrum, which is
one-third longer than the carapace, is unarmed above, except a minute spinule on the
carapace, just behind the frontal border; it exceeds the antennal scales by one-third of
its length. That part of the rostrum which is situated above the lateral carinze is low
and barely diminishes in height towards the tip; its upper margin is straight. The lower
part of the rostrum is proximally about three times as high as the upper and gradually
narrows towards the tip ; it is armed near the latter with three sharp teeth.

Antennal tooth small. Along the distal end of the lower border of the carapace seven
slender spines are observed, which diminish a little in length backward, and the foremost

60*

422 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

of them is placed at the pterygostomian angle. The tapering telson terminates in a
long slender tooth, which makes distinct angles with the posterior margin, and this margin
is little broader than the tooth is long; of the two movable spines on either side the
inner is three times as long as the outer and extends much beyond the median tooth.
The rounded, upper surface of the telson carries two pairs of small spinules, the anterior
pair somewhat nearer to the proximal than to the distal extremity.

External maxillipeds short, barely reaching beyond the insertion of the antennal
peduncles. Fingers of the first pair of legs shorter than the palm, the latter a little
thicker than the carpus. The first joint of the carpus of the second legs is about once
and a half as long as the third, and both are a little longer than the second, which is twice
as long as the third; the chela is just as long as the first and the third carpal joints
taken together, the fingers being a little longer than the palm. The three other legs are
slender and a little setose. The meropodites are armed with a sharp tooth near the distal
end of their lower margin; the propodites carry six or seven movable spinules along the
lower edge, which gradually grow longer and stronger towards the distal end; the
slender dactyli, measuring a little more than one-third of the propodites, terminate in
two strong claws, which are preceded on their lower margin by six movable spinules that
diminish in length towards the articulation. So the meropodites of the fifth pair are
1:4 mm. long and six times as long as broad ; the carpopodites are a little more than
half as long as the meropodites, the propodites as long as the meropodites, but nine times
as long as broad in the middle; the dactyli, finally, are 0°56 mm. long, measuring a
little more than one-third of the propodites.

The ova are numerous, small, 0°5 mm. long, and once and a half as long as broad.

Geographical Distribution.—Japan, Kadsiyama (Ortmann); Hakodate, Yokohama
(Doflein).

LATREUTES LAMINIROSTRIS, Ortmann.
Latreutes laminirostris, Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 506.

One egg-laden female from the Inland Sea of Japan; deep water.

As usual, Dr. Ortmann has not mentioned, in the work referred to, the length attained
by this remarkable species: the present female is 53 mm. long from the tip of the rostrum
to the end of the telson. The rostrum (14 mm.) is one-fourth longer than the carapace
(11 mm.). The tapering central axis of the rostrum runs at first straight forward, then
slightly upward, whereas the pointed tip is again curved downward. That part of the
rostrum which is situated above the central axis, and which is much lower than the
inferior part, is slightly arched and carries six small acute equal teeth, four
equidistant on the middle and two midway between them and the tip. Ortmann’s
typical specimen, also a female, was armed with nine teeth above. The lower edge
carries seven much smaller teeth nearly of the same length, of which the first is situated
midway between the distal end of the antennular peduncle and the first tooth of the
upper edge; the inferior edge regularly curves, posteriorly, upward toward the central
axis. The small spine on the carapace, which is a little larger than the upper teeth
of the rostrum, is placed once and a half as far from the posterior border of the carapace
as from the first tooth of the upper edge.

FROM THE INLAND SEA OF JAPAN. 423

The short eye-peduncles do not quite reach the extremity of the first joint of the
antennular peduncle; there is a sharp, forwardly directed spine on the upper side of
this extremity. Six sharp teeth or spinules occur at the pterygostomian angle of the
carapace, and there is a small antennal spine just below the orbits.

The sixth segment of the abdomen is almost twice as long as the fifth, but a little
shorter than the tapering telson, which is 7 mm. long; the postero-lateral angles are
acute. The telson, which is rounded above, carries two pairs of minute spinules, which
were overlooked by Ortmann; the anterior pair a little nearer to the proximal than to
the distal extremity, the posterior a little nearer to the latter than to the anterior pair.
The tip of the telson is not truncate (“abgestutzt’’), as is said in the original description,
but it ends in a sharp tooth, on either side of which are inserted, as usual, two movable
spines, of which the outer is just as long as the median tooth, the inner twice as long.
The lateral swimmerets are a little shorter than the telson.

The short external maxillipeds reach as far forward as the eye-peduncles. The second
joint of the carpus of the second legs is twice as long as the first, the third appears
a little shorter than the first ; the chela is nearly as long as the second joint, the fingers
slightly shorter than the palm.

Eggs very numerous, small.

This specimen is of a pale greenish colour, the gastric region more yellowish brown.

Geographical Distribution —Japan, Tanagava (Ortmann).

HIPPOLYSMATA, Stimpson.
HIPpoLysMATA VITTATA, Stimpson, (Pl. 33. figs. 49, 50.)
Hippolysmata vittata, Stimpson, in Proc. Acad. Nat. Sciences Philadelphia, 1860, p. 26.
Hippolysmata vittata, var. subtilis, Thallwitz, Decapoden-Studien, 1891, p. 22.
Nauticaris unirecedens, Spence Bate, Report on the ‘Challenger’ Macrura, 1888, p. 608, pl. 110. fig. 1.
Nee Hippolysmata vittata, var., de Man, in Archiv f. Naturg. 53 Jahrg. 1888, p. 494.

Two egg-bearing females and one young specimen from the Inland Sea of Japan.

Dr. W. I. Calman, of the British Museum, was so kind as to examine for me the single
typical specimen (2 ) of Nauticaris unirecedens, Sp. Bate, from Hong Kong, and con-
cluded that this species ought to be considered identical with Hippolysmata vittata,
Stimps., from the same locality ; this was also my supposition. ‘“ The type specimen of
Nauticaris unirecedens,” so wrote Dr. Calman to me, “is a little larger than is stated by
Spence Bate. I think it would measure about 29 mm. in length, but I cannot attempt
to straighten it for fear of damage. The postero-lateral angle of the fifth abdominal
segment is distinctly more produced and more acute than in the original figure, The
sixth segment is longer than the fifth (about 26:20). The flagella of the antennules are
wanting, and I cannot even find any fragments of them in the bottle. There are no
arthrobranchize on the perzeopods, but there are epipods on all except the last pair.
I have compared the specimen with Stimpson’s description of Hippolysmata vittata, and
I think it very likely, as you suggest, that it is the same species. At all events, I cannot
find any character which distinctly contradicts this supposition.”

4.24 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

The two egg-bearing females are respectively 33 mm. and 32 mm. long from the tip
of the rostrum to the end of the telson, the carapace of the former being 12 mm. long,
of the other 11°5 mm.; the young specimen is 22°5 mm. long, the carapace 8 mm. The
carapace, rostrum included, measures little more than one-third of the whole length.
The rostrum of the two larger specimens reaches to the distal end of the second joint of
the antennular peduncle, that of the youngest individual little beyond the middle of that
joint. In the largest specimen the toothing-formula of the rostrum, which closely

resembles fig. 1 of plate 110 of the ‘ Challenger’ Report, is se the first tooth, twice as
far distant from the second as the second from the third, is little larger than the second,

but the difference is not so great as on that figure. ‘The distance of the first tooth from _

the frontal border is a little more than one-third the length of the carapace (rostrum
excluded) ; two teeth are on the cephalothorax, the third is placed above the frontal
margin. The foremost tooth is as far distant from the penultimate as from the pointed
tip, which is situated on a somewhat lower level than the upper border of the carapace.
The four teeth of the lower margin occur on its distal half and are considerably smaller
than the upper ones ; the first is situated just below the middle of the penultimate tooth
of the upper border, the foremost tooth midway between the tip and the foremost one
of the upper margin. In the other ova-bearing specimen the toothing-formula is =
just as in the ‘Challenger’ specimen of Nauwticaris unirecedens, and three teeth are on
the carapace, the fourth immediately before the frontal border; the foremost tooth is
once and a half as far distant from the tip as from the penultimate tooth. Of the four
small teeth of the lower margin, the first is situated just below the base of the foremost
tooth of the upper margin, the two anterior in front of it. Except the first, the teeth
of the upper border are equidistant, like those of the lower margin.

The rostrum of the youngest specimen shows the formula * two teeth are on the
carapace, the third above the frontal border; the foremost tooth of the upper margin
is a little farther from the tip than from the penultimate; of the four very small
teeth of the lower margin, the first and the second are situated below the foremost
tooth of the upper margin, the third and the fourth in front of it. In this youngest
individual the tip of the rostrum is situated at the same level as the upper surface
of the carapace. The slender antennal tooth reaches the cornea of the eye-
peduncles; the pterygostomian spinule is small and sharp, though distinct in the
three specimens.

The abdomen is rounded and smooth. The postero-lateral angle of the fifth segment
terminates in a sharp point, much sharper than it appears in the quoted figure 1 of the
‘Challenger’ Report. The sixth segment is once and a half as long as the fifth; in
Spence Bate’s figure it appears shorter than the fifth, but, as is shown above, this figure
is inaccurate. The postero-lateral angle of the sixth segment is acute, but not movable.
The telson, which is not quite twice as long as the sixth segment, tapers posteriorly to
the obtuse posterior border, which is in the middle acute and which, fringed, like the
lateral margins, with ciliated setee, carries on either side two movable spinules, of which
the inner are much longer than the outer. The somewhat flattened upper surface carries

FROM THE INLAND SEA OF JAPAN. 425

two pairs of spinules. The uropods are little longer than the telson and carry no movable
spine at base.

The stout eye-peduncles reach a little beyond the middle of the first joint of the
antennular peduncle. The internal antennz are, in the largest specimen, 38 mm. long,
a little longer than the body; the peduncle agrees with Spence Bate’s description and
figure of Nauwtic. unirecedens; the sharp-pointed basal spine or stylocerite reaches barely
beyond the eye-peduncles; the second joint is half as long as the first and twice as long
as broad, the third half as long as the second. The two flagella, however, which are
just as long as the body, do not agree with the figure in the ‘ Challenger’ Report, nothing
is said about them in the text, and, as is shown above, they are lost in the type specimen
in the British Museum. I suppose, however, that they have been wrongly figured in the
Report. These flagella are of equal length, filiform, but the outer one is slightly
thickened at its base for a short distance (4°5 mm.), which is a little shorter than the
peduncle, and this thickened part is beset with olfactory sete.

The outer antennz are as long as the inner; the basal joint of the peduncle, which
reaches midway between the tip of the eye-peduncles and the distal end of the first joint
of the antennular peduncle, carries a small spine at its outer angle; the flagella, 36 mm.
long in the adult females, are a little longer than the body. The antennal scales barely
narrow distally, and the small spine which terminates the slightly concave outer margin
reaches barely beyond the truncate tip.

The pediform external maxillipeds project with half their terminal joint beyond the
antennal scales; the exopodite reaches a little beyond the middle of the antepenultimate
joint.

The legs of the first pair extend to the extremity of the antennal scales. The carpus is,
in the adult, a little shorter than the chele, but slightly longer than the palm, that is
once and a half as long as the fingers; the latter gape a little along their proximal half.
The elongate, filiform legs of the second pair project with their small chela and the last
joint of their carpus beyond the distal extremity of the basal thickened part of the outer
antennular flagella. The carpus is composed of 22 joints; the penultimate joint is
0°3 mm. long, those in the middle are slightly longer, viz. 0°36 mm., and the last joint
is twice as long as the penultimate. The chela, 1:22 mm. long, is twice as long as the
last joint of the carpus, and the palm is a little longer than the fingers.

The three other legs apparently agree with those of the ‘ Challenger’ specimen of

yautie. wnirecedens. The meri carry on their outer surface a few movable spinules,
those of the third pair, e. g., five; the propodites carry similar spinules along their
posterior margin in two rows, those of the third legs seven pairs; the dactyli, finally,
measure, in the legs of the third pair, one-fourth of the propodites and are armed with
six spines along their posterior margin, which gradually increase in length, the last being
the terminal claw.

The eggs are very numerous, ovate, 0°6 mm. long and 0°4 mm. broad.

I am indebted to Prof. Heller, of Dresden, for having been enabled to examine the
single type specimen of Thallwitz’s variety subtilis from Cebu: it proved to differ from
our specimens only by tts smaller size.

4.26 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

One of the five specimens described (Archiv f. Naturg. lili. p. 494) by me in 1888 as a
variety amboinensis, and which were collected by Dr. Brock at Amboina, is lying before me.
It proves now to be a different species from Hipp. vittata, Stimps., and it may hence-
forth bear the name of Hippolysmata amboinensis. The whole animal has a slenderer
appearance. The rostrum, the free part of which is almost as long as the carapace,
has a much slenderer form than that of Hipp. vittata, and the first tooth of the upper
margin is situated at one-fourth the length of the carapace from its frontal border.
The sixth segment of the abdomen is more elongate, alinost twice as long as the fifth,
and barely shorter than the telson. The peduncles of the inner antenneze and the
antennal scales are more elongate, slenderer, and the stylocerite is shorter than the
eyes, rudimentary. The legs are also slenderer. A more detailed description will be
published hereafter.

The following remarks about the single type specimen of MJerhippolyte orientalis,
Sp. Bate, captured by the ‘Challenger ’ Expedition at a depth of 800 fathoms off New
Guinea, will, I think, be weleome. Dr. Calman wrote me about it as follows :—‘* With
regard to Merhippolyte orientalis, 1 am sorry that I cannot give you many details.
The type specimen is in such an extremely bad state that no one except Mr. Spence
Bate would have thought for a moment of describing it as a new species! I attempted
to make a sketch of the rostrum etc., but I thought it was not worth the trouble. The eye-
peduncles are shorder than the first segment of the antennular peduncle, probably not
more than two-thirds of its length, though on account of the membranous consistency
of all the parts it is difficult to form an idea of the exact proportions. There is a well-
marked pterygostomial spinule on the carapace. The mandible has an incisor-process
and a three-segmented palp. There are apparently arthrobranchiz on the perzeopods.
All the perzopods are wanting. JI doubt very much whether it would ever be
possible to recognize the species again.”

According to Calman’s latest paper on the Hippolytidee (Ann. Mag. Nat. Hist. ser. 7,
vol. xvii. January 1906, p. 80), this species seems to be indeed a Werhippolyte.

Another species, which was described by me in 1892 and in 1902 under the name of
Merhippolyte orientalis (in Max Weber’s ‘ Decapoden des Indischen Archipels,’ p. 407,
and in Abhandl. Senckenb. naturforsch. Gesellschaft, Bd. xxv. p. 849), is certainly different
from that deep-sea species of the ‘Challenger’ Expedition; it is also different from
Hippolysmata vittata and Hippolysmata amboinensis, and may henceforth bear the name
of Hippolysmata kiikenthali, with some doubt as regards the genus, because the mandible
was not examined by me.

Geographical Distribution of Hippolysmata vittata, Stimps.—Hongkong (Stimpson and
Spence Bate); Cebu (Thaliwitz); Pulau Bidan, Penang (Lanchester).

FROM THE INLAND SEA OF JAPAN. 427

ALPHEUS, Fabr.

ALPHEUS BREVIROSTRIS (Olivier). (Pl. 83. figs. 51, 52.)

Palemon brevirostris, Olivier, in Encyclop. Méthod. t. viii. 1789, p. 664, pl. 319. fig. 4.
Alpheus brevirosiris, Milue-Edwards, Hist. Nat. Crust. ii. 1837, p. 350.

Alpheus rapax, de Haan, Fauna Japonica, Crust. 1849, p. 177, tab. 45. fig. 2.

Alpheus rapax, de Man, in Max Weber’s Zool. Ergebnisse, 1892, ii. p. 404.

Two egg-bearing females from the Inland Sea of Japan, deep water.

Before describing these two specimens, I wish to make some synonymical remarks
about the species of the ‘ drevirostris” section of this genus, to which section the two
females doubtless belong.

In his great work ‘ Les Alpheidz : Morphologie externe et interne etc., Ann. Sc. Nat.,
Zool. 8° sér. t. ix. 1899, p. 14, Prof. Coutiére writes :—‘ I] convient d’identifier A. mala-
baricus, de Haan, avec A. rapax, Fabricius (?), Spence Bate.” In my opinion this identifi-
cation is erroneous, and de Haan’s 4. malabaricus ought to be regarded as a proper species
that henceforth may bear the name of drevicristatus, under which name this species has
been figured by the author of the ‘Fauna Japonica.’ Before me are lying a typical specimen
of A. malabaricus, de Haan, and another of 4. rapax, de Haan, both from de Haan’s typical
collection in the Leyden Museum, both specimens inadry state. Through the kindness of
Prof. Déderlein, of Strassburg, I received also four specimens of an Alpheus from the Bay
of Tokyo, described by Dr. Ortmann under the name of A. malabaricus (in Zool. Jahrb.,
Syst. v. 1890, p. 481), and two, also from that Bay, of Ortmann’s 4. rapaz (Ll. c. p. 481).
The examination of these specimens proved that Ortmann’s 4. malabaricus is really the
same species as that which was described by de Haan under this name and figured under
the name of A. brevicristatus ; and furthermore that this species is no doubt different from
Spence Bate’s A. rapax (Report on the ‘Challenger’ Macrura, p. 552, pl. 99. fig. 1).
The rostrum of 4. malabaricus passes backward into a carina which is subacute and
strongly compressed between the eyes, but which soon broadens behind the corneze and
becoming obtuse and rounded passes into the surface of the gastric region; its shape is
therefore characteristic. The second joint of the antennular peduncle of Spence Bate’s
A. rapax is described as three times as long as the first, but in his fig. 1c it appears
little more than twice as long; the peduncle resembles therefore that of de Haan’s
malabaricus. In de Haan’s A. malabaricus the antennal scale barely extends beyond the
antennular peduncle, whereas in the fig. 1 ¢ of the ‘ Challenger’ Report it reaches much
beyond it. The telson of 4. brevicristatus appears broader in proportion to its length,
and the spinules of the posterior pair are situated closer together than those of the
anterior, whereas this is not the case in fig. 1 z.

Not only the chelipeds, but also the four other legs, present a slenderer shape than
those of A. brevicristatus. In both chelipeds the upper border of the merus is obduse
and quite unarmed, but in Spence Bate’s A. rapax the upper margin ends in a sharp tooth.

In the latter species the upper margin of the larger chela carries no trace of the transverse
SECOND SERIES.—ZOOLOGY, VOL. IX. 61

428 Dk. J. G. DE MAN ON CRUSTACEA CHIEFLY

furrow near the articulation of the dactylus characteristic of Alpheus brevicristatus,
and the carinze on the outer surface of the palm are neither described nor figured in
the ‘Challenger’ Report. The dactylus of the smaller chela of 4. brevicristatus has a
slenderer form. ‘These two species are therefore certainly different.

It is on the authority of Coutiere, who has compared the type of 4. brevirostris, Oliv.,
with the Leyden type of de Haan’s 4. rapax, that these two species are considered also
now by me as identical, though I may observe that in de Haan’s 4. rapax the upper
border of the larger chela presents xo trace at all of the transverse groove near the
articulation of the dactylus which is characteristic of A. brevirostris (vide Coutiére, J. ¢.
p. 280, fig. 281, and in Bull. Soc. Entom. France, 1898, p. 250, fig. 1), and that the
antennal scale reaches barely or not beyond the antennular peduncle. In my opinion it
would be preferable to consider A. rapax of Fabricius as identical with A. brevirosiris,
Oliv., for Fabricius’s description is fully applicable to the latter. The two specimens,
apparently both males, of 4. rapaa, received from the Museum of Strassburg, seem at
first sight to belong to two different species. The larger specimen, which is 65 mm. long
from the tip of the rostrum to the end of the telson, fully agrees with the Leyden type
of A. rapax, de Haan, and ought thus to be referred to A. brevirostris, Oliv. The other
specimen, however, 55 mm. long, belongs perhaps to that species which has been
described and figured by Spence Bate under the name of A. rapa (J. ¢. p. 552, pl. 99.
fig. 1). It fully agrees with it, except the antennal scale, which, though a little longer
than the peduncles of the outer and inner antenne, has a less slender shape, being
proximally broader in proportion to its length, whereas the terminal spine barely reaches
beyond the tip of the scale. As regards the shape of the antennal scale and the peduncles
of both pairs of antennze, this specimen agrees with that of 4. brevirostris, except that
in the latter the antennal peduncle extends a little beyond the tip of the scale. The
rostral carina, acute and strongly compressed between the eyes, does not reach so far
backward as in the other older specimen, but fades away soon behind the eyes. All the
legs are @ liltle slenderer than in the specimen of 4. brevirostris. Both the larger and
the smaller chela closely resemble those of Spence Bate’s A. rapax. The larger chela is
29 mm. long and 8°75 mm. broad, the palm being 17°5 mm. long, the fingers 11°5 mm. ;
the smaller chela is 28 mm. long, the fingers three times as long as the palm, and the
ereatest breadth of the chela, about in the middle, is almost one-fourth of its length.
The fact that the fingers are longer in proportion to the length of the palm than in the
‘Challenger’ species may be explained by the larger size of our specimen. Ortmann
referred both specimens to 4. rapax, de Haan = brevirostris, Oliv. (teste Coutiere) ;
perhaps he will eventually prove to be right, if the length of the fingers of the smaller
chela of A. brevirostris is shown to be so very variable.

The two egg-laden females from the Inland Sea of Japan are of equal size, adult,
55 mm. long from the tip of the rostrum to the end of the telson. The rostrum reaches
in one female almost to the distal end of the first joint of the antennular peduncle, in the
other only to the middle of this joint; it passes into a carina, which between the eyes is
sharp, strongly compressed; the upper edge, between the eyes slightly concave, runs
obliquely upward and, reaching the upper surface of the carapace, becomes obtuse, even

FROM THE INLAND SEA OF JAPAN. 429

a little flattened, and gradually fades away about on the middle of the cephalothorax.
The surface of the latter is punctate, the puncta being larger posteriorly.

The telson, 7°5 mm. long and 4 mm. broad at its base, resembles that of the ‘ Challenger’
rapax, but the spinules of the posterior pair stand closer together. The lateral swimmerets
extend a little beyond it. The second joint of the antennular peduncle is in one specimen
twice, in the other almost twice, as long as the first and almost three times as long as the
third; the flattened stylocerite ends in a sharp spinule which reaches to the extremity of
the first joint of the peduncle. The antennal scale, slightly longer than the antennular
peduncle, has the same form as in the adult male from Strassburg; it has a much
stouter shape than that of Spence Bate’s rapax-specimen, the scale being 7 mm. long and
2°75 mm. broad proximally; the terminal spine barely reaches beyond the tip of the
scale and its outer margin is slightly concave. The antennal peduncles, reaching only
to the middle of the third joint of the inner antennee, are shorter than the scales, whereas
in the adult male from the Strassburg Museum they reach a little beyond them. The
external maxillipeds reach to the end of the antennal scales.

In one specimen the larger cheliped is placed on the right side, in the other on the
left. The larger cheliped resembles that of the ‘ Challenger’ specimen of 4. rapax (J. ¢.
pl. 99. fig. 1%). The upper border of the merus terminates in a sharp tooth, the rather
sharp infero-internal edge is beset with very small teeth and ends in a much stronger
pointed tooth. The chela is in one specimen 20 mm. long and 5°5 mm. broad, the fingers
being 7°-4 mm. long. The larger chela of the other female is 17 mm. long, 5'4:mm. broad,
the fingers 7°25 mm. long. The outerand the inner surfaces of the larger chela are finely
granulated, except the distal half of the fingers which is smooth; the outer surface of the
fixed finger is slightly concave, that of the palm presents no trace of carine; the two
carinz on the upper border are distinct, the inner, fringed with long hairs and continued
to the carpal articulation, more than the outer, which fades away nearly on the middle
of the palm. The lower edge of the chela is also fringed with hairs internally, from the
carpal articulation to the tip of the immobile finger, and the hairs along the upper
border are continued to the tip of the dactylus.

The smaller cheliped (Pl. 83. figs. 51, 52) also much resembles that of the ‘ Challenger ’
rapax-specimen, but the immobile finger is distinctly broader at its base than the dactylus,
whereas in fig. 1 of pl. 99 the dactylus appears broader than the immobile finger. The
merus, as slender as on that figure, is armed with the same teeth as that of the larger
cheliped. The chela is strongly compressed. In the larger female it is 17°25 mm. long,
the palm 6:25 mm. long and 3°5 mm. broad; in the other specimen these numbers are
145 mm.,5°5mm.,and3°7mm. The fingers gape a little and are compressed, especially
the immobile, which at its flattened base is distinctly broader than the dactylus, whereas
both taper towards the pointed, crossing tips; their inner edges are hairy. The upper
and lower borders of the chela of the dactylus are fringed with long hairs on their inner
side,

The four following legs closely resemble those of the older specimen, 65 mm. long, of
Ortmann’s A. rapax from the Bay of Tokyo, mentioned above. The second joint of the
carpus of the second legs is 4 mm. long, @ little longer than the first (35 mm.). The

G1*

430 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

meri of the third pair are 10:25 mm. long and 1'9 mm. broad, 5} times as long as broad ;
those of the fourth pair are 7:5 mm. long and 1°75 mm. broad; those of the fifth are
6°75 mm. long and 1 mm. broad.

The ova are very numerous and small, 0°6-0'62 mm. long and somewhat less broad.

The red upper surface of the body is marked with symmetrically arranged spots and
strize of a white colour, but the fourth segment of the abdomen is adorned on either side
with a dark red-brown spot, which is quite characteristic. The red inner surface of the
larger chela with a few large white flecks near the upper and lower borders; similar
flecks occur also on the inner surface of the palm of the smaller chela, and its fingers are
almost entirely white.

The ova-bearing female from the river near Pare-Pare, Celebes (de Man, in Weber,
Zool. Ergebn. 1892, ii. p. 404), was also examined by me, and seemed to belong to the
same species as the two females from the Inland Sea of Japan. Its size is a little
smaller, the upper surface of the telson is a little more rounded, and, in consequence
of its smaller size, the rostral crest is not continued so far backward; but otherwise
there are no differences.

ALPHEUS JAPoNIcUS, Miers. (PI. 33. fig. 53.)
Alpheus japonicus, Miers, in Proc. Zool. Soc. 1879, p. 53.
Alpheus japonicus, Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 476, Taf. 36. fig. 14.
Alpheus longimanus, Spence Bate, Report on the ‘ Challenger’ Macrura, 1888, p. 551, pl. 98. fig. 4.

Two males from the Inland Sea of Japan and one egg-bearing female without
definite locality, but no doubt also from the Inland Sea. This Prawn is common in
8-15 fathoms.

The two male specimens are 45 mm. long; the female 40 mm. from the tip of the
rostrum to the end of the telson.

The sharp-pointed rostrum is little shorter than the visible part of the first joint of
the antennular peduncle and extends horizontally forward. The second joint of the
antennular peduncle is about once and a half as long as the visible part of the first, and
the third is shorter than the first ; the flattened and broad stylocerite ends distally in a
sharp spinule, which reaches almost to the distal end of the first joint of the antennular
peduncle. The small spinule on the basal joint of the antennal peduncle is placed on
the distal edge of its lower surface, and is therefore not visible from above. The antennal
scale, the outer margin of which is slightly concave, is as long as the peduncle of the
inner antennee and a little shorter than that of the outer. The telson, the lateral
margins of which are slightly prominent in the middle, carries two pairs of spinules.

The external maxillipeds are a little shorter than the antennal peduncles ; according
to Spence Bate, they should extend to a little beyond them.

In both male specimens the left cheliped is the larger. ‘The infero-internal margin
of the merus is fringed with hair and terminates in a sharp tooth; the infero-external
margin is very finely denticulate, and its upper border ends also in a sharp tooth. The
larger chela, which is a little more than three times as long as broad, agrees with the
quoted descriptions and figures ; the fingers are little more than half as long as the palm.

FROM THE INLAND SEA OF JAPAN. 431

The merus of the other cheliped agrees with the described one, but there is no tooth
at the far end of the upper border and that of the infra-internal border is also rudimentary.
The wrist is a little longer than that of the left cheliped.

This species now proves to belong to those of the “ edwardsi’’-section, in which the
dactylus of the smaller cheliped presents the “‘ Baleniceps ”-torm in the male, whereas
that of the female is simple. Miers makes no mention of this character, but only two
specimens were at his disposal, probably females. Spence Bate was able to examine both
males and females ; but the difference was nevertheless overlooked by him, and his fig. 4k’
represents apparently the smaller cheliped of a female. In both males the smaller chela
(Pl. 33, fig. 53) is a little longer than the larger, and in both the fingers are somewhat
shorter than the palm. The straight upper margin of the palm, which distinctly narrows
distally, terminates in an acute lobe a short distance behind the articulation of the
fingers, but the lower margin has no constriction at all; the longitudinal depressions on
the inner and on the outer side of the palm are distinct. Just as on the larger chela,
a sharp spine occurs on either side of the articulation of the dactylus. The dactylus has
the well-known “ Baleniceps”-form: two crests, which are beset with stiff setee and which
arise from the middle of the finger, run, on its outer and inner side, forward and upward,
and unite at a short distance behind the tip; looked at from above the dactylus appears
here somewhat broadened, whereas it is narrowest in the middle. The fingers shut close
together, and their hooked tips cross one another.

The larger cheliped of the female is wanting; the smaller agrees with the figure 4 k
in the ‘ Challenger’ Report. The upper border of the merus is unarmed, but there is a
sharp tooth or spine at the distal end of the infero-internal margin. ‘The fingers are
distinctly somewhat longer than the palm, and the slender tapering dactylus is simple,
without hairy carinze. Both the upper and the lower border of the palm are entire,
without a constriction or lobe behind the articulation of the fingers. The inner surface
of the larger chela is finely granulated; the granules are wanting on the triangular
depression, on the middle of the palm, and at the base of the immobile finger, except in
the middle ; the dactylus is smooth, except at its base. The granulation on the outer
surface is less distinct.

The carpus of the second pair of legs is 5-articulate ; the first joint is almost as long
as the three following taken together, the fifth is once and a half as long as the two
preceding, which are equal and the shortest of all. The fingers are a little longer than
the palm. The other legs are slender, smooth, unarmed.

Eggs very numerous and small.

The upper surface of the body and of the peduncles of the internal antenne, as also
the inner surface of the chelipeds, are reddish.

Coutiére (“‘ Les Alpheidze,’ in Ann. Se. Nat., Zool. 8° sér. ix. 1899, p. 35) is inclined
to regard this species also as a variety of 4. edwardsii, but I am not of that opinion.

This Prawn is named the Claweracker, because it makes a loud cracking noise with its
claws which can be heard under water; if the noise is made while the animal is being
handled it is instinctively dropped, owing to the slight shock received. The exertion is
so great that the end of the big claw is frequently cast off.

432 DR. J. G. DE MAN ON CRUSTACKEA CHIEFLY

Measurements in millimetres.

Length of the body from the tip of the rostrum to 3. é. On
WH EOLOL WIE WIKOMN 5 6 5 5 5 6 6 o ao » 2E 4.6 40

meng thyofthelcananace, sient tne aaa ete alent nn nS) 15 13

ibengthvortbeslarceriche aaa aa a ai an Su) 31

Greatest breadth of the palm of the larger chela . . 8 9

ISA GEN MEE 6 6 5 a 8 oe ee og) MO 10 oe

enethvotsthe small ericliclaaaaeme meu line nne anno 35 16°5

Greatest breadthiofihe salma nese ee 3°75 4°25 2

IVE Wes WR 5 5 6 6 o o o 6 3g o LG 15°5 9-5

Geographical Distribution.—Lat. 34° 6' N., long. 136° 15’ E., at 11 fathoms; lat.
35° 7’ N., long. 186° 55’ E., at 3 fathoms (Miers) ; off Yokorka, Japan, in from 5 to
20 fathoms, and off Kobé, Japan, depth 8 to 50 fathoms (Spence Bate); Bay of Tokyo and
Tanagava, Japan (Ortmann).

PEN XUS, Fabricius.

PEenauUS (METAPENXUS) LAMELLATUS, de Haan.

Peneus lamellatus, de Haan, Fauna Japonica, Crust. 1849, p. 193, tab. 46. figs. 4 & 5.

Peneus lamellatus, Miers, in Proc. Zool. Soe. 1878, p. 808.

Peneus lamellatus, Kishinouye, in Journ, Fish. Bureau, Tokyo, vol. viii. no. 1, 1900, p. 25, pl. 6.
fig. 1.

Parapeneus lamellatus, Rathbun, in Proc. U.S. Nat. Museum, xxvi. 1902, p. 38.

One adult female from the Inland Sea of Japan.

This beautiful specimen is 75 mm. long from tip of rostrum to the end of the telson.
The upper margin of the rostrum, which, extending just beyond the eyes, reaches as far
forward as the first joint of the antennular peduncle and as the setose scale or
prosartema, carries nine teeth; of these the first is placed immediately before the
middle of the carapace and the fifth above the frontal margin; the second is almost
as far from the fifth as from the first. When this female is compared with a
typical male from the Leyden Museum, lying before me, the lower margin of the
rostrum appears to run, in the Leyden type, a little more oblique than in the female,
and this is also the case with the teeth of the upper margin, especially with the first
and the second, so that these teeth are in the male a little more erect. I cannot decide
whether this is a sexual, a local, or an individual difference.

The upper margins of the first three teeth of the rostrum and of the carina on the
third to sixth segments of the abdomen are marbled with blue and yellow, and the hairs
with which the appendages of the body, excepting the eye-peduncles and the upper
antennee, are furnished are of a beautiful red colour.

The external maxillipeds, which reach to the tip of the antennal scales, are provided
with an exopodite that reaches almost to the middle of the carpopodite. The perzeopods
are also all furnished with a well-developed exopodite; the exopodites of the fifth pair

FROM THE INLAND SEA. OF JAPAN. 433

reach just beyond the ischium, whereas those of the first reach nearly to the distal end of
the merus.

Geographical Distribution—West coast of the island of Yezo, near Cape Sooga,
lat. 45° N. (de Haan}; Hizen, Nagasaki (Rathbun).

PEN £US (METAPENEUS) AKAYEBI, Rathbun. (PI. 33. fig. 54.)
Peneus velutinus, Spence Bate, ‘Challenger’ Macrura, 1888, p. 253 (part.), nee Pen. velutinus, Dana.
Peneus velutinus, Kishinouye, in Journ, Fish. Bureau, Tokyo, viii. no. 1, 1900, p. 26, pl. 6. fig. 2;
pl. 7. figs. 11, ll a, 116.
Parapeneus akayebi, Rathbun, in Proc. U.S. Nat. Museum, xxvi. 1902, p. 39.
? Metapeneus stridulans, W.-Mason, Alcock, in Ann. & Mag. Nat. Hist. ser. 7, xvi. 1903, p. 526.

One male and one female from the Inland Sea of Japan, caught in deep water.

According to the label, this species, which is very common, has curious pea-green
eyes, the body covered with red mottled spots.

The male is 57 mm. long from tip of rostrum to the end of the telson ; the carapace
with the rostrum is 17°5 mm. long, without the rostrum 10 mm.; the sixth segment of
the abdomen, measured on median line, appears to be 8°75 mm. long, 4°6 mm. broad
anteriorly, 3°7 mm. broad posteriorly. The rostrum, which reaches to the end of the
second joint cf the antennular peduncle, is horizontal and 1+6-toothed; the gastric
tooth is situated at the anterior fourth of the carapace, as in Kishinouye’s figure ;
according to Miss Rathbun, it should be situated in adult individuals “a little in front
of the anterior third”; the foremost tooth is a little farther from the tip than from
the penultimate. The telson, which is little longer than the sixth segment, is armed
with one immovable spiniform tooth, which is preceded by three movable spines.

The short flagella of the inner antennz are little more than twice as long as the
terminal joint of their peduncle; they are of equal length, but the upper is much
stouter than the gradually tapering lower flagellum, and they reach entirely beyond the
antennal scales.

The external maxillipeds extend almost to the tip of the rostrum; their exopodite
reaches to the middle of the merus.

All the thoracic legs carry an exopodite.
The female is 60 mm. long; the carapace, rostrum included, 19°5 mm., without it

11 mm.; the sixth segment of the abdomen is, measured on median line, 9°5 mm. long,
5 mm. broad anteriorly, 4 mm. posteriorly; the telson, 9°75 mm. long, has its lateral
margins armed as in the male. The very slightly ascending rostrum, which reaches to
the end of the second joint of the antennular peduncle, is 1+7-toothed; the three or
four anterior teeth decrease a little in size, and the anterior tooth is a little farther
from the tip than from the penultimate, whereas the gastric tooth is situated at the
anterior fourth as in the male.

The external maxillipeds and the thoracic legs egree with those of the male; those of

the fifth pair carry also an exopodite.

Both in the male and in the female the carapace carries a pair of stridulating-organs,

first mentioned by Dr. Alcock in his description of Ietapeneus stridulans (Ann. & Mag.

434 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

Nat. Hist. ser. 7, xvi. 1905, p. 526). Each organ consists of a smooth, quite glabrous
band, which, arising from the posterior end of the branchiostegite, curves forward, slightly
narrowing distally; this band carries 18-20 smooth transverse ridges that run parallel
with one another; they are broadest in the middle of the organ and gradually narrow
towards both extremities.

Alcock suggests that Peneus akayebi, Rathb., may prove to be identical with Meta-
peneus stridulans, W.-Mason, from the east coast of Bengal. I hesitate, however, to
identify them, because they belong to a section of this subgenus the species of which
are very closely related and chiefly distinguishable by such slight differences as the
shape of the sixth pleonic segment and the proportion of its length to that of the
carapace. Though Alcock’s description of Metap. stridulans agrees very well with
these specimens, these characters are not spoken of. I suppose even that Pen. akayebi
is a different species, for the stridulating-organ of Metap. stridulans is usually composed
of 5, occasionally of as many as 12, transverse ridges, whereas 18-20 are observed in
Pen. akayebi. Moreover, the closely related Pen. (Metapeneus) acclivis, Rathb., is also
furnished with these remarkable organs, and the description of Metap. stridulans
agrees very well with this species.

In Wetap. stridulans the “second abdominal tergum is medially carinated in less
than its posterior half,” and this carina is suleate ; in Pen. akayebi this carina is hardly
grooved, and runs 77 the middle of the tergum.

Geographical Distribution.—Inland Sea, Bay of Ise, Japan (Kishinouye) : Wakanoura,
Kii; Onomichi, Bingo; Kawatana; Hizen, Nagasaki; Mogi, near Nagasaki, Japan
(Rathbun).

PrenzuS (METAPENZUS) ACCLIVIS, Rathbun. (Pl. 33. fig. 55.)
Parapenaus acclivis, Rathbun, in Proc. U.S. Nat. Museum, xxvi. 1902, p. 41, figs. 12-14.

One female from the Inland Sea of Japan, captured at a depth between 5 and 20
fathoms or more.

This specimen is 80 mm. long from tip of rostrum to the extremity of the telson,
Measured on median line, the carapace appears to be 26°5 mm. long, the rostrum
included, and 16 mm. without it; the sixth segment of the abdomen is 11°3 mm. long
just seven-tenths as long as the carapace; the greatest width of this segment anteriorly
is 7°25 mm., whereas it is 5°5 mm. broad posteriorly. The telson is 14°5 mm. long, just
twice as long as the greatest width of the sixth segment anteriorly ; the telson, which
gradually tapers to the acuminate tip, is armed with one pair of immovable spines, which
are preceded by three pairs of strong movable ones ; the immovable spine on either side
is barely longer than the anterior movable one, and the two following grow gradually
longer, so that the third is three times as long as the anterior spine and slightly extends
beyond the immovable. The outer swimmerets hardly reach beyond the extremity of
the telson, the inner not at all.

When the peduncles of the inner antennz are placed immediately below the rostrum,
the latter appears to reach a little beyond the end of the second antennular segment,

FROM THE INLAND SEA OF JAPAN. 435

but in the usual horizontal position of the peduncle it does not appear to extend to the
extremity of the second joint. As regards its usual shape, the rostrum evactly agrees
with Miss Rathbun’s figure, but there are only six teeth on the free part ; the gastric
tooth is situated at the anterior fourth of the carapace. The six teeth of the free part
are equidistant and of equal size except the anterior, which appears distinctly smaller
than the preceding ; this anterior tooth is once and a half as far from the penultimate
as from the tip of the rostrum.

_ The stylocerite of the inner antennze reaches to the end of the first joint, which carries
a spine at the far end of its outer border; the short flagella, which extend beyond the
antennal scales, are nearly twice as long as the terminal joint of the antennular
peduncles ; they are subequal in length, but the upper is a little shorter. The basal
joint of the antennal peduncle has no spine at the outer angle or on the distal border
of the lower surface; the outer margin of the scale is very slightly arcuate proximally,
and the distal spine reaches as far forward as the laminar portion.

The last two joints of the external maxillipeds extend beyond the antennal peduncle,
reaching almost to the tip of the rostrum; their exopodite reaches to the end of the
merus-joint. The outer footjaws, as also the legs of the first and of the second pair, are
unispinose at base, but the legs of the first pair carry, moreover, a spine at the distal
end of the lower margin of their ischium. ‘The legs of the third pair attain the tip of
the antennal scales, those of the fifth extend with little more than their dactylopodites
’ beyond the antennal peduncles, while the legs of the fourth pair are but little shorter.
The terminal joints of the fifth legs are little more than half as long as their propodites.

The thelycum does not fully agree with the figure of the original paper; it is therefore
figured afresh (Pl. 33. fig. 55).

The stridulating-organ of this species much resembles that of Pen. (Metap.) akayebi,
Rathb., but there are only 13 or 14 ridges, which gradually narrow, like the organ itself,
towards the anterior end.

Geographical Distribution —Mogi, near Nagasaki (Rathbun).

Prnmus (PARAPENEOPSIS) TENELLUS, Sp. Bate.

Peneus tenellus, Spence Bate, Report on the ‘ Challenger’ Macrura, 1888, p. 270.

Peneus tenellus, Kishinouye, in Journ. Fish. Bureau, Tokyo, viii. no. 1, 1900, p. 22, pl. 6. fig. 3, pl. 7.
fig. 8A & B.

Peneus crucifer, Ortmann, in Zool. Jahrb., Syst. v. 1890, p. 451, Taf. 36. fig. 5 a, b.

One female from the Inland Sea of Japan, captured at a depth between 5 and 20
fathoms or more.

This specimen is not yet full-grown, being 52 mm. long from the tip of the rostrum
to the end of the telson, whereas, according to Kishinouye, the female attains a length of
75 mm. The body is described, both by Spence Bate and Kishinouye, as being smooth
and naked ; in the present female, however, the carapace is very finely scabrous, being
covered rather closely with minute spinules which are only 0:03-0:05 mm. long; still
smaller spinules occur also on the telson and perhaps here and there on the other
segments of the abdomen.

SECOND SERIES.—ZOOLOGY, VOL. IX. 62

436 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

This species is characterized by all the rostral teeth standing upon the free part
of the rostrum, there being none on the carapace. Of the eight teeth, the first, which
is situated upon the anterior border of the carapace, appears a little smaller than the
two following; these are also larger than the rest, which slightly decrease in size
distally ; the anterior tooth is about once and a half as far distant from the tip of the
rostrum as from the penultimate.

A longitudinal fissure proceeds from the anterior border of the carapace, immediately
above the antennal spine, backward, and extends along two-thirds of the length of the
carapace; it was first described by Kishinouye. There is also a third transverse suture
at the level of the third pair of legs. Stridulating-organs wanting.

The rather obtuse carina of the fourth segment of the abdomen arises at one-seventh
of its length from the anterior extremity. The telson is not “about half the length of
the lateral plates of the rhipidura,” but measures about two-thirds their length, the
basal joints included. The furrow on the upper surface reaches little beyond the middle
of the telson.

The stylocerite of the inner antennz reaches not quite so far forward as the antennal
peduncle, which extends to the middle of the cornez ; the antennular peduncle is 9 mm.
long, the flagella 7°2 mm. ‘The basal joint of the antennal peduncle carries a spine at
the outer angle ; the spine at the distal end of the very slightly arcuate outer margin
of the scale reaches not quite to the rounded extremity of the laminar portion. The
flagellum is almost twice as long as the body.

The external maxillipeds, which project with their terminal joint beyond the antennal
peduncles, reach to the middle of the scales.

The legs of the third pair extend to the distal extremity of the antennal peduncles
and are unarmed at base, and those of the fifth reach just beyond the rostrum, to the
terminal third part of the scales. (See additional Note B on page 454.)

Geographical Distribution—Bay of Kobe, Japan (Spence Bate); Inland Sea of Japan
and along the lower half of this Empire (Aishinouye) ; Maizuru, Japan (Orémann).

Pen aus (TRACHYPENZUS) CURVIROSTRIS, Stimpson. (PI. 33. figs. 56-58.)
Peneus curvirostris, Stimpson, in Proc. Acad. Nat. Sciences Philadelphia, 1860, p. 44.
Peneus curvirostris, Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 451, Taf. 36. fig. 4 a, 5.
Peneus curvirostris, Kishinouye, in Journ. Fish. Bureau, Tokyo, vol. viii. no. 1, 1900, p. 23, pl. 6. fig. 4.
Parapeneus curvirostris, Rathbun, in Proc. U.S. Nat. Mus. xxvi. 1902, p. 38.
Trachypeneus curvirostris, Alcock, in Ann, & Mag. Nat. Hist. ser. 7, vol. xvi. 1905, p. 523.
Peneus anchoralis, Spence Bate, ‘ Challenger’ Macrura, 1888, p. 258, pl. 35. fig. 1 (partim ?).

Two females from the Inland Sea of Japan, caught in deep water.

They are nearly equally long, one measuring 80 mm., the other 78 mm. from the tip
of the rostrum to the end of the telson. In both specimens carapace and abdomen are
tomentose and scabrous with short adjacent setee and microscopical spinules ; the set are
0:15-0°16 mm. long, the spinules 0:03-0:04 mm. The lower margin of the rostrum,
which reaches to the distal end of the second joint of the antennular peduncle, is distinetly
curved in its ascent upwards and fringed with long cilia. In the female, which is
80 mm. long, the upper margin is 1+7-toothed; the gastric or first tooth, which is

FROM THE INLAND SEA OF JAPAN. 437

situated still a little farther from the second than the second from the fourth, is a little
smaller than the second; the first and the second are placed upon the carapace, the
third reaches for the greater part of its length beyond the anterior margin of the
carapace; the second, third, and fourth teeth are a little larger than the following,
which decrease in size, and the anterior tooth is a little farther from the pointed tip of
the rostrum than from the penultimate. In the other specimen the formula is 1+8;
the second tooth is very little farther from the fifth than from the first or gastric
tooth, and the latter is also somewhat smaller than the second ; the following teeth are,
as in the other specimen, equidistant, the two or three foremost are a little smaller
than the preceding, and the anterior tooth is twice as far from the penultimate as
from the acuminate tip, which is directed horizontally forward ; two teeth are on the
carapace, the third is situated above the anterior border. The lateral carinze of the
rostrum do not reach beyond the anterior border of the carapace. The postrostral ridge
is distinct, and extends until quite near the posterior margin of the carapace. The
sulcus gastro-frontalis (Stimpson) is indistinct in both specimens, as are also the
cardiaco-branchial grooves ; the antennal and the gastro-hepatic sulci are well developed,
and the antero-lateral part of the cervical groove, situated just below the hepatic spine
and beginning, at some distance from the anterior border of the carapace, at the
posterior end of the antennal carina, is rather deep but short, being hardly once and a
half as long as its distance from the anterior border. The outer angle of the orbital
margin is produced into a sharp though small tooth; antennal and hepatic spines well
developed, the former a little the larger. Pterygostomian angle angular, though not
produced into a tooth or spine. Stridulating-organs wanting.

Characteristic of this species (Pl. 33, fig. 57) is a fissure about in the middle of the
lower margin of the first segment of the abdomen, distinctly visible in Spence Bate’s
fig. 1 of Peneus anchoralis, which is identical with Pen. curvirostris. In both specimens
the second segment of the abdomen carries a short median carina as far from the anterior
as from the posterior margin of this segment; the carina of the third segment reaches
from the posterior margin to the anterior fourth part, the carinze of the fourth and
fifth terminate in a narrow cleft at the posterior extremity, but the carina of the third
segment does not ; the carina of the sixth segment, which is little longer than broad, is, in
the specimen 78 mm. long, posteriorly more strongly curved backward than in the other.
The telson, which is very little longer than the sixth segment, but one-third shorter than
the outer swimmerets, terminates (fig. 58) in an acuminate pointed tip; it is deeply
grooved in the middle of the upper surface, and the lateral margins carry four movable,
very small spinules ; the foremost or first spinule is inserted at little more than one-third,
the posterior or fourth at one-seventh the length of the telson from the posterior
extremity, the second is inserted just midway between the first and the fourth, and the
third immediately in front of the fourth ; the fourth is twice as large as the third, and
the two anterior are a little larger than the third. Stimpson describes the telson as
similar to that of de Haan’s Pen. monoceros, where it is armed with three minute
spinules, and Kishinouye describes it likewise : Ortmann (Spengel, Zool. Jahrb., Syst. v.
1890, p. 447) was therefore apparently wrong when denying their existence altogether ;

62*

438 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

Spence Bate is also inaccurate, for (‘ Challenger’ Macrura, 1888, p. 259) he says “‘ that one
small spinule is visible on close inspection,” whereas (p. 261) he describes three spinules.

The antennular flagella are of equal length and only one-third shorter than their
peduncle, 7.e., the distance between its distal extremity and the anterior border of the
carapace ; the spine at the far end of the outer margin of the first joint of the peduncle
is slightly directed outward.

There is only a minute spinule on the distal border of the lower surface of the basal
joint of the outer antenne. The peduncle reaches as far forward as the stylocerite of
the inner antennee, z. e. to the middle of the cornez, and the flagellum is a little more
than twice as long as the body. The outer margin of the antennal scales, which reach as
far forward as the antennular peduncles, appears very slightly arched; the spine at the
distal end extends as far forward as the laminar portion, which in the specimen 78 mm.
long narrows a little more distally than in the other, an individual difference of
course.

The external maxillipeds, of which the terminal joint extends beyond the antennal
peduncles, reaching almost to the middle of the scales, are described by Stimpson as
“nudi’” on their outer surface; in our two specimens they are, however, distinctly
hairy ; the exopodite reaches to the distal end of the merus-joint of the endopodite.

The legs of the first and of the second pair are unispinose at base, their second joint
being armed with a spine; the third legs, which reach to the tip of the antennal scales, are
unarmed. 'The slender legs of the fifth pair reach to the tip of the eyes or slightly
beyond them ; their dactyli are little more than half as long as the propodites.

Both specimens seem to have copulated. The thelycum (PI. 33. fig. 56) agrees with
Ortmann’s and Kishinouye’s figures, in the female 78 mm. long: the amorphous gum-
like substance with which it is covered resembles Kishinouye’s figure 10¢; but in the
other it has a remarkable shape, appearing as a narrow, asymmetric, shield-like body,
somewhat pointed at the distal end and divided by a transverse suture in the middle;
it is of a whitish colour, whereas the lateral margins are of a pale violet (fig. 56).

Carapace and abdomen marked with innumerable small dots or points of a dark,
perhaps bluish, colour; they are quite well visible in Spence Bate’s figure 1 of Pen.
anchoralis.

Peneus anchoralis, Sp. Bate, was founded upon specimens from the Arafura Sea and
from Yokohama; those from the latter locality are no doubt identical with Pen. curvi-
rostris, Stimps., those from the Arafura Sea certainly belonged to Pen. granulosus, Hasw.,
but it is difficult to say whether Haswell’s species, though most closely related to
Pen. curvirostris, is indeed identical with it or not. The second segment of the abdomen
of Pen. granulosus seems to be destitute of the small crest near the centre of the upper
border which is characteristic of curvirostris; the apex of the telson is acute, but not
developed into a spine as in the Japanese species, and finally the lateral margins should
be armed in Pen. granulosus only with a single, weak spine. The last named difference
explains perhaps the fact that Spence Bate describes at one place the existence of one,
but at another page that of three spinules on the lateral margins of the telson.

Dr. Alcock’s suggestion that Pen. affinis (de Haan) = Pen. barbatus, de Haan, should

FROM THE INLAND SEA OF JAPAN. 439

also be identical with Peneus curvirostris is, no doubt, erroneous. The carapace of
Pen. barbatus is rounded and devoid of a postrostral ridge; the anterior margin carries a
spine below the basal joint of the outer antennze, which does not occur in Pen.
curvirostris; and in Pen. barbatus there is a faint ridge posterior to the hepatic spine,
no trace of which exists in Pen. curvirostris.

The external maxillipeds of de Haan’s species reach to the tip of the antennal scales
and the third legs are also unispinose.

Pen. monoceros of de Haan= Pen. ensis, de Haan, appears, on the contrary, most closely
related to Pen. curvirostris. The rostrum reaches, however, to the end of the antennal
scales, and from the hepatic spine a ridge goes backward to the posterior margin of the
carapace. The outer flagellum of the internal antenne is not longer than the eye-
peduncles. The third legs are also unispinose at base; and the second somite of the
abdomen seems to be devoid of a rudimentary crest on the upper border.

Pen. monoceros, Fabr., is also a different species (vide de Man, in Max Weber's
Zool. Ergebnisse, 1892, t. ii. p. 513, pl. 29. fig. 54).

Geogruphical Distribution —Simoda (Stimpson): Yokohama (‘ Challenger’): Arafura
Sea (Spence Bate): Kochi, Bays of Tokyo and Sagami (Ortmann) : Hakodate, Hokkaido ;
Aomori, Rikuoku; Hizen, Nagasaki (Rathbun): Pacific Coast of Japan from the Bay of
Awomori to Kagoshima (Kishinouye).

Stomatopoda.
CHLORIDELLA, Miers *.

CHLORIDELLA AFFINIS (Berthold).
Squilla affinis, Berthold, ‘ Reptilien aus Neu-Grenada und Crustaceen aus China,’ Géttingen, 1846, p. 26,
tab. 3. figs. 1, 2; Bigelow, in Proc. U.S. Nat. Museum, xvii. 1894, p. 538 (ubi synon.).
Chloridella affinis, Rathbun, in Proc. U.S. Nat. Museum, xxvi. 1992, p. 55.

One specimen from the Inland Sea of Japan.

In this specimen, which is 45 mm. long from the tip of the rostrum to the end of the
telson, the oblique corneal axis of the eyes is 3°2 mm. long, ¢. e. 0°07 of the length of
the body ; according to Bigelow it is comparatively a little shorter in adult individuals,
measuring here only 0°05 ot the length of the body.

The antero-lateral spines of the carapace reach just beyond the suture between the
latter and the rostrum, whereas in adult specimens they are shorter than it. The median
carina of the carapace is bifurcated for [} of its length. The rostrum carries on its
anterior half a feeble median ridge. The telson carries between the marginal spines, on
each side of the middle line, one lateral, nine intermediate, and four submedian denticles.
The submedian spines of the telson are probably provided with movable tips, which in
that case should be a juvenile character.

* This species and the following are provisionally placed in the genus Chloridella, Miers, but it appears to

me probable that it will prove necessary to create a new name for the genus including those species that were
referred by Bigelow to Squilla, J. C. Fabr.

440 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

CHLORIDELLA FASCIATA (de Haan).
Squilla fasciata, de Haan, Fauna Japonica, Crust. 1849, p. 224, tab. 51. fig. 4.
Chloridella fasciata, Rathbun, in Proc. U.S. Nat. Museum, xxvi. 1902, p. 54 (ubi synon.).

Two specimens from the Inland Sea of Japan.

This species seems to be rare, for neither Miers in 1880 nor Bigelow in 1894, when
describing this species, had specimens of it at their disposal, whereas only two were
collected by the ‘ Challenger’ Expedition, also in the Inland Sea of Japan.

The present specimens are respectively 51 mm. and 40 mm. long from the tip of the
rostrum to the end of the telson, about as long as those that were described by Brooks ;
de Haan’s single type specimen was 5 centim. long. The eyes are described by Brooks
as “nearly cylindrical” ; in our specimen, 40 mm. long, the peduncle and still more the
corneal axis are distinctly somewhat compressed, and the latter, which is directed somewhat
obliquely as in Chlor. affinis, Berthold, measures 0°045 of the length of the body.

The spiniform teeth at the antero-lateral angles of the carapace are directed a little
outward and reach almost as far forward as the suture between carapace and rostrum.

The tooth on the middle of the outer margin of the inner spine of the basal prolonga-
tion of the uropods is obtuse, though not rounded, and the inner border is armed with
twelve or thirteen sharp teeth, which slightly increase in length distally ; the ‘ Challenger ’
specimens presented here only seven or eight teeth. The terminal paddle of the
exopodite measures two-thirds the length of the first joint, the outer margin of which
is furnished with eight movable spines, which increase in size and in length distally.
According to Brooks, the paddle measured in the ‘ Challenger’ specimens half the length
of the proximal joint. The length of the telson, measured in the middle line, is three-
fourths its greatest width. The median crest, which ends posteriorly in a sharp tooth,
carries a small notch at one-fourth of its length from the base. Between the marginal
spines are observed on each side one lateral, eight intermediate, and four or five sub-
median denticles, which are all very sharp. There are, in the smaller specimen, on the left
side five, on the right four submedian denticles. According to the label, this species
presents a red colour above; in the larger specimen the carapace and abdominal terga
are mottled with minute dark points.

Geographical Distribution.—Japan (de Haan); Inland Sea of Japan, depth 15 fathoms,
bottom blue mud (Brooks).

B.—LAKE AT YUNNAN-FU, CHINA.
POTAMON, Savigny.

PARAPOTAMON, nov. subgen.

A new subgenus Parapotamon is proposed for those Potamonide that present the
general characters of Parathelphusa, but in which the fingers of both chelipeds are
spoon-shaped, excavated at the tips. In the typical species of Parathelphusa, Parath.
tridentata, H. M.-Edw., and Parath. sinensis, H. M.-Edw., the fingers are distally acute,
pointed, and the other species of this subgenus seem to agree with them as regards this
character. In the remarkable new species of River-Crab from Yunnan, however, that

FROM THE INLAND SEA OF JAPAN. AAT

was described last year by Dr. Calman as Parath. spinescens, both fingers of both
chelipeds are spoon-shaped: though in old males the fingers of the larger cheliped
become obtuse, gradually losing their spoon-like shape, as may be observed in some
species of Leptodius. Pot. spinescens becomes therefore the type of the new subgenus
Parapotamon.

PoTaMON (PARAPOTAMON) SPINESCENS, Calman.
Parathelphusa spinescens, Calman, in Ann. & Mag. Nat. Hist. ser. 7, vol. xvi. 1905, p. 156.

Four males and one female without eggs from the Yunnan-Fu Lake, China.

In the youngest male and in the female of the same size the antero-lateral margins are
armed on each side with five spiniform teeth; in the male, however, the last tooth on the
left side is rudimentary and a smaller granule is situated just before it. In the male the
right cheliped is somewhat larger than the left. The right chela, which is 24 mm. long,
is almost just as long as the length of the cephalothorax in the middle line; the fingers
are a little shorter than the palm and barely longer than the palm is high. Though the
dactylus is nearly straight, there is, however, a small hiatus between both fingers ; both
fingers carry fine punctuations, which are partly arranged in longitudinal rows. In
Calman’s somewhat larger male the dactylus was “slightly arched and very obscurely
furrowed.” The outer surface of the palm is guile smooth, finely punctate. The fingers
of the left chela are just as long asthe palm. The anterior border, articulating with the
chela of the upper surface of the carpus, carries a few small denticulations; otherwise
the upper surface is smooth, punctate. In the female the right chela is very slightly
larger than the left; the fingers are a little shorter than the palm and a little longer
than the latter is high. Fingers and palm of the left cheliped are equally long. The
fingers of both chelipeds are, in this male and in the female, spoon-shaped ; the margins
ot the spoon-shaped tips of the fingers are white. The three other males are of much
larger size ; in the largest the cephalothorax is 54 mm. broad, the antero-lateral margin
of the right side is armed with five teeth, that of the left with seven, the last being rudi-
mentary ; the first and second are grown together at the base, as are also the third and
fourth. (See Note C on page 454.)

Measurements in millimetres.

1 2. 3.
d OF
Greatest breadth of the cephalothorax at the level of the
penultimate antero-lateral teeth. . . . 5A 34:5 34
Length of the cepbalothorax in the middle sites paikuont ‘the
abdomen . . : : Myre! So sages 25 24
Distance between dis lama or bital ie Bes sti, eel iyo oh ome sic 21:5 21
breacimmomtbedrontal border . yo. . - 5 « » » « « 145 10 10
Wengthvot the Jarger (right) chela. . .... .. . . 54 24. 20

Nos. 1 and 2 the largest and the youngest males, No. 3 the female.
Both specimens were collected, together with the types described by Dr. Calman, in
the lake at Yunnan-Fu.

44.2 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

PoraMon (PARATHELPHUSA) ENDYMION, de Man. (PI. 33. figs. 59-63.)
Potamon (Parathelphusa) endymion, de Man, in Zoologischer Anzeiger, xxx. 1906, p. 35.

Two egg-laden females from the lake at Yunnan-Fu, China, that were collected
together with specimens of Potamon (Parapotamon) spinescens (Calman).

This species is related to Pot. spinescens, Calman, and Pot. (Parapthelphusa) lanzi,
Doflein, but is of smaller size.

Carapace three-fourths as long as broad, just as in Pot. spinescens, when the lateral
spines are included; the carapace presents its greatest width at about one-third of its
length from the frontal margin. The upper surface, flattened on its posterior half,
curves anteriorly towards the rather steeply deflexed front; it is also flattened trans-
versely, but the anterior branchial regions are somewhat swollen, more than in Potamon
spinescens, and they slope steeply down to the antero-lateral border, which is not the
case in Calman’s species. The two lateral furrows of the H-like figure on the middle of
the carapace are shallow, though distinct, but the transverse median part is hardly
discernible. On the outer side of each lateral furrow a transverse impression is
observed, bounding the inner part of the anterior branchial region posteriorly. The
posterior branchial regions are also somewhat inflated and separated by shallow
impressions from the anterior. The branchio-cardiac impressions are shallow, like
those between the intestinal and branchial areze. The distance between the external
orbital angles is little more than half the greatest width of the carapace (the spines
included), the proportion being in the larger specimen as 5:9, but in the other female
that distance is comparatively larger. As in Pot. (Parapotamon) spinescens, the post-
frontal crest is only represented by the two barely prominent, epigastric lobes; the
mesogastric furrow between them is, however, somewhat deeper, and, instead of being
rugose, these lobes are distinctly granular, each being beset with ten or twelve rounded
granules. An oblique, shallow, though interrupted furrow or impression defines the
gastric region laterally ; on each side of it is another impression, of which the outer one,
on the branchial region, is larger than the other. The gastric and epibranchial regions
slope anteriorly down, so that the upper surface is somewhat concave behind the orbits.
The gastric region is granular anteriorly, the granules being similar to those of the
epigastric lobes; the swollen, anterior branchial lobes are also granular anteriorly and
near the antero-lateral border, but the granules near the latter are somewhat larger. In
the younger female they are less distinct. The rest of the upper surface is smooth, very
finely punctate.

The front (Pl. 33, fig. 60) resembles that of Potamon spinescens, but ils upper surface is
distinctly granular ; the granules are, however, smaller than those of the epigastric lobes ;
in the larger specimen the anterior border is half as broad as the distance between the
external orbital angles. The finely granulated frontal margin is notched in the middle
line, nearly as in Potamon spinescens, and the external angles are rounded; it makes right
angles with the lateral margins, so that the latter run at first parallel and then curve
outward. The upper orbital margin is also granular in the middle and externally, but
in the younger specimen it is still smooth, like the frontal border. The external orbital

FROM THE INLAND SEA OF JAPAN. 448

angles are not sharp and reach almost as far forward as the front. The antero-lateral
margin is not shorter than the postero-lateral. The slightly convex outer edge of the
flattened, extraorbital tooth is beset with three or four subacute granules, but its upper
surface is smooth. The granules of the extraorbital tooth are followed, in the larger
female, by six or seven spiniform, sharp teeth, of which the last is much smaller than the
rest, which are subequal and nearly of the same size; in the other specimen there are
eight spines on the left and seven on the right side, those of the left being a little more
unequal. These teeth or spines are smooth and glabrous. The lateral spines of Pot.
(Parapotamon) spinescens are more acuminate and there are only five or six on each
side. The rounded postero-lateral margins are smooth, converge less strongly back-
ward than those of Pot. spinescens, and run almost parallel. The posterior margin of
the carapace is just half as broad as the latter 1s long and appears thus comparatively
broader than in Pot. spinescens; it is also much more concave than in that species.

The orbits are, in the larger specimen, a little more than half as broad as the frontal
border; as they have exactly the same measurements in both females, they appear
in the younger specimen comparatively larger. They are differently formed than in
Pot. spinescens. The orbits are move regularly oval, their outer margin being move
regularly curved (P\. 33. fig. 60); the lower margin is more distinctly crenulate than
the upper, and, as in Pot. (Parapotamon) spinescens, the inner angle is not produced at
all, but there is an internal suborbital lobe, inserted between the inner angle and the
basal joint of the outer antennze. In Pot. spinescens this lobe is semi-elliptical, obtuse,
its outer surface is concave, and it reaches beyond the middle of the orbital hiatus; in
Pot. endymion the lobe is subacute and considerably smaller, its length being only one-
third of the width of the orbital hiatus. The basal joint of the outer antennz has also a
different form in the two species ; in Pot. spinescens it is just as long as broad, it does not
reach the front, and its outer surface is slightly convex ; in this new species, however,
the basal joint is longer than broad, its outer surface is quite flat, and it narrows
distinctly towards the front, which it attains (fig. 61).

The subhepatic and subbranchial regions are smooth, but the pterygostomian area is
somewhat granular on its outer half.

The epistome, which is smooth, is w little longer in proportion to its breadth than in
Pot. spinescens. In that species the granulated posterior margin is not prominent in
the middle, but the granulated ridge, which extends from the middle backward into the
bueeal cavity, is very prominent. In Pot. endymion we observe just the contrary, the
granulated posterior margin of the epistome forms a prominent tooth in the middle line,
but the ridge into the buccal cavity is hardly distinguishable.

The external maxillipeds (fig. 62) are characteristic. The longitudinal groove on the
ischium, which in Pot. spinescens is well-marked, is wanting, or, at the utmost, a faint
trace of it is discernible; this joint is punctate, the puncta are a little larger and more
crowded near the inner border. The merus-joint, which in Pot. spinescens is once and
a half as broad as long, appears in Pot. endymion hardly broader than long; in the
larger female it is 2 mm. long, but only 2°4 mm. broad; it is quadrangular, the

straight inner border and the equally long antero-internal being much shorter than the

SECOND SERIES.—ZOOLOGY, VOL. IX. 63

44 A DR. J. GQ. DE MAN ON CRUSTACEA CHIEFLY

oblique outer border, which is nearly straight, though curving a little at each extremity ;
anteriorly the merus-joint is obtuse. On this joint the puncta are larger near the inner
border. The finely punctate exognath reaches barely as far forward as the inner border _
of the ischium, but in Pot. spinescens it extends almost to the middle of the merus.

The finely, though rather densely, punctate terminal joint of the oval abdomen is
regularly rounded, its posterior border is just twice as broad as this joint is long; the
penultimate joint is almost exactly as long as the terminal, the preceding grow
gradually shorter; the punctuations are very fine and rare.

In both females the right cheliped is somewhat larger than the left. The upper
margin of the merus carries a subterminal spine, which is preceded by a smaller one, and
several sharp granules; the lower margin carries also four or five, subacute, spiniform
teeth and there are a few granules on the anterior border. The lower surface of the
merus is wrarmed, presenting no tooth or spine near the carpal articulation. Just as
in Pot. (Parapotamon) spinescens and Pot. (Parathelphusa) lanzi, Dofl., the carpus is
armed internally with éwo unequal, pointed spines, of which the upper is the larger; its
upper surface is somewhat granulated, the granules being larger on the inner side. The
length of the larger chela is two-thirds of the greatest width of the carapace; the palm
is a little longer than the fingers and a little longer than its height at the articulation of
the latter. The rounded upper border of the palm and its outer surface are beset with
subacute granules, which are rather small and not very numerous ; the lower margin is
smooth, as also the inner surface. Characteristic is the immobile finger (Pl. 38. fig. 63)
of the Jarger cheliped. This finger is somewhat curved at its base, the prehensile
edge is here emarginate, whereas the lower border bulges somewhat out; between the
emargination and the subacute tip of the finger it carries eight or nine small, obtuse,
somewhat unequal teeth. As the dactylus is nearly straight, there is proximally a small
interspace between the fingers ; the dactylus carries on the proximal half of its upper
border six or seven obtuse granules, and it presents longitudinal rows of puncta, one of:
which runs on the middle of the upper border. The dactylus is provided with eleven or
twelve obtuse teeth, which diminish in size towards the tip. The immobile finger is
punctate and marked with a longitudinal furrow on its outer surface.

The fingers of the smaller chela are barely shorter than the palm, the immobile finger
is not emarginate at its base and scarcely bulges out; each finger carries about a dozen
low, obtuse or subacute teeth, which diminish in size towards the tip; the tips of the
fingers are pointed.

The ambulatory legs are smooth, somewhat punctate. The upper margin of the
merus is slightly denticulate and ends in a sharp, though small, subterminal tooth ;
these teeth are, however, on the last pair and in the younger specimen less distinct.
The upper border of the following two joints is also finely denticulate internally; the
lower border terminates, both at the outer and at the inner side, in a small sharp tooth —
and one or two smaller teeth occur on the middle of that border. The spinulose
dactylopodites are, on all the legs, distinctly longer than the propodites. The ambu-
latory legs of the younger individual are a little pubescent on the upper border of

the joints.

FROM THE INLAND SEA OF JAPAN. 4A

The eggs are not numerous, globular, large, their dianeter being 1°75-1'8 mm.

Pot. (Parathelphusa) endymion may easily be distinguished from Pot. (Parapotamon)
spinescens, Calman, by the shape of the carapace, by the granulatiops on the anterior
regions and on the chelipeds, by the more numerous teeth on the antero-lateral border,
by the shape of the orbits and of the epistome, by the characters of the outer footjaws, by
the remarkable shape of the immobile finger, by the extremities of the fingers, which are
not spoon-shaped, &ce.

Pot. (Parathelphusa) lanzi, Dofl. (Pl. 38. figs. 64, 65), of which several typical syecimens,
both males and females, were kindly sent me for examination by Dr. Doflein, of Munich,
is also a quite different species. The cephalothorax of this larger species is longer in
proportion to its breadth, it is somewhat more arched both longitudinally and from side
to side, the interregional furrows are deeper, the lateral parts of the postfrontal crest
much more distinct, the lateral margins of the front more oblique, the antero-lateral teeth,
the orbits, the epistome, the outer footjaws, all are different from Pot. endymion; the
chelipeds, the chele, and the tapering fingers are slenderer, their pointed tips more
acuminate, the immobile finger presents the usual form ; the characters of the ambulatory
legs finally are also different.

Measurements in millimetres of the two specimens of Pot. endymion.
No. 1. No. 2.

Greatest width of the carapace, the spinesincluded . . . . . 22°5 20
Length of the carapace, in the middle line, the abdomen yale 5 ly 15
Distance between the external orbital angles . . . . . . . . 126 12'3
Breadthyolthetrontal marci) 0s so 2 2 pss es ss sl GS 5°5
breadth Otsthespostenlormarein | 3) 6. sf). 5 « » «© » « Ste0 75
Breadth of the orbits . Dees ke GN age eo hy eo 3°25
Flere eOleuneOCblismeeet Eerste tebe) ey ts 3 fe) se S272 2°25
ene thotthellarcerchelases ta 8c 43 ty ee ee ak 3 ek el MASS 12

3 5 mpallinaye! Tai yia |: a oles tS Be FSIS: 7
Height of the palm at the artealatio of ‘the ee aed ait Va HOwS 5:3

Measurements in millimetres of four typical specimens of Potamon lanzi, Dofl., from
the lower River Han, China, a few days from Hankow (Museum of Munich).

3 3 2. 2
Greatest width of the carapace, spines included . . . . 35° 26 34-5 30
Length of the carapace in the middle line, without the
abdomen . . . « 2 SUN coy sia sy 29 22°3 29 24:5
Distance between the aie ctal orbital ANTICS Ge Mo ts ed 19°5 23°75 20°3
Breatsujotthefrontal margin... . . +... =~. #10 8 10 8:5
PEPSDOSUCRIOUMMATOIN GM ew 3 + ss 2 « LL 9 13°5 11:3
Breadth PMUETOLDIUSaie wraenee ey Siti CL ae 8 6 55 6 55
Height of the orbits . . . Pes tek! 3°2 3°6 35
Length of the antepenultimate joint of the sndeion:. Ses. Behe G 272
Fy » penultimate i - sk nh ape 3

63*

446 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

Measurements of four typical specimens of P. lanzi (continued).

3 é. 2. g.
Breadth of the anterior margin of the penultimate joint . 6 4°6
a 3 posterior 7 Jp es Aaa eS 6
Length of the terminal joint. 5 3°6 ae
is » larger chela 275 186 22°5 138
"3 Bs palm . TEENS Msn Aut 15:5 10 13 10
Height of the palm at the articulation of the fingers . 11:3 7 9°3 675

C.—DARJEELING, BENGAL.

PALZMON (PARAPALEMON ?) HENDERSONI, de Man. (Pl. 38. figs. 66-68.)
Palemon (Parapalemon?) hendersoni, de Man, in Ann. & Mag. Nat. Hist. ser. 7, vol. xvii. 1906, p. 405.

Three specimens from Darjeeling, fresh water, at a height of 2500 feet, collected by
Mr. J. A. Gammie.

Apparently a new species, which I have the pleasure of dedicating to Prof. J. R. Hender-
son, of Madras, author of ‘“‘A Contribution to Indian Carcinology,” and other useful papers.
This species is somewhat related to Pal. altifrons, Hend., from Delhi and Lahore, to
Pal. scabriculus, Heller, from Ceylon, and Pal. (Macrobrachium) latimanus, v. Martens.

The largest specimen is 61 mm. long from the tip of the rostrum to the end of the
telson; the carapace, rostrum included, is 25 mm. long, almost half the length of
the whole body. ‘The carapace is scabriculate on its anterior half, being closely covered
with minute spinules; the branchial regions are smooth, finely punctate, but on the upper
surface of the carapace the scabriculate area reaches nearly to the posterior border.
The rostrum, which is siort, reaching only to the middle of the penultimate joint of the
antennular peduncle, arises from the anterior third of the carapace, and its free part is
directed obliquely downward, so that the acute tip is situated at a much lower devel than
the surface of the carapace. The upper border (Pl. 33. fig. 66) is armed, in the two larger
specimens, with seven rather small and low teeth, in the third with six. The first tooth,
situated at one-fifth of the length of the carapace from the frontal border, is, in the
two larger specimens, twice as far from the second as are the following, which reach
to the tip; in the third specimen, however, the six teeth are equidistant. The first three
teeth are situated on the cephalothorax. The nearly straight lower border carries, in
the largest specimen, one single tooth not far from the tip and placed immediately
below the foremost tooth of the upper border; in the smallest specimen there is also one
tooth on the lower border, but it is situated between the foremost tooth of the upper
margin and the tip. In the third specimen, which is 52 mm. long, the lower border
carries ¢wo teeth, situated just bebind and before the foremost tooth of the upper border.

. 344, 344 343
The formule of the three specimens are therefore: “**» “$", and +“. The free part of

the rostrum is narrow, and that part which is situated above the lateral crest appears
in the middle of the free part but little higher than that below it.

The antennal spine is small and reaches barely beyond the frontal border. The hepatic
spine is extremely small and, in the two larger individuals, it is even wanting on the right

a

,

FROM THE INLAND SEA OF JAPAN, 4A7

side; it is placed rather far below the other, for their tips are twice as far from one
another as the hepatic spine from the frontal border of the carapace.

The telson, once and a half as long as the sixth segment and almost three times as long
as broad at its base, ends in an acute tooth; the inner of the two spines on either side
exceeds, as usual, the tip of the telson. Of the two pairs of spinules on the upper surface,
the anterior stand a little behind the middle ; the four spinules are, in the larger specimen,
arranged in a quadrant, but in the others the posterior pair are situated a little closer to
the anterior than the spinules of the anterior pair are distant from one another.

The eye-peduncles are small and reach barely beyond the middle of the first joint of
the peduncle of the inner antennz. The two outer flagella are united for a short
distance, which is barely as long as the last joint of the peduncle.

The external maxillipeds reach as far forward as the peduncles of the inner antenne.

The legs of the first pair extend, in the largest individual, with the distal fifth of
their carpus beyond the tip of the antennal scales, but in the following somewhat smaller
specimen by one-third of the carpus; those of the third specimen are lost. The carpus
is as long as the merus and one-third longer than the chela, their proportion being as
4:3; the fingers are nearly as long as the palm.

Unfortunately only in the largest specimen one leg of the second pair is present, in
the two others both are wanting. The remaining leg (Pl. 33. fig. 68) is the left and, as
I conclude from the size of the coxze, apparently the smaller. This leg is 48 mm. long,
twice as long as the carapace, the rostrum included, but a little shorter than the whole
body; one-fourth of the carpus extends beyond the antennal scales. The merus,
85 mm. long, when measured along its upper border, is cylindrical, but, it is somewhat
thickened distally; it is here 3-4 mm. thick, at the proximal end, however, 2 mm.;
this joint reaches as far forward as the peduncles of the inner antennz. The
carpus, 6°5 mm. long, is distinctly shorter than the merus; it regularly thickens
a little towards the distal end and, though generally cylindrical, appears very slightly
compressed, as this joint is 3°6 mm. broad at the distal end of its upper surface, but
3°25 mm. at that of its lateral side. The chela is 22 mm. long, three times as long as the
carpus, the palm is 10°25 mm. long, appearing very slightly shorter than the fingers.
The upper surface of the palm is 3°7 mm. broad at the articulation of the fingers,
3°6 mm. in the middle, and still a little less broad at the proximal extremity, being
therefore barely broader than the carpus; in a lateral view, however, palin and fingers
appear to narrow regularly from the carpal articulation to the tips of the fingers, the
palm being 3 mm. thick proximally and 2°3 mm. at the articulation of the dactylus.
The palm appears therefore also slightly compressed in the proportion of 3:4. Viewed
from above, the fingers do not appear to narrow towards their tips, which are strongly
curved inward ; they shut close together. The fingers are somewhat tomentose; the fixed
finger carries a very small, conical tooth at the end of the cutting-edge, ¢. e., at about
one-third of its length from the articulation, and between this tooth and the articulation
an elongate low prominence is observed which carries two or three small obtuse teeth.
The cutting-edge of the dactylus terminates also in a small, conical tooth just behind the
middle, and midway between this tooth and the articulation is a slightly larger, some-

448 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

what compressed, conical tooth. The ischium- and the merus-joints are closely beset
with minute spinules, which are a little larger on the lower border; carpus and palm
are also everywhere covered with similar minute spinules, but the larger spinules of the
lower border of the merus are wanting. The fingers are devoid of spinules, except a
few at their base.

The three following legs are moderately slender; those of the third pair reach to the
distal end of the antennal scales, the fourth are a little shorter, and the fifth extend but
little beyond the middle of the scales. The merus-joint of the third pair is somewhat
spinulose and setose on its lower border; on the two following legs the spinules and sete
become gradually less numerous ; the rest of the surface is nearly smooth. The carpo-
and propodites are also nearly smooth, but the latter are spinulose on their lower border.

Pal. altifrons, Hend., differs by the more numerous teeth on the upper border of the
rostrum, which has a different form, being vertically deep and having the apex placed
in the same horizontal line as the surface of the carapace. The second legs are practically
cylindrical, have a slenderer form, and the fingers are shorter than the palm.

Pal. scabriculus, Heller, is also distinguished by the more numerous teeth of the some-
what longer rostrum, six of which are placed on the carapace. The carpus of the second
legs is not shorter than the merus, and the fingers are much longer than the palm.

Pal. latimanus, v. Mart., finally, of which a male 75 mm. long, from the Island of
Halmahera, is lying before me, differs also by its vertically deeper and longer rostrum ;
the telson has a slenderer shape, and the characters of the second legs are different.
The carpus, indeed, is much thinner at its base, its shape being thus quite different; the
fingers are shorter than the palm, their cutting-edges shorter, their teeth much more
numerous; the palm, finally, is distinctly broader than the carpus.

These three species show, however, no doubt, still other differences of less importance.

D.—THURSDAY ISLAND, TORRES STRAITS.

Pen £2us (PENZUS) LATISULCATUS, Kishinouye, var.? (Pl. 33. fig. 69.)
Peneus latisulcatus, Kishinouye, in Journ. Fish Bureau, Tokyo, vol. vii. no, 1, 1900, p. 12, pl. 2. fig. 2,
pl. 7. fig. 2A.

One female, dredged in 5 fathoms, May 21st, at Thursday Island, Torres Straits.

This specimen agrees pretty well with Kishinouye’s description of Pen. latisuleatus
from Japan, except as regards the thelycum. This female is 87 mm. long from the tip
of the rostrum to the end of the telson. The rostrum reaches to the middle of the third
joint of the antennular peduncle and is armed above with ten teeth, below with one; the
first tooth, which stands just before the middle of the carapace, is a little more than
twice as far from the second as the second from the third, and the distance of the
foremost tooth from the tip is but little shorter than that between the two first teeth.
The distance (12°38 mm.) behind the posterior tooth is a little more than once and a
half as long as that (7°75 mm.) from this tooth to the orbit. The first four teeth are
on the upper surface of the carapace, and the fifth is situated just above the orbital

FROM THE INLAND SEA OF JAPAN. 449

margin. The single tooth of the lower margin is situated immediately below the fore-
most tooth of the upper. The dorsal carina, which is distinctly grooved and which
terminates abruptly at the distance of 1 mm. from the posterior margin, widens a little
backward; the also quite distinct, lateral furrows are much broader, nearly twice than
the median groove and reach as far backward as the latter.

The fifth and the sixth segments of the abdomen are carinate, the carina of the sixth
terminating in an acute tooth; on the outer surface of these segments are observed, just
below the middle, three short ridges which run parallel with one another, the lower
margin of the sixth terminates in a small acute tooth. The telson, little longer than the
sixth segment, is deeply grooved in the middle line, and the lateral margins carry on
the posterior half three small movable spines; the middle spine is twice as far from the
anterior as from the posterior, and the posterior is as far from the tip as from the
anterior spine.

The external maxillipeds reach to the distal end of the first joint of the antennular
peduncle. The legs of the first pair extend to the middle of the antepenultimate joint of
the outer footjaws, those of the second pair to the distal end of the penultimate joint,
those of the third pair finally reach by their fingers beyond the extremity of the external
maxillipeds. The subequal legs of the fourth and fifth pairs reach to the distal end of
the antennal peduncles. The basipodites of the legs of the first and second pairs are
armed on their inner border with a slender spine, and the arched lower margin of the
following joint terminates, in the first pair of legs, in a very smal/ acute tooth (no spine).

The thelycum (Pl. 33. fig. 69) is composed of two lateral walls, the outer surface of
which is flattened, triangular, and narrowing somewhat anteriorly ; the inner margins
of these plates are in contact along their posterior half, whereas they diverge along
the anterior. The two lateral walls lean anteriorly on an arched transverse piece,
situated between the coxze of the fourth pair of legs posteriorly and bounding the
cavity anteriorly ; this transverse piece carries anteriorly a concave protuberance, barely
as long as the transverse piece itself, and terminating anteriorly in a small subacute
tooth.

The thelycum of the typical Japanese Pen. latisulcatus differs apparently by the
lateral plates, which are in contact with each other nearly along their whole length,
and the protuberance has also a different form. Perhaps this species may therefore
eventually prove to be distinct, though I fully agree with the opinion of Lanchester
(in Proc. Zool. Soc. 1901, vol. ii. p. 571), ‘‘ that too little is known about the thelycum,
and its possibly seasonal varieties within the same species, to justify the founding of
a new variety on this character.”

Probably one female from Thursday Island should be referred to Pen. caniculatus,
Oliv., var. australiensis, Sp. Bate, but I hesitate to do so, as this variety is still
insufficiently known.

450 DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

E.—COAST OFF BAHIA.

PEN £US (PENZUS) BRASILIENSIS, Latr.

Peneus brasiliensis, Latreille, in Nouv. Dict. Hist. Nat. xxv. p. 256 (1817); Miers, in Proc. Zool. Soe.
1878, pp. 299, 306; von Martens, in Archiv f. Naturgeschichte, xxxviii. Jahrg. 1872, p. 140;
Rathbun, in U.S. Fish Commission Bulletin for 1900, vol. ii. p. 100.

Five young specimens, dredged in a depth of 2} fathoms off Bahia.

The largest specimen is 80 mm. long from the tip of the rostrum to the end of the
telson, the smallest measures 65 mm. In all the basipodite of the first and second pairs
of legs is armed with an acute spine, as also the ischium of the first pair; the third pair
of legs is unarmed. In each specimen the lower margin of the rostrum is armed with two
teeth ; in three specimens the anterior of these teeth is placed just below the foremost
tooth of the upper border, in the two others the posterior tooth is placed below it. In
two specimens the upper border is armed with nine teeth, in one with ten, in two with
eleven, and in all the first four teeth are placed upon the carapace.

According to Miss Rathbun, the carina on the fourth segment of the abdomen is very
sharp in adult individuals which are 165 mm. long ; in our younger specimens this carina
is still only faintly developed. Otherwise these specimens agree with the descriptions
in the references given above.

SICYONIA, H. M.-Edw.
SIcyonra scuLpra, H. M.-Edw., var. ?

Sicyonia sculpta, H. Milne-Edwards, Hist. Nat. Crust. ii. p. 409; Heller, Die Crustaceen des siidlichen
Europa, 18638, p. 291; Spence Bate, Report on the ‘ Challenger’? Macrura, p. 294, pl. 43. fig. 1.

Two females without eggs and one male were dredged off Bahia at a depth of
2+ fathoms.

Sicyonia sculpta, which inhabits the Mediterranean and Adriatic Seas, has also been
captured off St. Vincent, Cape Verde Islands, by the ‘Challenger’ Expedition, and the
‘Challenger’ specimen seemed, according to the author of that Report, to agree with
the Mediterranean species. When I now, however, compare the three specimens captured
off Bahia with a specimen (?) from Messina belonging to the Strassburg Museum (vide
Ortmann, in Spengel, Zool. Jahrb., Syst. v. 1890, p. 458), I observe indubitable
differences.

The two females are 40 mm. and 34 mm. long from the tip of the rostrum to the end
of the telson, the younger male measures 832mm. The specimen from Messina is 37 mm.
long. The principal differences are the following :—The rostrum of the Mediterranean
specimen projects almost horizontally forward, exactly as in the ‘ Challenger’ female
(Spence Bate, /. c. fig. 1), but the rostrum of the three American specimens is more
obliquely directed upward, the straight lower margin, indeed, making an angle.of about
30° with the upper border of the carapace. As regards the teeth on the latter and on
the rostrum, the American specimens agree with that from Messina, but the third tooth,

FROM THE INLAND SEA OF JAPAN. 451

which stands at the base of the rostrum, is much smaller than the two preceding on the
upper border of the carapace and than the corresponding tooth in the Mediterranean
specimen. The straight upper border of the rostrum is armed with three teeth of equal
size, of which the third or anterior, placed immediately behind the acute tip, is, in the
larger female a little farther from the second than the second from the first, whereas
in the two other specimens the second is a little farther from the first than from the
third.

The straight lower margin of the rostrum ends in a sharp tooth, and, exactly as in the
specimen from Messina, there are between this tooth and the tip, which is curved down-
ward, nearer to the tip than to the tooth, éwo other pointed teeth which are also curved
downward. According to Milne-Edwards the lower margin should carry only one tooth,
according to Heller one or two, according to Spence Bate also one; Spence Bate is here,
however, inaccurate, for he figures (J. ¢. fig. 1) two teeth below the tip. In the specimen
Jrom Messina, as well as in those that were captured off Bahia, there are six teeth
between the tip of the rostrum and the posterior margin of the carapace and three
teeth below the tip.

The abdomen agrees with that of the specimen from Messina, but the grooves, both
the transverse and the oblique, are in the American specimens much less deep and
shallower.

The third difference which I observe is presented by the first three pairs of legs, which
in the specimens caught off Bahia are a little slenderer.

If the differences described are, indeed, constant, the American species should form a
variety, for which the name americana is proposed.

SICYONIA CARINATA (Olivier).
Sicyonia carinata (Olivier), Spence Bate, Report on the ‘ Challenger’ Macrura, p. 294, pl. 43. figs. 2, 3.

Three young specimens, dredged off Bahia, 2} fathoms.

The largest specimen is 88 mm. long from tip of rostrum to the end of the telson.

These specimens fully agree, especially as regards the toothing of the rostrum, with
the above cited figures of the ‘Challenger’ Report, the upper border of the rostrum
carrying two teeth behind the acute tip and one immediately below it,

SECOND SERIES.—ZOOLOGY, VOL. IX. 64

Fig.

DR. J. G. DE MAN ON CRUSTACEA CHIEFLY

EXPLANATION OF THE PLATES.

The specimens figured are from the Inland Sea of Japan, unless otherwise stated. :

PuateE 31.
1. Lambrus (Oncodolambrus) predator, de Man. x 3.
2. 35 Cephalothorax viewed from in front, x 3. ‘a
Bb 2 fp Lower surface of the anterior part of the cephalothorax, x 10.
4, Asthenognathus inequipes, Stimpson. Inferior view of the cervical region, x 17. (The oblique
position of the figure is accidental.) mae
De 5 Left cheliped of the female, x 17 (the chela is turned a little backward, so that
the full height of the palm is not visible). ti
6. 5 Right leg of the antepenultimate pair, x 8} (all the joints covered with a dark
tomentum except the dactylus).
7. Leucosia rhomboidalis, de Haan. Abdomen ofthe male, x 3.
8. Arcania heptacantha (de Haan). Female, x 2.
u: 5 Abdomen of a young male, x 3.
10. BS Cheliped of the female, x 2.
ll. Arcania globata, Stimpson. x 3.
12. 5 Front and anterior part of the cephalothorax viewed from above, x 10.
13. 3 The same, lateral view, x 10.
14. Galathea acanthomera, Stimpson. External maxilliped of the right side of a male, x 17.
15. 55 Leg of the second pair of the same, x 10.
16. Crangon consobrinus, de Man. Anterior part of carapace and eye-peduncles, x 8.
Ne = Tip of rostrum, x 50.
18. 55 Antennal scale, without the sete, x 8.
19. a Chela, x 8.
PLATE 32.
20. Crangon cassiope, de Man, Antennal scale without the set, x 5.
2l. 5) Extremity of the scale, without the setz, x 25.
22. Bs External maxilliped, x 5.
23. D Chela, x 10.
24. by Left leg of the fifth pair, x 5.
25. 5 Dactylus of the same leg, x 10.
26. Leander longipes, Ortmann. ‘Telson of the egg-bearing female, x 5.
27. 55 Extremity of the telson, x 25.
28. A Right leg of the second pair, x 4.
29, a Toothing of the same leg, x 25.
30. 55 Teeth, more magnified, x 50.
31. Spirontocaris rectirostris (Stimpson). Egg-bearing female, x 3.
32. FF Antennal scale, without the sete, x 6.
33. oS Supposed male, x 3.
34. 55 Antennal scale of the male, x 6, also without the sete.
35. Spirontocaris propugnatriz, de Man. x 3.
36. 5 Extremity of the rostrum, x 6.
37. PA Part of the rostrum, where the teeth of the lower margin begin, x 6.
38. - Anterior part of the carapace, x 10.

FROM THE INLAND SEA OF JAPAN, 453

Fig. 39. Spirontocaris propugnatriz. Posterior half of abdomen, x 6.

Fig.

40.
41.
42.

43.

44,
45.
46.
47.

48.

49.

50.

pe Extremity of telson, x 50.
35 Extremity of left antennal scale, x 10, without the sete.
Spirontocaris alcimede, de Man. x 3. One of the specimens in which a pterygostomian
spinule occurs.

te

r

a Rostrum of another specimen, which is 3 toothed, x 6. (The rostrum of this figure

should point upward.)

55 Posterior half of the abdomen of the same individual, x 6.
. Eye-peduncles and both pairs of antenne of the same specimen, x 6.
a Leg of the second pair of the same specimen, x 10.

Spirontocaris pandaloides, Stimpson. Cephalothorax, rostrum, and antennal scale of an adult
specimen, X 3.

Fr Leg of the second pair of another individual, x 10.
PLATE 33.
Hippolysmata vittata, Stimpson. Cephalothorax, antennule, and antenne of an egg-bearing
female, x 3.
5 Terminal part of the abdomen of the same female, x 38.

51 & 52. Alpheus brevirostris (Olivier). Chelz of the smaller cheliped in the two egg-bearing

53.

54.

55.
56.
57.

58.
59.

60.
61.
62.
63.
64.

65.
66.

67.
68.
69.

females, viewed from the outer side, x 3.

Alpheus japonicus, Miers. Chela and carpus of the smaller cheliped of one of the two
males, viewed from the upper or inner side, x 2.

Peneus (Metapeneus) akayebi, Rathbun. Stridulating-organ of the female, on the right side
of the carapace, x 17.

Peneus (Metapeneus) acclivis, Rathbun. Thelycum, viewed from outer side, x 5.

Peneus (Trachypeneus) curvirostris, Stimpson. Thelycum, x 5.

3 First segment of the abdomen, lateral view, presenting the fissure near the lower
margin, X 3.
rf Telson of the same female, x 5.

Potamon (Parathelphusa) endymion, de Man. Egg-bearing female from the lake at
Yunnan-Fu, China, x 2.

., Anterior part of the cephalothorax, x 3.
an Lower side of the anterior part of the cephalothorax, x 3.
53 External maxillipeds of the same female, x 3.

9 Right (larger) chela of the same female, x 3.
Potamon (Parathelphusa) lanzi, Doflein, typical male specimen from the lower Han River,
China (belonging to the Museum of Munich), external maxilliped, x 3.
as Larger (right) chela of the same male, outer side, x 2.
Palemon (Parapalemon?) hendersoni, de Man, anterior part of the cephalothorax of the largest
specimen from Darjeeling, in which an hepatic spine is wanting, x 3.
rh Telson of the same specimen, x 3.
55 Left leg of the second pair of the same specimen, x 3.
Peneus (Peneus) latisulcatus, Kishinouye, var.? Thelycum, viewed from the outer side, x 6.

4.54, ON CRUSTACEA CHIEFLY FROM THE INLAND SEA OF JAPAN.

ADDITIONAL NOTES.

A.

(Page 418.) According to Lenz’s paper in Spengel’s Zool. Jabrb., Syst. xiv. 1901, p. 429, Bare
Island should be situated between Vancouver Island and the continent ; afterwards, however, Prof. Lenz
informed me that this very small island is situated close to the east coast of the northern island of New
Zealand, between lat. 40° and Cape Kidnappers.

B.

(Page 436.) Penaus tenettus. The thoracic legs seem to be devoid of epipodites, and the exopodite
of the fifth pair is rudimentary or wanting. Parapeneopsis acclivirostris, Alcock, has a longer rostrum,
recurved at the tip, the thelycum different, and the antennular flagella are shorter.

C.

(Page 441.) Poramon spinescens. The fingers of the smaller cheliped are also, in these adult males,
distinctly spoon-shaped, excavated at the tips, but those of the larger leg show a tendency to lose
this spoon-like shape, the fingers appearing obtuse at their tips. Whereas the fingers of the smaller
cheliped still shut nearly close together, those of the larger become gradually more gaping, and in the
largest specimen the dactylus is strongly arched and there is a large interspace between both fingers ;
the excavation of the tips of the fingers has become quite indistinct, though it is still perceptible.
The larger chela of this male is just as long as the cephalothorax is broad; the palm, measured horizon-
tally, appears once and half as long as the fingers, and the height of the palm near the articulation of
the dactylus is equal to the horizontal length of the fingers. Palm and fingers are quite smooth; the
dactylus carries 12 or 13 obtuse teeth of different size, of which the first, near the base, is larger than
the rest; the immobile finger is also armed with some obtuse, unequal teeth.

The chelipeds are yellow, but the upper surface of the carpus, the upper border of the palm and of
the dactylus, as also the upper end of the arm, are of a beautiful red.

[12th February, 1907. ]

XXXT.

Trans. Linn.Soc. Ser.2.Zo00!. Vol IX, P!

DE MAN.

BRO LR

J.T.RENNIE REID.LITH.EDIN?

J.G.DE MAN DEL.
CRUSTACEA FROM JAPAN.

DE Man . TRANS. LINN. Soc. SER.2. Zool. Vol.IX, Pl. XXXII

J.G. OE MAN DEL. J.T.RENNIE REID, LITH. EOIN®

CRUSTACEA FROM JAPAN

"
F :
;
‘
i
~ ¥
; h
: i iy - =
, m i /
; j
i r
‘ y
! : ‘
% 7 ' '
; 1 P
4 Si i
ASS .
, 1
‘ ‘
we
m
ion
j
‘
i
Seek sate f
t
F
an ae oD
‘ } :
5
- 4 ‘
¢ y
; :
, a
_ ‘
i \
; ‘ , ' i
1
.
4
f
‘ Py

DE Man. TRANS LINN. S0¢. SER.2. Z601.Vol.IX, Pl. XXXII]

J.G.DE MAN DEL. JTRENNIE REID, LITH. EDIN®

CRUSTACEA FROM JAPAN AND OTHER COUNTRIES.

: '
:
,
i
cm
;
i
‘
R
\
=
“ ‘
,
¥ ’
: ms
is
:
a
i t M4
p :
1
’
x
A ‘
'
;
¥ !
\
ye
tte fc
5,
: :
,
i
’
= 3
ae * lovee) a ‘
J Ve ite +
ry ¥ s i ‘ pa a

XII. On Cercococeus eremobius, gen. et sp. nov., an Aberrant Form of Coccide.
_By Houew Scorr, B.A. (Cantab.). (Communicated by J. J. Lasvur, W.A., F.R.S., F.L.S.)

(Plate 34.)

Read 18th April, 1907.

THE species of Coccid which I have to describe in this paper was found on a desert-
plant growing on the hill known as Djebel-el-Melah, “ the Mountain of Salt,” in one of
the southern spurs of the great central plateau of Algeria, where the mountains slope
down to the Algerian Sahara, a few miles north of the oasis of Biskra. The hill
consists of beds of rock-salt and gypsum overlaid by Cretaceous strata. At this place
was growing a specimen of a Cistaceous plant, Helianthemum kahiricum, Delile, found
throughout the Algerian Sahara, and having a general distribution from Syria to
Algeria: for this determination I am indebted to Dr. Stapf, of the Herbarium of the
Royal Gardens, Kew. The specimen is a dwarf woody shrub, reaching about four
inches from the ground. It bears on its twigs conspicuous white masses of a substance
somewhat like cotton-wool in appearance. The small oblong-lanceolate leaves of the
plant are, as in so many desert-plants, covered with hairs, which are very minute in this
case. These hairs give a greyish colour to the plant, but are not otherwise conspicuous,
whereas the wool-like masses are decidedly so, contrasting strongly with the leaves
and twigs, as will be seen in Pl. 34, fig. 1, which shows almost the whole plant.
The infected plant was collected on Dec. 25, 1906, by Mr. J. J. Lister, who found that
each of the wool-like masses contains a Coccid. Not having time to investigate the
matter himself, Mr. Lister very kindly handed the material to me for examination.

I owe my best thanks to Mr. Robert Newstead for kindly examining specimens, and
for pointing out the salient features and systematic position of the insect; also to
Dr. David Sharp for much help and advice in dealing with it. With this assistance it
has been possible to make out fairly satisfactorily its characters.

It is necessary to form a new genus of the subfamily Dactylopiine, closely allied to
Asterolecanium *, Signoret. The following is a diagnosis :—

CERCOCOCCUS, gen. nov.

Fem. adult. Corpus prolongatum, cauda distincti munitum. Lobi anales magni,
setigeri. Antenuz minute. Insectum omnino carens pedibus.
Maris puparium nullo modo translucens, extrinsecus filis longis. Mas incognitus.

* Astero-Lecanium (Targ.), Sign. (1868), Ann. Soc. ent. France, 1869, p. 101; Asterolecanium, Targ. 1869 [sec.
Fernald, Cat. Cocc. p. 49], Sign. (1868), Ann. Soc. ent. France, 1870, p. 276. .
SECOND SERIES.—ZOOLOGY, VOL. IX. 65

456 MR. HUGH SCOTT ON AN

CERCOCOCCUS EREMOBIUS, Sp. nov.

Fem. adult. Corpus subrotundum; supra fortiter convexum, ad latera obsolete trans-
versim sulcatum, tribus seriebus longitudinalibus tuberculorum perparum eleva-
torum, foveis numerosissimis fila albida emittentibus, insectum omnino tegentia ;
infra convexum. Cauda brevis, depressa.

Adult female. It will thus be seen that the most striking characters are the prolonga-
tion of the body to form a perfectly distinct tail terminated by large anal lobes, and the
presence of very numerous pits scattered over the dorsal surface, from which emerge
long curling threads, which together form the dense white covering that entirely
conceals the insect and its ovisac. The ovisac, too, is itself a highly remarkable
structure.

The body of the insect, considered apart from the tail, has a very rounded and convex
appearance. Dorsally there is a steep posterior slope down to the tail, which is depressed
and considerably below the general surface. Ventrally, the surface of the tail is
continuous with that of the trunk. No traces of asymmetry, such as are exhibited by
the adult females of some species of Lecanium, are visible. The colour of the specimens
preserved in formalin and afterwards transferred to spirit is orange, lighter on the
ventral surface. When freshly found they also appeared to be orange, though no close
examination was then made. They vary considerably in size; a large specimen
measures about 33 mm. long by 24 mm. broad, while a smaller one was only about
2+ mm. long. This is no doubt due to differences in the degree of maturity.

The transverse depressions at the sides of the body are somewhat vague. Posteriorly,
where the surface of the trunk slopes down to the tail, there are others extending right
across the body. The basal portion of the tail also shows a kind of segmentation, which
in the distal part becomes, dorsally, very obscure. The tubercles (Pl. 34. fig. 4, a, a) of
the median and lateral longitudinal rows are only very slightly elevated. They are
serially arranged with respect to the transverse furrows, one tubercle standing between
each two depressions. They will be mentioned again in connection with the excretory
products of the insect. A description of the pits (Pl. 34. fig. 4) scattered so densely over
the dorsal surface of the trunk is left till the peculiarities of the integument as a whole
are dealt with.

The ventral surface (Pl. 34. fig. 2), with the exception of a few minute setz only
visible under a high power of the microscope, is smooth, lacking the pits so numerous
on the dorsal. At either side it has two well-marked longitudinal furrows (Pl. 34.
fig. 2a, b), into the inner of which open the spiracles. An inconspicuous transverse
furrow leads from the inner to the outer longitudinal furrow, at the point where the
posterior spiracle is placed. Just in front of the anterior spiracle, the inner furrow
converges with and joins the outer. In front of this point of union the outer furrow
curves round towards the middle line, so that it approaches, but does not meet, its
fellow of the opposite side of the body.

In this anterior part of the furrow lies the antenna (PI. 34, figs. 2¢, 3), which consists
of a single short chitinous piece, round in surface view, bearing a short thick spine and
four or more setz.

ABERRANT FORM OF COCCID. 457

The surface of the trunk within the inner furrows is strongly convex, and at the
centre of it is the rostrum, sunk in a depression. The loop of the rostral filaments is
long. There is no trace of legs or feet, or of eyes.

The spiracles (PI. 34. fig. 2, d), as previously mentioned, open, two on either side of the
body, into the inner longitudinal furrows. Like the antennze, they are more strongly
chitinised and darker in colour than the surrounding integument. Each consists of a
shallow circular chamber, in the external floor of which is the orifice to the exterior,
while an opening in the internal floor leads into a tracheal trunk. In connection with
each spiracle is a curiously shaped structure, strongly chitinised, and very conspicuous
in specimens that have been emptied of their soft parts by treatment with caustic
potash. This structure is adjacent to the side of the stigmatic chamber remote from
the lateral margin of the body, and extends forwards and slightly towards the middle
line. In transverse sections it appears as a ridge projecting from the cuticle into the
interior of the body, and muscles extending from the dorsal surface are attached to it at
one point. On the side of the stigmatic chamber nearer the lateral margin of the body
is a group of cutaneous glands, which will be spoken of later.

The ventral surface of the posterior part of the trunk and of the tail shows very
distinct division into eight segments (Pl. 34. fig. 2). The anal orifice is surrounded by
eight sete (Pl. 34, fig. 9) rising from a slender chitinous ring, which has a beaded
appearance. Just anterior to this anal ring rise a large and a small pair of sete
in close proximity to each other (Pl. 34. fig. 9), and the integument of the terminal
segment in this region bears a number of very minute projections (Pl. 34. fig. 9, @).
Between the anal lobes is a shorter median lobe (Pl. 34. fig. 9, 2), dorsal to the anus.
Each anal lobe (Pl. 34. fig. 9, ¢) is large; it bears a long seta at its extremity, and five
smaller ones (the arrangement of which is shown in the figure) on its more proximal
portions. Every seta rises from a chitinised base, having the form of a cup with a
raised and thickened rim. The genital aperture is a somewhat transverse opening in the
ventral middle line, in the region of the furrow between the sixth and seventh segments ;
the integument immediately surrounding it shows extremely fine strive radiating from it
(Pl. 34. fig. 9, d).

Passing now to a consideration of the integument and the excretory products of the
insect, the most striking feature of all is the very numerous threads secreted by cutaneous
glands on the dorsal surface. These form a dense covering completely concealing not
only the body of the female, but also the ovisac in which it lies, and produce the
conspicuous wool-like masses on the infected plant. The covering of threads as a whole
looks opaque and white: but when highly magnified each thread is seen to be perfectly
transparent, colourless, and glass-like; cylindrical, curving, and sometimes with a
certain amount of longitudinal striation in the interior of its substance (Pl. 34. fig. 7).
The threads vary in thickness; the diameter of cross-section of a rather thick one
measured was about 12. These threads are insoluble in cold wax-dissolving reagents.
But Dr. Hopkins, to whom I am much indebted for examining them, states that they
are formed of a wax freely soluble in hot reagents. It dissolves in boiling absolute
alcohol, and separates on cooling into glistening plates of homogeneous appearance.

65*

458 MR. HUGH SCOTT ON AN

These crystals, which melt at 92° C., seem to represent the entire substance, which is
practically a pure chemical compound’: a mere trace remains in the alcohol from’
which they separated. The wax is clearly a different kind to that called coccerine,
described by C. Liebermann (Berichte der Deutschen chem. Gesell., Bd. xvii. p. 1975),
which melted at 106°.

These threads are produced by unicellular glands placed in pairs beneath pits in the
integument. Each pit (Pl. 34. figs. 4, 10 a, 11, 12, 17) is formed by an invagination of
the cuticle, but its walls are thinner than the cuticle covering the surface between the
pits (Pl. 84. fig.17). The area of the cross-section of the pit is about the same throughout.
On its floor there are two pore-plates lying close together, so that they appear somewhat
like a figure of eight (Pl. 34. fig. 10). The term “ pore-plate” is not meant to imply
the presence of any perforations in the plate, for I have seen no trace of any such.
It is used to indicate a specialised portion of the integument lying over a gland, and
through which the secretion of that gland passes to the exterior*. If there are no
perforations in these plates, the very interesting physiological question arises as to the
form in which the secretion is produced by the glandular cells, and the manner in which
it traverses the chitinous membrane between it and the exterior. Berlese (J. c.) considers
that in all insects the chitinous integument is uninterrupted by any perforations, so
that dermal secretions of all kinds must pass through an extremely thin membrane.’
Each pore-plate has a broad rim, sloping slightly downwards and inwards, like the
surface of a funnel, and more strongly chitinised than the surrounding cuticle; moreover,
the rim is transversely striated, and under a high magnification a dark dot is often
apparent in the centre of each of the strice. The strize do not appear in sections in any
way as pores perforating the rim. The outline of each pore-plate is in the form of
an oval flattened at one side, the flattened sides of the two ovals being adjacent, but not
quite contiguous. The outer margin of the flattened side of the rim has at its central
point a small concavity. The two concavities, being opposite to one another, leave a
slightly widened space (Pl. 34. fig. 10,0), in which there rise, between the two pore-
plates, two minute papillee (Pl. 34. figs. 1l a, 12,170). ‘These appear to be evaginations
of the thin cuticle between the rims; they are in a line at right angles to the long
diameter of the pair of pore-plates, and hence when the pair is seen—as it very frequently
is—from the side, the two minute papillae appear as a single one (Pl. 34, fig. 11, a).
However, under a high power, and in some transverse sections, the two can be seen,
quite distinct from each other. The space within the rim of each pore-plate is closed
by a membrane having an indistinctly dotted or mottled appearance.

These pore-plates are almost invariably in pairs, as described‘above. In a single case
there were seen three of them together, forming a roughly triangular figure.. They vary
considerably in size, larger and smaller pairs being interspersed over the dorsal surface
of the trunk ; but the two pore-plates of each individual pair are of equal size. They are
exceedingly numerous over the whole dorsal surface of the trunk, where they seem to
have no definite orientation, their long diameters lying in all directions. They are
-extremely conspicuous in preparations of specimens that have been treated with cane

* A, Berlese, ‘Gli Insetti,’ vol. i. p. 492 and footnote.

ABERRANT FORM OF COCCID. 459

potash. On the tail they are very scantily distributed, and only the smaller ones are
present. On the ventral surface of the insect they are almost absent ; there are a few
small pairs arranged more or less definitely on each of the segments of the posterior
region, and a small pair on each anal lobe.

Under each pore-plate is a single large glandular cell (Pl. 34. fig. 17, a), about 48 p
across in a direction parallel to the surface of the body. ‘The two cells under each pair
of pore-plates lie touching one another, and their contiguous sides are flattened. . In
transverse sections each pair may thus appear to bea bicellular gland. But each cell
has its own pore-plate and secretes its own thread, for two separate threads emerge from

each of the pits in the integument (PI. 34. fig. 14); therefore each pair of cells must be |

looked upon as consisting of two unicellular glands in close proximity to one another.

The substance of each of these glandular cells is divided into two portions. Imme- ~

diately under the pore-plate is a small well-defined part, much clearer and less deeply
staining than the rest. The remainder of the cell, which contains a large oval nucleus,
is much less clear and stains fairly deeply; this part has, in sections stained with
hematoxylin and orange G, the peculiarity, which it shares with the chitinous cuticle,
of taking up the latter stain, whereas its nucleus and the adjacent hypodermal and
subhypodermal tissues take up the hematoxylin. The arrangement of these pits and
glands with respect to the said hypodermis is worth noting. Beneath the hypodermis is
a loose layer of subhypodermal cells (PI. 34. fig. 17, e). The hypodermal layer is inter-
rupted by the invagination of the chitinous cuticle to form the walls of the pit, and
beneath the floor of the pit is represented only by the two glandular cells. On the other
hand, the subhypodermal layer, at the point where it abuts against the walls of the pit,
is invaginated so that it forms a loose capsule ensheathing the pair of glandular cells
(Pl. 34. fig. 17, /).

On the dorsal surface of the tail, beyond the area of the pairs of pore-plates described
above, and in the region of the fourth segment, there are two groups of curious structures,
one on either side of the middle line. They are pits in the integument, and when seen
in side view are like hollow inverted cones with obtuse and rounded apices (PI. 34.
fig. 10,d). They vary in shape, size, and arrangement. They are usually, but not
always, in pairs; their number varies in different individuals, and is not even always the
same in the two groups borne by one individual ; usually there are five or six pits in each
group. They are more strongly chitinised than the surrounding cuticle. The upper
part of their walls looks homogeneous, while the lower or apical portion has a eribriform
or sieve-like appearance. However, there do not seem to be any perforations as in a true
sieve-plate, but merely thinner and more lightly-staining areas lying in the meshes of a
network of thicker substance. It has not been possible to determine whether there are
any glands in connection with these remarkable structures, but their appearance suggests
that they may be essentially the same as the pore-plates already described, though they
differ from them in detail.

There. is another widely-distributed kind of cutaneous gland, characterised by the
possession of a long, narrow, chitinous duct opening on a level with the general surface
of the cuticle (Pl. 34. fig.15,a). The duct is a narrow tube perpendicular to the surface

4.60 MR. HUGH SCOTT ON AN

of the body, and having a straight or almost straight course for some distance. The first
part of this straight tube has delicate walls (Pl. 34, fig. 11.0), often shrivelled in prepared
specimens. The remainder has slightly thicker walls (fig. 11, ¢). At the end of the straight
tube there is a very curious sharp bend (PI. 34. figs. 11d, 16); just on the outer side of
the bend the walls of the duct are very slightly invaginated. Beyond the bend the
walls of the duct are thinner again, as in the first portion. Immediately beyond the
bend the duct is very narrow; it then broadens, and terminates in a slightly swollen
end, the cavity of which is increased by numerous minute rounded diverticula, giving to
the end the appearance of a morula (PI. 34, figs.11e, 15, 16). The portion of the duct
from the region of the bend to its termination is buried in the interior of a large
elandular cell, measuring about 34 across. The substance of this cell consists of a
well-defined inner clear portion, which stains very faintly, end an outer granular portion,
which stains fairly deeply. The granular part is usually very thin, except in one
region remote from the point of entry of the duct and containing the nucleus (Pl. 34,
fig. 16,a). The inner clear part sometimes shows refractive globules (Pl. 34. fig. 16),
more frequently striz radiating from the termination of the duct (Pl. 34. fig. 15, ).
How far these appearances are artificial it is impossible to say, but it seems probable
that the strive really represent a fibrillar structure of the protoplasm, such as has been
described and figured by Prof. A. Berlese * as existing in the interior or excretory
portion of many glandular cells in insects. There is a mass of protoplasm appearing to
be of the same consistency as the granular portion of the gland, and containing several
rather small nuclei, surrounding the duct at its point of entry into the gland-cell (Pl. 34,
fig. 15,¢). This mass, though in close contact with the gland-cell, is clearly marked off
from it; but does not itself, in the sections examined, always show definite division into
several sinall cells, though indications of such division are present. I have endeavoured
without success to determine whether these small cells are accessory glandular cells
pouring a secretion into the duct at its bend, as might be suspected from the curious
conformation of the latter; but there is no evidence that such is the case. The small
cells show no special glandular structure and no division whatever into two portions, as
do the large unicellular glands. ‘Traces of a loose capsule of subhypodermal tissue can
sometimes be seen surrounding the glands and their ducts.

The glands of this kind are numerous over the whole dorsal surface in the
spaces between the pits of the thread-producing glands, their ducts being shown in
fig. 10, f (Pl. 34). They are also numerous over areas where the thread-producing
glands are almost entirely absent, that is, on the dorsal surface of the tail and
the ventral surface of the trunk. They are more sparingly distributed on the ventral
surface of the tail. No excretory products have been seen in connection with their
orifices, so that it is not possible to state what part they play in the life-history of
the insect.

A third kind of cutaneous glands must be noticed. They are much more localised in
their distribution than the preceding, and are confined to the ventral surface. Their
pore-plates lie in seven transverse bands in the posterior region of the body, near’the

* Gli Insetti, vol. i. p. 498 & fig. 559.

ABERRANT FORM OF COCCID. 461

hind margins of the segments. The most anterior band is ill-defined, and its pore-
plates few. The next four bands are well-defined and continuous, consisting of two or
three irregular rows of pore-plates. The next band (PI. 34. fig. 9, e)—at the posterior
margin of the sixth segment—resembles these, but is interrupted in the middle line by
the genital aperture (Pl. 34. fig. 9,d). The band on the seventh segment merely
consists of a small group of pore-plates on either side of the middle line; a few pore-
plates are also present on the terminal segment. On the more anterior part of the
body these pore-plates are only present in small groups just external to each spiracle
and antenna.

These pore-plates have a distinct form of their own. As usual, they are more strongly
chitinised than the surrounding cuticle. They are not arranged in pairs. Each (PI. 34.
fig. 13, also fig. 9) is circular, with a broad rim, immediately within which is a circle of
dots. The space within these stains more deeply than the rest, but has a rather indistinct
central dot of more lightly-staining substance. The glands in connection with these
plates lie immediately below the chitinous derma. They are somewhat elongated,
narrower in the portion nearer the derma, and much smaller than any previously
described ; those in the posterior region of the body measure about 18 u in a direction
perpendicular to the body-surface. I cannot be quite certain whether those in the
posterior region consist of one or more cells. In connection with each spiracular group
of pore-plates there can be seen a number of closely-packed, elongated, somewhat pear-
shaped cells, each with a distinct nucleus.

Judging from analogy with allied forms, which fill their stigmatic grooves with wax *,
the position of some of the glands of the third kind with respect to the spiracles might
suggest that they may have this function. Moreover, on the inner surface of some of
the ovisacs can be seen four patches of white amorphous substance, corresponding
roughly with the positions where the spiracles would lie when the female was in the
ovisac ; but, though extremely probable, it is not certain that the amorphous substance
is wax. In connection with this, it may be mentioned that Berlese +, in describing
certain species of Lecaniuwm, figures plates which are like the pore-plates described
above, and which are found in the stigmatic grooves and belong to glands that secrete
wax into these grooves.

Ovisac (Pl. 34. fig. 5).—The white mass of threads arising from the dorsal surface
conceals not only the insect, but also the curious ovisac in which it lies. The ovisacs
are fixed to the twigs, for in the material examined the insects are always attached to
the woody stems, and not to the small leaves, of the plant. In the great majority of
specimens the ovisac has the form of a widely-open cup or basket; opaque, brownish-
yellow, with smooth inner surface, and the outer surface rough and bearing a number of
white threads similar to those arising from the dorsal surface of theinsect. It is closely
adapted to the shape of the creature’s body, and there is a deep impression in the margin
at one point, and sometimes a slight spout-like prolongation, in which the tail of the

* A. Berlese, ‘ Le Cocciniglie Italiane,’ pt. ii. pp. 132, 133 & 182, 183 [ex Riv. Pat. Veg. vol. iii. No. 1-8] ;
Newstead, ‘ Monograph of British Coccide,’ vol. i. p. 15,

+ ‘Le Cocciniglie Italiane,’ pt. ii. tay. 5. fig. 2a,

4.G2 MR. HUGH SCOTT ON AN

insect lies extended. Under the microscope there can be seen in the walls a structure
consisting, not of interlacing threads, but of branches anastomosing in all directions
(Pl. 34. fig. 6, «). They are colourless, transparent, and glassy in appearance, and very
minute, their thickness being less than half that of the white threads covering the insect’s
body. The skeleton that they form is best seen in the margin of the cup; in the rest of
the walls it is covered with opaque material (PI. 34. fig. 6, 0) of brownish and yellowish
colour, with felted masses of threads, and with some of the larger white threads so
characteristic of the insect. nm the denser portions the substance of the ovisac has to
some extent a radiate arrangement; irregular thicker parts run from the .base towards
the margin of the cup, and alternate with thinner and more translucent portions. The
ovisac, like the white threads, will not dissolve in cold chloroform, xylol, or ether. The
mode of its formation cannot be made out in the preserved material.

Such is the structure in the great majority of specimens. But in a very few of the
dried specimens the ovisac has proved to be completely closed except for an opening on the
somewhat spout-like prolongation corresponding to the tail of the insect. As mentioned
above, in the open cups there is a depression in the wall to accommodate the tail, and in
one case this depression was just arched over, so that it formed a round hole (Pl. 34,
fig. 5, @). This is probably an intermediate stage between the open depression and the
closed prolongation containing the tail. It appears as if, at a later stage in the life-
history than that attained by most of these specimens, more secretory material is added
to the open cup, so that the latter becomes a closed structure, completely shutting in the
female, as isthe case in the allied genus Asterolecanium*. One of these closed structures
was empty, the other contained a much shrivelled female. This latter one bore on its
‘outer surface a number of the white threads, but they formed a mass much smaller than
that of the threads covering the females seated in open cups.

In several of the specimens preserved in formalin, each of the tubercles of the dorsal
longitudinal! rows bears a small plate or flake of glass-like secretion, insoluble in cold wax-
dissolving reagents. ach of these flakes is attached at its centrai part to the surface of
the tubercle, and has in its peripheral portions exact impressions of the pairs of pore-
plates of the circumjacent integument. Through some of these impressions rise the
threads secreted from the pore-plates beneath. The whole appearance suggests that the
secretion has been poured out from glands on the tubercle, has flowed in all directions
from this central point, surrounding the bases of the threads emerging from the pits, and
has hardened into the transparent flake, receiving in the process impressions of the pits
and pore-plates of the integument. It is possible that, by further excretion, the separate
flakes on the tubercles may grow in extent till they unite and form a single complete
covering to the dorsal surface, continuous at the sides with the margins of the cup-like
ovisac, thereby transforming the latter into a closed structure. But there is no proof
that such is the case.

Male Puparium (PI. 34. fig. 8)—The puparia are attached to the woody stems of the
Helianthemum. Yach is about 13 mm. long, much smaller than the ovisac, elongate-

* Newstead, ‘ Monograph of British Coccidie,’ vol. ii, p. 150.

ABERRANT FORM OF COCCIDA:. 463

ovate, narrowed posteriorly, quite opaque, white or very pale greenish, thus differing in
colour from the ovisac. The inner surface of the walls is smooth, the outer surface
rough. Externally there are a number of white threads similar to those secreted by the
female; they are more numerous near the anterior end. Their presence in this structure
formed by the male is interesting. The microscopic composition of the puparium is
much like that of the felted parts of the ovisac, but there is no disposition into more
opaque and more translucent portions. The puparia are empty, each having a neat,
transverse slit at the posterior end, where the male has emerged (Pl. 34, fig. 8, @).

The mycelium of an Ascomycete fungus is found ramifying in the walls of some of the
puparia and ovisacs, and outside the bases of some of the latter. The dark masses of
spores formed in the mycelium appear as sooty-looking specks. Within one of the closed
ovisacs mentioned above were a number of dried and empty perithecia of the fungus.
Mr. R. H. Biffen, who has kindly examined the fungus, states that, although he has found
no ascospores, the form of the perithecia and spore-masses in the mycelium are strongly
conclusive of its being a species of Capnodium. Newstead * states that the honey-dew
secreted by British Coccids is almost invariably attacked, shortly after deposition, by a
fungus of the genus Meliola. This is allied to Capnodium. Berlese + also speaks of
fungi habitually accompanying Coccids on plants. It is therefore probable that the
fungus in question nourishes itself on certain excretory products of the Algerian
Coccid.

It should be mentioned that a single dried-up female specimen was found entangled
in the mass of threads belonging to another individual, and that it contained within its
shrivelled integument a number of oblong-ovoid bodies, also in a desiccated condition.
The specimen was one of those preserved in formalin, and must therefore have been dead
and desiccated at the time when the infected plant was found. Its condition unfor-
tunately makes it impossible to ascertain the nature of the ovoid bodies which it
contains.

Cercococcus eremobius was found in the desert, on a plant which is essentially a
desert-plant. The greatest peculiarity about the insect is the thick covering of white
threads, corresponding to which is the large development of cutaneous glands. An
analogy is suggested between this character and that of those numerous desert-plants
which are clothed with hairs, which serve them in good stead by preventing excessive
loss of moisture ; and it is possible that the covering of threads may benefit the insect in
the same manner that the hairs benefit the plants. It may be emphasised again that
they can hardly serve any purpose of protective similarity to surroundings, since they
form conspicuous white masses on the twigs. Since the locality where the creature was
found is by no means inaccessible, it is to be hoped that before long fresh supplies of
material will be obtained, including both sexes and all stages, so that the whole life-
history of this interesting insect can be elucidated.

* «Monograph of British Coccide,’ Ray Soe. vol. i. p. 19.
+ A. Berlese, ‘Le Cocciniglie Italiane,’ pt. i. p. 43 [ex Riv. Pat. Veg. vol. ii. No. 1-8}.

SECOND SERIES.—ZOOLOGY, VOL. IX. 66

464

Tone 1
Fig. 2.
Fig. 3.
Fig. 4
Fig. 5
Fig. 6
lv
Tig. 8
Fig. 9
Fig. 10.
Fig. 11.
Fig. 12
Fig. 18
Fig. 14
Fig. 15
Fig. 16.
Fig. 17.

ON AN ABERRANT FORM OF COCCIDA.

EXPLANATION OF PLATE 34.

. The specimen of Helianthemum kahiricum bearing Cercococcus eremobius. Nearly two-thirds

natural size.

Ventral view of adult female. , outer, 6, inner longitudinal furrow of the right side;
c, antenna; d, anterior spiracle.

Antenna in side view, showing the spine and two of the hairs.

. Dorsal view of adult female. a, a, tubercles of the median and lateral longitudinal rows. The

dots represent the numerous pits of the thread-producing glands.

. Ovisac on a twig. a, hole in which the tail of the insect lay.
. Portion of the wall of the ovisac, highly magnified. a@, anastomosing branches; 6, opaque

material.

. Part of one of the threads secreted from the dorsal surface, showing the longitudinally striated

appearance. To same scale as fig. 6.

. Male puparium on a twig. a, slit at posterior end.
. Ventral view of extremity of tail of adult female, showing the anal ring and orifice. a, minute

papille on the integument; 6, median lobe, dorsal to the anus; c, one of the anal lobes;
d, genital aperture ; e, transverse band of circular pore-plates on the sixth segment.

Portion of the dorsal integument of a specimen treated with caustic potash, highly magnified.
In the upper part are some of the numerous pairs of pore-plates. a, outline of floor of one
of the pits ; 4, space left between the two concavities in the margins of the rims; ¢, com-
mencement of the tail-region, where the pore-plate pairs are almost completely absent. In
the lower part of the figure are some of the pits with sieve-like walls; d, pit in side view;
e, one in surface view; jf, narrow ducts of glands of the second kind.

Two of the dorsal pits in side view, each with a pair of pore-plates at the bottom; a, papillve
between the pore-plates. In the middle, a duct of a gland of the second kind; 4, thin-walled
portion of straight tube ; ¢, its thicker-walled portion; d, bend in the duct; e, terminal

swelling.

. One of the dorsal pits and pairs of pore-plates, from a transverse section torn in such a way

as to show the pair of papillee between the pore-plates.

. A single circular pore-plate from one of the transverse ventral bands. To same scale as fig. 10.
. Part of the dorsal surface of a dried specimen, showing some of the pits, each with two threads

emerging from it; the greater part of the threads cut away.

. One of the glands of the second kind. a, aperture of duct on outer surface of body; 4, inner

secretory portion of gland-cell showing radial striz; c, accessory cells. _
Part of one of the same glands, more highly magnified. a, nucleus of glandular cell.
Transverse section through one of the pairs of thread-producing cells. The section only passes
through the nucleus of one gland. a, one of the cells; 6, papilla between the two pore-
plates ; c, cuticle of the general body-surface ; d, hypodermis ; e, subhypodermal layer, and
f, the sheath which it forms round the glands.

TRANS. LINN. Soc., Srr.2. Zoou. Vol. IX. Pu.34.

. d E Wilson, Cambridge
_ ' CERCOCOCCUS EREMOBIUS, sp.nov.
a
yy > vs ni -
oe rc. (

:
Pars
~*

XII. Observations on Australasian Polyclads.
By Professor W. A. HasweEtt, J7.A4., D.Se., F.RS., PLS.

(Plates 35-37.)

Read 6th June, 1907.

THs Polyelads of Australasia have hitherto received very little attention. Stimpson’s
* Descriptions of some new Marine Invertebrates” (22), published in 1855, contains
diagnoses of several members of the group obtained in Port Jackson. Schmarda in the
first volume of his ‘ Neue wirbellose Thiere,’ issued in 1859, published some observations
on several species from New South Wales and New Zealand. In Saville Kent's ‘ Great
Barrier Reef’ three species are referred to as Pseudoceros Kentii, n. sp., von Graff,
Pseudoceros dimidiatus, nu. sp., von Graff, and Prosthecereus flavomaculatus, n. sp., von
Graff; these were apparently named by von Graff, but I am not aware that any account of
them has been published. Woodworth (26) described a few species from the same locality.
Marianne Plehn (18) described a species from New Zealand and the Chatham Islands,
and another from Tasmania. T. W. Kirk (10) described two New Zealand species, and
T. F. Cheeseman (2) two more. Lastly, Laidlaw (11) described Leptoplana australis and
Stylochus vigilax from specimens in the collection of the British Museum, and recognized
a specimen of Cryptocelides Loveni, Bergendal, labelled ‘“ Port Phillip, J. B. Wilson.”

Of Stimpson’s descriptions it is impossible to make any use, and the same holds good
of Schmarda’s. Though, presumably, the forms from Port Jackson described by these
authors are the same as some of those dealt with in the following pages, it is quite
impossible, without the opportunity of examining the original specimens, to attain to any
certainty in this direction. Thus it is quite possible that Stimpson’s Déoncus badius was
the species here referred to as Leptoplana australis, Laidlaw, and at first I was disposed
to name it Leptoplana badia in order to retain the old name; but on reconsideration 1
came to the conclusion that it would be better to avoid any identifications of such a merely
conjectural kind, and to set Stimpson’s and Schmarda’s names aside altogether.

Lang’s comment on Stimpson’s paper may be quoted here :—** Die Diagnosen sind alle
sehr kiimmierlich. Bei dem giinzlichen Fehlen der Abbildungen werden deshalb die Arten
wohl kaum wieder mit Sicherheit zu erkennen sein. Ihre generische Zugehorigkeit
ist in den meisten Fallen nicht zu errathen” (17, p. 17). Of Schmarda’s descriptions the
same author observes :—‘* Leider sind die anatomischen Beobachtungen ausserst kiimmer-
lich und die Angaben iiber Fehlen oder Vorhandensein und Stellung der Augen, Lage
und Natur der Oeffnungen des Kérpers, Form des Pharynx etc. wohl nicht ganz
. Zuverlissig ; so dass vielen der beschriebenen Arten ihre Stellung im System niclit
mit Sicherheit angewiesen werden kann ” (op. cit. p. 19).

I have pleasure in acknowledging assistance received from Mr. R. Etheridge, Curator
SECOND SERIES.—ZOOLOGY, VOL. IX. 67

466 PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS,

of the Australian Museum, Sydney, and Prof. W. B. Benham, of Dunedin, N.Z., by both
of whom I was given the opportunity of examining specimens from the collections under
their charge, to Mr. Chas. Hedley for specimens collected at Masthead Island and
Cooktown, and to Mr. 8. J. Johnston, B.A., B.Sc., Senior Demonstrator of Biology,
Sydney University, for the specimens of Diplosolenia.

The most important morphological and physiological results embodied in the following
ave :—(1) the discovery of a new Planocerid (named Tr/py/locelis) with three reproductive
apertures; (2) the discovery of a Cryptocelis-like form with a genito-intestinal canal ;
(3) the evidence of a peculiar mode of copulation by localized perforation in the new
Australian genus Echinoplana.

The nomenclature of the parts of the female reproductive apparatus of the Polyclads
is somewhat confused, and, without any intention of dogmatizing as to homologies, it is
necessary that I should state here the terms used in the following descriptions, and the
sense in which they are employed. The term ovaries needs no comment. The oviducts
are the ducts by which the ova reach the uteri. The wfer/are the two elongated chambers
in which the fully-developed ova collect, and in which they undergo maturation and may
become fertilized. 'The ducts by which the ova pass out from the uteri are the wterine
ducts; very commonly the right and left uterine ducts unite to form a median uterine
duct. ‘The remainder of the apparatus consists of a median passage to the whole of which
I apply the term vagina. The part of this into which the female aperture directly leads
is the antrum femininum. This, or a part of it, may have its walls thickened to form a
bursa copulatrix. The part following on this is the ootype ; this is the region into which
the ducts of the shell-glands open. The ootype usually runs upwards and forwards, or
directly upwards, and bends sharply back to pass into the dorsal limb of the vagina,
the shell-gland ducts being frequently continued on this for some distance, and even on
the terminal portions of the uterine ducts. Into the dorsal limb of the vagina open the
uterine ducts, or median duct, as the case may be. Beyond this point the vagina may be
prolonged backwards. Sometimes it terminates in a median sac, the receptaculum
seminis ; rarely there are two receptacula, right and left. In a few cases the vagina
terminates behind by opening on the ventral surface of the body by a posterior female
eperture.

TRIPYLOCELIS TYPICA, n. g.,n. sp. (Plate 35.)

In the living condition this Polyclad is about 1:5 to 2 cm. in length, and in breadth
about 0°75 to 1 em.—the breadth not being more than half the length. The brain,
tentacles, and eyes are in the first fifth. The mouth is in front of the middle. The
male aperture is in the last third. Tie space between the male and female apertures
is about one-half of that between the latter and the posterior margin.

~The tentacles are in the form of elongated cones; they are not retractile into de-
pressions at their bases. The arrangement of the eyes (Pl. 85. fig. 2) is fairly constant.
Each tentacular group comprises some twelve to twenty. ‘Two or three small eyes are
usually to be detected in the tentacles above the level of the cthers. Of the remainder
there are rarely any situated directly over the brain; but they all, or nearly all, lie in

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS, 467

front of or behind it, those in front being the more numerous (about 50), arranged in
two parallel groups with a small space between them.

The colouring varies somewhat, but is never very pronounced. Some specimens are
almost colourless, but for the light green ramifications of the intestinal exca; but in
most there is a faint diffused brown tint due to the presence of minute dots of brown
pigment, which are most numerous in the region behind the pharynx. The ventral
surface is usually of a ight brown colour except in the more central parts, where the
pharynx and principal parts of the genital apparatus appear white. One of the most
striking features in the aspect of the living animal is the conspicuousness of the
intestine and its main branches owing to their presenting the appearance of narrow, dark
greenish lines on the dorsal aspect of the animal.

The mouth leads into the pharynx by a funnel-like passage. The pharynx has about
eight to ten pairs of lateral folds and the intestine gives off about the same number of
main diverticula. Pharynx and intestine are almost conterminous.

The two vasa deferentia (Pl. 35. fig. 1) join at about the point of union of the anterior
and middle thirds of the penis. The median duct thus formed runs forwards with an
almost straight course for some distance ; it is slightly dilated, and its wall is somewhat
thickened, so that it may be regarded as forming a median vesicula seminalis. Further
forwards it becomes narrower, and is thrown into a number of coils, in the ordinary,
retracted condition of the penis. Eventually, when it reaches a point a little distance
in front of the anterior end of the penis, it bends sharply round, and runs almost straight
back through the axis of the granule-gland papilla, where the ducts of the granule-
elands open into it (granule-gland reservoir).

The penis (figs. 1 & 3) is an elongated muscular cylinder, without spines or other
special chitinous developments, straight in a well-extended specimen, but more or less
bent in a specimen contracted in the direction of the long axis. In front it is quite
circular in cross-section ; further back it is more or less compressed. Its walls consist
of outer longitudinal and inner circular layers of muscular fibres of approximately equal
thickness, and of a layer of columnar epithelium bounding the lumen. Surrounding it
is a thick mass of retiform parenchyma.

Projecting backwards into the lumen of the penis from its anterior end is the conical
papilla (granule-gland papilla) perforated by the terminal part of the ejaculatory duct,
having the ducts of the granule-glands opening into it. This papilla is formed of an
involution of the muscular wall of the penis filled with the retiform tissue that
surrounds it.

The chief (anterior) female aperture leads into an ootype (fig. 1) of long-oval form
with greatly plicated walls. At its anterior end this bends back and passes into the
dorsal limb of the vagina. The latter runs backwards near the dorsal surface of the
body, and receives from below the unpaired ducts formed by the union of the right and
left uterine ducts. Instead of terminating blindly or expanding into a receptaculum
seminis, as in most other Polyclads, the vagina then bends downwards and opens on the
ventral surface some little distance behind the main female aperture (PI. 35, figs. 1, 5, & 6).

This posterior continuation of the oviduct has a thick: muscular wall; its epithelium is
67*

168 PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

vaised up into a number of longitudinal ridges. Behind the point where the median
uterine duct leaves it below a process of epithelium projects into the lumen; this may,
perhaps, act as a valve for preventing the passage of the eggs backwards to the posterior
female aperture.

The uteri are wide tubes containing, in the sexually active animal, numbers of ripe
eggs, together with quantities of spermatozoa. The eggs are all at the same stage. In
each is the usual centrally-placed spindle with four chromosomes.

This appears to be the fourth Polyclad genus in which a second female aperture has
been discovered. But in other respects the four forms have very little in common.
Lang’s genus Zrigonoporus *, of which two American species—TZ. foliwm and T. den-
driticus t—have been described hy Verrill, has no tentacles, has numerous scattered
cephalic eyes, and has a separate prostate with independent duct. Of the affinities of
Bergendalia, Laidlaw, little is known, but in that genus the second aperture leads into
the antrum femininum. Jaidlawia, Herzig, has the aperture in question on the dorsal
instead of the ventral side, and there is a well-developed receptaculum seminis. In
Polyporus, Plehn, there is a close correspondence in the relations of this aperture with
what we find in Tripydocelis, The single specimen of Polyporus found was not sexually
mature, so that little was ascertainable regarding the reproductive apparatus ; but there
are pores all round leading into the intestinal branches, and there are no eyes.

If we leave the occurrence of the third reproductive aperture out of account the closest
relationships of Tripylocelis are with the Planoceride. ‘The members of Lang’s group B
of the species which he referred to the genus Planocera are, apparently, the nearest
relatives. But the relationship is by no means very close, the differences in the male
reproductive apparatus being very considerable.

Tripylocelis typica is perhaps the commonest Polyclad that occurs between tidal limits
in Port Jackson. It is chiefly to be found in tidal pools among the thalli of Ulva and
can usually be obtained in considerable numbers by pulling the Algz to pieces and
shaking them out in a vessel of water. .

It is an extremely active form, swims vigorously, and on the surface of a solid object
is able to progress rapidly by advancing lateral lobes, which are able to adhere to the
surface, and are pushed forward from right and left sides alternately—the result being
a kind of “walking,” as distinguished from “‘ creeping ” or “ looping” locomotion.

The following is a definition of the genus Zripylocelis :—

Planoceridee with fairly broad, oval, leaf-shaped body, with conical non-retractile
tentacles. Brain-tentacles and eyes relatively further forward than in Planocera—in
the first fifth or thereabouts. Two groups of tentacular eyes, and smaller eyes just in
front of and behind the brain; no marginal eyes. ‘Three genital apertures. Male
aperture a considerable distance behind the pharyngeal sac. Principal female aperture
not far behind the male: second female aperture ou the ventral surface not far behind
the first. Penis muscular, without sheath and without chitinous parts. A small median

HT] Moz.
7 As pointed out by Laidlaw (14), the generic position of these is doubtful.

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS. 469

vesicula. Prostate-gland reservoir situated in the course of the ejaculatory duct. No
bursa copulatrix. Vagina continued backwards to open on the ventral surface by the
second female aperture.

The discovery by Herzig (9) of a Polyclad (named by him Zaédlawia) which has a
second female aperture situated not on the ventral but on the dorsal side, may be
regarded as of some importance, as it may help us towards a determination of the homo-
logies of the parts. A correspondence with the Trematode “ canal of Laurer” and with
the Cestode ‘‘ vagina” and receptaculum seminis obviously suggests itself. Bergendal
has recently described a Triclad with a second female aperture, and I have shown that
such an opening (dorsally situated) occurs in the dAcewla, so that an arrangement of this
kind would seem to be very widespread in the Platodes. In all such cases the object of
the arrangement is, most probably, to enable fresh supplies of spermatozoa to be taken
in without any interference with the passage outwards of the fertilized ova. But it may
be that in some cases the canal has lost its function, though still persisting, or may have
become adapted to other uses.

DIPLOSOLENIA JOHNSTONI, n. g., n. sp. (Plate 36, figs. 1 & 2.)

This is a very large Polyclad, one preserved specimen measuring 6 cm. in length and
3 em. in breadth. The mouth is somewhat in front of the middle—in the third
centimetre. The reproductive apertures are situated close together (or combined in one)
at the junction of the fourth and fifth centimetres. The tentacles are rather long and
of conical shape: they are placed a little behind the junction of the first and second
centimetres. The eyes are in two very compact groups, each comprising about 380, at
the bases of the tentacles, and two longitudinally extended less compact groups between
them, each of these also comprising about 30. The dorsal surface is almost black, with
a narrow light margin.

The pharynx has about twelve pairs of lateral folds.

In two of the specimens, which have the penis exserted and a considerable length of
the stylet projecting, the male and female apertures are separated by a distinct interval.
On the other hand, in the specimens with the penis retracted the parts of the body-wall
in the neighbourhood of the apertures become involuted to form a kind of common
antrum having but a single external opening.

The ventral part of the vagina is a long narrow passage the lumen of which has a
triangular cross-section in tie greater part of its extent. Its muscular investment is
thin, and it has an epithelium of long narrow cells. In the posterior part of its extent,
in the specimen sectioned, the lumen is filled entirely with the shell-gland secretion
which completely saturates the epithelium. Further forwards numerous spermatozoa
are. also present ; but in the anterior part of the passage these disappear and the lumen
is filled with the shell-gland secretion.

The vagina becomes bent back sharply on itself, the dorsal limb running back to a
point just over the base of the (exserted) stylet, where it bifurcates to form the ducts
of the two lateral receptacula seminis which run almost transversely outwards (fig. 2).
This dorsal part of the vagina is a narrow cylindrical tube, with an epithelium of

170 PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

comparatively short and broad cells and a thick muscular wall. Not far behind its
origin it receives on the ventral side a median duct formed by the union of the right
and left uterine ducts. In the lumen of these there are many spermatozoa.

The two receptacula seminis (fig. 1) are of great size, and when distended with sperms
become very conspicuous structures. Ina mounted specimen 30 mm. in length one of
them is 4mm. long. In the distended condition the stretched wall is very thin and its
structure is difficult of determination. But in the collapsed state the wall appears
relatively thick, and is found to have essentially the same structure as that of the
unpaired receptacle of Leptoplana australis. It has a thick muscular investment, and an
epithelium of large cells, each of which has a prominent rounded process at its inner
end, having the appearance of an exuding droplet of secretion.

Each lateral vas deferens is dilated to form an extensive elongated seminal vesicle, but
this does not appear specially thickened—at least in the distended condition it is in in the
specimen sectioned. From the two lateral vesiculee a pair of narrow ducts run inwards
and at the base of the penis unite to form the ejaculatory duct.

The penis is enclosed within an elongated sheath. The penis itself consists of a
very thin-walled chitinous tube enclosed in a thick layer of circularly-arranged fibres.
The tube or stylet is 5 or 6 mm. in length, tapering to a sharp point at its free end,
slightly dilated and funnel-shaped proximally. ‘The layer of circular fibres is continued,
somewhat reduced in thickness, over the portion of the penis which is protruded in the
specimen sectioned : its function must be to bring about peristaltic contractions of the
thin-walled chitinous tube and so of enclosed ducts. Within the tube is a layer of
longitudinal muscular fibres: internally the lumen is occupied by a core of parenchyma
in which run ejaculatory and prostatic ducts, the former towards the centre, the latter
towards the dorsal side.

The prostate is a median structure which extends forwards as far as a point a little in
front of the point of union of the oviducts. Its duct, at first narrow, widens, then
becomes narrower and sinuous, then expands into a channel devoid of epithelial lining,
which acts as a reservoir. This runs back for some distance parallel with the ventral
limb of the vagina and on its ventral side. This sinus or reservoir passes behind into a
narrow cylindrical duct with well-defined walls, which runs to the base of the penis and
traverses that organ throughout its length, running within the hollow stylet, parallel
with and dorsal to the ejaculatory duct, but remaining separate from the latter.

The following are the chief features which distinguish this genus :—There is a pair of
nuchal tentacles with groups of tentacular eyes: no marginal eyes. Reproductive
apertures closely approximated. Vagina long and narrow throughout, without bursa
copulatrix. A pair of large receptacula seminis. Duct of prostate separate from
ejaculatory duct throughout its entire length. A pair of vesiculze seminales in the
form of large dilatations of the right and left vasa deferentia. An elongated
penial stylet.

So far as I am aware, only two Polyclads are known to possess paired receptacula,
viz. Discocelis tigrina, Lang (“horseshoe-shaped gland”), and Leptoplana subviridis,
Plehn (Laidlaw). This character, together with the complete separation of prostatic

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS. 471.

and ejaculatory ducts, would suffice to distinguish Diplosolenia from all other
Planoceride.

Tdioplana, Woodworth (26), resembles Diplosolenia in the exceptional feature of the
complete separation of prostatic and ejaculatory ducts; but it has marginal eyes, has,
apparently, no penial stylet, and has a median vesicula.

Planocera has the ejaculatory and prostatic ducts uniting to form a common duct; its
penis is lined with chitinous spines, and the vagina gives rise to a bursa copulatrix: the
receptaculum seminis is unpaired.

Stylochoplana, again, has the prostate intercalated in the course of the ejaculatory
duct, has a median vesicula, a single receptaculum, and possesses a bursa copulatrix ;
while Stylochus has marginal eyes, the reproductive apertures situated close to the
posterior end, has the prostate separate, but with its duct uniting with the ejaculatory
duct, a single vesicula, and no receptaculum.

Paraplanocera, though it has paired vesicule, resembles Planocera in the character
of the penis and in the unpaired receptaculum ; it also has a muscular diverticulum of
the vagina of the nature of a bursa copulatrix.

LEPTOPLANA AUSTRALIS, Laidlaw. (Plate 36. figs. 3-5.)

This is, so far as my data go, the most widely distributed, as well as one of the largest,
of the Australasian Polyclads. It is one of the commonest of the Port Jackson species,
and was obtained also at Jervis Bay, on the southern part of the New South Wales coast.
It is by far the commonest species on the coast of Tasmania, and it extends also to
New Zealand.

Leptoplana australis may be identical with Dioncus badius of Stimpson (23), or with
D. oblengus of the same author, both of which were found in Port Jackson, and it may
also be the same as Polycelis australis of Schmarda (20), which was found on the New
South Wales coast. But the characters given by the authors named are not of a nature
to justify even generic determination. Thus Stimpson’s definition of the genus Dioncus
runs :—“ Corpus planum, dilatatum. Caput corpore continuum. Os subcentrale.
Ocelli numerosi, in umbones duos claros subdistantes dispositi. Maricole.” ‘The
description which he gives of D. badius is as follows :—‘ Body half as broad as long, of
a reddish-brown colour above, with a flake-white dust intermixed. Anteriorly there are
two prominent circular knobs, which contain, scattered over the entire surface, the very
numerous and minute eyes. Below the body is of a pale sepia colour, except the white
digestive organs, and the mouth is placed behind the centre. Length 1:5; breadth
0°75 inch.” D. oblongus is stated to ditfer from D. badius mainly in having a clear space
around the eyes on each knob.

Schmarda’s Polycelis australis may be this species, but the characters given and the
figure would not warrant an identification. The following is the description :—

“Der Kérper ist platt, linglich, vorn abgerundet und riickwirts kaum weniger
verschmichtigt. Die Farbe des Riickens ist dunkelbraun mit unterbrochener blasser
Mittellinie. Die Bauchseite ist réthlichbraun. Lange 30 mm., Breite 13 mm. Die
Augen stehen in zwei Gruppen am ende des ersten Sechstels, sie sind einander sehr

472 PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

eenihert und besitzen einen weissen Hof der sich nach vorn und auswirts in einen
kurzen Streifen fortsetzt welcher mit dem der anderen Seite divergirt. Die
Mundéffnung ist zwar auch hier wie bei der vorigen Species central; die Geschlechts-
ffaungen sind jedoch dem Centrum sehr viel naher geriickt und liegen im zweiten
Drittel des Korpers. Das Parenchym ist dicker und stirker, als es bei andern Polycelis
der Fall ist.”

Schmarda found his specimens on the coast of the [lawarra district, New South
Waies, and in Auckland Harbour, New Zealand.

Leptoplana australis reaches a large size; the largest specimens I have seen alive were
3 inches long. It is subject to great variation in colour. The larger specimens are
very dark—some almost black, with a lighter line round the edge. Smaller specimens
are much lighter—some with only a light general shade of brown, through which run the
branching mottled bands of olive-green that represent the ramifications of the intestine.

Quite conspicuous features of the upper surface are the two clear colourless rounded
knobs on which the ‘ tentacular” groups of eyes are borne. These are situated at about
the junction of the first fourth of the length of the body with the second. They are of
oval or elliptical outline, with the long axis directed forwards and outwards. Each of
them comprises about forty eyes. Smaller eyes are arranged in two groups separated by
a definite mesial space; they are more numerous than the tentacular eyes, and are almost
entirely in front of them.

The ventral surface is grey, the reproductive apparatus (or rather certain portions
of it) forming a more or less pronounced white pattern on it. The mouth is always a
little in front of the middle; the male reproductive aperture is about halfway between
the mouth and the posterior border, and the female aperture a short distance behind
the male.

The pharynx gives off about fifteen to seventeen pairs of diverticula, and the number
of main intestinal branches is about the same.

The vesicula seminalis is at the junction of the lateral vasa deferentia and the
ejaculatory duct. From this the ejaculatory duct runs forwards to the prostate
reservoir. The latter is a thick-walled, Jong, oval body, the appearance of which in
the entire specimen differs a good deal according to the way in which it lies. In an
end view it presents a remarkable wheel-like appearance which is not recognizable in
a lateral view. In sections this wheel-like effect is found to be due to the presence of a
ving of longitudinal recesses which open into the lumen at the end nearest the base of
the penis after traversing the wall of the organ throughout its length.

From the end of the prostate reservoir opposite that at which it enters the ejaculatory
duct runs forwards to the base of the penis. The latter is of great length. It encloses
a narrow, twice-curved, hollow, chitinous stylet, dilated into a funnel proximally, and
distally tapering to a fine point. In the two largest specimens it is sharply bent near
the apex.

The female aperture leads into the antrum femininum, whch runs upwards and
backwards as a wide passage to open into the ootype or shell-gland reservoir. The
latter is remarkably developed, much wider than the antrum, with a minutely folded

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS. 473

inner surface : it is produced backwards some distance behind the female aperture. Its
wall is very thick and muscular, as is that of the antrum, which may be regarded as
assuming the character of a bursa copulatrix, In front it narrows somewhat; at its
anterior end it bends sharply back as it passes into the narrow receptacular duct or
vagina. The right and left uterine ducts run almost transversely inwards from the
corresponding uteri, and unite to form a short unpaired duct which enters the vagina
just over the male aperture. Behind this junction the backward prolongation of the
vagina runs as a narrow tube on the left side of the ootype to open immediately behind
the posterior extremity of the latter into a very large sac (receptaculum seminis). The
anterior part of the duct is slightly constricted at regular intervals—the constrictions
producing a headed appearance. In some specimens this beaded appearance extends
throughout its length. The receptaculum itself is a sac with folded walls lined by a
large-celled columnar epithelium. Sometimes it appears collapsed and empty or nearly
so: more frequently it contains a great mass of spermatozoa together with granules or
droplets of a secretion evidently derived from the columnar cells. In some specimens
sperms occur throughout the length of the duct.

The entire reproductive system of ZL. australis has a very close resemblance to that
of L. fallax, Diesing, as described by Quatrefages (21). he chief differences appear
to be that in the latter species the penial stylet is coiled on itself, the vagina is sinuous,
and the accessory sac is unsymmetrically developed*. JZ. alcinoi and ZL. vitiea, as
figured and described by Lang (17), resemble Z. australis in the peculiar internal
structure of the prostate reservoir, but differ from it in other respects—notably in the
relatively slight development of the receptaculum seminis.

There is a considerable difference between individuals of Z. australis, when fixed, as
regards the length of the posterior prolongation of the vagina (duct of the receptaculum
seminis) and the size of the receptaculum itself. But it seems most probable that this
is due to differences in the condition of the parts and the degree of contraction which
they have undergone.

LL. australis occurs at a comparatively high level between tide-marks, and is to be
found by turning over stones.

In sections of one of the Port Jackson specimens I was interested to find in the
pharynx unmistakable fragments of an Enteropneust. This was the more remarkable
since no Enteropneust has ever been recorded as occurring in Port Jackson.

In the intestine of a Tasmanian specimen was the lingual ribbon of a Gastropod.

What may be a dwarf variety of this species is commen in Lyttleton Harbour, N.Z.
Preserved specimens are under 1 cm. in length. In the living condition it is
transparent, with some brown pigment on the dorsal surface, and is of very delicate
consistency, so that it is very difficult to obtain entire specimens. The eyes are much
fewer in number than in mature specimens of the ordinary Z. australis, but in
this respect there is a correspondence with immature specimens of that form. In

* In Quatrefages’s figure the lateral uterine ducts are represented as opening separately, and, moreover, as
opening, not into the vagina, but into the ootype, which is obviously an error.

SECOND SERIES.—ZOOLOGY, VOL. IX. 68

47 4: PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

small specimens of the latter, however, the reproductive apparatus is undeveloped,
whereas the small forms now under consideration are sexually mature. In the
reproductive system the main differences may be reduced to differences in proportions :
the penial stylet, granule reservoir, vesicula seminalis, antrum femininum, and recep-
taculum seminis have all the same general character as in ZL. australis. Perhaps the
most important difference is in the great relative width of the dorsal limb of the vagina
(or duct of the seminal receptacle).

From Lyttleton Harbour, from Waiheke, Auckland Harbour, and from Kaikoura,
I have specimens of Leptoplana differing littie from the Australian specimens which I
have referred to LZ. australis. The tentacular groups of eves are denser owing to the
larger size of the individual eyes, but in other respects there is a close correspondence.
In one of the specimens (from Kaikoura) the penial stylet, instead of tapering to a
fine point, ends in a truncated extremity provided with a circlet of hook-like processes ;
and in another (from Auckland Harbour) it is nearly straight and relatively short. But
such differences are probably merely individual variations.

Of the identity of this common Australian species with Laidlaw’s Teptga
australis (11) I have no great doubt, though the description given is not very full.
The arrangement of the eyes agrees fairly well; and the reference to the “ long nearly
straight stylet” of the penis, to the prostate divided into some six or seven longitudinal
chambers, as well as the allusion to Z. alcinot as an allied species, all seem to point
to this determination. The colour given, dark chocolate-brown, is unusual.

The British Museum specimens described by Laidlaw were collected in Port Phillip
by Dr. R. Lendenfeld.

It is a somewhat remarkable fact that Plehn (18) records the occurrence in French
Pass (northern New Zealand), and also in the Chatham Islands, of a species of
Leptoplana which corresponds in many respects with L. australis, but which has only
one genital aperture, like that author’s LZ. californica (19), and is not regarded by her
as distinct from the latter species. In the hope of finding something corresponding to
this, I have looked over my Australian, Tasmanian, and New Zealand specimens set
down as L. australis; but they all have the two separate apertures, and I have as yet
seen nothing corresponding with Plehn’s species.

MICROCELIS SCHAUINSLANDI, Plehn (18).

A. solitary specimen which I obtained at St. Helens, on the east coast of Tasmania,
resembles Plehn’s (18) species (also from Tasmania) in such points as are capable of
being made out. It has the same general very characteristic arrangement of the eyes
and posterior position of the pharynx, but the specimen was damaged and little can
be made of the reproductive apparatus. It was observed to be, like Cryptocelis, an
exceedingly sluggish form of unusually firm consistency. Its colour on the upper
surface was brown, very distinctly mottled.

Resembling the above in the posterior position of the pharynx, the marginal eyes,
the two separate but closely approximated reproductive apertures, and the presence of
a median receptaculum seiminis, is a New Zealand Polyclad which I have found under

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS. 475

stones in Lyttleton Harbour. But in this the numerous minute eyes over the brain-
region are (imperfectly) divided into two by a narrow space, and though they extend
forwards nearly tothe anterior margin, it is only as a relatively narrow band. Moreover,
the marginal eyes only extend over less than a half of the margin. The female
aperture leads vertically upwards to the ootype, which runs forwards a short distance
and narrows as it bends backwards. Into the dorsal limb on the ventral aspect the
uterine ducts open at a point nearly directly over the female aperture. Posteriorly
the vagina is preduced and terminates in a large «nd complicated receptaculum seminis.
The male apparatus was not distinguishable in the entire mounted specimen, and
sections are not available at present.

ECHINOPLANA CELERRIMA, 0. g., n. sp. (Plate 36. figs. 6 & 7; Plate 37. figs. 1-8.)

This is a rather small Polyclad, averaging about 1°5 cm. in length and 6 mm. in
breadth at the broadest part (towards the anterior end). The colour varies somewhat
and is never very pronounced. Usually the dorsal surface has a reddish-brown tint.
There are no tentacles. The eyes (Plate 37. fig. 1) are arranged in two somewhat
elongated groups, one on either side of the brain, each including about thirty. The
brain and the eyes are in the first fifth of the length of the body. The mouth is
distinctly behind the middie. The male aperture is about halfway between that and
the posterior end—at about the junction of the third fourth with the last. The female
aperture is a considerable distance behind the male, nearer the posterior end than to
the latter. In front of it is a peculiar, transversely corrugated area of the integument.
In front of the male aperture in the living specimen the penis is usually plainly
recognizable as a narrow elongated brown object.

The pharynx has twelve to fifteen pairs of lateral folds, and the number of pairs of
intestinal cxeca is about the same. The main intestine extends some distance iu front
of the anterior extremity of the pharynx.

The lateral vasa deferentia (Pl. 37. fig. 2) open into an elongated median vesicula
seminalis, which terminates in front in a very fine duct (duct of vesicula). This
traverses from before backwards a conical papilla projecting backwards into the lumea
of the granule reservoir from its anterior extremity and opens into the latter,

The granule reservoir is of great length: it has the form of a tube with muscular
walls, wider at its proximal or anterior end than at its distal or posterior, with three
sharp bends in its course. In its posterior part it presents about half a dozen slight
regular constrictions. Its muscular layers are of considerable thickness and _ its
epithelium is thrown into a series of longitudinal folds.

The ejaculatory duct, narrow and coiled where it leaves the granule reservoir, widens
posteriorly as it traverses the penis. The anterior narrow part has a very definite
cylindrical epithelium surrounded by a condensed layer of the muscular fibres of the
penis; many granule-gland ducts traverse the muscular liyers and perforate the cells
of the epithelium to open into the lumen. Posteriorly the duct soon loses its epithelium
and becomes beset with the horny teeth described below.

The penis consists of an enormously thick mass of muscular fibres occupying

the greater part of the vertical diameter of the body and about a tenth part of the
68*

176 PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

transverse. In the living animal, and to a still more marked degree in preserved
specimens, its position is indicated by a rounded elevation on the dorsal surface. There
is no enclosing sheath or sac, the more peripheral muscular fibres passing into the
muscular layers of the body-wall or into the layers of parenchyma-muscle that surround
the various neighbouring organs (intestinal branches, vasa deferentia, uteri). The
muscular fibres are some longitudinal in direction, some circular, some radial, but they
are not arranged in any definite layers or zones. At the external aperture the muscular
mass is quite continuous with the muscular layers of the body-wall. The entire lumen
of the penis (distal part of ejaculatory duct) is lined with horny spines or teeth.
These (Pl. 36. figs. 6 & 7) are largest in the neighbourhood of the external opening,
evadually decreasing in size anteriorly. The larger spines are slightly curved, pointed,
and have a shape comparable to that of the claw of a bird. In the smaller spines the
base is relatively more expanded than in the larger and the distal part more abruptly
curved. In the neighbourhood of the opening a process from the underlying tissue
projects into the cavity of the spine: further forwards this is not recognizable.

The female aperture leads into a narrow passage surrounded by a thick mass of
parenchyma with numerous muscular fibres. Through this mass run numerous shell-
gland ducts, and these perforate the epithelium of the passage in all parts except the
part immediately adjacent to the aperture, so that an antrum as distinct from an ootype
or shell-gland reservoir can hardly be said to exist. This part of the ootype gives off
laterally a number of very short and small diverticula, which have a fairly regular
arrangement. When it approaches near the dorsal surface of the body it expands in the
interior of a rounded prominence which projects dorsally in this region a little in front
of the female aperture. As it passes forwards it becomes narrower and gives off short
irregular diverticula. When it reaches the muscular mass of the penis projecting behind
the male aperture it passes to the left, and is continued forwards as a narrow diverticulum
(Pl. 37. fig. 3) for some distance beyond the male reproductive aperture. Not far from
its anterior extremity it receives on its dorsal side the common uterine duct. This
unsymmetrical anterior prolongation of the vagina has not a specially developed mus-
cular layer, so that it cannot be looked upon as a bursa copulatrix. On the other hand,
the ducts of the shell-glands open into it in much greater abundance than into the
central part of the ootype itself, and it is best looked upon asa prolongation of the latter.
The reflected portion, or dorsal limb of the vagina, produced backwards in most Polyclads
beyond the point at which it receives the uterine duct, and frequently leading to a
receptaculum seminis, is here entirely absent.

Clear evidence of the mode of action of the copulatory parts of the reproductive
apparatus is afforded by two of my series of sections. In one, a transverse series, there
is traceable a long narrow object running obliquely, on one side only, through the thick
mass of tissue referred to above as surrounding the vagina, the upper end lying near the
lumen of the latter. Traced downwards this body is found to run to the ventral surface,
where it terminates by perforating the epidermis of the corrugated area in front of the
female aperture, projecting slightly on the exterior. In front, between this body and
the lumen of the vagina, the tissue is unusually open and spongy, and in the interstices

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS, A477

are numerous spermatozoa, entirely absent in other parts. Moreover, in the adjoining
part of the lumen, and in that part alone, there are numerous spermatozoa. The long
narrow body is found, when examined under a high power, to consist of a strand of
globules of prostate secretion, or something indistinguishable from it in appearance,
mixed with spermatozoa. Its entire length is roughly about 0°6 mm.

There can be little doubt that we have here to do with a wound inflicted by the
formidably armed penis. The copulating individuals were applied together by their
ventral surfaces, the corrugated areas acting like the suckers of the Cotylea, as organs
of adhesion, when the penis of the one was driven in through the mass of tissue
surrounding the lumen of the vagina, nearly penetrating as far as the latter. The
spermatozoa and prostate secretion were then discharged and the penis withdrawn, a
plug of prostate secretion closing up the wound and thus preventing the escape of the
spermatozoa. The passage of the spermatozoa to the interior of the lumen is facilitated
by the fact that in the middle region of the vagina there are very few shell-gland ducts
passing inwards and perforating the epithelium.

In another specimen, cut into a series of longitudinal vertical sections, the same thing
appears. Here there is a large mass of spermatozoa in the substance of the wall of the
vagina, and this is connected with the ventral surface in front of the corrugated area
by a narrow cleft filled with prostate secretion mixed with spermatozoa, the plug of
this material projecting freely on the surface. But in this case the perforation has
actually passed through the epithelium of the vagina, and in this position a portion of
the mass of spermatozoa projects freely into the lumen.

Such a mode of copulation as this—by perforation of the body-wall ina definite locality
—has not been proved to occur in other Polyclads, and is certainly exceptional in that
class. Perhaps it may be found to occur in the case of other forms with chitinous penial
parts and a thick-walled bursa copulatrix. I have found a similar type of copulation to
characterize Prorhynchus (7) and Stratiodrilus (8).

The ova in the uterus are all in the stage with a centrally placed spindle and, usually,
a spermatozoon (rarely more than one) in various phases of transition in the cytoplasm.

Echinoplana is apparently more nearly allied to Leptoplana than to any described
genus. But it differs in several very important points from the members of that genus.
The entire structure of the male copulatory apparatus is widely different from what we
find in Leptoplana or in any related form. The same holds good of the corresponding
parts of the female reproductive apparatus. The complete absence of a reflected or
dorsal limb of the vagina is a very special feature, and the massive vagina with its
unsymmetrically placed anterior diverticulum is as characteristic, in its way, as the penis
with its array of teeth.

Paraplanocera, Laidlaw (15), has a similar diverticulum of the vagina, or, more strictly,
has a bursa which is in the form of a muscular diverticulum of the vagina; but it has
no other points of resemblance to the form now under consideration, though the penis
has small chitinous spines. Paraplanocera has tentacles, an independent prostate,
paired vesicule, and a receptaculum seminis.

Echinoplana may be defined as a Leptoplanid without tentacles or marginaleyes. Two

ATS PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

elongated groups of tentacular eyes. Mouth behind the middle of the ventral surface.
Separate male and female reproductive apertures. A median vesicula, between which
and the penis is an elongated prostate reservoir. A very large penis without sheath,
lined internally with numerous chitinous teeth. Ventral limb of vagina (antrum and
ootype) with very thick walls. No dorsal limb present. Single unsymmetrical diverti-
culum projecting forwards from vagina. No receptaculum seminis.

Echinoplana celerrima is one of the commonest of the Polyclads of Port Jackson. It
is characterized by great alertness and activity. In addition to the ordinary swimming
and creeping movements, it progresses like Tipylocelis, but much more actively, by a
kind of “ walking.” Lateral lobes of the extremely mobile body assume the function of
lateral appendages. It is interesting to note that a precisely similar mode of locomotion
was observed by Lang in Planocera Graffii, of which he writes—‘ Wenn Planocera
Graffii abwechselnd rechts und links Partien des vorderen Kérperrandes vorstreckt und
dann den Kérper nachzieht so sieht es beinahe aus wie wenn sie sich derselben als Fisse
bediente ”’ (17, p. 635).

ENTEROGONIA PIGRANS, n. g., n. sp. (Plate 37. fig. 4.)

. This Polyclad is of oval or elliptical outline, 1 em. in length in the preserved condition
and 5 or 6 mm. in breadth. It is a thickish form, of comparatively firm consistency,
remarkable for its extremely sluggish movements. The general colour is greenish or
dark grey on the dorsal side; when the living animal is examined with a simple lens,
this becomes resolved into innumerable spots of dark olive, very minute towards the
margin, larger towards the middle. The ventral surface is reddish grey, except where
the pharynx and main testicular ducts show white. In two of the specimens there is a
dark spot towards the posterior end—the appearance being produced by the intestinal
branches here being of a peculiarly dark colour. This does not appear to be constant ;
but when it does occur it probably is in some way associated with the occurrence of the
genito-intestinal passage referred to below.

The mouth is considerably behind the middle of the body, and, in the fixed specimen,
the reproductive apertures are situated very close together, and are nearer to the posterior
edge than to the mouth. There are numerous scattered minute eyes over the brain-
region, and between the latter and the anterior margin, as well as marginal eyes running
all round the periphery. The eyes over the brain-region are quite irregularly distributed,
and not in any way bilaterally grouped, a feature which would in itself distinguish the
present species from Lang’s Cestoplana alba and C. compacta (17, p. 472).

The male aperture leads into a nearly vertical antrum, the epithelium of which is
thickened and raised into ridges. Here are situated the unicellular glands corresponding
to the prostate glands. Into the antrum projects the penis in the form of a short
muscular papilla entirely devoid of chitinous parts. The ejaculatory duct, formed by
the unicn of the lateral vasa deferentia, is a sinuous tube which presents no appearance
of becoming thickened or dilated to form a vesicula seminalis.

The antrum femininum is a vertical chamber with a fairly thick muscular wall. The
ootype curves forwards and upwards from the antrum and bends sharply downwards and

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS. 479

backwards to form the dorsal limb of the vagina. The ootype is characterized by the
development of a peculiar spiral ridge of its epithelium. ‘The dorsal limb of the vagina,
after receiving on its ventral side the common duct formed by the union of the lateral
uterine ducts, rans backwards as a narrowing tube, which opens behind into the median
posterior branch of the intestine—a genito-intestinal passage being thus established.

The absence of distinct prostate glands, other than the glandular cells in the wall of
the antrum, and other features connect this form with Discocelis. But in that genus
there is a pair of lateral receptacula, and there is a common genital atrium. In Laidlaws
recently-created genus Thalamoplana (16) there are distinct male and female apertures ;
but there is a concentric receptaculum seminis, and the prostatic cells in the epithelium
of the antrum are raised on muscular ridges. icrocelis, Plehn (18), is also allied, but
has a large receptaculum. The occurrence of the genito-intestinal canal is of such
importance that it seems desirable to distinguish the Australian form from the members
of these allied genera.

The discovery of the genito-intestinal canal helps to connect more definitely the
receptaculum seminis of Polyclads with parts that occur in other Platodes. The corre-
spondence of the canal in question with the similarly-named canal in the Heterocotylea
cannot well be doubted; while the homology between the latter and the Laurer’s canal
of the Malacocotylea, though it may be open to question, seems to have the balance of
evidence in its favour*. If we accept this conclusion, we must regard as representing
Laurer’s canal in the Polyclads not only the genito-intestinal canal of Lnterogonia, but
the receptaculum seminis of the Acotylea in general (unpaired in most, paired in
Discocelis tigrina, Leptoplana subviridis, and Diplosolenia, with an opening on the
dorsal surface in Laidlawia), and the posterior female passage of Tiigonoporus and
Tripylocelis.

CESTOPLANA AUSTRALIS, n. sp. (Plate 37. fig. 5.)

I have only obtained a single specimen of a species of Cestoplana, which, superficially
at least, is very like the European species C. rubrocincta, Grube. It is a long and
narrow Polyclad, which, as Lang remarks, might readily be taken for a Nemertean ;
its length was 2 cm., its breadth 8 mm. ‘The upper surface is of a light neutral tint in
front, becoming reddish orange further back. Close to each Jateral border runs a band
of the most vivid vermilion, and a median band of the same runs along the whole
length. In front the Jateral bands bend inwards and unite with one anether some little
distance from the anterior extremity. Posteriorly the two lateral bands unite just in
front of the slight notch or depression at the posterior end, but the median band
terminates a short distance in front of this. The narrow space between the latera
band and the lateral border is almost colourless. There are very many very minute
eyes scattered over the anterior portion, with the exception of a zone round the margin.

The only external difference which I can detect between the Australian and European
species is that in the former the three longitudinal bands completely fuse, whereas in

* See Goto, 5, p. lot.

480 PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

the latter they do not. The specimen was immature and the reproductive system
undeveloped.
This somewhat aberrant Polyclad creeps, but never swims. As Lang remarks of the
European form, the anterior portion begins to move while the posterior is still at rest.
Found between tide-limits in Port Jackson (Woollahra Point).

PsEUDOCEROS (2) CARDINALIS, n. sp. (Plate 37. fig. 6.)

The length of the preserved specimens is 1 cm., the breadth 8 mm. The colour of
the upper surface in the living animal was bright scarlet.

The tentacles are very inconspicuous, being mere blunt lobes at the sides of the
anterior median notch. ‘The central group of eyes numbers about 150 altogether.
It is obscurely divided into two behind by a very narrow space, but. is undivided
in front. The tentacular eyes are difficult to count, owing to their being very closely
aggregated anteriorly ; but there appear to be about 100 on each tentacle, distributed
equally on the dorsal and ventral surfaces. The mouth is situated just below the brain.
‘The male reproductive aperture is at the beginning of the second sixth of the entire
length, and is only a short distance behind the month. The female aperture is about
one-sixth of the length behind this. he sucker is situated about the middle of the
length of the body; it has the form of a disk elevated above the general level of
the ventral surface.

The wall of the bell-shaped pharynx is devoid of the foldings characteristic of other
species of Psewdoceros.

‘he male apparatus is single. There is a conical penis, which contaius a chitinous
stylet ; there is a pear-shaped granule reservoir, and a large long-oval vesicula seminalis
thiice the length of the granule reservoir.

Two specimens were obtained together on an oar-weed brought up by the trawl in
Tron Cove River, Port Jackson.

Of the two specimens obtained, one was mounted entire, the other was cut into
sections. The latter was found to contain ripe ova in the uteri, but the testes were
immature and contained no spermatozoa, and the vesicula seminalis contained only a
eranular mass. There were no spermatozoa in the uteri; but in the parenchyma, near
the dorsal surface, directly over the ootype, was a large mass of mature spermatozoa
which must have been received by perforation of the penis of another individual, and
there was a similar mass somewhat further forwards *.

The nature of the pharynx distinguishes this from the described species of Pseudo-
ceros, with which in other respects it is nearly allied. The generic position of this and
also of the following species cannot be looked upon as definitely settled.

PsEUDOCEROS (?) LIMBATUS, 0. sp.

The length of the preserved specimen is 1:5 cm., the breadth 0°5 cm. I am indebted
to Mr. Alan McCulloch for a coloured sketch of the living animal, in which the upper
surface is light red with a well-defined marginal band of purple.

* These contained spermatozoa of two distinct kinds.

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS. 481

The tentacles are comparatively large, subtriangular, and to judge from Mr. Mc-
Culloch’s sketch, must, in the living animal, have extended well beyond the notch in
which they are situated. There are about 30-40 eyes in each tentacle and about 40
on each side above. The cerebral eyes are concentrated into a single dense clump.
The mouth is situated just behind the brain. Both of the reproductive apertures are
situated well in advance of the middle of the body, the male at about the junction of
the second and third sevenths, the female close behind it.

The sucker is somewhat in front of the middle. It is not an elevated disk as in
Pseudoceros cardinalis, but a shallow circular pit with a radial arrangement of its
muscular fibres. The pharynx is small and situated anteriorly, but is strongly plicated.

The penis, which is directed backwards, has a tubular chitinous stylet. There is a
small, oval, prostate reservoir and a large prostate. The male apparatus is situated
partly below, but mainly behind the pharynx. The female apparatus appears to be
quite simple.

But for the character of the pharynx, this form might perhaps be included in the
genus Prosthecereus; but a plicated pharynx does not appear to occur in any of
the Luryleptide, and, in spite of certain points of divergence from the other species,
1 think it best to refer it, for the present, to the genus Pseudoceros.

The only specimen was obtained by Mr. Chas. Hedley, F.L.S., on a reef at Masthead
Island, Capricorn Group, Queensland.

PROSTHECEREZUS ANOMALUS, N. sp.

The tentacles are of moderate size in the preserved specimen and in the form of
flattened cones. The cerebral eyes are arranged in two closely approximated groups
over the brain—about 30 in each group. There are about 100 tentacular eyes. The

mouth, which is very small, is situated immediately behind the brain. The male
aperture is just behind the pharynx; the female aperture is a little distance behind
the male. The sucker is situated at about the middle of the length,

SECOND SERIES. —ZOOLOGY, VOL. IX. 69

482 PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

The pharynx is cylindrical, about a sixth of the entire length. The intestine is fairly
wide, nearly half the length, with numerous pairs of czeca, the branches of which
anastomose.

The penis (text-fig., pe.) contains a very slender elongated chitinous stylet (st.). The
prostate (pr.) is relatively long and narrow.

The ootype is nearly vertical in position. Continuing it backwards is the wide
vagina; the latter bifurcates behind into two thick tubes, which pass almost transversely
outwards, each becoming divided into two. Each of these opens into a wide thin-
walled sac, and there are thus four of these—the receptacula seminis. From the vagina,
on its ventral side, is given off the median uterine duct, which soon bifurcates.

The female reproductive apparatus of this species thus appears to differ in a radical
way from that of the species of the genus described or figured by Lang. ‘The presence
of the four receptacula connected by their ducts with the vagina is, in fact, so far as I
am aware, a quite unique condition. In the other Co¢ylea in which they have been
found to occur the “accessory glands” are connected with the oviducts (see Lang,
pp. 297-800, pl. 28. figs. 1 & 3, pl. 24. fig. 1, &e.).

I have only one example of this interesting form—an old stained and mounted
specimen obtained in Port Jackson.

PROSTHIOSTOMUM MACULATUM, n. sp. (Plate 87, fig. 7.)

The largest specimen is about 2 cm. in length and 7 mm. in breadth in the preserved
condition. Both anterior and posterior ends are rounded. The general colour of the
dorsal surface is light brown with a few large darker spots. The “cerebral” eyes
are disposed in two elongated imperfectly separated groups completely united in front,
each comprising about 50 in a mature specimen. There are about 100 marginal or
submarginal eyes in front of these. The mouth is situated immediately behind the
cerebral eyes at the junction of the first and second sevenths of the length. The sucker
is a little behind the middle of the body; the reproductive apertures, a little in front of
this, are nearly in the middle.

The sucker is a pit with a narrow opening. The integument lining it has its epidermis
ereatly thickened, and is thrown into a number of radiating folds around a longitudinal
slit bordered with a number of minute papille.

The cylindrical pharynx is about 5 mm. in length (about a fourth of the total length)
and 1:5 mm. in diameter. There are about 12 pairs of intestinal ceca.

The structure of the reproductive apparatus agrees closely with that of the corre-
sponding parts in P. siphunculus, as described by Lang.

The antrwm masculinum extends in a vertical direction for a short distance, passes
slightly behind the male aperture, then runs forwards again, becoming strongly bent on
itself. At its anterior end is the papilla, from which the apex of the penial stylet
projects. The prostate reservoir is a small rounded dilatation of the ejaculatory duct.
The median vesicula is of great relative size with very muscular walls. The lateral
(accessory) vesiculze seminales are smaller, spherical bodies with very thick walls and
small lumina

PROF, W. A. HASWELL ON AUSTRALASIAN POLYCLADS. 483

The female aperture leads into an antrum which takes the form of a bursa copulatrix
with very thick walls. The ootype has a very contracted lumen with folded walls ;
it extends downwards to the neighbourhood of the ventral surface. The shell-glands
are very highly developed, extending through more than one-half of the length of
the body. 4

In a specimen which has been cut into longitudinal vertical sections there is beneath
the integument of the dorsal surface, in the region immediately behind the sucker, a
great mass of spermatozoa, and a similar mass on and around the sucker on the ventral
surface. The sections are imperfect, and though a fissure enclosing spermatozoa is to
be traced downwards from the dorsally situated mass, it is impossible to determine
to what extent this has been formed by post-mortem treatment. In any case there
is sufficient evidence of the occurrence here of an indirect form of copulation by
perforation of the integument.

In view of this observation it is of interest to note that Lang expresses a suspicion
that the structure of the parts in Prosthiostomum points to self-fertilization (17, p. 638).

From Prosthiostomum siphunculus, Delle Chiaje (sp.), and from P. dohrnii, Lang
(17), this Australian species is distinguished by the number and arrangement of the
eyes ; and similar differences distinguish it from Laidlaw’s (15) two species, P. elegans
and P. cooper, from the Maldives.

BIBLIOGRAPHY.

1. Brrcenpat.—Einige Bemerkungen tiber Cryptocelides Lovent. Kongl. Fysiogr. Sallskapets Hand-
lingar, Ny foljd, iv. 1892, 3*.

2. Cureseman, T. F.—On two new Planarians from Auckland Harbour. Trans, N.Z. Inst. xv. 1882.

3. Cottincwoop, C.—On thirty-one Species of Marine Planarians, &c. Trans. Linn. Soc., Zool. (2)
i. 1876.

4. Gambier, F. W.—British Marine Turbellaria. Quart. Journ. Micro. Sci. xxxiv. 1893.

5. Goro, S.—Studies on the Ectoparasitic Trematodes of Japan. Journ. Coll. Sci. Imp. Univ. Japan,
viii. 1894.

6. Grarr, L. voy.—Enantia spinifera der Repriisentant einer neuen Polycladen Familie. Mittheil.
naturwiss. Vereins fiir Steiermark, 1889.

7. Haswett, W. A.—On a Prorhynchid Turbellarian from Deep Wells in New Zealand. Quart.
Journ. Micro. Sci. vol. xl. n.s. 1898.

On a new Histriobdellid. Quart. Journ. Micro. Sci. vol. xliii. n.s. 1900.

g. Herzic, E. M.—Laidlawia trigonopora, nu. gen., n. sp. Zool. Anz. xxix. 1905

10. Kirx, T. W.—On some new Marine Planarians. Trans. N.Z. Inst. xiv. 1881.

11. Larptaw, F. F.—Notes on some Marine Turbellaria from Torres Straits and the Pacific. Mem.
and Proc. Manch. Lit. & Phil. Soc. xlviii. 1903. .

A Collection of Turbellaria Polycladida from the Straits of Malacca (Skeat Expedition).

P.Z. S. 1903, i. pp. 301-318, pl. 23.

* T have not been able to see this paper.

1S4 PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS.

13. Larptaw, F. F.—Turbellaria Polycladida of Zanzibar, collected by C. Crossland.—Part I. The
Acotylea. P.Z.S. 1903, ii. pp. 99-113, pl. 9.
Suggestions for a Revision of the Classification of the Polyclad Turbellaria. Mem. and

I4.

‘ Proc. Manch. Lit. & Phil. Soe. vol. xlviii. 1903.

15. The “ Marine Turbellaria.” Fauna and Geography Maldive and Laccadive Archipelagoes,
1902.

16, —— Report on the Polyclad Turbellaria collected by Professor Herdman at Ceylon in 1902.

Report Pearl Oyster Fisheries of the Gulf of Manaar, 1904. Suppl. Report No. 9.

17. Lane, Aw~—Die Polycladen. Fauna u. Flora des Golfes von Neapel, xi. 1884.

18. Prenn, M.—Ergebnisse einer Reise nach dem Siid]. Pacific (Schauinsland, 1896-1897). Polycladen.
Zool. Jahrb. Abth. f. Syst. &c. Bd. xii.

1g. —— Drei neue Polycladen. Jen. Zeitschr. Naturw. Bd. xxxi. 1897.

20. Scumarpa, L. K.—Neue wirbellose Thiere, Bd. i. 1859.

21. QuaTRreraces, A. pe.—Mémoires sur quelques Planariés marines. Ann, Sci. Nat. 8° série, Zool.
t. iv. 1845.

22. Stimpson, W.—Descriptions of some new Marine Invertebrata. Proc. Acad. Nat. Sci. Philad.
vol. vil. 1855.

Prodromus descriptionis animalium, ete. Proc. Acad. Nat. Sci. Philad. 1857.

24, SrumMER-TRauNrers, R. yon.—Tropische Polycladen. 1. Das Genus Thysanozoon, Grube.
Zeitschr. f. wiss. Zool. Bd. Ix. 1895.

25. Vurritt, A. E.—The Marine Planarians of New Zealand. Trans. Conn. Acad. viii. pp. 459-520,
pls. 40-44 (1888).

26. Woopwortu, W. McM.—Some Planarians from the Great Barrier Reef of Australia. Bull. Mus.
Comp. Zool. Hary. Coll. vol. xxxii. pp. 63-67.

EXPLANATION OF THE PLATES.

Lettering.

d., anterior diverticulum of ootype. ¢., ejaculatory duct. iné., intestinal ceca, m., mouth,
vot., ootype. p., penis. p.s., penis-sheath. p.st., penial stylet. pr., prostate reservoir or prostate
ducts. 7, receptaculum seminis. 7.d., duct of receptaculum. sf/.gld., shell-glands. wt, uterus.
ut.d., uterine duct. va., dorsal limb of the vagina. v.d., vasa deferentia. v.s., vesicula seminalis.

PuatrE 35,
Fig. 1. Tripylocelis typica. General view of the organization. x 15.
2 A. $5 Brain, eyes, and outline of tentacles.
3. 5 is Transverse section passing through the proximal part of the penis.
4 5 x) ‘Transverse section at the point where the lateral uterine ducts unite.
DIG x6 Two successive transverse sections passing through vagina and its ventral

aperture ( 2 ?).
5 > Diagrammatic lateral view of the reproductive ducts.

“1

PROF. W. A. HASWELL ON AUSTRALASIAN POLYCLADS. 155

PLATE 36.
Fig. 1. Diplosolenia johnstoni. General view of the reproductive ducts from the ventral aspect.
Vasa deferentia and ejaculatory duct shaded ; ootype dotted.
mE 5 6 Transverse section passing through the point of union of the ducts of
the receptacula seminis.
3. Leptoplana australis. General view of the reproductive ducts from the ventral aspect.
4. sa 3 Diagrammatic lateral view of the reproductive apparatus. Epithelia
dotted ; muscular layers shaded.
D: x e Eyes. The two oval bodies are the anterior appendages of the brain.
6&7. Echinoplana celerrima. Spines of the penis. x 500.

PLATE 37.

Fig. 1. Echinoplana celerrima. Brain and eyes; from stained and mounted specimen.
2. s 5 General view of the reproductive apparatus, from the ventral aspect.
3. 5 Transverse section passing through the point of union of the uterine

duct with the diverticulum of the ootype.

4. Enterogonia pigrans, Diagrammatic lateral view of the reproductive apparatus, showing the
genito-intestinal canal. Epithelial parts dotted; muscular layers shaded.

5. Cestoplana australis. Outline magnified, to show arrangement of vermilion bands.

6. Pseudoceros cardinalis. Ventral view of male reproductive ducts.

7. Prosthiostomum maculatum. Cerebral eyes.

SECOND SERIES.— ZOOLOGY, VOL. IX. 70

Trans. Linn. Soc., Ser. 2. Zoou.Vol.IX.PL 35.

r. oe

E. Wilson, Cambridge

AUSTRALASIAN POLYCLADS.

HASWELL. Trans. Linn. Soc., Szr.2. Zoou. Vol.[X.Pu. 36.

E.Wilson, Cambridge.

AUSTRALASIAN POLYCLADS.

+. ‘a
\
‘
‘
’
\
‘
!
f
'
i

,
a

i

ie '
wy

ay
ri
:
al
se ..

a
/

Trans. Linn. Soc., Szr.2.Zoon. Vol. [X.Pri. 37

fel

HASWELL.

SEA EL an TOO

a A pO
en aowasnws

Te)

AUSTRALASIAN POLYCLADS

[Norr.—Synonyms and Native Names are printed in italics.

INDEX.

to be used for the first time. ]

A star is added to names which appear

Abdomen, segments of, 358,
Acalyptrata, 345.
Accipitres, Linn., mentioned, 29,
Acidia heraclei, Zinn., ejaculatory
apodeme of, 356,
Acilius suleatus, Linn., 372.
Acipenser, Lini., mentioned, 47, 69,
BOs
Acmea, Eschsch., 277, 278, 284, 285.
corticata, Hutton, mentioned,
275, 281, 289.
fragilis, Chemnitz, mentioned,
275, 273, 281, 288.
_ galathea, Lam., mentioned, 275,
279, 280, 286, 288.
testudinalis, Miller, mentioned,
275, 276, 281, 282, 288, 289.
virginea, Willer, mentioned, 275,
276, 281, 289.
Aceela, mentioned, 469.
Aconophora flavipes, Germar, men-
tioned, 331.
obfuscata *, Buckton, 331; men-
tioned, 337.
Acotylea, mentioned, 479.
Acrobates, Desm., mentioned, 97,
118, 124-131, 198, 199, 204.
pulchellus, W. Zothschild, men-
tioned, 204.
pygmeus, Desm., mentioned, 169,
Hpyprymnus, Garrod, 86, 148, 145,
146, 177, 178.
rufescens, Gray, mentioned, 144,
Aischna cyanca, Miller, genitalia in,
370.
Ageniaspis, Dahlbom, 253 ; mentioned,
231, :
fuscicollis, Dahlbom, mentioned,
231, 263.

Agromyzidee, 368,
Albula, Gron., 68, 69, 70.
conorhynchus, Bl. Schn., men-
tioned, 68, 81.
Alepocephalus rostratus, 2isso, men-
tioned, 79.
Alestes, Miller & Troschel, men-
tioned, 51, 53, 54, 59, 73-77, 81.
dentex, Miiller § Troschel, men-
tioned, 49, 51.
Kotschyi, Heckel, mentioned, 51,
565, 56, 59, 65.
longipinnis, Giinther, mentioned,
51, 545 ovaries examined
(Rowntree), 74.
macrolepidotus, Cuv. g§ Val.,
mentioned, 51, 54, 56, 59, 62,
64.
nurse, Miller & Troschel, men-
tioned, 51-56, 59, 62, 64;
ovaries examined (Rowntree),
74, 80.
Alpheus, Fabr., 427.
brevicristatus, de Haan, men-
tioned, 427, 428.
brevirostris, Olivier, 427; men-
tioned, 388, 428, 453.
-Edwardsii, Aud., mentioned, 341.
japonicus, Miers, 430; men-
tioned, 388, 453.
longimanus, Spence Bate, 430,
malabaricus, de Haan, men-
tioned, 427.
rapac, de Haan, 427, 428, 429.
Amblypoda, mentioned, 116, 117.
Amiurus, Rafinesque, mentioned, 47,
49, 54, 61, 66-76.
nebulosus, Zesuenr, mentioned,
67.

SECOND SERIES.—ZOOLCGY, Vou. 1%

Amphiproviverra, Ameghino, men-
tioned, 108, 109.
Amphitherium, Blainv., mentioned,
180.
Amphitrite hastatodes, de Haan, 391.
hastatoides, de Man, 391.
Anacyrtus, Giinther, mentioned, 54,
60, 62, 64, 73, 76, 78.
gibbosus, Linn., mentioned, 4).
guatemalensis, Giinther, men-
tioned, 51.
microlepis, vinh., mentioned,
51, 76% ovaries examined
_ (Rowntree), 74, 80.
Anaitis, Duf., mentioned, 259.
Anchon albolineatum, Buckton, men-
tioned, 333, 337.
fuscum *, Buckton, 334; men-
tioned, 337.
strigatum *, Buckton, 333; men-
tioned, 338,
Ancistromesus, Dall, mentioned, 275,
278, 282, 286, 288.
Andrena aterrima, G‘rav., mentioned,
242,
Anguillide, 69.
Anomalon circumflexun, Girav., larva
of, 240.
Anopheles cinereus, Hoffm., men-
tioned, 342, 383.
Anostomus, Gron., mentioned, 54, 59,
62, 64, 73.
fasciatus, Spia, mentioned, 51.
Antechinomys, Arefft, 91, 95, 97,
98; mentioned, 166, 193, 194.
laniger, Gould, mentioned, 93,
97, 98, 167, 217.
Anthomyia, Meigen, mentioned, 343.
pallida, Zett., ovipositor of, 38s.

TL

488

Anthomyia pluvialis, Linn., 226;
ovipositor of, 364.
radicum, Linn., 226 ; mentioned,
348, 351, 353, 385; recep-
taculum of, 365, 386.
sulciventris, Zett., 226.

Anthomyiide, 225-227.

Anthomyiine, 226, 228.

Anthomyza pallida, Zett., ovipositor
of, 361, 386.

Aphelinine, mentioned, 233.

Aphid, mentioned, 336.

Apis mellifica, Linn., appendages of,
369; genitalia of, 372; ovipositor
or sting of, 249, 250.

Arachnida, genitalia in, 370.

Arcania, Leach, 398.

11-spinosa, de Haan, mentioned,
401.

globata, Stimps., 400;
tioned, 387, 388, 452.

heptacantha, de Haan, 398;
mentioned, 388, 399, 452.

quinquespinosa, W. Mason, men-
tioned, 399.

septemspinosa (Fabr.), Leach,
mentioned, 398, 399.

var. gracilis, Hend., men-

tioned, 399.

Archeopteryx, Meyer, mentioned, 23.

Architeenioglossa, mentioned, 271.

Arctia caja, Zinn., mentioned, 372.

Arthropoda, ancestral forms of, 378.

Artystone trysibia Schiédte found in
body-cavity of Anacyrtus micro-
lepis, Reinh., 76.

Asilide, 226, 341-346, 353, 358,
363, 364.

Asilus crabroniformis, Zinn., men-
tioned, 222; ejaculatory apodeme
in, 304.

Asio, Briss., mentioned, 7, 27, 29,
30, 34-42.

accipitrinus, Pallas, skull of, 35,
36, fig. 45.

madagascariensis, Smith, men-
tioned, 15, 35; skull of, 35.

major, Schl., skull of, 35.

nisuella, Daud., mentioned, 35.

otus, Linn., skull of, 35.

Asionide, cranium of, 3, 14, 29, 31,
39-43,

Asioning, 38, 43.

Asteiide, 368.

Asterolecanium, Signoret, mentioned,
455, 462.

men-

INDEX.

Asthenognathine, mentioned, 396.
Asthenognathus, St?mps., 392, 394.
ineequipes, Stimps., 592; men-
tioned, 387, 388, 452.

Athene, Bote, skeleton of, 23; men-
tioned, 36, 39, 42.

Auchenaspis biscutatus, Geoffr., 67.

Auchenoglanis (Auchenaspis) bi-
scutatus, Geoffr., 67.

Australasian Polyclads, Observations
on, by Prof. W. A. Haswell,
465-485.

Australian Marsupialia, On the Evo-
lution of; with Remarks on the
Relationships of the Marsupials in
general, by B. Arthur Bensley,
83-217.

Azelia Macquarti, Stag, 226.

Balioptera combinata, Linn., 226,

230; receptacula of, 366,
tripunctata, Fallén, receptacula
of, 366.

Barbus Bynni, Cuv. g Val., men-
tioned, 66.

Bellerophon expansus, Sowerby, men-
tioned, 288.

Bensley, B. Arthur, On the Evo-
lution of the Australian Marsu-
pialia; with Remarks on the Re-
lationships of the Marsupials in
general, 83-217.

Bergendalia, Laidlaw,
468.

Beris nigra, Meigen, abdomen of, 361,
362, 363, 384.

vallata, Forster, 353; abdomen
of the male of, 361, 384;
ductus ejaculatorius of, 384;
ejaculatory apodeme of, 359,
384; forcipes interiores of,
350, 384; ovipositor of, 362,
363 ; penis of, 344, 346.

Bettongia, Gray, mentioned, 101,
133, 148-146, 177, 178, 200.

mentioned,

cuniculus, Oygilb., mentioned,
144,
Gaimardi, Desm., 215; men-

tioned, 144, 146, 177.
Lesueuri, Quoy g& Gaim., men-
tioned, 144-146.
penicillata, Gray, 215; men-
tioned, 144-J47, 201.
Bettongiine, 143-147, 150, 200.

| Bibio, Geoffr., mentioned, 341, 358,

Bibio hortulanus, Zinn., appendages
of, 383; forceps inferior of, 383;
genitalia of, 360, 363, 383; ovi-
positor of, 364 ; penis of, 344-347,

351: receptaculum seminis of,
383.
marci, Zinn., mentioned, 344—
346,

Bibionide, 342-846, 349, 358, 360,
363.

Bithynis japonica, de Haan, men-
tioned, 411.

Olfersiz, Rathbun, 314.

Blastothrix sericea, Dul., 234 ; laying
eggs in the body of a coccid, 234.

Bombus, Zatr., mentioned, 250,

Bombylide, 226, 358, 364.

Borboride, 358, 366.

Borborus, Meiyen, mentioned, 366.

suillorum, Hal., 226.

Brachyodont Series
Dorcopsis,
151-154.

Brachytrupes, mentioned, 371.

Braula ceca, Nitz., 369.

Braulide, 369.

Brycon faleatus, Miiller § Trosch.,
mentioned, 51.

Bryconaethiops, Giinther, 54; men-
tioned, 53, 56, 59, 62, 64, 73.

microstoma, Giinther, mentioned,
51.
Bubo, Cuv., mentioned, 18, 30, 34,
37-42.

capensis, Smzth, mentioned, 27,
28, fig. 45, 46.

magellanicus, Gmelin,
tioned, 2, 24, 28.

maximus, Fleming, mentioned,
29.

perspicillatum, Zath., mentioned,
fig. 45.

Bubonide, 14.

Bubonine, 2, 3, 39.

Buckton, G. Bowdler, Observations
on some undescribedor little-known
Species of Hemiptera-Homoptera
of the Family Membracide, 329-
338.

Burramys, Broom, mentioned, 144,
147, 199-201.

(Dendrolagus,
Setonyx) (Bensley),

men-

Callimone, Spin., mentioned, 240.
Calliphora, Desv., mentioned, 219,
221, 222, 225, 227, 350,

Calliphora erythrocephala, Afeigen,
343; mentioned, 219, 230, 385;
appendages of, 353, 885; ex-
tremity &c. of, 385; ovipositor
of, 859-364; paraphalli, 348, 351;
penis of, 353, 385.

Caloprymnus, Zhomas, mentioned,
148, 147-157, 177, 178, 200.

campestris, Gould, 215, 216.

Caluromys, Winge, mentioned, 101,

182, 189-191, 209.
philander, Zinn., 215.
sp., 216.

Caprimulgi, skull of, 2.

Carapus fasciatus, Cuv., 67.

Carassius auratus, Bleeker, mentioned,
66.

Cardines, homology of the (Wesché),
229.

Carine, Kaup, mentioned, 27,34, 37,
40, 41.

Carobia, Quatr., mentioned, 259.

Catabomba pyrastri, Linn., men-
tioned, 350, 351; appendages of, |
385 ; ejaculatory apodeme of, 356 ; |
penis of, 345, 346, 385.

Cathartze, mentioned, 31.

Catostomus macrolepidotus, Lesueur,
mentioned, 66, 81.

Cecidomyiide, mentioned, 341, 366.

Centetes, ///., mentioned, 96, 120.

Centetidee, mentioned, 96, 119-121.

Centrotus, Fabr., mentioned, 334, 336.

cornutus, Zinn., mentioned, 329, |
336.
nectaris, Green, mentioned, 336,
333.
Ceratopogon, Meigen, 364.
obscurus, Winn., 367.
Ceratopsyllus jubatus, Way., 366.
Cercococcus, Scott *, 455,
eremobius, Scott*, 456; men-
tioned, 463, 464; found on
Helianthemum —_ kahiricum,
Delile, mentioned, 455, 464.
Cercopide, mentioned, 336.
Ceresa, Am. fA. Serv., mentioned, 331.
nitens, Buckton*, 332; men-
tioned, 337.
Cestoplana, Lung, mentioned, 479.
australis, //usw.*, 479; men-
tioned, 485.
rubrocincta, Grube, mentioned,
479.

Cheetocercus, Krefft, mentioned, 91, |

167, 168, 193. |

INDEX.

Chetocercus cristicauda, Krefft, 217 ;
mentioned, 94, 98, 164, 167, 194.
Chaleidide, 231, 248.
Chalcinopsis dentex, Gtinther, men-
tioned, 51.
Chalcinus brachypomus, Cuv. § Val.,
mentioned, 51.
Chalurius spurius, Fallén, 367, 386.
Characinide, On some Points in the
Visceral Anatomy of the, with an
Enquiry into the Relations of the
Ductus Pneumaticus in the Phy-
sostomi generally, by Walter §.
Rowntree, 47-81.
Chelyoida, Buckton, mentioned, 331.
Chionaspis aspidistre, Signoret, men-
tioned, 233.
Chirocentrus, Cuv., 69, 70.
dorab, Cuv., 69, 81.
Chironectes, Ji/., mentioned, 163,
182, 184.
Chironomide, 363, 364, 367.
Chironomus, Linn., 364; ovipositor
in, 358.
plumosus, Zinn., 363; ovipositor
of, 363.
riparius, Weiyen, ovipositor of,
363.
Chiton, Zinz., 278.
Chitonide, mentioned, 276.
Chloridella, Miers, 439.
affinis, Berthold, 388, 439.
fasciata, De Haan, 338, 440.
Chloromyia formosa, Scop., men-
tioned, 341.

Chloropide, 368.

Cheeropus, Oyilb., mentioned, 110,
114, 131, 176, 195.
castanotis, Gray, upper incisors
and eanine of, 105, 106, 111,
113, 114,113, 189, 216, 217.
Chrysichthys auratus, Geoffr., 67.
Chrysochloride, mentioned, 110, 119—
121.
Chrysochloris, Cuv., mentioned, 96,
119, 120.
Chrysops cecutiens, Zinn., mentioned,
221, 222,
Ciccaba, Wagl., mentioned, 35, 36,
39, 40.
Cimbex ariana, Kirby, genitalia of, 372.
Circaétus, Vierll., mentioned, 6.
Cirratulus atrocolluris, Grube, 265;
mentioned, 258.
capensis, Schmarda, 266; men-
tioned, 255.

489

Cirratulus tentaculatus meridionalis,
Montagu, 265; mentioned, 255.
Citharidium, Boulenger, mentioned,

53.
Ansorgii, Boulenger, mentioned,

52.
Citharinus, Cuv., mentioned, 47, 50—
54, 62, 73, 76, 78.
congicus, Cuv., mentioned, 52.
Geoftroyi, Cuv., mentioned, 49,

52, 65.

latus (Khrenb.), Miller §
Trosch., mentioned, 52, 57-
59, 65.

macrolepis, Boulenger, men-
tioned, 52, 64, Si.
Clarias lazera, Cuv. g Vul., men-
tioned, 67, 81.
Clupea, Linn., 69, 70.
harengus, Linn., 69.
sprattus, Linn., 69, 81.
Coccide, On Cercococeus eremobius*,
gen, et sp. noy.,an Aberrant Form
of, by Hugh Scott, 455-464.
Ceelopa, Metyen, 368.
Ccenolestes, Thomas, 114, 124, 125,
129, 208.
obscurus, J'homas, 215; men-
tioned, 124.
Coenosiin, 228.
Columbia transmontana, Zigenmann,
69.
Comptosia ocellata, New., ovipositor
of, 361; receptacula of, 365.
Comys, Forster, 233; bibliography of,
252.
bicolor, Toward, mentioned, 232.
infelix, Ambleton, On the Ana-
tomy and Development of,
a Hymenopterous Parasite
of Lecanium hemisphericum,
Targioni - Tozzetti, by Alice
L, Embleton, 231-254.
Condylarthra, mentioned, 209.
Conger conger, Linn., 69.
Conopide, 367.
Cordyluride, 342, 343, 348, 351,
361, 368; teeth of, 377.
Coregonus albus, Lesueur, 68, 70.
Cotylea, Lioy., mentioned, 482.
Crangon, Fabr., 405.
affinis, Ortm., 405; mentioned,
406,
alaskensis, Lochtngton, men-
tioned, 405, 406.
alba, Holmes, mentioned, 407.
(ihe

490

Crangon angusticauda, De Haan, |
408.

cassiope, de Man, 406; men-
tioned, 388, 407, 408, 452.

consobrinus, de Man, 405 ; men-
tioned, 388, 452.

Holmesi, Rathb., mentioned,
407.

propinquus, Stimps., mentioned,
406,

vulgaris, Fabr., mentioned, 405,
406, 407, 408.

Creodonta, mentioned, 120, 121.

Crustacea, On a Collection of, Deca-
poda chiefly
from the Inland Sea of Japan;
with Descriptions of New Species,
by Dr. J. G. de Man, 387-454.

Cryptocelides Loveni, Bergendal,
465 ; mentioned, 483.

Cryptocelis alba, Zang, mentioned,
478.

compacta, Lang, mentioned,
478.

Cryptocelis-like form, discovery of,
466, 474.

Culex, Zinn., cardines of, 229 ; cen-
tral organ of, 383; forceps of,
384; interior hook of, 383; ovi-
positor of, 358.

pipiens, Linn., mentioned, 222,
230, 342; genitalia of, 360;
maxilla and mandible of, 223,
230; penis of, 346.

Culicide, 226, 342-349, 358, 360,
367, 378.

Curimatus, Cuv., mentioned, 53, 59,
62, 65, 73, 76.

albula, Liitken, mentioned, 52.

cyprinoides, Cuv. g¢ Val., men-
tioned, 52.

dobula, Giinther, mentioned, 52,
57.

Gilberti, Quoy & Gaim., men-
tioned, 52; ovaries examined
(Rowntree), 74.

lineatus, Cuv. g Val.,
tioned, 52.

Cyclopodia Hopei, Westw., 369.

Cyclorhynchus planirostris, de Haan,
421,

Cyclorrhapha, 370.

Cyprinidx, mentioned, 48, 70.

Cyprinodon, Lac., 69.

calaritanus, Cuv. g Val., 69.

Cyprinodontide, 69, 70.

and Stomatopoda,

men-

INDEX.

Cyrtide, 367.
Cyrtoneura,
228.

Macq., mentioned,
stabnlans, Fullén, mentioned,
225, 227, 230.

Dactylopiine, mentioned, 455.

Dactylopsila, Gray, mentioned, 114,
125-133, 170, 188, 199.

Dasyuridx, mentioned, 90, 107, 117—
122, 142, 162, 163, 169, 179-181,
193, 194, 204-210.

Dasyurine, mentioned, 91, 97-105,
114, 124, 186-188.

Dasyuroides, Spencer, mentioned, 91,
168, 193, 194.

Byrnei, Stirl., mentioned, 94,
164, 167, 217.

Dasyurus, Geoffr., mentioned, 91,
94-99, 168, 171, 182-186, 193,
194, 206.

albopunctatus,
tioned, 206.

Byrnei, Stirl., 217; mentioned,
98.

Geoffroyi, Gould, 216; men-
tioned, 91-98, 164, 168,
194.

hallucatus, Gould, 217; men-
tioned, 91-98, 163-168, 194,
206,

maculatus, Kerr, mentioned, 91—
98, 163, 168, 194, 206, 214,
215.

viverrinus, Shaw, mentioned,
91, 94, 98, 112, 164, 168,
194, 215, 217.

Decatoma, Spin., mentioned, 240.

Dendrolagus, Mii/ler, mentioned, 143,
151-156, 178, 202.

Bennettianus, De Vis, 151;
Queensland form, 151-153.
Dorianus, Ramsay, 151; Pa-

puan form, 151, 153.

inustus, Miiller § Schlegel, 151,
178; Papuan form, 151-
153.

Lumholtzi, Collett, 151; Queens-
land form, 151-153.

ursinus, Schlegel § Muller, 217 ;
mentioned, 178.

Dermestes, Zinn., mentioned, 344.

Didelphyide, mentioned, 92. 97-105,
114-131, 163, 171, 173, 180-188,
192, 202-208.

Schlegel, men-

Didelphys, Zinn., mentioned, 101,
118, 171, 173, 180, 184, 192.
aureus, mentioned, 188,
[=Didelphys and Metachirus],
182.
Dilophus, Mergen, mentioned, 226.
albipennis, Meigen, labium of,
230; ovipositor of, 363.
febrilis, Linn., ovipositor of, 363,
364; receptaculum seminis
of, 364.
Dinocerites, genitalia of, 360.
Dioncus, Stimps., mentioned, 471.
badius, Stimps., mentioned, 465,
471.
oblongus, Stimps., mentioned,
471. :
Diplosolenia, Hasw.*, 469; men-
tioned, 466, 471, 479.
Johnstoni, Hasw.*, 469, 485.
Diprotodon, Owen, mentioned, 141,

158-160, 172-175, 180, 213-
216.
australis, Owen, mentioned, 213,
215, 216.

Diprotodontia, mentioned, 123, 207,
208. ,

Diprotodontide, 116, 125, 128, 142,
158-163, 173, 174, 192, 198, 203,
210. :

Diptera, differences in mouth of
males and females, 376.

Genitalia of both the Sexes

in, and their Relation to the

Armature of the Mouth (W.

Wesché), 339-386.

Labial and Maxillary Palpi

in (W. Wesché), 219-230. —

, mouth-parts in, diagram of,

375.

ovipositor in, 358,

—— Similarity of Appearance of
Genitalia and Mouth-parts in,
369-373. :

—— Table of Relationship of Geni-
talia and Mouth-armature in,
376. }

Dipterous larva, parasitic, mentioned,
242,

Discocelis, Schmidt, mentioned, 479.

tigrina, Lang, mentioned, 470,
479.

Distichodus, Miiller §- Troschel, men-

tioned, 52-54, 59-63, 78.
antonii, Schiliduis, mentioned,
52, 64, 81.

Distichodus niloticus, Miiller & Tro-
schel, mentioned, 52, 57, 58, 64.
Distivechurus, Peters, mentioned, 97,
118, 124-134, 184, 198, 199,

204,
pennatus, Peters, mentioned,
112, 116, 124, 126, 215,
216.

Docoglossa, On the Evolution of
Topographical Relations among
the, by H. J. Fleure, 269-290,

Dolichopodide, 341-349, 353, 358-
364.

Dolichopus, Latr., 342, 349, 351,
359,

crenatus, Osten Sacken, men-
tioned, ftnote 350.

festivus, Haliday, 384; men-
tioned, 350 ; ejaculatory duct
of, 384; forceps interior of,
384; ovipositor of, 364;
palpus genitalium of, 384;
penis of, 346; spinus titil-
latorius of, 384; theca of,
384.

griseipennis, Stannius, men-
tioned, 345 ; ejaculatory apo-
deme of, 356, 364, 3845; ovi-
positor of, 361, 362, 385;
penis of, 345, 346, 384;
receptacula of, 365 ; vas defe-
rens of, 384,

nobilitatus, Linn., penis of, men-
tioned, 345.

plumipes, Scop., 346, 384.

Dorcopsis, Schlegel & Miller, men-
tioned, 143, 151, 155, 156, 178,
202, 204.

luctuosa, D’ Albertis, 151; men-
tioned, 178; Papuan form,
151, 153, 215, 216.

Macleayi, Mtkl.-Macl., 151;
mentioned,178; Papuan form,
151, 153.

Muelleri, Schlegel, 151; men-
tioned, 178; Papuan form,
151, 153.

Dromeus, Ztanz., mentioned, 19.

Dromicia, Gray, 170; mentioned,
86, 118, 124-135, 150, 158, 159,
169, 199-206.

caudata, Milne-Hdw., men-
tioned, 204,

concinna, Gould, 215; men-
tioned, 116, 127, 1382, 134,
206.

INDEX.

Dromicia lepida, Thomas, mentioned, |

127, 132, 133, 1384, 150, 206.
nana, Desm., mentioned, 127,
132, 133, 134, 150, 169, 199,
206, 217.

Dromiciops, Z’homas, mentioned, 184,
185,

Drosophila funebris, Fabr., genitalia
of, 368.

Drosophilidie, 368.

Ductus Pneumaticus, Relations of the,
in the Physostomi (W. 8S. Ruwn-
tree), 47-81.

Duplicidentuta, Ill., mentioned,
159.

Dytiscus, Linn., 370.

Echinomyia fera, Zinn., mentioned,
343.

Echinoplana, Hasw.*, 475;
tioned, 466, 477.

celerrima, Husw.*, 475; men-
tioned, 478, 485.

Echinops, Mart., mentioned, 96,
120.

Ege-Guides and the Appendages of
the Ultimate Segment of the Ovi-
positor, 360.

men-

Ejaculatory apodeme, structure of, |

354; nomenclature of, 354.

Elamena (Trigonoplix) unguiformis,
Alcock, 396.

Elanoides, Vieill., mentioned, 29.

Elanus, Sav., mentioned, 32.

Electrophina, Buckton *, 331.

pacificata *, Buckton, 331 ; men-
tioned, 332, 337.
Elops, Zinn., 68, 70.
saurus, Linn., 68, 81.

Emarginula, Zam., mentioned,
269.

Embleton, Alice L., On the Anatomy
and Development of Comys infelix,
Embleton, a Hymenopterous Para-
site of Lecanium hemisphericum,
Targioni-Tozzetti, 231-254.

Empida stercorea, Zinn., mentioned,
841, 342.

Empidex, mentioned, 221-227, 341-
353, 358-367, 375.

Empis, Zinn., mentioned, 222, 224,
359.

chioptera, Fallén, 224, 230;
ovipositor of, 361.

491

Empis stercorea, Linn., forceps
superior of, 334; mouth-parts of,
3773 penis of, 346, 384.

tessellata, Fubr., labrum of,
386,

Encyrtine, mentioned, 232-235,

species parasitic upon :—
Aphides, 233.
Coleoptera, 233.
Hemiptera, 233.
Hymenoptera, 233.
Lepidoptera, 233.

Encyrtus fuscicollis, Dalm., men-
tioned, 231, 240, 242, 252; ovi-
positor of, 248 ; respiration of the
larva of, 242, 243.

Enterogonia, Hasw.*, 478; men-
tioned, 479.
pigrans, Hasw.*, 478; men-

tioned, 485.
Entylia fuscodorsa, Buckton*, 332 ;
mentioned, 337.
mesta, Buckton*, 332; men-
tioned, 3-57.
Epanorthide, mentioned, 124, 208,
209.
Ephydra, Ztond., mentioned, 224,
226, 229.
Ephydride, 345, 347, 355, 458;
labrum of, 374.
Ericulus, Geoffr., mentioned, 96,
120.
Erinaceide, mentioned, 116.
Erinaceus, Zinn., mentioned, 84.
Eriphyle aphroditois, Pallas, men-
tioned, 263.
capensis, Atnberg, 263; men-
tioned, 255.
Fristalis, Zatr., 222, 229,
arbustorum, Zinn., mentioned,
221.
intricarius, Linn., ovipositor of,
362.
tenax, Linn., mentioned, 221,
222, 230, 342-353, 356, 365,
385.
Erythrinus, Giron., mentioned, 52-
54, 61, 63, 64, 77.
uniteeniatus, Spiv, mentioned,
49, 51; ovaries examined in
(Rowntree), 74, 80.
Esox, Linn., 69.
lucius, Zund., mentioned, 69,
81.
Eugnathichthys, Boulenger, men-
tioned, 52, 53, 59, 62, 64, 73.

492

Eugnathichthys Ectveldii, Boulenger,
mentioned, 51.
macroterolepis, Boulenger, men-
tioned, 51.
Eulalia capensis,
mentioned, 255.
Lunia stragulum, Grube, mentioned,
ftnote 264.
Eupalemon elegans, de Man, men-
tioned, 296.
endehensis, de Man,
tioned, 317.
Foai, Coutiére, 306 ; mentioned,
291, 326.
lar, Fabr., 291; mentioned,
325.

Schmarda, 269 ;

men-

macrobrachion, JJerklots, 299;
mentioned, 291, 306, 309,
320; measurements of, 320,
325.
sundaicus,
306.
Eupelmus, Dalm., mentioned, 240.
Euphrosyne capensis, Ainberg, 256 ;
mentioned, 255, 267.
polybranchia, Schmarda, 256.

Heller, mentioned,

Euryleptide, mentioned, 481.
Eutheria, 84.

Falconiformes, skull of, 2, 14.

Fissurella, Brug., mentioned, 269.

Fissurellide, circulatory system,
general features of the, 272.

Flabelligera affinis, J. Sars, men-
tioned, 266.

luctator, Stimps., 266 ;
tioned, 255.

Fleure, H. J., On the Evolution of
Topographical Relations
the Docoglossa, 269-290,

Forsyth Major's Classification of the
Marsupials in general, 207-210.

Fucellia fucorum, Fallén, 351.

Fulgoride, mentioned, 336.

Fundulus robustus, Bean, 69.

men-

among

Gadus morrhua, Zinn., mentioned,
ftnote 77, 79.
Galathea, Fabr., 402.
acanthomera, Stimps., 402;
mentioned, 387, 388, 404,
452.
orientalis, Ortm., 402;
tioned, 404.

men-

INDEX.

Galaxias, Cuv., 69.
truttaceus, Cuv. g- Val., 69, 81.
Galaxiide, 70.
Galesaurus, Owen, mentioned, 99.
planiceps, Owen, mentioned,
213.

Gastrophilus equi, Fubr., 368; mouth-
parts of, 377.

Gastropods, The common Ancestor
of the Prosobranch (Fleure),
269.

Genitalia, The, of both the Sexes in
Diptera, and their Relation to the
Armature of the Mouth, by Walter
Wesché, 339-386.

in the Arachnida and Hydrach-

nide, 370.

homologies of the, in Peri-

planeta orientalis and Diptera,

Burm., 371.
and Mouth-armature in Diptera,

Table of Relationship of, 376.
and Mouth-parts, Further Re-
marks on the Relationship between
(Wesché), 3738.
Geomyzide, 361.
Geotrupes stercorarius, Zinn., 372.
Gisella, Bonaparte, mentioned, 1.
Glaucidium, oie, mentioned, 29, 36—
42,
Ridgwayi, Sharpe, skeleton of,
36.
Glossina, Wied., mentioned, 224,
341-351, 377, 378.
morsitans, Wesiw., 343; men-
tioned, 353, 365, 385.
pallidipes, Austen, 351; men-
tioned, 353, 354, 385.
palpalis, Des., mentioned, ftnote
342, 343, 348, 350-354, 365,
369; genitalia of, 372, 385.
tachinoides, Westw., mentioned,
345, 350-353, 385.
‘« Glue-glands,” 359.
Glycera africana, Arwidsson, 260.
conyoluta africana, Aeferstein,
260; mentioned, 255.
Goodea atripinnis, Jordan, 69.
Graphomyia maculata, Scop., 226,
230.
Grapsus (Platynotus)
de Haan, 392.
Grymexomys, Burm. [= Marmosa],
182.
Gymnarchus, Cuv., 68.
niloticus, Cuv., 68, $1.

depressus,

Gymnasio, Bonap.,
37-40, 42.

Gymnobelideus, MZ‘Coy, mentioned,
125, 127, 147, 198, 199, 206.

Gymnoscops, JZ'ristram, mentioned,
18, 27, 30.

insularis, Tristram, mentioned,
30.

Gymnotide, 48, 60, 70.

Gymnotus, Zinn., mentioned, 50.

Gymnura, Vig. g Horsf., mentioned,
84.

Gynoplista bella, Westw., mentioned,
342, 353 ; central organs of, 395,
384 ; forceps superior of, 384.

mentioned, 7,

Hematobia stimulans, Meigen,
receptacula of, 365, 377.

Hematopota crassicornis, Wahlb.,
ovipositor of, 363 ; receptacula of,
365.

pluvialis, Zinn., mentioned,
ftnote 219, 221; labium of,
221, 230; ovipositor of, 363 ;
receptacula of, 365,

Haliotide, mentioned, 286.

Haliotis, Zinn., mentioned, 269-272,
284, 285.

tuberculata, Zinn., mentioned,
287.
Halosaurus, Johns., 70.
macrochira, Giinther, 69.
Hamma, Buckton *, 330.
nodosa, Buckton *, 330; men-
tioned, 337.

Haswell, W. A., Observations on
Australasian Polyclads, 465-
485.

Helcion pellucidum, Zinn., men-
tioned, 286,

Heliodilus, Hdw., mentioned, 1.

Heliomyiide, 358.

Hellenus hastatoides (Fabr.),de Haan,
388, 391.

hastatoides, Alcock, 391.

Helomyza similis, Meigen, recep-
tacula of, 366.

Helophilus, Meigen, mentioned, 222,
229.

pendulus,
222, 230,

Hemilepidia erythrotenia, Schmarda,

258; mentioned, 255, 257, 267.
tuberculata, Schmarda, 257.
Hemiodus, Miiller, 76.

Linn., mentioned,

Hemirhamphus, Cuv., mentioned,
61.

Hemiurus [=Peramys], 182.

Heterocotylea, genito-intestinal canal
in, 479.

Heteroneuride, 368.

Hilaria cilipes, Meigen, ovipositor of,
360-362, 354; penis of, 345, 346,
384.

Hippobosea, Zinn., mentioned,

Hippoboscidee, 368.

Hippolysmata, Stimps., 423,

amboinensis, de Man,
tioned, 426.
Kiikenthali, de Man, mentioned,
426.
vittata, Stimps., 423;
tioned, 388, 426, 453.
var. subtilis, Thallwitz,
423; mentioned, 425.
Hippolyte pandaloides, Stimps., 418.
réectirostris, Stimps., 411.

Homalomyia canicularis, Linn., 225,

230.
manicata, Meigen, receptacula
of, 365, 386.

Homalomyiine, 225, 226, 228.

Homology of the Cardines ( Wesché),
229; of ovipositor, 341-348.

Huhua, Hodys., mentioned, 29, 37,
42.

nipalensis, Hodys., mentioned,
37 ; sternum of, 29.

Hydrachnide, genitalia in, 370.

Hydrellia griseola, Fallén, 229, 230 ;
genitalia of, 373; receptacula of,

366,

Hydrocyon, Cwv., mentioned, 53, 54,
59, 62, 73, 76.

brevis, Giinther, mentioned, 49,
51, 64, 80.

Forskalii, Cuv., mentioned, 49,
51, 52, 64; ovaries examined
(Rowntree), 74.

goliath, Boulenger, mentioned,
dl.

Hydrotwa, Desv., mentioned, 343,

dentipes, abr., ovipositor of,
359, 360, 364.

Hydrotea, Rond., 228,

dentipes, Habr., 225, 230.

924,

“a4

men-

men-

INDEX.

| Hyetodesia obscurata, Meiyen, men-

occulta, Meigen, mentioned,
227.
Hyetodesia, Rond., mentioned, 227,
228.
lucorum, Zett., 225, 230,

tioned, 343.
perdita, Metyen, 225, 230.
Hylemyia cardui, Meiyen, 226, 230.
cinerosa, Zett.(?), receptacula of,
365.
pullula, Zté., 226.
Hymenoptera, genitalia of, 372; ovi-
positor or sting of, 249.
Hyodon, Less., 70.
alosoides, Rafinesque, 69.
Hypophallus, 348.
Hypopharynx, 374.
Hypsauzhenia jugulata, Buckton *,
332; mentioned, 337.
Hypsiprymnodon, /tams., mentioned,
1383, 143-147, 177, 178, 200, 201.
moschatus, /tams., mentioned,
144, 150, 215, 217.
Hypsodont (the) Series (Lagor-
chestes, Lagostrophus, Onychogale,
Petrogale, Macropus), 154-158.
Hyracoidea, mentioned, 140.

Ibiceps rufipennis, Buckton*, 334;

mentioned, 338.
Ichthyoborinw, mentioned, 60, 63

lolad
tle

b]

Ichthyoborus, Giinther, mentioned,
52, 53, 54; 59, 62, 64, 73.
besse, Joannis, mentioned, 51,
80,
niloticus [lapsus calami=besse ],
Joannis, mentioned, 51, 80.

Ichthyomys, Zhomas, mentioned,
162.
Idioplana, Woodworth, mentioned,
471.

Iphis heptacantha, de Haan, 398,
399.

Japan, On a Collection of Crustacea,
chiefly from the Inland Sea of,
by Dr. J. G. de Man, 387-454.

Kangaroos, mentioned, 143, 154-
157.
Dorca, mentioned, 143.
Tree, mentioned, 143.
Ketupa, Zess., mentioned, 14, 18, 34,
36, 40-42.
Kleidos palmatus, Buckton *, 333.
vomeris, Buckton, mentioned,
333.

493

Labrum, composition of, 374; forcipes
superiores of, 37-4.
Lagisca, Malmgr., mentioned, 258.
Lagorchestes, Gould, mentioned, 143,
154-157, 178, 179, 202.
conspicillatus, Gould, mentioned,
157, 179.
hirsutus, Gould, 2165 ;
tioned, 157, 179.
leporoides, Gould, 216; men-
tioned, 157, 179.
Lagostrophus, Thomas, mentioned,
142, 154-157, 178, 179, 202.
fasciatus, Z'hom«s, mentioned,
Lig:
Laidlawia, Herzig, mentioned, 468,
469, 479, 483.
trigonopora, Herzig, mentioned,
483,
Lambrus, Leach, 388; mentioned,
387.
(Oncodolambrus) predator,
de Man, 389; mentioned,
388, 452.
(Parthenopoides) pteromerus,
Ortm., 391,
Lamproptera, Fuirmaire, mentioned,
333.
Laphria fulva, Zyger, testes of, 357.
Lasiops, Meigen, 225, 228,
ctenoctema, Aow., 226.
Latreutes, Stimps., 421.
acicularis, Ortm., 421
tioned, 388.
laminirostris, Ortm., 422; men-
tioned, 388.
planirostris, Ortm., 421; men-
tioned, 388.
Lauxania wnea, Mullén, ovipositor of,
364, 385, 386.
Leander (Desm.), Stimps., 409.
longipes, Ortm., 409; men-
tioned, 387, 388, 411, 452.
longirostris, de Man, 409, 411.
Ortmanni, Ltathbun, mentioned,
411.
pacificus, Stimps., mentioned,
410,
paucidens, de Haan, 409 ; men-
tioned, 388.

men-

3; men-

serratus, Penn., mentioned,
410.

styliferus, Milne-Edw., men-
tioned, 411.

treillianus, Jtisso, mentioned,

410,

49 4

Lebiasina, Cuv. § Val., mentioned,
52-54, 59-64, 73-77.
bimaculata, Cuv. § Val., men-

tioned, 49, 51,5; ovaries
examined in (Rowntree), 74,
80.

Lecanium, Burm., mentioned, 233,
456, 461.
hemisphericum, argioni-Toz-
zetti, On the Anatomy and
Development of Comys infelix,
Embleton, a Hymenopterous
Parasite of, by Alice L.
Embleton, 231-254.
var. filicum, Douglas, men-
tioned, 233, 235.
Leontis, Malmgren, mentioned, 262.
Lepidonotus clava, Montagu, men-
tioned, 236.
clava semitectus, Stimpson, 256;
mentioned, 255, 267.
semitectus, Stimpson, 256,
Lepidoptera, genitalia in, 372.
Lepidosteus, Agass., mentioned, 47,
69, 70, 72.
Leporinus, Spiv, mentioned, 52, 53,
59, 62, 64, 73.
Frederici, Bloch, mentioned, 51.
Leptide, 367.
Leptocentrus impunctus, Buckton™,
334; mentioned, 338.
Leptodius, JMdilne-Hdw., mentioned,
441.
Leptoplana, Zhr., mentioned, 477.
alcinoi, Lang, mentioned, 473,
474,
australis, Laidlaw, 471; men-
tioned, 465, 470, 472, 473,
474, 485,
badia, Stimps., mentioned, 465.

californica, Plehn, mentioned,
74.

fallax, Diesing, mentioned,
473.

subviridis, Plehn (Laidlaw),

mentioned, 47U, 479.
vitrea, Zang, mentioned, 473.
Leuciscus leuciscus, Zinn. men-
tioned, 66.
rutilus, Zinn., mentioned, 66,
81.
Leucosia, Fabr., 397.
maculata, Stimps., mentioned,
397.
rhomboidalis, de Haan, 397;
mentioned, 388, 452.

INDEX.

Leucothorax, Buckton *, 334.
villosa, Buckton*, 334; men-
tioned, 338.
Limnogale, Major, mentioned, 96.
Limnosina, mentioned, 366.
Limosina fuscipennis, Hal., 226.
lugubris, Hul., 226.
sylvatica, Meigen, 226.
Lipobranchus capensis, Willey *,
266; mentioned, 255, 267, 268.
Locusta viridissima, Linn., ovipositor
or sting of, 250.
Lonchea nigrimana, Meiyen, penis
of, 3847, 350, 351, 354.
Loncheide, 345, 349, 350, 359,

Qe

lite
Lonchopteride, 367.
Lottia viridula, Lamarck, mentivned,
275, 280, 289.
Loxocera albiseta, Schrank, genitalia
of, 368.
Lucanus cervus, Linn., 372.
Lucilia, . Desv., mentioned, 348,
359.
cesar, Linn., 351.
sericata, Meigen, »vipositor of,
360, 364,
Lumbriconereis capensis, Grube, 265;
mentioned, 255.

cavifrons, Grube, mentioned,
265.
coccinea, Lenier, 264; men-
tioned, 255, 265, 267,
268.
Diibeni, Atnberg, mentioned,
265.
nardonis, Grube, 265; men-

tioned, 255, 267.
Lupa, Leach, 391.
(Hellenus) hastatoides (/abr.),
de Haan, 391; mentioned,
388.
Lysidice capensis, Grube, 264; men-
tioned, 255, 267.

Maclovia iricolor capensis, Montagu,
264; mentioned, 255, 267.
Macrobrachium Ihering,
mentioned, 320, 3824.
jamaicensis, Herbst, var. ango-
lensis, de Man*, mentioned,
314, 324.
var. Vollenhovenii, Herk-
lots, 309; mentioned, 291 ;
measurements of, 322, 326.

Ortm.,

Macrobrachium latimanus, v. Mar-
tens, 206; mentioned, 291, 298,
299, 325, 446,

Olfersii, Wiegm., 314; men-
tioned, 291, 3815, 320;
measurements of, 323, 324,
326.

sp., 319; mentioned, 291, 324,
Macrodon, Muller, mentioned, 52-64,
72) Tos
trahira, Spiv, mentioned, 49,
51; ovaries examined in.
(Rowntree), 74, 80.
Macropodide, 116, 125, 128, 183-
1385, 142, 143, 160-163, 177-
180, 192, 198, 200, 211.
Macropodinz, mentioned, 144, 145,
150-160, 178, 200, 202.
Macropus, Shaw, molars of, 89, 148,
154, 178, 179, 201, 202,
agilis, Gould, mentioned, 157.
Bedfordi, Zhomas, mentioned,
157, 208.
Billardieri, Desm., mentioned,
156, 157.
brachyurus, Quoy & Gaim.,
mentioned, 151.
Coxeni, Gray, mentioned, 156,.
MAO’
dorsalis, Gray, mentioned, 178,
2038, 217.
Eugenii, Desm., mentioned, 155,
156, 157, 179, 2U2.
giganteus, Zimm., 156, 215,
215,
Greyi, Gray, mentioned, 156,
157, 203. '
irma, Jourd., mentioned, 156,
202. ,
magnus, Owen, mentioned, 156.
Parryi, Benn., mentioned, 157.
robustus, Gould, mentioned,
156,
ruficollis, Desm., mentioned, 157.
rufus, Desm., mentioned, 155,
156, 215, 216.
stigmaticus, Gould, mentioned,
156, 157. ,
Thetidis, Zess., mentioned, 156,

157. i
Wilcoxi, ‘Coy, mentioned,
156, 157.

Macroscelide, mentioned, 96, 116.

Major, Forsyth, ‘he Classification of
the Marsupials in general, 207-
210.

Malacocotylea, genito-intestinal canal
in, mentioned, 479.

Malapterurus electricus, Zinn., men-
tioned, 67, 81.
Mammalia, analogy in the, 378.
—— phylogenetic plan showing the
primary relationships of the, 84.
Man, J. G. de, On some Species of
the Genus Palemon, Fubr., from
Tahiti, Shanghai, New Guinea, and
West Africa, 291-327.

—— On a Collection of Crustacea,

Decapoda and Stomatopoda, chiefly |

from the Inland Sea of Japan;
with Descriptions of new Species,
387-454.
Marmosa, Gloger, mentioned, 181-
184, 191, 192, 200.
cinerea, 7emm., 215; mentioned,
182, 183, 185.

elegans, Waterh., mentioned,
183, 185.

murina, Zinn., mentioned, 183, |
184, 185.

pusilla, Desm., mentioned, 183, |
PATE,

rapposa, Thomas, mentioned,
183, 184, 185.

velutina, Wager, mentioned,
185.

velutissima, read velutina, Wag-
ner, mentioned, 185.
Marphysa, Quatrefages.
adenensis, Gravier, mentioned,
263, 264.
Belli, Audouin & Edwards, men-
tioned, 264.
capensis, Schmarda, 263; men-
tioned, 255, 267.
hemasoma, Quatrefages, men-
tioned, 263.
purcellana, Willey *, 263 ; men-
tioned, 255, 264, 267.
sanguinea hemasoma, Montagu,
263 ; mentioned, 255, 267.
Marsupialia (Australian), On the
Evolution of; with Remarks on
the Relationships of the Marsu-
pials in general, by B. Arthur
Bensley, 83-217.

Marsupials, Australian, the Adaptive |

Modifications of the Foot-structure

in the (Bensley), 162.

bibliography of, 211-214.

—_— —— identification of the stem-
form of the (Bensley), 179.

SECOND SERIES.—ZOOLOGY, VOL.

)

INDEX.

Marsupials, Australian, phylogenetic |
arrangement of the (Bensley), 192.

systematic arrangement
of the, 210-211.

Mastigonereis lutipalpa. Schmarda, |

261; mentioned, 260.
operta, Stimpson, 261; men-
tioned, 255, 267, 268.
podocirra, Schmarda, 262.
retrodentata, Quatrefages, 261.

Megascops, Aaup, 42.

Melanophora, Oken, mentioned, 227.

Melophagus oyinus, Linn., genitalia
of, 368, 369.

Membracidx, Observations on some
undescribed or little - known
Species of Hemiptera~Homoptera
of the Family, by G. Bowdler
Buckton, 329-338.

Membracis foliacea, Fubr., mentioned,
329.

micans, Buckton*, 330; men-
tioned, 337.

vergens, Buckton*, 330; men- |
tioned, 337. |

Merhippolyte orientalis, Spence Bate,
mentioned, 426. |

Meta segmentata, Clerck, genitalia |
in, 370.

Metachirus, Burm., mentioned, 182, |
184, 185, 187, 192.

opossum, Zinn., mentioned, 111, |
186, 215, 216; upper incisors |
and canine of, 105.
Metapenzeus acclivis, Rathbun, 434;
mentioned, 388, 453.
akayebi, itathbun, 483; men-
tioned, 388, 434, 435, 453.
lamellatus, de Haan, 432; men-
tioned, 388.
stridulans, W.- Mason,
mentioned, 434.

Metatheria, 84,

Michrochrysa polita, Zinn., 341. |

Micoureus griseus, Desm., men- |
tioned, 213. |

Micralestes, Boulenger, mentioned,
58, 62, 75.

acutidens, Peters, mentioned, 5],
65.

altus, Boulenger, mentioned, 51 ;
ovaries examined (Rowntree),
74.

Stormsi, Boulenger, mentioned,
51, 645; ovaries examined
(Rowntree), 74. |

433 ;

TX.

495

Microbiotheriidxe, Miocene of South
America, mentioned, 181, 208.

Microcelis, Plehn, mentioned, 479.

Schauinslandi, Plehn, 474.

Microhierax, Sharpe, mentioned, 29.

Microlestes, Plien, mentioned, 101.

Micropallas, Cowes, mentioned, 1.

Micropezide, 368.

Microschema mucronata, Buckton *,
330; mentioned, 337.

Milichiide, 368.

Monotremata, 84.

Morellia hortorum,
tioned, 343.

Mormyride, 68, 70.

Mounting specimens without pressure,
382-383.

Mouth-armature in Diptera, table
of relationship of .genitalia of,
376.

Mouth-parts in Diptera, diagram of,
375.

Multituberculata, mentioned, 209.

Musca, Linn., mentioned, 347, 348.

corvina, Fubr., 226.

domestica, JLinn., mentioned,
219, 226, 227, 230, 343, 386;
cardines in, 229; in copula,
349 ; ovipositor of, 347, 354—
362; teeth of, 377-382 ;
testes of, 357.

Muscidze, mentioned, 219-229, 341—
353, 3615; types in the, 359.

Muscidee Acalyptrate, 226, 341.

Mycetophilide, mentioned, 341-349,
358.

Mydeine,

Mydea impuneta, Fallén, 225, 230.

Myiocera carinifrons, Mallén, 226.

Fallén, men-

225, 228,

225, 2

| Myletes, Cuv., 73.

brachypomus, Cuv., mentioned,
51.

| Myopa buccata, Linn., 368.

Myra, Leach, 397.
carinata, 7’. Bell, mentioned, 397.
coalita, Hilgend., mentioned,
397.
dubia, Miers, mentioned, 397.
fugax, Fabr., 397; mextioned,
388.
pentacantha, Alcock, mentioned,
397.
Myrmecobiine, 91, 99.
Myrmecobius, Waterh., mentioned,
90-118, 131, 164, 168, 171, 180,
190-194, 207.
72

496

Myrmecobius fasciatus, Waterh.,
mentioned, 212, 214, 216; ante-
molar teeth of, 105, 106.

Myrmica rubra, Zinn., ovipositor or
sting in, 250, 252.

Nacella, Schum., mentioned, 275,
282.
Nacellide, mentioned, 276.
Nannocharax, Giinther, mentioned,
52, 65.
niloticus, Joarnis, mentioned,
52.
Nannostomus, vead Nanostomus,
Giinther, mentioned, 59.
lateralis, Boulenger, mentioned,
51.
Nauticaris unirecedens, Spence Bate,
423; mentioned, 424, 425.
Nautilus, Breyn., mentioned, 271.
Neanthes, Kinberg, 261.
acuminata, Hhlers, mentioned,
261.
capensis, Willey *, 261; men-
tioned, 255, 267, 268.
crucifera, Grube, mentioned,
261.
latipalpa _brevicirris,
mentioned, 261.
typica, Kindberg,
mentioned, 255, 267.
Necrophorus interruptes, Steph., 372.
Nematus, Jur., genitalia in, 372.
Nemopoda, MJacg., mentioned, 343,
348, 351, 352.
cylindrica,
226, 230.
Neoborus, Boulenger, mentioned, 59,
62, 64.
ornatus, Boulenger, mentioned,
51.
Nepa cinerea,
252.
Neptunus (Amphitrite) hastatoides,
de Man, 391.
(Hellenus) hastatoides, Alcock,
391.
Nereis brevicirris, Grube, mentioned,
261.
latipalpa, Schmarda, 260,
mendax, Stimpson, 262.
operta, Stimpson, 261,
Stimpsonis, Grube, 262.
Neuronia popularis, Fabr., genitalia
ot, 372.

Grube,

261;

Fabr.,

mentioned,

LTinn., mentioned,

INDEX.

Ninox, Hodgs., mentioned, 13, 24,
26, 35, 40-42,
connivens, Lath., mentioned, 32.
Norellia spinimana, Fallén, men-
tioned, 342, 348, 351, 385; ap-
pendages dec. of, 353, 385; ovi-
positor of, 361, 385, 386.
Notidanus, Cwu., 76.
Notiphila nigricornis, Stenhammar,
penis of, 345, 358, 386.
Notopteride, 70.
Notopterus, Lacép., 77.
borneensis, Bleck., 68, 81.
kapirat, Lacép., 68, 81.
Notoryctes, Stirling, mentioned, den-
tition of, 117-122, 171-173, 207.
typhlops, Stirling, mentioned,
212-217.
Notoryctide, mentioned, 110, 117,
163, 171, 192, 197, 198, 210.
Nototherium, Owen, mentioned, 158-
160.
Nyctala, Brehm, skull of the nest-
ling of, 2; mentioned, 24, 30-42.
Tengmalmi, Gmelin, skull of, 8,
45.
Nyctaline, 38, 39.
Nycteribia Dufourii, Westw., men-
tioned, 369.
Westwoodii, Kolenati, forceps of,
369.
Nycteribiidee, 369.

Ocypode (Elamene) wnguiformis, de
Haan, 396.
Odonata, genitalia in, 370.
(Kstride, 219, 368.
(Kstrus, Zinn., mentioned, 227.
Oliviera lateralis, Fabr., mentioned,
342, 349, 351; appendages &e.
of, 353, 375, 386.
Oncodolambrus, de Man, 388; men-
tioned, 387.
predator, de Man, 389; men-
tioned, 388, 452.
Onychogale, Gray, mentioned, 148,
154-157, 178, 202.
frenata, Gould, mentioned, 179.
lunata, Gould, mentioned, 179.
unguifera, Gould, mentioned,
179.
Ophicentrus minor, var., Buckton,
mentioned, 335,
serpentarius, Buckton*, 335;
mentioned, 338.

Ophyra leucostoma, Wied., men-
tioned, 225, 226, 228, 230; ovi-
positor of, 360, 364.

Orestias Oweni, Cu. g Val., men-
tioned, 69.

Ornithorhynchus, Blum., mentioned,
99, 101, 102.

Orphnephilidee, 367.

Ortalide, 345, 350, 358, 359, 372,
377.

Orthoptera, genitalia in, 370.

Oryzoryctes, Grand., mentioned, 120.

Osmerus, Zinn., 70.

eperlanus, Lacép., 68.

Ostariophyses, mentioned, 48.

Otinotus, Buckton, mentioned, 334.

Ouranorthus *, Buckton, 333.

palus*, Buckton, 333; ‘men-
tioned, 337.

Owls, A Contribution towards our
Knowledge of the Morphology of
the, by W. P. Pycraft, 1-46.

Oxyrhachis, Burm., Amy. et Serv.,
larvee of, mentioned, 336.

Pachyrrhina maculosa, Meigen, 353,
384.

, ejaculatory sac and apo-
deme of, 355, 384-886; penis of,
344-346, 353 ; theca and flagellum
of, 384.

Palemon, Fabr., On some Species of
the Geuus, from Tahiti, Shanghai,
New Guinea, and West Africa, by
Dr. J. G. de Man, 291-327, 387.

acanthurus, Wiegm., mentioned,
299, 300.

altifrons, Hend., mentioned, 446,
448,

asperulus, v. Martens, 293, 295,
296.

brevirostris, Olivier, 427.

Faustinus, Saussure, mentioned,
317.

Hendersoni, de Man, 446.

jamaicensis, Herbst, 309.

var. Vollenhovenii,
Aurivillius, 309.

japonicus, Ortm., mentioned, 411.

latimanus, v. Martens, 296 ;
mentioned, 446, 448.

longirostris, de Man, 409,

ornati, Oliv., 291.

Ortmanni, Rathbun, 409,

paucidens, Hilgd., mentioned,
306, 307, 308, 309, 327.

a

Palemon paucidens, de Haan, 409.

scabriculus, Heller, mentioned, |

446, 448.

spectabilis, Heller, mentioned,
292.

spinimanus, H. Milne-Edw.,314.

styliferus, H. Milne-Edw., 409.

Vollenhovenii, Herklots, 309.

(Eupalemon) elegans, de Man,
mentioned, 296.

(-——) endehensis, de Man,
mentioned, 317.

( ) Foai, Coutiére, 306;
mentioned, 291, 324, 326.

( ) lar, Fabr., 291, 325.

(—) macrobrachion, Herklots,
299; mentioned, 291, 300,
306, 307, 308, 309; mea-
surements of, 320, 324, 325.

) sundaicus, Heller, men-
tioned, 306.

(Macrobrachium) Iheringi,
Orim., mentioned, 320, 324,

) jamaicensis, Herbst, var.

angolensis, de Man*, men-

tioned, 314, 324.

) var. Vollenhovenii,

Herbst, 309; mentioned, 291;

measurements of, 322, 326.

) latimanus, v. Martens,

296; mentioned, 291, 298,

299, 446.

) Olfersii, Wiegm., 314;
mentioned, 291, 315, 320;
measurements of, 323, 324,
326.

(——) sp., 319; mentioned,
291, 324,

(Parapalemon ?) asperulus, v.
Martens, 293; mentioned,
291, 295, 296.

(——) Hendersoni, de Man,
446 ; mentioned, 388, 453.

) Horstii, de Man, men-

tioned, 296.

) javanicus, de Man, men-

tioned, 296.

Palloptera ustulata, Fallén, men-
tioned, 347.

Palpi genitalium, 350.

Palpi, Labial and Maxillary, in
Diptera, by W. Wesché, 219-230.

—— nomenclature of, 350 ; struc-
ture of, 350.

Paludina, Zam., mentioned, 273.

(

(

(

(

(

(
(

INDEX.

| Pandion, Sav., mentioned, 34, 35.

Pandora scutellaris, Meigen, 352.

Pangonia longirostris, Hardwicke,
penis of, 346.

Paranereis, Ainberg, mentioned, 262.

Parapalemon? asperulus, v. Martens,
293; mentioned, 291, 325.

Hendersoni, de Man, mentioned,

453.

Horstii, de Man, mentioned,
296.

javanicus, de Man, mentioned,
296.

Parapeneus acclivis, Rathbun, 434.
akayebi, Rathbun, 433.
curvirostris, Rathbun, 436.
lamellatus, Rathbun, 432.

Parapeneopsis acclivirostris, Alcock,

mentioned, 454.
tenellus, Spence Bate, 485 ;
mentioned, 388.
Paraphago, Boulenger, mentioned, 62,
65.
rostratus, Boulenger, mentioned,
51.

471, 477.

Man *, 440; mentioned, 387.
Hendersoni, de Man, 446.

tioned, 388, 442, 443,
Parathelphusa, Hdw., mentioned,
387.
endymion, de Man *, 442; men-
tioned, 388, 453.
Langi, Doflein, mentioned, 453.
sinensis, H, Milne-Edw., men-
tioned, 440.
spinescens, Calm., 441.
tridentata, H. Milne-Edw., men-
tioned, 440.
Parmenis, Malmgren, 258.
capensis, Willey *, 258; men-
tioned, 255, 267.

tioned, 258.
Parodon, Val., 73.
Parthenopoides, Miers, mentioned,
388.
pteromerus, Ortm., 391.
Parydra coarctata, Fallén, 345, 347,
350, 366; labrum of, 374, 377,
286.
Patella, Linn., 276, 278, 284.

Paraplanocera, Laidlaw, mentioned, |

Parapotamon, noy. subgenus, de |

spinescens, Calm., 441; men-

Ljungmani, Malmgren, men- |

497

Patella cxrulea, Zinn., mentioned,
275, 282, 289.
magellanica, Sowerh., mentioned,
275.
ornata, Dillwyn, mentioned, 275,
282.
radians, Gmelin, mentioned, 282.
vulgata, Zinn., mentioned, 275,
282, 286, 289.
Patellide, mentioned, 284.
Patina pellucida, Risso, mentioned,
79, 282,

Pegomyia bicolor, Wied., mentioned, .
226, 230; ovipositor of, 361, 366,
386.

Peneeus, Fabr., 432.

affinis, de Haan, mentioned,
438.

anchoralis, Spence Bate, 436 ;
mentioned, 437, 488.

barbatus, de Haan, mentioned,
438, 439.

crucifer, Ortm., 435.

curvirostris, Ortm., 4386; men-
tioned, 437, 439.

eusis, de Haan, mentioned, 439.

granulosus, Hasw., mentioned,
438.

lamellatus, de Ruan, 432.

monoceros, de Haan, mentioned,
437, 439.

tenellus, Spence Bate, 435;
mentioned, 454.

velutinus, Spence Bate, 433.

(Metapenzus) acclivis, Rathbun,
434; mentioned, 388, 453.

) akayebi, Rathbun, 433 ;
mentioned, 388, 434, 435,
453.

(——) lamellatus, de Haan,
432; mentioned, 388.

(Parapeneopsis) tenellus, Spence
Bate, 435 ; mentioned, 388.

(Penzus) brasiliensis, Zatr.,450;
mentioned, 388.

) eanaliculatus, Oliv., var.

australiensis, Spence Bate,

mentioned, 449.

) latisuleatus, Kishinouye,

var., 448; mentioned, 388,
449, 453.
(Trachypenseus) curvirostris,
Stimps., 436 ; mentioned,
388, 437, 439, 453.
Peralestes, Owen, mentioned, 180,

72*

(

(

(

L98

Perameles, Geoffr., mentioned, 85, |

86, 110, 111.
Barrowensis, 7omas,meutioned,
175, 196.
Bougainvillei, Quoy & Gain.,
molars of, 89; mentioned,

111-116, 122, 175, 176, 191- |

197, 215-217.

Broadbenti, Rams., mentioned,
aI bale

Cockerelli, Fams., mentioned,
111, 113, 174, 175, 191-196,
217.

Doreyana, Quoy § Gaim., men-
tioned, 111-116, 175, 191-
196, 215, 216.

Gunni, Gray, mentioned, 111-
114, 175, 176, 195-197.

longicaudata, Peters & Doria,

mentioned, 111.
macrura, Gould, mentioned, 111-
114, 175, 195, 196.
moresbyensis, tams., mentioned,
111, 114, 175, 195, 196.
nasuta, Greoff., mentioned, 111—
114, 175, 176, 191-196.
obesula, Shaw, upper incisors of,
105; mentioned, 111-116,
175, 191-197,.215, 216.
Raffrayana, Milne-Edw., men-
tioned, 111-114, 174, 175,
195, 196, 217.
Peramelide, 97, 102, 106-122, 142,
163, 174-198, 204-210.
Perameling, mentioned, 110-115.
Peramys, Lesson, mentioned, 120,
181-200.
americana, Miller, 215; men-
tioned, 183.
brevicaudata, Hrevl., 216.
domestica, Wagner, mentioned,
182.
Theringi, Thomas, 215; men-
tioned, 183.
Peratherium, Aym., mentioned, 89,
90, 97, 108, 181-208.
affinis, Gervais, mentioned,
ftnote 187.
arvernensis, Gervais, mentioned,
ftnote 187, 189.
Aymardi, Filhol, 215; men-
tioned, ftnote 189.
exilis, Gervais, mentioned,
189.
Lamandini, Filhol, mentioned,
ftnote 187.

INDEX.

Peratherium sp., 214.
Perca, Linn., mentioned, 50.
Percopside, 69, 70.
Percopsis guttatus, Agass., 69.
Perinereis mendax, Stimpson, 262 ;
mentioned, 255, 267, 268.
Periplaneta, Burm., mentioned, 371,
374.
orientalis, Zinn., mentioned, 348,
349; genitalia of, 371.
Petauroides, Zhomas, 128 ;
tioned, 185-137, 141, 198.
volans, V’homas, mentioned, 136,
140.
Petaurus, Shaw, mentioned, 86, 127—-
134, 143, 147, 198, 199.
australis, Shaw, mentioned, 127.
breviceps, Waterh., mentioned,
127, 131, 170.
Doreyana, Quoy ¢ Gaim., men-
tioned, 131.
sciureus, Shaw, 215 ;
tioned, 116, 127.
Petersius, Hilgendorf, mentioned, 53,
64, 75.

men-

men-

tioned, 51; ovaries examined
(Rowntree), 74.
Petrocephalus, Mare., mentioned, 68,
Mie
bane, Lacép., mentioned, 81.
Petrogale,
154-157, 178, 179, 202.

tioned, 51.
loricatus, Giinther, mentioned,
. ol.

Phalanger, Storr, mentioned, 114,
125, 126, 133, 142, 156, 162, 170,
191, 198-204.

breviceps, Thomas, mentioned,
132, 133.

celebensis, Gray, mentioned,
132, 133, 170.

leucippus, Thomas, mentioned,
133.

Leopoldinus, Boulenger, men- |

Gray, mentioned, 148, |

brachyotis, Gould, mentioned,
179.
concinna, Gould, mentioned, |
179.
inornata, Gould, mentioned, |
179.
penicillata, Gray, mentioned,
179.
Phago, Giinther, mentioned, 52, 59,
62, 65.
Boulengeri, Schilthuis, men-

| Phalanger lullule, Thomas, men-
tioned, 133, 170.
| maculatus, Gr'eoffr., mentioned,
133.
| melanotis, 7’homas, mentioned,
| 131, 133.
| orientalis, Pallas, 216; men-
tioned, 133, 170.
| Rothschildi, Thomas, mentioned,
133.
ursinus, Zemm., mentioned, 131,
132, 133.
Phalangeride, 97, 98, 109, 114, 116-
| 125, 135, 142, 163, 169-179, 192,
| 198-210.
Phalangerine,125~-134,142, 158,202.
Phascogale, Zenm., mentioned, 91,
94-97, 107, 168, 166, 171, 182-
| 186,198.
apicalis, Gray, mentioned, 97,
98, 165.
calura, Gould, mentioned, 96,
98, 124, 165.

Doris, U’homas, mentioned, 204.
| dorsalis, Peters §° Doria, men-
tioned, 98, 165, 186, 204.
flavipes, Waterh., 216; mention-

ed, 97, 98, 165, 166, 186-193,
206, 216.
Gunni, Gray, 206.
Macdonnellensis, Spencer, men-
tioned, 164.
minima, Geoffr., mentioned, 98,
165,190; a Tasmanian form,
206.
minutissima, Gould, mentioned,
93, 97, 98, 186.
nasuta, Geoffr., 206.
| obesula, Shaw, 206.
penicillata, Zemm., mentioned,
96, 98, 105, 124, 165, 182—
190.
Swainsoni, Waterh., mentioned,
98, 165, 186, 206.
Thorbeckiana, Schlegel, men-
tioned, 91, 97, 98, 165, 186,
204,
viverrinus, Geoffr., mentioned,
206.
Wallacei, Gray, mentioned, 91,
94, 97, 98, 165, 204,
Phascolarctine, 116, 125-142, 161,
202.
Phascolarctus, Blainv., mentioned,
125, 135, 136, 141, 161, 170-1738, .
181, 190-200,

|
4

ee eee Seem S”—S” SF

Phascolarctus cinereus, Gloldf., 215,
216.

Phascologale cristicauda, Arefft, men-
tioned, 213.

Phascolomyide, 116, 125, 142, 151-
173, 191-198, 203, 211.

Phascolomys, Geoffr., mentioned,
158, 159, 161, 170-174, 181.
Mitchelli, Owen, 215, 216,
217.
Philander, Briss. { = Caluromys|,

182.

|

Phito, Rond., 227.

Jardaria, Fabr., mentioned, 227;
wing of, 227.

melanocephala, Meigen, wing of,
227, 230.

Phitomyzide, ovipositor in, 359. |

Phorbia floccosa?, Wacq., 226.

Phoride, 227, 368.

Phorocera concinnata, Meigen, ovi-
positor of, 364.

serriventris, Rond., abdomen of,
386 ; ovipositor of, 364.

Photodilide, 14.

Photodilinz, 38.

Photodilus, Geoffr., cranium of, 3;
skeleton of, 2; mentioned, 7, 14,
36, 38, 40.

badius, Horsf., mentioned, 43, |
fig. 45, 46.

Phronia, Winn., mentioned, 339, 342, |
343.

Phycodromide, 368.

Phyllodoce, Sav., mentioned, 255,

sp., 259.

Physostomi, Relations of the Ductus
Pneumaticus in the (W. 8. Rown-
tree), 47-81.

}

Phyto melanocephala, Meigen, men-

tioned, 343.

Pimpla Fairmairei, Laboulbéne, men-
tioned, 240.

Pinnotheride, mentioned, 387.

Piophilide, 368.

Pipunculide, 367.

Placentalia, mentioned, 84;
structure in the, 162.

Plagiaulacide, 180,

Plagiaulax, Fuale., mentioned, 101,
180.

Planocera, Bluinv., mentioned, 468,
471.

foot-

Graffii, Lang, mentioned, 478.
Planocerid, discovery of a new, men-
tioned, 466.

INDEX.
Planoceridx, mentioned, 468, 471.
Platybema planirostre, Rathbun,
421.

Platychirus, St.-Farg., 370.
Platygrapsus, Stimps., 392.
depressus, de Haan, 392; men-
tioned, 388.
depressus, Ortm., 392.
Platylambrus, Stimps., mentioned,
388.

Platynereis striata, Schmarda, 262 ; |

mentioned, 255, 267, 268.
Platynotus depressus, de Haan, 392.
Platypezidee, 367.

| Pleurotomaria, Defr., mentioned, 270,

272, 286.
Pecilobothrus, Mik, mentioned, 342,
3al.
nobilitatus, Zinn., mentioned,
384; ovipositor of,-362; pa-
pilla of, 386.
Polietes, JZtond., mentioned, 227,
228,

lardaria, Fubr., mentioned, 225, |

230; ovipositor of, 359-364.
Pollenia rudis, Fubr., mentioned, 343,
348, 351, 353, 385.
Polocentrus caudatus, Buchkton*, 335;
mentioned, 338.
labatus, Buckion *, 335; men-
tioned, 338.

Polycelis australis, Schmarda, men- |

tioned, 471.

Polychzta, Littoral, from the Cape
of Good Hope, by Arthur Willey,
255-268.

Polyclads, Observations on Austra-
lasian, by Prof. W. A. Haswell,
465-485.

Polynoe attenuata, McIntosh, 257.

scolopendrina, 257 5
mentioned, 255, 258, 267.

Polyporus, Plehn, mentioned, 468.

Polyprotodontia, mentioned, 122,
207, 208.

Polypterus, Geoffr., mentioned, 47,
65, 66, 70, 72.

Savigny,

Poppea concinna, Fowler, mentioned,

331.
succinea, Buckton *, 331 ; men-
tioned, 337.
Portunus (Amphitrite) hastatodes, de
Haan, 391.

Potamogale, Du Chaillu, mentioned, |

120.

499

| Potamon, Savigny, 440.
Lanzi, Doflein,
445.
spinescens, Calm., 441; men-
tioned, 387, 442, 443, 454.
(Parapotamon)spinescens, Calni.,
441; mentioned, 388, 442,
443, 444, 445.
(Parathelphusa) endymion, de
Man, 442; mentioned, 388,
443-445, 453.
) Lanzi, Doflein, men-
tioned, 442, 444, 445, 453.
Potamonide, 440.
Potoroinee, 143-150, 200.
Potorous, Desm., 148, 147, 177, 178,
200, 201, 206.
apicalis, Gould, mentioned, 147—
150, 206.
Gilberti, Gould, mentioned, 147—

442, 444,

(

150, 206.
platyops, Gould, mentioned,
147-150, 201, 206.
rufus, Higg. & Pett., mentioned,
149,
dwarfed Tasmanian form,
mentioned, 149.
tridactylus, err, mentioned,
147-150, 206, 217.
Prochilodus, Agass., mentioned, 53,
57, 59, 62, 64, 73, 76.
lineatus, Cuv. g Val.,
tioned, 52, 57, 58, 81.
Procoptodon, Owen, mentioned, 143,
156, 160.
Properamelidee, 192.
Prorhynchus, Schultze,
477.
Prosthecerzeus, Schmarda, mentioned,
481.
anomalus*, Hasw., 481.
flavomaculatus, von Graff’, men-
tioned, 465.
Prosthiostomum, Stimps., mentioned,
483.

men-

mentioned,

Cooperi, Laidlaw, mentioned,
483.

Dohrnii, Lang, mentioned, 48:3.

elegans, Laidlaw, mentioned,
483.

maculatum *, Hasw., 482 ;
tioned, 485.
siphunculus, Delle Chiaje, men-
tioned, 482, 483.
Prostreptoneure, mentioned, 27 |,

men-

oro?

-i-.

Potamogalids, mentioned, 119, 121. | Prothylacinide, 109, 181.

500

Prothylacinus, Ameghino, mentioned,
108.
Protocalliphora groenlandica, Zett.,
ovipositor of, 360, 364.
Protopterus, Owen, mentioned, 47,
76.
annectens, Owen,
ftnote 70, 80.
Prototheria, 84.
Pseudoceros, Lang, mentioned, 480,
481.
cardinalis, Hasw.*, 480; men-
tioned, 481, 485.
dimidiatus, von Graff, mentioned,
465.
Kentii, von Graff, mentioned,
465.
limbatus, Hasw.*, 480.
Pseudochirus, Ogilby, 89, 125, 135,
141, 161, 170, 198-202.

mentioned,

Albertisi, Peters, mentioned, 137, |

140, ftnote 199.

Cooki, Desm., mentioned, 136,
140, ftnote 199.

Corinne, mentioned, 140, ftnote
199,

cupreus, mentioned, 140.

Dahli, Collett, mentioned, 136,
140.

Forbesi, Zhomas, mentioned,
136, 140,170; molars of, 89.

occidentalis, 7homas, mentioned, |

140.
peregrinus, Bodd., mentioned,
136, 187, 140, 215, 216.
Pseudonereis, Kinberg, 262,
anomala,
ftnote 262.
Psilide, 368.
Psilopa sipho,
349,
Psychodidee, 367.
Pteromaline larve, mentioned, 240.
Pteromalus, Swederus, mentioned,
240.
Ptychoptera albimana, Fabr., geni-
talia of, 360, 361, 367.
scutellaris, Meigen, mentioned,
342, 346, 351, 355, 384.
Pulex irritans, Linn., mentioned, 366.
Pulicide, 366.
Pulsatrix, Kaup, 36, 40, 42.
Pycraft, W. P., A Contribution to-
wards our Knowledge of the Mor-
phology of the Owls.—Part II.
Osteology, 1-46.

Wied.,

Gravier, mentioned, |

mentioned, |

INDEX.

Pyrrhulina, Cuv. & Val., mentioned,
54, 61, 63, 64.
semifasciata, Steindachner, ova-
ries examined in (Rowntree),
74.
Pyrrhulina unifasciata, »ead semifas-
ciata, Steindachner, mentioned, 51.

Receptacula seminis, 359, 364.

Rhynchocyclus planirostris, Miers,
421,

Rhyphide, 344, 358.

Rhyphus fenestralis, Scop., mentioned,
341, 344,

punctatus,
341, 344.

Rodentia, mentioned, 159.

Rowntree, W. S., On some Points in
the Visceral Anatomy of the Cha-
racinide, with an Enquiry into
the Relations of the Ductus Pneu-
maticus in the Physostomi gene-
rally, 47-81.

Rutilia splendida, Guérin, mentioned,
342; forcipes superiores of, 375.

Fabr., mentioned,

Sacculus ejaculatorius, structure of,
354.

Saissetia hemispheriea,
mentioned, ftnote 233.

Cockerell,

Salminus, Agass., mentioned, 53, 54, |

62, 64, 73,
maxillosus, Cuv. & Val., men-
tioned, 51, 80.
Salmo, Zinn., mentioned, 70.
fario, Linn., 68.
trutta, Linn., 68, 81.
Salmonide, 71.
Sarcodaces, Giinther, mentioned, 50-
538, 59-64, 73, 77, 78.
odoé, Bloch, mentioned, 49, 51,
81; ovaries examined (Rown-
tree), 74, 80.
Sarcophaga, ZLinn., mentioned, 343,
347, 350, 385.
carnaria, Linn., mentioned, 345,
348, 349, 351, 353, 385.
Sarcophagidee, 226,
Sarcophilus, Cuv., mentioned, 91-99,
168, 173, 180, 193, 194, 206.
latifrons, Fallén, 343.
ursinus, Harr., mentioned, 91,
94, 97, 98, 164, 214, 216,
217.
Scalibregma inflatum, Rathke, men-
tioned, ftnote 266,

Scatophaga, Meigen, mentioned, 224,
228, 340.
Scatophaga carnaria, Zinn., mention-
ed, 340.
litorea, Fallén, mentioned, 343,
351; appendages &e. of, 353,
385.
lutaria, Fabr., genitalia of, men-
tioned, 339, 349, 386.
stercoraria, Linn., genitalia of,
mentioned, 339; ovipositor
of, 361, 366, 377, 386.
Scatopse, G'eoffr., mentioned, 341.
notata, Zinn., mentioned, 342,
344-351, 355, 364; genitalia
of, 382, 383.
Sceloglaux, Kaup, mentioned, 1, 40.
Scenopinide, 367.
Scenopinus fenestralis, Linn., 367.
Schenomyza cinerella, Mallén, recep-
taculum of, 386.
Schilbe mystus, Zinn., mentioned,
67.
Sciara, Meigen, mentioned, 341, 342.
thome, Zinn., mentioned, 344,
349; genitalia &c. of, 383.
Sciomyza cinerella, Fullén, genitalia
of, 368.
Sciomyzide, 368.
Scissurella, @’Orb., mentioned, 269,
270, 272, 286.
Sclerocrangon, (. O. Sars, 408.
angusticauda, De Haan, 408 ;
mentioned, 388.
Scops, Mochr., mentioned, 18, 28, 34,
40, 42; skeleton of, 2.
asio, Linn., mentioned, 7.
rutilus, Pucher., mentioned, 7.
semitorques, Schlegel, mentioned,
ih
Scotiapex, Swazns., mentioned, 1, 36,
39.
Scotopelia, Bonaparte, mentioned, 1.
Scott, Hugh, On Cercococcus eremo-
bius *, gen. et sp. noy., an Aber-
rant Form of Coccide, 455-464.
Scurria, Koch, mentioned, 275, 282,
Scutellina galathea, Zam., mentioned,
275.
Seioptera vibrans, Zinn., receptacula
of, 366, 386.
Sepsidee, 343, 348, 353, 354, 358.
Sepsis cynipsea, Linn., mentioned,
348, 351, 352, 366, 385, 386.
Serrasalmo, Zacép., mentioned, 52,
59-63, 73, 78.

Serrasalmo humeralis, Cuv. & Val.,
mentioned, 51, 59, 64, 80.
piraya, Cuv., mentioned, 51, 59,
64,
Setonyx, Zesson, mentioned, 143,
202.
brachyurus, Quoy g: Gaim., men-
tioned, 179.
(= Macropus brachyurus, Quoy
g§ Gaim.), mentioned, 151.
Sicus ferrugineus, Linn., 368.
Sicyonia, H. Milne-Edw., 450.
carinata, Olivier, 451; men-
tioned, 388.
sculpta, H. Milne-Edw., var. ?,
450; mentioned, 388; cap-
tured off Bahia, 387.
Siluranodon, Bleek., mentioned, 67.
auritus, Gieoffr., 67.
Siluride, 48, 70, 71.
Simulide, 226, 363, 367.
Simulium, JZatr., mentioned, 222,
363, 369.
ornatum, Meigen, 367.
reptans, Linn.?, ovipositor of,
363, 367.
Siphona, Zeét., mentioned, 221.
geniculata, De Greer, 224, 226.
Sminthopsis, Thomas, mentioned, 91—
97, 164, 166, 184, 187, 193, 194.
crassicaudata, Gould, mentioned,
93, 98, 166, 167, 187, 216,
217.
hirtipes, Thomas, 98, 167, 217.
larapinta, Spencer, mentioned,
166, 167.

leucopus, Gray, mentioned, 92- |

98, 111, 112, 166, 186-193,
214-216.

Macdonnellensis, Spencer, men-
tioned, 98.

macroura, Gould, mentioned,
98.

murina, Waterh., mentioned, 97, |

98, 166, 190.
Solenodontide, 119.
Soricide, 96, 116, 124, 209.
Sparassodonta of Ameghino, men-
tioned, 90, 181, 208.
South American Miocene, men-
tioned, 194.
Speotyto, Gloger, mentioned, 7, 24—
29, 39-42; skeleton of, 2.
cunicularia, Molina, skull of,
15; mentioned 45, 46.
Spherocera denticulata, Meigen, 226.

INDEX,

Spheerocera subsultans, Fabr., men-
tioned, 226, 227, 230.
Spheerophoria scripta, Zinn., men-
tioned, 345, 353.
Sphongophorus, Fuirm., mentioned,
330.
Spiracle-bearing segment, 363.
Spirontocaris, Spence Bate, 411;
mentioned, 387.
alcimede, de Man, 416; men-
tioned, 388, 418, 453.
amabilis, Zenz, mentioned, 416,
418.
flexa, Rathbun, mentioned, 416,
417.
gracilis, Stimps., mentioned, 416,
417,
pandaloides, St¢mps., 418; men-
tioned, 387, 388, 414, 453.
propugnatrix, de Man, 414;
mentioned, 388, 416, 418,
453.
rectirostris, Stimps., 411; men-
tioned, 387, 388, 413, 452.
stylus, Stimps., mentioned, 416,

419.

tridens, Rathbun, mentioned,
416.

unalaskensis, Rathbun, mention-
ed, 416.

Squilla affinis, Berthold, 439.
fasciata, de Haan, 440.
Steatornis, Humb., 38.
Stemonyphantes lineatus,
genitalia in, 370.
Sternopygus carapus, Zinn., men-
tioned, 67, 81.
virescens, Valenc.,mentioned, 67.
Sthenelais fuliginosa capensis, Ola-
parede, 259; mentioned, 255.
Stomatopoda, 439.
Stomoxys, Geoff., mentioned, 225,
377, 378.
calcitrans,
343, 365.
Stratiodrilus, Haswell, mentioned,
471.
Stratiomyiide, 341-344, 353, 361-
364,
Stratiomys, Geoffr., mentioned, 359,
363.
chameleon, ovipositor of, 361
362.
Striges, skull of, mentioned, 2.
Strigide, skull of, 3, 14, 39-43.
Strix, Zinn., mentioned, 7, 18, 34-42.

Menge,

Zinn., mentioned,

,

501

Strix delicatula, Gould, mentioned,
27, 37.
flammea, Zinn., mentioned, 37,
45.
javanicus, Gmel., 38.
perlatus, Véerll., 38.
poensis, Fraser, 37.
pratincola, Bonaparte, 37.
Tengmalmi, Gmelin, mentioned,
43,
Stylochoplana, Stimps., mentioned,
471.
Stylochus, Zhr., mentioned, 471.
vigilax, Laidlaw, mentioned,
465.
Submentum, relation of the great
apodeme to the, 373-379.
Surnia, Dumér., rostrum of, 5; men-
tioned, 14, 27, 29, 37-42.
aluco, Zinn., mentioned, 5.
funerea, Heugl., mentioned, 5.
Synodontis gambiensis, Giinther §
Playf., 67.
Syritta, St.-Farg. et Serv., 363, 364.
pipiens, Linn., mentioned, 341—
353 ; ovipositor of, 362-365,
372, 385.
Syrnium, Sav., skeleton of, 2, 18-42.
aluco, Zinn., mentioned, 5, 36;
skull of, 1, 5, 45, 46.
seloputo, Gray, mentioned, 36,
uralense, Pallas, mentioned, 36.
Syrphide, mentioned, 221-227, 341-
367, 375.
Syrphus balteatus, De Geer, men-
tioned, 221.
pipiens, Zinn., mentioned, 342,
359.

Tabanide, 219, 226, 342, 344, 358,
| 363, 364, 378.
Tabanus, Linn., mentioned, 219, 221,
230, 369; ovipositor of, 358,
bovinus, Linn., ovipositor of,
361, 386; mentioned, 365.
bromius, Zinn., mentioned,
ftnote 219, 221, 222, 341-
346, 355; genitalia of, 386.

italicus, Meigen, mentioned,
ftnote 219.
sudeticus, Zeller, mentioned,

ftnote 219, 221, 222.
Tachina, Meigen, mentioned, 224,
Taloipa, Buckton*, 334.

tinctoria, Buckton*, 334; men-
tioned, 338,

502

Talpa, Linn., mentioned, 120.
Talpide, 110, 119, 121.
Tarsipedine, 125-134, 135, 163.
Tarsipes, Gerv., mentioned, 125, 163,
169, 170, 198.
rostratus, Gerv. g Verr., 216;
mentioned, 134.

Tegule, 227.

Tetragonopterus, Seba, mentioned,
54, 62, 76.

abramis, Jenyns, mentioned, 51,
53, 56, 65, ‘75; ovaries exa-
mined in (Rowntree), 74. ©

argentatus, Cuv., mentioned,
51, 55, 64.

fasciatus, Cuv., mentioned, 49,
51, 55, 56, 64.

maculatus, Zinn., mentioned,
49.

melanurus, Bloch, mentioned,
49,

multiradiatus, Giinther, men-
tioned, 51, 65.

Thalamoplana, Laidlaw, mentioned,
479.

Therevide, 367.

Thylacinine, 91, 107.

Thylacinus, Zemm., 90, 91, 107,
108, 109, 164, 169, 171, 180,
190, 198, 194, 205, 206, 209,

eynocephalus, Harr., mentioned,
107, 108, 190, 214, 216.

speleeus, Owen, 108, 214.

Thylacoleo, Owen, mentioned, 161,
180.

Thylacoleontide,
211.

Thylacomyinz, 110, 115.

161-163, 192,

Thylacomys, Owen, mentioned, 110, |

114, 117, 126, 127, 176, 195,
207.
lagotis, Reid, mentioned, 115,
118, 176, 189.

INDEX.

| Thylacomys leucura, Thomas, men-

tioned, 112,115, 116,176, 215-217.
minor, Spencer, mentioned, 115.

| Thymallus vulgaris, Zinn., 68, 70.

Thysanozoon, Grube, mentioned, 484.

Tipula, Zinn., mentioned, 226; ovi-
positor of, 358.

oleracea, Linn., mentioned, 341-
346, 351, 352, 355; genitalia
of, 360, 377; sense-organs
of, 384.

Tipulide, 341-349, 358, 360, 364.

Toxoneura muliebris, Harris, men-
tioned, 347-356, 366, 386.

Trachypenzus curvirostris, Stimps.,
436; mentioned, 388, 453.

curvirostris, Alcock, 436.

Tragopa triangulata, Buckton *, 331 ;
mentioned, 337.

Trapezoida, Buckton *, 335.

hirsuta, Buckton*, 335; men-
tioned, 338.

Trichocera hiemalis, De Geer, geni-

talia of, 360, 367.

Trichosurus, Zess., mentioned, 125-
134, 142, 156, 158, 170, 191-199;
molars of, 89.

vulpecula, Kerv, mentioned, 112,
116, 215, 216.

Triconodon, Ameghino, mentioned,
180, 181.

Trigonoplax, H. Milne-Edw., 396.

unguiformis, de Haan, 396;
mentioned, 388.

Trigonoporus, Lang, mentioned, 468,
A479.

dendriticus, Verrill, mentioned,
468.

folium, Verrill, mentioned, 468,

Tripylocelis*, Hasw., 466; men-
tioned, 468, 479.

typica*, Hasw., 466 ; mentioned,

| 468, 484,

Tritodynamia, Ortm., mentioned, 393,
396.
Japonica, Ortm., mentioned, 395,
396.
Tritomodon, mentioned, 180.
Trochide, 275.
Trochus, Linn., mentioned, 269, 270.
cinerarlus,

273.

Linn., mentioned,
Trypetide, 345 ; ovipositor of, 359.

Ulidia nigripennis, Linn., 386; ovi-
positor of, 362-364, 386; penis
of, 347-350, 386; receptacla of,
366.

Varicorhinus beso, Riippell, men-.

tioned, 66.
Vasa efferentia, 357.
Vesicule seminales, 357.
Vespa, Linn., cardines in, 229.
vulgaris, Linn., ovipositor or
sting in, 250.

Wallabies, mentioned, 154-157.
Large, 154-157.
Small, 154-157.

Wesché, Walter, The Labial and’

Maxillary Palpi in Diptera, 219-
230.

The Genitalia of both the
Sexes in Diptera, and their Rela-
tion to the Armature of the Mouth,
339-386.

Willey, Arthur, Littoral Polycheta
from the Cape of Good Hope, 255-
268.

Wyniardia bassiana, Spencer, men-
tioned, 200, 213.

Xenocharax, Giinther, mentioned, 53,
54, 59, 62, 63, 64, 73, 76, 78.
spilurus, Giinther, mentioned,
52, 57.

PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET,

Se a ee

Hg PE t Pi
2nd Ser, ZOOLOGY. ] 7 (VOL. IX. PART 1.

THE

TRANSACTIONS

\\° OF ,

THE LINNEAN SOCIETY OF LONDON.

A CONTRIBUTION TOWARDS OUR KNOWLEDGE) OF THE
MORPHOLOGY OF THE OWLS—PaRtT II. OSTEOLOGY.

BY

W..P. P¥CREART, A.LS,y. F.Z.8.

fon. DON:

PRINTED FOR THE LINNEAN SOCIETY

vee
BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.

‘t
SOLD arne SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

July 1903.

Py

et

. 4 \ 4 ’ ¢
ey Mates: Yo bmey fae

iS ORM i

LINNEAN SOCIETY OF LONDON,

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows 3; to

Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the
Zoological parts of these which have been published are as undermentioned (for the Botanical parts see Botanical

wrapper) :—
Seconp SrerrEs.—Zoonoey,

Walaaet When Price to the Price to
Published. Public. Fellows.
e5 i Gh ae Gy Gh,
iPart We ers cone dl Ze a, 018 0
Part Iu aksyé, Sang O) (3) 0 4 6
ly UES HA aa o Il YO) Song Ib il
Part IY, Aelie goao O 1G O soso 0 12) ©
Part We lisiGiseos Ouls) Or gota @ ilsi 7G
Part WS ice te lane OMe a0 Greg:
Part VU SiS: ase leo, eOmecrmelune70 0)
lieiam NAUUE DUC Geos JO) 0) oob5 Wal (0)
II. Part T1879. nee ea One. OMS 10
Part 1G CH Bas OS O oace Malik 6
liebe SIUGb Ite a Ibs} Ol Goad 1b ib
Part JA, eee woow UP Ye Goch, © & 8
Part We Wetses caoe O 3 O 555, 0 2B
Part Wils TIGER Moo IO O jcon We Oo
Part) VEL O883. es. OlNbn0) 7.4. 0 eo 9
leis NAUMAN Gaga tl) 9S) (ohne @ 2) 3}
Part IDG ERE Goce OM Sh O chant & 2
Part ENG S84 ey Old 16. OM om
Part Cet eae ONO aeoa WO Ye 3
lethye OCU EGIS Good t) GO O sig, O 2
iehyy DONE ee Soa Oe OM ooc4 2b &
Part: Xavi, S855 4...2) OG (Ones 04 1G
Parte, OXVieliSS5i men Ol aoOleet 0 er ao,
Jehg OQ ists Gag Oey Osun o tl). Sy) -e)
leiyn D\VOle Istts Saar) Bi Oe s5cq B'S}
12 vy QUle ists vemeg UO 3 (8 gag OM 2)
Itl. Part MD SBA es ade Ope elie tag
Part ie 1884. 2c. D2) One. Lae 0
Party UaeltSSin were ll Ol Oi rho matG
Part LVe 1885. 0) Oss ee ieag OO
Part We dist Bod UM COs at, OO
Part Wily ister Beas OP OO Sasa OW) ZL
IV. Part ey SSCs re ee LO nnecnean OMe ian()
Part TH: 1887 22 1 8 Op sa SO
Part IND Aelster ipa) iy O) Sane @ Ie o
V. Part itch oa eb) Oo aang O fo
Part JOM eee Bes Ws OW Ses. oe)
Parts “ill. 1889) lea iO eee m0
Part TWVe ASOT Ola Oe Ome nae)
Part Wee L890) cn 36) 20) ar Oe

Vales: When Price to the
Published. Public.
i esd
Wedeiag VAR UGE o55 @ 1] W@ 2
Part, Vide Sollee 0 tGmOme
PartiVllls 89D meee (0) its} (0
Part IX. 118925 5.52 0) 12) 10);
Part) eXe 189352 eal eon Ome
Part, XT) 18945... 0 2) Ge.
Wi Part SoA. oO mOmae
Parts Dl S04 ear, claolile Op
Part TM. 18945.... 0 10 0%
Barba lVenlS9GHin. seems
Part) “Ve L896i20 72 Ol 10s Ole
Part) VieS965. 2,9) OmSim Ome
Part VII. 1896. . 0 121047
Part VIII. 1897. . ON Zao.
VIEL Parte e896: ee Os Ons Ome
letigr es IMIG ISG aoe MI) Oe
leh NOE SIC Sooo OG Oy.
Part IV. 1898. . OA0T0:
Partie) VenlSoS aan OSS Ome
Part VI. 1898. . 018 0.
Part VII. 1899. . 018 0.
PartiVillb 1899) ese (ey
1p 1D. GEE oe LO @ ,
Rartay 2X<51'9 00 eG Ome
Part XG 005mm One
VIII. Part i 190055 0} 10) 0)
Part) > Ei S900Se a0 10) 10m
Part TE 900m O10) 20.
Part IY. 1901. . 014 0.
Part sieaVemloOleemers Oe roe nO
Part VieO0I 0 1O00ne
Part vill GON Ns. cel) RCeOme
Part Vilie902...... 0 4.10.
ParbaXewlo 02.0.7... Ol onmOme
Parte ae O0d. 5... 2 On Om.
Part xan 19038... oO Oke
Part oii 1903. 55 2 OM1L0) Ome
Part XIII. Index. (Jn preparation.)
IX. Part T1903. 2ear0! OO eee
Parti ul; 1903:82.).000) \SiaeOuen oe

Srconp Srrtns.—Zoonoey (continued).

Boo oc ¢ '

Soe Gorey CS Go Gr Sue Oo So a © OC OC Oo oo 6160'S 6

Price to
Fellows,

>

So

i
woo

i
ow

owe ON WHO OD VW

a

DH ORO ow

oo

SFRODTSOCAGHADAD HD SRATSOROAARSCF FCOSAOSAWSOSTDTCSCOCRS Ze

2nd Ser, ZOOLOGY. | (VOL. IX. PART 2.

THE

>}
i=]

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON.

ON SOME POINTS IN THE VISCERAL ANATOMY OF
THE CHARAOCINIDA, WITH AN ENQUIRY INTO THE
RELATIONS OF THE DUCTUS PNEUMATICUS IN
THE PHYSOSTOMI GENERALLY.

BY
WALTER S. ROWNTREE, BSc. F.LS.

LONDON:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.
SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

July 1908.

r
yh

i]

LINNEAN SOCIETY OF LONDON,

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s, to Fellows; to

Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoologioal and Botanical sections, The prices of the
Zoological parts of these which have been published are as undermentioned (for the Botanical parts see Botanical

wrapper) :—
Srconp SeRrEs.— ZooLoey,

Walane When Price to the Price to
Published. Public. Fellows.
ED th Oh a Ge Gh
I. Part I. 1875 ak 40 eae. ORES IO
Part WG SG Sooo ) GH UW) soca OD Zh G
Vghan IOI UEGy Gooa 4) fp OW) ce55 GL ih ©
Part lio Wo cece sy Gaon (0) e}
Part Wo lve coos OS Wacso Ws
Part VI. 1877 ee LD Se Omer O) 1G a6
Barty | VS S78... TO AB og leh 1X0)
Iepyan \WAINEE EER Gaoo 1h OW O spe 1 ©
II. Part lis USWA ao5ue i, 25 WD gé65 Oils} ©
Part II. 1881. Wyss TH Sa6o O Iu! @
ivan JODR EER Gong JG WW aada it il
Part DVS 882) raver OY Bases O & GB
Part Viel SS arc. OY Be OL Oars
Part Wie HEEBE acon ik O aaen Ol
lydia | AUG IneSBE Sha WO) 6) ( Saag @ 8)
legs WADI TBEBR ga56 8) OM aagn WY BY 8}
Part IDS USB coca O BO goon O B 8
Part EX USSA ec Ol 4 Osea OMROnEO
Part > Ob aketee Ss aaod UNO GW Goa OU ye a8
laie = OU eS sooo OW (WW coon WY 4b
Vine 200 ieee ooo OM @ Wsoée, O 2G
1a, OSI IUCREE Sano WS Wi oea5 @ Zhe G
Part) X6V-e1885. 250, ORAS 6 a Ome omn
reer DOIG Istsss page (ON Oo acaa W) gh ©
Parke xvii S86. ci.) On Ome Ono
Ter OUNL, Mette goa O) 2) (8 coon B @
Ti. Part DT, 1884. es, ela 10) Sena eomeG
Part WS T8B4 cee 2 Oe el ret)
lee UNDUE kets aang Ih Wi asan Jk 2 @
Tart ING iets sase 23 Osa g5 U) OA @
Part ie lctsve Same octet Sang UO) GG 0)
Part MAIS Welch oaae WY 1G Osoas Zh
TV. Part Me L886. 1 4) ONS OMlicinO
Part Li etter eer el Geto Oris TL il
Part PMT 1883s ce OMIGH Ones Onl 2iO
We death dy i ilfststs Soc QARE EO Seioa Oy OO
Part Di. S885 3, Oe be Ohare OmmouES
Part TM, US 89. "Sere lt oxfeeiOleer mee lOO
Part TVS S902 eee 0 M20 TORO no ae
Part Vii US90i2). 096 Osa err Olea G

Seconp Srrrns.—Zooroey (continued).

AVadiiie! When Price to the Price to
Published. Public, Fellows.
23 fh ah &: Seve
V. Part Vi. 18915. ... 0 12 10). -.0RNOmeG
Part VIT. 2890... 2.70 G. 10) 5... One
Part VIII. 1892. 0 8 0.... OGaS
Part IX. 1892. ..... 0 12° 0%... OmomeG
Party eae) He oar. lL 8 10... ... ee
Part XI. 1894..... 0 2 6... Oe
Vil Partie 94cm er 2 0 0... 0 0RG
(Partee etl 8 OAc 111 0 5...) 23m
Part IIT. 1894. .... 0 10 0 .. 2.0m
Jers IAW AISI aoc 1 4 0 .... OFISmaD
Part. Vi. 1896; ..... 0! 10) (0) 30a
Part VI. 1896, .... 0 8 0.5022) OGM
Party Ville SOGsre 0 12 0.2.2 ORS
Part VII. 1897. 4... 0 2c 6 22. 0meomee
VIL. Part T, 1896. 2... 0 10° (0° .250i ee
Part 1. 1897...... 10 12 0.325 (O=OmEG]
Part (iI. 1897. ..., @ 6 (0) 2.22 Ome
Part) TV. 1898. :.... 010) 0) 3). Omens
Part. Vi 1898..... 0 18 0%... 0mISmEG
Part VI. 1898. .... 0.13 0.2... 50 Moms
Part VII. 1899: ...... 0:18 10035. 5 Onlmne
Part VUIT.1899...°... 0 120) 2555) Om omen
Part) (LX 899s. 1 0 (0.42. 50gTS aa
Pari Xe l9008e 0 6 Op OeeaG
Part, “Xd; 1900aee. 0 2,9 Oem
VIII. Part P1900). .;...0 10 OO meee
Part) I 19005... 0: 10 (OO meta
Part IIS00N.:<. 0: 10) SOR Ome 7a
Part’: TVs 190k... 0 14° (0nseeeOMOMeS
Part WV. 1901..... 0 5 O220. 0) Saag
PartaayemloOl rset. 0 10) O82. (On 7ama
Part Wil. 1900. .:..|, 2) SalOieee ella
Part Vill, 1902. ..... 0 4) Omen OmmnO
Parte ex, 1902: ..... 0) > sO Oem
Part eek. 1908... 3 D (OOM Oem
Parb eel 1903.)4.... 10 CaO er eeO ma
Bartexit, 1903577... (OOP S0ie. Oma
Part XIII. Index. (In preparation.)
IX. Part T.. 1908: ... 0 SO MOR ee0 womens
Part) “20> 1905.5. 32, Ossi sO eo OMe

FOYE
2nd Ser, ZOOLOGY.) (VOL. IX. PART 3,

MSAD

TRANSACTIONS

OF

ON THE EVOLUTION OF THE AUSTRALIAN MARSUPIALIA; WITH
REMARKS ON THE RELATIONSHIPS OF THE MARSUPIALS IN
GENERAL.

BY

B. ARTHUR BENSLEY, B.A. (Tor.), Pa.D. (Col.), University of Toronto, Canada.

(Communicated by Prof. G. B. Howrs, D.Se., LL.D., F.R.S., Sec. Linn. Soc.)

Troon, D.O..N:

PRINTED FOR THE LINNEAN SOCIETY
BY TAYLOR AND FRANCIS, RED LION COURT, FLEET S'‘TREET.
SOLD AT TILE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW,

December 1908.

i
\
\
i
.
i
'
'
.
|
.
‘ =
' ~-
i
}
\
f
i
* '
q
'
' ‘
/
" '
i
a j
4 i
'
\
h 4
‘
|
.
ve
‘
-
~ >

LINNEAN SOCIETY OF LONDON.

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows ; to
Vols. 26-80, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the
Zoological parts of those which have been published are as undermentioned (for the Botanical parts see Botanical
wrapper) :—

Srconp Srrres.— ZooLoey, Second Srrres.—Zoonoey (continued).
When Price to the Price to Vol When Price to the Price to
Volume. Published. Public. Fellows. On este Published. Public. Fellows.
Line sd! £ 3s. dei fs. ds £ 8. ak
I. Part LvlS7is 2 cal 4 Ose ots 0| V. Part VII. 1891...-. 0 6° 0 -... 0 =a
Part & ID, 1875! o<.4 04/60 0.08 0) eel Part VIII, 1892....25:0° 8 0°... 0 .6ame
Part'\) TN. US76ie. StL ee Teo) Part IX. 1892. .... 0.12 “0 .... 0 “oan
Pact, AV S18ii. 2.) OG VOe ane O 1 2p0n Part X..1898. 4... 18 07222. 1
Party we We vl 8T74,13). SOS Oem en On Sar Part XT, 1894.)... 0-2 6 .... OMeQmmE
Parts 6 VLBI 7.5.91 2 10 ts oar OOO TEI a aoe & -0° Oe
Part. VIEVISTS.0..2 nly Oa ee ey a0. as Ee tee er
Part) VELL 1879. .ooat) 0.0 61680. ce an (gout cuignaates ee
. 6) ‘awiase, . . ‘
ie, Part TOUS TO. cone ly ae OMe. OM sanO Part. IV. 1896. ....-1 140. eos eae
Part WWE ashlee 5 (ls) Wo soon. Il Part’ “V.-1896. .... 0) WONsO) oe Omen
Part III. 1882.....1 8 0....1 1 0) Part VI. 1896... .. 0:8 70M; S00nvcame
Port IV. 1882..... OS 6)... 0 AG) Part VII. 1896. .... 0 12.0 2... 0" Olam
Part Y. 1882. .... Ore ia aaa Part VIII, 1897.....2 10.2 Glee One
ee eee ope eats Ay 2 VII. Part I. 1896 010 0 0) S7eneo
Part’ Ville US83y Me 0 by OM. ce.. Oise 69) ae : es ie f
Taek We dees sip Baar ee tes Part IL. mie FER 12 On eS 0
Part: Ike 1888.-.5..10: 8° 0... 0 BB ae = pan te a ay. i :
Part? “kalSSd, co tO a6 eae. ONO uMa a » TOES... ; f
ESE : anaen Part. Ve 1898/0... 001800) Semen eee
sy ee Sat ny t i a Part VI. 1898. .... 0 18.0 252,00 mmme
Part XIII. 1884 iy BO sose a) al al Part: Vil. 1899)... Oel8) tO OS
Part SXGiNe 1885. Pres NOME tae) ‘ 4 0 4 6 ONS SIO Sees 0 2 One a OF29 0
Part XeVE 1885. ses 0 4 6 5 0 3 6 Part TX. 1899. Bisicy ill 0) 0 f 3 0 15 0
r .
Part XVI. 1885. 0 <5 LOWS Oceans Ba SUS Gon : Osea 3 °
ines & : ;
Part XV, A886... 2010) Bye0 ew Or® Part « 201000 0
PartXVUT. 1888. .... 092 6)..50 0) 200) vinn, Panty 1s 1900s 10 On
III. Part THURS 3. af lao ae Part © 1, 1900s.2.. 0 10’ 02% 0 S7aee
Part. Tl. 1884:..... 112 “O22 iia Part III. 1900. .... 0.10. “0 +2 2%...0; aa
Part | TI T885.5.) 3 10)-0 eo alo eee Part IV. 1901..... O14 0.2.6 OsDGhme
Part IV. 1885..... 0 9830 <0 ae Part. “Wi GONG. : 0 5 0 ...c0sae
Part. SV. 1887.42 (ai 048 110) p.2ae. COG ROM Part VI. 1901. .... 0.10 0: 0
Part’ Vil 1888.02. 0.6 0%... 0 eon Part VII. 1901. .... 1.8 0.2 aaa
TV) Part, Go" PSG) Lod Oe Ogee Part VIII. 1902. .... 0) 4) 0) e-em
Part TMASe eee Our ial) Park SaxXeeig02) ~ 5. (Oy vay. (0) 5 et) 3.9
Part) (ULE 1888) J25.70016 0) the O leon Part =X. 1908..... 1 0 0.... 015 0
art XI. 1903. .... 6 Ok oe ae aes
Vibert. ibtess.oc wold uo; aeeroMolna os is aaa ane : Kee
Ss 2 arb AOU LOUD: 52. Sroka f
Bert SMUT SCE: See 00a Oo) er a Parb MIT Indes. |... 10) laa nmeanS
Part 110: -889...04 ala W shee algo)
Part. TV; 1890... ,..°0 12 0! ..2.0) 9 "20> See aa eee. 0 9. Oe iene
Part“) Weu1890.-.. eaOueG iO. eee es Part II. 1903..... 0.8 04. .s0moe
Part VI. 1891, 2.5, OS" 20 See Oe ae Part ILI. 1903. .... 1 4 0ns.ce Selene

eae
2nd Ser, ZOOLOGY. | | (VOL. IX. PART 4

THE

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON,

THE LABIAL AND MAXILLARY PALPI IN DIPTERA.
BY

WALTER WESCHE, F.R.M.S.

(Communicated by Grorce Masser, F.L.S.)

Ley ND. O N:

PRINTED FOR THE LINNEAN SOCIETY
BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.
SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

February 1904.

\

LINNEAN SOCIETY OF LONDON.

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows; to
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the
Zoological parts of those which have been published are as undermentioned (for the Botanical parts see Botanical

wrapper) :—
Srconp Srrres.— Zooroey. Szconp Surtus.—Zooroey (continued).

When Price to the Price to When Price to the Price to
Viet: Published. Publie. Fellows, | Volume. Published. Publie, Fellows.
Be OG eh: Ee tah ee Sees £ 132d
eeRant Th, HIGGR caso 2k 4h WO nose @ 1S © Y. Part Villy USoieee.. (08 16) 0 sre Oe
Part 1 We goes OY MS O ooo0,M 4 o Rart VilliiaiS925 ee. 0. 8 0 ....¢0 Pomme
Te ON IW cose tb SS) Masao ll a w Part “Xe Wiso2 reer 0.12 0 225. (0) 3SReo
lee IN iG soso O UG W soca Oe Oo Part, iX3 1898.05.00 P85 100 sae ele
Part We Wee booet sie Sou5 Willey Part XU, 1894. R270) 296: 2 Oe
Part Woe alee conse ll. 2 W) pono WO) We © VI. Part T, 1894. S202 2) 0) (Ore TORO
Paya WOK Wee sooo Ib KS O seo Were @ Partet Wee SO eel eelelee Ome 3a
Part VELL S79. 1 +. 1 0-0 .... 0.1570 Part I. 1894. ....50°0 07.9 OV yam
IJ. Part eS Ogee tl 4 Omi. 0) lSae0 Part LY. 1896. .5.. 14 0S. aOmsan
Part Ih, Testis pean Wily Wrggas Oy ly @ Part V. 1896s. 22. 0210) 0) 33 4200s
indy MMM, ase eee de ik WP acoo tl © Part Vi. 18962... 4 0 SiO eee. LORsOmeD
Pantie GV Sol sa. cn OPO sooner G Part WIL. 1896. 2.2 (0) 112) 0) Se OO me
Part Vidi882. es OeiSgn0) Mee 0 eee a ou Part VIM. 1897.02.25) Ol) 2, 1G ace ae
Part” Wil 1883) <..0 1) 0) 403-2. 0815105) SyiieePart sy ol aSOGhee se 0; 10a Operon ames
lm NAG UBB os Oo) O sacs @ 3) Part Ut. 1897.5... 0/12) 90m ODOM
Ieee WANT, EB, Go55 © Bs OW sepa @ 2 Part IL 1897. 2220 0lG. Ones lO lme ans
Pant. (ik 88. oe. 0) 938 207... One go Part IV. 1898..... 0 10 70.4. :. 00) 2iaine
Paris i aL 8Sd. yj... (0 aC.) 280 ane Part V. 1898...... 0 18 0,22. Ooms
line = Pig Icey eon ee Oegans Wa Part “Vil U898-....20013) (Oe ce Oaommna
Part XII. 1885.....0 6 0....0.4 6 Part VII. 1899. .... 0 18 (0% .> (0igaae
Part XU (Sed: ea. 00 6 20) =e nO ee 76 Part VIE. M1899)... OOS 10 ew Ome
Part XIV. 1885.....0 6 0....0 4 6| Part [X. 1899...2. 2) 0.00.) .80 bee
ines Vg Uisteby booe 0 12S WGecq OW sk © Parti) OX. 19008... 210) iGO, en O meets
ie DQG Uy oases) soa, 0) BY Pants SX. 1900s oe 0D One me

IB FAH OOM Nighy scan OW oS O tose O 2 2
z VIII. Part IMCL rg ONO Soom Wi a
Parva MiSSSveeetec 0 Gene: ORAZ Part’ TE 1900 eo SOO
III. Part T. A884. .ie5.. LSA 0 58 leo ge) = ey IER CN oe Ol Cask, O 7
Part TD: W884 5 te 2 Ob eel eee Part LV= LO01 ee OIA) (ON tO MO mes
Part PUTS Soper Ome Ole, poe Het 24 (5) Part Vi. 90a Ouro Ole (0) 8 9
Part IV. 1885.....0 8 0....0 6 9 Part VI. 1900.../'0 10 “0.9 Oumee
Part ='V. 1887.....0 8 0....0 6 0 Part Wi. 190is 0001-8 0s ie
Part VI. 1688.....0 6 0....0 4 6 - Part VIIL 1902,./.. 0 4°02. 20msiag
IV. Part WeolSSG ieee ly 45 Om ey OntSaan) Parka axe MlO0R aoe... 0! 2on, Olen 0 meme
Part TROLS Biffete teense fle cS Omecnater tel and eam) Part) exeul008. 20. 1) 0) 10 eee sO oe)
lepine ABU absietienog (Iie (Gaon Wie) Partin kee 908s... ec: OLGLNO errr Omen
V. Part emtss se ee (oar) toe = § (Oy 3) ie Pantene i}Oe eiges OMOE Oe: PaO uf 6
Part iets) sees (Nay (Ol ey (0) Sh O Part XDD Undex:: .....:0 4279" foe eee
Part) “TLLS89r ae oo 20 aes ole OO IX. Part 1, 908...) 10) 49) UO Sent Oreo
Part: Ve W899 059. Oud 2 Oe tO eo ao Part) gl. 1903.20 2200S) HO nee eo meO
Part IV. 890.0 ONG a0 ee ee maek ri Part, DD. 1908... .edl 4 fOr. oD
Part, | - Wily SON ft O20 eevee an) Part WV. W904 ice 0 uO eetnen tO maemo

ToS.
2nd Ser. ZOOLOGY.) [VOL. IX. PART 5,

THE

TRANSACTIONS

OF

OWN aE ECE
ANATOMY AND DEVELOPMENT OF COMYS INFELIX, EMBLETON,
A HYMENOPTEROUS PARASITE OF LECANIUM HEMISPHARICUM.

BY

ALICE L. EMBLETON, B.Sc.,

1851 EXHIBITION SCIENCE RESEARCH SCHOLAR, ASSOCIATE OF THE UNIVERSITY OF WALES (CARDIFF COLLEGE).

(Communicated by Dr. Davin Suarp, F.R.S., F.L.S.)

Teen DON:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.
SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

June 1904.

LINNEAN SOCIETY OF LONDON.

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s, to Fellows; to
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the
Zoological parts of those which have been published are as undermentioned (for the Botanical parts see Botanical
wrapper) :—

Srconp Srerres.— Zootoey. Seconp Serres.—Zoo oer (continued).
When Price to the Price to When Price to the Price to
Volume. Published. Public. Fellows, | Volume. Published. Public. Fellows.
S$ s: ad. £23. «d: Li ies vid. &* eas
LPart ay) “ee 1S7ba0 1s 40 ei aD V. Part VIII. 1892. .... 0 8-0 .... 0 Can
Part pile IS7bok 6 0: b Ones eG Part IX. 1892..... O12 0 .<..20 [Oe
Bart. , MUL yO le Bal 0u eee tO Part — X. 1893, 7.220 8° 0\7-5 Slee
Part | LV, 187%. 3 20nlG) (Ome eOloen0 Part XI.) 1894,.52..90 2 65 ee Oeane
Part) V.87%.....¢-0/18.)0.... 20013 06
Part Vi. 1877. .0.. 1. 20... Oe 6 woe ae ie ee :
Part, Vile 1878... nce 16 0 eee 7) 10 ae ere a7 th ceo a
Purt VIEL 1879, ..4.01 0) 0.2.5 0915 40 ae ety ee
Part. IV. 1896.)... 4 4 0... 0.qeme
TE Part (yi 1879: 2... Ieee Te ee tet ens emt Part ¥. 1896, (.... ¢' 20 0°) 2. onaeae
Part) Mel8e. fe 0 5 0. Onde Part VI. 1896:....20 8) 0.0 Moen
Part III. 1882..... 18 00s aioe lea lO Part VIL 1896..... 0 12 0 .... 0 SUMO
Part IV. 1882.....0 7 6....0 5 6 Part VIII. 1897. .... 0 2-6 ....,0:a2man
Pact py. (Veal Sa ecee 0-810... sO pee? 8
Part | *VIo1883-,.0) 10-0). &.womip10,|) Vio Pant | ly iegee ses 010 02... 0 “7g
Bake evi ses neo oS 470 Oe Part 11, 1897s . 22. 0: 12-20... 5 eee
area Vill 1083.0... d0e08 10, ee 108 eS Part IH. 1897.....0 6 0....0 4 6
PA GET Jest otha s Cowen Parh TV. 1898.) .. 0:10 0.2.) oyanme
Pat Xo 1esl iO Ae OMe Part Y.-1898;....0 18 0°....5 0 aeme
Parti Xl otSs4cec is, 0ul0 M0 Us etna na Part VI. 1898.....0138 0....0 9 9
Part. OX aliSe5ane tO a Gu Ole ee OmeG Part Vill. 1899: 22.4. (0OS1SieO my sO mle
Dy SOUNy WEL 8) BR @ cs50 ) 4 OG Part VIII. 1899. .... OAD 4082 SOO eo
Ae NOH Gey anes OF WO ..-. t 2b. @ Part, (LEX: 1899... had: VOR OP SOs lm
Part). XiVo S85. oe 0) 4 Oe Paes Part X.1900.-....0 6 0....0 4 6
Port) XVE VBS5. 2, 10 wb 0 se 0 ane Part XI. 1900.....0 2 9....0 2 0
Part XVIL, 2886. 27 00-8 .0) 4s... O82 78) nar Part ce T300 i ant ee
Part XVIIL. 1888,".. 4-00" 2 56) + Ore ORO Part IL 1900..... 010. 0°.....0. ame
III. Part T1884 8, ld Or h5 ANG Part IMT, 1900:.. 22 0°10- 0°... One
Part, UT, AS64."..19) ol 1 Os Ome elas ae Part 1V..1901,°.2 2.0 14 0 ....pug0™—e
Pat’ W885... /. 110. 0. ..eed eonta Part: V.180l,2-.5 05 0... Opec
Part- Ve 1885:.0nee 0.280", cat0 abe ae ‘Part Vin J190Ic.. 0 10 Om. eee
Part. ViCS87; .. 220.8) 0) ae eee Part VIIDAS0Ine... 1 8 OLe
Part: | WIP WU SSS 52,0 6° One ee crete Part VILL. 1902. .... 0: 4 OS eeOmee een
HV. Part) IPSSGs 4 id 4 0 re Oe Part 1X, 1902. .... 0 5 JOT: eres
Part TUS Gy ee een) Omen eral RO Parts) xem O0 Baek ocual 00. SEO) 15> 10)
Part’ Ir 1888y...32.0 16 07... - ONO Part - RE. 1908... ... 0. 6) 2 Oe
NivPadt ) Waiee a... 01 eke mee (Oe eee Pat ee
Part... 11, 1868), 2% 0:05 Ol. Ow eee Rett Se ater + Oo
Part UI. 1889. 65... 15-7 0..0.°7° 0" 01) UK Parte 903.) 2.0) a0 en
Part’) ‘TV. 1890... 6. DD-100. eno aoa . Part 41s, 1903;;....,0 8’ 0 eGo
Part Vi 18902 e8 20216" 0a, AO ene Part Vill. 1903). 2: Hie Serie: (ess: ° (0,
Part) AVL. Soe e0 al? ar ee Oe Part) V2 904) 26 10.16 OSes a0 eens
Part VIL, 1891. 22700, 16 NO nO ee Part “Vs 1904.4... 0 16-20 spew ONeene

2nd Ser, ZOOLOGY. | (VOL. IX. PART 6.

THE

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON.

- om: yt ATS
QvoP? Ne
TE a
7 maT >) eee
LITTORAL POLYCHATA FROM THE CAPE OF GOOD HOPI. s(t
BAYS Cg ea ¢
a a
ARTHUR WILLEY, D.Sc., F.R.S., Colombo Museum, Ceylon. Se

(Communicated by Dr. W. G. Riprwoop, F.L.S.)

EON DON:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.
SOLD A’ THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

July 1904.

‘
‘
'
»
+
‘
|
} \
‘
; -
.
! s
1
4
d '
’ .
/
'
7 .
‘ \
‘
oo
; .
' .
i
i
'
’ -
?
‘
4
‘
;
‘
1 .
'

2nd Ser, ZOOLOGY. |} (VOL. IX. PART 7,

THE

TRANSACTIONS |

OF

THE LINNEAN SOCIETY OF LONDON.

ON THE EVOLUTION OF TOPOGRAPHICAL RELATIONS = <<.
AMONG THE DOCOGLOSSA, is

BY
Hi. J. FLEURE, D.Sc., Fellow of the University of Wales. SY %
(Communicated by Professor W. A. Herpman, F.R.S., F.L.S.)

HeO) NCD’ O: Ni:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.
SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

November 1904.

t
-
o
fl
‘
\
‘
f
‘
f
'
,
iets
‘
A ;
: i
’ i ; :
Vrs ot
’
sal iv
¥ € af
‘i, ’ F
re fey nat
are |p i " ;
i 1 7, y
bi] Ren nf
, ania) 2) i yA ena
Z b TN ; a H ns
y . ud) 4
a ; et mas ! the Ki
i
- ‘hg
{ tO ay) od
m ; Gr; Ne

LINNEAN SOCIETY OF LONDON.

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows; to
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the
Zoological parts of those which have been published are as undermentioned (for the Botanical parts see Botanical

wrapper) :—

Srconp Srrres.— Zooroey. | Srconp Srrtus.—Zooroey (continued),
Woluite When Price to the Price to Wales When Price to the Price to
i Published. Public. Fellows. Published. Public. Fellows.
BED ehesd: £ s. d.| £ s. d. £ os. d.
eePant Weekes I dk OP ooo 5 WT © V. Part Xisb03>.... . Semone ee ce
Part WS USGS oo MS poe, WY eh Parh sXIedRo4. . . Oo OmGem 0 2a0
Thine OTS ASW Gogo UE EO S556 1 a © Wit. ant T. 1894, ....'2) -0) sOn ee OnE
Bethe ALN aT OC) eae ean Part J, 1894)... ..°1 1 (ee ee
Bart. Vs 1874 O18 510 tie OLE NS Part IM. 1894..:.. 0.100 =. Olja
Barty VES 0 toni) 120. eee 16) 5G Part IV. 1896. .... 1/4 ODS eOelenmm
Part VII. 1878.....116 0....1 7 0 Part V. 1896.".... 0 10 <0 25-50 mien
arg MUU ISO. oe 210) Oden 0 01 LO NNO) Part VI. 1896..... 0 Pe.
Ie Pact of) ple 1870 een Lead ON eal and| Part VIL. 1896..... 0.12.0 2. -onRouee
Part) Ue 48Siane 00S: OP ae. sO. $y) PartVITI. 1897. .....0. 2 6 6.0. s0nweae
cesta ante 1 8 0.2110) vitipart | 1.1806, 2.). 0:10(0nesel
Part IV. 1882.....0 7 6....0 5 6 Part IL. 1897....: 0°12 0... 0 a.m
Part —_-V. 1882..... 0 3 0....0 2 3| Part IIT, 1897..... 0° 6 0) 2 0
Paxton, Wi; BS: See 1/05 20855 .5.90)/ 157, 0 Part IV. 1898. .... 0 10’ 0... 0 yam
Tie USS eehe oats Ue aeceer ie Part V. 1898. .... 0 18) 0°... 0 dame
Ips eNlblenlcey 4 hte ila) Relea Oia) pie Wr Hegel olaS ioe Seen
Le aie ete cr ee Part VII. 1899.....018 0.... 018 6
ER eo Mec aes) ne oye et Part VILL 1899...... 012 0... 0 Game
Paxiee WI SOds a. tc 20 TONIC. FOR AC Bald) See 88a, Ln Gio ea
se ONE el eR AU tera Part . X. 1900: .....0' 6» Oa. , one
Be eee ics 20. aut aie, fue ree Part XI. 1900.....0 2 9....0 2 0
Teray NG alstsa 5 al OMG WOM ss OA SG) i
Park SOV, USSR cci0 de Bl hue na i tlle Pent LOO a0 1 ORNO REE come 7 6
Part XVI. 1885: :... 0 Gr 0 3.9) re oe nen pa Rs ; err ; : :
Part XVII. 1886.....0 3 0....0 2 8 Pash Lob ateaoeen
Part XVIII. 1888. .....0 2°6....0 2 6 Pan toons Oe 0
IT, Part oUeed a mele AOn cule ome er ec
Part. >. Te 1984.0 alo 10 4 elo Part VIL 190124 1 8. 0 ee
Part Te @885e. 0.0 dO, 107. ae a Ane mae Pat VII 10RD Lo)
ze ef Se as : ee . ; Al Part 1X. 1902. .... 0 5 0.20 e0eme
Bi aw nay gee Part) X, 1008 6,. 1.0’ “0: eee amee
: prime oe Part XI. 1903.....0 6 0....0,.4 6
IV. Part Te WSEG heeded Oe ge Cano Part Kit. 1903..... 010.0. 0 gna we
Bare) Te LBC el eo) ta aaa Part XIII. Index. ..0 2 9...,0 2 8
Part. TUL, 1888... 0 16 0 2 ..°0. 02:00) Re a
Vie sPart Te ANSSS5 eis oO Ve 20 eee 7 OO NO) Part. e903) ....2..0) Sen One OUROMEC
idem Scie ee Pat A Poe cae Ua 4) Part TIE. 1903.,.... 1 400) ee Odeme
Part IGS 889. o... il Ma 0 + . il 0 me Part IV. 1904s aoe 0 6 0 2 A 0 4 6
Part UV ISS Olea, fOr Oatceiens Oana: 0} Tyan Vi 904. 2-22 OV Ga NO ee O ee
SE Pere es Oe bal eu ae honteaey Part VI. 1904.....0 6 0....0 4 6
Part. “VI. 1891), 20 12° © 4. 8 0 Past VIL. 1904.2. 0 6 Oe
Bart. Vale AS Oise el nOm nO mera Ome: 6 | (In Progress.)
Part; SV S92 are 3 “Wr scca OW, C0) | X. Part (1, 1904) So 08 Olea Omens
Part » 0X 5:1892) 7530. 2 eee Seno Part, 11, 1904. ce Oe Oem 0

2nd Ser. ZOOLOGY. |

(VOL. IX. PART 8,
shall,

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON.

ON SOME SPECIES OF THE GENUS PALAMON, Far. FROM TABI,
SHANGHAI, NEW GUINEA, AND WEST AFRICA,

wit VY =)
=\ ti y 4 as /
BY Gysasee
Dr. J. G. pe MAN, of Ierseke (Holland). Y «
(Communicated by Rev. T. R. R. Sressrnc, M.A., F.R.S.,.Sec.L.S.)

EO Ns DO. N :

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET,

SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

December 1904.

q os ce

rit

qe

LINNEAN SOCIETY OF LONDON.

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. ‘The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows; to
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the
Zoological parts of those which have been published are as undermentioned (for the Botanical parts see Botanical

wrapper) :—
Srconp Srrres.—Zoonoey, Srconp Srrtes.—Zoooey (continued).
of When Price to the Price to Vol When Price to the Price to
Volume. Published: Public. Fellows. olume. Published! Public, Fellows.
Es. Wels & os: id: £ sg. ide £ sss
I. Part Is Mei oeca Al Zh) aay OS Ow V. Part XI. 1894..... 0 2 “ 2 0 2a ;
Soh ae er
Fart Me T8182. 0" 650) 11.270 6) 9 vi. Pareeinicon 2 > Glo
Party “TU D876... ol SO et SEO) BME Se ein (ee 1
mn IOs WS boon OME W conn O12 O Part “TL 1894....010 0... ana
Part)” V. 1877ase...0 18) 0's 2041S 6 Pi ee 1 4 60
Park Vio UST 75ers! Dies O 16a kG pee a 01000. Oem
Part «Vile 1878. ..a, 116) Ouest 79-0 Foe ahsiaii. k 0... 6 ee ;
Part VIII. 1879..... 1 0 0 .3.. 0 15:0 RR er ane a.
Hi. Part ‘I. 1879..... 1 4 0 .... 018 0 Part VIII. 1897....: 0 2 6 .... 0 agmmal
Dn Ws ) i= ) ‘
BE MNS os NG O11. ©) vit Part i 1896... 0 10k 0) soe
Part)’ Willa 1880) ae, eS ia Ua wane Mit an Beh a eee: Re oe
. de aave =< . .
Part IV. 1882..... OU 26h 0" ee eh mri ae 6. ae
e . ONT ENe, awel ais . .
nia OM ts tear OO) ONS ee eee Part IV. 1898.....010 0 ..../0 same
in aaa ase tiae Meh ay pgee OE A Part. V. 1898..... 0 18 0 .:., <Q
Part) Vill WISss0 mua 100 GeO eben O es ao Se SOs ae ee
ca ee come ee 4 eg: : ; Part VII. 1899..... 018 0 .... Meme
Pach gwaese eo ehea aie es Part VIII, 1899..-.. 012 0 .....0 Sem
a ee OG eis one ae Ste Part [X.°1899. ..).71.0 0... ORSae
Pant «Xi teed. 2. 6, 10 nO oe RO ae ee = aoa o ae
Pare KEL, Wehr. 0 6.10 1 eee Sk are Fa
Part ° XI, 1900; .... 0-2 9 .....0 SoMa
Part XII. 684......0 6° 0 7°. 0 # 6) nee a.
Part XIV, 1885......0 6 0....0 4 6| VUE Part 1. 1900..... Nae 0
Part) KV. WSsGH 0A Oi Oe ale Part ca ees cae é :
Part KWL AIBS5. ce wor, One a oo Part ee Re 0
Part XVII, 1886. ..:. 0 3 0 ....0 9.2 ae z Eaaehaes 5 Oe
Part X VILL I888. . 0. Om2e. 60,1) 00 eeo Bet he TOT ec 2 0 eee
eam fae <a Part VI. 1901; .....0 10 0 24.0", 7am
ie eau tence Part’ VII.-1901.0.2.. 1/8 (0) eee
ae : a cies es Part VIII. 1902. .... 0 4.0 2 0eewe
Part, TLE Weg5: 02 1/10) 207 we, a ee Se ae eee
Hoan IN Us, one U 6) W.eo0 0 B O oe = eae Rise, * coe tee tana
abs , ees nel
zat ies ce : ei a : : Part’ XI, 1908..... 0 6 © eo Ontame
Bes NOLS R ES Jor: came ae Part XII. 1903. .... 010 0.0.0 Zune
IV. Part Ife 1886. milenatle 1 4 0) 3 . 0 18 0 Part Xa Index. te 0 9 9 F : 0 D) 3
ea mere: a dees ; IX: Part) of 1908. ..... 0 9a Que aoe eEaee
Bart ET PBB cee one LO Og ee am Part) Ils 1903. .....0: 82 0. 5am Gman
V. Part —_—*‘T.:1888..... 012 0....0 9 0 Parte Ml 1903; . <2 1 4.) ¢.suovistee
Part. . ID, 18882:...0 5 0 ...,0> 38 pea tero0d. . elpeee ae ae ae
Part INE 1889. eeee 1 7 0 5 . 1 0 0 Part We 1904. tes. 0 6 0 x 4 0 4 6
Part IV. 1890. aos 0 12 0 Oo a 0 9 0 Part VI. 1904. a 0 6 0 , E 0 4 6
gen Wile Muweake yg SMe eM os Part WEL.1904. ..:./0 600 2a gee
Pe A Noe AB OTe ere Part VITT. 1904.....010 0....0 7 6
Part) Wi. 189illesac. Ol i627 0 mayo Ome ameG (In Progress.)
Part VIII. 1892. .... 0 ABUL bee hemecan X. Part 1, 1904. |. 0" S00 ee
Part’ 1X. te99=...) 0 12 ,"0ls MOM O NAO. Part Tl; 1904... 200 8) c0 eee
Part -- %, "1898.02: 1. Se Oa aan

2nd Ser, ZOOLOGY.) [VOL. IX. PART 9.

THE

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON.

OBSERVATIONS ON SOME UNDESCRIBED OR LITTLE-KNOWN SPECIES
OF HEMIPTERA-HOMOPTERA OF THE FAMILY MEMBRACIDA,

BY

G. BOWDLER BUCKTON, F.R.S., F.L.S.

fe OF Ne oN:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.
SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

July 1905.

.
, \
F
4 ‘
d
-
s
| é
i
'
'
i
'
.
* ‘
\
' ‘
on F
\
, f
‘
'
’
’
” ’
(
4 .
Oy is
; 4 \
f
aN
' ‘ox
*
f vi i}
dik
4 y J if

LINNEAN SOCIETY OF LONDON.

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows; to
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the
Zoological parts of those which have been published are as undermentioned :—

Sreconp Sertes.— Zoonoey. Srconp Srrtes.—Zoonoey (continued).
Vouune. When Price to the Ete to Waigtnes When Price to the rican
ublished. Public. Fellows. Published. Public. Fellows.
£ s. d. os ih. Sands Ly Smee
I. Part Tes 75e ee. lie 4 Ov. 5.c90T1 899018 “Va, Parte wale 1604; o/s pameme 1 10
Part. @Ule 1875,.¥.4..00 16) Ob pe 404 aol] Part , T1894: ..... 1:11 9 O05 mes
Parts LDL AS Oscoda) aSe <0 ee alee Part. IE. 1894. .... 0 10 Ox S200 Game
Party UV. 87%2, tee OG Or eae ok O lou) Part IV.°1896..... 1 4 0... Ogu
Part’, Wi. 1877. 0 18-0143 0 sla Part V.1896.....-0 10 0 .... 0 7am
an NOG Wey sage 1 PPO) Sono WIG & Part Vi. 1896...-. 0 8 10.222. 00 tomo
Parts VI 1878) oy Lele Omseeeda Bye Part VIL. 1896) ....20 12 0 ... 90mm
Parbavilllc 1879. 20s 1 0 v0) ee tb aeO Part VIM. 1897.25-. 0. 2.6 2.) 02m
i. ae a on nee mane ; i ate 3| VE. Pacts sue ioe ene 0.10” 0 2°) .Ona ame
: 5 ae a oa Part, ID. 1897. ..:. 0-12 0° .24210"'9mme
Fe Neo rata te ea Sea Part III, 1897.....0 6 0....0 4 6
ol NESS so URE rates Da De Part IV. 1898..... 010 0.... 0 yue
Part ¢ Vo1882. ..2; NYAGS BAM atanpog AU 2h
Dra aca ae ern need prea eae Part V. 1898. .... 018 (0° .).2 Ogee
EE MEO Lhe eS Part VI. 1898. .... 013 JO.) 220) ae
d : Part VIL, 1899...... 018, 0 ..:2 Oui
Part) Ville sess nos. 04 30-0 none es
a a Ae Me i oe AAR Part VIII. 1899..... 012 0....0°9 @
2 Part , IX. 1899...... 1 0.0 =. ..0Gfomee
Part. eXMS84e 0.310. 46. 08 eS
Part GRA See 0) 0110: 10. LO res Bart eg 000 he Eee a an
Oe Sa RIC, ite Sela Uae Part XI: 1900; ...:70 2°°O . 2.702
Part XIU, 1884, .... 0 6 © 1..5-0) 4 “6|-VIIE Part ~ 1900.4. 2.70.40) 90 Ome
Part, XLV.1885. ..4. © 16: 0 22076 Part 31<1900....... 0 10 20%. -.) Ogee
Part. CV. 1885. ...8, Oct" 6 455 0's 46 Part ITI; 1900, .... 010 0.24. 0877aue
Part. KVESTS8o 10 1D nO. Oe ae Part IV. 1901. .... 0 14) 0-4... (OUOmEE
Part XOVUL. 1886: 2....20 5, 0 Suet Oreo Part’, “V. 1901. .. 2.0095 +0! 2.2.0 aes
Part XVIUL 1888, ...5..0) 2-6 50 280 Part) “VI. 1901.2... 0: 1050... 0) yee
IL. Part 1, 1884 on, Pola ec Ou seer ero Part VII. 1901.....1 8 0 .... 1 18
Pant 1G Weve eo) Go. a ae © Part VII. 1902... .. OF 4S (0) See Ole oma
Park W0. 1885/%. 4 ol! 100. le eee Part IX. 1902.....0 5 0....0 3 9
Part INY SR no Ye Oo 8 Oo.6. 0 & O Part’ XeV1908S- eee (0. OU SeODa
Part V. 18870... ON 8) Ouse OMmnOMEO) Part XI. 190345. 0. 6 (0 S20
Peme Aibokeeee ae (GO) conn @ 2b @ Part sXuliey 19 03sec (0) 10) SO pert eer eet
TV. Part > Me 188604... kee e0 eee me Part XU. Index. .. 0 2 9....0 2 8
Part I. 1887.......1 8 O...<: 1) 1) 0)" 7x, Part | eeatses.,.. 0 9 S0geeenOnnoee
Part) WU aS88s te.) OG) Omer nOnel nO. Party ie 1903..,... 0 Se eeOmOmO
V. Part T; 1888.2. 2;..0,12) (ON... eOmeOmnO Part III. 1903. .... 1 4600 Oe
Party ILMi6esr 720 5 ie eee Part 40yo 1904. -... 0 (6s00ie ner Oeeene
Part / TL 689. 22... dy 7 One en Part — \Wool004, 5... 60 56e008 ase nO ee
Part TV. 1890. «2-012 0)... 0 Oma Ports! WikeO04. 6.6220 8 Ores
Part » Vs, 1890. .3.../0)6) 70m. = OmearmG Part. Wi, 1904... 2 O06) 0 nO eben
lemupe) OG UM acs WIRY ao OO Parhiviill, 1904. \...2. 10s10) Oe Oe mo
Part VIL 80.0, bg0s 060 ae Oe Part) Ik 904. 0. (Ones Ol ee Oe
Part VIII. 1892. .... OA dhs 0 @ C CE af
Part 0X, 1802. 2.2 202 WO oe X. Part, 1190455.. 0 Be oe
Part. |X. S932 eel esi OP eel eanlencO Rarte sul O04A ea nOmo EO (0)
Parb™ OXd,. 1804.02. .60! (2 16a oe Part ILL. 1905. ...2 09) Olee 0G ws

2nd Ser, ZOOLOGY. | (VOL. IX. PART 10.

THE

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON.

THE GENITALIA OF BOTH THE SEXES IN DIPTERA, AND THEIR
RELATION TO THE ARMATURE OF THE MOUTH.

BY
WALTER WESCHE, F:R.M.S.

(Communicated by Joux Horxtnson, F.L.S.)

ieee N DON:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.
SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW,

July 1906.

¢
\
,
\
= 4
\
2 ar
\
'
A
“
\
;
, , ie
\
\
)
P 1
4
aa
‘
\
'
j
alt 4
Pe
j ey

“~ ~*~ '
- ae .
i .
’ fs .
> - . ’ bo
. : -
o tf = ew
.
: =
‘
= ‘
?
2
— - > ; :
; .
‘
° ‘
7 4
. : oo
Re os . - « =
. :
a - ~ ot a 8s
J oar
z r a A
P rif"! **
: ~- *,
* e * 14 =. Se
i -
- oad Lee, .
- . { ’ >
S r

LINNEAN SOCIETY

OF LONDON,

MEMORANDA CONCERNING TRANSACTIONS.

The First Series of the Transactions, containing both Betanical and Zoologic

in 30 Vols., and a few entire sets are still for sale.
obtained at the original prices.
Vols. 26-30, 4s. to the public, ‘and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections.
Zoological parts of those which have been published are as undermentioned :—

Srconp Srrres.— ZooLoey.

a ant When Price to the Price to
Published. Public. Fellows.

© 48. Gs figs idl

I; Parts T.=VIM; 1875-79) 8 10°04. 16 6
AL Pant Oy 0, 879" eee oe ee roa sen
Parte PUR 1881.62. abe 0 see Onl) mall
Part il.ghesoe eee. 1) 38 One 8 1 SD
Part IV. 1882..... 0 57-46"... 0b oR
Part V. 1882. .... 0) 8 0ha 0) 2 38
Part’ ¢ VIL 1888, aoa: 1.10 “O9-ee Onda ON
Tore, Vil. 1883..-.,. 0) Ba) eee oe
Bart VIG e883. yes 0. OO, Lemme eS
art) “EXeMIS6s see OF 3. Wi .rke a0) coe
Daits @ Oks S84. cen cw, Aug: eeeOneS a6
Pork, 2XTs US! .20, 10 BO. 08 a7 x6
Parte MIG US85.089. 0 6 eee Oo) oe
Part Xe 684, 2.7.90) POLO a eeOg od BEG)
Part XIV. 1885. .... 0 6 0 «4.140 4°6!
ithe ONG IGS deo WW db G8 on OBB
Part XVI. 1885. Oe EM ay eee)
Pict, RVI SUS8O: 2. 210.8) WO, HOU RD eS
Parte Vile 1888, ao Os 26 ee ano LO)

III. Part Tie hctck Meee Wome Ral ie te ei 255

Parts eye 1804, 2.201 2) 1, eemetieng)|
Pont) SW W1885y.27. 1410 Oe ows
Part IV. 1885. .... 0 98) s0): See On 16am,
Patt =) | V.A88y. eed Ee Sie0) Sao OO)
Part | Wil.) 1S88.(40 5 O06 MO a ee Omeanel
IV. Part T, W68b.0. 15 MAS oe One
Paw; W887... >... 0 See, aoe ieee
Part: » WTAaSS8h 2 5 Olle e0er-- OmloanO)

V. Part Le 2888. 431/40) 128 10. ee ee)
Port “1, 18882.5.00 4b 0 ee Oren
Part TIT. 1889. .... Ls 7-20) 2s ee
Part IV. 1890..... WEE Wana. 0) Oo
Part || VC 189012. 4. 0° (60) omen
Part, * 9 Vip 891." .. 22 0124.0) oe OU OO
Part. ViTs 1891s... 40. 1610's 20s Os eae
Part VIII. 1892. .... 0.38) On. eo eG aD
Part, TX: 18925+< 5.019002 oe oenD
Part). Xe61803." 20. Sy Bee ae Glen)
Part. “XT, 1804.!...... 0796. ae. 05 10
VI. Part Il, 1894. Wi, 582.0) aD see ea a nD)
Part, 00. 1804) ...7. Jl Gils 0), 2 mio as
Paré. TL US8450; 4010) Oceans
Part | TV, 1896... 27°50 -seeeOmseno

for)

Thane! When Price to the
Published. Public.
ED ith

Vil Ranta VewSo6s 28-0) 110
Meena Wil DUI Boag Osh O 5
Part Willig 9Geee nO
lean UU USE sas OF BD B,

VII. Part Ie Me sooo (io) (0)
de IDL, ISG sp a5 Oe
de, JU0K, Ws sos OE
Part IV. 1898. .... 0 10
Jeti NYG SOY = 55 Oils)
Parte avi eiS 9S see Omics
Partie Villas oor nO mls
Rant Veli s9 9a ee Ole
Party LX elS990 lO
ens NG NGO og OBO)
ide MOE UNOS .54 W) 2) 8)

VIII. Part He SOS Soom (0) 10)
arte elle 900 See ONO)
Part) ie 90 Ose 0 10
Part IV. 1901... 0 14
lean Wo WO sao OF
ede Wil WOE goo WO
lahdy AYO IS soso
Part Van a 902. 5-2 0 4
arte sXe sl O02 2 er OMeS,
Part X. 1903. afl 0
ley 20 UROBE 5445 OG
Part 2G 1905855280) 10
Part XIII. Index. Our

IX. Part ie UG BY, 2s 55 OEY)
Part Diydg0sne:.. 0) 8
Part) SUS Stee. 1 2
Part diVe 19045... 10) 6
Part = WralGos> <2. 106
Part vile QO he =.) (OlaK6
Part Villa 1904.25. 7 0).46
Part iVall 904, = 52. 0 10
Panty ake 905: . ee Oeeia
Partppexe O06) an mOsm

(In Progress.)

X. Part TLQ0A 20S:
Part see 904 5 se Oh te
Hetty WWOIG AIOSS ome lS)
Party LV. 9055 ee 0) 00
Wart! tei SOO Gi euemO amu

Srconp Surtes.—Zoonoey (continued).

i=)

IS) (Stes) Sie) ei eS fe Serie! Sie) Se Soe I a)

See) KON OS 1S SS So Oo So oS

SPS aay)
(3) Sy (=a)

may be

nical al contributions, has been completed
i Only certain single volumes, or parts to complete sets,
The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows; to

The prices of the

Price to
Fellows.

bo Oo OD 1

~I

Te Oo

Is v7 bt

10

18

Ca Se a

ce

bo

Wa 1D

d.
6

OW OO DOD OD oo ©

OS OO Gy oenen SCO 6 Waa SO 6 S.o

amp oS Ww

Co

2nd Ser. ZOOLOGY. | (VOL. IX

THE

TRANSACTIONS

OF

. PART fi.

THE LINNEAN SOCIETY OF LONDON,

ON A COLLECTION OF CRUSTACEA, DECAPODA AND STOMATOPODA,

CHIEFLY FROM THE INLAND SEA OF JAPAN;.
WITH DESCRIPTIONS OF NEW SPECIES.

BY
Dr. J. G. pe MAN, of Terseke (IHolland).

(Communicated by Rev. T. R. R. Stessrnc, M.A., F.R.S., F.L.S.)

SS
ae = ss
& AGO
Bua
qn
Wl

D A
4. fy
ir mi |

aul.

by Gahap-O Ns

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.

SOLD At THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

March 1907.

RAR
\ WOOD
CAN

a

'
Las

LINNEAN SOCIETY OFS TON DOmy:

MEMORANDA CONCERNING TRANSACTIONS,

The First Series of the Transactions, containing both Botanical and Zoological contributions. has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows; to
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections.
Zoological parts of those which have been published are as undermentioned :—

The prices of the

Srconp Srrres.— Zooey, Srconp Srrtrs.—Zooiocy (continued),

When Price to the Price to Vol When Price to the Price to
Volume. Published. Public. Fellows. Suna: Published. Public. Fellows,
25 3 (th KE Ge (dh a2), dh £ (sume
TParts Vile lS7o—19 SOO 6) 6s) Vets Pant Te aS00" =. 0 10) 10: 2S Uae
II. PartsI.—XVIII.1879-88. 717 0....518 5| Part Il. 1900. .... 0:10" 0.7.0 ata
Part eal90 Ome 0 0) sve OA
UL Parte) Ty 1BRds a) Oy oe ae ia Ves ” ie
Part IV. 1901, _ tee 010 6
Part WS altekeee oo IL I) OW sgee 1 ff my
3 Part ‘VEs9 0s eee (0) Sy (0) 0 Sas
Wende WU, Tete, 3 Th TO) @) . Bee x q :
4 i Bart’) Vile 9 Oe OMlOMeO 0 FIG
PartielVewlS Gasser OS 0M OG 0 : :
ae lade WIM TOM, oo, OS = i EO)
Pact eer eb tenga i Ste Part VILL. 1902 0 4 0 0 3 6
art . 1902 é
P Wt WHS o5a5 OW @ M OG ab i 4 ‘
ante Viele Part 1X. 1902. .... 05 10 Sanc0 -oenG
IV. Part We det, seco dl ah 0) 5 =, (0) alist (0) | Part ox, IGOR, . 52. i 0). 0) 015 0
Bart) | MUP MGS Goo tens O80" te LaeleOl Rant) eX 1903s 0 @ O....0 4 @
Part INE 1888. eteviey =| 0 16 0 e 4 (0) 12 0] Part SOUL, 1903. iyo’ 0 10 0 : 0 4” 6
V. Part Hl, Wee shoe O 12 O 0 9 0] Part XIII. Index. .. 0 2 9 0) 2s
Part It alesis Goss @ SO. 0 3 9! IX. Part UES soe6 O 8 Os 0 69
lade JPN Test), Soa5 NO OM oan, © Judy IU GOES 644, 0 8 @ ...,0 8 @
Part IN TIEIO, nooo O12 Ooo. © 8) Jamin INO USO aoe al 48) Co. ) Sw
Part Wo WE coos OW coun M eh 1H) Part) JV. 19040 22 10.6) 40) eo een
Dart NOE Te pho OILY OW see (Wh 0) Part Vi LOO OMG! 0) 10: eee
Jem; WANG IRS, so5e O @ © coon M 2 @) lin Wie eH peed © OG OM so5, 26 &
lebygy WIIG JES ooee Oy S(O Gaon OW 8 O leiiae UNE alsye ee 0) OW) oe.
Bartle EX 1802: i. ce 0120 ....0 9 0) Part VIEL. 1904)... 0 0 20) One
Part X. 1893 of ale Bi AO are ele SOM ites Ibe Uy 45, 0) & O ..., 0 © &
lend OS CE oo OB OB cosa WB 0 | Parb Xe. 190682... ON 12) 0a O mC)
Vil Part TI. 1894. ,2 0 a5. 1 1 @ ide GS WEE ooen 1) 1) @ s35.,0 © &
(In Progress.)
Part I: W894. ellie FOR Seep ees st
X. Part E1904.) Ogee Dae
Part III, 1894.....010 0....0.7 6| aon auncae nae ; é :
| ar 6 e ot isnatfes oO : 5
edn OYE IB soon Jb 2! W s555 0 WS O) c
‘ 5 Part JL, 1905. ....0060) 0,2... Omemee
Part We Mee cooo. WN WY aan O WB e 3
é : leva Ie URS ohn OW) 1) WW G45, Wy
lyn NAS UC apo MO ash © caso ® GO Part VV. 1906 07 6 053
ar . ae, © ‘e)\e: d Ds .
Zac VME, USNS aac 2 Rants 9
ae ae Pe ae : : Z : Parh VI 1906 0.0 5.0... SOnSmee
ae bead ets oe ee Wages ee eae Part: VIL. 19065....50 8 0 2)..70nome
Walle Pant Me WSN Gann (0) ako) (0) - On 7anG (In Progress.)
Part IL /1897 ...4.0 12" 0°. 5. 0) Oemu
Wap WOE WES sean © GB OW soon O LG
Part TNS ARS See WO IO O coun OY
Part We dlstiss qasoaks: (nag, it) IB 6
Part Wikjulsthes 4a55 1B WM s556 0° 9
Part Vl 1899) 2.52.0) 118) 0 ae Ol SiG
leg WOOT ait ecg 0) ee 0) 655, © & ©
Bart) TEX 189955 1. tO) Ort O lyme)
Part X. 1900. .... Oo} 0) Goa ah
Part > XO 003 2s 0) 2a Oe)

2nd Ser. ZOOLOGY.) (VOL. IX. PART 12.

THE

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON,

ON CERCOCOCCUS EREMOBIUS, GEN, ET SP. N,,
AN ABERRANT FORM OF COCCIDZ. /

CONF
BY Kone =

HUGH SCOTT, B.A. (Cantab.).
(Communicated by J. J. Lisver, M.A., F.R.S., F.L.S.)

GROEN e DON:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.

r

SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND co., PATERNOSTER-ROW.

July 1907.

LINNEAN SOCIETY OF LONDON.

MEMORANDA CONCERNING TRANSACTIONS,

The First Series of the Transactions, containing both Botanical and Zoological contributiors, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. he price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows; to
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections.
Zoological parts of those which have been published are as undermentioned :-—

The prices of the

Srconp Srrres.— Zoorey. Second Serrns.—Zooroey (continued).

When Price to the Price to When Price to the Price to
Volume. Published. Public. Fellows. | Volume. Published. Publie, Fellows.
on Seas Sseaceen| oe oe dh se 1S) as
I. Parts I.—VIM. 1875-79. 8 10 O . 7% Bi! W2UUL, Want I U0, caso OO © ..,.W 7
II. Parts 1.—XVIII.1879-88. 717 0 .... 518 5 Fart: Th 00 010° 0
Part IIT. 1900;.5... 0 10 0... 0 me
III. Part Te 1584... A 20a eS : : ;
pee Th feed) ~~ 1S ORS Lm Part mee WO i we “ ie y ate Oh OME
Part Vi 100le 0 520 404 Ome
Part) We 188i. ee 1 On Oe Oa : > E
: JT MME TOS 5555 O10) Oo... 0 Fe
Ze ene ie Ine SUE ora cae Part, VIL, 1001, a 8 100) Sa
ry ie LEO Neh h 0 ONE Uae Part VIII. 1902. .... 0-4 0. io eoano
kine Sc es seals 1) Part 1X. 1902. .... 0.5 0%. Ao) soma
IV. Part TUSS62%c we, ll 4 Ones 25 10 1S an Part “X.1903..... 10 0 <eOnienee
Part," T8872 J 8 02 Pa 6 Part XI 1903:....°0 @o 00 2 = Gmeuae
Part JONG ietste aong Ws W oe. (0) 1 0) ms NIUE WOW, ...., O 10) © eee
V. Part M1S88.2 an 0 120 010 0) ino OU0G Iki 5.10) 2 2... 0 2 &
Part WG West oan O S OW ance) & 9} IX. Part We MBOBS os ec 0 (9,205.2), OGG
Burt! UBS) heel Ome clemrOme| Part. I0) 1903)).2 4/0, 0801009) NO Gm
Rart® TVs 0890) ean0s 2010. a) ono) Part “111. 1903... 4 40.6), Ome
Part) Ve 89040..20e 6) 100. 60 et Part: LV. 1904,...10 “60 |. 0) me
Part NOL JG seco O 1D O52. 54 0 ©) 0 Part We MOE Se Oe Ww - 0 4 6
Bart ViLVseley.- 0-6 0 2-10 4) 6! Parte VI 10045, . 6-2 0596 = 0a OME |
Part VIII. 1892. .... OS 80 4.5.20 16604 Parb VU 1G0k se 06s Ogee ee |
Part ‘ 1X. 1892..... Omws0. =, 0) O50 | Part VILL, 190¢,%. 0.10 (0 9. 0. ayaa
Part X. 1893. .... Regt Ser Os Sees Lat BN Part, TX. 1905... 50) <6. 0"... Ones
Part, XU, 1894 2 0) 9 6,. wae 20 Part. X. 1906. .... 012° 0) |. OMEOENO
VitseParb (1 1994022820.) 0 salons Part. XT 1907-.-2.. 0 1210) ae sO One
Part Il. 1894. Lil @ . L 9 8 Part Xl) 1907. .-3. 0 0 OD ies
. (In Progress.)
rank TUT 88H eons O10), 0 Sache) aes mr ee 0 Se
gMcliges (oa meerna tere te < Pars TI. 1904). ...1088) 02-0 Gam
Part v. 1896. .... 0 10 Caen 0 arene Part, TOT. 1903" ot OMepn Oe Enea
Pe meee cane oR ae 0) Part ,TV.1905,.....4010)-0 ..../0) syaN@
Part VIL 1896..... 012 0... 0 9 0 Part, Vo av0es, toa a | toe
eee eases feat, Mace 2 O Pert VI. 1906.....0/8 0 ....0 2:98
VII. Part =I. 1896..... 010 0....0 7 6 Part VU. (S07 Gnd: 0°... avon
Part © Wl 189700 290,19) O09 BO (In Progress.) :
Part’ TM D807: -..-10%-16) 0% Oe eae
Part IN. 1898.44.) Oa. 0 5 Omens
Pact “V 1898%.,.- 5 OC1e 70.22 weOnIGenG
Part’ VI, 1898)..,. 013° © 2.0) Ouro
Part VIL. 1899. 1352.0 18.0 se nGnioeee
Part VIE. 1899... .. 0 12 0 42.10) S010
Port UX. 1899. .-.20 0016-90, OMISEnO
Part — &X.. 19C0.;.< 2.10. 66% 0). MOnManG 2
Pact: "Xi, 1900; 1), 02.99) een

2nd Ser, ZOOLOGY. |

LVOL. IX. PART 13.
THE

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON,

BY

Pror. W. A. HASWELL, M.A., D.Sc. F.R.S., F.LS.

EOeN D-O N:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.
SOLD Al’ THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

August 1907.

LINNEAN SOCIETY OF LONDON.

MEMORANDA CONCERNING TRANSACTIONS,

The First Series of the Transactions, containing both Botanical and Zoological contributions, has been completed
in 30 Vols., and a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may be
obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows; to
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The Second Series of the Transactions is divided into Zoological and Botanical sections.
Zoological parts of those which haye been published are as undermentioned :—

The prices of the

Srconp Srrres.— Zoonoey. Sxrconp Surtus.—Zoonoey (continued).
When Price to the Price to Vol When Price to the Price to
Volume. Published Public. Fellows. ue Published. Public. Fellows,
25 Gh Ss ids sae £. s.aids
I eet RNA ies 3) I) W ooo, 7% G)) WINUL Tree TALS 00: 5 2 10) LOM On Oe
IL. PartsI.—XVIII.1879-88. 7 17 0 .... 518 5] Part I. 1900.....010 0....0 7 &
Part III. 1900. .... 0 10 << 0G
Mi Park) SS eyes POY eae ene gies a oe
§ a4 Part TV 19015. 0120) a Om OME
Part Te 8845556 ale. 0) 2 Od) : a :
oa i | lei AVE JON Sean) Gy W Soo, 0) 8
Lenin IS SSIS Soa Hl MO) 0) Ce 26) S
* A : dee We Ue ees IK 0) 555 OF
Part lig Wests, oan WY ot W hos OY OG . :
ae sree Sahm Ot leno Part. * VEE 290i eel Si Ole eee
ar 0 OOUe secs) A fe S Cah
ae eae ote Part VILL, 1902; 14.0 4 (0)-eee OuemnG
- vag tea a Part TX.'1902.,...00) 5 10 22 0 See
LY. Part Il, 1886. eee a it 4 0 - 18 0) Part xe 1908. ’ ae il 0 (0) ; . 0 15 0
Barb 7 JOE SSeS Obey Lt ae Part ‘KI 1903....2 0 6 © ...05eaee
IRart ea Soon an O mel One Ole 012504) Part XII. 1903.....010 0....0 7 6
V. Part Moo altetsteh ounec Om 0 OOO) Part XIII. Index. OP .2 92 ...50) 2S
Part TEMUISSS anaes OY 5 @). 0 3 D9) Tx Part I. 1903. = ONTOP MOK. Ons)
Parte Vee IsS oer sae Th eye Oka i 0) (0) Rart) luiewi903% 22-5 0) 3S) 0 OOM
Win De SENOS sces @ 2) 0) 6.9 0 Part M1903). 60 100 man
Part Wo WES acco OG OW 0 4 6 Part [V. 1904. 225 10) 16) (00 eee One
ieydk WARS BIG Deas Ol WF 0) hon O 0) Part “Va l90ds). .. 09 U6) (0) 2 a OmeeLnG
hyde NIUE, TheSIS Song OP 1G Oe 45 0 4 6 ivan ale Wee) GY OW) ae, @) ete
Ieehdn WOU, eee sang Oeics 0) oo oy (3) (0) Part “Vile 19 0:25 3550" 16) Ole Oe
iin IPSS URRY Sy, (ie) 5 0) 8) leven AUG WE eo OO W) os.) 7
Part Dee tei totes 1" Gtoh. (OMe ek all, (Oy Part) UXeu905:,. 2.12 0)" G0 sO ee ee
lend NOG INSEE Boa Ub 2G a O 2 Part; 8X5 1906; .. =. (0) 12105 see Om Omen
VI. Part bake OF a OO I iy Part XS LOOT cs O20 ee
Part ME NEC a al bg 3 Part XU. 19075... 0) Se (0M ae Ueoeey
letindg AW alee og 0) alt) (0) We le CONE Me sco O 0) cook © LO
(1n Progress.) 7
lett, JENS USING, Se Ol EO a OMS SOn
oe : X. Park T1904...) 008) 10 so ommaes
Part Ws st, Goon (0) TO) (0) #5
Part VI. 1896 ete jwcenta Parte i 1904. 22 Ole s) 0 - 0 6 0
; RE SE ata, Seat eccageied Part’ LL, 1905:%) 22 -mOms0% (On em On mma
ena AVI, USC Aces ON 5 0) ee) . :
ee , 2 Part, IVs 1905s eeremOiel(O) (Ohare pae Ome aan
Part Vil. 1807s i. 10162, “Gee 70 Geen : y
Part Ve giS0G32 e072 1G ae OmeaD mets
VII. Part Li stkay 5a (0) 1X0) 00) Oe 78) Part Viod906 8.10) 3 Onn eeOn ONES
Part I. 1897, 722. Ol2, 0 = O97 0 Part Vil 19075..5...0° 3 0) sane Ono
Vethan WU Ase eee OY (6) 0) » Ome Ga (In Progress.) :
Part IV. 1898.....010 0 0 FT 1G) EXEL, Part, “SO 7aeeae 2 (8. 10me eee lene)
Part We Wests, sc a5 10) Ws} 0) Wy et 3,
Part Val ISO Ray (Q) is}, (0) 5 (0 8) 8)
Part Vil. 1899: <2 750) ie) 0 7 OMS ee6
loin NAOT iets 5555 <0) key (0) 5 0) 8)
Teebane TOK ARE Ao Il) 5 0) iss 0)
Part DE LENO ee. C0) Kop (0) 0) 456
Part X02 190052... 0) 25 9 Sl 2 0)

2nd Ser, ZOOLOGY. | (VOL. IX. PART 14.

THE

TRANSACTIONS

OF

THE LINNEAN SOCIETY OF LONDON,

tae TITLEPAGE, CONTENTS, AND INDEX. “AiG Ar

Loum: WO) N.:

PRINTED FOR THE LINNEAN SOCIETY

BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET.

SOLD AT THE SOCIE'TY’S APARTMENTS, BURLINGTON-HOUSE, PICCADILLY, W.,

AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW.

October 1907.

ai en

an.

The First Series of the Meaneanienes containing both Botanical and Zoological contributions, has been completed

LINNEAN

SOCIETY OF LONDON.

in 30 Vols., and a few entire sets are still for sale.
obtained at the original prices.
Vols. 26-30, 4s. to the public, and 3s. to Fellows.

The price of the Index to Vols.

Only certain aoe volumes, or parts to complete sets, may be
-25 is 8s. to the public, and 6s. to Fellows;

MEMORANDA CONCERNING TRANSACTIONS.

The Second Series of the Transactions is divided into Zoological and Botanical sections.
Zoological parts of those which have been published are as undermentioned : =

Volume.

III. Part
Part
Part
Part
Rant
Part
Part
Part
Part
Patt
Part
Part
Part
Part
Part,
Part
Part

IVY

Part
Part
Part
Part
Part
Part
Part
Part
Part
Part
Part
. Part
Part.
Part
Part
Part
Part
Part
Part
Part
Part

ME

Srconp Srrirs.— Zoonney,

When Price to the Price to
Published. Public. Fellows.
Leesa. eG 5.
I. Parts I.-VIII. 1875-79. 8 10 0 Ole? AO
Te Least, Tetley (songs fy leh 6)
Succes 114 0 ale 5 AG
TN USS k oe 1A) bee eat)
INNES eh 1; Boe 110 O 2216
IV. 1885 ORO SP Om wee OF a6,
V. 1887 sl) RAO) = 10) (ayy 0)
VI. 1888 0 6 0 (Qp ) AD is)
Tepleeceees One 018 0
IM Stee ook 1-48). -0, PS yo)
TMS SSSs een 016 0 Qpi2) 10)
LE RUSS ieee ORI Oe ere Om oO)
JES istetsh A ay OB. On aH)
WSS Ole ae 17 Oa Ais SAU? 0) 300)
IV. 1890 LO a0 i 8 0)
NES Oo Ons ere OweG Oe cues Om 42 16
NY Fes) ttt 9) (ea ees, a0 ee! 2 Oe (yf) 0)
Vile SS 0! G90 0 AL 6
VILE SO 2s eo OO SSe0 (Oe {oh 49)
VN WB9DS ve SON dS Ole. 0 9
SOS 5 oes, le Lom) ep tia el Ba 0)
eli GO Ae ci cry Oona OR 2 0)
WS AES als gray 24, (OAD 5 te ae
DW. W894 ee Oh Ss 1B
MM BOA.) 52. OO) 50 Diese
TVS IGs sehet lice et £0) matt) alts)
Weal S96 aces OMLOM Ota dO) fie 10
Vin lBo 6s ce) 85 400 ce ce OG 0
Ville SOGH Saas OM aa OF are OOO)
Vil SO Tees. COM ob 1G be, Ore ee)
Pel SOGH. Sn OmlOr 20 5 Wor ee Aa
TR NSS is, ae oO 2 10 20,29 ©
TIS 7e OF 6 Ol ee Ove 6
DV WOO eae eLOn NO acta gn 0
Vili898) a0 anOne) 10in 013 6
Vi SOB coc NO hehe Ile. (Ofer e) AS]
VIL AUS 99 nee, OMS 0 PONS G
VL TSO9 eee OO Ole ie onto)
DXB OOo. et alien) Ool Sr n0)
XG ALO OOS son) emote 0 4°76
RT OOO wl ee ad O2es0

Part

Volume. Hae
£
VIII. Part MF AYO Sa se 0)
Parts, J900; eee O
Parte lis O00: meee ame0:
Parte Ve, WO Oller 0)
Part © We 190i sen 0
Parte, AVAL 9 Oimeew cm 0)
Pant Vee eOOleme nae
Part VEU GOR s ous 0
Partine Xe 902s eee)
Pant exe 903 ee eel
Bart eNts9083e 2. 0
Part “xo G03h sa. 0
Part XIII. Index. .. 0
IX. Part We ISOBY Bats 40)
lehudg 0G USE a5
1h UO URE RSS so A
Parts Ves 904s 2 20
Part SVE 90+ oa. a0)
Part. Wil 904. 5. 0
Part VII. 1904. .... 0
Part Vill 29025 2... 0
Part SEXelO0 ba nen
Parte OO Gea. ne
Part. XMEIOT a. oe 0
Par Ge xee lO Oi eer)
Part Xi AGO 7 eos 00
Part XIV. 1907. Index.0
X. Part We L904 sero)
Part II. 1904. .... 0
Part TLE i9050 ei. 0
Part DV; 1905. 424.110
Part) Vii 190655. 5.0.0
Part,” VilSsl906i ee nO
Partie Velle wl GO vem
Part Vill io Owen eenO

(In Progress.)

XII. Part

Af

NELOTG 5 Qecemal

10
10
10
14

5
10

10

Price to the
Publie.

8 ds

Sl =r}

SIO ORS one ono Oo SSNS ©

Se) (SSS) Sa Sea a at

—y

Price to
Fellows.
Sh
AG
Te
a 36
010 6
(ya 8)
7296
I)
a 0
oy. AS)
15 0
0 4 6
a 6
2 te
6 9
6 0
18 0
4 6
4 6
4 6
4 6
OWwe/gn6
4 6
9 0
9 0,
ied
4 6
Ale}
ie:
6 0
6) 9
iP A8)
5 3
ie)
23
3 (0
Xo)

to

The prices of the

wn

i

BAS

Be ¥4 4
: , - ss ~
ooo BS " ” . |
ce van erie. he
; hg ecu “t
~ aA a “
t Tape
. ¥ .
~~ Fe ar of

Poti oy
ata ky FF wy
>. : | | \e a