BULLETIN

OF THE

MUSEUM OF COMPARATIVE ZOOLOGY

AT

HARVARD COLLEGE, IN CAMBRIDGE.

VOL. XXXII.

CAMBRIDGE, MASS., U.S.A.
1898 - 1899.

Reprinted with the permission of the original publisher
KRAUS REPRINT CORPORATION
New York
1967

Printed in U.S.A.

CONTENTS.

PAGE
No. 1.— Studies from the Newport Marine Laboratory. XLI. On Dacry-
LOMETRA. By A. Acassız and A. G. Mavır. (18 Plates.) April, 1898 . 1
No. 2.— On some Mrpusæ from AusrtraLIA. By A. Acassiz and A. G.
Mayer. (3 Plates.) April, 1898 13
No. 3.— The GorpraceaA of certain American Collections. With particular
Reference to the North American Fauna. By T. H. MONTGOMERY, JR.
(16 Plates.) April, 1898 , , s e 21
No. 4. — Some PLANARIANS from the Great Barrier Reef of Australia. By
61

W. MoM. Woodworth. (1 Plate.) April, 1898 . . . .

No. 5 — Reports on the Drepaina OPERATIONS off the West Coast of
Central America to the Galapagos, etc., by the U. S. Fish Commission
Steamer “ Albatross.” XXIII. Preliminary Report on the Ecnını. By
A. Acassız. (13 Plates and Chart.) June, 1898

No. 6.— The Nervous System of Nereis virens Sars. A Study in Com-
parative Neurology. By J.J. Hamaker. (5 Plates.) July, 1898

No. 7.— On Remains of STRUTHIOLITHUS CHERSONENSIS from Northern
China, with Remarks on the Distribution of Struthious Birds. By ©. R.
Eastman. (1 Plate.) July, 1898 ......

No. 8.— Reports on the Drepaine OPERATIONS off the West Coast of
Sentral America to the Galapagos, etc., by the U. S. Fish Commission
Steamer “Albatross.” XXIV. Preliminary Report on BRANCHIOCERI-
ANTHUS Urorouus, a New Type of Actinian. By E. L. Mark. (3 Plates.)
a O Ae ne ee o.

No. 9. —AcaLreens from the Fısı Ispanos. By A. Acassız and A. G.
Mayer. (17 Plates.) February, 1899

No, 10.— Reports on the Results of Dredging, under the Supervision of
ALEXANDER Agassiz, in the Gulf of Mexico and the Caribbean Sea, and
on the East Coast of the United States, 1877 to 1880, by the U. S. Coast
Survey Steamer “Blake,” Lieut. Commander C. D. SIGSBER, U: 9. Na
and Commander J. R. Barrzerr, U. S. N., Commanding. XXXVIII.
Étude Monographique des PLEUROTOMAIRES ACTUELS. Par E. L. BOUVIER
et H, Fiscuer. (4 Plates)... .

69

191

Bulletin of the Museum of Comparative Zoology
AT HARVARD COLLEGE.
von XXXII. No. 1.

STUDIES FROM THE NEWPORT MARINE LABORATORY.

COMMUNICATED BY ALEXANDER AGASSIZ,

No. XLI.

ON DACTYLOMETRA.

3Y ALEXANDER AGASSIZ AND ALFRED GOLDSBOROUGH MAYER.

Wirun THIRTEEN PLATES.

CAMBRIDGE, MASS., U.S.A.
PRINTED FOR THE MUSEUM.

APRIL, 1898.

|
|
|

No. 1. — Studies from the Newport Marine Laboratory. Commu-
nicated by ALEXANDER AGASSIZ.

XLI,
On Dactylometra.

By ALEXANDER Agassiz AND ALFRED GOLDSBOROUGH MAYER.

Tux genus Dactylometra belongs to the Pelagidæ. At the present
time there are four genera of this family known, and they may be dis-

tinguished as follows : —

(1) Pelagia, 8 tentacles, 16 marginal lappets.
(2) Chrysaora, 24 tentacles, 32 marginal lappets.
(3) Dactylometra, 40 tentacles, 48 marginal lappets.

(4) Melanaster, 24 tentacles, 48 marginal lappets.

Thirteen species of Pelagia and eight of Chrysaora are known, and
they are found distributed among all of the great oceans of the world.
There are only two species of Dactylometra, and they are found along
the Atlantic coasts of North and South America. The genus Mela-
naster is represented by but one species, M. Mertensii (L. Agassiz, Cont.
Nat. Hist. U. S., 1862, Vol. IV. pp. 126, 166). It was described and
figured by Brandt,! and is found in the North Pacific.

The genus Dactylometra consists of Pelagidæ with 40 tentacles (3
large and 2 small ones between each successive pair of marginal sense
organs), and with 48 marginal lappets (6 between each successive pair
of sense organs),

In Dactylometra quinquecirra the bell is high, being almost hemi-
spherical in shape. In mature medusæ there are five tentacles between
each successive pair of marginal sense organs (Figs. 2-4, 6). Three of
these tentacles, the primary and secondary (I, II, Il, Fig. 6), arise from
the clefts between the lappets, and the other two, which we will call
tertiary tentacles (LIT, III, Fig. 6), are generally seen arising from the

1 Brandt, J. F., 1838; Mem. Acad. d. St. Pétersbourg, 6 Série, Tom. IV. p.
885, Pls. XVI. and XVII.
XXXII — NO. l.

2 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

sub-umbrella floor of the ocular lappets ; for even in very large medusæ
the ocúlar lappets exhibit only a slight notch, as is seen in oc, oc, Fig-
ure 6, and in Figure 26, Plate X., and only in a very few of the largest
and most mature medusæ do we find the ocular lappets divided so that
the tertiary tentacles arise from the cleft. This latter condition is shown
in Figure 19, Plate IX. It is interesting to notice that these tertiary
tentacles do not make their appearance until the medusa is almost
mature, when the bell has attained a diameter of 130 mm., and even
then many of them fail to develop into anything more than mere fila-
ments of very short length, such as are shown in Figure 4, Plate IV.,
and Figure 27, Plate IX. Indeed they never grow to a greater length
than is seen in Figure 6, which represents their condition in a medusa
measuring 190 mm. across the disk. These rudimentary tentacles stand
in striking contrast with the primary and secondary ones, which, when
fully extended, attain a length of from three to four times the diameter
of the disk.

In common with all other Pelagidæ, there are eight marginal sense
organs in Dactylometra ; four of these occupy the primary, and four the
secondary radii. They are set into little niches in the edge of the bell,
and project downwards from the oral surface. Figure 7, Plate VIL,
represents the sense organ as seen from the oral side, and Figure 8 is a
somewhat diagrammatic longitudinal section, showing the so called
“olfactory pit” (opt), which projects downwards from the upper surface
of the bell just above the region of the sense organ. As was long ago
pointed out by L. Agassiz, these sense organs are morphologically noth-
ing more than little hollow tentacles, the entoderm of which contains
a mass of otolythic concretions (con, Figs. 7, 8, and 9).

The mouth opening (M, Figs. 6, 31, Plates VI., IX.) is cruciform,
and occupies the centre of the oral surface of the disk. It is surrounded
by four oral fringes, or palps, which occupy the secondary radii, and
when fully expanded attain a length of about three or four times the
diameter of the disk itself. Fewkes! is mistaken in stating that the
oral appendages are “ of two kinds, four of which are quite long, floating
gracefully along after the medusa as it swims in the water, The re-
maining oral appendages being shorter, more ruffled, and confined to the
immediate vicinity of the mouth.” He was evidently deceived by
observing a peculiar state of contraction of the oral fringes, for their
shape is constantly changing. Sometimes one sees them as beautifully

1 Fewkes, J. W., 1881; Studies of the Jelly Fishes of Narragansett Bay, p. 178
Bull. Mus. Comp. Zoöl., Vol. VIII. pp. 142-182, 10 Plates.

|
|

|

AGASSIZ AND MAYER: DACTYLOMETRA. 3

fimbricated pennants trailing in long graceful curves far behind the
medusa, and at other times they are drawn up into a shapeless mass
about the medusa’s mouth. <A view of the free extremity of one of the
oral fringes is given in Figure 30, Plate XI.; and a good idea of their
general structure and appearance may be obtained from Figures 1, 2, and
4. Small wart-like protuberances (Fig. 29, Plate XI.) are found scat-
tered thickly over the outer surface of the oral fringes; they consist of
clusters of nematocysts and red colored pigment granules.

The plan of the stomach is represented in Figure 15, Plate VIII. It
occupies a large space in the midst of the umbrella, and consists of a
wide central cavity from which radiate outwards sixteen simple pockets
(r, e”, r, 7”, ete). These pockets are separated from one another by
sixteen radiating partitions or septw (rp, Figs. 6, 15, and 33), which
join the upper and lower walls of the umbrella together. These radial
partitions contain numerous muscle fibres, by the contraction of which
the rhythmical movements of the bell are produced. From an inspec-
tion of Figure 15 it will be seen that eight of the pockets (r,r, r) of
the stomach lead out into the sense organs, and that the eight others
(r’, r’, 7°) lead out into the tentacles ; indeed, the tentacles are hollow
throughout almost their entire length, and their entoderm is ciliated
exactly as is that of the stomach itself,

The genital products are contained in four radially situated infold-
ings of the oral wall of the stomach, and their position is marked upon
the oral floor of the disk itself by the four deeply sunken sub-genital
pits (gpt, Figs. 6, 14, 17, 18, 23, and 25). Figures 14 and 23 are
somewhat diagrammatic sections of the medusa, intended to show the
manner in which the genital organs fold inwards into the cavity of the
stomach. Figure 14, Plate VIII., represents a section through the centre
of the sub-genital pit and the pocket of the stomach ; while the section
represented in Figure 23, Plate X., passes through the edge of the sub-
genital pit, and through the centre of one of the septal partitions. A
view looking down upon one of the genital organs, the exumbrella being
removed, is given in Figure 17, Plate IX. In this figure part of the gen-
ital epithelium is represented as torn away in order to show the opening
of the sub-genital pit (gpt) lying below. It will be seen from an in-
spection of Figures 14, 16, 17, and 23 that the genital organs are fur-
nished with numerous gastric cirri (ge) which project inwards into the
savity of the stomach. A view looking down upon the surface of the
ovary is given in Figure 34, Plate XI., where we see the immature eggs

of various sizes attached by peduncles to the ovarian wall. A section of

4 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

one of these immature eggs, drawn from a specimen killed in Flemming’s
fluid and stained in Kleinenberg’s haematoxylin, is given in Figure 11,
Plate VJI. The protoplasm is finely granular, the nucleus large and
vesicular, and the nucleolus contains about half a dozen deeply staining
highly refractive granules. Figures 12 and 13, Plate VII., represent
young scyphostomæ, one of which possesses two and the other four
tentacles. The older one is 1.5 mm. in height.

The medusa is provided with two well differentiated sets of radially
arranged muscle fibres. The principal set is found in the septæ of the
oral surface of the disk (mso, Figs. 6 and 33, Plates VI. and XL),
and the other set is found in the exumbrella me, me’, Fig. 6, and also
Figs. 2-4), and alternates in position with the set in the oral septe ; so
that there are sixteen radial strands of muscle fibres in the sub-umbrella
alternating with sixteen strands in the exumbrella. It will be seen that
of the sixteen exumbrella muscle strands, eight (me, Fig. 6) go to the
sense organs, and eight (me’) to the primary tentacles.

Clusters of nematocysts are found in the numerous wart-like pro-
tuberances (Figs. 24 and 29, Plates X. and XI.) which are thickly
scattered over the exumbrella surface of the disk, the palps, and the
tentacles. These protuberances are thickly clustered near the centre
of the disk, where they appear as little hemispherical projections above
the general surface ; near the outer edges of the disk, however, they are
elongate in shape, and at the extreme edge they are again hemispherical
(see Figs. 19, 26, 28, and 32). If a weak solution of pierie acid
in 50% alcohol is allowed to permeate the sea water in which the
medusa is living the nematocysts are exploded with great energy.
Ordinarily they then present the appearance shown in Figure 21, Plate
IX., but occasionally one finds one resembling that shown in Figure 22,
where the main shaft of the thread is tightly coiled in a right-handed
helix, and a small ellipsoidal mass of protoplasm (p) is borne upon the
free extremity. It is very difficult to imagine how such a thread could
be turned outward in the ordinary manner, and it is probable that the
extreme stimulation caused by the picric acid produced an abnormal
discharge of the nematocysts. The nematocysts which exhibited this
peculiar structure were invariably immature in development, and were
only about one tenth as numerous as the ordinary normal ones.

An idea of the color of the medusa may be obtained from Plates II.
and III. In some individuals the general color of the disk is yellowish
with a bluish opalescence, while in others it is decidedly pinkish ; and it
is interesting to notice that when the pinkish ones have been confined in

AGASSIZ AND MAYER: DACTYLOMETRA. 5

an aquarium for a few days they lose their bright color and fade into a
dull yellow. Indeed, if the medusa be kept in an aquarium without
food for about six weeks, its disk shrinks to about one eighth of its
former diameter, and all color fades away until it becomes transparent.
The aboral surface of the disk is thickly sprinkled over with light ochre-
yellow spots, caused by protuberances bearing masses of nettling cells ;
and in addition to these there are sixteen radially arranged regions of
reddish brown spots, lying just above the radial septe of the stomach
(see Figs. 2, 3, and 5). These reddish regions usually extend for about
half way from the periphery of the disk towards the centre, and a closer
examination shows that they are caused by highly refractive rosin
colored pigment granules which are contained in the protoplasm of the
epithelial cells of the disk (see Fig. 20, Plate IX.). In this respect
indeed, according to McKendrick,! they are similar to the red pigment
spots of Chrysaora. The male genital organs are usually decidedly pink
in color, while the ovaries are often yellowish, or ashy gray. The bands
of muscles in the radial septæ are of a glistening white. The ectoderm
of the tentacles is ochre-yellow, and the entoderm is often pink. The
oral fringes are often yellowish, with a delicate shade of opalescent blue
in certain lights. In other individuals, however, they are of a delicate
shade of pink, and they are always sprinkled over with red colored
pigment spots similar to those of the aboral surface of the disk. The
marginal sense organs are usually white in color.

The habits of this medusa have already been described by A. Agassiz,”
who says: “It is somewhat strange that almost all of the Medusæ which
have been observed were found in the brightest sunshine only, or in very
dark nights. Early in the morning, and until about ten o'clock, even
on clear days, meduse do not make their appearance, while from eleven
until one or two o’clock they can be caught in abundance. After this
time they disappear gradually, and late in the afternoon it is rare to see
a single jelly-fish. Between nine and ten o’clock they come to the sur-
face again; and that hour, in fact, is one of the most favorable for col-
lecting, in spite of the darkness.” When the medusæ are confined in
the stale water of an aquarium they often assume the position shown in
Plate V., where the disk is flattened out to an extreme degree, the palps
are spread out over the floor of the aquarium, and the tentacles droop
listlessly downward. In this position the medusa often remains

1 McKendrick, J. G.: Coloring Matter of Medusa. Journ. of Anat. and Physiol.,

1881, Vol. XV. pp. 261-264.
2 North Am. Acal., 1865, p. 49.

6 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

motionless for a long time, and one is almost forced to compare it with
a round flat table supported upon numerous legs.

This medusa makes its appearance on the southern coast of New Eng-
land at about the first of August. The diameter of the disk is then
about 50 mm., and there are only three tentacles between each succes-
sive pair of eye spots, so that it might well be mistaken for a Chrysaora,
were one unacquainted with its future development. The medusæ
increase both in size and numbers until about the middle of Septem-
ber, after which time they begin to disappear, although a few may still
be found as late as the middle of October. The diameter of the disk of
the full grown medusæ is often as great as 250 mm. In common with
many other Discophoræ this species seems to prefer the relatively im-
pure water of bays and narrow estuaries. It is very common in the
upper reaches of Narragansett Bay near Tiverton, while in the relatively
purer water of the lower bay it is generally rarer.

The medusa has also been found at Nantucket (Desor); Naushon
(A. Agassiz); Bermudas (A. 8. Bickmore); between the Bermudas and
the Azores (J. Drayton) ; and a well marked southern variety from Beau-
fort, North Carolina (W. K. Brooks).

Two species of fish have been found to accompany this medusa; one
of these is a Clupeoid, and the other is the young of the common
Butterfish (Stromateus triacanthus). These fish constantly crowd about
the medusa, and so persistent are they in following the jelly-fish that
they often allow themselves to be dipped up in the net along with their
companion.

The relations between the fish and the medusa, however, are far from
symbiotic, for the fish gorge themselves with fragments of the tentacles
and oral fringes, which they tear off from time to time. The medusa, on
the other hand, is not wholly unavenged, for every now and then it suc-
ceeds in stinging to death and devouring one of its persecutors. The
fish which possess this curious habit are rarely more than an inch in
length.

An interesting abnormal specimen of Dactylometra quinquecirra was
found at Tiverton, Rhode Island, in September, 1896. In this individ-
ual there were three oral fringes, six genital organs and sub-genital pits,
twelve marginal sense organs, forty-cight marginal lappets, and thirty-
six tentacles (three between each successive pair of marginal sense
organs).

|
|
|

~

AGASSIZ AND MAYER: DACTYLOMETRA.

Dactylometra lactea L. Agassiz.
Plates XII. and XIIL, and Fig. 10, Plate VII.
Chrysaora lactea F. Eschscholtz, 1829; Syst. der Acal., p. 81, Taf. VII. Fig. 3.
Dactylometra lactea L. Agassiz, 1862; Cont. Nat. Hist. of U. S., Vol. IV. pp.
125, 126, and 166.
Dactylometra lactea E. Haeckel, 1879; Das Syst. der Medusen, p. 517.
Dactylometra lactea R. von Lendenfeld, 1884; Proc. Linnean Soc. New South
Wales, Vol. IX. p. 271.

Dactylometra lactea is the type species of the genus Dactylometra,
which was established by L. Agassiz.!

We shall confine ourselves to describing the differences which exist be-
tween this species and its near ally D. quinquecirra. Figure 35, Plate
XII., shows a young Dactylometra lactea in the stage where there are
but four lappets, and five tentacles between each successive pair of sense
organs. In the mature individual (Fig. 36, Plate XIII), on the other
hand, there are six lappets and five tentacles between cach two sense
organs. The tentacles of the mature individuals arise from the notches
between the lappets. The primary tentacles when expanded may stretch
out to a length of from two to three times the diameter of the bell. The
secondary tentacles, however, are only about one half, and the tertiary
about one quarter, as long as the primary ones.

The general color of the medusa is a milky white ; and the bell is
sprinkled over with light ochre-yellow spots, which are clustered thickly
about the aboral pole. The genital organs are slightly yellowish, and
a faint purplish iridescence is seen playing over the palps. The sense
organs are of a brilliant white. A view of one of these sense organs as
seen from the under side of the bell is given in Figure 10, Plate VII.

Our figures were drawn from life, from specimens obtained in Havana
harbor, Cuba, on February 22, 1893, while on the expedition to the
Bahamas in the yacht “Wild Duck.”

It is probable that this medusa is the species found by Eschscholtz in
the Bay of Rio Janeiro, Brazil.

The habits of this medusa are remarkable for their extreme regularity ;
during the morning hours not one of them is to be found, while at about
four o’elock in the afternoon they suddenly appear in large numbers,

and remain swimming near the surface until long after nightfall. One

1 Cont. Nat. Hist. of U. S., 1862, Vol. IV. p. 125.

8 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

cannot but wonder how these creatures can live, and thrive, in the dis-
gustingly impure water of Havana harbor. It is, however, well known

that the Discophoræ in general seem to prefer the relatively impure

waters of bays to that of the open ocean. Dactylometra lactea is much
smaller than its northern ally, D. quinquecirra. Fully developed speci-
mens measure only 65 mm. across the disk, while the young individual
figured upon Plate IX. was only 40 mm. in diameter.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

AGASSIZ AND MAYER: DACTYLOMETRA. 9

EXPLANATION OF PLATES.

PLATE I.

Dactylometra quinquecirra, from a photograph.

PLATE IL

Side view of Dactylometra quinquecirra, } natural size. Specimen found
at Tiverton, Rhode Island.

PLATE IIT.

Aboral view of a mature male Dactylometra quinquecirra, measuring
190 mm. in diameter. ‘Tiverton, Rhode Island.

PLATE IV.

Oral view of Dactylometra quinquecirra. One of the oral fringes is cut
away in order to expose to view one of the sub-genital pits (see gpt,
Figs. 5, 13, 16, 17, 22, and 24). From a specimen found at Tiverton,

Rhode Island, measuring 180 mm. in diameter.

PLATE V.

Side view of a Dactylometra quinquecirra which has been confined for
a considerable length of time in the stale water of an aquarium. The
bell is greatly flattened and expanded, and the oral fringes and ten-
tacles lie spread out over the aquarium floor, The specimen here
represented is in the “ Chrysaora” stage of development; i. e. there

are only three tentacles, and four lappets between each successive pair

of sense organs.
PLATE VI.

Oral view of an octant of the disk of a mature Dactylometra quinquecirra ;

natural size. The oral fringes are cut off in order to show the shape of
the mouth opening (4); gen, genital organs; gpt, sub-genital pit; me
and me’, muscles of exumbrella; mso, muscles of sub-umbrella ; rp, radial
partitions of stomach; oc, ocular lappets; ¢, tentacular lappets; I, H,
III, primary, secondary, and tertiary tentacles respectively.

10

Fig. 7.
Fig. 8.
Fig. 9
Fig, 10.
Fig. 11,

Figs. 12

Fig. 14.

"an
an

Fig.

Fig. 16.

Fig. 18.
Fig. 19.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

PLATE VII.

Sense organ of Dactylometra quinquecirra, as seen from the oral side of
the disk.

Longitudinal section of the sense organ. opt, “olfactory” pit; con, oto-
lythic concretions ; ect, ectoderm; ent, entoderm.

Otolythic concretions from the sense organ of Dactylometra quinquecirra.

Oral view.of sense organ of Dactylometra lactea.

Section of immature egg from ovary of Dactylometra quinquecirra.

and 13. Young scyphostome of Dactylometra quinquecirra, with two and
four tentacles respectively.

PLATE VIII.

Longitudinal section of Dactylometra quinquecirra through the centre of
two of the sub-genital pits. ge, gastric cirri; gpt, sub-genital pit.
Plan of the stomach of Dactylometra quinqueeirra. r, 7’, 7, 7”, pockets of

the stomach; rp, rp, radial partitions of stomach.

PLATE IX.
Dactylometra quinquecirra.

View looking down upon a small portion of the genital organs to show
gastric cirri (yc).

View looking down upon one of the genital organs, the exumbrella being
removed, and part of the genital epithelium being torn away in order to
expose the opening of the sub-genital pit (gpt).

Side view of oral fringes facing a primary radius.

Aboral view of ocular lappets and tertiary tentacles. Showing the ten-
tacle arising from a cleft in the lappet.

Section 6.6u thick, through one of the red pigment spots of the exumbrella
surface of the disk. The pigment appears as small highly refractive
rosin colored granules in the protoplasm of the epithelial cells. Nema-
tocystic capsules of various sizes are situated between the epithelial cells.

An exploded nematocyst.

A young and probably immature nematocyst exploded under the stimulus
of picric acid.

PLATE X.
Dactylometra quinquecirra,

Longitudinal section through edge of sub-genital pit and centre of one of
the radial septe of the stomach.

One of the wart-like protuberances of the exumbrella surface of the disk,
showing pigment granules and clusters of nematocysts.

Side view of oral fringes seen facing a secondary radius.

View of ocular lappet and tertiary tentacle, showing a slight cleft in the
lappet.

Enlarged view of the tertiary tentacle shown in Figure 26,

Oral view of primary tentacle. The tentacle is hollow.

|
|
|
|

Fig.

Fig. 2

AGASSIZ AND MAYER: DACTYLOMETRA. 11

PLATE XI.
Dactylometra quinquecirra.

One of the wart-like protuberances, consisting of nematocystic capsules
and pigment cells, found upon the oral fringes.

. Free extremity of one of the oral fringes.

Oral fringes cut off to show shape of mouth opening (M).

. Aboral view of primary tentacle showing also dark-red pigment spots

on disk.

. Oral view of radial septum (rp) of the stomach. mso, muscles in radial

septum.
View looking down upon genital epithelium of ovary showing mode of
attachment of eggs of various sizes.

PLATE XII.

. Side view of a young Dactylometra lactea. There are only 4 lappets be-

tween each successive pair of sense organs, instead of 6, as in the mature
medusa. 1%, natural size.

PLATE XIII.

. Oral view of a mature Dactylometra lactea. Natural size.

Dactylometra. Plate, I.

CHILD, PHOTO ARTOTYPE E. BIERSTADT, N. Y.

ope

Bulletin of the Museum of Comparative Zoölogy
AT HARVARD COLLEGE.
Vout. XXXII. No. 2.

ON SOME MEDUSA FROM AUSTRALIA.

By ALEXANDER AGASSIZ AND ALFRED GOLDSBOROUGH MAYER.

Wiru THREE PLATES.

CAMBRIDGE, MASS., U. S. A.:
PRINTED FOR THE MUSEUM.

Avrin, 1898.

No. 2. — On some Medusa from Australia. By ALEXANDER
AGASSIZ AND ALFRED GOLDSBOROUGH MAYER.

Tun meduse described in the following paper were obtained while
accompanying Mr. Agassiz during his recent visit to the Great Barrier
Reef of Australia in April and May, 1896. Our course lay along the
Queensland coast between the Great Barricr Reef and the mainland as
far north as Lizard Island. Unfortunately, the season of the south-
east Monsoon is far from favorable for collecting pelagic animals, as
the winds blow a brisk gale almost incessantly and the water is much
disturbed.

The few hauls of the surface net were all very similar, and brought to
light large numbers of Sagittee and Copepods, and a few Appendicularia,
Doliolum, and Decapod larve. The only Cœlenterates found were
several specimens of a Rhegmatodes, and a Mertensia. Two Dis-
cophoræ were found during our cruise. One of these is a new species,
for which we propose the name Desmonema rosea, and the other is
Crambessa mosaica Haeckel.

Desmonema rosea, nov. sp.
Plate I. Fig. 1.

The genus Desmonema was established by L. Agassiz.1 It contains Cyaneidæ
with eight sense organs, and numerous tentacles which are arranged in eight
bunches arising from the sub-umbrella. The tentacles of each of these bunches
are arranged, one after the other, in a single row. The margin of the bell
possesses eight principal lappets and sixteen to thirty-two secondary lappets.

An oral view of Desmonema rosea is given in Plate I, Figure 1. The bell is
rather flat, being about twice as broad as it is high. The eight primary lappets
are separated from one another by deep clefts, which extend inwards for about
a quarter of the distance from the margin of the bell towards the centre. There
are thirty-two small, smoothly rounded, secondary lappets.

The eight marginal sense organs are sunken in long narrow niches lying in
the oral floor of the sub-umbrella.

The tentacles are arranged in eight crescent-shaped rows, lying between and
alternating with the eight primary lappets of the disk. There is but a single
row of tentacles in each of these crescents.

! Agassiz, L., 1862; Contrib. to Nat. Hist. of U. S., Vol. IV. p. 166.
VOL. XXXII, — NO. 2,

16 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The muscles of the oral surface of the disk are very conspicuous, and are
arranged in sixteen bi-forked bundles, containing both circular and radial
muscle fibres. The bundles lying adjacent to the tentacles are about twice as
broad as those that lie near the sense organs.

The genital organs protrude from the oral surface of the disk as four com-
plexly fimbricated sacs. The four oral fringes are very wide, and their free
edges are sharply folded.

The general color of the substance of the disk is a delicate opalescent yellow,
reminding one of the medieval glass of Venice. “The muscle system of the oral
wall of the bell is of a delicate pink, as are also the genital organs. The oral
fringes are of a most exquisite and delicate rose-color, and the entoderm of the
tentacles is port-wine colored. The sense organs bear an intense orange pig-
ment. The diameter of the disk is 180 mm. The medusa was found swim-
ming in great numbers in Largs Bay, near Adelaide, South Australia, on May
29, 1896.

This species is similar in some respects to Cyanea Muelleriantho, that was
described by Haacke! from the Gulf of St. Vincent. C. Muellerianthe is,
however, smaller than Desomnema rosea ; the shape of the marginal lappets
and muscle bands is different, and, above all, there are several rows of ten-
tacles in each crescent, instead of a single row, as is characteristic of the
genus Desmonema. It also bears some resemblance to Cyanea annaskala, von
Lendenfeld.?

Crambessa mosaica Haeckel.
Plates II. and II.

Cephea mosaica Quoy et Gaimard, 1824; Voyage de l’Uranie, Zoologie, p. 569,
Plate 85, Fig. 3.

Rhizostoma mosaica F. Eschscholtz, 1829; System der Acalephen, p. 59. T. H.
Huxley, 1849; Phil. Trans. Roy. Soc., pp. 422, 482, Plate 38 (Figs. 26, 27),
Plate 39 (Figs. 28-34).

Cutostylus mosaicus L. Agassiz, 1862; Contrib. to Nat. Hist. of U. S., Vol. IV. p.
152. Grenacher und Noll, 1876; Abhandl. Senckenberg. Ges., Bd. X.,p. 58.

Catostylus Wilkesii L. Agassiz, 1862; Contrib. to Nat. Hist. of U. S., Vol. IV. p. 152

Crambessa mosaica E. Haeckel, 1879; Das System der Medusen, p. 622. R. von
Lendenfeld, 1883; Zeitschrift fiir Wissen. Zool., Bd. XXXVIII p. 635. R. von
Lendenfeld, 1884; Proc. Linn. Soc. New South Wales, Vol. IX. Part IL p.299.
R. von Lendenfeld, 1884; Proc. Linn. Soc. New South Wales, Vol. IX. p. 926;
Meduse of the Australian Seas, Part L p. 30, Sydney, 1887. Zeit. fiir Wissen.
Zool., 1888, Bd. XLVII. Heft 2, pp. 213, 231-242, Pls. 19, 21, 23-27.

Although so much has been written concerning the anatomy and histology
of this medusa, no figure of it has as yet been given, if we except the sketch
by Quoy and Gaimard in the Voyage de PUranie, Plate 85, Fig. 3. We therc-

1 Haacke, W., 1887; Jenaische Zeitschrift, Bd. XX. pp. 605-614, Plate 96.
2 R. von Lendenfeld, 18823 Zeit. Wiss. Zool, Bd. 37, p. 465, PI 27-33

AGASSIZ AND MAYER: MEDUSA) FROM AUSTRALIA. 17

fore give a figure on Plate II., and a few details of its structure on Plate III.,
and also a brief description of the Medusa.

When fully expanded the bell is rather flat, being five or six times as broad
as it is high. The aboral surface is thickly covered with small granular
he marginal lappets are very

papilla, which give it a roughened appearance. T
numerous, and their number is not very constant, but there are usually 16
between each pair of sense organs, and as there are 8 sense organs, it would
seem that the normal number of lappets is 128.

An aboral view of one of the marginal sense organs is given in Figure 3,
Plate III. An excellent figure of a longitudinal section has been given by
von Lendenfeld (88, p. 269, Fig. 66).

Four thick pillars extend downwards from the ventral surface of the bell,
and support the brachial disk, or subgenital porticus, as it is often called. The
brachial disk, in turn, bears the eight mouth-arms (von Lendenfeld, ’88, p.
239, Taf. 19, Fig. 10). A drawing of one of these mouth-arms is given in
Figure 5, Plate III.; and it is lettered to correspond with von Lendenfeld’s
Figure 36, Plate 23. The short, simple, upper portion of the arm is indicated
by e, and a, 6, and d show the three wings of the lower arm; a being
ventral, and band d dorsal. A cross section of the lower portion of the arm
taken at niveau ss, Figure 5, is given in Figure 6. Its lettering is similar to
that of Figure 5. A view of the terminal portion of one of the mouth-arms
showing the suctorial mouths, surrounded by double rows of small tentacles, is
given in Figure 4. In life these tentacles keep in incessant motion, and by
this means small particles of food are swept into the numerous suctorial
mouths which open at intervals between the rows of tentacles. Good de-
scriptions of the mouth-arms will be found in the papers of Grenacher and
Noll (76),! and of Hamann (82).2

The color of this medusa is normally cobalt-blue, but, as was discovered by
von Lendenfeld (84, p. 925), a species of Zoöxanthella commonly infests it,
forming dense clusters throughout the jelly ; and when this is the case the
blue color is lost, and the medusa changes to a brown color, varying from that
of white bread to that of coffee. Our figure (Plate II.) shows one of these in-
fested medusw, and it will be seen that the only trace of the normal color is
found in a faint blue line marking the uppermost regions of the suetorial
mouths of the mouth-arms. In the estuary of the Brisbane River on May 21 we
saw a great number of these meduse nearly every one of which was of a deep
cobalt-blue, while now and then one was seen almost white in color, and still
others showed intermediate stages between the deep blue and the white. We
found the white or slightly brownish medusa in the Hawkesbury River near
Sydney on April 4; in the harbor of Cairns, Queensland, April 27; and in the
Brisbane River, May 21. We also found a small dark brown or coffee-colored

1 H, Grenacher und F. C. Noll, 1876; Abhand. d. Senckenberg. Naturf. Gesell.,
Vol. X. p. 146, Plates T., TIL-VIL.
2 O. Hamann, 1882; Jen. Zeit. fiir Naturwis., Vol. XV. pp. 243-285, 3 Plates

18 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

specimen at Lark Opening, near Cooktown, Queensland, on May 4. The
medusa has been found by von Lendenfeld in the harbors of Sydney and
Melbourne. In Melbourne Harbor the specimens are blue, while in Sydney
they are universally brown or coffee-colored (von Lendenfeld, ’88, p. 241).
The medusa is evidently common all along the eastern coast of Australia,

where it congregates in large numbers in the harbors and brackish estuaries.
Von Lendenfeld found a small species of fish, Trichiurus declivis Jenyns, in
“symbiosis” with the medusa in Sydney Harbor. We found the same species
accompanying the meduse collected in Cairns Harbor, Queensland.
The diameter of the disk of full grown meduse is about 250 mm.

|
|

Fig.

Fig.

Fig. 3.
Fig.

Fig.
Fig.

AGASSIZ AND MAYER: MEDUSÆ FROM AUSTRALIA, 19

EXPLANATION OF THE PLATES.

PLATE I.

Desmonema rosea; oral view ; $ natural size.

PLATE II.

Side view of Crambessa mosaica ; % natural size.

PLATE III.
Crambessa mosaica.

Aboral view of marginal sense organ.

Side view of the terminal portion of one of the mouth-arms, highly magni-
fied, showing the suctorial mouths and the furrows bordered by rows of
tentacles.

Side view of a mouth-arm.

Cross section of one of the mouth-arms.

Bulletin of the Museum of Comparative Zodlogy
AT HARVARD COLLEGE,
Vou. XXXII. No. 8.

THE GORDIACEA OF CERTAIN AMERICAN
COLLECTIONS.

WITH PARTICULAR REFERENCE TO THE NORTH AMERICAN FAUNA.

Br Tuomas H. MONTGOMERY, Jr.

Wirn FIFTEEN PLATES.

CAMBRIDGE, MASS., U.S. A.:
PRINTED FOR THE MUSEUM.
APRIL, 1898.

No. 3.— The Gordiacea of certain American Collections with par-
ticular Reference to the North American Fauna, By THOMAS
H. MONTGOMERY, JR.

Ir has been the intention of the author to pursue certain anatomical
studies on Gordius and its allies, but this object could not be immedi-
ately carried out owing to lack of material, and also to the difficulty
encountered in determining the American species. Accordingly it seemed
advisable to describe the American species systematically before entering
upon an anatomical study of them. It is very apparent that the North
American Fauna is very rich in species of Gordiacea, and yet the inves-
tigation of these interesting forms has been almost wholly neglected
by American zoölogists, Joseph Leidy being thus far almost our only
writer upon them; and more species are known from South than from
North America.

The material for this study was mainly derived from three collections.
The Leidy collection, which is the property of the University of Pennsyl-
vania, is the richest of these, and contains some of Leidy’s types; my
thanks are due to Dr. Charles W. Stiles of the Smithsonian Institution
for having kindly forwarded this collection. I would express my obliga-
tions to Dr. Alexander Agassiz for permission to examine the excellent
collection of the Museum of Comparative Zodlogy at Harvard ; and to
Dr. W. M. Woodworth for his trouble in sending me this collection. My
thanks are further due to Dr. Stiles for specimens from his own and
from the Smithsonian collection ; and for other specimens to the following
gentlemen : Prof. E. G. Conklin, University of Pennsylvania ; Prof. Thos.
H. Morgan, Bryn Mawr College; Mr. E. G. Vanatta, Academy of Natural
Sciences, Philadelphia; and to Mr. Satterthwaite of Westtown, Pa.
Further, I would express my thanks to the curators of the Academy
of Natural Sciences, Philadelphia, for the opportunity to examine the
specimens in this collection. In a previous paper, to appear in April,
1898, in Spengel’s “ Zool. Jahrbiicher,” I described two new exotic spe-
cies from American collections,

VOL. XXXII. — NO. 8, 1

24 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Certain of the new species here described were based upon the exam-
ination of a single specimen, but of most of them several individuals
were studied. It is, in my opinion, less adducive to confusion of nomen-
clature to preliminarily separate aberrant forms; even though only single
specimens are known, than to class them all under one name. This
group of worms is difficult of study, the external specific characters are
few, and it appears that it is a group in which many of the species are
in process of transformation, judging from their amount of variability.
Then there are sometimes sexual differences, as well as differences at
various times of life. Bearing these points in view, I have laid par-
ticular stress upon the following systematic characters: the superficial
markings of the cuticle, which have been shown by Villot to present
exceedingly valuable and reliable characters, though even the surface of
the cuticle is in many forms subject to individual variation ; the form of
the posterior end in both sexes ; and, though this is not as reliable as
the preceding character, the form of the anterior end. The presence of
median dorsal or ventral grooves along the surface of the body is a less
reliable character, and still less so is the coloration. The males are
more easily determined than the females; the form of the tail lobes in
Gordius and the presence or absence of hairs or spines in their neighbor-
hood are of much importance ; but in the genius Chordodes the form of
the posterior end of the male is much more uniform than in the males of
the former genus.

A historical review of the previous researches upon the American
species will be followed by the descriptions of the species examined.

A. HISTORICAL REVIEW.

1847. Creplin described Chordodes parasitus, n. sp., from a Brazilian
Acanthoditis, but his brief description is altogether insufficient for pur-
poses of identification.

1849. Gay described from Valparaiso, Chile, a Gordius (G. chilensis)
as follows : “ Gordius gracilis, cinereo-fuscus, obscurus ; capite nigro . . .
es de color pardo morenusco oscuro, son la estremidad anterior del cuerpo
6 la region cefálica negra ; en los manchos la porcion posterior se bifurca
mucho.” I agree with Camerano (’90) that this diagnosis is wholly
insufficient.

1850. Leidy briefly refers a Gordius seen by him to the European
G. aquaticus.

MONTGOMERY: GORDIACEA. 25

1851. Leidy describes G. varius, n. sp., which is distinguished from
@. aquaticus in that “the caudal extremity of the female is trifurcated,
while that of the European species is blunt. The length is from 4 to
12 inches. Of this species there are several varieties from different
localities, which may upon further examination prove to be distinct spe-
cies. , . . A second species of Gordius was obtained by Professor Baird
from a spring in Essex County, New York. It is much more deli-
cate than the former, and from 5 to 7 inches long. The female caudal
extremity is blunt. The male caudal extremity is bifurcate and fringed
with peculiar epidermoid appendages. For this second species the name
Gordius lineatus was proposed.” In the same year Diesing writes of
G. chordodes, n. sp. : “Corpus longissimum teretiusculum crassum rigi-
dum, fusco-brunneum. Caput apice rotundatum. .. . Habitaculum.
Acanthodis glabrata: in cavo abdominis, in Brasilia (Beschke). Spe-
cimen identicum femineum ex aqua in Brasilia hausit (Natterer).”
Diesing’s species is not sufficiently characterized.

1853. Baird mentions two new species from North America, of which
the first is a good species: G. platyura, n. sp. : “Body of a uniform dull
white color, quite smooth, with a depressed line on one side throughout
its whole length, obscurely ringed at unequal distances, narrower at the
anterior extremity and terminating in a broad, flattish tail, which is
slightly bifid. Length of animal 32 inches, breadth of middle of body
about 4 a lino; tail 1 line broad. Jamaica?” Q, fasciatus, n. sp. :
“Epidermis granulated. . . . Body smooth, skin prettily shagreened
with very fine lines crossing each other in opposite directions, of a light
color banded with broad patches of dark brown. Anterior extremity
smaller than posterior, and roughened with raised circular ridges, which
extend for about three lines, and as well as posterior extremity of a very
dark color, almost black. Length 114 inches, breadth about 1 milli-
meter.” This was a female from North America. I regard this descrip-
tion of G. fasciatus as wholly inadequate for purposes of identification,
and on this account judge that Römer (’96) is in error in placing it
as a Synonyme of @. aquaticus Linn, Leidy (°53) mentions Gordit of
a milk-white color as very common in grasshoppers in the vicinity of
Philadelphia.

1855. Möbius gives a good description of Chordodes puosus, n. sp.
from Venezuela ; though he confused the head end with the posterior
extremity: length 212 mm., “er . . . nahm aber während neun Tagen,
die er noch im Wasser, sich träge bewegend, lebte, um 259 mm. zu, so
dass seine Linge, als er todt war, 471 mm. betrug”; tail end swollen,

26 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

cloacal opening large, circular, Head laterally compressed, somewhat
pointed ; cuticle with irregular prominences, which are figured. Or, in
the words of his own diagnosis: “Corpus nigrum verrucosum, medio
cylindricum, utrinque attenuatum depressumque, linea ventrali et dorsali,
quarum parti caudali fasciculi pilorum insiti. Caput ellipsoideum, con-
cavitate frontali. Extremitas caudalis trigona, apice rotundata.” This
is a good species.

1856. Leidy gives a much fuller description of his @. varius. He
also describes G. robustus, n. sp., and places synonymous with it G. seta
Müll., and G. lineatus Leidy.

1857. The same author mentions a collection of 48 Gordii, 525 miles
west of Fort Riley, Kansas.

1858. Leidy notes the occurrence of an embryo of G. varius in Lum-
briculus limosus.

1861. Diesing regards G. platyurus and fasciatus Baird as good
species; as also G. varius Leidy, with which he considers synonymous
the “ G. aquaticus” of Leidy, '51. He considers Leidy’s lineatus and
robustus as doubtful synonymes of G. seta Müll. (aquaticus Linn.). Die-
sing also describes G. subspiralis, n. sp.: “Corpus maris brunneum,
feminæ antrorsum attenuatum, late brunneum, nitidum, iridescens.
Caput annulo obscure brunneo cinctum. Extremitas caudalis maris
subspiralis crucibus furc® terminalis divergentibus, incurvatis, levibus,
plica membranacea semilunari ad basin junctis, feminæ obtusa, subcom-
pressa. . ... In palude cum Siredonibus, copiose, in territorio Kansas
(Hammond).”

1866. Schneider regards G. varius Leidy as a possible synonyme of
G. gratianopolensis Charvet (G. tricuspidatus Meissn.).

1874. Villot regards the following as good species: G. lineatus,
robustus, and varius of Leidy; G. subspiralis and G. chordodes of Diesing ;
and G. chilensis Blanchard. He describes the following new species from
America: — G. æneus, n. sp.: “Extrémité antérieure tronquée, légère-
ment renflée. Ouverture ano-génitale du mâle entourée d’un anneau
brun. Lobes bien développés, séparés par un assez large intervalle.
Brun bronzé. (Les jeunes individus sont d'un blanc jaunâtre uniforme.)
Epiderme divisé en losanges par un réseau de lignes saillantes oblique-
ment croisées”; from Cumana, Venezuela. G. reticulatus, n. 8p.: “ Ex-
trómitó antérieure terminée en pointe aiguë. Diamêtre du corps allant
en grossissant de l’extrémité antérieure à l'extrémité postérieure, qui
se termine en point tronqúee. Ouverture ano-génitale large. Brun
marron. Une bande dorsale et une bande ventrale d’un brun plus

MONTGOMERY : GORDIACEA. 2

foncé. Epiderme aréolé. Aréoles formant un réseau à mailles irré-
gulières et inégales, ayant en moyenne 10 millièmes de millimètre.
Une bordure simple de petites papilles autour des aréoles. . . . Cali-
fornie.” G. prismaticus, n. sp.: “Gréle et aplati. Extrémité posté-
rieure bilobée. Brun pâle. Epiderme aréolé. Aréoles prismatiques et
hexagonales, ayant environ 10 milliemes de millimétre de haut sur 6
millièmes de millimètre de large. Quelques papilles très-petites et très-
espacées. . . . Nouvelle-Grenade. Prairie du plateau de Bogota, par
2,600 mètres d'altitude.” He also describes G. deshayesi, n. sp. from
Venezuela.

1879. Weyenbergh gives inadequate descriptions of the following new
species from South America: G. tenuis, dubius, and acridiorum. In the
same year Leidy gives a good description of his G. robustus, based on
specimens from Maryland.

1881. Oerley regards the following American forms as good species :
G. fasciatus and platyurus Baird, G. eneus and trilobus of Villot.

1885. Jeffrey-Bell mentions G. verrucosus Baird, from Vera Paz, Gua-
temala, and from Costa Rica.

1887. Villot considers G. subspiralis Dies. as a synonyme of G. aqua-
ticus Duj.; and adds: “Tl se peut aussi que le Gordius robustus de
Leidy et mon Gordius reticulatus, établis sur des échantillons recueillis
en Amérique, ne soient que des synonymes du Gordius violaceus de
Baird.

1890. Camerano considers as doubtful species C. chilensis Gay, tenuis
and dubius Weyenbergh, and parasitus (Creplin), and considers that
G. acridiorum Weyenbergh is either a Mermis or Filaria. According to
him Villot’s three species, G. æneus, deshayesi, and prismaticus, are ten-
able, and he gives a description of a female specimen of E, eneus.

1892. Camerano describes a male of @. paranensis, n. sp. from Pal-
meira (Parana).

1893. The same writer says of the occurrence of G. verrucosus Baird
in North America, as reported by Jeffrey-Bell, “Je crois qu’il serait
nécessaire de faire une révision des spécimens américains rapportées a
cette espèce.” He also mentions a specimen of G. varius Leidy, from
Monterey, Mexico, and states “mais on peut douter ... que ces
caractères soient suffisants pour distinguer le G. varius du G. tricuspi-
datus L. Dufour ( G. gratianopolensis Diesing, Villot).” Janda describes
a male of Chordodes brasiliensis, n. sp. from Brazil.

1894. Goeldi mentions the occurrence of the preceding species in
Brazil. Camerano describes the following new species from Paraguay

28 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.
and the Argentine Republic: G. alfredi, danielis, and peracce ; and also
describes female specimens of prismaticus Villot and paranensís Camer.

1895. Camerano describes males of G. latastei, n. sp. from Santiago,
Chile. Römer describes Chordodes variopapillatus, n. sp. from Brazil; a
female of G. violaceus Baird from Arizona, a male of G. aquaticus Linn.
from Brazil, and a female of the same species from Chili.

1896. Römer, in his excellent revision of the Gordiacea, regards the
following American species as tenable: G. platyurus Baird, eneus Villot,
paranensis Camer., Chordodes brasiliensis Janda, Chordodes pilosus
Mobius, and © variopapillatus Rómer; the following as synonymes
of G. aquaticus Linn. : robustus and lineatus Leidy, fasciatus Baird, sub-
spiralis Diesing; G. reticulatus Villot as a synonyme of G. violaceus
Baird ; and the following species as untenable: G. chilensis Blanchard,

‚prismatieus Villot, Chordodes parasitus Creplin, G. parasitus Diesing,
G. deshayesi Villot, and G. verrucosus Baird. According to Römer,
Gordius varius Leidy becomes Ohordodes varius Leidy. In this year
Camerano describes Chordodes balzani, n. sp. from Bolivia.

1897. Camerano (’97") mentions specimens of G. obesus Camer. from
Santiago and Gualaquiza, and of Chordodes bouvieri Villot from Guala-
quiza ; and also describes males of ©. feste, n. sp. from Cuenca. In a
second paper (97”) he describes O. talensis, n. sp. from Bolivia; he notes
G. alfredi Camer. from Tala, G. varius Leidy from Bolivia, and O. peracce
from Tucuman ; and he further describes male and female specimens of

O. brasiliensis Janda from S. Lorenzo and Tala,

CRITIQUE,

There is considerable confusion in regard to the tenability of certain
of the preceding American species, mainly due to insufficient diagnoses,
so that it is necessary to relinquish some of them. Those species which
are insufficiently described, and hence untenable, are in my opinion the
following: Chordodes parasitus Creplin ('47), Gordius chilensis Gay
(Blanchard) (49), G. aquaticus Leidy ('50), G. chordodes Diesing ('51),
G. tenuis, dubius, and acridiorum of Weyenbergh (79). G. fasciatus
Baird ('53) seems also inadequately described. The description of G. sub-
spiralis Diesing is sufficient for identification, though in the descriptive
part of this paper we shall find it to be possibly synonymous with @.
aquaticus robustus (Leidy). Leidy’s description (’51) of his G. lineatus
is really insufficient; but this species is nevertheless tenable, since I
had opportunity to examine the type specimens, and have found them
to differ from any other species of Gordius, G. reticulatus Villot (’74)

MONTGOMERY : GORDIACEA. 29

seems to be also untenable, though it may correspond to @. violaceus
Baird, as is supposed by Villot (87) and by Romer (96); Villot's
G. deshayesi, eneus, and prismaticus seem to me to be also doubtful,
though the last two may be preliminarily retained, since Camerano
has found further specimens which appear to agree with them. Römer
has given good reasons to show that G. verrucosus Baird is too poorly
described to be tenable; and there is the greater reason for relinquishing
this species, since Camerano (’93) has described under the same name
a form which appears to be a Chordodes ; hence this name must be
dropped from the American fauna, if not altogether from the nomen-
clature of the Gordiacea. Leidy’s (56, ’79) descriptions of his G.
robustus are sufficient for purposes of identification, though in the
descriptive part of our paper I shall rank this form as a subspecies of the
European G. aquaticus Linn.

There remain then the following species from the American continents
which appear tenable: Gordius varius Leidy, G. robustus Leidy, E. line-
atus Leidy, G. platyurus Baird, E. eneus Villot, E. paranensis Camer.,
G. alfredi Camer., @. danielis Camer., G. latastei Camer., @. obesus
Camer. ; Chordodes pilosus Möbius, ©. brasiliensis Janda, © peracce
Camer., ©. variopapillatus Römer, ©. balzani Camer., O. bouvieri Villot,
C. feste Camer., and ©. talensis Camer. Thus all of the species described
by Camerano appear tenable, but since most of his descriptions are unac-
companied by figures the reidentification of them is rendered very difh-
cult, The following species are to be regarded as doubtful, if not
even needing elimination: G. fasciatus Baird, G\ prismaticus Villot,
G. deshayesi Villot, and G. verrucosus Baird.

It seems to be questionable whether the specimens of @. violaceus
Baird and @. aquaticus Linn., described by Römer from Arizona and
South America respectively, really belong to these species; but this
point will be more fully criticised in our description of G. aquaticus
robustus (Leidy), and of the two new species, @. platycephalus and G.
densareolatus.

In the following descriptions, unless otherwise specified, it will be
understood that the cuticle has been examined on surface preparations
and sections in Canada balsam.

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

B. DESCRIPTIONS OF THE SPECIES EXAMINED.

1. Gordius aquaticus robustus (Leidy).
Plates 1, 2, and Figs. 13, 16-19 of Plate 3.

G. robustus Leidy, 756, ’79.

G. subspiralis Diesing, ’61.

G. robustus Leidy, Villot, 74.

? G. violaceus Baird, Villot, ’87.

? G. aquaticus Linn., Römer, ’95, ’96.

(Leidy’s 79, types; Leidy coll. no. 5056, Coningo, Maryland.)

Form. The male more slender than the female, in both sexes the body of
approximately the same diameter in its whole extent. Head end (Figs. 2, 6, 9,
10, 11, 17) usually obtuse, especially in the male (Figs. 2, 6, 10), sometimes
somewhat conical in the female (Fig. 17). Occasionally a slight neck constric-
tion is present. Usually no median longitudinal grooves are to be seen. In
some of the larger specimens, especially the females, the whole body is very
much flattened, and these are apparently individuals which have discharged
their ova. The posterior end of the female is obtusely truncated (Figs. 7, 8),
with a faint vertical groove on the termiual aspect; the cloacal aperture is
terminal, and lies in this groove. Figures 9, 10, show depression on the
terminal face of the head.

The posterior end of the male is spirally inrolled (a character of the males
of all the Gordiacea examined), and is furcate. The tail lobes (Figs. 1, 3-5, 19)
are short, nearly cylindrical on cross section except that they are somewhat
concave on their medio-ventral surface, and divergent. The cloacal open-
ing is round and situated anterior to the point of bifurcation of the lobes.
On the ventral side of the anterior ends of the tail lobes is situated a more or
less crescent-shaped transverse cuticular ridge, with posteriorly directed con-
cave edge. This sharp cuticular ridge is postcloacal. Short branching hairs
occur on the surface of the tail lobes as elsewhere on the surface of the body,
but no spicules ; and there is no particular arrangement of the hairs in the
vicinity of the cloacal aperture

Cuticle (Figs. 12, 13, 16). True areoles are absent. In most of the speci-
mens, and especially well marked in the males, the surface of the cuticle is
marked by very fine intersecting lines, and, at greater distances apart, by
broader intersecting raised ridges, which are strictly parallel to the finer lines
which lie in the rhombic spaces demarcated by them, and which themselves
are formed of bundles of fine lines. These larger ridges are seen with low
powers of magnification, but higher powers are necessary in order to detect the
system of finer lines which lie between them. On the cuticle, especially abun-
dant at the ends of the body, are also seen short, thick, and branching hairs,

MONTGOMERY: GORDIACEA. öl

Color. Variable according to age, and apparently also according to locality.
The body varies from a yellowish white to a yellowish brown, or a light
chocolate-brown ; the males are usually darker than the females. The tip of
the head is white, and behind ita broad reddish brown ring is to be seen in
most specimens (not present in some females from Kansas). In all the females
a more or less intense reddish brown ring immediately encircles the cloacal
aperture ; and in some specimens there is a narrower, lighter ring outside of
and separated from the former. In the male a similar dark ring encircles the
cloacal opening, at a little distance from it ; and a spot of deep brown may lie
at the posterior edge of this ring : the posteloacal cuticular ridge is brown, its
posterior edge a much darker reddish brown.

Dimensions. Largest male, 655 mm. ; greatest diameter, 1.3 mm. Largest
female, length 705 mm. ; greatest diameter, 1.9 mm, The males are more
slender and usually somewhat shorter than the females. The individuals from
the western United States (Montana, Kansas) averaged considerably longer
than any eastern specimens examined.

Comparison. This species has been placed by me as a subspecies of @. aqua-
ticus Linn., since the differences do not warrant ranking it as a separate species.
It differs from the true European aquaticus in these points : the presence of a
dark ring around the female cloacal aperture, the absence of a row of hairs
around the male cloacal aperture, and the absence of white spots (“helle
Flecken ”) on the cuticle.

Especial Diagnostic Characters. The approximately equal diameter of the
whole body, with the obtuse truncation of the two ends; the absence of true
areoles, and the presence of short hairs on the cuticle ; the presence of a trans-
verse posteloacal cuticular ridge in the male, the posterior edge of which is
darkest in color, and the absence of a line of hairs around the cloacal aperture.

Jeographical Distribution. I have seen specimens from Maryland, Massachu-
setts, District of Columbia, New York, Maine, Pennsylvania, Montana, and
Kansas ; and Leidy mentions its occurrence at the Bay of Fundy. The speci-
mens of G. aquaticus” mentioned by Römer (95), from South America, are
probably specimens of our subspecies.

2. G. aquaticus difficilis, n. subsp.
Figs. 14, 15, Plate 3.

(Type: 1 male, Leidy coll. 5100, Roan Mt., N. Carolina.)

Form. General form as in the preceding subspecies, cylindrical with the
greatest diameter posteriorly, head (Fig. 14) rounded. Posterior end (Fig. 15)
as in the preceding, but a parabolic line of hairs curves around the cloacal
aperture, the posterior ends of this line of hairs situated upon the latero-ventral
surfaces of the tail lobes, at about the plane of anterior bifurcation of the
latter.

Cuticle. At the tip of the head there are small, round or polygonal promi-
nences or areolw, of slight elevation, and of a deeper brown color than the sur-

32 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

rounding portions of the cuticle. These are densely massed together at the
tip of the head, and a few isolated areolæ occur along the sides of the body
near the head. Similar areolæ, but of larger size, are found also on the dorso-
lateral surfaces of the tail lobes. The cuticle is elsewhere marked by minute
intersecting lines, much as in the preceding species ; when studied in alcohol
the cuticle appears to be areolated in its whole extent, but sections show that
the apparent areole are nothing but slight elevations between the bundles of
intersecting lines, the supposed areolæ themselves being striated by fine lines,

Color, Remarkably iridescent. Whole body a light chocolate-brown color,
the tip of the head the same. There is a dark ring immediately around the
cloacal aperture. The whole surface of the posteloacal cuticular ridge is of a
uniform brown color, without darker posterior edge,

Dimensions. Length, 70 mm. ; greatest diameter, .5 mm.

Comparison. This form resembles, but seems to be distinct from G. aquati-
cus robustus. It differs from the latter as follows : in the presence of areolæ
on the head and tail lobes; in the circumeloacal line of hairs ; in the uniform
dark brown color of the posteloacal cuticular ridge ; in the absence of a
dark ring around the neck, and of a white tip to the head. These characters
seem to warrant placing it, preliminarily at least, as a new subspecies.

3. G. lineatus Leidy.
Figs. 20-31, Plate 4.

G. lineatus Leidy, ’51.

G. robustus Leidy, ’56.

2G. seta Müll., Diesing, "61.

G. lineatus Leidy, Villot, 74.
G. aquaticus Linn., Römer, ’96.

(Types: Leidy coll. 5008, Essex County, New York, 1851.)

Form. Head end (Figs. 20, 21) rounded, not constricted from the body ;
body cylindrical, somewhat narrowed anteriorly. Median grooves absent,
Posterior end (Figs. 22, 23, 28) not swollen, obtusely truncated in the female.

Males somewhat more slender than the female, and somewhat flattened.
Tail lobes (Figs. 24-26) rather long and divergent, their distal ends curved
inwards (ventrad). Cloacal opening elongate, above the lobes, Cuticular
spicules of an elongate conical form on the median surfaces of the tail lobes,
though not on the distal ends of the lobes. Two rows of rather long branch-
ing hairs on the ventral surface of the body; one row on each side of the
median line, each row extending from a little in front of the cloacal opening to
a little behind the point of bifurcation of the tail lobes ; the hairs are longest
in the middle of each line, Tail lobes concave on their medio-ventral surfaces.

Cuticle (Figs. 29-31). Areolated, areolæ closely opposed without inter-
vening spaces ; rectangular or polygonal in outline, frequently elongated in
the direction of the body axis, and with a tendency to group themselves into

MONTGOMERY : GORDIACEA. oo

longitudinal rows which form ridges on the body surface. The areole vary
considerably in form and size.

Color. A pale transparent yellowish white, the female of a deeper buff
color. In the female the cloacal opening is immediately surrounded by a
narrow, reddish brown ring.

Dimensions. Length of largest male, 278 mm. ; greatest diameter, .6 mm.
Length of largest female, 283 mm. ; greatest diameter, .8 mm. The females
are a little longer and broader than the males, but both sexes are very slender,

Comparison. This species stands closest to (Y. violaceus Baird, but differs
from it in the form of the tail lobes, and in the arrangement of the spicules on
them, as well as by its very slender form.

Particular Diagnostic Characters. Very slender and short, of a pale yellowish
or buff color. Areola small, close together, with a tendency to form longitu-
dinal ridges. A line of long hairs to each side of the cloacal aperture in the
male, and spicules on the tail lobes.

Geographical Distribution. New York, Maryland, and one specimen secured
by me in a spring in Chester County, Pennsylvania. Leidy” type specimens
were also found in a spring.

4. G. densareolatus, n. sp.
Figs. 32-33, Plate 4; Plate 5.

(Types: Leidy coll. 5063, Fort Bridger, Wyoming.)

Form of Female. Head end (Figs. 34, 35) conical, terminally rounded,
the terminal portion slightly constricted off; mouth opening terminal. An-
terior portion of the body narrower than the middle and posterior portions.
With more or less pronounced dorsal and ventral median lines. Posterior end
(Figs. 38, 39) slightly widened horizontally, obtusely truncated, with a shal-
low vertical groove on its posterior aspect, in the middle of which the cloacal
opening lies.

Form of Male. Generally similar to but more slender than the female.
The tail lobes (Figs. 36, 37) are short, thick, asymmetrical, and divergent.
The cloacal aperture is small, circular, and immediately enveloped by a dark
ring ; it is situated anterior to the tail lobes, on the ventral surface of the body.
On the antero-ventral surface of the tail lóbes is an integumentary (not purely
cuticular) ridge of slight elevation, the two arms of this ridge converging and
joining just behind the cloacal aperture. The ventro-median surfaces of the
tail lobes are concave. From the cloacal aperture, and embracing it, there
extends cephalad, for a distance about equal to the length of the shorter tail
lobe, a comparatively wide groove on the ventral surface of the body; at each
antero-lateral edge of this groove lies a rounded prominence or ridge. The
ventro-median surfaces of the tail lobes and the posteloacal integumentary
ridge are covered with short conical cuticular spicules, which extend cephalad
to each side of the cloacal aperture.

34 BULLETIN: [MUSEUM OF COMPARATIVE ZOOLOGY.

Cuticle (Figs. 32, 33). Areolated; the areolæ variable in size and form,
usually elongate-oval or irregularly pentagonal in outline. Their longitudinal
axis is usually nearly perpendicular to that of the body, and they tend to pro-
duce transverse rows or chains, in each of which rows some of the areol are
confluent. Narrow, shallow grooves separate neighboring parallel rows of
areolæ ; and beneath the areolæ a system of fine intersecting lines is seen.
No interareolar bristles or hairs are present.

Color. In the female (three specimens examined) the extreme tip of the
head is white, this is followed by a light buff ring, and immediately behind
the latter a broad transverse reddish brown ring. The cloacal aperture is
immediately surrounded by a thin black ring, and around the latter is a much
broader circular area of a reddish brown color. The rest of the body is a deep
yellowish brown in one specimen; a light chocolate color in the other speci-
men. Color of the male (a single specimen) similar to that of the female,
but darker, a deep chocolate color; a nearly black ring surrounds the cloacal
opening, while the postcloacal integumentary ridge is slightly lighter than the
surrounding parts.

Dimensions. Male, length, 290 mm. ; greatest diameter, 1.1 mm. Female,
length of largest specimen, 395 mm., ; greatest diameter, 1.7 mm.

Comparison. This species is quite similar to the European G. tatrensis
Janda, but differs from it in that all the arcole are of a dark color and
there are no groups of areoles forming white spots; and in the male the tail
lobes are shorter and thicker, and there is no “knorriges, glattes, dreiwandiges
Höfchen ” around the cloacal aperture, such as is described by Janda (93).
It also differs from G. violaceus Baird in the manner of distribution of the |
spicules on the tail lobes, and in the confluence of the areolæ. Itis however
most closely allied to G. platycephalus, n. sp. ; these resemblances will be dis-
cussed under the heading of that species.

Especial Diagnostic Characters. The dense arrangement of the irregular
areolæ, which have a tendency to produce transverse rows, and the tendency
to confluence of the arcoles ; the comparatively robust form of the body; the
short, thick tail lobes in the male, with the conical spines on their ventral sur-
face, the obscure posteloacal integumentary ridge, and the ventral depression
within which the cloacal aperture lies.

Geographical Distribution. The type specimens (2 males, 1 female) are
from Fort Bridger, Wyoming; and another female from South Montana (coll. |

Acad. Nat. Sci. Philadelphia).

5. G. platycephalus, n. sp. |

Plate 6; Figs. 46-49, Plate 7. |

(Type of female : coll. Acad. Nat. Sci. Phila., Guatemala. Type of male:
coll. Acad. Nat. Sci. Phila., South Montana.)

Form of Female. Anterior portion of the body attenuated and flattened in

all.specimens, and the head constricted from the body. The head (Figs. 42, 6,

MONTGOMERY: GORDIACEA. 35

c, 45) is of slightly greater diameter than the part of the body immediately
preceding, tip of the head rounded. or truncated; in one specimen. a groove on
the ventral surface of the head, Body either flattened or cylindrical, with
deep dorsal and ventral median grooves. In all females except one the pos-
terior end (Figs. 42 a, 43, 44) of the body is somewhat constricted for a dis-
tance of about 6 mm., but the extreme posterior end is usually swollen, some-
what knob-shaped ; on lateral view this end appears obliquely truncated, the
posterior end of the body has a vertical groove on its terminal aspect, and the
cloacal aperture is not exactly terminal, but somewhat ventral.

In the male the anterior portion (Fig. 40) of the body is not attenuated,
though it is slightly flattened horizontally; the head is elongate-oval in out-
line, of greater diameter than the part immediately preceding, and is terminally
rounded. The body like that of the female, but more slender. Tail lobes
(Fig. 41) rather slender and long, asymmetrical, their distal ends curved
ventro-mediad ; they are nearly cylindrical, flattened only on the median. sur-
face. The large, elongate cloacal aperture is situated on the ventral surtace of
the body, and separated from the anterior point of bifurcation of the tail lobes
by a distance equal to half the length of the tail lobes. This aperture does not
lic in a groove ; nor do spicules nor long hairs occur near it or on the tail lobes,
but only minute, short hairs.

Cuticle (Figs. 46-49). Areolated ; the arcole slightly smaller than those
of G. densareolatus, more or less of the same size, and either irregularly
polygonal or somewhat elongate in outline, and then usually elongated in the
direction of the transverse axis of the body. The areole are usually well
separated from one another, except in the median line, and show no tendency
to produce confluent rows. Small interareolar groups of small bristles occur
in most of the individuals, these bristles varying in number and form.

Color. Brown, varying in shade, but never very intense ; tip of head lighter,
and a more or less pronounced dark ring around the neck. In the male an
obscure brown ring immediately surrounds the cloacal aperture,

Dimensions. Length of male, 216 mm.; greatest diameter, 1 mm. Length
of largest female, 335 mm.; greatest diameter, 1.4 mm.

Comparison. In the configuration of the cuticle this species is most closely
allied to G. violaceus Baird, and to G. densareolatus mihi. The males of these
three species are very different, however, in regard to the arrangement of the
spicules on the posterior end, such spicules being absent in platycephalus.
This character does not serve to distinguish the females of these species how-
ever, though the flattening of the anterior portion of the body is diagnostic of
platycephalus. But I am wholly at a loss to classify one female from Montana
in the collection of the Acad. Nat. Sci.: it has the flattened head of platyceph-
alus, with the confluent areolæ of densareolatus ; its color is a deep buff, with a
narrow black ring immediately around the mouth, but with no dark ring
around the neck; the shape of the posterior end and the deep median grooves
of the body resemble platycephalus, so that on the whole I should be inclined
to consider it as platycephalus. But might not this specimen be a hybrid be-

36 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

tween these two species? Though the females of violaceus, densarcolatus, and
platycephalus are so similar, the males are nevertheless very different in regard
to the form of their posterior ends, so that these species may be regarded as
genetically related, and we must consider that the course of modification which
they have undergone has influenced the males more than the females. This
view of the question cannot be regarded as bizarre, since in other groups of
animals also the males are in some cases far more dissimilar than the females.

Lispecial Diagnostic Characters. The flattening of the anterior end, and the
constriction of the head from the body ; the slight enlargement of the posterior
end in the female; the absence of spicules on the tail lobes in the male; the
presence of small, rounded-polygonal cuticular areoles, which are as a rule well
separated from one another,

Geographical Distribution. Guatemala, South Montana, Pennsylvania,
Bridger Basin, Fort Laramie. 1 male and 7 females examined.

6. G. platyurus Baird.
Figs. 50-52, Plate 7.

G. platyura Baird, ’58.

G. platyurus Baird, Diesing, ’61,
G. platyurus Baird, Villot, 74.
G. platyurus Baird, Oerley, '81.
G. platyurus Baird, Römer, ’96.

(1 female examined : Leidy coll. 5096.)

Form. Very massive, flattened dorso-ventrally, with broad dorsal and ven-
tral grooves, which do not extend quite to the posterior end. Head end (Fig.
52) conical, Largest diameter posteriorly. The posterior end (Fig. 51) is
dorso-ventrally flattened, expanded, wider than the preceding portion of the
body, spatulate in form, with shallow dorsal and ventral depressions. The
cloacal aperture is terminal.

Cuticle (Fig. 50). With fine intersecting lines, much as in G. aquaticus
Linn. Here and there bundles of elevated lines are demarcated from the finer
lines, and these bundles, which are parallel to the finer lines, deliminate
rhomboid-shaped spaces.

Color. A light yellowish buff, somewhat iridescent, Extreme tip of head
white, behind which is a faint brownish ring. Posterior tip of the body a light
yellowish white.

Dimensions. Length, 540 mm.; greatest diameter of body, 2.4 mm. ; great-
est transverse diameter of tail, 2.8 mm.

No locality is marked for this specimen ; the only other specimen known,
the type in the British Museum, is labelled “ Jamaica?” so that it is not yet
proved that this species is American. It seems to me probable that it does not
come from the North American continent, since otherwise there would proba-
bly be numerous examples of this large species extant. Its massive form and
the spatulate shape of the posterior end are good diagnostic characters.

MONTGOMERY: GORDIACEA. ot

7. G. leidyi, n. sp.
Figs. 53-55, Plate 7; Figs. 56-59, Plate 8,

(1 female, type: Leidy coll. 5089, no data.)

Form. Head (Fig. 53) conically pointed, obtusely rounded at the tip, where
the mouth opening is terminal and forms a slight projeetion. Body eylindri-
cal, with deep dorsal and ventral grooves, the ventral one not quite in the
median plane. Posterior portions of the body for a distance of about 12 mm.
slightly flattened horizontally, and of slightly less diameter than the middle of
the body. The dorsal groove of the head extends nearly from the head to the
posterior end of the body ; the ventral groove ends about 5 mm. in front of
the posterior end. Posterior end (Figs. 56-58) truncated, almost vertically.
Cloacal aperture (Fig. 57) terminal, nearly in the centre of the disk which
forms the distal face of the posterior end. This aperture is placed upon a
round, elevated papilla, the latter sunk in a depression of the distal face of the
body. To each side of the cloacal papilla is a short vertical, elevated integu-
mentary ridge, while below the depression in which the cloacal papilla lies is a
transverse ridge, On the dorsal side of the posterior end (Fig. 56) is a nearly
U-shaped integumentary fold, to each side of which is an elongated pit or
depression, The latter pits lie respectively on the dorso-lateral sides of the
posterior end of the body, and extend posteriorly as far as the vertical tegu-
mentary ridges which form the lateral boundaries of the depression in which
the cloacal papilla is situated. The distal end of the dorsal U-shaped fold
forms a transverse ridge bounding dorsally the depression in which the cloacal
papilla lies. The dorso-median groove of the body extends on the surface of
this fold to the posterior end of the latter, the groove being broadest at this
point, Thus the posterior end of the body is vertically truncated, with a con-
cave distal face ; in the centre of this depression, situated on the summit of a
slightly elevated papilla, lies the cloacal aperture ; the outlines of this terminal
concavity of the body form nearly a square, its boundary being two short
vertical ridges, a transverse ventral ridge, and a transverse dorsal ridge which
is the distal end of a U-shaped fold of the integument situated on the dorsal
side of the end of the body. These relations are somewhat complicated, and
may be best understood by reference to the Figures 56-58,

Cuticle. Aveolated ; areolæ (Figs. 54, 55) only slightly elevated, on cross
section they show no hyaline summit, irregularly polygonal in outline ; they
are usually elongated in the line of the transverse axis of the body, and show
a tendency to form short and interrupted rows or chains, contiguous areoles in
such rows being confluent. The areolæ are separated only by narrow spaces.
The cuticle is also marked by intersecting elevated lines (Fig. 59) placed at
considerable but varying distances apart, these lines being easily visible on the
alcoholic specimen. Interareolar bristles are absent,

Color. Head end light yellowish brown, the rest of the body a deep yellow-
ish brown color, with an obscure dusky brown ring on the head. In the

VOL. XXXII. — NO, 8. 2

38 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

median line of the dorsal groove of the body are two narrow, parallel stripes
of an intense reddish brown color, which are in contact with one another in
the median line. In the ventral median groove of the body are two parallel
stripes of about the same color as those in the dorsal groove, but with this
difference, that they are separated from one another by a distance about equal
to the diameter of either of the lines, and that the two ventral stripes are of
not quite the same diameter. Both the dorsal and ventral stripes of color
disappear near the two ends of the body.

Dimensions. Length, 295 mm.; greatest diameter, 1.5 mm.

This species is sharply distinguished from all other species of the genus
known to me by the peculiar form of the posterior end, and by the colored
stripes lying within the median grooves of the body. Unfortunately the
locality is not given for the specimen. I have named it in honor of the pioneer
student of the American Gordiacea, Joseph Leidy.

3. G. agassizi, n. sp.
Figs. 63-66, Plate 9.

(Type, 1 male : Harvard collection no. 296, Sandwich Isl.)

Form. Body much flattened dorso-ventrally, without well marked median
grooves; the body is flattened in such a way that the dorsal side is rounded,
the ventral side concave, so that a cross section would show the body to be
somewhat sickle-shaped. Head end likewise dorso-ventrally flattened. The
particular characteristic of this species, however, is that the broad plane of the
neck does not coincide with the broad plane of the head, but is nearly vertical
to it (Fig. 63). Thus the flattening of the head and of the body lies in approx-
imately the same plane, while the neck region (for a distance of about .9 mm.)
is twisted around through an angle of nearly 180 degrees, and hence the broad
plane of the neck is nearly vertical to that of the head and to that of the body.
Hence in viewing the head end either from the dorsal (Fig. 63) or the ventral
aspect, the neck appears like a short slender thread, This relation has prob-
ably been produced by a torsion of the neck region through an angle of nearly
180 degrees. The outline of the head, viewed from the flattened surface, has
somewhat the shape of an unbarbed arrow-head, broadest posteriorly and with
rounded tip ; the mouth opening is large, transversely elongated, and situated
at the termino-ventral margin of the head. The neck inserts itself along an
elevated ridge which cireumscribes the posterior portion of the head. On the
dorsal surface of the head (Fig. 63), to each side of the median line, are found
just behind this elevated ridge a number of small ridges which are parallel to
one another (i. e. those on the same side of the ridge are parallel), and these
are obliquely disposed to the large ridge. These relations are difficult to
describe, but may be understood by comparing the figure.

The tail lobes (Figs. 65, 66) are bent ventrad, nearly at right angles to the
posterior end of the body region proper. Hach tail lobe is much flattened
laterally, and is very short; it is somewhat rounded on the lateral, and corre-

MONTGOMERY: GORDIACEA. 39

spondingly concave on the median side. Viewed from the side (Fig. 65)
each shows a more or less conical outline, broadest at the proximal end,
rounded at the distal end. The dorso-median margin is slightly thickened.
The flattened planes of the two lobes are not parallel, their dorso-median
margins being much closer together than the ventro-median; the proximal ends
of their dorso-median margins are in contact, while the ventro-median margins
are farthest apart proximately (Fig. 66). The median plane between the
two lobes is vertical to the flattened plane of the posterior end of the body
proper. The tail lobes may well be termed leaf-shaped,

Cuticle (examined in alcohol only, since it seemed inadvisable to section the
single specimen at hand). The surface (Fig. 64) viewed with low powers
of the microscope, shows very plainly a system of deep intersecting lines, be-
tween which lie slightly elevated areolæ of rhombic or rhomboid outline.
There are necessarily two systems of parallel lines; and in one of these two
systems the lines tend to occur in pairs.

Color. The body is a uniform rufous-brown color, the tail lobes somewhat
lighter. The mouth region of the head is yellowish, behind which follows a
zone of a nearly black color; the posterior portion of the head is but little
darker in color than the body.

Dimensions. Length, 158 mm. ; greatest diameter of head, 1 mm.; greatest
diameter of body, 1.5 mm.

Comparison. This species may be very sharply distinguished from any
other Gordiacean yet described, by the torsion of the neck through an angle of
nearly half a circle, and by the extreme flattening of the tail lobes. It is the
only species known from the Sandwich Islands.

9. G. capitosulcatus, n. sp.
Figs. 67-69, Plate 9; Fig. 70, Plate 10.

(Type, 1 male: Harvard coll. 1466 a, Cuba.)

Form. The body is dorso-ventrally slightly flattened, with slight dorsal
and ventral median grooves. The head (Figs. 67, 69) is flattened laterally,
higher than broad, and is separated from the body by a slight constriction
(neck). The head has the greatest diameter at the anterior end, where it is
obliquely truncated, the dorsal margin projecting slightly farther forwards
than the ventral. The terminal aspect (Fig. 67) of the head is concave, the
large mouth opening situated in the median line, nearer the ventral than the
dorsal margin of the head. At each anterior dorso-lateral margin of the head
is a ring-shaped prominence (Fig. 69), which surrounds a pit-like depression.
It would be impossible to determine the structural significance of these pits
without sectioning the head, but this was not permitted on the single speci-
men examined,

The tail lobes (Fig. 70) are slightly divergent; each is terminally rounded,
nearly cylindrical on cross section, but concave on the median surface. The
lobes are long and slender, and apparently bear no hairs or spicules. The

40 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

cloacal aperture is situated on the ventral surface of the body, a short distance
in front of the point of union of the two lobes.

Cuticle (Fig. 68). With small elevated areoles, situated close together.
The areoles are somewhat variable in size and form, but are mostly rounded-
polygonal in outline. Their surface is not smooth, as in the other species of
the genus examined by me, but with irregular short tubercles. Interareolar
bristles are apparently absent.

Color. Pitch-black to the naked eye, but with a brownish tinge when
viewed with the microscope. The margins of the ring-shaped prominences of
the head are of a whitish color.

Dimensions. Length, 165 mm.; greatest diameter, 9 mm.

Especial Diagnostic Characters, The presence of depressions on the anterior
dorso-lateral margins of the head ; the roughened surface of the areoles; the
intense black color.

Comparison. This form differs from G. violaceus, densarcolatus, and platy-
cephalus by the roughened surface of the areoles, the coloration, and the
presence of the pits on the head. It has no close resemblance to any of the
species described by Camerano from South America, and on the whole appears
to be a well defined species.

10. G. paranensis Camer.
Wigs. 71-74, Plate 10.

G. paranensis Camerano, ’92, 04.
G. paranensis Camer., Römer, *96.

(1 female, 5 males: Harvard coll. no. 1478, Casabianca, Chile.)

Form of Female. Body somewhat flattened dorso-ventrally, without well
marked median grooves, Head cbnical, concave on the terminal aspect,
mouth terminal ; head not constricted from the body. Posterior end (Fig. 71)
truncated, with a circular depression on its terminal aspect, in which the
cloacal opening lies ; this posterior end of the body is round on cross section,
while the immediately preceding portion of the body is much flattened dorso-
ventrally.

Form of Male. Body more slender than in the female, and with more or
less well marked median grooves. In one male (Fig. 74) the head end is
conical with rounded tip; in the others it is separated from the body by a
slight constriction, and is terminally truncated, the terminal face concave
(Fig. 73) ; on this terminal aspect of the head in one specimen is a vertical,
median ridge, to each side of which is a depression. A conical and a truncate
form of head being found in different specimens of this species would lead to
the conclusion that the truncate form, in which the anterior aspect of the head
is concave, is probably due to a muscular contraction of the tip of the head.
The tail lobes (Fig. 72, a, b) are comparatively short and thick, flattened on
their proximo-median surfaces as well as on their dorso-lateral surfaces,

MONTGOMERY: GORDIACEA. 41

Apparently neither hairs nor spicules occur in the vicinity of the lobes. At
the anterior point of union of the tail lobes, on the ventral surface of the body,
is situated a V-shaped ridge, each arm of which is placed on the ventro-median
margin of the corresponding tail lobe, the two arms of the V converging, and
joining at an angle cephalad at the anterior point of union of the tail lobes.
Anterior to this ridge is situated a broad and deep depression. At the deepest
part of this depression, i, e. in the medio-ventral line just anterior to the ridge
described, is situated the cloacal aperture; this aperture lies on the summit of
a slightly elevated papilla. ‘The anterior margin of the depression which sur-
rounds the cloacal papilla forms a sharp ledge, irregularly semicircular in out-
line (the opening of the semicircle directed caudad); this sharp ledge, forming
the anterior and lateral margin of the cloacal depression, is not elevated above
the level of that portion of the ventral surface of the body which lies anterior
to the depression. Along this ledge are arranged a row of short, thick hairs.
Accordingly, we find on the ventral surface of the posterior end of the male a
narrow V-shaped ridge at the base of the tail lobes; anterior to this a large and
deep depression, in the centre of the posterior part of which the cloacal papilla
is situated; and the anterior and lateral boundary of this depression, formed by
a nearly semicircular sharp ledge.

Cuticle. There are no areoles, but a system of broad intersecting, oblique
lines, between which are much finer intersecting lines. Short hairs also occur
here and there, but sparsely. The cuticle is thus very similar to that of G.
aquaticus robustus (Leidy).

Color. Body of a dull olive-brown color. Tip of the head yellowish white,
the remainder of the head reddish brown, varying in shade. The posterior end
of the female is yellowish ; the pit in which the clpacal aperture lies is of a
deep reddish brown color, there being thus a disk of this color immediately
around the aperture. Inthe male the postcloacal cuticular ridge is reddish
brown, its posterior edge black,

Dimensions. Length of female, 470 mm.; greatest diameter, 1.8 mm.
Length of largest male, 340 mm. ; greatest diameter, 1,3 mm.

Especial Diagnostic Characters. The presence of a posteloacal ridge in the
male, and the presence of a sharp precloacal ledge which bounds the cloacal
depression; the intersecting lines of the cuticle.

Comparison. These specimens seem to agree wholly with Camerano’s
description of the species. This form is most closely allied to G. aquaticus
robustus (Leidy).

Geographical Distribution. Asuncion, Paraguay; Palmeira (Parana); Casa-
bianca, Chile.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

11. G. violaceus Baird.
Figs. 60-62, Plate 8; Figs. 75-77, Plate 11.

G. violaceus Baird, ’53. (For the synonymy of descriptions of specimens from
other localities than America, cf, Römer, ’96.)

2 G. reticulatus Villot, 74 (from California).

2 G. violaceus Baird, Villot, 87.

2 G. violaceus Baird, Römer, ’95 (from Arizona).

2 G. violaceus Baird, Romer, *96.

(1 female, Harvard coll. no. 1465, California ; 1 female, Harvard coll. no.
1666 c, Cuba.)

Description of the Califormian Specimen. Body cylindrical without median
lines. Posterior and especially the anterior portions of the body somewhat
narrower than the middle; head (Fig. 61) of a rounded conical form, not con-
stricted from the body, mouth terminal. Posterior end (Fig. 60) of the same
diameter as the immediately preceding portion of the body, obtusely rounded
terminally; the small, round cloacal aperture is terminal. Cuticle (Fig. 62)
areolated: the brownish areole vary somewhat in size, are irregularly polyg-
onal, and do not form rows but are well separated from one another; a few
small interareolar bristles are present. Color: a clear chocolate-brown, head
paler, a deep reddish brown ring around the mouth. Length, 130 mm. ; greal-
est diameter, 9 mm.

Description of the Cuban Specimen. Body nearly cyclindrical, with dorsal
and ventral median grooves. Head cyclindrical, narrower than the portion of
the body immediately preceding, terminally truncated (Fig. 76) ; this plane of
truncation is slightly convex, the mouth situated in its centre. The anterior
portion of the body is gradually attenuated, of less diameter than the middle
portion. Posterior end of the body (Fig. 75) obliquely truncated, in that the
dorsal margin projects farther caudad than does the ventral margin; the pos-
terior end of the body is flattened on its ventral surface. In the medio-ventral
line is a shallow narrow groove, which passes dorsad on the terminal aspect of
the body, this groove being deepest at the dorso-terminal margin of the body.
Within this groove lies the terminal cloacal aperture, Cuticle (Fig. 77
areolated : areolæ low, irregularly rounded in outline, smooth superficially,
close together; the areolæ are very little darker than the inter-areolar spaces;
between them lie thick, conical hairs, which are higher than the areole.
Color a uniform grayish brown, head lighter; the mouth is surrounded by a
narrow reddish brown ring, and the vertical groove at the posterior end of the
body is also of this color. Length, 112 mm.; greatest diameter, 1 mm.

Comparison, There is some doubt in my mind whether these specimens
«should be attributed to G. violaceus Baird, but they certainly come closer to this
species than to any other, and until further specimens are examined from these
localities may best be placed under this species. But we know that platy-
cephalus and densareolatus come very close to violaceus in the structure of the

MONTGOMERY: GORDIACEA. 43

cuticle, but differ markedly in the form and armature of the posterior end of
the male; and so it may be that the males of these ‚two specimens, when dis-
covered, may also be found to differ from the males of violaceus. But it would
be inadvisable to classify these two females as a new species until the males
are known.

The Californian form agrees very closely with the G. reticulatus of Villot
(74), also from California, but the male of the latter is likewise unknown, so
that reticulatus must still be regarded as a doubtful species. Villot (87) and
Römer (’95, 96) hold the view that reticulatus may be synonymous with viola-
ceus. The males of specimens from all these localities must: first be examined
before we can decide whether the true violaceus really occurs in America, or
whether a subspecies or different species, distinct also from platycephalus and
densareolatus does not take its place. Hence these two doubtful female
specimens from Cuba and California, may only preliminarily be placed under
violaceus.

PARAGORDIUS, n. gen. (cf. the Appendix).

(Type of the genus: Gordius varius Leidy, *51, 56.)

Generic Characters. The cloaca in the adult female is remarkably long
(Fig. 86), nearly half an inch in length, and the caudal ganglion (Fig.
79, N. Gl.) is in direct connection with the cloacal epithelium, and at no
point with the epidermis. The male is characterized by the absence of a
cloacal musculature (Fig. 78). The trilobation of the posterior end of the
female (Vigs. 88-90) possibly also furnishes a true generic character.

Thus Paragordius differs anatomically more widely from Gordius and Chor-
dodes, than the last two do from each other; for in both of the last two the
female cloaca is very short, usually a fraction of a millimeter, and the caudal
ganglion is never in contact with the cloacal epithelium, and in these also
the male always possesses a cloacal musculature (Fig. 18). I am inclined to
suppose that the European Gordius tricuspidatus (Dufour) (G. gratianopolensis
Dies.) should be placed in this new genus, since its female has also a triloba-
tion of the posterior end. But unless the latter species be found to show also
the anatomical generic characters given above, it must be kept separate from
Paragordius, since it is doubtful whether the mere trilobation of the posterior
end constitutes a good generic character, for we find in the female of G. tolo-
sanus Duj. a tendency to bilobation of the posterior end of the body. I have
had no opportunity to examine @. tricuspidatus, and find no description of the
anatomical structures at issue, so that the generic position of this European
species must still remain doubtful, though it certainly should not be placed
under Chordodes, as Romer (96) has done, since the male has the typical bilo-
bation of the posterior end shown by all true Gordiz. I quite agree with
Janda (93) that the shallow ventral groove on the posterior end of the males
of Chordodes is one of the important characters of the genus, since I have found
this typical form of the posterior end in all male Chordodes examined by me,

44. BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

and it has been found characteristic for all the males of Chordodes which have
been heretofore described. The characters of these three genera of Gordiacea
may be compared as follows.

Gordius. Posterior end of the male bilobed, a cloacal musculature present;
posterior end of the female rarely much swollen, never cleft (except in G. tolo-
sanus, where it is deeply grooved rather than cleft); the caudal ganglion in the
female not in contact with the cloacal epithelium, the cloaca very short; cu-
ticula marked with intersecting lines or with low areola, never with high tuber-
cles or papillæ, though frequently with short hairs or conical processes (spicules),

Chordodes. Posterior end of the male not bilobed, but only with a compara-
tively shallow groove on the ventral surface; posterior end of the female usu-
ally swollen, never cleft ; the caudal ganglion in the female not in contact with
the cloacal epithelium, the cloaca very short; the male with a special cloacal
musculature; cuticle usually marked with high tubercles or papille, and
apparently always with hyaline club-shaped processes, which are very differ-
ent from the interareolar hairs of Gordius.

Paragordius. Posterior end of the male bilobed, no cloacal musculature ;
posterior end of the female trilobed ; the caudal ganglion in the female is in
contact with the cloacal epithelium, and the cloaca is long; cuticle as in Gor-
dius, except that the papilla are enveloped in a hyaline layer, which forms the
external layer of the cuticle (Fig. 91, a).

These three genera appear to be very natural groups, and each is to be dis-
tinguished by the union of certain characters, rather than by the presence of
any single character. Of the three, Gordius occupies an intermediate position,
with relations on the one hand to Chordodes, on the other to Paragordius,
though it shows the greater affinity to Chordodes ; while there are no good
characters in common between Chordodes and Paragordius. Thus Gordius
might be regarded as the more primitive parent form, from which the others
have differentiated ; but I reserve a discussion of this point for a subsequent
contribution.

The following preliminary note on the caudal ganglion of Paragordius (in
the female) may be of anatomical interest. This ganglion lies in contact with
the cloacal epithelium at the anterior point of bifurcation of the two lateral
tail lobes (Fig. 79). The posterior margin of this ganglion forms a thin verti-
cal lamina, which may be in contact with that portion of the epidermis lying
between the two lateral lobes, but it certainly does not terminate in contact
with the epidermis of the ventral surface of the body. The ventral nerve
chord anterior to the caudal ganglion lies in the ventro-median line between the
intestine and the longitudinal musculature of the body wall, as in both Gordius
and Chordodes, as far as the latter genera have been examined. But there are
certain small nerves which take their origin from the antero-dorsal margin of the
caudal ganglion, and these nerves lie directly beneath the cloacal epithelium.
These relations were studied on sections of two females; and anterior and pos-
terior nerves may be distinguished with reference to the course which they pur-

sue. The anterior nerves, which are of greater diameter than the posterior

MONTGOMERY : GORDIACEA. 45

ones, varied in number in the two specimens sectioned, In the one, two
nerves arise from the dorso-lateral margin of the ganglion, and may be traced
cephalad for a number of sections; they diverge only slightly; in the other
specimen there is, in addition to the two lateral anterior nerves, also an un-
paired median nerve of greater diameter than the other two, which bifurcates
at its anterior end. The posterior nerves arise a couple of sections behind the
anterior ones, and are two in number (one on each side of the median line),
though in one of the specimens there appeared to be two on one side and one
on the other; these posterior nerves pass caudad, diverging in their course, and
may be traced into the lateral tail lobes, where they divide into still smaller
nerves.

In one male sectioned an elongated cuticular penis was present in the cloaca,
this being only the second case of a penis being observed in a Gordiid, the other
case having been observed by Vejdovsky. But the description of these inter-
esting anatomical details must be postponed to a later paper.

12. Paragordius varius (Leidy).
Figs. 78-85, Plate 11; Figs. 86-93, Plate 12,

Gordius varius Leidy, ’51, 56, ?58.

. varius Leidy, Diesing, 1861.

. gratianopolensis Charvet, Schneider, ’66.
G. varius Leidy, Villot, 174.

G. trilobus Villot, Oerley, ’81.

G. varius Leidy, Camerano, *93.
Chordodes varius Leidy, Römer, ’96.

8 Q

(Leidy’s original type specimens have apparently not been preserved.)

Form of the Female. The anterior and posterior portions of the body are
narrower than the middle, the decrease in diameter being very gradual; the
anterior is narrower than the posterior end. The head end (Figs. 83-85) is
obliquely truncated in such a way that the antero-ventral margin projects
farther forward than does the antero-dorsal; this truncated plane, which
forms the terminal aspect of the head, is very nearly flat. The mouth lies
near the ventral edge of the truncated plane. The posterior end (Figs. 88-90)
is trilobed, there being one dorso-median and two latero-ventral lobes; these
lobes have no cuticular spines on their surface, and in the great majority of the
numerous specimens examined are of equal length. Two specimens in the
Harvard collection were exceptions to this equality in length : in one the dorsal
lobe was slightly longer than the others, in the other slightly shorter. But the
dorsal lobe is narrower than the others, and further differs from the latter in
having an elevated median ridge on its ventral surface, so that on cross section
it appears triradiate (Figs. 80, 81). The lateral lobes are crescent-shaped on
cross section. The cloacal aperture, wholly hidden by these lobes, lies at their
base and between them, so that the cuticle and epidermis of the inner surface
of the lobes are directly continuous with the cuticle and epithelium of the

46 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.
cloaca (Fig. 86). The lobes may be either parallel or divergent, and hence
are probably movable.

Form of the Male. The anterior end (Fig. 82) as in the female, but the
body more slender. The tail lobes (Fig. 87) are comparatively long and
slender, cylindrical in shape, and obtusely rounded terminally. Small conical
spicules occur on the medio-ventral surfaces of the anterior half of the lobes,
and short hairs on their anterior surfaces. The elongate cloacal aperture lies in
the medio-ventral line of the body, anterior to the tail lobes.

Cuticle. On cross section (Fig. 91, a) an outer thin hyaline layer is seen, and
an inner, much thicker fibrous layer, Embedded in the hyaline layer are small
lozenge-shaped bodies, which stain more deeply than any other portion of the
cuticle, and which correspond to the areolæ seen on surface views. The exter-
nal surface of the hyaline layer of the cuticle is marked by short conoidal
processes of the same structure as the hyaline matrix; these are not seen on
surface views. On surface view (Figs. 91,5-93) the cuticle appears areolated:
the areola are small, variable in size and form, and irregularly arranged.
Sometimes they occur in groups, sometimes in interrupted longitudinal rows ;
their arrangement varies both in different individuals as well as on different
portions of the same individual. The areoles are irregularly polygonal in
outline. In one female larger brown-colored areoles were present in addition
to the smaller, lightly colored ones; the former were mainly arranged in the
form of longitudinal ridges, and were irregularly star-shaped in outline.

Color. Color usually lighter in the females than in the males, varying from
%a light brown or yellowish to a dark brown (the larger individuals usually
darker). The tip of the head is white or a pale brownish; just behind there
is a dark ring of color, usually rusty brown or even black, rarely pale ; this
ring is darkest at its anterior edge, and darker on the dorsal than on the ven-
tral side of the body. At least a trace of this ring is to be seen on all speci-
mens when mature, though the intensity of its coloration is very variable.

Dimensions. Length of largest male seen, 350 mm.; greatest diameter, .9 mm.
Length of largest female, 290 mm.; greatest diameter (of a flattened individual),
2mm. The males are more slender and average considerably shorter than
the females,

Especial Diagnostic Characters. The trilobation of the posterior end of the
female, the long and cylindrical tail lobes of the male, the oblique truncation
of the head end, and the usually very dark colored ring around the head,
render this species very easy of identification.

Comparisons. This species has the greatest affinity to Gordius (Para-
gordius ?) tricuspidatus (Dufour); but it differs from it in that there are no
spicules or spines upon the tail lobes of the female, and in that the dorsal is
narrower than the lateral lobes; further, in varius the areoles of the cuticle
are frequently arranged into rows or groups.

Geographical Distribution. I have examined specimens from the following
localities: New York, Maine, Massachusetts, New Jersey, Pennsylvania, Vir-
ginia, Kansas, California, and Guatemala; and it has been observed by others

MONTGOMERY: GORDIACEA. 47

in Mexico, Peru, and Bolivia. It appears to have a very extensive range, and
it and G. aquaticus robustus are the most abundant forms in the northeastern
portion of the United Stabes.

Genus CHORDODES (Creplin) Möbius.

13. C. morgani, n. sp.
Fig. 94, Plate 12; Figs. 95-100, Plate 13.

(1 female, type, in my possession, from Maryland; a second female from
Iowa in the Harvard coll. no. 1470.)

‚Description of the Type Specimen. Form: Perfectly eylindrical without
median lines, Anterior end gradually attenuated, head (Fig. 98) much nar-
rower than the posterior end, rounded. Tail end (Fig. 97) swollen, obtuse
posteriorly, the swelling most pronounced on the ventral aspect. Cuticle:
With three kinds of prominences (Figs. 99, 100): (1) Larger tubercles which
are about twice as high as broad, nearly circular on cross section, and rounded
apically; these bear no hairs, and are distributed at nearly equal distances on
the surface of the cuticle, with only a slight tendency to arrange themselves
into disjointed groups. (2) Smaller tubercles, which are pointed at the apex
and more or less conical in form; these vary considerably in height, but are
never more than a quarter the height of the preceding kind. Each bears on its
summit a single delicate hair. These tubercles are arranged quite densely on
the surface of the cuticle, and the larger ones among them are grouped closely
around the tubercles of the Ist order; in the median line of the body they are
more numerous, especially the larger ones of them, which form groups be-
tween as well as around the tubercles of the Ist order. (3) Delicate slender
hyaline processes, frequently club-shaped, which occur only sparsely, and are a
little higher than the first kind of tubercles. Color; A uniform yellowish
brown. Dimensions: length, 222 mm. ; greatest diameter of body, 1.1 mm.

Description of the Second Specimen. Form: Anterior end pointed, and head
tip (Fig. 94) rounded as in the preceding, but on the ventral surface of the
head, to each side of the median line, is a short longitudinal groove. Body
nearly cylindrical, with narrow but deep median grooves; on a portion of
the surface there are likewise irregular longitudinal grooves. Posterior end
(Figs. 95, 96) swollen, though of less diameter than the body at its middle
point; this distal swelling is of greater diameter than the dorso-ventrally
flattened portion of the body which immediately precedes it. The posterior
end is truncated terminally, and near tie centre of this terminal aspect
(Fig. 96) is situated the cloacal aperture at the middle point of a vertical
ridge, to each side of which is a groove. On the dorso-lateral sides of the
posterior end larger and deeper grooves are situated. Color: A uniform dull
chocolate-brown, the terminal aspect of the head somewhat lighter in color.

48 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Cuticle: as in the preceding specimen, Dimensions: length, 158 mm. ;
greatest diameter of body, 9 mm.

Especial Diagnostic Characters. The peculiarities of the cuticular tubercles,
and the uniform coloration.

Comparisons. The papillæ of the cuticle have some resemblance to those of
C. hamatus Römer, from West Africa. But the arrangement and form of the
papilla of the latter form is not quite the same, judging from Römer’s (96)
description of them: “Die Haut ist mit Papillen bedeckt von der Form
kleiner Hiigel und spitzer Zacken. Sie sind ganz niedrig; ihre Form ist
nicht gleichmässig, auch ihre Entfernung von einander nicht die gleiche, aber
sie haben im allgemeinen denselben Habitus. . . . Kopfende des Weibchens
stark zugespitzt.”

Thus far only two specimens observed, from Iowa and Maryland, respectively.

I have thé pleasure of naming this species in honor of my friend Dr. Thos.
H. Morgan, of Bryn Mawr College, who kindly gave me the first specimen
seen.

14. C. puerilis, n. sp.
Figs. 101-105, b, Plate 13,

(Type, 1 male: Leidy coll. no. 5071, from a cockroach, A second male se-
cured by me in Chester County, Pennsylvania.)

Form, Anterior portion of the body more slender than the posterior.
Head end (Figs. 101, 102) dorso-ventrally flattened, obliquely truncated ter-
minally, mouth opening terminal. Middle and posterior portions of the
body horizontally flattened in the larger specimen, cylindrical in the smaller.
Posterior end of the body (Fig. 103) narrower than the preceding part, almost
cylindrical, terminally rounded; a median groove is present on its ventro-
terminal end, and to each side of this groove the integument forms a slightly
elevated ridge.

Cuticle. With four kinds of prominences (Fig. 105, 6): (1) the largest tuber-
cles, usually of a rounded-conical shape, but vary somewhat in length (the
length is usually one third greater than the largest diameter, which is at the
base). On the rounded apex occur short, rather thick hairs, terminally
pointed, from 5 to 10 hairs to each tubercle. (2) Long hyaline, slender pro-
cesses, which vary considerably in form, but are usually either finger-shaped or
club-shaped ; these are the highest and least abundant of all cuticular promi-
nences, and are devoid of hairs. (3) These most abundant tubercles are usu-
ally conical in shape, and from one fourth to one half the length of the first
kind; each bears on its summit a single strong hair, which is slightly longer
than the hairs of the 1st kind of tubercles. (4) The smallest tubercles are not
quite as high as the preceding kind, are hemispherical, and without hairs. On
surface views of the cuticle (Figs. 104, 105, a) the various kind of tubercles are
seen to be closely arranged together, without any regular distribution into
groups. All these tubercles are very small, and may be distinguished clearly
only on thin sections studied with the y immersion lens of Zeiss,

MONTGOMERY: GORDIACEA. 49

Color. A more or less deep chocolate-brown, somewhat lighter on the
anterior end; head (not merely its terminal tip) almost white.

Dimensions. Length of the larger individual, 212 mm. ; greatest diameter of
body, 1 mm.

Especial Diagnostic Characters. The three kinds of minute papillæ, which
are not arranged into particular groups, furnish the chief diagnostic character,

Comparisons. This species differs in the characters of its cuticle from all
foreign species of the genus. These characters appear to resemble those de-
scribed by Camerano (?97, b) for C. talensis ; but the description of this author
is difficult to understand, and is without figures, so that it seems justifiable to
class our species as new, at least until further descriptions and figures of
talensis are published.

15. C. gordioides, n. sp.
Figs, 106, 107, Plate 13; Figs. 108-110, Plate 14.

(Types, males and 1 female: coll. Acad. Nat. Sci. Phila., Hayden’s Survey,
S. Montana.)

Form of Male. Anterior end slender, cylindrical, attenuated, head (Fig. 108)
somewhat conical with obtusely truncated tip; mouth terminal. Body for
the most part cylindrical, without well marked median grooves; thickest in
the posterior fourth. Posterior end, for the distance of 10 mm., consider-
ably narrower than the immediately preceding portion, and slightly flat-
tened dorso-ventrally. Cloacal aperture on the ventral surface (Fig. 110),
anterior to the posterior tip of the body; just anterior to this aperture is a
slightly elevated semilunar ridge. In the medio-ventral line of the body,
behind the cloacal aperture, is a shallow groove, to each side of which is a
longitudinal prominence. The posterior end of the body is rounded termi-
nally.

Form of Female. General form as in the male, the body thickest in the mid-
dle, narrowed anteriorly, more or less flattened posteriorly. Posterior end
(Fig. 109, a, b) enlarged, somewhat spherical, constricted off from the imme-
diately preceding portion of the body; cloacal aperture terminal, nearer the
dorsal than the ventral side.

Cuticle. With low flattened tubercles, but little higher than the areole of
Gordius densareolatus. On surface views (Fig. 107) these appear small, ovoid
or rounded-polygonal in outline, but they vary considerably in shape and some-
what in size, and are often much elongated. They have a tendency to arrange
themselves into parallel rows, contiguous rows being well separated from one
another; the direction of these rows is slightly oblique to the transverse axis of
the body, and their component tubercles are more or less confluent. Light
colored lines also are seen on the surface of the cuticle, the distances which sepa-
rate these lines being variable; these lines, the optical representations of shallow
grooves, demarcate rhomb-shaped portions of the cuticle, On cross sections
(Fig. 106) the tubercles are seen to be low, and usually flattened apically, with

50 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

rounded margins ; occurring at intervals, but sparsely, are club-shaped slender
hyaline processes such as are characteristic of the genus, and these are about
double the height of the tubercles ; and also small inter-tnbercular groups of
short, spiniform hairs.

Color, In the males the head is yellowish white, with a narrow black ring (ap-
pearing like a spot) immediately around the mouth; the rest of the body a deep
chocolate-color, or rufous-brown, with evidences of lighter brown bands and
spots on the posterior end, though these spots were absent in one individual;
in the smallest male the body color was a clear yellowish brown. In the
single female specimen a black ring immediately surrounded the mouth open-
ing; but the rest of the body is a clear yellow, the head and posterior tip of
the body somewhat lighter.

Dimensions. Length of largest male, 215 mm.; greatest diameter, 1.3 mm.
Length of female, 150 mm.; greatest diameter, .8 mm.

Diagnostic Characters. The presence of flattened tubercles, resembling the
cuticular areoles of the genus Gordius, none of which have a light spot on the
surface, and which tend to arrange themselves into irregular oblique rows, and
the presence of inter-tubercular groups of small hairs, serve to distinguish this
species.

Comparisons. This species resembles most closely C. occidentalis, n. sp.
(9. v.). It also bears some resemblance to C. moluccanus Römer (96), but
differs from the latter in the form and more especially the arrangement of
the tubercles of the cuticle, and in the swelling of the posterior end of the
female.

The cuticle of this species resembles that of a Gordius rather than that of a
Chordodes, except that it shows the typical hyaline processes of the latter
genus ; and since its cuticle thus unites characters of these two genera, the
specific name gordioides is suggested.

16. C. occidentalis, n. sp.
Figs. 111-114, Plate 14; Figs. 115-117, Plate 15.

(Type 1 male: Harvard coll. no. 1469, San Francisco, Cal. A second male :
Harvard coll. no. 1481, Rio Gila, Arizona.)

Form. Whole body much flattened dorso-ventrally, with the exception of
the posterior end, in the type without, in the second specimen with, shallow
median grooves ; anterior and posterior ends narrower than the middle portion
of the body, though the anterior end is attenuated only for a distance of about
lem. Head (Figs. 111, a, b) flattened, conical, the tip rounded or else obtusely
truncated, and then the dorsal margin projects farther forward than does the
ventral. Mouth terminal. In the second specimen (Figs. 114, a, b) a Y-shaped
ridge is situated on the terminal aspect of the head, the mouth placed at the
point of union of the three arms of the Y; the unpaired arm of this Y-shaped
ridge runs from the mouth mediad and dorsal, the paired arms latero-ventrad.

MONTGOMERY: GORDIACEA. 51

Posterior end of the body is nearly cylindrical, somewhat flattened ventrally,
and terminally either rounded or obliquely truncated. On the medio-ventral
surface of the posterior end (Figs. 112, 113, a, 6) is a shallow groove, which
extends from the cloacal aperture caudad to.the distal end of the body, and to
each side of this groove is a longitudinal ridge of slight elevation.

Cuticle. On surface view (in Canada balsam) two kinds of low, flattened
tubercles or areoles are to be seen (Figs. 115, 117): (1) The larger (those of
greater diameter) are darker in color, and either elongate (in one specimen) or
rounded-polygonal in outline. In the median line of the body they are smaller
and more densely arranged than elsewhere. In one specimen (the type, Fig.
115) these areoles were non-confluent ; but in the other they show a tendency
to group themselves in interrupted, transverse rows, and consequently are
more elongate in form than in the former (type) specimen. On the surface of
some.of the larger areoles is seen a small, circular clear spot, in the centre of
which appears a small granule ; sections show that this spot is a pit on the
surface of the tubercle, which is nearly filled with a small rounded-conical
process ; those tubercles on the lateral surfaces of the body which contain such
clear spots, and they are few in number, are usually dumbbell-shaped in outline,
and their clear spots are smaller than those of the median tubercles, in which
they occur more frequently. (2) Smaller, lighter colored tubercles, much
more variable in form and size than the preceding, and which are irregularly
arranged between the former kind. In one of the specimens a system of oblique
lines is seen on the surface of the cuticle, and these lines are peculiar in that
they do not lie between rows of areoles, but appear to run. right across their
surface (Fig. 117).

On transverse section of the cuticle two kinds of tubercles are seen, corre-
sponding to the two kinds seen on surface views (Fig. 116): (1) Low tuber-
cles of greater diameter, which are flattened apically, and have no projections.
(2) Tubereles of smaller diameter, very irregular in form, and usually of
slightly less elevation than the preceding; these correspond to the smaller,
lighter tubercles seen on surface views. The apex of these is not flattened, but
more or less irregularly rounded ; from the summit project upwards short
conical or spiniform processes, which are exceedingly variable in form, some-
times cleft or pectinate terminally, though most of them are largest at the base
and pointed at the apex. Rarely is there only a single process to a tubercle :
as a rule there are a number, and on the tubercles of the dorso-median line of
the body they are more numerous than elsewhere. In addition to these two
kinds of tubercles are seen on sections, though only sparsely, hyaline club-
shaped processes.

Color of type specimen : black, at the anterior end of the body with a
reddish tinge ; the tip of the head yellowish white. The second specimen was
of a deep rufous-brown color, lighter at the anterior tip of the body, and
blackish at the posterior end,

Dimensions. Length of larger individual (type), 255 mm.; greatest diameter,
1,5 mm.

52 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Especial Diagnostic Characters. The presence of two kinds of slightly elevated
tubercles on the cuticle: tubercles of greater diameter with smooth summit, a
few of which have an apical clear spot ; and tubercles of smaller diameter and
usually less elevation, the surface of which is not smooth, but with more or less
numerous conical or spiniform short processes,

Comparisons. This species is more closely allied to C. gordioides than to
any other, but differs from it in the following points : the presence of tubercles
whose surface bears short processes, the presence of clear spots on the surface of
some of the smooth tubercles, and the absence of groups of inter-tubercular
hairs ; there are also differences in coloration, such as the absence of a black
ring around the mouth aperture, and the much darker color of the body.

17. C. cubanensis, n. sp.
Figs. 118-123, Plato 15.

(Type of Female : Harvard coll. 1466, Cuba. Type of Male: Harvard coll.
1466 d, Cuba.)

Form of Male. Body nearly cylindrical, without well marked median
grooves; anterior end slightly attenuated. Head (Fig. 123) very narrow,
truncated apically, slightly concave on the terminal aspect; on the ventral
surface alone is there a constriction separating the head from the body. Pos-
terior end (Fig. 122) flattened on the ventral surface, with a median groove
behind the cloacal aperture ; this groove is broadest and deepest at the ventro-
terminal point of the body; just behind the cloacal aperture is a transverse
semilunar ridge.

Form of Female. Larger and more robust than the male, the shape of the
anterior portion of the body otherwise similar. Head (Fig. 120) rounded at
the apex, mouth nearer the ventral than the dorsal margin of the head. Pos-
terior portion of the body (Fig. 121) narrower than the middle, but the
extreme distal end is swollen, and of greater diameter than the part immedi-
ately preceding. This posterior end is obliquely truncated, the dorso-terminal
margin projecting farther caudad than the ventro-terminal; the swelling is.
most pronounced on the ventral side.

Cuticle (Fig. 119). With papille of three kinds, besides hyaline processes ;
though it is difficult to distinguish sharply between these kinds since they
seem to intergrade. The hyaline processes have the same form as in most
other species of the genus, and are slender club-shaped processes, slightly
swollen and rounded at the apex ; they are about the height of: the highest
papille. lst Kind of Papille: The smallest, least elevated, and most numer-
ous of all occur close together between the groups of larger papillae, and also
in the latter groups. They are pyramidal or somewhat elongate-conical on
lateral view, pointed at the apex which bears a single (rarely two) long, deli-
cate, usually curved spine; this spine is thickest at the base and sometimes
recurved at the tip. A. modification of these tubercles attains nearly the height
of the largest papilla. 2d Kind of Papillæ: These, the second in point of

MONTGOMERY : GORDIACEA. 53

numerical abundance, occur usually in groups close together. They are much
larger than the preceding, round on cross section, usually considerably longer
than broad, slightly thickened at the base but with nearly parallel sides, and
with rounded summit. In a few cases the sides of these papilla: may be some-
what denticulate, On the margin of the rounded or flattened apex occur from
three to six short thick spines, which are broadest at the base and pointed
at the tip ; these spines curve upwards and outwards, and are shorter and
thicker than those of the preceding category of papilla. These papille vary
considerably in form, and sometimes are nearly square in lateral outline.
3d Kind of Papillæ: These are of the same form and size as the preceding,
but are less numerous (the groups which they compose are smaller), and differ
from them in the absence of spines on their summits; the apex of most of them
is elliptically rounded, rarely flattened, and then somewhat denticulate with
short conical processes; some of these papilla are expanded at the summit.

On surface views (Fig. 118) of the cuticle, seen with a low power of magnifi-
ation, only the second and third kinds of papillæ are seen, and they appear as
small brown disks with a clearer central point. The cuticle of the male differs
from that of the female only in that the larger kinds of papillæ are less
abundant, and the groups formed by them smaller.

Color, The males are uniform pitch-black, the head somewhat lighter ; in
the female the whole body is deep black.

Dimensions, Length of largest male, 165 mm.; greatest diameter, 1 mm.
Length of female, 280 mm.; greatest diameter of body, 1.6 mm.; greatest
diameter of tail swelling, 1 mm.

Especial Diagnostic Characters. The dense arrangement of the usually
pencil-shaped papille, and the union of the larger of them into large and
irregular groups, together with the black color of the body, serve to distinguish
this form.

Comparisons. In color this species resembles C. brasiliensis Janda, and
C. feste Camer. ; but it differs from the former as well as from C. morgani, n. Sp»,
in the form and arrangement of the papille. In the structure of the cuticular
papillæ it also differs from feste, judging from Camerano’s description: but
unfortunately most of the South American Gordiacea described by this author
have not been figured, and for a clear understanding of the form and arrange-
ment of areoles and papille figures are absolutely necessary.

Postscript, — The preceding descriptions of body form, color, and dimen-
sions have been based entirely upon a study of alcoholic specimens, with the
consequence that the particulars in regard to coloration and dimensions could
be only approximately ascertained, since the action of the aleohol would prob-
ably produce shrinkage, and certainly obscures the brightness of the coloration.
Formaline would probably be a superior fluid for the preservation of museum
specimens. For purposes of histological fixation rapidly penetrating fluids are
necessary, such as picro-formaline or picro-nitric acid,

VOL. XXXII. — NO. 3. 3

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

APPENDIX.

Since the preceding was sent to press (on the 16th August, 1897) I have re-
ceived from Professor Camerano the following important contribution by him :
“Monografia dei Gordii” (Accademia Reale delle Scienze di Torino, 1897).
This paper furnishes descriptions and revisions of all known species of Gordi-
acea, with illustrations of all those described by Camerano, and is especially
«valuable on this account from the systematic standpoint. To the already
known genera Gordius and Chordodes, this author adds two others, namely,
Paragordius and Parachordodes. Paragordius embraces, according to Camerano,
G. tricuspidatus (Dufour), G. emeryi Cam., G. stylosus Linstow, and G. varius
Leidy. He characterizes it as follows: “ L’ estremità posteriore del 4 è bifor-
cata al di là dell’ apertura postcloacale con lobi profondamente separati fra
loro: non vi è lamina cutanea posteloacale. L’ estremità posteriore della @ è
divisa in tre lobi posteloacali profondamente separati fra loro i quali circondano
V apertura cloacale. Lo strato cuticolare esterno presenta delle formazioni
areolari di una sola seria e pochissimo sporgenti, irregolarmente disposte: non
vi sono granuli o tubercoli rifrangenti interareolari.’ By an unusual coinci-
dence, I had in the preceding pages proposed the same name, Paragordius, to
include Leidy’s species Gordius varius, so that independently of one another

Jamerano and I have founded a new genus, and given it the same name for
the same species, By the rules of priority in nomenclature, however, Camerano’s
publication having appeared first, the genus must stand Paragordius Camerano,
not Montgomery. But the diagnostic given by Camerano for this new species
is not very well chosen, for in it the only character of possible genetic value
mentioned is the trilobation of the posterior end in the female; this character
is of doubtful generic value, since in Gordius tolosanus Duj. the posterior end
of the female is somewhat lobed, though in this case the lobation is gener-
ally regarded as only a specific character! I think the characters described
above by me as diagnostic of the new genus, —namely, the structure of the
cloaca in the female and the absence of a cloacal musculature in the male, — are
of higher value, and accordingly should constitute the diagnostic of the genus.
The second new genus of Camerano is characterized thus (Parachordodes) :
“Tstremitá posteriore del & biforcata al di là apertura cloacale, con lobi pro-
fondamente separati fra loro: nessuna lamina-cutanea posteloacale. D’ es-
tremità posteriore della 9 è intiera coll’ apertura cloacale mediana collocata in
un,solco dorso ventrale più o meno profondo. Strato cuticolare esterno meno

MONTGOMERY: GORDIACEA. 55

complicato che nel genere Chordodes: ora con una sola sorta di formazioni
areolari, ora con formazioni areolari di due sorta: le une pit basse e chiare: e
le altre un po’ più elevate e scure che stanno intorno allo sbocco dei canaletti
che attraversano gli strati cuticolari: fra le areole spesso vi sono granuli o
tubercoli rifrangenti.” Parachordodes thus differs from Gordius merely in re-
gard to cuticular structures, and 1 cannot consider that such differences warrant
the separation of a new genus. For in the Gordiacea the cuticular differences
have little more than specific value, as is well shown by the fact that the
cuticle may show marked differences in different individuals of the same
species. If more important differences than these be subsequently determined,
then, and not until then, does it seem justifiable to me to recognize Parachor-
dodes as a well defined genus; and hence it should for the present be retracted
into Gordius. (A preliminary to this monograph was published by Camerano
in the Zool. Anzeiger for August, 1897, with the title “Nuova classificazione
dei Gordii.”) In this paper are given figures of the cuticle of Chordodes
talensis Cam., which represent its structure as quite different from that of my
new species, C. puerilis.

Here may also be mentioned certain papers on American Gordiacea which
had been omitted in the preceding pages.

Girard (1851, “Historical Sketch of Gordiacea,” Proc. Acad. Nat. Sci.
Philadelphia, 5) mentions a specimen caught at Richmond, Virginia, and
several collected in Oregon by the U. S. Exploring Expedition.

Sanford (1853, “On some Points in the History of Gordius,” Proc, Amer.
Assoc. Sci.) collected some specimens from crickets (Gryllidæ).

Thompson (1853, “ History of Vermont,” Burlington) states that Gordii are
very common in still waters and mud in that State.

White (1859, “ Gordius trifurcatus, n. sp.,” Proc. Boston Soc. Nat. Hist., 7)
gives the following description of this species, which may possibly be referable
to the Q of Paragordius varius (Leidy): “ Male. Length, 5 inches; diameter,
4 line; shape uniformly cylindrical; head obtusely conical; posterior end
divided into two long and narrow lobes, and one shorter and broader lobe, in-
curved and fringed with short thick hairs. At base of larger lobe is the genital
opening, from which the spermatozoa are seen escaping with extremely long
tails. Color uniformly light brown. . . . It does not answer to either of the
two described by Dr. Leidy.”

Clementi (1869, “ Hair Snakes,” Canadian Entomologist) found Gordii (?)
in a large spider.

Leidy (1870, “ The Gordius, or Hair-worm,” Amer, Entomol. and Botanist, 2)
gives a good description and some figures (the only figures published by him
of members of this group) of his G. varius. He considers the males of this
form, as previously described by him, to be really of two different species: in
the one (varius) the forks of the tail are thick, with a crescentic fold above the
genital pore. In the other males (@. longilobatus, n. sp.) “the forks of the tail
are two or three times the length of the thickness of the body, and the forks do
not include at their base a crescentic fold as in the former.” (The male here

56 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

termed “varius” is really G. aquaticus robustus, while “longilobatus” is the true
Paragordius varius.)

Leidy (1871, “ Notice of some Worms,” Proc, Acad. Nat. Sci. Philadelphia)
describes as “Gordius lacustris” several specimens from Kansas; but “lacus-
tris” is certainly a misprint for “robustus,” since the papers to which he refers
for the preceding name make no mention of it.

Garman (1886, “ Amblystoma and Gordius,” Science Observer, Boston Sci.
Soc., cited by Camerano) mentions a Gordius found in Amblystoma.

MONTGOMERY : GORDIACEA. 57

LITERATURE LIST.

Baird, W.
"53, Descriptions of some new Species of Entozoa, from the Collection of

the British Museum. Proc. Zoól. Soc. London, 21.

Camerano, L.
790. Intorno ad una specie di Gordius (G. æneus Villot) raccolta dal Sig.
G. B. Anselmo in Venezuela e intorno alle specie di quaesto genere fino
ad ora descritte dell? America meridionale. Ann. Mus. Civ. Genova,
(2) 10.
Camerano, L.
92. Descrizione di una nuova specie del genere Gordius di Palmeira (Parana)
raccolta dal Dott, Franco Grillo. Ibid.
Camerano, L.
93. Sur quelques Gordiens nouveaux ou peu connus. Bull. Soc. Zool.
France, 18.
Camerano, L.
"94. Viaggio del dottor Alfredo Borelli nella Republica Argentina e nel
Paraguay. Boll. Mus. zool. Anat, comp. Torino,
Camerano, L.
’95. Description d'une nouvelle espöce de Gordius du Chili. Actes de la
Soc. Sci. du Chili, 5.
Camerano, L.
796. Descrizione di una nuova specie di Gordio del Basso Beni (Bolivia)
raccolta dal Prof. L. Balzan, Ann. Mus. Civ. Genova, 16 (36).
Camerano, L.
97". Viaggio del Dr. Enrico Festa nella Repubblica dell” Ecuador e regioni
vicine. Boll. Mus. Zool. ed. Anat, comp. Univ. Torino, 12.
Camerano, L.
197%. Viaggio del Dott. Alfredo Borelli nel Chaco boliviano e nella Repub-
blica Argentina, Gordü. Ibid.
Creplin, C. H.
47. In Froriep's Notizen' a. d. Gebiete d. Natur- und Heilkunde, Weimar,

58 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Diesing, K. M.
51. Systema helminthum. Vol. 2. Vindob.

Diesing, K. M.
"61. Revision der Nomatoden. Sitzb. k. k. Akad. Wiss. Wien, 42.

Gay, C.
48. Historia fisica y politica de Chile, ete. Zoologia, Tom. 3. Paris.
Goeldi, E. A.

794. Breva noticia acerca de alguns Vermes interessantes do Brazil. Bol
Mus. Paraense, 1.
Janda, J.
193. Beiträge zur Systematik der Gordiiden. Spengel’s Zool. Jahrb., Abth.
f. System., 7.
Jeffrey-Bell.
185. Note on a Nematode Worm obtained by M. H. H. Johnston on Kili-
manjaro. Proc. Zoól. Soc. London.
Leidy, J.
50. Note on the Development of Gordius aquaticus. Proc. Acad. Nat. Sci.
Philadelphia, 5.

Leidy, J.
"51. Gordiacee. Ibid.
Leidy, J.
53, A Flora and Fauna within Living Animals. Smithsonian Contribu-
tions, 5.
Leidy, J.

"56. A Synopsis of Entozoa and some of their Ecto-congeners observed by
the Author. Proc. Acad. Nat, Sci. Philadelphia, 8.
Leidy, J.
57. (No title.) Ibid.
Leidy, J.
"58. Contributions to Helminthology. Ibid.
Leidy, J.
79. On Gordius and some Parasites of the Rat. Ibid., and Ann. Mag. Nat.
Hist. 3.
Mobius, K.
55. Chordodes pilosus, ein Wurm aus der Familie der Gordiaceen. Zeit.
wiss. Zool., 6.
Montgomery, T. H. Jr.
’98. Descriptions of two new Exotic Species of Chordodes. Spengel’s Zool.
Jahrb.
Oerley, L.
’81. On Hair-worms in the Collection of the British Museum. Ann. Mag
Nat. Hist., (5) 8.

MONTGOMERY : GORDIACEA. 59

Rómer, F.
195. Die Gordiiden des Naturhistorischen Museums in Hamburg. Spengel’s

Zool. Jahrb., Abth. f. System., 8.

Rómer, F.
96. Beitrag zur Systematik der Gordiiden. Abh. Senckenberg. naturforsch.

Ges., 23.

Schneider, A.
"66, Monographie der Nematoden. Berlin.

Villot, A.
74, Monographie des Dragonneaux (Genre Gordius, Dujardin). Arch.

Zool. gén. et expér., 3.
Villot, A.
87. Révision des Gordiens. Ann. Sci. Nat, (7), L.

Weyenbergh, D. H.
79. Descripciones de nuevos Gusanos. Bol. Acad. Nac. Cienc, Argentina, 3.

MONTGOMERY. — Gordiacea,

EXPLANATION OF THE PLATES.

AU the figures have been drawn with the aid of the camera lucida, unless other-
wise specified. The microscope and lenses of Zeiss were used; almost all the
contour figures of the body have been drawn with oc. 2, 0bj. A; and the majority
of the surface figures of the cuticle with oc. 4, obj. C.

PLATE 1.
Gordius aquaticus robustus.

Fig. 1, Male, ventral view of posterior end (A, 4).
Fig. 2. Idem, head end (A, 2).

Fig. 3. Male, oblique ventral view of posterior end (Harvard coll. 290, A, 2).

Fig. 4. Lateral view of the preceding, from the left side.

Fig. 5. Male, ventral view of posterior end (A, 2).

Fig, 6. Male, ventral view of head end; the transverse curved line shows the
posterior extension of the white area (A, 2).

MONTGOMERY. — Gordiacea.

PLATE 2.
G. aquaticus robustus.

Female, ventral view of posterior end ; a portion of the brown ring around
the anus is to be seen (A, 2).

Female, posterior end, showing the ring of brown color around the cloacal
aperture (A, 2).

Male, antero-ventral view of the head end, showing the grooves on the
surface (A, 2).

Idem, lateral view.

Female, head end (A, 2).

Surface view of the cuticle, Canada balsam, the finest lines drawn only in

one portion (C, 4).

MONTGOMERY. — Gordiacea,

PLATE 3.

G. aquaticus robustus. Surface view of the cuticle of a male, the intra-
rhombic intersecting lines not reproduced (C, 4).

G. aquaticus difficilis, n. subsp. Male, head end, from the type (A, 2).

Idem, oblique lateral view of the posterior end (A, 2).

Figs. 16-19, @. aquaticus robustus.

Male, surface view of the cuticle, Canada balsam (C, 4).

Female, head end (A, 2).

Male, cross section through the body in the plane of the cloacal aperture.
Cut., cuticula; Æpi., epidermis ; Musc., longitudinal musculature of the
body wall; Cl. Musc., cloacal musculature ; Ch. t., connective tissue ;
N., branches of the ventral nerve chord; Cl. Ap., cloacal aperture
(C, 2).

Male, ventral view of posterior end (A, 2).

Fig

Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.

as
Fig.
Fig.

Fig.

MONTGOMERY. — Gordiacea, :

208,
29%,

30.
31.

PLATE 4.

Figs. 20-31, Gordius lineatus Leidy.
Female, antero-ventral view of head end (A, 2).
Male (type), head end (A, 2).
Female (type), lateral view of posterior end (A, 2).
Female, posterior end, the cloacal epithelium apparently extruded (A, 2).
Male (type), lateral view of the posterior end (A, 4).
Idem, ventral view (A, 2).
Male, oblique lateral view of posterior end (A, 2).
Male, ventral view of posterior end (A, 2).
Female (type), termino-ventral view of posterior end, showing the colored
ring around the cloacal aperture (A, 2).
Male (type), transverse section of the cuticle (immers. 43, 2).
Idem, surface view (C, 4).
Male, surface view of cuticle (C, 4).
Female, transverse section of cuticle (immers. 45, 2).

Figs. 32, 33, @. densareolatus, n. Sp.

Female, surface view of cuticle (C, 4).
i ransverse section of the cuticle of the same specimen (immers. 17, 2).

Fig.
Fig.
Fig.
Fig.

Fig. $
Fig. ¢

MONTGOMERY. — Gordiacea,

PLATE 5,
G. densareolatus, n. sp.

Female, head end (A, 2).

Female (type), head end (A, 2).

Male (type), oblique lateral view of posterior end (A, 2).
Idem, ventral view.

Female (type), posterior end (A, 2).

Female, posterior end rendered transparent by cedar oil, the intestine and
a portion of the atrium and oviducts are seen. The transverse line
represents the edge of the extruded cloacal epithelium (A, 2).

~~

MONTGOMERY, — Gordiacea,

PLATE 6.
G. platycephalus, n. sp.

Male (type), head end (A, 2).

Idem, ventral view of posterior end (A, 2).

Female: «a, oblique ventral view of posterior end; b, dorsal view of head
end; c, lateral view of head end (A, 2).

Female (type), ventral view of posterior end (A, 2).

Idem, lateral view.

Female (type), head end seen in its broadest plane of expansion (A, 2).

LO

MOoNTGomERY, — Gordiacea.

PLATE 7.

Figs. 46-49, G. platycephalus, n. sp.
Female, surface view of cuticle (C, 4).
Transverse section of the cuticle of the same individual (immers. ;
Female (type), surface view of cuticle (C, 4).
Female, transverse section of cuticle (immers. yz, 2).
Male (type), transverse section of cuticle (immers. gy, 2).

Figs. 50-52, (7. platyurus Baird. Female.

Surface view of cuticle, the finest intersecting lines reproduced at only

one place (C, 4).
Posterior end, free hand drawing. X about 4.
Head end (A, 2).

Figs. 58-55, G. leidyi, n. sp. Female.
Dorsal view of the head (A, 2).
Transverse section of the cuticle (immers. +z, 2)
Surface view of cuticle (C, 4).

1

2).

Y

46?

460

|

50.

fe)

o

Fig.
Fig.
Fig.

Fig.

Fig.

a.
Fig.

Fig.

MONTGOMERY. — Gordiacea.

60,
61.
62.

56.
57.
58,

59.

PLATE 8.

Figs. 56-59, G. leidyi, n. sp. Female (type).
Dorsal view of posterior end (A, 2).
Idem, termino-ventral view.
Idem, oblique lateral view.
Surface view of cuticle (C. 4).
Figs. 60-62, G. violaccus Baird (?) Female (Harvard coll. no. 1465)
Posterior end (A, 2).
Head end (A, 2).
Surface view of cuticle (C, 4).

Fig.

Fig

MONTGOMERY

65.

e Gordiacen.

PLATE 9,

Figs. 63-66, Œ. agassizi, n. sp. Male (type).
Dorsal view of anterior end (A, 2).
Surface view of cuticle (seen in alcohol, A, 4).
Posterior end of the body, the tail lobes seen from the right side, the pre-
ceding portion of the body from the latero-ventral aspect (A, 2).
Posterior end, tail lobes seen from the termino-dorsal aspect, the preced
ing portion of the body from the ventral (A, 2).
Figs, 67-69, G. capitosuleatus, n. sp. Male (type).
Terminal aspect of head end, its ventral edge towards the bottom of the
page (A, 2).
Surface view of cuticle (C, 4)
1 (A, 2)

Dorsal view of head ent

66

69

Fig. 7

Fig. 7
Fig.

Fig.

MONTGOMERY, — Gordiacea,

73.

PLATE 10.

G. capitosulcatus, n. sp. Male (type), oblique ventral view of the posterior
end (A, 2).
Figs. 71-74, G. paranensis Camer.
Female, termino-ventral view of the posterior end (A, 2).
Male, posterior end: «, oblique ventral view; 0, oblique lateral view
(A, 2).
Male, head end (A, 2).
Male, head end (A, 4).

MONTGOMERY. — (rordiacea.

PLATE 11.

Figs. 75-77, G. violaceus Baird (?). Female, Harvard coll. no. 1466 c.

g. 82.

Latero-ventral view of posterior end (A, 2).

Head end (A, 2).

Transverse section of cuticle (immers. 4, 2).

Figs. 78-85, Paragordius varius (Leidy).

Male, transverse section through the plane of the cloacal aperture. Cut.,
cuticle; Æpi., epidermis; Musc., longitudinal musculature of the body
wall; Par., parenchym; N., branches of the ventral nerve chord; Ci.
Ap., cloacal aperture (C, 2).

Female, transverse section through the body in the plane of the caudal
ganglion. Cl. Epi., cloacal epithelium ; N. GL, ganglion of the ventral
nerve chord ; the other lettering as in the preceding figure (C, 2).

Female, cross section through the tail lobes near their proximal ends; the
median lobe is on the dorsal side (A, 2).

Female, cross section of tail lobes near their distal ends (A, 2).

Male, oblique dorsal view of head (A, 2).

Female, head end: a, ventral view; b, dorsal view, showing two colored
stripes. The colored ring around the neck is shown in these figures
(A, 2).

Female, ventral view of head end; the transverse line shows the posterior
limit of the white area (A, 2).

Female, lateral view of head end; the transverse line shows the posterior

limit of the white area (A, 2).

Cut.

#5

Epi.

Muse.

76.

En

|
|

ne

330

Ya

Fig.

Fig.
Fig.
Fig.
Fig.
Fig. |

Fig. |
Fig. {
Fig.

MONTGOMERY. — Gordiacea.

PLATE 12.

Figs. 86-93, Paragordius varius (Leidy).

Optical median section of the posterior end of the female, from a recon-
struction of sections (A, 2), to show the characteristic length of the
cloaca. The omitted portion x-y has about ten times the length of
the portion x-z.. T. L, tail lobes; Cut., cuticle; CL, cloaca; Cl. Ap.,
cloacal aperture; Atr., atrium (uterus); Ov. D., left oviduct; Int., in-
testine; Rec. S., receptaculum seminis.

Male, oblique ventral view of posterior end (A, 2).

Female, posterior end from the dorsal side (A, 2).

Female, oblique dorsal view of the posterior end (A, 2).

Female, dorsal view of the posterior end (A, 2).

Cuticle of a female : a, transverse section (immers. 5, 2); b, surface view
(C, 4).

Female, surface view of cuticle (C, 4).

Male, surface view of cuticle (C, 4).

Chordodes morgani, n. sp., female (Harvard coll. 1470). Head end from the

left side (A, 2).

o a eee a
Y ee a ;

> x N
à $ >
ios
è <
>
à
A a €
A ma ee E à = = xN $
= S
EE ee E E ela Sa ¿E <

6.

MONTGOMERY.

Gordiacea,

PLATE 13.

Pigs. 95-100, Chordodes morgani, n. sp., female; 95, 96, from Harvard coli. 1470;
97-100, from the type specimen.
Fig. 95. Posterior end from the right side (A, 2),
Fig. 96. Termino-lateral view of the posterior end (A, 2).
Fig. 97. Lateral view of the posterior end, right side (A, 2).
Fig. 98, Head end (A, 2).
Fig. 99. Transverse section of the cuticle (immers. q}, 2).
Fig. 100. Surface view of the cuticle, from the side of the body (C, 4)

Figs. 101-105, b, €. puerilis, n. sp., male. Figs. 101-104, from the type specimen ;

105, a and b, from a specimen in my possession.

Fig. 101. Head,end, dorsal view (A, 2).

Fig. 102. Idem, lateral view.

Fig. 103. Oblique ventral view of the posterior end (A, 2).

Fig. 104. Surface view of the cuticle (C, 4).

Fig. 105. Cuticle: a, surface view (C, 4); b, transverse section (immers. 4, 2).
Figs. 106, 107, C. gordioides, n. sp., male (type).

Fig. 106. ‘Transverse section of the cuticle (immers. q, 2).

Fig. 107. Surface view of cuticle, the areoles reproduced on only a portion of the

surface (C, 4).

m
9

MONTGOMERY. — Gordiacea,

PLATE 14.

Figs. 108-110, C. gordioides, n. sp. (types).
Male, head end (A, 2).
Female, posterior end: a, ventral view; b, lateral view (A, 2).
Male, ventral view of posterior end (A, 2).

Figs. 111-114, C. occidentalis, n. sp., males.
Figs. 111, 112, from the type specimen.
Head end: a, lateral view; b, dorsal view (A, 2).
Lateral view of posterior end (A, 2).
Posterior end: a, oblique dorsal view ; b, from the left side (A, 2).
Head end: a, termino-ventral view; b, from the left side (A, 2).

JON
ie
Ha

1090

Fig.

Fig.

MONTGOMERY, — Gordiacea,

PLATE 15.

Figs. 115-117, C. occidentalis, n. sp., males, the first two from the type

115.

116.

ML

specimen.
Surface view of the cuticle, a line passing through x, x shows the median
plane of the body (C, 4).
Transverse section through the cuticle in the dorso-median plane of the
body (immers. 4, 2).
Surface view of the cuticle; the arrow denotes the line of the transverse
axis of the body (C, 4).

Figs. 118-123, C. cubanensis, n. sp. (types).
Female, surface view of cuticle (C, 4).
Female transverse section of cuticle (immers. qz, 4).
Female, head end (A, 2).
Female, posterior end from the left side (A, 2).
Male, oblique ventral view of the posterior end (A, 2).
Male, head end from the right side (A, 2).

|
HE
À
HR

{ |
{
y! V À N N :
4 119

Bulletin of the Museum of Comparative Zoology
AT HARVARD COLLEGE.
Vou. XXXII. No. 4.

SOME PLANARIANS FROM THE GREAT BARRIER REEF
OF AUSTRALIA.

By W. MoM. WOODWORTH.

Wiru ONE PLATE.

CAMBRIDGE, MASS., U. S.A.:
PRINTED FOR THE MUSEUM.

Arrır, 1898.

No. 4. — Some Planarians from the Great Barrier Reef of
Australia. By W. MoM. WoobwoktH,

Pseudoceros devisii, sp. nov.
Fig. 1.

General color, bright orange-yellow, with a broad marginal band of deeper
orange and a prominent median dorsal ridge along which the pigment is denser
than over the general surface of the body, though not so deep in color as the
marginal band. Length, 33 mm.; greatest breadth, 16 mm. From the Brisbane
River near Brisbane, April, 1896.

The single specimen of this species, was given to me by Mr. ©, W. de Vis,
the curator of the Brisbane Museum, in whose honor it is named. The draw-
ing for the accompanying figure (Fig. 1) was prepared by Dr. A. G. Mayer,
but unfortunately the specimen was destroyed before it could be preserved, or
studied in detail. The color and brilliancy of its markings, however, distin-
guish it from any described species.

Idioplana australiensis, gen. nov., sp. nov.
Figs. 2-5.

General color, bluish to yellowish cream when seen on black background,
more reddish on white background. Small spots of dark reddish brown uni-
formly distributed over the dorsal surface except at extreme margin, which
is free from pigment. Ventral surface without pigment, white. Slightly
translucent ; pharynx and male organs indistinctly visible, the vasa deferentia
together with the penis forming a Y-shaped figure (Fig. 2). Expanded ante-
riorly, with a deep median fold or notch. Anterior or expanded portion pro.
vided with marginal eye-spots extending backwards to a distance about one
third the total length of the animal from the anterior end. No marginal eye-
spots in anterior median notch, Two tentacles, situated about one sixth the
total length from the anterior end ; anterior faces of tentacles provided with
eye-spots, and a scattered group of eye-spots over the brain region extending
as far forwards as a line joining the tentacles (Fig. 3). Length 50 mm.;
greatest breadth 22 mm. A very sluggish form. Only one specimen, taken
on the reef at Hope Island, May 12.

The genus Idioplana differs from other Planoceridæ chiefly as regards the
sexual organs. The closely approximated sexual openings lie in the anterior

VOL. XXXII. — NO, 4.

64 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

part of the posterior third of the body, and in the preserved specimen the
male gonopore was at a distance of less than 1 mm. from the female open-
ing. There is a large muscular penis enclosing a spacious prostate gland
(“ Körnerdrüse”), which communicates with the ductus ejaculatorius at the tip
of the organ, the two ducts opening to the exterior almost together (Fig. 5).
The ciliated ductus ejaculatorius extends along the ventral wall of the penis,
at the posterior end of which it divides into two vesiculæ seminales, convo-
luted tubes with muscular walls (not well shown in the diagram, Fig. 4). The
vasa deferentia are two large convoluted canals of nearly uniform calibre which
connect with the vesiculæ seminales at the root of the penis. There are also
two smaller posterior seminal canals which unite with the main anterior canals
immediately before these join the vesicule seminales (Fig. 4). As the speci-
men was in a late stage of sexual activity, no trace of the testes could be found,
and the vasa deferentia were only partly filled with spermatozoa.

The female gonopore leads into a spacious bursa, from which the vagina,
with diminishing calibre, leads upward and backward over the male organ,
and receives the oviducts which open into it opposite to one another at a point
above the posterior limits of the penis. The canal, which is ciliated through-
out, and which from now on is reduced in diameter and uniform in calibre,
passes forwards and downwards over the penis, reaching nearly to the ventral
wall, and then bends upwards and backwards again, passing over the vagina
and terminating a little posterior to the female gonopore in a vesicle or
enlargement, ‘‘accessorische Blase” (Figs. 4 and 5).

The terminology employed in the description of the sexual organ of this
species is not that which is customarily used in descriptive anatomy of Poly-
clads, but corresponds to the terminology applied to Triclads. The so called
“accessory vesicle” I believe to be directly homologous with the uterus of Tri-
clads, and that fertilization takes place in it. Although the single specimen of
the species under consideration was far advanced sexually, there was still a
packet of spermatozoa in the uterine vesicle, and Plehn ! has figured the uterine
vesicle of Latocestus atlanticus containing both ova and spermatozoa.

Diposthus corallicola, gen. nov., sp. nov.
Figs. 6-11.

Color, yellowish rose; very opaque except at margin, which is bluish and
very translucent. Of the internal organs the position of the pharynx only can
be seen as a lighter ragged median streak, Two closely approximated groups
of eye-spots about one tenth the total length of the animal from the anterior
end. Two pointed conical tentacles close to anterior margin. Length, 15-40
mm. ; greatest breadth, 6-13 mm. Abundant under coral rock on reef at Hope
Island, May 12. A very active form.

1 Plehn, M. Neue Polycladen gesammelt von Herr Kapitan Chiercha, etc. Jen.
Zeitschr. f. Naturw., Bd. XXX. p. 160, Taf. XI. Fig. 10, 1896.

WOODWORTH: SOME PLANARIANS FROM AUSTRALIA. 65

The peculiarities of the male sexual organs of this form have necessitated
the establishment of a new family for its reception (Diposthidæ). The salient
feature is the separation of the penis and prostate gland into two distinct
organs, both of which are doubtless intromittent. The prostate gland
(“ Kornerdriise”) occupies the male genital atrium together with the penis,
and lies posterior to it directly over the gonopore. It is slightly larger than
the penis, and both organs are nearly pendent or perpendicular in position
(Figs. 8, 9, and 11). The prostate gland is provided with a heavy layer of
circular muscles, which together with the epithelial covering of the organ
decrease in thickness toward the free end, and at the very tip are entirely

+ . op . . . .
lacking. I could not satisfy myself that there was an opening at this point,

nor could I demonstrate any distinct lumen, which is explicable, possibly, by
the fact that all of the specimens were in a late stage of sexual activity, all
traces of testes, vasa deferentia, ovaries, and oviducts having disappeared, Two
kinds of nuclei occur in the prostate gland, deeply staining nuclei which are
found chiefly under the zone of circular muscles, and more lightly staining
granular ones, which are accumulated at the free end of the organ (Fig. 9).
The latter kind I am inclined to look upon as belonging to the glandular cells
of the prostate. There are two large vesicule seminales the ducts of which
unite to form the ductus ejaculatorius (Fig. 11). The seminal vesicles were
filled with spermatozoa. The female gonopore opens into a large atrium, the
walls of which are thrown into folds, and into which numerous unicellular
glands, the shell glands, open. The walls of the chamber are also highly
muscular, to function doubtless as a bursa copulatrix. Two of the specimens
that were sectioned showed remains of uterine vesicles. One specimen exhib-
ited two pairs of vesicles, and in the other there were three vesicles on one side
of the body and two on the other side, with a third duct which ended abruptly,
indicating the atrophy of the third vesicle of that side, In the specimen with
but two pairs of uterine vesicles the vesicule seminales had also disappeared,
but their ducts could be traced hack for a considerable distance from the penis.
The full number of uterine vesicles can only be determined from material in an
earlier stage of sexual activity.

The uterine vesicles cannot be compared exactly with the uterine glands fig-
ured by Lang? for Oligocladus sanguinolenta, for in that species they are
only indirectly connected with the uterus by means of the oviducts. They
are better comparable with the uterine vesicles of Uteriporus vulgaris of Ber-
gendal,? and I look on the uterine vesicles of Diposthus as different chambers
of a multipartite uterus, having at least six such parts. The vesicles of each
side of the body communicate with a common duct, the ducts of opposite sides
uniting to enter the female genital atrium (Fig, 11). As evidence that the

1 Lang, A. Die Polycladen, Fauna u. Flora des Golfes y. Neapel, Monog. XL,
Taf, XXI. Fig. 8, 1884,

2 Bergendal, D Studier öfver Turbellarier. II, Om Byggnaden af Uteriporus,
ete, Lunds Univ. Ars-Skrift, Fys, Sállsk. Handlingar, Bd. VII. 1896.

66 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

uterine vesicles are but separate chambers of a compound organ, is the simul-
taneous occurrence in these chambers of both ova and spermatozoa. Figure
10 represents a section through one of the uterine vesicles, showing an ovum
surrounded by a dense mass of filaments, which in every way resemble the
spermatozoa found in the seminal vesicles of the same specimen. Lang (op. cit.,
p- 297) speaks of skeins of fine filaments resembling spermatozoa in the acces-
sory vesicles, and of “lumps” which he doubted not were fragments of eggs
that had found their way in there,

|

Big: 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Hig. T
Fig. 8
Fimo
Fig. 10

Fig. 11

WOODWORTH : SOME PLANARIANS FROM AUSTRALIA. 67

EXPLANATION OF THE PLATE,

ABBREVIATIONS.
dt. ejac. Ductus ejaculatorius. ut. Uterine vesicles.
gl. pr. Prostate gland. vag. Vagina,
gl. sh, Shell gland. v.d. Vasa deferentia.
ov. Ovum. ves. Uterine vesicles.
ov'dt. Oviduct v. sem. Vesicule seminales.
pe. Penis. d Male gonopore.
sp'2. Spermatozoa. $ Female gonopore.

. Pseudoceros devisii, sp. nov. Drawn from life by A. G. Mayer. X 2.

. Idioplana australiensis, gen. nov., sp. nov. From life; slightly enlarged.

. Idioplana australiensis. To show the arrangement of the tentacular and
epi-cerebral eye-spots. X 4.

. Idioplana australiensis. Diagram of sexual organs. X 10.

. Idioplana australiensis. A drawing from five longitudinal consecutive
sections to show the course of the sexual ducts. X 20.

. Diposthus corallicola, gen. nov., sp. nov. Drawn from life; natural size.

. Diposthus corallicola. Anterior end of corrosive sublimate preparation to
show the arrangement of tentacular eye-spots. X 4.

. Diposthus corallicola. Longitudinal section through gonopores. X 47.

. Diposthus corallicola. Enlarged drawing of prostate gland, from same
section. X 170.

. Diposthus corallicola. Diagram of sexual organs. X 50.

. Diposthus corallicola, Section through a uterine vesicle. X 160.

Barrier Reef Planarians.

|

Bulletin of the Museum of Comparative Zoölogy
AT HARVARD COLLEGE.
VoL. XXXII. No. 5.

REPORTS ON THE DREDGING OPERATIONS OFF THE WEST COAST OF
CENTRAL AMERICA TO THE GALAPAGOS, TO THE WEST COAST
OF MEXICO, AND IN THE GULF OF CALIFORNIA, IN CHARGE OF
ALEXANDER AGASSIZ, CARRIED ON BY THE U. S. FISH COMMIS-
SION STEAMER “ALBATROSS,” DURING 1891, LIEUT. COMMANDER
Z. L. TANNER, U. 8. N., COMMANDING,

XXIII.

PRELIMINARY REPORT ON THE ECHINI.

By ALEXANDER AGASSIZ.

[Published by Permission of MARSHALL MCDONALD and GEORGE M, BOWERS,

U.S. Fish Commissioners. |

With THIRTEEN PLATES
,

AND A CHART OF THE ROUTE OF THE “ ALBATROSS.”

CAMBRIDGE, MASS, U.S. A.:
PRINTED FOR THE MUSEUM.

June, 1898.

No. 5. — Reports on the Dredging Operations off the West Coast
of Central America to the Galapagos, to the West Coast of
Mexico, and in the Gulf of California, in charge of ALEXANDER
AGASSIZ, carried on by the U. S. Fish Commission Steamer
“ Albatross,” during 1891, LIEUT. COMMANDER Z. L. TANNER,
U. S. N., Commanding.

XXIII.
Preliminary Report on the Echini. By ALEXANDER Acassız.

Tue following brief descriptions, accompanied with figures of the
d ’ 5
more interesting species, collected during the “ Albatross ”

of 1891, are published to prevent possible confusion in the names

Expedition

adopted for the new species of sea-urchins, which will appear on the
Plates preparing for the final Report.

As regards the distribution of Echini in the Pacific, we have at the
present day a condition of things very similar to that which must have
prevailed in the Atlantic when the species of Echini living in the Crag
and in the Maltese beds had their representatives in the West Indies,
having, as has been suggested, found their way from the Mediterranean
along the shores of an ancient continent. Some of the species living
on the west coast of Central America have a very extended geographi-
cal distribution in the Pacific, and yet no one claims that this great
range has been brought about by their migration along the shores
of a continent, or continental islands, existing between Panama and
the Sandwich Islands or the Marquesas.

The great equatorial current gives us a cause fully efficient to effect
such a wide distribution, and that in a comparatively short time. While
undoubtedly many of the species of Echini have no pelagic Plutei, and
are 80 to speak viviparous, or carry their young for a considerable period,
yet we should remember that young Echini, even after they have
assumed the characters of the adult, are capable of being transported

VOL. XXXII. — NO. 5. al

72 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.
long distances by currents. It is not an uncommon thing to find the
young of Arbacia, of Strongvlocentrotus, and of Echinarachnius on our
coasts floating about, and they are not unfrequently caught in the sur-
face townet. The same holds good for many species of Starfishes and of
Ophiurans, as well as of Holothurians. In Florida I have canght in the
same way the young of Cidaris and Hipponoé, and of Toxopneustes, and
of many other species of Starfishes and Ophiurans. These young Echino-
derms all float, and may be carried very long distances during the period
in which they still have the huge embryonic tentacles characteristic of
their younger stages, when the ambulacral feet are entirely out of pro-
portion in size to the rest of the test, and the young thus possess a great
floating capacity when their suckers are expanded,

They retain these suckers for a considerable period of time, during
which they can be transported very great distances. There is no other
explanation for the identity of the littoral marine fauna of the Bermudas
than that the young and embryos of the Echinoderms and Polyps of the
West Indies have been carried northward fully six hundred miles by
the Gulf Stream at a rate of from one to three miles a day, and have
finally settled in the Bermudas.

We can well imagine an equatorial current taking during Miocene and
Eocene periods the young of the Echini flourishing in the Crag and in the
Mediterranean, and in the southern extension of that fauna perhaps only
from the Cape Verd Islands, and bringing them to the shores of North-
ern South America or into the Caribbean Sea. That stretch is but
little longer than the stretch which we know is annually traversed by
Acalephs, Pteropods, Fishes, and Annelids, along the course of the Gulf
Stream from the Straits of Florida to Narragansett Bay, and to the
southern shores of Cape Cod and the adjacent islands.

The existence of a continent or of intervening islands does not seem
to me necessary to explain the similarity of the Echinid fauna of former
timos on both sides of the Atlantic or Pacific. The causes now at work
appear to me sufficient to explain their relationship, when we take into
account what is known of the efficient transporting agency of equatorial
or other oceanic streams for the Pluteus or the young stages of Echini
during a considerable period of their post-embryonie life.

We should also remember that, even with cur imperfect knowledge of
the bathymetrical range of Echini, the range in depths of many genera
is known to be very great, as will be seen from an examination of the
lists given in the “Challenger” Reports and from the depths obtained

by this Expedition. Among these I may mention those having a great

>

AGASSIZ: PRELIMINARY REPORT ON THE ECHINI. te

geographical distribution, as well as a wide bathymetrical range.! This
will serve to show the extent to which many species can slowly migrate
upon the bottom, even at a very considerable distance from land or con-
tinental or insular slopes, when living in the track of a great equatorial
current which supplies them with a constant and abundant supply of
food.

DESMOSTICHA, HAECKEL.
CIDARIDÆ, Mttt.
GONIOCIDARID At, Hazroreı.

Dorocidaris panamensis A. Aa.
Plate I ; Plate II. Fig. 1.

The test of this species is greatly flattened; the primary radioles are short, in
many specimens not longer than the diameter of the test; they are compara-
tively slender but with a coarser granulation than in the Atlantic species (D.
papillata). ‘The abactinal system is also smaller, the anal system more pentag-
onal, and the genital plates more elongate than in the D. papillata, and the
actinal spines are smaller than in that species,

Station No. 3367, off Cocos Island, 100 fathoms.

rs “ 3368, off Cocos Island, 66 fathoms.
Li “ 3378, off Galera Point, 112 fathoms.
ik “ 3397, off Galera Point, 85 fathoms.

Goniocidaris Doederleini A. Aa.
Plate III. Fig. 1.
The nearest ally of this species is G. canaliculata, from which it is readily
distinguished by the greater flatness of the test and the very slender primary

tenera. Range in fathoms. Genera. Range in fathoms.
+. Dorogidane ite- au e O Eppan e ee A al
LOLA ue vw LE A Shen ON
GOD... 0 holy fs) Pourtalesia, 4. u a + 0880
SALIA u a A a a Homolampas. . . +. . 1600
Pododidaria +... 11.7 1076 A eg OU
Colopleurus. . 4. . . 1898 WCHINOCATGIUM + s « . 2075
Aspidodiadema . . . . 1800 HOM... à « 400
Dermatodiadema . . . 800 STISSODEIG u... ey 0 0008
PROMO DO ARO ss « 000)
Temnechinus , , . . 600 a ee. ONU
Diigonocidaris. y 400 dt 5 so 4000 008
Holinusan ern edo PA NV a N a 0

Spherechinus . . . . 400 a er‘

74 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

radioles, When alive the bare spaces of the test are of a deep brownish violet,
the primary spines reddish, and the miliaries greenish with brownish longitu-
dinal bands. The sutures of the genital and ocular plates are marked by a
sharp violet line and the genital system is comparatively bare and free from
miliaries.

Station No. 3369, off Cocos Island, 52 fathoms.

Two new species of Porocidaris were dredged by the “ Albatross.”

Porocidaris Milleri A. Ac.
Plate IV.

This species is closely allied to P. elegans collected by the “ Challenger.” Its
radioles differ from those of the Challenger species in having finer serrations.
The actinal system is marked for the prominent development of the interam-
bulacral plates, while in P., elegans the actinal system is paved with ambulacral
plates (Agassiz, Chall. Echini, Plate III. Fig. 3). The primary tubercles of
this species are not crenulate as they are in the other species of the genus, but
the species possesses the remarkable primary actinal radioles which are so char-
acteristic of the genus. The median interambulacral suture is bare of miliaries,
while in P, elegans the presence of the closely packed miliaries conceals the
suture.

Station No, 9, off Cape Mala, 465 fathoms.

di “ 3360, off Cape Mala, 1672 fathoms.

“ “ 3381, off Malpelo Island, 1772 fathoms. |
pe “3399, off Galera Point, 1740 fathoms.
66 [77

3415, off Acapulco, 1879 fathoms,

Porocidaris Cobosi A. Ac.
Plate III. Figs. 2-5.

This species is rendily recognized by its comparatively small actinal and
abactinal system. The primary radioles recall those of the Atlantic species
P. Sharreri. In the ambulacral zone the primary miliaries are larger than
those surrounding the serobicular area of the primary tubereles in the interam-
bulacral area, The primary mammary bosses are perforated and crenulated ;
the scrobicular area is narrow.

Station No. 3404, off Chatham Island, 385 fathoms.

SALENID/E, Acass.
Salenia miliaris A. Ac.
Plate IL Figs 2-4.
This species can at once be distinguished from its Pacifie congeners by the

great size of the anal system, which is irregularly hexagonal aud covered with

mr

AGASSIZ: PRELIMINARY REPORT ON THE ECHINI 10

a comparatively larger number of plates than in the other pacific species. The
primary radioles are marked for the great development of the milled ring.
The primary ambulacral tubercles are small and the two vertical rows are sepa-
rated by a wide band crowded with minute miliaries. Some of the primary
radioles are curved at the extremity and their great length is very striking. In
a specimen measuring 12 mm, in diameter, the radioles were slightly over
60 mm. in length.
Station No. 3357, off Mariato Point, 782 fathoms.

4 “ 3360, on way to Cocos Island, 1672 fathoms.
4 “ 3361, on way to Cocos Island, 1471 fathoms,
a “ 3362, on way to Cocos Island, 1175 fathoms.

. “ 3376, South of Malpelo Island, 1132 fathoms.
T « 3380, off Malpelo Island, 899 fathoms.

# “ 3407, Galapagos Islands, 885 fathoms.
de “3411, Galapagos Islands, 1189 fathoms.
a “3413, Galapagos Islands, 1260 fathoms.

ARBACIADA, PETERS.

At Station 3382 in 1793 fathoms we dredged a single specimen of a species
constituting a new genus (Dialithocidaris), and one which we may consider as
the Pacific representative of Podocidaris of the West Indies. Tam inclined to
consider as also belonging to this genus Podocidaris prionigera A. Ag, which
when described was referred with considerable doubt to the genus Podocidaris.

DIALITHOCIDARIS, A. Ac.

The genus is marked by the great size of the genital and ocular plates of the
apieal system; by the width of the interambulacral area, by the peculiar
linear arrangement of the large interambulacral miliaries along the median line
parallel with the horizontal sutures of the upper interambulacral plates. The
plates nearer the ambitus and on the actinal surface each carry two primary
tubereles. The ambulacral plates carry one primary tubercle. The sutures of
the abactinal coronal plates are somewhat sunken and bare, as in some species

of Goniocidaris.

Dialithocidaris gemmifera A. Ao.
Plate V. Figs. 1, 2.

There are only four anal plates in the single specimen we dredged. The
genital and ocular plates are crowded with irregularly arranged sessile spines,
either globular or elubshaped. The madreporite is well developed, The ac-
tinal system is marked by ten large elliptical plates placed in the extension of
the ambulacral system. The longest primary radioles are 8 mm. in length,

76

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

flattened, fluted and serrated on the edges, and are comparatively smaller and
more slender than in species of the allied genus Podocidaris. The diameter of

the test of this specimen was 21 mm. when alive, The color of test and spines |

is yellowish brown. |

DIADEMATIDÆ, PETERS. i
DERMATODIADEMA, A. Ac,

This genus holds the same relation to Aspidodiadema which Echinothrix
holds to Diadema. It differs from it in having only small secondary tubercles
in the ambulacral areas while in Aspidodiadema there are large primary tuber-

cles in the actinal region, as in Hemicidaris.

Marked for its high test, the five large plates surrounding the anal opening, |
o ? o O d |
the great width of the ambulacral area, and the stout primary interambulacral

Dermatodiadema globulosum A. Aa. |
Plate V. Figs. 3, 4. |

radioles, and the deep furrows of the scrobicular area. The longest speci-
men collected measured 22 mm, in diameter, the largest primary radioles are
nearly twice the length of the diameter of the test.

Station No. 3381, north of Malpelo Island, in 1772 fathoms.

143

This species has a comparatively large apical system. The anal system is
covered by a large number of small plates, a ring of seven or eight somewhat
larger plates surrounding the anal opening. The greater number of specimens
were about 20 mm. in diameter, the apical system measuring 14 mm. across.
It is probable that Aspidodiadema antillarum, collected by the “ Blake,’
have to be transferred to Dermatodiadema, the lack of material inducing me
to associate it with Aspidodiadema, although it possesses the miliary primary
ambulacral tubercles characteristic of the genus Dermatodiadema. Test dark

claret color, primary radioles pinkish.
Station No.

66

3398, off Galera Point, in 1573 fathoms.

Dermatodiadema horridum A. Ac.
Plate V. Figs. 5-7.

’ will

3362, between Cocos Island and Mainland, in 1175 fathoms.
3363, between Cocos Island and Mainland, in 978 fathoms.
3364, between Cocos Island and Mainland, in 902 fathoms.
3375, south of Malpelo Island, in 1201 fathoms.

3376, near Malpelo Island, in 1132 fathoms.

3381, north of Malpelo Island, in 1772 fathoms.

33

?
3400, from Galera Point to Galapagos, in 1322 fathoms.
3413, northwest of Culpepper Island, Galapagos, in 1360 fathoms.

398, off Galera Point in 1573 fathoms.
)

AGASSIZ: PRELIMINARY REPORT ON THE ECHINI. Ul

ECHINOTHURIDÆ, Wyv. Tuoms.

Phormosoma panamensis A. Ac.

The specimens of this species were all in a poor condition ; they are allied to
Ph. tenuis, but this species is interesting in having on the actinal side the char-
acters of Phormosoma most decidedly developed, while on the abactinal side the
great elongation of the ambulacral plates and the arrangement of the coronal
plates resemble the structural features of Asthenosoma. Better preserved
specimens may show this species to belong to a new genus of the family
intermediate between Phormosoma and Asthenosoma.

Station No. 3374, southwest of Malpelo Island, in 1825 fathoms.

Phormosoma hispidum A. Ac.
Plate VI., Plate VII,

This species is the Pacific representative of the Caribbean and Northern
Atlantic Ph. uranus; it has like it an extensive geographical range, but in
comparatively deeper water. The largest specimens collected measured 201 mm.
in diameter. The test is marked for the great width of the ambulacral sys-
tem at the ambitus, which is nearly as wide as the adjoining interambulacral
system. The large primary radioles of the actinal edge are slightly curved and
tipped. On the abactinal side of the ambitus the coronal plates both ambu-
lacral and interambulacral carry only one small secondary tubercle on the
distal extremity of the plate with a few irregularly scattered miliaries. The
outer primary row of interambulacral tubercles extends on half the coronal
plates, the inner row only on two or three of them.

PETALOSTICHA, HAECKEL
SPATANGIDÆ, AGass.
POURTALESI A, A. Ac.

Pourtalesia Tanneri A. Ac.
Plate VII. Figs. 1, 2.

This species is closely allied to P. laguncula (Challenger Eehini, Plate
XXII” Figs, 7-15, Plate XX XT. Figs. 1-11), but differs from it in the shape
of the test, which is less hottle-shaped, the greater height of the anterior ex-
tremity of the test, the concentration of the primary tubercles on the sides of
the test in more or Jess parallel rows in a triangular space extending from the
anal system to the lower angle of the ambitus, and the greater length of the

78 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

primary spines on the flanks of the test. The actinal side of the test is flatter
than in P. laguncula and the proboscis runs at a less angle from the plane of
that surface. The color of the test is pinkish ; the radioles are white.
Station No. 3411, between Bindloe and Wenham Islands, in 1189 fathoms. |
a “ 3431, off Altata, Gulf of California, in 995 fathoms.

PLEXECHINUS, A. Ac.

Only two specimens of this interesting genus were collected. This genus is
peculiar as combining some of the features of Urechinus and Pourtalesia. Tt
has the flush actinostome of the former genus, and at the apex the widely sepa-
rated bivium and trivium of Pourtalesia, with simple ambulacral pores, a slightly
developed anal proboscis, a sunken anal system, and a well developed broad
subanal fasciole.

Plexechinus cinctus A. Aa.
Plate VIII. Figs. 3, 4. |

Seen in profile, the test slopes very gradually from the bivium towards the
rounded anterior extremity. The posterior end is truncated, deeply scooped
out above the broad anal proboscis for the reception of the anal system. The
anal proboscis is rounded, curved back anteriorly, and is gradually lost in the
keel of the actinal plastron. The anal proboscis is banded by a wide fasciole,
extending far beyond the posterior edge of the anal system to the keel of the
actinal plastron. On the actinal side the posterior ambulacral areas are broad
and bare. At the anterior extremity towards the ambitus the test is closely
covered by primary tubercles; they become smaller as they pass beyond the
ambitus and cover the abactinal area, and are quite regularly scattered over the
test with somewhat distant minute miliaries in the intertubercular spaces,
The anal system is transversely elliptical, strengthened by eight large trape-
zoidal plates surrounding the central anal opening.

Station No. 3424, off Tres Marias, Gulf of California, in 676 fathoms.

Echinocrepis setigera A. Ac.
Plate XIII.

Although a number of fragments of this species were collected, only a single
fairly complete specimen was obtained from Station 3399. It meastured 99 mm.
in length, 51 mm. in greatest width, and 53 mm. in height. The test was of
chocolate color; the primary spines from 20 to 22 mm. long were pinkish,
This species is at once distinguished from E. euneata by the great elongation
of the odd interambulacrum and the more rounded outline of the test, which is
far less angular than that of E. cuneata (Challenger Echini, Plate XX VIT. Figs.
1-5). The anterior ambulacral groove is but slightly depressed above the
ambitus; the actinal pouch is entirely on the oral surface, and the anterior ex-

AGASSIZ: PRELIMINARY REPORT ON THE ECHINI. 79

tremity does not pass, as in E. cuneata, into the anterior ambulacral groove.
The actinal system is protected by five large plates.
Station No. 3398, off Galera Point, in 1573 fathoms.
de “3399, off Galera Point, in 1740 fathoms,
bie “ 3415, off Acapulco, in 1879 fathoms.

Urechinus giganteus A. Ao.
Plate VIII. Figs. 7, 8.

At Station No. 3431 two specimens of Urechinus were collected, much larger
than those of any of the species dredged by the “ Challenger” or the “ Blake,”
the largest specimen measuring 93 mm, in length, 80 mm. in width, and 39 mm.
in height. This species is also remarkable for the great length of the primary
spines, which in proportion to the length of the test are nearly as long as those
of Linopneustes, Above the ambitus the test is covered by minute miliaries, scat-
tered uniformly over the surface of the coronal plates. The primary tubercles
are placed uniformly over the abactinal part of the test both in the ambulacral
and interambulacral areas. They are somewhat more closely placed near the
ambitus and in the interambulacral areas of the actinal surface they are closely
packed, forming a distinct actinal plastron in the odd interambulacrum.
The phyllodes extend to the sixth or seventh pore from the actinostome. It is
interesting to note that there are quite well developed “ bourrelets ” separating
the ambulacral areas, a feature characteristic of such genera as Conolampas,
Conoclypus, Echinolampas, and the like.

Station No. 3431, off Altata, Gulf of California, in 995 fathoms.

Cystechinus Loveni A. Ac.
Plate IX.

The test of this species is comparatively stout; it is closely allied to Cyste-
chinus Wyvillii, but can at once be distinguished from it by its transverse anal
system (it is longitudinal both im O. Wyvillii and ©. elypeatus), by the great
size of the actinal interambulacral plate of the actinal plastron, and by the struc-
ture of the apical system which is intermediate between that of O. clypeatus
and ©. Wyvillii (Challenger Echini, Plate XXIX.” Fig. 1, and Plate XX XV.»
Fig. 10).

Station No. 3415, off Acapulco, in 1879 fathoms.

Cystechinus Rathbuni A. Aa.
Plate X.

A number of specimens in excellent condition were dredged from the two
localities at which this species was procured. The test of this species is thinner
than in ©. Loveni, but somewhat thicker than in ©. vesica. A few of the speci-
mens came up which had preserved their shape, so we are able to give figures

80 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

of the outlines of the group of Cystechinus allied to ©. vestca. The specimens
collected varied from a brilliant dark violet to a light claret color.

The anal system is very large, placed well above the ambitus, On the actinal
side, the primary tuberculation of the test is smaller and less crowded than in
C. vesica, There are four genital pores in all the specimens collected; another
feature distinguishing it from ©. vesica.

Station No. 3360, southwest of Mariato Point, in 1672 fathoms.

“ 3374, southwest of Malpelo Island, in 1823 fathoms.

ANANCHYTIDZE, ALB. Gras.
PHRISSOCYSTIS, A. Ac.

This genus is allied to Paleeotropus and Palæobrissus in having like them a
simple ambulacral system, without even the slight trace of petaloid arrangement
found in Palæobrissus, The petals are perhaps even more like those of Cyste-
chinus, Echinocrepis, Gonicopatagus, Calymne, and the like. At the actino-
some, however, the phyllodes attain an extraordinary development, recalling
those of Paleopneustes and Linopneustes. The apical system is compact,
similar to that of the genera last named, and the primary spines are long and
curved, recalling those of the same genera,

Phrissocystis aculeata A. Aa.
Plate XII, Figs, 1-7.

Although not a single complete specimen of this species was obtained, a
sufficient number of larger fragments were collected to enable us to reconstruct
fairly accurately this interesting sea-urchin. Its general facies must have
been similar to the species of Linopneustes and of Paleopnenstes. The test
is thin, covered with large distant primary radioles of uniform size. The
outline of the test must have been somewhat conical, the apex slightly ec-
centric posteriorly, the oral plastron very prominent and flanked by the wide
bare areas of the posterior ambulacra. The apical system compact, with four
large genital openings ; the madreporic body covers the whole of the posterior
part of the apical system, and surrounds the anterior genital pores. The ocular
plates are small but distinct, with large pores. The smaller plates of the
abactinal part of the ambulacral areas are bare, the larger plates carry first one,
and at the ambitus there are from eight to ten distant primaries. In the odd
interambulacrum there is a slight, median furrow extending a short distance
from the apex to the anal system. The whole test is covered with distant
miliaries irregularly seattered over the coronal plates. The anal system is
transversely elliptical and protected by a number of irregularly shaped plates;
the phyllodes are greatly developed. The largest specimens must have been
100 mm, in length, and 50 mm. in height, and from 80 to 90 mm. in width
across the anterior half of the posterior interambulacral area.

Station No, 3366, in 1067 fathoms.

AGASSIZ: PRELIMINARY REPORT ON THE ECHINI. 81

Homolampas hastata A. Aa.
Plate XI. Fig. 1.

This species is distinguished by the great distinctness and sharpness of its
peripetalous fasciole, which is somewhat pentagonal in outline, and surrounds
the few primary tubercles found near the apical extremity. The plates of both
of the ambulacral and interambulacral areas of the abactinal part of the test are
covered by small secondaries and miliaries; this tuberculation is closer than
in any other species of the genus; the posterior extremity is more vertically
truncated, and the anal groove not so pronounced as in H. fulva. The anal
system is pyriform, its greatest diameter being transverse, as in H. fragilis.
There are only two of the larger primary tubercles in the lateral interambu-
lacra carrying long curved spines.

Station No, 3363, northeast of Cocos Island, in 976 fathoms.

3 “ 3365, northeast of Cocos Island, in 1010 fathoms.
s “ 3376, South of Malpelo, in 1132 fathoms.

BRISSINA Gray.

Aérope fulva A. Ac.
Plate VII. Figs. 5, 6.

This species is readily distinguished from the other species of the genus by
its proportionally greater length, the lateral flattening of the test, and the
pointed anal rostrum. The tuberculation is also closer, and the primary spines
are slender in comparison with the stouter spines of Aérope rostrata. The anal
plastron is elongate, triangular, closely packed with primaries, the actinostome
longitudinally elliptical. The anal system is somewhat pyriform, with the
anal opening in the posterior part of the anal system.

Station No. 3361, on way to Cocos Island from Mariato Pt., in 1471 fathoms.

À “ 3362, on way to Cocos Island from Mariato Pt., in 1175 fathoms.
i “ 3381, north of Malpelo Island, in 1772 fathoms.
t “ 3398, off Galera Point, in 1573 fathoms,

s “ 3399, off Galera Point, in 1740 fathoms.

Schizaster latifrons A. Ao.
Plate XI. Figs. 2, 3.

At Station 3431, in 995 fathoms, were collected specimens of a species of
Schizaster remarkable for the great development of the anterior extremity, the
breadth of the anterior ambulacrum, and the short posterior pair of ambulacral
petals, This species belongs to the Schizaster group, of which S. Philippii is a
well known representative.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Schizaster Townsendi A. Aa.

A large series of this species was collected during our cruise from a number
of localities. It is marked by the flatness of the test and the great width of
all the lateral ambulacra, the small size of the anal system, the close primary
tuberculation of the actinal plastron, which is in striking contrast to the bare
actinal surface.

Station No. 3394. Gulf of Panama. 511 fathoms.
113 143 3419. (13 113 [11 [14
[73 “cc 3494. 114 [23 674 [23
m “ 3425. ne 7 al tt
“4, 3426. a “ 146, .
a “ 3431. n ” 995. . *
is “3436. & ue OOS sit

tf “ 3437. 50 miles south of Guaymas. 628 “

Periaster tenuis A. Ac.
Plate XI. Figs. 6, 7.

This species is much flatter and less globular than the species from the Gulf
of Mexico (P. limicola). It has no anal fasciole; the peripetalous fasciole is
wide, with prominent miliaries; the actinal plastron is elongated, and the
tuberculation of the test close. In life the color of the test is light brown.

Station No. 3381. Off Galera Point. 1772 fathoms.

Fi ‘6 3398. s ry us 1573 “e
pei ‘6 3999. da a y 1743 di

Brissopsis columbaris A. Ac.

The Pacific Brissopsis is readily distinguished from the Atlantic species by
the great length of the lateral ambulacra, the flatness of the test, and the great
width of the area enclosed by the subanal fasciole. The anal extremity of the
test is also more sloping than in the European species, and characterized by
the great size of the anal system. The great variation found in specimens of
the genus both on the Atlantic and Pacific sides of the Isthmus of Panama
leads me to think that there has been some confusion in referring tó Brissopsis
such elongate types of Spatangoids as I figure on Plate XX VI. Figure 7 of the
Blake Echini.* I shall refer to this again in my final Report on the Albatross

Echini.
Station No. 3353. 695 fathoms.
ge “ 3356. Off Mariato Point. 546 #
e “ 3382. Off Mala Point. 1793 ff
“3394. Panama Bay. 611 ff

1 Mem. Mus. Comp. Zoöl., Vol. X., No. 1, 1883,

AGASSIZ: PRELIMINARY REPORT ON THE ECHINI. 83

Toxobrissus pacificus A. Ac.
Plate XI. Figs. 4, 5.

There occurs in the Pacific a Spatangoid which has been regarded as allied
to Brissus. Specimens of it are known to me from the Sandwich Islands and
from Zanzibar. A species closely allied to the above mentioned specimens has
been dredged off Point Mala, at Station No, 3355, in 182 fathoms. I am in-
clined at present to place these specimens in the genus Toxobrissus of Desor.
The species dredged by the Albatross are marked for the flatness of the test,
the confluence of the posterior ambulacra along the median line for nearly
half their length, the great width of the posterior extremity of the test, the
large and uniform size of the posterior ambulacral plates on the actinal side
of the test, as well as the small size of the actinal plastron.

SPATAGODESMA, A. Ac.
Plate XII. Fig. 8.

From Station 2769, during the voyage of the “Albatross” from New York
to San Francisco, were obtained specimens of a small species of Spatangoid,
in which the character of the apical fasciole differs widely from that of any
Spatangoid known to me, It possesses a broad elliptical fasciole encircling
both the ambulacra and the anal system. A transverse band divides the fas-
ciole into two areas, one enclosing the anal system and the other becoming the
peripetalous fasciole. Such a fasciole is unknown to me, and among the young
Spatangoids I have had oceasion to examine nothing similar exists, The near-
est approach to the fasciole of this genus, for which I propose the name Spata-
godesma, seems to exist in the young of Agassizia, in which an imperfect
subanal fasciole branches off from the peripetalous fasciole. The actinostome
is still quite central, and no prominent posterior labrum is as yet developed in
the largest specimen, which is about 7 mm. in length. I have not yet been able
to satisfy myself of the relations of this interesting Spatangoid.

Fig.
Fig.
Fig.
Fig.

Fig.
Fig.
Fis.
Tig.
Fig.

Fig.

Fig.

Fig.
Fig.

Fig.

Fig.

Fig.
Fig.
Fig.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

EXPLANATION OF THE PLATES.

PLATE I.

Dorocidaris panamensis A. Ag., from the abactinal side.

. Dorocidaris panamensis A, Ag., from the actinal side.

All figures natural size.

PLATE Il

Dorocidaris panamensis A. Ag., facing the odd anterior ambulacrum,

. Salenia miliaris A. Ag., abactinal view.

Salenia miliaris A. Ag., partly denuded.

. Saienia miliaris A, Ag., facing the odd anterior ambulacrum.

All figures natural size.

PLATE II.

Goniocidaris Doederleini A. Ag., from the abactinal side.
Porocidaris Cobosi A. Ag., in profile.

.Porocidaris Cobosi A. Ag., from the actinal side.

Porocidaris Cobosi A. Ag., facing the posterior interambulacrum.
Porocidaris Cobosi A. Ag., from the abactinal side.
All figures natural size.
PLATE IV.
Porocidaris Milleri A. Ag., (female) test, seen from the abactinal side.
Porocidaris Milleri A. Ag., from the actinal side,
All figures natural size.
PLATE V.

Dialithocidaris gemmifera A, Ag., from the abactinal side.

. Dialithocidaris gemmifera A. Ag., facing the odd anterior ambulacrum,
. Dermatodiadema globulosum A. Ag., from the abactinal side.

Dermatodiadema globulosum A. Ag., facing the odd anterior ambulacrum.
Dermatodiadema horridum A. Ag., from the abactinal side.
Dermatodiadema horridum A. Ag., in profile.

. Dermatodiadema horridum A. Ag., facing the left anterior ambulacrum.

All figures natural size.

Fig.

Fig.

Fig.

Fig.

Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.

Fig.

Fig.
Fig.

Fig.
Fig.

Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.

2.

AGASSIZ: PRELIMINARY REPORT ON THE ECHINI. 85

PLATE VI.

. Phormosoma hispidum A. Ag., from the abactinal side, odd anterior ambu-

lacrum on the left.

. Phormosoma hispidum A. Ag., (a larger specimen,) from the abactinal side,

odd anterior ambulacrum on the left.
All figures natural size.

PLATE VII.

. Phormosoma hispidum A. Ag., (same as Fig. 1, Pl. VI.) from the actinal

side, odd anterior ambulacrum on the left.
Phormosoma hispidum A. Ag., from the actinal side, odd anterior ambula-

crum on the left.
All figures natural size.

PLATE VII.

. Pourtalesia Tanneri A. Ag., from the abactinal side.

Pourtalesia Tanneri A. Ag., in profile.
Plexechinus cinctus A. Ag., in profile.
Plexechinus cinctus A. Ag., from the abactinal side.
Aérope fulva A. Ag., from the abactinal side.
Aérope fulva A. Ag,, in profile.
Urechinus giganteus A. Ag., from the abactinal side.
Urechinus giganteus A. Ag., in profile.

All figures natural size.

PLATE IX.
Cystechinus Loveni A. Ag., in profile.
Cystechinus Loveni A. Ag., from the abactinal side.
All figures natural size.

PLATE X.
Cystechinus Rathbuni A. Ag., facing the anal system.
Cystechinus Rathbuni A. Ag., from the abactinal side.

All figures natural size.

PLATE XL
Homolampas hastata A. Ag., from the abactinal side.
Schizaster latifrons A. Ag., from the abactinal side.
Schizaster latifrons A. Ag., from the actinal side.
Toxobrissus pacificus A. Ag., from the abactinal side.
Toxobrissus pacificus A. Ag., from the actinal side.
Periaster tenuis A. Ag., from the abactinal side.

. Periaster tenuis A. Ag., from the actinal side.

All figures natural size.

VOL. XXXII. —NO. 5. 2

> © D a

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

PLATE XII.

. Phrissocystis aculeata A. Ag., apical part of test.

. Phrissocystis aculeata A. Ag., odd interambulacral area.
. Phrissocystis aculeata A. Ag., odd ambulacrum.

. Phrissocystis aculeata A. Ag., left posterior ambulacrum.
. Phrissocystis aculeata A. Ag., part of left side of test.

Phrissocystis aculeata A. Ag., anal system.

. Phrissocystis aculeata A. Ag., actinal system.

Figs. 1-7 natural size.

. Fasciole of Spatagodesma. $.

PLATE XIII.

. Echinocrepis setigera A. Ag., apical part of test.

Echinocrepis setigera A. Ag., odd anterior ambulacrum from above.

. Echinocrepis setigera A. Ag., odd anterior ambulacrum from below.
. Echinocrepis setigera A. Ag., posterior interambulacrum from above.

Echinocrepis setigera A. Ag., anal proboscis in profile.

. Echinocrepis setigera A. Ag., anal system from below.

All figures natural size.

PLATE A.

CHART SHOWING THE TRACK OF THE “ ALBATROSS.”

2088

«a

4

ALBATROSS Ex

Agassiz AIBATROSS CRUISE. 1891.

ES Xe
o

Bulletin of the Museum of Comparative Zoology
AT HARVARD COLLEGE.
Vou. XXXII. No. 6.

THE NERVOUS SYSTEM OF NEREIS VIRENS SARS.

A STUDY IN COMPARATIVE NEUROLOGY.

By J. I. HAMAKER.

Wirn Five PLATES.

CAMBRIDGE, MASS., U.S. Bad
PRINTED FOR THE MUSEUM,

June, 1898.

|
|
|

No. 6.— The Nervous System of Nereis virens Sars. À Study
in Comparative Neurology. By J. I. HAMAKER.!

CONTENTS.
Page Page
Introduotlotie ui: sil Wiles ed e. Peripheral Fibres . . . 105

Methode can. Dart cle Donnin 0.106

Part dl. Destripon,.« «u... De Ws. MODAN. cp ar an UO
I, @opograpny. vos. 0. . UL ee a POOO Liste « + . « LOY
a. Brain 02) 3. Brain. >. ¿A a8

. “Mushroom Bodies” . . . 109
- Optic-Gangllod: w una A
. Ventral Nerve Cord. . . . 112

b. Sub-esophageal Ganglion 93
c. Typical Body Segment . 94
d. Parapodial System. . . 95

DIS 0 »

2. Protective Tissue . . , . 96 ‘-DONITOSOMES A AUS
Oy USA A we ee VC fF NGEVO PIDU WoW yo se ke
4. Ciliated Groove . . . . . 100 ti Giant Pibres: 277 “HS
5. Ventral Nerve Cord. . . . 100 D. Fibres of Set À v ao ein ‘M6
6. Nerve Fibres . x , , . . 102 O: Fibres:of Set-B w O

Se Gant Fibres ows LOR d. Fibres of Set C . . . . 118

Dv, BDIGH OL SOLAS su 22 LUO MSUMMANS ut una in sue. LAS
C Eibresotset so’, . , 108 Bibliography . +. . 5... 0... 200
d. Fibres of Set C . . . . 105! Explanation of Plates . . . . . 124

INTRODUCTION.

ALTHOUGH 80 much has been written on the nervous system of repre-
sentatives of all the chief groups of the metazoa, we are as yet far from
thoroughly understanding the action of the myo-neural mechanism of
any animal. It is true, much light has been thrown upon the subject
during the last decade through the use of the newer methods of investi-
gation; but the many valuable facts that have been established are as
yet so disconnected, that they can scarcely be said to be more than stig-
gestive. In no case has the myo-neural system of an animal been at all
completely worked out. In the higher animals the complexity of this
system makes such a task almost impossible. At any rate, the most

1 Contributions from the Zoölogical Laboratory of the Museum of Comparative

Zoölogy at Harvard College, under the direction of E. L. Mark, No. XCI.
VOL. XXXII. — NO, 6, 1

90 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

promising method of approaching the subject would seem to be through
the simpler forms. Some of the homodynamously segmented animals
offer advantages in this respect which no other forms do.

The following considerations may be noted as applying particularly
to the polychaete annelid, Nereis, as a favorable subject for this kind of
study :—(1) Nereis has a sharply centralized nervous system, consisting
of well defined ganglia, which occupy very definite relations to the parts
they control. (2) The histological elements of the nervous system are
highly differentiated and constant in their relationships. (3) A typical
body segment of the animal is simple in structure, having a compara-
tively small number of muscles. (4) There is almost no serial differen-
tiation of the body segments, excepting in a few of the anterior and the
anal metameres ; hence it is necessary to determine the structure of one
segment only in order to know the structure of nearly the whole animal.
(5) Since there are only a few muscles, the movements of the animal are
limited in number, and may readily be analyzed and classified. (6) Phy-
siological experiments may be performed with more than usual facility,
because the worms are hardy and live well in the aquarium. To these
considerations may be added the eminently practical one that Nereis
may be easily obtained in unlimited quantities. With such material
the problem before us seems to be presented in a comparatively simple
form. By these considerations I was led to take up the study of the
nervous system of Nereis. Some of the results obtained are given in
what follows.

Methods.

The material used for intra-vitam staining was the atokal form of
Nereis virens Sars. It was obtained from the muddy banks of the
Charles River at a place where the water at low tide contains less than
0.3% salt. The worms found here may be transferred to fresh water
without suffering serious injury. For ordinary histological preparations
both atokal and epitokal forms were used. The latter were obtained
from the mud flats of Lynn Harbor at the mouth of the Saugus River.

Before killing, the specimens were narcotized with chloral hydrate or
alcohol, and the intestine cleaned by forcing a stream of water through
it. The body was always opened to allow rapid penetration, and some-
times the intestine was removed, Two methods of fixing and staining
were employed: either fixing in corrosive sublimate and staining in iron-
hamatoxylin, or fixing and staining by osmic acid. The value of corro-
sive sublimate as a fixing agent is well known, and I obtained excellent

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 91

results by fixing in a 1% to 5% solution of acetic acid saturated with
corrosive sublimate. Preparations fixed in this way were stained on the
slide with Heidenhain’s iron-hematoxylin. The osmic acid method of
vom Rath (95) gives results in many ways equally good, and for some
purposes, such as tracing nerves, even better than the corrosive and
iron-hæmatoxylin method. I found a mixture in the following propor-
tions of osmic acid, picric acid, acetic acid, and platinic chloride very

satisfactory : —
Omicidi 270. MEA AAC, LA ec
Piorio acid, sat. aq. sol. , , , 100%
Pinto hlorida 24. wo A 200%
AO LIO AI. se A, 166

The results obtained are not at all uniform in quality, since the rate of
preeipitation of the osmium by the pyroligneous acid seems to vary.
The value of successful preparations, however, counterbalances the
capriciousness of the method. The results obtained by these two
methods agree in almost every particular, even to the relative intensity
of the stains in the various tissues.

For intra-vitam staining the following method proved most success-
ful. Specimens of Nereis having about seventy segments were injected
with a concentrated solution of methylen blue in normal salt solution.
They were then laid, ventral side uppermost, in a moist chamber for
about two hours, after which the stain began to appear in the sub-
esophageal ganglion, From this region the stain gradually penetrated
eaudad, and when it was thought to have reached its optimum, it was
fixed by Bethe’s (95) ammonium molybdate method. The objects were
then embedded in paraffine and cut.

PART I. DESCRIPTION.
1. TOPOGRAPHY.

The central nervous system of Nereis virens is well developed.
Throughout the entire length of the body the ventral nerve cord ex-
hibits a sharp differentiation of ganglia and longitudinal connectives.
The ganglia are segmentally arranged and constant in position; the
nerves are regularly arranged in metameric groups of five pairs each
(Plate 1, Fig. 8). The ventral cord lies deeper than the hypodermis,
from which it is partially separated by the circular muscle bundles.

92 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,
The circular muscles do not form a continuous sheet, but consist of
small bundles which lie partially embedded in the hypodermis. Some
of these muscles cross the mid-ventral line external to the nerve, thus
causing a partial separation of cord and hypodermis, Between the
muscle bundles, however, the neurilemma of the nerve cord is in contact
with the hypodermis. The brain also lies deeper than the hypodermis,
from which it is suspended by a narrow membrane lying in the median
plane.
a. Brain,

The form of the brain is roughly that of a trapezoid (Plate 1, Fig. 1,
ceb.), the anterior pair of eyes marking approximately the extremities of
the longer one of the parallel sides, while the posterior pair marks the
limits of the shorter one. The anterior angles of the trapezoid are
drawn out toward the palps, thus making the anterior margin of the
brain slightly concave. The dorsal aspect of the brain is broadly cor-
date, the re-entrant angle being at the anterior side. Fourteen pairs of
nerves arise from the brain by distinct roots. As they are arranged
symmetrically, it will not be necessary to describe both sides. Beginning
anteriorly at the median line, and numbering and describing the nerves
of one side in order, there is first near the median line a group of three
nerves (I, II, IIT), which arise near together.

The first nerve (I) passes forward, then downward, and finally back-
ward along the dorsal wall of the proboscis; the second (Il) goes
directly forward to the antenna; the third (III) runs forward along the
dorsal wall of the head.

At the anterior lateral angle of the brain there is another group of
three nerves (IV, V, VI). The fourth nerve (IV) divides into two
branches, one going to the ventro-median wall of the palp, the other to
the dorso-median wall of the same organ. The fifth nerve (V) extends
ventrally to the proboscis; the sixth (VI) is the large sensory trunk
of the palp; and the seventh (VII) arises from the brain laterally,
between the group just described and the anterior eye of the same side;
it passes forward along the lateral wall of the palp.

The eighth, ninth, and tenth nerves are the three roots of the circum-
œsophageal commissure. They unite in the commissural ganglion, which
lies a short distance ventral to the anterior eye. The eighth (VIII) is
a small nerve arising near the seventh, passing out parallel with it, and
then turning down into the ganglion. The ninth nerve (IX) arises
laterally from the brain, passes out directly beneath the eye, and then
bends down to the commissural ganglion. The tenth (X) arises from the

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 93

ventral edge of the brain immediately ventral to the ninth, and passes
out directly to the commissural ganglion.

The eleventh (XT) and twelfth (XII) nerves are the two optic nerves.
They converge from the eyes toward the centre of the brain.

The thirteenth (XIII) nerve arises back of the posterior eye, and
goes directly to the ciliated groove. The fourteenth (XIV) is a rather
diffuse nervous connection between the brain and the dorsal surface of
the head. The region innervated lies nearly midway between the pos-
terior eye and median plane, but slightly nearer the latter.

Besides these fourteen paired nerves (I-XIV) there is a single median
nervous connection between the dorsal surface of the head and the
brain. This is similar to the diffuse fourteenth nerve, but is smaller
and lies slightly anterior to it. Its position is shown in Figure 1.

From the commissural ganglion a nerve (a) passes forward to the
proboscis, where it unites with the fifth nerve of the brain (V) in a
ganglion. Another nerve (8) passes backward along the side of the
head. Four or five small connectives, not shown in Figure 1, unite
the commissural ganglion with the optic ganglion, which lies in con-
tact with the ventral side of the anterior eye. Lastly, the circum-
Ͼsophageal commissure passes from the commissural ganglion around the
cesophagus to the sub-cesophageal ganglion, traversing on its way a
ganglion which lies beneath the anterior pair of tentacular cirri. From
this anterior cirrus ganglion two large nerves go each to an anterior
cirrus, and, from the anterior side of it, a smaller one (8) to the pro-
boscis. On its posterior side the anterior cirrus ganglion is connected
by a small nerve (0) with the posterior cirrus ganglion, which lies
beneath the posterior pair of tentacular cirri. The latter ganglion gives
off two large nerves, one to each of the two posterior tentacular cirri,
and also sends. a nerve (4) backward along the side of the head. The
posterior cirrus ganglion is connected with the sub-cesophageal ganglion
by a large nerve trunk (n. pa-coms.), which lies posterior to and parallel
with the circum-osophageal commissure. This trunk gives off several
branches from a region midway between the posterior cirrus ganglion
and the sub-œsophageal ganglion.

b. Sub-esophageal Ganglion.
The sub-cesophageal ganglion (gn. sb-oe., Fig. 1) lies on the posterior
boundary of the cephalic segment. It gives off three pairs of nerves
besides the commissures and the nerves of the posterior cirri. Begin-

94 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

ning at the median plane, the first pair of nerves (y) goes to the
esophagus; they pass forward near the median line, and then bend
backward to be distributed to the ventral wall of the csophagus. The
second and third pairs are respectively the commissures and the nerves
of the posterior cirri; the other two pairs pass out laterally and forward
to the walls of the cephalic segment. The ganglia of the first three or
four body segments are displaced backward, as compared with those of
the typical segment, and are consequently crowded together, thus mak-
ing them appear as a single ganglion. Each segment in this region
receives the typical number of nerves, though in regard to size the
nerves of each segment do not bear quite the same relation one to
another as they do in a typical segment.

c. Typical Body Segment.

The ganglia of the ventral chain (Plate 1, Fig. 8) are about half as
long as the segment, and, if the segmentation of the longitudinal mus-
culature be taken to indicate the boundaries of the metameres, they lie
intersegmentally, with at least two thirds their length in the posterior
segment. The longitudinal connectives are enclosed in a single sheath,
so that there appears to be but one connective, At the intersegmental
plane there is a constriction of the investing tissue of the ganglion, but
this does not extend to the nervous part. Five larger pairs of nerves
are given off from each ganglion, and there are, besides, many smaller
ones, which consist of only a few fibres each. A few of the latter are
given off ventrally, but most of them pass dorsally from the mid-dorsal
line of the nerve cord, and probably are distributed chiefly to the
digestive tube.

The paired nerves are most conveniently characterized with reference
to their position in the segment. Beginning at the anterior end, there
is close behind the intersegmental constriction of the ganglion the first
segmental nerve (I, Fig. 8). This nerve is rather slender and passes
out at right angles to the ventral nerve cord. It lies external to the

longitudinal muscles, and is partly buried in the hypodermis (1', Fig. 4).
It may be traced with little change in calibre to the dorsal base of the
parapodium, where it is lost either in the circular parapodial muscles,
or in the hypodermal plexus, or most likely in both.

The second segmental nerve (11) is the largest of the five, and is the
parapodial trunk. It leaves the segmental ganglion near its posterior
end and passes diagonally backward across and external to the longi-

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 95

tudinal muscles to the parapodial ganglion, which lies in the ventral
base of the parapodium between the longitudinal muscles and the pore
of the segmental organ.

The third segmental nerve (III) is very small. It arises from the
posterior end of the segmental ganglion and passes diagonally outward
and backward to the posterior base of the parapodium, Like all the
paired segmental nerves, it lies embedded in the hypodermis throughout
its length. Next in size to the parapodial trunk is the fourth nerve
(IV). It arises from the anterior third of the succeeding ganglion, and
hence lies in the posterior part of the segment, near the intersegmental
line, where the longitudinal muscles are attached to the hypodermis.
It runs parallel with the first nerve (1) of the following segment, but
continues in a direct course around the body almost to the mid-dorsal
line.

The fifth segmental nerve (V) is very slender. It lies close behind
the fourth nerve, and very near the intersegmental plane. It extends
as far as the base of the parapodium,

d. Parapodial System.

The innervation of the parapodium (Figs. 5, 8) is almost wholly from
the parapodial ganglion, from which four nerves radiate toward the
periphery. The most anterior one (1) is very slender and usually passes
out in front of the pore of the segmental organ, It goes to the retractor
muscles of the anterior side of the parapodium. The second parapodial
nerve (2) is comparatively large, and innervates the ventral ramus.
Near the ganglion it divides into an anterior and a posterior branch.
The anterior one passes along the anterior wall of the ventral ramus
to the anterior setigerous lobe. The other runs along the posterior wall
to the posterior setigerous lobe, giving off a branch to the ventral cirrus
and another to the inferior ligula of the ventral ramus.

The third parapodial nerve (3) passes outward and upward on the pos-
terior wall of the parapodium. About half way up the side of the para-
podium it gives off a branch to the glandular region of the dorsal base
of the parapodium. Then other branches are sent to the anterior setige-
rous lobe of the dorsal ramus, the posterior setigorous lobe, the superior
ligula, and the dorsal cirrus.

A fourth parapodial nerve (4) goes to the posterior base of the

parapodium.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

2. Protective TISSUE.

The protective tissue of the central nervous system is of two kinds ;
an outer membrane, the neurilemma, and within this a spongy tissue,
the neuroglia. The neurilemma is better developed in the anterior than
in the posterior region of the animal, and better in the older epitokal
individuals than in the atokal forms. In the cephalic segment it is very
thick and forms a capsule around the brain (n’lem., Plate 2, Fig. 9),
and it also envelops the nerves from the brain. A tissue similar in
toxture lines the dorsal wall of the head, there taking the place of a
basement membrane (Plate 3, Fig. 20). The brain capsule is continuous
with this lining of the wall of the head along the mid-dorsal line, and
also around the nervous connections that unite the brain with the poste-
rior dorsal surface of the head. It likewise serves as a place of attach-
ment for some of the muscles of the head, as the neurilemma of the
ventral cord does for some of the diagonal muscles ; but elsewhere the
capsule is free from the wall of the head, being suspended in the cavity
of the cephalic lobe.

The neurilemma of the brain is continuous with that of the ventral
cord along the circum-csophageal commissures. Except where it is
pierced by nerves, the neurilemma of the cord (Plate 2, Fig. 18) forms
a closed tube, whose walls are thickest at the anterior end. Along its
dorsal side the wall of the neurilemma tubo is continuous with the
tunica intima of the ventral longitudinal blood-vessel by means of a
narrow membrane which connects the two tubes throughout their
entire lengths (Fig. 18).

In structure the neurilemma is uniformly fibrous; it does not stain in
iron-hamatoxylin or osmic acid. On the outer surface of the brain cap-
sule there is a layer of nuclei which may possibly belong to the neuri-
lemma. I have not clearly seen such nuclei on the neurilemma of the
ventral cord. The connective tissue of the muscles is continuous with
the neurilemma in many places, and resembles it in every respect.

The neuroglia of the brain is a network of delicate fibres with numer-
ous small clongated nuclei located at the nodes of the network. This
tissue lines the inner surface of the brain capsule, from which it pene-
trates into all parts of the brain except the neuropil and the masses of
small nuclei connected with it. The neuroglia of the ventral cord is like
that of the brain in texture, and it surrounds and penetrates the nervous
structures of the cord in the same way as in the brain. The portion
immediately surrounding the cord, however, is somewhat differentiated

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 97

from the more peripheral part (Plate 2, Fig. 18). Its fibres are coarser
and more compact, and they take a circular direction around the cord.
Between the successive ganglia the fibres of this inner layer also pass
in between the three longitudinal connectives. A few neuroglia nuclei
lie scattered about among the nerve fibres of the cord.

3, BRAIN.

The brain of Nereis consists essentially of a central mass of interwoven
fibres with a few irregalar masses of neuropil, and a peripheral layer of
cells loosely arranged in symmetrical groups. The cells have undergone
a remarkable degree of differentiation, and the cells in each pair of groups
have distinct characteristics. There are at least six distinct classes of
cells; while a number of the larger cells are arranged symmetrically in
pairs, each pair having individual peculiarities of form and structure,
The six classes, however, are not characterized by form and structure
alone. Indeed, it is the difference in chemical reaction that is most
distinctive, and calls for more than passing mention. The classes are
as follows: —

(1) In front, on each side of the brain, between the anterior median
and the anterior lateral groups of nerves, there lie two masses of exceed-
ingly small nuclei (Plate 3, Figs. 21, 24). The larger ventral mass is
approximately crescentic in a transverse section of the brain with the
concave side of the crescent lateral and partially embracing a mass of
neuropil. The smaller, dorsal mass of cells is also crescentic in trans-
verse section, but with the concave side toward the median plane. This
crescent also partially embraces a mass of neuropil. In preparations
stained with osmic acid or iron-hematoxylin the cells of these masses
show only very faint traces of cell boundaries. The nuclei stain very
deeply, and always contain a number of granules of various sizes. The
nuclei are about 6p in diameter and are arranged in rows radiating from
the neuropil (Fig. 21). They are set very closely in the rows, and are
often almost in contact. The spaces between the rows are wider toward
the neuropil, while toward the convex side of the mass the arrangement
is more compact and the alignment of the nuclei almost disappears.
The spaces between the rows of nuclei have a fine fibrous appearance, as
though fibres from the cells passed ont to the neuropil. Cells similar to
these occur in other parts of the brain, e. g. a small group at the posterior
border of the brain, just median to the root of the nerve which runs to
the ciliated groove (XIII). Cells slightly larger, but otherwise like

98 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

these, occur scattered over the dorsal surface of the brain, and a group
of them lies around the root of the fourteenth nerve (XIV).

(2) At the posterior lateral border of the brain there is a group of
spindle-shaped cells, which extends backward and outward along the root
of the thirteenth nerve (XIII) as far as the point where the nerve
pierces the capsule (Plate 2, Figs, 9, 15). These cells do not stain very
deeply, and do not show very sharply defined cell boundarics, although
the spindle form can be distinctly seen. Similar cells give rise to the
fibres that form the fourteenth nerve and pass out to the dorsal sur-
face of the head.

(3) Along the internal border of the last group there lies a third very
peculiar class of cells (Figs. 9, 12). Unlike those of the two preceding
classes, the cells of this class have a very definite cell boundary. They
are comparatively few in number, and are narrowly pear-shaped with the
process extending toward the centre of the brain. In preparations
stained to best advantage for other structures these cells are so deeply
stained that only in a few cases can the nucleus of the cell be scen at all.
The cell process also stains so deeply that it appears in strong contrast
with the other fibres.

(4) In the same transverse section as the last, but nearer the median
plane, is another group of pear-shaped cells (Figs. 9, 11). These are
larger and proportionately broader than the last, and stain very differ-
ently from them. In iron-heematoxylin the cytoplasm does not become
blue, but takes on a brownish color. Its structure is almost homogene-
ous excepting an irregular network of a few coarse fibres which usually
centres about the nucleus and does not extend throughout the body
of the cell, The processes of these cells also go toward the centre of
the brain.

(5) At the side of the brain beneath the nerve of the anterior eye
there lies a group of cells which seem to have no direct connection with
the brain except that of being enclosed in the brain capsule. The cells
are rather large and spherical, and send their processes out along the
ninth nerve (IX) of the brain toward the commissural ganglion, A few
of the cells lie scattered along the dorsal side of the ninth nerve, and
make this group of cells continuous with the group which lies beneath
the anterior eye, and which we have called the optic ganglion. The cells
of both groups have the same general appearance (Fig. 19). The cyto-
plasm presents no special peculiarities. There is no cell membrane and
the limits of the cell are very indistinct, because there are very few
granules at the periphery. Each cell is surrounded by neuroglia fibres

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 99

arranged loosely in concentric layers. The inner fibres seem to be em-
bedded in the cytoplasm of the cell.

(6) Although the remaining cells (Figs. 16, 17) present a great vari-
ety of size, form, and detail of structure, there is a gencral similarity
which permits of classing them together as a group distinct from those
we have described thus far. They have the pear shape and the granu-
lar eytoplasm which are characteristics of most of the cells of the
ventral ganglia. But beyond this there is little that is common to
all the cells of the group. In the posterior half of the brain there are
several pairs of very large cells which have the characteristics of this
group, and in addition a very striking one of their own. The nucleus
lies in the narrow end of the cell, and is surrounded by the granular
cytoplasm. At the other end of the cell, there is a large vacuolar space
containing a number of deeply staining bodies of irregular form, em-
bedded in an indistinct coagulum. Other cells have very finely granular
substance occupying a similar position, the granules being much
smaller and staining less deeply than those of the body of the cell.
In these cases the nucleus shows no signs of degeneration. In some
cells (Fig. 16) the cortical part of the cytoplasm is penetrated by narrow
lamellæ, which, when viewed from the surfaces of the cell, present the
appearance of a honeycomb structure.

There is another structure within the brain capsule which is very
strange, and for which I cannot account. It consists of a considerable
number of spheroidal cavities (Fig. 10), containing a substance which
assumes several forms. The cavities are arranged in two symmetrical
groups, one on each side of the brain (Fig. 9), extending around and
between the fibres of the fourteenth nerve, and backward and outward
to the root of the thirteenth nerve. The cavities, which are surrounded
by neuroglia, vary somewhat in size, the average being about equal to
that of the average nerve cell of the brain. Each usually contains a
number of spherical granules, sometimes of nearly uniform sizo, some-
times differing much in this respect. They are stained in iron-hema-
toxylin, but in preparations treated with osmic acid thoy are yellow,
Sometimes the cavities are filled with an almost homogeneous substance ;
at other times, however, the substance only partially fills the cavities, and
assumes an irregular stellate form. In a few cases there are doubtful
indications of a nucleus.

These structures cannot be due to degeneration of nerve cells, because
they are very regular in the place of their occurrence, and there is no
indication of degenerated fibre tracts. The contents of the cavities are

100 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

not fat, for they are not blackened by osmic acid, nor do they seem to
be pigment, since the granules are comparatively large and at the same
time quite variable in size. Racowitz (95) found amceboid cells de-
positing pigment in the region of the ciliated groove, but I have no
reason to think the condition in the present instance is due to the same
instrumentality.

4. CILIATED: GROOVE.

The ciliated grooves lie on the posterior margin of the cephalic lobe
behind the posterior eyes (Plate 3, Fig. 20). A fold of the anterior
margin of the cephalic segment extends forward a short distance over
the cephalic lobe, to which it is attached in the median plane, and also
at a point just lateral to the eyes, On the ventral wall of the pockets
thus formed there is an elongate eminence or ridge about 0.5 mm. long
and 0.2 mm. wide, with the long axis transverse to the body. Along
the crest of this ridge lies the ciliated groove. The ridge is formed by a
thickening of the non-ciliate hypodermis, the cells of which are very long
and arranged with their distal ends directed toward the crest of the
ridge, thus producing in cross section the figure of an inverted V. The
space between the cells which form the ridge is occupied by the ciliated
cells, The cuticula over these cells is very thin, and the cilia project
through it, forming a narrow band along the bottom of the groove. The
nuclei of the ciliated cells lie much deeper than those of the non-ciliate
cells on either side of them. The organ is innervated by the nerve
XIII. There are no glands in the hypodermis of this region, but the
overhanging fold of the cephalic segment is richly supplied with them,

5. VENTRAL Nerve Corp.

The structure of the ventral nerve cord can best be set forth by de-
scribing transverse and longitudinal sections of it. A transverse section
between ganglia (Plate 5, Fig. 31) shows that there are three longitu-
dinal connectives; two large lateral and symmetrical ones, and a small,
more dorsal median one. Each connective is enveloped by the fibres of
the inner layer of neuroglia, which thus separates the three connectives.
In preparations successfully fixed in either the corrosive sublimate or
osmic acid mixtures, the cross section of the connectives shows nothing
but the circular outlines of the cut nerve fibres, with their contents and
a few neuroglia fibres penetrating the connective from the sides. The
fibres vary greatly in size, from the large giant fibres, which are one

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 101

third the diameter of a lateral connective, to the smallest, which how-
ever are large enough to enable one to distinguish the circular outline of
the sheath and its contents.

There are three giant fibres, one in each connective. Those in the
lateral connectives are much larger than the median one.

On the median side of each of the paired connectives, close beneath
the median connective, there is another very large fibre which, in some
regions of the body, is but little smaller than the median giant fibre
(Plate 5, Fig. 31). These fibres, which I shall call set A, also stain very
lightly, but they show no traces of a network.

The numerous fibres which constitute the remaining portions of the
connectives stain more deeply. Most of them show no differentiation,
but frequently the larger ones are more intensely stained in the centre
than at the periphery.

In longitudinal sections of the connectives (Plate 3, Fig. 26), the
fibres appear as parallel bands separated by crinkled lines, — the fibre
sheaths folded by a slight longitudinal contraction of the animal at the
time of fixation. Many of the larger fibres, excepting set A, often show
a darker central band corresponding to the darker centre of the trans-
verse section. A few nuclei are scattered among the fibres of the
‘connectives.

Transverse sections through the ganglia of the ventral chain present a
single central filrous mass bordered ventrally and laterally by ganglionic
cells. Bundles of neuroglia fibres pierce the central mass at intervals
along the median plane, and divide the ganglion into symmetrical halves.
The greater part of the fibrous mass consists of longitudinal fibres, but
there are many fibres which traverse the ganglion in other directions.

The cells of the ventral ganglia do not vary as much in size, form, and
structure as do those of the brain ; however, besides the uniformly gran-
ular ones of various sizes and shapes (Plate 2, Figs. 13, 14, 4), corre-
sponding to those of class six in the brain, there are some cells (Fig.
14, B) which stain very lightly, and the cytoplasm of which is homo-
geneous with the exception of a few coarse granules of very limited dis-
tribution. There are only a few pairs of these cells in each ganglion ;
one of the pairs belongs to the fibres of set 4, and these are among the
larger cells of the ganglion.

The coarse irregular granules of the cells last described occupy the
middle of the cell, where they are arranged in the form of a hollow
sphere, at the centre of which there is a round deeply staining granule
(Fig. 14, B). This structure is undoubtedly what has been described as

102 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

a centrosome. It is not confined to this class of cells, but in good prep-
arations occurs with such frequency that it may be said to exist in all
the cells of the ventral cord (Fig. 13). The nucleus is always eccen-
tric, and frequently, though not always, flattened. There are often two,
three, or more centrosomes in a single cell. In one instance there were
ten. In the cells whose cytoplasm is granular the centrosome does not
appear as distinctly as it does in the others. However, when the stain-
ing has been successful, there appears at the centre of the cell a mass of
granules which are larger and stain more deeply than those of the rest
of the cell. The granules of this mass are arranged in the form of hol-
low spheres, the contents of which are destitute of all granules excepting
the single round body at the centre.

6. Nerve FIBRES.
a. Giant Fibres.

There are three giant fibres which traverse the ventral cord through-
out its entire length (Plate 2, Fig. 18, Plate 5, Fig. 31); the pair of
extremely large ones, which lie one on the outer side of each of the
paired connectives, and the smaller unpaired one lying in the median
connective, All these have the same peculiarities of structure. With
the methods employed they stain very lightly and appear almost homo-
geneous. On close examination, however, the section of the fibre is seen
to be made up of a small number of polygonal areas marked off by an
indistinct network (Plate 1, Fig. 3). This network apparently owes its
existence simply to the presence of discrete masses of protoplasm, the
boundaries of which give the appearance of a network. In longitudinal
sections the giant fibres show the same structure, except that the polygo-
nal areas are clongated in the direction of the axis of the fibre. When
these fibres are stained in methylen blue, the stain is precipitated at the
borders of the areas, producing a finely granular network in a homoge-
neous field of blue,

The paired fibres may be traced forward into the cireum-Ͼsophageal
connectives to a point between the anterior cirrus ganglión and the
commissural ganglion, where they divide into a number of small branches.
The branches cannot be distinguished from other large fibres of the con-
nective, but they appear to pass through the commissural ganglion to
the optic ganglion. The fibres which connect the commissural and optic
ganglia are processes of the cells of the optic ganglion, but since I was
unable to trace a fibre continuously from the optic ganglion until it

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 103

united with the giant fibre, I cannot be sure that there is such a con-
nection. I have found no other cells connected with these giant fibres.

The median giant fibre divides in the sub-cesophageal ganglion into
several branches, which continue forward parallel with one another along
the median plane. One of them I was able to trace to one of a group of
large cells lying between the ventral ends of the circum-csophageal con-
nectives. The other cells of the group are connected with similar fibres,
but I could trace only one continuously from the cell to the giant fibre.

The three giant fibres extend back into the last segment of the body
without branching or changing their relative sizes or positions. “Occa-
sionally the median fibre in passing through a ganglion divides and
allows the passage of a bundle of fibres between the two parts, which
then immediately reunite, and the fibre continues on as before. This
condition occurs frequently, but appears to be wholly accidental, since it
is very irregular in the frequency of its occurrence, as well as in the size
of the loop produced, and also in the relative sizes of the two divisions
of the fibre. In one instance I found a similar condition in one of the
lateral giant fibres, but it was not very well marked.

The giant fibres are pierced by many smaller ones, which pass directly
through them (Plate 1, Fig. 2). In the case of the lateral giant fibres
this occurs most frequently near the places where the segmental nerves
are given off from the ganglion. Sometimes the small fibres branch
within the large one, the branches then continuing on through the giant
fibre. In preparations stained with osmic acid, the small fibres stain
much more deeply than the giant fibres, thus becoming very distinct.
In a part of a methylen blue preparation which had not taken the stain,
the small fibres traversing the giant fibres could be readily seen because
they were more highly refractive than the giant fibre.

I cannot say that in successive segments the giant fibres are pierced
by corresponding sets of smaller fibres, but there is at least one set
which regularly traverses the giant fibre on passing out into the fourth
(IV) and fifth (V) segmental nerves. - This fibre will be described as
set B.

b. Fibres of Set A.

Along the inner margin of the lateral connectives there lies a set of
fibres (Plate 2, Fig. 18, Plate 5, Fig. 31) which in transverse section
are almost as large as the median giant fibre, and resemble it in their
resistance to stains. They differ from giant fibres, however, in the fol-
lowing particulars (compare Plate 4, Fig. 27, A):— (1) They are ar-
ranged segmentally, one pair of fibres originating in each segment ;

VOL. XXXII, — NO. 6. 2

104 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

(2) Each fibre is connected with a single cell; (3) They do not extend
through more than two segments; (4) They are not pierced by other
fibres, nor (5) do they show the reticulum found in giant fibres ; (6) They
are branched. The cell (Plate 4, Fig. 27) of which this fibre is a process
lies on the ventral side of the ganglion near the origin of the third seg-
mental nerve (III). The general direction of the process is forward, but
at the outset it crosses and recrosses the median plane, decussating twice
with its companion of the other side, one decussation being immediately
behind and the other in front of the origin of the second (IT) segmental
nerve. After the second crossing the two fibres run side by side close
beneath the median giant fibre, until they pass the first point of decus-
gation of a similar set of fibres in the next anterior segment. Here they
diverge and apparently break up into fibrillations or branches too small
to be traced in preparations stained in the ordinary way. I have not
succeeded in staining this fibre with methylen blue, This system is well
developed in every segment from the last one of the tail to within twenty
segments of the head, where the fibre gradually becomes smaller until,
in the first three or four segments, it cannot be distinguished among
the other fibres of the cord.

c. Fibres of Set B.

Next in size come the cells and fibres of set B (Plate 4, Figs. 27, B,
28). The cells lie ventrally about midway between the origin of the
first (I) and second (II) segmental nerves. From each cell a process
extends forward and gradually rises into the middle of the ganglion.
Opposite the origin of the fourth (IV) segmental nerve, the fibre turns
squarely across the ganglion, running parallel to its mate, with which
it decussates, and then divides into two branches, both of which go to
the periphery ; one through the fourth (IV), the other through the fifth
(V) segmental nerve. The two fibres of a pair lie in contact for some
distance where they cross from one side of the ganglion to the other
(Plate 1, Figs. 6, 7), and they anastomose at several points along the
line of contact (Plate 4, Fig. 28). The fibres of sets A and B are in-
timately associated at the point where they cross each other (Plate 1,
Fig. 6, Plate 3, Figs. 22, 23), for they are not only in contact, but the
smaller fibres lie in a deep indentation on the larger one. The relation
of fibre B to the lateral giant fibre is still more intimate. Immediately
after branching, one or both branches pass directly through the lateral
giant fibre before passing out of the ganglion (Plate 1, Fig. 2). Some-
times one branch may pass around the giant fibre, but still be in con-

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 105

tact with it, while the other branch passes directly through it. Some-
times the penetrating branch, instead of passing through the middle of
the giant fibre, goes so far to one side that it does not become free from
the sheath of the giant fibre, but is still wholly embedded in its
substance.

d. Fibres of Set O.

The next fibre system (Plate 4, Figs. 27, 30), set O, is apparently
centripetal, since no cell was found connected with it, and since what ap-
pears to be the main fibre, entering the cord from the fourth segmental
nerve (IV), immediately divides, forming the characteristic Y of centrip-
etal fibres. One of the branches runs directly back and ends in fibrillations
opposite the second nerve (II) of the succeeding segment. The other
branch runs forward, and ends in a similar way opposite the second seg-
mental nerve (II) of its own segment. Near its origin the second
branch gives off a third which runs diagonally backward and across the
ganglion, ending in a position symmetrical to the ending of the first
branch, Since the counterpart of each of these three branches is found
on the opposite side of the nerve cord, there must be six branches ending
in each segment, on either side three, all of which are opposite the sec-
ond segmental nerve (Fig. 27, II). The ends of the fibres are enlarged,
and give off a few fibrillations. The three endings of each side of the
body lie side by side, and are connected with one another by several
ladder-like anastomoses (Plate 4, Figs. 29, 30). The fibres of this set
are rather large, and lie almost wholly on the ventral side of the cord.
The third or decussating branches, however, are rather slender, and in
crossing the ganglion first curve up and then down. Where the two
fibres cross each other they are always in contact.

e. Peripheral Fibres.

The following are some of the fibres found in the parapodial ganglion
(compare Plate 1, Figs. 5, 8, Plate 5, Fig. 39): (a) Fibres entering the
ganglion from the second (II, Figs, 5, 39) segmental nerve pass through
the ganglion and out either by the first (1) or by the fourth (4) para-
podial nerves. (5) Fibres entering from the segmental nerve divide
into two branches, one of which passes out through the second (2), the
other through the third (3) parapodial nerve. Neither of these classes
of fibres gives off fibrillations in the ganglion. (c) A third kind of fibre
enters the parapodial ganglion from the segmental nerve, and ends in
the ganglion in fibrillations.

106 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The second (2) and third (3) parapodial nerves contain both motor
and sensory fibres. In Figure 32 the motor fibres are shown, and in
Figure 33 the sensory fibres of the third parapodial nerve. The motor
fibres turn back along the muscles that move the sete, and are lost
among the muscle fibres. The cells of the sensory fibres lie far beneath
the hypodermis. They send a process either to the hypodermis, or to
the tissue around the openings through which the sete project. At the
latter place the fibres apparently end in fibrillations. Figure 37 repre-
sents a sensory cell of the anterior wall of the parapocium. The periph-
eral process of this cell enlarges just beneath the cuticula into a small
knob, from which a fine prolongation extends out through the cuticula
Figure 38 represents a similar cell and nervous process in the posterior
wall of the parapodium. In Figure 35 is seen a sensory cell from the
base of the parapodium, and in Figure 36 one from the side of the body
near the fourth segmental nerve.

Figure 34 shows the manner in which the motor fibres end in the
longitudinal muscles, and Figure 40 shows'the bushy endings of the
fibres around the glands of the hypodermis between the bundles of
circular muscles.

PART II. DISCUSSION.

1. TOPOGRAPHY.

In methylen blue preparations it is usually not easy to determine the
relation of the stained fibres to other organs, because of the difficulty of
seeing structures which are not stained. For this reason I first made
a study of the topography of the nervous system, tracing the nerves
with considerable detail in preparations made by vom Rath’s method.
By this means nerves consisting of but a few fibres can be traced through
serial sections. The account of the topography given in Part I. is
more minute, but otherwise agrees in the main with that given by
Quatrefages ('50) for Nereis. There is one important point, however,
in which I cannot agree with Quatrefages. He states that the segmen-
tal nerve which he designates by the letter o (Planche 3) passes forward
through the dissepiment to the preceding segment, thus making a ner-
vous connection between two segments, in addition to that of the ventral
nerve cord, From the diagram (Plate 1, Fig. 8) it will be seen that there
is no segmental nerve passing from one segment to another in N.
virens. The three nerves (I, IV, V) that arise near the intersegmental
plane pass out parallel with that plane, two anterior to it and one

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 107

posterior. The segmentation of the longitudinal muscles is marked by
an interdigitation of the fibres of ono segment with those of the next.
Those interdigitations lie in the plane of the constriction of the body
which gi.es the external appearance of segmentation. The line of
attachment of the longitudinal muscles to the hypodermis (Plate 1,
Fig. 4) and the constriction in the protective tissue of the segmental
ganglion (Plate 1, Fig. 8) also lie in this plane, which, as will be seen
from Figures 4 and 8, thus separates the fifth (V’) and first (1) seg-
mental nerves throughout their length. The segmental dissepiment is
concave anteriorly. Its ventral median edge is attached in the con-
striction of the segmental ganglion, and is therefore in the intersegmental
plane. But its lateral border is attached to the hypodermis, between
the dorsal and ventral longitudinal muscles, anterior to the interseg-
mental plane and even anterior to the position of the fourth (IV’) seg-
mental nerve in that region (Fig. 4). Hence, if the position of the
dissepiment were taken to determine the boundary of segments, the
fourth (IV’) and fifth (V”) segmental nerves would appear to pass
backward from the segment in which they arise to the one succeeding it.
But I have found no segmental nerve passing forward through the
dissepiment as described by Quatrefages, nor indeed passing out of the
segment in either direction, if we determine the boundary of segments
by the segmentation of the musculature,

When compared with other annelids, we find that Nereis presents a
generalized condition with respect to its nervous system. It indeed agrees
very well with the description given by Lang (88-94) of the nervous
system typical of Chætopods. In comparison with other Polychætes,
however, Nereis shows a rather high degree of development, indicated
by the deep position and elaborate protective tissue of the vent ral nerve
cord. In the majority of Polychætes the ventral nervous system lies
embedded in the hypodermis, or intimately connected with it. Ina
few genera, however, such is not the case. Wawrzik (92) shows that in
Hermione and Aphrodite the ventral cord is entirely free from the hy-
podermis, and in this respect he classes these genera with the Oligo-
chwtes. Nereis would also belong to this class, since the ventral cord
lies internal to the circular muscles, as it does in the Oligochætes.

9, PROTEOTIVE TISSUE.

The nature and origin of the protective envelopes of the nervous sys-
tem of Polychætes have been the subject of considerable discussion.
The differences of opinion are probably due chiefly to real differences in

108 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

the animals studied. There is not much doubt, however, concern-
ing the origin of the inner spongy layer, the neuroglia. Jourdan (84)
showed that the enveloping tissue of the central nervous system was
intimately connected with the cells of the hypodermis. Rohde (87)
called this tissue “Subcuticularfasergewebe,” and described it as a
development of the basal processes of the cells of the hypodermis.
Wawrzik (92) made a comparison of a large number of Polychætes, and
found that in all those in which the ventral cord was connected with
the hypodermis the neuroglia was an integral part of the hypodermis
cells. Haller (89) denies the existence of the condition described by
Rohde ('87) for Polynoé, since he found that the nerve cord was sur-
rounded by a membrane which separated the neuroglia from the hypo-
dermis. However it may be in this case, there certainly cannot be a
connection between the hypodermis and neuroglia in such forms as
Hermione, Aphrodite, and Nereis, in which these structures are clearly
separated. But the condition found in so many other genera indicates
that the neuroglia is derived from the ectoderm along with the nervous
elements.

The neurilemma is apparently found only in those forms in which
the nerve cord is free from the hypodermis. But even when present it
may be so thin as to be readily overlooked. Such is sometimes the case
at the posterior end of Nereis. On the other hand, it becomes very
thick around the brain of Nereis, sometimes reaching a thickness of
fully 100%. Friedländer ('88) and Graber (80) call this structure
cuticular. Haller considers it simply the matted fibres of the neuroglia.
Racowitza (’96) states that muscle fibres, as well as the neuroglia, con-
tribute to make up the neurilemma. Where muscle fibres are attached
to the outer surface of the neurilemma, or neuroglia fibres to its inner
surface, membrane and fibre shade insensibly into each other, so as to
suggest their structural identity. But, as has been shown above, as
well as by other writers, the neurilemma in its reaction to stains is
‘very different from either muscle or neuroglia. Whatever may be the
weight of this evidence, it is clear that the neurilemma, the connective
tissue of the muscles, and the tunica intima of the ventral longitudinal
blood-vessel have the same structure, and must be derived from the
same source. That source is most likely the mesoderm.

3. BRAIN.

Although the brain of Nereis gives rise to so many nerves, it is
small and simple when compared with the brain, for example, of the

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 109

decapod Crustacea. In the latter, the fibrous part is relatively much
greater, and the fibres are collected into small bundles forming numerous
commissures between the various parts of the brain. Since the number
and size of the nerves leaving the brain of decapod Crustacea is small
compared with the size of the brain, the increase in the fibrous sub-
stance of the brain must be due to a greater development of the associa-
tion fibres of all kinds, including not only fibres which lie wholly within
the brain, but also those branches of centripetal and centrifugal fibres
which bring the various parts of the brain into relation with one another.
This condition is apparently correlated with the increased development
of the “ mushroom bodies ” in Arthropods, as we shall see below.

4. “Musuroom Boprzs.”

The compact masses of small nuclei that lie in the anterior part of
the brain of Nereis (Plate 3, Figs. 24, 21) have been described by a
number of writers, who have, however, usually expressed considerable
doubt concerning their significance.. Ehlers (68) and Schröder (86)
describe this structure under the name “ Nervenkórner.” Rohde (87)
calls a similar structure in Polynoë and other Polychætes “ Nerven-
kernen.” Retzius ('95) refers to it’ as a “Haufen groben Korner,”
which he says are larger about the periphery of the mass. He thinks
the larger granules may be cells, but doubts the cellular character of
the smaller ones. His preparations were stained in methylen blue, but
showed no processes connected with the nuclei. Haller (’89) discusses
the nature of these structures at some length, and describes the elements
as small multipolar ganglion cells. He calls the mass a “ Tentakel-
ganglion,” and supposes it to be connected with the sense organs of the
antenne. Racowitza (’96) applies to it a similar term, “ganglion anten-
naire,” but he does not mean to indicate thereby that the ganglion has
any direct connection with the antenna. Haller objects to Rohde’s ap-
plication of the descriptive term “* Hutpilz ” to these ganglia “ weil sie
sehr leicht zu einer Verwechselung mit den hutpilzförmigen Körpern
am Hirn der Insecten veranlassen dürfte, mit denen aber diese Ganglien
nichts Homologes aufweisen können.”

Notwithstanding this statement of Haller, I think there are good
reasons for considering this organ as in some degree homologous with
the mushroom bodies of the insect brain. The resemblance between the
two appears more strongly, if we compare both with a corresponding
structure in the brain of the crayfish. On the anterior lateral border
of the brain of this Crustacean there is a triangular mass of small cells

110 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

which Krieger (79) designates as gz, In my own preparations of the
brain of the crayfish I find that this ganglion resembles the “ ganglion
antennaire” of Annelids in the following points. In both, (1) such
ganglia are confined to the brain, no similar structure occurring in the
ventral cord. (2) The ganglion is intimately associated with the masses
of neuropil, which also occur nowhere but in the brain. (3) The small
size of the nuclei and the meagre cytoplasm distinguish these cells
from the other cells of the brain. (4) There is a peculiar arrangement
of the cells in rows radiating from the neuropil. According to the de-
scription given by Kenyon (’96), the mushroom bodies of the honey bee
exhibit the same peculiarities. The chief difference to be found in the
three cases is the relative size of the nuclear and the neuropil masses,
and in the arrangement of the two parts. In Nereis the nuclear mass
partially surrounds the neuropil, whereas in the insect the relation of
the two parts is reversed, the neuropil partly enveloping the nuclear
mass, The crayfish presents an intermediate condition in this respect,
The nuclear elements do not stain readily in methylen blue, —a condi-
tion also found by Allen and Bethe in Crustacea, and by Retzius in
Nereis ; but in the bee Kenyon obtained impregnations of the cells by
the Golgi method. His preparations show that the cells of theso ganglia
send processes into the neuroglia, where they end in dendrites almost as
complex as those found in the brain of Vertebrates. Since in the worm
there is relatively little neuropil, the dendrites of the associated cells
will probably be found to be less well developed. Kenyon’s supposition
that the intelligence of the insect is to be accounted for by the com-
plexity of the relations between the nervous elements made possible by
these association fibres seems quite plausible; and if we apply the
same argument to the worm, we may suppose its low intelligence to be
in part correlated with the small amonnt of neuropil, or, in other words,
the limited development of the association fibres.

Aside from the cells of this ganglion and those connected with the
ciliated groove, the brain of Nereis contains about as many cells as a
typical ganglion of the ventral chain. If we compare the brain with
the ganglia of the ventral chain, or if we compare the central nervous
system of Annelids with that of Arthropods, the only structural condi-
tion to be found which warrants the supposition that it is correlated
with the supposed psychic functions of the brain is the mushroom body
and the related development of association fibres. This correlation has
often been pointed out for insects, and I think we may extend the ob-
servation to decapod Crustacea and Annelids.

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 111

Racowitza shows that those Polychætes which lack antennæ also lack
the “ ganglion antennaire.” He does not prove, however, that the cells
of this ganglion may not be present in the brain, and therefore does not
exclude the possibility that the ganglion may be present in a diffuse
form.

5. Optic GANGLION.

The condition of the optic ganglion in Noreis virens is of interest,
because it serves to explain what have hitherto appeared to be unac-
countable differences between several species of Nereis. Carrière (’85,
pp: 33-35) described this ganglion for N. cultrifera, and Retzius (95)
found it in N. diversicolor. On the other hand, Carriére says there is
no such ganglion in a species from Norderney which he examined, and
Graber (’80) and Haller (’89) also failed to find it in Nereis coste. It
seemed strange that a central ganglion, like this, should exhibit such
will-of-the-wisp peculiarities in passing from one species to another so
closely related to it. I think, however, that the condition of this gan-
glion in N. virens shows clearly what becomes of the ganglion when it
disappears from its place beneath the anterior eye, as in N. coste. In
N. virens the ganglion evidently lies partly beneath the eye and partly
within the brain capsule. A few scattering cells show the path the gan-
glion has taken in its migration inward or outward. It is not only the
great similarity in the appearance of the cells and the contiguity of the
two parts that makes this view seem probable, but also the cells of both
groups send their processes to the commissural ganglion and neither part
appears to be directly connected with the brain. It is not apparent what
is the relation of the ganglion to the anterior eye. Carrière thought
the ganglion formed part of the connection between the eye and the
brain, but this cannot be, for later writers agree that the anterior eye
as well as the posterior is innervated directly from the brain.

The posterior end of the brain deserves more careful study than I
have as yet been able to give it; I shall therefore simply call attention
to a few facts. Five of the six kinds of cells described for the brain are
to be found in the posterior part, and of these five three are not found
elsewhere. Moreover these three are the most peculiar ones, — those of
the second, third, and fourth classes. This portion of the brain is partly
separated from the remainder of it, and is intimately connected with
the surface at the ciliated grooves and at the dorsal sensory regions
through the thirteenth (XIII) and fourteenth (XIV) nerves. Perhaps

112 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the whole is to be considered a complex sensory organ, analogous to the
olfactory organ of Vertebrates in its intimate relation with the brain.
Retzius shows that the sensory fibres of the ciliated groove are processes
from bipolar cells of this region. The fibres of the fourteenth pair of
nerves are the processes of cells similar in form and position to the
bipolar cells of the thirteenth nerve.

6. VENTRAL Nerve Corp.

The structure of the ventral nerve cord has been well described for
Lumbricus by Friedländer (94), and Hatschek (89-91) has given a
good figure of a transverse section of the ventral cord of Sigalion. Most
writers, however, have not succeeded in preparing the ventral cord so as
to show clearly that the connectives consist wholly of longitudinal fibres.
There is nowhere in the ventral cord a neuropil in the sense of that
which is found in the brain. There are small masses of fibrillations in
the ganglia, of course, but they simply fill up the interstices between the
fibres, and never occur in masses large enough to produce the punctate
appearance peculiar to the neuropil of the brain.

The paucity of nuclei among the fibres of the cord will not permit one
to regard the fibre sheaths as composed of the expansions of non-nervous
cells. In the decapod Crustacea the fibre sheaths are nucleate, and in
the case of the sheath of giant fibres the nuclei are go numerous that
the sheath may be described as a flat endothelium. In Nereis, however,
the sheath must be a product of the fibre itself.

7. CENTROSOMES.

Since Lenhossék (’95) announced the discovery of the centrosome in
the adult nerve cells of the frog, there have appeared a number of papers
describing similar structures in Reptiles (Buehler, 95), Cyclostomes
(Schaffer, ’96), Molluscs (McClure, ’96), and Worms (Lewis, 796). Hei-
denhain (97) summarizes the evidence and gives a bibliography. Dahl-
gren (’97) describes what he calls a centrosome artifact in the spinal
ganglia of the dog. This artifact, he says, is produced by the formation
of a erystal of corrosive sublimate in the cell. In Nereis I find the
best demonstrations of centrosomes in preparations that have been fixed
in corrosive sublimate, but they also occur in preparations fixed in the
osmic acid mixture of vom Rath. I think there is no reason for con-
sidering the phenomenon an artifact in this case. I will simply call
attention in this connection to two facts that were mentioned previously ;

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 113

first, the general occurrence of the centrosome in the cells of the ventral
ganglia, and, secondly, the large number of centrosomes that may occur
in a single cell. I have no explanation to offer for the latter condition.
Since the structures appear only under special conditions of staining,
and since I had only one preparation of the brain stained in iron-hæma-
toxylin, I am not in a position to say whether the centrosome occurs in
the brain or not, even though I failed to find it in the preparations
I had at hand.

8. Nerve FIBRES.

a. Giant Fibres.

The literature concerning giant fibres is voluminous, and extensive
bibliographies on the subject may be found in the works of Eisig (87)
and Friedländer (88, 794). I shall concern myself here with only a few
of the many points in which those fibres have given rise to discussion.
It has been frequently demonstrated that they are the processes of cells,
and they have been taken by many writers to be nervous in function,
but some authors still doubt that that is their nature ; Lenhossék (92),
for example, has recently expressed the conviction that they are not.
The most serious objection that has been urged against their nervous
nature is the absence of evidence that they are related to other ner-
vous structures, either by fibrillations within the cord or by centrifugal
branches.

I think there is sufficient reason for maintaining that in Nereis virens
the fibres of set B serve as branches for the lateral giant fibres. I
therefore believe that the function of the latter is to transmit nervous
impulses like ordinary nerve fibres.

Tho most peculiar feature of giant fibres is that they are often con-
nected with more than one cell. In 1881, Spengel (81) arrived at the
conclusion that in Halla there was a fusion of giant fibres, but he bad
no direct evidence. Rohde (’87), however, shows conclusively that at
least one giant fibre in the ventral cord of Sthenelais is formed by the
union of the processes of two cells. These lio in the brain and send
their processes through the circum-wsophageal connectives to the sub-
esophageal ganglion, where they fuse and whence they continue as a
single fibre throughout the entire length of the animal. Friedländer
(88) found that the lateral giant fibres of the earthworm are connected
with a number of cells in the posterior segments of the animal. This
discovery was confirmed by Cerfontaine (92), who also found that the
median fibro is connected with several cells at the anterior end of the

114 BULLETIN: MUSEUM OF VOMPARATIVE ZOOLOGY.

body. In Rhynchelmis, too, according to Vejdovsky (88, '92), the giant
fibres are connected with a number of cells and in such a way that
each might well be considered a bundle of fibres. Finally, Lewis (96)
describes in a Moldanid a giant fibre which is connected with a large
number of cells. There is not yet sufficient evidence to show whether
the giant fibres of Chætopods are more frequently multicellular or uni-
cellular, but there can be no doubt that they are often multicellular.

The giant fibres of Crustacea have not been so well investigated as
those of Chætopoda, but in Homarus, at least, each giant fibre is the
process of a single large cell, according to the description of Allen (94).

Our present knowledge of the giant fibres (in the sense in which I use

the term) might be summarized in the following way. The giant fibres
of Annelids and Crustacea are much larger than ordinary fibres, and ex-
‘tend for long distances through the central nerve cord ; they are con-
nected either with one very large cell or with the processes of several
cells, and they give off neither fibrillations nor branches. In some cases,
as in Lumbricus, there are anastomosing bars, or connections, between
two giant fibres; in others, the giant fibres may divide or they may fuse
with one another, but in no case is there an ending corresponding to the
fibrillations of other nerve fibres by which the giant fibres might be put
in connection with other nervous structures. In Nereis, however, there
is a very intimate connection between the lateral giant fibres and the
centrifugal branches of set B, as I have shown, and by this system of
connections the giant fibres are put in relation with every segment of the
body.

What the function of such giant fibres may be is readily conceivable,
and I believe the true explanation has already been offered by several
writers. Vignal (’83) suggested that their purpose was to bring about
a more direct connection of the nervous system as a whole than is done
by less extensive fibres. Friedländer’s experiments on the earthworm
show that, when the ventral cord is severed, the sudden longitudinal
contraction of the body can no longer be brought about. Friedländer
argues that, since these fibres are the only ones, so far as we know, that
pass through the entire length of the animal, it is reasonable to suppose
they are the ones that conduct the stimulus for this contraction,

In Nereis I have frequently noted a sudden longitudinal contraction
where there was apparently no stimulus except the passing of a shadow.
I have not yet had the opportunity to test this further, to determine if
the stimulus proceeded from the eyes, but I found that no tactile stim-

ulus was sufficient to produce such a sudden and general longitudinal

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 115

contraction. When the habits of the animal are considered, it is pos-
sible to understand what the function of such a contraction brought
about by the stimulus of light might be. The worm lives in the mud
in burrows, and frequently rests with the anterior end above the surface,
while the remainder of the body is in fhe burrow. Under such circum-
stances the longitudinal contraction would cause the animal to retreat
into the burrow, for longitudinal contractions are in general accompanied
by the pointing of all the parapodia towards that end of the body from
which tho stimulus comes. For example, if the stimulus is applied at
the anterior end, the parapodia are all thrown forward, and the longitu-
dinal contraction of the body immediately follows. This will cause the
anterior end to move towards the tail while the latter remains station-
ary, since the position of the parapodia prevents movement of the body
in the opposite direction. Now, if the shadow cast by a predatory
animal were to bring about this movement, the mechanism would be of
vital importance to the worm. Perhaps the importance of the function
and the great extent of the movement brought about help to account
for the large development of the giant fibres. The objection may be
urged that since the phenomena which I have described for Nereis have
not been found elsewhere, they cannot be of general importance, even if
the condition be admitted for Nereis. But the exceptional conditions
under which such phenomena can be observed render it probable that
they may have been overlooked even when present.

It must be remembered that, in order to demonstrate the passage of
one fibre through another, there must be a differential staining of the
substance of the two fibres. Only in preparations fixed and stained by
the method of vom Rath, and not in all of these, have I obtained such a
differentiation. Successful preparations, however, leave no doubt con-
cerning the actual relation of the fibres, for I have carefully compared
series of sections cut in each of the three cardinal planes, and always
with the same result,

If, then, the giant fibres are nervous in function, the neuron theory
of Waldeyer (91) will require considerable modification. The nervous
element is not always unicellular, but may consist of a number of cells
united in function. The nervous connection between fibres is not always
through fibrillations; it may be directly between the axis cylinders
themselves,

b. Fibres of Set A.

Since little is known about the relations of the fibres of set A to other
fibres, we cannot say much about their probable function, Nevertheless,

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

116

there are several facts which point to a connection with the forward
locomotion of the animal. The worm advances by a rhythmical move-
ment of the parapodia, which begins at the posterior end and passes
toward the head. With this movement there is usually associated a
serpentine motion of the body, which also passes from behind forward,
Both movements are less vigorous near the head, and the serpentine dis-
appears entirely between the twentieth and tenth segments. The size
of the fibres of set A in a given region corresponds to the degree of
activity of the locomotor movements of that region. Whether this
fact is more than a mere coincidence I cannot say, but it would seem
to be so. Besides, if there is a causal relation between the condition of
these fibres and the locomotor movements, we may even account for the
enormous size of the fibres on the ground of their functional importance.
Another evidence of this correlation is the serial arrangement of the
fibres, which may be connected with the progressive character of the
motor excitation, and with the postero-anterior disposition of each fibre,
the latter corresponding to the direction of the movement.

Although these speculations concerning the function of giant fibres
are purely tentative, they may serve as a basis for physiological
experiments.

c. Fibres of Set B.

In describing the fibres of set B (Plate 4, Figs. 27, 28) I merely
mentioned the fact of an anastomosis between the axis cylinders of the
components of each pair. I wish here to discuss the subject more fully.
The description of these fibres was by no means based wholly on meth-
ylen-blue preparations. Indeed, all the facts, excepting that of anasto-
mosis, were demonstrated on serial sections before an impregnation by
methylen-blue was obtained. The fibres are so large that they can
easily be traced through serial sections. This fact is important in con-
sidering the value of the evidence for anastomosis.

I have carefully examined seventeen pairs of these in serial sections
cut in one or the other of the three cardinal planes of the body, and in
addition eight pairs stained in methylen-blue and examined before cut-
ting. Where the fibres of a pair crossed the ganglion they were always in
contact with each other, and, with one exception, they ran parallel for a
considerable distance, In the exceptional case the fibres crossed each
other at an angle of about ten degrees, which still allowed a line of con-
tact equal in length to one fourth the width of the ganglion. The fibres
usually cross the ganglion at right angles to its longitudinal axis, but
in one instance they crossed at an angle of about sixty degrees (Plate 1,

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 117

Fig. 7). Thus, one of the fibres partially retraces its course in order to
maintain a course parallel with its fellow. Sagittal sections (Plate 3,
Figs. 22, 23) show that the fibres are always flattened on their apposed
faces. That part of the sheaths which forms the dividing wall is usually
very thin, and in some cases seems to be wholly wanting. In the prepa-
rations which are best preserved, however, the dividing wall can always
be seen. I have not been able to demonstrate satisfactorily anastomoses
in preparations made by the more usual histological methods. In
methylen-blue preparations the fibres do not appear to be in contact, but
this is due to a shrinking of the axis cylinder within the sheath produced
during the fixing of the stain. The anastomoses, however, do exist, and
are clearly shown in methylen-blue preparations (Plate 4, Fig. 28) ; they
proceed from small elevations on the opposed faces of the fibres. From
what has gone before, it is evident that the anastomosing bars simply
pierce the thin membrane that separates the two fibres, and that they
practically lie wholly within the fibre sheaths. Hence they cannot be
rogarded as fibrillations fused by the action of the methylen-blue. The
fibrillæ of the axis cylinder pass out into the anastomosing bars, but
whether they pass completely across from one fibre to the other I can-
not say. There is, however, a distinct interdigitation of the fibrille of
the opposite fibres. The appearance of the preparations gives one the
impression that there has been a breaking of the fibrillæ of the anasto-
moses due to the shrinking of the fibres, The anastomoses are not
always as evident as they are in the case reproduced in Figure 28, but
there is always some indication of them. This may‘ consist simply of the
pointed elevations arranged in pairs opposite each other on the fibres.

Since the cells of set B are situated in a central organ, they are prob-
ably motor, and since the fibres are united in bilaterally symmetrical
pairs, they probably act in concert. Such animals as Annelids differ
from more complex organisms in that many of their movements are in
unison on the two sides of the body. The longitudinal contractions and
expansions of the body are examples. In Nereis the movement of all the
parapodia backward or forward, when the animal is touched at one end
or the other, is another instance. When such movements are 80 fre-
quent and of such vital importance, one may well expect to find an inti-
mate association of the related nerve fibres.

Allen (96) describes decussating nerve elements in the abdominal
ganglia of the lobster so closely united that he was unable to resolve
them into their constituent parts. He finds, however, that similar ele-
ments in the thorax are not so intimately related. At another place he

118 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

makes the statement that these elements of the abdomen innervate the
abdominal muscles, while those of the thorax go to the ambulatory ap-
pendages. The reason for the difference in the arrangement of the nerve
elements will be immediately perceived. The muscles of the abdomen
act bilaterally in unison, hence the union of the associated nerve elements.
In the case of the ambulatory appendages there is little movement in
unison, hence the corresponding independence of the fibres concerned.

d. Fibres of Set ©,

Lenhossék (’92) makes the general statement relative to the sensory
fibres of the earthworm, that they do not cross the ventral nerve cord,
but end in fibrillations on the side from which they enter the cord.
The fibre © of Nereis is an interesting exception to this rule. Concern-
ing the anastomoses of this system I need say but little. The fibres are
80 large, the anastomoses so numerous and distinct, and the fibre in such
excellent condition for study, that there is small chance for error. There
is no vacuolation of the fibre nor other evidence to lead one to conclude
that there has been a fusion of fibrillations in the manner suggested by
Cajal (’96). I have seen no evidence of anastomosis between fibres
except those of set Band set O, and here the anastomosis is always
between fibres of the same set.

I wish to call attention to one more point relative to these fibres.
The small decussating branches cross the ganglion by a sinuous course,
and yet where they cross each other they are invariably in contact.
Why this should be so is difficult to say, unless the function of the fibres
necessitates such contact. A similar relationship is also to be found be-
tween fibres of other sets, as in the case of the fibres of sets A and B, as
described above. Although physiologists do not recognize contact be-
tween axis cylinders as a means of bringing fibres into functional re-
lation, it seems to me quite probable that such a relation exists in
some Cases,

SUMMARY.

1. The central nervous system of Nereis virens occupies a deeper
position than does that of most Polycheetes. It is separated from the
hypodermis by the circular muscles, and is enveloped by an elaborate
protective tissue.

2. The protective tissue consists of two parts ; an inner spongy layer,
the neuroglia, of ectodermic origin, and an outer sheath, the neurilemma,
of mesodermic origin,

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 119

3. The “mushroom bodies” of insects and decapod Crustacea are
represented in the brain of Nereis by the anterior masses of small
nuclei.

4. The optic ganglion, which in some species of Nereis lies beneath
the anterior eye, may in other species lie within the brain capsule.

5. There is no neuropil in the ventral nerve cord.

6. There are three longitudinal connectives between each two succes-
sive ganglia of the ventral nerve cord, one small median and two larger
lateral ones.

7. The sheaths of the nerve fibres of the ventral cord have no nuclei,
and hence must be a product of the fibres themselves.

8. The nerve cells of the ventral cord commonly have one or more
centrosomes.

9, The giant fibres are nervous in function, and are put into relation
with peripheral organs through ordinary centrifugal fibres.

10. The giant fibres give off no fibrillations, and nervous relation with
other fibres is established directly between the axis cylinders.

11. Certain decussating fibres are always united in pairs by anasto-
moses between the axis cylinders where they cross each other.

12. Certain centripetal fibres of the same set are always united by
anastomoses between the ends of the branches.

13. Contact between axis cylinders may possibly be one of the means

of bringing nerve fibres into functional relation with each other.

In conclusion, I wish to acknowledge my indebtedness to Professor
E. L. Mark for kindly advice and assistance rendered me in many ways
while pursuing my studies in the Zöological Laboratory of Harvard Uni-
versity. I gladly avail myself of this opportunity to express to him my

sincere thanks.

o

VOL. XXXII. — NO. 6, 3

120 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

BIBLIOGRAPHY.

Allen, E. J.
94, Studies on the Nervous System of Crustacea. Quart. Jour. Mier. Sci.,

Vol. 36, N. S., pp. 461-498, Pls. 35-38.

Allen, E. J.
96. Studies on the Nervous System of Crustacea. Quart. Jour. Mier. Sci.,

Vol. 39, N. S., pp. 33-50, Pl. 4.
Bethe, A.
195. Studien über das Centralnervensystem von Carcinus Maenas, nebst
Angaben über ein neues Verfahren der Methylenblaufixation. Arch. f.
mikr. Anat., Bd. 44, Heft 4, pp. 579-622, Taf. 34-36.
Buehler, A.

95, Protoplasma-Structur in Vorderhirnzellen der Eidechse. Verhandl. phys.
med. Gesellsch. zu Würzburg, N. F., Bd. 29, No. 6, pp. 209-252,
Taf. 3-5,

Also separate, 44 pp., 3 Taf., 1895,
Cajal, Ramon y.

196. Nouvelles contributions à ’étude histologique de la rétine et à la ques-
tion des anastomoses des prolongements protoplasmiques. Jour, Anat, et
Physiol., Tome 32, No. 5, pp. 481-543, Pls. 12-15.

Carriére, J.

"85. Die Sehorgane der Thiere vergleichend-anatomisch dargestellt. München

und Leipzig. vi-+ 205 pp», 147 Abbildg.
Cerfontaine, P.

92. Contribution à l'étude du système nerveux central du Lombric terrestre.

Bull. Acad. roy. Belgique, Série 3, Tome 23, pp. 742-752, Pls. 1, 2.
Dahlgren, U.

97. A Centrosome Artifact in the Spinal Ganglion of the Frog. Anat. An-

zeiger, Bd. 13, No. 4, 5, pp. 149-151, 2 Figs.
Ehlers, E.

64-68. Die Borstenwürmer (Annelida chætopoda) nach systematischen und

anatomischen Untersuchungen dargestellt. Leipzig, Engelmann, xx + 748

pp», 24 Taf.

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 121

Eisig, H.

"87. Monographie der Capitelliden des Golfes von Neapel und der an-
grenzenden Meeres-Abschnitte nebst Untersuchungen zur vergleichenden
Anatomie und Physiologie. Fauna u. Flora des Golfes von Neapel.
Monographie XVI., Berlin, xxvi + 906 pp., 37 Taf.

Friedlánder, B.

"88. Beiträge zur Kenntnis des Centralnervensystems von Lumbricus. Zeit.

f. wiss. Zool., Bd. 47, Heft 1, pp. 47-84, Taf. 9, 10.
Friedlander, B.

89. Ueber die markhaltigen Nervenfasern und Neurochorde der Crustaceen
und Anneliden. Mitth. Zool. Stat. Neapel, Bd. 9, Heft 2, pp. 205-265,
Taf. 8.

Friedlander, B.

"94, Altes und Neues zur Histologie des Bauchstranges des Regenwurms

Zeit. f. wiss. Zool., Bd. 58, Heft 4, pp. 661-693, Taf. 40.
Friedlinder, B.

95. Ueber die Regeneration herausgeschnittener Theile des Centralnerven-
systems von Regenwürmern. Zeit. f. wiss. Zool., Bd. 60, Heft 2, pp. 249-
283, Taf. 13, 14.

Graber, V.

’79. Morphologische Untersuchungen über die Augen der freilebenden mari-
nen Borstenwürmer, Arch. f. mikr. Anat., Bd. 17, pp. 243-323, Taf.
18-20.

Haller, B.

’89. Beiträge zur Kenntniss der Textur des Central-Nervensystems höherer
Würmer. Arbeit. Zool. Inst. Wien, Tome 8, Heft 2, pp. 175-312, Taf.
16-20.

Hatschek, B.

"89-91. Lehrbuch der Zoologie, u. s. w. Lieferungen 1-3. Jena, G. Fischer.

432 pp. [unfinished].
Heidenhain, M.

’97. Ueber die Mikrocentren mehrkerniger Riesenzellen, sowie über die Cen-
tralkörperfrage im Allgemeinen. Morphol. Arbeiten [Schwalbe], Bd. 7,
Heft 1, pp. 225-280,

Jourdan, E.

’84. Le cerveau de PEunice Harassii ct ces rapports avec l’hypoderme,

Comptes Rendus Acad. Sci. Paris, Tome 98, pp. 1292-1294.
Kenyon, F. C.

"96. The Brain of the Bee. Jour. Comp. Neurology, Vol. 6, No. 3, pp.

133-210, Pls. 14-22.
Krieger, K. R.

’80. Ueber das Centralnervensystem des Flusskrebses. Zeit. f. wiss. Zool.,

Bd. 33, Heft 4, pp. 527-594, Taf. 31-33.

122 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

aa

Lang, A.
"88-94. Lehrbuch der vergleichenden Anatomie der wirbellosen Thiere.

Jena, G. Fischer, 1888-94, xvi + 1198 pp., 251 Abbildg.

Lenhossék, M. von.
292. Ursprung, Verlauf und Endigung der sensibeln Nervenfasern bei Lum-

bricus. Arch. f. mikr. Anat., Bd. 39, Heft 1, pp. 102-136, Taf. 5.

Lenhossék, M. von.
95, Centrosom und Sphäre in den Spinalganglienzellen des Frosches. Sitzb.

phys.-med. Gese sch. Würzburg, Jahr. 1895, No. 5-7, pp. 79-105.
Lenhossék, M. von.
95%, Centrosom und Sphäre in den Spinalganglienzellen des Frosches. Arch-
f. mikr Anat., Bd. 46, Heft 2, pp. 345-369, Taf. 16, 17.
Lewis, M.
96, Centrosome and Sphere in Certain of the Nerve Cells of an Invertebrate.
Anat. Anzeiger, Bd. 12, No. 12, 18, pp. 291-299, 11 Figs. Sept. 2.
McClure, C. F. W.
°96, On the Presence of Centrosomes and Attraction Spheres in the Ganglion
Cells of Helix Pomatia, with Remarks upon the Structure of the Cell Body.
Princeton Coll. Bull., Vol. 8, No. 2, pp. 38-41.
Quatrefages, A i de.
50. Études sur les types inféricures de Pembranchement des Annelés. Ann.
Sci. Nat., Série 3, Zool., Tome 14, pp. 281-289.
Racowitza, E. G.
795. Sur le rôle des Amibocytes chez les Annélides polychètes. Comptes
Rendus Acad. Sci. Paris, Tome 120, No. 8, pp. 464-467.

Racowitza, E. G.
96. Le lobe céphalique et l’encéphale des annélides polychètes. (Anatomie,

morphologie, weg Arch. Zool. Exp., Série 3, Tome 4, No. 1, 2,
pp. 133-343, Pls.

Rath, O. vom.

195. Zur Conservirungstechnik. Anat. Anzeiger, Bd. 11, No. 9, pp. 280-
288.
Retzius, G.

91. Zur Kenntniss des centralen Nervensystems der Würmer. Biol. Unters.,
N. F., Bd. 2, pp- 1-28, Taf. 1-10.
Retzius, G.
295. Zur Kenntniss des Gchirnganglions und des sensiblen Nervensystems
der Polychäten. Biol. Unters., N. F., Bd. 7, pp. 6-11, Taf. 2, 3.
Rohde, E.
187. Mistologische Untersuchungen über das Nervensystem der Polychäten.

Zool. Beiträge [Schneider], Bd. 2, Heft 1, pp. 1-81, Taf, 1-7.

HAMAKER: NERVOUS SYSTEM OF NEREIS VIRENS. 123

Sars, M.

35, Beskrivelser og iagttagelser over nogle merklige eller nye i Havet ved
den Bergenske Kyst lebende Dyr, af Polypernes, Acalephernes, Radia-
ternes, Annelidernes og Molluskernes Classer, etc. Bergen. 8l pp»
15 pls.

Schaffer, J.

96. Ueber einen neuen Befund von Centrosomen in Ganglien- und Knorpel-
zellen. Sitzb. Akad. Wissensch. Wien, math.-naturw. Cl., Bd. 105, Heft
2, Abth. 3, pp. 21-28, 1 Taf.

Schréder, G.

"86. Anatomisch-histologische Untersuchung von Nereis diversicolor, O. Fr.

Mill. Inaug.-Dissertation. Rathenow. Carl Koppel. 43 pp», L Tal.
Spengel, J. W.

’81. Oligognathus Bonellim, eine schmarotzende Eunice. Mitth. Zool. Stat.

Neapel, Bd. 3, Heft 1 u. 2, pp. 15-52, Taf. 2-4.
Vejdovsky, F.

"88-92. Entwicklungsgoschichtliche Untersuchungen. Prag, iv + 401 pp»,

u. Atlas mit 32 Taf.
Vignal, W.

’83. Recherches histologiques sur les centres nerveux de quelques invertébrés.

Arch. Zool. Exp., Série 2, Tome 1, No. 2, pp. 267-412, Pls. 15-18.
Vom Rath, O. See Rath, O. vom.
Waldeyer, W.

"91. Ueber einige neuere Forschungen im Gebiete der Anat nie des Central-

nervensystems. Deutsch, med. Wochenschrift, Jahre. 1891, No. 44.
Also separate. G. Thieme, Leipzig, 1891, 64 pp», 10 Figs.
Wawrzik, E.

292. Ucher das Stützgewebe des Nervensystems der Chaetopoden.

träge [Schneider], Bd. 3, Heft 2, pp. 107-127, Taf. 14-19.

Zool. Bei-

124

a.
anastm.
RR
ceb.
cir.

cir. ta,
cl. etl,
el. sns.

con't, lg. l.

c so,
cta.
dřsep.
Jor.
Sor. mot.
foran.
gn. coms.
gn. pa pd.
gn. sb-e.

coms, CrC-@,

con't. lg. m.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

EXPLANATION OF PLATES.

ABBREVIATIONS.
Anterior. gn. sg.
Anastomosis. h’drm.
Axis cylinder. mu.
Brain. mu. ere,
Dirrus. mu. lg.
Tentacular cirrus. mu. ob,
Jiliated cell. n.
Sensory cell. ned,
Circum-cesophageal com- n’gli.
missure. nl.
Lateral longitudinal con- wlem.
nective. n’ pil.
Median longitudinal con- n. pa-coms,
nective. oc.
Centrosome. p
Cuticula. po. sq.
Dissepiment. rel,
Fibrillations, sq. ce.
Motor fibres. set.
Nerve fibres. sul. cil.
Commissural ganglion. tu. for.
Parapodial ganglion. Weds

Sub-osophageal ganglion,

va, SNJ.

All Figures, except 1, 5, 8, 27, and 39, were outlined with the camera lucida.
Figures 28-30 and 82-41 were drawn from methylen-blue preparations.

Segmental ganglion.
Hy podermis.

Muscle fibres.
Circular muscles.
Longitudinal muscles.
Oblique muscles.
Nerve.

Giant fibre.
Neuroglia.

Nucleus.
Neurilemma.
Neuropil.
Para-commissural nerve.
Eye.

Posterior.

Pore of segmental organ.
Reticulum.

Cephalic segment,
Seta.

Jiliated groove.

Fibre sheath.

Tunica intima.
Blood-vessel,

Fig. 1.

Fig. 2.

Hamaker, — Nervous Syst. Nereis.

PLATE 1,

Diagram showing the disposition of the herves of the brain and sub-
esophageal ganglion in dorsal aspect. In order to show the commissural
ganglion and its nerves, the right anterior eye has not been indicated;
I, V, a, B, y, nerves to the proboscis; II, nerve to the antenna; III, IV,
VII, 6, e, É n, nerves to the muscles and the general surface of the head;
0, commissure between the anterior and posterior cirrus ganglia; VI,
nerve to the palp; VIII, IX, X, nerves from the brain to the commis-
sural ganglion; XI, XI, optic nerves; XIII, nerve of the ciliated
groove; XIV, three openings in the dorsal surface of the brain capsule,
through which loose bundles of nerve fibres pass to the integument of
the mid-dorsal region of the cephalic lobe.

Para-sagittal section of a giant fibre to show the passage through it of a
fibre of set B (compare Fig. 27). In this case the branching of the fibre
B takes place within the giant fibre, and the axis cylinder of fibre B is
shrunken.

Cross section of a lateral giant fibre, to show the reticulum.

Frontal section of the body wall between two parapodia, to show the rela-
tive positions of nerves IV’, V’, and I’, and of the attachment of the
longitudinal muscles and the dissepiment.

Diagram of posterior aspect of part of a cross section, showing the dispo-
sition of the parapodial nerves. The second and third parapodial nerves
(compare Fig. 8) are designated by 2 and 8 respectively.

Frontal section of a segmental ganglion, showing the intimate relation
between the fibres of sets A and B (compare Fig. 27).

Section similar to that in Fig. 6 showing the relation of the decussating
parts of fibres B; also showing exceptional oblique course across the
ganglion.

Diagram showing the disposition of the segmental and parapodial nerves
of a typical segment. J, II, III, IV, V, the five segmental nerves num-
bered from in front backward. 1, 2, 3, 4, the four parapodial nerves.

HAMAKER. — Nervous Syst. Nereis,

PLATE 2.

Fig. 9. Transverse section through the posterior end of the brain.

Fig. 10. Pigment (?) from the posterior part of the brain.

Figs. 11 and 12. Nerve cells of the fourth and third classes of the brain respectively.

Fig. 18. A group of four ganglionic cells of a segmental ganglion, in frontal
section, to show centrosomes.

Fig. 14. A, ordinary ganglionic cell. B, one of the large cells of set B (compare
Fig. 27). The cytoplasm is not granular and takes little stain, ex-
cepting the large irregular granules around the centrosomes,

Fig. 15. Brain nerve cell of the second class.

Figs. 16 and 17. Brain nerve cells of the sixth class.

Fig. 18. Transverse section through the posterior end of a segmental ganglion
from the region of the fifteenth segment. It shows the connection
between the neurilemma and the tunica intima of the ventral blood-
vessel; also the position of the ventral nerve cord relative to the hypo-
dermis and the circular muscles.

Fig. 19. Brain nerve cell of the fifth class.

u,

tte,

Hamaker, — Nervous Syst. Nereis,

PLATE 3.

Fig. 20. Parasagittal section of the cephalic segment tangent to the lateral surface

of the posterior eye, to show the ciliated groove.

Fig. 21. Enlarged view of part of Figure 24, to show the arrangement of the
nuclei of the “mushroom body.”

Figs. 22 and 23. Para-sagittal sections of fibre A (compare Fig. 27), showing
relation to fibres B, and also contact of fibres B with each other.
The median giant fibre also appears in Fig. 23.

Fig. 24. Transverse section through the anterior part of the brain, showing the
“mushroom body.”

Fig. 25 is omitted.

Fig. 26. Frontal section of a longitudinal connective.

HAMAKER. — Nervous Syst. Nereis,

Fig. 27.

Fig. 28.
Fig. 29.

Fig. 30.

PLATE 4.

Diagram to illustrate the fibre systems of sets A, B, and C, in two succes-
sive ganglia, as projected on the frontal plane.

A pair of anastomosing fibres of set B.

Fibres of set C, showing anastomosis between a posterior and a decus-
sating branch.

Fibres of set C, showing anastomosis between an anterior and a posterior

branch.

a

Hamaker. — Nervous Syst. Nereis,

PLATE 5.

Transverse section of the longitudinal connectives of the ventral cord.

Frontal section of the ventral ramus of a parapodium (compare Fig. 8),
showing motor (?) elements in the posterior branch of the second
parapodial nerve.

A section similar to that in Figure 32, showing sensory elements in the
second parapodial ne, ve.

Motor fibres and endings in the longitudinal muscles.

Sensory cell from the base of a parapodium.

Sensory cell from the side of the body near the fourth segmental nerve.

Sensory nerve termination from the anterior wall of the parapodium.

Sensory cell from the posterior wall of the parapodium,

Diagram to show the course of fibres in the parapodial ganglion.

Fibres of the “sub-hypodermal plexus” ending among the glands of the
bypodermis.

A nerve fibre showing the spiral arrangement of the fibrillæ, and also the

shrinking of the axis cylinder from its sheath.

fo Z 4 aig
f

Bulletin of the Museum of Comparative Zoölogy
AT HARVARD COLLEGE,
Vou. XXXII. No. 7.

ON REMAINS OF STRUTHIOLITHUS CHERSONENSIS
FROM NORTHERN CHINA,

WITH REMARKS ON THE DISTRIBUTION OF
STRUTHIOUS BIRDS.

By ©. R. EASTMAN.

Wirn ONE PLATE.

CAMBRIDGE, MASS., U. S.A.:
PRINTED FOR THE MUSEUM.

Avaust, 1898.

No. T. — On Remains of Struthiolithus chersonensis from North-
ern China, with Renarks on the Distribution of Struthious
Birds, By ©. R. EASTMAN.

In the year 1857, or thereabouts, a remarkable fossil egg was discoy-
ered in the Government of Cherson, in South Russia. The circumstances
of its being brought to light were peculiar, and its subsequent history
is instructive enough to repay a brief recapitulation, which we give as
follows. ;

During a spring freshet, a small stream occupying an old watercourse
excavated a hollow below a milldam near the village of Malinowka, in the
Chersonesus. Some peasants happening by at the time observed floating
on the surface of the pool an egg-shaped object, which they lost no time
in capturing. A neighboring frecholder acquired the specimen, and from
him it passed into the possession of his nephew, a man by the name of
Dobrowolsky, who offered it for sale to various Russian institutions.
Declined by all on account of the exorbitäntly high price demanded for
it (1,000 roubles), it was preserved in the family ‘for over twenty-five
years, until through a deplorable mishap it was shattered into nearly
forty pieces. Some of the fragments, however, were obtained by Pro-
fessor W. von Nathusius, who examined them microscopically, and
declared that they indicated a very close relationship to the common
ostrich.!

But long before the destruction of this unique specimen, namely, in
the year 1872, Professor A. Brandt of Charkow, to whom it was sub-
mitted by the owner, had the forethought to take a plaster cast of the
fossil, and at the same time prepared a minute description of it. We
are indebted to him for our principal knowledge of this egg, as well as
for the news of the fatality that ultimately overtook it.

1 Zoolog. Anzeiger, Vol. IX. No. 214, p. 47 (1886).

2 Bull. de Acad. Imp. des Sei. St. Petersburg, Vol. XVIIL pp. 158-161 (1873);
Mélanges Biol, Vol. VII. pp. 730-735; Ibis, [3], Vol. IV. pp. 4-7 (1874); Zoolog.
Anzeiger, Vol. VIII. No, 191, p. 191 (1885).

VOL. XXXII, — NO, 7, 1

128 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The matrix from which the specimen was washed out is stated by
Brandt to have been a reddish brown friable loam, overlying a bed of
crystalline gypsum. So symmetrical was the shell in outline that it
was wellnigh impossible to determine which end corresponded to the
acute, and which to the obtuse pole of ordinary birds’ eggs.’ Being
perfectly intact, little more could be done than to note its general ex-
ternal appearance, and ascertain its weight and dimensions. The weight,
which fell a trifle short of two Russian pounds, was found to be dispro-
portionate to the thickness, but is probably to be accounted for by the
mineral infiltration observed by Nathusius. According to the latter,
portions of the interior were lined with a crystalline deposit having «
thickness of 1.8 cm. in places, and there was also a quantity of loose cal-
careous matter within the ovulite, supposed to represent the fossilized
membrana teste.

The capacity of the egg, that is to say, its cubic contents, was deter-
mined by Brandt indirectly from the displacement of a plaster cast
in water, which amounted to upwards of 2200 c.cm. Allowing say
125 c.cm. for the volume of the shell substance itself, the actual capacity
of the interior is seen to be about 2075 c.cm., as indicated in the table
on page 133. Now, as the shell of the largest known ostrich egg has
only two thirds this capacity, it is plain that the fossil egg must be the
legacy of a larger bird than the ostrich, and very likely one differing in
other respects as well as size. That the egg belonged to some Struthious
bird is shown in a convincing manner by the microscopic structure,
which is eminently characteristic of the group.” But in the absence of
direct skeletal evidence, such as might have been afforded by associated
bones, it seems inadmissible to refer so huge an egg to the same genus
as the living ostrich. Therefore we aro inclined to dissent from the
proposition of Nathusius to abandon the genus Struthiolithus, which
was very properly created for its reception by Brandt, and agree with

1 On the cause and significance of polar deformation of egg shells, compare the
following suggestive papers: Ryder, J. A., The Mechanical Genesis of the Form of
the Fowl’s Egg (Proc. Amer. Philos. Soc., Vol. XX XT. (1893), pp. 203-209); Wick-
mann, H., Die Lage des Vogeleies im Eileiter vor und während der Geburt (Journ.
für Ornithologie, Vol. XLIV. (1896), pp. 81-92).

2 Nathusius, W. v., Ueber die Eischalen von Aepyornis, Dinornis, Apteryx, etc.
(Zeitschrift für wissensch. Zool., Vol. XXI. (1871), pp. 380-856); Ueber die char-
acteristischen Unterscheidungszeichen verschiedener Straussen-eier (Journ. für
Ornithol., Vol. XXIII. (1885), pp. 165-178) ; Hutton, F. W., On the Microscopical
Structure of the Egg Shell of the Moa (Trans. New Zealand Inst., Vol. IV. pp.
166, 167), 1872.

EASTMAN: REMAINS OF STRUTHIOLITHUS CHERSONENSIS. 129

the latter in retaining it, provisionally at least, or until information
is at hand concerning the creature itself. The specific title applied
by Brandt to the ovulite and its as yet unknown parent bird is
S. chersonensis.

According to Eichwald, fossil avian remains are extremely scarce in
Russia. An instance is reported by von Nordmann, however, where
certain bones were recovered from Tertiary deposits near Odessa, but no
hint is given as to their probable affinities. The Pliocene of the Siwalik
Hills in India, as is well known, has yielded ostrich remains which
indicate a species (S. asiaticus) apparently closely related to S. camelus ;
and other fragments, described as S. karatheodoris, have been found in
the Lower Pliocene of Samos.

Up to the present time no further examples, either of egg shells or
of the skeleton of S. chersonensis have come to light; and as already
remarked, all that remains of the unique type are the fragments said to
be still preserved in the St. Petersburg Museum. We have, therefore,
no little satisfaction in being able to announce the discovery of a second
perfect specimen, which has recently found its way to this country from
China. The configuration of the shell and much of its surface detail
are shown on the accompanying plate, which has been reproduced from
photographs. The history of the new specimen is as follows. Four or
five years ago, a Chinese farmer, while working at the foot of a bank of
earth about six meters high, dug out what he considered to be a pair of
“ dragon’s eggs.”? One was broken, the other entire, and, presuming
the latter to have some commercial value, he took it with him to Kal-
gan, and disposed of it to Rev. William P. Sprague, one of the American
Board missionaries residing there. Rev. James H. Roberts, a brother
missionary who has also spent many years in China, was present when
the egg was sold, and on revisiting this country last spring brought the
specimen with him on behalf of Mr. Sprague, to be offered for sale
to some scientific institution. Eventually it was purchased for the
Museum of Comparative Zoölogy, where it is now deposited.

The Chinese workman who found the egg was well known to the
servants of the missionaries as a man living in Yao Kuan Chuang. This
is a small village in the district of Hsi Ning, about fifty miles south-
southwest from Kalgan by road, but somewhat nearer in a straight line, as
that region is very mountainous. Subsequently Mr. Sprague visited the

1 The circumstance of two eggs being found together accords well with Owen’s
suggestion that the oviposition of the moa was probably in pairs. Cf. Extinct
Wingless Birds of New Zealand, Vol. I. p. 820, 1879.

130 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

exact spot where the eggs were dug up, in company with the man who
found them, and thns satisfied himself of the authenticity of the dis-
covery. The fragments of the second specimen were unfortunately not
preserved, and as Mr. Sprague was in doubt whether the perfect one
was indeed an egg or perhaps only a geode, he cut a hole about 1.5 cm.
in diameter at one end, to ascertain if there were crystals on the inside.
But on reflecting light into the interior, the walls were seen to present
the same general appearance as the external surface, and a loose calca-
reous mass, partly in the form of powder and partly flakes that appear
to have become scaled off from the inner surface, was found within the
cavity. This mass is still preserved in the same condition as when
found, and weighs 18.1 grams. Possibly it represents in part the
calcified shell membranes, such as have been found fossilized in certain
‚moa eggs. Examined with a pocket lens, the flakes present no appear-
ance of having an organized structure.

3efore proceeding to a description of the egg, it may be of interest
to note some of the topographical features of the region as depicted by
Mr. Roberts in the following sketch, which he was kind enough to draw
up at the writer’s request : —

“The city of Kalgan numbers about 100,000 inhabitants, and lies 140 miles
northwest of Peking, Chita. Having lived there ever since 1880, I am famil-
iar with the city, and all the surrounding section of country. The village of
Yao Kuan Chuang I have frequently visited on my preaching tours to Yi
Chou and Hsi Ning. The Hsi Ning (Western Repose) valley extends from
W.S. W. to E. N. E., being from seven to ten miles wide near its eastern end,
but at a distance of more than thirty miles from that end it begins to widen
gradually toward the west. Yao Kuan Chuang is situated about twenty miles
from the eastern end, and two miles from the mountains on the north,
Through the valley, from west to east, flows the Sang Kan (Mulberry Dry),
which is the largest river in the region northwest of Peking. It has no bridges,
except in winter, but is fordable in certain places. The Hu Liu (Pot Flow)
River, flowing northeast past Yü Chou and then north, joins the Sang Kan at a
point three miles cast of Yao Kuan Chuang. The Hsi Ning valley, except at
its western end and where the Hu Liu River comes in, is walled with moun-

ains several hundreds of feet above the river. The elevation above sea level,
at the junction of the Sang Kan and Hu Lin, is shown by the barometer to he
about 2740 feet. The mountains present the appearance of bare rock, gullied
out in former times by glacial ice, and so steep that grass cannot grow on the
larger part of their surface. The rocks are stratified, tilted up at a high angle,
and contain a large amount of mica. In one place sheets of mica are taken out
for commercial purposes, to be made into window panes.
“There are extensive formations of loess in the Yü Chou and Msi Ning

EASTMAN: REMAINS OF STRUTHIOLITHUS CHERSONENSIS. 131

valleys, from fifty to a hundred feet deep; but the loess, or yellow clay, has
been worn away from the greater part of the land north of the Sang Kan River,
so that sand and gravel predominate. Wherever irrigated by a stream from
the mountains, the land becomes very good and fertile, Near the eastern end
of the Hsi Ning valley the Sang Kan River enters a narrow gorge, through
which it flows ten or thirteen miles to the lower valley of the Pao An. Hach
of these valleys was once a lake, walled in by the mountains around them.
The evidence of this is unmistakable. As to the Hsi Ning valley, the Chinese
say it is recorded in their histories that it was occupied by a lake until about
the year 1000 A. D., when the waters cut through the mountains, and the lake
was drained off. The recent date assigned to this event, and the general accu-

acy of Chinese history, would seem to make the story credible, while the
configuration of the land shows that such an occurrence must have taken place.
Now the valley is tilled very carefully, and villages of a hundred families are
interspersed only two or three miles apart. The soil is mostly poor and grav-
elly, but the stones in it, heated by the sun, radiate their heat at night, and
the mountain range on the north is frequently visited with rain; yet the
climate on the whole is rather arid.

“ Just north of Yao Kuan Chuang is a remarkable seam of red rock [erup-
tive dike ?] eighty feet thick, intersecting the mountains in a vertical plane.
It is very conspicuous in contrast to the brown-colored mountain, and is alleged
by the natives to be the trail of a serpent, Eight miles west of Yao Kuan
Chuang is a ridge composed of at roungled pebbles and rock fragments
[esker ?], which extends from the foot of the northern mountains to the river,
a distance of about two miles. Farther west are a large number of craters, the
widest being about two miles in diameter, and containing numerous smaller ones.
Seventeen miles southwest of Yao Kuan Chuang, on the opposite side of the
river, is Fu T’ou Chiang, a market for coal, which is brought in large blocks
on mules from the southern mountains. This coal is of two kinds, one that
smokes, and one that does not; and the chief peculiarity of both kinds is that,
if any part of a block gets on fire, the whole will slowly consume away, leaving

only white ashes. A lump of it as big as a man’s

ist, if covered with ashes,
will keep a fire all night.”

Reference to the works of von Richthofen! and Pumpelly? on the geol-
ogy of China shows that the above account is in substantial agreement
with the descriptions of the surrounding region, as furnished by these
authors. Both of them comment on the desiceation that has taken place
within comparatively recent times, and note the traces of former shorc
lines along the mountain sides. The lower part of the Sang Kan flows
through a synclinal valley, and higher up in its course it drains a number
of loess basins, from one of which our fossil was derived. According to

1 Richthofen, F. Fr. von, China, Vols. L.-IIT. Berlin, 1883.

2 Pumpelly, R., Across America and Asia. 2d edition, New York, 1872.

132 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

von Richthofen,' the superficial deposits of these basins were laid down
over the bottoms of isolated salt lakes having no outlet, and were after-
wards partially buried by alluvial detritus. The moderate depth to
which ravines have cut through the ancient lake beds, and the straight
narrow gorge of the Sang Kan, through which the waters were drained off,
corroborate the belief that this event took place at no very remote period,
At all hazards, it is certain that no earlier age can be assigned to the
gravel beds from which the fossil egg was exhumed than the Pleistocene,

Turning our attention now to the specimen itself, we notice that it
presents almost exactly the same appearance as the «Russian egg de-
scribed by Brandt, a comparison with which has been facilitated by
means of a plaster cast very kindly presented to the Museum by this
author, In size, symmetry of contour, and all outward respects, the
two are remarkably alike. The only noticeable difference is in respect
to weight, but this is readily explained by the mineral infiltration already
referred to in the case of the type, and the absence of such replacement
in the present example. Assuming that none of the internal calcareous
mass has been lost, the Chinese egg must have weighed originally about
337 grams, as against “nicht ganz zwei russische Pfunde” for Brandt’s
specimen, ‘The weight of our individual, as given in the subjoined table,
is for the empty shell, perforated at one end in the manner already de-
scribed. The mutilation cannot be said to be altogether regrettable,
however, since it permits of an inspection of the interior, and facilitates
a measurement of the volume (equal to the loss of weight in water),
cubic contents, and specific gravity of the object.

Measurements of a number of other eggs were made in like manner,
and the results are tabulated herewith for the sake of comparison. In
the case of the South American ostrich, an average-sized and also a
large-sized individual were selected from a collection of over two dozen
contained in the Museum of Comparative Zodlogy. The dimensions of
Epyornis and Dinornis egos are taken from the literature of these
genera, excepting No. 3 of the list, which were derived from a plaster
cast obtained from the Paris Museum of Natural History. For these
two genera the cubic contents have been calculated from the formula of
an ellipsoid of revolution (since no data are at hand to show that they
have ever been ascertained by direct experiment) and, not being cor-
rected for the thickness of the shell nor for its departure from the figure

1 Loc, cit, p. 344 et seq. Also Vol. I. p. 110, quoted in Whitney’s “Climatic
Changes” as follows: “It may be stated as a certainly ascertained fact, that
in Central Asia a dry climate has prevailed for a long time.”

RSONENSIS.

ni

u

CH!

Y
=
=
=
=
E
=|
O
=
=
=)
gc
E
un

MAINS OF

N
u

RE

EASTMAN:

TABLE SHOWING COMPARATIVE DIMENSIONS OF THE EGG SHELLS OF STRUTHIOUS BIRDS.

Name of Species.

Æpyornis maximus Geoff.

Dinornis sp. .
Struthiolithus chersonensis Bdt.

“e “ ce

Struthio camelus Linn.

“e ce ce

Rhea darwinii Gould .

+
ce ce [73

Rhea americana (Linn.) .
[23 se «€
Casuarius bennetti Gould
pl australis Wall.
Dromeus nove hollandiæ (Lath.)

Gallus gallus Linn. .

Teurpngräuorg

cm.
35.1
34.0
32.0
31.0
25.2
18.0
118.0
16.38
16.0
13.5
12.9
13.8
13.8
14 | 14.0
15 12.1

116/113 |
l17| 55 |

sxy
O8IDASUVAL,
RADAR) LINE)
oem
"SOX Y JO OLY

|
|

cm.
24.5

92.5
23.0 | 84.0
23.4 |84.8
lırs | — [55.9 |1.42/1.78/2
115.0 [52.0 (46.0 |120 2.65)

ek: 51.35. it 1.22 | |2.20

=
B

76.8
71.0
72.0
73.4

92.1 |
85.0 1.51 | 3.65
1.39 | 3.45
133| —

2.556
2.471 |
13.50 | 46.00 42.40 1.19 2.001 — |
9.45 35.85 29.90 | 1.43 oe

Poa as
9.00 34.60 28.40 11.43) 11.00 12.

59 |

9.13| | 35.90| | 28.70 | 1.51 | 1.00 | 2.361

H | | |
| 8.65 32.80 27.15 ER
8.08 30.04 25.40 1.40 1.19 2.405 5
4.30 15.55 19.40 1.98 0.39) 2.193
j | |

310.05 | 12

| 310. 95 | 125

79.89

77.831 4

9.95 | 37.35 31.251130) | 1.00 |2.230 106.64

74.43 |

73.86

| 65.41

OWO A

Cubic

Contents.

c.cm.
11,035.8
9,012.5
8,863.5
8,887.8
4,180.6

2,075.+

1,896.90 |
1,423.63,
| 1,350.00

Remarks.

Original in Brit. Museum, No. 41,847.
ng ‘ Paris Museum.

* Paris Museum.

Plaster cast in M C.Z.

Figured by Owen and Rowley.

| Type specimen.

Chinese example.

Largest ostrich egg in M. C. Z.
Brandt's largest ostrich egg.
From Sandy Point, Patagonia.
From *

Brazil.

New Britain.

3! From Australia.

7 From Victoria, Australia.

50.00

Medium-sized hen’s egg.

134 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

of a perfect ellipsoid, are somewhat in excess of the true values. The
capacity of all the other eggs was found by weighing them when filled
with water, and reducing the results; hence thé table can be relied
upon as being more accurate than that based upon the computations of
Geoffrey St.-Hilaire.

This author asserts that the egg of Æpyornis is equal in capacity to
6 ostrich eggs, or 12 of the nandu, 16.5 of the cassowary, 17 of the emeu,
or 148 of the common fowl, assuming the latter to hold on the average
60 c.cm. One arbitrary standard is as good as another, and, as the hen's
egg for which measurements are given (No. 19 of the table) happened
to contain exactly 50 c.cm., we may adopt it as our unit; whence it
appears that a medium-sized emeu’s egg is equal to 7 such units ; a casso-
wary’s (O. bennetti) 11; a nandu’s 10 to 12; an ostrich’s 27 ; a moa’s 84;
and the largest known Æpyornis egg no less than 220. It may be worth
mentioning in this connection, for the benefit of those interested, that
the eggs of Æpyornis and Dinornis have been sold at prices ranging
between £100 and £200, and the price asked for the type of Struthio-
lithus equals about $770 of our money. These fabulous prices are de-
pendent, of course, upon the scarcity of the objects ; for although egg
fragments of the two first named genera are tolerably abundant, the
number of perfect specimens all told is less than a score.

But to return to the description of Struthiolithus. It is probable that
the egg shell was only partially embedded in the soil when found, the
evidence for this being that the greater portion of the surface is in-
crusted, more or less granulated, and otherwise affected by atmospheric
erosion. The loast weathered side is that shown in Figure 2, on which
several arcas are to be observed where the original shell has remained
unaltered. Some discoloration has been brought about through the
agency of iron oxide, aud grains of ferruginous sand still adhere to the
shell in places, or are even partially embedded in the crust. This side
of the shell is of a brownish yellow color, somewhat darker than the
opposite or more weathered side, shown in Figure 1. Numerous fine
pittings are to be seen over the greater part of the periphery, some of
which may be due to destructive agencies, but the majority of them are
clearly to be regarded as the round terminal pores of air canals. To

1 Comptes Rendus, Vol. XXXII. (1851), p. 102. Cf. also Comptes Rendus,
Vols. XX XIX. (1858), p. 833, and LXV, (1864), p. 476; Proc. Zoolog. Soc. London,
1852, p. 9, and 1867, pp. 892-991; Ibis, [2], Vol. IV. (1868), p. 65; Ornithological
Miscellany (G. D. Rowley), Vol. IIL. p. 287 (1878); Owen’s Extinct Wingless
Birds of New Zealand, Vol. I. pp. 817-820 (1879),

EASTMAN: REMAINS OF STRUTHIOLITHUS CHERSONENSIS. 135

determine their precise relations it would be necessary to sacrifice a
portion of one of the best preserved arcas for the purpose of making
a thin section; but as no such area is contiguous to the aperture cut
at the upper end by Mr. Sprague, no further incisions have been at-
tempted. It is doubtful in any case whether a section would show more
than has already been ascertained from Nathusius’s study of the type
specimen, which merely proved that the air canals terminated in a
similar fashion as in Struthio camelus. But the variations in the struc-
ture of egg shells among different birds, or even in different parts of the
same egg, are so considerable,’ that we are averse to depending upon
this method for accurate systematic identification. In the opinion of
the writer, the utmost we are warranted in affirming with regard to the
relationship of Struthiolithus is, that it probably was very like the living
ostrich, but not necessarily a member of the same genus; hence the
propriety of retaining Brandt’s name in a tentative sense is apparent.

The occurrence of fossil ostrich remains in the loess of such widely
separated regions as Northern China and Russia has a direct bearing
upon the distribution of Struthious birds. It enables us to speak
positively with regard to the former extension of the Struthionidæ over
Eur-Asia and Africa since the Pliocene, and gives rise to some infer-
ences, within duly circumscribed bounds, regarding the past history of
Raft-breasted birds in general. It is necessary to distinguish between
what can be affirmed of the ostrich group, properly speaking, and what
we can assume with more or less plausibility concerning the rest of the
so-called Ratite. For, if it were possible to recognize the latter as a
natural division, embracing forms genetically related to one another, or
all derived from a common ancestral type (that is to say, a “Ratite ”
type), then we should be warranted in establishing a single hypothesis of
distribution for all branches of the Ratite. But the best modern +orni-
thological opinion holds that the division into Ratite and Carinate is
unnatural, since the differences between existing species of Raft-breasted
birds are nearly as great as between any of the Ratite and Carinatæ.?

1 Blasius, R., Ueber die Bildung, Structur, und systematische Bedeutung der
Eischale der Vogel. Leipzic, 1867, pp. 48.

2 Cf. Fürbringer, M., Untersuchungen zur Morphologie und Systematik der
Vögel, Vol. II. p. 682. Amsterdam, 1888. Also (on the relations of Gastornis)
under same caption in Biol. Centralblatt, Vol. XIX. pp. 573-587 (1898). On the
taxonomic relations of Rhea, loc. cit. (1888), p. 1442: of Hesperornis, Ornith. Mo-
natsber. deutsch. Vereins z. Schutze der Vogelwelt, Vol. XV. p. 488 (1890).

136 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

For our present purpose, this question as to the homogeneous or
compound character of the Ratite is of prime importance, involving as
it does the problem of common or multiple origin for its several mem-
bers, and hence the key note of their distribution. If we admit, for
instance, as has been suggested by some, that the South American
ostrich was derived from the tinamous or some other Carinate stock,
then it must be altogether excluded from the list of forms that migrated
into the western hemisphere by means of a former land connection.
But if, on the other hand, structural resemblances are sufficient to point
to a genetic relationship with the ostrich or other Struthious type,
then its occurrence in Patagonia can be accounted for in no other way
than on the hypothesis of a land migration.

Or again, in the case of the remarkable Stereornithes, from the early
Tertiary of Argentina and Patagonia, which were considered by Ame-
ghino, Gadow, and for a time also by Lyddeker, as ancestral forms of
Ratite, — if they could be shown really to have Struthious affinities it
would be a simple matter to connect them and the modern Whea with
Diatryma from the Eocene of New Mexico ; further, with the Tertiary an-
cestors of Struthiolithus, Struthio proper, and the moas of New Zealand ;
and perhaps finally with the little known Gastornithide from the London
and French Eocene. Indeed, much stress was laid by Lyddeker on the
resemblances between Gastornis and the leading genus of the Stereor-
nithes, Phororachos. But however attractive such a theory might seem
at first glance, we are obliged to renounce it as illusory in view of recent
destructive criticisms at the hands of such excellent anatomists as F. A,
Lucas, C. W. Andrews, and others, who have caused even Ameghino
and Lyddeker to recede from their original opinions.

The last named author,! writing in 1893, placed Gastornis, Brontornis,
and Phororhachos unhesitatingly among the Zatit@, as the latter are
commonly understood. He refers to “the modern German [Fürbring-
ers] view that the Ratite form a compound group, of which the
various sections have been independently derived from several perfectly
distinct Carinate ancestors, and that their mutual resemblances to one
another are solely owing to the effects of adaptation” ; but his own
personal opinion is expressed in the following words: “I confess,
however, that the supposed Anserine affinities of Gastornis appear far
from clear to me, while I always feel that the great difficulty in admitting
the multiple origin of the Ratite is that, if this had been the case, there

1 Lyddeker, R., On the Extinct Birds of Argentina (Ibis, [6], Vol. V. pp. 46, 47).
1893:

EASTMAN: REMAINS OF STRUTHIOLITHUS CHERSONENSIS. 137

would have been far less structural similarity to one another among the
various groups than we find to prevail.”

Dr. Ameghino's final conclusion is that the Stereornithes appear to
show that the division of the class of birds into Ratite and Carinate is
not fundamental, — a view which has been especially advocated by Mr.
Lucas among American ornithologists. The latter,’ in a critical review
of Ameghino’s work, speaks as follows : “ Apparently the main reasons
for comparing such forms as Phororhachos and Brontornis with the
Struthiones is because they are large and extinct, when, as a matter of
fact, mere size is no reason for supposing a bird related to an ostrich ;
while the pelvis of Phororhachos, with its aborted pubis, shows that this
genus at least is very many removes from any Struthious bird. Nei-
ther is Gastornis, with its primitive type of skull, any relation of the
Stereornithes.”

Precisely the same attitude is displayed by Mr. Andrews,? who denies
that the Stereornithes have anything in common with Struthious birds
or with the Gastornithide, neither are they by any possibility descended
from Hesperornis. Says he: “The Stereornithes seem to be a hetero-
geneous group of birds in all of which the wings were reduced and the
bulk increased through the operation of some peculiar local conditions ;
for instance, the land which they inhabited may have been an island on
which no large carnivorous animals occurred. A similar example is
offered by New Zealand, where the Dinornithide, Apteryx, Aptornis, and
Onemiornis (all flightless birds of large size, and belonging to distinct
orders) were formerly found. Indeed, there seems'to be no reason why
at any time, from the late Secondary period onward, and in any region,
similar groups of flightless birds might not have arisen under favorable
circumstances. The Gastornithide may be another instance of such.
In most cases such specialized races die out without leaving any descend-
ants when the peculiar conditions to which they have become adapted
pass away ; but the modern Ratitæ may be survivors of one or several
ancient groups of such flightless birds.”

Still more explicit are Mr. Lucas's views as to the nature and origin
of the so-called Struthious birds, which he defines as “those generalized
birds which through some special adaptation to their surroundings or
freedom from enemies have been able to survive to the present day.
Rhea and Struthio are typical examples of this. While it would be

1 The Auk, New Series, Vol. XIII. (1896), pp. 62, 63.
2 Remarks on the Stereornithes, a Group of Extinct Birds from Patagonia
(Ibis, [7], Vol. IL p. 12). 1896.

138 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

assuming too much to say that the limit of size has been attained by
any bird of flight, it is yet very evident that the difficulties of flight
increase very rapidly with increase of size, —hence the correlation be-
tween gigantic stature and loss of flight. It is a noteworthy fact that
the forms attaining the maximum size in their respective groups are
generally flightless, e. g. Onemiornis, Notornis, Didus, etc. Since flight-
less forms have originated in comparatively recent times wherever the
conditions were favorable, so undoubtedly they arose in the past, and
only amid unusually favorable conditions and stable environment could
these ancient flightless forms persist. That existing ‘Ratite’ birds
were long ago differentiated from the parent stock, or that they arose
independently, is indicated by the great differences between forms sepa-
rated by considerable stretches of water. In view of the parallel devel-
opment of the horse and rhinoceros in Europe and America, it would
hardly seem necessary to suppose a unity of origin for Struthious birds ;
moreover the paleontological history of the class is so fragmentary that
phylogenetic arrangements of the birds can be regarded as little more
than guesses. As to the characters of the Ratite, the absence of a keel
to the sternum and the slight angle between the scapula and coracoid are
purely degenerate features without the slightest taxonomic value; and
the ‘ Ratite’ type of skull is a generalized skull having resemblances to
that of the reptiles. The characters in which Hesperornis resembles the
ostrich are generalized characters, such as one would be surprised not to
find in so early a bird; its shoulder girdle is unique among birds and
decidedly reptilian, while the foot is the most highly specialized swim-
ming foot known. That this bird is the direct descendant of any land
bird is incredible. As for the tinamous, their skull and pelvis of a very
generalized type prevents us from regarding them as recent derivatives ;
they are in fact * hold-overs’ in a region noted for the number of curious
forms it contains, indicating the persistence of a few very old species in

: , : : non te 1
the midst of a more advanced yet not strictly modern fauna,

1 The above quoted remarks are from some notes which Mr. Lucas very kindly
took the pains to write out by way of comment on the present paper before it was
finally prepared for press; and the writer has great pleasure here in acknowl-
edging his indebtedness to this source for many helpful criticisms and suggestions.
Reference should be made also to Mr. Lucas’sreview of Professor Thompson’s
paper “On the Systematic Position of Hesperornis,” published in The Auk, Vol.
VIII. p. 304 (1891), as well as to the comments of Dr. J. A. Allen in the same
journal, Vol. XV, p. 70 (1898), which brings the literature of Zesperornis down to
date. For a copious bibliography of the distribution of recent birds, see Mr. P. L,
Sclater's address before the Second Ornithological Congress at Budapest, May,
1891.

EASTMAN; REMAINS OF STRUTHIOLITHUS CHERSONENSIS. 139

Strong enough arguments, we think, have been put forward to show
that the theory of a common origin of the Ratite is untenable, and
hence no single hypothesis of distribution is able to account for the facts
of their distribution. We cannot imagine a race of ostriches sprung
from Hesperornis or anything of like nature in the Cretaceous, spreading
over the whole earth in the Tertiary, and then, as decay set in, leaving
its fragments scattered in remote corners of the globe. But the prob-
lems presented by the alternative theory, that of multiple origin, are
none the less interesting or important, although decidedly more compli-
cated. To seek the nearest Carinate affinities for the different sections
of Ratite separately ; to develop the paleontological history of each
more fully; and to inquire into the physical and biological conditions
which led to their insulation, perpetuation, and differentiation in various
provinces, — these are only a few of the points that invite an extended
investigation. Some of the problems have already been touched upon,
notably as regards the origin of the moas, the South American ostrich,
and Æpyornis ; and we may profitably turn our attention for a moment
in this direction, beginning with Æhea.

No one can deny that the physical resemblances between Rhea and
Struthio are very great; in fact, the popular term “South American os-
trich ” is an obvious commentary on their similarity. Although both
genera are regarded as typical of distinct families, and are even com-
monly placed in separate suborders, yet, if one were asked to specify the
nearest living ally of the African ostrich, he would unhesitatingly point
to Rhea. Only two interpretations of structural resemblances are pos-
sible: either they indicate direct genetic relationships, or we have here
a most remarkable case of convergence. Now in this instance we confess
to sharing Dr. Lyddeker’s prejudice against the theory of parallel devel-
opment, already quoted, as there is too striking a coincidence in the
forms produced to be explained as the result of adaptation during recent
times in two widely distant regions of the globe. The natural and only
logical plan would be to assume blood relationship between Jvhec and
Struthio as a matter of course, until it is proved that by no possibility
could they have been derived from the same ancestral stock. Now if
Rhea had different progenitors from the ostrich, we are in utter ignorance
as to what they were like, as no other descendants remain. That there is
anything in common between Ahea and the tinamous we cannot believe
for a moment, in view of the different organization of the latter. Hence,
Captain Hutton's theory, which derives both Rhea and the moas from

a tinamou-like ancestor which crossed into Australia and New Zealand

140 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

by means of an imaginary Antarctic continent, must be relegated on
both biological and geological evidence to the same category as the
Lemurian hypothesis.

Rhea still enjoys a comparatively wide distribution in South America,
and its remains have been found in the bone caverns of Brazil. If the
evidence of Diatryma in New Mexico means anything at all, it would
point to a connection between a fossil North American and the existing
South American ostrich. It is true that the late Tertiary yields no
evidence of Struthious birds in North America. But it is also true that
until the discovery of Struthiolithus under the shadow of the Great Wall
in China, no one could have suspected the whole intervening territory
between Northeastern Asia, South Russia, and Africa to have been in
comparatively recent times inhabited by true ostriches. The palæonto-
logical record is from the nature of things very deficient in the case of
land birds, and many gaps,can only be filled on indirect evidence. One
such gap is now partially filled by the occurrence of Struthiolithus in
Northern China. A race having the constitutional vigor and numerical
force to establish itself in this latitude, — and in a mountainons region
as well, where the struggle for existence is always intensified by a larger
number of enemies than are found on the plains, to say nothing of the
rigors of winter, — must have been able to penetrate still further north-
ward, and might readily have accompanied the mammals that migrated
across the land bridge formerly connecting the palænretie and nearctic
regions.

In a word, if we can predicate any blood relationship between the
African and South American ostriches, it is certain that the latter could
have reached its present habitat in no other way than along the route
marked by Struthio camelus, S. karatheodoris and S. asiaticus, Struthio-
lithus, Diatryma, and the Rhea of Brazilian bone caverns. If any will
presume to deny a relationship between Struthio and Rhea, they are
confronted with these difficulties: to explain how two separate deriva-
tives from Carinato birds should come to present such marvellous
similarity to one another through the operation of purely fortuitous
conditions, and to point out a lineage for Rhea connecting it more
closely with Carinates than with the ancestors of Struthio. Sceptically
inclined individuals are welcome to regard Rhea as one of the “waifs
and strays of a lost avifauna left by the sea of time stranded on the
shores of the present,” but we personally prefer the more positive
view, which connects the New and Old World ostriches in the manner

indicated.

EASTMAN: REMAINS OF STRUTHIOLTTIIUS CHERSONENSIS. 141

Turning now to the advent of the moa into New Zealand, and of
Epyornis upon the island of Madagascar, we note that previous writers
have essayed in various ways to meet the following dilemma. If these
birds migrated from what is now the mainland prior to the Tertiary, why
have not their remains been found in strata older than the Pliocene ?
Or, if the islands remained inaccessible to them until the late Tertiary,
how was the passage finally accomplished by wingless birds? And what
is still more to the point, why were they in the latter event unaccompa-
nied by placental mammals? Of the two principal theories put forward
to explain the facts, neither, in the opinion of the present writer, suffi-
ciently accounts for all the difficulties. In the case of the moas, Mr.
Wallace * supposes that their ancestors cither flew or swam across straits
impassable to contemporaneous mammals. But Captain Hutton prep-
erly takes exception to this view, on the ground that their Ratite
characters are due to their being unable to fly; moreover, the oldest
known moas were entirely without wing bones, and possessed a very
rudimentary shoulder girdle. As to the alternative explanation that
they crossed by swimming, Captain Hutton? remarks as follows: “ But,
although the emeu and the rhea are both said to take readily to water,
many placental mammals do the same, and it is very unlikely that the
Struthious birds should twice have swum across the same straits — once
from the Oriental to the Australian region, and again from the Austra-
lian to New Zealand — which were impassable to mammals. There are
also other reasons for doubting the northern origin of the Australasian
Ratite.”

Now, as most persons are aware, both biological and geological evi-
dence go to show that New Zealand has been separated from Australia,
and Madagascar from Africa, at all events ever since the dawn of the
Tertiary, and probably since the latter part of the Cretaceous. Accord-
ingly, Captain Hutton supposes, after having shown the improbability
of wingless birds either flying or swimming across straits, that the moas
have had a different origin from the rest of the Ratite, — an opinion
from which we find no reason to dissent, although rejocting his hypothe-
sis of a tinamou-like ancestor in the Eocene, and a submerged Antarctic
continent. Although the oldest known remains of the Dinornithide are
of Miocene age, and no one has attributed to the family an earlier origin
than the Eocene, yet the occurrence of moa remains on the continent of

1 Wallace, A. R., Island Life, 2d edition, p. 481, 1892.

2 Hutton, F. W., The Moas of New Zealand (Trans. New Zealand Inst., Vol.
XXIV. p. 147), 1891.

VOL. XXXII. — NO, 7. 2

142 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

Australia (Pliocene of Queensland) furnishes ground for believing that
a form of wingless bird arose in Australasia at some time previous
to the detachment of the present New Zealand and adjacent islands
from the mainland. If the stock really had an uninterrupted Tertiary
history, it is of course immeasurably removed from any existing Carinate
forms, or other Ratite. The moas are supposed to have enjoyed their
period of culmination during the Pliocene, when they flourished prodi-
giously and covered the land, but thereafter they suffered great mortality.
Their extinction was certainly not caused by the encroachment of natu-
ral enemies, any more than was the case with Æpyornis; nor can their
decline be reasonably attributed to physical or climatic changes of which
we have no evidence. Their decadence not being traceable to external
influences, we can only interpret it as the result of some inherent cause
or causes, — taken together with the retarded action of natural selec-
tion, — such as are frequently seen to follow in the wake of hypertropby
among various groups.

We cannot pass from this subject without calling attention to Captain
Hutton’s remarkably ingenious explanation of the crowding together of
so many varieties of Struthious birds in’ the limited area of New Zea-
land, and the unequal distribution of species between the two islands.
What appear at first sight to be unparalleled or anomalous features of
distribution are all consistently explained on the theory of a simple
order of geographical changes, namely, alternating elevation and subsi-
dence of land masses. Two periods of subsidence and one of elevation
are sufficient to account for all the phenomena, according to Captain
Hutton’s hypothesis. His interpretation, with which Mr. Wallace
heartily concurs, is concisely summarized by the latter author as fol-
lows: “First, we must suppose a land connection with some country
inhabited by Struthious birds, from which the ancestral forms might be
derived ; secondly, a separation into many considerable islands, in which
the various distinct species might become differentiated ; thirdly, an
elevation bringing about the union of these islands to unite the distinct
species in one area; and fourthly, a subsidence of a large part of the
area, leaving the present islands with the various species crowded
together.” !

To rovert finally and in few words to the origin of Æpyornis, it is plain
that, if its ancestors reached the island from Africa as flightless birds,

the migration must have taken place not later than the Focene, since no

mammals initiated since the Cretaceous are found in Madagascar, The
1 Loc. city p. 419.

BASTMAN: REMAINS OF STRUTHIOLITHUS CHERSONENSIS. 143

lemurs, insectivora, rodents, and reptiles now inhabiting the island bear
the stamp of great antiquity. It is true that „Zpyornis remains, so far
as known, are confined to deposits of supposed Pleistocene age, and the
material itself is inconsiderable.* The deficiency may be partially ac-
counted for owing to lack of exploration, and imperfection of the palæ-
ontological record, especially as concerns land birds. But if it be
objected that a continuous Tertiary history implies a higher degree of
specialization and greater specific variation than we have any evidence
of, we must not forget that here the action of natural selection was
more or less suspended, owing to the abundance of food, absence of
carnivorous or other powerful enemies, and generally mild conditions.
The same causes produced the same effects upon the ancestors of Æpy-
ornis as upon the moas, of which Captain Hutton affirms, “ Under such
favorable circumstances the conditions of life were very easy, and the
birds grew larger and fatter, more sluggish and more stupid,” — until,
in fact, they became effete and were finally extirpated,

On the assumption that the forerunners of Æpyornis have inhabited
Madagascar ever since its separation from the mainland, it is not surpris-
ing that this genus should have to stand by itself as the representative
of a distinct suborder. Some resemblances to the Struthio-Rhea branch

of the Ratite are observable, to be sure, — even the egg shells having a
remarkably similar structure, as shown by Nathusius, — and anatomists
like Fürbringer, Milne-Edwards, and Grandidier have endeavored to
show even greater similarity to Dromeus and Casuarius, which are
regarded as the most primitive of existing Aatite. But all are agreed

that the differences far exceed the resemblances, being in fact funda-
mental. To us it seems that the amount of divergence from other
known types, living and fossil, coupied with the slower rate of variation
affecting insular forms, and, above all, the generalized characters ob-
served in Æpyornis, go to show that this bird is far from being a modern
derivative, but is the last of a very ancient race. We see no impropri-
ety in supposing that its flightless ancestors inhabited the Madagascar
region when it was still a part of Africa, aud that other descendants
may have migrated northward into Europe during the early Tertiary.

L Andrews, ©. W., Note on a nearly complete Skeleton of Æpyornis from Mad
agascar (Geol. Mag., Dec. 4, Vol. IV. pp. 241-250), 1897. Burckhardt, R., Ueber
Æpyornis (Palaeont, Abhandl. Dames und Kayser, Vol. IL Heft 2), 1893.

144 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

DESCRIPTION OF THE PLATE.

Fig. 1. Struthiolithus chersonensis Brandt. From superficial deposits in the neigh-
borhood of Kalgan, China. Obverse, or more weathered aspect, with
prominent pittings. X $.

Fig. 2. Reverse aspect of same specimen, showing uncorroded areas. Millimeter
scale at bottom of the figures. X §.

(Reproduced from photographs by Mr. C. H. Currier, of Boston.)

EASTMAN -- STRUTHIOLITHUS.

ME MELIOTYPE PRINTING CO., BOSTON

Bulletin of the Museum of Comparative Zodlogy
AT HARVARD COLLEGE.
VoL. XXXII. No. 8.

REPORTS ON THE DREDGING OPERATIONS OFF THE WEST COAST OF
CENTRAL AMERICA TO THE GALAPAGOS, TO THE WEST COAST
OF MEXICO, AND IN THE GULF OF CALIFORNIA, IN CHARGE OF
ALEXANDER AGASSIZ, CARRIED ON BY THE U. S. FISH COMMIS-
SION STEAMER “ALBATROSS,” DURING 1891, LIEUT. COMMANDER
Z. L. TANNER, U. S. N., COMMANDING.

XXIV.

PRELIMINARY REPORT ON BRANCHIOCERIANTHUS URCEOLUS,
A NEW TYPE OF ACTINIAN.

By E. L. Mark.

[Published by Permission of MARSHALL MCDONALD and GEORGE M, BOWERS,

U. S. Fish Commissioners. |

Wirt THREE PLATES.

CAMBRIDGE MASS, U.S.A.:
PRINTED FOR THE MUSEUM.
Avaust, 1898.

No. 8. — Report on the Dredging Operations off the West Coast
of Central America to the Galapagos, to the West Coast of Mex-
ico, and in the Gulf of California, in charge of ALEXANDER
AGASSIZ, carried on by the U. S. Fish Commission Steamer
“ Albatross,” during 1891, LIEUT. COMMANDER Z. L. TANNER,
U.S. N., Commanding.

XXIV.

Preliminary Report on Branchiocerianthus urceolus, a new Type of
Aetinian. By E. L. Markt

ONE of the most interesting of the many new forms brought up in the
dredgings of the Albatross Expedition under Mr. Agassiz in 1891 was «
deep-sea actinian which bore so strong a resemblance to Cerianthus in
its general appearance that the sketches made at the time are marked
“new Cerianthus.” An examination of the superficial characters shows,
however, as Mr. Agassiz (91, p. 187) at once recognized, that this new
form differs in important points from the genus Cerianthus, and may
indeed require the erection of a new family for its reception. Its two
most striking morphological features are a pronounced bilateral sym-
metry and the possession of an incomplete circle of branching gill-like
organs. The latter peculiarity I utilize in proposing for it the new
generic name Branchiocerianthus.

A considerable number of specimens were taken at cach of two hauls
in the Gulf of Panama, not far from Cape Mala, at Stations 3385 and
3389, the depths being respectively 286 fathoms and 210 fathoms, and
the bottom being in both instances green mud. The colors, as shown by
the colored sketches drawn at the time by Mr. Westergren, were brick-
red for the column, deep carmine for the marginal tentacles, and rose-
pink for the oral disk, and all the structures arising from it, including
the oral tentacles. The specimens were all preserved in strong alcohol.

1 Contributions from the Zoölogical Laboratory of the Museum of Comparative
Zovlogy at Harvard College, E. L. Mark Director, No. XCII.

VOL. XXXII. — NO. 8.

148 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The animal as sketched in the fresh condition by Mr. Agassiz consists
of a vase- or pitcher-shaped “ flower” surmounting a long stalk, from
which it is clearly marked off by a broad shallow constriction. The
outlines are extremely graceful, presenting in front and back views
(Plate 1) a symmetrical vase-like figure with flaring lips. A side view
(Plate 2) shows that one margin of the disk is much higher than the
other, so that in this aspect the upper end of the animal resembles the
lip of a broad-mouthed pitcher. The disk, when spread by pinning out
the marginal tentacles, has an oval outline; but in the natural state the
lateral margins of the disk are folded in symmetrically from either side,
so as almost to touch at a point a little below the middle of the oval.
This bending in of the margins of the disk produces at the upper end of
the animal a sort of eccentric funnel-shaped depression, which, however,
does not lead to the mouth opening, but to the outside again, at tho
lower margin of the oral disk. The fancied resemblance of the animal
to a little pitcher, which this side view presents, has suggested the
specific name adopted, — urceolus.

In this actinian we may recognize two fairly well marked regions, —
the “flower” or calyx and the stalk. The former embraces the upper
part of the column in addition to the disk proper and its appendages ;
the latter is the remaining part of the column with its swollen base. A
slight and somewhat oblique constriction, often emphasized by a con-
spicuous line, indicates the place of transition from one to the other.

The oral disk is go modified in form as to bear little resemblance to
that of Cerianthus. When in preserved specimens the marginal ten:
tacles are pinned out under slight tension, the disk has, as already
stated, an oval outline, the longitudinal and transverse diameters of
which in the specimens measured range between 25 mm. (longitudinal)
by 15 mm, and 38 mm. by 30 mm. The surface of the disk is very
oblique to the longitudinal axis of the column. The marginal ten-
tacles are interrupted at the edge of the disk nearest the base. This
edge I shall, for convenience, designate as posterior, the opposite pole of
the long axis of the disk anterior. I adopt these designations without
the intention of advocating now any of the theoretical views concerning
the broader questions of comparative morphology in Actinozoa.

The marginal tentacles have a deep carmine color in the living animal,
becoming in alcohol a dark maroon or claret color; they vary in number
from 85 to 97.. I am not yet certain that there is an unpaired anterior
tentacle, as in Cerianthus, but believe it probable that there is, because
there are usually an uneven number of tentacles, and the method of

MARK: BRANCHIOCERIANTHUS URCEOLUS. 149

their formation is apparently the samo as in Cerianthus. They are sym-
metrically arranged at the margin of the disk, the hiatus at the posterior
margin giving the line which connects their several insertions the form
of an elongated horseshoe. The shortest, and evidently youngest, ten-
tacles occupy the ends of the two arms of the horseshoe, and are there-
fore near the posterior margin of the disk; but the tentacles do not
increase regularly in length toward the anterior margin. There are
usually one or two pairs of posterior tentacles that are quite short ;
but the two of a pair are not necessarily of equal length ; the more an-
terior ones are successively longer and longer until a maximum length
of about 125 mm. (in alcoholic material) is reached in about the tenth
pair from the posterior margin. From this region forward there is a
gradual and not great diminution in the length of the tentacles up to
and including the anterior ones. This gives to the whole disk with its
marginal tentacles, when pinned out, an appearance somewhat resembling
a common palm-leaf fan. The marginal tentacles are so crowded that
they are much flattened at their bases, and overlap one another, as
shown in the figure on Plate 3; but the shorter tentacles near the
posterior margin of the disk (one or two pairs) are usually separated
from the next more anterior ones by an appreciable distance.

The middle region of the disk is raised invo a nearly cylindrical oral
tube, which in cross section is oval, and bears at and near its summit
the oral tentacles ; below these there is usually a slight constriction. The
free end of this oral tube occupies a plane perpendicular to the long axis
of the column, i. e. it is not oblique, like the disk from which it rises.
Owing to the obliquity of the disk to this axis, one side of the oral tube
— the posterior — is much longer than the opposite side. For the same
reason the posterior face of the oral tube passes almost imperceptibly
into the disk whereas the lateral and especially the anterior faces
make with the disk an angle. Viewed from the upper end, the oral tube
in the fresh specimen presents an oval outline, the oval being nearly
twice as long in the antero-posterior direction as in the transverse.
There is no marked difference between the two angles of the mouth; but
sections may disclose the presence of a siphonoglyph. The oral tentacles
outnumber the marginal ones, there being about 130 of them. They are
arranged roughly in quincunx, and occupy four or five rows on the outer
surface of the upper end of the oral tube. They vary in size, the largest
being (in alcoholic material) about 30 to 35 mm. in length, and 0.5 mm.
in diameter at the base. These, too, are so crowded as to be much
flattened at their origin.

150 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

From the oral disk immediately outside the base of the oral tube
there arises a row of branching thin-walled organs, which are probably
to be regarded as gills. Like the marginal tentacles, this row of gills is

interrupted at the posterior side of the disk, and consequently has the
form of a horseshoe. There are usually 16 pairs of gills and an odd
one, making 33 in all; but in one specimen there were only 27 in all, |
and in another there were 37. Mr. Agassiz has sketched one individual
with only 21. There is considerable variability in the size of the gills ;
the posterior ones, though usually smaller than the anterior ones, are not
always so. A few small gills are sometimes intercalated between those
of larger size. Though arranged approximately in a single row, there is
some irregularity in their position, especially toward the anterior end of
the disk, where they are more crowded than near the posterior end. In
fresh specimens they are of a rose-pink color, but in alcohol this color is
lost.

Each gill consists of a single cylindrical, or somewhat flattened hol-
low stalk, terminating in quite regular dichotomously forking branches.
The stalk rises abruptly from the surface of the disk, sometimes being
slightly constricted at its base; it has a length of from 4 mm. to
8 mm., and a fairly uniform diameter of from 0.5 mm. to 1 mm. The
forking may extend to the production of branches of the ninth or tenth
order. The terminal branches, of which there may be nearly 500 to a

gill, end blindly with rounded tips. In alcoholic material the branches
are often varicose, owing to distention with coagulated contents. In
the living condition they are probably of much more nearly uniform |
calibre. When not contracted the gills rise above the oral orifice, and
even above the oral tentacles themselves.

tadial canals are traceable running across the disk from the base of
the oral tube to the bases of the marginal tentacles, before reaching
which many of them fork, each of the branches communicating with the
lumen of a single tentacle.

The outer surface of the calyx constitutes the upper part of the
column; its height is greatest in front, diminishing to practically zero
behind. Above the constriction which marks the transition from it to
the stalk proper, it gradually expands to meet the margin of the oblique
oral disk. It is marked with fine longitudinal dark lines alternating
with lighter ones, as in the rest of the column.

The column resembles that of Cerianthus in being elongated, cylindri-
cal, and enlarged at its basal end. Below the constriction which marks

the boundary of cup and stalk it presents a spindle-shaped enlargement.
5

MARK: BRANCHIOCERIANTHUS URCEOLUS. 191

In the alcoholic specimens which I have measured the stalk below the
constriction varies from 105 mm. to 200 mm. in length, and from 3 mm.
to 5 mm. in diameter. The enlarged bulb-like end is from 6 mm, to
8 mm., or even more, in diameter. But in preserved specimens the
stalk is much folded lengthwise, so that the direct measurement of its
diameter is not reliable. A portion of the stalk of a medium sized indi-
vidual cut out, slit open lengthwise, and pinned out, measured 22 mm.,
so that the corresponding diameter would be about 7 mm. The bulbous
enlargement reaches about twice that diameter. The column is smooth,
except for very minute appendages at the basal end, and marked by
narrow longitudinal brownish lines about 0.2 mm. broad, alternating
with lighter colored brownish or buff ones about three times as wide
(0.6 mm.). The darker lines in the fresh specimen are bright brick-red.
Individuals differ much in the depth of color, which ‘in all is less in the
bulbous region, where the distinction between the light and dark lines
almost disappears. The wall of the stalk, though not very thick, is quite
rigid ; that of the bulb is somewhat thinner. The bulb terminates in
a small, nipple-like elevation ; but a terminal pore, if present, must be
vory minute, for I have been unable to detect one by examination of
specimens in toto. The bulb often contains a large quantity of the
shells of Foraminifera.

The outer surface of the lower part of the bulb is provided with small
tapering filamentous appendages, which are 50u to 75u in diameter at
their basos, but diminish abruptly to less than half that diameter and
then taper gradually to a point. They vary in length from 0.5 mm. to
1 mm., rarely more in alcoholic material, and are rather evenly distrib-
uted over the surface at distances of 0:5 or 0.3 mm. apart; but they are
not arranged in any definite pattern. They remind one of the peculiar
appendages of the egg shell in Fundulus heteroclitus, especially of those
of the half grown ovarian eggs (cf. Eigenmann, 790, Plate I. Fig. 6).

The lower end of the column, its bulbous enlargement, is invested
by a case which extends upward at least for a tenth of the length of the
column, and is open at its lower end. It is, of course, the secreted
product of the lower and of the column, but it has a more complicated
structure than is common in sheath-inhabiting actinians. In an indi-
vidual whose stalk was 200 mm. long the case was about 25 mm. long,
exelusive of the peculiar hair-like appendages which it bears. The
lower three fifths of the caso differs from the upper two fifths in possess-
ing very numerous long hair-like hollow appendages. The upper two

fifths may be divided into two zones of about equal breadth. The secre-

152 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

tion of the upper zone is simply corrugated and folded longitudinally ;
that of the lower exhibits thickened patches a millimeter or more in
diameter and fairly closely set. Of these patches there are three or four
rows running around the column. The secretion of all three regions has
a fibrous appearance due to the fine, sinuous corrugations which run
around the column, and are the expression of differences in the thickness
of the secretion. The effect is as though the secretions had been formed
as rings, more or less complete, which had been crowded downward by
successive additions above. In the region which bears the hairs the
same appearance is shown, under a low magnification, by the secretion
which constitutes the hair. This is a hollow thin-walled cylinder, which
gradually diminishes in size from its base, where its diameter is about
0.2 mm., to its free end. The hairs attain a length of 25 mm. or more,
and are so tough that they form for the actinian a means of secure
anchorage in the mud. The corrugations of the wall of the hair which
run around it are not due to folds of the secreted substance, for optical
longitudinal sections of the hairs show that the inner surface of the tube
is smooth. When the animal is thrown into alcohol this case, with its
tuft of matted hairs, is readily detached from the column, but a careful
examination of those case secretions which are not thus artificially sepa-

rated shows that there are at intervals exceedingly minute filaments
running out from the surface of the animal to the inner surface of the
case, and I am convinced that these are the minute filaments described
as arising at regular intervals from the surface of the bulbous portion
of the column. I believe that a single filament is enclosed in each of

the hollow hair-like appendages of the case, and that the form and size |
of these hairs is, in part at least, determined by the presence and shape |
of the filaments of the column. The lower part of the case is much
wrinkled lengthwise, and projects to some distance (5 or 10 mm.) below
the base of the column. It has the appearance of having once sur-
rounded the enlarged end of the column and having been slipped back-
ward, its elasticity causing it. to become wrinkled like the mouth of a
meal-sack that is tied. The hair-like processes surround this open end
of the case and project beyond it.

MARK: BRANCHIOCERIANTHUS URCEOLUS. 158

PAPERS CITED.

Agassiz, A.

'91. Three Letters from Alexander Agassiz to the Hon. Marshall McDonald,
United States Commissioner of Fish and Fisheries, on the Dredging
Operations off the West Coast of Central America to the Galapagos, etc.
Bull. Mus. Comp. Zoól, Harv. Coll., Vol. 21, No. 4, pp. 185-200. June.

Eigenmann, C. H.
'90. On the Egg Membranes and Micropyle of some Osseous Fishes. Bull.
Mus. Comp. Zoól. Harv. Coll., Vol. 19, No. 2, pp. 129-154, Pls. 1-3.
March.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

EXPLANATION OF PLATES,

The drawings are by Magnus Westergren, from sketches by Alexander Agassiz.

PLATE 1.
Upper end of Branchiocerianthus urceolus, seen from the side (posterior) bearing
the youngest marginal tentacles. Fresh specimen. Magnified 21,
PLATE 2.

Side (slightly oblique) view of same.

PLATE 3.
Upper end of an alcoholic specimen of same, viewed from the same side as in

Plate 1, but as it appears when the tentacles are pinned out so as to expose the
surface of the oral disk,

ALBATROSS Ex, 1891.

BRANOHIOOBRIANTHUS, PLATE 1.

A. M, Wostergron del,

Folsom & Sunorgren photo,

ALBATROSS Ex. 1891. BRANORIOCERIANTHUS. PLATE 2;

AM, Westergron del, Folsom & Sunorgren photo,

ALBATROSS Ex. 1891, BRANCHIOOBRIANTHUS. PLATE 8.

A.M, Westergron del, Folsom & Sunergren photo,

Bulletin of the Museum of Comparative Zoölogy
AT HARVARD COLLEGE.
Von. XXXII. No. 9.

ACALEPHS FROM THE FIJI ISLANDS.

By ALEXANDER AGASSIZ AND ALFRED GOLDSBOROUGH MAYER.

Wira SEVENTEEN PLATES.

CAMBRIDGE, MASS, U.S. A. :
PRINTED FOR THE MUSEUM.

FEBRUARY, 1899.

No.9.— Acalephs from the Fiji Islands. By ALEXANDER
AGASSIZ and ALFRED GOLDSBOROUGH MAYER.

INTRODUCTION.

Durine a visit to the Fiji Islands, extending from November 7, 1897,
to January 13, 1898, we devoted considerable time to a study of the
Marine pelagic fauna. We made use only of open tow-nets, and while
the majority of our hauls were made upon the surface, a number were
also made at depths varying from twenty-five to one hundred fathoms,
Most. of these deeper hauls were made at a station three to five miles
south of the entrance of Suva Harbor, and it was remarkable that the
tows drawn from one hundred fathoms were far richer, both in number
and variety of species, than were those made at twenty-five to fifty
fathoms. No precise conclusions can, however, be drawn concerning the
bathymetrical distribution of marine organisms from these results, for,
owing to the fact that the mouth of the tow-net was constantly open,
we cannot state the depth from which any given animal may have come ;
and, moreover, the near proximity of the land, and the extremely com-
plex currents and eddies that are so characteristic of this region of
Coral Reefs, probably play a far more important part in the distribution
of pelagic organisms than does the mere fact of depth. The breakers
are constantly forcing the ocean water over the reefs into the shallow
lagoons, from which it finds access again into the sea through the
openings in the walls of the reef, so that in some instances one finds «
strong current, that is in great measure independent of the tide, con-
stantly flowing outward from the lagoon into the open ocean. Such

currents acting in conjunction with the tidal flow cause extensive eddies

that result in a very ununiform distribution of pelagic life.
Besides our hauls in the neighborhood of Suva we made others off
Taviuni, Kimbombo, Vanua Mbalavu, Ngamia, Wailangilala, Totoya,

rm

Mbatiki, Kandavu, and Nukulan Islands. The hauls were rich in

Crustacea, but deficient in the number of Worms, Eel 1 larva,

ferior when com-

and Medusee. Indeed, our success was remarkably in
VOL. XXXII. — NO. 9. 1

158 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.
pared with what we should expect to find in the sub-tropical regions of
the Atlantic Gulf Stream. We obtained seven genera of Scyphomedusæ,
twenty of Hydromedusæ, nine of Siphonophoræ, and two Ctenophora.

With the exception of two Rhizostomæ, all of the genera are repre-
sented by species found in the Atlantic Ocean. Indeed, the affinity
between the Meduse of the Fiji Islands and those of the West Indies is
remarkably close, and in six cases we are unable to distinguish any
specific differences between the Fijian species and well known Atlantic
forms, and therefore venture to assert that they are specifically identical.
The following table illustrates the Atlantic distribution of these Medusæ
that are found also in the Fiji Islands.

1 Halitiara formosa, Fewkes. Dry Tortugas Islands, Florida,

2 Pandea violacea, nov. sp. Dry Tortugas Islands, Florida.

3 Rhegmatodes floridanus, L. Agassiz. Bahamas, Gulf of Mexico,

4 Mginella dissonema, Haeckel. Canary Islands; Dry Tortugas
Islands.

5 Agalma Pourtalesii, nov. sp. Dry Tortugas Islands, Florida,

6 | Abyla (Abylopsis) quincunx, Chun. Tropical Atlantic, Dry Tortugas
Islands, Florida,

In the following ten genera the Fijian form is represented in the
Atlantic Ocean by a very closely allied species: Linerges, Nausithoé,
Tamoya, Cunina, Aglaura, Gonionemus, Mitrocoma, Tiaropsis, Eutima,
and Sphæronectes. Indeed, the acalephan fauna of the Fiji Islands,
if one excepts the Rhizostomæ, is more closely related to that of the
Dry Tortugas Islands, Florida, than is the latter to the fauna of the
Mediterranean Sea.

In this connection it is interesting to notice that A. Agassiz has
shown that the deep sea fauna of the Gulf of Mexico and the Caribbean
Sea is far more closely allied to that of the Pacific than it is to that of
the Atlantic, and this is accounted for upon the supposition that before
the Cretaceous period the Gulf of Mexico and the Caribbean were in freer
communication with the Pacific than with the Atlantic. Again, in
1892,? the same author found that in nearly all the groups of deep sea

gassiz, A., 1892, Bull. Mus. Comp. Zoöl., Vol. XXIII. pp. 74-82.

|
|
|
|

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 159

Fishes, Crustaceans, Mollusks, Worms, Echinoderms, and Polyps dredged
off the west coast of Central America there were familiar West Indian
types, or east coast forms.

A recent survey of the region of the Isthmus of Panama has been
made by Hill,t who considers that it is impossible to make any serious
deductions concerning the relations of North and South America during
epochs preceding the Jurassic period, owing to the lack of data. He
also concludes (p. 261) that the waters of the Atlantic and Pacific were
probably as completely separated by a great continental land barrier in
Cretaceous times as they are to-day, a proposition fully as tenable as the
opposite hypothesis that they were united. If the marine passage ever
existed across the I“thmus of Panama, or elsewhere in Tropical America,
it must have been during the later Eocene period, and this strait was
probably of a shallow and restricted character, and had finally disap-
peared before the close of the Miocene period.

If straits of considerable width and depth have ever connected the
waters of the Tropical Atlantic with those of the Pacific, it is probable
that the great Equatorial current would pour through them from the
Atlantic into the Pacific, and thus the Pacific Ocean would become im-
pregnated with Atlantic species. Once having gained access to the
Pacific, the westerly Equatorial set would soon distribute the pelagic
animals widely over the ocean,

It is interesting to notice that while so many characteristic types of
Tropical Atlantic Medusæ are also found in the Pacific, the most emi-
nently characteristic Tropical Pacific genera, the Rhizostome have re-
n the Atlantic.? This, indeed, is what we

markably few analogues
should expect as a result of the westerly set of the great Equatorial
enrrent that would freely sweep animals from the Atlantic into the
Pacific, but would in like measure hinder the entrance of Pacific forms
into the Atlantic.

Of the thirty-eight species of Acalephs found by us in the Fiji Islands,

twenty-six are new to science,

1 Hill, R. T., 1898, The Geological History of the Isthmus of Panama, etc.
3ull. Mus. Comp. Zovl., Geological Series, Vol. NN VIII. No. 5, pp. 151-285,
Plates L-XIX.

2 In this connection, see Lendenfeld, R. von, 1884, The Geographical Distribu-
tion of Australian Seyphomedusæ, Proc. Linn. Soc, New South Wales, Volv Ta.
pp. 421-433. Also, Vanhoffen, E., 1888, Bibliotheca Zool., Bd. I. Heft 3, p. 46,
Map.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

DESCRIPTION OF SPECIES.

I. HYDROMEDUS i.

Halitiara formosa, Fuwxrs,

Halitiara formosa, Fewxes, J. W., 1882, Bull. Mus. Comp. Zoöl., Vol. IX, p. 276,
Plate 1V. Fig. 2.

This Medusa is found in both Atlantic and Pacific Oceans. It is exceedingly
abundant throughout the summer at the Dry Tortugas Islands, Florida, and
we also found it in considerable numbers late in December in Suva Harbor,
Viti Levu Island, Fiji Islands. The entoderm of the proboscis and tentacle
bulbs of the male is usually dull brown or drab, while in the female it is bright
green.

Pandea violacea, nov. sp.

This Medusa is common throughout the summer at the Dry Tortugas Islands,
near the entrance to the Gulf of Mexico. We also obtained several specimens
in Suva Harbor, Fiji Islands, early in January, 1898. We hope soon to pre-
sent a figure of the Medusa in a paper upon “ Medusa from the Dry Tortugas,”
that is to be published in the Bulletin of the Museum of Comparative
Zoölogy. At present we will content ourselves with a description of the
animal.

Generic Characters. Pandea, Lesson, 1843. Tiaridæ with numerous tenta-
cles, 8-16 or more arranged in a single row. The outer surfaces of the tentacle
bulbs bear ocelli, There is no peduncle to the proboscis. The upper edges
of the proboscis are bound to the 4 radial tubes by means of 4 mesenteries.
There are 4 simple gonads, with smooth outer surface.

Specific Characters. The bell is 4 mm. in height, and is pear-shaped. The
bell walls are of only moderate thickness. There are 32 tentacles. Tight of
these are large, being about three times as long as the bell height ; and 24 are
rudimentary. The bulbs of the large tentacles are hollow. Each and every
tentacle bulb bears a single ocellus, making in all 32 ocelli. The velum is
well developed. The proboscis is flask-shaped, its proximal portion being dis-
tended by the 4 genital glands, The lips are simple and cruciform. There
are 4 straight radial tubes and a broad circular tube. The color of the ento-
derm of the proboscis and tentacle bulbs is delicate pink or light purplish
brown. A green streak runs along the outer surface of the entoderm of each
of the 4 radial canals. The ocelli of the tentacle bulbs are purple in color.

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 161

Pennaria vitrea, nov. sp.
Plate 1, Figs. 1, 2.

Generic Characters. Pennaria, Goldfuss, 1820. Four rudimentary marginal
tentacles that are rednced to mere basal bulbs, No ocelli. Four radial tubes.
Gonads within the proboscis. No peduncle nor oral appendages to the
proboscis.

Specific Characters. The bell is 3 mm. in height; and the walls are thick
and rigid. There are 4 rudimentary tentacle bulbs. The velum is not promi-
nent. The 4 radial canals are straight and narrow. The proboscis in the
female specimen (Fig. 1) was retracted within the cavity of the bell, but in the
male (Fig. 2) it was flask-shaped, and projected for a little distance beyond
the velar opening of the bell. These conditions, however, may be merely
different states of contraction and not constant differences ; but, as we observed
only two individuals, one a male and the other a female, we cannot be certain
upon this point. The ova are large and pyriform, and are grouped in 4 radi-
ally arranged clusters within the proboscis. The mouth opening of the pro-
boscis is very simple, and there are no fimbriations or appendages.

Prominent cireular muscles were observed in the ectoderm of the cavity of
the bell in the female, but these were not seen in the case of the male. It is
possible, however, that they become apparent only during certain states of con-
traction. In the female the ove and tentacular bulbs were flesh-colored, and
the entoderm of the mouth of the proboscis was green. In the male the ten-
tacular bulbs were green, the entoderm of the proboscis pink, and the lips
green.

Found off Kimbombo Island, November 25, and off Mbatiki Island, Decem-
ber 5, 1897.

Cyteeis vulgaris, nov. sp.
Plate 2, Figs. 3, 4, 5.

Generic Characters. Cytæis, Eschscholtz, 1829. Margellid® with simple
unbranched oral tentacles upon the proboscis, and with 4 radially situated
marginal tentacles,

Specific Characters. ‘The bell is thin and slightly pyriform in shape, and is
3 mm. in height, The 4 marginal tentacles are somewhat stiff,.and are usu-
ally seen curled upward in a semicircular curve. The tentacle bulbs are large
and prominent. The velum is well developed. There are 4 thin straight
radial tubes. The proboscis (Figs. 3, 4) is pyriform, and possesses a slightly
developed peduncle. Sixteen simple oral tentacles surround the mouth. The
entodermal cells of these tentacle are disk-shaped, and their free ends are armed
each with a battery of nematocysts (Fig. 5). The mouth opening is simple,
and, excepting for the oral tentacles, there are no other appendages. The color
of the entoderm of the tentacle bulbs is dark salmon-red, as is also the entoderm

162 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

of the proboscis near the peduncle, ‘he remainder of the entoderm of the
proboscis and of the tips of the marginal tentacles is green.

Young Medusa. The youngest Medusa observed by us had only 4 oral ten
tacles. The red color of the basal bulbs of the marginal tentacles extended for
a considerable distance upward along the radial tubes. The bell was 1 mm. in
height.

This form is very abundant all over the Fiji Islands in November and De-
cember. We came across a great swarm of these Medusæ within the crater of

Totoya Island.
Bougainvillea fulva, nov. sp.
>
Plate 2, Fig. 6.

Generic Characters. Bougainvillea, Lesson, 1843. Margellidæ with dendriti-
cally branching oral tentacles, and with 4 radially arranged bunches of mar-
ginal tentacles, Proboscis wide and quadratic in cross section. Gonads
developed in the ectoderm of the proboscis,

Specific Characters. The bell is pyriform, and 2.5 mm. in height. The bell
walls are of moderate thickness, There are 4 radially arranged bunches of
marginal tentacles, each one of which consists of 3 tentacles. A dark pur-
ple ocellus is found in the ectoderm of the tentacle bulbs at the base of each
tentacle. The velum is well developed, There are 4 straight narrow radial
canals. The proboscis is wide and quadratic in cross section. There are 4
oral tentacles, each one of which branches dendritically twice. The gonads
are found in the ectoderm of the proboscis. The entoderm of the tentacle
bulbs and of the proboscis is flesh-colored.

Single specimen, obtained in towing with an open net at 100 fathoms, three

miles south of the mouth of Suva Harbor, December 16, 1897.

Laodicea marama, nov. sp.
Plate 3, Figs. 7, 8.

Generic Characters. Laodicea, Lesson, 1848. Leptomeduse without otocysts,
and with 4 simple unbranched radial canals, upon which lie the gonads, Ten-
tacles numerous. The tentacle bulbs are well developed, and many of them
possess ectodermal ocelli, There are clubs and cirri upon the bell margin,
between the tentacles.

Specific Characters, The bell is quite flat and disk-shaped, and is 5.5 mm. in
diameter. There are about 50 long, slender, flexible tentacles, each one of which
possesses a well developed basal bulb. A single dark purple ocellus is found
upon the inner side of the bulb of most of the tentacles. This ocellus is situ-
ated in the ectoderm, There are numerous clubs and cirri (Fig. 8) between
the tentacles, The velum is prominent, There are 4 narrow radial canals, in
the upper portions of which, near to the proboscis, the gonads are found. The
proboscis is short and slender, and the lips not prominent, The entoderm of

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 163

the radial tubes and of the basal bulbs of the tentacles is opaque and slightly
bluish in color. The entoderm of the proboscis is often green.

This species is common in Suva Harbor in December. It is closely allied to
Laodicea ulothrix, Haeckel, of the Canary Islands and Bahamas.

Laodicea fijiana, nov. sp.
Plate 3, Figs. 9, 10.

The umbrella is thin and bell-shaped, and 6 mm. in
diameter. There are about 70 long, slender, marginal tentacles, the free ends
of which are usually coiled in a close helix. These tentacles are very fragile,
and in the adult Medusa most of them are found to have been broken off near
to the basal bulbs. In the young jelly-fish, however, they are usually seen in
a perfect condition. A single dark brown pigment spot, or ocellus, is found in
the ectoderm on the inner side of the basal bulbs of about three quarters of the
tentacle. There are no cirri between the tentacles, but there are about 8 clubs
The lower portions

Specific Characters.

that occupy this situation. The velum is well developed.
of the 4 radial tubes, near to the circular canal, are straight and narrow; but
their upper portions, near to the proboscis, exhibit complex diverticula. The
gonads are situated upon this complexly developed portion of the radial canals;
and in the female (Figs. 9, 10) the ova are prominent, and project outward in
grape-like clusters over the surface of the genital organs. The proboscis is
short and quadratic in cross section, and there are 4 well developed fimbriated
lips. The entoderm of the proboscis and radial and circular canals is opaque
and slightly blue in color,
Jommon at Suva, Viti Levu Island, in December.

Eutimeta levuka, nov. sp.
Plate 9, Figs. 30, 31.

Generic Characters. Eutimeta, Haeckel, 1879. Leptomeduse with 8 otocysts,
and 8 tentacles, and with marginal cirri. The proboscis is borne upon a long
peduncle. The gonads are found upon the 4 radial canals,

Specific Characters. The bell is thin, and flatter than a hemisphere. It is
8 mm. in diameter. There are 8 well developed hollow tentacles. Four of
these are about as long as the diameter of the bell, and the 4 others are only
about one half of this length. Small lateral cirri (see Fig. 31) are found upon
the sides of these tentacles, Tn addition to the large tentacles already described,
there are 24 small papille upon the bell margin, and these are flanked by
lateral cirri exactly as are the large tentacles. There are 8 otocysts, each one
of which contains 3-5 otoliths. The velum is large.’ The 4 radial canals are
straight and narrow, and the gonads are found upon them near to the circular
vessel. The peduncle of the proboscis is slender, and 14 times as long as the

164 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.
diameter of the bell. The proboscis is flask-shaped, and there are 4 flanged
lips. The genital organs, tentacles, and proboscis are slightly opaque and
bluish in color.

Several specimens, Suva Harbor, January, 1898.

Staurodiscus nigricans, nov. sp.
Plate 4, Figs. 11, 12,

Generic Characters. Staurodiscus, Haeckel, 1879. Leptomeduse without oto-
cysts, but with clubs between the tentacles, There are 4 radial canals, each of
which gives rise to a pair of lateral canals. The gonads are situated upon the
canals,

Specific Characters. The bell is thin and flexible, and is 14 mm. in diameter.
There are 12 short tentacles having well developed basal bulbs. Between each
successive pair of tentacles there are 7 knot-like protuberances upon the bell
margin, each one of which bears a sensory club. The velum is insignificant in
size. Each of the 4 radial canals gives rise to a pair of lateral branches ; and
thus 12 eanals reach the circular vessel. The gonads are developed upon these
canals (see Fig. 11). The proboscis is very short, and the mouth opening wide.
The gelatinous substance of the bell has a brownish tinge. The genital organs,
basal bulbs of the tentacles, and the proboscis are light sepia in color. The
bulbs of the sensory clubs are dark brown,

Single specimen, captured in an open tow-net that was drawn from 100
fathoms, three miles south of the entrance of Suva Harbor, December 11,
1897.

This form differs from the two Atlantic species of Staurodiscus described by
Haeckel (1879, Syst. der Medusen, pp. 145, 146) chiefly in that the side branches
of the 4 radial canals reach the circular canal, whereas in the Atlantic forms

they end blindly.

Gonionemus suvaensis, nov. sp.
Plate 5, Figs. 14-16.

Generic Characters. Gonionemus, A. Agassiz, 1865. Trachomeduse, not
Leptomedusæ (see Haeckel, 1879, Syst. der Medusen, p. 146), with numer-
ous sucker bearing tentacles that are all similar each to each. There are oto-
cysts situated between the tentacles. The gonads are sinusoidally folded, and
are situated upon the 4 radial canals.

tion, and the lips are prominent.
Specific Characters. The bell is flat and of moderate thickness. It is about
three times as broad as it is high, and is 8 mm. in diameter, There are about
70 long, stiff tentacles. The proximal portion of each tentacle is straight, but
near the free end there is a small sucking disk, and beyond this the tentacle
makes a sharp bend (see Figs. 14, 15). There are about 16 otocysts, 4 in

The proboscis is cruciform in cross sec-

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 165

each quadrant. The velum is well developed. There are 4 straight radial
tubes, upon the lower portions of which, near to the circular tube, the gonads
are situated. The gonads are folded in a sinusoidal curve alternately to the
right and Jeft of the radial tube. The proboscis is cruciform in cross section
(see Fig, 16), and the lips are prominent. Green pigment spots are found in
the ectoderm of the basal bulbs of the tentacles, and also upon the radial tubes
close to their junction with the proboscis (Fig. 16). The entoderm of the
radial tubes in the region of the gonads is tinged with green. The ectoderm
of the bell margin is of a delicate rose color, and the proboscis and gonads are
brown.

This Medusa was common in Suva Harbor late in December. It is more
closely allied to the species described by Murbach + from Wood’s Hole, Massa-
chusetts, than it is to Gonionemus vertens, A. Agassiz,? of the Gulf of Georgia,
Washington.

Aglaura prismatica, Maas.
Plate 4, Fig. 13.

Aglaura prismatica, Maas, O., 1897. Mem. Mus. Comp. Zoöl. at Harvard Coll,
Vol. XXIII. No. 1, p. 24, PL III. Figs. 4, 5.

Lessonia radiata? Eydoux, F., et Souleyet, L., 1841-52, Voyage de la Bonite,
Vol. II. p. 643, Zoöphytes, Pl. II. Fig. 16.

Feneric Characters. Aglaura, Péron and Lesueur, 1809. Aglauridæ in which
the 8 gonads are situated upon the pedunele of the proboscis, directly over the
points of juncture of the 8 radial canals with the gastric portion of the pro-
boscis. There are numerous tentacles, and 8 otocysts.

Specific Characters. The bell is about 3 mm. in height, and about as broad
y thin. The side walls

as it is high. The walls, although rigid, are exceedingl;
of the bell are vertical, and the top is quite flat with, however, a slight apical

projection. There are about 60 tentacles that are so fragile

broken off short in every specimen observed by us. There are €

E.
Ka

a single otolith. The velum is very and powerful, anc

J ra P,
dusa is enabled to dart throu

means of its rapid movements that the

water. The 8 radial tubes are straight, and very narro

flask-shaped, and provided with a well developed

prominent cruciform lips. The 8 sausage-shaped
from the sides of the peduncle at the point of
with the ric portion of the probos« The

proboscis are usually brownish red in color. DOMIE
almost transparent. The Medusa was very common
Maas, 1897, has figured a Medusa from the ( i
to be identical with our Fijian form. The diffe
1 Murbach, L., 1895, Journ. Morphol., Boston, Vol. XT. p. 498.

2 Agassiz, A., 1865, North American Acalepha, p. 123, Figs. 197-200.

166 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

those of Maas may readily be accounted for by the circumstance that his figures
were drawn from preserved specimens that were probably distorted by con-
traction, A very closely allied species is found at the Dry Tortugas Islands,
Florida,

Liriope hyalina, nov. sp.
Plate 9, Fig. 32.

Generic Characters. Liriope, Lesson, 1848. Geryonide with 4 gonads upon
the 4 radial canals. The circular canal is simple, and without blind, centripe-
tal branches. There are 8 permanent tentacles; 4 of these are long, hollow,
and radially situated, and 4 are short, solid, and interradial. Eight otocysts ;
4 radial, and 4 interradial,

Specifie Characters. The bell is about 14 times as broad as high, and the sides
are straight and sloping. It is 6.5 mm. in diameter. ‘There are 8 tentacles;
4 of these are radial, and are about as long as the diameter of the bell. They
are hollow, and are covered with rings of nematocysts. The other 4 tentacles
are interradial and very short, and are carried curled sharply upward.

There are 8 otocysts (4 radial and 4 interradial), each containing a single
spherical otolith. The velum is prominent. The radial canals are wide in the
neighborhood of the circular vessel, where the gonads are found. In the upper
portions of their length, however, near the proboscis, they are straight and
slender. The proboscis projects for a considerable distance beyond the velar
opening. The mouth opening is surrounded with nematoeysts. This Medusa
is extremely hyaline, excepting that the entoderm near the mouth of the pro-
boscis is slightly rose colored.

This form was found off Tavinni Island and in Suva Harbor. It is closely
allied to Liriope seutigera, MeCrady, of Charleston Harbor and the West
Indies,

Æginella dissonema, HAECKEL.
Æginella dissonema, Haeckel, B., 1879, Syst. der Medusen, p. 840, Taf. XX. Fig. 16.

This Medusa was found by us in the Fiji Islands. Haeckel describes it from
the Canary Islands, and we have found it at the Dry Tortugas, Florida. Ina
‚paper that will soon be published in this Bulletin, we hope to present a figure

OL 16
1 Cunina octonaria, McCrapy.

Cunina octonarta, MeCrady, J., 1857, Gymn. Charleston Harbor, p. 109, Pl. XIL,
Figs. 4, 5. Also Proc. Elliot Soc., Vol. I. Pl. 1V.-VIL

Several specimens of a Cunina that is closely allied if not identical with
Cunina octonaria of Charleston Harbor, South Carolina, were found by us in

the Fiji Islands early in January, 1898. The Fijian form may be slightly less

|
|
i
|
|
|
|
i

|
|
i
j

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 167

highly colored than MeCrady’s species; the entoderm of the proboscis exhibit-
ing no trace of the sage-green color that usually characterizes the Atlantic form.
On this account we are in doubt as to whether the two forms are identical
species.
Oceania pacifica, nov. sp.
Plate 5, Fig. 17.

Generic Characters. Oceania, Péron and Lesueur, 1809. Eucopidæ with nu-
merous otocysts scattered irregularly between the numerous tentacles. No
marginal cirri. The 4 gonads are developed upon the 4 radial canals. "There
is no peduncle to the proboscis.

Specifice Characters. The bell is of moderate thickness, but is very flexible.
It is hemispherical in shape, and is 6 mm. in diameter. There are 16 thin
flexible tentacles of moderate length. The tentacle bulbs are large. There are
two otocysts between each successive pair of tentacles, and each of these oto-
cysts contains a single spherical otolyth. The velum is well developed. There
are 4 straight, narrow, radial canals, in the middle regions of which the gonads
are developed. The proboscis is short, and possesses 8 simple lips. The ento-
derm of the tentacle bulbs, of the proboscis, and of the radial tubes in the
region of the gonads is emerald green.

Several specimens found at Suva and at Nukulau Island.

Oceania ambigua, nov. sp.
Plate 6, Figs. 18, 19.

The bell is pyriform, and 4 mm. in diameter. The gelatinous substance is
very thick. There are 16 short tentacles with large basal bulbs, There are
either one or two otocysts between each successive pair of tentacles. Each
otocyst contains a single spherical otolith, The velum is well developed, There
are 4 straight, moderately wide radial canals. The gonads are situated upon
these canals near to the proboscis. The proboscis is flask-shaped, and there
are four simple lips.

The entodermal axis of each tentacle is brown in color, and the ectoderm of
the proboscis, gonads, and tentacles is green.

Single specimen found at Suva, January 4, 1898.

Clytia polynesiæ, nov. sp.
l Plate 6, Fig. 20.
Generic Characters. Clytia, L. Agassiz, 1862. Ewcopidæ with 16 tentacles
alternating with 16 otocysts. Gonads upon the 4 radial canals. No peduncle

to the proboscis.
Specific Characters. The bell is thin and hemispherical in shape, and is
pect pe,

5 mm. in diameter. There are 16 short, sharply coiled, marginal tentacles,

168 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

with large basal bulbs. Sixteen otocysts, each containing a single spherical
otolith, alternate with the 16 marginal tentacles. The velum is well devel-
oped. There are 4 straight, narrow, radial canals, upon the lower region of
which, near the circular canal, the gonads are developed. In the female the
ova are very large and prominent. The proboscis is small and flask-shaped,
and there are 4 curved lips. The entoderm of the proboscis, tentacle bulbs,
and radial canals in the region of the gonads is green.
Several specimens found in Suva Harbor early in January.

Tiaropsis rosea, nov. sp.
Plate 7, Figs. 21, 2%

Generic Characters. Tiaropsis, L. Agassiz, 1849. Eucopidæ with 8 otocysts
and 8 pigment spots (2 in each quadrant). There are numerous otoliths within
each otocyst. The gonads are developed upon the 4 radial canals. There is
no peduncle to the proboscis.

Specific Characters: Young Medusa, The bell in the young Medusa ob-
served by us was ellipsoidal in shape, and 2.5 mm. in height. There were 4
well developed marginal tentacles that were coiled in,a close helix. The bulbs
of these tentacles were large. In addition to the tentacles mentioned above,
there were 4 small protuberances upon the bell margin that probably repre-
sented the beginnings of other tentacles. There were 8 marginal sense organs
(Figs. 21, 22), 2 in each quadrant. These consisted (Fig. 22) of an entodermal
pigment spot, and a number of otoliths contained in an open fold of the velum,
A cross section (Fig. 23) of the sense organ of Tiaropsis diademata, L. Agassiz,
will serve to illustrate their structure. In Figure 23 the entoderm is repre-
sented in brown and the ectoderm in gray; the section is taken through the
bell margin. It will be seen that the pigment spot (pg.) is situated within the
entoderm upon the inner side of the circular tube (ct.); and that the otoliths
(ot.) are ectodermal and enclosed within an open fold of the velum (vel.).

There were 4 straight broad, radial tubes. The proboscis was small, and
there were 4 prominent, fimbriated lips. The gonads had not yet made their
appearance. A number of immature specimens were found in Suva Harbor
early in January.

A species that is closely allied, if. not identical, with this Medusa is found at

the Dry Tortugas Islands, Florida.

Mitrocoma mbengha, nov. sp.
Plate 8, Wigs. 24, 25.

Generic Characters. Mitrocoma, Haeckel, 1864. Eucopidæ with numerous
open otocysts, and numerous tentacles and cirri, Gonads upon the 4 radial

canals. No peduncle to the proboscis.

|
|
|

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 169

Specific Characters. The bell is slightly flatter than a hemisphere, and is
9 mm. in diameter. There are 16 short tentacles with large basal bulbs. In
addition to these there are about 80 short cirri upon the bell margin. There
are 32 otocysts (2 between each successive pair of tentacles). The otocysts
each contain 5-9 otoliths. The velum is well developed. There are 4 straight,
narrow radial canals, upon the lower regions of which the gonads are devel-
oped. The proboscis is short and flask-shaped, and cruciform in cross section.
There are 4 prominent, fimbriated lips. The entoderm of the tentacle bulbs
and of the proboscis, and the ectoderm of the gonads, are dull yellow. The
entoderm of the radial canals is grass-green,

Suva Harbor, Fiji Islands, in January.

Polycanna purpurostoma, nov. sp.
Plate 8, Figs. 26-28.

Generic Characters, Polycanna, Haeckel, 1879. Æquoridæ with numerous
simple radial canals. Proboscis very wide; the mouth opening surrounded by
numerous small lappets. Gonads upon the radial canals.

Specific Characters. The bell is lens-shaped, and about three times as broad
as itis high. It is about 30 mm. in diameter. The gelatinous substance of the
bell is very thick, so that the cavity is shallow. There are 16 short tentacles
with wide basal bulbs. In addition to these there are about 120 very small
tentacles, or papillæ, upon the margin of the bell (see Fig. 28). There are
about 100 otocysts, each one of which contains two small, spherical otoliths.
The velum is prominent. There are about 100-120 straight, narrow radial
tubes, upon the lower halves of which the gonads are developed. The pro-
boscis is very wide and shallow, and the mouth opening large; there are 100-
120 small lappets surrounding the mouth (M., Fig. 27). The entoderm of the
basal bulbs of the tentacles is flesh colored. The gonads are slaty blue; and
the proboscis, especially in young Medusa, is pink. In old individuals it is
usually hyaline.

Jommon among the Fiji Islands in December.

Rhegmatodes floridanus, L. Acassız.

Rhegmatodes floridanus, Agassiz, L., 1862, Cont. Nat. Hist, U. S. A., Vol. IV. p. 361.

A Medusa that we are unable to distinguish from Rhegmatodes floridanus
of the Bahamas and Gulf of Mexico was met with occasionally in the Fiji
Islands.

Hirene kambara, nov. sp.
Plate 8, Fig. 29.

Generic Characters. Eirene, Eschscholtz, 1829, Eucopide with numerous
otocysts and tentacles, and sometimes also marginal cirri. There is a distinct

170 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

peduncle to the proboscis. The gonads are developed upon a restricted portion
of the 4 radial canals.

Specific Characters. The bell is flat with sloping sides, and is 8 mm. in
diameter. There are about 32 very small, slender tentacles, having well de-
veloped basal bulbs. There are 64 otocysts, 2 between each successive pair of
tentacles. Each otocyst contains a single spherical otolith. The velum is dis-
tinct. The 4 radial canals are straight and narrow, and the gonads occupy
their lower portions. There is a distinct peduncle. ‘The proboscis is simple,
ss 4 curved lips. The entoderm of the basal bulbs of the tentacles

and pos
and the proboscis are turquoise in color.

Single specimen found in Suva Harbor, December 29, 1897. This form
differs from Eutima pyramidalis,! L. Agassiz, of the West Indies, in that the
peduncle is smaller and the proboscis larger than in the Atlantic form.

II. SCYPHOMEDUS 4A.
Tamoya, sp.

A single specimen of Tamoya, in a very imperfect condition, was found by
us in towing with an open net at 100 fathoms, three miles south of the entrance

of Suva Harbor.
Nausithoé punctata, var. pacifica, nov. var.

Nausithoé punctata, Kölliker, A., 1853, Zeit. fiir Wissen. Zool., Bd. IV. p. 323.

This Medusa is extremely close to if not identical with Nausithoé punctata,
Kölliker. It appears to differ from the latter, however, in that the brownish
yellow spots upon the ectoderm of the ex-umbrella are not so prominent. As
this difference appears to be constant, we consider the Fijian form to be a close
variety of Nausithoë punctata. Nausithoë punctata is found in the Mediter-
ranean, and we have taken it at the Dry Tortugas Islands, Florida.

Linerges aquila, HAECKEL,
Plate 10, Figs. 33, 34.

Linerges aquila, Haeckel, E., 1879, Syst. der Medusen, p. 496.
Tn the absence of a figure and an accurate description hy Haeckel, we are
far from certain that the species about to be deseribed by us is identical with
L. aquila of the east coast of Madagascar,
Generic Characters. Tinerges, Taeckel, 1
quadrangular proboscis without mouth arms, and with simple quadratic lips.

880. Discomeduse with simple

1 We hope soon to publish a figure of Butima pyramidalis in the Bulletin of the

Museum.

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 171

There are 8 marginal sense organs, 8 tentacles, and 16 marginal lappets. There
are 16 broad radial pouches and branched sac-shaped canals within the lap-
pets. There is no ring canal. ‘There are 4 horseshoe-shaped gonads, the con-
vex proximal arches of which are divided into two adjacent wings by means of
a median septum. There are 48-52 (in this species 52) wart-like, hollow pro-
tuberances upon the floor of the sub-umbrella.

Specific Characters: Adult Medusa, The bell is 13 mm. high, and 16 mm.
in diameter. The side walls are straight and vertical, and the top flat. There
are 16 marginal lappets that are about twice as wide as they are long. There
are 8 small tentacles that are about 14 times as long as the lappets. Eight
marginal sense organs alternate with the tentacles. The entoderm of these sense
organs contains a spherical mass of otoliths (Fig. 34). The 4 gonads are crescent-
shaped, the two horns of each crescent being separated by means of a median
partition. There are 16 gastro-vascular pouches, and no marginal ring canal,
Projecting inward from the floor of the sub-umbrella into the bell cavity there
are 52 hollow wart-like protuberances. These are arranged in three rows.
The most proximal row contains 4, the middle row 16, and the most distal
row 32 of these protuberances. The proboscis is quadrangular in cross section,
and there are 4 flanging lips. The general color of the Medusa is brown.
There are 8 rows of dark brown pigment spots running longitudinally down
the inner surface of the lips of the proboscis. Separated areas of brown ento-
dermal cells are found in the distal portions of the 16 gastro-vascular pouches,

Young Medusa. Ephyre of this Medusa were common among the Fiji
Islands in November and December. They resemble the adult in general color,
but the umbrella is very flat and disk-shaped. The youngest specimen observed
by us was 2.5 mm. in diameter.

We observed the ephyræ of this Medusa off Taviuni, Ngamia, Wailangilala,
and Vanua Mbalavu Islands, and in Suva Harbor. We also came across a
large swarm of the adults in the lagoon of Wailangilala Atoll on November 20.

This Medusa differs from Linerges mercurius, Haeckel, of the West Indies in
that there are 52 wart-like protuberances upon the sub-umbrella, instead of 48,
as in the Atlantic species. Also the brown colored entodermal cells approach
nearer to the stomach, and their color is perhaps slightly duller than in the
West Indian form.

Aurelia vitiana, nov. sp.
Plate 10, Fig. 35.

Generic Characters. Aurelia, Péron and Lesueur, 1809. Discomeduse with a
simple central mouth, and 4 mouth arms or palps. The radial canals are nar-
row and branched, and there is a ring canal. There are 8 marginal sense
organs, and 8 broad, sometimes bipartite, marginal lappets, each of which
bears on its dorsal side, some distance away from the umbrella margin, a row
of numerous short tentacles, that alternate with as many dorsal lappets.

Specific Characters. The bell is hemispherical and 80 mm. in diameter.
gelatinous substance is quite thick, 'There are 8 simple marginal lappets

VOL, XXXII. — NO, 9. 2

The

172 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

that bear upon their dorsal surfaces, at a slight distance from the bell margin,
a row of numerous short tentacles. The 8 marginal sense organs are large and
deeply set within niches between the marginal lappets. Sixteen narrow chy-
miferous tubes radiate outwards from the stomach cavity, Eight of these are
straight and unbranched, and go to the middle of the marginal lappets. The
8 others give off side branches that anastomose. They go to the marginal
sense organs (see Fig. 35). The mouth arms, or palps, are short and narrow,
and do not protrude beyond the bell margin. Their free edges are lined by «
row of numerous short slender tentacles. The 4 gonads are horseshoe-shaped,
and the subgenital pits are wide and open. The gonads, palps, and tentacle
bulbs of the adult Medusa are of a delicate lilac. The bell is hyaline. Young
Meduse lack the lilac color, and are quite transparent.

This species was common upon the surface in Suva Harbor in the early
morning hours in December, when the water was smooth and calm. A slight
ripple seemed to be sufficient to cause them to sink out of sight.

RHIZOSTOM 4.

Cephea dumokuroa,! nov. sp.
Plates 11, 12, Figs. 36-39.

Generic Characters. Cephea, Péron and Lesueur, 1809. Discomeduse without
tentacles, and without a central mouth opening. There are 4 sub-genital cavi-
ties, and ventral suction cusps upon the 8 mouth arms, The mouth arms give
off short branches that are simple and not dichotomous. There are 8 ocular
canals, and numerous other radial canals. There are 8 marginal sense organs.
: Characters. The umbrella is di:

peripheri are vertical. It about 300 mm. in diameter. A large dome, cov-

shaped, and the walls near the

Spe

vrotuberances, arises from the centre of the aboral sur-

ered wit

There are about 20 protuberances upon thedome. About
| the remainder are much smaller. There are 8 mar-

1 ar aroe. AT th
>56 e large, and the

are deeply sunken within niches upon the bi IL margin

i\e entodermal chore of these sense organs terminates in a mass
?) oranules. There are about 9 very short marginal

the umbrella. There are 8 short, thick, mouth arms,

in

+h give rise to numerous short branches covered with

surround these suction mouths.

suction mouths. Rows of very small tentacles
There are 4 sub-genital pits (ypt., Fig. 38). general idea of the internal

' the medusa is given by Figure 38.

structure
cavity lying within the great central dome of the umbrella. It is completely
separated from the sub-genital porticus (s. por.) by means of the double mem
branous folds « rans (ov.). Thirty-two tubes (see Fig. 37) radi

f the genital or;

ian name for this species.

1 Dumokuro is the native F

3 The stomach (S) is a wide

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 173

ate outward from the stomach cavity. Four of these tubes are wider than
the remainder, and lead into 4 of the marginal sense organs. All of the radial
vessels are connected one with another by means of a broad ring of anastomos-
ing canals that occupies the peripheral zone of the umbrella. The 4 genital
organs (ov., Fig. 38) are stretched upon 8 fleshy radii, and serve to separate
the stomach cavity (S) from the cavity of the sub-genital porticus (s. por.).
Eight chymiferous tubes spring from the cavity of the stomach, and enter the
8 mouth arms, where they send off side branches to the suction mouths. These
tubes also give off side branches that samify and anastomose within the gelati-
nous lower floor of the sub-genital porticus. - None of these tubes, however,
enter the cavity of the sub-genital porticus, that is thus sepa rated entirely
from the gastro-vascular system of the Medusa. The general color of the gelati-
nous substance of the Medusa is blue. There are numerous deep blue streaks
upon the apex of the central dome, and a deep blue band runs Around the
outer surface of the Medusa just above the mouth arms. This band is broad
and biforked in the regions of the mouth arms. The chymiferous tubes in the
mouth arms and the radial tubes of the umbrella are deep blue. The tenta-
cles of the suctorial mouths and the anastomosing chymiferous vessels of the
umbrella are coffee colored.

We suddenly came upon a swarm of these Meduse off Vanua Mbalavu
Tsland on November 25, and in all our subsequent voyaging through the Fiji
group we never saw another specimen. The Medusæ were accompanied by a
number of small fish.

Pseudorhiza Thocambaui,' nov. sp.
Plate 13, Figs. 40-44,

Generic Characters2 Pseudorhiza, von Lendenfeld, 1884. Rhizostomæ with
a single sub-genital cavity. There is a central mouth opening upon the lower
side of the brachial disk. There are 8 mouth arms, the inner or ventral sides
of which contain a deep groove. The mouth arms bifurcate, and there is a
single long filament that arises from the place of bifurcation of each mouth arm,
making in all 8 filaments, The canal system consists of 16 main radial canals
and a ring canal. Centrifugally from the ring canal there is an anastomosing
network of canals, and centripetally there are between two adjacent main radial
canals about 10 canals running from the ring canal inwards radially.

Specific Charact

rs. The bell is hemispherical and 32 mm. in diameter.

There are 8 marginal sense organs, the sensory portion of which consists of a

spherical chore of dark brown pigment granules, surrounded by a mass of small

transparent otoliths (Fig. 41). There are 9 small marginal lappets between

each successive pair of marginal sense organs. The brachial disk, upon which
1 Named after Thocambau, king of Fiji.

2 See Lendenfeld, R. von, 1884, Proc. Linn. Soc. New South Wales, Vol. IX
p. 292. Also, 1887, Descrip. Catalogue Medusæ, Australian Museum, p. 23.

174 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the mouth arms are borne, is attached to the lower floor of the umbrella by
means of 4 pillars. There is a central mouth opening upon the lower floor of
the brachial disk. Right mouth arms arise from the brachial disk. Near their
distal ends they bifurcate, and a single long filament arises from the place of
bifurcation of each arm. Thus there are in all 8 of these filaments. These are
each about 20 mm. in length, and their surface is covered with wart-like pro-
tuberances, Numerous suctorial mouths are found upon the inner and lower
sides of the mouth arms. They are surrounded by a great, number of small
tentacles, forming a row around each mouth opening. In addition to these
small tentacles there are numerous club-shaped papille between the suctorial
mouths. The surface of these papilla is covered with clusters of nematocysts.
The 4 gonads are V-shaped, the apex of the V being pointed inwards toward
the centre of the disk. The canal system has been described under “Generic
Characters.” The color of the gelatinous substance of the umbrella is dull blue,
The sub-umbrella, brachial disk, and mouth arms are dull green, and the 8 long
filaments are deep blue. The 8 radial canals’ that run to the 8 marginal sense
organs are green in color, The remaining radial canals, huwever, are almost
colorless. There are a number of white spots upon the peripheral portions of
the exumbrella, and the wart-like protuberances upon the 8 filaments are also
white.

The medusa was common in Suva Harbor in December. It swims with
great rapidity by means of an incessant contraction and expansion of the um-
brella, This movement is accomplished by the action of a powerful system of
circular muscles in the sub-umbrella,

Young Medusa (Figs. 42-44). A young ephyra of this species was captured
in Suva Harbor on January 11, 1898. The bell was 5 mm. in diameter and
qnite flat and disk-shaped. There were 8 marginal sense organs. The central
mass of dark brown pigment granules of the sense organ was developed, but
the peripheral shell of transparent granules had not yet made its appearance
(compare Figs, 41 and 44). There were 24 marginal lappets, the 16 ocular
lappets being about twice as long as the 8 intermediate lappets. There were
16 radial pouches from the stomach. Eight of these went to the sense organs,
and 8 to the intermediate lappets. The sub-genital porticus was already pres-
ent, and the brachial disk was suspended from the floor of the sub-umbrella
by means of 4 gelatinous pillars exactly as in the adult. The ephyra pos-
sessed only a simple central mouth opening, having 4 cruciform lips. The
margins of the lips were lined with a row of short, slender tentacles, with
knob-like ends exactly like those that surround the suctorial months on the
mouth arms of the adult Medusa. No trace of the genital organs could be
detected, but the gastric cirri were represented by 12 short filaments (3 in each
quadrant). The color of the ephyra was very similar to that of the adult.

Were it not for the sub-genital porticus and brachial disk, this little Rhizos-
toma would resemble, in all respects, the young of the Semostomæ. We have
observed the ephyra of an allied genus, Stomolophus meleagris, L. Agassiz,
that was in a slightly more advanced stage than the one here figured; and in

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 175

this individual the central mouth was still used for the capture of food. It
seems probable that in those genera of Rhizostomæ that possess a single sub-
genital space (the Monodenmia of Haeckel), the brachial disk has become
separated from the floor of the sub-umbrella by the enlargement and final
coalescence internally of the 4 sub-genital pits of the Semostome.

Cassiopea ndrosia,! nov. sp.
Plate 14, Figs. 45, 46.

Generic Characters. Cassiopea, Péron and Lesueur, 1809. Discomedusæ
without tentacles and without central mouth. There are 8 pinnately or tri-
chotomously branching mouth arms, the lower or ventral surface of which is
occupied by numerous suction mouths and vesicles. There are 4 sub-genital
cavities, and 4 gonads. There are 12 or more marginal sense organs, and
numerous anastomosing radial canals.

Specific Characters. The bell is flat and disk-shaped, and 50 mm. in diameter.
The number of marginal sense organs appears to be variable. In the specimen
here figured there were 22, but in another individual there were only 18.
Whatever their number may be, they are situated at equal distances one from
another. There are 4 small sub-genital ostia. There are 8 pinnately branch-
ing mouth arms, each arm being about 30 mm. in length. Their ventral sur-
face is occupied by a great number of suction mouths surrounded by small
tentacles, In addition to these there are also a large number of leaf-shaped
vesicles scattered among the suction mouths. These vesicles are more numer-
ous near the centre than they are at the free ends of the arms. The umbrella
possesses two powerful sets of radial muscle bands. Twenty-two of these (in
specimens with 22 marginal sense organs) are situated in the floor of the sub-
umbrella, and radiate outwards toward the marginal sense organs.” An equal
number are situated in the exumbrella, and alternate in position with the
set in the sub-umbrella. The muscle bands of the floor of the sub-umbrella
usually appear opaque and white in color, especially in states of contrac-
tion. The muscles of the exumbrella, on the other hand, are deep blue-
green. The general color of the umbrella is ashy brown. A large spearhead-
shaped white spot is situated upon the radius of each sense organ at a little
distange inward from the bell margin. Thus in a medusa with 22 marginal
sense organs there are 22 of these large spots. The apex of each of these spots
is directed outward toward the sense organ. Tn addition to these large spots
there are usually 4 short white radial streaks between each pair of adjacent
sense organs. There are numerous white spots upon the inner portions of the
sub-umbrella, The upper fleshy portions of the 8 mouth arms are white or
slightly grayish, The tentacles surrounding the suction mouths are deep brown
in color, and the leaf-shaped vesicles are olive-green.

1 Ndrosi is the native Fijian name for this species.

176 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

We found a specimen of this Medusa in Suva Harbor, November 14, 1897.
And we also came upon a small swarm of them at Komo Island on November
27, 1897.

On January 3, 1898, we observed an individual of this species resting upon
the muddy bottom of Suva Harbor. The aboral surface of the umbrella was
pressed against the bottom, while the mouth arms and oral surface were
uppermost. In this position it remained quiescent for more than an hour,
merely waving its mouth arms in a slow, sweeping manner. Indeed, its ap-
pearance reminded one far more of an Actinian than of a Medusa. It is inter-
esting to observe that the common Cassiopea frondosa of the West Indies, and
also Cassiopea Mertensii, Brandt, of the Caroline Islands possess similar habits,

The number of marginal sense organs in this species is certainly quite vari-
able, and is greater than has as yet been observed in any other species of the
genus Cassiopea. We prefer, however, on account of its close resemblance in
all other respects, to place it in the genus Cassiopea.

III. CTENOPHOR4,

Eucharis grandiformis, nov. sp.
Plate 15, Figs. 47, 48.

The body of this Ctenophore is 135 mm. in length, and our figures represent
the natural size of the animal. The lappets in this species are by no means so
wide and voluminous as in the Mediterranean Eucharis multicornis,! Esch-
scholtz. The whole outer surface of the body and lappets is covered with
numerous long papillæ. The auricles are long, and are often carried coiled in
a close helix (see Fig. 48). A pair of long unbranched tentacles arise from
either side of the body near the region of the mouth. The ciliated plates are
very numerous and close together, and are arranged in 8 rows, The oral sense
organ is situated at the bottom of a deep cleft. The windings of the canal sys-
tem through the lappets are far less complex than is the case in E. multicornis.
The ciliated plates, gastric cavity, and chymiferous tubes are cinnamon-yellow
in color. All other parts of the animal are hyaline. This Ctenophore is ex-
tremely delicate in structure, the least touch being sufficient to tear the tissues
of the animal.

We found them in considerable numbers, floating near the surface, in Suva
Harbor, early in the mornings of December when the water was perfectly
calm. The least ripple caused them to sink to an unknown depth.

1 Compare our figures with those of E. multicornis by Chun, C., 1880, Fauna
und Flora des Golfe von Nezpel, I. Monographie, p. 206, Taf. V. Figs. 1-3,

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 177

Beroë australis, nov. sp.
Plate 16, Figs. 49, 50.

The animal is 40 mm. in length, and the body is compressed laterally, one
side being about three times as broad as the other. Our figure exhibits the
broad side. The aboral otocyst is surrounded by a figure 8-shaped row of
branched papillæ. The long axis of this row lies in the plane of the broad
side of the Ctenophore. There are 8 rows of ciliated plates that are bordered
by masses of stellate pigment cells (see Fig. 50). Fight radial canals run be-
neath the ciliated plates, and in addition to these there are two wide, straight
lateral tubes (l., Fig. 49). All of these tubes communicate with the circular
canal (c., Fig. 49) ; and are also put into further communication, one with an-
other, by means of a simple network of connecting tubes (k., k., Fig. 49). The
month opening is very wide and deep. The pigment cells of the ciliated plates,
and of the papille surrounding the aboral sense organ, are erimson-lake in
color. The gelatinous substance of the animal is quite transparent.

Many specimens of this Ctenophore were found in Suva Harbor in December.
The animal swims with remarkable rapidity by means of the movements of its
combs of cilia.

IV. SIPHONOPHORZ.

Physalia utriculus, Escnscnorrz.

Physalia utriculus, Eschscholtz, F., 1829, Syst. der Acalephen, p. 163, Taf. XIV.
Figs. 2, 8.

Physalia australis, Lesson, R. P., 1850, Voy. de la Coquille, Zoophytes, p. 38,
PI. V. Fig. 1.

Physalia utrleulus, Haeckel, E., 1888, Challenger Report, Zoólogy, Vol. XXVIII.
p. 351.

Several specimens of this Siphonophore were captured by us among the
Fiji Islands. It is well represented by Lesson’s figure.

Sphæronectes Köllikeri, Huxrer.
Diplophysa Köllikeri, HAECKEL.

S. Köllikeri, Plate 16, Figs. 51, 52. D. Köllikeri, Plate 7, Fig. 53.

Spheroneetes Köllikeri, Huxley, T. H., 1859, Oceanic Hydrozoa, Ray Soc. Publica-
tion, p. 50, Pl. III. Fig. 4.

Spheronectes Köllikert, Haeckel, E., 1888, Challenger Report, Zoölogy, Vol.
XXVIII. pp. 130, 861.

Diplophysa Kollikeri, Haeckel, E, 1888, Challenger Report, Zoology, Vol. XXVIII.
pp: 107, 180, 359.

178 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Generic Characters (Polygastric generation = Sphæronectes). Polygastric
Calyconecte with a single, rounded, edgeless, subspherical, swimming bell.
There is a complete tubular hydroecium on the ventral side of the swimming
bell, from the inner apex of which arises the long tubular hydrosoma. The
groups of units are eudoxiform, and separated by free internodes. Each feed-
es a single covering scale.

ing polyp poss

Generic Characters (Monogastric generation = Diplophysa). Monogastric
Calyconectæ, representing a single group of units, consisting in a feeding
polyp with tentacle and covering scale, and a fertile gonophore that serves also
as a swimming bell. The covering scale is hemispherical, or subspherical, and
possesses a simple ovate or cylindrical canal (phyllocyst) within its ventral
axis.

Spheronectes Köllikeri. — The swimming bell is about three quarters of a
sphere in shape, and is about 10 mm, in diameter. The cavity of the bell is
shallow; and there is a large and p@verful velum, by means of the movements
of which the animal is enabled to swim. The bell possesses a circular canal,
and 4 narrow, somewhat crooked, radial canals (r, 7, r, r, Fig. 51). These
communicate, by means of the narrow duct (c), with the gastro-vascular cavity
of the hydrosoma (h). There is a straight spindle-shaped vacuolated vesicle (f)
buried within the gelatinous substance of the nectophore. The hydrosoma (4)
arises from the inner end of a long, narrow invagination of the outer wall of
the swimming bell (the hydræcium). The order of appearance of the various
organules upon the hydrosoma is shown in Figure 52, The first to develop
are the feeding polyps (p); then follow, in order, the tentacles (1), the cover-
ing scale (c. s.), and gonophore (g). The hydrosoma attains a length of about
50 mm., and there are numerous groups of units (cormidia) separated hy free
internodes. One of these groups of units that has very recently become sepa-
rated from the hydrosoma, and is therefore in the Diplophysa stage, is repre-
sented, highly magnified, in Plate 17, Figure 53.

Diplophysa Köllikeri (Plate 17, Fig. 53). — This is merely the free Mono-
gastric or Eudoxia form of Sphæronectes Köllikeri, and consists of a single
group of units that has become separated from the hydrosoma of the latter
animal, and leads an independent existence. The covering scale (o. s.) is thick
and hemispherical in shape. It contains a simple ovate canal or phyllocyst
(phe.). There is a single feeding polyp (p), a single tentacle (1), and a gono-
phore (g). The tentacle gives rise to many small filamentous side branches,
which terminate in nematocystic bulbs. The gonophore serves also as a swim-
ming bell, and its manubrium will become much larger than is shown in
Figure 58, and will contain the genital products. The tube ad. is the means
by which the animal was once attached to the hydrosoma of S. Köllikeri. This
tube soon atrophies. The entoderm of the feeding polyps of the manubrium
of the gonophore and of the nematocystic bulbs of the tentacles is rich yellow
or orange. This animal, in the Sphæronectes stage, was met with in various
places among the Fiji Islands. Our drawings are derived from a specimen
found upon the surface of Suva Harbor, December 12, 1897.

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 179

This species is very closely related to Sphæronectes gracilis, Claus, of the
Mediterranean and Tropical Atlantic! We have taken $. gracilis at the Dry

Tortugas Islands, Florida.
Le) ’

Diphyopsis angustata, HAECKEL.
Plate 17, Fig. 54.

Diphyes angustata, Eschscholtz, F., 1829, Syst. der Acalephs, p. 186, Taf. 12,
Fig. 6.

Diphyopsis angustata, Haeckel, E., 1888, Challenger Report, Zoöl., Vol. XXVIII.
pp. 152, 363.

Generic Characters. Diphyopsis, Haeckel, 1888. Diphyide with two angu-
lar, slenderly pyramidal swimming bells of similar form and subequal size,
one placed behind the other. The first swimming bell possesses a complete
infundibular hydroeium on its ventral side. The groups of units upon the
hydrosoma are Eudoxiform, and are separated by free internodes. Each feed-
ing polyp possesses a covering scale. The covering scales are spathiform, with

a deep ventral groove.

Specific Characters. The specimens found by us conform to the definition of

the genus Diphyopsis, with the notable exception that we observed no posterior
swimming bell, such as is figured by Haeckel (1888, Plate XXXIII.) in Di-
phyopsis compressa. It is possible that no such structure exists in Diphyopsis
angustata, but knowing the ease with which swimming bells are broken off and
lost, we hesitate to make such a statement. The anterior swimming bell (the only
one observed) is 37 mm. in length. The bell cavity is long and spindle-shaped,
and terminates in a long, narrow neck, in the end of which there are usually a
number of vacuolated cells containing green pigment (cp., Fig. 54). There is
a well developed velum, the sudden contractions of which cause the animal to
shoot rapidly through the water. There is a circular vessel within the swim-
ming bell, and also two long curving side tubes (r, r, Fig. 54). These com-
municate by means of a long duct (c) with the gastro-vascular cavity of the
hydrosoma. The cavity of the hydracium is about 17 mm. in depth. There
is a long spindle-shaped mass of vacuolated cells (/) extending from the inner
apex of the hydrocium into the gelatinous substance of the swimming bell.
The hydrosoma also arises from the inner apex of the hydraeium. Tt often.
attains a length of 100 mm., and there are numerous groups of units (cor-
midia) found upon it. These are separated one from another by free inter-
nodes. The cormidia all arise from the ventral side of the hydrosoma. The
first to appear are the feeding polyps. Then follow, in order, the tentacles and
the covering scales. No gonads or swimming bells were seen. Each covering
The tentacles give

scale possesses a deep open groove along its ventral side.
off simple lateral filaments that terminate in nematocystic bulbs. The ento-

1 See Chun, C., 1892, Abhandl. d. Senckenb. naturf, Ges., Bd. XVIII. p. 84,
Fig. 5.

180 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

derm of the feeding polyps is rose colored, and the nematocystic bulbs of the
tentacles are orange-yellow.

This is the largest species of Diphyid known at the present time.

Common among the Fiji Islands in December, 1897.

Abyla quincunx, Caux.

Abylopsis quincunz, Chun, C., 1888, Sitzungs Berichte Akad. Wissen. Berlin,
p. 1160.

Abyla (Abylopsis) quincunz, Chun, C., 1897, Verhandl, Deutsche Zool. Gesell., p. 71,
Fig. 13.

A large number of specimens of a polygastrie Cal yconectes that appears to be
identical with Abyla quincunx, Chun, of the Tropical Atlantic, were found by
us among the Fiji Islands. Huxley (1859, p. 58) mentions having found this
form in the Indian and Pacific Oceans, and he describes it under the name of
Abyla pentagona. Several specimens of the Eudoxia form of this species (see
Aglaisma quincunx, Chun, 1888) were found by us while in the Fiji Islands.
This Fiji form appears to be identical with one observed by us at the Dry
Tortugas Islands, Florida.

Agalma Pourtalesii, nov. sp.

We obtained this beautiful new Siphonophore near the mouth of Suva Har-
bor, Viti Levu Island. We have also found a number of specimens that appear
to be identical in all respects with the Fijian form, at the Dry Tortugas
Islands, Florida. We hope soon to present a number of figures of this animal
that is now in preparation,

33

in a paper upon “Medusæ from the Tortugas
and will appear in the Bulletin, At present we will content ourselves by pre-
senting a description of the species.

Generic Characters. Agalma, Eschscholtz, 1825. Siphosome short and rigid,
about as long as the nectosome. The whole siphosome is densely covered with
thick prismatic bracts. The dactylozooide and feeding polyps are thickly
scattered along the stem of the siphosome. The gonostyles spring from the
internodes between the dactylozooide and feeding polyps. The tentilla are
tricornuate, with a terminal ampulla and two paired horns.

Specific Characters. The entire animal is about 25 mm. in length. The
feeding polyps, tasterns, gonostyles, and tentacles all spring from the ventral
side of the siphosome. The float is of small size, and is balloon-shaped, and the
pneumatopore is surrounded by radially arranged streaks of dark red pigment,
The swimming bells are dovetailed alternately on either side of the nectosome,
so that their velar openings are found on two diametrically opposite sides of the
nectosome. The siphosome is densely covered upon all sides with thick pris-
matic bracts, The feeding polyps are somewhat stouter in shape than the
tasterns, but in other respects are quite similar to them in appearance, The

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 181

tentacles arise from the bases of the tasterns and feeding polyps. Each tentacle
gives off a number of lateral branches which terminate in a coiled nematocyst
battery, an ampulla, and two paired horns. Both male and female gonostyles
are found upon the same siphosome, and they arise from the side of the sipho-
some between the tasterns and feeding polyps. The male gonostyles are long
and slender, while the female are short and stout, and contain each one of them
a considerable number of ove. Both male and female gonostyles are borne
upon long, slender filaments that are highly contractile. The color of the ento-
derm of the hydrosoma, swimming bells, feeding polyps, and tasterns is rose-
red or pink, The nematocyst batteries upon the terminal portion of the
tentacles are dark red. ‘The gonads and bracts are colorless.

182 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

LIST OF SPECIES.

Hydromedusce.4

* /Eginella dissonema, Haeckel
*Aglaura prismatica, Maas
Bougainvillea fulva, nov. sp.
Clytia polynesiæ, nov. sp.
*Cunina octonaria, McCrady ?
Cytæis vulgaris, nov. sp.
Eirene kambara, nov. sp. +
*Butimeta levuka, nov. sp.
*Gonionemus suvaensis, NOV. Sp.
*Halitiaria formosa, Fewkes
Laodicea fijiana, nov. sp. .
Laodicea marama, nov. sp.
Liriope hyalina, nov. sp.
*Mitrocoma mbengha, nov. sp.
Oceania ambigua, nov. sp.
Oceania pacifica, nov. sp. +
*Pandea violacea, nov. sp. +
Pennaria vitrea, nov. sp.
Polycanna purpurostoma, nov. sp: .
*Rhegmatodes floridanus, L. Agassiz .
Staurodiscus nigricans, nov. sp.
*Tiaropsis rosea, nov. sp.

Seyphomeduse.

Aurelia vitiana, nov. sp.
Cassiopea ndrosia, nov. sp.
Cephea dumokuroa, nov. sp. +

*Linerges aquila, Haeckel ?

*Nausithoé punctata, var. pacifica, nov. var.
Pseudorhiza Thocambaui, nov. sp.
Tamoya, sp.

Ctenophore.

Beroé australis, nov. sp.
Eucharis grandiformis, nov. sp.

1 Medusæ whose names are preceded by an asterisk are represent
Fiji Islands and in the Atlantic Ocean by identical or very closely allied species.

ed both in the

AGASSIZ AND MAYER: ACALEPHS FROM

*Abyla quincunx, Chun
*Agalma Pourtalesii, nov. sp. .
*Aglaisma quincunx, Chun
Diphyopsis angustata, Haeckel
*Diplophysa Kollikeri, Haeckel
Physalia utriculus, Eschscholtz
*Spheronectes Köllikeri, Huxley

Siphonophore.

THE FIJI ISLANDS.

183

184 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

LITERATURE CITED.

Agassiz, A.
1865. North American Acalephæ. 234 pp., 360 Figs.

Agassiz, A.
1883. Report of Dredging by U. S. Coast Survey Steamer “Blake.” Mem.
Mus. Comp. Zoöl. Harvard Coll., Vol. X. Part I., Report on Echini.
126 pp., 32 Plates.
Agassiz, A.
1892. General Sketch of the Expedition of the “ Albatross,” from February
to May, 1891. Bull. Mus. Comp. Zool. Harvard Coll., Vol. XXIII.
No. 1. pp. 1-90, 22 Plates.
Agassiz, L.
1849. Contributions to the Natural History of the Acalephæ of North
America. Mem. American Academy, Vol. IV. 2 Parts. pp. 221-
374, 16 Plates.

Asassiz, L.
1862. Cont. Nat. Hist. U. S., Vol. IV. 380 pp., Plates XX.-XXXV.

Brandt, J. F.
1838, Ueber Schirmquallen. Mem. Acad. Impériale des Sci. de St. Péters-
bourg, Sci. Naturelles, Ser. 6, Tom. IV. pp. 237-411, Taf. I-XXXI.
Chun, C.
1880. Die Ctenophoren des Golfes von Neapel. Fauna und Flora des Golfes
von Neapel, Monographie I. 813 pp, 18 Tafeln,
Chun, C.
1888. Bericht über eine nach den Canarischen Inseln im Winter 1887-88
ausgeführte Reise. Sitzungs Berichte Akad. Wissen. Berlin. pp.
1141-1173.

Chun, C.
1892. Die Canarischen Siphonophoren. Abhandl. Senckenb. naturf. Gesell.,
Bd. XVIII. pp. 57-144, 5 Tafeln.
Chun, C.
1897, Ueber den Bau und Morpholog. Auffassung der Siphonophoren.
Deutsche Zool. Gesell. Verhandl., pp. 48-111, Figs. in Text.

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 185

Claus, C.
1874. Die Gattung Monophyes und ihr Abkómmling Diplophysa. Schrift.

Zool. Inst. Wien, p. 25, Taf. IV.
Eschscholtz, F.
1825. Bericht über die Zoologische Ausbeute der Reise. Okens Isis.

Eschscholtz, F.
1829. System der Acalephen. 190 pp., 16 Tafeln.
Eydoux, F., et Souleyet, L.
1841-52. Voyages autour du monde sur la “Bonite.” 2 Tom., 101 Pls.
Paris.
Fewkes, J. W.
1882. Notes on Acalephs from the Tortugas. Bull. Mus. Comp. Zoöl.,
Vol. IX. No. 7, pp. 251-289, 7 Plates.
Goldfuss, A. G.
1820. Handbuch der Zoologie. Nürnberg. Bd. I, II. 4 Tafeln.

Haeckel, E.
1879-80. Syst. der Medusen. 672 pp., 40 Tafeln.
Haeckel, E.
1888. Siphonophore. Challenger Report, Zoöl., Vol. XXVIII. 380 PP»,
50 Plates.
Hill, R. T.
1898. The Geological History of the Isthmus of Panama, ete. Bull. Mus.
Comp. Zoöl., Geological Series, Vol. XXVIII. No. 5, pp. 151-285,
19 Plates.
Huxley, T. H.
1859. Occanic Hydrozoa. Ray Soc. 143 pp., 12 Plates,
Kölliker, A.
1853, Bericht tiber einige im Herbste 1852 in Messina angestellte vergleich-
end-anatomische Untersuchungen, 11. Quallen. Zeit. fiir Wissen.

Zool., Bd. IV. pp. 315-329.

Lendenfeld, R. von.
1884. Geographical Distribution of Australian Scyphomedusæ. Proc. Linn.
Soc. New South Wales, Vol. IX. pp. 421-433.

Lendenfeld, R. von.
1884. Seyphomeduse of the Southern Hemisphere. Proc. Linn. Soc. New
South Wales, pp. 155-169, 259-306, Plates TIL and IV.
Lendenfeld, R. von.
1887. Descriptive Catalogue of Scyphomeduse of Australian Sens. 2 Parts,
pp. 1-32, 1-49, 1 Plate.
Lesson, R. P.
1830. Voyage de la Coquille. 135 pp., 1 vol. of Plates. Paris.

186 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Lesson, R. P.
1836. Mémoire sur la famille des Béroides. Annal. Sci. Nat, Tom. V.
pp. 235-266.
Lesson, R. P.
1843. Histoire Naturelle des Zoophytes, Acalephes. Paris. 596 pp., 12
Plates.
Maas, O.

1897. Reports on Exploration off West Coast of Mexico, etc., by Fish Com-
mission Steamer “ Albatross,” during 1891. Mem. Mus. Comp.
Zoöl., Vol. XXIII. No. 1, Die Medusen, pp. 1-92. 15 Plates.
McCrady, J.
1856. Descrip. Oceania (Turritopsis) nutricula, nov. sp., and the Embryo-
logical History of a singular Medusan Larva, found in the Cavity
of its Bell. Proc. Elliot Soc., Vol. I. pp. 55-90, Plates 4-7.
McCrady, J.
1857. Gymnophthalmata of Charleston Harbor. Proc. Elliot Soc., Vol. I.
pp- 103-221, Plates 8-12.
Murbach, L.
1895. Preliminary Note on the Life History of Gonionemus. Journ. Mor-
phol., Vol. XI. pp. 493-496.
Péron, F., et Lesueur, C. A.
1807. Voyage de découvertes aux terres Australes, Mollusques et
Zoophytes.
Péron, F., et Lesueur, C. A.
1809. Tableau caractères génériques et specifiques de toutes les espèces de
Méduses. Annal. Mus. Hist. Nat., Tom. XIV. pp. 325-366.
Paris.

Quoy, J. R. C., et Gaimard, P.
1824. Voyage de Uranie et de la Physieienne. Zoologie. Paris.

Quoy, J. R. C., et Gaimard, P.
1833, Voyage de l’Astrolabe. Zoologie, Tom. IV.

Vanhöffen, E.
1888. Untersuch. über Semacostome und Rhizostome Medusen. Bibliotheca

Zoól., Bd. I. Heft 3, pp. 51. „6 Tafeln, 1 Karte.

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS.

ig. 18,
ig. 19,
x. 20.

VOL, XXXII, — NO, 9, 3

Sa go

EXPLANATION OF THE PLATES.

PLATE 1.

Pennaria vitrea, nov. sp., female.

ha se E male.
PLATE 2.
Cytæis vulgaris, nov. sp.
he à Proboscis.
x à One of the oral tentacles of the proboscis.

Bougainvillea fulva, nov. sp.

PLATE 8.

Laodicea marama, nov. sp.

e té Edge of bell.
Laodicea fijiana, nov sp.
4 8 Edge of bell.

PLATE 4.

Staurodiscus nigricans, nov. sp.
s s Edge of bell.
Aglaura prismatica, Maas.

PLATE 5.
Gonionemus suvaensis, nov. sp.
w a Edge of bell.
n = Aboral view of the proboscis.

Oceania pacifica, nov. sp.

PLATE 6.

Oceania ambigua, nov. sp.
Mi fé Oral view of a quadrant of the bell.

Clytia polynesiæ, nov. sp.

187

Fig.

38.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

PLATE 7.

Tiaropsis rosea, nov. sp.

5 «Marginal sense organ.

Section through the marginal sense organ of Tiaropsis diademata,
L. Agassiz: ent., entoderm; ect., ectoderm; ct., circular vessel, cut
across; vel., velum; pg., pigment spot in the entoderm of the cir
cular vessel; ot., otolith enclosed within an open pocket of the
velum.

PLATE 8.

Mitrocoma mbengha, nov. sp.

if E Edge of bell.

Polycanna purpurostoma, nov. sp.
n # Oral view of a portion of the bell. M., mouth.
de 2 Edge of bell showing otocysts and tentacles.

Eirene kambara, nov. sp.

PLATE 9.
Eutimeta levuka, nov. sp.
% x Edge of bell.
Liriope hyalina, nov. sp.

PLATE 10.
Linerges aquila, Haeckel.
de er Marginal sense organ.
Aurelia vitiana, nov. sp.

PLATE 11.

Cephea dumokuroa, nov. sp. Side view.
e i Aboral view of the disk.

PLATE 12.

Cephea dumokuroa. Section to show the internal structure: gpt., sub-
genital pit; m. $. 0., marginal sense organ; ov., genital organ; r, one
of the 8 thick gelatinous pillars that support the membranes of the
genital organs; these radial pillars correspond in position with the
8 mouth arms; S, stomach cavity; s. por., sub-genital porticus; rt.,
radial chymiferous vessel; t, chymiferous tube in the mouth arm.

Cephea dumokuroa: Marginal sense organ.

PLATE 13.

Pseudorhiza Thocambaui, nov. sp. Adult Medusa.

$ da Marginal sense organ of adult Medusa.
a 4 Side view of a young Ephyra.
oH pá Oral view of a young Ephyra.

Marginal sense organ of young Ephyra.

AGASSIZ AND MAYER: ACALEPHS FROM THE FIJI ISLANDS. 189

PLATE 14.

Fig. 45. Cassiopea ndrosia, nov. sp. Side view.
Fig. 46. Es En Aboral view of disk,

PLATE 15.

Fig. 47. Eucharis grandiformis, nov. sp. View of narrow side.
Fig. 48. SS y View of broad side. (Figure uncolored.)

PLATE 16.

Fig. 49. Beroë australis, nov. sp. View of broad side: l, lateral tube; k, side
connecting tubes; c, circular tube.
Fig. 50. Beroé australis, nov. sp. View of ciliated plates and of stellate pigment
cells.
Fig. 51. Sphæronectes Köllikeri, Huxley : c, duct connecting radial vessels with
the gastro-vascular cavity of the hydrosoma; c.s., covering scale ;
J, vacuolated vesicle ; y, gonophore ; A, hydrosoma ; p, feeding polyp;
r,r, r,r, radial vessels of the swimming bell; ¢, tentacle.
Sphæronectes Köllikeri, Huxley. Enlarged view of the hydrosoma.
Lettering similar to Figure 51.

ar
bo

Fig.

PLATE 17.

Fig. 53. Diplophysa Köllikeri, Haeckel. The free Eudoxia form of Spheronectes

Köllikeri: ad., tube by means of which the group was once attached
to the hydrosoma of S. Kollikeri; c. s., covering scale; g, swimming
bell-gonophore; p, feeding polyp; phc., phyllocyst of the covering
scale; t, tentacle. x
ig. 54. Diphyopsis angustata = Diphyes

angustata, Eschscholtz: c, duct con-

necting the chymiferous vessels of the swimming bell with the gastro-

vascular cavity of the hydrosoma; c.p., oleocyst; f, vacuolated

5

ve

sicle; A, hydrosoma; feeding polyp; v, v, ascending vessels of

ntacle,

the swimming bell; ?,

ay
à 4
y

di,
oll

.. >

un)

ame

FIJI ACALEPHS. PLATE 9.

IIA NE aan a7 e

30 ES >,
> g
42.

A FH}

x 7 | Ore. yé
(24704 j daat! JETA query
O À \

mma

18

w

=
> Wa. a 2090900

HAS

aan

FIVIIIGIS à

y

GM. del

À
A

x

>

Bulletin of the Museum of Comparative Zoölogy
AT HARVARD COLLEGE.
Vor. XXXII. No. 10.

REPORTS ON THE RESULTS OF DREDGING, UNDER THE SUPERVISION
OF ALEXANDER AGASSIZ, IN THE GULF OF MEXICO AND THE
CARIBBEAN SEA, AND ON THE EAST COAST OF THE UNITED
STATES, 1877 TO 1880, BY THE U. S. COAST SURVEY STEAMER
“BLAKE,” LIEUT.-COMMANDER C. D. SIGSBEE, U. S.N., AND COM-
MANDER J. R. BARTLETT, U. S.N., COMMANDING.

[Published by Permission of CARLILE P, PATTERSON and HENRY $, PRITCHETT, Superin-
tendents U, S. Coast and Geodetic Survey,]

XXXVIII.
ÉTUDE MONOGRAPHIQUE
DES

PLEUROTOMAIRES ACTUELS.

Par E. L. Bouvier er H. FISCHER.

AVEC QUATRE PLANCHES.

CAMBRIDGE, MASS., U.S. A. :

PRINTED FOR THE MUSEUM.
Serr. 1899.

No. 10. — Reports on the Results of Dredging, under the Supervision
of Alexander Agassiz, in the Gulf of Mexico and the Caribbean
Sea, and on the Last Coast of the United States, 1877 to 1880,
by the U. S. Coast Survey Steamer “ Blake,” Lieut.-Commander
©. D. Sigsbee, U. S. N., and Commander J. R. Bartlett, U.S. N.,
Commanding.

[Published by Permission of CARLILE P, PATTERSON and Henry S. PRIT-
CHET, Superintendents of U. S. Coast and Geodetic Survey. ]

XXXVIII.

Etude Monographique des Pleurotomaires actuels. Par E. L. Bouvier et
H. Fiscuer.

Lu mémoire que nous presentons aux zoologistes est une sorte de
Monographie où nous nous sommes efforcés de réunir les faits les plus
importants relatifs à l’histoire des Pleurotomaires actuels. Il se com-

h à SA ew Š
pose de deux parties bien distinctes, l’une purement historique, où nous
avons condensé les observations de nos prédécesseurs et les renseigne-
ments que nous avons pu recueillir sur les exemplaires connus de ce
curieux genre; lautre anatomique et essentiellement consacrée à nos

recherches personnelles. Si la première partie est destinée à rendre des

services aux hommes de science et à faciliter lours recherches, la seconde,
croyons-nous, jettera quelque lumière sur les animaux qui nous occupent
et sur l’évolution des Gastéropodes. Les Pleurotomaires, en effet, sont
les plus anciens de tous les Mollusques aujourd’hui connus ; on les
rencontre dans les couches fossilifères les plus primitives, et l’on était en
droit de supposer, même avant de connaître leur animal, qu'ils servi-
raient quelque jour à rattacher les Gastéropodes archaïques aux formes
qui leur ont donné naissance. Ces prévisions furent confirmées en
partie quand, à la suite des heureux dragages entrepris par le “ Hassler”
et le “Blake,” M. Dall put examiner des animaux de ce genre et se
convaincre qu’ils avaient deux branchies bipectinées symétriques, deux
orifices rénaux et un anus situé sur la ligne médiane dorsale. Il restait
toutefois à pousser plus loin les recherches et à faire, pour les organes
vitaux, ce que M. Dall avait réalisé pour la morphologie,
VOL. XxX11. — No. 10. 1

194 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Ce desideratum, formulé par beaucoup de zoologistes, sera satisfait,
dans une certaine mesure, par la publication du présent mémoire. Grace
à la libéralité éclairée et à la générosité scientifique de M. Alexandre
Agassiz, nous avons eu la fortune, que beaucoup nous envieront, sans
doute, de posséder un des animaux du Pleurotomaria Quoyana recueillis
par le “ Blake.” C'est le premier animal du genre dont on ait fait
Panatomie; nous avons eu toute liberté pour Vétudier; il nous a été
livré généreusement, sans condition aucune, et c'est pourquoi nous avons
réussi, bien qu’il fat incomplet, á en tirer un parti qu’on ne trouvera pas
sans intérêt, nous l’espérons, du moins.

Quelle que soit, du reste, la valeur des observations qui vont suivre,
nous en rapportons tout le mérite à l'éminent directeur des campagnes
du “Blake,” à M. Alexandre Agassiz. A l’auteur de tant de belles dé-
couvertes, au naturaliste qui a su retrouver dans les océans les restes de
faunes qu’on croyait éteintes, nous sommes heureux de dédier ce mémoire,
en lui présentant l'hommage de notre respectueuse reconnaissance.

PREMIERE PARTIE.
État Actuel de nos Connaissances sur les Pleurotomaires.

PLEUROTOMARIA (DEFRANCE), Sowerpy, 1821,

Le terme générique Pleurotomaria n’est pas apparu subitement dans
la nomenclature zoologique et il semble avoir été assez généralement
employé avant d’être décrit et publié. C'est en juin 1821,” que nous
trouvons la première indication du nouveau genre, dans les “Tableaux
systématiques ” de Férussac. Il s’agit de la simple citation suivante :

“Genre VI, PLEUROTOMAIRE, Pleurotomaria ? Defrance.” Le point
d'interrogation placé après Pleurotomaria se rapporte probablement au
nom latinisó Pleurotomaria (qui pouvait avoir différentes terminaisons),
traduction du nom Plewrotomaire de Defrance ; il est peu probable, en
effet, que le baron de Férussac ait eu quelques doutes en attribuant la
paternité du genre à Defrance, car les auteurs suivants wont pas la

1 Nous remercions également M. Agassiz d’avoir bien voulu nous permettre de
publier en France ce travail qui fait partie des “Reports on the Results of Dredg-
ng, under the Supervision of Alexander Agassiz, . . . by the U. S. Coast Survey
Steamer, “Blake.” ”

2 Cette date de publication nous est fournie par Dall (81, 77). D’autres auteurs

indiquent soit 1821, soit 1822,

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 195

moindre hésitation à ce sujet. L'ouvrage en question fait partie de
“Histoire naturelle générale et particulière des Mollusques terrestres
et fluviatiles,” etc., du père Férussac dont la publication a commencé
en 1819.

En décembre, 1821, James Sowerby ('21) dit quelques mots du genre
Pleurotomaria sans en nommer l’auteur; ces quelques mots suffisent
d’ailleurs pour le caractériser :

“ Trochus Gibsi. . . . the upper part (of the shell) is smooth, except
the concentric band, upon which are semicircular strie, indicating a
sinus in the outer lip... .

“This and several scher shells, hitherto called Trochi, with the band
around the spire, may more properly belong to the genus Pleurotomaria,
which I may be induced at some future period to adopt.”

Cette description, d’après M. le professeur Dall,’ confère la priorité du
genre à James Sowerby.

Il est certain que la citation faite par le baron de Férussac ne peut
pas être considéré comme suffisante pour établir le genre; au contraire,
la description de James Sowerby met en évidence les caractères essentiels
du Pleurotomaria ; mais il est beaucoup plus douteux que James Sowerby
ait entendu décrire ‘explicitement ce genre comme résultant de ses re-
cherches personnelles. Dans le passage que nous venons de citer, l’auteur
parle, en effet, du Pleurotomaria comme d’un genre bien connu, familier
à tous les zoologistes de cette époque et dont il est superflu, par consé-
quent, de nommer l’auteur. Il ajoute qu’il pourra être conduit ulté-
rieurement à Vadopter; cette dernière expression suffirait presque à
prouver que le paléontologiste anglais n’est pas l’auteur du genre en
question,

Deux ans plus tard (sept. 1823),? Defrance publie (23) les figures de
deux espèces, avec la légende suivante :

2. Pleurotomaire ornée (Defr.).
3. Pleurotomaire tuberculeuse (Defr.).

En 1824, Defrance ne fait que citer.le genre Pleurotomaire dans son
“Tableau des corps organisés fossiles” (24) ; enfin, en 1826, il donne
la description du genre et cite cing espèces :

Pleurotomaria tuberculosa Defr., Pl. anglica Defr., (Trochus anglicus
et similis Sow.) ; Pl. granulata Defr. (Trochus granulatus Sow.); Pl.

1 Dall ("81 et ’91, 396).

2 Voir, pour la date de publication des planches du vingt-sixitme cahier de
“Atlas du Dictionnaire des sciences naturelles, la bibliographie de la France,” ou
“Journal général de l'imprimerie et de la librairie ” -Pillet aîné, Paris, 1828.

196 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

ornata Defr. (Trochus ornatus Sow.) ; Pl. elongata Defr. (Antrochus
elongatus ? Sow.).

J. Sowerby (31), dans un fascicule de son “Genera of Recent and
Fossil Shells,” publié fin 1830 ou commencement de 1831,* donne la
diagnose du genre et figure deux espèces (Pleurotomaria reticulata et
Pl. elongata). Il ajoute: “On the genera distinguished by a more or
less deep fissure or notch in the upper part of the outer lip, the Pleuro-
tomaria of Defrance has no canal...” Ce passage de Sowerby lui-
méme reconnaît formellement la priorité à Defrance. Une deuxième
fois, en 1844, Sowerby (44), cite Defrance comme Pauteur du genre.

C’est à l’aide des documents précédents que nous devons attribuer la
priorité soit à Defrance, soit à Sowerby. Il nous semble que, dans le
cas actuel, l'application stricte des lois de priorité est impossible ; la
citation des “Tableaux systématiques ” doit être écartée ; la description
de Trochus Gibsi par J. Sowerby (21), avec les remarques qui l’accom-
pagnent renferment, au contraire, une bonne définition ; mais il y aurait
encore là matière à discussion, car le nouveau terme Pleurotomaria n’est
pas proposé à titre définitif par l’auteur, qui se contente d'annoncer qu’il
Vadoptera peut-être plus tard.

La première définition réellement complète et indiscutable du nouveau
genre est celle qu'a donnée Defrance en 1826, précédée par la planche
publiée en 1823. Il faut donc choisir entre les deux notations: ‚Pleuro-
tomaria J. Sowerby 1821 et Plewrotomaria Defrance 1826 : la premiere a
l'avantage de la priorité, mais sa valeur n'étant pas absolument démontrée,
la question nous paraît insoluble, comme nous le disions plus haut,

Tl faut donc renoncer à faire ici l'application: étroite des règles de la
nomenclature. Il nous semble que l’idée première du genre nouveau re-
vient à Defrance, qui l'avait reconnu depuis un certain temps déjà, sans
le publier ; Sowerby eut naturellement connaissance de la nouvelle coupe
générique grâce aux rapports qu’il entretenait avec Defrance ; on lit, en
effet, dans la préface des “Tableaux des corps organisés fossiles,” ? que
ce dernier avait obtenu de J. Sowerby la communication d’un certain
nombre de fossiles; on comprend très bien, dans ces conditions, que
Sowerby, en 1821, ait parlé du nouveau genre, sans d’ailleurs se pro-
noncer complètement sur sa valeur, et qu’il ne Pait adopté définitive-

1 Voir, pour la date de publication du trente deuxième cahier de cet ouvrage :
D. Sherborn, “On the Dates of Sowerby’s Genera of Recent and Fossil Shells”
(Annals and Magazine of Natural History, Vol. XIII, 1894, p. 870; R. B. Newton,
Syst. List Edwards Collection (Catal. Brit. Mus., 1891, p. 821).

2, Defrance (724, x).

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 197

ment, en 1831, qu’aprés la description de Defrance, et en lui rendant
entière justice. Il nous semble donc tout indiqué de nous ranger à
Popinion de Sowerby lui-même et de considérer Defrance comme le véri-
table auteur du genre; nous croyons avoir ajouté quelques arguments
à Pappui de cette manière de voir, que nous partageons avec H. Wood-
ward (85). En résumé, nous admettrons comme références originales,
à la fois la citation de Férussac (21), prouvant que le genre était adopté
par Defrance en 1821, et le passage du “Mineral Conchology ” de J.
Sowerby (’21), évidemment inspiré par Defrance, et qui met en évidence
les caractères du nouveau genre. Nous adöpterons donc la notation
suivante :

Pleurotomaria Defrance (Sowerby) [?21, x].

En terminant ce chapitre, nous adressons nos vifs remerciements &
MM. Deniker, de Paris, Sherborn et Gude, de Londres, qui nous ont
fourni de précieux renseignements sur les dates de publication des pre-
miers ouvrages où il est question du genre Plewrotomaria.

Le genre Pleurotomaria est largement représenté dans les terrains
sédimentaires. En 1885, Etheridge et H. Woodward ('85) ont fait le
recensement des espèces fossiles et en ont signalé 1156: ce nombre a été
un peu augmenté depuis.

La première espèce connue se rencontre dans le cambrien inférieur
(zone à Olenellus) des États-Unis, c’est-à-dire dans les plus anciens sédi-
ments fossilifères : c'est le Pleurotomaria (Raphistoma) Attleborensis Shaler
et Foerste (’88), décrit d’après un fragment trouvé à North Attleborough,
Mass. Les Pleurotomaires sont donc contemporains des plus anciens
Trilobites et des plus anciens Brachiopodes, qui ont été longtemps con-
sidérés comme les plus vieux animaux.

A l’époque silurienne, les espèces sont déjà très nombreuses, surtout
en Amérique; leur nombre augmente encore au carbonifère et atteint
son maximum pendant la période jurassique (367 espèces d’après Wood-
ward). La diminution commence à se faire sentir au crétacé. Les sédi-
ments tertiaires ne renferment que très peu d’espèces, et, pendant
longtemps, on n’en connaissait que quelques unes dans l’éocène ; l'ab-
sence complète du genre à partir du miocène avait même fait supposer
que les Pleurotomaires étaient complètement et définitivement éteints :
ceci explique le vif intérêt qu’a excité la découverte de la première espèce
vivante.! Cette lacune est maintenant en partie comblée. E. Vincent (90)

1 Mörch avait cru retrouver à l’état vivant un autre genre fort intéressant. Voir
la description de Murchisonia (Murchisonella) spectrum, de Vile Saint-Thomas (Malak.
Blätter, 1885, T. XXII, p. 184); mais cette forme est actuellement placée par les
auteurs près des Turbonilla.

198 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

signale, en 1896, treize espèces éocènes ; H. Crosse (82, 6) en indique
deux dans la miocène: Pleurotomaria Sismondai Goldfuss du miocène de
Bünde et Pl. tertiaria MacCoy d'Australie. Deux autres espèces ont
été trouvées dans les tufs madréporiques quartenaires de la Guateloupe :
Pleurotomaria Fischeri Mayer ms et Pl. Duchassaingi Schramm (69) ;
malheureusement elles n’ont pas été clairement définies. Nous ajoute-
rons à cette liste une forme des terrains tertiaires récents d'Italie : Pleu-
rotomaria gigas Borson et une espèce miocene de Vile Santa-Maria
(Açores) : Pl. atlantica Cotta.

COMPARAISON DES EsPhcEs VIVANTES AVEC LES Esphors FossILES.

Les espèces vivantes sont rangées dans les deux sections Æntemnotro-
chus P. Fischer et Perotrochus P. Fischer (85). Les Entemnotrochus
(Pleurotomaria Adansoniana et Pl. Rumphit), caractérisés par la position
élevée (supra- médiane) du sinus et par leur ombilic, se relient a une
série de formes crétacées rangées dans la section Leptomaria. Une
espèce éocène le (Leptomaria) landinensis Vincent (96), du landénien
inférieur belge, présente aussi quelques rapports avec le Pl. Adansoniana.
En outre, il existe dans le tertiaire de l’Italie septentrionale un véritable
Entemnotrochus, le Pleurotomaria gigas Borson, Dans un travail récent
(97), Sacco signale, en effet, les rapports étroits qui unissent cette espèce
au Pl. Rumphii ; d'autre part, elle a aussi des affinités avec l’espèce mio-
cène de Bünde, Pl. Sismondai Goldfuss. Sacco n’est même pas éloigné
d'admettre que Pleurotomaria Rumphii provienne directement de PR
gigas et celui-ci de Pl. Sismondat D’après Pauteur, le Pleurotomaria
atlantica Cotta présente également les plus grands rapports avec les
Entemnotrochus, et est peut-être même identique à Pl. gigas.

Les Perotrochus (Pleurotomaria Quoyana et Pl. Beyrichii), caractérisós
par la situation inframédiane du sinus et par absence d’ombilic se relient
à quelques formes jurassiques ? qui semblent appartenir à cette section.

Les deux sections actuellement vivantes du genre Pleurstomaria sont
donc représentées à l’état fossile, mais il n’est pas encore possible de
préciser la filiation de chaque espèce.

Espèces VIVANTES DE PLEUROTOMARIA.

Les espèces vivantes du genre sont actuellement au nombre de quatre :
le lecteur trouvera plus loin, dans la liste des spécimens, Vindication des
descriptions et des figures originales, à propos de chaque type; nous

1 Sacco ('97) et Borson (’21). 2 P, Fischer (85, 850).

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 199

nous contenterons done ici de donner leurs caractéres différentiels et

leur habitat.

Ces espèces sont, par ordre de découverte :

Pleurotomaria Quoyana P. Fischer et Bernardi, 1856.

Pleurotomaria Adansoniana Crosse et P. Fischer, 1861.

Pleurotomaria Beyrichii Hilgendorf, 1877.

Pleurotomaria Rumphii Schepman, 1879.

Ces quatre espéces doivent ¿tre réparties en deux sections, qui ont
déjà été définies par H. Crosse en 1882 (’82, 8), mais sans être nommées.
En 1885, P. Fischer ('85) a proposé pour ces deux sections les noms
Entemnotrochus et Perotrochus.

Nous résumons en un tableau les caractéres differentiels les plus
évidents qui permettent de reconnaître ces espèces, d’ailleurs très
distinctes :

Dernier tour fortement strié longitudinalement
et portant des granulations au voisinage de
la suture. Coloration: flammules rouges sur
fond rosé.

Espèce de très grande taille, habitant les An-

SECTION 1: Entemnotrochus. tilles.

Sinus situó anden du mis Pleurowmaria Adansoniana, Cr. et Fisch.
lieu du dernier tour; un | Sculpture très atténuée sur le dernier tour ;
ombilic.! entaille beaucoup plus étroite que dans
l'espèce précédente. Coloration : flammules
passant du rouge orangé au rouge carmin et
au violet clair, sur fand jaune blanchátre.

Espèce de taille gigantesque, provenant des
Moluques. Pleurotomaria Rumphii, Schep.

Sculpture très accentuée formée de cordons
spiraux rendus subnoduleux par leur croise-
ment avec des stries longitudinales. Colora- `
tion : flammules rouge vif sur fond jaune
clair.

SECTION 11: Perotrochws, Espéce de grande taille, habitant les mers du

Sinus situé au-dessous du Japon. . . Pleurotomaria Beyrichii, Hilg.

ig va Run Seulpture beaucoup plus fine et plus régulière

que dans l'espèce précédente. Coloration :

taches ou flammules obscures rouge brunátre
sur fond rosé.

Espèce de taille médiocre, habitant les Antilles.

Pleurotomaria Quoyana, Fisch et Bern.

1 Le sinus fournit un excellent caractère pour ces deux sections ; il est très long
chez les Entemnotrochus, beaucoup plus court chez les Perotrochus.

200 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

Recensement des Exemplaires Connus de Pleurotomaires
Actuels.

Les Pleurotomaires actuels figurent encore aujourd’hui parmi les
grandes raretés conchyliologiques; le nombre des exemplaires trouvés
jusqu’à présent déposés soit dans des musées, soit dans des collections
particulieres, dépasse á peine une vingtaine. Il est certain que ce nom-
bre s’accroitra fortement par la suite, lorsque les conditions d’existence
de ces animaux seront mieux connues; mais en attendant ce moment,
peut-être encore lointain, il nous a semblé utile de recueillir tous les
renseignements possibles sur ces premiers spécimens et d'indiquer ceux
qui ont été figurés. Nos recherches nous ont été grandement facilitées
par MM. Crosse et Dautzenberg, de Paris; MM. les professeurs Dall, de
Washington; Döderlein, de Strasbourg; E. von Martens, de Berlin ;
Smith et Woodward, de Londres; ainsi que par MM. Damon, de Wey-
mouth; Fulton et Sowerby, de Londres; Schneider, de Bâle, auxquels
nous adressons nos bien vifs remerciements.

Il existe au moins vingt et un spécimens distincts dont voici l’énumé-
ration.

1. Pleurotomaria Quoyana (Type). — C'est en 1855 que le premier
exemplaire de Pleurotomaria actuel a été recueilli par le commandant
Beau “sur une nasse mouillée à une grande profondeur, à plusieurs
milles du rivage de Marie-Galante, entre cette île et la Dominique.” *
La coquille était intacte ; Panimal et Popercule manquaient. P. Fischer
et Bernardi ('56) ont décrit et figuró ce premier spécimen et signalé
Vimportance exceptionnelle de sa découverte. Cette belle coquille fit
ensuite partie de la collection de M. Rolland du Roquan; à la mort de ce
dernier, elle fut vendue successivement à M. Moitessier, à M. R. Damon,
et enfin, en 1872, pour le prix relativement faible de 25
Mrs. de Burgh? dont la collection, après sa mort, est échue à Miss de
Burgh, qui en est actuellement propriétaire.

2. Pleurotomaria Adansoniana (Tyrx). — La seconde espèce connue
a été décrite et figurde par H. Crosse et P. Fischer (61)), d’après un
exemplaire incomplet qui gisait ignoré dans la collection du docteur
Commarmand. Cette coquille a été achetée, en 1858, par M. H. Crosse,
qui la possède encore actuellement.?

uinéegs à
guinees à

1 H. Crosse et P. Fischer (’61), p. 155.

2 H. Crosse et P. Fischer ('61, 155); H. Crosse (’82, 16); Cooker (’95, 122).

3 Ce spécimen vient d’être légué à l’un de nous par le savant et regretté Directeur
du “Journal de Conchyliologie,” décédé pendant l’impression de ce mémoire.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 201

3. Pieurotomaria Quoyana. — Cet exemplaire est le premier qu’on
ait recueilli avec Panimal; il a été capturé par L. Agassiz, en décembre
1871, lors de Pexpédition du “ Hassler,” près des Barbades, par 100
brasses environ de profondeur. Quelques détails concernant son organi-
sation ont été publiés, en 1872, par L. Agassiz (72) et reproduits, peu
de temps après, dans une note de E. von Martens ('72).! L’opercule
manquait. Ce spécimen est déposé dans les collections du Museum of
Comparative Zoölogy, à Cambridge (U. S.).

4. Pleurotomaria Beyrichii (Tyrer). — L'espèce a été décrite, en
1877, par Hilgendorf (77), d’apres un exemplaire en médiocre état,
acheté & Enoshima (Japon) & un marchand de coquilles locales. Cet
échantillon a été, plus tard, figuró par E. von Martens (’80, Pl. VIII);
il fait, actuellement, partie des collections du musée de Berlin.

5. Pleurotomaria Quoyana. — Exploration du “Blake,” station 290
(1879), au large des Barbades, 73 brasses. Cet exemplaire avait louver-
ture un peu brisée ; il était pourvu de l'animal, qui a été étudié par Dall
(89, 397, PI. XXXI, Fig. 1). La coquille est actuellement déposée au
U. S. National Museum de Washington ('89, 397).

6. Pleurotomaria Quoyana. — Exploration du “ Blake,” st. 296 (1879)
au large des Barbades, 84 brasses.? Exemplaire en parfait état, avec
Panimal et Popercule, déposé au Museum of Comparative Zoölogy de
vambridge (U. S.) : c'est celui qui a été étudié dans le présent mémoire.

7. Pleurotomaria Adansoniana. — Exploration du “Blake,” st. 278
(1879), au large des Barbades, 69 brasses. Un individu mort, brisé,
déposé au Museum of Comparative Zoölogy de Cambridge.

8. Pleurotomaria Adansoniana. — Exploration du “ Blake,” st. 276
(1879), au large des Barbades, 94 brasses. Individu avec animal, déposé
au U. S. National Museum de Washington, figuré par Dall (89, Pl.
XXXVII, Fig. 4).

Dimensions : diamètre maximum, 88 millimètres ; hauteur maximum,
70 millimètres.

9. Pleurotomaria Adansontana. —= Exploration du “ Blake,” st. 291
(1879), au largo des Barbados, 200 brasses. C’est un superbe échan-
tillon, pourvu de Panimal; il a été déposé au Museum of Comparative
Zoölogy de Cambridge.

1 Voir aussi H. Crosse et P. Fischer (”72); H. Crosse (76); A. Agassiz ('88),
Vol. II, p. 69.

2 Les figures publicos par Agassiz (88) et par Dall (’89) se rapportent aux spéci-
mens de Pleurotomaria Quoyana et de Pl. Adansoniana, dont il est maintenant

question.

202 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Dimensions: diametre maximum, 130 millimétres ; hauteur maximum,
130 millimetres.

10. Pleurotomaria Rumphii (Type). — Le type de la quatrième espèce
a été trouvé dans un lot de coquilles des Moluques appartenant au Jardin
zoologique de Rotterdam; c'est le plus grand de tous les exemplaires
connus de Pleurotomaires vivants. Ila été décrit, en 1879, par Schep-
man (’79), figuré trois ans après par le même auteur (82), et, plus
tard, par Sowerby (’87) et par A. Pilsbry.!

Dimensions: diamétre maximum, 190 millimétres; hauteur, 170
millimetres,

11. Pleurotomaria Beyriehii. — Un exemplaire en partie brisé, acheté
en 1881 & Enoshima (Japon) par le docteur Dóderlein dans une des
nombreuses boutiques oü. sont en vente, A titre de curiosités, des ani-
maux marins de diverses espèces (Hyalonema Sieboldi, Fusus pagoda,
Crustacés, etc.). Cet exemplaire appartient au docteur Döderlein,
actuellement professeur à Strasbourg.

12. Pleurotomaria Beyrichii. — Un très bel exemplaire, en parfait
état, acheté à Enoshima, en avril 1881, par le docteur Dóderlein à un
pêcheur japonais qui l'avait rapporté, avec d’autres espèces, de Misaki
(Japon). Ce spécimen a été communiqué ultérieurement à M. Schneider,
de Bâle, puis acheté, pour la somme de 950 marks, par Pætel, qui Va
légué, avec sa collection, au musée de Berlin?

Dimensions: diamètre, 89 millimètres; hauteur, 83 millimètres ;
longueur de la fissure à l'ouverture, 33 millimètres.

13. Pleurotomaria Adansoniana. — Cet exemplaire, qui était habité
par un Pagure, fut recueilli dans une nasse à homard, à l’îlet au Fajou,
dans le grand cul-de-sac de la Pointe-A-Pitre (Guadeloupe), à 150 brasses
de profondeur. Il a été figuré, en 1882, par H. Crosse (82, 12, PL I,
Fig. 1-2) et déposé au musée Lherminier, à la Pointe-à-Pître.

14. Pleurotomaria Beyrichii. — Un bel exemplaire très frais, acquis, en
1882 ou 1883, par le docteur Gottsche à Enoshima (Japon), a été acheté
ensuite par M. R. Damon, de Weymouth, pour la somme de 500 marks,
puis décrit et figuré par Woodward (85). Miss Farrington, de Preston,
s’est rendue propriétaire de cette belle coquille et l’a léguée, avec ses
collections, à sa sœur qui la possède encore actuellement.

1 Pilsbry (90), Pl. LVII, Fig. 13, 14, et H. Crosse (’80 et 821).

2 Nous croyons devoir rectifier une annotation de Woodward (’85, au bas de
la page 435), qui peut prêter à confusion. L’auteur, d’après une communication
du docteur Gottsche, signale les deux exemplaires du docteur Döderlein comme
incomplets et ne présentant pas la fissure, ce qui n'est pas exact, car l’exemplaire
du mugée de Berlin est intact.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 203

15. Pleurotomaria Beyrichit. — Un bel exemplaire, acheté à Londres
par M. Sowerby, sans indication de provenance, se trouve dans une col-
lection particulière en Angleterre.

16. Pleurotomaria Beyrichii. -— Un exemplaire, acheté par M. Fulton
à un négociant du Japon, fait aujourd’hui partie d'une collection par-
ticulière de New-York.

17. Pleurotomaria Beyrichit. — Un autre spécimen, ayant la même
origine que le précédent, a été acquis, en 1895, par le U. S. National
Museum de Washington.!

Dimensions : diamètre maximum, 65 millimètres ; hauteur, 60 milli-
mètres.

18. Pleurotomaria Quoyana. — Exploration de U. S. Fish C. S. “ Al-
batross,” st. 2354 (1885?), au large des côtes du Yucatan, prés Arrow-
smith Bank, 130 brasses. Un individu mort, en bon état, & part une cas-
sure à Pouverture; déposé au U. S. National Museum de Washington.

Dimensions : diamétre maximum, 48 millimötres; hauteur, 40 milli-
mètres.

19. Pleurotomaria Adansoniana. — Ce magnifique exemplaire, le plus
grand et le plus beau de l'espèce, a été remarqué, en 1890, par Lechmeer
Guppy,” à l’exposition de la Trinité et de Tobago. Il a été acheté par
M. R. Damon et figuré par lui (91), puis acquis pour la somme de 55
livres sterling par le British Museum de Londres, où il est exposé.

20. Pleurotomaria Beyrichii. — Un bel exemplaire, avec l'animal, a
été pêché, en 1897, au Japon, au large de Boshu, sur le banc Okinosé ;
il fut communiqué aussitôt à M. Alan Owston, de Yokohama, et au pro-
fesseur Mitsukuri, de Tokio, qui put l’observer vivant.® Ce spécimen
est maintenant dans les collections du British Museum de Londres.

21. Pleurotomaria Beyrichit. — Un bel exemplaire a été acquis tout
récemment par M. Dautzenberg, de Paris.

En résumé, les exemplaires connus se répartissent de la manière
suivante :

Pleurotomaria Quoyana, 5; Pl. Adansoniana, 6; Pl. Rumphü, 1;
Pl. Beyrichii, 9.

Il est fort probable que cette liste n'est pas complete, surtout en ce
qui concerne le Pleurotomaria Beyrichii dont un ou deux exemplaires,
autres que ceux mentionnés, existent encore en Europe. Il semble

`

même que cette dernière espèce soit appelé à devenir moins rare que les

1 Pilsbry (95), W. E. Collinge (’94).
2 Lechmere Guppy (’90 et *91), Pilsbry (’92).
3 Mitsukuri (297).

204 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

autres, car les pécheurs japonais en capturent de temps en temps.
Souhaitons que les spécimens soient un jour assez nombreux pour
permettre d’entreprendre une étude anatomique compléte de ces formes
si intéressantes.

Résumé de nos Connaissances sur l’Aspect Extérieur et
l'Organisation de l’Animal.

Une courte description de l’animal a été donnée par Agassiz (72).
Plusieurs figures, prises sur Panimal vivant par M. J.-H. Blake, tant de
Pleurotomaria Quoyana que de Pl. Adansoniana, ont été publiées par
M. Dall,’ qui a complété, sur bien des points, la description précédente ;
récemment enfin, Mitsukuri (?97) a publié une courte note sur l’animal
vivant du Pl, Beyrichii.

1° Pleurotomaria Quoyana et Pl. Adansoniana. — Dall a surtout dé-
crit en détail le Pleurotomaria Quoyana, et les lignes suivantes se rap-
portent spécialement à cette espèce, à moins d'indication contraire ;
les deux formes diffèrent d’ailleurs assez peu.

Coloration générale, téguments. — Le corps est brun rouge; de fines
lignes sombres transversales s'étendent sur la tête, plus serrées en avant
des tentacules qu’en arrière, Le bord du manteau est plus sombre que le
pied. La tête, en arrière des tentacules, présente une surface rugueuse ct
ridée transversalement ; le reste du corps, à Pexception de la sole pédicuse,
est finement granuleux ou papilleux.

Chez le Pleurotomaria Adansoniana, la surface du corps est moins
rugueuse que chez Pl. Quoyana.

Pied. — Le pied, très développé, est presque deux fois plus long que
la coquille. Sa partie postérieure porte un opercule corné multispiré,
semblable à celui des Trochidés,? mais de petite dimension relativement
à la taille de la coquille. En arrière du lobe operculigère se trouve une
surface triangulaire limitée de chaque côté par les prolongements pos-
térieurs des lobes épipodiaux ; cette surface est légèrement tuberculeuse
chez le Pleurotomaria Quoyana, tandis que chez le Pl. Adansoniana, elle
est sillonnée de fortes rides transversales et creusée d’un sillon médian ;
d’après M. Dall, est cette différence qui fournit le caractère le plus net
pour séparer les deux espèces d’après l’animal.

Epipodium. — Il est large, mince, muni, comme les bords du man-
teau, d'une rangée de petites papilles courtes; il est dépourvu des
1 Dall (89, PL XXIX, Fig. 1; Pl. XXX, Fig. 1-6).

2 Dall (89, Pl. XXXII, Fig. 10).

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 205

filaments et prolongements qu’on observe chez les Trochidés et chez les
Scissurellidés. Sur Panimal vivant, l’épipodium est intimement appliqué
sur le manteau, qu'il double en contournant le bord de la coquille et se
comporte done tout autrement que chez les Trochidés, dont I’épipodium
s'épanouit librement dans l’eau. Celui du Pleurotomaria Adansoniana
est moins développé que celui du Pl. Quoyana.

Tête. — La tête est pourvue de deux tentacules droits et pointus et
prolongée antérieurement par un gros muffle cylindrique, dont lex-
trémité, arrondie en forme de disque, porte la bouche au centre et,
inférieurement, un sillon médian. Les yeux, situés A la base des ten-
tacules, sont ronds, petits et noirs ; ils montrent une perforation centrale,
de sorte que l’eau de mer occupe la place du cristallin absent et pénètre
dans la coupe formée par le fond de l'œil,

Manteau. — Ses bords sont garnis de papilles, surtout le long de len-
taille qui correspond à la fissure de la coquille. Lorsque l’animal est en
extension, les deux lèvres de cette entaille s’accolent, sauf A Pextré-
mité interne du sinus, ot il reste une ouverture ayant le róle d'un
siphon.

Branchies. — Les branchies n’ont été observées que chez le Pleuro-
tomaria Adansoniana ; elles sont au nombre de deux, situées à l’intérieur
do la cavité palléale, près de la soudure du manteau avec le corps, et
leur ensemble forme deux séries presque symétriques de feuillets aplatis,
déposés de chaque côte d’un bourrelet longitudinal, à parois minces, con-
tenant le vaisseau branchial, et qui s'étend parallèlement à l’entaille du
manteau.’ L'extrémité antérieure du bourrelet est libre sur une petite
distance et forme une pointe où les fouillets branchiaux diminuent de
grandeur. Chacune des deux branchies contient environ trois cent
soixante feuillets. Dans la région où le bourrelet devient libre, on
observe un renflement des téguments, et, à côté de lui, un petit organe
hémisphérique qui est peut-être sensoriel (osphradium ?).

Appareil digestif. — M. Dall a décrit la mâchoire et la radule dont
il sera question plus loin; il a observé, chez le Pleurotomaria Adan-
soniana, les restes d'un jabot à parois très minces. L’extrémité de
Pintestin se recourbe fortement en S avant de s'ouvrir à l’anus.?

Reins (?). — Immédiatement au-dessous de Vintestin se trouve une
glande ou bien une paire de glandes en contact sur la ligne mé-
diane qui paraissent s'ouvrir, près de Pextrémité postérieure de len-
taille palléale, par deux orifices obliques et symétriques. M. Dall

1 Dall (’89, p. 401; p. 434, Fig F; Pl. XXX, Fig. 2).
2 Dall (89, Pl. XXX, Fig. 8),

206 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

suppose que ces deux glandes et ces deux orifices sont les reins avec
leurs pores exereteurs.

Enfin, M. Dall (89, Pl. XXX, Fig. 3), figure, prös des orifices ré-
naux ou supposés tels, une autre paire de pores sur lesquels il ne
s'explique pas; peut-être faut-il y voir des pores génitaux.

2° Pleurotomaria Beyrichii. — Ce que nous savons sur cette espèce
se réduit à fort peu de chose : d’après M. Mitsukuri (97), la sole pé-
dieuse est jaune-paille ; les côtés du pied sont marqués de taches et
de traînées carmin foncé sur fond orange; le muffle est rouge-carmin.
Le tentacule gauche de l’individu observé avait une petite branche près
de la pointe. Les deux lobes épipodiaux se relèvent jusque sur la co-
quille, et il semble même, d’après la description de l’auteur japonais,
que ces lobes épipodiaux présentent un développement considérable.

DEUXIÈME PARTIE.
Etude Anatomique du Pleurotomaria Quoyana.

L'animal que nous avons eu entre les mains avait été extrait de sa
coquille, que nous n'avons, par conséquent, pas pu examiner, L’éti-
quette jointe à l'échantillon portait les indications suivantes :

“U.S. Coast Survey, C. P. Patterson, Supt. Carribean Islands Ex-
ploration. U.S, C. S. S. ‘Blake,’ Alex. Agassiz, 1878-1879.

“ Pleurotomaria Quoyana.

“No. 296. Depth 73 fathoms, Barbados.

“This is the best of the specimens and (except the operculum) is
exactly as when it was taken from the shell.”

L’animal a été représenté tel qu'il nous parvint dans la figure 1,
Pl. I, du présent mémoire. Arraché avec effort de sa coquille, il avait
perdu le tortillon tout entier, les branchies, la plus grande partie du
manteau et des viscères, à l'exception de la radule et d'une portion
déchirée de Poesophage et de la masse buccale. Le pied se trouvait
intact, mais il était recroquevillé, comme le montre la figure ; le muscle
columellaire, les tentacules, les yeux, une partie des parois de la ca-
vité antérieure du corps et une faible partio de la région inféricure
du manteau n'étaient pas trop endommagés et se prétaient encore
parfaitement à des recherches anatomiques; l’épipodium n’était pas
apparent au premier abord et nous píimes croire un instant qu'il faisait
défaut ; mais un examen ultérieur plus attentif nous a permis de le voir,

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 207

contracté par Valcool, à la partie supérieure du pied. Il paraît peu
développé et s'arróte assez loin des tentacules.

Pour plus de détails sur la morphologie externe, nous renvoyons
& nos figures et aux mémoires de M. Dall, que nous avons résumés
plus haut.

DESCRIPTION DE LA RADULE.

M. Dall (89, 399, Pl. XXXI, Fig. 1) a décrit et figuré en partie la
radule du Pleurotomaria Quoyana: en plus de la dent impaire, il dis-
tingue trois séries de dents: 1° les laterales, au nombre de vingt-six,
plus ou moins aplaties; 2° les uneini, ayant un à trois denticules
presque aussi longs que la pointe principale ; les uncini situés anté-
rieurement (uncini majores) sont plus forts et ont plus de denticules
que les suivants (uncini minores); on compte dix-huit à vingt de ces
uncini; 3% les laterales minores, très nombreuses, petites, transpa-
rentes et si rapprochées qu’il est presque impossible de les compter.
M. Dall pense qu’il en existe quarante à cinquante; elles sont minces
et spatuliformes, l’auteur n’a observé sur aucune de ces dents les
touffes qu'il a décrites chez le Pleurotomaria Adansoniana. La formule

c - q
radulaire serait donc R E + 5) + ( Em + >)
5740) 3—1 0

La lettre R désigne la dent rachidienne (impaire) ; les chiffres in-
scrits en dénominateur indiquent le nombre de cuspides qui caractérise
chaque dent.

Nous allons compléter la description de M. Dall en y apportant
quelques modifications.

La radule des Pleurotomaires est extraordinairement compliquée ; il
existe peu de Mollusques qui présentent d'aussi importantes modifi-
cations le long d’une série transversale ; mais ces modifications pré-
sentent un remarquable caractére de continuité qu’on ne trouve á un
pareil degré chez aucun Diotocarde; nous reviendrons d’ailleurs en
détail sur cette particularité spéciale aux Pleurotomaires,

Les rangées transversales n’ont nullement une direction transverse,
mais sont repliées en forme de V. La figure 12 (Pl. III) représentant
la radule, vue par dessous, montre cette disposition d’ensemble; il est
clair que de ce côté les dents ne peuvent pas être aperques ; seules leurs
surfaces d'insertion sont visibles. Nous avons représenté une demi-rangée
(Pl. III, Fig. 1) & un plus fort grossissement, également vue par des-
sous; on remarque que la surface d'insertion de chaque dent varie de
forme et de grandeur depuis la ligne médiane jusqu’à Vextrémité ; c’est

VOL, XXXII — No. 10. 2

208 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

à la limite du premier et du deuxième tiers que la base des dents pré-
sente le plus de force,

Nous avons pu établir le nombre exact des dents d’une rangée; il
existe une dent médiane impaire, et, de chaque côté, cent dix-sept dents
paires; on trouve, en outre, au-delà de la cent dix-septième dent, la
base rudimentaire d’une cent dix-huitième. La forme de ces dents est
très variée, et, pour la commodité de la description, nous diviserons les
dents paires en cinq groupes, qui n’ont pas d’ailleurs de limites précises :
1° dents centrales; 2° dents lamelleuses ; 3° dents à crochets (ou uneini) ;
4° dents à brosses; 5° dents flabelliformes.

Nous désignerons chaque dent par son numéro d’ordre & partir de la
dent impaire (exclusivement). Les dents ayant méme numéro d’ordre
dans deux rangées différentes sont toujours rigoureusement identiques ;
nous avons constaté le fait très minutieusement, notamment dans les ré-

' gions de transition entre deux groupes consécutifs, où Pon pouvait s’atten-
dre à des variations individuelles; il n’en est rien; la vingt-neuvième
dent, par exemple, a toujours identiquement la forme figurée.

La disposition générale des dents présente quelques particularités
dignes d’être signalées; les deux moitiés ne se correspondent pas exacte-
ment, mais il y a un léger chevauchement (Pl. III, Fig. 2 et Fig. 12);
en outre, la surface où s’attachent les dents (épithélium lingual) est loin
d’être plane ; la dent impaire et ses voisines sont insérées sur une forte
saillie longitudinale, de chaque cóté de laquelle on trouve une forte
dépression ott sont les dents lamelleuses triangulaires. Au-delá de cette
dépression court une nouvelle saillie qui supporte les grosses dents
à crochet. Ces dispositions se voient facilement en examirant la radule
par dessous,

Dent impaire. — La dent impaire (Pl. TIT, Fig. 4, 5, 6), de petites
dimensions,’ est presque couchée sur ’épithélinm lingual; elle est con-
stituée par une lame vaguement quadrangulaire, située dans le plan
médian, et prolongée antérieurement ? par une pointe assez aiguë ; elle
s'insère sur l’épithélium lingual suivant une bande étroite et allongée
dans le sens de la ligne médiane (cette bande d'insertion est représentée
obliquement, vue de trois quarts, dans la figure 5); son bord postérieur,
très incliné, porte, de chaque côté, deux expansions minces et foliacées,
visibles à droite et à gauche dans la figure 5.

1 Les dents figurées isolément sont toutes figurées au même grossissement de
70 diamètres : ce sont toutes des dents du côté gauche.

2 Dans les pages qui suivent, nous appellerons extrémité antérieure d’une demi-
rangée de la radule, celle qui se trouve sur la ligne médiane; par exemple, la figure
2 représente la moitié antérieure de la demi-rangée gauche.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 209

Dents centrales paires. — De part et d’autre de la dent impaire, les
premières dents que l’on rencontre sont de beaucoup plus grande taille
que la dent impaire et que les suivantes ; elles sont insérées (Fig. 3) sur
une ligne & peu prés transversale; on peut les appeler les dents centrales
paires en limitant cette région, arbitrairement d’ailleurs, aux trois pre-
mières. Elles sont larges et aplaties (Fig. 7, 8, 9), c’est la deuxième
qui présente la largeur maximum; elles portent, du cóté externe, une
crête de renforcement qui se prolonge latéralement et inférieurement
par une expansion mince et foliacée (Fig. 3), homologue de celles de la
dent médiane, mais diminuant rapidemént de grandeur. Leur surface
d'insertion est concave (Fig. 7-9).

On remarquera qu’il existe une assez brusque différence de forme et
de taille entre la dent impaire et la première centrale : celle-ci n’a
point l'équivalent de la pointe que possède la dent impaire. Au con-
traire, la quatrième dent ne diffère de la troisième que par le rétrécisse-
mont de son extrémité supérieure. La quatrième établit donc une
transition très ménagée entre les dents centrales, et les dents lamel-
lenses ; nous n'avons, d’ailleurs, attribué à ces distinctions qu'une valeur
absolument artificielle.

Dents lamelleuses. — A partir de la troisième dent inclusivement, la
direction générale de la rangée change et, au lieu d’être transversale,
s'éloigne de laxe suivant en angle aigu d’environ 30 degrés. La taille
des dents diminue légèrement jusqu’à la sixième ou septième (Fig. 2),
en même temps que la forme tend à devenir triangulaire ; entre la sep-
tième et la vingt-cinquième la taille reste à peu près constante, puis
s'accroît rapidement au-delà. Les petites expansions folincées qui se
trouvent à la base des dents centrales persistent sur les dents lamel-
leuses, mais disparaissent sur la vingt-neuviéme, La huitième et la
vingt-huitième dent sont représentées sur les figures 3 (Pl. III) et 2
(PIS EVE

Dents à crochets. — La vingt-neuvième dent établit la transition entre
les dents lamelleuses triangulaires et les dents à crochets. Elle pré-
sente encore une forme générale triangulaire (Pl. TV, Fig. 1, dent du
milieu) et son extrémité est assez mince; d’autre part, vue de côté
(Fig. 3), elle présente une remarquable analogie de forme avec la dent
suivante (Fig. 4) et montre très nettement le début de deux cuspides ;
il est impossible de rapporter cette dent aux précédentes plutót qu’aux
suivantes, car elle a des caractères exactement intermédiaires.

Les dents à crochet qui suivent perdent assez rapidement la forme
triangulaire, tout en restant aplaties latéralement ; elles s’allongent,

210 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

s'incurvent, et leur extrémité se munit de cuspides. La trentième (Fig.
4) est bicuspide, mais elle possède une faible indication d'une troisième
cuspide qui va se développer sur les suivantes. La trente et unième
(Fig. 6) et la trente-deuxieme sont trés caractérisées comme dents tri-
cuspides ; on observe que le bord externe et le bord interne présente l’un et
l’autre bourrelet d’épaississement : le bourrelet interne, du côté concave,
forme deux des cuspides que nous appellerons cuspides internes, tandis que
le bourrelet externe, du côté convexe, se termine par la cuspide externe.
Dans la région comprise entre les deux bourrelets, la dent est mince et
aplatie. La trente-troisième (Fig. 7) ne porte plus que deux cuspides :
la cuspide interne inférieure a avorté, et sa place n’est plus indiquée
que par une légère ondulation du bord, dont on ne trouve même plus de
trace dans la trente-quatrième. Les dents numérotées de 32 à 35 sont
les plus fortes de toutes; ce sont aussi celles dont la base présente les
plus grandes dimensions (Pl. III, Fig. 1).

A partir de la trente-sixième, la partie recourbée de la dent diminue
d'épaisseur; en même temps, le bourrelet externe de renforcement
prend de moins en moins d'importance dans la partie terminale, en sorte
que la cuspide externe, toujours médiocrement saillante, même à la
trente et unième dent où elle présente son maximum de développement,
est de plus en plus en retrait avec la cuspide interne ; en somme, la trente-
quatrième dent prend déjà Paspect d’une dent unicuspide, aspect qui ne
fait que s'accentuer (Pl. 4, Fig. 21) et devient très marqué pour la
quarante et unième ; toutefois la cuspide externe, quoique pratiquement
‘invisible, existe toujours morphologiquement comme terminsison émous-
sée du bourrelet externe, toujours présent.

La quarante-deuxiéme dent (Pl. IV, Fig. 9) est encore unicuspido,
mais elle montre, sous la forme d'un léger renflement, le rudiment d'une
nouvelle cuspide interne, qui se développe rapidement sur les dents
suivantes (voir Fig. 11) et qui paraît se former au même point où avait
disparu la cuspide interne inférieure des trentième, trente et unième et
trente-deuxième dents, Les quarante-troisième et quarante-quatriémo
dents deviennent donc bicuspides, mais sont fondamentalement du méme
type tricuspide que la trente et unieme, en tenant compte de ce que
nous venons de dire sur la cuspide externe. Cette structure ne tarde
pas à se compliquer: la quarante-cinquième dent (Pl. IV, Fig. 11)
montre, en effet, entre ces deux cuspides internes, le début d'une troi-
siéme cuspide interne (cuspide moyenne) qui va se développer sur los
dents suivantes (ex : 48° dent, Fig. 25), qui deviendront donc tricuspides,
mais dun type différent des triscuspides rencontrées antérieurement,

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 211

puisque les trois cuspides dépendent du bourrelet interne. On voit que
la série des dents A crochet est loin d'être uniforme. Les dernières
d’entre elles s’amincissent sensiblement.

Dents à brosses. — La quarante-huitième dent fait la transition entre
les dents á crochet et les dents à brosse. Le bourrelet externe est tou-
jours présent ; mais son extrémité distale (qui équivaut à la cuspide ex-
terne de la trente et unième dent) s'étant de plus en plus retirée de la
pointe, n’arrive plus qu'au niveau de la cuspide interne inférieure et
se termine par une sorte de petit épaississement irrégulier, qui se garnit
de minuscules baguettes sétiformes, dont on n’observe qu’une trace à
peine perceptible sur la quarante-huitième dent, mais qui forment déjà
une petite touffe sur la cinquantième (Pl. IV, Fig. 13). Ces soies
sont transparentes avec une vague structure transverse ; leur diamètre
mesure environ .002 millimètre A quelque distance’ de leur insertion.
Leur extrémité distale est assez pointue. Sur la cinquante-septième
dent, ces soies sont bien développées et atteignent l'extrémité libre de
la dent, qu’elles dépassent même sur les suivantes : elles se courbent
en enveloppant l'extrémité de la dent de chaque côté, ainsi que son bord
convexe (Pl. IV, Fig. 27).

La forme générale des dents & brosse n’est plus celle des dents á
crochet : leur courbure est beaucoup plus faible (Fig. 15) et tend à
s’atténuer encore en approchant de l'extrémité de la série (Fig. 16);
leur tige est très amincie.

Les dernières dents à brosse se modifient sensiblement : leurs cus-
pides sont moins saillantes (ex: 103° dent, fig. 16); une expansion
foliacée, latérale et externe, se développe tout du long de la tige ; elle
est déjà sensible sur la cent troisième dent et devient extrêmement
développée vers la cent dixième (Pl. IV, Fig. 17) et sur les suivantes ;
cette expansion va même former la totalité des dents flabelliformes, par
suite de l’atrophie graduelle de la tige et de la brosse.

Les cuspides paraissent disparaître un peu avant la brosse : nous les
avons encore aperques sur Ja cent neuvième dent; la brosse existe
jusqu’à la cent onzieme inclusivement, mais en s'atrophiant rapidement.

Dents flabelliformes. — La cent douzième dent ne présente plus trace
de brosse: une petite ineurvation du bord indique seulement sa place.
Les dernières dents, jusqu’à la cent dix-septième (Pl. IV, Fig. 5) sont
minces, aplaties, spatuliformes, un peu recourbées ; leurs bases sont tel-
lement petites qu’elles prennent une disposition divergente rappellant
assez celle des feuillets d'un éventail, dont les lames un peu courbes

iraient en décroissant lentement. On voit sur la figure, juxtaposée à la

212 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

base de la derniére dent, la base rudimentaire d'une cent dix-huitióme
dent non développée.

Plaques basales accessoires. — Dans toute la région moyenne de chaque
demi-série, la base des dents est accompagnée d'une petite pièce acces-
soire fortement adhérente à l’épithélium lingual: on Papergoit en ex-
aminant la radule par sa face inférieure (Pl. ILI, Fig. 1) ou encore en
observant Pépithélium lingual par transparence, après avoir enlevé
chaque dent. La première plaque s’observe sur la trente-quatrième
dent; il y a déjà, à vrai dire, sur la trente-troisième, une pièce analogue,
dun peu plus grande dimension, mais qui paraît encore soudée à la
base. A partir de la trente-quatrième dent, cette plaque diminue un
peu de dimension, puis augmente à partir de la trente-neuvième ; entre
la quarante et unième et la soixante-quatorzième, elle conserve à peu
près la même taille et la même forme: elle est ovale et disposée obli-
quement ; puis elle diminue rapidement et disparaît après la quatre-
vingt-uniéme dent. En résumé, la radule présente une dent impaire,
quelques dents centrales transversales, une série de dents lamelleuses,
puis des dents à crochet, d’abord bicuspides, ensuite tricuspides, de nou-
veau bicuspides, ensuite unicuspides, puis tricuspides, mais suivant une
disposition différente, et, enfin, des dents flabelliformes. Il existe des
transitions ménagées entre ces divers types de dents, sauf entre la dent
impaire et la première centrale.

COMPARAISON DE LA RADULE DES PLEUROTOMARIA AVEC CELLE DES
AUTRES DIOTOCARDES,

La structure de la radule du Pleurotomaria Quoyana est absolument
spéciale et on ne trouve rien d’analogue chez les autres Scutibranches ;
ces derniers montrent, en effet, une tendance très nette à la division
d’une série transverse en régions spécialisées ; on y distingue presque
toujours une région centrale formée par la dent impaire et un petit nom-
bre de dents paires (généralement cinq de chaque côté), et deux régions
marginales (une à droite et une à gauche), constituées par des dents
alongécs et recourbées en forme de crochet; entre la région centrale et
chaque région marginale se trouve assez souvent une dent latérale dif-
ferenciée, tantôt rudimentaire, tantôt, au contraire, très forte, et qui vient
encore accentuer la séparation des deux régions, déjà bien marquée par
les différences de structure et de grandeur des dents.!

1 Ce schéma est parfois modifié; la dent latérale n'est pas toujours différenciée
En outre, les premières marginales sont parfois différentes des suivantes; mais,
dans aucun cas, on n’observe de dispositions comparables à celles des Pleuroto-

maires,

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 210

Chez le Pleurotomaria, il y a sans doute des variations considérables
entre les diverses dents d’une rangée, mais il est impossible de tracer des
limites précises entre les régions qu’on peut y distinguer, car les dents se
modifient de Pune à l’autre (sauf la dent impaire et sa voisine) par des
transitions ménagées ; on voit donc la différence profonde qui distingue
les Pleurotomaires des autres Diotocardes.

En outre, apres les dents que nous avons appelées centrales, il existe
une longue série de dents lamelleuses triangulaires, mal séparées des
précédentes et dont il est impossible de trouver l’équivalent chez aucun
autre Diotocarde ; si nous les rapportions aux dents centrales, il faudrait
admettre, pour ces dernières, un nombre voisin de vingt huit, c’est-à-dire
près de six fois le nombre (ciuq) qu'on observe chez les autres Dioto-
cardes. D’autre part, il semble impossible de les comparer aux dents
à crochet; elles constituent done une série bien spéciale aux Pleuro-
tomaires.

Il est possible que ces dents aient avorté chez les autres Diotocardes;
ainsi s'expliquerait la brusque difference de structure entre les dents
centrales et les dents marginales de ces derniers. L’atrophie de la dent
latérale chez certains genres de Trochidés,! où une simple lamelle rudi-
mentaire sépare les centrales des marginales semble appuyer cette manière
de voir et prouve, en tous cas, que des dents peuvent disparaître dans
cette région de la radule. Malheureusement, cette hypothèse, quelque
satisfaisante qu’elle soit, n'est pas étayée par un nombre suflisant de faits.

La présence des dents à brosse signalée, par Dall, chez le Pleuroto-
maria Adansoniana ct, par nous, chez le Pl. Quoyana, paraît également
caractéristique du genre.

Il ost intéressant de noter que la radule des Seissurella n’a aucun
rapport avec celle des Pleurotomaires, mais se rattache facilement au
type qu’on observe chez les autres Diotocardes.

On voit que Pétude de la radule du Pleurotomaria présente un très
grand intérêt; cet organe possède, en effet, au moins trois caractères
particuliers à ce genre; transitions ménagées entre les dents d’une
série, présence des dents lamelleuses triangulaires, dents à brosses.
Le premier de ces caractères est capital, car il est en rapport avec un
état d'organisation primitif par rapport à celui des autres Prosobranches,
chez lesquels les dents de la radule sont groupées en régions spécialisées et
dépourvues de termes de transition. Pour mieux faire comprendre notre
pensée, en employant une comparaison d’ailleurs tout artificielle, nous
pouvons dire qu’on observe, chez les Pleurotomaires, quelque chose

1 Troschel, Gebiss der Schnecken, Bd, IT.

214 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

d’analogue & ce qui se passe chez les Mammiferes fossiles les plus anciens,
dont la dentition est remarquablement homogéne et présente des tran-
sitions ménagées entre les dents spécialisées, incisives, canines, prémo-
laires, etc., qui sont beaucoup mieux séparées chez les Mammifères plus
récents.

MÂCHOIRES.

Les mâchoires ont déjà été décrites par M. Dall (89). Ce sont deux
pièces cornées minces, très faibles, et qui n’ont probablement qu’un rôle
fonctionnel très restreint; nous donnons la figure de Pune d'elles (Pl.
III, Fig. 2). Du bord le plus épais partent de nombreuses stries irrégu-
lières et serrées, qui couvrent plus de la moitié de la surface.

ORGANES DES SENS.

Gil. — L’œil a été décrit par Dall (89, 398) et il en a été question dans
un autre passage de ce mémoire. Nous avons vérifié, à l’aide de coupes,
que c’est un œil très simple, ouvert à l'extérieur et tapissé à l’intérieur
par la rétine; son diamètre mesure environ un demi-millimètre ; l’orifice
extérieur est un peu plus étroit. Nous aurions voulu savoir si sa cavité
intérieure est absolument vide comme chez les Nautiles, ou bien si elle
contient une substance transparente plus ou moins molle, faisant fonction
de cristallin rudimentaire, comme chez les Trochus ; mais le médiocre état
de conservation ne nous a pas permis de résoudre sûrement la question.
D’après M. Dall, il wy a pas de cristallin et l’eau de mer pénètre libre-
ment dans la cavité. Notons toutefois une grande analogie de forme et
de structure avec l'œil des Zrochus.*

Cette disposition générale de l’œil est certainement très primitive.

Otocystes. — Les deux otocystes (PI. III, Fig. 10) sont situés en avant
et un peu au-dessus de la grosse commissure palléo-pédieuse : ils sont
ovoïdes, leur plus grand diamètre mesure .55 millimètres. Leurs parois
sont épaisses et laissent voir par transparence une cavité bourrée d'oto-
lithes hyalins, de dimensions très inégales, comparables par conséquent
à ceux qui ont été décrits par M. de Lacaze-Duthiers (72, 141) chez
les Patella; les petits (2 p à 7 p) sont exactement sphériques; les
gros (jusqu'à 15 p et 20 u) sont quelquefois sphériques, mais plus fré-
quemment ovoïdes allongés ou bosselés ; on ne saurait mieux les comparer
qu'aux concrétions qui se déposent concentriquement autour de plusieurs
centres d'attraction voisins ; deux, trois ou davantage; la surface extérieure,
au lieu d’être une sphère, est composée de plusieurs fragments de sphère

1 Pelseneer (94, 59).

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 215

qui correspondent chacun A un centre d’attraction; il semble done que
les gros otolithes bosselés soient formés par Paccroissement de deux ou
trois petits qui se sont accidentellement soudés.

Le nerf acoustique, après avoir quitté l’otocyste, remonte le long de la
corne antérieure des cordons scalariformes en adhérant très intimement
à la membrane conjonctive de cette corne ; nous n'avons pas pu le suivre
jusqu'aux ganglions cérébroïdes.

Le grand nombre et l'inégalité des otolithes, ainsi que leur structure
très simple, viennent encore accentuer les caractères primitifs des Pleuro-
tomaires ; on sait, en effet, que, chez les formes spécialisées, les otolithes
ont une taille uniforme, une structure plus ou moins compliquée et con-
stante et que leur nombre subit une réduction parfois considérable.

SYSTÈME NERVEUX. !

Par tous ses caractères essentiels, le système nerveux du Pleurotomaria
Quoyana ressemble & celui des autres Prosobranches diotocardes, surtout
à celui des Fissurellidés et des Trochidés ; ses ganglions cérébroïdes, ses
centres stomato-gastriques, sa commissure viscórale et ses cordons palléo-
pédieux sont construits sur le même type; différences, qui sont toutes de
détails, ont d’ailleurs leur importance et rapprochent, A notre avis, le
systeme nerveux des Pleurotomaires de celui des Amphineures.

1° Centres cérébroïdes (Fig. 2, 3, 8, 9, 10,0). — Les centres cérébroïdes
occupent la même place et ont la même forme que ceux des Trochidés ;
toutefois leur commissure est sensiblement plus large, ce qui tient A un
état de condensation moins avancé des cellules ganglionnaires. Cet état
primitif des centres cérébroides est rendu manifeste par l'étude des nerfs
labiaux supérieurs (m!) et latéraux (m?) ; ces nerfs, en effet, wont pas
leur origine dans les ganglions, mais, comme dans l’Haliotide sur la com-
missuro (c), au voisinage de la partie renflée des ganglions. Cette der-
nièro est triangulaire et, comme de coutume, se prolonge latéralement et
en-dessous, sur les côtés de la masse buccale, pour former une longue
saillie labiale (L); les connectifs cérébro-pédieux (%1) et cérébro-
palléaux (A?) naissent côte à côte sur le bord des ganglions, en arrière
de cette saillie.

Sur le bord antérieur des ganglions on voit naître, de haut en bas,

1 Nous présentons nos vifs remerciements à M. le professeur de Lacaze-Duthiers,
qui a fait recueillir pous nous, à Roscoff, de magnifiques Chiton fasciularis, à M. le
professeur Perrier, qui nous a communiqué des Turbo radiatus recueillis par M.
Jousseaume dans la mer Rouge. Comme on le verra plus loin, ces Mollusques
nous ont été fort utiles pour interpréter le système nerveux des Pleurotomaires.

216 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

trois nerfs labiaux (m, m,* m5) aussi puissants que les deux premiers ;
le troisième se détache de la saillie labiale dans sa région basilaire qui
est très renflée ; plus inféricurement, cette saillie émct encore un autre
nerf labial (m°), puis se rétrécit beaucoup et devient alors purement
fibreuse ; elle forme alors la commissure labiule (c?) qui ne présente rien
de particulier.

Sur la face externe des ganglions cérébroïdes prend naissance le puis-
sant nerf tentaculaire (t'); il émet, près de sa base, une petite brahche
nuquale et, un peu plus haut, lo nerf optique (f). Uva sans dire que ce
dernier est simplement accolé au nerf tontaculaire, mais qu'il n’est point
fusionné avec lui.

99 Stomato-gastrique. — Comme chez tous les Diotocardes, le système
nerveux stomato-gastrique a son origine sur le bord postérieur de la
saillie labiale (Z1) à une faible distance de sa baso; comme dans ces
derniers aussi, ses connectifs (k) sont allongés, remontent latéralement
sous les muscles dans les flancs de la masse buccale, et envoient, chemin
faisant, quelques ramuscules nerveux (s°) à ce dernier organe.

La partie ganglionnaire (Fig. 12) présente la forme normale caracté-
ristique du groupe, celle d’un fer à cheval. Mais ici, le fer à cheval ne
présente pas les renflements prononcés qu’on observe chez les Trochidés
et même chez certains Chitons (Fig. 6 et 7), et qui indiquent d jà un
commencement de condensation ganglionnaire ; à son origine connectivale,
il est un peu dilaté, mais, partout ailleurs, son diamètre est sensiblement
le même.

Les nerfs stomato-gastriques nous ont paru plus gros et moins nom-
breux que chez les autres Diotocardes ; nous n'avons pas pu tous les
suivre, en raison du mauvais état de la masse buccale, mais nous croyons
avoir observé l'origine de tous, soit sur l’animal, soit en préparations
microscopiques. En dehors des ramuscules issus des connectifs, ils sont
de chaque côté au nombre de trois : deux (s*, st) se détachent du ren-
flement antérieur et correspondent vrai-semblablement aux nerfs que
Pun de nous a désignés par les lettres s* et s* dans le Turbo (Bouvier,
187, Fig. 5), l'autre (s) se détache du bord postérieur du fer à cheval, au
voisinage de son millieu, et doit être l'équivalent des nerfs s'et s? du
même animal, Kun (sê) des deux nerfs antérieurs se dirige en arrière
et correspond aux trois nerfs (s*) des Turbo; il est très volumineux
et paraît renfermer des cellules nerveuses. Le nerf postérieur se bifurque
peu après son origine.

3° Connectifs issus des ganglions cérébroïdes et commissure viscérale
(Fig. 3, 8, 9, 10, 11). — Issus du bord postéro-inférieur des ganglions

|
|
|
|

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 217

cérébroides, le connectif cérébro-palléal (k*) et lo connectif cérébro-pédieux
(1) descendent, en divergeant, dans la dépression profonde produite en
avant par la cavité du corps (Fig. 3); le connectif cérébro-palléal est
en arrière et le connectif cérébro-pédieux en avant ; le premier est plus
puissant que le second.

A droite les deux connectifs atteignent leur maximum d’écartement
vers le tiers supérieur de leur longueur ; à gauche ils divergent sur une
plus grande étendue, mais, comme du cóté droit, se rapprochent ensuite
peu à peu, et viennent aboutir dans les cornes proximales (Fig. 9 et 10, 01)
que forment avant leur réunion les cordons palléo-pédieux. Chemin
faisant, les connectifs envoient quelques branches nerveuses aux parois
du corps; les connectifs cérébraux-pédieux en émettent au moins deux
(e, b, W) et les connectifs cérébro-palléaux une (e, e+). Ces derniers,
d’ailleurs, présentent le caractère tout particulier de donner naissance,
sur leur trajet, à la branche correspondante de la commissure viscórale.
A droite, la branche sus-intestinale (h) se détache du connectif cérébro-pal-
léal (&?) au point où celui-ci s'écarte le plus du connectif cérébro-pédioux,
c’est-à-dire à une faible distance du ganglion cérébroïde droit (C). A
gaucho, la branche sous-intestinale (h *) se détache plus bas du connectif
céróbro-palléal (A!) un peu au-dessous du millieu de ce dernier. Quoiqu'il
eu soit, on observe ici ce fait unique, chez les Gastéropodes, d’une commis-
sure viscérale issue, non point des centres palléaux, mais des connectifs
qui réunissent les ganglions cérébroïdes à ces centres.

4° Cordons palléo-pédieux (Fig. 4, 8, 9, 11). — Chez les autres Gas-
téropodes diotocardes (voir Fig. 5 et 13), chaque cordon palléo-pédieux
se prolonge proximalement, au-dessus de la grosse commissure, par une
corne ganglionnaire palléale (0°) qui se continue avec les connectifs (#1, °)
issus du cerveau, et la commissure viscérale vient aboutir dans une autre
corne (Cy, Cd.) qui provient de la bifurcation de la première, Dans le
Pleurotomaria Quoyana, il n’en est plus de même : la corne (C1) de chaque
cordon, fort grosse et très allongée, ne se bifurque pas et ne donne pas
naissance à la commissure viscérale ; elle se dirige obliquement en arrière
et en haut et les connectifs (k', k*) issus du terveau viennent aboutir à son
extrémité supérieure,

D'ailleurs, ces cornes ganglionnaires ne sont pas exclusivement de na-
ture palléale ; sur leur face externe (Fig. 4 et 11) elles sont creusées
d'un sillon profond (s) qui les divise en deux rubans superposés ; le ruban
supérieur (Ps) continue exactement le connectif cérébro-palléal (Æ?)
correspondant, et le ruban inférieur (Pr) le comnoctif cérébro-pédieux
(%*). Il est done manifeste que les cornes du Pleurotomaire, au lieu

218 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

d’ötre exclusivement palléales, sont de nature mixte, que la bande supé-
ricure est de nature palléale et l'inférieure de nature pédieuse. C’est ce
que prouve, du reste, l’étude des nerfs qui émanent de cette région ; les
nerfs (a, al, a”) issus du ruban supérieur se rendent, en effet, dans les
muscles qui forment le plancher de la cavité antérieure du corps, et qui
sont nettement columellaires, ceux qu'émet le ruban inférieur, s’enfoncent,
au contraire, dans la musculature du pied ou dans la paroi du corps.

Or, s'il est impossible de contester la nature mixte des cornes proxi-
males des cordons pédieux, il faut en conclure que les cordons sont
mixtes, comme les cornes elles-mémes. En effet, le profond sillon que
nous avons constaté sur la face externe des cornes se prolonge manifeste-
ment, sinon sur toute la longueur des cordons, au moins sur une grande
étendue de ces derniers; il apparaît même, mais beaucoup moins nette-
ment, sur leur face interne (Fig. 11). Comme les cornes proximales,
chaque cordon se trouve dés lors partagé en un ruban supérieur qui
continue le ruban palléal des cornes et en un ruban inférieur qui
continue leur ruban pédieux.

Ainsi, les deux rubans de chaque cordon, en effet, ne sont certaine-
ment pas de nature identique, du moins au point de vue de leurs rap-
ports et de leurs fonctions. La preuve en est encore dans la position
des commissures qui les relient nn côté à l’autre et dans la nature des
nerfs qu'ils émettent.

Les cordons se dirigent d'avant en arriere dans le pied (Fig. 9), oü
ils figurent presque un ovale très peu convexe. Au point où ils se
continuent avec les cornes, ils sont unis par une commissure puissante
(Fig. 11, 2), à la fois ganglionnaire et fibreuse, qui s’étend entre les deux
rubans de chaque côté. Plus en arrière se trouvent d'autres commis-
sures exclusivement fibreuses; nous en avons pu réparer neuf ;? mais
leur nombre est certainement plus considérable. Elles sont moins régu-
lièrement transversales que la première et émettent des ramuscules qui
se rendent dans la sole pédieuse. Ces commissures sont done vraiment
de nature pédieuse; d’ailleurs, dans le Pleurotomaire, comme dans les
Fissurolles, les Haliotides et les Trochidés, elles paraissent indépendantes
des rubans supérieurs et rattachent l’un à l’autre les rubans inférieurs.
Ce fait prouve, à notre avis, que les deux rubans de chaque côté ne sont
pas de même nature.

Un autre fait vient encore rendre plus évidente cette manière de voir,

1 Ces commissures sont représentées dans la figure 9; en arrière de la neuvième,
les coupes nous en ont fait voir cinq autres, ce qui porterait à quatorze le nombre
des commissures accessoires.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 219

Comme dans les autres Diotocardes á cordons divisés en deux parties
par un sillon, le ruban supérieur et le ruban inférieur émettent des
nerfs qui n’ont pas la même distribution. Ainsi que lont montré
M. de Lacaze Duthiers dans l’Haliotide (°59), M. Boutan (86) dans la
Fissurelle et le Parmophora, M. Béla Haller dans divers Diotocardes
(84, 20, Fig. 2), les rubans supérieurs envoient des nerfs (a°, a?, a?*,) en
avant dans les parois du corps, plus en arrière dans la partie supérieure
du pied, qui est de nature columellaire, ainsi que dans Pépipodium qui
Vorne de ses franges. Toutefois, dans notre Pleurotomaire, nous n’avons
pu suivre ces nerfs jusqu'à l’épipodium, réduit et recroquevillé, de
Vanimal. Entre les nerfs antérieurs issus des rubans supérieurs des
cordons, et ceux qui proviennent du connectif cérébro-palléal ou du
ruban qui lui fait suite dans les cornes, il y a tous les passages et le
champ de distribution est le même, comprenant les parois latérales du
corps et tout ce qui, de près ou de loin, fait partie de la région collu-
mellaire. Tout autre est la distribution des nombreux nerfs issus des
rubans inférieurs des cordons; ces nerfs (p) se rendent dans la sole
inférieure du pied et sont manifestement de nature pédieuse; en
outre, comme chez les autres Diotocardes, les deux nerfs les plus an-
térieurs (p?) sont très gros et se rendent dans la partie antérieure du
pied.

En résumant ce qui précède, on voit: 1° que les cordons palléo-pédieux
sont le prolongement direct de leurs cornes antérieures ; 2° que les deux
rubans qui les constituent de chaque côté sont aussi les prolongements
directs des rubans des cornes ; 3° que les rubans supérieurs sont la con-
tinuation des connectifs cérébro-palléaux et les rubans inférieurs la con-
tinuation des connectifs cérébro-pédieux ; 4° que, dans les cordons comme
dans les cornes, les rubans supérieurs envoient des nerfs aux parois du
corps, au muscle columellaire on à ses dépendances, tandis que les rubans
inférieurs donnent surtout des nerfs exclusivement pédieux ; 5° enfin
que les rubans supérieurs paraissent indépendants l’un de Pautre, en
arrière de la commissure antérieure, tandis que les rubans inférieurs
sont réunis par de nombreuses commissures d’où partent des rameaux
pédieux.

En conséquence, nous croyons pouvoir dire que les cordons, comme
les cornes qui les terminent, sont de nature mixte, que leur partie
inférieure est pédieuse, et que leur partie supérieure est palléale, en
donnant au mot palléal une signification étendue et que nous préciserons
plus loin,

La structure mixte des cordons est rendue tres manifeste par la

220 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

présence de certains nerfs (pn) qu'on voit se détacher sur leur face
externe et qui sont mixtes dans leur nature, comme le montrent les
coupes, ou, plus simplement, un examen à la loupe. Ces nerfs sont
surtout communs et faciles & observer dans la partie antéricure des
cordons; quand on examine ceux-ci par la face externe, on voit que
leur sillon est interrompu, de distance en distance, par des saillies
linéaires obliques qui prennent leur origine dans les rubans supérieurs,
se dirigent vers les rubans inférieurs, et, après les avoir atteints, se
fusionnent, pour former un nerf, avec une racine nerveuse 6manée de ces
derniers. Ces nerfs mixtes se rendent dans le plan moyen du pied, et,
bien que nous n’ayons pu suivre leur trajet jusqu’au bout, doivent avoir
une distribution mixte, certains de leurs ramuscules se rendant à la
partie supérieure et columellaire du pied, les autres dans sa partie infé-
rieure ou pédieuse. L'existence de ces nerfs prouve, une fois de plus,
que des parties contiguës, mais de nature différente, peuvent confondre
leurs organes d'innervation tout en restant distinctes dans leur origine
et dans leur distribution. Il en est de même des nerfs acoustiques :
dans le Pleurotomaria Quoyana, par exemple, les octocystes (0) sont situées,
comme dans les autres Diotocardes, sur le bord antérieur de la grosse
commissure pédieuse (4), où un tractus fibreux les réunit, aussi les nerfs
acoustiques (0!) se confondent-ils avec les connectifs cérébro-palléaux, et
il nous a même été impossible, dans notre animal, de les suivre jusqu'aux
ganglions cérébroïdes, où ils prennent pourtant leur origine. Cette loi
a été rendue parfaitement évidente par M. de Lacaze-Duthiers (72) en
ce qui concerne les nerfs acoustiques des Gastéropodes, mais l'étude du
système nerveux des Mollusques en fournirait sans difficultés d’autres
exemples.

5° Commissure viscérale, nerfs palléaux (Fig. 3, 8, 9, 11). — Ainsi que
nous l'avons fait remarquer plus haut, les branches de la commissure
viscérale ont une origine connectivale qu’on n’observe nulle part ailleurs
chez les Gastéropodes, et qui donne au système nerveux du Pleurotomaria
Quoyana un aspect des plus caractéristiques. La branche sus-intestinale
(h) se détache du connectif cérébro-palléal (4?) à une faible distance du
ganglion cérébroïde droit; elle se dirige en arrière en suivant la paroi
dorsale du corps, puis se recourbe progressivement à gauche et passe
au-dessus du jabot de Panimal. Dans notre animal, cet organe était
en mauvais état et réduit à ses parois déchirées et affaissées ; au point
où s’effectua la rupture quand on arracha le Mollusque de la coquille,
s'étaient produites des lésions qui, heureusement, n’avaient pas enlevé
la ‘branche nerveuse; nous avons pu la suivre jusqu'au point où

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 2e

les parois dorsales du corps cessaient d'existor, comme l'indique la
figure.

Quant à la branche sous-intestinale (h") elle se détache beaucoup plus
bas du connectif cérébro-palléal gauche (4°), nous l'avons vu se diriger à
droite sous le tube digestif. muis elle était rompue à peu près au niveau
de ce dernier.

En tous cas, nos recherches sont suffisantes pour établir que les Pleu-
rotomaires possèdent, comme les autres Prosobranches, une commissure
viscérale croisée.

Nous avons ¿tó moins heureux dans notre étude des nerfs palléaux
proprement dits, et c’est une lacune que nous sommes les premiers à
déplorer. Ordinairement ces nerfs sont volumineux et d’une dissection
des plus faciles, au moins à leur origine; ici, nous ne sommes pas
certains de les avoir aperçus ; à droite, nous ne voyons rien qui puisse
leur correspondre ; mais, à gauche, nous considérons comme nerf palléal
une branche nerver.e (Fig. 8, m) qui se détache du connectif cérébro-
palléal presque au même point que la commissure viscérale. Au reste,
comme on peut s’en convaincre en jetant un coup d'œil sur la figure 1,
qui représente l'animal tel qu’il nous fut communiqué, c'est à peine si
notre Pleurotomaire présentait les restes de la partie inférieure du man-
teau, et nous nous demandons si les déchirures n'auraient pas arraché,
jusqu'à leur racine, les grands nerfs palléaux. Si Pon considère cette
hypothèse comme improbable, et nous convenons qu’elle en a toutes les
apparences, il faut admettre que nous n'avons pas su apercevoir, au
moins à droite, les vraies origines palléales. On sait que, chez les autres
Diotocardes, les grands nerfs palléaux se détachent des cornes palléales
supérieures ou dans le voisinage de celles-ci.

De la branche sus-intestinale de la commissure viscérale, nous n'avons
vu naître qu’un nerf assez grêle (Fig. 3 et 8, d), qui se rendait en arrière
dans les tégumonts dorsaux. De la branche sous-intestinale naissaient
deux nerfs columellaires postérieurs (l, 1), dont Pun était très volumineux.
Entre le point où elle se détache du connectif cérébro-palléal gauche et
la corne palléo-pédieuse correspondante, cette branche est très volumi-
neuse; nous avons vu qu'elle émet, au premier de ces points, un nerf
assez puissant (Fig. 8 et 9, m), qui plonge, à gauche, dans les parois du
corps, et qu’on doit considérer, à notre avis, comme un nerf palléal
gauche,

1 Peut-être faut-il considérer comme les nerfs palléaux deux gros nerfs (Fig.
11, a2, al2) qui naissent des rubans eupérieurs un peu en arrière de la commissure
palléo-pédieuse.

222 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

NATURE DES Centres NERVEUX DU Prep CHEZ LES GASTÉROPODES.

Les faits qui précèdent nous paraissent propres à jeter la lumière sur
un point de l’histoire naturelle des Mollusques qui a soulevé les plus
nombreuses et les plus vives controverses, nous voulons parler de la
nature des cordons ganglionnaires du pied des Diotocardes et de l’éten-
due que présentent les centres palléaux de ces Gastéropodes. Sans
vouloir remonter au début de cette controverse, ce qui n’aurait aucune
utilité, nous dirons que M. de Lacaze-Duthiers (59, '90), et, à sa suite,
MM. Wegmann (’84) et Boutan (86, '88, "90, ’98) considèrent les cordons
du pied des Diotocardes comme formés par la fusion, suivant leur
longueur, des centres palldaux et pédieux de chaque côté, tandis que
MM. Spengel (81), Béla Haller (84, ’89, 94), Thiele (90) et Pelseneer
(87, *88, 90, 91), les regardent comme franchement pédieux et con-
sidèrent comme centres palléaux les cornes ganglionnaires qui les précè-
dent et qui donnent naissance à la commissure viscérale et aux nerfs
pallóaux. Abstraction faite de la nature de l’Epipodium, sur laquelle
nous reviendrons plus loin, les arguments essentiels fournis par les
premiers sont les suivants : 1° les cordons du pied sont divisés en deux
rubans superposés par un sillon externe; 2° le ruban inférieur est en
relation avec le connectif cérébro-pedieux et le ruban supérieur, plus ou
moins directement, avec le connectif cérébro-palléal et les nerfs palléaux.
Les arguments de la partie adverse sont essentiellement tirés de lexis-
tence d’un vrai ganglion palléal chez les Diotocardes, ganglion qui se
trouve dans la région proximale des cordons, en avant de la grande
commissure antérieure ; si l’on admettait, disent-ils avec M. Pelseneer,
que ces ganglions se continuent dans le ruban supérieur des cordons, il
faudrait admettre aussi que les ganglions palléaux sont commissurés, ce
qui n’existe pas chez les Mollusques. Les arguments de second ordre
seront signalés plus loin.

Pour notre part, nous croyons que la première théorie est l'expression
exacte de la vérité, mais que la seconde est loin de s’en éloigner, au fond,
autant qu'on pourrait le croire. Commençons d’abord par préciser, aussi
exactement que possible, le sens qu'on donne communément au terme
de palléal où à son équivalent, celui de pleural, en anatomie comparée
des Mollusques.

Si l'on s’en tenait à l'étymologie même des mots palléal ou pleural, on
devrait réserver les noms de ganglions palléaux ou de ganglions pleuraux
aux ganglions qui président à l’innervation du pallium on manteau et

i 4 0 + A
des parois du corps des Mollusques. Mais, en réalité, ces termes ont dú,

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 228

par la force des choses, recevoir une signification plus large. Chez les
Gastéropodes, oü les ganglions palldaux (ganglions pleuraux) sont bien
distincts — et c'est le cas chez tous les Monotocardes —ces ganglions
donnent naissance aux nerfs du manteau ou vrais nerfs palléaux, A cer-
tains nerfs des parois antérieures du corps et à certains nerfs des muscles
columellaires. Quil y ait ou non communauté d’origine entre ces der-
nióres parties, il est certain que les nerfs qui les innervent présentent des
caractères communs, non seulement à cause de leur origine dans un même
centre, mais aussi en raison des anastomoses qwils contractent fréquem-
ment ou de leur champ de distribution, qui est mixte: le plus souvent, en
effet, sinon toujours, on voit les nerfs palléaux envoyer des rameaux dans
les parties du muscle columellaire et des parois du corps qu’ils traversent.
Quoi qu’il en soit, il ressort de ce qui précède que le champ d'innervation
des centres palléaux appelés aussi centres pleuraux, est des plus vastes et
s'ótend bien au-delà du manteau proprement dit ou des parois du corps.

1° Pleurotomatres. — Ceci dit, revenons au Pleurotomaire. L'intérêt
essentiel de ce curieux Gastéropode, c’est qu'il ne présente pas, comme
les autres Diotocardes, de corne palléale distincte en avant des cordons
ganglionnaires du pied; de sorte que, si l’on interprétait au sens étroit
la théorie de M. Haller et de M. Pelsencer, il faudrait dire que le Pleu-
rotomaire est dépourvu de centres palléaux. A priori, cette interprétation
est inexacto, car notre Mollusque a un manteau, un muscle columellaire
et des parois du corps aussi développées que les Haliotis ou les Trochus,
de sorte qu’il doit avoir comme eux, c'est évident, des ganglions et des
nerfs palléaux. Aussi bien, nous voulons rejeter cette interpretation,
que seul, peut-être, M. Thiele! pourrait accepter, et nous croyons entrer
tout a fait dans les vues de MM. Haller et Pelseneer, en disant que les
centres palléaux sont formés, en partie du moins, par le ruban supérieur
(Fig. 4 et 11, Ps) de la corne unique, qui prolonge proximalement de
chaque cóté les cordons pédieux. Nous avons vu, en effet, que ce ruban
est le prolongement direct du connectif cérébro-palléal (42) et qu’il émet,
par l'intermédiaire de ce connectif, la commissure viscérale, des nerfs
pariétaux, des nerfs palléaux proprement dits et des nerfs columellaires.
Quant au ruban inférieur (Pi), qui se continue par le connectif cérébro-

! Thiele ('92), en effet, ne considère point comme des centres palléaux les
cordons latéraux des Chitons, mais il les assimile au plexus épipodial de l’Haliotide,
sans d’ailleurs en donner de raison sérieuse. Pour lui, les ganglions palléaux se
développeraient ultérieurement chez les Prosobranches, Cette opinion, qu’aucun
auteur n’a pu adopter, est en contradiction manifeste avec les faits signalés dans
le présent mémoire.

VOL. XXXIIL — NO, 10. 3

224 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

pédieux (4), aucun fait ne saurait nous permettre de le considérer
comme palléal, et ses connexions indiquent manifestement qu’il est de
nature pédieuse.

Nous voici, dès lors, en présence de cornes ganglionnaires (ec!) qui sont
mixtes á tous égards, palldales dans leur ruban supéricur et pédicuses dans
leur ruban inférieur. Or, chacune de ces cornes est le prolongement
d’un cordon ganglionnaire du pied, et chacun des rubans qui les compose
est le prolongement du ruban correspondant de chaque cordon. Si les
cornes sont palléales dans leur partie supérieure et pédieuses dans leur
partie inférieure, tout semble dès lors indiquer qu’il en est de même des
cordons. En fait, c'est ce que montre également l'étude des nerfs émis
parles rubans qui constituent chacun d'eux. Ainsi qu’on le savait
depuis longtemps, et comme on peut s'en convaincre par les figures qu’a
données M. Béla Haller ('84, 38, Fig. 4), le pied des Diotocardes et des
autres Gastéropodes se compose de deux parties: Pune, supérieure, qui
est formée par les fibres musculaires antérieures du muscle columellaire
et qui recouvre les cordons ; l'autre, inférieure, qui constitue la sole
pédieuse. Or les nerfs issus du ruban supérieur se rendent dans la
partie supérieure du pied, et, comme il est facile de s’en convaincre par la
dissection des nerfs les plus antérieurs des cordons, dans sa partie colu-
mellaire ; tandis que les nerfs émis par le ruban inférieur se rendent en
divers points de la sole pédieuse. En d’autres termes, le ruban supérieur
des cordons innerve la zone columellaire comme le ruban supérieur des
cornes, dont il est impossible, d’ailleurs, de le séparer ; il est dès lors,
comme lui, de nature palléale.

Ainsi dans le Pleurotomaire, chaque cordon du pied se compose d’un
cordon de nature palléale et d’un cordon de nature pédieuse, de sorte
que ces centres ganglionnaires sont mixtes, sinon de leur naissance &
leur extrémité, au moins sur une grande partie de leur étendue.

Ceci étant établi, et la grosse commissure des cordons réunissant aussi
bien les rubans supérieurs que les rubans inférieurs, il faut conclure de
ce qui précède que les centres palléaux sont commissurés au même titre que
les centres pédieux. C’est un fait contre lequel, dans le Pleurotomaire, du
moins, il nous paraît difficile de s'élever. Il n’est donc pas juste de
prétendre, avec M. Pelsencer (88, 91) que les ganglions palléaux ne sont

1 Il est possible, en effet, que la partie palléale ne s’étende pas jusqu’au bout des
cordons, et c’est ce que sembleraient indiquer les centres pédieux des Fissurellidés,
Pour être fixé sur ce point, il faudrait étudier l’exacte distribution des fibres du
muscle columellaire et connaître le point précis où s’arréte le sillon en arriere.
Cette observation s’applique à tous les Diotocardes.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 225

jamais commissurés ; sans doute, la commissure qui réunit les ganglions
nest jamais distincte de la commissure pédieuse, mais les recherches sur
la structure intime des centres nerveux des Gastéropodes permettent
d'établir qu'il existe, dans la commissure pédieuse, des fibres, qui se ren-
dent aux ganglions palléaux transversales. Au reste, l'exemple du
Pleurotomaire suffirait, à lui seul, pour montrer qu'il en est ainsi,
même si l’on admettait que la partie palléale de ce mollusque se limite
au ruban supérieur des cornes. Dans ce cas, en effet, le centre palléal
des cornes s’arréterait au niveau de la grosse commissure des cordons,
et, comme le montrent les coupes, recevrait quand méme des fibres de
cette commissure.

2° Fissurellides, Haliotidés, Trochidés. — Le système nerveux des
Diotocardes primitifs diffère surtout de celui des Pleurotomaires, par
la présence de deux cornes proximales (Fig. 5 et 13), à l'extrémité an-
térieure de chaque cordon du pied, l’une supérieure (Cg, Od), que chacun
s'accorde & reconnaitre comme palléale et qui sert de point de départ aux
branches de la commissure viscérale ; l’autre inférieure (C1), à laquelle
aboutissent les connectifs cérébro-palléaux et cérébro-pédieux. Ces deux
cornes ont été désignées par M. Béla Haller (84, 17) sous le nom de centres
pleuro-cérébraux, le méme auteur donne le nom de centre pleural ou com-
missural, c'est-à-dire de centre palléal,* à chaque corne supérieure ; mais,
comme il passe outre sur les cornes inférieures, on doit croire, sinon qu’il
les tient tout entières pour pleurales, du moins qu’il ne leur attribue,
à aucun degré, la nature pédieuse. Au surplus, il est nécessaire de rap-
peler que les partisans de la théorie des cordons pédieux simples ne
cherchent pas à caractériser la nature des cornes inférieures; M. Béla
Haller les a bien figurées, mais M. Pelseneer, qui a consacré le plus de
travaux & la question, ne les mentionne méme pas, et, pourtant, elles
sont très développées dans les Fissurelles ("86, Pl. XXXV), dans les
Cemoria (Haller, ’94, Fig. 142, voir Pl. II, Fig. 13), un peu moins dans
certains Turbo (Fig. 95) et Trochus ; dans les Haliotides, elles sont à
peine sensibles. Or, toutes les fois qu’elles existent, ces cornes ser-
vent, comme dans les Pleurotomaires, de point d’arrivée aux connectifs
cérébro-palléaux et cérébro-pédieux, et, comme chez ces derniers aussi,
elles sont parcourues par un sillon longitudinal et divisées en deux
rubans qui continuent ceux des cordons du pied. Ces rubans et ce sillon
se voient surtout bien dans la Fissurelle, ainsi qu’il résulte des figures de

1 Spengel (’81) a donné le nom de ganglions pleuraux aux centres que nous
appellons ganglions palléaux; M. von Ihering les nomme aussi ganglions commis-

surauz (717).

226 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

M. Boutan (’86, Fig. 10, Pl. XXXV), et, comme ils ressemblent en
tout A ceux des Pleurotomaires, comme les cordons offrent avec eux
les mémes rapports, il y a licu, croyons-nous, de leur accorder la méme
signification.

Toutefois, il existe ici une différence importante, qui est précisément
la cause de la controverse qui nous occupe; uno partie de la substance
palléale des cornes et des cordons s’est accumulée à l’origine de la com-
missure viscéralo pour former des ganglions palléaux déja distincts sous
la forme de cornes supéricures, en méme temps qwune partie de la sub-
stance pédieuse des cornes inférieures se concentrait dans le ruban infé-
rieur des cordons. Dela le développement réduit des cornes inférieurs
des -Fissurelles, Turbo, Trochus, etc., qui sont beaucoup moindres que les
cornes uniques des Pleurotomaires.

Ainsi, chez les animaux qui nous occupent, le système ganglionnaire
palléal tend à se localiser en deux masses: l’une antérieure, qui s'isole
et émettra des nerfs pour les parties palléales de la région antérieure du
corps; l’autre, en contact sur toute sa longueur avec les centres pédieux,
qui desservira une partie des parois du corps, et surtout la partie du
muscle columellaire qui forme les plans supérieurs du pied. Aux nerfs
issus de cette partie palléale située dans le pied, M. Béla Haller a donné
le nom de nerfs latéraux, ils se rendent, dit-il, dans les parois du corps,
et les plus antérieurs dans le muscle columellaire ('84, 21).

39 Patellidés, Néritidés, Cyclophoridés, Paludinidés, Cypræidés. — Le
processus de concentration, dont nous venons de voir un exemple dans
les formes précédentes, s'accentuo davantage encore chez les Patellidés,
chez les Néritidés et chez certains Monotocardes primitifs (voir p. 238,
Fig. E). Les ganglions palléaux destinés aux parties palléales de la

Es fo

région antérieure du corps (manteau, partie du muscle columellaire la
plus voisine de la coquille, une partie des parois du corps) se séparent de
plus en plus des cordons et forment les ganglions palléaux proprement
dits ; ils isolent ainsi du ruban palléal supérieur, qui restera logé dans le
pied, au voisinage immédiat des parties qu’il innerve (parois les plus an-
térieures du corps, parties du muscle columellaire situées dans le pied).
Dès lors, les rubans pédieux et palléaux de chaque cordon, destinés à
innerver des parties conticués se fusionnent et se concentrent de plus en
plus, leur sillon de séparation s’efface, et Pon arrive ainsi à des cordons
simples en apparence, mixtes en réalité,
donné le nom de cordons pédieux, bien qu’ils renferment une partie pal-

auxquels tous les auteurs ont

léale et une partie pédieuse. Ces cordons ne diffèrent guère de ceux des

Diotocardes primitifs que par Patrophie de leurs cornes inférieures, Visole-

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. oa

ment plus grand de leurs cornes supérieures appelées ganglions palléaua,
la disparition de leur sillon et la moindre quantité de cellules palléales
qwils renferment. Pour le reste, tout est semblable, et la fusion dont
leurs deux parties constituantes ont été le sióge n’est rien autre chose
quo le rósultat de la concentration de centres ganglionnaires contigus,
concentration dont les exemples abondent dans Phistoire des Gas-
téropodes,!
4° Autres Gastéropodes. — Des formes précédentes on passe sans difti-
culté, par une concentration plus grande encore, aux Gastéropodes
dépourvus de cordons ganglionnaires dans le pied (voir p. 238, Fig. 1).
Les ganglions palléaux proprement dits se séparent nettement de la
région pédieuse, et les cordons de cette dernière se condensent d'arrière
en avant pour former les ganglions pedieux, arrondis ou ovoïdes, des
divers auteurs. Mais ces prétendus ganglions pédieux sont mixtes, en
réalité; ils innervent, comme on sait, les parties pédicuses et columel-
laires de pied et n’ont point perdu toute relation avec les ganglions pal-
léaux proprement dits. Ces derniers, en effet, comme chez le Pleuroto-
maire et les Diotocardes, sont commissurés à travers les ganglions et la
commissure pédieuse. Il y a longtemps qu’on s’était aperçu de ce fait,
et Walter en a donné un bel exemple dans ses études micrographiques
sur le système nerveux de la Lymnde (63, Taf. IV, Fig. 1); on trouve
quelque chose d'analogue dans le travail de M. Gamault (87, Fig. 32)
sur le Cyclostome.
En résumé, nous pensons, avec MM. de Lacaze-Duthiers et Boutan,

ie les cordons du pied des Fissurelles, des Haliotides, des Zurdo et des

ql

semblable & celui des Chitons

Son Cquiv:

» fait nest point rare dans inner

ion palle

exagérer que de transformer en cordons palléa

de M. Boutan n’e

Patellidé L’interpre

ce n’est pas entre les Patellides et les Fissurellides que

Helcion, mais bien entre les Chitons et les Pleurotomaires

228 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Troques, sont palléaux dans leur partie supérieure, pédieux dans leur
moitié inférieure; avec MM. Haller, Pelseneer, Thiele, ete., qu'ils sont
les homologues des cordons pédicux des Patelles, des Nérites, des Cyclo-
phores, des Paludinés, des Cyprées, et des ganglions pédieux de tous les
autres Gastéropodes, abstraction faite de la masse ganglionnaire palléale
un peu plus abondante qu’ils renferment. Nous différons des uns et des
autres en attribuant une nature mixte, à la fois palléale et pédieuse, aux
centres ganglionnaires (ganglions pédieux de tous les auteurs), condensés
ou non, qui envoient des nerfs à la masse musculeuse complexe qu'on
désigne sous le nom de pied chez tous les Gastéropodes. Les Pleuroto-
maires se distinguent de tous les autres Mollusques de la classe en ce que
la masse ganglionnaire palléale ne tend pas encore à isoler sa partie
antérieure, et reste en relation, dans toute son étendue, avec la masse

ranelionnaire pédieuse.
o

NATURE DE L'ÉPIPODIUM DES PROSOBRANCHES.

La comparaison des Diotocardes avec les Pleurotomaires et les Mono-
tocardes nous ayant conduit à établir la nature mixte des cordons ou des
ganglions qui se rendent au pied des Gastéropodes, il nous est facile
maintenant d'interpréter comme il convient la frange pariétale connue
sous le nom d’épipodium qui borde la partie supérieure du pied, chez
la plupart des Diotocardes et chez certains Monotocardos tels que la
Janthine.

Il est un fait bien établi par tous les auteurs, c'est que les nerfs de
Pépipodium sont précisément les nerfs latéraux de MM. Béla Haller
(84, 20) qui naissent du ruban supérieur du cordon palléo-pédieux.
M. de Lacaze-Duthiers l’a, depuis longtemps, montré dans l’Haliotide
(59), plus récemment, M. Boutan a observé la même disposition chez
la Fissurelle, (86) et M. Béla Haller (’84, 94) chez les Diotocardes qu’il
a étudi s.

M. Pelsencer (90, 91), qui a consacré plusieurs mémoires à cette
question, ne met pas en doute ce fait, mais il en conteste la valeur en
objectant que les nerfs de la partie antérieure de Pépipodium chez les
Diotocardes, partent du connectif cérébro-pédieux et que, parmi ceux de
la partie postéricure, certains sont mixtes et prennent à la fois leur origine
dans les deux rubans des cordons. Pour ce qui est du premier point,
l'observation de M. Pelseneer est exacte, mais n’a pas Pimportance que
lui attribue cet auteur. Si Pépipodium “était palléal, dit M. Pelseneer,
il est bien évident que, dans cette région, il devrait être innervé partiel-
lement par le connectif cérébro-pleural.” Cet argument no nous parait

|
|

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 229

pas fondé, car les connectifs sont formés de fibres nerveuses, et ne
peuvent dès lors, donner naissance à des nerfs. Il s'agirait de savoir si
les fibres épipodiales qui se détachent des connectifs cérébro-pédieux
prennent naissance dans le ruban supérieur ou dans le ruban inférieur
du cordon et, jusqu'ici, personne ne nous a fixés sur ce point. Mais si
Von observe que les nerfs épipodiaux postérieurs se détachent tous du ru-
ban supérieur, il y a des raisons pour croire que ceux issus des connectifs
cérébro-pédieux prennent aussi leur origine dans le même ruban et que
s'ils émergent des connectifs précédents, c’est que ceux-ci sont externes
par rapport aux connectifs cérébro-palléaux et, par conséquent, plus
rapprochés de l’épipodium. Il ne faut pas oublier, en effet, que les nerfs
se rendent, par la voie la plus directe, à leur champ de distribution par-
ticulier et, ici, cette voie ne saurait être que celle des connectifs cérébro-
pédieux. Si, comme l’affirme M. Thiele (’92), certains nerfs épipodiaux
paraissent se d tacher des ganglions cérébroïdes, leur origine réelle devrait
être, à notre avis, dans le ruban supérieur des cordons.

Quant aux nerfs épipodiaux qui, d’après M. Pelseneer, prendraient
leur origine à la fois dans les deux rubans de chaque cordon, il y a lieu
de croire que ce sont des nerfs mixtes, dont certaines branches se rendent
à Vépipodium et d’autres dans les parties avoisinantes de la sole pédieuse.
La dissection de toutes les branches d’un nerf du pied étant des plus
pénibles, il ne sera pas facile de donner des arguments pour ou contre
cette manière de voir; mais ce qu’il y a de bien certain, c'est que, chez
les Diotocardes très primitifs, tels que les Pleurotomaires, les nerfs
mixtes sont beaucoup moins concentrés que dans les autres formes du
groupe, et présentent encore au dehors deux racines, l’une qui se détache
du ruban inférieur, l’autre qui émerge du ruban supérieur.

Chez les Janthines, ainsi que l’un de nous (Bouvier, ’86, 91, Pl. III,
Fig. 2) Pa établi, les nerfs de l’épipodium naissent tous des deux ganglions
pédieux et paraissent dès lors être de nature exclusivement pédieuse.
Mais cette apparence n’est point d'accord avec la réalité: les nerfs de
l’epipodium des Janthines sont les plus supérieurs du pied, comme ceux
de la partie columellaire pédieuse, et sont dès lors de nature palléale
comme ces derniers. Cela revient à dire, comme nous l'avons établi
précédemment, que les ganglions du pied sont de nature mixte, à la fois
palléaux et pédieux.

Nous laissons de côté les formations latérales des Patellidés du genre
Lottia (Helcion) qui sont innervées par le cordon pédieux. M. Pelseneer
(90, 152) tient ces formations pour épipodiales, mais M. Béla Haller
(94, 73-76) et tout récemment M. Boutan (98) ont établi qu’elles sont

230 BULLETIN: MUSEUM OF COMPARATIVE ZOÜLOGY.

de nature glandulaire et sans homologie aucune avec l’épipodium. Fát-
elle méme épipodiale, son innervation par les vordons du pied ne prouve-
rait nullement la nature exclusivement pédieuse de l’épipodium, car les
cordons du pied sont, nous Pavons dit, à la fois palléaux et pedieux.

De ce qui précède, il résulte qu’on doit, avec MM. de Lacaze-Duthiers
et Boutan, considérer l'épipodium des Prosobranches comme étant de
nature palléale, en donnant à ce mot la signification que nous lui avons
attribuée précédemment. Tant qu'on n’aura pas établi qu’il provient
d’un dédoublement du manteau, on ne pourra pas dire que l’épipodium
a la même origine et la même nature que ce dernier, mais il est au moins
pleural comme le manteau lui-même, comme le muscle columellaire et
comme les régions des parois du corps qu'innervent les ganglions palléaux
isolés,

Un mot encore, pour en finir avec cette question de l’épipodium.
D’après MM. Béla Haller et Pelseneer, la division des cordons du pied
des Diotocardes en deux rubans superposés serait due à la présence des
nerfs épipodiaux qui se détachent de la partie supérieure des cordons,
tandis que les autres nerfs du pied prennent naissance dans leur partie
inférieure.

Pour nous, cette s paration des deux sortes de nerfs montre tout sim-
plement que les deux rubans des cordons sont des centres ganglionnaires
distincts et la preuve, c’est que le sillon qui les sépare n’est nulle part
plus accentué que chez le Pleurotomaria Quoyana, espèce dont l'épipodium
est, comme on sait, très peu développé, Si la séparation des rubans
atteint son maximum chez les Prosobranches les plus primitifs, en dépit
de leur épipodium très réduit, c’est que, selon toute vraisemblance, ces
rubans ne sont autre chose que des centres ganglionnaires primitivement
séparés qui se sont rapprochés peu à peu, et ont Jini par se fusionner suivant
leur longeur. Au début, la ligne de démarcation des deux sortes de centres
s’est trouvée indiquée par un sillon large et profond, puis la concentration
s’accentuant, le sillon a disparu peu à peu, en même temps que s'isolart
en avant UNE PARTIE de la substance ganglionnaire palléale sous la forme
de ganglions palléaux distincts.

ORIGINE DU Sysrème NERVEUX DES ProsoBRANCHES,

Si Pon se demande maintenant quels sont les Mollusques dont le
systéme nerveux a pu, en se modifiant, servir de point de départ à celui
des Diotocardes, on est conduit à considérer les Amphineures, et princi-
palement les Chitonidés, comme étant les formes qui se rapprochent le
plus de ces Mollusques,

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 231

Système nerveux des Ohitonidés et des Diotocardes. — Comme les Dio-
tocardes, en effet, les Chitonidés (voir p. 233, Fig. A) sont pourvus de
cordons ganglionnaires palléaux et de cordons pédieux, qui présentent
entre eux et avec les ganglions voisins les mémes relations essentielles.
Ainsi que nous avons pu nous en convaincre par la dissection du Chiton
Jfascicularis Poli et de 1 Acanthopleura Savatieri Rochebr., les cordons
ganglionnaires pédieux des Chitonidés s'étendent parallèlement sur toute
la longueur du pied, ils émettent les nerfs pédieux proprement dits et
sont mis en relation par des commissures transversales plus ou moins
irrégulières dont certaines se bifurquent parfois et envoient des branches
dans la sole pédiense. Ces faits ont été mis en évidence par M. Béla
Haller (82), mais nos observations diffèrent de celles qu’a publiées cet
auteur en deux points qui méritent d’être signalés : le premier, c'est
que la commissure pédieuse antérieure est, comme l’a montré M. von
Ihering (77, 45, Fig. 4), plus volumineuse que les autres et régulière-
ment transversale ; le second, c'est que chaque cordon pédieux se rat-
tache aux centres cérébroides du même côté par un prolongement qui
s’atténue régulièrement d’arrière en avant. Ce prolongement anasto-
motique est muni de cellules nerveuses comme les cordons, mais, con-
trairement aux observations de M. Béla Haller (82, Fig. 1) sur le Chiton
siculus Gray, il est plus gróle que les cordons, surtout dans sa partie
antérieure qui se rattache aux centres cérébroïdes ; c'est un connectif
cérébro-pédieux encore riche en cellules ganglionnaires, et, si Pon rap-
proche de ce fait existence d’une commissure pédieuse grosse et régu-
lièrement transversale, on trouve que les cordons pédieux des Chitonidés
sont semblables, & tous égards, aux rubans pédieux des Diotocardes,
surtout à ceux des Pleurotomaires et des Fissurelles, cordons qui se
prolongent en avant de la grosse commissure, comme on sait et se con-
tinuent progressivement avec les connectifs cérébro-pédieux.

Mêmes homologies entre les cordons palléaux des Chitonidés et les
rubans palldaux des Pleurotomaires. Ils donnent naissance, comme
eux, aux nerfs du manteau et des muscles columellaires, et, comme eux
aussi, envoient des branches aux viscères; bien plus, chaque cordon
palléal se met en relation avec le cordon pédieux du méme cóté par des
anastomoses transverses trés nombreuses. La présence de ces anasto-
moses est d’une grande importance pour éclaircir le problème qui nous
occupe; elle montre que les cordons pédienx et palléaux des Chitonidés
ne sont pas plus indépendants que ceux des Diotocardes, et qu'il suffira
d’un raccourcissement progressif des anastomoses palléo-pédieuses pour
que les cordons pédicux et palléaux de chaque côté arrivent à se con-

232 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

fondre et forment des cordons palléo-pédieux semblables & ceux des Dio-
tocardes. Les anastomoses palléo-pédieuses sont trés nombreuses chez
certains Amphineures vermiformes ot elles ont été mises en évidence
par M. Hubrecht (’81, ’82) dans la Proneomenia Sluiteri Hubr. ; M. von
Thering en avait trouvé une de chaque cóté dans le Chiton salamander
(77, Fig. 47) et M. Béla Haller dans le Chiton fascicularis (82, 11,
14); mais, depuis, M. Plate (95°, ’96) et M. Thiele (95) ont mis
en évidence, chez divers Chitonidés, un ensemble d’anastomoses aussi
riche que celui des Proneomenta, et nous avons observé le méme fait
dans le Chiton fascicularis.

Plus frappantes encore, si c’est possible, sont les ressemblances et les
homologies qui existent dans la partie antérieure du système nerveux chez
les Chitonidés et les Diotocardes. Les centres cérébroides des Chitons,
il est vrai, sont moins condensés et émettent des nerfs sur toute leur
partie commissurale, mais on sait que la concentration des mêmes gan-
glions est loin d’être complète chez les Diotocardes et que Y Haliotide,
à ce point de vue, ressemble singulièrement aux Chitons. Au resto, les
homologies qui nous occupent ne portent pas seulement sur la structure
générale du système, mais s'étendent jusqu'à ses détails: les saillies
labiales des Chitonidés ont la móme forme et la méme origine que celles
des Diotocardes, et, comme chez ces derniers, donnent naissance à la
commissure labiäle (Fig. A, c*) et à un certain nombre de nerfs labiaux ;
le systéme stomato-gastrique a la méme forme caractéristique dans les
deux groupes et se trouve au méme ótat de concentration imparfaito, il
occupe la méme position sur les parois de la masso buccale et, fait plus
remarquable encore, prend ses origines (k) au m@me point sur la saillie
labiale des centres cérébroïdes. D’après M. Béla Haller (94, Fig. 1),
on retrouverait méme, chez les Patellidés, les ganglions subradulaires
qu’on observe chez tous les Chitons.

Ayant montré les homologies étroites qui existent entre le systöme
nerveux des Chitonidés et celui des Diotocardes, il nous resto à inter-
préter les dissemblances qu’on observe dans le méme système, entre les
deux groupes. La plus importante, celle qui, à vrai dire, domine et
comprend toutes les autres, c’est la forme et l’étendue particulière du
système ganglionnaire palléal chez les Chitonidés. Grave à la position
du manteau qui, sous la forme d’une frange épaisse, entoure le corps des
Chitonidés et passe juste au-dessus de la bouche et de Panus, lo système
ganglionnaire palléal des Chitonidés forme un anneau allongé parfaite-
ment clos qui passe lui-même, comme le manteau, au-dessus de la bouche
et de Panus. Produit par une concentration nerveuse très pen accen-

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 233

tuée, cet anneau palldal a conservé des relations étroites avec les centres
nerveux voisins : en arrière de la région céphalique, nous avons vu qu’il

se rattachait aux cordons pédieux
par de nombreuses anastomoses ;
dans la région céphalique, à cause
de la superposition du manteau et
des lèvres, l’anneau s’est fusionné

“avec les centres cérébroïdes et ne

s’en distingue plus que par un sillon
(Fig. A, s.). M. von Ihering (77,
44, Fig. 4) s’est parfaitement rendu
compte de cette coalescence et,
avant tout autre, s’est efforcé de
mettre en évidence la nature com-
plexe des centres ganglionnaires an-
térieurs des Chitons. Nous ne voy-
ons pas pourquoi M. Béla Haller

(182, 4) a essayé do combattre une

opinion aussi juste ; il suffit de jeter
un coup d’eil sur le système ner-
veux d'un Chiton pour acquérir la
conviction que M. von Ihering a dit
vrai. Bien plus, les figures de M.
3óla Haller (82, Fig. 32) établis-
sent elles-mêmes manifestement que
le système nerveux céphalique des
Chitonidés se compose de deux cen-
tres différents dont l’un innerve ex-
clusivement le manteau, tandis que
l’autre envoie une rangée de nerfs
aux lobes céphaliques et une se-
conde à la région labiale. Nous
faisons allusion à la coupe repré-
sentée dans la figure 32 du mémoiro
de M. Béla Haller sur les Chitons
de ’Adriatique ; quant à la figure 1
du même travail, elle donne l’idée
la plus fausse des relations qui ex-
istent entre les centres antérieurs

Fie. A.— Système nerveux du Chiton
Jascicularis. — P nerfs palléaux; s sil-
lon séparant les centres cérébroides de
la partie antériéure des centres pal-
léaux ; & origine du stomato-gastrique ;
R ganglions subradulaires. (Dans
cette figure et dans les cing suivantes
les centres cérébroïdes sont marqués de
traits verticaux ||], les centres pédiaux
de traits horizontaux = et les centres
palléaux de ponctuations..... )

ct la partie libre des cordons palléaux, car les nerfs des lobes céphaliques

234 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

(nerfs 2 de Pauteur) y paraissent naitre du ruban qui continue les cor-
dons palléaux, tandis que les nerfs palléaux (nerfs 1 de l’auteur) et les
nerfs labiaux prendraient leur origine dans la partie des centres anté-
rieurs qui continue les cordons pédieux. C’est tout le contraire de ce
que Pon observe en réalité, quand on étudie les Chitonidés.* Au reste,
des traces manifestes de la disposition particulière au système nerveux
de ces derniers se rencontrent chez les Gastéropodes ; dans ces derniers,
en effet, des fibres nerveuses de la commissure cérébroide traversent les
centres cérébroïdes eux-mêmes, et, par les connectifs cérébro-palléaux, se
rendent aux ganglions palléaux (voir Walter, ’63, Pl. IV, Fig. 1; de
Nabias, ’94, 14, PL IV, Fig. 79).

Une autre différence entre le systeme nerveux palléal des Chitonidés
et celui des Diotocardes, c’est l'absence, chez les premiers, de toute com-
missure viscérale. Comme les Diotocardes, les Chitonidés émettent des
nerfs branchiaux et viscéraux issus du systéme palléal, mais ces nerfs ne
se rattachent pas à une anse viscérale fermée et vont séparément se
rendre aux organes qu'ils desservent. C’est, nul ne le conteste, un état
primitif auquel a dû faire suite un état où certains nerfs viscéraux s'ana-
stomosaient d'un côté à l’autre par-dessous l'intestin; toutefois, on n’a
rien observé jusqu'ici, chez les Chitonidés, qui montrât le début de
ces anastomoses. M. von Ihering avait cru trouver cette commissure
primitive, dans le Chiton cinereus (77, Fig, 4); mais il avait pris pour
elle la commissure subradulaire, dont les origines ne sont pas, comme

it, sur les cordons palléaux. Depuis, M. Haller a signalé dans

il le croy
le Chiton fascicularis (82) et dans le Ch. magnificus (94%) une paire
de nerfs palléaux antérieurs qui viendraient se fusionner dans des gan-

Ipi "FA ara Aa. Vautri
la partie anrerreure de reston

glions s

gue portant ura

ouvrage

1

18 qu'elle «

: ‘ 1 A
nals bien aes Centres cere

2 trouvent sur le bord

de la saillie de nombreuses anastomoses transversalc

"$ cordons pédieux aux cordons palléaux.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 400

une anse sous-intestinale anastomotique, qui reliait entre eux les cordons
palléaux, Les Chitonidés sont assez variables dans leur structure (Voy.
Plate, 96, 176) pour qu’on puisse espérer y découvrir quelque jour les
débuts de cette anse commissuralo,

Modifications progressives du système nerveux des Amphineures et
des Gastéropodes. — Puisqu'il est impossible de contester les homologies
profondes qui existent entre le système nerveux des Chitonidés et celui
des Gastéropodes primitifs, il nous faut indiquer maintenant le pro-
cessus par lequel a dû passer le premier pour former le second, Quelle
que soit l’idée que l’on se fasse du système branchial des Chitonidés, ou
des formes voisines qui ont servi de point de départ aux Diotocardes, on
doit admettre, avec M. Bütschli (87) et les autres zoologistes que ce
système était réduit à deux branchies chez les Diotocardes primitifs,
que ces deux branchies étaient situées en arrière à droite et à gauche de
Panus, qu’elles étaient innervées par Panse viscórale primitive dont nous
avons parló plus haut, et que Pensemble de co complexe anal s'est dé-
placé progressivement à droite et a fini par se placer en avant du cóté
dorsal, dans une cavité palléale située en arrière de la région céphalique.
On sait que le résultat de ce déplacement a ¿té la torsion en 8 de chiflre
de la commissure viseérale chez les Prosobranches, et (ainsi que Pun de
nous Va montré (Bouvier, 92, *93) et comme M. Grobben (94) et
M. Pelsencer (94) Pont reconnu dans la suite) que les Gastéropodes
à système nerveux orthoneure dérivent des Prosobranches diotocardes
par un retour, vers sa position primitive, du complexe anal, et par la
distorsion de la commissure viscérale qui a été la conséquence de ce
déplacement rétrograde.

Ceci dit, nous allons examiner les divers stades qu'a dû suivre le sys-
tome nerveux chitonien des Gastéropodes primitifs, des Prærhipidoglosses
comme les appelle M. Plate (95), pour arriver à se transformer en un
système nerveux typique de Gastéropode.

Premier stade, (Wie. B).— Le manteau recule en arrière de la tête, qui
fait de plus en plus saillie en avant, ce qui a pour conséquence de séparer
complètement des centres cérébroïdes, les cellules ganglionnaires palléales
qui formaient un ruban sur leur bord dans les Chitons ; il se forme ainsi,
de chaque côté, un connectif cérébro-palléal à peu près dépourvu de
cellules ct les cordons palléaux, grossis dans leur partie antériouro, se pro-
longent plus ou moins sur ces connectifs. A Pautre extrémité du corps,
la chambre qui renferme les denx branchies devient de plus en plus pro-
fonde, les cordons palléaux s'arrêtent à l'extrémité antérieure de cette

chambre dont les bords sont innervés par deux nerfs anastomoses qui

250 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY.

sont les prolongements, à peu près dépourvus de cellules nerveuses, des
cordons palléaux. Telle est l’origine des nerfs palléaux primitifs ; des
branches nerveuses moins puissantes partent aussi de la commissure
viscérale pour se rendre dans le manteau. En même temps, la coquille
gagne en importance, les muscles columel-
laires se développent et se superposent au
pied, dont ils constituent la partie supé-
3 e rieure. Les muscles columellaires et pédi-

P Ip eux forment, de plus en plus, un ensemble
4 morphologique distinct, et, par un raccour-

eissement progressif des anastomoses pal-
léo-pédieuses, les centres ganglionnaires de
ces muscles se rapprochent de plus en
plus. — La forme orthoneure, qui réalise

r

ce type, ne nous est pas connue; inter-
médiaire entre les Chitonidés et les Dio-

f
/
EEE
A
D
\

E tocardes, elle devait avoir un système

E
UAE
i

nerveux peu concentré et différait par
conséquent beaucoup du Prærhipidoglosse,
à ganglions condensés, tel que Pa conçu
M. Plate.

Deuxième stade (Fig. C). — La cavité
palléale est plus profonde et vient peu à

LE

peu se placer dorsalement en arrière de
la tête, ce qui amène le croisement de la
commissure viscérale et le déplacement en
avant de ses racines, qui sont issues,
comme on sait, des cordons palléaux.

Grâce au processus de condensation qui

JA continue, ces derniers ont fini par se fu-

Fic. B. — Système nerveux sionner, suivant leur longueur, avec les

hypothétique intermédiaire cordons pédieux, pour former, de chaque

entre celui des Chitons et

celui des Pleurotomaires. —
P nerfs palléaux ; A anus,

côté, entre les muscles pédieux et une
partie des muscles columellaires, un cor-
don unique où ils apparaissent, l’un sous
forme de ruban palléal, Pautre sous la forme de ruban pédieux séparé
du premier par un profond sillon. Au reste, la condensation ganglion-
naire, dans le sens de la longueur, est encore fort peu avancée, ct les
deux rubans se prolongent en une corne, en avant de la grosse commis-
sure des cordons, jusqu'aux connectifs cérébro-pédieux et cérébro-palléaux.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 237

Lo Pleurotomaria Quoyana se trouve précisément au stade que nous
venons de signaler.

Troisième stade (Fig. D). — Ce stade ne diffère du précédent que par
la condensation plus grande de tous les centres nerveux, et, dans bien
des cas, par l’atrophie de la branchie droite des Pleurotomaires. Du
ruban supérieur de chaque corne des cordons, et probablement des par-
ties voisines de ces derniers, se sépare une corne dorsale exclusivement

Ÿ 7
PEN KP
e AW It
Fra. C.— Système nerveux du Fic. D.— Système nerveux
Pleurotomaria Quoyana. dialyneure des Trochidés.

palléale ; des deux cornes qui se trouvent, dès lors, en avant de chaque
cordon, celle qui est située du cóté dorsal tend de plus en plus & grandir
et à sisoler, tandis que la corne ventrale se réduit progressivement, tant
par le passage de sa partie palléale dans la corne dorsale, que par la
condensation en arrière du ruban pédieux qui la compose. En même
temps s’atténue le sillon qui sépare les deux rubans de chaque cordon.
Les Fissurellidés, Haliotidés et Trochidés se trouvent à ce stade : leurs

238 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY,

nerfs palléaux d’origine commissurale sont assez peu importants, mais
s'anastomosent déjà avec les nerfs pallgaux primaires,

Quatrième stade (Fig. E). — Les rubans palléaux et pédieux de chaque
cordon se fusionnent complètement et l’on ne voit plus trace des sillons
qui les séparent ; les cornes ganglionnaires dorsales forment des ganglions
palléaux plus ou moins isolés et les cornes ventrales ont disparu. Les
Patellidés, les Néritidés, les Hélicinidés, les Paludinidés, les Cyclo-
phores et les Cyprées sont à ce stade.

Fic. E. — Système nerveux Fia F.— Système nerveux zygoneure
presque zygoneure du Cyclo- du Triton.

phorus.

Cinquième stade (Fig. F). — Les cordons palléo-pédieux se conden-
sent suivant leur longueur et forment des ganglions palléo-pédieux ovoïdes,
plus connus sous le nom de ganglions pédieux ; les cornes palléales
deviennent des ganglions palléaux très distincts qui se rapprochent de
plus en plus des centres cérébroïdes, du moins dans la majorité des cas.
En même temps, on voit, chez les Prosobranches, les nerfs palléaux
| secondaires contracter des anastomoses de plus en plus étroites avec les
| nerfs palléaux primaires correspondants, ce qui conduit à la zygoneurie.

Chez les Gastéropodes orthoneures, le complexe anal revient du côté
droit, et la distorsion du système nerveux se produit à divers degrés.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 239

Telle est, croyons-nous, l’idée la plus nette et la plus exacte que Pon
puisse se faire de l’évolution du système nerveux chez les Gastéropodes.
Muni de cordons palléo-pédieux profondément sillonnés et dépourvus de
cornes palléales distinctes le Pleurotomaria Quoyana se trouve à un stade,
inconnu jusqwici, qui permet de concevoir tous les autres et qui rap-
proche singulièrement les Gastéropodes archaiques des Amphineures,
dont ils paraissent tirer leur origine.

Au reste, ce n’est point seulement par son systéme nerveux que le
Pleurotomaria Quoyana se place, avant tout autre, en téte des Mol-
lusques Gastéropodes. Avec ses branchies presque rigoureusement
symétriques et égales, ses orifices rénaux (et génitaux ?) également pairs
et symétriques, sa radule A régions mal délimitées, ses yeux en cupule
ouverte A Pextóriour, et ses nombreux otolithes toujours petits et de
dimensions inégales, il fait songer aux formes les plus primitives du
groupe et permet de concevoir, mieux que par le passé, les formes plus
primitives encore dont les Diotocardes sont issus.

Il est à présumer que des recherches anatomiques entreprises sur
d'autres espèces du genre viendront combler les lacunes nombreusés que
laisse forcément notre travail, et étendre plus sûrement le champ de
généralisation que nous avons ouvert. A ce propos, qu’il nous soit per-
mis de signaler aux zoologistes l'intérêt tout particulier qui s’attache aux
origines de la commissure viscérale chezles Pleurotomaires ; situées sur
les connectifs cérébro-palléaux dans le Pleurotomaria Quoyana, elles
sont certainement fort curieuses et ne ressemblent guère à celles des
autres Gastéropodes. Mais avant de chercher la raison de ces origines
remarquables et d'en tirer parti pour expliquer l’origine des Gastéro-
podes, il faut être bien certain qu’elles ne sont pas spécifiques et se-
condaires, mais primitives et caractéristiques du genre. C’est un point
important, sur lequel nous fixeront bientôt, il y a lieu de Pespérer,
d’autres anatomistes.

VOL. XXXII. — NO, 10 4

240 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

INDEX BIBLIOGRAPHIQUE.

1° PARTIE SPÉCIALE (HISTORIQUE).
Férussac.

’21. Tableaux systématiques des animaux Mollusques, Pl. XXXIV. 1821 (?).
Cet ouvrage fait partie de l Histoire générale et particulière des Mollusques
terrestres et fluviatiles, publié à partir de 1819.

Sowerby, J.
21, Mineral Conchology, Vol. III. p. 139, T. CCLXXVIII, décembre 1821.
Borson.

"21. Saggio di orittografia piemontese (Memorie della Reale Accademia delle

Scienze di Torino, T. XXVI, 1821, Pl. IL. Fig. 1).
Defrance.

728. Dictionnaire des sciences naturelles, planches, 2 partie; Règne organisé.
Zoologie, conchyliologie et malacologie, par M. Ducrotay de Blainville, 26°
cahier, Pl. LXXXVI, Fig. 2 et 3.

Defrance.
24. Tableau des corps organisés fossiles, p. 114.
Defrance.

726. Dictionnaire des sciences naturelles, par plusieurs professeurs du Jardin
du roi et des principales écoles de Paris, T. XLI, 1826, p. 381 (Pleuroto-
maria).

Sowerby, J.
"31. The Genera of Recent and Fossil Shells, Part XXXII (fin de l’année 1830
ou commencement de 1831). Pleurotomaria, avec une planche coloriée.
Sowerby, J.
44. Mineral Conchology, Vol. VII, p. 69, T. DCXL, novembre 1844.
Fischer, P., et Bernardi.
"56. Description d'un Pleurotomaire vivant (Journ, de Conchyl., T. V, 1856,
p. 160-166, Pl. V, Fig. 1-3).
Adams, H., et A.
| "58. The genera of Recent Mollusca, London, Vol. II, p. 630.
| Ryckholt, De.
| 60. Revue des genres qui composent la famille des Haliotidæ (Journ. de
conchyl., T. VIII, 1860, p. 183).

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 241

Crosse; H, et Fischer, P.
"61. Observations sur le genre Pleurotomaire et Description d’une deuxième

espèce vivante appartenant au même genre (Journ. de conchyl,, T. IX,
1861, p. 155-167; Pl. V, Fig. 1 et 2).
Schramm.
"69. Catalogue des Coquilles et des Crustacés de la Guadeloupe, 2° édition,
Basse-Terre, 1869, p. 9.
Woodward, S. P.
70. Manuel de conchyliologie, Paris, 1870 (traduction Alois Humbert),
p- 281, 560.
Agassiz, L.
’72. New-York Weekly Tribune, Feb. 14, 1872.
Crosse, H., et Fischer, P. j
72. Journ. de conchyl., T. XX, p. 287 (nouvelles).
Martens, E. von.
272. Lebende Pleurotomarien (Nachbl. der deutsch. Malak. Ges., T. IV,
1872, p. 55).
Reeve.
74. Conch. Icon. Pleurotomaria, Fig. 1-2.
Crosse, H.
76. Journ. de conchyl., T. XXIV, p. 216 (nouvelles).
Hilgendorf.
77. Sitzungsber. der Ges. Naturforsch. Freunde zu Berlin, 1877. Sitzung
von 20 márz, p. 72-73.
Schepman.
79. Wine neue recente Pleurotomaria (Tijdschr. der Ned. Dierk. Vereen.,
Deel TV, 1879, p. 162-167).
Crosse, H.
’89. Journ. de conchyl., T. XXVIII, 1880, p. 203-204 (nouvelles).

Martens, E. von.
'80. Conchyl. Mith., Vol. I, p. 33, Pl. VIII, Cassel, 1880.
Dall.
'81. Bull. Mus. of Comp. Zoól. at Harvard College, Vol. IX, 1884.
Crosse, H.
"82. Les Pleurotomaires de l’époque actuelle (Journ. de conchyl., T. XXX,
1882, p. 5-22, Pl. 1, Fig. 1-2).
Crosse, H.
822, Note additionnelle sur le Pleurotomaria Rumphii, Schepman (Journ. de
conchyl., Vol. XXX, 1882, p. 183-184).
Schepman.
’82. Conchyliologische Bijdragen (Tijdsch. d. Ned. Dierk. Vereen., Deel VI,
1882, p. 20-23, Pl. II, Fig. 1-3).

242

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Tryon, G. W.
’83. Structur. and System. Conchology, Vol. II, p. 318, Philadelphia, 1885.
Fischer, P.
85. Manuel de conchyliologie, p. 849-851, 1885.
Woodward.
785. On recent and fossil Pleurotomariide (Geological Magazine, October,
1885, p. 433-439, Pl. XI. Fig. 1).
786. Journ. de conchyl., T. XXXIV. p- 106, 1886 (bibliographie).
Sowerby.
°87. Thesaurus Conchyliorum, T. V, 1887, p. 185, T. CDXC, Fig. 1-2.
Agassiz, Alexander.
"88. Three Cruises of the Blake, Vol. II. (Bull. of the Mus. of Comp. Zool.
at Harvard College in Cambridge, Vol. XV. p. 69; Fig. 288 et 289).
Shaler and Foerste.
°88. Preliminary Description of North Attleborough Fossils (Bull. of the
Mus. of Comp. Zoöl., Vol. XVI. p. 30-31, Pl. IL. Fig. 11).
Dall.
89. Explorations of United States Coast Survey Steamer “ Blake.” Report
on the Mollusca, Part IL. Gastropoda and Scaphopoda (Bull. of the Mus.
of Comp. Zoöl. at Harvard College, Vol. XVIII. 1889, p. 396-403,
Pl. XXIX, XXX; XXXL, XXXIJ. (partim).)
Lechmere Guppy.
90. On a Specimen of Pleurotomaria from Tobago, Trinité, 1890, broch.
in-8° de 2 pages.
Pilsbry, A.
"90. Manual of Conchology, Structural and Systematic, Vol. XII. 1890,
p. 99-72, Pl. LVL-LVIL |
Crosse, H.
791. Journ. de conchyl., T. XX XIX. 1891, p. 68, p. 360 (nouvelles).
Damon.
’91.— The Conchologist, Vol. I. 1891, une planche hors texte.
Lechmere Guppy.
"91. Ona specimen of Pleurotomaria from Tobago (Proc. Zoöl. Soc. London,
1891, p. 484-485).
Pilsbry,
‘92. The Nautilus, Vol. V. 1892, p. 119 (General Notes).
Collinge, W. E.
94. The Journal of Malacology, Vol. IIL. 1894 (notes), p. 20.
Cooke, A. H.
"95. The Cambridge Natural History, Molluscs, London, 1895.
Pilsbry. M
795. Nautilus, Vol. VIIL 1895, p. 131 (notes).

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 243

Vincent, E.
796. Note préliminaire sur le Pleurotomaria (Soc. roy. Malac. de Belgique,
procés verbal de la séance du 12 décembre 1896, Pl. LXXVII, Fig. 1).
Mitsukuri.
"96. A living Specimen of Pleurotomaria Beyrichii (Annotationes zoologicæ
japonenses, Vol. I. Tokyo, 1897, p. 67).
Sacco.
97. I Molluschi dei terreni terziarii del Piemonte e della Liguria, Parte
XXII. 1897, p 2.
Bouvier, E.-L., et Fischer, H.
97. Sur l’organisation et les affinités des Pleurotomaires (Comptes rendus de
l'Académie des sciences, mars 1897).
Bouvier, E.-L., et Fischer, H.
98. Sur l’organisation des Pleurotomaires (Comptes rendus de l’Académie
des sciences, mai 1898).

2° Partin GÉNÉRALE (ANATOMIQUE).

Lacaze-Duthiers, H. de.
759. Mémoire sur le système nerveux de PHaliotide (Annales des sciences
naturelles, Zoologie, 4° sér., T. XII, 1859).
Walter, G.
63. Mikrosocopische Studien über das Central-nervensystem wirbelloser
Thiere, Bonn, 1863.
Lacaze-Duthiers, H. de.
"72. Otocystes ou capsules auditives des Mollusques (Gastéropodes) [Archives
de zoologie expérimentale, 1re sér., T. I, 1872].
Thering, H. von.
"77. Vergleichende Anatomie des Nervensystems und Phylogeme aer Mol-
lusken, Leipzig, 1877.
Spengel.
"81. Die Geruchsorgane und das Nervensystem der Mollusken (Zeitschr. wiss.
Zool., Bd. XXXV, 1881).
Hubrecht, A,
"81. Proneomenia Sluiteri (Niederl. Arch. Zool, Suppl. I, 1881).
Béla Haller.
"82. Die Organisation der Chitonen der Adria (Arbeit. Zool. Institut Wien,
E DV, L882).
Hubrecht.
83. Contribution à la morphologie des Amphineures (Bull. scient. du Nord
ot do la Belgique, 5° année, 1882).
Béla Haller.
"84. Untersuchungen über marine Rhipidoglossen. Erste Studie (Morpholog.

Jahrbuch, Bd. IX, 1884).

244 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

Wegmann.

784. Contribution à l’histoire naturelle des Haliotides (Archives de zoologie
expérimentale, 2° sér., T. I, 1884).

Boutan, L.

786. Recherches sur Panatomie et le développement de la Fissurelle (Archives

de zoologie expérimentale, 2° sér., T. III bis, 1886).
Bouvier, E.-L.

786. Contribution à Vétude des Prosobranches pténoglosses (Bull. Soc.

malacol. de France, 1886).
Béla Haller.

86. Untersuchungen über marine Rhipidoglossen. II. Textur des Central-

nervensystems und seiner Hüllen (Morphol. Jahrb., Bd. XI, 1886).
Bouvier, E.-L.

87. Système nerveux, morphologie générale et classification des Gastéropodes
prosobranches (Annales des sciences naturelles, Zoologie, 7° sér., T. ITI,
1887).

Biitschli, O.

’87. Bemerkungen über die wahrscheinliche Herleitung der Asymmetrie der
Gastropoden, spec. der Asymmetric im Nervensystem der Prosobranchiaten
(Morph. Jahrbuch, Bd. XII, 1887).

Garnault, P.

87. Recherches anatomiques et histologiques sur le Cyclostoma elegans, Bor-

deaux, 1887.
Pelseneer, P.

’87, Sur la valeur morphologique de l’épipodium des Gastropodes rhipido-
glosses (Comptes rendus de l’Académie des sciences, T. CV, n° 14, 1887,
et Bull. scient. de la France et de la Belgique, 1887).

Boutan, L.

#88. Contribution à l’étude de la masse nerveuse ventrale (cordons palléaux-
viscéraux) et de la collerette de la Fissurelle (Archives de zoologie expéri-
mentale, 2° sér., T. VI, 1838).

Pelseneer, P.

188. Sur Pépipodium des Mollusques (Bulletin scientifique de la France et de

la Belgique, 3° sér., 17° année, 1888).
Béla Haller.

89. Réplique à M. Boutan (Sur Pépipodium des Mollusques) [Bulletin

scientifique de la France eb de la Belgique, 3° sér., T. I, 1889].
Lacaze-Duthiers, H. de.
90. De la valeur relative de quelques procédés d’investigation en anatomie
comparée (Archives de zoologie expérimentale, 2° sér., T. VIII, 1890).
Pelseneer, P.
190. Sur Pépipodium des Mollusques (deuxiéme note) [Bulletin scientifique
de la France et de la Belgique, T. XXII, 1890].

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 245

Thiele, J.
"90. Ueber Sinnesorgane der Seitenlinie und das Nervensystem der Mollusken
(Zeits. wiss. Zool., Bd. XLIX, 1890).

Boutan, L.
’90. Le système nerveux du Paramophorus (Scutus) dans ses rapports avec
le manteau, la collerette (manteau inférieur) et le pied (Revue biologique
du Nord, T. II, 1890).
Pelseneer, P.
"91. Sur l’épipodium des Mollusques (troisième note) [Bulletin scientifique
de la France et de la Belgique, T. XXIII, 1891].

Bouvier, E.-L.

"92. Observations sur les Gastéropodes opisthobranches de la famille des
Actéonidés (Comptes rendus de la Société philomathique de Paris, 24
décembre 1892).

Thiele, J.

"92. Beiträge zur Kenntniss der Mollusken. I. Ueber das Epipodium (Zeits.

wiss. Zool. Bd. LIII, 1892).
Bouvier, E.-L.

’93. Sur la distorsion des Gastéropodes hermaphrodites (Comptes rendus de

la Société philomathique de Paris, 14 janvier 1893).

Grobben, K.
"94. Zur Kenntniss der Morphologie, der Verwandtschaftsverháltnisse und des
Systems der Mollusken (Sitz. Ber. Akad. Wien, Bd. CIII, p. 61-86).

Béla Haller.

94. Studien über Docoglosse und Rhipidoglosse Prosobranchier, nebst
Bemerkungen über die phyletischen Beziehungen der Mollusken un-
tereinander, Leipzig, 1894.

Béla Haller.

"94%, Beiträge zur Kenntniss der Placophoren (Morphol. Jahrb., Bd. XXI,

1894).

Nabias, B. de.
"94. Recherches histologiques et organologiques sur les centres nerveux des
Gastéropodes, Bordeaux. 1894.
Pelseneer, P.
"94. Recherches sur divers Opisthobranches (Mém. Cour. des savants
étrangers. Académie de Bruxelles, T. LIII 1894).
Pelseneer, P,
?94* Introduction à l’&tude des Mollusques, Bruxelles, 1894.

Plate, L.
"95. Bemerkungen über die Phylogenie und die Entstehung der Asymmetrie
der Mollusken (Zoolog. Jahrb., Bd. IX, 1895).

246 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Plate, L.
95%. Ueber den Bau des Chiton aculeatus (Sitz. Ber. Ges. Naturf. Freunde,

Berlin, 1895, p. 154-164).
Thiele, Jie
"95. Ueber die Verwandtschaftsbezichungen der Amphineuren (Biol. Central-
blatt, Bd. XV, 1895).
Plate, L.
196. Ueber einige Organisationsverhältnisse der Chitonen (Sitz. Ber. Ges.
Naturf. Freunde, Berlin, 1896, p. 42-50).
Plate L.
96", Ueber die Organisation einiger Chitonen (Verhandl. deutsch. Zool.
Gesellsch., 6° Jahresversammlung, 1896).
Bouvier, E.-L., et Fischer, H.
’96. Sur l’organisation et les affinités des Pleurotomaires (Comptes rendus de
P Académie des sciences, mars 1897).
Thiele, Ju
97. Beiträge zur Kenntniss der Mollusken. Ueber Hautdrüsen und Derivate
(Zeits. wiss. Zool., 1897).
Boutan, L.
98. L'organe glandulaire périphérique de l’Helcion pellucidum Lin. (Archives
de zoologie expérimentale, 3* sér., Vol. V, 1898).
Bouvier, E.-L., et Fischer, H.
98. Sur Porganisation des Pleurotomaires (Comptes rendus de l'Académie
des sciences, mai 1898).

rcs see

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS. 247

EXPLICATION DES PLANCHES.

Lettres Communes.

Organes divers. — Ao, aorte antérieure; F, pied; M, masse buccale; M’, mufle;
o, otocyste; œ, œil; t, tentacule; U, manteau; V, muscle columellaire.

Ganglions. — B, ganglions buccaux (stomato-gastrique), C, ganglions cérébroïdes ;
L, saillie labiale; Cd, ganglion palléal droit ou corne palléale supérieure droite ;
Cg, ganglion palléal gauche ou corne palléale supérieure gauche; C4, corne pal-
léale unique des Pleurotomaires; Ps, partie palléale des centres nerveux du pied;
Pi, partie pédieuse des centres nerveux du pied; s, sillon palléo-pédieux,

Commissures, connectifs, nerfs. — c, commissure cérébroïde ; cl, commissure labiale ;
h, branche sus-intestinale de la commissure viscérale; 41, branche sous-intesti-
nale de la commissure viscérale; %, connectif córébro-buccal ; 41, connectif céré-
bro-pédieux ; /:?, connectif cérébro-palléal ; à, commissure palléo-pédieuse ; J, nerf
optique; o”, nerf acoustique; ml, m2, m3,.... nerfs labiaux; s1, s2, s3.... nerfs
buccaux; (, nerf tentaculaire; p, nerfs pédieux; p’, nerfs pédieux antérieurs ;
pu, nerfs mixtes (palléo-pedieux).

PLANCHE I.

=

Exemplaire de Pleurotomaria Quoyana, tel qu'il nous a été communiqué,
vu du côté gauche. 5, sole pédieuse.
Fig. 2. Région céphalique du même, ouverte sur la ligne médiane dorsale,

Fig.

Fig. 3. Cavité antérieure du corps du même exemplaire ouverte sur la ligne mé-
diane dorsale ; la masse buccale est enlevée. S, sole pédieuse ; d, nerf
issu de la branche sus-intestinale de la commissure viscérale.

Fig. 4. Partie antérieure du cordon pallée-pédieux gauche, vue par sa face ex-

terne. a, al, a?, nerfs issus de la partie palléale.

Fig. 5. Partie antérieure du cordon palléo-pédieux droit du Turbo radiatus Gme-
lin, face externe; a, grand nerf palléal droit; al, a2, a3, autres nerfs
issus de la partie palléale du cordon.

Fig. 6 et 7. Branches gauche et droite du système nerveux stomato-gastrique dans
le Chiton fascicularis ; les cellules nerveuses se concentrent de chaque
côté en deux ganglions.

248 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY.

PLANCHE II. |

Fig. 8. Méme préparation que celle de la Figure 3, mais le lobe antérieur du pied
ayant été fendu et la cavité antérieure largement ouverte, on peut voir |
les parties nerveuses qui avoisinent le pied. v, tronc vasculaire qui
paraît plonger dans le pied en s'y bifurquant ; e, b, 0’, nerfs pariétaux qui
se détachent des connectifs cérébro-palléaux ; c et e”, nerfs pariétaux ou
columellaires issus des connectifs cérébro-palléaux ; l, V, nerfs columel-
laires postérieurs ; d, nerf des parois dorsales.

Fig. 9. Même préparation que la précédente, mais la sole pédieuse a été étalée et
en partie disséquée du côté ventral pour mettre à découvert les cordons
palléo-pédieux. Lettres accessoires comme pour la Figure 8.

Fig. 10. Ganglion cérébroïde droit du Pleurotomaria Quoyana, vu par la face
interne.

Fig. 11. Figure indiquant la disposition anatomique et les connexions des parties
antéricures des cordons palléo-pédieux. La préparation est vue sous le
méme aspect (face inférieure) que celle de la Figure 9, seulement les
cordons ont été disposés de telle sorte que celui de gauche est vu par sa
face externe, tandis que celui de droite est vu par sa face externe en
avant de la grosse commissure palléo-pédieuse i, par sa face interne en
arrière a, al, a/2, nerfs émis par le ruban palléal du cordon gauche;
a’, a 2, a’3, nerfs émis par le ruban palléal du cordon droit. Les autres
lettres accessoires comme dans la Figure 8.

Fig. 12. Système stomato-gastrique du Pleurotomaria Quoyana.

Fig. 18, Partie antérieure des cordons palléo-pédieux de la Cemoria noachina, L.,
d’après M. Béla Haller (94, Fig. 142). C”, cornes inférieures ou palléo-
pédieuses ; Cy, Cd, cornes supérieures purement palléales (ganglions
palléaux gauche et droit).

PLANCHE III.

Fig. 1. Une demi-rangée de gauche de la radule, vue par-dessous. X 50.

Fig. 2. Dent impaire et quarante-trois premières dents du côté gauche. X 50.

Fig. 3. Région antérieure d’une rangée; dent impaire, les trois premieres dents

droites; les huit premières dents gauches. X 70.

Fig. 4. Dent impaire, vue de profil. La ligne d'insertion est à droite. X 70.

Fig. 5. Dent impaire, vue par dessous et un peu de trois quarts. X 70.

Fig. 6. Dent impaire montrant, vue de trois quarts, la surface antérieure pourvue

de ses deux expansions foliacées. X 70.

Fig. 7. Premiere dent gauche, du cóté externe. x 70.

Fig. 8. Deuxième dent gauche; du côté externe. 70.

Fig. 9. Troisième dent gauche, du côté interne. X 70.

Fig. 10. Les deux otocystes, avec l’origine des nerfs acoustiques. X 20.

Fig. 11. Groupe d’otolithes. X 500.

Fig. 12. Fragment de la radule, vu par dessous, limité à la longueur d’une rangée,
montrant le mode d'insertion des dents. Sur les bords extrêmes, les
téguments ont été enlevés sans laisser apercevoir, dans leur partie libre,
les dernières dents flabelliformes, X 20.

Une mâchoire. X 8.

u

oi dl Aiea
f Tran

Pu

JE

PR”

Me;

“es al
“Seca eres i

Ss
N

ee

pe
—

al

Mier det,

Fig.
Fig.
Fig.
Fig.
Fig.

se

Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.

BOUVIER ET FISCHER: PLEUROTOMAIRES ACTUELS,

PLANCHE IV.

. Groupe des trois dents 28, 29 et 30, du côté gauche. X 70.

. Vingt-huitième dent gauche. X 70.

. Vingt-neuvième dent gauche. X 70.

. Trentième dent gauche. X 70.

. Groupe des huit dernières dents (110-117) du côté gauche. X 70.

Trente et unième dent gauche. X 70.

. Trente-troisième dent gauche. X 70.
. Trente-septième dent gauche. X 70.
. Quarante-deuxième dent gauche. X 70.

Extrémité supérieure de la cent neuvième dent gauche. X 260.
Quarante-cinquième dent gauche. X 70.

. Quarante-huitieme dent gauche. X 70,
. Extrémité antérieure de la cinquantième dent gauche. X 140.

Extrémité antérieure de la soixante-sixième dent gauche. X 140,
Soixante-sixième dent gauche. X 70.

. Cent troisième dent gauche. X 70.
. Cent neuvième dent gauche. X 70,

249

wi 0