The skeletal system of Necturus maculatus Rafinesque

gOL 3 eennvates
668
C277W54

£503
REPT

Raby» 4
ee Mfate

Y OF NATURAL HISTORY; =

_ VOLUME 5, NUMBER 9. |

tw
1

SYSTEM OF NECTURUS MACULATUS Rafinesque.

¥ ' thy

By HARRIS HAWTHORNE WILDER, Pn. D.

‘

2
Re AEROS ECU 1.3

PUBLISHED BY THE SOCIETY.

MIKO Lay,

r TAGE
a

ah oe hat (a ae
Td. Acta

ETAL SysTEM OF NEectTuRUS MACULATUS Rafinesque.

By Harris Hawrnorne Wiper, Pu. D.

(Read February 19, 1902.)

INTRODUCTION.

eleben says, “Es wiire im Interesse einer besseren Fundamentirung der verglei-

iden Anatomie héchst wiinschenswert, wenn bald zu den wenigen ausfiihrlichen

rs....eines Urodelen....einiger Reptile und niederen Siiuger, besonders Insecti-
n, kimen.... Wer wagt es?” The following paper is an attempt to carry out the
restion of the above quotation as far as it concerns a Urodele; and it is hoped that,

eulosa in Europe. This exploitation of the species as a laboratory animal has led to
he development of means for its capture in quantity for such purposes, and it is thus by
0 means so uncommon an object of study in the laboratories of Europe as it was a few

ars ago."

ranting that the second of the above alternatives has the most in its favor, the
would be limited to one of the Proteidae, a family represented by three generic

Necturus is supplied at present at most of the American universities at $10.00 per hundred, and although just at
there is an attempt being made by dealers to advance the price, especially of those exported to Europe, such con-

1

388 HARRIS HAWTHORNE WILDER ON

groups, which are partially or wholly subterranean in their habits and thus modified in
various degrees to conform to this environment. Proteus is an inhabitant of certain
caves in Austria, and shows, among other troglodytic characters, reduction of eyes, loss of
pigment, and a pronounced attenuation of body and limb, the latter combined with a loss
of toes. These aberrant characteristics are still more marked in the recently discovered
Typhlomolge, which occurs in a subterranean river in the state of Texas. The third and
last member of the group, Necturus, occurs in the Great Lakes and other open waters,
and, corresponding to its freer life, shows the least modification. It remains by day
beneath stones or in subterranean burrows constructed by itself in the mud, and becomes
active at night, swimming through the water in search of prey. Its body and legs are
robust, and do not show the slender proportions seen in Proteus and Typhlomolge. Its
eyes, though small, are functional, and it is well pigmented, especially above. It is thus
at the same time a representative of the lowest group of Urodeles, and one not extremely
modified.

The genus Necturus is exclusively North American and includes but one common
species, V. maculatus Raf., although the series of synonyms, as well as some variability of
marking, have combined to give the general impression that there is a large number.

Aside from NV. maculatus Raf., Cope enumerates a second species, V. punctatus
Gibbes. of rare occurrence and found only in the waters of North and South Carolina.

Necturus maculatus, the subject of this memoir, is a widely distributed form and
occurs in abundance in the Great Lakes and their tributaries, and throughout the entire
Mississippi valley, as well as in many of the river systems of the Atlantic slope, and those
flowing into the Gulf of Mexico. Its abundance and extensive distribution have led the
United States governinent to conduct experiments relative to its introduction as a “ food
fish,” thus far with satisfactory results. In form and appearance it is similar to that of a
larval salamandrid, and is subcylindrical, body slightly and head markedly depressed,
with three large and bushy external gills upon each side, having two gill slits between
them, and with four decidedly robust legs, each with four toes. The tail, which does not
exceed half the length of the body, is compressed and surrounded dorsally, caudally, and
ventrally by a caudal fin, rounded in outline at the tip. Sexually mature adults seem to
be fairly constant in size, varying in total length between about 27 cm, and 34 cm.

The color markings of the integument are due to scattered pigment cells, the varia-
tions in the frequency and size of which result in the formation of all possible shades
between no color at all and dense black spots. As these cells are very large, and in most
places evident to the unaided eye, the principle is seen upon close inspection to be similar
to that of a modern half tone illustration, which a careful scrutiny resolves into a series of

black dots, although in the latter case the darker shades are due to increased size, and in

NECTURUS MACULATUS. 589

the former to an increased number of the elements involved. The ground color of the
skin, as may be seen on the ventral side, which is mainly free from pigment, is yellowish
or pinkish, but over the sides and back varying degrees of pigmentation produce different
shades of slate color, in places with a distinct bluish or purplish tint. The deepest pig-
mentation is that of the dorsal side where certain densely pigmented areas form irregu-
larly rounded black spots upon a bluish-slate background, thus suggesting the specific
name of “ maculatus.” *

The skeleton consists of several independently separable portions and, like that of
other Urodeles, contains a large amount of unossified hyaline cartilage. By far the
largest portion consists of the vertebral column, the skull, to which is attached the hyo-
branchial complex, and the posterior extremities, which are attached by the ilia to a single
sacral vertebra. The two halves of the shoulder girdle are free from one another and
from the rest of the skeleton, and they, with their corresponding free limbs, form two dis-
tinct skeletal parts. The remaining skeletal elements are the nasal and optic capsules,
the two laryngo-tracheal cartilages, and the series of rudimentary sternebra which lie in
the mid-ventral thoracic region. All of these latter parts are wholly cartilaginous and
entirely disconnected from other parts of the skeleton.

In arranging the descriptive material of this memoir, the above practical division of
the skeleton has been taken into consideration as well as the more usual morphological
one, and it has seemed best to arrange the subject in the order following. In this the
vertebral column will be first treated, together with the ribs and sternum. The visceral
arches and the free sense capsules will be treated with the skull, and the limbs and their
girdles will appear last. This brings the parts together in their topographical relation-
ships and will be found more practical than a wholly morphological division.

THE VERTEBRAL CoLUMN.

General Description.

The vertebral column, as is the case with fishes, shows little regional differentiation,
since the limbs are too small and weak either to modify the motions of the trunk by the
muscles attached to them, or to bring their places of attachment into prominence as
points of leverage or support. The only gain in this respect over the condition seen in

1 Just as this manuscript leaves my hands, I have received from Mr, Alexander Nielsen, of Venice, Erie county, Ohio,
an extensive dealer in Necturus, a specimen having a totally different coloring from the usual one, and Mr. Nielsen, who
has caught thousands of specimens, writes that it is the first of the kind he has ever seen. The ground color of this speci-

men is alight reddish buff, with no suggestion of the usual dark slate color. The back is covered with dark brown spots,
smaller than in the normal forms. In form and size it closely resembles the common species,

390 HARRIS HAWTHORNE WILDER ON

fishes is the direct attachment of the pelvic girdle, which by the medium of a specialized
pair of ribs, becomes articulated to a single vertebra, usually the 19th. This vertebra,
the sacral, lies naturally in the cloacal region and it seems a matter of doubt whether, as
in higher forms, to consider it the boundary between the trunk and the tail, or whether
as in fishes, to limit the latter region to those vertebrae which bear closed haemal arches,
the first of which is usually the fourth vertebra posterior to the sacral one, 7. e., the 23d.
The first of these alternatives seems the more natural and open to the fewest objections,
for, while it places the first few caudal vertebrae in the same list structurally with those
of the trunk, from which, indeed, they are practically indistinguishable, it avoids the
equally great embarrassment of leaving the same number as a nameless and anomalous
group intervening between the sacral and caudal regions, a relationship unlike anything
occurring elsewhere among vertebrates.

The statements given above, that the two vertebrae which have a definite distinction,
the sacrum and the “first haemal arch vertebra,” are usually the 19th and 23d respee-
tively, suggest variation in this order, a matter which has been made the subject. of
papers by G. H. Parker (96) and Bumpus (’97), and has been referred to by Waite (97).

Regarding the sacral vertebra, according to the first two authors it was the 19th in
65% of 127 specimens examined, the 20th in 27% and in the remaining 8% it was placed
“obliquely,” that is, with an attachment to the 19th vertebra upon one side and the 20th
upon the other. This obliquity is usually sinistro-dextral, 7. ¢., with the left sacral rib
in advance of the right (left, 19th; right, 20th). Out of eight such oblique specimens
examined by Bumpus, seven were sinistro-dextral, and only one dextro-sinistral.
Parker's two specimens were both sinistro-dextral.

Waite describes three oblique specimens, which show for the most part additional
abnormalities. In one of these the right sacral rib was on the 19th vertebra and the left
on the 20th, thus making the direction dextro-sinistral. In the other two cases the right
rib was on the 18th, a case not found by the other authors named, and the left on the
19th, again dextro-sinistral.

Waite also figures a case with two normal sacral ribs on the 19th vertebra, and an
additional smaller rib upon the right side of the 20th, also attached to the ilium.

A similar variation occurs in the position of the Ist haemal arch, although a careful
comparison by Bumpus has shown that variation here is entirely independent of that of
the sacral vertebra. Out of 98 specimens examined with reference to this, the 1st haemal
arch was borne on the 22d vertebra in 11 specimens, on the 23d in 82, and on the 24th
in 5,

The variation in the total number of vertebrae, which is considerable, is seen to be

mainly that of the caudal region, as it has been shown that there is a variation of but two

NECTURUS MACULATUS. 391

vertebrae at the sacrum, and but three at the first haemal arch. The results of counting

the vertebrae in 100 specimens are expressed by Bumpus in the following table :—

No. of vertebrae. No. of specimens.
45 : : : : 2
44 § : : 9
45 : ; : ; : 21
46 ; : : : : 18
47 14
48 16
49 11
50 5
51 : 4

100

By this it will be seen that the average total number is 45 or 46.

A Typical Vertebra.

In order to understand the structure of the separate vertebrae, any trunk vertebra
except one of the first two or three may be selected as a type and its parts studied in
detail, after which the differences seen in other regions may be noted, and a few special
vertebrae selected as worthy of individual study. The 16th vertebra has been selected
for this purpose, and four views are given of it in text figures 1-4. These figures were

drawn from a dried vertebra prepared by macer-

Centram, wit
TAA, ArT. Concavily Sus
aN : ie ant. Zy9@ Pophy
y oi) b loirh Qytic. Facet.
y

y) wf
Re
Wie?

ation in caustic potash and hence lack the cartilag-

inous portions and other related soft parts referred

to in the text.
This vertebra consists primarily of a ventrally

: M2 >
situated centrum or body; a neural arch, the dorsal le) 2 So
“hi ; Ne) 2 NBN
aspect of which is broadened out into a broad, flat i" 2 EH, eee
c a Yi 8 P/F blewrape bhy-

plate; and upon each side a complicated transverse a Siat carblage
\ ophysis.
\ Diapophy

process directed backwards and bearing a short rib

which articulates with it in two places.

Socket for cartilagi-
of Neéurapophysis

The centrum is in the form of a slender hour

ye
median process. evs tip

o = 71 Cr ale aS ea a =
glass, its ends marked by iS) deep ne shaped Fig. 1. 16th vertebra; dorsal aspect. X 3.

depressions, thus making the entire vertebra con-

392 HARRIS HAWTHORNE WILDER ON

spicuously amphicoelous. In the recent state
. . . : . int. 2ygapophy sis.

these depressions are filled in with portions of the :

original notochord which become restricted to

these intervertebral segments by the develop-

ment of the osseous vertebrae. Corresponding to ry |

yt

the shape of the space included by two approxi-

mated ends of vertebrae, these notochordal seg-
> Socket for
pace nae
Cartilage

ments are in the shape of two cones placed base
to base, the place where the two bases meet being

the line of separation between the two vertebrae ae ,
lost Zygapophysis , Newrapophyses

involved. pee it oh ried
The neural arch is much flattened dorsally, ty,

and its sides, the neural laminae, become so much erica

obscured by fusions with the elements of the Fig. 2. 16th vertebra; ventral aspect. X 3.

transverse process that they are seen distinctly

only when the vertebra is viewed from one end. The main bulk of the arch consists of

a flat neural plate, which in a dorsal view of the vertebra nearly conceals the rest.

This plate is approximately a rectangle with sides a little incurved. Anteriorly its cor-

ners become prolonged into a pair of rounded anterior zygapophyses, and posteriorly it
develops a pair of somewhat similarly

Fossa.

gi ta shaped posterior zygapophyses, between
Gnt. 2ygebobh = : Se ZB Neurapoph.

which extends a short, median process.

Paula r5serent As the posterior zygapophyses of each
Socket for Pheurape-

Gt erly physial Cartilage vertebra overlap the anterior ones of

of Centrum. )
the next succeeding, the articular sur-

v7 * Post. vertebral
Sheath for plewrape- fossa.

magia cacelase faces of the former lie upon the ventral,
Wig. 8. 16th wertepene Wet dateraliaspeotemcet: and those of the latter upon the dorsal

aspect. At the blunt posterior point of
the posteriorly directed median process, there is a small cup-shaped depression filled, in
the recent state, with a nodule of cartilage which forms the tip of the neural spine, the
osseous continuation of which extends along the median line of the neural plate, becom-
ing obsolete anteriorly.

The term “ transverse process” is a convenient name by which to distinguish certain
composite and irregular masses placed upon either side of the centrum and bearing the
proximally forked ribs. Each transverse process consists essentially of two cartilaginous
rods, dorsal and ventral, encased in an irregular bony mass. Of these two rods, the dor-

sal one, the diapophysis, proceeds from the side of the neural arch, while the ventral

NECTURUS MACULATUS.

(Sh)
eo)
oo

one, the pleurapophysis, is associated

with the centrum. The investing bony sat Pees ce niet
: . i Neurapep
mass is formed secondarily about these Pest 2ygobebhs

Diaper

rods by periostosis and consists essen-

_7 Dorsal Laces

tially of two simple investing sheaths
SacweF For
Cartilage

and two flattened laminae, of which Lateral Coming —

of Neural arch

one is vertical and dorsal, connecting

the two sheaths, and the other horizon- peatey.

Fossa. Post Comcevily
of Centrum

tal and ventral, applied to the ventral
side of the pleurapophysial sheath and Fig. 4. 16th vertebra; posterior end view. X83.
spreading out proximally over nearly

the entire length of the centrum, with which it becomes fused. This latter lamina is
perforated by the ventral foramen, situated immediately posterior to the pleurapophysis
and transmitting the collateral vertebral vessels. As the dorsal lamina meets the ventral
one nearly at right angles, they form with the side of the centrum, two fossae, the anterior
and the posterior vertebral fossae. The posterior fossa is much the deeper and communi-
cates with the region ventral to the vertebral column by means of the ventral foramen
just mentioned.

In the larva the cartilaginous diapophysis and pleurapophysis are directly contin-
uous proximally with a lateral mass, the “rib bearer” (Rippentriiger) of Gippert, while
distally they unite to form the cartilaginous rib, the whole mass being at this time a con-
tinuous piece. The process of ossification within and around the rib bearer and the sides
of the vertebrae, and especially the growth of the osseous sheaths about these rods, gradu-
ally cut them off from one another and restrict them proximally into tapering points, so
that in a macerated adult vertebra, from which all the cartilage has been removed, the
moulds of the parts in question are seen as very deep and conical pockets.

Later ossification separates the free rib from the encasing sheaths of the transverse
process, and joints are thus formed between the two bodies. The part of the rib thus
segmented from the diapo hysis forms the tubercular process and the part once in con-
nection with the pleurapophysis becomes the capitular process, both processes retaining
throughout life an articular connection with the rods from which they originally separated.

Embryological investigation has rendered it probable that the pleurapophysis and
its associated capitular head represent the primary condition, and that the cartilage of the
rib bearer becomes dorsally extended for increase of support, developing later the second-
ary or tubercular attachment. Géppert has shown this secondary growth of the tuber-
cular connection in the case of Triton and some other Urodeles, a growth which proceeds

from the ribs and follows along a line of connective tissue until it reaches the rib bearer

594 HARRIS HAWTHORNE WILDER ON

at its dorsal prolongation, applied to the side of the neural arch. As he expresses it in
his summary, “Die dorsale Spange ist ein verlingertes Tuberculum, das im Dienst einer
ausgiebigeren Befestigung der Rippe steht.” Unfortunately he fails to find this dorsal
connection in Necturus, and states that “der dorsale Theil des Quertortsatzes tritt nicht
so deutlich als ein Balken hervor wie bei Salamandra und entbehrt auch hier der distalen
Héhlung, da die dorsale Rippenspange den direkten Anschluss an den Querfortsatz nicht
erreicht.”

As shown below, and in several of my figures, this “distale Héhlung” at the tuber-
cular articulation of the transverse process is often very evident, and through the
medium of the enclosed cartilaginous rod, articulates with the tubercular process of the
free rib. Gippert’s description (96, p. 398) of the condition in Salamandra corresponds
so completely with that of the adult Necturus that it could well be substituted for the
description given in this work. The “* Abweichungen ” which he finds in Necturus do not
exist.

A summary of the main features of a typical vertebra and its relations to other parts
of the skeleton may be given as follows : —

1. Osseous elements. Centrum; neural arch, including dorsal plate and lateral
laminae; transverse process, including sheaths of rib bearers, and the associated dorsal
and ventral laminae.

2. Cartilages and other soft parts. Two pairs of lateral rods or rib bearers; a neural
spine; intervertebral notochordal cones.

3. Articular surfaces. Two for the centrum; two anterior and two posterior zygapo-
physes; two tubercular and two capitular articulations for the corresponding heads of the

ribs.

Comparison with Other Vertebrae.

A study of the separate features brought out in the above description, as they occur
in the rest of the vertebral column, yields the following results. These, for the sake of
brevity, are expressed somewhat in the form of a table in which a few obvious abbrevia-
tions are employed. Thus the vertebrae are expressed by consecutive numbers, begin-
ning the enumeration with the atlas, which is 1; S is the sacral vertebra (19 or 20),
and H is the first one bearing a haemal arch (22 to 24).

Neural and haemal spines. —The neural spine forms a prominent ridge in 1; some-
what depressed and rounded at the end in 2. The neural spine of 3 shows the maximum

of size in the trunk vertebrae and is here the most erect, 7. e., makes the greatest angle

NECTURUS MACULATUS. 395

with the longitudinal axis of the vertebra. From there as far as §, it is gradually smaller
and more depressed. After 8 it becomes again larger and more erect, so that the spine of
21 is about equal to that of 5, and that of 22 surpasses it. It rapidly lengthens as far as
26 to 27, where it reaches its maximum of length and is about equal to the haemal spine;
from then on, both neural and haemal spines become gradually reduced as the vertebrae
diminish in size.

A haemal spine appears suddenly on the 22d to 24th vertebra, usually the 25d, an
indication of it sometimes being found upon the previous vertebra in the form of a thin
bony bridge which stretches across the ventral face of the vertebra from the outer angle
of the transverse process to the centrum, thus covering the posterior portion of the ven-
tral lamina. In one case noted, the vertebra just anterior to H bore upon one side a
slender process, 4 to 5 mm. long, evidently representing an incomplete haemal arch.
The first haemal spine is of about the maximum size, and, after three or four vertebrae
in which this size is maintained, the haemal spines begin to diminish with the gradual
reduction of the vertebrae, as in the case of the neural spines. The neural and haemal
spines of the same vertebrae closely resemble one another, but, in the first few vertebrae
that possess both, the former take a slightly more erect position while the latter ie more
nearly parallel to the axis of the vertebrae. Towards the end of the tail both neural and
haemal spines sometimes appear bifurcated at the tip, or even double as far as their base.
In a given vertebra this may affect either spine alone or both spines.

Zygapophyses. — In 1, the anterior zygapophyses are wanting and the posterior are
near together. From 2 to S both sets are divergent and almost identical, save that the
entire vertebrae at about 15 to 17 are a little wider than elsewhere and hence the
zygapophyses are here farther apart. After 8 the vertebrae rapidly become narrow, and,
beginning with H, both sets approach very near together and are reduced in size. They
become obsolete at about 26, beyond which the vertebrae articulate solely by means of
the vertebral centra.

Ventral foramina.—These do not appear in the three anterior vertebrae and are
usually of small size in 4. From then on, as far as H, there is usually a pair in each
vertebra, although occasionally the foramen of one side may be converted into a notch
by a deficiency in the ventral lamina. In H, where the roots of the haemal spine form
thin laminae lying over the region of these foramina, openings occur not only in the ven-
tral laminae as usual, but also in the laminae connected with the haemal spine, and the
two sets of foramina are accurately superimposed. Beyond H the reduction of the entire
transverse process renders the ventral foramina unnecessary, but in the two or three
vertebrae succeeding H the new foramina through the base of the haemal spine may be

present, although usually appearing upon but one side.

396 HARRIS HAWTHORNE WILDER ON

Transverse processes.— These processes, quite rudimentary in 1, are represented in
2 by the dorsal lamina alone, while the ventral lamina appears but slightly indicated by
a curving in of the ventral margin of the former. <A continuation of this method pro-
duces in 3 a fair indication of a ventral lamina, which in 4 has become of normal size and
appearance. In vertebra 2 to 7, approximately, the dorsal laminae are set perpendicu-
larly, beyond which they grow more and more oblique. In the sacral vertebra they
become again nearly perpendicular, beyond which they rapidly become obsolete but
remain more nearly perpendicular. As they degenerate they separate into two portions,
a dorsal and a ventral, corresponding to the diapophysis and pleurapophysis (Géppert’s
“Balken”). This division oceurs at approximately the 25th vertebra, after which the
dorsal or tubercular portion disappears in 2 to 3 more vertebrae, while the ventral or
capitular portion persists for a few vertebrae farther.

Attachment of ribs. — As shown above, the ribs are bifurcated at their proximal ends,
the two being known as the dorsal or tubercular and the ventral or capitular attachments.
These relations of the soft parts are plainly indicated in dried specimens by the presence
and relative development of the sheaths, the laminae of which furnish a set of moulds
giving the exact condition of the cartilages, and thus the following items, although
obtained for the most part from dried specimens, will doubtless be found to be wholly
reliable. Vertebra 1 is entirely without ribs. In 2, 3, and 4, both tubercular and capitu-
lar ends of the ribs are of large caliber and both are attached to cartilaginous rods. In 2
the capitular sheath is exceedingly small and the tubercular one a little larger. In 3 the
tubercular sheath is quite large and the capitular one a little smaller, and in 4 both are
large, the capitular sheath being slightly in excess of the other. From 5 to S the capitu-
lar sheath is always present but there is no distinct tubercular one, this attachment being
one of simple osseous contact. In the sacral vertebra there are two sheaths, one for each
attachment. They are of about equal size and are larger than in any other vertebra.
These two attachments of the sacral rib are figured and described by Géppert (96,
p- 402), but, since he has failed to find them in the other vertebrae and has even stated
that a tubercular attachment does not exist, he is under the necessity of accounting for
them in some other way. This he does by describing a process (Hicker) which grows
out of the side of the rib bearer, and “dient der Befestigung der kurzen oberen Spange
der Sakralrippe.” Although these words are noncommittal, the implication is, that this
“ Hocker” and the connection thus formed are a new formation and not homologous with
the dorsal or tubercular connection. This leads to the inevitable conclusion that the
dorsal or tubercular connections of the sacral ribs are not homologous with those of the
others, a conclusion which a moment’s consideration of an entire adult skeleton would

render invalid. The vertebrae which intervene between the true sacral vertebra and the

NECTURUS MACULATUS. 397

one bearing the first haemal arch, vary considerably in their relation to costal elements.
The first of these vertebrae in many cases bears a small pointed rib with a capitular
attachment alone, but this seems to be more usual in cases in which the sacral vertebra is
the 19th. In one such case with the first haemal arch upon the 22d vertebra, the 20th
bore small ribs and upon the 21st appeared very minute capitular sockets, which, as this
was a macerated specimen, might have originally borne rib rudiments. There seems to
be much variation in these as well as in other particulars in this very variable region, but

the specimens examined have been too few in number to give any general conclusions.

Study of Special Vertebrae.

Atlas.—The main characteristics are the reduction in total length to about one half
that of a normal trunk vertebra, the peculiar shape of the centrum, the rudimentary con-
dition of the transverse process, and the conspicuous and projecting neural arch. The
posterior end of the centrum is normal but anteriorly it broadens rapidly to form a wide
wedge, the base of which is directed forwards and bears the two condyles. These are
slightly concave, elliptical surfaces, inclined a little outward, and meeting in the median
line where a small but conspicuous tubercle is formed that fits into a space upon the ven-
tral side of the skull between the two exoccipitals. The ventral aspect of the centrum is
somewhat hollowed and a little roughened for muscular attachment, a peculiarity shared
by the 2d vertebra. The neural arch is large and conspicuous. Its sides are flattened
and obliquely set, and it ends in the mid-dorsal line in a rounded edge or keel, the poste-
rior end of which bears a small socket for the cartilaginous tip of the neural spine, as in
other vertebrae. The two posterior zygapophyses are large and heavy. The transverse
processes are rudimentary and very variable, even upon the two sides of the same
vertebra. Their most constant parts are a usually bifurcated piece projecting obliquely
backwards in the same position as, and probably homologous with, the dorsal lamina of
the succeeding vertebrae ; and a much shorter spine, placed ventral to this and also pro-
jecting backwards. The larger process is pierced by an obliquely directed foramen which
communicates with the neural canal and transmits the internal carotid artery and the first
pair of spinal nerves.'. Between this process and the short ventral spine there is a narrow
but very deep fossa. By the bifurcation of the original transverse process and the addi-
tion of the spine ventral to it, there are formed three backwardly directed processes which

1 Hoffmann suggests that the phenomenon of a pair of spinal nerves boring through the substance of a vertebra in an

animal in which the nerve exits are normally intervertebral. is an indication that the “atlas” really represents two fused
vertebrae, the line of separation being marked by the nerve and its foramen.

398 HARRIS HAWTHORNE WILDER ON

are very variable in their mutual relations. Occasionally all three will be distinct and of
about the same size, but usually the ventral spine is much smaller than the other two.
There may also be varying degrees of bifurcation of the genuine transverse process, or it
may present one or more small processes in addition to the usual three.

Vertebra bearing the first haemal arch. — This is normally the 23d in the series, but
it may be the 22d or even frequently the 24th. After the sacrum, the neural plates rap-
idly narrow, and, by the time the vertebra in question is reached, this part has become
about as narrow as the base of the neural spine, so that the latter appears in direct con-
tinuation with it, the posterior zygapophyses appearing as lateral processes borne upon
the sides of the spine. This increase of the apparent length of the neural spine renders
it strikingly similar to the haemal spine, and thus is seen the first suggestion of that
dorso-ventral bilaterality which becomes so marked in the more posterior caudal vertebrae.
The transverse process is thin and poorly developed, and the occasional bifurcation of its
free end gives.an indication of its division into tubercular and capitular portions as seen
in the vertebrae immediately posterior to this. The haemal arch, which is a new element
and thus gives distinction to this and the following vertebrae, begins as a pair of low, thin
ridges springing from the ventral side of the centrum and uniting across the mid-ventral
line near the posterior limit of the vertebra in such a way as to frame in a haemal
foramen just ventral to the centrum, which, with the foramina of the succeeding verte-
brae, forms the haemal canal. The ventral laminae are perforated as usual by the two
ventral foramina, and, as they are nearly covered by the two ridges that form the roots
of the haemal arch, a pair of foramina occurs also in these latter opposite the usual
ones. In most instances the first haemal arch appears, as it were, suddenly and perfectly
developed, and the vertebra immediately preceding the one that bears it usually shows no
suggestion of such a structure. Occasionally, however, a rudiment of the foot or root of
the arch appears upon one or both sides of the previous vertebra in the form of a thin
plate lying parallel to the ventral lamina, and extended across from centrum to outer end
of the transverse process, and, in one instance observed, a slender haemal process appeared
upon the left side, which projected back over the following vertebra and plainly repre-
sented one side of a rudimentary haemal arch. There was no trace of this upon the
other side.

The caudal vertebrae posterior to the above, with especial reference to the terminal
ones. —The most conspicuous characteristic of the caudal vertebrae posterior to the above
is the presence of nearly equal neural and haemal arches running out into long spines
with their free ends pointed obliquely backwards. The reduction of the transverse
process, which begins just posterior to the sacrum, progresses rapidly after passing the

first haemal arch. The process first becomes divided into its two portions, the dorsal one

NECTURUS MACULATUS. 399

of which survives but two or three vertebrae, while the ventral one continues for two or

three more. Pari passu with the loss of the transverse processes, the zygapophyses

become reduced and finally disappear.

With the loss of these lateral elements the

vertebrae become narrowed from side to side, and, as this appearance is further increased

by the great extension of neural and haemal arches and spines, the entire column becomes

strongly compressed laterally, thus corresponding to the change in external form which is

so apparent in the living animal. In this more characteristically caudal region (circa

between 28 to 40) the hour glass shape of
the centra becomes more apparent and
in the reduction which follows towards
the tip, this element becomes more and
more conspicuous as the main portion of
the vertebrae. Thus, after about the
40th vertebra, rapid reductions take
place in the arches, both neural and
haemal, and the terminal one or two
vertebrae consist of centra alone, the
posterior end of the final one being
rounded and without the characteristic
cup. In this loss of arches it seems that
the neural arch disappears first, as in all
the cases examined several of the last
vertebrae consisted of body and haemal
arch alone, but the specimens examined
were too few in number to establish this
asalaw. From about the point at which
the vertebrae first become laterally com-
pressed, or more exactly, after the loss of
the zygapophyses, there is shown an
almost complete dorso-ventral bilateral
symmetry, so that in an isolated vertebra
it becomes extremely hard to distinguish
between the neural and haemal arches.
There is, however, a slight difference
which, though difficult of formulation,
may be perceived after a little study

and seems to consist mainly in a some-

(a)

Fig. 5. Caudal vertebrae ; (a) and (b) two examples of ter-
minal vertebrae ; (c) instances of double haemal spines ; (d)
and (e) two successive vertebrae in the same series showing
doubling of neural spines. The arrows in (d) pass through

the neural and haemal canals.

400 HARRIS HAWTHORNE WILDER ON

what greater size of the neural canal in comparison with the haemal, and a consequent
greater height of the anterior portion of the neural arch, which in this region becomes
prolonged and tubular.

A common phenomenon in these last caudal vertebrae, beyond about the 38th, is
an antero-posterior bifurcation or doubling of either the neural or the haemal spines or
both. Frequent examples of this are noticed in the accompanying figure (fig. 5), in
which appear several grades of this malformation, from a bifurcation of the free end to
what seems to be a complete doubling of the spine. This phenomenon is referred to

by Bumpus (’97) and figured in some of his radiograph illustrations.

Sternum.

The several cartilaginous rudiments which represent this part in Necturus are some-
what difficult of detection and thus entirely escaped the attention of the earlier investi-
gators. They consist of a number of thin cartilages found in several successive myocom-

mata of the pectoral region and confined mainly to the area covered by the overlapping

A B C D E f:

Fig. 6. Sternal pieces of six different individuals, showing variation.

epicoracoids. As will be seen by text figure 6, which represents the sternal elements
of six normal individuals, the cartilaginous pieces show great individual difference in shape
and size, and are capable of some variation in the myocommata involved. The largest
segmental portion or sternebruin is that of the 4th myocomma which is in general an
irregularly triangular or often bat-shaped piece. It lies just posterior to the epicoracoids,
in the angle formed by their overlapping edges, and is sufficiently large and superficial to
give rise to a part of the fibers of the pectoralis muscle. This muscular attachment as

well as its position relative to the shoulder girdle fixes its identity as the homologue of

NECTURUS MACULATUS. 401

the rhomboidal sternal plate of the higher Urodela, and it would seem probable that this
latter has resulted from the development of this piece alone. The sternebrum associated
with the 3d myocomma forms a long and slender open V, which may or may not be
continuous with the sternebrum of the 4th. Sternal elements are often found both in the
2d and in the 5th myocomma, usually as a pair of cartilages, or as a unilateral piece.
Of these, the one associated with the 5th myocomma is much the more frequent, being
found in rather more than half of the specimens thus far examined, either upon one side

or both. Sternal elements in the 2d myocomma are rarely seen.

Ribs.

These have already been partly described with the transverse processes of the ver-
tebrae. In the larva they are cartilaginous and are directly continuous with the two
cartilaginous rods enclosed in the osseous transverse process, and are thus bifurcated at
their proximal end. During the later process of ossification the
ribs become ossified as separate pieces and the cartilage remain-
ing between the proximal ends of the ribs and the sheaths of the
vertebral rods becomes divided across, thus furnishing two sets of “és
articular cartilages. By the shape of the first two ribs in the ies
adult, it would seem that both of the rods are equally involved and bs
simply fuse to form the ribs, but farther down the ventral or capit- NS
ular attachment retains its full size and importance, while the am
dorsal one, although more in the same line with the free end, FS
becomes much reduced, and in many-vertebrae its attachment is
merely ligamentous. Thus, described anatomically, the ribs of the ae
2d and the 3d vertebrae possess almost equal capitular and tuber-
cular heads, and show a tendency to become bifurcated at their age
tips. In the ribs of vertebra 4 to 6 the two heads are still almost
equal but the distal part of the rib is a simple rod; and farther =

“SS

: . yh
down, from about the 8th vertebra on, the capitular head forms
the main attachment, and meets at a decided angle the straight Fig. 7. Ribs taken from
piece formed by the free end and the tubercular head. The sacral EI Tee A Ly

and placed so as to show

rib possesses two very large and equal heads, but is not bifurcated

the anterior aspect. The

distally, and beyond the sacrum the rib element is a variable tubercular head is above
quantity (v. sup.). It would thus seem, judging from purely puck tle. captiulan below,

The numbers refer to the

anatomical evidence, that the condition described by Géppert sttebea to which cashiean

as characteristic of Necturus is not a universal one applicable to belongs. X 2.

402 HARRIS HAWTHORNE WILDER ON

all of the vertebrae, but is restricted to a certain region, approximately that of vertebra
8 to 18.

It would seem important to investigate the development of transverse process and
rib in certain of the other vertebrae, for example the 2d and the 4th. Regarding the
sacral vertebra and its rib, as stated previously, Géppert’s conclusion, which does not
allow a complete homology with the rest of the vertebral column, seems anatomically

improbable.

Tur SKULL AND VISCERAL SKELETON.

Anatomically, the bones of the head consist of (1) the skull in the restricted sense,
which includes the cranium, the ear capsules, and the upper jaw pieces, (2) the nasal and
optic capsules, and (3) the visceral skeleton, represented by the mandible, the hyobran-
chial apparatus, and the laryngeal cartilages. In point of origin the elements are three
in number: (1) the primitive cartilage, still seen in the sense capsules, the primordial
cranium, and the most of the visceral skeleton; (2) cartilage bones, or ossifications of
localized portions of the primary cartilage; and (3) dermal bones of integmental origin,
investing the surface of the cartilage in skull and mandible.

Although there are osseous elements from two sources, the proportion of unossified
cartilage left in the adult head is very great and compares in this respect with the condi-
tion seen in the Chondrostei rather than with that of other fishes or of most amphibians.
This large amount of cartilage is very suggestive of an embryonic or immature form, and
suggests either the possibility which has already been frequently expressed, that Necturus
is in a sense a permanent larva, as is the case with the Axolotl, or that it is an exceed-
ingly primitive form, perhaps nearer the fishes than any other amphibian.

An orderly treatment of this complex bit of anatomy is extremely difficult and may
perhaps be facilitated by the presentation of a classification of the subject, which will be
adhered to in the later description.

GENERAL DeEscrIPTION.
Tur CHONDROCRANIUM.
Tut Ossrous ELEMENTS.
(a.) Otie and occipital regions.
1. Pro-otie.
2. Opisthotic.
3. Exoccipital.
4. Quadratum.

5 and 6, Paraquadratum and operculum.

NECTURUS MACULATUS 403

(b.) Brain case.
1. Frontal.
2. Parietal.
8. Parabasal.
(c.) Upper jaw.
1. Maxillary arch.
a. Premaxillary.
2. Palatopterygoid arch.
a. Vomer.
b. Palatopterygoid.
Tur ViIsckRAL SKELETON.
(a.) General morphology of the arches.
(b.) Mandible.
1. Dentale.
2. Angulare.
3. Spleniale.
(c.) Hyobranchial apparatus.
(d.) Suspensorial relations of the hyoid.
Tur FREE SENSE CaAPsULEs.
(a.) Nasal capsule.
(b.) Optic capsule.
Tue TEETH.

ComMPARISON OF NOMENCLATURE.

General Description of the Skull and its Parts.

The term “skull” is used here in its most restricted sense and does not include either
the mandible, the hyobranchial apparatus or the nasal and optic capsules. These elements
are for the most part easily detachable from the firmly consolidated skull, and are thus, in
the anatomical sense, not to be included with it. A single exception may be made in the
case of the nasal capsule with regard to which the usage of authors differs, since, although
anatomically distinct in Necturus, it belongs genetically to the primordial skull. Wieders-
heim, for example, who is correct from the morphological standpoint, figures it in his
drawing of the skull, while both W. K. Parker (in Proteus) and Huxley fail even to find
it. In this paper it will be omitted from consideration at present to receive special treat-
ment later on.

The skull, then, denuded of all its extraneous elements, will present an appearance
much as is given in figures 2 and 3 (plate 63). The general outline suggested a pentagon

to Huxley, but in order to see it as such, one should imagine the ligaments restored that

404 HARRIS HAWTHORNE WILDER ON

stretch from the outer end of the premaxillaries to the quadrates. The five angles of
the pentagon may be named from the bones that compose them, (1) premanillary, (2
and 3) two quadrate, and (4 and 5) two opisthotic ; and the five sides in a similar way
may be termed (1) occipital, (2 and 3) the paraquadrate, and (4 and 5) the palato-ptery-
void. When dissected or macerated sufficiently to show its parts, it gives the distinct and
correct impression of a cartilaginous structure overlaid by flat dermal bones laid on and
overlapping “like shingles,” to borrow Parker’s apt comparison, and showing in places
certain intervals through which the primordial skull and its ossifications are distinctly
seen. When viewed dorsally (pl. 63, fig. 2) the most conspicuous dermal elements are
the frontals and the parietals, which form almost the entire roof, and at the anterior end
of the former appear the little premaxillaries. Of the maxillary arch, which normally
extends from the outer end of the premaxillary to the quadrate, there is no trace, other
than a firm ligament when in the recent state, but the inner arch, parallel to this, is well
represented by the vomer and the palato-pterygoid, the outer edges of which are seen in
the figure. The only other dermal bones visible from above are the paraquadrates which
are seen extending along the postero-lateral sides of the pentagon. The dorsal aspect dis-
plays also parts of four cartilage bones: the quadrates forming the antero-lateral angles,
the pro-otics and opisthotics in the interval between the parietals and paraquadrates, and
the exoccipitals which border the foramen magnum and form the condyles. Upon the
ventral side (pl. 63, fig. 3) the most conspicuous bone is the huge parabasal, a dermal bone
nearly covering the roof of the mouth, anterior and lateral to which are seen the denti-
gerous premaxillaries, vomers, and palato-pterygoids. The paraquadrates form a part of
the outer margin and a conspicuous process from them is attached to a similar process that
projects from a very small oval bone, the operculum, which closes the opening into the
otic capsule. The same four cartilage bones seen from the dorsal side appear here, three
of which strengthen the otic region while the fourth forms the suspensorium for the
attachment of the mandible.

The Chondrocranium.

If, now, as may be easily accomplished in a macerated skull, the dermal or “ shingle”
bones be carefully removed, there remains a very curious piece of cartilage suggestive of
an inner framework, the primordial skull, or chondrocranium. The cartilage bones, pre-
viously noticed and now entirely uncovered, are very evidently the result of localized
processes of ossification within this, which have taken place in what was once a continuous
cartilaginous mass, the homologue of the wholly cartilaginous skull of the present
Selachians.

NECTURUS MACULATUS. 405

The chondrocranium consists essentially of a delicate framework which runs around
the entire cranial cavity and with which certain other cartilaginous elements are associ-
ated in varying degrees of intimacy of relation. The framework consists in the embryo
of a pair of lateral bars, placed parallel to one another and divisible into an anterior ele-
ment, the trabecula, and a posterior, or parachordal element. In the adult these names
are retained as designations for the regions corresponding to those elements, the trabecu-
lar region being that anterior to the otic capsules, while the part associated directly with
' the capsules themselves is the parachordal region. Posteriorly, between the otic capsules
the frame is completed by a pair of cartilaginous bands or arches, of which one, the supra-
occipital arch,’ passes dorsal to the nerve cord, and the other, the basi-occipital arch, lies
ventral to this latter part. According to Miss Platt, the supra-occipital arch is really the
neural arch of a vertebra anterior to the atlas, which has fused with the skull to increase
its strength. At the anterior end of the skull the frame is completed by the development
of an internasal plate which connects the convergent trabeculae. Anterior to this the
free ends of the trabeculae are continued a short distance in the form of small rostral
processes.

Of the originally external elements, the two large otic capsules are the most
intimately related and are completely fused with the posterior portion of the primary part
and thus form the largest and most voluminous portion of the chondrocranium. Anterior
to these are the semi-detached quadrate cartilages, connected with the sides of the tra-
becular frame by a narrow process, the trabeculo-quadrate isthmus. Still farther forward
and separated from the latter by an interval in which are situated the eye and its acces-
sory organs, appears a pair of distinct cartilages, attached to the trabeculae by connective
tissue. These are the ante-orbital processes plainly representing a rudiment of the sub-
ocular arch, or pterygo-quadrate process, so enormously developed in the frog and other
Anura, an homology suggested by W. K. Parker’s term of “ ethmo-palatine process.”

The other cartilaginous parts topographically associated with the chondrocranium are
the nasal and optic capsules, of which the former lies directly upon the anterior end of
the completed skull, and barely comes in contact with the chondrocranium, while the
latter is completely isolated. These parts will receive separate treatment later.

The ossifications of the chondrocranium (= “cartilage bones”) are few in number in
comparison with those of most Urodeles and consist of but four pairs: (1) the quadrates,
used to form a strong support for the mandible, (2) the exoccipitals, forming the condyles

1Miss Platt objects to the term “‘occipitale superius” for the dorsal cartilaginous arch on the ground that the supra-
occipital is not an amphibian bone and prefers “interoccipitale,” as suggested to her by Gaupp. This term, however, leaves
out of account the ventral arch, unless we use the terms “interoccipitale dorsale ” and ‘ interoccipitale ventrale ” which are
cumbersome and possess the disadvantage of the substitution of new terms for old and readily understood ones. Moreover,
a portion of the Amniote supra-occipital is preformed in cartilage,and it seems more than probable that there is at least a
partial homology between it and the cartilaginous arch found here. The same may also be said of the basi-occipital piece.

406 HARRIS HAWTHORNE WILDER ON

and thus strengthening the important articulation of the head with the vertebral column,
and (3 and 4) two ossifications of the otic capsule. These, of which the anterior is the
pro-otic, and the posterior the epi- and opisthotic combined, are in the form of hollow cups
which fit over the anterior and posterior ends of the otic capsule, looking, to borrow W.
K. Parker’s vivid simile (in Proteus), “as if they were ready to dehisce transversely like
the pyxidium of the pimpernel (Anagallis) .”

The Osseous Elements.

The otic and occipital regions.—1. pro-otic. This very irregular bone consists
essentially of a conical cap to which are added four projecting processes, one upon the
inner and three upon
the outer aspect. The
inner process, the ala, is
flat and wing-like and is

appled along the outer

edge of the base of the

VENTRAL External Lateral. trabecula. Of the three
(a) ( 6) processes which project
Cenat fo Fos for ome from the external face
P. q. Vv ant Nort Semi-€ Conas of Gn vert Sarai
ale & N aa of the bone, the two an-
Y hg. d . - Groove for ¢
cs ae ee ORE ort terior ones, dorsal and
ay) 7
* ventral quadrate proc-
ou ; esses, possess concave
ee and roughened articular
VENTR. =
Saris surfaces to recelve cor-
DORSAL INTERNA B
(c) (d) responding processes of
ey :
the quadrate cartilage.
Fig. 8. Four views of the right pro-otic. p. q. d., dorsal quadrate process ; p. q. V., | ; 8
ventral quadrate process ; f. sty. m., stylo-mastoid foramen. The ventral process

may be considered an
extension of the ala, and it receives a rounded protuberance situated in the middle of the
posterior margin of the cartilaginous quadrate. The dorsal quadrate process forms the
external dorsal edge of the bone and displays a more elongated articular surface for the
reception of the otic process of the quadrate. The remaining external process has no
relation with other skeletal parts and its probable use is that of attachment for some

important muscle or ligament, as it is very constant in appearance and always well

NECTURUS MACULATUS. 407

developed, but this point must be left for later investigation. It runs parallel to and very
near the ventral quadrate process, and between the two there is thus formed a deep notch,
at the bottom of which is seen the outer opening of the stylo-mastoid canal, that transmits
the facialis nerve. This canal enters the pro-otic at its inner posterior margin, traverses
almost its entire ventral wall and emerges by the notch just described.

Internally the pro-otic is hollowed out for the reception of the anterior portion of the
membranous labyrinth. At the junction of its internal and dorsal edges there is a canal
for the transmission of the anterior vertical semi-circular canal, the ampulla of which les
in a deep but narrow excavation at the apex of the cone. The anterior portion of the
horizontal semi-circular canal is lodged in a groove upon the outer side of the internal
excavation, just beneath the dorsal quadrate process.

The pro-otic may come in contact with three bones, the actual condition varying
with the age and degree of development. Its ventral surface is always overlapped by the
parabasal, and is usually somewhat roughened over the region of contact. The parietal
partly overlaps it dorsally. In large and mature specimens the anterior border of the
operculum may touch it upon the ventral side.

2. opistHoTic. This bone consists of a hollow cup, forming the posterior lateral
angle of the skull, and investing the posterior part of the otic capsule. It is irregularly
conical in shape, being somewhat flattened like a triangular pyramid, and thus presents

externally for examina-

Canal for
horigental
™ Semi-c Canal

a
\

tion three nearly flat sur-

4 faces, a dorsal, a ventral,

eo and an internal lateral,
cele Canal py nae facing the occipital con-
Pern: dyle. Of these surfaces,
- : —" Mastoid :

mastoid Position 05 Baers the dorsal is the flattest,

Procers. Post. vert. .
Pent Cicer: the ventral is quite con-

VENTRAL DORSAL INTERNAL

vex, and the internal lat-

Fig. 9. Three views of the right opisthotic. 3. Taken from a specimen eral is marked by a deep

smaller than in the case above. : .
groove along which lies
; the vagus group of cra-
nial nerves. The apex of the cone or pyramid is directed backwards and a little out-
wards, and terminates in a slightly elevated process, which may be called the mastoid for

convenience, since it has some analogy with the mammalian process of that name.

Internally the bone presents two canals, a large central depression, and in the bottom
of the latter a still deeper recess into which the canals open. The two bony canals, which

are excavated in the dorsal wall of the hollow cone, may be distinguished as the lateral

408 HARRIS HAWTHORNE WILDER ON

and median, and lodge, respectively, the horizontal and posterior vertical semi-circular
canals of the membranous labyrinth. The ampullae of these two canals lie at their
posterior union and are situated in the deep recess. The large central depression forms a
portion of the median chamber which contains the membranous vestibule with its large
otolith. The opisthotic comes in contact with two bones, the paraquadrate and the exoc-
cipital. The former merely overlaps with its posterior portion the outer margin of the
opisthotic, but the exoccipital forms, at least im large specimens, a definite union through
the medium of distinct processes projecting from the two bones.

The side to which a given opisthotic bone belongs is best determined by the canals
and the deepest recess. The canals lie nearest to the dorsal surface and the recess is on
the internal side.

3. EXOCCIPITAL. This pair of bones, the bodies of which form the occipital condyles,
embraces and defines the foramen magnum and, with the addition of dorsal and ventral
eartilagmous arches, entirely encloses it. Each bone consists of a body and two flat
processes, the supra- and basi-occipital alae. The body is practically identical with the

condyle, and its posterior aspect forms

erie dae a rounded surface for articulation with the
A ala.
ola atlas. The dorsal or supra-occipital ala

is nearly perpendicular to the rest of the

bone and to the floor of the skull and is

Basi-occipital
QLa. @

set obliquely, at an angleo about 45° with

both longitudinal and transverse axes of

the head, so that, when viewed from above,

POSTE RIOR. veer the median occipital region forms a con-
on a ote eee Ae EC Ce 3. ; : } ‘ os
Hig, 10: ‘Two viewsiof the riphtexoceipital 1 spicuous re-entrant angle. This ala arises

from the body as a narrow stalk which rap-
idly widens, and thus resembles a triangle resting upon its apex. The dorsal margin, or
base of the triangle, is grooved for the reception of the cartilaginous arcus supra-occipi-
talis which spans the median interval between this ala and its opposite.

The ventral or basi-occipital ala lies exactly in the horizontal plane, coincident with
that of the parabasale, and is applied so closely to this latter bone by its ventral surface
that the separation is often attended with some little difficulty. Both W. K, Parker and
Wiedersheim found the exoccipitals in Proteus actually a part of the parabasal, or, as the
latter expresses it, ‘‘ synostotisch verbunden.” This is nearly the case in Necturus, but, as
the two bones are really separable in all cases tried, it is possible that this may be the
actual condition in Proteus also. The ventral ala is triangular in shape but is attached to

the body of the bone by a side and not an angle, and, when viewed from the ventral side,

NECTURUS MACULATUS. 409

forms with the condyle a right-angled triangle, the right angle facing the middle line and
the hypothenuse forming the lateral side. The small interval formed between the oppos-
ing right angles of the two exoccipitals is spanned by a ventral cartilaginous arch, the
arcus basi-occipitalis, which, with the supra-occipital arch and the two bones under consid-
eration, completes the enclosure of the foramen magnum.

This basi-occipital cartilage is of greater extent in young and larval animals, and the
ventral alae result from the gradually increasing ossification of its lateral ends. As this
process continues towards the center, the inner angles of the bone nearly touch one
another in old adults, while the cartilaginous arch is correspondingly diminished.

The exoccipital comes into contact with four bones: (1) the parabasale, which is
closely applied to the basi-occipital ala, (2) the opisthotic, which, in adults, touches both
the body and the dorsal margin of the supra-occipital ala, (5) the parietal, a bit of the
posterior margin of which becomes applied to a corresponding portion of the dorsal mar-
gin of the supra-occipital ala, within its contact with the opisthotic, and (4) the atlas,
which articulates with the condyles by a movable joint.

The natural relations of an isolated exoccipital are best learned from the flat ventral
surface. The condyle is posterior and the long straight edge is internal.

4. quapRatuM. This is primarily a cartilaginous element, associated with the
primordial skull, and representing the proximal (posterior) end of the palato-pterygo-
quadrate arch, the functional upper jaw of the Selachians. In the Anura this arch is
entire, but in Necturus it is represented by its two ends alone: the antorbital process,
which represents its anterior, and the quadratum, its posterior portion.

Functionally it serves as a “suspensorium,” or piece interposed between the skull
and the mandible, and forming an articular surface for the latter. This joint occurs at
its outer anterior angle, and that region of the quadrate becomes ossified, plainly to give
strength to this very important joint. It thus happens that there is in the adult, both an
osseous and a cartilaginous quadrate, the former being situated externally and the latter
towards the median line.

The irregular shape of the quadrate taken as a whole may best be seen by a reference
to figures 10 and 11 (plate 64). The external osseous portion consists of an anterior
articular process of very hard bone, fitted with an articular socket to receive the rounded
cartilaginous knob (articulare) of the mandible, and a hollow trough-like posterior process
fitted over the outer edge of the cartilage much as in the case of many of the dermal
bones. The cartilaginous portion consists of a flattened plate which attaches by its broad-
est side to the bony portion and tapers down to a narrow isthmus as it approaches the
skull. Contact with the latter is formed by means of anterior and posterior extensions of

the isthmus, which become applied to the outer side of the trabecula just in front of the

410 HARRIS HAWTHORNE WILDER ON

otic capsule. It is difficult to determine, in a piece of this shape, the location of the proc-
esses cited by authors in their description of the quadrate of other Amphibia. The pos-
terior lateral extension which runs along the side of the bone is probably the otic process,
and the narrowed part, or its posterior prolongation, the pedicel. Of the two processes
which form the isthmus and become applied to the trabecula, the anterior one may possi-
bly be the ascending process. A slight angle seen in the anterior margin is doubtless the
rudiment of the “ cartilaginous pterygoid” found in most Urodeles, the extension of which
to the antorbital process would form the palato-pterygoid arch which is wanting here.
Aside from its attachment to the trabecula, the quadrate, osseous and cartilaginous, enters
into more or less complete attachment to four bones. As described above, two processes
of the pro-otic form quite definite articulations with the quadrate cartilage, the ventral
one receiving an articular surface formed by a thickened piece of cartilage in the middle
of the posterior margin, and the dorsal one being applied along the inner edge of the otic
process. The paraquadrate overlaps it externally, and the outer posterior corner of the
palato-pterygoid fits into a groove in the inner side of the articular process and overlaps a
raised area of the quadrate cartilage. The cartilaginous articulare of the mandible forms
a movable articulation with it.

In determining the position of an isolated osseous quadrate it may be remembered
that the side showing the hollow groove is
internal, that the larger end is anterior, and

that the broader, plainer surface is dorsal.
5 and 6. PARAQUADRATUM AND OPERCU-

LuM. ‘These elements, of which the first is a

5
5
x
r
°
cc
‘
S

dermal bone, and the second an ossification

= =
aS 5
. ys : :
ie ay re os of a detached portion of the otic capsule, are
eMcular procecs

se termsetesne Pore ee ae closely connected topographically and joined

to one another by | strong ligament which

_-- Colawellar ---...... # .
aaa Process of
Opercutum

unites processes in each bone mutually
directed toward the other. The paraquadra-
tum, the shape of which suggested a boom-
erang to Huxley, attached in its normal man-
ner to the little discoidal operculum, presents
the appearance given in figure 11. When in
INTERNAL place upon the skull, the paraquadrate lies
along the outer side of the otic region, form-

EXTERNAL.

Fig.11. Right paraquadratum and operculum. ;
Two views. X 3. ing the two sides of the pentagonal outline

of the skull designated above as “ paraquad-

NECTURUS MACULATUS. 411

rate,” while the operculum fits tightly and exactly into the fenestra ovalis, a large
lateral opening in the cartilaginous otic capsule.

The most conspicuous character of the paraquadrate is the short and blunt opercular
process, which projects from the inner edge of the curved bone at approximately the
middle and thus divides it into two nearly equal portions, the anterior or quadrate,
and the posterior, or opisthotic halves. These two halves are nearly flat, or like very
shallow troughs, the planes of which are set nearly at right angles to one another; and
when in place, the anterior portion is placed nearly perpendicularly to the skull over-
lapping the quadratum along the side, while the posterior portion is nearly horizontal and
covers the outer part of the dorsal surface of the opisthotic.

The paraquadrate is always connected with three bones and may possibly come in
contact with two more. By its opercular process it forms a definite articulation with the
columellar process of the operculum, and its two flat portions are applied to the outer sur-
face of the quadrate and opisthotic. Aside from these, its anterior and posterior ends
may touch the outer corners of the palato-pterygoid and parietal respectively. If an
isolated paraquadrate be held so that the opercular process is directed downwards, it is in
its normal position and the position of the planes of the anterior and posterior portions
will serve to locate it.

The operculum fits something like a stove lid into the fenestra ovalis, an oval open-
ing in the cartilaginous otic capsule. It consists of a flattened oval base or body bearing
upon its outer surface an irregular columellar process. By means of this process it articu-
lates with the opercular process of the paraquadrate and normally touches no other bone,
although in old animals the pro-otic bone may enlarge sufficiently to come in contact with
its anterior edge. The columellar process is directed upwards and a little forwards, and
will thus give the proper orientation for the bone.

The homologies and, consequently, the nomenclature of these two bones have been a
matter of much uncertainty and varient treatment among authors. For the first of these
I have selected the term ‘“ paraquadratum ” on the authority of Gaupp, who has proposed
it as at least a provisional term to indicate the dermal encasing piece associated in the
amphibians with the quadrate. He apparently inclines to the belief that this element
may prove homologous with the mammalian tympanicum, a term by which he designates
the piece in his revision of the ‘‘ Anatomie des Frosches.”” The operculum has become so
universally identified with the stapes of higher forms that in the first writing of the manu-

script for this work the word “ stapes”

was used and its probable homology stated. I was
led to a change of this view by the examination of a set of slides of a larval Necturus
of 44 mm. in which the part in question arises as a semi-detached bit of the cartilaginous

otic capsule, precisely as was seen by Stéhr in Triton and Siredon. This was equally

412 HARRIS HAWTHORNE WILDER ON

evident and presented a similar appearance in a series taken from a larva of 26 mm. This
observation, which seems perfectly clear and unassailable, and which corroborates the
results given by Stéhr in other Urodeles, is absolutely at variance with the statement of
Miss Platt, who says that “it arises independently ” and identifies as its anlage a mass of
cells which, in an embryo of 19 mm., is situated just in front (outside) of the fenestra
ovalis, and between it and the end of the ceratohyal. It is possible that this anlage
may represent a true columella or hyomandibular cartilage, which may later become
fused with the true operculum to form its projecting external process, although in
the larvae of 26 mm. and 44 mm., the series which I have examined, I can find no
trace of such a double origin of the parts in question.

The brain case.— This is formed in great part by five dermal bones, the two frontals,
two parietals, and the parabasale, of which the first four form its dorsal, lateral, and ante- .
rior walls, and the parabasale its floor. The box formed by these is deficient poste-
riorly, where it is completed by the otic capsules and other cartilaginous elements with
their ossified areas. Above, each
frontal is so completely welded to
ined “Vv its accompanying parietal by means
; of interlocking splints that they practi-
cally form one bone while the two sets
a6Na x PB (= uncinate proces of united pairs meet in the median

of Frontal ) : 5 A
line by means of somewhat thickened

flat edges, forming symphyses. Upon

Ht es ties °F the ventral surface of this roof, both
rabeeulae. (ouficle
stavicte) bones end down extensive processes

iB

sae eae Ae that lie just within the trabecular
arch of the primordial skull and make
an extensive ridge in the form of a
long narrow U, the loop being directed

Pine Gite anteriorly. This ridge forms the
ridge is_ reingorced

front and sides of the brain case and

by @ precess from é Vie 22>

reteset: G/ f {e “: at its ventral edge is everywhere
in contact with the parabasale.

The result of this smgular forma-

Fig. 12. The roof of the brain case seen from beneath. 3.
: : : , 5
This is composed of the two frontals and the two parietals, upon tion is that in Necturus the lateral

the ventral side of which is developed a series of processes which walls of the brain case are formed of

together form a U-shaped ridge that forms a front and sides to : +p
solid dermal bone reinforced exter-
the case. Contact surfaces for other bones are designated by an

x and the abbreviations of the bone involved. nally by cartilage while in most Uro-

NECTURUS MACULATUS. 413

deles the walls are formed by the cartilaginous trabeculae or in part by an ossification in
these elements, the orbito-sphenoid. A foramen for the olfactory nerve is formed at
the junction of the frontal and parietal contributions to this ridge; other nerves, as
the trigeminus and facialis, emerge from the cranial cavity just posterior to it, but the
nerves to the eye muscles and the opticus pass through tiny foramina, running very
obliquely through the process itself.

The floor of the brain case is formed by the very extensive parabasal, which receives
upon its dorsal surface the lateral processes projecting from the roof, and reinforces them
posteriorly by a low ridge topographically continuous with them. The details of the
separate bones are as follows : —

1. rrontaL. The frontals form a little more than a third of the dorsal surface of
the skull. They lie in contact with one another for about two thirds of their length,

diverging anteriorly to form a pair of short

premaxillary processes, and posteriorly to (ie fh Premaxiltary
: R f rocess if Process
form the longer and thinner parietal proc- “i \ p)'\

esses. A single frontal, isolated from its

surroundings, resembles a flat and quite

OLfactery
Process.

Fa
£4
Fen, ih Unenate

irregular splinter of bone from the under
(ventral) side of which hangs a partially
detached plate directed backwards, the
processus uncinatus of Wiedersheim,’ which
forms the anterior portion of the U-shaped

ridge described above. The notch enclosed

Margin im Contact with opposite Frontat,

between this process and the main body of

the bone transmits the olfactory nerve and

es——

is converted into a foramen by the addition DORSAL

of the antero-lateral process of the parietal

tie tal—Protess,
rela
<=
™
~4
4
>
>
iF

Parietal process

and the trabecula. A conspicuous proc-

Rave

ess upon the outer margin of the bone Gin ee ee ere
in this region, directed backwards, and surfaces with other bones are designated by an x.
seeming to belong to the uncinate process
rather than to the main body of the frontal, assists also in the formation of the olfactory
foramen and may be termed the olfactory process.

Upon the ventral side of the bone are seen two roughened ridges for the attachment
of other bones. Of these the more anterior is a curved ridge connecting the olfactory and
premaxillary processes and serving for the attachment of the vomer; the other involves

the ventral surface of the uncinate process and comes in contact with the parabasal.

414 HARRIS HAWTHORNE WILDER ON

Between them is a flattened depression, the internasal fossa, which lodges the internasal
plate of the chondrocranium. The dorsal side possesses few features of interest. “At the
anterior end is a longitudinal groove in which rests the ascending process of the premaxil-
lary and farther back is an oblique temporal ridge, extending from the olfactory notch to
the middle of the interfrontal suture, and serving as the anterior limit of the temporalis
muscle.

2. PARIETAL. This is the most irregular bone of the skull, and the largest, with the
exception of the parabasal. It possesses extensions in several directions, but it is so irreg-

ular that the number

Antero-Lateral of definite processes,

Process

Antero-laterat
Process

of which Wiedersheim
enumerates five, is
somewhat arbitrary.
median Of these, the most

Antero- Process)

a median important function-
Orbito-Sphenoid Process
Process. = ally as well as ana-
a PBL '' 3 : .
& tomically, is the lon-
= - . .
Quad rate 3 Alerts gitudinal ridge upon
ay Ir .
Peer 3 bis the ventral side, the
Y .
: edge of which comes
Pro-otic € Pp ty 4 :
Precess. £ ny in contact with the
‘ Process.
parabasal, and which
functionally replaces
an orbito-sphenoid and
serves as the side
z Opisth oti =
Opisthot of the brain case.
Process ~ VENTRAL DORSAL hee
This may be called
Fig. 14. Two views of right parietal. > 3. Contact surfaces with other bones

are designated by an x. the orbito-sphenoid
process, as sugges-
tive of its function,
and this process together with the uncinate process on the frontals completes the U-
shaped ridge mentioned above. This ridge is perforated by several very slanting foramina
for the transmission of the nerves of the eye muscles, but the exact identification of these
must be left for later investigation. The irregularly curved and very thin dorsal surface
may be conveniently divided into a nearly flat anterior portion and a decidedly convex
posterior portion, taking its shape from the otic capsule which it covers. From the

anterior portion proceed three processes, an antero-median, an antero-lateral, and a quad-

NECTURUS MACULATUS. 415

rate. The first two receive between them the parietal process of the frontal, and the
antero-lateral one, which is much the larger, reaches the olfactory region and assists in
the formation of the olfactory foramen. This process bears upon its ventral side a large
portion of the orbito-sphenoid process. The quadrate process is small and flat, and pro-
jects over that part of the quadrate cartilage designated above as the “isthmus.” The
outline of the posterior portion is practically a square, of which the two outer corners are
obvious, the posterior one being considerably prolonged. These may be termed respec-
tively the pro-otic and opisthotic processes in reference to the bones which they partially
overlap. This portion forms a thin outer covering for the otic capsule and corresponds to
it in shape, being convex above and concave beneath. It forms the otic fossa of the ven-
tral surface.

There are several prominent muscular ridges upon the dorsal surface of the parietal,
the most important of which are several irregular ones upon the posterior portion and a
median one formed between the two bones and prolonged into a median intermuscular
septum.

3. PARABASAL. ‘This is the flattest and most extensive bone of the skull and forms
nearly the whole of the floor
of the brain case, and at the
same time the roof of the
mouth. It is nearly in the

shape of a parallelogram with

Position
of . .
trabecule. 3 i

rounded corners, but isa little
broader in the otic region
and becomes somewhat nar-
rowed anteriorly. It is al-
most without special features
other than the impressions
made by the bones which
come in contact with it, the
frontals and parietals upon its

dorsal, and the vomers upon

a YN
? Area in QE
lm dcgntane =

its ventral surface. Upon Exo

the dorsal surface at the DORSAL
VENTRAL

posterior end there is a pair

. : Fig. 15. Two views of the parabasal taken from different individuals. The
of low lateral ridges which 4

view of the dorsal side is 4 times enlarged, that of the ventral, being from a
reinforce the orbito-sphenoid much larger individual, but 24. Contact surfaces with other bones are desig-

processes of the parietal and nated By, a2,

416 HARRIS HAWTHORNE WILDER ON

converge at the posterior end. There is necessarily a slight depression between them,
but anything asmarked as the depression described by Wiedersheim and likened to a
“sella turcica,” I have failed to find. Probably this outline would be much more noticeable
in_cross sections, which were studied extensively by this author, and in a young animal it
is likely that it would be proportionally more marked. The parabasal normally comes into
contact with ten bones, eight upon the dorsal surface and two upon the ventral. Dorsally
there are found in order, the two frontals, and the two parietals, by their uncinate and
orbito-sphenoid processes respectively ; the two pro-otics, and the two exoccipitals, the
latter almost anchylosed with it mm old adults. Upon the ventral side are the two vomers.
The palato-pterygoids and the opisthotics come very near the parabasal, but form no defi-
nite area of contact.

The upper jaw.— Under this term may be included all the dentigerous bones of the
skull proper, as the teeth of all these oppose those upon the mandible. In this group
both of the typical arches are represented, the maxillary and the palato-quadrate, or the
“inner” and “ outer arches” of W. K. Parker, the former by the premaxillaries and the
latter by the vomers and palato-pterygoids. Each of these arches bears a row of teeth,
and, when the mouth is closed, the single row on the mandible is received between the two
upper rows. The internal or palatine row is longer than the external or maxillary.

1. THE MAXILLARY ARCH. This arch is represented in most Urodeles by both pre-
maxillaries and maxillaries, the latter extensive and ending in long, backward projecting
processes which, in the recent state, are attached by strong ligaments to the quadrates,
thus making the arch a complete one. In Necturus it suffers much reduction and is
represented by the premaxillaries alone. The maxillary fails completely, and no rudi-
ment of it seems to be present at any developmental stage. Hyrtl describes a dried
specimen with a small but tooth-bearing mawil-
lary upon one side. Neither Huxley nor Wie-

dersheim could find a trace of it, and in an

examination of more than fifty skulls I have
never seen it. The specimen described by

Hyrtl must have been either an abnormal case

Sea0Gd Suipuarsy

or one of mistaken identity.
a. Premavillary.— This bone consists of

DORSAL MEMT RAL a slightly curved alveolar portion, to the inner
Fig. 16. Two views of right premaxillary. x 3. end of which an ascending process is added at

nearly aright angle. The alveolar portion is
placed in an antero-lateral position, ventral to the large nasal capsule, which lies obliquely

across it and forms the anterior part of the outer margin of the skull. The ascending

NECTURUS MACULATUS. ALT

process is in the form of a flattened splint and is superposed upon the dorsal surface
of the anterior end of the frontal, lying in a shallow groove formed for its reception.
This is the only contact between the premaxillary and any of the other bones of the
skull; but the extreme tips of the rostral cartilages may touch its ventral surface, and
the two premaxillaries almost touch each other at their antero-internal angles.

2. THE PALATO-PTERYGOID ARCH. This arch in Urodeles consists typically of three
dermal elements, vomer, palatine, and pterygoid, but owing to fusion or loss of one or
more of the elements, it usually appears in a modified form. In some Urodeles the first
two of the typical elements fuse and form a broad vomero-palatine, while in others, as is
the case in Necturus, the vomer is distinct and the second and third elements fuse to form
a palato-pterygoid. Usually in the vicinity of the pterygoid there is a part of the primitive
cartilaginous palato-pterygoid arch around which the dermal bones may be supposed to
have originated, and this is denominated the “ cartilaginous pterygoid.” In a few cases
(Desmognathus, Batrachoseps) the osseous pterygoid fails, leaving the vomero-palatine
as the sole dermal representative of the arch.

a. Vomer.—The main portion of this bone is somewhat triangular in shape and
forms the extensive, flattened palatine portion. Added to this upon its outer border
appears the alveolar portion, set at right angles to the main part and slightly curved.
The palatine portion forms the greater part of the anterior third of the roof of the mouth
but the two bones do not meet in the mid-
dle line. The inner borders, forming the
longest side of the triangle, are quite near
one another anteriorly but diverge posteri-
orly, and in the space thus left appears the
parabasale, which also extends along the | ny cya
dorsal side of the palatine process of the yy °F PB.
vomer, thus doubling the bony roof of the
palate over a large extent of the anterior
surface. The parabasal does not, how-
ever, extend anteriorly as far as the vom-
erine teeth, and there is thus left a consid-
erable interval, bounded by the above

mentioned teeth, the inner borders of the

vomers, and the anterior border of the DORSAL.
¢ ; ; VENTRAL.

parabasal, where there is no bony roof. ; a i

: Fig. 17. Two views of right vomer. 3. Contact sur-

This opening is nearly square, and through faces with other bones are designated by an x.

it, in the prepared skull, there may be

418 HARRIS HAWTHORNE WILDER ON

seen the internasal plate of the primordial skull. Viewed from the ventral side, the two
vomers may be said to overlap the parabasal, and the areas of contact are clearly marked
upon the dorsal surface of the former bone and the ventral surface of the latter. The
frontal bone comes in contact with the vomers dorsally, and the articular surface for this
forms a long, narrow ridge, running along the lateral border above the row of teeth. At
the extreme anterior end, the two vomers come in contact with each other, the symphysis
being marked by a few jagged, irregular projections. The postero-lateral border of the
palatine portion is recurved and exhibits an articular surface for the anterior end of the
palato-pterygoid.

b. Palato-pterygoid. — This is a flattened bone, somewhat in the form of a narrow
parallelogram, but with the anterior end rounded. It is set in the skull obliquely to the
longitudinal axis and extends from the vomer to the quadrate, articulating with both and
with no other bones. Its alveolar portion is small and confined to a short curved ridge

upon the outer border at the anterior end. The space
ay between the inner border of this bone and the outer border
of the parabasal is in the form of a very narrow triangle
and is filled in the recent state by a firm membrane. The

Alveolat pottion

orientation of this bone may be easily made by the teeth
which are upon the ventral side, at the anterior end, and

along the outer margin.

The Visceral Skeleton.

AL TRAL . .
Pee eas General morphology of the arches.— The seven visceral

Fige15." Two views of right pal: arches inherited from the Selachians, and surviving in most

ato-pterygoideum. X 4, (from small f
higher vertebrates, are represented here, although one of

specimen). Contact surfaces with
other bones are designated by an x. them (the 6th) is vestigial, and appears in the adult as
merely an intermuscular septum. Their disposition is as

follows : —
1. This arch, forming the functional jaws of the Selachians, possesses typically
a dorsal and a ventral segment, of which the former is represented by the cartilaginous
palato-quadrate arch, and the latter by Meckel’s cartilage of the mandible. Both of these
arches, in most higher vertebrates, become encased by dermal bones and lose their
identity more or less completely in the adult, the palato-quadrate arch suffering more
in this particular than the other. In Necturus the anterior and posterior ends of the
palato-quadrate arch survive as the antorbital process and the quadrate cartilage

(including its ossification) respectively, and the median portion is unrepresented. The

NECTURUS MACULATUS. 419

dermal bones connected with this arch are the vomer and the palato-pterygoid. These
as well as the quadrate have been previously described. The ventral segment is
represented by a very complete pair of Meckel’s cartilages, encased by three dermal bones,
the dentale (dentary), spleniale (operculum), and angulare. The proximal, or posterior
end of this cartilage forms a rounded articular surface which articulates with the mandib-
ular process of the quadrate, and forms an element often ossified in amphibians, and then
known as the articulare.

2. The second visceral arch consists in the Selachians of a dorsal segment, the hyo-
mandibular, and a ventral segment, the true hyoid. The latter piece is always present
and well developed in Urodeles, and in Necturus consists of two well developed cartilages
upon each side, but opinions differ widely concerning the fate of the hyomandibular.
The views which connect it with the stapes of mammals and the columella of reptiles seem
to have much support, but any connection between it and the Urodele operculum is evi-
dently disproven by the origin of this latter piece from the side of the otic capsule. It
thus seems safe to assert that the dorsal or hyomandibular segment of the second visceral
arch has no skeletal representative in Necturus.

3-5. These are the three gill arches which guard and regulate the two gill slits
situated between them upon either side of the neck, support the three pairs of
integumental branchiae, and furnish attachment for the muscles which regulate them.
This association of external branchiae, belonging to the integumental system, with the
three gill arches, which in fishes support the internal endodermic gills, has led in the past
to a confused suggestion of homology between these genetically distinct structures. The
gill arches are represented here by twelve cartilages, five on each side and two in the
median line, and all of these pieces are joined with the four of the hyoid arch to form the
hyobranchial apparatus.

6. This is the rudiment mentioned above, and represented by an intermuscular
septum.

7. This arch is represented by a single pair of cartilages which guard the entrance
to the trachea and extend along its sides; the cartilagines laterales, or laryngo-tracheal
cartilages.

The parts will now be taken up in detail.

The mandible—The mandible consists of the four separate elements mentioned
above :’ Meckel’s cartilage and the three dermal bones, dentale, spleniale, and angulare.
The cartilage is well encased by the dermal pieces, but its surface is exposed along a part
of the internal aspect, as well as over a large portion of the proximal, or articular end,
where it forms the joint. In the larva the cartilages of the two sides meet at the mid-
ventral line, and receive the protection of the dentalia upon their outer side alone, but as

420 HARRIS HAWTHORNE WILDER ON

these latter bones develop, they expand somewhat at their meeting with one another and
form a strong bony symphysis, suppressing the median ends of the cartilage. The dentale
is the largest of the dermal bones and covers the entire outer side of the mandible from
symphysis to posterior angle. It forms nearly the anterior third of the inner surface, and
two thirds of the lower edge, forming, with the lower border of the angulare, the fora-
men mandibulare. The angulare is next in size and is the main bone of the inner sur-
face, extending along the posterior two thirds of the mandible. It forms the posteriorly
directed angular process, and a small part of the bone appears in this region upon the
outer surface. The spleniale is very much reduced in Necturus and is in the form of a
little oval scale, set somewhat on the inner side, filling an interval between the angulare
and the dentale. It is dentigerous and bears a few (5 to 7) teeth, which form a row not
exactly continuous with that of the dentale, and opposed to that of the palato-pterygoid
in the upper jaw. ‘The details of the above osseous elements follow.

1. pentTaLe. This bone gives the contour to the jaw and follows quite closely the
general outlines of the head, and thus the posterior part is nearly straight while the ante-
rior third curves somewhat abruptly inwards. It is quite thin and its upper and lower

edges are curved inwards, making its outer surface somewhat convex and the inner con-

INWER ASPECT

Fig. 19. Internal view of right dentale. x 3. Contact surfaces with other bones are designated by an x.

cave. At the anterior third so much of the lower edge curves in that it forms a fairly
broad and flat submental surface involving about half of the total width. The dentale
bears a row of teeth which occupy nearly the anterior half. They are inserted along the
inner aspect, about half their length appearing above the upper edge of the bone. Just
posterior to this row of teeth the upper margin rises a little to form a very rudimentary
ramus which receives the insertion of the very stout tendon of the masseter muscle.
Internally, the dentale is chiefly characterized by its concavity, which lodges Meckel’s
vartilage and forms a deep and narrow groove anteriorly which broadens out posteriorly
to the full width of the bone.

The two dentalia form a strong symphysial articulation with one another at their

NECTURUS MACULATUS. 42]

anterior ends, a union which, in a macerated skull, often holds long after the separation of
the other components of the jaw. The recurved ventral margin of the dentale becomes
closely applied upon the inner side of the mandible to the corresponding edge of the
angulare and usually completes the formation of the intermandibular foramen. The
spleniale comes in contact with it by a portion of its outer surface.

2. ANGULARE. This bone is a curved splint, broad behind and tapering anteriorly
into an extremely fine and sharp point. Its outer aspect is concaved to receive Meckel’s
cartilage, and this surface, together with the inner surface of the dentale, forms a nearly
complete canal for the protection of the cartilage just named. The most solid portion of
the bone is that which forms its
lower margin, and this part, when Outer Aspect
articulated with the dentale, forms
a direct continuation of the flat sub-
mental surface of the latter. Poste-
riorly it ends in a blunt, rounded, gk
angular process which furnishes at-
tachment to the digastric muscle.

Along the external aspect of this
heavy portion there run a sharp
edge for articulation with the ven-
tral edge of the dentale, and in this,
at about its posterior third, is either

a deep notch or a complete fora-

For mandib

men, the former being much the :
Groove sor
more usual. This is reinforced by neckesicars tags
the edge of the dentale and forms Inwen Aspect
the mandibular foramen, through Fig. 20. Two views of right angulare. Xx 3. Contact surfaces

- . 5 with other bones are designated by an x.
which the mylo-hyoid branch of the

fifth nerve reaches the intermandib-
ular region. The remaining, or dorsal part of the bone forms a flattened and much
curved wing ending in a sharp upper edge, the rounding outline of which forms a coro-
noid process which receives the insertion of the temporalis muscle. Along the anterior
slope of this process the upper margin is applied to the lower margin of the spleniale.
This bone is easily oriented, since the more solid base is ventral and the slight curve
of the main axis follows the curve of the jaw.
3. SPLENIALE. Next to the operculum of the otic capsule, this is the smallest bone

of the skull, and by a singular coincidence, many authors have given it the name “ oper-

422 HARRIS HAWTHORNE WILDER ON

culare,” thus producing a confusion of terms. The spleniale is a thin scale of oval shape,
presenting a slightly concave inner, and a slightly convex outer surface. Along the side
of the former is attached a row of teeth, and the outer surface shows a longitudinal divi-
sion into an area covered by the den-
tale and a free area, which contributes
a little to the formation of the outer

surface of the mandible.

The oval shape and lack of detail

INNER ASPECT
‘ Outer ASPECT

Fig. 21. Two views of right spleniale. X 8. Contact sur- oipthis hove render is oe
faces with other bones are designated by an x. P. ant., processus ficult, but its outer and inner surfaces
anterior. are very evident, and in most cases

the anterior end is prolonged into a
point, while the posterior end is rounded. In its natural position it rests by its lower mar-
gin upon the upper edge of the angulare just anterior to the coronoid process, and the
ventral half of its outer surface is closely applied to the dentale.

The hyobranchial apparatus. — This consists of a series of sixteen pieces, all but
one being entirely cartilaginous, and representing four of the original visceral arches.
Their arrangement is so easily seen from plate 65, figure 12, that a verbal description of
the parts would seem almost superfluous. The system is seen to contain two median
pieces, universally designated as the first and second basibranchials, and referred to the
first and second branchial arches respectively, of which they form the middle pieces or
copulae. Of these, the second is ossified in the adult and its free posterior end usually
terminates in a rounded extremity, although an occasional individual shows a division at
the end into a two or three forked form, such as occurs normally in Siren. The hyoid
arch, which is attached to the anterior end of the first basibranchial, consists of two pieces,
an inner hypohyal and an outer ceratohyal. The first branchial arch is the best devel-
oped of all the branchial arches, consisting of two nearly equal pieces, cerato- and
epi-branchiale 1, the first of which is directly connected with both the first and second
basibranchial and with its opposite. The second branchial arch consists of a smaller
epibranchial, and a rudimentary ceratobranchial, reduced to a nodule of cartilage lying
upon the inner side of the distal end of the first ceratobranchial. An epibranchial smaller
than the previous one is the only remnant of the third branchial arch.

In life the distal ends of the three epibranchials support the external integumental
branchiae and furnish attachment for some of their muscles, a circumstance which has
often misled investigators as to the true homology of these purely integumental organs,
since the location suggests a definite phylogenetic relation to the internal gill system.

In the intervals between these epibranchials there occur, even in the adult Necturus, two

NECTURUS MACULATUS. 493

open gill slits guarded by rudiments of gill rakers, much as in the Axolotl, and as these
occur, the one between the first and the second arches, and the other between the second
and the third, and consequently in the intervals between the integumental branchiae, the
suggestion of connection between the two systems is still more misleading. It may be
remembered in this connection that there is a similar external gill upon the shoulder
girdle in the Dipnoan, Protopterus, the relation of the two being wholly topographical
and without morphological significance. Hence the physiological moments which have
developed the external branchiae in this place are plainly the utilization of the gills and
their muscular mechanism for support and motion, as well as the currents of fresh water
which can be driven through the gill slits in such a way as to bathe the respiratory
fringes.

Since the attempt of Fischer (’64) to account for a fourth branchial arch by supposing
that the first consists of two fused arches, no good suggestion concerning branchial arches
posterior to the third had been made up to about ten years ago, since when a series of
investigations by Gegenbaur (29’), Géppert (94), and myself (92), have rendered it
probable that the laryngo-tracheal cartilages of Amphibia are modified portions of the
fifth arch, perhaps the epibranchials. A fourth arch, attached to the hyobranchial
apparatus, is present in several of the lower Urodeles, (Siren, Amphiuma, Crypto-
branchus), and its apparent absence in Necturus has awakened some little speculation.
The rudiment of this was finally discovered by Géppert, who found it to consist of a
raphe of connective tissue, separating two muscles normally belonging to the branchial
system, and in which, in the larva, he discovered a few cartilage cells. This view I can
corroborate by investigations upon the same point at the same time, and which had led
me independently to the same conclusions previous to the appearance of the work of
Géppert. This raphe is given in the figure referred to (pl. 65, fig. 12), and there can now
be little doubt that it represents the missing fourth epibranchial. Concerning the struc-
ture of the metamorphosed fifth branchial arches (laryngo-tracheal cartilages), they con-
sist normally of a single pair of somewhat curved pieces, applied to the sides of the glottis
and short trachea, and were first described by Henle under the name of cartilago lateralis.
They are individually very variable and often have notches, foramina, or detached pieces,
but consist essentially of an anterior flat piece in the form of a triangle, from the poste-
rior internal angle of which depends a curved tracheal process extending posteriorly. A
more careful description of these cartilages and figures of several varying forms are given
in the author’s article on the amphibian larynx (96).

Suspensorial relations of the hyoid.— The distal end of the ceratohyal enters into a
ligamentous connection with various parts of the skull and mandible and thus bears a

close relation to the suspensorial apparatus of the mandible, a region which has such an

424 HARRIS HAWTHORNE WILDER ON

important morphological bearing that a description of it is worthy of special consideration.
Huxley (74) was the first to describe carefully this region in Necturus, and according to
this author there are three important ligaments in this region: (1) a mandibulo-hyoid
between the angular process of the mandible and the distal end of the ceratohyal; (2) a
hyo-suspensorial between this latter point and the quadrate; and (3) a suspensorio-
stapedial, from the quadrate to the operculum. A series of dissections of several very
large individuals, under the most favorable conditions as to preservation, has led me to
modify somewhat the account given by Huxley, and to substitute the imterpretation

graphically given in the accompany-
The distal end

of the ceratohyal is traversed by a

ing figure (fig. 22).

Soo deep groove in which runs the tendon
of the digastric muscle, and which

Membrane.

Groove for
tendon of ‘
Digastric Tauscle, ri

forms an outer and an inner lip.
At the posterior end of the inner lip
there is developed a rather promi-

nent process. There are two distinct

Lig. md-hy

extern, ff mandibulo-hyoid ligaments, external
Lig. ma-hy and internal, the one from the outer
intern.

lip of the digastric groove to the

Fig. 22. Relations of suspensorium, jaw and hyoid, taken from
the right side. X 3. Lig. hyo-sus., ligamentum hyo-suspenso- lower corner of the angular proces:
riale; lig. md.-hy. extern.,. ligamentum mandibulo-hyoideum and the other from the inner lip of
externum ; lig. md.-hy. intern., ligamentum mandibulo-hyoideum the groove and from its process to

internum ; membrane, that between skull and ceratohyal, form- °
the angular process just above the

The hyo-

suspensorial ligament is a narrow but

ing posteriorly a thickened band, extending between operculum
attachment of the other.

and ceratohyal.
very strong and distinct band extending between the middle of the inner lip of the
digastric groove and the outer side of the quadrate bone. In spite of repeated careful
dissections I could not find any definite “ stapedio-suspensorial ” ligament, but instead of
this a rather soft and somewhat indefinite sheet of connective tissue extending along
the entire inner lip of the ceratohyal and attached along a corresponding length of the
skull, involving the operculum, a bit of the cartilaginous otic capsule, and a part of the
quadrate. At its posterior edge its fibers become stronger and thicker, and thus resemble
a hyo-stapedial ligament, but as this part is directly continuous with the remainder of the
sheet and is by no means as definite as the other genuine ligaments, it can hardly be
described as a definite part. This loose sheet of connective tissue, which, as a matter of

fact, adheres very closely to the columellar process of the operculum, is undoubtedly the

NECTURUS MACULATUS. 425

part described by Huxley as the stapedio-suspensorial ligament. Certainly no definite
ligament, the fibers of which extend between the operculum and the quadrate, can be
demonstrated.

In this connection it is interesting to compare the work of other writers on the sub-
ject, who have naturally been influenced by the description given by Huxley. W. K.
Parker (°77, pl. 28, fig. 5) has figured this region in detail in Proteus. This shows the
same broad connective tissue band first described, but no definite stapedio-suspensorial
ligament. This band, which appears identical with that which I have described in
Necturus, is lettered “st. sl.,” but no mention is made of it in the text. The two other
ligaments are as in Necturus, if one counts the two mandibulo-hyoid ligaments as one.
Wiedersheim (77) accepts Huxley’s statements and quotes them (paraphrased, p. 435).
He speaks of a lig. stapedio-suspensoriale in Proteus, but in his figure 19, in which he
refers to this ligament as “Prop,” he really points out the bony connection between
operculum and paraquadratum, showing that he has confused the two, perhaps for the
moment. His figure of Proteus is in reality like W. K. Parker’s, and well expresses the
condition in Necturus.

The Free Sense Capsules.

The nasal capsules and the sclera of the eyes receive a more or less complete rein-
forcement of cartilage, which is structurally a skeletal part, and may be considered under
the heading given above.

The nasal capsules are somewhat in the shape of round-bottom flasks with necks
curved outwards, and they lie upon the sides of the anterior angle of the skull, extending
from the alveolar processes of the premaxillaries to the antorbital processes. The nostrils
or anterior nares appear at the end of the outwardly curved necks and are hence diver-
gent from one another and situated upon the sides of the blunt snout with a wide inter-
space. The posterior nares lie underneath, near the rounded posterior end, and in close
connection with the anterior margin of the antorbital processes which are curved some-
what around them. The sides and roof of each capsule are covered by a delicate reti-
cular or fenestrated cartilage of a soft consistency and rather difficult of demonstration as
cartilage.

A very similar structure is found in Proteus, where it was first described by Leydig,
in 1853, (quoted by Wiedersheim, 77) and the form in Necturus appears to have been
first shown by Wiedersheim, in 1877, who both describes and figures it with great clear-

426 HARRIS HAWTHORNE WILDER ON

ness, but may perhaps have given in his drawing somewhat too great regularity to the
fenestrations. Both Huxley (Necturus) and W. K. Parker (Proteus) appear to have
entirely overlooked the structure or at least not to have recognized its skeletal nature.
In the adult it appears like a shell covering the dorsal half of the capsule, its edge being
entire and running along the lateral outline. Its surface is perforated by openings in the
form of irregular parallelograms approximately arranged in two longitudinal rows and
converting the entire structure into a delicate cartilaginous network. At the anterior end
a narrow loop appears to extend across the ventral side, thus enclosing the nostril. By an
examination of the larvae it would seem that this structure at first consists of a single
longitudinal rod applied to the inner side of the capsule, after which transverse rods
develop between the folds of the nasal mucous membrane. Miss Platt (97) has given a
brief description of this in a larva of 46 mm., and Winslow (98) has both described and
figured it at the same stage. In the printed description of both authors the transverse
lateral processes develop or project directly from the main longitudinal rod, but in Win-
slow’s figure 16 certain of the transverse rods appear entirely separate from the longi-
tudinal one as though they had arisen independently of it. The later development of this
capsule and the manner in which the original elements finally weave themselves into the
complicated network seen in the adult, are points which have not as yet been investigated.

The term “optic capsule” seems almost too formal a word for the thin cartilaginous
ring which appears in the sclera and surrounds the eye. It is hardly demonstrable to the
unaided eye but is clearly seen in sections. Miss Platt found it in a larva of 46 mm. to
be “ three cells wide and one cell deep,” and in a small adult, of which I have sections, it
is from about 15 to 20 cells wide and about 4 cells deep in the thickest portion, tapering
towards each edge. The drawing of it given in plate 64, figure 6, is purely diagrammatic
and obtained from sections, as I have not succeeded in preparing an eye so that it could be
seen directly. According to Miss Platt, both nasal and optic capsules arise independently.

The Teeth.

The teeth of Necturus are arranged in two parallel rows in the upper jaw and in a
single row in the lower, those of the latter fitting into the interval between the two rows
of the former, their mutual arrangement being such that the teeth on the dentale oppose
those of the vomer, those on the spleniale oppose those on the palato-pterygoid, while the
premaxillary teeth form an unbroken outer row shutting over all the rest. At the junc-
ture between the vomerine and palato-pterygoid teeth of the inner, upper row, and again
between the dental and splenial teeth of the lower row there are slight breaks in the con-
tinuity of the rows, and usually a perceptible diastema or gap. Several of the authors

NECTURUS MACULATUS. 427

have given a definite number of teeth for each dentigerous bone, and Cope (’g9) has even
employed the number of teeth as a diagnostic between species, but as a matter of fact the
number is quite inconstant, and it seems probable that the rows are added to upon their
outer or posterior ends as the animals become older. The result of a series of enumera-
tions of different specimens may be given in the following tables.

TasLe A.

Teeth enumerated in entire specimens.

Designation Total |
of Side. | Premaxillary. | Vomer. | Palato-pterygoid. Dentale. Spleniale. ona Total
specimen. side. | teeth.
5
Rt 12 12 5) 14 6 49
es L, 12 13 | 5 15 5 50. | 22
Rt 10 LON 4) 14 6 45 :
: L u iia 6 Cie 5 Ag) 28
Rt 10 10. =|} 5 qo 6 45
ao if u 11 6 15 5 dey. 23
Rt ) 10 6 13 5 43
u i 9 11 6 13 5 44 | 87
Rt 14 15 8 15 il 59
E L Lb 15 3 16 S 62 | 121
F Rt 12 13 8 18 ——P —
if re) 12 7 17 s = | a
Taste B.

Teeth counted in isolated bones.

PMX. Ve Teale, DENT. SPL.
14 14 7 20 | 7
Ty ah ai} 8 18 uf
13 13 8 i | 7
15 15 9 Tome,
15 1G) 8 ilge | 6
16 6 AG eh. 7
15 6 18
| ig 17
average ® | 12.5— | 13.0 | 6.6 15.9+ | 6.2—

1 These small bones were unfortunately lost in this specimen.

2The specimens which furnished these isolated bones were generally larger than those used for Table A, which
accounts for the greater number of teeth.

3% this average are included the statistics from both tables.

498 HARRIS HAWTHORNE WILDER ON

The teeth are pleurodont, 7. e., situated upon the inner side of the tooth-bearing
ridge, and possess a root and a crown, the division between the two being coincident topo-
graphically with the edge of the bone upon the outside of a row of teeth. Thus the root,
which is formed simply of bone substance, and is unprotected by enamel, is guarded by
the ridge of bone, while the enameled crowns project beyond this protection. The root
is made of a hard bony substance or dentine, and is directly continuous with the bone
which bears it, arising from it by two lateral supports, which leave between them the
characteristic hollow seen in all of the figures that represent teeth as seen from the inner
aspect. Each hollow is really a foramen, which forms the entrance to the pulp canal of
the crown and serves in life to transmit the special nerves and blood vessels which supply
the tooth. The crown is in the shape of a slender cone, with a tip that is often set off a
little from the rest by a slight constriction. The apex is pointed and slightly recurved,
and is colored at the very point a deep amber color, similar to that found upon the
exposed outer surface of the incisors of beavers and many other rodents.

Comparison of Nomenclature.

That the numerous authors who have written upon the Urodele skull have not been
in accord with regard to the nomenclature of the elements concerned is to be expected
when we consider both the time covered by their investigations, and the fact that grounds
for an accurate homologizing of the parts have been sadly lacking, especially previous to
the epoch making years of the ’60’s and ’70’s. An attempt to arrange the synonyms
used by various authors in designating the parts of the skull of Necturus and a few allied
forms is given in the following table. The two last columns give respectively the terms
used in this work and the abbreviations by which they are designated in the figures.

NECTURUS MACULATUS.

TaBLe C,

Comparison of nomenclature of important parts of the skull.

429

HYRTL, 1865.
(Cryptobranchus)

HUXLEY, 1874.

WIEDERSHEIM, 1877.

ossa frontalial
ossa parietalia

os sphenoideum
basilare

ossaintermaxillaria

os maxillare supe-
rius (?)”

os pterygoideum
[Here without
the palatine. ]

ossa mastoidea

ossa petrosa

Os occipitis
partes laterales

os tympanicum

operculum

maxilla inferior {

lingulae cartilagin-
eae

cartilago ethmoidea

saccus yestibuli

parva lamina car-
tilaginea _ ossis
occipitis

os occipitis — pars
basilaris

frontal (bones)!
parietal (bones)
parasphenoid

premaxillae
vomers

palato-pterygoid

squamosals
pro-otic (ossifica-
tions)

epi-otics including
opisthotics

exoccipitals

quadrate (ossifica-
tion)

stapes

dentary

splenial
trabeculae

internasal fusion

of trabeculae

antorbital
esses

proc-

auditory capsules

frontalia
parietalia
parasphenoid

premaxillaria
vomer!

pterygo-palatinum

tympanicum sive
squamosum

pro-otische Ver- )
knocherung s.
regio pro-otica
regio epi-otica
regio opistho-
tica
regio occipi-
talis latera-
lis
kleine, am Ende
des Quadrat-knor-
pels auftretende
Verknécherung

operculum mit
columella

dentale

angulare

operculare

Trabekel; Schadel-
balken

unpaare Platte

vorderste Ausliau-
fer der Trabekel

Antorbital-fortsatz

Labyrinth-kapsel;
pars cartilaginea

supra-occipitale
Knorpel-platte

basi-occipitale
Knorpel-platte

areqdi990-oso.ajod

an ossified bridge
of cartilage

arch

W. K. PARKER, 1877. GAUPP, 1895. | WILDER, 1903. Abbrevi-
(Proteus) | (also frog, 1896.) | ations.
sl
frontals frontal |F
parietals | parietal pe
parasphenoid | parabasale parabasal | PB
premaxillaries | premaxillary PMX
vomers | vomer Vv
|
pterygo-palatines | | palato-ptery- PEE
| | goid
|
squamosals | paraquadra- paraquadrate | PQ
tum? |
tympanicum |
pro-otic pro-otic PO
epi-otic including | opisthotic 00
opisthotic
feebly developed exoccipital EXO
occipital floor
suspensorium, Os- quadrate Q
sified part
stapes | operculum O
|
dentary dentale DENT
articulare | angulare ANG
spleniale SPL
trabeculae trabeculae tr
internasal cartilage | internasal plate | in
trabecular cornua rostral proc-|r
esses
ethmo-palatine ele- antorbital proc- |-ao
ment, or antorbi- ess
tal element
ear capsules otic capsule ot
super-occipital supra-occipital | so
band arch
basi-oceipital bo

1 The form given here is taken direct from the text of the author cited, and may be in one case the plural and in another

the singular form, or may be used as an adjective.

These differences are of course without significance.

2 Gaupp, in suggesting the term “paraquadratum ” leaves open a possible, or even probable, homology with the tympani-

cum of mammals, and in the frog has so named the corresponding bone.

430 HARRIS HAWTHORNE WILDER ON

THE APPENDICULAR SKELETON.

Shoulder Girdle.

The shoulder girdle consists of a pair of thin plates, almost entirely cartilaginous,
wrapped about the sides of the body near the anterior end of the trunk, and entirely dis-
connected from other parts of the body skeleton and from one another. At about the
middle of each plate is situated the glenoid fossa for the reception of the head of the
humerus, and from this as a center there radiate three processes or lobes, one extending
dorsally and two ventrally. The dorsal or scapular extension is narrow at its origin but
broadens out towards its free end into a hatchet-shaped piece which extends as far dor-
sally as the transverse processes of the vertebrae. The narrow part of this extension
becomes ossified to form a scapula, in shape something like the diaphysis of a shaft bone,
with a constricted middle portion and two broadened ends; beyond this ossification the
hatchet-shaped piece remains as a cartilaginous suprascapula. Of the two ventral exten-
sions, the anterior, or procoracoid, is long and narrow and directed nearly anteriorly, while
the posterior, or coracoid, forms an almost circular flat plate closely applied to the myo-
tomic muscles on the ventral side of the thoracic region and extends so far beyond the
median line that the coracoid of one side considerably overlaps the other, the left being
usually the ventral or superficial one.

The cartilage, which is thin in most regions, is considerably thickened about the
glenoid fossa both to strengthen the region and to allow sufficient depth for the reception
of the head of the humerus. Externally, the thickened portion forms a definite ridge or
lip nearly surrounding the fossa, being deficient only for a small space upon the antero-
medial aspect.

Midway between the glenoid fossa and the re-entrant angle formed between procor-
acoid and coracoid is seen a small foramen coracoideum through which passes the supra-
coracoid nerve on its way to supply the muscles upon the ventral surface of the shoulder
girdle. C.K. Hoffmann has pointed out that in the Anura this nerve lies in the interval
between the procoracoid and coracoid, while here it bores through the cartilage, and that
the result is brought about in the former case by a deepening of the incision between the
two elements in question far enough to include the region of the foramen.

When the shoulder girdle is in its proper relationship to the body, the coracoid
extends from the second to the fourth myocomma, and hence the largest of the sternebra,
the one connected with the fourth myocomma, is situated exactly at the point at which
the posterior margins of the two coracoids diverge from one another, a relation precisely
similar to that of the sternal plate of the higher Urodela and of the arciferous Anura (e. ¢.,

NECTURUS MACULATUS. 431

Bombinator). This gives a suggestion of homology between the later amphibian sternum
and the sternebrum of the fourth myocomma alone, the shape and extent of which fre-
quently remind one of the well known rhomboid plate of such a form as Salamandra.

The procoracoid extends anteriorly as far as the first myocomma and its free anterior
end is frequently covered by the transverse fold formed by the posterior border of the

intermandibular muscle (M. intermaxillaris of authors) .

Pelvic Girdle.

The pelvic girdle consists of a flat ventral plate, the pubo-ischium, and two lateral
pieces, the ilia, attached to the sacral vertebra by means of a pair of ribs. The ventral
plate has an elongated pentagonal outline similar to that shown by the skull but with its
longitudinal axis still more prolonged. The anterior median angle is especially tapering
and extends along the mid-ventral line of the abdomen but shows no trace of an epipubic
cartilage (cartilago ypsiloides) as in so many Urodeles. The ilia are attached along the
posterior lateral sides and at their bases are situated the acetabula for the reception of the
heads of the femora. As in the case of the skull, the posterior margin is slightly incurved
and its outer angles (corresponding in position to the mastoid processes of the opisthotics)
are somewhat prolonged and tuberculate, forming the tuberosities of the ischia. At about
the middle of the pubo-ischium are seen two small obturator foramina which may be used
as indicative of the boundary between the pubic and ischiadic elements which are here
otherwise unmarked. A pair of osseous areas situated in the posterior half, and which
develop and increase in size during growth, plainly represent the osseous ischia. The
growth of these is well marked by concentric lines. The middle portion of the ilum is
also ossified, the bone being a little curved and with a rounded dorsal, and a broad and flat
ventral end.

The ventral face of the pubo-ischium shows a slight muscular ridge along the middle
line, and the dorsal or inner face is considerably excavated to form a pubo-ischiadic fossa
for the accommodation of some of the viscera. The floor of the acetabulum is usually
broken through by an acetabular foramen which leads into this last mentioned fossa.

The lateral view (pl. 65, fig. 13) shows the manner in which the ilium is attached to
the sacral rib. Instead of meeting end to end and forming a definite joint, the cartilag-
inous ends of the two are prolonged and tapering and are applied to the sides of one
another and held in place by firm connective tissue. The relations of this attachment to
the vertebrae have been considered above under the vertebral column. In two cases

which I have seen, a second ligamentous attachment appeared upon one side, extending

432 HARRIS HAWTHORNE WILDER ON

from the rib previous to the sacral rib to the dorsal end of the ilium; and in one of these
which I have had the opportunity of examining more closely, the sacral rib on the right
is normal and attached to what is probably the 19th vertebra, while the sacral rib upon
the left side proceeds from the next posterior vertebra (20th). In addition to this, a
strong but narrow ligament proceeds from the free end of the previous rib (19th) and
inserts upon the dorsal end of the ilium anterior to the attachment of the sacral rib.
As this fragment came from a student’s preparation, the remainder of which had been
lost, the exact determination of the vertebrae cannot be made, but judging from the
other cases of oblique attachment which have been reported, the numbering is undoubt-
edly as given. ‘This case resolves itself into one of “ oblique attachment,” such as have
been reported by G. H. Parker, Waite, and Bumpus, the obliquity being dextro-sinistral

(see above, under Vertebral Column).

The Free Limbs.

The serial homology between the fore and hind limbs in Necturus is very striking,
and in so primitive an animal, perhaps the most primitive one possessing a cheiropte-
rygium, points to a fundamental similarity of origin. It is thus of greater morphological
interest than correspondences in such modified structures as the paddles of Ichthyosaurus
where it is likely that the similarity is a secondary modification due to a similar method
of use. This resemblance, which is apparent externally, is still more emphasized by the
skeletal parts, and it seems incongruous to find such similar free limbs attached to such
different girdles. Each limb terminates in four digits, the lost member being generally
considered to be the first, and the phalangeal formula, 2—2-5-2, is the same in both
manus and pes."

With the exception of carpus and tarsus, which are wholly cartilagimous, the limb
bones of the adult consist each of a bony shaft running through the middle, and two
cartilaginous epiphyses, a slight exception being the terminal phalanges which are without
epiphyses at the distal end. The shaft, or diaphysis, ossifies perichondrially, as always in
the Amphibia, and forms a tubular sheath of bone, thick and constricted in the middle of
its length and tapering at the ends to thin edges, the whole mass being something like
the centrum of a biconcave vertebra considerably prolonged in the direction of its length.
The epiphyses always remain purely cartilaginous, and never obtain calcareous deposits as
in the frog.

‘On this point Cope (°89, p. 25-28) seems to have made a singular mistake, giving the phalangeal formula for N.
maculatus as 1-3-8-2 for the manus, and 1-2-2-2 for the pes. For the rare Carolina form, N. punctatus, he gives the
formula 2-2-3-2 for both limbs, which corresponds to the normal condition in N. maculatus and is undoubtedly character-
istic of the entire genus.

— ~

NECTURUS MACULATUS. 433

The lmbs when in the normal resting or swimming position are directed backwards,
and are held in such a way that in the hind limbs the soles face inwards, while in the
fore limbs the dorsal surface of the manus is ventral, and the palm dorsal. This is apt to
cause considerable difficulty in orienting the parts, but if the action of the limbs be
watched in a living animal it will be seen that this position is due, not so much to a
torsion, as to a swinging of the entire limb at the shoulder. The position assumed by the
fore limb is the easier to understand, and in this it is evident that the limb when extended
forward as in walking, rests with the palm down and the dorsal surface of the manus up,
but that when the entire limb is swung from the shoulder so as to change its direction and
point backwards, the positions of palm and dorsum are necessarily reversed. In the hind
limb there is some tendency to counteract this by a torsion of the limb about its own axis
and thus in a trailing hind foot the sole is turned somewhat inward.

Humerus.— The humerus, like the other long bones of the limbs, consists of an
osseous shaft and two cartilaginous epiphyses. The usual cylindrical shape is retained
only at the middle while the ends are both strongly flattened, and at right angles to
one another, the proximal end being flattened laterally, and the distal end dorso-ventrally.
The most prominent part of the proximal epiphysis is the head, which fits into the glenoid
fossa of the shoulder girdle and forms the characteristic ball-and-socket joint. It is sub-
spherical in shape, and is slightly affected by the general flattening of the entire region.
Its ventral face is prolonged into a sharp ridge which is continued by the osseous
diaphysis and forms one of the most distinctive features of the bone. This is the crista
ventralis (crista deltoidea of Ecker) and serves for the insertion of most of the ventral
shoulder muscles. This crest is highest near the junction of cartilage and bone and
rapidly recedes, so that at the middle of the diaphysis no trace remains.

The distal end, which is broadened laterally and flattened dorso-ventrally, bears an
extensive median furrow running around the end and up both surfaces, and dividing it
into two masses which may be distinguished as the external (lateral) and the internal
(median) condyles, although they include parts which in higher forms are distinct from
the condyles themselves. The external condyle is somewhat the larger and is ball-shaped,
fitting into the socket in the head of the radius. It also serves as a point of origin for the
extensor muscles of the forearm and hand. Having these two functions, it plainly corre-
sponds to both the external condyle and the capitulum of higher animals (epicondylus
medialis and eminentia capitata of the frog; Gaupp).

The internal condyle is more nearly the exact homologue of the part of that name in
higher forms as it gives origin to the flexor muscles of forearm and hand, but does not
participate as directly in the formation of the elbow joint. The groove between the two
condyles lies in the greater sigmoid notch of the ulna, and the olecranon of the ulna and

454 HARRIS HAWTHORNE WILDER ON

the capitulum of the radius fit respectively into its dorsal and ventral continuations
during forced extension and flexion of the forearm. These slight grooves or depressions,
of which the ventral is, perhaps, a little stronger than the dorsal, are thus the first sugges-

tions of the olecranar and cubital fossae which develop in higher forms.

The humerus is distinguished from the femur by its sharp crista ventralis which pro- .

jects farther than that of the femur and is differently shaped. This same feature, which
marks the ventral side, and the large and spherical external condyle serve to orient the
humerus.

Femur.— The general shape of the femur is like that of the humerus, but without
much flattening at the proximal end. Here it shows two projections for muscular attach-
ment: (1) a distinct process, mostly formed by bone, but tipped with a cartilage which in
the adult is independent of the main epiphysis, and (2) a ridge or crest upon the side of
the head, involving also a part of the shaft. It is somewhat difficult to compare these two
distinct features with the single crista ventralis of the humerus, but in accordance with
position it would seem that the bony process corresponds most closely with that part, and
may thus bear the same name, while the ridge which lies more upon the external lateral
aspect may be designated as the crista lateralis. As this crest is similar in appearance,
though not in position, to the proximal cartilaginous part of the crista ventralis of the
humerus, it is possible that these two parts found in the femur correspond to the single
process of the humerus, and that in the former a difference in motion and consequently
in muscular insertion has separated the two parts and diverted one of them from its origi-
nal position. Such an explanation is, however, not in accord with the almost universally
greater differentiation of the anterior limb, and one would expect to find in the femur
the more primitive condition. Hoffmann distinguishes in the Urodeles a single projec-
tion referred to as crista femoris, a name applied by Gaupp to a similar part in the frog,
but the femur of this latter animal is so unlike that of Necturus that homologies of parts
are uncertain. It hence seems better to refer to the two processes found in Necturus by
the two terms given above, crista ventralis and crista lateralis, terms which accurately
express the position of the parts, and one of which suggests a justifiable homology with
the humerus of the same animal.

The distal joimt is imperfectly divisible into the external and internal condyles,
although there is no distinct groove between them. The internal condyle is larger and
longer and serves as the main point of origin for both flexor and extensor muscles of the
lower leg and foot. The smaller external condyle articulates with the proximal end of
the fibula.

The osseous crista ventralis, separated from the head by a notch, is sufficient to dis-
tinguish femur from humerus and to mark the ventral side of the bone. This and the

more projecting internal condyle of the distal epiphysis will complete the determination.

NECTURUS MACULATUS. 435

Antibrachium and manus ; crus and pes.— These parts in the two limbs are almost
indistinguishable from one another, as may be seen by a comparison of figures 14 and 15
(plate 66), which were drawn from the right anterior and posterior limbs of the same
individual by the aid of a camera lucida. Corresponding to a difference in use of elbow
and knee, the proximal ends of ulna and radius are different from those of tibia and fibula,
but otherwise the parts correspond as closely as do consecutive parapodia of a polychaet-
ous Annelid.

The ulna shows a greater sigmoid notch and the radius a capitulum, both for the
reception of parts of the distal epiphysis of the humerus. Furthermore, these bones are
a little longer than are the tibia and fibula. Carpus and tarsus are exact duplicates in
the two limbs and consist of two proximal bones, a centrale, and three bones of the distal
row. In the proximal row the intermedium is fused with the outer element (ulnare or
fibulare), leaving between them a foramen for the transmission of an artery.’ In the dis-
tal row, digits IL and III have each a distinct element, while those corresponding to IV
and V are fused. Four digits, II-V, are represented in each case, each with a well devel-
oped metacarpal or metatarsal and nine phalanges, three in the fourth digit and two in

each of the others. Three phalanges in the third digit, also, is a not uncommon anomaly.

LITERATURE.

Bumpus, H. C.
°97. A contribution to the study of variation. (Skeletal variations of Necturus maculatus Raf.) Journ. of morphol.,
vol, 12, p. 455-484, pl. A-C.
Cope, E. D.
89. The Batrachia of North America. Bull. U.S. nat. mus., no. 34, 525 pp., 119 figs., 86 pls.
Fischer, J. G.
°64. Anatomische abhandlungen tiber die Perennibranchiata und Derotremen, 4°. Taf. 1-8.
Gadow, H.
97. On the evolution of the vertebral column of Amphibia and Amniota. Phil. trans. royal soc. London, vol. 187,
series B, p. 1-57, fig. 1-56.
Gaupp, E.
792. Beitriige zur morphologie des schiidels. 1. Morphol. arbeiten, bd. 2, p. 275-481, taf. 13-16.
94a. Beitriige zur morphologie des schidels. 2. Das hyo-branchial skelet der Anuren und seine umwandlung. Mor-
phol. arbeiten, bd. 5, h. 3, p. 899-488, fig. 1, taf. 18-19.
94b. Beitriize zur morphologie des schidels. 3. Zur vergleichenden anatomie der schlitengegend am knochernen
wirbelthier-schiidel. Morphol. arbeiten, bd. 4, h. 1, p. 77-130, tat. 6-7.
96. Die anatomie des frosches. (New edition of the well known book of Ecker and Wiedersheim.) 1896,
Gegenbaur, C.
92. Die epiglottis. Vergleichend-anatomische studie. 4°: p. 1-69, fig. 1-14, taf. 1-2.

'The artery in the hand which runs through the foramen carpale is the art. cubitalis which thus reaches the dorsal sur-
face and is distributed to the digits, In the foot a similar function is performed by the continuation of the art. femoralis.

436 HARRIS HAWTHORNE WILDER ON

Goppert, E.
°94. Die kehlkopfmuskulatur der Amphibien. Morphol. jahrb., bd. 22, h. 1, p. 1-78, fig. 1-9, taf. 1-2.
°96. Die morphologie der amphibienrippen. Festschrift zum siebenzigsten geburtstage v. Carl Gegenbaur, bd. 1, p.
393-435, fig. 1-10, taf. 1-2. Leipzig.
°98. Der kehlkopf der Amphibien und Reptilien. 1 Theil: Amphibien. Morphol. jahrb., bd. 26, h. 2, p. 282-329,
fiz, 1-5, taf. 8-11.
:01. Beitriige zur vergleichenden anatomie des kehlkopfes und seiner umgebung, mit besonderer beriicksichtigung der
Monotremen. Aus Semon: Zoologische forschungsreise in Australien und dem Malayischen archipel. 4°: p. 535-
634, fig. 1-52, taf. 17-21.
Hoffmann, C. K.
°73—78. Amphibien. In Bronn’s Klassen und ordnungen des thierreichs, bd. 6, abth. 2. Leipzig u. Heidelberg.
Huxley, T. H.
°74. On the structure of the skull and of the heart of Menobranchus lateralis. Proc. zool. soc. London, 1874, p. 186—
204, pl. 29-82.
Hyrtl, J.
65. Cryptobranchus japonicus. Vienna.
Parker, G. H.
796. Variations in the vertebral column of Necturus. Anatom. anzeiger, bd. 11, p. 711-717, fig. 1-2.
Parker, W. K.
°77. On the structure and development of the skull in the urodelous Amphibia. Part 1. Phil. trans. royal soc. Lon-
don, vol. 167, pt. 2, p. 529-597, pl. 21-29.
(In the above article, the description of Proteus is the one of especial value in this place.)
Platt, J. B.
97. The development of the cartilaginous skull and of the branchial and hypoglossal musculature in Necturus. Mor-
phol. jahrb., bd. 25, h. 8, p. 377-464, pl. 16-18.
Stohr, Ph.
°80. Zur entwicklungsgeschichte des urodelenschadels. Zeitschr. f. wissenschaftl. zool., bd. 33, h. 4, p. 477-526, taf.
29-30. <
Waite, F.C. :
°97. Variations in the brachial and lumbo-sacral plexi of Necturus maculosus [= maculatus] Rafinesque. Bull. mus.
comp. zool., vol. 31, p. 69-92, pl. 1-2.
Wiedersheim, R.
°77. Das kopfskelet der Urodelen. Morphol. jahrb., bd. 3, h. 8, p. 352-448, fig. 1, taf. 19-23.
Wilder, H. H.
*92. Studies in the phylogenesis of the larynx. Anatom. anzeiger, bd. 7, no. 18, p. 570-580, fig. 1-3.
°96. The amphibian larynx. Zool. jahrb., abtheil. f. anat. u. ontog., bd. 9, h. 2, p. 273-818, fig. A—D, pl. 19-21.
Winslow, G. M.

798. The chondrocranium in the Ichthyopsida. Tufts college studies, no. 5, p. 147-201, pl. 1-4.

NECTURUS MACULATUS. A37

EXPLANATION OF PLATES.

(The outlines of all the figures were drawn with a camera, three times the natural size in the majority of the cases.
Since, however, the originals varied considerably in size, the resulting figures cannot be considered as proportionate to one
another. The figures of the chondrocranium, for example (pl. 63, figs. 4, 5), were drawn from a very small specimen and
enlarged four times, but the resulting figures are almost exactly comparable with those of the entire skull (pl. 63, figs. 2, 3)
drawn from a medium sized specimen at a magnification of but three diameters. A large skull at two diameters would give
about the size of figures 2 and 3.)

ABBREVIATIONS USED.

(The abbreviations given in the following list include those used in the figures, as well as those employed in the text

at various places.)

1. Bones.

A —atlas.

ANG — angulare.

BB — basibranchiale 2

CS — costa sacralis.

DENT — dentale.

EXO — exoccipitale.

F — frontale.

FB — fibula.

FM — femur (diaphysis).
H—“Ist haemal arch vertebra” (see text).
HM — humerus (diaphysis).
IL — ilium.

IS — ischium.

MC — metacarpalia.

MT — metatarsalia.

OO — opisthoticum.

OP — operculum.

P — parietale.

PB — parabasale.

PH — phalanges.
PMX — premaxillare.
PO — pro-oticum.
PPT — palato-pterygoideum.
PQ — paraquadratum.
Q— quadratum.
R—radius.

S— sacrum.

SC — scapula.

SPL — spleniale.

T— tibia.

UL— ulna.

V — vomer.

2. Other Designations.

Abbreviation. Name. Location.
a. b-o. — arcus basi-occipitalis.

art. —articulare (Meckel’s cartilage).

a. S—-O. — arcus supra-occipitalis.

b. b. 1. — basibranchiale 1.

br. — bronchus.

ce. —carpalia.

caps. nas. —capsula nasalis.

caps. opt. — capsula optica.

caps. ot. — capsula otica.

cart. lat. — cartilago lateralis.

cart. Mk. — cartilago Meckelii.

c. b. 1-2. — ceratobranchiale 1 et 2.

cent. — centrale.

c. h. —ceratohyale.

cond. ext. — condylus externus. H and FM.
cond. int.— condylus internus. H and FM.
cond. oc. — condylus occipitalis. EXO.

Abbreviation. Name. Location
cor.— coracoideum.

er. lat.— crista lateralis FM.

er, mus.— crista muscularis. Pelvic girdle.

er. temp.— crista temporalis. F.
er. ventr.— crista ventralis. H and FM.
e. b. 1-4.— epibranchialia 1-4.

epiph. ec. s.— epiphysis costae sacralis.

epiph. il.— epiphysis ilii.

. acet.— foramen acetabulare.

. carp.— foramen carpale.

cor.— foramen coracoideum.

. mand.— foramen mandibulare.

. obtur.— foramen obturatorium.

. Sty.-m.— foramen stylo-mastoideum.

rh Rh Fh FR Fh Fh rh

. tars.— foramen tarsale.
fib.— fibulare.
fn. oval.— fenestra ovalis.

438 HARRIS HAWTHORNE WILDER ON

Abbreviation. Name. Location.
foss. acet.— fossa acetabularis.
foss. glen.— fossa glenoidalis.
foss. olec.— fossa olecrani.
foss. pub.-isch.— fossa pubo-ischiadica.
h.-h.— hypohyale.
int.— intermedium.
isch.—ischium (cartilaginous).
isth.— isthmus.
lig. hyo-sus,— ligamentum hyo-suspenso-
riale.
lig. md-hy. ext:— ligamentum mandibulo-
hyoideum externum.
lig. md-hy. int.—ligamentum mandibulo-
hyoideum internum.
marg. ventr. —margo ventralis.
nar. ant.— naris anterior.

p. a-l.— proc. antero-lateralis. Ve

p. a-m.—proc. antero-medianus. BP:

p. a-o.— proce. antorbitalis.

p. ang.— proce. angularis. ANG.
p. ant.— proc. anterior. SPL.
p. artie.— proc. articularis. Q.

p. asc.— proc. ascendens. q.

p. colum.— proc. columellaris. OP.
p. cor.— proc. coronoideus. ANG.
p- mast.— proc. mastoideus. 00.

p. 00.— proc. opisthoticus. 12

Of anterior naris.

Abbreviation. Name.
. opere.— proc. opercularis.

. ot.— proce. oticus.

. par.— proce. parietalis.

. pmx.— proce. premaxillaris.

. po. — proe. pro-oticus.

. pt. — proc. pterygoideus.

. g.— proc. quadratus.

. q. d—proc. quadratus dorsalis.

. r.— proce. rostralis.

. uncin.— proc. uncinatus.
pe.— procoracoideum.

ped.— pediculus.

pl. int.-nas— planum internasale.
pub.— pubis.

pul.— pulmo.

q.— quadratum (cartilage).
rad.— radiale.

ss.— suprascapula.

syim.— symphysis mandibulae.
t.— tarsalia.

tib.— tibiale.

trab.— trabecula.

trach.— trachea.

tub. isch.— tuberositas ischii.
uln.— ulnare.

bof tof ef lof tes Lette} bef le} le} el tel

PLATE 62.

. orb-sph.— proce. orbito-sphenoidalis.

. q. Y.— proc. quadratus ventralis.

Location.

PQ.
12,
Q+4.
F.

F.

P.

q.

P.
PO.
PO.

F.

Figs. la, 1b. Vertebral column. x 1!. The view is a dorsal one as far as the sacrum, beyond which it becomes twisted

so as to expose the right side of the flattened caudal region.

The break between vertebrae 34 and 35 indicates that the

detached piece was drawn from a second specimen, but one in which the identity of the vertebrae was known.

A —atlas ; S—sacrum; H —the first vertebra showing a complete haemal arch.

PLATE 63.

Fig. 2. Dorsal view of skull with nasal capsules removed.

Fig. 3. Ventral view of same.

Fig. 4. Dorsal view of chondrocranium with its cartilage bones.

Fig. 5. Ventral view of same.

PLATE

Fig. 6. Nasal capsule in place ; dorsal view.
Fig. 7. Internal aspect of left mandible.

External view of same.

Fig. 10. Dorsal view of suspensorium (quadratum).

Fig. 11. Ventral view of same.

8
Fig. 9. Details of the otic capsule and suspensorium ; ventral view of right side.

64.

Some of the parts are slightly sepa-

Fig. 20.

Fig. 21.

NECTURUS MACULATUS.

PLATE 65.

. Ventral view of head and neck regions, showing visceral arches.

. Lateral view of pelvie girdle, with attachment to vertebral column.
. External lateral view of right humerus.

. External lateral view of right femur.

PLATE 66.

. Ventral view of right shoulder girdle.
. Lateral view of right shoulder girdle.
. Forearm and manus ; dorsal view of right.

Lower leg and pes ; dorsal view of right.

PLATE 67.

Ventral view of pelvic girdle.

Dorsal view of pelvic girdle.

- Printed, January, 1903.

439

MEMOIRS Bost. Soc. Nat. Hist. VoL. 5. PEATEGZ,

1b la

1L. H. Wilder, del.

WILDER, SKELETAL SYSTEM OF NECTURUS.

Feet ay tt are i es TIN IE oe eee ee ee NBN Gree Fe Lee

_ Mewotrs Bost. Soc.Nat HIST. VOLS. PLATE 63.

er

trab/(-Sloor of.
olfactory fossa.)

’
Caps. ot.

Heliotype Printing Co Boston.

WILDER, SKELETAL SYSTEM OF NECTURUS.

|
|

PLATE 64.

oe
ay

es
ag

es 9)
e ee
i 2
Sahl ie
a

IM OF NE

STE

7X

ETAL ¢

SKEL

WILDER,

— HA Witderdel.

ite}

= 5

= 4xa puor 3

i 4ut4uoo i ="

a, qury « A : :
s 7

)

WILDER, SKELETAL SYSTEM OF NECTURUS.

Meare soydida

sMoirs Bost Soc.Nat Hist. VoL.5.

a

Soc Nat. Hist.VoL5.

= Ftars,

potype Prenting OaBostin.

2 Sen TE RS tS Se a a ee me ee LE REE ee AAA A lt OS A A ELIE R ET COLE TE A ie

Heljorype Prnuing Ca Busters,

WaSTqny.
Yast qny

“QaID $f"

- ~<OSTU 49
499” Sssof EUS

unqqos

,SKELETAL System orf NEcTuRuS.

WItDER

Ay
ty in

ul